From d7eecb483cc29e929bbc5515b8def830d7fc6ad2 Mon Sep 17 00:00:00 2001 From: Daniel Mack Date: Thu, 28 Jan 2010 16:13:01 +0800 Subject: jfs_dmap.[ch]: trivial typo fix: s/heigth/height/g MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Daniel Mack Signed-off-by: Dave Kleikamp Cc: Jiri Kosina Cc: André Goddard Rosa Cc: jfs-discussion@lists.sourceforge.net diff --git a/fs/jfs/jfs_dmap.c b/fs/jfs/jfs_dmap.c index d9b031c..7e19d2f 100644 --- a/fs/jfs/jfs_dmap.c +++ b/fs/jfs/jfs_dmap.c @@ -195,7 +195,7 @@ int dbMount(struct inode *ipbmap) bmp->db_maxag = le32_to_cpu(dbmp_le->dn_maxag); bmp->db_agpref = le32_to_cpu(dbmp_le->dn_agpref); bmp->db_aglevel = le32_to_cpu(dbmp_le->dn_aglevel); - bmp->db_agheigth = le32_to_cpu(dbmp_le->dn_agheigth); + bmp->db_agheight = le32_to_cpu(dbmp_le->dn_agheight); bmp->db_agwidth = le32_to_cpu(dbmp_le->dn_agwidth); bmp->db_agstart = le32_to_cpu(dbmp_le->dn_agstart); bmp->db_agl2size = le32_to_cpu(dbmp_le->dn_agl2size); @@ -287,7 +287,7 @@ int dbSync(struct inode *ipbmap) dbmp_le->dn_maxag = cpu_to_le32(bmp->db_maxag); dbmp_le->dn_agpref = cpu_to_le32(bmp->db_agpref); dbmp_le->dn_aglevel = cpu_to_le32(bmp->db_aglevel); - dbmp_le->dn_agheigth = cpu_to_le32(bmp->db_agheigth); + dbmp_le->dn_agheight = cpu_to_le32(bmp->db_agheight); dbmp_le->dn_agwidth = cpu_to_le32(bmp->db_agwidth); dbmp_le->dn_agstart = cpu_to_le32(bmp->db_agstart); dbmp_le->dn_agl2size = cpu_to_le32(bmp->db_agl2size); @@ -1440,7 +1440,7 @@ dbAllocAG(struct bmap * bmp, int agno, s64 nblocks, int l2nb, s64 * results) * tree index of this allocation group within the control page. */ agperlev = - (1 << (L2LPERCTL - (bmp->db_agheigth << 1))) / bmp->db_agwidth; + (1 << (L2LPERCTL - (bmp->db_agheight << 1))) / bmp->db_agwidth; ti = bmp->db_agstart + bmp->db_agwidth * (agno & (agperlev - 1)); /* dmap control page trees fan-out by 4 and a single allocation @@ -1459,7 +1459,7 @@ dbAllocAG(struct bmap * bmp, int agno, s64 nblocks, int l2nb, s64 * results) * the subtree to find the leftmost leaf that describes this * free space. */ - for (k = bmp->db_agheigth; k > 0; k--) { + for (k = bmp->db_agheight; k > 0; k--) { for (n = 0, m = (ti << 2) + 1; n < 4; n++) { if (l2nb <= dcp->stree[m + n]) { ti = m + n; @@ -3606,7 +3606,7 @@ void dbFinalizeBmap(struct inode *ipbmap) } /* - * compute db_aglevel, db_agheigth, db_width, db_agstart: + * compute db_aglevel, db_agheight, db_width, db_agstart: * an ag is covered in aglevel dmapctl summary tree, * at agheight level height (from leaf) with agwidth number of nodes * each, which starts at agstart index node of the smmary tree node @@ -3615,9 +3615,9 @@ void dbFinalizeBmap(struct inode *ipbmap) bmp->db_aglevel = BMAPSZTOLEV(bmp->db_agsize); l2nl = bmp->db_agl2size - (L2BPERDMAP + bmp->db_aglevel * L2LPERCTL); - bmp->db_agheigth = l2nl >> 1; - bmp->db_agwidth = 1 << (l2nl - (bmp->db_agheigth << 1)); - for (i = 5 - bmp->db_agheigth, bmp->db_agstart = 0, n = 1; i > 0; + bmp->db_agheight = l2nl >> 1; + bmp->db_agwidth = 1 << (l2nl - (bmp->db_agheight << 1)); + for (i = 5 - bmp->db_agheight, bmp->db_agstart = 0, n = 1; i > 0; i--) { bmp->db_agstart += n; n <<= 2; diff --git a/fs/jfs/jfs_dmap.h b/fs/jfs/jfs_dmap.h index 1a6eb41..6dcb906 100644 --- a/fs/jfs/jfs_dmap.h +++ b/fs/jfs/jfs_dmap.h @@ -210,7 +210,7 @@ struct dbmap_disk { __le32 dn_maxag; /* 4: max active alloc group number */ __le32 dn_agpref; /* 4: preferred alloc group (hint) */ __le32 dn_aglevel; /* 4: dmapctl level holding the AG */ - __le32 dn_agheigth; /* 4: height in dmapctl of the AG */ + __le32 dn_agheight; /* 4: height in dmapctl of the AG */ __le32 dn_agwidth; /* 4: width in dmapctl of the AG */ __le32 dn_agstart; /* 4: start tree index at AG height */ __le32 dn_agl2size; /* 4: l2 num of blks per alloc group */ @@ -229,7 +229,7 @@ struct dbmap { int dn_maxag; /* max active alloc group number */ int dn_agpref; /* preferred alloc group (hint) */ int dn_aglevel; /* dmapctl level holding the AG */ - int dn_agheigth; /* height in dmapctl of the AG */ + int dn_agheight; /* height in dmapctl of the AG */ int dn_agwidth; /* width in dmapctl of the AG */ int dn_agstart; /* start tree index at AG height */ int dn_agl2size; /* l2 num of blks per alloc group */ @@ -255,7 +255,7 @@ struct bmap { #define db_agsize db_bmap.dn_agsize #define db_agl2size db_bmap.dn_agl2size #define db_agwidth db_bmap.dn_agwidth -#define db_agheigth db_bmap.dn_agheigth +#define db_agheight db_bmap.dn_agheight #define db_agstart db_bmap.dn_agstart #define db_numag db_bmap.dn_numag #define db_maxlevel db_bmap.dn_maxlevel -- cgit v0.10.2 From 17d9ddc72fb8bba0d4f67868c9c612e472a594a9 Mon Sep 17 00:00:00 2001 From: "Pallipadi, Venkatesh" Date: Wed, 10 Feb 2010 15:23:44 -0800 Subject: rbtree: Add support for augmented rbtrees Add support for augmented rbtrees in core rbtree code. This will be used in subsequent patches, in x86 PAT code, which needs interval trees to efficiently keep track of PAT ranges. Signed-off-by: Venkatesh Pallipadi LKML-Reference: <20100210232343.GA11465@linux-os.sc.intel.com> Signed-off-by: Suresh Siddha Signed-off-by: H. Peter Anvin diff --git a/Documentation/rbtree.txt b/Documentation/rbtree.txt index aae8355..221f38b 100644 --- a/Documentation/rbtree.txt +++ b/Documentation/rbtree.txt @@ -190,3 +190,61 @@ Example: for (node = rb_first(&mytree); node; node = rb_next(node)) printk("key=%s\n", rb_entry(node, struct mytype, node)->keystring); +Support for Augmented rbtrees +----------------------------- + +Augmented rbtree is an rbtree with "some" additional data stored in each node. +This data can be used to augment some new functionality to rbtree. +Augmented rbtree is an optional feature built on top of basic rbtree +infrastructure. rbtree user who wants this feature will have an augment +callback function in rb_root initialized. + +This callback function will be called from rbtree core routines whenever +a node has a change in one or both of its children. It is the responsibility +of the callback function to recalculate the additional data that is in the +rb node using new children information. Note that if this new additional +data affects the parent node's additional data, then callback function has +to handle it and do the recursive updates. + + +Interval tree is an example of augmented rb tree. Reference - +"Introduction to Algorithms" by Cormen, Leiserson, Rivest and Stein. +More details about interval trees: + +Classical rbtree has a single key and it cannot be directly used to store +interval ranges like [lo:hi] and do a quick lookup for any overlap with a new +lo:hi or to find whether there is an exact match for a new lo:hi. + +However, rbtree can be augmented to store such interval ranges in a structured +way making it possible to do efficient lookup and exact match. + +This "extra information" stored in each node is the maximum hi +(max_hi) value among all the nodes that are its descendents. This +information can be maintained at each node just be looking at the node +and its immediate children. And this will be used in O(log n) lookup +for lowest match (lowest start address among all possible matches) +with something like: + +find_lowest_match(lo, hi, node) +{ + lowest_match = NULL; + while (node) { + if (max_hi(node->left) > lo) { + // Lowest overlap if any must be on left side + node = node->left; + } else if (overlap(lo, hi, node)) { + lowest_match = node; + break; + } else if (lo > node->lo) { + // Lowest overlap if any must be on right side + node = node->right; + } else { + break; + } + } + return lowest_match; +} + +Finding exact match will be to first find lowest match and then to follow +successor nodes looking for exact match, until the start of a node is beyond +the hi value we are looking for. diff --git a/include/linux/rbtree.h b/include/linux/rbtree.h index 9c29541..8e33a25 100644 --- a/include/linux/rbtree.h +++ b/include/linux/rbtree.h @@ -110,6 +110,7 @@ struct rb_node struct rb_root { struct rb_node *rb_node; + void (*augment_cb)(struct rb_node *node); }; @@ -129,7 +130,9 @@ static inline void rb_set_color(struct rb_node *rb, int color) rb->rb_parent_color = (rb->rb_parent_color & ~1) | color; } -#define RB_ROOT (struct rb_root) { NULL, } +#define RB_ROOT (struct rb_root) { NULL, NULL, } +#define RB_AUGMENT_ROOT(x) (struct rb_root) { NULL, x} + #define rb_entry(ptr, type, member) container_of(ptr, type, member) #define RB_EMPTY_ROOT(root) ((root)->rb_node == NULL) diff --git a/lib/rbtree.c b/lib/rbtree.c index e2aa3be..15e10b1 100644 --- a/lib/rbtree.c +++ b/lib/rbtree.c @@ -44,6 +44,11 @@ static void __rb_rotate_left(struct rb_node *node, struct rb_root *root) else root->rb_node = right; rb_set_parent(node, right); + + if (root->augment_cb) { + root->augment_cb(node); + root->augment_cb(right); + } } static void __rb_rotate_right(struct rb_node *node, struct rb_root *root) @@ -67,12 +72,20 @@ static void __rb_rotate_right(struct rb_node *node, struct rb_root *root) else root->rb_node = left; rb_set_parent(node, left); + + if (root->augment_cb) { + root->augment_cb(node); + root->augment_cb(left); + } } void rb_insert_color(struct rb_node *node, struct rb_root *root) { struct rb_node *parent, *gparent; + if (root->augment_cb) + root->augment_cb(node); + while ((parent = rb_parent(node)) && rb_is_red(parent)) { gparent = rb_parent(parent); @@ -227,12 +240,15 @@ void rb_erase(struct rb_node *node, struct rb_root *root) else { struct rb_node *old = node, *left; + int old_parent_cb = 0; + int successor_parent_cb = 0; node = node->rb_right; while ((left = node->rb_left) != NULL) node = left; if (rb_parent(old)) { + old_parent_cb = 1; if (rb_parent(old)->rb_left == old) rb_parent(old)->rb_left = node; else @@ -247,8 +263,10 @@ void rb_erase(struct rb_node *node, struct rb_root *root) if (parent == old) { parent = node; } else { + successor_parent_cb = 1; if (child) rb_set_parent(child, parent); + parent->rb_left = child; node->rb_right = old->rb_right; @@ -259,6 +277,24 @@ void rb_erase(struct rb_node *node, struct rb_root *root) node->rb_left = old->rb_left; rb_set_parent(old->rb_left, node); + if (root->augment_cb) { + /* + * Here, three different nodes can have new children. + * The parent of the successor node that was selected + * to replace the node to be erased. + * The node that is getting erased and is now replaced + * by its successor. + * The parent of the node getting erased-replaced. + */ + if (successor_parent_cb) + root->augment_cb(parent); + + root->augment_cb(node); + + if (old_parent_cb) + root->augment_cb(rb_parent(old)); + } + goto color; } @@ -267,15 +303,19 @@ void rb_erase(struct rb_node *node, struct rb_root *root) if (child) rb_set_parent(child, parent); - if (parent) - { + + if (parent) { if (parent->rb_left == node) parent->rb_left = child; else parent->rb_right = child; - } - else + + if (root->augment_cb) + root->augment_cb(parent); + + } else { root->rb_node = child; + } color: if (color == RB_BLACK) -- cgit v0.10.2 From be5a0c126ad1dea2128dc5aef12c87083518d1ab Mon Sep 17 00:00:00 2001 From: "venkatesh.pallipadi@intel.com" Date: Wed, 10 Feb 2010 11:57:06 -0800 Subject: x86, pat: Preparatory changes in pat.c for bigger rbtree change Minor changes in pat.c to cleanup code and make it smoother to introduce bigger rbtree only change in the following patch. The changes are cleaup only and should not have any functional impact. Signed-off-by: Venkatesh Pallipadi LKML-Reference: <20100210195909.792781000@intel.com> Signed-off-by: Suresh Siddha Signed-off-by: H. Peter Anvin diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c index ae9648e..628e507 100644 --- a/arch/x86/mm/pat.c +++ b/arch/x86/mm/pat.c @@ -30,6 +30,8 @@ #include #include +#include "pat_internal.h" + #ifdef CONFIG_X86_PAT int __read_mostly pat_enabled = 1; @@ -53,19 +55,15 @@ static inline void pat_disable(const char *reason) #endif -static int debug_enable; +int pat_debug_enable; static int __init pat_debug_setup(char *str) { - debug_enable = 1; + pat_debug_enable = 1; return 0; } __setup("debugpat", pat_debug_setup); -#define dprintk(fmt, arg...) \ - do { if (debug_enable) printk(KERN_INFO fmt, ##arg); } while (0) - - static u64 __read_mostly boot_pat_state; enum { @@ -132,17 +130,6 @@ void pat_init(void) #undef PAT -static char *cattr_name(unsigned long flags) -{ - switch (flags & _PAGE_CACHE_MASK) { - case _PAGE_CACHE_UC: return "uncached"; - case _PAGE_CACHE_UC_MINUS: return "uncached-minus"; - case _PAGE_CACHE_WB: return "write-back"; - case _PAGE_CACHE_WC: return "write-combining"; - default: return "broken"; - } -} - /* * The global memtype list keeps track of memory type for specific * physical memory areas. Conflicting memory types in different @@ -159,14 +146,6 @@ static char *cattr_name(unsigned long flags) * memtype_lock protects both the linear list and rbtree. */ -struct memtype { - u64 start; - u64 end; - unsigned long type; - struct list_head nd; - struct rb_node rb; -}; - static struct rb_root memtype_rbroot = RB_ROOT; static LIST_HEAD(memtype_list); static DEFINE_SPINLOCK(memtype_lock); /* protects memtype list */ @@ -349,6 +328,64 @@ static int free_ram_pages_type(u64 start, u64 end) return 0; } +static int memtype_check_insert(struct memtype *new, unsigned long *new_type) +{ + struct memtype *entry; + u64 start, end; + unsigned long actual_type; + struct list_head *where; + int err = 0; + + start = new->start; + end = new->end; + actual_type = new->type; + + /* Search for existing mapping that overlaps the current range */ + where = NULL; + list_for_each_entry(entry, &memtype_list, nd) { + if (end <= entry->start) { + where = entry->nd.prev; + break; + } else if (start <= entry->start) { /* end > entry->start */ + err = chk_conflict(new, entry, new_type); + if (!err) { + dprintk("Overlap at 0x%Lx-0x%Lx\n", + entry->start, entry->end); + where = entry->nd.prev; + } + break; + } else if (start < entry->end) { /* start > entry->start */ + err = chk_conflict(new, entry, new_type); + if (!err) { + dprintk("Overlap at 0x%Lx-0x%Lx\n", + entry->start, entry->end); + + /* + * Move to right position in the linked + * list to add this new entry + */ + list_for_each_entry_continue(entry, + &memtype_list, nd) { + if (start <= entry->start) { + where = entry->nd.prev; + break; + } + } + } + break; + } + } + if (!err) { + if (where) + list_add(&new->nd, where); + else + list_add_tail(&new->nd, &memtype_list); + + memtype_rb_insert(&memtype_rbroot, new); + } + return err; +} + /* * req_type typically has one of the: * - _PAGE_CACHE_WB @@ -364,9 +401,8 @@ static int free_ram_pages_type(u64 start, u64 end) int reserve_memtype(u64 start, u64 end, unsigned long req_type, unsigned long *new_type) { - struct memtype *new, *entry; + struct memtype *new; unsigned long actual_type; - struct list_head *where; int is_range_ram; int err = 0; @@ -423,42 +459,7 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type, spin_lock(&memtype_lock); - /* Search for existing mapping that overlaps the current range */ - where = NULL; - list_for_each_entry(entry, &memtype_list, nd) { - if (end <= entry->start) { - where = entry->nd.prev; - break; - } else if (start <= entry->start) { /* end > entry->start */ - err = chk_conflict(new, entry, new_type); - if (!err) { - dprintk("Overlap at 0x%Lx-0x%Lx\n", - entry->start, entry->end); - where = entry->nd.prev; - } - break; - } else if (start < entry->end) { /* start > entry->start */ - err = chk_conflict(new, entry, new_type); - if (!err) { - dprintk("Overlap at 0x%Lx-0x%Lx\n", - entry->start, entry->end); - - /* - * Move to right position in the linked - * list to add this new entry - */ - list_for_each_entry_continue(entry, - &memtype_list, nd) { - if (start <= entry->start) { - where = entry->nd.prev; - break; - } - } - } - break; - } - } - + err = memtype_check_insert(new, new_type); if (err) { printk(KERN_INFO "reserve_memtype failed 0x%Lx-0x%Lx, " "track %s, req %s\n", @@ -469,13 +470,6 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type, return err; } - if (where) - list_add(&new->nd, where); - else - list_add_tail(&new->nd, &memtype_list); - - memtype_rb_insert(&memtype_rbroot, new); - spin_unlock(&memtype_lock); dprintk("reserve_memtype added 0x%Lx-0x%Lx, track %s, req %s, ret %s\n", @@ -937,28 +931,40 @@ EXPORT_SYMBOL_GPL(pgprot_writecombine); #if defined(CONFIG_DEBUG_FS) && defined(CONFIG_X86_PAT) /* get Nth element of the linked list */ -static struct memtype *memtype_get_idx(loff_t pos) +static int copy_memtype_nth_element(struct memtype *out, loff_t pos) { - struct memtype *list_node, *print_entry; + struct memtype *list_node; int i = 1; - print_entry = kmalloc(sizeof(struct memtype), GFP_KERNEL); - if (!print_entry) - return NULL; - - spin_lock(&memtype_lock); list_for_each_entry(list_node, &memtype_list, nd) { if (pos == i) { - *print_entry = *list_node; - spin_unlock(&memtype_lock); - return print_entry; + *out = *list_node; + return 0; } ++i; } + return 1; +} + +static struct memtype *memtype_get_idx(loff_t pos) +{ + struct memtype *print_entry; + int ret; + + print_entry = kzalloc(sizeof(struct memtype), GFP_KERNEL); + if (!print_entry) + return NULL; + + spin_lock(&memtype_lock); + ret = copy_memtype_nth_element(print_entry, pos); spin_unlock(&memtype_lock); - kfree(print_entry); - return NULL; + if (!ret) { + return print_entry; + } else { + kfree(print_entry); + return NULL; + } } static void *memtype_seq_start(struct seq_file *seq, loff_t *pos) diff --git a/arch/x86/mm/pat_internal.h b/arch/x86/mm/pat_internal.h new file mode 100644 index 0000000..6c98780 --- /dev/null +++ b/arch/x86/mm/pat_internal.h @@ -0,0 +1,28 @@ +#ifndef __PAT_INTERNAL_H_ +#define __PAT_INTERNAL_H_ + +extern int pat_debug_enable; + +#define dprintk(fmt, arg...) \ + do { if (pat_debug_enable) printk(KERN_INFO fmt, ##arg); } while (0) + +struct memtype { + u64 start; + u64 end; + unsigned long type; + struct list_head nd; + struct rb_node rb; +}; + +static inline char *cattr_name(unsigned long flags) +{ + switch (flags & _PAGE_CACHE_MASK) { + case _PAGE_CACHE_UC: return "uncached"; + case _PAGE_CACHE_UC_MINUS: return "uncached-minus"; + case _PAGE_CACHE_WB: return "write-back"; + case _PAGE_CACHE_WC: return "write-combining"; + default: return "broken"; + } +} + +#endif /* __PAT_INTERNAL_H_ */ -- cgit v0.10.2 From 9e41a49aab88a5a6c8f4875bf10a5543bc321f2d Mon Sep 17 00:00:00 2001 From: "Pallipadi, Venkatesh" Date: Wed, 10 Feb 2010 15:26:07 -0800 Subject: x86, pat: Migrate to rbtree only backend for pat memtype management Move pat backend to fully rbtree based implementation from the existing rbtree and linked list hybrid. New rbtree based solution uses interval trees (augmented rbtrees) in order to store the PAT ranges. The new code seprates out the pat backend to pat_rbtree.c file, making is cleaner. The change also makes the PAT lookup, reserve and free operations more optimal, as we don't have to traverse linear linked list of few tens of entries in normal case. Signed-off-by: Venkatesh Pallipadi LKML-Reference: <20100210232607.GB11465@linux-os.sc.intel.com> Signed-off-by: Suresh Siddha Signed-off-by: H. Peter Anvin diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile index 06630d2..a4c7683 100644 --- a/arch/x86/mm/Makefile +++ b/arch/x86/mm/Makefile @@ -6,6 +6,7 @@ nostackp := $(call cc-option, -fno-stack-protector) CFLAGS_physaddr.o := $(nostackp) CFLAGS_setup_nx.o := $(nostackp) +obj-$(CONFIG_X86_PAT) += pat_rbtree.o obj-$(CONFIG_SMP) += tlb.o obj-$(CONFIG_X86_32) += pgtable_32.o iomap_32.o diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c index 628e507..9510111 100644 --- a/arch/x86/mm/pat.c +++ b/arch/x86/mm/pat.c @@ -130,65 +130,7 @@ void pat_init(void) #undef PAT -/* - * The global memtype list keeps track of memory type for specific - * physical memory areas. Conflicting memory types in different - * mappings can cause CPU cache corruption. To avoid this we keep track. - * - * The list is sorted based on starting address and can contain multiple - * entries for each address (this allows reference counting for overlapping - * areas). All the aliases have the same cache attributes of course. - * Zero attributes are represented as holes. - * - * The data structure is a list that is also organized as an rbtree - * sorted on the start address of memtype range. - * - * memtype_lock protects both the linear list and rbtree. - */ - -static struct rb_root memtype_rbroot = RB_ROOT; -static LIST_HEAD(memtype_list); -static DEFINE_SPINLOCK(memtype_lock); /* protects memtype list */ - -static struct memtype *memtype_rb_search(struct rb_root *root, u64 start) -{ - struct rb_node *node = root->rb_node; - struct memtype *last_lower = NULL; - - while (node) { - struct memtype *data = container_of(node, struct memtype, rb); - - if (data->start < start) { - last_lower = data; - node = node->rb_right; - } else if (data->start > start) { - node = node->rb_left; - } else - return data; - } - - /* Will return NULL if there is no entry with its start <= start */ - return last_lower; -} - -static void memtype_rb_insert(struct rb_root *root, struct memtype *data) -{ - struct rb_node **new = &(root->rb_node); - struct rb_node *parent = NULL; - - while (*new) { - struct memtype *this = container_of(*new, struct memtype, rb); - - parent = *new; - if (data->start <= this->start) - new = &((*new)->rb_left); - else if (data->start > this->start) - new = &((*new)->rb_right); - } - - rb_link_node(&data->rb, parent, new); - rb_insert_color(&data->rb, root); -} +static DEFINE_SPINLOCK(memtype_lock); /* protects memtype accesses */ /* * Does intersection of PAT memory type and MTRR memory type and returns @@ -216,33 +158,6 @@ static unsigned long pat_x_mtrr_type(u64 start, u64 end, unsigned long req_type) return req_type; } -static int -chk_conflict(struct memtype *new, struct memtype *entry, unsigned long *type) -{ - if (new->type != entry->type) { - if (type) { - new->type = entry->type; - *type = entry->type; - } else - goto conflict; - } - - /* check overlaps with more than one entry in the list */ - list_for_each_entry_continue(entry, &memtype_list, nd) { - if (new->end <= entry->start) - break; - else if (new->type != entry->type) - goto conflict; - } - return 0; - - conflict: - printk(KERN_INFO "%s:%d conflicting memory types " - "%Lx-%Lx %s<->%s\n", current->comm, current->pid, new->start, - new->end, cattr_name(new->type), cattr_name(entry->type)); - return -EBUSY; -} - static int pat_pagerange_is_ram(unsigned long start, unsigned long end) { int ram_page = 0, not_rampage = 0; @@ -328,64 +243,6 @@ static int free_ram_pages_type(u64 start, u64 end) return 0; } -static int memtype_check_insert(struct memtype *new, unsigned long *new_type) -{ - struct memtype *entry; - u64 start, end; - unsigned long actual_type; - struct list_head *where; - int err = 0; - - start = new->start; - end = new->end; - actual_type = new->type; - - /* Search for existing mapping that overlaps the current range */ - where = NULL; - list_for_each_entry(entry, &memtype_list, nd) { - if (end <= entry->start) { - where = entry->nd.prev; - break; - } else if (start <= entry->start) { /* end > entry->start */ - err = chk_conflict(new, entry, new_type); - if (!err) { - dprintk("Overlap at 0x%Lx-0x%Lx\n", - entry->start, entry->end); - where = entry->nd.prev; - } - break; - } else if (start < entry->end) { /* start > entry->start */ - err = chk_conflict(new, entry, new_type); - if (!err) { - dprintk("Overlap at 0x%Lx-0x%Lx\n", - entry->start, entry->end); - - /* - * Move to right position in the linked - * list to add this new entry - */ - list_for_each_entry_continue(entry, - &memtype_list, nd) { - if (start <= entry->start) { - where = entry->nd.prev; - break; - } - } - } - break; - } - } - if (!err) { - if (where) - list_add(&new->nd, where); - else - list_add_tail(&new->nd, &memtype_list); - - memtype_rb_insert(&memtype_rbroot, new); - } - return err; -} - /* * req_type typically has one of the: * - _PAGE_CACHE_WB @@ -459,7 +316,7 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type, spin_lock(&memtype_lock); - err = memtype_check_insert(new, new_type); + err = rbt_memtype_check_insert(new, new_type); if (err) { printk(KERN_INFO "reserve_memtype failed 0x%Lx-0x%Lx, " "track %s, req %s\n", @@ -481,7 +338,6 @@ int reserve_memtype(u64 start, u64 end, unsigned long req_type, int free_memtype(u64 start, u64 end) { - struct memtype *entry, *saved_entry; int err = -EINVAL; int is_range_ram; @@ -505,46 +361,7 @@ int free_memtype(u64 start, u64 end) } spin_lock(&memtype_lock); - - entry = memtype_rb_search(&memtype_rbroot, start); - if (unlikely(entry == NULL)) - goto unlock_ret; - - /* - * Saved entry points to an entry with start same or less than what - * we searched for. Now go through the list in both directions to look - * for the entry that matches with both start and end, with list stored - * in sorted start address - */ - saved_entry = entry; - list_for_each_entry_from(entry, &memtype_list, nd) { - if (entry->start == start && entry->end == end) { - rb_erase(&entry->rb, &memtype_rbroot); - list_del(&entry->nd); - kfree(entry); - err = 0; - break; - } else if (entry->start > start) { - break; - } - } - - if (!err) - goto unlock_ret; - - entry = saved_entry; - list_for_each_entry_reverse(entry, &memtype_list, nd) { - if (entry->start == start && entry->end == end) { - rb_erase(&entry->rb, &memtype_rbroot); - list_del(&entry->nd); - kfree(entry); - err = 0; - break; - } else if (entry->start < start) { - break; - } - } -unlock_ret: + err = rbt_memtype_erase(start, end); spin_unlock(&memtype_lock); if (err) { @@ -593,7 +410,7 @@ static unsigned long lookup_memtype(u64 paddr) spin_lock(&memtype_lock); - entry = memtype_rb_search(&memtype_rbroot, paddr); + entry = rbt_memtype_lookup(paddr); if (entry != NULL) rettype = entry->type; else @@ -930,22 +747,6 @@ EXPORT_SYMBOL_GPL(pgprot_writecombine); #if defined(CONFIG_DEBUG_FS) && defined(CONFIG_X86_PAT) -/* get Nth element of the linked list */ -static int copy_memtype_nth_element(struct memtype *out, loff_t pos) -{ - struct memtype *list_node; - int i = 1; - - list_for_each_entry(list_node, &memtype_list, nd) { - if (pos == i) { - *out = *list_node; - return 0; - } - ++i; - } - return 1; -} - static struct memtype *memtype_get_idx(loff_t pos) { struct memtype *print_entry; @@ -956,7 +757,7 @@ static struct memtype *memtype_get_idx(loff_t pos) return NULL; spin_lock(&memtype_lock); - ret = copy_memtype_nth_element(print_entry, pos); + ret = rbt_memtype_copy_nth_element(print_entry, pos); spin_unlock(&memtype_lock); if (!ret) { diff --git a/arch/x86/mm/pat_internal.h b/arch/x86/mm/pat_internal.h index 6c98780..4f39eef 100644 --- a/arch/x86/mm/pat_internal.h +++ b/arch/x86/mm/pat_internal.h @@ -9,8 +9,8 @@ extern int pat_debug_enable; struct memtype { u64 start; u64 end; + u64 subtree_max_end; unsigned long type; - struct list_head nd; struct rb_node rb; }; @@ -25,4 +25,22 @@ static inline char *cattr_name(unsigned long flags) } } +#ifdef CONFIG_X86_PAT +extern int rbt_memtype_check_insert(struct memtype *new, + unsigned long *new_type); +extern int rbt_memtype_erase(u64 start, u64 end); +extern struct memtype *rbt_memtype_lookup(u64 addr); +extern int rbt_memtype_copy_nth_element(struct memtype *out, loff_t pos); +#else +static inline int rbt_memtype_check_insert(struct memtype *new, + unsigned long *new_type) +{ return 0; } +static inline int rbt_memtype_erase(u64 start, u64 end) +{ return 0; } +static inline struct memtype *rbt_memtype_lookup(u64 addr) +{ return NULL; } +static inline int rbt_memtype_copy_nth_element(struct memtype *out, loff_t pos) +{ return 0; } +#endif + #endif /* __PAT_INTERNAL_H_ */ diff --git a/arch/x86/mm/pat_rbtree.c b/arch/x86/mm/pat_rbtree.c new file mode 100644 index 0000000..9063f40 --- /dev/null +++ b/arch/x86/mm/pat_rbtree.c @@ -0,0 +1,271 @@ +/* + * Handle caching attributes in page tables (PAT) + * + * Authors: Venkatesh Pallipadi + * Suresh B Siddha + * + * Interval tree (augmented rbtree) used to store the PAT memory type + * reservations. + */ + +#include +#include +#include +#include +#include +#include +#include + +#include +#include + +#include "pat_internal.h" + +/* + * The memtype tree keeps track of memory type for specific + * physical memory areas. Without proper tracking, conflicting memory + * types in different mappings can cause CPU cache corruption. + * + * The tree is an interval tree (augmented rbtree) with tree ordered + * on starting address. Tree can contain multiple entries for + * different regions which overlap. All the aliases have the same + * cache attributes of course. + * + * memtype_lock protects the rbtree. + */ + +static void memtype_rb_augment_cb(struct rb_node *node); +static struct rb_root memtype_rbroot = RB_AUGMENT_ROOT(&memtype_rb_augment_cb); + +static int is_node_overlap(struct memtype *node, u64 start, u64 end) +{ + if (node->start >= end || node->end <= start) + return 0; + + return 1; +} + +static u64 get_subtree_max_end(struct rb_node *node) +{ + u64 ret = 0; + if (node) { + struct memtype *data = container_of(node, struct memtype, rb); + ret = data->subtree_max_end; + } + return ret; +} + +/* Update 'subtree_max_end' for a node, based on node and its children */ +static void update_node_max_end(struct rb_node *node) +{ + struct memtype *data; + u64 max_end, child_max_end; + + if (!node) + return; + + data = container_of(node, struct memtype, rb); + max_end = data->end; + + child_max_end = get_subtree_max_end(node->rb_right); + if (child_max_end > max_end) + max_end = child_max_end; + + child_max_end = get_subtree_max_end(node->rb_left); + if (child_max_end > max_end) + max_end = child_max_end; + + data->subtree_max_end = max_end; +} + +/* Update 'subtree_max_end' for a node and all its ancestors */ +static void update_path_max_end(struct rb_node *node) +{ + u64 old_max_end, new_max_end; + + while (node) { + struct memtype *data = container_of(node, struct memtype, rb); + + old_max_end = data->subtree_max_end; + update_node_max_end(node); + new_max_end = data->subtree_max_end; + + if (new_max_end == old_max_end) + break; + + node = rb_parent(node); + } +} + +/* Find the first (lowest start addr) overlapping range from rb tree */ +static struct memtype *memtype_rb_lowest_match(struct rb_root *root, + u64 start, u64 end) +{ + struct rb_node *node = root->rb_node; + struct memtype *last_lower = NULL; + + while (node) { + struct memtype *data = container_of(node, struct memtype, rb); + + if (get_subtree_max_end(node->rb_left) > start) { + /* Lowest overlap if any must be on left side */ + node = node->rb_left; + } else if (is_node_overlap(data, start, end)) { + last_lower = data; + break; + } else if (start >= data->start) { + /* Lowest overlap if any must be on right side */ + node = node->rb_right; + } else { + break; + } + } + return last_lower; /* Returns NULL if there is no overlap */ +} + +static struct memtype *memtype_rb_exact_match(struct rb_root *root, + u64 start, u64 end) +{ + struct memtype *match; + + match = memtype_rb_lowest_match(root, start, end); + while (match != NULL && match->start < end) { + struct rb_node *node; + + if (match->start == start && match->end == end) + return match; + + node = rb_next(&match->rb); + if (node) + match = container_of(node, struct memtype, rb); + else + match = NULL; + } + + return NULL; /* Returns NULL if there is no exact match */ +} + +static int memtype_rb_check_conflict(struct rb_root *root, + u64 start, u64 end, + unsigned long reqtype, unsigned long *newtype) +{ + struct rb_node *node; + struct memtype *match; + int found_type = reqtype; + + match = memtype_rb_lowest_match(&memtype_rbroot, start, end); + if (match == NULL) + goto success; + + if (match->type != found_type && newtype == NULL) + goto failure; + + dprintk("Overlap at 0x%Lx-0x%Lx\n", match->start, match->end); + found_type = match->type; + + node = rb_next(&match->rb); + while (node) { + match = container_of(node, struct memtype, rb); + + if (match->start >= end) /* Checked all possible matches */ + goto success; + + if (is_node_overlap(match, start, end) && + match->type != found_type) { + goto failure; + } + + node = rb_next(&match->rb); + } +success: + if (newtype) + *newtype = found_type; + + return 0; + +failure: + printk(KERN_INFO "%s:%d conflicting memory types " + "%Lx-%Lx %s<->%s\n", current->comm, current->pid, start, + end, cattr_name(found_type), cattr_name(match->type)); + return -EBUSY; +} + +static void memtype_rb_augment_cb(struct rb_node *node) +{ + if (node) + update_path_max_end(node); +} + +static void memtype_rb_insert(struct rb_root *root, struct memtype *newdata) +{ + struct rb_node **node = &(root->rb_node); + struct rb_node *parent = NULL; + + while (*node) { + struct memtype *data = container_of(*node, struct memtype, rb); + + parent = *node; + if (newdata->start <= data->start) + node = &((*node)->rb_left); + else if (newdata->start > data->start) + node = &((*node)->rb_right); + } + + rb_link_node(&newdata->rb, parent, node); + rb_insert_color(&newdata->rb, root); +} + +int rbt_memtype_check_insert(struct memtype *new, unsigned long *ret_type) +{ + int err = 0; + + err = memtype_rb_check_conflict(&memtype_rbroot, new->start, new->end, + new->type, ret_type); + + if (!err) { + new->type = *ret_type; + memtype_rb_insert(&memtype_rbroot, new); + } + return err; +} + +int rbt_memtype_erase(u64 start, u64 end) +{ + struct memtype *data; + + data = memtype_rb_exact_match(&memtype_rbroot, start, end); + if (!data) + return -EINVAL; + + rb_erase(&data->rb, &memtype_rbroot); + return 0; +} + +struct memtype *rbt_memtype_lookup(u64 addr) +{ + struct memtype *data; + data = memtype_rb_lowest_match(&memtype_rbroot, addr, addr + PAGE_SIZE); + return data; +} + +#if defined(CONFIG_DEBUG_FS) +int rbt_memtype_copy_nth_element(struct memtype *out, loff_t pos) +{ + struct rb_node *node; + int i = 1; + + node = rb_first(&memtype_rbroot); + while (node && pos != i) { + node = rb_next(node); + i++; + } + + if (node) { /* pos == i */ + struct memtype *this = container_of(node, struct memtype, rb); + *out = *this; + return 0; + } else { + return 1; + } +} +#endif -- cgit v0.10.2 From b3ac891b67bd4b1fc728d1c784cad1212dea433d Mon Sep 17 00:00:00 2001 From: Luca Barbieri Date: Wed, 24 Feb 2010 10:54:22 +0100 Subject: x86: Add support for lock prefix in alternatives The current lock prefix UP/SMP alternative code doesn't allow LOCK_PREFIX to be used in alternatives code. This patch solves the problem by adding a new LOCK_PREFIX_ALTERNATIVE_PATCH macro that only records the lock prefix location but does not emit the prefix. The user of this macro can then start any alternative sequence with "lock" and have it UP/SMP patched. To make this work, the UP/SMP alternative code is changed to do the lock/DS prefix switching only if the byte actually contains a lock or DS prefix. Thus, if an alternative without the "lock" is selected, it will now do nothing instead of clobbering the code. Changes in v2: - Naming change - Change label to not conflict with alternatives Signed-off-by: Luca Barbieri LKML-Reference: <1267005265-27958-2-git-send-email-luca@luca-barbieri.com> Signed-off-by: H. Peter Anvin diff --git a/arch/x86/include/asm/alternative.h b/arch/x86/include/asm/alternative.h index 3b5b828..55fee12 100644 --- a/arch/x86/include/asm/alternative.h +++ b/arch/x86/include/asm/alternative.h @@ -28,12 +28,14 @@ */ #ifdef CONFIG_SMP -#define LOCK_PREFIX \ +#define LOCK_PREFIX_HERE \ ".section .smp_locks,\"a\"\n" \ _ASM_ALIGN "\n" \ - _ASM_PTR "661f\n" /* address */ \ + _ASM_PTR "671f\n" /* address */ \ ".previous\n" \ - "661:\n\tlock; " + "671:" + +#define LOCK_PREFIX LOCK_PREFIX_HERE "\n\tlock; " #else /* ! CONFIG_SMP */ #define LOCK_PREFIX "" diff --git a/arch/x86/kernel/alternative.c b/arch/x86/kernel/alternative.c index 2589ea4..80b222e 100644 --- a/arch/x86/kernel/alternative.c +++ b/arch/x86/kernel/alternative.c @@ -244,7 +244,8 @@ static void alternatives_smp_lock(u8 **start, u8 **end, u8 *text, u8 *text_end) if (*ptr > text_end) continue; /* turn DS segment override prefix into lock prefix */ - text_poke(*ptr, ((unsigned char []){0xf0}), 1); + if (**ptr == 0x3e) + text_poke(*ptr, ((unsigned char []){0xf0}), 1); }; mutex_unlock(&text_mutex); } @@ -263,7 +264,8 @@ static void alternatives_smp_unlock(u8 **start, u8 **end, u8 *text, u8 *text_end if (*ptr > text_end) continue; /* turn lock prefix into DS segment override prefix */ - text_poke(*ptr, ((unsigned char []){0x3E}), 1); + if (**ptr == 0xf0) + text_poke(*ptr, ((unsigned char []){0x3E}), 1); }; mutex_unlock(&text_mutex); } -- cgit v0.10.2 From 9c76b38476b18c45f97098a10b0176b321eba3ea Mon Sep 17 00:00:00 2001 From: Luca Barbieri Date: Wed, 24 Feb 2010 10:54:23 +0100 Subject: x86-32: Allow UP/SMP lock replacement in cmpxchg64 Use the functionality just introduced in the previous patch: mark the lock prefixes in cmpxchg64 alternatives for UP removal. Changes in v2: - Naming change Signed-off-by: Luca Barbieri LKML-Reference: <1267005265-27958-3-git-send-email-luca@luca-barbieri.com> Signed-off-by: H. Peter Anvin diff --git a/arch/x86/include/asm/cmpxchg_32.h b/arch/x86/include/asm/cmpxchg_32.h index ffb9bb6..8859e12 100644 --- a/arch/x86/include/asm/cmpxchg_32.h +++ b/arch/x86/include/asm/cmpxchg_32.h @@ -271,7 +271,8 @@ extern unsigned long long cmpxchg_486_u64(volatile void *, u64, u64); __typeof__(*(ptr)) __ret; \ __typeof__(*(ptr)) __old = (o); \ __typeof__(*(ptr)) __new = (n); \ - alternative_io("call cmpxchg8b_emu", \ + alternative_io(LOCK_PREFIX_HERE \ + "call cmpxchg8b_emu", \ "lock; cmpxchg8b (%%esi)" , \ X86_FEATURE_CX8, \ "=A" (__ret), \ -- cgit v0.10.2 From 86a8938078a8bb518c5376de493e348c7490d506 Mon Sep 17 00:00:00 2001 From: Luca Barbieri Date: Wed, 24 Feb 2010 10:54:24 +0100 Subject: lib: Add self-test for atomic64_t This patch adds self-test on boot code for atomic64_t. This has been used to test the later changes in this patchset. Signed-off-by: Luca Barbieri LKML-Reference: <1267005265-27958-4-git-send-email-luca@luca-barbieri.com> Signed-off-by: H. Peter Anvin diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 25c3ed5..3676c51 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -1054,6 +1054,13 @@ config DMA_API_DEBUG This option causes a performance degredation. Use only if you want to debug device drivers. If unsure, say N. +config ATOMIC64_SELFTEST + bool "Perform an atomic64_t self-test at boot" + help + Enable this option to test the atomic64_t functions at boot. + + If unsure, say N. + source "samples/Kconfig" source "lib/Kconfig.kgdb" diff --git a/lib/Makefile b/lib/Makefile index 347ad8d..4af4786 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -99,6 +99,8 @@ obj-$(CONFIG_GENERIC_CSUM) += checksum.o obj-$(CONFIG_GENERIC_ATOMIC64) += atomic64.o +obj-$(CONFIG_ATOMIC64_SELFTEST) += atomic64_test.o + hostprogs-y := gen_crc32table clean-files := crc32table.h diff --git a/lib/atomic64_test.c b/lib/atomic64_test.c new file mode 100644 index 0000000..4ff649e --- /dev/null +++ b/lib/atomic64_test.c @@ -0,0 +1,158 @@ +/* + * Testsuite for atomic64_t functions + * + * Copyright © 2010 Luca Barbieri + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ +#include +#include + +#define INIT(c) do { atomic64_set(&v, c); r = c; } while (0) +static __init int test_atomic64(void) +{ + long long v0 = 0xaaa31337c001d00dLL; + long long v1 = 0xdeadbeefdeafcafeLL; + long long v2 = 0xfaceabadf00df001LL; + long long onestwos = 0x1111111122222222LL; + long long one = 1LL; + + atomic64_t v = ATOMIC64_INIT(v0); + long long r = v0; + BUG_ON(v.counter != r); + + atomic64_set(&v, v1); + r = v1; + BUG_ON(v.counter != r); + BUG_ON(atomic64_read(&v) != r); + + INIT(v0); + atomic64_add(onestwos, &v); + r += onestwos; + BUG_ON(v.counter != r); + + INIT(v0); + atomic64_add(-one, &v); + r += -one; + BUG_ON(v.counter != r); + + INIT(v0); + r += onestwos; + BUG_ON(atomic64_add_return(onestwos, &v) != r); + BUG_ON(v.counter != r); + + INIT(v0); + r += -one; + BUG_ON(atomic64_add_return(-one, &v) != r); + BUG_ON(v.counter != r); + + INIT(v0); + atomic64_sub(onestwos, &v); + r -= onestwos; + BUG_ON(v.counter != r); + + INIT(v0); + atomic64_sub(-one, &v); + r -= -one; + BUG_ON(v.counter != r); + + INIT(v0); + r -= onestwos; + BUG_ON(atomic64_sub_return(onestwos, &v) != r); + BUG_ON(v.counter != r); + + INIT(v0); + r -= -one; + BUG_ON(atomic64_sub_return(-one, &v) != r); + BUG_ON(v.counter != r); + + INIT(v0); + atomic64_inc(&v); + r += one; + BUG_ON(v.counter != r); + + INIT(v0); + r += one; + BUG_ON(atomic64_inc_return(&v) != r); + BUG_ON(v.counter != r); + + INIT(v0); + atomic64_dec(&v); + r -= one; + BUG_ON(v.counter != r); + + INIT(v0); + r -= one; + BUG_ON(atomic64_dec_return(&v) != r); + BUG_ON(v.counter != r); + + INIT(v0); + BUG_ON(atomic64_xchg(&v, v1) != v0); + r = v1; + BUG_ON(v.counter != r); + + INIT(v0); + BUG_ON(atomic64_cmpxchg(&v, v0, v1) != v0); + r = v1; + BUG_ON(v.counter != r); + + INIT(v0); + BUG_ON(atomic64_cmpxchg(&v, v2, v1) != v0); + BUG_ON(v.counter != r); + + INIT(v0); + BUG_ON(!atomic64_add_unless(&v, one, v0)); + BUG_ON(v.counter != r); + + INIT(v0); + BUG_ON(atomic64_add_unless(&v, one, v1)); + r += one; + BUG_ON(v.counter != r); + + INIT(onestwos); + BUG_ON(atomic64_dec_if_positive(&v) != (onestwos - 1)); + r -= one; + BUG_ON(v.counter != r); + + INIT(0); + BUG_ON(atomic64_dec_if_positive(&v) != -one); + BUG_ON(v.counter != r); + + INIT(-one); + BUG_ON(atomic64_dec_if_positive(&v) != (-one - one)); + BUG_ON(v.counter != r); + + INIT(onestwos); + BUG_ON(atomic64_inc_not_zero(&v)); + r += one; + BUG_ON(v.counter != r); + + INIT(0); + BUG_ON(!atomic64_inc_not_zero(&v)); + BUG_ON(v.counter != r); + + INIT(-one); + BUG_ON(atomic64_inc_not_zero(&v)); + r += one; + BUG_ON(v.counter != r); + +#ifdef CONFIG_X86 + printk(KERN_INFO "atomic64 test passed for %s+ platform %s CX8 and %s SSE\n", +#ifdef CONFIG_X86_CMPXCHG64 + "586", +#else + "386", +#endif + boot_cpu_has(X86_FEATURE_CX8) ? "with" : "without", + boot_cpu_has(X86_FEATURE_XMM) ? "with" : "without"); +#else + printk(KERN_INFO "atomic64 test passed\n"); +#endif + + return 0; +} + +core_initcall(test_atomic64); -- cgit v0.10.2 From a7e926abc3adfbd2e5e20d2b46177adb4e313915 Mon Sep 17 00:00:00 2001 From: Luca Barbieri Date: Wed, 24 Feb 2010 10:54:25 +0100 Subject: x86-32: Rewrite 32-bit atomic64 functions in assembly This patch replaces atomic64_32.c with two assembly implementations, one for 386/486 machines using pushf/cli/popf and one for 586+ machines using cmpxchg8b. The cmpxchg8b implementation provides the following advantages over the current one: 1. Implements atomic64_add_unless, atomic64_dec_if_positive and atomic64_inc_not_zero 2. Uses the ZF flag changed by cmpxchg8b instead of doing a comparison 3. Uses custom register calling conventions that reduce or eliminate register moves to suit cmpxchg8b 4. Reads the initial value instead of using cmpxchg8b to do that. Currently we use lock xaddl and movl, which seems the fastest. 5. Does not use the lock prefix for atomic64_set 64-bit writes are already atomic, so we don't need that. We still need it for atomic64_read to avoid restoring a value changed in the meantime. 6. Allocates registers as well or better than gcc The 386 implementation provides support for 386 and 486 machines. 386/486 SMP is not supported (we dropped it), but such support can be added easily if desired. A pure assembly implementation is required due to the custom calling conventions, and desire to use %ebp in atomic64_add_return (we need 7 registers...), as well as the ability to use pushf/popf in the 386 code without an intermediate pop/push. The parameter names are changed to match the convention in atomic_64.h Changes in v3 (due to rebasing to tip/x86/asm): - Patches atomic64_32.h instead of atomic_32.h - Uses the CALL alternative mechanism from commit 1b1d9258181bae199dc940f4bd0298126b9a73d9 Changes in v2: - Merged 386 and cx8 support in the same patch - 386 support now done in assembly, C code no longer used at all - cmpxchg64 is used for atomic64_cmpxchg - stop using macros, use one-line inline functions instead - miscellanous changes and improvements Signed-off-by: Luca Barbieri LKML-Reference: <1267005265-27958-5-git-send-email-luca@luca-barbieri.com> Signed-off-by: H. Peter Anvin diff --git a/arch/x86/include/asm/atomic64_32.h b/arch/x86/include/asm/atomic64_32.h index 03027bf..2a934aa 100644 --- a/arch/x86/include/asm/atomic64_32.h +++ b/arch/x86/include/asm/atomic64_32.h @@ -14,109 +14,193 @@ typedef struct { #define ATOMIC64_INIT(val) { (val) } -extern u64 atomic64_cmpxchg(atomic64_t *ptr, u64 old_val, u64 new_val); +#ifdef CONFIG_X86_CMPXCHG64 +#define ATOMIC64_ALTERNATIVE_(f, g) "call atomic64_" #g "_cx8" +#else +#define ATOMIC64_ALTERNATIVE_(f, g) ALTERNATIVE("call atomic64_" #f "_386", "call atomic64_" #g "_cx8", X86_FEATURE_CX8) +#endif + +#define ATOMIC64_ALTERNATIVE(f) ATOMIC64_ALTERNATIVE_(f, f) + +/** + * atomic64_cmpxchg - cmpxchg atomic64 variable + * @p: pointer to type atomic64_t + * @o: expected value + * @n: new value + * + * Atomically sets @v to @n if it was equal to @o and returns + * the old value. + */ + +static inline long long atomic64_cmpxchg(atomic64_t *v, long long o, long long n) +{ + return cmpxchg64(&v->counter, o, n); +} /** * atomic64_xchg - xchg atomic64 variable - * @ptr: pointer to type atomic64_t - * @new_val: value to assign + * @v: pointer to type atomic64_t + * @n: value to assign * - * Atomically xchgs the value of @ptr to @new_val and returns + * Atomically xchgs the value of @v to @n and returns * the old value. */ -extern u64 atomic64_xchg(atomic64_t *ptr, u64 new_val); +static inline long long atomic64_xchg(atomic64_t *v, long long n) +{ + long long o; + unsigned high = (unsigned)(n >> 32); + unsigned low = (unsigned)n; + asm volatile(ATOMIC64_ALTERNATIVE(xchg) + : "=A" (o), "+b" (low), "+c" (high) + : "S" (v) + : "memory" + ); + return o; +} /** * atomic64_set - set atomic64 variable - * @ptr: pointer to type atomic64_t - * @new_val: value to assign + * @v: pointer to type atomic64_t + * @n: value to assign * - * Atomically sets the value of @ptr to @new_val. + * Atomically sets the value of @v to @n. */ -extern void atomic64_set(atomic64_t *ptr, u64 new_val); +static inline void atomic64_set(atomic64_t *v, long long i) +{ + unsigned high = (unsigned)(i >> 32); + unsigned low = (unsigned)i; + asm volatile(ATOMIC64_ALTERNATIVE(set) + : "+b" (low), "+c" (high) + : "S" (v) + : "eax", "edx", "memory" + ); +} /** * atomic64_read - read atomic64 variable - * @ptr: pointer to type atomic64_t + * @v: pointer to type atomic64_t * - * Atomically reads the value of @ptr and returns it. + * Atomically reads the value of @v and returns it. */ -static inline u64 atomic64_read(atomic64_t *ptr) +static inline long long atomic64_read(atomic64_t *v) { - u64 res; - - /* - * Note, we inline this atomic64_t primitive because - * it only clobbers EAX/EDX and leaves the others - * untouched. We also (somewhat subtly) rely on the - * fact that cmpxchg8b returns the current 64-bit value - * of the memory location we are touching: - */ - asm volatile( - "mov %%ebx, %%eax\n\t" - "mov %%ecx, %%edx\n\t" - LOCK_PREFIX "cmpxchg8b %1\n" - : "=&A" (res) - : "m" (*ptr) - ); - - return res; -} - -extern u64 atomic64_read(atomic64_t *ptr); + long long r; + asm volatile(ATOMIC64_ALTERNATIVE(read) + : "=A" (r), "+c" (v) + : : "memory" + ); + return r; + } /** * atomic64_add_return - add and return - * @delta: integer value to add - * @ptr: pointer to type atomic64_t + * @i: integer value to add + * @v: pointer to type atomic64_t * - * Atomically adds @delta to @ptr and returns @delta + *@ptr + * Atomically adds @i to @v and returns @i + *@v */ -extern u64 atomic64_add_return(u64 delta, atomic64_t *ptr); +static inline long long atomic64_add_return(long long i, atomic64_t *v) +{ + asm volatile(ATOMIC64_ALTERNATIVE(add_return) + : "+A" (i), "+c" (v) + : : "memory" + ); + return i; +} /* * Other variants with different arithmetic operators: */ -extern u64 atomic64_sub_return(u64 delta, atomic64_t *ptr); -extern u64 atomic64_inc_return(atomic64_t *ptr); -extern u64 atomic64_dec_return(atomic64_t *ptr); +static inline long long atomic64_sub_return(long long i, atomic64_t *v) +{ + asm volatile(ATOMIC64_ALTERNATIVE(sub_return) + : "+A" (i), "+c" (v) + : : "memory" + ); + return i; +} + +static inline long long atomic64_inc_return(atomic64_t *v) +{ + long long a; + asm volatile(ATOMIC64_ALTERNATIVE(inc_return) + : "=A" (a) + : "S" (v) + : "memory", "ecx" + ); + return a; +} + +static inline long long atomic64_dec_return(atomic64_t *v) +{ + long long a; + asm volatile(ATOMIC64_ALTERNATIVE(dec_return) + : "=A" (a) + : "S" (v) + : "memory", "ecx" + ); + return a; +} /** * atomic64_add - add integer to atomic64 variable - * @delta: integer value to add - * @ptr: pointer to type atomic64_t + * @i: integer value to add + * @v: pointer to type atomic64_t * - * Atomically adds @delta to @ptr. + * Atomically adds @i to @v. */ -extern void atomic64_add(u64 delta, atomic64_t *ptr); +static inline long long atomic64_add(long long i, atomic64_t *v) +{ + asm volatile(ATOMIC64_ALTERNATIVE_(add, add_return) + : "+A" (i), "+c" (v) + : : "memory" + ); + return i; +} /** * atomic64_sub - subtract the atomic64 variable - * @delta: integer value to subtract - * @ptr: pointer to type atomic64_t + * @i: integer value to subtract + * @v: pointer to type atomic64_t * - * Atomically subtracts @delta from @ptr. + * Atomically subtracts @i from @v. */ -extern void atomic64_sub(u64 delta, atomic64_t *ptr); +static inline long long atomic64_sub(long long i, atomic64_t *v) +{ + asm volatile(ATOMIC64_ALTERNATIVE_(sub, sub_return) + : "+A" (i), "+c" (v) + : : "memory" + ); + return i; +} /** * atomic64_sub_and_test - subtract value from variable and test result - * @delta: integer value to subtract - * @ptr: pointer to type atomic64_t - * - * Atomically subtracts @delta from @ptr and returns + * @i: integer value to subtract + * @v: pointer to type atomic64_t + * + * Atomically subtracts @i from @v and returns * true if the result is zero, or false for all * other cases. */ -extern int atomic64_sub_and_test(u64 delta, atomic64_t *ptr); +static inline int atomic64_sub_and_test(long long i, atomic64_t *v) +{ + return atomic64_sub_return(i, v) == 0; +} /** * atomic64_inc - increment atomic64 variable - * @ptr: pointer to type atomic64_t + * @v: pointer to type atomic64_t * - * Atomically increments @ptr by 1. + * Atomically increments @v by 1. */ -extern void atomic64_inc(atomic64_t *ptr); +static inline void atomic64_inc(atomic64_t *v) +{ + asm volatile(ATOMIC64_ALTERNATIVE_(inc, inc_return) + : : "S" (v) + : "memory", "eax", "ecx", "edx" + ); +} /** * atomic64_dec - decrement atomic64 variable @@ -124,37 +208,97 @@ extern void atomic64_inc(atomic64_t *ptr); * * Atomically decrements @ptr by 1. */ -extern void atomic64_dec(atomic64_t *ptr); +static inline void atomic64_dec(atomic64_t *v) +{ + asm volatile(ATOMIC64_ALTERNATIVE_(dec, dec_return) + : : "S" (v) + : "memory", "eax", "ecx", "edx" + ); +} /** * atomic64_dec_and_test - decrement and test - * @ptr: pointer to type atomic64_t + * @v: pointer to type atomic64_t * - * Atomically decrements @ptr by 1 and + * Atomically decrements @v by 1 and * returns true if the result is 0, or false for all other * cases. */ -extern int atomic64_dec_and_test(atomic64_t *ptr); +static inline int atomic64_dec_and_test(atomic64_t *v) +{ + return atomic64_dec_return(v) == 0; +} /** * atomic64_inc_and_test - increment and test - * @ptr: pointer to type atomic64_t + * @v: pointer to type atomic64_t * - * Atomically increments @ptr by 1 + * Atomically increments @v by 1 * and returns true if the result is zero, or false for all * other cases. */ -extern int atomic64_inc_and_test(atomic64_t *ptr); +static inline int atomic64_inc_and_test(atomic64_t *v) +{ + return atomic64_inc_return(v) == 0; +} /** * atomic64_add_negative - add and test if negative - * @delta: integer value to add - * @ptr: pointer to type atomic64_t + * @i: integer value to add + * @v: pointer to type atomic64_t * - * Atomically adds @delta to @ptr and returns true + * Atomically adds @i to @v and returns true * if the result is negative, or false when * result is greater than or equal to zero. */ -extern int atomic64_add_negative(u64 delta, atomic64_t *ptr); +static inline int atomic64_add_negative(long long i, atomic64_t *v) +{ + return atomic64_add_return(i, v) < 0; +} + +/** + * atomic64_add_unless - add unless the number is a given value + * @v: pointer of type atomic64_t + * @a: the amount to add to v... + * @u: ...unless v is equal to u. + * + * Atomically adds @a to @v, so long as it was not @u. + * Returns non-zero if @v was not @u, and zero otherwise. + */ +static inline int atomic64_add_unless(atomic64_t *v, long long a, long long u) +{ + unsigned low = (unsigned)u; + unsigned high = (unsigned)(u >> 32); + asm volatile(ATOMIC64_ALTERNATIVE(add_unless) "\n\t" + : "+A" (a), "+c" (v), "+S" (low), "+D" (high) + : : "memory"); + return (int)a; +} + + +static inline int atomic64_inc_not_zero(atomic64_t *v) +{ + int r; + asm volatile(ATOMIC64_ALTERNATIVE(inc_not_zero) + : "=a" (r) + : "S" (v) + : "ecx", "edx", "memory" + ); + return r; +} + +static inline long long atomic64_dec_if_positive(atomic64_t *v) +{ + long long r; + asm volatile(ATOMIC64_ALTERNATIVE(dec_if_positive) + : "=A" (r) + : "S" (v) + : "ecx", "memory" + ); + return r; +} + +#undef ATOMIC64_ALTERNATIVE +#undef ATOMIC64_ALTERNATIVE_ #endif /* _ASM_X86_ATOMIC64_32_H */ diff --git a/arch/x86/lib/Makefile b/arch/x86/lib/Makefile index cffd754f..05d686b 100644 --- a/arch/x86/lib/Makefile +++ b/arch/x86/lib/Makefile @@ -26,11 +26,12 @@ obj-y += msr.o msr-reg.o msr-reg-export.o ifeq ($(CONFIG_X86_32),y) obj-y += atomic64_32.o + lib-y += atomic64_cx8_32.o lib-y += checksum_32.o lib-y += strstr_32.o lib-y += semaphore_32.o string_32.o ifneq ($(CONFIG_X86_CMPXCHG64),y) - lib-y += cmpxchg8b_emu.o + lib-y += cmpxchg8b_emu.o atomic64_386_32.o endif lib-$(CONFIG_X86_USE_3DNOW) += mmx_32.o else diff --git a/arch/x86/lib/atomic64_32.c b/arch/x86/lib/atomic64_32.c index 824fa0b..540179e 100644 --- a/arch/x86/lib/atomic64_32.c +++ b/arch/x86/lib/atomic64_32.c @@ -6,225 +6,54 @@ #include #include -static noinline u64 cmpxchg8b(u64 *ptr, u64 old, u64 new) -{ - u32 low = new; - u32 high = new >> 32; - - asm volatile( - LOCK_PREFIX "cmpxchg8b %1\n" - : "+A" (old), "+m" (*ptr) - : "b" (low), "c" (high) - ); - return old; -} - -u64 atomic64_cmpxchg(atomic64_t *ptr, u64 old_val, u64 new_val) -{ - return cmpxchg8b(&ptr->counter, old_val, new_val); -} -EXPORT_SYMBOL(atomic64_cmpxchg); - -/** - * atomic64_xchg - xchg atomic64 variable - * @ptr: pointer to type atomic64_t - * @new_val: value to assign - * - * Atomically xchgs the value of @ptr to @new_val and returns - * the old value. - */ -u64 atomic64_xchg(atomic64_t *ptr, u64 new_val) -{ - /* - * Try first with a (possibly incorrect) assumption about - * what we have there. We'll do two loops most likely, - * but we'll get an ownership MESI transaction straight away - * instead of a read transaction followed by a - * flush-for-ownership transaction: - */ - u64 old_val, real_val = 0; - - do { - old_val = real_val; - - real_val = atomic64_cmpxchg(ptr, old_val, new_val); - - } while (real_val != old_val); - - return old_val; -} -EXPORT_SYMBOL(atomic64_xchg); - -/** - * atomic64_set - set atomic64 variable - * @ptr: pointer to type atomic64_t - * @new_val: value to assign - * - * Atomically sets the value of @ptr to @new_val. - */ -void atomic64_set(atomic64_t *ptr, u64 new_val) -{ - atomic64_xchg(ptr, new_val); -} -EXPORT_SYMBOL(atomic64_set); - -/** -EXPORT_SYMBOL(atomic64_read); - * atomic64_add_return - add and return - * @delta: integer value to add - * @ptr: pointer to type atomic64_t - * - * Atomically adds @delta to @ptr and returns @delta + *@ptr - */ -noinline u64 atomic64_add_return(u64 delta, atomic64_t *ptr) -{ - /* - * Try first with a (possibly incorrect) assumption about - * what we have there. We'll do two loops most likely, - * but we'll get an ownership MESI transaction straight away - * instead of a read transaction followed by a - * flush-for-ownership transaction: - */ - u64 old_val, new_val, real_val = 0; - - do { - old_val = real_val; - new_val = old_val + delta; - - real_val = atomic64_cmpxchg(ptr, old_val, new_val); - - } while (real_val != old_val); - - return new_val; -} -EXPORT_SYMBOL(atomic64_add_return); - -u64 atomic64_sub_return(u64 delta, atomic64_t *ptr) -{ - return atomic64_add_return(-delta, ptr); -} -EXPORT_SYMBOL(atomic64_sub_return); - -u64 atomic64_inc_return(atomic64_t *ptr) -{ - return atomic64_add_return(1, ptr); -} -EXPORT_SYMBOL(atomic64_inc_return); - -u64 atomic64_dec_return(atomic64_t *ptr) -{ - return atomic64_sub_return(1, ptr); -} -EXPORT_SYMBOL(atomic64_dec_return); - -/** - * atomic64_add - add integer to atomic64 variable - * @delta: integer value to add - * @ptr: pointer to type atomic64_t - * - * Atomically adds @delta to @ptr. - */ -void atomic64_add(u64 delta, atomic64_t *ptr) -{ - atomic64_add_return(delta, ptr); -} -EXPORT_SYMBOL(atomic64_add); - -/** - * atomic64_sub - subtract the atomic64 variable - * @delta: integer value to subtract - * @ptr: pointer to type atomic64_t - * - * Atomically subtracts @delta from @ptr. - */ -void atomic64_sub(u64 delta, atomic64_t *ptr) -{ - atomic64_add(-delta, ptr); -} -EXPORT_SYMBOL(atomic64_sub); - -/** - * atomic64_sub_and_test - subtract value from variable and test result - * @delta: integer value to subtract - * @ptr: pointer to type atomic64_t - * - * Atomically subtracts @delta from @ptr and returns - * true if the result is zero, or false for all - * other cases. - */ -int atomic64_sub_and_test(u64 delta, atomic64_t *ptr) -{ - u64 new_val = atomic64_sub_return(delta, ptr); - - return new_val == 0; -} -EXPORT_SYMBOL(atomic64_sub_and_test); - -/** - * atomic64_inc - increment atomic64 variable - * @ptr: pointer to type atomic64_t - * - * Atomically increments @ptr by 1. - */ -void atomic64_inc(atomic64_t *ptr) -{ - atomic64_add(1, ptr); -} -EXPORT_SYMBOL(atomic64_inc); - -/** - * atomic64_dec - decrement atomic64 variable - * @ptr: pointer to type atomic64_t - * - * Atomically decrements @ptr by 1. - */ -void atomic64_dec(atomic64_t *ptr) -{ - atomic64_sub(1, ptr); -} -EXPORT_SYMBOL(atomic64_dec); - -/** - * atomic64_dec_and_test - decrement and test - * @ptr: pointer to type atomic64_t - * - * Atomically decrements @ptr by 1 and - * returns true if the result is 0, or false for all other - * cases. - */ -int atomic64_dec_and_test(atomic64_t *ptr) -{ - return atomic64_sub_and_test(1, ptr); -} -EXPORT_SYMBOL(atomic64_dec_and_test); - -/** - * atomic64_inc_and_test - increment and test - * @ptr: pointer to type atomic64_t - * - * Atomically increments @ptr by 1 - * and returns true if the result is zero, or false for all - * other cases. - */ -int atomic64_inc_and_test(atomic64_t *ptr) -{ - return atomic64_sub_and_test(-1, ptr); -} -EXPORT_SYMBOL(atomic64_inc_and_test); - -/** - * atomic64_add_negative - add and test if negative - * @delta: integer value to add - * @ptr: pointer to type atomic64_t - * - * Atomically adds @delta to @ptr and returns true - * if the result is negative, or false when - * result is greater than or equal to zero. - */ -int atomic64_add_negative(u64 delta, atomic64_t *ptr) -{ - s64 new_val = atomic64_add_return(delta, ptr); - - return new_val < 0; -} -EXPORT_SYMBOL(atomic64_add_negative); +long long atomic64_read_cx8(long long, const atomic64_t *v); +EXPORT_SYMBOL(atomic64_read_cx8); +long long atomic64_set_cx8(long long, const atomic64_t *v); +EXPORT_SYMBOL(atomic64_set_cx8); +long long atomic64_xchg_cx8(long long, unsigned high); +EXPORT_SYMBOL(atomic64_xchg_cx8); +long long atomic64_add_return_cx8(long long a, atomic64_t *v); +EXPORT_SYMBOL(atomic64_add_return_cx8); +long long atomic64_sub_return_cx8(long long a, atomic64_t *v); +EXPORT_SYMBOL(atomic64_sub_return_cx8); +long long atomic64_inc_return_cx8(long long a, atomic64_t *v); +EXPORT_SYMBOL(atomic64_inc_return_cx8); +long long atomic64_dec_return_cx8(long long a, atomic64_t *v); +EXPORT_SYMBOL(atomic64_dec_return_cx8); +long long atomic64_dec_if_positive_cx8(atomic64_t *v); +EXPORT_SYMBOL(atomic64_dec_if_positive_cx8); +int atomic64_inc_not_zero_cx8(atomic64_t *v); +EXPORT_SYMBOL(atomic64_inc_not_zero_cx8); +int atomic64_add_unless_cx8(atomic64_t *v, long long a, long long u); +EXPORT_SYMBOL(atomic64_add_unless_cx8); + +#ifndef CONFIG_X86_CMPXCHG64 +long long atomic64_read_386(long long, const atomic64_t *v); +EXPORT_SYMBOL(atomic64_read_386); +long long atomic64_set_386(long long, const atomic64_t *v); +EXPORT_SYMBOL(atomic64_set_386); +long long atomic64_xchg_386(long long, unsigned high); +EXPORT_SYMBOL(atomic64_xchg_386); +long long atomic64_add_return_386(long long a, atomic64_t *v); +EXPORT_SYMBOL(atomic64_add_return_386); +long long atomic64_sub_return_386(long long a, atomic64_t *v); +EXPORT_SYMBOL(atomic64_sub_return_386); +long long atomic64_inc_return_386(long long a, atomic64_t *v); +EXPORT_SYMBOL(atomic64_inc_return_386); +long long atomic64_dec_return_386(long long a, atomic64_t *v); +EXPORT_SYMBOL(atomic64_dec_return_386); +long long atomic64_add_386(long long a, atomic64_t *v); +EXPORT_SYMBOL(atomic64_add_386); +long long atomic64_sub_386(long long a, atomic64_t *v); +EXPORT_SYMBOL(atomic64_sub_386); +long long atomic64_inc_386(long long a, atomic64_t *v); +EXPORT_SYMBOL(atomic64_inc_386); +long long atomic64_dec_386(long long a, atomic64_t *v); +EXPORT_SYMBOL(atomic64_dec_386); +long long atomic64_dec_if_positive_386(atomic64_t *v); +EXPORT_SYMBOL(atomic64_dec_if_positive_386); +int atomic64_inc_not_zero_386(atomic64_t *v); +EXPORT_SYMBOL(atomic64_inc_not_zero_386); +int atomic64_add_unless_386(atomic64_t *v, long long a, long long u); +EXPORT_SYMBOL(atomic64_add_unless_386); +#endif diff --git a/arch/x86/lib/atomic64_386_32.S b/arch/x86/lib/atomic64_386_32.S new file mode 100644 index 0000000..5db07fe --- /dev/null +++ b/arch/x86/lib/atomic64_386_32.S @@ -0,0 +1,175 @@ +/* + * atomic64_t for 386/486 + * + * Copyright © 2010 Luca Barbieri + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include +#include +#include + +/* if you want SMP support, implement these with real spinlocks */ +.macro LOCK reg + pushfl + CFI_ADJUST_CFA_OFFSET 4 + cli +.endm + +.macro UNLOCK reg + popfl + CFI_ADJUST_CFA_OFFSET -4 +.endm + +.macro BEGIN func reg +$v = \reg + +ENTRY(atomic64_\func\()_386) + CFI_STARTPROC + LOCK $v + +.macro RETURN + UNLOCK $v + ret +.endm + +.macro END_ + CFI_ENDPROC +ENDPROC(atomic64_\func\()_386) +.purgem RETURN +.purgem END_ +.purgem END +.endm + +.macro END +RETURN +END_ +.endm +.endm + +BEGIN read %ecx + movl ($v), %eax + movl 4($v), %edx +END + +BEGIN set %esi + movl %ebx, ($v) + movl %ecx, 4($v) +END + +BEGIN xchg %esi + movl ($v), %eax + movl 4($v), %edx + movl %ebx, ($v) + movl %ecx, 4($v) +END + +BEGIN add %ecx + addl %eax, ($v) + adcl %edx, 4($v) +END + +BEGIN add_return %ecx + addl ($v), %eax + adcl 4($v), %edx + movl %eax, ($v) + movl %edx, 4($v) +END + +BEGIN sub %ecx + subl %eax, ($v) + sbbl %edx, 4($v) +END + +BEGIN sub_return %ecx + negl %edx + negl %eax + sbbl $0, %edx + addl ($v), %eax + adcl 4($v), %edx + movl %eax, ($v) + movl %edx, 4($v) +END + +BEGIN inc %esi + addl $1, ($v) + adcl $0, 4($v) +END + +BEGIN inc_return %esi + movl ($v), %eax + movl 4($v), %edx + addl $1, %eax + adcl $0, %edx + movl %eax, ($v) + movl %edx, 4($v) +END + +BEGIN dec %esi + subl $1, ($v) + sbbl $0, 4($v) +END + +BEGIN dec_return %esi + movl ($v), %eax + movl 4($v), %edx + subl $1, %eax + sbbl $0, %edx + movl %eax, ($v) + movl %edx, 4($v) +END + +BEGIN add_unless %ecx + addl %eax, %esi + adcl %edx, %edi + addl ($v), %eax + adcl 4($v), %edx + cmpl %eax, %esi + je 3f +1: + movl %eax, ($v) + movl %edx, 4($v) + xorl %eax, %eax +2: +RETURN +3: + cmpl %edx, %edi + jne 1b + movl $1, %eax + jmp 2b +END_ + +BEGIN inc_not_zero %esi + movl ($v), %eax + movl 4($v), %edx + testl %eax, %eax + je 3f +1: + addl $1, %eax + adcl $0, %edx + movl %eax, ($v) + movl %edx, 4($v) + xorl %eax, %eax +2: +RETURN +3: + testl %edx, %edx + jne 1b + movl $1, %eax + jmp 2b +END_ + +BEGIN dec_if_positive %esi + movl ($v), %eax + movl 4($v), %edx + subl $1, %eax + sbbl $0, %edx + js 1f + movl %eax, ($v) + movl %edx, 4($v) +1: +END diff --git a/arch/x86/lib/atomic64_cx8_32.S b/arch/x86/lib/atomic64_cx8_32.S new file mode 100644 index 0000000..e49c4eb --- /dev/null +++ b/arch/x86/lib/atomic64_cx8_32.S @@ -0,0 +1,225 @@ +/* + * atomic64_t for 586+ + * + * Copyright © 2010 Luca Barbieri + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include +#include +#include + +.macro SAVE reg + pushl %\reg + CFI_ADJUST_CFA_OFFSET 4 + CFI_REL_OFFSET \reg, 0 +.endm + +.macro RESTORE reg + popl %\reg + CFI_ADJUST_CFA_OFFSET -4 + CFI_RESTORE \reg +.endm + +.macro read64 reg + movl %ebx, %eax + movl %ecx, %edx +/* we need LOCK_PREFIX since otherwise cmpxchg8b always does the write */ + LOCK_PREFIX + cmpxchg8b (\reg) +.endm + +ENTRY(atomic64_read_cx8) + CFI_STARTPROC + + read64 %ecx + ret + CFI_ENDPROC +ENDPROC(atomic64_read_cx8) + +ENTRY(atomic64_set_cx8) + CFI_STARTPROC + +1: +/* we don't need LOCK_PREFIX since aligned 64-bit writes + * are atomic on 586 and newer */ + cmpxchg8b (%esi) + jne 1b + + ret + CFI_ENDPROC +ENDPROC(atomic64_set_cx8) + +ENTRY(atomic64_xchg_cx8) + CFI_STARTPROC + + movl %ebx, %eax + movl %ecx, %edx +1: + LOCK_PREFIX + cmpxchg8b (%esi) + jne 1b + + ret + CFI_ENDPROC +ENDPROC(atomic64_xchg_cx8) + +.macro addsub_return func ins insc +ENTRY(atomic64_\func\()_return_cx8) + CFI_STARTPROC + SAVE ebp + SAVE ebx + SAVE esi + SAVE edi + + movl %eax, %esi + movl %edx, %edi + movl %ecx, %ebp + + read64 %ebp +1: + movl %eax, %ebx + movl %edx, %ecx + \ins\()l %esi, %ebx + \insc\()l %edi, %ecx + LOCK_PREFIX + cmpxchg8b (%ebp) + jne 1b + +10: + movl %ebx, %eax + movl %ecx, %edx + RESTORE edi + RESTORE esi + RESTORE ebx + RESTORE ebp + ret + CFI_ENDPROC +ENDPROC(atomic64_\func\()_return_cx8) +.endm + +addsub_return add add adc +addsub_return sub sub sbb + +.macro incdec_return func ins insc +ENTRY(atomic64_\func\()_return_cx8) + CFI_STARTPROC + SAVE ebx + + read64 %esi +1: + movl %eax, %ebx + movl %edx, %ecx + \ins\()l $1, %ebx + \insc\()l $0, %ecx + LOCK_PREFIX + cmpxchg8b (%esi) + jne 1b + +10: + movl %ebx, %eax + movl %ecx, %edx + RESTORE ebx + ret + CFI_ENDPROC +ENDPROC(atomic64_\func\()_return_cx8) +.endm + +incdec_return inc add adc +incdec_return dec sub sbb + +ENTRY(atomic64_dec_if_positive_cx8) + CFI_STARTPROC + SAVE ebx + + read64 %esi +1: + movl %eax, %ebx + movl %edx, %ecx + subl $1, %ebx + sbb $0, %ecx + js 2f + LOCK_PREFIX + cmpxchg8b (%esi) + jne 1b + +2: + movl %ebx, %eax + movl %ecx, %edx + RESTORE ebx + ret + CFI_ENDPROC +ENDPROC(atomic64_dec_if_positive_cx8) + +ENTRY(atomic64_add_unless_cx8) + CFI_STARTPROC + SAVE ebp + SAVE ebx +/* these just push these two parameters on the stack */ + SAVE edi + SAVE esi + + movl %ecx, %ebp + movl %eax, %esi + movl %edx, %edi + + read64 %ebp +1: + cmpl %eax, 0(%esp) + je 4f +2: + movl %eax, %ebx + movl %edx, %ecx + addl %esi, %ebx + adcl %edi, %ecx + LOCK_PREFIX + cmpxchg8b (%ebp) + jne 1b + + xorl %eax, %eax +3: + addl $8, %esp + CFI_ADJUST_CFA_OFFSET -8 + RESTORE ebx + RESTORE ebp + ret +4: + cmpl %edx, 4(%esp) + jne 2b + movl $1, %eax + jmp 3b + CFI_ENDPROC +ENDPROC(atomic64_add_unless_cx8) + +ENTRY(atomic64_inc_not_zero_cx8) + CFI_STARTPROC + SAVE ebx + + read64 %esi +1: + testl %eax, %eax + je 4f +2: + movl %eax, %ebx + movl %edx, %ecx + addl $1, %ebx + adcl $0, %ecx + LOCK_PREFIX + cmpxchg8b (%esi) + jne 1b + + xorl %eax, %eax +3: + RESTORE ebx + ret +4: + testl %edx, %edx + jne 2b + movl $1, %eax + jmp 3b + CFI_ENDPROC +ENDPROC(atomic64_inc_not_zero_cx8) -- cgit v0.10.2 From 339d3261aa3eb0e12f68ef868e042c1ca03628f7 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Sat, 6 Feb 2010 09:42:39 +0100 Subject: x86/amd-iommu: Remove double NULL check in check_device dev was tested just above, so drop the second test. Signed-off-by: Julia Lawall Signed-off-by: Joerg Roedel diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index adb0ba0..2c4a501 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -118,7 +118,7 @@ static bool check_device(struct device *dev) return false; /* No device or no PCI device */ - if (!dev || dev->bus != &pci_bus_type) + if (dev->bus != &pci_bus_type) return false; devid = get_device_id(dev); -- cgit v0.10.2 From 5d214fe6e808a8caa9cb6f610c0190d3f50ac570 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Mon, 8 Feb 2010 14:44:49 +0100 Subject: x86/amd-iommu: Protect IOMMU-API map/unmap path This patch introduces a mutex to lock page table updates in the IOMMU-API path. We can't use the spin_lock here because this patch might sleep. Signed-off-by: Joerg Roedel diff --git a/arch/x86/include/asm/amd_iommu_types.h b/arch/x86/include/asm/amd_iommu_types.h index ba19ad4..5e46e78 100644 --- a/arch/x86/include/asm/amd_iommu_types.h +++ b/arch/x86/include/asm/amd_iommu_types.h @@ -21,6 +21,7 @@ #define _ASM_X86_AMD_IOMMU_TYPES_H #include +#include #include #include @@ -237,6 +238,7 @@ struct protection_domain { struct list_head list; /* for list of all protection domains */ struct list_head dev_list; /* List of all devices in this domain */ spinlock_t lock; /* mostly used to lock the page table*/ + struct mutex api_lock; /* protect page tables in the iommu-api path */ u16 id; /* the domain id written to the device table */ int mode; /* paging mode (0-6 levels) */ u64 *pt_root; /* page table root pointer */ diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 2c4a501..b97f2f1 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -2327,6 +2327,7 @@ static struct protection_domain *protection_domain_alloc(void) return NULL; spin_lock_init(&domain->lock); + mutex_init(&domain->api_lock); domain->id = domain_id_alloc(); if (!domain->id) goto out_err; @@ -2456,6 +2457,8 @@ static int amd_iommu_map_range(struct iommu_domain *dom, iova &= PAGE_MASK; paddr &= PAGE_MASK; + mutex_lock(&domain->api_lock); + for (i = 0; i < npages; ++i) { ret = iommu_map_page(domain, iova, paddr, prot, PM_MAP_4k); if (ret) @@ -2465,6 +2468,8 @@ static int amd_iommu_map_range(struct iommu_domain *dom, paddr += PAGE_SIZE; } + mutex_unlock(&domain->api_lock); + return 0; } @@ -2477,12 +2482,16 @@ static void amd_iommu_unmap_range(struct iommu_domain *dom, iova &= PAGE_MASK; + mutex_lock(&domain->api_lock); + for (i = 0; i < npages; ++i) { iommu_unmap_page(domain, iova, PM_MAP_4k); iova += PAGE_SIZE; } iommu_flush_tlb_pde(domain); + + mutex_unlock(&domain->api_lock); } static phys_addr_t amd_iommu_iova_to_phys(struct iommu_domain *dom, -- cgit v0.10.2 From 04e856c072b84042bb56c487c2868638bb3f78db Mon Sep 17 00:00:00 2001 From: Chris Wright Date: Wed, 17 Feb 2010 08:51:20 -0800 Subject: x86/amd-iommu: Pt mode fix for domain_destroy After a guest is shutdown, assigned devices are not properly returned to the pt domain. This can leave the device using stale cached IOMMU data, and result in a non-functional device after it's re-bound to the host driver. For example, I see this upon rebinding: AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0000 address=0x000000007e2a8000 flags=0x0050] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0000 address=0x000000007e2a8040 flags=0x0050] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0000 address=0x000000007e2a8080 flags=0x0050] AMD-Vi: Event logged [IO_PAGE_FAULT device=02:00.0 domain=0x0000 address=0x000000007e2a80c0 flags=0x0050] 0000:02:00.0: eth2: Detected Hardware Unit Hang: ... The amd_iommu_destroy_domain() function calls do_detach() which doesn't reattach the pt domain to the device. Use __detach_device() instead. Cc: stable@kernel.org Signed-off-by: Chris Wright Signed-off-by: Joerg Roedel diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index b97f2f1..0c04254 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -2298,7 +2298,7 @@ static void cleanup_domain(struct protection_domain *domain) list_for_each_entry_safe(dev_data, next, &domain->dev_list, list) { struct device *dev = dev_data->dev; - do_detach(dev); + __detach_device(dev); atomic_set(&dev_data->bind, 0); } -- cgit v0.10.2 From 3551a708f35fc712af43aeb7f541512c5cfc4936 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Mon, 1 Mar 2010 13:52:19 +0100 Subject: x86/amd-iommu: Report errors in acpi parsing functions upstream Since acpi_table_parse ignores the return values of the parsing function this patch introduces a workaround and reports these errors upstream via a global variable. Signed-off-by: Joerg Roedel diff --git a/arch/x86/kernel/amd_iommu_init.c b/arch/x86/kernel/amd_iommu_init.c index 9dc91b4..feaf471 100644 --- a/arch/x86/kernel/amd_iommu_init.c +++ b/arch/x86/kernel/amd_iommu_init.c @@ -138,9 +138,9 @@ int amd_iommus_present; bool amd_iommu_np_cache __read_mostly; /* - * Set to true if ACPI table parsing and hardware intialization went properly + * The ACPI table parsing functions set this variable on an error */ -static bool amd_iommu_initialized; +static int __initdata amd_iommu_init_err; /* * List of protection domains - used during resume @@ -391,9 +391,11 @@ static int __init find_last_devid_acpi(struct acpi_table_header *table) */ for (i = 0; i < table->length; ++i) checksum += p[i]; - if (checksum != 0) + if (checksum != 0) { /* ACPI table corrupt */ - return -ENODEV; + amd_iommu_init_err = -ENODEV; + return 0; + } p += IVRS_HEADER_LENGTH; @@ -920,11 +922,16 @@ static int __init init_iommu_all(struct acpi_table_header *table) h->mmio_phys); iommu = kzalloc(sizeof(struct amd_iommu), GFP_KERNEL); - if (iommu == NULL) - return -ENOMEM; + if (iommu == NULL) { + amd_iommu_init_err = -ENOMEM; + return 0; + } + ret = init_iommu_one(iommu, h); - if (ret) - return ret; + if (ret) { + amd_iommu_init_err = ret; + return 0; + } break; default: break; @@ -934,8 +941,6 @@ static int __init init_iommu_all(struct acpi_table_header *table) } WARN_ON(p != end); - amd_iommu_initialized = true; - return 0; } @@ -1211,6 +1216,10 @@ static int __init amd_iommu_init(void) if (acpi_table_parse("IVRS", find_last_devid_acpi) != 0) return -ENODEV; + ret = amd_iommu_init_err; + if (ret) + goto out; + dev_table_size = tbl_size(DEV_TABLE_ENTRY_SIZE); alias_table_size = tbl_size(ALIAS_TABLE_ENTRY_SIZE); rlookup_table_size = tbl_size(RLOOKUP_TABLE_ENTRY_SIZE); @@ -1270,12 +1279,19 @@ static int __init amd_iommu_init(void) if (acpi_table_parse("IVRS", init_iommu_all) != 0) goto free; - if (!amd_iommu_initialized) + if (amd_iommu_init_err) { + ret = amd_iommu_init_err; goto free; + } if (acpi_table_parse("IVRS", init_memory_definitions) != 0) goto free; + if (amd_iommu_init_err) { + ret = amd_iommu_init_err; + goto free; + } + ret = sysdev_class_register(&amd_iommu_sysdev_class); if (ret) goto free; -- cgit v0.10.2 From 8f4f202b335144bf5be5c9e5b1bc9477ecdae958 Mon Sep 17 00:00:00 2001 From: Luca Barbieri Date: Fri, 26 Feb 2010 12:22:40 +0100 Subject: lib: Only test atomic64_dec_if_positive on archs having it Currently atomic64_dec_if_positive() is only supported by PowerPC, MIPS and x86-32. Signed-off-by: Luca Barbieri LKML-Reference: <1267183361-20775-1-git-send-email-luca@luca-barbieri.com> Signed-off-by: H. Peter Anvin diff --git a/lib/atomic64_test.c b/lib/atomic64_test.c index 4ff649e..0effcac 100644 --- a/lib/atomic64_test.c +++ b/lib/atomic64_test.c @@ -112,6 +112,7 @@ static __init int test_atomic64(void) r += one; BUG_ON(v.counter != r); +#if defined(CONFIG_X86_32) || defined(CONFIG_MIPS) || defined(CONFIG_PPC) || defined(_ASM_GENERIC_ATOMIC64_H) INIT(onestwos); BUG_ON(atomic64_dec_if_positive(&v) != (onestwos - 1)); r -= one; @@ -124,6 +125,9 @@ static __init int test_atomic64(void) INIT(-one); BUG_ON(atomic64_dec_if_positive(&v) != (-one - one)); BUG_ON(v.counter != r); +#else +#warning Please implement atomic64_dec_if_positive for your architecture, and add it to the IF above +#endif INIT(onestwos); BUG_ON(atomic64_inc_not_zero(&v)); -- cgit v0.10.2 From d7f6de1e9c4a12e11ba7186c70f0f40caa76f590 Mon Sep 17 00:00:00 2001 From: Luca Barbieri Date: Fri, 26 Feb 2010 12:22:41 +0100 Subject: x86: Implement atomic[64]_dec_if_positive() Add support for atomic_dec_if_positive(), and atomic64_dec_if_positive() for x86-64. atomic64_dec_if_positive() for x86-32 was already implemented in a previous patch. Signed-off-by: Luca Barbieri LKML-Reference: <1267183361-20775-2-git-send-email-luca@luca-barbieri.com> Signed-off-by: H. Peter Anvin diff --git a/arch/x86/include/asm/atomic.h b/arch/x86/include/asm/atomic.h index 8f8217b..706c694 100644 --- a/arch/x86/include/asm/atomic.h +++ b/arch/x86/include/asm/atomic.h @@ -246,6 +246,29 @@ static inline int atomic_add_unless(atomic_t *v, int a, int u) #define atomic_inc_not_zero(v) atomic_add_unless((v), 1, 0) +/* + * atomic_dec_if_positive - decrement by 1 if old value positive + * @v: pointer of type atomic_t + * + * The function returns the old value of *v minus 1, even if + * the atomic variable, v, was not decremented. + */ +static inline int atomic_dec_if_positive(atomic_t *v) +{ + int c, old, dec; + c = atomic_read(v); + for (;;) { + dec = c - 1; + if (unlikely(dec < 0)) + break; + old = atomic_cmpxchg((v), c, dec); + if (likely(old == c)) + break; + c = old; + } + return dec; +} + /** * atomic_inc_short - increment of a short integer * @v: pointer to type int diff --git a/arch/x86/include/asm/atomic64_64.h b/arch/x86/include/asm/atomic64_64.h index 51c5b40..4d6e2cd 100644 --- a/arch/x86/include/asm/atomic64_64.h +++ b/arch/x86/include/asm/atomic64_64.h @@ -221,4 +221,27 @@ static inline int atomic64_add_unless(atomic64_t *v, long a, long u) #define atomic64_inc_not_zero(v) atomic64_add_unless((v), 1, 0) +/* + * atomic64_dec_if_positive - decrement by 1 if old value positive + * @v: pointer of type atomic_t + * + * The function returns the old value of *v minus 1, even if + * the atomic variable, v, was not decremented. + */ +static inline long atomic64_dec_if_positive(atomic64_t *v) +{ + long c, old, dec; + c = atomic64_read(v); + for (;;) { + dec = c - 1; + if (unlikely(dec < 0)) + break; + old = atomic64_cmpxchg((v), c, dec); + if (likely(old == c)) + break; + c = old; + } + return dec; +} + #endif /* _ASM_X86_ATOMIC64_64_H */ diff --git a/lib/atomic64_test.c b/lib/atomic64_test.c index 0effcac..58efdab 100644 --- a/lib/atomic64_test.c +++ b/lib/atomic64_test.c @@ -112,7 +112,7 @@ static __init int test_atomic64(void) r += one; BUG_ON(v.counter != r); -#if defined(CONFIG_X86_32) || defined(CONFIG_MIPS) || defined(CONFIG_PPC) || defined(_ASM_GENERIC_ATOMIC64_H) +#if defined(CONFIG_X86) || defined(CONFIG_MIPS) || defined(CONFIG_PPC) || defined(_ASM_GENERIC_ATOMIC64_H) INIT(onestwos); BUG_ON(atomic64_dec_if_positive(&v) != (onestwos - 1)); r -= one; -- cgit v0.10.2 From 9efbcd590243045111670c171a951923b877b57d Mon Sep 17 00:00:00 2001 From: Luca Barbieri Date: Mon, 1 Mar 2010 19:55:45 +0100 Subject: lib: Fix atomic64_add_unless test atomic64_add_unless must return 1 if it perfomed the add and 0 otherwise. The test assumed the opposite convention. Reported-by: H. Peter Anvin Signed-off-by: Luca Barbieri LKML-Reference: <1267469749-11878-2-git-send-email-luca@luca-barbieri.com> Signed-off-by: H. Peter Anvin diff --git a/lib/atomic64_test.c b/lib/atomic64_test.c index 58efdab..ee8e6de 100644 --- a/lib/atomic64_test.c +++ b/lib/atomic64_test.c @@ -104,11 +104,11 @@ static __init int test_atomic64(void) BUG_ON(v.counter != r); INIT(v0); - BUG_ON(!atomic64_add_unless(&v, one, v0)); + BUG_ON(atomic64_add_unless(&v, one, v0)); BUG_ON(v.counter != r); INIT(v0); - BUG_ON(atomic64_add_unless(&v, one, v1)); + BUG_ON(!atomic64_add_unless(&v, one, v1)); r += one; BUG_ON(v.counter != r); -- cgit v0.10.2 From 6e6104fe085026e6ef82cc5cc303d6c8ceb7e411 Mon Sep 17 00:00:00 2001 From: Luca Barbieri Date: Mon, 1 Mar 2010 19:55:46 +0100 Subject: x86-32: Fix atomic64_add_unless return value convention atomic64_add_unless must return 1 if it perfomed the add and 0 otherwise. The implementation did the opposite thing. Reported-by: H. Peter Anvin Signed-off-by: Luca Barbieri LKML-Reference: <1267469749-11878-3-git-send-email-luca@luca-barbieri.com> Signed-off-by: H. Peter Anvin diff --git a/arch/x86/lib/atomic64_386_32.S b/arch/x86/lib/atomic64_386_32.S index 5db07fe..a2f847c 100644 --- a/arch/x86/lib/atomic64_386_32.S +++ b/arch/x86/lib/atomic64_386_32.S @@ -133,13 +133,13 @@ BEGIN add_unless %ecx 1: movl %eax, ($v) movl %edx, 4($v) - xorl %eax, %eax + movl $1, %eax 2: RETURN 3: cmpl %edx, %edi jne 1b - movl $1, %eax + xorl %eax, %eax jmp 2b END_ diff --git a/arch/x86/lib/atomic64_cx8_32.S b/arch/x86/lib/atomic64_cx8_32.S index e49c4eb..d0e37b1 100644 --- a/arch/x86/lib/atomic64_cx8_32.S +++ b/arch/x86/lib/atomic64_cx8_32.S @@ -180,7 +180,7 @@ ENTRY(atomic64_add_unless_cx8) cmpxchg8b (%ebp) jne 1b - xorl %eax, %eax + movl $1, %eax 3: addl $8, %esp CFI_ADJUST_CFA_OFFSET -8 @@ -190,7 +190,7 @@ ENTRY(atomic64_add_unless_cx8) 4: cmpl %edx, 4(%esp) jne 2b - movl $1, %eax + xorl %eax, %eax jmp 3b CFI_ENDPROC ENDPROC(atomic64_add_unless_cx8) -- cgit v0.10.2 From 97577896f6b9c056fa0a5e9f6a608110cb3dcd33 Mon Sep 17 00:00:00 2001 From: Luca Barbieri Date: Mon, 1 Mar 2010 19:55:47 +0100 Subject: lib: Fix atomic64_add_unless return value convention atomic64_add_unless must return 1 if it perfomed the add and 0 otherwise. The generic implementation did the opposite thing. Reported-by: H. Peter Anvin Confirmed-by: Paul Mackerras Signed-off-by: Luca Barbieri LKML-Reference: <1267469749-11878-4-git-send-email-luca@luca-barbieri.com> Signed-off-by: H. Peter Anvin diff --git a/lib/atomic64.c b/lib/atomic64.c index 8bee16e..a21c12b 100644 --- a/lib/atomic64.c +++ b/lib/atomic64.c @@ -162,12 +162,12 @@ int atomic64_add_unless(atomic64_t *v, long long a, long long u) { unsigned long flags; spinlock_t *lock = lock_addr(v); - int ret = 1; + int ret = 0; spin_lock_irqsave(lock, flags); if (v->counter != u) { v->counter += a; - ret = 0; + ret = 1; } spin_unlock_irqrestore(lock, flags); return ret; -- cgit v0.10.2 From 25a304f277ad70166eeae25a4958d2049005c33a Mon Sep 17 00:00:00 2001 From: Luca Barbieri Date: Mon, 1 Mar 2010 19:55:48 +0100 Subject: lib: Fix atomic64_inc_not_zero test atomic64_inc_not_zero must return 1 if it perfomed the add and 0 otherwise. The test assumed the opposite convention. Signed-off-by: Luca Barbieri LKML-Reference: <1267469749-11878-5-git-send-email-luca@luca-barbieri.com> Signed-off-by: H. Peter Anvin diff --git a/lib/atomic64_test.c b/lib/atomic64_test.c index ee8e6de..f7bb706 100644 --- a/lib/atomic64_test.c +++ b/lib/atomic64_test.c @@ -130,16 +130,16 @@ static __init int test_atomic64(void) #endif INIT(onestwos); - BUG_ON(atomic64_inc_not_zero(&v)); + BUG_ON(!atomic64_inc_not_zero(&v)); r += one; BUG_ON(v.counter != r); INIT(0); - BUG_ON(!atomic64_inc_not_zero(&v)); + BUG_ON(atomic64_inc_not_zero(&v)); BUG_ON(v.counter != r); INIT(-one); - BUG_ON(atomic64_inc_not_zero(&v)); + BUG_ON(!atomic64_inc_not_zero(&v)); r += one; BUG_ON(v.counter != r); -- cgit v0.10.2 From f3e83131469e29032a700217aa394996107b8fc5 Mon Sep 17 00:00:00 2001 From: Luca Barbieri Date: Mon, 1 Mar 2010 19:55:49 +0100 Subject: x86-32: Fix atomic64_inc_not_zero return value convention atomic64_inc_not_zero must return 1 if it perfomed the add and 0 otherwise. It was doing the opposite thing. Signed-off-by: Luca Barbieri LKML-Reference: <1267469749-11878-6-git-send-email-luca@luca-barbieri.com> Signed-off-by: H. Peter Anvin diff --git a/arch/x86/lib/atomic64_386_32.S b/arch/x86/lib/atomic64_386_32.S index a2f847c..4a5979a 100644 --- a/arch/x86/lib/atomic64_386_32.S +++ b/arch/x86/lib/atomic64_386_32.S @@ -153,13 +153,12 @@ BEGIN inc_not_zero %esi adcl $0, %edx movl %eax, ($v) movl %edx, 4($v) - xorl %eax, %eax + movl $1, %eax 2: RETURN 3: testl %edx, %edx jne 1b - movl $1, %eax jmp 2b END_ diff --git a/arch/x86/lib/atomic64_cx8_32.S b/arch/x86/lib/atomic64_cx8_32.S index d0e37b1..71e080d 100644 --- a/arch/x86/lib/atomic64_cx8_32.S +++ b/arch/x86/lib/atomic64_cx8_32.S @@ -212,14 +212,13 @@ ENTRY(atomic64_inc_not_zero_cx8) cmpxchg8b (%esi) jne 1b - xorl %eax, %eax + movl $1, %eax 3: RESTORE ebx ret 4: testl %edx, %edx jne 2b - movl $1, %eax jmp 3b CFI_ENDPROC ENDPROC(atomic64_inc_not_zero_cx8) -- cgit v0.10.2 From a5c9161f27c3e1ae6c0094d262f03a7e98262181 Mon Sep 17 00:00:00 2001 From: "H. Peter Anvin" Date: Mon, 1 Mar 2010 11:49:23 -0800 Subject: x86, atomic64: In selftest, distinguish x86-64 from 586+ The x86-64 implementation of the atomics is totally different from the i586+ implementation, which makes it quite confusing to call it "586+". Also fix indentation, and add "i" for "i386" and "i586" as used elsewhere in the kernel. Signed-off-by: H. Peter Anvin Cc: Luca Barbieri LKML-Reference: <1267005265-27958-4-git-send-email-luca@luca-barbieri.com> diff --git a/lib/atomic64_test.c b/lib/atomic64_test.c index f7bb706..65e482c 100644 --- a/lib/atomic64_test.c +++ b/lib/atomic64_test.c @@ -144,14 +144,16 @@ static __init int test_atomic64(void) BUG_ON(v.counter != r); #ifdef CONFIG_X86 - printk(KERN_INFO "atomic64 test passed for %s+ platform %s CX8 and %s SSE\n", -#ifdef CONFIG_X86_CMPXCHG64 - "586", + printk(KERN_INFO "atomic64 test passed for %s platform %s CX8 and %s SSE\n", +#ifdef CONFIG_X86_64 + "x86-64", +#elif defined(CONFIG_X86_CMPXCHG64) + "i586+", #else - "386", + "i386+", #endif - boot_cpu_has(X86_FEATURE_CX8) ? "with" : "without", - boot_cpu_has(X86_FEATURE_XMM) ? "with" : "without"); + boot_cpu_has(X86_FEATURE_CX8) ? "with" : "without", + boot_cpu_has(X86_FEATURE_XMM) ? "with" : "without"); #else printk(KERN_INFO "atomic64 test passed\n"); #endif -- cgit v0.10.2 From 4daa2a8093ecd1148270a1fc64e99f072b8c2901 Mon Sep 17 00:00:00 2001 From: "Pallipadi, Venkatesh" Date: Wed, 24 Feb 2010 13:43:55 -0800 Subject: x86, pat: In rbt_memtype_check_insert(), update new->type only if valid new->type should only change when there is a valid ret_type. Otherwise the requested type and return type should be same. Signed-off-by: Venkatesh Pallipadi LKML-Reference: <20100224214355.GA16431@linux-os.sc.intel.com> Tested-by: Jack Steiner Signed-off-by: H. Peter Anvin diff --git a/arch/x86/mm/pat_rbtree.c b/arch/x86/mm/pat_rbtree.c index 9063f40..07de4cb 100644 --- a/arch/x86/mm/pat_rbtree.c +++ b/arch/x86/mm/pat_rbtree.c @@ -223,7 +223,9 @@ int rbt_memtype_check_insert(struct memtype *new, unsigned long *ret_type) new->type, ret_type); if (!err) { - new->type = *ret_type; + if (ret_type) + new->type = *ret_type; + memtype_rb_insert(&memtype_rbroot, new); } return err; -- cgit v0.10.2 From 372e22ef0a87d5fc10d387791f9f19721115820c Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Mon, 1 Mar 2010 22:06:25 -0800 Subject: x86, docbook: Fix errors from x86 headers merger Fix docbook errors for x86 headers that were recently merged: docproc: arch/x86/include/asm/atomic_32.h: No such file or directory docproc: arch/x86/include/asm/io_32.h: No such file or directory Signed-off-by: Randy Dunlap LKML-Reference: <20100301220625.bad46c19.randy.dunlap@oracle.com> Cc: Brian Gerst Signed-off-by: H. Peter Anvin diff --git a/Documentation/DocBook/device-drivers.tmpl b/Documentation/DocBook/device-drivers.tmpl index f9a6e2c..1b2dd4f 100644 --- a/Documentation/DocBook/device-drivers.tmpl +++ b/Documentation/DocBook/device-drivers.tmpl @@ -45,7 +45,7 @@ Atomic and pointer manipulation -!Iarch/x86/include/asm/atomic_32.h +!Iarch/x86/include/asm/atomic.h !Iarch/x86/include/asm/unaligned.h diff --git a/Documentation/DocBook/deviceiobook.tmpl b/Documentation/DocBook/deviceiobook.tmpl index 3ed8812..c1ed6a4 100644 --- a/Documentation/DocBook/deviceiobook.tmpl +++ b/Documentation/DocBook/deviceiobook.tmpl @@ -316,7 +316,7 @@ CPU B: spin_unlock_irqrestore(&dev_lock, flags) Public Functions Provided -!Iarch/x86/include/asm/io_32.h +!Iarch/x86/include/asm/io.h !Elib/iomap.c -- cgit v0.10.2 From ced918eb748ce30b3aace549fd17540e40ffdca0 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Wed, 17 Feb 2010 16:47:10 +0000 Subject: i8253: Convert i8253_lock to raw_spinlock i8253_lock needs to be a real spinlock in preempt-rt, i.e. it can not be converted to a sleeping lock. Convert it to raw_spinlock and fix up all users. Signed-off-by: Thomas Gleixner Acked-by: Ralf Baechle Acked-by: Dmitry Torokhov Acked-by: Takashi Iwai Cc: Jens Axboe LKML-Reference: <20100217163751.030764372@linutronix.de> diff --git a/arch/mips/include/asm/i8253.h b/arch/mips/include/asm/i8253.h index 032ca73..48bb823 100644 --- a/arch/mips/include/asm/i8253.h +++ b/arch/mips/include/asm/i8253.h @@ -12,7 +12,7 @@ #define PIT_CH0 0x40 #define PIT_CH2 0x42 -extern spinlock_t i8253_lock; +extern raw_spinlock_t i8253_lock; extern void setup_pit_timer(void); diff --git a/arch/mips/kernel/i8253.c b/arch/mips/kernel/i8253.c index ed5c441..9479406 100644 --- a/arch/mips/kernel/i8253.c +++ b/arch/mips/kernel/i8253.c @@ -15,7 +15,7 @@ #include #include -DEFINE_SPINLOCK(i8253_lock); +DEFINE_RAW_SPINLOCK(i8253_lock); EXPORT_SYMBOL(i8253_lock); /* @@ -26,7 +26,7 @@ EXPORT_SYMBOL(i8253_lock); static void init_pit_timer(enum clock_event_mode mode, struct clock_event_device *evt) { - spin_lock(&i8253_lock); + raw_spin_lock(&i8253_lock); switch(mode) { case CLOCK_EVT_MODE_PERIODIC: @@ -55,7 +55,7 @@ static void init_pit_timer(enum clock_event_mode mode, /* Nothing to do here */ break; } - spin_unlock(&i8253_lock); + raw_spin_unlock(&i8253_lock); } /* @@ -65,10 +65,10 @@ static void init_pit_timer(enum clock_event_mode mode, */ static int pit_next_event(unsigned long delta, struct clock_event_device *evt) { - spin_lock(&i8253_lock); + raw_spin_lock(&i8253_lock); outb_p(delta & 0xff , PIT_CH0); /* LSB */ outb(delta >> 8 , PIT_CH0); /* MSB */ - spin_unlock(&i8253_lock); + raw_spin_unlock(&i8253_lock); return 0; } @@ -137,7 +137,7 @@ static cycle_t pit_read(struct clocksource *cs) static int old_count; static u32 old_jifs; - spin_lock_irqsave(&i8253_lock, flags); + raw_spin_lock_irqsave(&i8253_lock, flags); /* * Although our caller may have the read side of xtime_lock, * this is now a seqlock, and we are cheating in this routine @@ -183,7 +183,7 @@ static cycle_t pit_read(struct clocksource *cs) old_count = count; old_jifs = jifs; - spin_unlock_irqrestore(&i8253_lock, flags); + raw_spin_unlock_irqrestore(&i8253_lock, flags); count = (LATCH - 1) - count; diff --git a/arch/x86/include/asm/i8253.h b/arch/x86/include/asm/i8253.h index 1edbf89..fc1f579 100644 --- a/arch/x86/include/asm/i8253.h +++ b/arch/x86/include/asm/i8253.h @@ -6,7 +6,7 @@ #define PIT_CH0 0x40 #define PIT_CH2 0x42 -extern spinlock_t i8253_lock; +extern raw_spinlock_t i8253_lock; extern struct clock_event_device *global_clock_event; diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c index 031aa88..c4f9182 100644 --- a/arch/x86/kernel/apm_32.c +++ b/arch/x86/kernel/apm_32.c @@ -1224,7 +1224,7 @@ static void reinit_timer(void) #ifdef INIT_TIMER_AFTER_SUSPEND unsigned long flags; - spin_lock_irqsave(&i8253_lock, flags); + raw_spin_lock_irqsave(&i8253_lock, flags); /* set the clock to HZ */ outb_pit(0x34, PIT_MODE); /* binary, mode 2, LSB/MSB, ch 0 */ udelay(10); @@ -1232,7 +1232,7 @@ static void reinit_timer(void) udelay(10); outb_pit(LATCH >> 8, PIT_CH0); /* MSB */ udelay(10); - spin_unlock_irqrestore(&i8253_lock, flags); + raw_spin_unlock_irqrestore(&i8253_lock, flags); #endif } diff --git a/arch/x86/kernel/i8253.c b/arch/x86/kernel/i8253.c index 23c1679..2dfd315 100644 --- a/arch/x86/kernel/i8253.c +++ b/arch/x86/kernel/i8253.c @@ -16,7 +16,7 @@ #include #include -DEFINE_SPINLOCK(i8253_lock); +DEFINE_RAW_SPINLOCK(i8253_lock); EXPORT_SYMBOL(i8253_lock); /* @@ -33,7 +33,7 @@ struct clock_event_device *global_clock_event; static void init_pit_timer(enum clock_event_mode mode, struct clock_event_device *evt) { - spin_lock(&i8253_lock); + raw_spin_lock(&i8253_lock); switch (mode) { case CLOCK_EVT_MODE_PERIODIC: @@ -62,7 +62,7 @@ static void init_pit_timer(enum clock_event_mode mode, /* Nothing to do here */ break; } - spin_unlock(&i8253_lock); + raw_spin_unlock(&i8253_lock); } /* @@ -72,10 +72,10 @@ static void init_pit_timer(enum clock_event_mode mode, */ static int pit_next_event(unsigned long delta, struct clock_event_device *evt) { - spin_lock(&i8253_lock); + raw_spin_lock(&i8253_lock); outb_pit(delta & 0xff , PIT_CH0); /* LSB */ outb_pit(delta >> 8 , PIT_CH0); /* MSB */ - spin_unlock(&i8253_lock); + raw_spin_unlock(&i8253_lock); return 0; } @@ -130,7 +130,7 @@ static cycle_t pit_read(struct clocksource *cs) int count; u32 jifs; - spin_lock_irqsave(&i8253_lock, flags); + raw_spin_lock_irqsave(&i8253_lock, flags); /* * Although our caller may have the read side of xtime_lock, * this is now a seqlock, and we are cheating in this routine @@ -176,7 +176,7 @@ static cycle_t pit_read(struct clocksource *cs) old_count = count; old_jifs = jifs; - spin_unlock_irqrestore(&i8253_lock, flags); + raw_spin_unlock_irqrestore(&i8253_lock, flags); count = (LATCH - 1) - count; diff --git a/drivers/block/hd.c b/drivers/block/hd.c index 5116c65..b9868ad 100644 --- a/drivers/block/hd.c +++ b/drivers/block/hd.c @@ -165,12 +165,12 @@ unsigned long read_timer(void) unsigned long t, flags; int i; - spin_lock_irqsave(&i8253_lock, flags); + raw_spin_lock_irqsave(&i8253_lock, flags); t = jiffies * 11932; outb_p(0, 0x43); i = inb_p(0x40); i |= inb(0x40) << 8; - spin_unlock_irqrestore(&i8253_lock, flags); + raw_spin_unlock_irqrestore(&i8253_lock, flags); return(t - i); } #endif diff --git a/drivers/input/gameport/gameport.c b/drivers/input/gameport/gameport.c index 7e18bcf..46239e4 100644 --- a/drivers/input/gameport/gameport.c +++ b/drivers/input/gameport/gameport.c @@ -59,11 +59,11 @@ static unsigned int get_time_pit(void) unsigned long flags; unsigned int count; - spin_lock_irqsave(&i8253_lock, flags); + raw_spin_lock_irqsave(&i8253_lock, flags); outb_p(0x00, 0x43); count = inb_p(0x40); count |= inb_p(0x40) << 8; - spin_unlock_irqrestore(&i8253_lock, flags); + raw_spin_unlock_irqrestore(&i8253_lock, flags); return count; } diff --git a/drivers/input/joystick/analog.c b/drivers/input/joystick/analog.c index 1c0b529..4afe0a3 100644 --- a/drivers/input/joystick/analog.c +++ b/drivers/input/joystick/analog.c @@ -146,11 +146,11 @@ static unsigned int get_time_pit(void) unsigned long flags; unsigned int count; - spin_lock_irqsave(&i8253_lock, flags); + raw_spin_lock_irqsave(&i8253_lock, flags); outb_p(0x00, 0x43); count = inb_p(0x40); count |= inb_p(0x40) << 8; - spin_unlock_irqrestore(&i8253_lock, flags); + raw_spin_unlock_irqrestore(&i8253_lock, flags); return count; } diff --git a/drivers/input/misc/pcspkr.c b/drivers/input/misc/pcspkr.c index ea4e1fd..f080dd3 100644 --- a/drivers/input/misc/pcspkr.c +++ b/drivers/input/misc/pcspkr.c @@ -30,7 +30,7 @@ MODULE_ALIAS("platform:pcspkr"); #include #else #include -static DEFINE_SPINLOCK(i8253_lock); +static DEFINE_RAW_SPINLOCK(i8253_lock); #endif static int pcspkr_event(struct input_dev *dev, unsigned int type, unsigned int code, int value) @@ -50,7 +50,7 @@ static int pcspkr_event(struct input_dev *dev, unsigned int type, unsigned int c if (value > 20 && value < 32767) count = PIT_TICK_RATE / value; - spin_lock_irqsave(&i8253_lock, flags); + raw_spin_lock_irqsave(&i8253_lock, flags); if (count) { /* set command for counter 2, 2 byte write */ @@ -65,7 +65,7 @@ static int pcspkr_event(struct input_dev *dev, unsigned int type, unsigned int c outb(inb_p(0x61) & 0xFC, 0x61); } - spin_unlock_irqrestore(&i8253_lock, flags); + raw_spin_unlock_irqrestore(&i8253_lock, flags); return 0; } diff --git a/sound/drivers/pcsp/pcsp.h b/sound/drivers/pcsp/pcsp.h index 1e12307..4ff6c8c 100644 --- a/sound/drivers/pcsp/pcsp.h +++ b/sound/drivers/pcsp/pcsp.h @@ -16,7 +16,7 @@ #include #else #include -static DEFINE_SPINLOCK(i8253_lock); +static DEFINE_RAW_SPINLOCK(i8253_lock); #endif #define PCSP_SOUND_VERSION 0x400 /* read 4.00 */ diff --git a/sound/drivers/pcsp/pcsp_input.c b/sound/drivers/pcsp/pcsp_input.c index 0444cde..b5e2b54 100644 --- a/sound/drivers/pcsp/pcsp_input.c +++ b/sound/drivers/pcsp/pcsp_input.c @@ -21,7 +21,7 @@ static void pcspkr_do_sound(unsigned int count) { unsigned long flags; - spin_lock_irqsave(&i8253_lock, flags); + raw_spin_lock_irqsave(&i8253_lock, flags); if (count) { /* set command for counter 2, 2 byte write */ @@ -36,7 +36,7 @@ static void pcspkr_do_sound(unsigned int count) outb(inb_p(0x61) & 0xFC, 0x61); } - spin_unlock_irqrestore(&i8253_lock, flags); + raw_spin_unlock_irqrestore(&i8253_lock, flags); } void pcspkr_stop_sound(void) diff --git a/sound/drivers/pcsp/pcsp_lib.c b/sound/drivers/pcsp/pcsp_lib.c index e1145ac..f6a2e72 100644 --- a/sound/drivers/pcsp/pcsp_lib.c +++ b/sound/drivers/pcsp/pcsp_lib.c @@ -65,7 +65,7 @@ static u64 pcsp_timer_update(struct snd_pcsp *chip) timer_cnt = val * CUR_DIV() / 256; if (timer_cnt && chip->enable) { - spin_lock_irqsave(&i8253_lock, flags); + raw_spin_lock_irqsave(&i8253_lock, flags); if (!nforce_wa) { outb_p(chip->val61, 0x61); outb_p(timer_cnt, 0x42); @@ -74,7 +74,7 @@ static u64 pcsp_timer_update(struct snd_pcsp *chip) outb(chip->val61 ^ 2, 0x61); chip->thalf = 1; } - spin_unlock_irqrestore(&i8253_lock, flags); + raw_spin_unlock_irqrestore(&i8253_lock, flags); } chip->ns_rem = PCSP_PERIOD_NS(); @@ -158,10 +158,10 @@ static int pcsp_start_playing(struct snd_pcsp *chip) return -EIO; } - spin_lock(&i8253_lock); + raw_spin_lock(&i8253_lock); chip->val61 = inb(0x61) | 0x03; outb_p(0x92, 0x43); /* binary, mode 1, LSB only, ch 2 */ - spin_unlock(&i8253_lock); + raw_spin_unlock(&i8253_lock); atomic_set(&chip->timer_active, 1); chip->thalf = 0; @@ -178,11 +178,11 @@ static void pcsp_stop_playing(struct snd_pcsp *chip) return; atomic_set(&chip->timer_active, 0); - spin_lock(&i8253_lock); + raw_spin_lock(&i8253_lock); /* restore the timer */ outb_p(0xb6, 0x43); /* binary, mode 3, LSB/MSB, ch 2 */ outb(chip->val61 & 0xFC, 0x61); - spin_unlock(&i8253_lock); + raw_spin_unlock(&i8253_lock); } /* -- cgit v0.10.2 From bc078e4eab65f11bbaeed380593ab8151b30d703 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Tue, 2 Mar 2010 16:01:10 +0100 Subject: oprofile: convert oprofile from timer_hook to hrtimer Oprofile is currently broken on systems running with NOHZ enabled. A maximum of 1 tick is accounted via the timer_hook if a cpu sleeps for a longer period of time. This does bad things to the percentages in the profiler output. To solve this problem convert oprofile to use a restarting hrtimer instead of the timer_hook. Signed-off-by: Martin Schwidefsky Signed-off-by: Robert Richter diff --git a/drivers/oprofile/oprof.c b/drivers/oprofile/oprof.c index dc8a042..b336cd9 100644 --- a/drivers/oprofile/oprof.c +++ b/drivers/oprofile/oprof.c @@ -253,22 +253,26 @@ static int __init oprofile_init(void) int err; err = oprofile_arch_init(&oprofile_ops); - if (err < 0 || timer) { printk(KERN_INFO "oprofile: using timer interrupt.\n"); - oprofile_timer_init(&oprofile_ops); + err = oprofile_timer_init(&oprofile_ops); + if (err) + goto out_arch; } - err = oprofilefs_register(); if (err) - oprofile_arch_exit(); + goto out_arch; + return 0; +out_arch: + oprofile_arch_exit(); return err; } static void __exit oprofile_exit(void) { + oprofile_timer_exit(); oprofilefs_unregister(); oprofile_arch_exit(); } diff --git a/drivers/oprofile/oprof.h b/drivers/oprofile/oprof.h index cb92f5c..47e12cb 100644 --- a/drivers/oprofile/oprof.h +++ b/drivers/oprofile/oprof.h @@ -34,7 +34,8 @@ struct super_block; struct dentry; void oprofile_create_files(struct super_block *sb, struct dentry *root); -void oprofile_timer_init(struct oprofile_operations *ops); +int oprofile_timer_init(struct oprofile_operations *ops); +void oprofile_timer_exit(void); int oprofile_set_backtrace(unsigned long depth); int oprofile_set_timeout(unsigned long time); diff --git a/drivers/oprofile/timer_int.c b/drivers/oprofile/timer_int.c index 333f915..dc0ae4d 100644 --- a/drivers/oprofile/timer_int.c +++ b/drivers/oprofile/timer_int.c @@ -13,34 +13,94 @@ #include #include #include +#include +#include +#include #include #include "oprof.h" -static int timer_notify(struct pt_regs *regs) +static DEFINE_PER_CPU(struct hrtimer, oprofile_hrtimer); + +static enum hrtimer_restart oprofile_hrtimer_notify(struct hrtimer *hrtimer) +{ + oprofile_add_sample(get_irq_regs(), 0); + hrtimer_forward_now(hrtimer, ns_to_ktime(TICK_NSEC)); + return HRTIMER_RESTART; +} + +static void __oprofile_hrtimer_start(void *unused) +{ + struct hrtimer *hrtimer = &__get_cpu_var(oprofile_hrtimer); + + hrtimer_init(hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); + hrtimer->function = oprofile_hrtimer_notify; + + hrtimer_start(hrtimer, ns_to_ktime(TICK_NSEC), + HRTIMER_MODE_REL_PINNED); +} + +static int oprofile_hrtimer_start(void) { - oprofile_add_sample(regs, 0); + on_each_cpu(__oprofile_hrtimer_start, NULL, 1); return 0; } -static int timer_start(void) +static void __oprofile_hrtimer_stop(int cpu) { - return register_timer_hook(timer_notify); + struct hrtimer *hrtimer = &per_cpu(oprofile_hrtimer, cpu); + + hrtimer_cancel(hrtimer); } +static void oprofile_hrtimer_stop(void) +{ + int cpu; + + for_each_online_cpu(cpu) + __oprofile_hrtimer_stop(cpu); +} -static void timer_stop(void) +static int __cpuinit oprofile_cpu_notify(struct notifier_block *self, + unsigned long action, void *hcpu) { - unregister_timer_hook(timer_notify); + long cpu = (long) hcpu; + + switch (action) { + case CPU_ONLINE: + case CPU_ONLINE_FROZEN: + smp_call_function_single(cpu, __oprofile_hrtimer_start, + NULL, 1); + break; + case CPU_DEAD: + case CPU_DEAD_FROZEN: + __oprofile_hrtimer_stop(cpu); + break; + } + return NOTIFY_OK; } +static struct notifier_block __refdata oprofile_cpu_notifier = { + .notifier_call = oprofile_cpu_notify, +}; -void __init oprofile_timer_init(struct oprofile_operations *ops) +int __init oprofile_timer_init(struct oprofile_operations *ops) { + int rc; + + rc = register_hotcpu_notifier(&oprofile_cpu_notifier); + if (rc) + return rc; ops->create_files = NULL; ops->setup = NULL; ops->shutdown = NULL; - ops->start = timer_start; - ops->stop = timer_stop; + ops->start = oprofile_hrtimer_start; + ops->stop = oprofile_hrtimer_stop; ops->cpu_type = "timer"; + return 0; +} + +void __exit oprofile_timer_exit(void) +{ + unregister_hotcpu_notifier(&oprofile_cpu_notifier); } -- cgit v0.10.2 From cf07d2ea43e5c22149435ee9002cb737eac20eca Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 28 Feb 2010 23:20:19 -0500 Subject: nfsd4: simplify references to nfsd4 lease time Instead of accessing the lease time directly, some users call nfs4_lease_time(), and some a macro, NFSD_LEASE_TIME, defined as nfs4_lease_time(). Neither layer of indirection serves any purpose. Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 4bc22c7..ed12ad4 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -455,7 +455,7 @@ static struct rpc_program cb_program = { static int max_cb_time(void) { - return max(NFSD_LEASE_TIME/10, (time_t)1) * HZ; + return max(nfsd4_lease/10, (time_t)1) * HZ; } /* Reference counting, callback cleanup, etc., all look racy as heck. diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 3a20c09..cc9164a 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -44,7 +44,7 @@ #define NFSDDBG_FACILITY NFSDDBG_PROC /* Globals */ -static time_t lease_time = 90; /* default lease time */ +time_t nfsd4_lease = 90; /* default lease time */ static time_t user_lease_time = 90; static time_t boot_time; static u32 current_ownerid = 1; @@ -2560,9 +2560,9 @@ nfs4_laundromat(void) struct nfs4_stateowner *sop; struct nfs4_delegation *dp; struct list_head *pos, *next, reaplist; - time_t cutoff = get_seconds() - NFSD_LEASE_TIME; - time_t t, clientid_val = NFSD_LEASE_TIME; - time_t u, test_val = NFSD_LEASE_TIME; + time_t cutoff = get_seconds() - nfsd4_lease; + time_t t, clientid_val = nfsd4_lease; + time_t u, test_val = nfsd4_lease; nfs4_lock_state(); @@ -2602,7 +2602,7 @@ nfs4_laundromat(void) list_del_init(&dp->dl_recall_lru); unhash_delegation(dp); } - test_val = NFSD_LEASE_TIME; + test_val = nfsd4_lease; list_for_each_safe(pos, next, &close_lru) { sop = list_entry(pos, struct nfs4_stateowner, so_close_lru); if (time_after((unsigned long)sop->so_time, (unsigned long)cutoff)) { @@ -2672,7 +2672,7 @@ EXPIRED_STATEID(stateid_t *stateid) { if (time_before((unsigned long)boot_time, ((unsigned long)stateid->si_boot)) && - time_before((unsigned long)(stateid->si_boot + lease_time), get_seconds())) { + time_before((unsigned long)(stateid->si_boot + nfsd4_lease), get_seconds())) { dprintk("NFSD: expired stateid " STATEID_FMT "!\n", STATEID_VAL(stateid)); return 1; @@ -3976,7 +3976,7 @@ nfsd4_load_reboot_recovery_data(void) unsigned long get_nfs4_grace_period(void) { - return max(user_lease_time, lease_time) * HZ; + return max(user_lease_time, nfsd4_lease) * HZ; } /* @@ -4009,7 +4009,7 @@ __nfs4_state_start(void) boot_time = get_seconds(); grace_time = get_nfs4_grace_period(); - lease_time = user_lease_time; + nfsd4_lease = user_lease_time; locks_start_grace(&nfsd4_manager); printk(KERN_INFO "NFSD: starting %ld-second grace period\n", grace_time/HZ); @@ -4036,12 +4036,6 @@ nfs4_state_start(void) return 0; } -time_t -nfs4_lease_time(void) -{ - return lease_time; -} - static void __nfs4_state_shutdown(void) { diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index c458fb1..f61bd73 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -1899,7 +1899,7 @@ nfsd4_encode_fattr(struct svc_fh *fhp, struct svc_export *exp, if (bmval0 & FATTR4_WORD0_LEASE_TIME) { if ((buflen -= 4) < 0) goto out_resource; - WRITE32(NFSD_LEASE_TIME); + WRITE32(nfsd4_lease); } if (bmval0 & FATTR4_WORD0_RDATTR_ERROR) { if ((buflen -= 4) < 0) diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 0f0e77f..8bff674 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -1203,8 +1203,6 @@ static ssize_t write_maxblksize(struct file *file, char *buf, size_t size) } #ifdef CONFIG_NFSD_V4 -extern time_t nfs4_leasetime(void); - static ssize_t __write_leasetime(struct file *file, char *buf, size_t size) { /* if size > 10 seconds, call @@ -1224,8 +1222,7 @@ static ssize_t __write_leasetime(struct file *file, char *buf, size_t size) nfs4_reset_lease(lease); } - return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%ld\n", - nfs4_lease_time()); + return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%ld\n", nfsd4_lease); } /** diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h index e942a1a..b463093 100644 --- a/fs/nfsd/nfsd.h +++ b/fs/nfsd/nfsd.h @@ -82,7 +82,6 @@ int nfs4_state_init(void); void nfsd4_free_slabs(void); int nfs4_state_start(void); void nfs4_state_shutdown(void); -time_t nfs4_lease_time(void); void nfs4_reset_lease(time_t leasetime); int nfs4_reset_recoverydir(char *recdir); #else @@ -90,7 +89,6 @@ static inline int nfs4_state_init(void) { return 0; } static inline void nfsd4_free_slabs(void) { } static inline int nfs4_state_start(void) { return 0; } static inline void nfs4_state_shutdown(void) { } -static inline time_t nfs4_lease_time(void) { return 0; } static inline void nfs4_reset_lease(time_t leasetime) { } static inline int nfs4_reset_recoverydir(char *recdir) { return 0; } #endif @@ -229,6 +227,8 @@ extern struct timeval nfssvc_boot; #ifdef CONFIG_NFSD_V4 +extern time_t nfsd4_lease; + /* before processing a COMPOUND operation, we have to check that there * is enough space in the buffer for XDR encode to succeed. otherwise, * we might process an operation with side effects, and be unable to @@ -247,7 +247,6 @@ extern struct timeval nfssvc_boot; #define COMPOUND_SLACK_SPACE 140 /* OP_GETFH */ #define COMPOUND_ERR_SLACK_SPACE 12 /* OP_SETATTR */ -#define NFSD_LEASE_TIME (nfs4_lease_time()) #define NFSD_LAUNDROMAT_MINTIMEOUT 10 /* seconds */ /* -- cgit v0.10.2 From e46b498c84163e86e2627c30bca298c968664f65 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 1 Mar 2010 19:21:21 -0500 Subject: nfsd4: simplify lease/grace interaction The original code here assumed we'd allow the user to change the lease any time, but only allow the change to take effect on restart. Since then we modified the code to allow setting the lease on when the server is down. Update the rest of the code to reflect that fact, clarify variable names, and add document. Also, the code insisted that the grace period always be the longer of the old and new lease periods, but that's overly conservative--as long as it lasts at least the old lease period, old clients should still know to recover in time. Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index cc9164a..eb8d124 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -45,7 +45,7 @@ /* Globals */ time_t nfsd4_lease = 90; /* default lease time */ -static time_t user_lease_time = 90; +static time_t nfsd4_grace = 90; static time_t boot_time; static u32 current_ownerid = 1; static u32 current_fileid = 1; @@ -2551,6 +2551,12 @@ nfsd4_end_grace(void) dprintk("NFSD: end of grace period\n"); nfsd4_recdir_purge_old(); locks_end_grace(&nfsd4_manager); + /* + * Now that every NFSv4 client has had the chance to recover and + * to see the (possibly new, possibly shorter) lease time, we + * can safely set the next grace time to the current lease time: + */ + nfsd4_grace = nfsd4_lease; } static time_t @@ -3973,12 +3979,6 @@ nfsd4_load_reboot_recovery_data(void) printk("NFSD: Failure reading reboot recovery data\n"); } -unsigned long -get_nfs4_grace_period(void) -{ - return max(user_lease_time, nfsd4_lease) * HZ; -} - /* * Since the lifetime of a delegation isn't limited to that of an open, a * client may quite reasonably hang on to a delegation as long as it has @@ -4005,18 +4005,14 @@ set_max_delegations(void) static int __nfs4_state_start(void) { - unsigned long grace_time; - boot_time = get_seconds(); - grace_time = get_nfs4_grace_period(); - nfsd4_lease = user_lease_time; locks_start_grace(&nfsd4_manager); printk(KERN_INFO "NFSD: starting %ld-second grace period\n", - grace_time/HZ); + nfsd4_grace); laundry_wq = create_singlethread_workqueue("nfsd4"); if (laundry_wq == NULL) return -ENOMEM; - queue_delayed_work(laundry_wq, &laundromat_work, grace_time); + queue_delayed_work(laundry_wq, &laundromat_work, nfsd4_grace * HZ); set_max_delegations(); return set_callback_cred(); } @@ -4123,17 +4119,11 @@ nfs4_recoverydir(void) /* * Called when leasetime is changed. * - * The only way the protocol gives us to handle on-the-fly lease changes is to - * simulate a reboot. Instead of doing that, we just wait till the next time - * we start to register any changes in lease time. If the administrator - * really wants to change the lease time *now*, they can go ahead and bring - * nfsd down and then back up again after changing the lease time. - * - * user_lease_time is protected by nfsd_mutex since it's only really accessed + * nfsd4_lease is protected by nfsd_mutex since it's only really accessed * when nfsd is starting */ void nfs4_reset_lease(time_t leasetime) { - user_lease_time = leasetime; + nfsd4_lease = leasetime; } -- cgit v0.10.2 From f958a1320ff7a1e0e861d3c90de6da12a88839dc Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 1 Mar 2010 19:43:02 -0500 Subject: nfsd4: remove unnecessary lease-setting function This is another layer of indirection that doesn't really buy us anything. Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index eb8d124..4471046 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -4115,15 +4115,3 @@ nfs4_recoverydir(void) { return user_recovery_dirname; } - -/* - * Called when leasetime is changed. - * - * nfsd4_lease is protected by nfsd_mutex since it's only really accessed - * when nfsd is starting - */ -void -nfs4_reset_lease(time_t leasetime) -{ - nfsd4_lease = leasetime; -} diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 8bff674..6738e9d 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -1219,7 +1219,7 @@ static ssize_t __write_leasetime(struct file *file, char *buf, size_t size) return rv; if (lease < 10 || lease > 3600) return -EINVAL; - nfs4_reset_lease(lease); + nfsd4_lease = lease; } return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%ld\n", nfsd4_lease); -- cgit v0.10.2 From f013574014816c7a557b3c52233f3620463f0b9b Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 1 Mar 2010 19:32:36 -0500 Subject: nfsd4: reshuffle lease-setting code to allow reuse We'll soon allow setting the grace period, so we'll want to share this code. Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 6738e9d..9c73cac 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -1203,26 +1203,36 @@ static ssize_t write_maxblksize(struct file *file, char *buf, size_t size) } #ifdef CONFIG_NFSD_V4 -static ssize_t __write_leasetime(struct file *file, char *buf, size_t size) +static ssize_t __nfsd4_write_time(struct file *file, char *buf, size_t size, time_t *time) { /* if size > 10 seconds, call * nfs4_reset_lease() then write out the new lease (seconds) as reply */ char *mesg = buf; - int rv, lease; + int rv, i; if (size > 0) { if (nfsd_serv) return -EBUSY; - rv = get_int(&mesg, &lease); + rv = get_int(&mesg, &i); if (rv) return rv; - if (lease < 10 || lease > 3600) + if (i < 10 || i > 3600) return -EINVAL; - nfsd4_lease = lease; + *time = i; } - return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%ld\n", nfsd4_lease); + return scnprintf(buf, SIMPLE_TRANSACTION_LIMIT, "%ld\n", *time); +} + +static ssize_t nfsd4_write_time(struct file *file, char *buf, size_t size, time_t *time) +{ + ssize_t rv; + + mutex_lock(&nfsd_mutex); + rv = __nfsd4_write_time(file, buf, size, time); + mutex_unlock(&nfsd_mutex); + return rv; } /** @@ -1248,12 +1258,7 @@ static ssize_t __write_leasetime(struct file *file, char *buf, size_t size) */ static ssize_t write_leasetime(struct file *file, char *buf, size_t size) { - ssize_t rv; - - mutex_lock(&nfsd_mutex); - rv = __write_leasetime(file, buf, size); - mutex_unlock(&nfsd_mutex); - return rv; + return nfsd4_write_time(file, buf, size, &nfsd4_lease); } extern char *nfs4_recoverydir(void); -- cgit v0.10.2 From efc4bb4fdd09c11f5558446e584a494c6feb43c7 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Tue, 2 Mar 2010 11:04:06 -0500 Subject: nfsd4: allow setting grace period time Allow explicit configuration of the grace period time as well as the lease period time. Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 4471046..6edfe23 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -45,7 +45,7 @@ /* Globals */ time_t nfsd4_lease = 90; /* default lease time */ -static time_t nfsd4_grace = 90; +time_t nfsd4_grace = 90; static time_t boot_time; static u32 current_ownerid = 1; static u32 current_fileid = 1; diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 9c73cac..7ab70ff 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -45,6 +45,7 @@ enum { */ #ifdef CONFIG_NFSD_V4 NFSD_Leasetime, + NFSD_Gracetime, NFSD_RecoveryDir, #endif }; @@ -69,6 +70,7 @@ static ssize_t write_ports(struct file *file, char *buf, size_t size); static ssize_t write_maxblksize(struct file *file, char *buf, size_t size); #ifdef CONFIG_NFSD_V4 static ssize_t write_leasetime(struct file *file, char *buf, size_t size); +static ssize_t write_gracetime(struct file *file, char *buf, size_t size); static ssize_t write_recoverydir(struct file *file, char *buf, size_t size); #endif @@ -90,6 +92,7 @@ static ssize_t (*write_op[])(struct file *, char *, size_t) = { [NFSD_MaxBlkSize] = write_maxblksize, #ifdef CONFIG_NFSD_V4 [NFSD_Leasetime] = write_leasetime, + [NFSD_Gracetime] = write_gracetime, [NFSD_RecoveryDir] = write_recoverydir, #endif }; @@ -1261,6 +1264,21 @@ static ssize_t write_leasetime(struct file *file, char *buf, size_t size) return nfsd4_write_time(file, buf, size, &nfsd4_lease); } +/** + * write_gracetime - Set or report current NFSv4 grace period time + * + * As above, but sets the time of the NFSv4 grace period. + * + * Note this should never be set to less than the *previous* + * lease-period time, but we don't try to enforce this. (In the common + * case (a new boot), we don't know what the previous lease time was + * anyway.) + */ +static ssize_t write_gracetime(struct file *file, char *buf, size_t size) +{ + return nfsd4_write_time(file, buf, size, &nfsd4_grace); +} + extern char *nfs4_recoverydir(void); static ssize_t __write_recoverydir(struct file *file, char *buf, size_t size) @@ -1352,6 +1370,7 @@ static int nfsd_fill_super(struct super_block * sb, void * data, int silent) [NFSD_MaxBlkSize] = {"max_block_size", &transaction_ops, S_IWUSR|S_IRUGO}, #ifdef CONFIG_NFSD_V4 [NFSD_Leasetime] = {"nfsv4leasetime", &transaction_ops, S_IWUSR|S_IRUSR}, + [NFSD_Gracetime] = {"nfsv4gracetime", &transaction_ops, S_IWUSR|S_IRUSR}, [NFSD_RecoveryDir] = {"nfsv4recoverydir", &transaction_ops, S_IWUSR|S_IRUSR}, #endif /* last one */ {""} diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h index b463093..7237776 100644 --- a/fs/nfsd/nfsd.h +++ b/fs/nfsd/nfsd.h @@ -228,6 +228,7 @@ extern struct timeval nfssvc_boot; #ifdef CONFIG_NFSD_V4 extern time_t nfsd4_lease; +extern time_t nfsd4_grace; /* before processing a COMPOUND operation, we have to check that there * is enough space in the buffer for XDR encode to succeed. otherwise, -- cgit v0.10.2 From e7b184f199fd3c80b618ec8244cbda70857d2779 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Tue, 2 Mar 2010 11:18:40 -0500 Subject: nfsd4: document lease/grace-period limits The current documentation here is out of date, and not quite right. (Future work: some user documentation would be useful.) Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfsctl.c b/fs/nfsd/nfsctl.c index 7ab70ff..413cb8e 100644 --- a/fs/nfsd/nfsctl.c +++ b/fs/nfsd/nfsctl.c @@ -1208,9 +1208,6 @@ static ssize_t write_maxblksize(struct file *file, char *buf, size_t size) #ifdef CONFIG_NFSD_V4 static ssize_t __nfsd4_write_time(struct file *file, char *buf, size_t size, time_t *time) { - /* if size > 10 seconds, call - * nfs4_reset_lease() then write out the new lease (seconds) as reply - */ char *mesg = buf; int rv, i; @@ -1220,6 +1217,18 @@ static ssize_t __nfsd4_write_time(struct file *file, char *buf, size_t size, tim rv = get_int(&mesg, &i); if (rv) return rv; + /* + * Some sanity checking. We don't have a reason for + * these particular numbers, but problems with the + * extremes are: + * - Too short: the briefest network outage may + * cause clients to lose all their locks. Also, + * the frequent polling may be wasteful. + * - Too long: do you really want reboot recovery + * to take more than an hour? Or to make other + * clients wait an hour before being able to + * revoke a dead client's locks? + */ if (i < 10 || i > 3600) return -EINVAL; *time = i; -- cgit v0.10.2 From 4abc14a733f9002c05623db755aaafdd27fa7a91 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Wed, 20 Jan 2010 14:52:23 +0100 Subject: iommu-api: Rename ->{un}map function pointers to ->{un}map_range The new function pointer names match better with the top-level functions of the iommu-api which are using them. Main intention of this change is to make the ->{un}map pointer names free for two new mapping functions. Signed-off-by: Joerg Roedel diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index adb0ba0..59cae7c 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -2515,8 +2515,8 @@ static struct iommu_ops amd_iommu_ops = { .domain_destroy = amd_iommu_domain_destroy, .attach_dev = amd_iommu_attach_device, .detach_dev = amd_iommu_detach_device, - .map = amd_iommu_map_range, - .unmap = amd_iommu_unmap_range, + .map_range = amd_iommu_map_range, + .unmap_range = amd_iommu_unmap_range, .iova_to_phys = amd_iommu_iova_to_phys, .domain_has_cap = amd_iommu_domain_has_cap, }; diff --git a/drivers/base/iommu.c b/drivers/base/iommu.c index 8ad4ffe..f4c86c4 100644 --- a/drivers/base/iommu.c +++ b/drivers/base/iommu.c @@ -83,14 +83,14 @@ EXPORT_SYMBOL_GPL(iommu_detach_device); int iommu_map_range(struct iommu_domain *domain, unsigned long iova, phys_addr_t paddr, size_t size, int prot) { - return iommu_ops->map(domain, iova, paddr, size, prot); + return iommu_ops->map_range(domain, iova, paddr, size, prot); } EXPORT_SYMBOL_GPL(iommu_map_range); void iommu_unmap_range(struct iommu_domain *domain, unsigned long iova, size_t size) { - iommu_ops->unmap(domain, iova, size); + iommu_ops->unmap_range(domain, iova, size); } EXPORT_SYMBOL_GPL(iommu_unmap_range); diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index 4173125..a714e3d 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -3714,8 +3714,8 @@ static struct iommu_ops intel_iommu_ops = { .domain_destroy = intel_iommu_domain_destroy, .attach_dev = intel_iommu_attach_device, .detach_dev = intel_iommu_detach_device, - .map = intel_iommu_map_range, - .unmap = intel_iommu_unmap_range, + .map_range = intel_iommu_map_range, + .unmap_range = intel_iommu_unmap_range, .iova_to_phys = intel_iommu_iova_to_phys, .domain_has_cap = intel_iommu_domain_has_cap, }; diff --git a/include/linux/iommu.h b/include/linux/iommu.h index 3af4ffd..0f18f37 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -36,10 +36,10 @@ struct iommu_ops { void (*domain_destroy)(struct iommu_domain *domain); int (*attach_dev)(struct iommu_domain *domain, struct device *dev); void (*detach_dev)(struct iommu_domain *domain, struct device *dev); - int (*map)(struct iommu_domain *domain, unsigned long iova, - phys_addr_t paddr, size_t size, int prot); - void (*unmap)(struct iommu_domain *domain, unsigned long iova, - size_t size); + int (*map_range)(struct iommu_domain *domain, unsigned long iova, + phys_addr_t paddr, size_t size, int prot); + void (*unmap_range)(struct iommu_domain *domain, unsigned long iova, + size_t size); phys_addr_t (*iova_to_phys)(struct iommu_domain *domain, unsigned long iova); int (*domain_has_cap)(struct iommu_domain *domain, -- cgit v0.10.2 From cefc53c7f494240d4813c80154c7617452d1904d Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Fri, 8 Jan 2010 13:35:09 +0100 Subject: iommu-api: Add iommu_map and iommu_unmap functions These two functions provide support for mapping and unmapping physical addresses to io virtual addresses. The difference to the iommu_(un)map_range() is that the new functions take a gfp_order parameter instead of a size. This allows the IOMMU backend implementations to detect easier if a given range can be mapped by larger page sizes. These new functions should replace the old ones in the long term. Signed-off-by: Joerg Roedel diff --git a/drivers/base/iommu.c b/drivers/base/iommu.c index f4c86c4..cf7cbec 100644 --- a/drivers/base/iommu.c +++ b/drivers/base/iommu.c @@ -107,3 +107,34 @@ int iommu_domain_has_cap(struct iommu_domain *domain, return iommu_ops->domain_has_cap(domain, cap); } EXPORT_SYMBOL_GPL(iommu_domain_has_cap); + +int iommu_map(struct iommu_domain *domain, unsigned long iova, + phys_addr_t paddr, int gfp_order, int prot) +{ + unsigned long invalid_mask; + size_t size; + + size = 0x1000UL << gfp_order; + invalid_mask = size - 1; + + BUG_ON((iova | paddr) & invalid_mask); + + return iommu_ops->map_range(domain, iova, paddr, size, prot); +} +EXPORT_SYMBOL_GPL(iommu_map); + +int iommu_unmap(struct iommu_domain *domain, unsigned long iova, int gfp_order) +{ + unsigned long invalid_mask; + size_t size; + + size = 0x1000UL << gfp_order; + invalid_mask = size - 1; + + BUG_ON(iova & invalid_mask); + + iommu_ops->unmap_range(domain, iova, size); + + return gfp_order; +} +EXPORT_SYMBOL_GPL(iommu_unmap); diff --git a/include/linux/iommu.h b/include/linux/iommu.h index 0f18f37..6d0035b 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -60,6 +60,10 @@ extern int iommu_map_range(struct iommu_domain *domain, unsigned long iova, phys_addr_t paddr, size_t size, int prot); extern void iommu_unmap_range(struct iommu_domain *domain, unsigned long iova, size_t size); +extern int iommu_map(struct iommu_domain *domain, unsigned long iova, + phys_addr_t paddr, int gfp_order, int prot); +extern int iommu_unmap(struct iommu_domain *domain, unsigned long iova, + int gfp_order); extern phys_addr_t iommu_iova_to_phys(struct iommu_domain *domain, unsigned long iova); extern int iommu_domain_has_cap(struct iommu_domain *domain, @@ -108,6 +112,18 @@ static inline void iommu_unmap_range(struct iommu_domain *domain, { } +static inline int iommu_map(struct iommu_domain *domain, unsigned long iova, + phys_addr_t paddr, int gfp_order, int prot) +{ + return -ENODEV; +} + +static inline int iommu_unmap(struct iommu_domain *domain, unsigned long iova, + int gfp_order) +{ + return -ENODEV; +} + static inline phys_addr_t iommu_iova_to_phys(struct iommu_domain *domain, unsigned long iova) { -- cgit v0.10.2 From 67651786948c360c3122b8a17cb1e59209d50880 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Thu, 21 Jan 2010 16:32:27 +0100 Subject: iommu-api: Add ->{un}map callbacks to iommu_ops This patch adds new callbacks for mapping and unmapping pages to the iommu_ops structure. These callbacks are aware of page sizes which makes them different to the ->{un}map_range callbacks. Signed-off-by: Joerg Roedel diff --git a/drivers/base/iommu.c b/drivers/base/iommu.c index cf7cbec..55d37e4 100644 --- a/drivers/base/iommu.c +++ b/drivers/base/iommu.c @@ -119,6 +119,9 @@ int iommu_map(struct iommu_domain *domain, unsigned long iova, BUG_ON((iova | paddr) & invalid_mask); + if (iommu_ops->map) + return iommu_ops->map(domain, iova, paddr, gfp_order, prot); + return iommu_ops->map_range(domain, iova, paddr, size, prot); } EXPORT_SYMBOL_GPL(iommu_map); @@ -133,6 +136,9 @@ int iommu_unmap(struct iommu_domain *domain, unsigned long iova, int gfp_order) BUG_ON(iova & invalid_mask); + if (iommu_ops->unmap) + return iommu_ops->unmap(domain, iova, gfp_order); + iommu_ops->unmap_range(domain, iova, size); return gfp_order; diff --git a/include/linux/iommu.h b/include/linux/iommu.h index 6d0035b..5a7a3d8 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -36,6 +36,10 @@ struct iommu_ops { void (*domain_destroy)(struct iommu_domain *domain); int (*attach_dev)(struct iommu_domain *domain, struct device *dev); void (*detach_dev)(struct iommu_domain *domain, struct device *dev); + int (*map)(struct iommu_domain *domain, unsigned long iova, + phys_addr_t paddr, int gfp_order, int prot); + int (*unmap)(struct iommu_domain *domain, unsigned long iova, + int gfp_order); int (*map_range)(struct iommu_domain *domain, unsigned long iova, phys_addr_t paddr, size_t size, int prot); void (*unmap_range)(struct iommu_domain *domain, unsigned long iova, -- cgit v0.10.2 From b146a1c9f7f1feeacf840fa1ba197a99593cea15 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Wed, 20 Jan 2010 17:17:37 +0100 Subject: VT-d: Change {un}map_range functions to implement {un}map interface This patch changes the iommu-api functions for mapping and unmapping page ranges to use the new page-size based interface. This allows to remove the range based functions later. Signed-off-by: Joerg Roedel diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index a714e3d..371dc56 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -3626,14 +3626,15 @@ static void intel_iommu_detach_device(struct iommu_domain *domain, domain_remove_one_dev_info(dmar_domain, pdev); } -static int intel_iommu_map_range(struct iommu_domain *domain, - unsigned long iova, phys_addr_t hpa, - size_t size, int iommu_prot) +static int intel_iommu_map(struct iommu_domain *domain, + unsigned long iova, phys_addr_t hpa, + int gfp_order, int iommu_prot) { struct dmar_domain *dmar_domain = domain->priv; u64 max_addr; int addr_width; int prot = 0; + size_t size; int ret; if (iommu_prot & IOMMU_READ) @@ -3643,6 +3644,7 @@ static int intel_iommu_map_range(struct iommu_domain *domain, if ((iommu_prot & IOMMU_CACHE) && dmar_domain->iommu_snooping) prot |= DMA_PTE_SNP; + size = PAGE_SIZE << gfp_order; max_addr = iova + size; if (dmar_domain->max_addr < max_addr) { int min_agaw; @@ -3669,19 +3671,19 @@ static int intel_iommu_map_range(struct iommu_domain *domain, return ret; } -static void intel_iommu_unmap_range(struct iommu_domain *domain, - unsigned long iova, size_t size) +static int intel_iommu_unmap(struct iommu_domain *domain, + unsigned long iova, int gfp_order) { struct dmar_domain *dmar_domain = domain->priv; - - if (!size) - return; + size_t size = PAGE_SIZE << gfp_order; dma_pte_clear_range(dmar_domain, iova >> VTD_PAGE_SHIFT, (iova + size - 1) >> VTD_PAGE_SHIFT); if (dmar_domain->max_addr == iova + size) dmar_domain->max_addr = iova; + + return gfp_order; } static phys_addr_t intel_iommu_iova_to_phys(struct iommu_domain *domain, @@ -3714,8 +3716,8 @@ static struct iommu_ops intel_iommu_ops = { .domain_destroy = intel_iommu_domain_destroy, .attach_dev = intel_iommu_attach_device, .detach_dev = intel_iommu_detach_device, - .map_range = intel_iommu_map_range, - .unmap_range = intel_iommu_unmap_range, + .map = intel_iommu_map, + .unmap = intel_iommu_unmap, .iova_to_phys = intel_iommu_iova_to_phys, .domain_has_cap = intel_iommu_domain_has_cap, }; -- cgit v0.10.2 From fcd95807fb61e67d602610e7ff7129ed769e9fee Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Mon, 11 Jan 2010 16:38:18 +0100 Subject: kvm: Change kvm_iommu_map_pages to map large pages This patch changes the implementation of of kvm_iommu_map_pages to map the pages with the host page size into the io virtual address space. Signed-off-by: Joerg Roedel Acked-By: Avi Kivity diff --git a/virt/kvm/iommu.c b/virt/kvm/iommu.c index 80fd3ad..11692b9 100644 --- a/virt/kvm/iommu.c +++ b/virt/kvm/iommu.c @@ -32,12 +32,30 @@ static int kvm_iommu_unmap_memslots(struct kvm *kvm); static void kvm_iommu_put_pages(struct kvm *kvm, gfn_t base_gfn, unsigned long npages); +static pfn_t kvm_pin_pages(struct kvm *kvm, struct kvm_memory_slot *slot, + gfn_t gfn, unsigned long size) +{ + gfn_t end_gfn; + pfn_t pfn; + + pfn = gfn_to_pfn_memslot(kvm, slot, gfn); + end_gfn = gfn + (size >> PAGE_SHIFT); + gfn += 1; + + if (is_error_pfn(pfn)) + return pfn; + + while (gfn < end_gfn) + gfn_to_pfn_memslot(kvm, slot, gfn++); + + return pfn; +} + int kvm_iommu_map_pages(struct kvm *kvm, struct kvm_memory_slot *slot) { - gfn_t gfn = slot->base_gfn; - unsigned long npages = slot->npages; + gfn_t gfn, end_gfn; pfn_t pfn; - int i, r = 0; + int r = 0; struct iommu_domain *domain = kvm->arch.iommu_domain; int flags; @@ -45,31 +63,62 @@ int kvm_iommu_map_pages(struct kvm *kvm, struct kvm_memory_slot *slot) if (!domain) return 0; + gfn = slot->base_gfn; + end_gfn = gfn + slot->npages; + flags = IOMMU_READ | IOMMU_WRITE; if (kvm->arch.iommu_flags & KVM_IOMMU_CACHE_COHERENCY) flags |= IOMMU_CACHE; - for (i = 0; i < npages; i++) { - /* check if already mapped */ - if (iommu_iova_to_phys(domain, gfn_to_gpa(gfn))) + + while (gfn < end_gfn) { + unsigned long page_size; + + /* Check if already mapped */ + if (iommu_iova_to_phys(domain, gfn_to_gpa(gfn))) { + gfn += 1; + continue; + } + + /* Get the page size we could use to map */ + page_size = kvm_host_page_size(kvm, gfn); + + /* Make sure the page_size does not exceed the memslot */ + while ((gfn + (page_size >> PAGE_SHIFT)) > end_gfn) + page_size >>= 1; + + /* Make sure gfn is aligned to the page size we want to map */ + while ((gfn << PAGE_SHIFT) & (page_size - 1)) + page_size >>= 1; + + /* + * Pin all pages we are about to map in memory. This is + * important because we unmap and unpin in 4kb steps later. + */ + pfn = kvm_pin_pages(kvm, slot, gfn, page_size); + if (is_error_pfn(pfn)) { + gfn += 1; continue; + } - pfn = gfn_to_pfn_memslot(kvm, slot, gfn); - r = iommu_map_range(domain, - gfn_to_gpa(gfn), - pfn_to_hpa(pfn), - PAGE_SIZE, flags); + /* Map into IO address space */ + r = iommu_map(domain, gfn_to_gpa(gfn), pfn_to_hpa(pfn), + get_order(page_size), flags); if (r) { printk(KERN_ERR "kvm_iommu_map_address:" "iommu failed to map pfn=%lx\n", pfn); goto unmap_pages; } - gfn++; + + gfn += page_size >> PAGE_SHIFT; + + } + return 0; unmap_pages: - kvm_iommu_put_pages(kvm, slot->base_gfn, i); + kvm_iommu_put_pages(kvm, slot->base_gfn, gfn); return r; } @@ -189,27 +238,47 @@ out_unmap: return r; } +static void kvm_unpin_pages(struct kvm *kvm, pfn_t pfn, unsigned long npages) +{ + unsigned long i; + + for (i = 0; i < npages; ++i) + kvm_release_pfn_clean(pfn + i); +} + static void kvm_iommu_put_pages(struct kvm *kvm, gfn_t base_gfn, unsigned long npages) { - gfn_t gfn = base_gfn; + struct iommu_domain *domain; + gfn_t end_gfn, gfn; pfn_t pfn; - struct iommu_domain *domain = kvm->arch.iommu_domain; - unsigned long i; u64 phys; + domain = kvm->arch.iommu_domain; + end_gfn = base_gfn + npages; + gfn = base_gfn; + /* check if iommu exists and in use */ if (!domain) return; - for (i = 0; i < npages; i++) { + while (gfn < end_gfn) { + unsigned long unmap_pages; + int order; + + /* Get physical address */ phys = iommu_iova_to_phys(domain, gfn_to_gpa(gfn)); - pfn = phys >> PAGE_SHIFT; - kvm_release_pfn_clean(pfn); - gfn++; - } + pfn = phys >> PAGE_SHIFT; + + /* Unmap address from IO address space */ + order = iommu_unmap(domain, gfn_to_gpa(gfn), PAGE_SIZE); + unmap_pages = 1ULL << order; - iommu_unmap_range(domain, gfn_to_gpa(base_gfn), PAGE_SIZE * npages); + /* Unpin all pages we just unmapped to not leak any memory */ + kvm_unpin_pages(kvm, pfn, unmap_pages); + + gfn += unmap_pages; + } } static int kvm_iommu_unmap_memslots(struct kvm *kvm) -- cgit v0.10.2 From cbb9d729f3433c9c2660b01dc52e6deb89488886 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Fri, 15 Jan 2010 14:41:15 +0100 Subject: x86/amd-iommu: Make iommu_map_page and alloc_pte aware of page sizes This patch changes the old map_size parameter of alloc_pte to a page_size parameter which can be used more easily to alloc a pte for intermediate page sizes. Signed-off-by: Joerg Roedel diff --git a/arch/x86/include/asm/amd_iommu_types.h b/arch/x86/include/asm/amd_iommu_types.h index ba19ad4..5e8da56 100644 --- a/arch/x86/include/asm/amd_iommu_types.h +++ b/arch/x86/include/asm/amd_iommu_types.h @@ -172,6 +172,34 @@ (~((1ULL << (12 + ((lvl) * 9))) - 1))) #define PM_ALIGNED(lvl, addr) ((PM_MAP_MASK(lvl) & (addr)) == (addr)) +/* + * Returns the page table level to use for a given page size + * Pagesize is expected to be a power-of-two + */ +#define PAGE_SIZE_LEVEL(pagesize) \ + ((__ffs(pagesize) - 12) / 9) +/* + * Returns the number of ptes to use for a given page size + * Pagesize is expected to be a power-of-two + */ +#define PAGE_SIZE_PTE_COUNT(pagesize) \ + (1ULL << ((__ffs(pagesize) - 12) % 9)) + +/* + * Aligns a given io-virtual address to a given page size + * Pagesize is expected to be a power-of-two + */ +#define PAGE_SIZE_ALIGN(address, pagesize) \ + ((address) & ~((pagesize) - 1)) +/* + * Creates an IOMMU PTE for an address an a given pagesize + * The PTE has no permission bits set + * Pagesize is expected to be a power-of-two larger than 4096 + */ +#define PAGE_SIZE_PTE(address, pagesize) \ + (((address) | ((pagesize) - 1)) & \ + (~(pagesize >> 1)) & PM_ADDR_MASK) + #define IOMMU_PTE_P (1ULL << 0) #define IOMMU_PTE_TV (1ULL << 1) #define IOMMU_PTE_U (1ULL << 59) diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 59cae7c..4170031 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -730,18 +730,22 @@ static bool increase_address_space(struct protection_domain *domain, static u64 *alloc_pte(struct protection_domain *domain, unsigned long address, - int end_lvl, + unsigned long page_size, u64 **pte_page, gfp_t gfp) { + int level, end_lvl; u64 *pte, *page; - int level; + + BUG_ON(!is_power_of_2(page_size)); while (address > PM_LEVEL_SIZE(domain->mode)) increase_address_space(domain, gfp); - level = domain->mode - 1; - pte = &domain->pt_root[PM_LEVEL_INDEX(level, address)]; + level = domain->mode - 1; + pte = &domain->pt_root[PM_LEVEL_INDEX(level, address)]; + address = PAGE_SIZE_ALIGN(address, page_size); + end_lvl = PAGE_SIZE_LEVEL(page_size); while (level > end_lvl) { if (!IOMMU_PTE_PRESENT(*pte)) { @@ -751,6 +755,10 @@ static u64 *alloc_pte(struct protection_domain *domain, *pte = PM_LEVEL_PDE(level, virt_to_phys(page)); } + /* No level skipping support yet */ + if (PM_PTE_LEVEL(*pte) != level) + return NULL; + level -= 1; pte = IOMMU_PTE_PAGE(*pte); @@ -806,31 +814,36 @@ static int iommu_map_page(struct protection_domain *dom, unsigned long bus_addr, unsigned long phys_addr, int prot, - int map_size) + unsigned long page_size) { u64 __pte, *pte; - - bus_addr = PAGE_ALIGN(bus_addr); - phys_addr = PAGE_ALIGN(phys_addr); - - BUG_ON(!PM_ALIGNED(map_size, bus_addr)); - BUG_ON(!PM_ALIGNED(map_size, phys_addr)); + int i, count; if (!(prot & IOMMU_PROT_MASK)) return -EINVAL; - pte = alloc_pte(dom, bus_addr, map_size, NULL, GFP_KERNEL); + bus_addr = PAGE_ALIGN(bus_addr); + phys_addr = PAGE_ALIGN(phys_addr); + count = PAGE_SIZE_PTE_COUNT(page_size); + pte = alloc_pte(dom, bus_addr, page_size, NULL, GFP_KERNEL); + + for (i = 0; i < count; ++i) + if (IOMMU_PTE_PRESENT(pte[i])) + return -EBUSY; - if (IOMMU_PTE_PRESENT(*pte)) - return -EBUSY; + if (page_size > PAGE_SIZE) { + __pte = PAGE_SIZE_PTE(phys_addr, page_size); + __pte |= PM_LEVEL_ENC(7) | IOMMU_PTE_P | IOMMU_PTE_FC; + } else + __pte = phys_addr | IOMMU_PTE_P | IOMMU_PTE_FC; - __pte = phys_addr | IOMMU_PTE_P; if (prot & IOMMU_PROT_IR) __pte |= IOMMU_PTE_IR; if (prot & IOMMU_PROT_IW) __pte |= IOMMU_PTE_IW; - *pte = __pte; + for (i = 0; i < count; ++i) + pte[i] = __pte; update_domain(dom); @@ -877,7 +890,7 @@ static int dma_ops_unity_map(struct dma_ops_domain *dma_dom, for (addr = e->address_start; addr < e->address_end; addr += PAGE_SIZE) { ret = iommu_map_page(&dma_dom->domain, addr, addr, e->prot, - PM_MAP_4k); + PAGE_SIZE); if (ret) return ret; /* @@ -1005,7 +1018,7 @@ static int alloc_new_range(struct dma_ops_domain *dma_dom, u64 *pte, *pte_page; for (i = 0; i < num_ptes; ++i) { - pte = alloc_pte(&dma_dom->domain, address, PM_MAP_4k, + pte = alloc_pte(&dma_dom->domain, address, PAGE_SIZE, &pte_page, gfp); if (!pte) goto out_free; @@ -1711,7 +1724,7 @@ static u64* dma_ops_get_pte(struct dma_ops_domain *dom, pte = aperture->pte_pages[APERTURE_PAGE_INDEX(address)]; if (!pte) { - pte = alloc_pte(&dom->domain, address, PM_MAP_4k, &pte_page, + pte = alloc_pte(&dom->domain, address, PAGE_SIZE, &pte_page, GFP_ATOMIC); aperture->pte_pages[APERTURE_PAGE_INDEX(address)] = pte_page; } else @@ -2457,7 +2470,7 @@ static int amd_iommu_map_range(struct iommu_domain *dom, paddr &= PAGE_MASK; for (i = 0; i < npages; ++i) { - ret = iommu_map_page(domain, iova, paddr, prot, PM_MAP_4k); + ret = iommu_map_page(domain, iova, paddr, prot, PAGE_SIZE); if (ret) return ret; -- cgit v0.10.2 From 24cd772315c19e4d9409d0d21367ec1ebab3149f Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Tue, 19 Jan 2010 17:27:39 +0100 Subject: x86/amd-iommu: Make iommu_unmap_page and fetch_pte aware of page sizes This patch extends the functionality of iommu_unmap_page and fetch_pte to support arbitrary page sizes. Signed-off-by: Joerg Roedel diff --git a/arch/x86/include/asm/amd_iommu_types.h b/arch/x86/include/asm/amd_iommu_types.h index 5e8da56..b150c74 100644 --- a/arch/x86/include/asm/amd_iommu_types.h +++ b/arch/x86/include/asm/amd_iommu_types.h @@ -200,6 +200,12 @@ (((address) | ((pagesize) - 1)) & \ (~(pagesize >> 1)) & PM_ADDR_MASK) +/* + * Takes a PTE value with mode=0x07 and returns the page size it maps + */ +#define PTE_PAGE_SIZE(pte) \ + (1ULL << (1 + ffz(((pte) | 0xfffULL)))) + #define IOMMU_PTE_P (1ULL << 0) #define IOMMU_PTE_TV (1ULL << 1) #define IOMMU_PTE_U (1ULL << 59) diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 4170031..503d312 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -776,28 +776,47 @@ static u64 *alloc_pte(struct protection_domain *domain, * This function checks if there is a PTE for a given dma address. If * there is one, it returns the pointer to it. */ -static u64 *fetch_pte(struct protection_domain *domain, - unsigned long address, int map_size) +static u64 *fetch_pte(struct protection_domain *domain, unsigned long address) { int level; u64 *pte; - level = domain->mode - 1; - pte = &domain->pt_root[PM_LEVEL_INDEX(level, address)]; + if (address > PM_LEVEL_SIZE(domain->mode)) + return NULL; + + level = domain->mode - 1; + pte = &domain->pt_root[PM_LEVEL_INDEX(level, address)]; - while (level > map_size) { + while (level > 0) { + + /* Not Present */ if (!IOMMU_PTE_PRESENT(*pte)) return NULL; + /* Large PTE */ + if (PM_PTE_LEVEL(*pte) == 0x07) { + unsigned long pte_mask, __pte; + + /* + * If we have a series of large PTEs, make + * sure to return a pointer to the first one. + */ + pte_mask = PTE_PAGE_SIZE(*pte); + pte_mask = ~((PAGE_SIZE_PTE_COUNT(pte_mask) << 3) - 1); + __pte = ((unsigned long)pte) & pte_mask; + + return (u64 *)__pte; + } + + /* No level skipping support yet */ + if (PM_PTE_LEVEL(*pte) != level) + return NULL; + level -= 1; + /* Walk to the next level */ pte = IOMMU_PTE_PAGE(*pte); pte = &pte[PM_LEVEL_INDEX(level, address)]; - - if ((PM_PTE_LEVEL(*pte) == 0) && level != map_size) { - pte = NULL; - break; - } } return pte; @@ -850,13 +869,48 @@ static int iommu_map_page(struct protection_domain *dom, return 0; } -static void iommu_unmap_page(struct protection_domain *dom, - unsigned long bus_addr, int map_size) +static unsigned long iommu_unmap_page(struct protection_domain *dom, + unsigned long bus_addr, + unsigned long page_size) { - u64 *pte = fetch_pte(dom, bus_addr, map_size); + unsigned long long unmap_size, unmapped; + u64 *pte; + + BUG_ON(!is_power_of_2(page_size)); + + unmapped = 0; + + while (unmapped < page_size) { + + pte = fetch_pte(dom, bus_addr); + + if (!pte) { + /* + * No PTE for this address + * move forward in 4kb steps + */ + unmap_size = PAGE_SIZE; + } else if (PM_PTE_LEVEL(*pte) == 0) { + /* 4kb PTE found for this address */ + unmap_size = PAGE_SIZE; + *pte = 0ULL; + } else { + int count, i; + + /* Large PTE found which maps this address */ + unmap_size = PTE_PAGE_SIZE(*pte); + count = PAGE_SIZE_PTE_COUNT(unmap_size); + for (i = 0; i < count; i++) + pte[i] = 0ULL; + } + + bus_addr = (bus_addr & ~(unmap_size - 1)) + unmap_size; + unmapped += unmap_size; + } + + BUG_ON(!is_power_of_2(unmapped)); - if (pte) - *pte = 0; + return unmapped; } /* @@ -1054,7 +1108,7 @@ static int alloc_new_range(struct dma_ops_domain *dma_dom, for (i = dma_dom->aperture[index]->offset; i < dma_dom->aperture_size; i += PAGE_SIZE) { - u64 *pte = fetch_pte(&dma_dom->domain, i, PM_MAP_4k); + u64 *pte = fetch_pte(&dma_dom->domain, i); if (!pte || !IOMMU_PTE_PRESENT(*pte)) continue; @@ -2491,7 +2545,7 @@ static void amd_iommu_unmap_range(struct iommu_domain *dom, iova &= PAGE_MASK; for (i = 0; i < npages; ++i) { - iommu_unmap_page(domain, iova, PM_MAP_4k); + iommu_unmap_page(domain, iova, PAGE_SIZE); iova += PAGE_SIZE; } @@ -2506,7 +2560,7 @@ static phys_addr_t amd_iommu_iova_to_phys(struct iommu_domain *dom, phys_addr_t paddr; u64 *pte; - pte = fetch_pte(domain, iova, PM_MAP_4k); + pte = fetch_pte(domain, iova); if (!pte || !IOMMU_PTE_PRESENT(*pte)) return 0; -- cgit v0.10.2 From f03152bb7d0a74f409ad63ed36916444a7493d72 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Thu, 21 Jan 2010 16:15:24 +0100 Subject: x86/amd-iommu: Make amd_iommu_iova_to_phys aware of multiple page sizes This patch extends the amd_iommu_iova_to_phys() function to handle different page sizes correctly. It doesn't use fetch_pte() anymore because we don't know (or care about) the page_size used for mapping the given iova. Signed-off-by: Joerg Roedel diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 503d312..52e44af 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -2556,17 +2556,22 @@ static phys_addr_t amd_iommu_iova_to_phys(struct iommu_domain *dom, unsigned long iova) { struct protection_domain *domain = dom->priv; - unsigned long offset = iova & ~PAGE_MASK; + unsigned long offset_mask; phys_addr_t paddr; - u64 *pte; + u64 *pte, __pte; pte = fetch_pte(domain, iova); if (!pte || !IOMMU_PTE_PRESENT(*pte)) return 0; - paddr = *pte & IOMMU_PAGE_MASK; - paddr |= offset; + if (PM_PTE_LEVEL(*pte) == 0) + offset_mask = PAGE_SIZE - 1; + else + offset_mask = PTE_PAGE_SIZE(*pte) - 1; + + __pte = *pte & PM_ADDR_MASK; + paddr = (__pte & ~offset_mask) | (iova & offset_mask); return paddr; } -- cgit v0.10.2 From 468e2366cdb80cf8a691b8bc212260cfbdbd518e Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Thu, 21 Jan 2010 16:37:36 +0100 Subject: x86/amd-iommu: Implement ->{un}map callbacks for iommu-api This patch implements the new callbacks for the IOMMU-API with functions that can handle different page sizes in the IOMMU page table. Signed-off-by: Joerg Roedel diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 52e44af..0e068c9 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -2552,6 +2552,33 @@ static void amd_iommu_unmap_range(struct iommu_domain *dom, iommu_flush_tlb_pde(domain); } +static int amd_iommu_map(struct iommu_domain *dom, unsigned long iova, + phys_addr_t paddr, int gfp_order, int iommu_prot) +{ + unsigned long page_size = 0x1000UL << gfp_order; + struct protection_domain *domain = dom->priv; + int prot = 0; + + if (iommu_prot & IOMMU_READ) + prot |= IOMMU_PROT_IR; + if (iommu_prot & IOMMU_WRITE) + prot |= IOMMU_PROT_IW; + + return iommu_map_page(domain, iova, paddr, prot, page_size); +} + +static int amd_iommu_unmap(struct iommu_domain *dom, unsigned long iova, + int gfp_order) +{ + struct protection_domain *domain = dom->priv; + unsigned long page_size, unmap_size; + + page_size = 0x1000UL << gfp_order; + unmap_size = iommu_unmap_page(domain, iova, page_size); + + return get_order(unmap_size); +} + static phys_addr_t amd_iommu_iova_to_phys(struct iommu_domain *dom, unsigned long iova) { @@ -2587,6 +2614,8 @@ static struct iommu_ops amd_iommu_ops = { .domain_destroy = amd_iommu_domain_destroy, .attach_dev = amd_iommu_attach_device, .detach_dev = amd_iommu_detach_device, + .map = amd_iommu_map, + .unmap = amd_iommu_unmap, .map_range = amd_iommu_map_range, .unmap_range = amd_iommu_unmap_range, .iova_to_phys = amd_iommu_iova_to_phys, -- cgit v0.10.2 From 12c7389abe5786349d3ea6da1961cf78d0c1c7cd Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Thu, 21 Jan 2010 11:50:28 +0100 Subject: iommu-api: Remove iommu_{un}map_range functions These functions are not longer used and can be removed savely. There functionality is now provided by the iommu_{un}map functions which are also capable of multiple page sizes. Signed-off-by: Joerg Roedel diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 0e068c9..d8da998 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -2506,52 +2506,6 @@ static int amd_iommu_attach_device(struct iommu_domain *dom, return ret; } -static int amd_iommu_map_range(struct iommu_domain *dom, - unsigned long iova, phys_addr_t paddr, - size_t size, int iommu_prot) -{ - struct protection_domain *domain = dom->priv; - unsigned long i, npages = iommu_num_pages(paddr, size, PAGE_SIZE); - int prot = 0; - int ret; - - if (iommu_prot & IOMMU_READ) - prot |= IOMMU_PROT_IR; - if (iommu_prot & IOMMU_WRITE) - prot |= IOMMU_PROT_IW; - - iova &= PAGE_MASK; - paddr &= PAGE_MASK; - - for (i = 0; i < npages; ++i) { - ret = iommu_map_page(domain, iova, paddr, prot, PAGE_SIZE); - if (ret) - return ret; - - iova += PAGE_SIZE; - paddr += PAGE_SIZE; - } - - return 0; -} - -static void amd_iommu_unmap_range(struct iommu_domain *dom, - unsigned long iova, size_t size) -{ - - struct protection_domain *domain = dom->priv; - unsigned long i, npages = iommu_num_pages(iova, size, PAGE_SIZE); - - iova &= PAGE_MASK; - - for (i = 0; i < npages; ++i) { - iommu_unmap_page(domain, iova, PAGE_SIZE); - iova += PAGE_SIZE; - } - - iommu_flush_tlb_pde(domain); -} - static int amd_iommu_map(struct iommu_domain *dom, unsigned long iova, phys_addr_t paddr, int gfp_order, int iommu_prot) { @@ -2616,8 +2570,6 @@ static struct iommu_ops amd_iommu_ops = { .detach_dev = amd_iommu_detach_device, .map = amd_iommu_map, .unmap = amd_iommu_unmap, - .map_range = amd_iommu_map_range, - .unmap_range = amd_iommu_unmap_range, .iova_to_phys = amd_iommu_iova_to_phys, .domain_has_cap = amd_iommu_domain_has_cap, }; diff --git a/drivers/base/iommu.c b/drivers/base/iommu.c index 55d37e4..6e6b6a1 100644 --- a/drivers/base/iommu.c +++ b/drivers/base/iommu.c @@ -80,20 +80,6 @@ void iommu_detach_device(struct iommu_domain *domain, struct device *dev) } EXPORT_SYMBOL_GPL(iommu_detach_device); -int iommu_map_range(struct iommu_domain *domain, unsigned long iova, - phys_addr_t paddr, size_t size, int prot) -{ - return iommu_ops->map_range(domain, iova, paddr, size, prot); -} -EXPORT_SYMBOL_GPL(iommu_map_range); - -void iommu_unmap_range(struct iommu_domain *domain, unsigned long iova, - size_t size) -{ - iommu_ops->unmap_range(domain, iova, size); -} -EXPORT_SYMBOL_GPL(iommu_unmap_range); - phys_addr_t iommu_iova_to_phys(struct iommu_domain *domain, unsigned long iova) { @@ -119,10 +105,7 @@ int iommu_map(struct iommu_domain *domain, unsigned long iova, BUG_ON((iova | paddr) & invalid_mask); - if (iommu_ops->map) - return iommu_ops->map(domain, iova, paddr, gfp_order, prot); - - return iommu_ops->map_range(domain, iova, paddr, size, prot); + return iommu_ops->map(domain, iova, paddr, gfp_order, prot); } EXPORT_SYMBOL_GPL(iommu_map); @@ -136,11 +119,6 @@ int iommu_unmap(struct iommu_domain *domain, unsigned long iova, int gfp_order) BUG_ON(iova & invalid_mask); - if (iommu_ops->unmap) - return iommu_ops->unmap(domain, iova, gfp_order); - - iommu_ops->unmap_range(domain, iova, size); - - return gfp_order; + return iommu_ops->unmap(domain, iova, gfp_order); } EXPORT_SYMBOL_GPL(iommu_unmap); diff --git a/include/linux/iommu.h b/include/linux/iommu.h index 5a7a3d8..be22ad8 100644 --- a/include/linux/iommu.h +++ b/include/linux/iommu.h @@ -40,10 +40,6 @@ struct iommu_ops { phys_addr_t paddr, int gfp_order, int prot); int (*unmap)(struct iommu_domain *domain, unsigned long iova, int gfp_order); - int (*map_range)(struct iommu_domain *domain, unsigned long iova, - phys_addr_t paddr, size_t size, int prot); - void (*unmap_range)(struct iommu_domain *domain, unsigned long iova, - size_t size); phys_addr_t (*iova_to_phys)(struct iommu_domain *domain, unsigned long iova); int (*domain_has_cap)(struct iommu_domain *domain, @@ -60,10 +56,6 @@ extern int iommu_attach_device(struct iommu_domain *domain, struct device *dev); extern void iommu_detach_device(struct iommu_domain *domain, struct device *dev); -extern int iommu_map_range(struct iommu_domain *domain, unsigned long iova, - phys_addr_t paddr, size_t size, int prot); -extern void iommu_unmap_range(struct iommu_domain *domain, unsigned long iova, - size_t size); extern int iommu_map(struct iommu_domain *domain, unsigned long iova, phys_addr_t paddr, int gfp_order, int prot); extern int iommu_unmap(struct iommu_domain *domain, unsigned long iova, @@ -104,18 +96,6 @@ static inline void iommu_detach_device(struct iommu_domain *domain, { } -static inline int iommu_map_range(struct iommu_domain *domain, - unsigned long iova, phys_addr_t paddr, - size_t size, int prot) -{ - return -ENODEV; -} - -static inline void iommu_unmap_range(struct iommu_domain *domain, - unsigned long iova, size_t size) -{ -} - static inline int iommu_map(struct iommu_domain *domain, unsigned long iova, phys_addr_t paddr, int gfp_order, int prot) { -- cgit v0.10.2 From 8bf8c376ab2eefaf0386f4e003e720e1434fa43d Mon Sep 17 00:00:00 2001 From: Dmitry Monakhov Date: Wed, 3 Mar 2010 06:28:06 +0300 Subject: blkdev: fix merge_bvec_fn return value checks v2 merge_bvec_fn() returns bvec->bv_len on success. So we have to check against this value. But in case of fs_optimization merge we compare with wrong value. This patch must be included in b428cd6da7e6559aca69aa2e3a526037d3f20403 But accidentally i've forgot to add this in the initial patch. To make things straight let's replace all such checks. In fact this makes code easy to understand. Signed-off-by: Dmitry Monakhov Signed-off-by: Jens Axboe diff --git a/fs/bio.c b/fs/bio.c index dc17afd..d89e019 100644 --- a/fs/bio.c +++ b/fs/bio.c @@ -555,7 +555,7 @@ static int __bio_add_page(struct request_queue *q, struct bio *bio, struct page .bi_rw = bio->bi_rw, }; - if (q->merge_bvec_fn(q, &bvm, prev) < len) { + if (q->merge_bvec_fn(q, &bvm, prev) < prev->bv_len) { prev->bv_len -= len; return 0; } @@ -608,7 +608,7 @@ static int __bio_add_page(struct request_queue *q, struct bio *bio, struct page * merge_bvec_fn() returns number of bytes it can accept * at this offset */ - if (q->merge_bvec_fn(q, &bvm, bvec) < len) { + if (q->merge_bvec_fn(q, &bvm, bvec) < bvec->bv_len) { bvec->bv_page = NULL; bvec->bv_len = 0; bvec->bv_offset = 0; -- cgit v0.10.2 From c2282adbdea3548ae0271c1b1b2deec5f56ad224 Mon Sep 17 00:00:00 2001 From: FUJITA Tomonori Date: Mon, 8 Mar 2010 09:11:07 +0100 Subject: Documentation: fix block/biodoc.txt dma mapping description - It looks incorrect to use blk_rq_map_sg with pci_map_page here about DMA mappings. dma_map_sg? - better to use dma_map_page instead of pci_map_page. http://marc.info/?l=linux-kernel&m=126596737604808&w=2 Signed-off-by: FUJITA Tomonori Signed-off-by: Jens Axboe diff --git a/Documentation/block/biodoc.txt b/Documentation/block/biodoc.txt index 6fab97e..508b5b2 100644 --- a/Documentation/block/biodoc.txt +++ b/Documentation/block/biodoc.txt @@ -1162,8 +1162,8 @@ where a driver received a request ala this before: As mentioned, there is no virtual mapping of a bio. For DMA, this is not a problem as the driver probably never will need a virtual mapping. -Instead it needs a bus mapping (pci_map_page for a single segment or -use blk_rq_map_sg for scatter gather) to be able to ship it to the driver. For +Instead it needs a bus mapping (dma_map_page for a single segment or +use dma_map_sg for scatter gather) to be able to ship it to the driver. For PIO drivers (or drivers that need to revert to PIO transfer once in a while (IDE for example)), where the CPU is doing the actual data transfer a virtual mapping is needed. If the driver supports highmem I/O, -- cgit v0.10.2 From 8b408fe4f853dcfa18d133aa4cf1d7546b4c3870 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Mon, 8 Mar 2010 14:20:07 +0100 Subject: x86/amd-iommu: Use helper function to destroy domain In the amd_iommu_domain_destroy the protection_domain_free function is partly reimplemented. The 'partly' is the bug here because the domain is not deleted from the domain list. This results in use-after-free errors and data-corruption. Fix it by just using protection_domain_free instead. Cc: stable@kernel.org Signed-off-by: Joerg Roedel diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 0c04254..b06f29e 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -2380,9 +2380,7 @@ static void amd_iommu_domain_destroy(struct iommu_domain *dom) free_pagetable(domain); - domain_id_free(domain->id); - - kfree(domain); + protection_domain_free(domain); dom->priv = NULL; } -- cgit v0.10.2 From 51a5006afcd13176276f0935ee57c4cc6f210e83 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 8 Mar 2010 21:45:19 +0900 Subject: sh: Merge clkdev API updates. This rolls in the remainder of the clkdev API bits from the ARM tree. This can more or less be used verbatim, so we just copy it over and nuke our local version. Signed-off-by: Paul Mundt diff --git a/arch/sh/include/asm/clkdev.h b/arch/sh/include/asm/clkdev.h new file mode 100644 index 0000000..5645f35 --- /dev/null +++ b/arch/sh/include/asm/clkdev.h @@ -0,0 +1,35 @@ +/* + * arch/sh/include/asm/clkdev.h + * + * Cloned from arch/arm/include/asm/clkdev.h: + * + * Copyright (C) 2008 Russell King. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * Helper for the clk API to assist looking up a struct clk. + */ +#ifndef __ASM_CLKDEV_H +#define __ASM_CLKDEV_H + +struct clk; + +struct clk_lookup { + struct list_head node; + const char *dev_id; + const char *con_id; + struct clk *clk; +}; + +struct clk_lookup *clkdev_alloc(struct clk *clk, const char *con_id, + const char *dev_fmt, ...); + +void clkdev_add(struct clk_lookup *cl); +void clkdev_drop(struct clk_lookup *cl); + +void clkdev_add_table(struct clk_lookup *, size_t); +int clk_add_alias(const char *, const char *, char *, struct device *); + +#endif diff --git a/arch/sh/include/asm/clock.h b/arch/sh/include/asm/clock.h index 11da4c5..4b19179 100644 --- a/arch/sh/include/asm/clock.h +++ b/arch/sh/include/asm/clock.h @@ -45,13 +45,6 @@ struct clk { struct cpufreq_frequency_table *freq_table; }; -struct clk_lookup { - struct list_head node; - const char *dev_id; - const char *con_id; - struct clk *clk; -}; - #define CLK_ENABLE_ON_INIT (1 << 0) /* Should be defined by processor-specific code */ diff --git a/arch/sh/kernel/Makefile b/arch/sh/kernel/Makefile index 02fd3ae..650b92f 100644 --- a/arch/sh/kernel/Makefile +++ b/arch/sh/kernel/Makefile @@ -11,7 +11,7 @@ endif CFLAGS_REMOVE_return_address.o = -pg -obj-y := debugtraps.o dma-nommu.o dumpstack.o \ +obj-y := clkdev.o debugtraps.o dma-nommu.o dumpstack.o \ idle.o io.o io_generic.o irq.o \ irq_$(BITS).o machvec.o nmi_debug.o process.o \ process_$(BITS).o ptrace_$(BITS).o \ diff --git a/arch/sh/kernel/clkdev.c b/arch/sh/kernel/clkdev.c new file mode 100644 index 0000000..29cd802 --- /dev/null +++ b/arch/sh/kernel/clkdev.c @@ -0,0 +1,162 @@ +/* + * arch/sh/kernel/clkdev.c + * + * Cloned from arch/arm/common/clkdev.c: + * + * Copyright (C) 2008 Russell King. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * Helper for the clk API to assist looking up a struct clk. + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +static LIST_HEAD(clocks); +static DEFINE_MUTEX(clocks_mutex); + +/* + * Find the correct struct clk for the device and connection ID. + * We do slightly fuzzy matching here: + * An entry with a NULL ID is assumed to be a wildcard. + * If an entry has a device ID, it must match + * If an entry has a connection ID, it must match + * Then we take the most specific entry - with the following + * order of precidence: dev+con > dev only > con only. + */ +static struct clk *clk_find(const char *dev_id, const char *con_id) +{ + struct clk_lookup *p; + struct clk *clk = NULL; + int match, best = 0; + + list_for_each_entry(p, &clocks, node) { + match = 0; + if (p->dev_id) { + if (!dev_id || strcmp(p->dev_id, dev_id)) + continue; + match += 2; + } + if (p->con_id) { + if (!con_id || strcmp(p->con_id, con_id)) + continue; + match += 1; + } + if (match == 0) + continue; + + if (match > best) { + clk = p->clk; + best = match; + } + } + return clk; +} + +struct clk *clk_get_sys(const char *dev_id, const char *con_id) +{ + struct clk *clk; + + mutex_lock(&clocks_mutex); + clk = clk_find(dev_id, con_id); + mutex_unlock(&clocks_mutex); + + return clk ? clk : ERR_PTR(-ENOENT); +} +EXPORT_SYMBOL(clk_get_sys); + +void clkdev_add(struct clk_lookup *cl) +{ + mutex_lock(&clocks_mutex); + list_add_tail(&cl->node, &clocks); + mutex_unlock(&clocks_mutex); +} +EXPORT_SYMBOL(clkdev_add); + +void __init clkdev_add_table(struct clk_lookup *cl, size_t num) +{ + mutex_lock(&clocks_mutex); + while (num--) { + list_add_tail(&cl->node, &clocks); + cl++; + } + mutex_unlock(&clocks_mutex); +} + +#define MAX_DEV_ID 20 +#define MAX_CON_ID 16 + +struct clk_lookup_alloc { + struct clk_lookup cl; + char dev_id[MAX_DEV_ID]; + char con_id[MAX_CON_ID]; +}; + +struct clk_lookup *clkdev_alloc(struct clk *clk, const char *con_id, + const char *dev_fmt, ...) +{ + struct clk_lookup_alloc *cla; + + cla = kzalloc(sizeof(*cla), GFP_KERNEL); + if (!cla) + return NULL; + + cla->cl.clk = clk; + if (con_id) { + strlcpy(cla->con_id, con_id, sizeof(cla->con_id)); + cla->cl.con_id = cla->con_id; + } + + if (dev_fmt) { + va_list ap; + + va_start(ap, dev_fmt); + vscnprintf(cla->dev_id, sizeof(cla->dev_id), dev_fmt, ap); + cla->cl.dev_id = cla->dev_id; + va_end(ap); + } + + return &cla->cl; +} +EXPORT_SYMBOL(clkdev_alloc); + +int clk_add_alias(const char *alias, const char *alias_dev_name, char *id, + struct device *dev) +{ + struct clk *r = clk_get(dev, id); + struct clk_lookup *l; + + if (IS_ERR(r)) + return PTR_ERR(r); + + l = clkdev_alloc(r, alias, alias_dev_name); + clk_put(r); + if (!l) + return -ENODEV; + clkdev_add(l); + return 0; +} +EXPORT_SYMBOL(clk_add_alias); + +/* + * clkdev_drop - remove a clock dynamically allocated + */ +void clkdev_drop(struct clk_lookup *cl) +{ + mutex_lock(&clocks_mutex); + list_del(&cl->node); + mutex_unlock(&clocks_mutex); + kfree(cl); +} +EXPORT_SYMBOL(clkdev_drop); diff --git a/arch/sh/kernel/cpu/clock.c b/arch/sh/kernel/cpu/clock.c index 83da5de..9ded1bc 100644 --- a/arch/sh/kernel/cpu/clock.c +++ b/arch/sh/kernel/cpu/clock.c @@ -10,10 +10,6 @@ * * Modified for omap shared clock framework by Tony Lindgren * - * With clkdev bits: - * - * Copyright (C) 2008 Russell King. - * * This file is subject to the terms and conditions of the GNU General Public * License. See the file "COPYING" in the main directory of this archive * for more details. @@ -30,6 +26,7 @@ #include #include #include +#include #include #include @@ -398,56 +395,6 @@ long clk_round_rate(struct clk *clk, unsigned long rate) EXPORT_SYMBOL_GPL(clk_round_rate); /* - * Find the correct struct clk for the device and connection ID. - * We do slightly fuzzy matching here: - * An entry with a NULL ID is assumed to be a wildcard. - * If an entry has a device ID, it must match - * If an entry has a connection ID, it must match - * Then we take the most specific entry - with the following - * order of precidence: dev+con > dev only > con only. - */ -static struct clk *clk_find(const char *dev_id, const char *con_id) -{ - struct clk_lookup *p; - struct clk *clk = NULL; - int match, best = 0; - - list_for_each_entry(p, &clock_list, node) { - match = 0; - if (p->dev_id) { - if (!dev_id || strcmp(p->dev_id, dev_id)) - continue; - match += 2; - } - if (p->con_id) { - if (!con_id || strcmp(p->con_id, con_id)) - continue; - match += 1; - } - if (match == 0) - continue; - - if (match > best) { - clk = p->clk; - best = match; - } - } - return clk; -} - -struct clk *clk_get_sys(const char *dev_id, const char *con_id) -{ - struct clk *clk; - - mutex_lock(&clock_list_sem); - clk = clk_find(dev_id, con_id); - mutex_unlock(&clock_list_sem); - - return clk ? clk : ERR_PTR(-ENOENT); -} -EXPORT_SYMBOL_GPL(clk_get_sys); - -/* * Returns a clock. Note that we first try to use device id on the bus * and clock name. If this fails, we try to use clock name only. */ -- cgit v0.10.2 From 2e733b3f84fa9c2ae60513c5f7b56d599ed2ae02 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 8 Mar 2010 21:46:37 +0900 Subject: sh: Support early clkdev allocations. early platform devices and the like may need to set up clock aliases, which require an allocation at a time well before the slab allocators are available. The clock framework comes up after bootmem, so using bootmem as a fallback should be sufficient. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/clkdev.c b/arch/sh/kernel/clkdev.c index 29cd802..defdd6e 100644 --- a/arch/sh/kernel/clkdev.c +++ b/arch/sh/kernel/clkdev.c @@ -20,6 +20,9 @@ #include #include #include +#include +#include +#include #include #include @@ -103,12 +106,16 @@ struct clk_lookup_alloc { char con_id[MAX_CON_ID]; }; -struct clk_lookup *clkdev_alloc(struct clk *clk, const char *con_id, - const char *dev_fmt, ...) +struct clk_lookup * __init_refok +clkdev_alloc(struct clk *clk, const char *con_id, const char *dev_fmt, ...) { struct clk_lookup_alloc *cla; - cla = kzalloc(sizeof(*cla), GFP_KERNEL); + if (!slab_is_available()) + cla = alloc_bootmem_low_pages(sizeof(*cla)); + else + cla = kzalloc(sizeof(*cla), GFP_KERNEL); + if (!cla) return NULL; -- cgit v0.10.2 From 881245dcff29df992d8431392a41fb81549129f9 Mon Sep 17 00:00:00 2001 From: William Cohen Date: Tue, 9 Mar 2010 09:26:04 +0100 Subject: Add DocBook documentation for the block tracepoints. This patch adds a simple description of the various block tracepoints available in the kernel. Signed-off-by: William Cohen Acked-by: Randy Dunlap Signed-off-by: Jens Axboe diff --git a/Documentation/DocBook/tracepoint.tmpl b/Documentation/DocBook/tracepoint.tmpl index 8bca1d5..e8473ea 100644 --- a/Documentation/DocBook/tracepoint.tmpl +++ b/Documentation/DocBook/tracepoint.tmpl @@ -16,6 +16,15 @@ + + William + Cohen + +
+ wcohen@redhat.com +
+ + @@ -91,4 +100,8 @@ !Iinclude/trace/events/signal.h + + Block IO +!Iinclude/trace/events/block.h + diff --git a/include/trace/events/block.h b/include/trace/events/block.h index 5fb7273..d870a91 100644 --- a/include/trace/events/block.h +++ b/include/trace/events/block.h @@ -40,6 +40,16 @@ DECLARE_EVENT_CLASS(block_rq_with_error, __entry->nr_sector, __entry->errors) ); +/** + * block_rq_abort - abort block operation request + * @q: queue containing the block operation request + * @rq: block IO operation request + * + * Called immediately after pending block IO operation request @rq in + * queue @q is aborted. The fields in the operation request @rq + * can be examined to determine which device and sectors the pending + * operation would access. + */ DEFINE_EVENT(block_rq_with_error, block_rq_abort, TP_PROTO(struct request_queue *q, struct request *rq), @@ -47,6 +57,15 @@ DEFINE_EVENT(block_rq_with_error, block_rq_abort, TP_ARGS(q, rq) ); +/** + * block_rq_requeue - place block IO request back on a queue + * @q: queue holding operation + * @rq: block IO operation request + * + * The block operation request @rq is being placed back into queue + * @q. For some reason the request was not completed and needs to be + * put back in the queue. + */ DEFINE_EVENT(block_rq_with_error, block_rq_requeue, TP_PROTO(struct request_queue *q, struct request *rq), @@ -54,6 +73,17 @@ DEFINE_EVENT(block_rq_with_error, block_rq_requeue, TP_ARGS(q, rq) ); +/** + * block_rq_complete - block IO operation completed by device driver + * @q: queue containing the block operation request + * @rq: block operations request + * + * The block_rq_complete tracepoint event indicates that some portion + * of operation request has been completed by the device driver. If + * the @rq->bio is %NULL, then there is absolutely no additional work to + * do for the request. If @rq->bio is non-NULL then there is + * additional work required to complete the request. + */ DEFINE_EVENT(block_rq_with_error, block_rq_complete, TP_PROTO(struct request_queue *q, struct request *rq), @@ -95,6 +125,16 @@ DECLARE_EVENT_CLASS(block_rq, __entry->nr_sector, __entry->comm) ); +/** + * block_rq_insert - insert block operation request into queue + * @q: target queue + * @rq: block IO operation request + * + * Called immediately before block operation request @rq is inserted + * into queue @q. The fields in the operation request @rq struct can + * be examined to determine which device and sectors the pending + * operation would access. + */ DEFINE_EVENT(block_rq, block_rq_insert, TP_PROTO(struct request_queue *q, struct request *rq), @@ -102,6 +142,14 @@ DEFINE_EVENT(block_rq, block_rq_insert, TP_ARGS(q, rq) ); +/** + * block_rq_issue - issue pending block IO request operation to device driver + * @q: queue holding operation + * @rq: block IO operation operation request + * + * Called when block operation request @rq from queue @q is sent to a + * device driver for processing. + */ DEFINE_EVENT(block_rq, block_rq_issue, TP_PROTO(struct request_queue *q, struct request *rq), @@ -109,6 +157,17 @@ DEFINE_EVENT(block_rq, block_rq_issue, TP_ARGS(q, rq) ); +/** + * block_bio_bounce - used bounce buffer when processing block operation + * @q: queue holding the block operation + * @bio: block operation + * + * A bounce buffer was used to handle the block operation @bio in @q. + * This occurs when hardware limitations prevent a direct transfer of + * data between the @bio data memory area and the IO device. Use of a + * bounce buffer requires extra copying of data and decreases + * performance. + */ TRACE_EVENT(block_bio_bounce, TP_PROTO(struct request_queue *q, struct bio *bio), @@ -138,6 +197,14 @@ TRACE_EVENT(block_bio_bounce, __entry->nr_sector, __entry->comm) ); +/** + * block_bio_complete - completed all work on the block operation + * @q: queue holding the block operation + * @bio: block operation completed + * + * This tracepoint indicates there is no further work to do on this + * block IO operation @bio. + */ TRACE_EVENT(block_bio_complete, TP_PROTO(struct request_queue *q, struct bio *bio), @@ -193,6 +260,14 @@ DECLARE_EVENT_CLASS(block_bio, __entry->nr_sector, __entry->comm) ); +/** + * block_bio_backmerge - merging block operation to the end of an existing operation + * @q: queue holding operation + * @bio: new block operation to merge + * + * Merging block request @bio to the end of an existing block request + * in queue @q. + */ DEFINE_EVENT(block_bio, block_bio_backmerge, TP_PROTO(struct request_queue *q, struct bio *bio), @@ -200,6 +275,14 @@ DEFINE_EVENT(block_bio, block_bio_backmerge, TP_ARGS(q, bio) ); +/** + * block_bio_frontmerge - merging block operation to the beginning of an existing operation + * @q: queue holding operation + * @bio: new block operation to merge + * + * Merging block IO operation @bio to the beginning of an existing block + * operation in queue @q. + */ DEFINE_EVENT(block_bio, block_bio_frontmerge, TP_PROTO(struct request_queue *q, struct bio *bio), @@ -207,6 +290,13 @@ DEFINE_EVENT(block_bio, block_bio_frontmerge, TP_ARGS(q, bio) ); +/** + * block_bio_queue - putting new block IO operation in queue + * @q: queue holding operation + * @bio: new block operation + * + * About to place the block IO operation @bio into queue @q. + */ DEFINE_EVENT(block_bio, block_bio_queue, TP_PROTO(struct request_queue *q, struct bio *bio), @@ -243,6 +333,15 @@ DECLARE_EVENT_CLASS(block_get_rq, __entry->nr_sector, __entry->comm) ); +/** + * block_getrq - get a free request entry in queue for block IO operations + * @q: queue for operations + * @bio: pending block IO operation + * @rw: low bit indicates a read (%0) or a write (%1) + * + * A request struct for queue @q has been allocated to handle the + * block IO operation @bio. + */ DEFINE_EVENT(block_get_rq, block_getrq, TP_PROTO(struct request_queue *q, struct bio *bio, int rw), @@ -250,6 +349,17 @@ DEFINE_EVENT(block_get_rq, block_getrq, TP_ARGS(q, bio, rw) ); +/** + * block_sleeprq - waiting to get a free request entry in queue for block IO operation + * @q: queue for operation + * @bio: pending block IO operation + * @rw: low bit indicates a read (%0) or a write (%1) + * + * In the case where a request struct cannot be provided for queue @q + * the process needs to wait for an request struct to become + * available. This tracepoint event is generated each time the + * process goes to sleep waiting for request struct become available. + */ DEFINE_EVENT(block_get_rq, block_sleeprq, TP_PROTO(struct request_queue *q, struct bio *bio, int rw), @@ -257,6 +367,14 @@ DEFINE_EVENT(block_get_rq, block_sleeprq, TP_ARGS(q, bio, rw) ); +/** + * block_plug - keep operations requests in request queue + * @q: request queue to plug + * + * Plug the request queue @q. Do not allow block operation requests + * to be sent to the device driver. Instead, accumulate requests in + * the queue to improve throughput performance of the block device. + */ TRACE_EVENT(block_plug, TP_PROTO(struct request_queue *q), @@ -293,6 +411,13 @@ DECLARE_EVENT_CLASS(block_unplug, TP_printk("[%s] %d", __entry->comm, __entry->nr_rq) ); +/** + * block_unplug_timer - timed release of operations requests in queue to device driver + * @q: request queue to unplug + * + * Unplug the request queue @q because a timer expired and allow block + * operation requests to be sent to the device driver. + */ DEFINE_EVENT(block_unplug, block_unplug_timer, TP_PROTO(struct request_queue *q), @@ -300,6 +425,13 @@ DEFINE_EVENT(block_unplug, block_unplug_timer, TP_ARGS(q) ); +/** + * block_unplug_io - release of operations requests in request queue + * @q: request queue to unplug + * + * Unplug request queue @q because device driver is scheduled to work + * on elements in the request queue. + */ DEFINE_EVENT(block_unplug, block_unplug_io, TP_PROTO(struct request_queue *q), @@ -307,6 +439,17 @@ DEFINE_EVENT(block_unplug, block_unplug_io, TP_ARGS(q) ); +/** + * block_split - split a single bio struct into two bio structs + * @q: queue containing the bio + * @bio: block operation being split + * @new_sector: The starting sector for the new bio + * + * The bio request @bio in request queue @q needs to be split into two + * bio requests. The newly created @bio request starts at + * @new_sector. This split may be required due to hardware limitation + * such as operation crossing device boundaries in a RAID system. + */ TRACE_EVENT(block_split, TP_PROTO(struct request_queue *q, struct bio *bio, @@ -337,6 +480,16 @@ TRACE_EVENT(block_split, __entry->comm) ); +/** + * block_remap - map request for a partition to the raw device + * @q: queue holding the operation + * @bio: revised operation + * @dev: device for the operation + * @from: original sector for the operation + * + * An operation for a partition on a block device has been mapped to the + * raw block device. + */ TRACE_EVENT(block_remap, TP_PROTO(struct request_queue *q, struct bio *bio, dev_t dev, @@ -370,6 +523,17 @@ TRACE_EVENT(block_remap, (unsigned long long)__entry->old_sector) ); +/** + * block_rq_remap - map request for a block operation request + * @q: queue holding the operation + * @rq: block IO operation request + * @dev: device for the operation + * @from: original sector for the operation + * + * The block operation request @rq in @q has been remapped. The block + * operation request @rq holds the current information and @from hold + * the original sector. + */ TRACE_EVENT(block_rq_remap, TP_PROTO(struct request_queue *q, struct request *rq, dev_t dev, -- cgit v0.10.2 From a636ee7fb35b731ba2b331f6294e809bb6be09c8 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Tue, 9 Mar 2010 06:57:53 +0000 Subject: driver core: Early dev_name() support. Presently early platform devices suffer from the fact they are unable to use dev_xxx() calls early on due to dev_name() and others being unavailable at the time ->probe() is called. This implements early init_name construction from the matched name/id pair following the semantics of the late device/driver match. As a result, matched IDs (inclusive of requested ones) are preserved when the handoff from the early platform code happens at kobject initialization time. Since we still require kmalloc slabs to be available at this point, using kstrdup() for establishing the init_name works fine. This subsequently needs to be tested from dev_name() prior to the init_name being cleared by the driver core. We don't kfree() since others will already have a handle on the string long before the kobject initialization takes place. This is also needed to permit drivers to use the clock framework early, without having to manually construct their own device IDs from the match id/name pair locally (needed by the early console and timer code on sh and arm). Signed-off-by: Paul Mundt Acked-by: Greg Kroah-Hartman diff --git a/drivers/base/platform.c b/drivers/base/platform.c index 1ba9d617..d2d4926 100644 --- a/drivers/base/platform.c +++ b/drivers/base/platform.c @@ -1239,6 +1239,25 @@ static int __init early_platform_driver_probe_id(char *class_str, } if (match) { + /* + * Set up a sensible init_name to enable + * dev_name() and others to be used before the + * rest of the driver core is initialized. + */ + if (!match->dev.init_name) { + char buf[32]; + + if (match->id != -1) + snprintf(buf, sizeof(buf), "%s.%d", + match->name, match->id); + else + snprintf(buf, sizeof(buf), "%s", + match->name); + + match->dev.init_name = kstrdup(buf, GFP_KERNEL); + if (!match->dev.init_name) + return -ENOMEM; + } if (epdrv->pdrv->probe(match)) pr_warning("%s: unable to probe %s early.\n", class_str, match->name); diff --git a/include/linux/device.h b/include/linux/device.h index 1821928..241b96b 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -451,6 +451,10 @@ struct device { static inline const char *dev_name(const struct device *dev) { + /* Use the init name until the kobject becomes available */ + if (dev->init_name) + return dev->init_name; + return kobject_name(&dev->kobj); } -- cgit v0.10.2 From 214a607a4f5bf5a14dab0304ba350cdaf3916795 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Wed, 10 Mar 2010 16:26:25 +0900 Subject: clocksource: Use dev_name() universally across the SH drivers. There is no need to copy in the name from the sh timer config now that dev_name() is available early. We prefer the dev_name() variant for consistent naming. Signed-off-by: Paul Mundt diff --git a/drivers/clocksource/sh_cmt.c b/drivers/clocksource/sh_cmt.c index 578595c..a1d3811 100644 --- a/drivers/clocksource/sh_cmt.c +++ b/drivers/clocksource/sh_cmt.c @@ -149,13 +149,12 @@ static void sh_cmt_start_stop_ch(struct sh_cmt_priv *p, int start) static int sh_cmt_enable(struct sh_cmt_priv *p, unsigned long *rate) { - struct sh_timer_config *cfg = p->pdev->dev.platform_data; int ret; /* enable clock */ ret = clk_enable(p->clk); if (ret) { - pr_err("sh_cmt: cannot enable clock \"%s\"\n", cfg->clk); + dev_err(&p->pdev->dev, "cannot enable clock\n"); return ret; } @@ -278,7 +277,7 @@ static void sh_cmt_clock_event_program_verify(struct sh_cmt_priv *p, delay = 1; if (!delay) - pr_warning("sh_cmt: too long delay\n"); + dev_warn(&p->pdev->dev, "too long delay\n"); } while (delay); } @@ -288,7 +287,7 @@ static void sh_cmt_set_next(struct sh_cmt_priv *p, unsigned long delta) unsigned long flags; if (delta > p->max_match_value) - pr_warning("sh_cmt: delta out of range\n"); + dev_warn(&p->pdev->dev, "delta out of range\n"); spin_lock_irqsave(&p->lock, flags); p->next_match_value = delta; @@ -450,7 +449,7 @@ static int sh_cmt_register_clocksource(struct sh_cmt_priv *p, cs->resume = sh_cmt_clocksource_resume; cs->mask = CLOCKSOURCE_MASK(sizeof(unsigned long) * 8); cs->flags = CLOCK_SOURCE_IS_CONTINUOUS; - pr_info("sh_cmt: %s used as clock source\n", cs->name); + dev_info(&p->pdev->dev, "used as clock source\n"); clocksource_register(cs); return 0; } @@ -496,13 +495,11 @@ static void sh_cmt_clock_event_mode(enum clock_event_mode mode, switch (mode) { case CLOCK_EVT_MODE_PERIODIC: - pr_info("sh_cmt: %s used for periodic clock events\n", - ced->name); + dev_info(&p->pdev->dev, "used for periodic clock events\n"); sh_cmt_clock_event_start(p, 1); break; case CLOCK_EVT_MODE_ONESHOT: - pr_info("sh_cmt: %s used for oneshot clock events\n", - ced->name); + dev_info(&p->pdev->dev, "used for oneshot clock events\n"); sh_cmt_clock_event_start(p, 0); break; case CLOCK_EVT_MODE_SHUTDOWN: @@ -543,7 +540,7 @@ static void sh_cmt_register_clockevent(struct sh_cmt_priv *p, ced->set_next_event = sh_cmt_clock_event_next; ced->set_mode = sh_cmt_clock_event_mode; - pr_info("sh_cmt: %s used for clock events\n", ced->name); + dev_info(&p->pdev->dev, "used for clock events\n"); clockevents_register_device(ced); } @@ -600,12 +597,12 @@ static int sh_cmt_setup(struct sh_cmt_priv *p, struct platform_device *pdev) /* map memory, let mapbase point to our channel */ p->mapbase = ioremap_nocache(res->start, resource_size(res)); if (p->mapbase == NULL) { - pr_err("sh_cmt: failed to remap I/O memory\n"); + dev_err(&p->pdev->dev, "failed to remap I/O memory\n"); goto err0; } /* request irq using setup_irq() (too early for request_irq()) */ - p->irqaction.name = cfg->name; + p->irqaction.name = dev_name(&p->pdev->dev); p->irqaction.handler = sh_cmt_interrupt; p->irqaction.dev_id = p; p->irqaction.flags = IRQF_DISABLED | IRQF_TIMER | IRQF_IRQPOLL; @@ -613,7 +610,7 @@ static int sh_cmt_setup(struct sh_cmt_priv *p, struct platform_device *pdev) /* get hold of clock */ p->clk = clk_get(&p->pdev->dev, cfg->clk); if (IS_ERR(p->clk)) { - pr_err("sh_cmt: cannot get clock \"%s\"\n", cfg->clk); + dev_err(&p->pdev->dev, "cannot get clock\n"); ret = PTR_ERR(p->clk); goto err1; } @@ -628,17 +625,17 @@ static int sh_cmt_setup(struct sh_cmt_priv *p, struct platform_device *pdev) p->clear_bits = ~0xc000; } - ret = sh_cmt_register(p, cfg->name, + ret = sh_cmt_register(p, (char *)dev_name(&p->pdev->dev), cfg->clockevent_rating, cfg->clocksource_rating); if (ret) { - pr_err("sh_cmt: registration failed\n"); + dev_err(&p->pdev->dev, "registration failed\n"); goto err1; } ret = setup_irq(irq, &p->irqaction); if (ret) { - pr_err("sh_cmt: failed to request irq %d\n", irq); + dev_err(&p->pdev->dev, "failed to request irq %d\n", irq); goto err1; } @@ -653,11 +650,10 @@ err0: static int __devinit sh_cmt_probe(struct platform_device *pdev) { struct sh_cmt_priv *p = platform_get_drvdata(pdev); - struct sh_timer_config *cfg = pdev->dev.platform_data; int ret; if (p) { - pr_info("sh_cmt: %s kept as earlytimer\n", cfg->name); + dev_info(&pdev->dev, "kept as earlytimer\n"); return 0; } diff --git a/drivers/clocksource/sh_mtu2.c b/drivers/clocksource/sh_mtu2.c index 4c8a759..8ce3555 100644 --- a/drivers/clocksource/sh_mtu2.c +++ b/drivers/clocksource/sh_mtu2.c @@ -118,13 +118,12 @@ static void sh_mtu2_start_stop_ch(struct sh_mtu2_priv *p, int start) static int sh_mtu2_enable(struct sh_mtu2_priv *p) { - struct sh_timer_config *cfg = p->pdev->dev.platform_data; int ret; /* enable clock */ ret = clk_enable(p->clk); if (ret) { - pr_err("sh_mtu2: cannot enable clock \"%s\"\n", cfg->clk); + dev_err(&p->pdev->dev, "cannot enable clock\n"); return ret; } @@ -193,8 +192,7 @@ static void sh_mtu2_clock_event_mode(enum clock_event_mode mode, switch (mode) { case CLOCK_EVT_MODE_PERIODIC: - pr_info("sh_mtu2: %s used for periodic clock events\n", - ced->name); + dev_info(&p->pdev->dev, "used for periodic clock events\n"); sh_mtu2_enable(p); break; case CLOCK_EVT_MODE_UNUSED: @@ -221,13 +219,13 @@ static void sh_mtu2_register_clockevent(struct sh_mtu2_priv *p, ced->cpumask = cpumask_of(0); ced->set_mode = sh_mtu2_clock_event_mode; - pr_info("sh_mtu2: %s used for clock events\n", ced->name); + dev_info(&p->pdev->dev, "used for clock events\n"); clockevents_register_device(ced); ret = setup_irq(p->irqaction.irq, &p->irqaction); if (ret) { - pr_err("sh_mtu2: failed to request irq %d\n", - p->irqaction.irq); + dev_err(&p->pdev->dev, "failed to request irq %d\n", + p->irqaction.irq); return; } } @@ -273,12 +271,12 @@ static int sh_mtu2_setup(struct sh_mtu2_priv *p, struct platform_device *pdev) /* map memory, let mapbase point to our channel */ p->mapbase = ioremap_nocache(res->start, resource_size(res)); if (p->mapbase == NULL) { - pr_err("sh_mtu2: failed to remap I/O memory\n"); + dev_err(&p->pdev->dev, "failed to remap I/O memory\n"); goto err0; } /* setup data for setup_irq() (too early for request_irq()) */ - p->irqaction.name = cfg->name; + p->irqaction.name = dev_name(&p->pdev->dev); p->irqaction.handler = sh_mtu2_interrupt; p->irqaction.dev_id = p; p->irqaction.irq = irq; @@ -287,12 +285,13 @@ static int sh_mtu2_setup(struct sh_mtu2_priv *p, struct platform_device *pdev) /* get hold of clock */ p->clk = clk_get(&p->pdev->dev, cfg->clk); if (IS_ERR(p->clk)) { - pr_err("sh_mtu2: cannot get clock \"%s\"\n", cfg->clk); + dev_err(&p->pdev->dev, "cannot get clock\n"); ret = PTR_ERR(p->clk); goto err1; } - return sh_mtu2_register(p, cfg->name, cfg->clockevent_rating); + return sh_mtu2_register(p, (char *)dev_name(&p->pdev->dev), + cfg->clockevent_rating); err1: iounmap(p->mapbase); err0: @@ -302,11 +301,10 @@ static int sh_mtu2_setup(struct sh_mtu2_priv *p, struct platform_device *pdev) static int __devinit sh_mtu2_probe(struct platform_device *pdev) { struct sh_mtu2_priv *p = platform_get_drvdata(pdev); - struct sh_timer_config *cfg = pdev->dev.platform_data; int ret; if (p) { - pr_info("sh_mtu2: %s kept as earlytimer\n", cfg->name); + dev_info(&pdev->dev, "kept as earlytimer\n"); return 0; } diff --git a/drivers/clocksource/sh_tmu.c b/drivers/clocksource/sh_tmu.c index 961f5b5..3423945 100644 --- a/drivers/clocksource/sh_tmu.c +++ b/drivers/clocksource/sh_tmu.c @@ -106,13 +106,12 @@ static void sh_tmu_start_stop_ch(struct sh_tmu_priv *p, int start) static int sh_tmu_enable(struct sh_tmu_priv *p) { - struct sh_timer_config *cfg = p->pdev->dev.platform_data; int ret; /* enable clock */ ret = clk_enable(p->clk); if (ret) { - pr_err("sh_tmu: cannot enable clock \"%s\"\n", cfg->clk); + dev_err(&p->pdev->dev, "cannot enable clock\n"); return ret; } @@ -228,7 +227,7 @@ static int sh_tmu_register_clocksource(struct sh_tmu_priv *p, cs->disable = sh_tmu_clocksource_disable; cs->mask = CLOCKSOURCE_MASK(32); cs->flags = CLOCK_SOURCE_IS_CONTINUOUS; - pr_info("sh_tmu: %s used as clock source\n", cs->name); + dev_info(&p->pdev->dev, "used as clock source\n"); clocksource_register(cs); return 0; } @@ -276,13 +275,11 @@ static void sh_tmu_clock_event_mode(enum clock_event_mode mode, switch (mode) { case CLOCK_EVT_MODE_PERIODIC: - pr_info("sh_tmu: %s used for periodic clock events\n", - ced->name); + dev_info(&p->pdev->dev, "used for periodic clock events\n"); sh_tmu_clock_event_start(p, 1); break; case CLOCK_EVT_MODE_ONESHOT: - pr_info("sh_tmu: %s used for oneshot clock events\n", - ced->name); + dev_info(&p->pdev->dev, "used for oneshot clock events\n"); sh_tmu_clock_event_start(p, 0); break; case CLOCK_EVT_MODE_UNUSED: @@ -323,13 +320,13 @@ static void sh_tmu_register_clockevent(struct sh_tmu_priv *p, ced->set_next_event = sh_tmu_clock_event_next; ced->set_mode = sh_tmu_clock_event_mode; - pr_info("sh_tmu: %s used for clock events\n", ced->name); + dev_info(&p->pdev->dev, "used for clock events\n"); clockevents_register_device(ced); ret = setup_irq(p->irqaction.irq, &p->irqaction); if (ret) { - pr_err("sh_tmu: failed to request irq %d\n", - p->irqaction.irq); + dev_err(&p->pdev->dev, "failed to request irq %d\n", + p->irqaction.irq); return; } } @@ -378,12 +375,12 @@ static int sh_tmu_setup(struct sh_tmu_priv *p, struct platform_device *pdev) /* map memory, let mapbase point to our channel */ p->mapbase = ioremap_nocache(res->start, resource_size(res)); if (p->mapbase == NULL) { - pr_err("sh_tmu: failed to remap I/O memory\n"); + dev_err(&p->pdev->dev, "failed to remap I/O memory\n"); goto err0; } /* setup data for setup_irq() (too early for request_irq()) */ - p->irqaction.name = cfg->name; + p->irqaction.name = dev_name(&p->pdev->dev); p->irqaction.handler = sh_tmu_interrupt; p->irqaction.dev_id = p; p->irqaction.irq = irq; @@ -392,12 +389,12 @@ static int sh_tmu_setup(struct sh_tmu_priv *p, struct platform_device *pdev) /* get hold of clock */ p->clk = clk_get(&p->pdev->dev, cfg->clk); if (IS_ERR(p->clk)) { - pr_err("sh_tmu: cannot get clock \"%s\"\n", cfg->clk); + dev_err(&p->pdev->dev, "cannot get clock\n"); ret = PTR_ERR(p->clk); goto err1; } - return sh_tmu_register(p, cfg->name, + return sh_tmu_register(p, (char *)dev_name(&p->pdev->dev), cfg->clockevent_rating, cfg->clocksource_rating); err1: @@ -409,11 +406,10 @@ static int sh_tmu_setup(struct sh_tmu_priv *p, struct platform_device *pdev) static int __devinit sh_tmu_probe(struct platform_device *pdev) { struct sh_tmu_priv *p = platform_get_drvdata(pdev); - struct sh_timer_config *cfg = pdev->dev.platform_data; int ret; if (p) { - pr_info("sh_tmu: %s kept as earlytimer\n", cfg->name); + dev_info(&pdev->dev, "kept as earlytimer\n"); return 0; } -- cgit v0.10.2 From ec0ffe2ee0e0fb9da4409d86bfd72636450f32df Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Wed, 10 Mar 2010 16:28:16 +0900 Subject: sh: Kill off all timer name clobbering. Now that dev_name() can be used early, we no longer require a static string. Kill off all of the superfluous timer names. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/sh2/setup-sh7619.c b/arch/sh/kernel/cpu/sh2/setup-sh7619.c index 114c7ce..9b96457 100644 --- a/arch/sh/kernel/cpu/sh2/setup-sh7619.c +++ b/arch/sh/kernel/cpu/sh2/setup-sh7619.c @@ -128,7 +128,6 @@ static struct platform_device eth_device = { }; static struct sh_timer_config cmt0_platform_data = { - .name = "CMT0", .channel_offset = 0x02, .timer_bit = 0, .clk = "peripheral_clk", @@ -138,7 +137,6 @@ static struct sh_timer_config cmt0_platform_data = { static struct resource cmt0_resources[] = { [0] = { - .name = "CMT0", .start = 0xf84a0072, .end = 0xf84a0077, .flags = IORESOURCE_MEM, @@ -160,7 +158,6 @@ static struct platform_device cmt0_device = { }; static struct sh_timer_config cmt1_platform_data = { - .name = "CMT1", .channel_offset = 0x08, .timer_bit = 1, .clk = "peripheral_clk", @@ -170,7 +167,6 @@ static struct sh_timer_config cmt1_platform_data = { static struct resource cmt1_resources[] = { [0] = { - .name = "CMT1", .start = 0xf84a0078, .end = 0xf84a007d, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh2a/setup-mxg.c b/arch/sh/kernel/cpu/sh2a/setup-mxg.c index 8f669dc..5301450 100644 --- a/arch/sh/kernel/cpu/sh2a/setup-mxg.c +++ b/arch/sh/kernel/cpu/sh2a/setup-mxg.c @@ -115,7 +115,6 @@ static DECLARE_INTC_DESC(intc_desc, "mxg", vectors, groups, mask_registers, prio_registers, NULL); static struct sh_timer_config mtu2_0_platform_data = { - .name = "MTU2_0", .channel_offset = -0x80, .timer_bit = 0, .clk = "peripheral_clk", @@ -124,7 +123,6 @@ static struct sh_timer_config mtu2_0_platform_data = { static struct resource mtu2_0_resources[] = { [0] = { - .name = "MTU2_0", .start = 0xff801300, .end = 0xff801326, .flags = IORESOURCE_MEM, @@ -146,7 +144,6 @@ static struct platform_device mtu2_0_device = { }; static struct sh_timer_config mtu2_1_platform_data = { - .name = "MTU2_1", .channel_offset = -0x100, .timer_bit = 1, .clk = "peripheral_clk", @@ -155,7 +152,6 @@ static struct sh_timer_config mtu2_1_platform_data = { static struct resource mtu2_1_resources[] = { [0] = { - .name = "MTU2_1", .start = 0xff801380, .end = 0xff801390, .flags = IORESOURCE_MEM, @@ -177,7 +173,6 @@ static struct platform_device mtu2_1_device = { }; static struct sh_timer_config mtu2_2_platform_data = { - .name = "MTU2_2", .channel_offset = 0x80, .timer_bit = 2, .clk = "peripheral_clk", @@ -186,7 +181,6 @@ static struct sh_timer_config mtu2_2_platform_data = { static struct resource mtu2_2_resources[] = { [0] = { - .name = "MTU2_2", .start = 0xff801000, .end = 0xff80100a, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh2a/setup-sh7201.c b/arch/sh/kernel/cpu/sh2a/setup-sh7201.c index 4ccfeb5..5b07ab5 100644 --- a/arch/sh/kernel/cpu/sh2a/setup-sh7201.c +++ b/arch/sh/kernel/cpu/sh2a/setup-sh7201.c @@ -318,7 +318,6 @@ static struct platform_device rtc_device = { }; static struct sh_timer_config mtu2_0_platform_data = { - .name = "MTU2_0", .channel_offset = -0x80, .timer_bit = 0, .clk = "peripheral_clk", @@ -327,7 +326,6 @@ static struct sh_timer_config mtu2_0_platform_data = { static struct resource mtu2_0_resources[] = { [0] = { - .name = "MTU2_0", .start = 0xfffe4300, .end = 0xfffe4326, .flags = IORESOURCE_MEM, @@ -349,7 +347,6 @@ static struct platform_device mtu2_0_device = { }; static struct sh_timer_config mtu2_1_platform_data = { - .name = "MTU2_1", .channel_offset = -0x100, .timer_bit = 1, .clk = "peripheral_clk", @@ -358,7 +355,6 @@ static struct sh_timer_config mtu2_1_platform_data = { static struct resource mtu2_1_resources[] = { [0] = { - .name = "MTU2_1", .start = 0xfffe4380, .end = 0xfffe4390, .flags = IORESOURCE_MEM, @@ -380,7 +376,6 @@ static struct platform_device mtu2_1_device = { }; static struct sh_timer_config mtu2_2_platform_data = { - .name = "MTU2_2", .channel_offset = 0x80, .timer_bit = 2, .clk = "peripheral_clk", @@ -389,7 +384,6 @@ static struct sh_timer_config mtu2_2_platform_data = { static struct resource mtu2_2_resources[] = { [0] = { - .name = "MTU2_2", .start = 0xfffe4000, .end = 0xfffe400a, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh2a/setup-sh7203.c b/arch/sh/kernel/cpu/sh2a/setup-sh7203.c index 3136966..433ac53 100644 --- a/arch/sh/kernel/cpu/sh2a/setup-sh7203.c +++ b/arch/sh/kernel/cpu/sh2a/setup-sh7203.c @@ -234,7 +234,6 @@ static struct platform_device scif3_device = { }; static struct sh_timer_config cmt0_platform_data = { - .name = "CMT0", .channel_offset = 0x02, .timer_bit = 0, .clk = "peripheral_clk", @@ -244,7 +243,6 @@ static struct sh_timer_config cmt0_platform_data = { static struct resource cmt0_resources[] = { [0] = { - .name = "CMT0", .start = 0xfffec002, .end = 0xfffec007, .flags = IORESOURCE_MEM, @@ -266,7 +264,6 @@ static struct platform_device cmt0_device = { }; static struct sh_timer_config cmt1_platform_data = { - .name = "CMT1", .channel_offset = 0x08, .timer_bit = 1, .clk = "peripheral_clk", @@ -276,7 +273,6 @@ static struct sh_timer_config cmt1_platform_data = { static struct resource cmt1_resources[] = { [0] = { - .name = "CMT1", .start = 0xfffec008, .end = 0xfffec00d, .flags = IORESOURCE_MEM, @@ -298,7 +294,6 @@ static struct platform_device cmt1_device = { }; static struct sh_timer_config mtu2_0_platform_data = { - .name = "MTU2_0", .channel_offset = -0x80, .timer_bit = 0, .clk = "peripheral_clk", @@ -307,7 +302,6 @@ static struct sh_timer_config mtu2_0_platform_data = { static struct resource mtu2_0_resources[] = { [0] = { - .name = "MTU2_0", .start = 0xfffe4300, .end = 0xfffe4326, .flags = IORESOURCE_MEM, @@ -329,7 +323,6 @@ static struct platform_device mtu2_0_device = { }; static struct sh_timer_config mtu2_1_platform_data = { - .name = "MTU2_1", .channel_offset = -0x100, .timer_bit = 1, .clk = "peripheral_clk", @@ -338,7 +331,6 @@ static struct sh_timer_config mtu2_1_platform_data = { static struct resource mtu2_1_resources[] = { [0] = { - .name = "MTU2_1", .start = 0xfffe4380, .end = 0xfffe4390, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh2a/setup-sh7206.c b/arch/sh/kernel/cpu/sh2a/setup-sh7206.c index 0648735..8caecb4 100644 --- a/arch/sh/kernel/cpu/sh2a/setup-sh7206.c +++ b/arch/sh/kernel/cpu/sh2a/setup-sh7206.c @@ -194,7 +194,6 @@ static struct platform_device scif3_device = { }; static struct sh_timer_config cmt0_platform_data = { - .name = "CMT0", .channel_offset = 0x02, .timer_bit = 0, .clk = "peripheral_clk", @@ -204,7 +203,6 @@ static struct sh_timer_config cmt0_platform_data = { static struct resource cmt0_resources[] = { [0] = { - .name = "CMT0", .start = 0xfffec002, .end = 0xfffec007, .flags = IORESOURCE_MEM, @@ -226,7 +224,6 @@ static struct platform_device cmt0_device = { }; static struct sh_timer_config cmt1_platform_data = { - .name = "CMT1", .channel_offset = 0x08, .timer_bit = 1, .clk = "peripheral_clk", @@ -236,7 +233,6 @@ static struct sh_timer_config cmt1_platform_data = { static struct resource cmt1_resources[] = { [0] = { - .name = "CMT1", .start = 0xfffec008, .end = 0xfffec00d, .flags = IORESOURCE_MEM, @@ -258,7 +254,6 @@ static struct platform_device cmt1_device = { }; static struct sh_timer_config mtu2_0_platform_data = { - .name = "MTU2_0", .channel_offset = -0x80, .timer_bit = 0, .clk = "peripheral_clk", @@ -267,7 +262,6 @@ static struct sh_timer_config mtu2_0_platform_data = { static struct resource mtu2_0_resources[] = { [0] = { - .name = "MTU2_0", .start = 0xfffe4300, .end = 0xfffe4326, .flags = IORESOURCE_MEM, @@ -289,7 +283,6 @@ static struct platform_device mtu2_0_device = { }; static struct sh_timer_config mtu2_1_platform_data = { - .name = "MTU2_1", .channel_offset = -0x100, .timer_bit = 1, .clk = "peripheral_clk", @@ -298,7 +291,6 @@ static struct sh_timer_config mtu2_1_platform_data = { static struct resource mtu2_1_resources[] = { [0] = { - .name = "MTU2_1", .start = 0xfffe4380, .end = 0xfffe4390, .flags = IORESOURCE_MEM, @@ -320,7 +312,6 @@ static struct platform_device mtu2_1_device = { }; static struct sh_timer_config mtu2_2_platform_data = { - .name = "MTU2_2", .channel_offset = 0x80, .timer_bit = 2, .clk = "peripheral_clk", @@ -329,7 +320,6 @@ static struct sh_timer_config mtu2_2_platform_data = { static struct resource mtu2_2_resources[] = { [0] = { - .name = "MTU2_2", .start = 0xfffe4000, .end = 0xfffe400a, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh3/setup-sh7705.c b/arch/sh/kernel/cpu/sh3/setup-sh7705.c index 7b892d6..5e45fb7 100644 --- a/arch/sh/kernel/cpu/sh3/setup-sh7705.c +++ b/arch/sh/kernel/cpu/sh3/setup-sh7705.c @@ -124,7 +124,6 @@ static struct platform_device rtc_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x02, .timer_bit = 0, .clk = "peripheral_clk", @@ -133,7 +132,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xfffffe94, .end = 0xfffffe9f, .flags = IORESOURCE_MEM, @@ -155,7 +153,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0xe, .timer_bit = 1, .clk = "peripheral_clk", @@ -164,7 +161,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xfffffea0, .end = 0xfffffeab, .flags = IORESOURCE_MEM, @@ -186,7 +182,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1a, .timer_bit = 2, .clk = "peripheral_clk", @@ -194,7 +189,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xfffffeac, .end = 0xfffffebb, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh3/setup-sh770x.c b/arch/sh/kernel/cpu/sh3/setup-sh770x.c index bc0c4f6..9adb853 100644 --- a/arch/sh/kernel/cpu/sh3/setup-sh770x.c +++ b/arch/sh/kernel/cpu/sh3/setup-sh770x.c @@ -157,7 +157,6 @@ static struct platform_device scif2_device = { #endif static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x02, .timer_bit = 0, .clk = "peripheral_clk", @@ -166,7 +165,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xfffffe94, .end = 0xfffffe9f, .flags = IORESOURCE_MEM, @@ -188,7 +186,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0xe, .timer_bit = 1, .clk = "peripheral_clk", @@ -197,7 +194,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xfffffea0, .end = 0xfffffeab, .flags = IORESOURCE_MEM, @@ -219,7 +215,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1a, .timer_bit = 2, .clk = "peripheral_clk", @@ -227,7 +222,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xfffffeac, .end = 0xfffffebb, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh3/setup-sh7710.c b/arch/sh/kernel/cpu/sh3/setup-sh7710.c index 0845a3a..2e581e3 100644 --- a/arch/sh/kernel/cpu/sh3/setup-sh7710.c +++ b/arch/sh/kernel/cpu/sh3/setup-sh7710.c @@ -127,7 +127,6 @@ static struct platform_device scif1_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x02, .timer_bit = 0, .clk = "peripheral_clk", @@ -136,7 +135,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xa412fe94, .end = 0xa412fe9f, .flags = IORESOURCE_MEM, @@ -158,7 +156,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0xe, .timer_bit = 1, .clk = "peripheral_clk", @@ -167,7 +164,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xa412fea0, .end = 0xa412feab, .flags = IORESOURCE_MEM, @@ -189,7 +185,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1a, .timer_bit = 2, .clk = "peripheral_clk", @@ -197,7 +192,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xa412feac, .end = 0xa412feb5, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh3/setup-sh7720.c b/arch/sh/kernel/cpu/sh3/setup-sh7720.c index a718a62..9a006e1 100644 --- a/arch/sh/kernel/cpu/sh3/setup-sh7720.c +++ b/arch/sh/kernel/cpu/sh3/setup-sh7720.c @@ -130,7 +130,6 @@ static struct platform_device usbf_device = { }; static struct sh_timer_config cmt0_platform_data = { - .name = "CMT0", .channel_offset = 0x10, .timer_bit = 0, .clk = "peripheral_clk", @@ -140,7 +139,6 @@ static struct sh_timer_config cmt0_platform_data = { static struct resource cmt0_resources[] = { [0] = { - .name = "CMT0", .start = 0x044a0010, .end = 0x044a001b, .flags = IORESOURCE_MEM, @@ -162,7 +160,6 @@ static struct platform_device cmt0_device = { }; static struct sh_timer_config cmt1_platform_data = { - .name = "CMT1", .channel_offset = 0x20, .timer_bit = 1, .clk = "peripheral_clk", @@ -170,7 +167,6 @@ static struct sh_timer_config cmt1_platform_data = { static struct resource cmt1_resources[] = { [0] = { - .name = "CMT1", .start = 0x044a0020, .end = 0x044a002b, .flags = IORESOURCE_MEM, @@ -192,7 +188,6 @@ static struct platform_device cmt1_device = { }; static struct sh_timer_config cmt2_platform_data = { - .name = "CMT2", .channel_offset = 0x30, .timer_bit = 2, .clk = "peripheral_clk", @@ -200,7 +195,6 @@ static struct sh_timer_config cmt2_platform_data = { static struct resource cmt2_resources[] = { [0] = { - .name = "CMT2", .start = 0x044a0030, .end = 0x044a003b, .flags = IORESOURCE_MEM, @@ -222,7 +216,6 @@ static struct platform_device cmt2_device = { }; static struct sh_timer_config cmt3_platform_data = { - .name = "CMT3", .channel_offset = 0x40, .timer_bit = 3, .clk = "peripheral_clk", @@ -230,7 +223,6 @@ static struct sh_timer_config cmt3_platform_data = { static struct resource cmt3_resources[] = { [0] = { - .name = "CMT3", .start = 0x044a0040, .end = 0x044a004b, .flags = IORESOURCE_MEM, @@ -252,7 +244,6 @@ static struct platform_device cmt3_device = { }; static struct sh_timer_config cmt4_platform_data = { - .name = "CMT4", .channel_offset = 0x50, .timer_bit = 4, .clk = "peripheral_clk", @@ -260,7 +251,6 @@ static struct sh_timer_config cmt4_platform_data = { static struct resource cmt4_resources[] = { [0] = { - .name = "CMT4", .start = 0x044a0050, .end = 0x044a005b, .flags = IORESOURCE_MEM, @@ -282,7 +272,6 @@ static struct platform_device cmt4_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x02, .timer_bit = 0, .clk = "peripheral_clk", @@ -291,7 +280,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xa412fe94, .end = 0xa412fe9f, .flags = IORESOURCE_MEM, @@ -313,7 +301,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0xe, .timer_bit = 1, .clk = "peripheral_clk", @@ -322,7 +309,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xa412fea0, .end = 0xa412feab, .flags = IORESOURCE_MEM, @@ -344,7 +330,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1a, .timer_bit = 2, .clk = "peripheral_clk", @@ -352,7 +337,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xa412feac, .end = 0xa412feb5, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4/setup-sh4-202.c b/arch/sh/kernel/cpu/sh4/setup-sh4-202.c index b9b7e10..223d066 100644 --- a/arch/sh/kernel/cpu/sh4/setup-sh4-202.c +++ b/arch/sh/kernel/cpu/sh4/setup-sh4-202.c @@ -31,7 +31,6 @@ static struct platform_device scif0_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -40,7 +39,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -62,7 +60,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -71,7 +68,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -93,7 +89,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -101,7 +96,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002f, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4/setup-sh7750.c b/arch/sh/kernel/cpu/sh4/setup-sh7750.c index ffd79e5..2f337b1 100644 --- a/arch/sh/kernel/cpu/sh4/setup-sh7750.c +++ b/arch/sh/kernel/cpu/sh4/setup-sh7750.c @@ -66,7 +66,6 @@ static struct platform_device scif1_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -75,7 +74,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -97,7 +95,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -106,7 +103,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -128,7 +124,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -136,7 +131,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002f, .flags = IORESOURCE_MEM, @@ -163,7 +157,6 @@ static struct platform_device tmu2_device = { defined(CONFIG_CPU_SUBTYPE_SH7751R) static struct sh_timer_config tmu3_platform_data = { - .name = "TMU3", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -171,7 +164,6 @@ static struct sh_timer_config tmu3_platform_data = { static struct resource tmu3_resources[] = { [0] = { - .name = "TMU3", .start = 0xfe100008, .end = 0xfe100013, .flags = IORESOURCE_MEM, @@ -193,7 +185,6 @@ static struct platform_device tmu3_device = { }; static struct sh_timer_config tmu4_platform_data = { - .name = "TMU4", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -201,7 +192,6 @@ static struct sh_timer_config tmu4_platform_data = { static struct resource tmu4_resources[] = { [0] = { - .name = "TMU4", .start = 0xfe100014, .end = 0xfe10001f, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4/setup-sh7760.c b/arch/sh/kernel/cpu/sh4/setup-sh7760.c index a16eb36..fa43da0 100644 --- a/arch/sh/kernel/cpu/sh4/setup-sh7760.c +++ b/arch/sh/kernel/cpu/sh4/setup-sh7760.c @@ -187,7 +187,6 @@ static struct platform_device scif3_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -196,7 +195,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -218,7 +216,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -227,7 +224,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -249,7 +245,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -257,7 +252,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002f, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7343.c b/arch/sh/kernel/cpu/sh4a/setup-sh7343.c index 45eb1bf..fdbd7d0 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7343.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7343.c @@ -207,7 +207,6 @@ static struct platform_device jpu_device = { }; static struct sh_timer_config cmt_platform_data = { - .name = "CMT", .channel_offset = 0x60, .timer_bit = 5, .clk = "cmt0", @@ -217,7 +216,6 @@ static struct sh_timer_config cmt_platform_data = { static struct resource cmt_resources[] = { [0] = { - .name = "CMT", .start = 0x044a0060, .end = 0x044a006b, .flags = IORESOURCE_MEM, @@ -239,7 +237,6 @@ static struct platform_device cmt_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "tmu0", @@ -248,7 +245,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -270,7 +266,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "tmu0", @@ -279,7 +274,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -301,7 +295,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "tmu0", @@ -309,7 +302,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002b, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7366.c b/arch/sh/kernel/cpu/sh4a/setup-sh7366.c index c494c19..d90fe0a 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7366.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7366.c @@ -169,7 +169,6 @@ static struct platform_device veu1_device = { }; static struct sh_timer_config cmt_platform_data = { - .name = "CMT", .channel_offset = 0x60, .timer_bit = 5, .clk = "cmt0", @@ -179,7 +178,6 @@ static struct sh_timer_config cmt_platform_data = { static struct resource cmt_resources[] = { [0] = { - .name = "CMT", .start = 0x044a0060, .end = 0x044a006b, .flags = IORESOURCE_MEM, @@ -201,7 +199,6 @@ static struct platform_device cmt_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "tmu0", @@ -210,7 +207,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -232,7 +228,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "tmu0", @@ -241,7 +236,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -263,7 +257,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "tmu0", @@ -271,7 +264,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002b, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7722.c b/arch/sh/kernel/cpu/sh4a/setup-sh7722.c index fd7e363..83ae9ad 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7722.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7722.c @@ -401,7 +401,6 @@ static struct platform_device jpu_device = { }; static struct sh_timer_config cmt_platform_data = { - .name = "CMT", .channel_offset = 0x60, .timer_bit = 5, .clk = "cmt0", @@ -411,7 +410,6 @@ static struct sh_timer_config cmt_platform_data = { static struct resource cmt_resources[] = { [0] = { - .name = "CMT", .start = 0x044a0060, .end = 0x044a006b, .flags = IORESOURCE_MEM, @@ -436,7 +434,6 @@ static struct platform_device cmt_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "tmu0", @@ -445,7 +442,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -470,7 +466,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "tmu0", @@ -479,7 +474,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -504,7 +498,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "tmu0", @@ -512,7 +505,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002b, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7723.c b/arch/sh/kernel/cpu/sh4a/setup-sh7723.c index 85c61f6..68446b3f 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7723.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7723.c @@ -211,7 +211,6 @@ static struct platform_device veu1_device = { }; static struct sh_timer_config cmt_platform_data = { - .name = "CMT", .channel_offset = 0x60, .timer_bit = 5, .clk = "cmt0", @@ -221,7 +220,6 @@ static struct sh_timer_config cmt_platform_data = { static struct resource cmt_resources[] = { [0] = { - .name = "CMT", .start = 0x044a0060, .end = 0x044a006b, .flags = IORESOURCE_MEM, @@ -246,7 +244,6 @@ static struct platform_device cmt_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "tmu0", @@ -255,7 +252,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -280,7 +276,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "tmu0", @@ -289,7 +284,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -314,7 +308,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "tmu0", @@ -322,7 +315,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002b, .flags = IORESOURCE_MEM, @@ -347,7 +339,6 @@ static struct platform_device tmu2_device = { }; static struct sh_timer_config tmu3_platform_data = { - .name = "TMU3", .channel_offset = 0x04, .timer_bit = 0, .clk = "tmu1", @@ -355,7 +346,6 @@ static struct sh_timer_config tmu3_platform_data = { static struct resource tmu3_resources[] = { [0] = { - .name = "TMU3", .start = 0xffd90008, .end = 0xffd90013, .flags = IORESOURCE_MEM, @@ -380,7 +370,6 @@ static struct platform_device tmu3_device = { }; static struct sh_timer_config tmu4_platform_data = { - .name = "TMU4", .channel_offset = 0x10, .timer_bit = 1, .clk = "tmu1", @@ -388,7 +377,6 @@ static struct sh_timer_config tmu4_platform_data = { static struct resource tmu4_resources[] = { [0] = { - .name = "TMU4", .start = 0xffd90014, .end = 0xffd9001f, .flags = IORESOURCE_MEM, @@ -413,7 +401,6 @@ static struct platform_device tmu4_device = { }; static struct sh_timer_config tmu5_platform_data = { - .name = "TMU5", .channel_offset = 0x1c, .timer_bit = 2, .clk = "tmu1", @@ -421,7 +408,6 @@ static struct sh_timer_config tmu5_platform_data = { static struct resource tmu5_resources[] = { [0] = { - .name = "TMU5", .start = 0xffd90020, .end = 0xffd9002b, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7724.c b/arch/sh/kernel/cpu/sh4a/setup-sh7724.c index e7fa2a9..abc98b0 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7724.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7724.c @@ -485,7 +485,6 @@ static struct platform_device veu1_device = { }; static struct sh_timer_config cmt_platform_data = { - .name = "CMT", .channel_offset = 0x60, .timer_bit = 5, .clk = "cmt0", @@ -495,7 +494,6 @@ static struct sh_timer_config cmt_platform_data = { static struct resource cmt_resources[] = { [0] = { - .name = "CMT", .start = 0x044a0060, .end = 0x044a006b, .flags = IORESOURCE_MEM, @@ -520,7 +518,6 @@ static struct platform_device cmt_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "tmu0", @@ -529,7 +526,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -554,7 +550,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "tmu0", @@ -563,7 +558,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -588,7 +582,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "tmu0", @@ -596,7 +589,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002b, .flags = IORESOURCE_MEM, @@ -622,7 +614,6 @@ static struct platform_device tmu2_device = { static struct sh_timer_config tmu3_platform_data = { - .name = "TMU3", .channel_offset = 0x04, .timer_bit = 0, .clk = "tmu1", @@ -630,7 +621,6 @@ static struct sh_timer_config tmu3_platform_data = { static struct resource tmu3_resources[] = { [0] = { - .name = "TMU3", .start = 0xffd90008, .end = 0xffd90013, .flags = IORESOURCE_MEM, @@ -655,7 +645,6 @@ static struct platform_device tmu3_device = { }; static struct sh_timer_config tmu4_platform_data = { - .name = "TMU4", .channel_offset = 0x10, .timer_bit = 1, .clk = "tmu1", @@ -663,7 +652,6 @@ static struct sh_timer_config tmu4_platform_data = { static struct resource tmu4_resources[] = { [0] = { - .name = "TMU4", .start = 0xffd90014, .end = 0xffd9001f, .flags = IORESOURCE_MEM, @@ -688,7 +676,6 @@ static struct platform_device tmu4_device = { }; static struct sh_timer_config tmu5_platform_data = { - .name = "TMU5", .channel_offset = 0x1c, .timer_bit = 2, .clk = "tmu1", @@ -696,7 +683,6 @@ static struct sh_timer_config tmu5_platform_data = { static struct resource tmu5_resources[] = { [0] = { - .name = "TMU5", .start = 0xffd90020, .end = 0xffd9002b, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7757.c b/arch/sh/kernel/cpu/sh4a/setup-sh7757.c index e75edf5..25eddf2 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7757.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7757.c @@ -63,7 +63,6 @@ static struct platform_device scif4_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -72,7 +71,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xfe430008, .end = 0xfe430013, .flags = IORESOURCE_MEM, @@ -94,7 +92,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -103,7 +100,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xfe430014, .end = 0xfe43001f, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7763.c b/arch/sh/kernel/cpu/sh4a/setup-sh7763.c index 7f6b0a5..a91263e 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7763.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7763.c @@ -131,7 +131,6 @@ static struct platform_device usbf_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -140,7 +139,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -162,7 +160,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -171,7 +168,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -193,7 +189,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -201,7 +196,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002f, .flags = IORESOURCE_MEM, @@ -223,7 +217,6 @@ static struct platform_device tmu2_device = { }; static struct sh_timer_config tmu3_platform_data = { - .name = "TMU3", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -231,7 +224,6 @@ static struct sh_timer_config tmu3_platform_data = { static struct resource tmu3_resources[] = { [0] = { - .name = "TMU3", .start = 0xffd88008, .end = 0xffd88013, .flags = IORESOURCE_MEM, @@ -253,7 +245,6 @@ static struct platform_device tmu3_device = { }; static struct sh_timer_config tmu4_platform_data = { - .name = "TMU4", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -261,7 +252,6 @@ static struct sh_timer_config tmu4_platform_data = { static struct resource tmu4_resources[] = { [0] = { - .name = "TMU4", .start = 0xffd88014, .end = 0xffd8801f, .flags = IORESOURCE_MEM, @@ -283,7 +273,6 @@ static struct platform_device tmu4_device = { }; static struct sh_timer_config tmu5_platform_data = { - .name = "TMU5", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -291,7 +280,6 @@ static struct sh_timer_config tmu5_platform_data = { static struct resource tmu5_resources[] = { [0] = { - .name = "TMU5", .start = 0xffd88020, .end = 0xffd8802b, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7770.c b/arch/sh/kernel/cpu/sh4a/setup-sh7770.c index 86d681e..83a27b0 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7770.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7770.c @@ -165,7 +165,6 @@ static struct platform_device scif9_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -174,7 +173,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -196,7 +194,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -205,7 +202,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -227,7 +223,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -235,7 +230,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002f, .flags = IORESOURCE_MEM, @@ -257,7 +251,6 @@ static struct platform_device tmu2_device = { }; static struct sh_timer_config tmu3_platform_data = { - .name = "TMU3", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -265,7 +258,6 @@ static struct sh_timer_config tmu3_platform_data = { static struct resource tmu3_resources[] = { [0] = { - .name = "TMU3", .start = 0xffd81008, .end = 0xffd81013, .flags = IORESOURCE_MEM, @@ -287,7 +279,6 @@ static struct platform_device tmu3_device = { }; static struct sh_timer_config tmu4_platform_data = { - .name = "TMU4", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -295,7 +286,6 @@ static struct sh_timer_config tmu4_platform_data = { static struct resource tmu4_resources[] = { [0] = { - .name = "TMU4", .start = 0xffd81014, .end = 0xffd8101f, .flags = IORESOURCE_MEM, @@ -317,7 +307,6 @@ static struct platform_device tmu4_device = { }; static struct sh_timer_config tmu5_platform_data = { - .name = "TMU5", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -325,7 +314,6 @@ static struct sh_timer_config tmu5_platform_data = { static struct resource tmu5_resources[] = { [0] = { - .name = "TMU5", .start = 0xffd81020, .end = 0xffd8102f, .flags = IORESOURCE_MEM, @@ -347,7 +335,6 @@ static struct platform_device tmu5_device = { }; static struct sh_timer_config tmu6_platform_data = { - .name = "TMU6", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -355,7 +342,6 @@ static struct sh_timer_config tmu6_platform_data = { static struct resource tmu6_resources[] = { [0] = { - .name = "TMU6", .start = 0xffd82008, .end = 0xffd82013, .flags = IORESOURCE_MEM, @@ -377,7 +363,6 @@ static struct platform_device tmu6_device = { }; static struct sh_timer_config tmu7_platform_data = { - .name = "TMU7", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -385,7 +370,6 @@ static struct sh_timer_config tmu7_platform_data = { static struct resource tmu7_resources[] = { [0] = { - .name = "TMU7", .start = 0xffd82014, .end = 0xffd8201f, .flags = IORESOURCE_MEM, @@ -407,7 +391,6 @@ static struct platform_device tmu7_device = { }; static struct sh_timer_config tmu8_platform_data = { - .name = "TMU8", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -415,7 +398,6 @@ static struct sh_timer_config tmu8_platform_data = { static struct resource tmu8_resources[] = { [0] = { - .name = "TMU8", .start = 0xffd82020, .end = 0xffd8202b, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7780.c b/arch/sh/kernel/cpu/sh4a/setup-sh7780.c index 02e792c..f7c48e9 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7780.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7780.c @@ -49,7 +49,6 @@ static struct platform_device scif1_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -58,7 +57,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -80,7 +78,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -89,7 +86,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -111,7 +107,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -119,7 +114,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002f, .flags = IORESOURCE_MEM, @@ -141,7 +135,6 @@ static struct platform_device tmu2_device = { }; static struct sh_timer_config tmu3_platform_data = { - .name = "TMU3", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -149,7 +142,6 @@ static struct sh_timer_config tmu3_platform_data = { static struct resource tmu3_resources[] = { [0] = { - .name = "TMU3", .start = 0xffdc0008, .end = 0xffdc0013, .flags = IORESOURCE_MEM, @@ -171,7 +163,6 @@ static struct platform_device tmu3_device = { }; static struct sh_timer_config tmu4_platform_data = { - .name = "TMU4", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -179,7 +170,6 @@ static struct sh_timer_config tmu4_platform_data = { static struct resource tmu4_resources[] = { [0] = { - .name = "TMU4", .start = 0xffdc0014, .end = 0xffdc001f, .flags = IORESOURCE_MEM, @@ -201,7 +191,6 @@ static struct platform_device tmu4_device = { }; static struct sh_timer_config tmu5_platform_data = { - .name = "TMU5", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -209,7 +198,6 @@ static struct sh_timer_config tmu5_platform_data = { static struct resource tmu5_resources[] = { [0] = { - .name = "TMU5", .start = 0xffdc0020, .end = 0xffdc002b, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7785.c b/arch/sh/kernel/cpu/sh4a/setup-sh7785.c index 1fcd88b..05e9308 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7785.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7785.c @@ -117,7 +117,6 @@ static struct platform_device scif5_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "tmu012_fck", @@ -126,7 +125,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -148,7 +146,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "tmu012_fck", @@ -157,7 +154,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -179,7 +175,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "tmu012_fck", @@ -187,7 +182,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002f, .flags = IORESOURCE_MEM, @@ -209,7 +203,6 @@ static struct platform_device tmu2_device = { }; static struct sh_timer_config tmu3_platform_data = { - .name = "TMU3", .channel_offset = 0x04, .timer_bit = 0, .clk = "tmu345_fck", @@ -217,7 +210,6 @@ static struct sh_timer_config tmu3_platform_data = { static struct resource tmu3_resources[] = { [0] = { - .name = "TMU3", .start = 0xffdc0008, .end = 0xffdc0013, .flags = IORESOURCE_MEM, @@ -239,7 +231,6 @@ static struct platform_device tmu3_device = { }; static struct sh_timer_config tmu4_platform_data = { - .name = "TMU4", .channel_offset = 0x10, .timer_bit = 1, .clk = "tmu345_fck", @@ -247,7 +238,6 @@ static struct sh_timer_config tmu4_platform_data = { static struct resource tmu4_resources[] = { [0] = { - .name = "TMU4", .start = 0xffdc0014, .end = 0xffdc001f, .flags = IORESOURCE_MEM, @@ -269,7 +259,6 @@ static struct platform_device tmu4_device = { }; static struct sh_timer_config tmu5_platform_data = { - .name = "TMU5", .channel_offset = 0x1c, .timer_bit = 2, .clk = "tmu345_fck", @@ -277,7 +266,6 @@ static struct sh_timer_config tmu5_platform_data = { static struct resource tmu5_resources[] = { [0] = { - .name = "TMU5", .start = 0xffdc0020, .end = 0xffdc002b, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7786.c b/arch/sh/kernel/cpu/sh4a/setup-sh7786.c index 7e58532..542ced8 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7786.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7786.c @@ -117,7 +117,6 @@ static struct platform_device scif5_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -126,7 +125,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffd80008, .end = 0xffd80013, .flags = IORESOURCE_MEM, @@ -148,7 +146,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -157,7 +154,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffd80014, .end = 0xffd8001f, .flags = IORESOURCE_MEM, @@ -179,7 +175,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -187,7 +182,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffd80020, .end = 0xffd8002f, .flags = IORESOURCE_MEM, @@ -209,7 +203,6 @@ static struct platform_device tmu2_device = { }; static struct sh_timer_config tmu3_platform_data = { - .name = "TMU3", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -217,7 +210,6 @@ static struct sh_timer_config tmu3_platform_data = { static struct resource tmu3_resources[] = { [0] = { - .name = "TMU3", .start = 0xffda0008, .end = 0xffda0013, .flags = IORESOURCE_MEM, @@ -239,7 +231,6 @@ static struct platform_device tmu3_device = { }; static struct sh_timer_config tmu4_platform_data = { - .name = "TMU4", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -247,7 +238,6 @@ static struct sh_timer_config tmu4_platform_data = { static struct resource tmu4_resources[] = { [0] = { - .name = "TMU4", .start = 0xffda0014, .end = 0xffda001f, .flags = IORESOURCE_MEM, @@ -269,7 +259,6 @@ static struct platform_device tmu4_device = { }; static struct sh_timer_config tmu5_platform_data = { - .name = "TMU5", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -277,7 +266,6 @@ static struct sh_timer_config tmu5_platform_data = { static struct resource tmu5_resources[] = { [0] = { - .name = "TMU5", .start = 0xffda0020, .end = 0xffda002b, .flags = IORESOURCE_MEM, @@ -299,7 +287,6 @@ static struct platform_device tmu5_device = { }; static struct sh_timer_config tmu6_platform_data = { - .name = "TMU6", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -307,7 +294,6 @@ static struct sh_timer_config tmu6_platform_data = { static struct resource tmu6_resources[] = { [0] = { - .name = "TMU6", .start = 0xffdc0008, .end = 0xffdc0013, .flags = IORESOURCE_MEM, @@ -329,7 +315,6 @@ static struct platform_device tmu6_device = { }; static struct sh_timer_config tmu7_platform_data = { - .name = "TMU7", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -337,7 +322,6 @@ static struct sh_timer_config tmu7_platform_data = { static struct resource tmu7_resources[] = { [0] = { - .name = "TMU7", .start = 0xffdc0014, .end = 0xffdc001f, .flags = IORESOURCE_MEM, @@ -359,7 +343,6 @@ static struct platform_device tmu7_device = { }; static struct sh_timer_config tmu8_platform_data = { - .name = "TMU8", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -367,7 +350,6 @@ static struct sh_timer_config tmu8_platform_data = { static struct resource tmu8_resources[] = { [0] = { - .name = "TMU8", .start = 0xffdc0020, .end = 0xffdc002b, .flags = IORESOURCE_MEM, @@ -389,7 +371,6 @@ static struct platform_device tmu8_device = { }; static struct sh_timer_config tmu9_platform_data = { - .name = "TMU9", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -397,7 +378,6 @@ static struct sh_timer_config tmu9_platform_data = { static struct resource tmu9_resources[] = { [0] = { - .name = "TMU9", .start = 0xffde0008, .end = 0xffde0013, .flags = IORESOURCE_MEM, @@ -419,7 +399,6 @@ static struct platform_device tmu9_device = { }; static struct sh_timer_config tmu10_platform_data = { - .name = "TMU10", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -427,7 +406,6 @@ static struct sh_timer_config tmu10_platform_data = { static struct resource tmu10_resources[] = { [0] = { - .name = "TMU10", .start = 0xffde0014, .end = 0xffde001f, .flags = IORESOURCE_MEM, @@ -449,7 +427,6 @@ static struct platform_device tmu10_device = { }; static struct sh_timer_config tmu11_platform_data = { - .name = "TMU11", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -457,7 +434,6 @@ static struct sh_timer_config tmu11_platform_data = { static struct resource tmu11_resources[] = { [0] = { - .name = "TMU11", .start = 0xffde0020, .end = 0xffde002b, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh4a/setup-shx3.c b/arch/sh/kernel/cpu/sh4a/setup-shx3.c index 780ba17..ba545b5 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-shx3.c +++ b/arch/sh/kernel/cpu/sh4a/setup-shx3.c @@ -70,7 +70,6 @@ static struct platform_device scif2_device = { }; static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -79,7 +78,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = 0xffc10008, .end = 0xffc10013, .flags = IORESOURCE_MEM, @@ -101,7 +99,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -110,7 +107,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = 0xffc10014, .end = 0xffc1001f, .flags = IORESOURCE_MEM, @@ -132,7 +128,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -140,7 +135,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = 0xffc10020, .end = 0xffc1002f, .flags = IORESOURCE_MEM, @@ -162,7 +156,6 @@ static struct platform_device tmu2_device = { }; static struct sh_timer_config tmu3_platform_data = { - .name = "TMU3", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -170,7 +163,6 @@ static struct sh_timer_config tmu3_platform_data = { static struct resource tmu3_resources[] = { [0] = { - .name = "TMU3", .start = 0xffc20008, .end = 0xffc20013, .flags = IORESOURCE_MEM, @@ -192,7 +184,6 @@ static struct platform_device tmu3_device = { }; static struct sh_timer_config tmu4_platform_data = { - .name = "TMU4", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -200,7 +191,6 @@ static struct sh_timer_config tmu4_platform_data = { static struct resource tmu4_resources[] = { [0] = { - .name = "TMU4", .start = 0xffc20014, .end = 0xffc2001f, .flags = IORESOURCE_MEM, @@ -222,7 +212,6 @@ static struct platform_device tmu4_device = { }; static struct sh_timer_config tmu5_platform_data = { - .name = "TMU5", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -230,7 +219,6 @@ static struct sh_timer_config tmu5_platform_data = { static struct resource tmu5_resources[] = { [0] = { - .name = "TMU5", .start = 0xffc20020, .end = 0xffc2002b, .flags = IORESOURCE_MEM, diff --git a/arch/sh/kernel/cpu/sh5/setup-sh5.c b/arch/sh/kernel/cpu/sh5/setup-sh5.c index e7a3c1e..c84ca20 100644 --- a/arch/sh/kernel/cpu/sh5/setup-sh5.c +++ b/arch/sh/kernel/cpu/sh5/setup-sh5.c @@ -68,7 +68,6 @@ static struct platform_device rtc_device = { #define TMU2_BASE (TMU_BASE + 0x8 + (0xc * 0x2)) static struct sh_timer_config tmu0_platform_data = { - .name = "TMU0", .channel_offset = 0x04, .timer_bit = 0, .clk = "peripheral_clk", @@ -77,7 +76,6 @@ static struct sh_timer_config tmu0_platform_data = { static struct resource tmu0_resources[] = { [0] = { - .name = "TMU0", .start = TMU0_BASE, .end = TMU0_BASE + 0xc - 1, .flags = IORESOURCE_MEM, @@ -99,7 +97,6 @@ static struct platform_device tmu0_device = { }; static struct sh_timer_config tmu1_platform_data = { - .name = "TMU1", .channel_offset = 0x10, .timer_bit = 1, .clk = "peripheral_clk", @@ -108,7 +105,6 @@ static struct sh_timer_config tmu1_platform_data = { static struct resource tmu1_resources[] = { [0] = { - .name = "TMU1", .start = TMU1_BASE, .end = TMU1_BASE + 0xc - 1, .flags = IORESOURCE_MEM, @@ -130,7 +126,6 @@ static struct platform_device tmu1_device = { }; static struct sh_timer_config tmu2_platform_data = { - .name = "TMU2", .channel_offset = 0x1c, .timer_bit = 2, .clk = "peripheral_clk", @@ -138,7 +133,6 @@ static struct sh_timer_config tmu2_platform_data = { static struct resource tmu2_resources[] = { [0] = { - .name = "TMU2", .start = TMU2_BASE, .end = TMU2_BASE + 0xc - 1, .flags = IORESOURCE_MEM, -- cgit v0.10.2 From c7ed1ab3f7b2fe0dedebf34cbf40bf12cb5ae48b Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Wed, 10 Mar 2010 18:35:14 +0900 Subject: serial: sh-sci: clkdev updates for MSTP gating. This fixes up some of the I/D/F clock ambiguity in the sh-sci driver. The interface clock in most cases just wraps back to the peripheral clock, while the function clock wraps in to the MSTP bits. As the logic was somewhat inverted, this cleans that up, and also enables all CPUs with SCI MSTP bits to match function clocks through clkdev lookup. As a result, this gets rid of the clk string abuse on the sh side, and the clock string will be killed off once the ARM code has had a chance to sync up. This also enables MSTP gating on CPUs like 7786 which had never wired it up before. Impacted CPUs are primarily all SH-Mobiles, SH7785, and SH7786. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7343.c b/arch/sh/kernel/cpu/sh4a/clock-sh7343.c index 2c16df3..c541301 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7343.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7343.c @@ -159,10 +159,10 @@ static struct clk mstp_clks[] = { MSTP("rwdt0", &r_clk, MSTPCR0, 13, 0), MSTP("mfi0", &div4_clks[DIV4_P], MSTPCR0, 11, 0), MSTP("flctl0", &div4_clks[DIV4_P], MSTPCR0, 10, 0), - MSTP("scif0", &div4_clks[DIV4_P], MSTPCR0, 7, 0), - MSTP("scif1", &div4_clks[DIV4_P], MSTPCR0, 6, 0), - MSTP("scif2", &div4_clks[DIV4_P], MSTPCR0, 5, 0), - MSTP("scif3", &div4_clks[DIV4_P], MSTPCR0, 4, 0), + SH_CLK_MSTP32("sci_fck", 0, &div4_clks[DIV4_P], MSTPCR0, 7, 0), + SH_CLK_MSTP32("sci_fck", 1, &div4_clks[DIV4_P], MSTPCR0, 6, 0), + SH_CLK_MSTP32("sci_fck", 2, &div4_clks[DIV4_P], MSTPCR0, 5, 0), + SH_CLK_MSTP32("sci_fck", 3, &div4_clks[DIV4_P], MSTPCR0, 4, 0), MSTP("sio0", &div4_clks[DIV4_P], MSTPCR0, 3, 0), MSTP("siof0", &div4_clks[DIV4_P], MSTPCR0, 2, 0), MSTP("siof1", &div4_clks[DIV4_P], MSTPCR0, 1, 0), diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7366.c b/arch/sh/kernel/cpu/sh4a/clock-sh7366.c index 91588d2..e10c64f 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7366.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7366.c @@ -163,9 +163,9 @@ static struct clk mstp_clks[] = { MSTP("rwdt0", &r_clk, MSTPCR0, 13, 0), MSTP("mfi0", &div4_clks[DIV4_P], MSTPCR0, 11, 0), MSTP("flctl0", &div4_clks[DIV4_P], MSTPCR0, 10, 0), - MSTP("scif0", &div4_clks[DIV4_P], MSTPCR0, 7, 0), - MSTP("scif1", &div4_clks[DIV4_P], MSTPCR0, 6, 0), - MSTP("scif2", &div4_clks[DIV4_P], MSTPCR0, 5, 0), + SH_CLK_MSTP32("sci_fck", 0, &div4_clks[DIV4_P], MSTPCR0, 7, 0), + SH_CLK_MSTP32("sci_fck", 1, &div4_clks[DIV4_P], MSTPCR0, 6, 0), + SH_CLK_MSTP32("sci_fck", 2, &div4_clks[DIV4_P], MSTPCR0, 5, 0), MSTP("msiof0", &div4_clks[DIV4_P], MSTPCR0, 2, 0), MSTP("sbr0", &div4_clks[DIV4_P], MSTPCR0, 1, 0), diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7722.c b/arch/sh/kernel/cpu/sh4a/clock-sh7722.c index 15db6d5..0732c8b 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7722.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7722.c @@ -164,9 +164,9 @@ static struct clk mstp_clks[] = { SH_HWBLK_CLK("cmt0", -1, R_CLK, HWBLK_CMT, 0), SH_HWBLK_CLK("rwdt0", -1, R_CLK, HWBLK_RWDT, 0), SH_HWBLK_CLK("flctl0", -1, P_CLK, HWBLK_FLCTL, 0), - SH_HWBLK_CLK("scif0", -1, P_CLK, HWBLK_SCIF0, 0), - SH_HWBLK_CLK("scif1", -1, P_CLK, HWBLK_SCIF1, 0), - SH_HWBLK_CLK("scif2", -1, P_CLK, HWBLK_SCIF2, 0), + SH_HWBLK_CLK("sci_fck", 0, P_CLK, HWBLK_SCIF0, 0), + SH_HWBLK_CLK("sci_fck", 1, P_CLK, HWBLK_SCIF1, 0), + SH_HWBLK_CLK("sci_fck", 2, P_CLK, HWBLK_SCIF2, 0), SH_HWBLK_CLK("i2c0", -1, P_CLK, HWBLK_IIC, 0), SH_HWBLK_CLK("rtc0", -1, R_CLK, HWBLK_RTC, 0), diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7723.c b/arch/sh/kernel/cpu/sh4a/clock-sh7723.c index 50babe0..4d4c123 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7723.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7723.c @@ -177,12 +177,12 @@ static struct clk mstp_clks[] = { SH_HWBLK_CLK("dmac1", -1, B_CLK, HWBLK_DMAC1, 0), SH_HWBLK_CLK("tmu1", -1, P_CLK, HWBLK_TMU1, 0), SH_HWBLK_CLK("flctl0", -1, P_CLK, HWBLK_FLCTL, 0), - SH_HWBLK_CLK("scif0", -1, P_CLK, HWBLK_SCIF0, 0), - SH_HWBLK_CLK("scif1", -1, P_CLK, HWBLK_SCIF1, 0), - SH_HWBLK_CLK("scif2", -1, P_CLK, HWBLK_SCIF2, 0), - SH_HWBLK_CLK("scif3", -1, B_CLK, HWBLK_SCIF3, 0), - SH_HWBLK_CLK("scif4", -1, B_CLK, HWBLK_SCIF4, 0), - SH_HWBLK_CLK("scif5", -1, B_CLK, HWBLK_SCIF5, 0), + SH_HWBLK_CLK("sci_fck", 0, P_CLK, HWBLK_SCIF0, 0), + SH_HWBLK_CLK("sci_fck", 1, P_CLK, HWBLK_SCIF1, 0), + SH_HWBLK_CLK("sci_fck", 2, P_CLK, HWBLK_SCIF2, 0), + SH_HWBLK_CLK("sci_fck", 3, B_CLK, HWBLK_SCIF3, 0), + SH_HWBLK_CLK("sci_fck", 4, B_CLK, HWBLK_SCIF4, 0), + SH_HWBLK_CLK("sci_fck", 5, B_CLK, HWBLK_SCIF5, 0), SH_HWBLK_CLK("msiof0", -1, B_CLK, HWBLK_MSIOF0, 0), SH_HWBLK_CLK("msiof1", -1, B_CLK, HWBLK_MSIOF1, 0), SH_HWBLK_CLK("meram0", -1, SH_CLK, HWBLK_MERAM, 0), diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7724.c b/arch/sh/kernel/cpu/sh4a/clock-sh7724.c index 6707061..bcf2b97 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7724.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7724.c @@ -194,12 +194,12 @@ static struct clk mstp_clks[] = { SH_HWBLK_CLK("rwdt0", -1, R_CLK, HWBLK_RWDT, 0), SH_HWBLK_CLK("dmac1", -1, B_CLK, HWBLK_DMAC1, 0), SH_HWBLK_CLK("tmu1", -1, P_CLK, HWBLK_TMU1, 0), - SH_HWBLK_CLK("scif0", -1, P_CLK, HWBLK_SCIF0, 0), - SH_HWBLK_CLK("scif1", -1, P_CLK, HWBLK_SCIF1, 0), - SH_HWBLK_CLK("scif2", -1, P_CLK, HWBLK_SCIF2, 0), - SH_HWBLK_CLK("scif3", -1, B_CLK, HWBLK_SCIF3, 0), - SH_HWBLK_CLK("scif4", -1, B_CLK, HWBLK_SCIF4, 0), - SH_HWBLK_CLK("scif5", -1, B_CLK, HWBLK_SCIF5, 0), + SH_HWBLK_CLK("sci_fck", 0, P_CLK, HWBLK_SCIF0, 0), + SH_HWBLK_CLK("sci_fck", 1, P_CLK, HWBLK_SCIF1, 0), + SH_HWBLK_CLK("sci_fck", 2, P_CLK, HWBLK_SCIF2, 0), + SH_HWBLK_CLK("sci_fck", 3, B_CLK, HWBLK_SCIF3, 0), + SH_HWBLK_CLK("sci_fck", 4, B_CLK, HWBLK_SCIF4, 0), + SH_HWBLK_CLK("sci_fck", 5, B_CLK, HWBLK_SCIF5, 0), SH_HWBLK_CLK("msiof0", -1, B_CLK, HWBLK_MSIOF0, 0), SH_HWBLK_CLK("msiof1", -1, B_CLK, HWBLK_MSIOF1, 0), diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7785.c b/arch/sh/kernel/cpu/sh4a/clock-sh7785.c index d997f0a..6061c73 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7785.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7785.c @@ -88,12 +88,12 @@ struct clk div4_clks[DIV4_NR] = { static struct clk mstp_clks[] = { /* MSTPCR0 */ - SH_CLK_MSTP32("scif_fck", 5, &div4_clks[DIV4_P], MSTPCR0, 29, 0), - SH_CLK_MSTP32("scif_fck", 4, &div4_clks[DIV4_P], MSTPCR0, 28, 0), - SH_CLK_MSTP32("scif_fck", 3, &div4_clks[DIV4_P], MSTPCR0, 27, 0), - SH_CLK_MSTP32("scif_fck", 2, &div4_clks[DIV4_P], MSTPCR0, 26, 0), - SH_CLK_MSTP32("scif_fck", 1, &div4_clks[DIV4_P], MSTPCR0, 25, 0), - SH_CLK_MSTP32("scif_fck", 0, &div4_clks[DIV4_P], MSTPCR0, 24, 0), + SH_CLK_MSTP32("sci_fck", 5, &div4_clks[DIV4_P], MSTPCR0, 29, 0), + SH_CLK_MSTP32("sci_fck", 4, &div4_clks[DIV4_P], MSTPCR0, 28, 0), + SH_CLK_MSTP32("sci_fck", 3, &div4_clks[DIV4_P], MSTPCR0, 27, 0), + SH_CLK_MSTP32("sci_fck", 2, &div4_clks[DIV4_P], MSTPCR0, 26, 0), + SH_CLK_MSTP32("sci_fck", 1, &div4_clks[DIV4_P], MSTPCR0, 25, 0), + SH_CLK_MSTP32("sci_fck", 0, &div4_clks[DIV4_P], MSTPCR0, 24, 0), SH_CLK_MSTP32("ssi_fck", 1, &div4_clks[DIV4_P], MSTPCR0, 21, 0), SH_CLK_MSTP32("ssi_fck", 0, &div4_clks[DIV4_P], MSTPCR0, 20, 0), SH_CLK_MSTP32("hac_fck", 1, &div4_clks[DIV4_P], MSTPCR0, 17, 0), diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7786.c b/arch/sh/kernel/cpu/sh4a/clock-sh7786.c index af69fd4..85e8d57 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7786.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7786.c @@ -87,12 +87,12 @@ struct clk div4_clks[DIV4_NR] = { static struct clk mstp_clks[] = { /* MSTPCR0 */ - SH_CLK_MSTP32("scif_fck", 5, &div4_clks[DIV4_P], MSTPCR0, 29, 0), - SH_CLK_MSTP32("scif_fck", 4, &div4_clks[DIV4_P], MSTPCR0, 28, 0), - SH_CLK_MSTP32("scif_fck", 3, &div4_clks[DIV4_P], MSTPCR0, 27, 0), - SH_CLK_MSTP32("scif_fck", 2, &div4_clks[DIV4_P], MSTPCR0, 26, 0), - SH_CLK_MSTP32("scif_fck", 1, &div4_clks[DIV4_P], MSTPCR0, 25, 0), - SH_CLK_MSTP32("scif_fck", 0, &div4_clks[DIV4_P], MSTPCR0, 24, 0), + SH_CLK_MSTP32("sci_fck", 5, &div4_clks[DIV4_P], MSTPCR0, 29, 0), + SH_CLK_MSTP32("sci_fck", 4, &div4_clks[DIV4_P], MSTPCR0, 28, 0), + SH_CLK_MSTP32("sci_fck", 3, &div4_clks[DIV4_P], MSTPCR0, 27, 0), + SH_CLK_MSTP32("sci_fck", 2, &div4_clks[DIV4_P], MSTPCR0, 26, 0), + SH_CLK_MSTP32("sci_fck", 1, &div4_clks[DIV4_P], MSTPCR0, 25, 0), + SH_CLK_MSTP32("sci_fck", 0, &div4_clks[DIV4_P], MSTPCR0, 24, 0), SH_CLK_MSTP32("ssi_fck", 3, &div4_clks[DIV4_P], MSTPCR0, 23, 0), SH_CLK_MSTP32("ssi_fck", 2, &div4_clks[DIV4_P], MSTPCR0, 22, 0), SH_CLK_MSTP32("ssi_fck", 1, &div4_clks[DIV4_P], MSTPCR0, 21, 0), diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7343.c b/arch/sh/kernel/cpu/sh4a/setup-sh7343.c index fdbd7d0..6f83949 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7343.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7343.c @@ -21,7 +21,6 @@ static struct plat_sci_port scif0_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 80, 80, 80, 80 }, - .clk = "scif0", }; static struct platform_device scif0_device = { @@ -37,7 +36,6 @@ static struct plat_sci_port scif1_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 81, 81, 81, 81 }, - .clk = "scif1", }; static struct platform_device scif1_device = { @@ -53,7 +51,6 @@ static struct plat_sci_port scif2_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 82, 82, 82, 82 }, - .clk = "scif2", }; static struct platform_device scif2_device = { @@ -69,7 +66,6 @@ static struct plat_sci_port scif3_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 83, 83, 83, 83 }, - .clk = "scif3", }; static struct platform_device scif3_device = { diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7366.c b/arch/sh/kernel/cpu/sh4a/setup-sh7366.c index d90fe0a..1d897ed 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7366.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7366.c @@ -23,7 +23,6 @@ static struct plat_sci_port scif0_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 80, 80, 80, 80 }, - .clk = "scif0", }; static struct platform_device scif0_device = { diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7722.c b/arch/sh/kernel/cpu/sh4a/setup-sh7722.c index 83ae9ad..847a276 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7722.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7722.c @@ -174,7 +174,6 @@ static struct plat_sci_port scif0_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 80, 80, 80, 80 }, - .clk = "scif0", }; static struct platform_device scif0_device = { @@ -190,7 +189,6 @@ static struct plat_sci_port scif1_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 81, 81, 81, 81 }, - .clk = "scif1", }; static struct platform_device scif1_device = { @@ -206,7 +204,6 @@ static struct plat_sci_port scif2_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 82, 82, 82, 82 }, - .clk = "scif2", }; static struct platform_device scif2_device = { diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7723.c b/arch/sh/kernel/cpu/sh4a/setup-sh7723.c index 68446b3f..187462b 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7723.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7723.c @@ -26,7 +26,6 @@ static struct plat_sci_port scif0_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 80, 80, 80, 80 }, - .clk = "scif0", }; static struct platform_device scif0_device = { @@ -42,7 +41,6 @@ static struct plat_sci_port scif1_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 81, 81, 81, 81 }, - .clk = "scif1", }; static struct platform_device scif1_device = { @@ -58,7 +56,6 @@ static struct plat_sci_port scif2_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 82, 82, 82, 82 }, - .clk = "scif2", }; static struct platform_device scif2_device = { @@ -74,7 +71,6 @@ static struct plat_sci_port scif3_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIFA, .irqs = { 56, 56, 56, 56 }, - .clk = "scif3", }; static struct platform_device scif3_device = { @@ -90,7 +86,6 @@ static struct plat_sci_port scif4_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIFA, .irqs = { 88, 88, 88, 88 }, - .clk = "scif4", }; static struct platform_device scif4_device = { @@ -106,7 +101,6 @@ static struct plat_sci_port scif5_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIFA, .irqs = { 109, 109, 109, 109 }, - .clk = "scif5", }; static struct platform_device scif5_device = { diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7724.c b/arch/sh/kernel/cpu/sh4a/setup-sh7724.c index abc98b0..ee97890 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7724.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7724.c @@ -213,7 +213,6 @@ static struct plat_sci_port scif0_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 80, 80, 80, 80 }, - .clk = "scif0", }; static struct platform_device scif0_device = { @@ -229,7 +228,6 @@ static struct plat_sci_port scif1_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 81, 81, 81, 81 }, - .clk = "scif1", }; static struct platform_device scif1_device = { @@ -245,7 +243,6 @@ static struct plat_sci_port scif2_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 82, 82, 82, 82 }, - .clk = "scif2", }; static struct platform_device scif2_device = { @@ -261,7 +258,6 @@ static struct plat_sci_port scif3_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIFA, .irqs = { 56, 56, 56, 56 }, - .clk = "scif3", }; static struct platform_device scif3_device = { @@ -277,7 +273,6 @@ static struct plat_sci_port scif4_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIFA, .irqs = { 88, 88, 88, 88 }, - .clk = "scif4", }; static struct platform_device scif4_device = { @@ -293,7 +288,6 @@ static struct plat_sci_port scif5_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIFA, .irqs = { 109, 109, 109, 109 }, - .clk = "scif5", }; static struct platform_device scif5_device = { diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7785.c b/arch/sh/kernel/cpu/sh4a/setup-sh7785.c index 05e9308..76afe57 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7785.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7785.c @@ -25,7 +25,6 @@ static struct plat_sci_port scif0_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 40, 40, 40, 40 }, - .clk = "scif_fck", }; static struct platform_device scif0_device = { @@ -41,7 +40,6 @@ static struct plat_sci_port scif1_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 44, 44, 44, 44 }, - .clk = "scif_fck", }; static struct platform_device scif1_device = { @@ -57,7 +55,6 @@ static struct plat_sci_port scif2_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 60, 60, 60, 60 }, - .clk = "scif_fck", }; static struct platform_device scif2_device = { @@ -73,7 +70,6 @@ static struct plat_sci_port scif3_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 61, 61, 61, 61 }, - .clk = "scif_fck", }; static struct platform_device scif3_device = { @@ -89,7 +85,6 @@ static struct plat_sci_port scif4_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 62, 62, 62, 62 }, - .clk = "scif_fck", }; static struct platform_device scif4_device = { @@ -105,7 +100,6 @@ static struct plat_sci_port scif5_platform_data = { .flags = UPF_BOOT_AUTOCONF, .type = PORT_SCIF, .irqs = { 63, 63, 63, 63 }, - .clk = "scif_fck", }; static struct platform_device scif5_device = { diff --git a/drivers/serial/sh-sci.c b/drivers/serial/sh-sci.c index 980f394..054a842 100644 --- a/drivers/serial/sh-sci.c +++ b/drivers/serial/sh-sci.c @@ -83,8 +83,8 @@ struct sci_port { /* Interface clock */ struct clk *iclk; - /* Data clock */ - struct clk *dclk; + /* Function clock */ + struct clk *fclk; struct list_head node; struct dma_chan *chan_tx; @@ -803,7 +803,7 @@ static int sci_notifier(struct notifier_block *self, (phase == CPUFREQ_RESUMECHANGE)) { spin_lock_irqsave(&priv->lock, flags); list_for_each_entry(sci_port, &priv->ports, node) - sci_port->port.uartclk = clk_get_rate(sci_port->dclk); + sci_port->port.uartclk = clk_get_rate(sci_port->iclk); spin_unlock_irqrestore(&priv->lock, flags); } @@ -814,21 +814,17 @@ static void sci_clk_enable(struct uart_port *port) { struct sci_port *sci_port = to_sci_port(port); - clk_enable(sci_port->dclk); - sci_port->port.uartclk = clk_get_rate(sci_port->dclk); - - if (sci_port->iclk) - clk_enable(sci_port->iclk); + clk_enable(sci_port->iclk); + sci_port->port.uartclk = clk_get_rate(sci_port->iclk); + clk_enable(sci_port->fclk); } static void sci_clk_disable(struct uart_port *port) { struct sci_port *sci_port = to_sci_port(port); - if (sci_port->iclk) - clk_disable(sci_port->iclk); - - clk_disable(sci_port->dclk); + clk_disable(sci_port->fclk); + clk_disable(sci_port->iclk); } static int sci_request_irq(struct sci_port *port) @@ -1557,10 +1553,10 @@ static struct uart_ops sci_uart_ops = { #endif }; -static void __devinit sci_init_single(struct platform_device *dev, - struct sci_port *sci_port, - unsigned int index, - struct plat_sci_port *p) +static int __devinit sci_init_single(struct platform_device *dev, + struct sci_port *sci_port, + unsigned int index, + struct plat_sci_port *p) { struct uart_port *port = &sci_port->port; @@ -1581,8 +1577,23 @@ static void __devinit sci_init_single(struct platform_device *dev, } if (dev) { - sci_port->iclk = p->clk ? clk_get(&dev->dev, p->clk) : NULL; - sci_port->dclk = clk_get(&dev->dev, "peripheral_clk"); + sci_port->iclk = clk_get(&dev->dev, "sci_ick"); + if (IS_ERR(sci_port->iclk)) { + sci_port->iclk = clk_get(&dev->dev, "peripheral_clk"); + if (IS_ERR(sci_port->iclk)) { + dev_err(&dev->dev, "can't get iclk\n"); + return PTR_ERR(sci_port->iclk); + } + } + + /* + * The function clock is optional, ignore it if we can't + * find it. + */ + sci_port->fclk = clk_get(&dev->dev, "sci_fck"); + if (IS_ERR(sci_port->fclk)) + sci_port->fclk = NULL; + sci_port->enable = sci_clk_enable; sci_port->disable = sci_clk_disable; port->dev = &dev->dev; @@ -1609,6 +1620,7 @@ static void __devinit sci_init_single(struct platform_device *dev, #endif memcpy(&sci_port->irqs, &p->irqs, sizeof(p->irqs)); + return 0; } #ifdef CONFIG_SERIAL_SH_SCI_CONSOLE @@ -1758,8 +1770,11 @@ static int sci_remove(struct platform_device *dev) cpufreq_unregister_notifier(&priv->clk_nb, CPUFREQ_TRANSITION_NOTIFIER); spin_lock_irqsave(&priv->lock, flags); - list_for_each_entry(p, &priv->ports, node) + list_for_each_entry(p, &priv->ports, node) { uart_remove_one_port(&sci_uart_driver, &p->port); + clk_put(p->iclk); + clk_put(p->fclk); + } spin_unlock_irqrestore(&priv->lock, flags); kfree(priv); @@ -1785,7 +1800,9 @@ static int __devinit sci_probe_single(struct platform_device *dev, return 0; } - sci_init_single(dev, sciport, index, p); + ret = sci_init_single(dev, sciport, index, p); + if (ret) + return ret; ret = uart_add_one_port(&sci_uart_driver, &sciport->port); if (ret) -- cgit v0.10.2 From ca037701a025334e724e5c61b3b1082940c8b981 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 2 Mar 2010 19:52:12 +0100 Subject: perf, x86: Add PEBS infrastructure This patch implements support for Intel Precise Event Based Sampling, which is an alternative counter mode in which the counter triggers a hardware assist to collect information on events. The hardware assist takes a trap like snapshot of a subset of the machine registers. This data is written to the Intel Debug-Store, which can be programmed with a data threshold at which to raise a PMI. With the PEBS hardware assist being trap like, the reported IP is always one instruction after the actual instruction that triggered the event. This implements a simple PEBS model that always takes a single PEBS event at a time. This is done so that the interaction with the rest of the system is as expected (freq adjust, period randomization, lbr, callchains, etc.). It adds an ABI element: perf_event_attr::precise, which indicates that we wish to use this (constrained, but precise) mode. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: <20100304140100.392111285@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 1d665a0b..0c03d5c 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -31,45 +31,6 @@ static u64 perf_event_mask __read_mostly; -/* The maximal number of PEBS events: */ -#define MAX_PEBS_EVENTS 4 - -/* The size of a BTS record in bytes: */ -#define BTS_RECORD_SIZE 24 - -/* The size of a per-cpu BTS buffer in bytes: */ -#define BTS_BUFFER_SIZE (BTS_RECORD_SIZE * 2048) - -/* The BTS overflow threshold in bytes from the end of the buffer: */ -#define BTS_OVFL_TH (BTS_RECORD_SIZE * 128) - - -/* - * Bits in the debugctlmsr controlling branch tracing. - */ -#define X86_DEBUGCTL_TR (1 << 6) -#define X86_DEBUGCTL_BTS (1 << 7) -#define X86_DEBUGCTL_BTINT (1 << 8) -#define X86_DEBUGCTL_BTS_OFF_OS (1 << 9) -#define X86_DEBUGCTL_BTS_OFF_USR (1 << 10) - -/* - * A debug store configuration. - * - * We only support architectures that use 64bit fields. - */ -struct debug_store { - u64 bts_buffer_base; - u64 bts_index; - u64 bts_absolute_maximum; - u64 bts_interrupt_threshold; - u64 pebs_buffer_base; - u64 pebs_index; - u64 pebs_absolute_maximum; - u64 pebs_interrupt_threshold; - u64 pebs_event_reset[MAX_PEBS_EVENTS]; -}; - struct event_constraint { union { unsigned long idxmsk[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; @@ -88,17 +49,29 @@ struct amd_nb { }; struct cpu_hw_events { + /* + * Generic x86 PMC bits + */ struct perf_event *events[X86_PMC_IDX_MAX]; /* in counter order */ unsigned long active_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; unsigned long interrupts; int enabled; - struct debug_store *ds; int n_events; int n_added; int assign[X86_PMC_IDX_MAX]; /* event to counter assignment */ u64 tags[X86_PMC_IDX_MAX]; struct perf_event *event_list[X86_PMC_IDX_MAX]; /* in enabled order */ + + /* + * Intel DebugStore bits + */ + struct debug_store *ds; + u64 pebs_enabled; + + /* + * AMD specific bits + */ struct amd_nb *amd_nb; }; @@ -112,12 +85,24 @@ struct cpu_hw_events { #define EVENT_CONSTRAINT(c, n, m) \ __EVENT_CONSTRAINT(c, n, m, HWEIGHT(n)) +/* + * Constraint on the Event code. + */ #define INTEL_EVENT_CONSTRAINT(c, n) \ EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVTSEL_MASK) +/* + * Constraint on the Event code + UMask + fixed-mask + */ #define FIXED_EVENT_CONSTRAINT(c, n) \ EVENT_CONSTRAINT(c, (1ULL << (32+n)), INTEL_ARCH_FIXED_MASK) +/* + * Constraint on the Event code + UMask + */ +#define PEBS_EVENT_CONSTRAINT(c, n) \ + EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVENT_MASK) + #define EVENT_CONSTRAINT_END \ EVENT_CONSTRAINT(0, 0, 0) @@ -128,6 +113,9 @@ struct cpu_hw_events { * struct x86_pmu - generic x86 pmu */ struct x86_pmu { + /* + * Generic x86 PMC bits + */ const char *name; int version; int (*handle_irq)(struct pt_regs *); @@ -146,10 +134,6 @@ struct x86_pmu { u64 event_mask; int apic; u64 max_period; - u64 intel_ctrl; - void (*enable_bts)(u64 config); - void (*disable_bts)(void); - struct event_constraint * (*get_event_constraints)(struct cpu_hw_events *cpuc, struct perf_event *event); @@ -162,6 +146,19 @@ struct x86_pmu { void (*cpu_starting)(int cpu); void (*cpu_dying)(int cpu); void (*cpu_dead)(int cpu); + + /* + * Intel Arch Perfmon v2+ + */ + u64 intel_ctrl; + + /* + * Intel DebugStore bits + */ + int bts, pebs; + int pebs_record_size; + void (*drain_pebs)(struct pt_regs *regs); + struct event_constraint *pebs_constraints; }; static struct x86_pmu x86_pmu __read_mostly; @@ -293,110 +290,14 @@ static void release_pmc_hardware(void) #endif } -static inline bool bts_available(void) -{ - return x86_pmu.enable_bts != NULL; -} - -static void init_debug_store_on_cpu(int cpu) -{ - struct debug_store *ds = per_cpu(cpu_hw_events, cpu).ds; - - if (!ds) - return; - - wrmsr_on_cpu(cpu, MSR_IA32_DS_AREA, - (u32)((u64)(unsigned long)ds), - (u32)((u64)(unsigned long)ds >> 32)); -} - -static void fini_debug_store_on_cpu(int cpu) -{ - if (!per_cpu(cpu_hw_events, cpu).ds) - return; - - wrmsr_on_cpu(cpu, MSR_IA32_DS_AREA, 0, 0); -} - -static void release_bts_hardware(void) -{ - int cpu; - - if (!bts_available()) - return; - - get_online_cpus(); - - for_each_online_cpu(cpu) - fini_debug_store_on_cpu(cpu); - - for_each_possible_cpu(cpu) { - struct debug_store *ds = per_cpu(cpu_hw_events, cpu).ds; - - if (!ds) - continue; - - per_cpu(cpu_hw_events, cpu).ds = NULL; - - kfree((void *)(unsigned long)ds->bts_buffer_base); - kfree(ds); - } - - put_online_cpus(); -} - -static int reserve_bts_hardware(void) -{ - int cpu, err = 0; - - if (!bts_available()) - return 0; - - get_online_cpus(); - - for_each_possible_cpu(cpu) { - struct debug_store *ds; - void *buffer; - - err = -ENOMEM; - buffer = kzalloc(BTS_BUFFER_SIZE, GFP_KERNEL); - if (unlikely(!buffer)) - break; - - ds = kzalloc(sizeof(*ds), GFP_KERNEL); - if (unlikely(!ds)) { - kfree(buffer); - break; - } - - ds->bts_buffer_base = (u64)(unsigned long)buffer; - ds->bts_index = ds->bts_buffer_base; - ds->bts_absolute_maximum = - ds->bts_buffer_base + BTS_BUFFER_SIZE; - ds->bts_interrupt_threshold = - ds->bts_absolute_maximum - BTS_OVFL_TH; - - per_cpu(cpu_hw_events, cpu).ds = ds; - err = 0; - } - - if (err) - release_bts_hardware(); - else { - for_each_online_cpu(cpu) - init_debug_store_on_cpu(cpu); - } - - put_online_cpus(); - - return err; -} +static int reserve_ds_buffers(void); +static void release_ds_buffers(void); static void hw_perf_event_destroy(struct perf_event *event) { if (atomic_dec_and_mutex_lock(&active_events, &pmc_reserve_mutex)) { release_pmc_hardware(); - release_bts_hardware(); + release_ds_buffers(); mutex_unlock(&pmc_reserve_mutex); } } @@ -459,7 +360,7 @@ static int __hw_perf_event_init(struct perf_event *event) if (!reserve_pmc_hardware()) err = -EBUSY; else - err = reserve_bts_hardware(); + err = reserve_ds_buffers(); } if (!err) atomic_inc(&active_events); @@ -537,7 +438,7 @@ static int __hw_perf_event_init(struct perf_event *event) if ((attr->config == PERF_COUNT_HW_BRANCH_INSTRUCTIONS) && (hwc->sample_period == 1)) { /* BTS is not supported by this architecture. */ - if (!bts_available()) + if (!x86_pmu.bts) return -EOPNOTSUPP; /* BTS is currently only allowed for user-mode. */ @@ -995,6 +896,7 @@ static void x86_pmu_unthrottle(struct perf_event *event) void perf_event_print_debug(void) { u64 ctrl, status, overflow, pmc_ctrl, pmc_count, prev_left, fixed; + u64 pebs; struct cpu_hw_events *cpuc; unsigned long flags; int cpu, idx; @@ -1012,12 +914,14 @@ void perf_event_print_debug(void) rdmsrl(MSR_CORE_PERF_GLOBAL_STATUS, status); rdmsrl(MSR_CORE_PERF_GLOBAL_OVF_CTRL, overflow); rdmsrl(MSR_ARCH_PERFMON_FIXED_CTR_CTRL, fixed); + rdmsrl(MSR_IA32_PEBS_ENABLE, pebs); pr_info("\n"); pr_info("CPU#%d: ctrl: %016llx\n", cpu, ctrl); pr_info("CPU#%d: status: %016llx\n", cpu, status); pr_info("CPU#%d: overflow: %016llx\n", cpu, overflow); pr_info("CPU#%d: fixed: %016llx\n", cpu, fixed); + pr_info("CPU#%d: pebs: %016llx\n", cpu, pebs); } pr_info("CPU#%d: active: %016llx\n", cpu, *(u64 *)cpuc->active_mask); @@ -1333,6 +1237,7 @@ undo: #include "perf_event_amd.c" #include "perf_event_p6.c" +#include "perf_event_intel_ds.c" #include "perf_event_intel.c" static int __cpuinit @@ -1465,6 +1370,32 @@ static const struct pmu pmu = { }; /* + * validate that we can schedule this event + */ +static int validate_event(struct perf_event *event) +{ + struct cpu_hw_events *fake_cpuc; + struct event_constraint *c; + int ret = 0; + + fake_cpuc = kmalloc(sizeof(*fake_cpuc), GFP_KERNEL | __GFP_ZERO); + if (!fake_cpuc) + return -ENOMEM; + + c = x86_pmu.get_event_constraints(fake_cpuc, event); + + if (!c || !c->weight) + ret = -ENOSPC; + + if (x86_pmu.put_event_constraints) + x86_pmu.put_event_constraints(fake_cpuc, event); + + kfree(fake_cpuc); + + return ret; +} + +/* * validate a single event group * * validation include: @@ -1529,6 +1460,8 @@ const struct pmu *hw_perf_event_init(struct perf_event *event) if (event->group_leader != event) err = validate_group(event); + else + err = validate_event(event); event->pmu = tmp; } diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 84bfde6..1144641 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -470,42 +470,6 @@ static u64 intel_pmu_raw_event(u64 hw_event) return hw_event & CORE_EVNTSEL_MASK; } -static void intel_pmu_enable_bts(u64 config) -{ - unsigned long debugctlmsr; - - debugctlmsr = get_debugctlmsr(); - - debugctlmsr |= X86_DEBUGCTL_TR; - debugctlmsr |= X86_DEBUGCTL_BTS; - debugctlmsr |= X86_DEBUGCTL_BTINT; - - if (!(config & ARCH_PERFMON_EVENTSEL_OS)) - debugctlmsr |= X86_DEBUGCTL_BTS_OFF_OS; - - if (!(config & ARCH_PERFMON_EVENTSEL_USR)) - debugctlmsr |= X86_DEBUGCTL_BTS_OFF_USR; - - update_debugctlmsr(debugctlmsr); -} - -static void intel_pmu_disable_bts(void) -{ - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); - unsigned long debugctlmsr; - - if (!cpuc->ds) - return; - - debugctlmsr = get_debugctlmsr(); - - debugctlmsr &= - ~(X86_DEBUGCTL_TR | X86_DEBUGCTL_BTS | X86_DEBUGCTL_BTINT | - X86_DEBUGCTL_BTS_OFF_OS | X86_DEBUGCTL_BTS_OFF_USR); - - update_debugctlmsr(debugctlmsr); -} - static void intel_pmu_disable_all(void) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); @@ -514,6 +478,8 @@ static void intel_pmu_disable_all(void) if (test_bit(X86_PMC_IDX_FIXED_BTS, cpuc->active_mask)) intel_pmu_disable_bts(); + + intel_pmu_pebs_disable_all(); } static void intel_pmu_enable_all(void) @@ -531,6 +497,8 @@ static void intel_pmu_enable_all(void) intel_pmu_enable_bts(event->hw.config); } + + intel_pmu_pebs_enable_all(); } static inline u64 intel_pmu_get_status(void) @@ -547,8 +515,7 @@ static inline void intel_pmu_ack_status(u64 ack) wrmsrl(MSR_CORE_PERF_GLOBAL_OVF_CTRL, ack); } -static inline void -intel_pmu_disable_fixed(struct hw_perf_event *hwc) +static void intel_pmu_disable_fixed(struct hw_perf_event *hwc) { int idx = hwc->idx - X86_PMC_IDX_FIXED; u64 ctrl_val, mask; @@ -560,68 +527,7 @@ intel_pmu_disable_fixed(struct hw_perf_event *hwc) (void)checking_wrmsrl(hwc->config_base, ctrl_val); } -static void intel_pmu_drain_bts_buffer(void) -{ - struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); - struct debug_store *ds = cpuc->ds; - struct bts_record { - u64 from; - u64 to; - u64 flags; - }; - struct perf_event *event = cpuc->events[X86_PMC_IDX_FIXED_BTS]; - struct bts_record *at, *top; - struct perf_output_handle handle; - struct perf_event_header header; - struct perf_sample_data data; - struct pt_regs regs; - - if (!event) - return; - - if (!ds) - return; - - at = (struct bts_record *)(unsigned long)ds->bts_buffer_base; - top = (struct bts_record *)(unsigned long)ds->bts_index; - - if (top <= at) - return; - - ds->bts_index = ds->bts_buffer_base; - - perf_sample_data_init(&data, 0); - - data.period = event->hw.last_period; - regs.ip = 0; - - /* - * Prepare a generic sample, i.e. fill in the invariant fields. - * We will overwrite the from and to address before we output - * the sample. - */ - perf_prepare_sample(&header, &data, event, ®s); - - if (perf_output_begin(&handle, event, - header.size * (top - at), 1, 1)) - return; - - for (; at < top; at++) { - data.ip = at->from; - data.addr = at->to; - - perf_output_sample(&handle, &header, &data, event); - } - - perf_output_end(&handle); - - /* There's new data available. */ - event->hw.interrupts++; - event->pending_kill = POLL_IN; -} - -static inline void -intel_pmu_disable_event(struct perf_event *event) +static void intel_pmu_disable_event(struct perf_event *event) { struct hw_perf_event *hwc = &event->hw; @@ -637,10 +543,12 @@ intel_pmu_disable_event(struct perf_event *event) } x86_pmu_disable_event(event); + + if (unlikely(event->attr.precise)) + intel_pmu_pebs_disable(hwc); } -static inline void -intel_pmu_enable_fixed(struct hw_perf_event *hwc) +static void intel_pmu_enable_fixed(struct hw_perf_event *hwc) { int idx = hwc->idx - X86_PMC_IDX_FIXED; u64 ctrl_val, bits, mask; @@ -689,6 +597,9 @@ static void intel_pmu_enable_event(struct perf_event *event) return; } + if (unlikely(event->attr.precise)) + intel_pmu_pebs_enable(hwc); + __x86_pmu_enable_event(hwc); } @@ -762,6 +673,13 @@ again: inc_irq_stat(apic_perf_irqs); ack = status; + + /* + * PEBS overflow sets bit 62 in the global status register + */ + if (__test_and_clear_bit(62, (unsigned long *)&status)) + x86_pmu.drain_pebs(regs); + for_each_set_bit(bit, (unsigned long *)&status, X86_PMC_IDX_MAX) { struct perf_event *event = cpuc->events[bit]; @@ -791,22 +709,18 @@ done: return 1; } -static struct event_constraint bts_constraint = - EVENT_CONSTRAINT(0, 1ULL << X86_PMC_IDX_FIXED_BTS, 0); - static struct event_constraint * -intel_special_constraints(struct perf_event *event) +intel_bts_constraints(struct perf_event *event) { - unsigned int hw_event; - - hw_event = event->hw.config & INTEL_ARCH_EVENT_MASK; + struct hw_perf_event *hwc = &event->hw; + unsigned int hw_event, bts_event; - if (unlikely((hw_event == - x86_pmu.event_map(PERF_COUNT_HW_BRANCH_INSTRUCTIONS)) && - (event->hw.sample_period == 1))) { + hw_event = hwc->config & INTEL_ARCH_EVENT_MASK; + bts_event = x86_pmu.event_map(PERF_COUNT_HW_BRANCH_INSTRUCTIONS); + if (unlikely(hw_event == bts_event && hwc->sample_period == 1)) return &bts_constraint; - } + return NULL; } @@ -815,7 +729,11 @@ intel_get_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event { struct event_constraint *c; - c = intel_special_constraints(event); + c = intel_bts_constraints(event); + if (c) + return c; + + c = intel_pebs_constraints(event); if (c) return c; @@ -864,8 +782,6 @@ static __initconst struct x86_pmu intel_pmu = { * the generic event period: */ .max_period = (1ULL << 31) - 1, - .enable_bts = intel_pmu_enable_bts, - .disable_bts = intel_pmu_disable_bts, .get_event_constraints = intel_get_event_constraints, .cpu_starting = init_debug_store_on_cpu, @@ -915,6 +831,8 @@ static __init int intel_pmu_init(void) if (version > 1) x86_pmu.num_events_fixed = max((int)edx.split.num_events_fixed, 3); + intel_ds_init(); + /* * Install the hw-cache-events table: */ diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c new file mode 100644 index 0000000..0d994ef --- /dev/null +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -0,0 +1,557 @@ +#ifdef CONFIG_CPU_SUP_INTEL + +/* The maximal number of PEBS events: */ +#define MAX_PEBS_EVENTS 4 + +/* The size of a BTS record in bytes: */ +#define BTS_RECORD_SIZE 24 + +#define BTS_BUFFER_SIZE (PAGE_SIZE << 4) +#define PEBS_BUFFER_SIZE PAGE_SIZE + +/* + * pebs_record_32 for p4 and core not supported + +struct pebs_record_32 { + u32 flags, ip; + u32 ax, bc, cx, dx; + u32 si, di, bp, sp; +}; + + */ + +struct pebs_record_core { + u64 flags, ip; + u64 ax, bx, cx, dx; + u64 si, di, bp, sp; + u64 r8, r9, r10, r11; + u64 r12, r13, r14, r15; +}; + +struct pebs_record_nhm { + u64 flags, ip; + u64 ax, bx, cx, dx; + u64 si, di, bp, sp; + u64 r8, r9, r10, r11; + u64 r12, r13, r14, r15; + u64 status, dla, dse, lat; +}; + +/* + * Bits in the debugctlmsr controlling branch tracing. + */ +#define X86_DEBUGCTL_TR (1 << 6) +#define X86_DEBUGCTL_BTS (1 << 7) +#define X86_DEBUGCTL_BTINT (1 << 8) +#define X86_DEBUGCTL_BTS_OFF_OS (1 << 9) +#define X86_DEBUGCTL_BTS_OFF_USR (1 << 10) + +/* + * A debug store configuration. + * + * We only support architectures that use 64bit fields. + */ +struct debug_store { + u64 bts_buffer_base; + u64 bts_index; + u64 bts_absolute_maximum; + u64 bts_interrupt_threshold; + u64 pebs_buffer_base; + u64 pebs_index; + u64 pebs_absolute_maximum; + u64 pebs_interrupt_threshold; + u64 pebs_event_reset[MAX_PEBS_EVENTS]; +}; + +static void init_debug_store_on_cpu(int cpu) +{ + struct debug_store *ds = per_cpu(cpu_hw_events, cpu).ds; + + if (!ds) + return; + + wrmsr_on_cpu(cpu, MSR_IA32_DS_AREA, + (u32)((u64)(unsigned long)ds), + (u32)((u64)(unsigned long)ds >> 32)); +} + +static void fini_debug_store_on_cpu(int cpu) +{ + if (!per_cpu(cpu_hw_events, cpu).ds) + return; + + wrmsr_on_cpu(cpu, MSR_IA32_DS_AREA, 0, 0); +} + +static void release_ds_buffers(void) +{ + int cpu; + + if (!x86_pmu.bts && !x86_pmu.pebs) + return; + + get_online_cpus(); + + for_each_online_cpu(cpu) + fini_debug_store_on_cpu(cpu); + + for_each_possible_cpu(cpu) { + struct debug_store *ds = per_cpu(cpu_hw_events, cpu).ds; + + if (!ds) + continue; + + per_cpu(cpu_hw_events, cpu).ds = NULL; + + kfree((void *)(unsigned long)ds->pebs_buffer_base); + kfree((void *)(unsigned long)ds->bts_buffer_base); + kfree(ds); + } + + put_online_cpus(); +} + +static int reserve_ds_buffers(void) +{ + int cpu, err = 0; + + if (!x86_pmu.bts && !x86_pmu.pebs) + return 0; + + get_online_cpus(); + + for_each_possible_cpu(cpu) { + struct debug_store *ds; + void *buffer; + int max, thresh; + + err = -ENOMEM; + ds = kzalloc(sizeof(*ds), GFP_KERNEL); + if (unlikely(!ds)) { + kfree(buffer); + break; + } + per_cpu(cpu_hw_events, cpu).ds = ds; + + if (x86_pmu.bts) { + buffer = kzalloc(BTS_BUFFER_SIZE, GFP_KERNEL); + if (unlikely(!buffer)) + break; + + max = BTS_BUFFER_SIZE / BTS_RECORD_SIZE; + thresh = max / 16; + + ds->bts_buffer_base = (u64)(unsigned long)buffer; + ds->bts_index = ds->bts_buffer_base; + ds->bts_absolute_maximum = ds->bts_buffer_base + + max * BTS_RECORD_SIZE; + ds->bts_interrupt_threshold = ds->bts_absolute_maximum - + thresh * BTS_RECORD_SIZE; + } + + if (x86_pmu.pebs) { + buffer = kzalloc(PEBS_BUFFER_SIZE, GFP_KERNEL); + if (unlikely(!buffer)) + break; + + max = PEBS_BUFFER_SIZE / x86_pmu.pebs_record_size; + + ds->pebs_buffer_base = (u64)(unsigned long)buffer; + ds->pebs_index = ds->pebs_buffer_base; + ds->pebs_absolute_maximum = ds->pebs_buffer_base + + max * x86_pmu.pebs_record_size; + /* + * Always use single record PEBS + */ + ds->pebs_interrupt_threshold = ds->pebs_buffer_base + + x86_pmu.pebs_record_size; + } + + err = 0; + } + + if (err) + release_ds_buffers(); + else { + for_each_online_cpu(cpu) + init_debug_store_on_cpu(cpu); + } + + put_online_cpus(); + + return err; +} + +/* + * BTS + */ + +static struct event_constraint bts_constraint = + EVENT_CONSTRAINT(0, 1ULL << X86_PMC_IDX_FIXED_BTS, 0); + +static void intel_pmu_enable_bts(u64 config) +{ + unsigned long debugctlmsr; + + debugctlmsr = get_debugctlmsr(); + + debugctlmsr |= X86_DEBUGCTL_TR; + debugctlmsr |= X86_DEBUGCTL_BTS; + debugctlmsr |= X86_DEBUGCTL_BTINT; + + if (!(config & ARCH_PERFMON_EVENTSEL_OS)) + debugctlmsr |= X86_DEBUGCTL_BTS_OFF_OS; + + if (!(config & ARCH_PERFMON_EVENTSEL_USR)) + debugctlmsr |= X86_DEBUGCTL_BTS_OFF_USR; + + update_debugctlmsr(debugctlmsr); +} + +static void intel_pmu_disable_bts(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + unsigned long debugctlmsr; + + if (!cpuc->ds) + return; + + debugctlmsr = get_debugctlmsr(); + + debugctlmsr &= + ~(X86_DEBUGCTL_TR | X86_DEBUGCTL_BTS | X86_DEBUGCTL_BTINT | + X86_DEBUGCTL_BTS_OFF_OS | X86_DEBUGCTL_BTS_OFF_USR); + + update_debugctlmsr(debugctlmsr); +} + +static void intel_pmu_drain_bts_buffer(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct debug_store *ds = cpuc->ds; + struct bts_record { + u64 from; + u64 to; + u64 flags; + }; + struct perf_event *event = cpuc->events[X86_PMC_IDX_FIXED_BTS]; + struct bts_record *at, *top; + struct perf_output_handle handle; + struct perf_event_header header; + struct perf_sample_data data; + struct pt_regs regs; + + if (!event) + return; + + if (!ds) + return; + + at = (struct bts_record *)(unsigned long)ds->bts_buffer_base; + top = (struct bts_record *)(unsigned long)ds->bts_index; + + if (top <= at) + return; + + ds->bts_index = ds->bts_buffer_base; + + perf_sample_data_init(&data, 0); + data.period = event->hw.last_period; + regs.ip = 0; + + /* + * Prepare a generic sample, i.e. fill in the invariant fields. + * We will overwrite the from and to address before we output + * the sample. + */ + perf_prepare_sample(&header, &data, event, ®s); + + if (perf_output_begin(&handle, event, header.size * (top - at), 1, 1)) + return; + + for (; at < top; at++) { + data.ip = at->from; + data.addr = at->to; + + perf_output_sample(&handle, &header, &data, event); + } + + perf_output_end(&handle); + + /* There's new data available. */ + event->hw.interrupts++; + event->pending_kill = POLL_IN; +} + +/* + * PEBS + */ + +static struct event_constraint intel_core_pebs_events[] = { + PEBS_EVENT_CONSTRAINT(0x00c0, 0x1), /* INSTR_RETIRED.ANY */ + PEBS_EVENT_CONSTRAINT(0xfec1, 0x1), /* X87_OPS_RETIRED.ANY */ + PEBS_EVENT_CONSTRAINT(0x00c5, 0x1), /* BR_INST_RETIRED.MISPRED */ + PEBS_EVENT_CONSTRAINT(0x1fc7, 0x1), /* SIMD_INST_RETURED.ANY */ + PEBS_EVENT_CONSTRAINT(0x01cb, 0x1), /* MEM_LOAD_RETIRED.L1D_MISS */ + PEBS_EVENT_CONSTRAINT(0x02cb, 0x1), /* MEM_LOAD_RETIRED.L1D_LINE_MISS */ + PEBS_EVENT_CONSTRAINT(0x04cb, 0x1), /* MEM_LOAD_RETIRED.L2_MISS */ + PEBS_EVENT_CONSTRAINT(0x08cb, 0x1), /* MEM_LOAD_RETIRED.L2_LINE_MISS */ + PEBS_EVENT_CONSTRAINT(0x10cb, 0x1), /* MEM_LOAD_RETIRED.DTLB_MISS */ + EVENT_CONSTRAINT_END +}; + +static struct event_constraint intel_nehalem_pebs_events[] = { + PEBS_EVENT_CONSTRAINT(0x00c0, 0xf), /* INSTR_RETIRED.ANY */ + PEBS_EVENT_CONSTRAINT(0xfec1, 0xf), /* X87_OPS_RETIRED.ANY */ + PEBS_EVENT_CONSTRAINT(0x00c5, 0xf), /* BR_INST_RETIRED.MISPRED */ + PEBS_EVENT_CONSTRAINT(0x1fc7, 0xf), /* SIMD_INST_RETURED.ANY */ + PEBS_EVENT_CONSTRAINT(0x01cb, 0xf), /* MEM_LOAD_RETIRED.L1D_MISS */ + PEBS_EVENT_CONSTRAINT(0x02cb, 0xf), /* MEM_LOAD_RETIRED.L1D_LINE_MISS */ + PEBS_EVENT_CONSTRAINT(0x04cb, 0xf), /* MEM_LOAD_RETIRED.L2_MISS */ + PEBS_EVENT_CONSTRAINT(0x08cb, 0xf), /* MEM_LOAD_RETIRED.L2_LINE_MISS */ + PEBS_EVENT_CONSTRAINT(0x10cb, 0xf), /* MEM_LOAD_RETIRED.DTLB_MISS */ + EVENT_CONSTRAINT_END +}; + +static struct event_constraint * +intel_pebs_constraints(struct perf_event *event) +{ + struct event_constraint *c; + + if (!event->attr.precise) + return NULL; + + if (x86_pmu.pebs_constraints) { + for_each_event_constraint(c, x86_pmu.pebs_constraints) { + if ((event->hw.config & c->cmask) == c->code) + return c; + } + } + + return &emptyconstraint; +} + +static void intel_pmu_pebs_enable(struct hw_perf_event *hwc) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + u64 val = cpuc->pebs_enabled; + + hwc->config &= ~ARCH_PERFMON_EVENTSEL_INT; + + val |= 1ULL << hwc->idx; + wrmsrl(MSR_IA32_PEBS_ENABLE, val); +} + +static void intel_pmu_pebs_disable(struct hw_perf_event *hwc) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + u64 val = cpuc->pebs_enabled; + + val &= ~(1ULL << hwc->idx); + wrmsrl(MSR_IA32_PEBS_ENABLE, val); + + hwc->config |= ARCH_PERFMON_EVENTSEL_INT; +} + +static void intel_pmu_pebs_enable_all(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + + if (cpuc->pebs_enabled) + wrmsrl(MSR_IA32_PEBS_ENABLE, cpuc->pebs_enabled); +} + +static void intel_pmu_pebs_disable_all(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + + if (cpuc->pebs_enabled) + wrmsrl(MSR_IA32_PEBS_ENABLE, 0); +} + +static int intel_pmu_save_and_restart(struct perf_event *event); +static void intel_pmu_disable_event(struct perf_event *event); + +static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct debug_store *ds = cpuc->ds; + struct perf_event *event = cpuc->events[0]; /* PMC0 only */ + struct pebs_record_core *at, *top; + struct perf_sample_data data; + struct pt_regs regs; + int n; + + if (!event || !ds || !x86_pmu.pebs) + return; + + intel_pmu_pebs_disable_all(); + + at = (struct pebs_record_core *)(unsigned long)ds->pebs_buffer_base; + top = (struct pebs_record_core *)(unsigned long)ds->pebs_index; + + if (top <= at) + goto out; + + ds->pebs_index = ds->pebs_buffer_base; + + if (!intel_pmu_save_and_restart(event)) + goto out; + + perf_sample_data_init(&data, 0); + data.period = event->hw.last_period; + + n = top - at; + + /* + * Should not happen, we program the threshold at 1 and do not + * set a reset value. + */ + WARN_ON_ONCE(n > 1); + + /* + * We use the interrupt regs as a base because the PEBS record + * does not contain a full regs set, specifically it seems to + * lack segment descriptors, which get used by things like + * user_mode(). + * + * In the simple case fix up only the IP and BP,SP regs, for + * PERF_SAMPLE_IP and PERF_SAMPLE_CALLCHAIN to function properly. + * A possible PERF_SAMPLE_REGS will have to transfer all regs. + */ + regs = *iregs; + regs.ip = at->ip; + regs.bp = at->bp; + regs.sp = at->sp; + + if (perf_event_overflow(event, 1, &data, ®s)) + intel_pmu_disable_event(event); + +out: + intel_pmu_pebs_enable_all(); +} + +static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct debug_store *ds = cpuc->ds; + struct pebs_record_nhm *at, *top; + struct perf_sample_data data; + struct perf_event *event = NULL; + struct pt_regs regs; + int bit, n; + + if (!ds || !x86_pmu.pebs) + return; + + intel_pmu_pebs_disable_all(); + + at = (struct pebs_record_nhm *)(unsigned long)ds->pebs_buffer_base; + top = (struct pebs_record_nhm *)(unsigned long)ds->pebs_index; + + if (top <= at) + goto out; + + ds->pebs_index = ds->pebs_buffer_base; + + n = top - at; + + /* + * Should not happen, we program the threshold at 1 and do not + * set a reset value. + */ + WARN_ON_ONCE(n > MAX_PEBS_EVENTS); + + for ( ; at < top; at++) { + for_each_bit(bit, (unsigned long *)&at->status, MAX_PEBS_EVENTS) { + if (!cpuc->events[bit]->attr.precise) + continue; + + event = cpuc->events[bit]; + } + + if (!event) + continue; + + if (!intel_pmu_save_and_restart(event)) + continue; + + perf_sample_data_init(&data, 0); + data.period = event->hw.last_period; + + /* + * See the comment in intel_pmu_drain_pebs_core() + */ + regs = *iregs; + regs.ip = at->ip; + regs.bp = at->bp; + regs.sp = at->sp; + + if (perf_event_overflow(event, 1, &data, ®s)) + intel_pmu_disable_event(event); + } +out: + intel_pmu_pebs_enable_all(); +} + +/* + * BTS, PEBS probe and setup + */ + +static void intel_ds_init(void) +{ + /* + * No support for 32bit formats + */ + if (!boot_cpu_has(X86_FEATURE_DTES64)) + return; + + x86_pmu.bts = boot_cpu_has(X86_FEATURE_BTS); + x86_pmu.pebs = boot_cpu_has(X86_FEATURE_PEBS); + if (x86_pmu.pebs) { + int format = 0; + + if (x86_pmu.version > 1) { + u64 capabilities; + /* + * v2+ has a PEBS format field + */ + rdmsrl(MSR_IA32_PERF_CAPABILITIES, capabilities); + format = (capabilities >> 8) & 0xf; + } + + switch (format) { + case 0: + printk(KERN_CONT "PEBS v0, "); + x86_pmu.pebs_record_size = sizeof(struct pebs_record_core); + x86_pmu.drain_pebs = intel_pmu_drain_pebs_core; + x86_pmu.pebs_constraints = intel_core_pebs_events; + break; + + case 1: + printk(KERN_CONT "PEBS v1, "); + x86_pmu.pebs_record_size = sizeof(struct pebs_record_nhm); + x86_pmu.drain_pebs = intel_pmu_drain_pebs_nhm; + x86_pmu.pebs_constraints = intel_nehalem_pebs_events; + break; + + default: + printk(KERN_CONT "PEBS unknown format: %d, ", format); + x86_pmu.pebs = 0; + break; + } + } +} + +#else /* CONFIG_CPU_SUP_INTEL */ + +static int reseve_ds_buffers(void) +{ + return 0; +} + +static void release_ds_buffers(void) +{ +} + +#endif /* CONFIG_CPU_SUP_INTEL */ diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index 80acbf3..42307b5 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -203,8 +203,9 @@ struct perf_event_attr { enable_on_exec : 1, /* next exec enables */ task : 1, /* trace fork/exit */ watermark : 1, /* wakeup_watermark */ + precise : 1, /* OoO invariant counter */ - __reserved_1 : 49; + __reserved_1 : 48; union { __u32 wakeup_events; /* wakeup every n events */ -- cgit v0.10.2 From 69fef0d2e2c2c049ef4207a52e78b50d527bd85a Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 4 Mar 2010 13:57:24 +0100 Subject: perf: Add attr->precise support to raw event parsing Minimal userspace interface to the new 'precise' events flag. Can be used like "perf top -e r00c0p" which will use PEBS to sample retired instructions. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: <20100304140100.468665803@chello.nl> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c index 05d0c5c..a201445 100644 --- a/tools/perf/util/parse-events.c +++ b/tools/perf/util/parse-events.c @@ -656,6 +656,10 @@ parse_raw_event(const char **strp, struct perf_event_attr *attr) return EVT_FAILED; n = hex2u64(str + 1, &config); if (n > 0) { + if (str[n+1] == 'p') { + attr->precise = 1; + n++; + } *strp = str + n + 1; attr->type = PERF_TYPE_RAW; attr->config = config; -- cgit v0.10.2 From caff2befffe899e63df5cc760b7ed01cfd902685 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 3 Mar 2010 12:02:30 +0100 Subject: perf, x86: Implement simple LBR support Implement simple suport Intel Last-Branch-Record, it supports all hardware that implements FREEZE_LBRS_ON_PMI, but does not (yet) implement the LBR config register. The Intel LBR is a FIFO of From,To addresses describing the last few branches the hardware took. This patch does not add perf interface to the LBR, but merely provides an interface for internal use. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: <20100304140100.544191154@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 0c03d5c..1badff6 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -48,6 +48,8 @@ struct amd_nb { struct event_constraint event_constraints[X86_PMC_IDX_MAX]; }; +#define MAX_LBR_ENTRIES 16 + struct cpu_hw_events { /* * Generic x86 PMC bits @@ -70,6 +72,14 @@ struct cpu_hw_events { u64 pebs_enabled; /* + * Intel LBR bits + */ + int lbr_users; + void *lbr_context; + struct perf_branch_stack lbr_stack; + struct perf_branch_entry lbr_entries[MAX_LBR_ENTRIES]; + + /* * AMD specific bits */ struct amd_nb *amd_nb; @@ -159,6 +169,13 @@ struct x86_pmu { int pebs_record_size; void (*drain_pebs)(struct pt_regs *regs); struct event_constraint *pebs_constraints; + + /* + * Intel LBR + */ + unsigned long lbr_tos, lbr_from, lbr_to; /* MSR base regs */ + int lbr_nr; /* hardware stack size */ + int lbr_format; /* hardware format */ }; static struct x86_pmu x86_pmu __read_mostly; @@ -1237,6 +1254,7 @@ undo: #include "perf_event_amd.c" #include "perf_event_p6.c" +#include "perf_event_intel_lbr.c" #include "perf_event_intel_ds.c" #include "perf_event_intel.c" diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 1144641..44f6ed4 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -480,6 +480,7 @@ static void intel_pmu_disable_all(void) intel_pmu_disable_bts(); intel_pmu_pebs_disable_all(); + intel_pmu_lbr_disable_all(); } static void intel_pmu_enable_all(void) @@ -499,6 +500,7 @@ static void intel_pmu_enable_all(void) } intel_pmu_pebs_enable_all(); + intel_pmu_lbr_enable_all(); } static inline u64 intel_pmu_get_status(void) @@ -674,6 +676,8 @@ again: inc_irq_stat(apic_perf_irqs); ack = status; + intel_pmu_lbr_read(); + /* * PEBS overflow sets bit 62 in the global status register */ @@ -848,6 +852,8 @@ static __init int intel_pmu_init(void) memcpy(hw_cache_event_ids, core2_hw_cache_event_ids, sizeof(hw_cache_event_ids)); + intel_pmu_lbr_init_core(); + x86_pmu.event_constraints = intel_core2_event_constraints; pr_cont("Core2 events, "); break; @@ -857,13 +863,18 @@ static __init int intel_pmu_init(void) memcpy(hw_cache_event_ids, nehalem_hw_cache_event_ids, sizeof(hw_cache_event_ids)); + intel_pmu_lbr_init_nhm(); + x86_pmu.event_constraints = intel_nehalem_event_constraints; pr_cont("Nehalem/Corei7 events, "); break; + case 28: /* Atom */ memcpy(hw_cache_event_ids, atom_hw_cache_event_ids, sizeof(hw_cache_event_ids)); + intel_pmu_lbr_init_atom(); + x86_pmu.event_constraints = intel_gen_event_constraints; pr_cont("Atom events, "); break; @@ -873,6 +884,8 @@ static __init int intel_pmu_init(void) memcpy(hw_cache_event_ids, westmere_hw_cache_event_ids, sizeof(hw_cache_event_ids)); + intel_pmu_lbr_init_nhm(); + x86_pmu.event_constraints = intel_westmere_event_constraints; pr_cont("Westmere events, "); break; diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c new file mode 100644 index 0000000..ea3e99e --- /dev/null +++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c @@ -0,0 +1,228 @@ +#ifdef CONFIG_CPU_SUP_INTEL + +enum { + LBR_FORMAT_32 = 0x00, + LBR_FORMAT_LIP = 0x01, + LBR_FORMAT_EIP = 0x02, + LBR_FORMAT_EIP_FLAGS = 0x03, +}; + +/* + * We only support LBR implementations that have FREEZE_LBRS_ON_PMI + * otherwise it becomes near impossible to get a reliable stack. + */ + +#define X86_DEBUGCTL_LBR (1 << 0) +#define X86_DEBUGCTL_FREEZE_LBRS_ON_PMI (1 << 11) + +static void __intel_pmu_lbr_enable(void) +{ + u64 debugctl; + + rdmsrl(MSR_IA32_DEBUGCTLMSR, debugctl); + debugctl |= (X86_DEBUGCTL_LBR | X86_DEBUGCTL_FREEZE_LBRS_ON_PMI); + wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctl); +} + +static void __intel_pmu_lbr_disable(void) +{ + u64 debugctl; + + rdmsrl(MSR_IA32_DEBUGCTLMSR, debugctl); + debugctl &= ~(X86_DEBUGCTL_LBR | X86_DEBUGCTL_FREEZE_LBRS_ON_PMI); + wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctl); +} + +static void intel_pmu_lbr_reset_32(void) +{ + int i; + + for (i = 0; i < x86_pmu.lbr_nr; i++) + wrmsrl(x86_pmu.lbr_from + i, 0); +} + +static void intel_pmu_lbr_reset_64(void) +{ + int i; + + for (i = 0; i < x86_pmu.lbr_nr; i++) { + wrmsrl(x86_pmu.lbr_from + i, 0); + wrmsrl(x86_pmu.lbr_to + i, 0); + } +} + +static void intel_pmu_lbr_reset(void) +{ + if (x86_pmu.lbr_format == LBR_FORMAT_32) + intel_pmu_lbr_reset_32(); + else + intel_pmu_lbr_reset_64(); +} + +static void intel_pmu_lbr_enable(struct perf_event *event) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + + if (!x86_pmu.lbr_nr) + return; + + WARN_ON(cpuc->enabled); + + /* + * Reset the LBR stack if this is the first LBR user or + * we changed task context so as to avoid data leaks. + */ + + if (!cpuc->lbr_users || + (event->ctx->task && cpuc->lbr_context != event->ctx)) { + intel_pmu_lbr_reset(); + cpuc->lbr_context = event->ctx; + } + + cpuc->lbr_users++; +} + +static void intel_pmu_lbr_disable(struct perf_event *event) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + + if (!x86_pmu.lbr_nr) + return; + + cpuc->lbr_users--; + + BUG_ON(cpuc->lbr_users < 0); + WARN_ON(cpuc->enabled); +} + +static void intel_pmu_lbr_enable_all(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + + if (cpuc->lbr_users) + __intel_pmu_lbr_enable(); +} + +static void intel_pmu_lbr_disable_all(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + + if (cpuc->lbr_users) + __intel_pmu_lbr_disable(); +} + +static inline u64 intel_pmu_lbr_tos(void) +{ + u64 tos; + + rdmsrl(x86_pmu.lbr_tos, tos); + + return tos; +} + +static void intel_pmu_lbr_read_32(struct cpu_hw_events *cpuc) +{ + unsigned long mask = x86_pmu.lbr_nr - 1; + u64 tos = intel_pmu_lbr_tos(); + int i; + + for (i = 0; i < x86_pmu.lbr_nr; i++, tos--) { + unsigned long lbr_idx = (tos - i) & mask; + union { + struct { + u32 from; + u32 to; + }; + u64 lbr; + } msr_lastbranch; + + rdmsrl(x86_pmu.lbr_from + lbr_idx, msr_lastbranch.lbr); + + cpuc->lbr_entries[i].from = msr_lastbranch.from; + cpuc->lbr_entries[i].to = msr_lastbranch.to; + cpuc->lbr_entries[i].flags = 0; + } + cpuc->lbr_stack.nr = i; +} + +#define LBR_FROM_FLAG_MISPRED (1ULL << 63) + +/* + * Due to lack of segmentation in Linux the effective address (offset) + * is the same as the linear address, allowing us to merge the LIP and EIP + * LBR formats. + */ +static void intel_pmu_lbr_read_64(struct cpu_hw_events *cpuc) +{ + unsigned long mask = x86_pmu.lbr_nr - 1; + u64 tos = intel_pmu_lbr_tos(); + int i; + + for (i = 0; i < x86_pmu.lbr_nr; i++, tos--) { + unsigned long lbr_idx = (tos - i) & mask; + u64 from, to, flags = 0; + + rdmsrl(x86_pmu.lbr_from + lbr_idx, from); + rdmsrl(x86_pmu.lbr_to + lbr_idx, to); + + if (x86_pmu.lbr_format == LBR_FORMAT_EIP_FLAGS) { + flags = !!(from & LBR_FROM_FLAG_MISPRED); + from = (u64)((((s64)from) << 1) >> 1); + } + + cpuc->lbr_entries[i].from = from; + cpuc->lbr_entries[i].to = to; + cpuc->lbr_entries[i].flags = flags; + } + cpuc->lbr_stack.nr = i; +} + +static void intel_pmu_lbr_read(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + + if (!cpuc->lbr_users) + return; + + if (x86_pmu.lbr_format == LBR_FORMAT_32) + intel_pmu_lbr_read_32(cpuc); + else + intel_pmu_lbr_read_64(cpuc); +} + +static int intel_pmu_lbr_format(void) +{ + u64 capabilities; + + rdmsrl(MSR_IA32_PERF_CAPABILITIES, capabilities); + return capabilities & 0x1f; +} + +static void intel_pmu_lbr_init_core(void) +{ + x86_pmu.lbr_format = intel_pmu_lbr_format(); + x86_pmu.lbr_nr = 4; + x86_pmu.lbr_tos = 0x01c9; + x86_pmu.lbr_from = 0x40; + x86_pmu.lbr_to = 0x60; +} + +static void intel_pmu_lbr_init_nhm(void) +{ + x86_pmu.lbr_format = intel_pmu_lbr_format(); + x86_pmu.lbr_nr = 16; + x86_pmu.lbr_tos = 0x01c9; + x86_pmu.lbr_from = 0x680; + x86_pmu.lbr_to = 0x6c0; +} + +static void intel_pmu_lbr_init_atom(void) +{ + x86_pmu.lbr_format = intel_pmu_lbr_format(); + x86_pmu.lbr_nr = 8; + x86_pmu.lbr_tos = 0x01c9; + x86_pmu.lbr_from = 0x40; + x86_pmu.lbr_to = 0x60; +} + +#endif /* CONFIG_CPU_SUP_INTEL */ diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index 42307b5..ab4fd9e 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -467,6 +467,17 @@ struct perf_raw_record { void *data; }; +struct perf_branch_entry { + __u64 from; + __u64 to; + __u64 flags; +}; + +struct perf_branch_stack { + __u64 nr; + struct perf_branch_entry entries[0]; +}; + struct task_struct; /** -- cgit v0.10.2 From ef21f683a045a79b6aa86ad81e5fdfc0d5ddd250 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 3 Mar 2010 13:12:23 +0100 Subject: perf, x86: use LBR for PEBS IP+1 fixup Use the LBR to fix up the PEBS IP+1 issue. As said, PEBS reports the next instruction, here we use the LBR to find the last branch and from that construct the actual IP. If the IP matches the LBR-TO, we use LBR-FROM, otherwise we use the LBR-TO address as the beginning of the last basic block and decode forward. Once we find a match to the current IP, we use the previous location. This patch introduces a new ABI element: PERF_RECORD_MISC_EXACT, which conveys that the reported IP (PERF_SAMPLE_IP) is the exact instruction that caused the event (barring CPU errata). The fixup can fail due to various reasons: 1) LBR contains invalid data (quite possible) 2) part of the basic block got paged out 3) the reported IP isn't part of the basic block (see 1) Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Masami Hiramatsu Cc: "Zhang, Yanmin" Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: <20100304140100.619375431@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/perf_event.h b/arch/x86/include/asm/perf_event.h index db6109a..a9038c9 100644 --- a/arch/x86/include/asm/perf_event.h +++ b/arch/x86/include/asm/perf_event.h @@ -136,6 +136,25 @@ extern void perf_events_lapic_init(void); #define PERF_EVENT_INDEX_OFFSET 0 +/* + * Abuse bit 3 of the cpu eflags register to indicate proper PEBS IP fixups. + * This flag is otherwise unused and ABI specified to be 0, so nobody should + * care what we do with it. + */ +#define PERF_EFLAGS_EXACT (1UL << 3) + +#define perf_misc_flags(regs) \ +({ int misc = 0; \ + if (user_mode(regs)) \ + misc |= PERF_RECORD_MISC_USER; \ + else \ + misc |= PERF_RECORD_MISC_KERNEL; \ + if (regs->flags & PERF_EFLAGS_EXACT) \ + misc |= PERF_RECORD_MISC_EXACT; \ + misc; }) + +#define perf_instruction_pointer(regs) ((regs)->ip) + #else static inline void init_hw_perf_events(void) { } static inline void perf_events_lapic_init(void) { } diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 1badff6..5cb4e8d 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -29,6 +29,41 @@ #include #include +/* + * best effort, GUP based copy_from_user() that assumes IRQ or NMI context + */ +static unsigned long +copy_from_user_nmi(void *to, const void __user *from, unsigned long n) +{ + unsigned long offset, addr = (unsigned long)from; + int type = in_nmi() ? KM_NMI : KM_IRQ0; + unsigned long size, len = 0; + struct page *page; + void *map; + int ret; + + do { + ret = __get_user_pages_fast(addr, 1, 0, &page); + if (!ret) + break; + + offset = addr & (PAGE_SIZE - 1); + size = min(PAGE_SIZE - offset, n - len); + + map = kmap_atomic(page, type); + memcpy(to, map+offset, size); + kunmap_atomic(map, type); + put_page(page); + + len += size; + to += size; + addr += size; + + } while (len < n); + + return len; +} + static u64 perf_event_mask __read_mostly; struct event_constraint { @@ -1550,41 +1585,6 @@ perf_callchain_kernel(struct pt_regs *regs, struct perf_callchain_entry *entry) dump_trace(NULL, regs, NULL, regs->bp, &backtrace_ops, entry); } -/* - * best effort, GUP based copy_from_user() that assumes IRQ or NMI context - */ -static unsigned long -copy_from_user_nmi(void *to, const void __user *from, unsigned long n) -{ - unsigned long offset, addr = (unsigned long)from; - int type = in_nmi() ? KM_NMI : KM_IRQ0; - unsigned long size, len = 0; - struct page *page; - void *map; - int ret; - - do { - ret = __get_user_pages_fast(addr, 1, 0, &page); - if (!ret) - break; - - offset = addr & (PAGE_SIZE - 1); - size = min(PAGE_SIZE - offset, n - len); - - map = kmap_atomic(page, type); - memcpy(to, map+offset, size); - kunmap_atomic(map, type); - put_page(page); - - len += size; - to += size; - addr += size; - - } while (len < n); - - return len; -} - static int copy_stack_frame(const void __user *fp, struct stack_frame *frame) { unsigned long bytes; diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 44f6ed4..7eb78be 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -547,7 +547,7 @@ static void intel_pmu_disable_event(struct perf_event *event) x86_pmu_disable_event(event); if (unlikely(event->attr.precise)) - intel_pmu_pebs_disable(hwc); + intel_pmu_pebs_disable(event); } static void intel_pmu_enable_fixed(struct hw_perf_event *hwc) @@ -600,7 +600,7 @@ static void intel_pmu_enable_event(struct perf_event *event) } if (unlikely(event->attr.precise)) - intel_pmu_pebs_enable(hwc); + intel_pmu_pebs_enable(event); __x86_pmu_enable_event(hwc); } diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index 0d994ef..50e6ff3 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -331,26 +331,32 @@ intel_pebs_constraints(struct perf_event *event) return &emptyconstraint; } -static void intel_pmu_pebs_enable(struct hw_perf_event *hwc) +static void intel_pmu_pebs_enable(struct perf_event *event) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct hw_perf_event *hwc = &event->hw; u64 val = cpuc->pebs_enabled; hwc->config &= ~ARCH_PERFMON_EVENTSEL_INT; val |= 1ULL << hwc->idx; wrmsrl(MSR_IA32_PEBS_ENABLE, val); + + intel_pmu_lbr_enable(event); } -static void intel_pmu_pebs_disable(struct hw_perf_event *hwc) +static void intel_pmu_pebs_disable(struct perf_event *event) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + struct hw_perf_event *hwc = &event->hw; u64 val = cpuc->pebs_enabled; val &= ~(1ULL << hwc->idx); wrmsrl(MSR_IA32_PEBS_ENABLE, val); hwc->config |= ARCH_PERFMON_EVENTSEL_INT; + + intel_pmu_lbr_disable(event); } static void intel_pmu_pebs_enable_all(void) @@ -369,6 +375,70 @@ static void intel_pmu_pebs_disable_all(void) wrmsrl(MSR_IA32_PEBS_ENABLE, 0); } +#include + +#define MAX_INSN_SIZE 16 + +static inline bool kernel_ip(unsigned long ip) +{ +#ifdef CONFIG_X86_32 + return ip > PAGE_OFFSET; +#else + return (long)ip < 0; +#endif +} + +static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + unsigned long from = cpuc->lbr_entries[0].from; + unsigned long old_to, to = cpuc->lbr_entries[0].to; + unsigned long ip = regs->ip; + + if (!cpuc->lbr_stack.nr || !from || !to) + return 0; + + if (ip < to) + return 0; + + /* + * We sampled a branch insn, rewind using the LBR stack + */ + if (ip == to) { + regs->ip = from; + return 1; + } + + do { + struct insn insn; + u8 buf[MAX_INSN_SIZE]; + void *kaddr; + + old_to = to; + if (!kernel_ip(ip)) { + int bytes, size = min_t(int, MAX_INSN_SIZE, ip - to); + + bytes = copy_from_user_nmi(buf, (void __user *)to, size); + if (bytes != size) + return 0; + + kaddr = buf; + } else + kaddr = (void *)to; + + kernel_insn_init(&insn, kaddr); + insn_get_length(&insn); + to += insn.length; + } while (to < ip); + + if (to == ip) { + regs->ip = old_to; + return 1; + } + + return 0; +} + static int intel_pmu_save_and_restart(struct perf_event *event); static void intel_pmu_disable_event(struct perf_event *event); @@ -424,6 +494,11 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) regs.bp = at->bp; regs.sp = at->sp; + if (intel_pmu_pebs_fixup_ip(®s)) + regs.flags |= PERF_EFLAGS_EXACT; + else + regs.flags &= ~PERF_EFLAGS_EXACT; + if (perf_event_overflow(event, 1, &data, ®s)) intel_pmu_disable_event(event); @@ -487,6 +562,11 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) regs.bp = at->bp; regs.sp = at->sp; + if (intel_pmu_pebs_fixup_ip(®s)) + regs.flags |= PERF_EFLAGS_EXACT; + else + regs.flags &= ~PERF_EFLAGS_EXACT; + if (perf_event_overflow(event, 1, &data, ®s)) intel_pmu_disable_event(event); } diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h index ab4fd9e..be85f7c 100644 --- a/include/linux/perf_event.h +++ b/include/linux/perf_event.h @@ -294,6 +294,12 @@ struct perf_event_mmap_page { #define PERF_RECORD_MISC_USER (2 << 0) #define PERF_RECORD_MISC_HYPERVISOR (3 << 0) +#define PERF_RECORD_MISC_EXACT (1 << 14) +/* + * Reserve the last bit to indicate some extended misc field + */ +#define PERF_RECORD_MISC_EXT_RESERVED (1 << 15) + struct perf_event_header { __u32 type; __u16 misc; -- cgit v0.10.2 From 1676b8a077c352085d52578fb4f29350b58b6e74 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 4 Mar 2010 14:19:36 +0100 Subject: perf-top: Show the percentage of successfull PEBS-fixups Use the PERF_RECORD_MISC_EXACT information to measure the success rate of the PEBS fix-up. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: <20100304140100.694233760@chello.nl> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index 31f2e59..c051833 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -410,6 +410,7 @@ static double sym_weight(const struct sym_entry *sym) static long samples; static long userspace_samples; +static long exact_samples; static const char CONSOLE_CLEAR[] = ""; static void __list_insert_active_sym(struct sym_entry *syme) @@ -450,6 +451,7 @@ static void print_sym_table(void) int counter, snap = !display_weighted ? sym_counter : 0; float samples_per_sec = samples/delay_secs; float ksamples_per_sec = (samples-userspace_samples)/delay_secs; + float esamples_percent = (100.0*exact_samples)/samples; float sum_ksamples = 0.0; struct sym_entry *syme, *n; struct rb_root tmp = RB_ROOT; @@ -457,7 +459,7 @@ static void print_sym_table(void) int sym_width = 0, dso_width = 0, max_dso_width; const int win_width = winsize.ws_col - 1; - samples = userspace_samples = 0; + samples = userspace_samples = exact_samples = 0; /* Sort the active symbols */ pthread_mutex_lock(&active_symbols_lock); @@ -488,9 +490,10 @@ static void print_sym_table(void) puts(CONSOLE_CLEAR); printf("%-*.*s\n", win_width, win_width, graph_dotted_line); - printf( " PerfTop:%8.0f irqs/sec kernel:%4.1f%% [", + printf( " PerfTop:%8.0f irqs/sec kernel:%4.1f%% exact: %4.1f%% [", samples_per_sec, - 100.0 - (100.0*((samples_per_sec-ksamples_per_sec)/samples_per_sec))); + 100.0 - (100.0*((samples_per_sec-ksamples_per_sec)/samples_per_sec)), + esamples_percent); if (nr_counters == 1 || !display_weighted) { printf("%Ld", (u64)attrs[0].sample_period); @@ -954,6 +957,9 @@ static void event__process_sample(const event_t *self, return; } + if (self->header.misc & PERF_RECORD_MISC_EXACT) + exact_samples++; + if (event__preprocess_sample(self, session, &al, symbol_filter) < 0 || al.filtered) return; -- cgit v0.10.2 From 8db909a7e3c888b5d45aef7650d74ccebe3ce725 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 3 Mar 2010 17:07:40 +0100 Subject: perf, x86: Clean up IA32_PERF_CAPABILITIES usage Saner PERF_CAPABILITIES support, which also exposes pebs_trap. Use that latter to make PEBS's use of LBR conditional since a fault-like pebs should already report the correct IP. ( As of this writing there is no known hardware that implements !pebs_trap ) Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: <20100304140100.770650663@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 5cb4e8d..7b5430b 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -154,6 +154,17 @@ struct cpu_hw_events { #define for_each_event_constraint(e, c) \ for ((e) = (c); (e)->cmask; (e)++) +union perf_capabilities { + struct { + u64 lbr_format : 6; + u64 pebs_trap : 1; + u64 pebs_arch_reg : 1; + u64 pebs_format : 4; + u64 smm_freeze : 1; + }; + u64 capabilities; +}; + /* * struct x86_pmu - generic x86 pmu */ @@ -195,7 +206,8 @@ struct x86_pmu { /* * Intel Arch Perfmon v2+ */ - u64 intel_ctrl; + u64 intel_ctrl; + union perf_capabilities intel_cap; /* * Intel DebugStore bits @@ -210,7 +222,6 @@ struct x86_pmu { */ unsigned long lbr_tos, lbr_from, lbr_to; /* MSR base regs */ int lbr_nr; /* hardware stack size */ - int lbr_format; /* hardware format */ }; static struct x86_pmu x86_pmu __read_mostly; diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 7eb78be..246c072 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -835,6 +835,16 @@ static __init int intel_pmu_init(void) if (version > 1) x86_pmu.num_events_fixed = max((int)edx.split.num_events_fixed, 3); + /* + * v2 and above have a perf capabilities MSR + */ + if (version > 1) { + u64 capabilities; + + rdmsrl(MSR_IA32_PERF_CAPABILITIES, capabilities); + x86_pmu.intel_cap.capabilities = capabilities; + } + intel_ds_init(); /* diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index 50e6ff3..5e40294 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -342,7 +342,8 @@ static void intel_pmu_pebs_enable(struct perf_event *event) val |= 1ULL << hwc->idx; wrmsrl(MSR_IA32_PEBS_ENABLE, val); - intel_pmu_lbr_enable(event); + if (x86_pmu.intel_cap.pebs_trap) + intel_pmu_lbr_enable(event); } static void intel_pmu_pebs_disable(struct perf_event *event) @@ -356,7 +357,8 @@ static void intel_pmu_pebs_disable(struct perf_event *event) hwc->config |= ARCH_PERFMON_EVENTSEL_INT; - intel_pmu_lbr_disable(event); + if (x86_pmu.intel_cap.pebs_trap) + intel_pmu_lbr_disable(event); } static void intel_pmu_pebs_enable_all(void) @@ -395,6 +397,12 @@ static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) unsigned long old_to, to = cpuc->lbr_entries[0].to; unsigned long ip = regs->ip; + /* + * We don't need to fixup if the PEBS assist is fault like + */ + if (!x86_pmu.intel_cap.pebs_trap) + return 1; + if (!cpuc->lbr_stack.nr || !from || !to) return 0; @@ -589,34 +597,26 @@ static void intel_ds_init(void) x86_pmu.bts = boot_cpu_has(X86_FEATURE_BTS); x86_pmu.pebs = boot_cpu_has(X86_FEATURE_PEBS); if (x86_pmu.pebs) { - int format = 0; - - if (x86_pmu.version > 1) { - u64 capabilities; - /* - * v2+ has a PEBS format field - */ - rdmsrl(MSR_IA32_PERF_CAPABILITIES, capabilities); - format = (capabilities >> 8) & 0xf; - } + char pebs_type = x86_pmu.intel_cap.pebs_trap ? '+' : '-'; + int format = x86_pmu.intel_cap.pebs_format; switch (format) { case 0: - printk(KERN_CONT "PEBS v0, "); + printk(KERN_CONT "PEBS fmt0%c, ", pebs_type); x86_pmu.pebs_record_size = sizeof(struct pebs_record_core); x86_pmu.drain_pebs = intel_pmu_drain_pebs_core; x86_pmu.pebs_constraints = intel_core_pebs_events; break; case 1: - printk(KERN_CONT "PEBS v1, "); + printk(KERN_CONT "PEBS fmt1%c, ", pebs_type); x86_pmu.pebs_record_size = sizeof(struct pebs_record_nhm); x86_pmu.drain_pebs = intel_pmu_drain_pebs_nhm; x86_pmu.pebs_constraints = intel_nehalem_pebs_events; break; default: - printk(KERN_CONT "PEBS unknown format: %d, ", format); + printk(KERN_CONT "no PEBS fmt%d%c, ", format, pebs_type); x86_pmu.pebs = 0; break; } diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c index ea3e99e..4f3a124 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c +++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c @@ -53,7 +53,7 @@ static void intel_pmu_lbr_reset_64(void) static void intel_pmu_lbr_reset(void) { - if (x86_pmu.lbr_format == LBR_FORMAT_32) + if (x86_pmu.intel_cap.lbr_format == LBR_FORMAT_32) intel_pmu_lbr_reset_32(); else intel_pmu_lbr_reset_64(); @@ -155,6 +155,7 @@ static void intel_pmu_lbr_read_32(struct cpu_hw_events *cpuc) static void intel_pmu_lbr_read_64(struct cpu_hw_events *cpuc) { unsigned long mask = x86_pmu.lbr_nr - 1; + int lbr_format = x86_pmu.intel_cap.lbr_format; u64 tos = intel_pmu_lbr_tos(); int i; @@ -165,7 +166,7 @@ static void intel_pmu_lbr_read_64(struct cpu_hw_events *cpuc) rdmsrl(x86_pmu.lbr_from + lbr_idx, from); rdmsrl(x86_pmu.lbr_to + lbr_idx, to); - if (x86_pmu.lbr_format == LBR_FORMAT_EIP_FLAGS) { + if (lbr_format == LBR_FORMAT_EIP_FLAGS) { flags = !!(from & LBR_FROM_FLAG_MISPRED); from = (u64)((((s64)from) << 1) >> 1); } @@ -184,23 +185,14 @@ static void intel_pmu_lbr_read(void) if (!cpuc->lbr_users) return; - if (x86_pmu.lbr_format == LBR_FORMAT_32) + if (x86_pmu.intel_cap.lbr_format == LBR_FORMAT_32) intel_pmu_lbr_read_32(cpuc); else intel_pmu_lbr_read_64(cpuc); } -static int intel_pmu_lbr_format(void) -{ - u64 capabilities; - - rdmsrl(MSR_IA32_PERF_CAPABILITIES, capabilities); - return capabilities & 0x1f; -} - static void intel_pmu_lbr_init_core(void) { - x86_pmu.lbr_format = intel_pmu_lbr_format(); x86_pmu.lbr_nr = 4; x86_pmu.lbr_tos = 0x01c9; x86_pmu.lbr_from = 0x40; @@ -209,7 +201,6 @@ static void intel_pmu_lbr_init_core(void) static void intel_pmu_lbr_init_nhm(void) { - x86_pmu.lbr_format = intel_pmu_lbr_format(); x86_pmu.lbr_nr = 16; x86_pmu.lbr_tos = 0x01c9; x86_pmu.lbr_from = 0x680; @@ -218,7 +209,6 @@ static void intel_pmu_lbr_init_nhm(void) static void intel_pmu_lbr_init_atom(void) { - x86_pmu.lbr_format = intel_pmu_lbr_format(); x86_pmu.lbr_nr = 8; x86_pmu.lbr_tos = 0x01c9; x86_pmu.lbr_from = 0x40; -- cgit v0.10.2 From 7e1a40dda619b0483fbe0740494ed2c2a1f05289 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 4 Mar 2010 12:38:03 +0100 Subject: perf, x86: Expose the full PEBS record using PERF_SAMPLE_RAW Expose the full PEBS record using PERF_SAMPLE_RAW Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: <20100304140100.847218224@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index 5e40294..ef56f05 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -457,6 +457,7 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) struct perf_event *event = cpuc->events[0]; /* PMC0 only */ struct pebs_record_core *at, *top; struct perf_sample_data data; + struct perf_raw_record raw; struct pt_regs regs; int n; @@ -479,6 +480,12 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) perf_sample_data_init(&data, 0); data.period = event->hw.last_period; + if (event->attr.sample_type & PERF_SAMPLE_RAW) { + raw.size = x86_pmu.pebs_record_size; + raw.data = at; + data.raw = &raw; + } + n = top - at; /* @@ -521,6 +528,7 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) struct pebs_record_nhm *at, *top; struct perf_sample_data data; struct perf_event *event = NULL; + struct perf_raw_record raw; struct pt_regs regs; int bit, n; @@ -562,6 +570,12 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) perf_sample_data_init(&data, 0); data.period = event->hw.last_period; + if (event->attr.sample_type & PERF_SAMPLE_RAW) { + raw.size = x86_pmu.pebs_record_size; + raw.data = at; + data.raw = &raw; + } + /* * See the comment in intel_pmu_drain_pebs_core() */ -- cgit v0.10.2 From 30a813ae035d3e220a89609adce878e045c49547 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 4 Mar 2010 13:49:21 +0100 Subject: x86: Move MAX_INSN_SIZE into asm/insn.h Since there's now two users for this, place it in a common header. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Masami Hiramatsu Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: <20100304140100.923774125@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/insn.h b/arch/x86/include/asm/insn.h index 96c2e0a..88c765e 100644 --- a/arch/x86/include/asm/insn.h +++ b/arch/x86/include/asm/insn.h @@ -68,6 +68,8 @@ struct insn { const insn_byte_t *next_byte; }; +#define MAX_INSN_SIZE 16 + #define X86_MODRM_MOD(modrm) (((modrm) & 0xc0) >> 6) #define X86_MODRM_REG(modrm) (((modrm) & 0x38) >> 3) #define X86_MODRM_RM(modrm) ((modrm) & 0x07) diff --git a/arch/x86/include/asm/kprobes.h b/arch/x86/include/asm/kprobes.h index 4ffa345..5478825 100644 --- a/arch/x86/include/asm/kprobes.h +++ b/arch/x86/include/asm/kprobes.h @@ -24,6 +24,7 @@ #include #include #include +#include #define __ARCH_WANT_KPROBES_INSN_SLOT @@ -36,7 +37,6 @@ typedef u8 kprobe_opcode_t; #define RELATIVEJUMP_SIZE 5 #define RELATIVECALL_OPCODE 0xe8 #define RELATIVE_ADDR_SIZE 4 -#define MAX_INSN_SIZE 16 #define MAX_STACK_SIZE 64 #define MIN_STACK_SIZE(ADDR) \ (((MAX_STACK_SIZE) < (((unsigned long)current_thread_info()) + \ diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index ef56f05..72453ac 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -379,8 +379,6 @@ static void intel_pmu_pebs_disable_all(void) #include -#define MAX_INSN_SIZE 16 - static inline bool kernel_ip(unsigned long ip) { #ifdef CONFIG_X86_32 -- cgit v0.10.2 From 3adaebd69557615c1bf0365ce5e32d93ac7d82af Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 5 Mar 2010 12:09:29 +0100 Subject: perf, x86: Fix silly bug in data store buffer allocation Fix up the ds allocation error path, where we could free @buffer before we used it. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: <20100305154128.813452402@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index 72453ac..a67fff1 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -127,10 +127,8 @@ static int reserve_ds_buffers(void) err = -ENOMEM; ds = kzalloc(sizeof(*ds), GFP_KERNEL); - if (unlikely(!ds)) { - kfree(buffer); + if (unlikely(!ds)) break; - } per_cpu(cpu_hw_events, cpu).ds = ds; if (x86_pmu.bts) { -- cgit v0.10.2 From 3c44780b220e876b01e39d4028cd6f4205fbf5d6 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 4 Mar 2010 21:49:01 +0100 Subject: perf, x86: Disable PEBS on clovertown chips This CPU has just too many handycaps to be really useful. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: <20100305154128.890278662@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 7b5430b..335ee1d 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -197,6 +197,7 @@ struct x86_pmu { void (*put_event_constraints)(struct cpu_hw_events *cpuc, struct perf_event *event); struct event_constraint *event_constraints; + void (*quirks)(void); void (*cpu_prepare)(int cpu); void (*cpu_starting)(int cpu); @@ -1373,6 +1374,9 @@ void __init init_hw_perf_events(void) pr_cont("%s PMU driver.\n", x86_pmu.name); + if (x86_pmu.quirks) + x86_pmu.quirks(); + if (x86_pmu.num_events > X86_PMC_MAX_GENERIC) { WARN(1, KERN_ERR "hw perf events %d > max(%d), clipping!", x86_pmu.num_events, X86_PMC_MAX_GENERIC); diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 246c072..224c952 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -792,6 +792,32 @@ static __initconst struct x86_pmu intel_pmu = { .cpu_dying = fini_debug_store_on_cpu, }; +static void intel_clovertown_quirks(void) +{ + /* + * PEBS is unreliable due to: + * + * AJ67 - PEBS may experience CPL leaks + * AJ68 - PEBS PMI may be delayed by one event + * AJ69 - GLOBAL_STATUS[62] will only be set when DEBUGCTL[12] + * AJ106 - FREEZE_LBRS_ON_PMI doesn't work in combination with PEBS + * + * AJ67 could be worked around by restricting the OS/USR flags. + * AJ69 could be worked around by setting PMU_FREEZE_ON_PMI. + * + * AJ106 could possibly be worked around by not allowing LBR + * usage from PEBS, including the fixup. + * AJ68 could possibly be worked around by always programming + * a pebs_event_reset[0] value and coping with the lost events. + * + * But taken together it might just make sense to not enable PEBS on + * these chips. + */ + printk(KERN_WARNING "PEBS disabled due to CPU errata.\n"); + x86_pmu.pebs = 0; + x86_pmu.pebs_constraints = NULL; +} + static __init int intel_pmu_init(void) { union cpuid10_edx edx; @@ -856,6 +882,7 @@ static __init int intel_pmu_init(void) break; case 15: /* original 65 nm celeron/pentium/core2/xeon, "Merom"/"Conroe" */ + x86_pmu.quirks = intel_clovertown_quirks; case 22: /* single-core 65 nm celeron/core2solo "Merom-L"/"Conroe-L" */ case 23: /* current 45 nm celeron/core2/xeon "Penryn"/"Wolfdale" */ case 29: /* six-core 45 nm xeon "Dunnington" */ -- cgit v0.10.2 From 74846d35b24b6efd61bb88a0a750b6bb257e6e78 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 5 Mar 2010 13:49:35 +0100 Subject: perf, x86: Clear the LBRs on init Some CPUs have errata where the LBR is not cleared on Power-On. So always clear the LBRs before use. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: <20100305154128.966563424@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 224c952..c135ed7 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -767,6 +767,20 @@ static __initconst struct x86_pmu core_pmu = { .event_constraints = intel_core_event_constraints, }; +static void intel_pmu_cpu_starting(int cpu) +{ + init_debug_store_on_cpu(cpu); + /* + * Deal with CPUs that don't clear their LBRs on power-up. + */ + intel_pmu_lbr_reset(); +} + +static void intel_pmu_cpu_dying(int cpu) +{ + fini_debug_store_on_cpu(cpu); +} + static __initconst struct x86_pmu intel_pmu = { .name = "Intel", .handle_irq = intel_pmu_handle_irq, @@ -788,8 +802,8 @@ static __initconst struct x86_pmu intel_pmu = { .max_period = (1ULL << 31) - 1, .get_event_constraints = intel_get_event_constraints, - .cpu_starting = init_debug_store_on_cpu, - .cpu_dying = fini_debug_store_on_cpu, + .cpu_starting = intel_pmu_cpu_starting, + .cpu_dying = intel_pmu_cpu_dying, }; static void intel_clovertown_quirks(void) diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c index 4f3a124..dcec765 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c +++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c @@ -53,6 +53,9 @@ static void intel_pmu_lbr_reset_64(void) static void intel_pmu_lbr_reset(void) { + if (!x86_pmu.lbr_nr) + return; + if (x86_pmu.intel_cap.lbr_format == LBR_FORMAT_32) intel_pmu_lbr_reset_32(); else -- cgit v0.10.2 From a562b1871f7f7d2f3a835c3c1e07fa58d473cfb7 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 5 Mar 2010 16:29:14 +0100 Subject: perf, x86: Robustify PEBS fixup It turns out the LBR is massively unreliable on certain CPUs, so code the fixup a little more defensive to avoid crashing the kernel. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: <20100305154129.042271287@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index a67fff1..e7ac517 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -399,10 +399,23 @@ static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) if (!x86_pmu.intel_cap.pebs_trap) return 1; + /* + * No LBR entry, no basic block, no rewinding + */ if (!cpuc->lbr_stack.nr || !from || !to) return 0; - if (ip < to) + /* + * Basic blocks should never cross user/kernel boundaries + */ + if (kernel_ip(ip) != kernel_ip(to)) + return 0; + + /* + * unsigned math, either ip is before the start (impossible) or + * the basic block is larger than 1 page (sanity) + */ + if ((ip - to) > PAGE_SIZE) return 0; /* @@ -420,7 +433,7 @@ static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) old_to = to; if (!kernel_ip(ip)) { - int bytes, size = min_t(int, MAX_INSN_SIZE, ip - to); + int bytes, size = MAX_INSN_SIZE; bytes = copy_from_user_nmi(buf, (void __user *)to, size); if (bytes != size) @@ -440,6 +453,10 @@ static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) return 1; } + /* + * Even though we decoded the basic block, the instruction stream + * never matched the given IP, either the TO or the IP got corrupted. + */ return 0; } -- cgit v0.10.2 From cc7f00820b2f3be656569c41158d9323e425bcfe Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 8 Mar 2010 17:51:33 +0100 Subject: perf, x86: Avoid double disable on throttle vs ioctl(PERF_IOC_DISABLE) Calling ioctl(PERF_EVENT_IOC_DISABLE) on a thottled counter would result in a double disable, cure this by using x86_pmu_{start,stop} for throttle/unthrottle and teach x86_pmu_stop() to check ->active_mask. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index e7ac517..a7401e4 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -461,7 +461,6 @@ static int intel_pmu_pebs_fixup_ip(struct pt_regs *regs) } static int intel_pmu_save_and_restart(struct perf_event *event); -static void intel_pmu_disable_event(struct perf_event *event); static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) { @@ -528,7 +527,7 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) regs.flags &= ~PERF_EFLAGS_EXACT; if (perf_event_overflow(event, 1, &data, ®s)) - intel_pmu_disable_event(event); + x86_pmu_stop(event); out: intel_pmu_pebs_enable_all(); @@ -603,7 +602,7 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) regs.flags &= ~PERF_EFLAGS_EXACT; if (perf_event_overflow(event, 1, &data, ®s)) - intel_pmu_disable_event(event); + x86_pmu_stop(event); } out: intel_pmu_pebs_enable_all(); -- cgit v0.10.2 From 8f4aebd2be9892bf8fb79a2d8576d3f3ee7f00f6 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Sat, 6 Mar 2010 13:26:11 +0100 Subject: perf, x86: Fix pebs drains I overlooked the perf_disable()/perf_enable() calls in intel_pmu_handle_irq(), (pointed out by Markus) so we should not explicitly disable_all/enable_all pebs counters in the drain functions, these are already disabled and enabling them early is confusing. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index a7401e4..66c6962 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -476,18 +476,16 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) if (!event || !ds || !x86_pmu.pebs) return; - intel_pmu_pebs_disable_all(); - at = (struct pebs_record_core *)(unsigned long)ds->pebs_buffer_base; top = (struct pebs_record_core *)(unsigned long)ds->pebs_index; if (top <= at) - goto out; + return; ds->pebs_index = ds->pebs_buffer_base; if (!intel_pmu_save_and_restart(event)) - goto out; + return; perf_sample_data_init(&data, 0); data.period = event->hw.last_period; @@ -528,9 +526,6 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) if (perf_event_overflow(event, 1, &data, ®s)) x86_pmu_stop(event); - -out: - intel_pmu_pebs_enable_all(); } static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) @@ -547,13 +542,11 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) if (!ds || !x86_pmu.pebs) return; - intel_pmu_pebs_disable_all(); - at = (struct pebs_record_nhm *)(unsigned long)ds->pebs_buffer_base; top = (struct pebs_record_nhm *)(unsigned long)ds->pebs_index; if (top <= at) - goto out; + return; ds->pebs_index = ds->pebs_buffer_base; @@ -604,8 +597,6 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) if (perf_event_overflow(event, 1, &data, ®s)) x86_pmu_stop(event); } -out: - intel_pmu_pebs_enable_all(); } /* -- cgit v0.10.2 From 4807e3d5dc7bb7057dd6ca3abb09f3da2eb8c323 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Sat, 6 Mar 2010 13:47:07 +0100 Subject: perf, x86: Fix PEBS enable/disable vs cpuc->enabled We should never call ->enable with the pmu enabled, and we _can_ have ->disable called with the pmu enabled. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index 66c6962..9ad0e67 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -338,7 +338,7 @@ static void intel_pmu_pebs_enable(struct perf_event *event) hwc->config &= ~ARCH_PERFMON_EVENTSEL_INT; val |= 1ULL << hwc->idx; - wrmsrl(MSR_IA32_PEBS_ENABLE, val); + WARN_ON_ONCE(cpuc->enabled); if (x86_pmu.intel_cap.pebs_trap) intel_pmu_lbr_enable(event); @@ -351,7 +351,8 @@ static void intel_pmu_pebs_disable(struct perf_event *event) u64 val = cpuc->pebs_enabled; val &= ~(1ULL << hwc->idx); - wrmsrl(MSR_IA32_PEBS_ENABLE, val); + if (cpuc->enabled) + wrmsrl(MSR_IA32_PEBS_ENABLE, val); hwc->config |= ARCH_PERFMON_EVENTSEL_INT; -- cgit v0.10.2 From 2df202bf7520eaffcbfb07e45dfa3cfb0aeee2c0 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Sat, 6 Mar 2010 13:48:54 +0100 Subject: perf, x86: Fix LBR enable/disable vs cpuc->enabled We should never call ->enable with the pmu enabled, and we _can_ have ->disable called with the pmu enabled. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c index dcec765..0145f99 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c +++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c @@ -69,7 +69,7 @@ static void intel_pmu_lbr_enable(struct perf_event *event) if (!x86_pmu.lbr_nr) return; - WARN_ON(cpuc->enabled); + WARN_ON_ONCE(cpuc->enabled); /* * Reset the LBR stack if this is the first LBR user or @@ -93,9 +93,10 @@ static void intel_pmu_lbr_disable(struct perf_event *event) return; cpuc->lbr_users--; - BUG_ON(cpuc->lbr_users < 0); - WARN_ON(cpuc->enabled); + + if (cpuc->enabled && !cpuc->lbr_users) + __intel_pmu_lbr_disable(); } static void intel_pmu_lbr_enable_all(void) -- cgit v0.10.2 From d329527e47851f84b1e7944ed9601205f35f1b93 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 8 Mar 2010 13:57:14 +0100 Subject: perf, x86: Reorder intel_pmu_enable_all() The documentation says we have to enable PEBS before we enable the PMU proper. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index c135ed7..d3e2424 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -487,6 +487,8 @@ static void intel_pmu_enable_all(void) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + intel_pmu_pebs_enable_all(); + intel_pmu_lbr_enable_all(); wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, x86_pmu.intel_ctrl); if (test_bit(X86_PMC_IDX_FIXED_BTS, cpuc->active_mask)) { @@ -498,9 +500,6 @@ static void intel_pmu_enable_all(void) intel_pmu_enable_bts(event->hw.config); } - - intel_pmu_pebs_enable_all(); - intel_pmu_lbr_enable_all(); } static inline u64 intel_pmu_get_status(void) -- cgit v0.10.2 From 12ab854d744f04bfc5c6c4db723b7e31fc03eb29 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Sat, 6 Mar 2010 18:57:38 +0100 Subject: perf, x86: Deal with multiple state bits for pebs-fmt1 Its unclear if the PEBS state record will have only a single bit set, in case it does not and accumulates bits, deal with that by only processing each event once. Also, robustify some of the code. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index 9ad0e67..b4680da 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -538,6 +538,7 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) struct perf_event *event = NULL; struct perf_raw_record raw; struct pt_regs regs; + u64 status = 0; int bit, n; if (!ds || !x86_pmu.pebs) @@ -561,13 +562,22 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) for ( ; at < top; at++) { for_each_bit(bit, (unsigned long *)&at->status, MAX_PEBS_EVENTS) { - if (!cpuc->events[bit]->attr.precise) + event = cpuc->events[bit]; + if (!test_bit(bit, cpuc->active_mask)) continue; - event = cpuc->events[bit]; + WARN_ON_ONCE(!event); + + if (!event->attr.precise) + continue; + + if (__test_and_set_bit(bit, (unsigned long *)&status)) + continue; + + break; } - if (!event) + if (!event || bit >= MAX_PEBS_EVENTS) continue; if (!intel_pmu_save_and_restart(event)) -- cgit v0.10.2 From ad0e6cfe2a2a61d7b5530188e571d508146cb43b Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Sat, 6 Mar 2010 19:49:06 +0100 Subject: perf, x86: Fix silly bug in intel_pmu_pebs_{enable,disable} We need to use the actual cpuc->pebs_enabled value, not a local copy for the changes to take effect. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index b4680da..2423694 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -333,11 +333,10 @@ static void intel_pmu_pebs_enable(struct perf_event *event) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); struct hw_perf_event *hwc = &event->hw; - u64 val = cpuc->pebs_enabled; hwc->config &= ~ARCH_PERFMON_EVENTSEL_INT; - val |= 1ULL << hwc->idx; + cpuc->pebs_enabled |= 1ULL << hwc->idx; WARN_ON_ONCE(cpuc->enabled); if (x86_pmu.intel_cap.pebs_trap) @@ -348,11 +347,10 @@ static void intel_pmu_pebs_disable(struct perf_event *event) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); struct hw_perf_event *hwc = &event->hw; - u64 val = cpuc->pebs_enabled; - val &= ~(1ULL << hwc->idx); + cpuc->pebs_enabled &= ~(1ULL << hwc->idx); if (cpuc->enabled) - wrmsrl(MSR_IA32_PEBS_ENABLE, val); + wrmsrl(MSR_IA32_PEBS_ENABLE, cpuc->pebs_enabled); hwc->config |= ARCH_PERFMON_EVENTSEL_INT; -- cgit v0.10.2 From b83a46e7da4a948cc852ba7805dfb1a392dec861 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 8 Mar 2010 13:51:12 +0100 Subject: perf, x86: Don't reset the LBR as frequently If we reset the LBR on each first counter, simple counter rotation which first deschedules all counters and then reschedules the new ones will lead to LBR reset, even though we're still in the same task context. Reduce this by not flushing on the first counter but only flushing on different task contexts. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c index 0145f99..f278136 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c +++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c @@ -72,12 +72,11 @@ static void intel_pmu_lbr_enable(struct perf_event *event) WARN_ON_ONCE(cpuc->enabled); /* - * Reset the LBR stack if this is the first LBR user or - * we changed task context so as to avoid data leaks. + * Reset the LBR stack if we changed task context to + * avoid data leaks. */ - if (!cpuc->lbr_users || - (event->ctx->task && cpuc->lbr_context != event->ctx)) { + if (event->ctx->task && cpuc->lbr_context != event->ctx) { intel_pmu_lbr_reset(); cpuc->lbr_context = event->ctx; } @@ -93,7 +92,7 @@ static void intel_pmu_lbr_disable(struct perf_event *event) return; cpuc->lbr_users--; - BUG_ON(cpuc->lbr_users < 0); + WARN_ON_ONCE(cpuc->lbr_users < 0); if (cpuc->enabled && !cpuc->lbr_users) __intel_pmu_lbr_disable(); -- cgit v0.10.2 From 7645a24cbd01cbf4865d1273d5ddaa8d8c2ccb3a Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 8 Mar 2010 13:51:31 +0100 Subject: perf, x86: Remove checking_{wr,rd}msr() usage We don't need checking_{wr,rd}msr() calls, since we should know what cpu we're running on and not use blindly poke at msrs. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 335ee1d..e24f637 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -29,6 +29,17 @@ #include #include +#if 0 +#undef wrmsrl +#define wrmsrl(msr, val) \ +do { \ + trace_printk("wrmsrl(%lx, %lx)\n", (unsigned long)(msr),\ + (unsigned long)(val)); \ + native_write_msr((msr), (u32)((u64)(val)), \ + (u32)((u64)(val) >> 32)); \ +} while (0) +#endif + /* * best effort, GUP based copy_from_user() that assumes IRQ or NMI context */ @@ -821,14 +832,15 @@ void hw_perf_enable(void) static inline void __x86_pmu_enable_event(struct hw_perf_event *hwc) { - (void)checking_wrmsrl(hwc->config_base + hwc->idx, + wrmsrl(hwc->config_base + hwc->idx, hwc->config | ARCH_PERFMON_EVENTSEL_ENABLE); } static inline void x86_pmu_disable_event(struct perf_event *event) { struct hw_perf_event *hwc = &event->hw; - (void)checking_wrmsrl(hwc->config_base + hwc->idx, hwc->config); + + wrmsrl(hwc->config_base + hwc->idx, hwc->config); } static DEFINE_PER_CPU(u64 [X86_PMC_IDX_MAX], pmc_prev_left); @@ -843,7 +855,7 @@ x86_perf_event_set_period(struct perf_event *event) struct hw_perf_event *hwc = &event->hw; s64 left = atomic64_read(&hwc->period_left); s64 period = hwc->sample_period; - int err, ret = 0, idx = hwc->idx; + int ret = 0, idx = hwc->idx; if (idx == X86_PMC_IDX_FIXED_BTS) return 0; @@ -881,8 +893,8 @@ x86_perf_event_set_period(struct perf_event *event) */ atomic64_set(&hwc->prev_count, (u64)-left); - err = checking_wrmsrl(hwc->event_base + idx, - (u64)(-left) & x86_pmu.event_mask); + wrmsrl(hwc->event_base + idx, + (u64)(-left) & x86_pmu.event_mask); perf_event_update_userpage(event); @@ -987,7 +999,7 @@ void perf_event_print_debug(void) pr_info("CPU#%d: fixed: %016llx\n", cpu, fixed); pr_info("CPU#%d: pebs: %016llx\n", cpu, pebs); } - pr_info("CPU#%d: active: %016llx\n", cpu, *(u64 *)cpuc->active_mask); + pr_info("CPU#%d: active: %016llx\n", cpu, *(u64 *)cpuc->active_mask); for (idx = 0; idx < x86_pmu.num_events; idx++) { rdmsrl(x86_pmu.eventsel + idx, pmc_ctrl); diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index d3e2424..971dc6e 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -525,7 +525,7 @@ static void intel_pmu_disable_fixed(struct hw_perf_event *hwc) rdmsrl(hwc->config_base, ctrl_val); ctrl_val &= ~mask; - (void)checking_wrmsrl(hwc->config_base, ctrl_val); + wrmsrl(hwc->config_base, ctrl_val); } static void intel_pmu_disable_event(struct perf_event *event) @@ -553,7 +553,6 @@ static void intel_pmu_enable_fixed(struct hw_perf_event *hwc) { int idx = hwc->idx - X86_PMC_IDX_FIXED; u64 ctrl_val, bits, mask; - int err; /* * Enable IRQ generation (0x8), @@ -578,7 +577,7 @@ static void intel_pmu_enable_fixed(struct hw_perf_event *hwc) rdmsrl(hwc->config_base, ctrl_val); ctrl_val &= ~mask; ctrl_val |= bits; - err = checking_wrmsrl(hwc->config_base, ctrl_val); + wrmsrl(hwc->config_base, ctrl_val); } static void intel_pmu_enable_event(struct perf_event *event) -- cgit v0.10.2 From d80c7502ff63aa0d99d8c0c5803d28bbef67a74e Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 9 Mar 2010 11:41:02 +0100 Subject: perf, x86: Fixup the PEBS handler for Core2 cpus Pull the core handler in line with the nhm one, also make sure we always drain the buffer. Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index 2423694..1bfd59b 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -472,20 +472,39 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) struct pt_regs regs; int n; - if (!event || !ds || !x86_pmu.pebs) + if (!ds || !x86_pmu.pebs) return; at = (struct pebs_record_core *)(unsigned long)ds->pebs_buffer_base; top = (struct pebs_record_core *)(unsigned long)ds->pebs_index; - if (top <= at) + /* + * Whatever else happens, drain the thing + */ + ds->pebs_index = ds->pebs_buffer_base; + + if (!test_bit(0, cpuc->active_mask)) return; - ds->pebs_index = ds->pebs_buffer_base; + WARN_ON_ONCE(!event); + + if (!event->attr.precise) + return; + + n = top - at; + if (n <= 0) + return; if (!intel_pmu_save_and_restart(event)) return; + /* + * Should not happen, we program the threshold at 1 and do not + * set a reset value. + */ + WARN_ON_ONCE(n > 1); + at += n - 1; + perf_sample_data_init(&data, 0); data.period = event->hw.last_period; @@ -495,14 +514,6 @@ static void intel_pmu_drain_pebs_core(struct pt_regs *iregs) data.raw = &raw; } - n = top - at; - - /* - * Should not happen, we program the threshold at 1 and do not - * set a reset value. - */ - WARN_ON_ONCE(n > 1); - /* * We use the interrupt regs as a base because the PEBS record * does not contain a full regs set, specifically it seems to @@ -545,12 +556,11 @@ static void intel_pmu_drain_pebs_nhm(struct pt_regs *iregs) at = (struct pebs_record_nhm *)(unsigned long)ds->pebs_buffer_base; top = (struct pebs_record_nhm *)(unsigned long)ds->pebs_index; - if (top <= at) - return; - ds->pebs_index = ds->pebs_buffer_base; n = top - at; + if (n <= 0) + return; /* * Should not happen, we program the threshold at 1 and do not -- cgit v0.10.2 From 63fb3f9b2312e131be5a0a2dddb63f2fb123db9b Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 9 Mar 2010 11:51:02 +0100 Subject: perf, x86: Fix LBR read-out Don't decrement the TOS twice... Signed-off-by: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: paulus@samba.org Cc: eranian@google.com Cc: robert.richter@amd.com Cc: fweisbec@gmail.com LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c index f278136..df4c98e 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c +++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c @@ -129,7 +129,7 @@ static void intel_pmu_lbr_read_32(struct cpu_hw_events *cpuc) u64 tos = intel_pmu_lbr_tos(); int i; - for (i = 0; i < x86_pmu.lbr_nr; i++, tos--) { + for (i = 0; i < x86_pmu.lbr_nr; i++) { unsigned long lbr_idx = (tos - i) & mask; union { struct { @@ -162,7 +162,7 @@ static void intel_pmu_lbr_read_64(struct cpu_hw_events *cpuc) u64 tos = intel_pmu_lbr_tos(); int i; - for (i = 0; i < x86_pmu.lbr_nr; i++, tos--) { + for (i = 0; i < x86_pmu.lbr_nr; i++) { unsigned long lbr_idx = (tos - i) & mask; u64 from, to, flags = 0; -- cgit v0.10.2 From ba7e4d13fc7e25af1d167d40e6f028298dfc55ad Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Sat, 6 Jun 2009 13:58:12 +0200 Subject: perf, x86: Add INSTRUCTION_DECODER config flag The PEBS+LBR decoding magic needs the insn_get_length() infrastructure to be able to decode x86 instruction length. So split it out of KPROBES dependency and make it enabled when either KPROBES or PERF_EVENTS is enabled. Cc: Peter Zijlstra Cc: Masami Hiramatsu Cc: Frederic Weisbecker Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index e984403..e1240f6 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -58,6 +58,9 @@ config X86 select HAVE_ARCH_KMEMCHECK select HAVE_USER_RETURN_NOTIFIER +config INSTRUCTION_DECODER + def_bool (KPROBES || PERF_EVENTS) + config OUTPUT_FORMAT string default "elf32-i386" if X86_32 diff --git a/arch/x86/lib/Makefile b/arch/x86/lib/Makefile index 419386c..cbaf8f2 100644 --- a/arch/x86/lib/Makefile +++ b/arch/x86/lib/Makefile @@ -20,7 +20,7 @@ lib-y := delay.o lib-y += thunk_$(BITS).o lib-y += usercopy_$(BITS).o getuser.o putuser.o lib-y += memcpy_$(BITS).o -lib-$(CONFIG_KPROBES) += insn.o inat.o +lib-$(CONFIG_INSTRUCTION_DECODER) += insn.o inat.o obj-y += msr.o msr-reg.o msr-reg-export.o -- cgit v0.10.2 From caa0142d84ceb0fc83e28f0475d0a7316cb6df77 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Sat, 6 Jun 2009 13:58:12 +0200 Subject: perf, x86: Fix the !CONFIG_CPU_SUP_INTEL build Fix typo. But the modularization here is ugly and should be improved. Cc: Peter Zijlstra Cc: Frederic Weisbecker Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index 1bfd59b..c59678a 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -661,7 +661,7 @@ static void intel_ds_init(void) #else /* CONFIG_CPU_SUP_INTEL */ -static int reseve_ds_buffers(void) +static int reserve_ds_buffers(void) { return 0; } -- cgit v0.10.2 From 8a03ae2a5baed3df09e5643615bdd853fc142a09 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 29 Jan 2010 20:39:07 +0000 Subject: block: drbd: Convert semaphore to mutex The bm_change semaphore is semantically a mutex. Convert it to a real mutex. Signed-off-by: Thomas Gleixner Signed-off-by: Philipp Reisner diff --git a/drivers/block/drbd/drbd_bitmap.c b/drivers/block/drbd/drbd_bitmap.c index b61057e..f58e765 100644 --- a/drivers/block/drbd/drbd_bitmap.c +++ b/drivers/block/drbd/drbd_bitmap.c @@ -66,7 +66,7 @@ struct drbd_bitmap { size_t bm_words; size_t bm_number_of_pages; sector_t bm_dev_capacity; - struct semaphore bm_change; /* serializes resize operations */ + struct mutex bm_change; /* serializes resize operations */ atomic_t bm_async_io; wait_queue_head_t bm_io_wait; @@ -114,7 +114,7 @@ void drbd_bm_lock(struct drbd_conf *mdev, char *why) return; } - trylock_failed = down_trylock(&b->bm_change); + trylock_failed = !mutex_trylock(&b->bm_change); if (trylock_failed) { dev_warn(DEV, "%s going to '%s' but bitmap already locked for '%s' by %s\n", @@ -125,7 +125,7 @@ void drbd_bm_lock(struct drbd_conf *mdev, char *why) b->bm_task == mdev->receiver.task ? "receiver" : b->bm_task == mdev->asender.task ? "asender" : b->bm_task == mdev->worker.task ? "worker" : "?"); - down(&b->bm_change); + mutex_lock(&b->bm_change); } if (__test_and_set_bit(BM_LOCKED, &b->bm_flags)) dev_err(DEV, "FIXME bitmap already locked in bm_lock\n"); @@ -147,7 +147,7 @@ void drbd_bm_unlock(struct drbd_conf *mdev) b->bm_why = NULL; b->bm_task = NULL; - up(&b->bm_change); + mutex_unlock(&b->bm_change); } /* word offset to long pointer */ @@ -295,7 +295,7 @@ int drbd_bm_init(struct drbd_conf *mdev) if (!b) return -ENOMEM; spin_lock_init(&b->bm_lock); - init_MUTEX(&b->bm_change); + mutex_init(&b->bm_change); init_waitqueue_head(&b->bm_io_wait); mdev->bitmap = b; -- cgit v0.10.2 From 938179b4f8cf8a4f11234ebf2dff2eb48400acfe Mon Sep 17 00:00:00 2001 From: Dimitri Sivanich Date: Fri, 5 Mar 2010 11:42:03 -0600 Subject: x86: Improve Intel microcode loader performance We've noticed that on large SGI UV system configurations, running microcode.ctl can take very long periods of time. This is due to the large number of vmalloc/vfree calls made by the Intel generic_load_microcode() logic. By reusing allocated space, the following patch reduces the time to run microcode.ctl on a 1024 cpu system from approximately 80 seconds down to 1 or 2 seconds. Signed-off-by: Dimitri Sivanich Acked-by: Dmitry Adamushko Cc: Avi Kivity Cc: Bill Davidsen LKML-Reference: <20100305174203.GA19638@sgi.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/microcode_intel.c b/arch/x86/kernel/microcode_intel.c index 85a343e..3561702 100644 --- a/arch/x86/kernel/microcode_intel.c +++ b/arch/x86/kernel/microcode_intel.c @@ -343,10 +343,11 @@ static enum ucode_state generic_load_microcode(int cpu, void *data, size_t size, int (*get_ucode_data)(void *, const void *, size_t)) { struct ucode_cpu_info *uci = ucode_cpu_info + cpu; - u8 *ucode_ptr = data, *new_mc = NULL, *mc; + u8 *ucode_ptr = data, *new_mc = NULL, *mc = NULL; int new_rev = uci->cpu_sig.rev; unsigned int leftover = size; enum ucode_state state = UCODE_OK; + unsigned int curr_mc_size = 0; while (leftover) { struct microcode_header_intel mc_header; @@ -361,9 +362,15 @@ static enum ucode_state generic_load_microcode(int cpu, void *data, size_t size, break; } - mc = vmalloc(mc_size); - if (!mc) - break; + /* For performance reasons, reuse mc area when possible */ + if (!mc || mc_size > curr_mc_size) { + if (mc) + vfree(mc); + mc = vmalloc(mc_size); + if (!mc) + break; + curr_mc_size = mc_size; + } if (get_ucode_data(mc, ucode_ptr, mc_size) || microcode_sanity_check(mc) < 0) { @@ -376,13 +383,16 @@ static enum ucode_state generic_load_microcode(int cpu, void *data, size_t size, vfree(new_mc); new_rev = mc_header.rev; new_mc = mc; - } else - vfree(mc); + mc = NULL; /* trigger new vmalloc */ + } ucode_ptr += mc_size; leftover -= mc_size; } + if (mc) + vfree(mc); + if (leftover) { if (new_mc) vfree(new_mc); -- cgit v0.10.2 From 6f4edd69e40aba4f45bf9558c1e9a950d79ab4e4 Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Wed, 10 Mar 2010 14:44:58 -0600 Subject: x86, UV: Clean up UV headers for MMR definitions Update UV mmr definitions header file. Eliminate definitions no longer needed. Move 2 definitions from tlb_uv.c into the header file where they belong. Signed-off-by: Jack Steiner LKML-Reference: <20100310204458.GA28835@sgi.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/uv/uv_mmrs.h b/arch/x86/include/asm/uv/uv_mmrs.h index 2cae46c..b2f2d2e 100644 --- a/arch/x86/include/asm/uv/uv_mmrs.h +++ b/arch/x86/include/asm/uv/uv_mmrs.h @@ -1,4 +1,3 @@ - /* * This file is subject to the terms and conditions of the GNU General Public * License. See the file "COPYING" in the main directory of this archive @@ -15,13 +14,25 @@ #define UV_MMR_ENABLE (1UL << 63) /* ========================================================================= */ +/* UVH_BAU_DATA_BROADCAST */ +/* ========================================================================= */ +#define UVH_BAU_DATA_BROADCAST 0x61688UL +#define UVH_BAU_DATA_BROADCAST_32 0x0440 + +#define UVH_BAU_DATA_BROADCAST_ENABLE_SHFT 0 +#define UVH_BAU_DATA_BROADCAST_ENABLE_MASK 0x0000000000000001UL + +union uvh_bau_data_broadcast_u { + unsigned long v; + struct uvh_bau_data_broadcast_s { + unsigned long enable : 1; /* RW */ + unsigned long rsvd_1_63: 63; /* */ + } s; +}; + +/* ========================================================================= */ /* UVH_BAU_DATA_CONFIG */ /* ========================================================================= */ -#define UVH_LB_BAU_MISC_CONTROL 0x320170UL -#define UV_ENABLE_INTD_SOFT_ACK_MODE_SHIFT 15 -#define UV_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHIFT 16 -#define UV_INTD_SOFT_ACK_TIMEOUT_PERIOD 0x000000000bUL -/* 1011 timebase 7 (168millisec) * 3 ticks -> 500ms */ #define UVH_BAU_DATA_CONFIG 0x61680UL #define UVH_BAU_DATA_CONFIG_32 0x0438 @@ -604,6 +615,68 @@ union uvh_lb_bau_intd_software_acknowledge_u { #define UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE_ALIAS_32 0x0a70 /* ========================================================================= */ +/* UVH_LB_BAU_MISC_CONTROL */ +/* ========================================================================= */ +#define UVH_LB_BAU_MISC_CONTROL 0x320170UL +#define UVH_LB_BAU_MISC_CONTROL_32 0x00a10 + +#define UVH_LB_BAU_MISC_CONTROL_REJECTION_DELAY_SHFT 0 +#define UVH_LB_BAU_MISC_CONTROL_REJECTION_DELAY_MASK 0x00000000000000ffUL +#define UVH_LB_BAU_MISC_CONTROL_APIC_MODE_SHFT 8 +#define UVH_LB_BAU_MISC_CONTROL_APIC_MODE_MASK 0x0000000000000100UL +#define UVH_LB_BAU_MISC_CONTROL_FORCE_BROADCAST_SHFT 9 +#define UVH_LB_BAU_MISC_CONTROL_FORCE_BROADCAST_MASK 0x0000000000000200UL +#define UVH_LB_BAU_MISC_CONTROL_FORCE_LOCK_NOP_SHFT 10 +#define UVH_LB_BAU_MISC_CONTROL_FORCE_LOCK_NOP_MASK 0x0000000000000400UL +#define UVH_LB_BAU_MISC_CONTROL_CSI_AGENT_PRESENCE_VECTOR_SHFT 11 +#define UVH_LB_BAU_MISC_CONTROL_CSI_AGENT_PRESENCE_VECTOR_MASK 0x0000000000003800UL +#define UVH_LB_BAU_MISC_CONTROL_DESCRIPTOR_FETCH_MODE_SHFT 14 +#define UVH_LB_BAU_MISC_CONTROL_DESCRIPTOR_FETCH_MODE_MASK 0x0000000000004000UL +#define UVH_LB_BAU_MISC_CONTROL_ENABLE_INTD_SOFT_ACK_MODE_SHFT 15 +#define UVH_LB_BAU_MISC_CONTROL_ENABLE_INTD_SOFT_ACK_MODE_MASK 0x0000000000008000UL +#define UVH_LB_BAU_MISC_CONTROL_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHFT 16 +#define UVH_LB_BAU_MISC_CONTROL_INTD_SOFT_ACK_TIMEOUT_PERIOD_MASK 0x00000000000f0000UL +#define UVH_LB_BAU_MISC_CONTROL_ENABLE_DUAL_MAPPING_MODE_SHFT 20 +#define UVH_LB_BAU_MISC_CONTROL_ENABLE_DUAL_MAPPING_MODE_MASK 0x0000000000100000UL +#define UVH_LB_BAU_MISC_CONTROL_VGA_IO_PORT_DECODE_ENABLE_SHFT 21 +#define UVH_LB_BAU_MISC_CONTROL_VGA_IO_PORT_DECODE_ENABLE_MASK 0x0000000000200000UL +#define UVH_LB_BAU_MISC_CONTROL_VGA_IO_PORT_16_BIT_DECODE_SHFT 22 +#define UVH_LB_BAU_MISC_CONTROL_VGA_IO_PORT_16_BIT_DECODE_MASK 0x0000000000400000UL +#define UVH_LB_BAU_MISC_CONTROL_SUPPRESS_DEST_REGISTRATION_SHFT 23 +#define UVH_LB_BAU_MISC_CONTROL_SUPPRESS_DEST_REGISTRATION_MASK 0x0000000000800000UL +#define UVH_LB_BAU_MISC_CONTROL_PROGRAMMED_INITIAL_PRIORITY_SHFT 24 +#define UVH_LB_BAU_MISC_CONTROL_PROGRAMMED_INITIAL_PRIORITY_MASK 0x0000000007000000UL +#define UVH_LB_BAU_MISC_CONTROL_USE_INCOMING_PRIORITY_SHFT 27 +#define UVH_LB_BAU_MISC_CONTROL_USE_INCOMING_PRIORITY_MASK 0x0000000008000000UL +#define UVH_LB_BAU_MISC_CONTROL_ENABLE_PROGRAMMED_INITIAL_PRIORITY_SHFT 28 +#define UVH_LB_BAU_MISC_CONTROL_ENABLE_PROGRAMMED_INITIAL_PRIORITY_MASK 0x0000000010000000UL +#define UVH_LB_BAU_MISC_CONTROL_FUN_SHFT 48 +#define UVH_LB_BAU_MISC_CONTROL_FUN_MASK 0xffff000000000000UL + +union uvh_lb_bau_misc_control_u { + unsigned long v; + struct uvh_lb_bau_misc_control_s { + unsigned long rejection_delay : 8; /* RW */ + unsigned long apic_mode : 1; /* RW */ + unsigned long force_broadcast : 1; /* RW */ + unsigned long force_lock_nop : 1; /* RW */ + unsigned long csi_agent_presence_vector : 3; /* RW */ + unsigned long descriptor_fetch_mode : 1; /* RW */ + unsigned long enable_intd_soft_ack_mode : 1; /* RW */ + unsigned long intd_soft_ack_timeout_period : 4; /* RW */ + unsigned long enable_dual_mapping_mode : 1; /* RW */ + unsigned long vga_io_port_decode_enable : 1; /* RW */ + unsigned long vga_io_port_16_bit_decode : 1; /* RW */ + unsigned long suppress_dest_registration : 1; /* RW */ + unsigned long programmed_initial_priority : 3; /* RW */ + unsigned long use_incoming_priority : 1; /* RW */ + unsigned long enable_programmed_initial_priority : 1; /* RW */ + unsigned long rsvd_29_47 : 19; /* */ + unsigned long fun : 16; /* RW */ + } s; +}; + +/* ========================================================================= */ /* UVH_LB_BAU_SB_ACTIVATION_CONTROL */ /* ========================================================================= */ #define UVH_LB_BAU_SB_ACTIVATION_CONTROL 0x320020UL @@ -681,334 +754,6 @@ union uvh_lb_bau_sb_descriptor_base_u { }; /* ========================================================================= */ -/* UVH_LB_MCAST_AOERR0_RPT_ENABLE */ -/* ========================================================================= */ -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE 0x50b20UL - -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_OBESE_MSG_SHFT 0 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_OBESE_MSG_MASK 0x0000000000000001UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_DATA_SB_ERR_SHFT 1 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_DATA_SB_ERR_MASK 0x0000000000000002UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_NACK_BUFF_PARITY_SHFT 2 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_NACK_BUFF_PARITY_MASK 0x0000000000000004UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_TIMEOUT_SHFT 3 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_TIMEOUT_MASK 0x0000000000000008UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_INACTIVE_REPLY_SHFT 4 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_INACTIVE_REPLY_MASK 0x0000000000000010UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_UPGRADE_ERROR_SHFT 5 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_UPGRADE_ERROR_MASK 0x0000000000000020UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_REG_COUNT_UNDERFLOW_SHFT 6 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_REG_COUNT_UNDERFLOW_MASK 0x0000000000000040UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_REP_OBESE_MSG_SHFT 7 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MCAST_REP_OBESE_MSG_MASK 0x0000000000000080UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REQ_RUNT_MSG_SHFT 8 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REQ_RUNT_MSG_MASK 0x0000000000000100UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REQ_OBESE_MSG_SHFT 9 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REQ_OBESE_MSG_MASK 0x0000000000000200UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REQ_DATA_SB_ERR_SHFT 10 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REQ_DATA_SB_ERR_MASK 0x0000000000000400UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REP_RUNT_MSG_SHFT 11 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REP_RUNT_MSG_MASK 0x0000000000000800UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REP_OBESE_MSG_SHFT 12 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REP_OBESE_MSG_MASK 0x0000000000001000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REP_DATA_SB_ERR_SHFT 13 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REP_DATA_SB_ERR_MASK 0x0000000000002000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REP_COMMAND_ERR_SHFT 14 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_REP_COMMAND_ERR_MASK 0x0000000000004000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_PEND_TIMEOUT_SHFT 15 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_UCACHE_PEND_TIMEOUT_MASK 0x0000000000008000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_REQ_RUNT_MSG_SHFT 16 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_REQ_RUNT_MSG_MASK 0x0000000000010000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_REQ_OBESE_MSG_SHFT 17 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_REQ_OBESE_MSG_MASK 0x0000000000020000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_REQ_DATA_SB_ERR_SHFT 18 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_REQ_DATA_SB_ERR_MASK 0x0000000000040000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_REP_RUNT_MSG_SHFT 19 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_REP_RUNT_MSG_MASK 0x0000000000080000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_REP_OBESE_MSG_SHFT 20 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_REP_OBESE_MSG_MASK 0x0000000000100000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_REP_DATA_SB_ERR_SHFT 21 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_REP_DATA_SB_ERR_MASK 0x0000000000200000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_AMO_TIMEOUT_SHFT 22 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_AMO_TIMEOUT_MASK 0x0000000000400000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_PUT_TIMEOUT_SHFT 23 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_PUT_TIMEOUT_MASK 0x0000000000800000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_SPURIOUS_EVENT_SHFT 24 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_MACC_SPURIOUS_EVENT_MASK 0x0000000001000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_IOH_DESTINATION_TABLE_PARITY_SHFT 25 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_IOH_DESTINATION_TABLE_PARITY_MASK 0x0000000002000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_GET_HAD_ERROR_REPLY_SHFT 26 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_GET_HAD_ERROR_REPLY_MASK 0x0000000004000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_GET_TIMEOUT_SHFT 27 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_GET_TIMEOUT_MASK 0x0000000008000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_LOCK_MANAGER_HAD_ERROR_REPLY_SHFT 28 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_LOCK_MANAGER_HAD_ERROR_REPLY_MASK 0x0000000010000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_PUT_HAD_ERROR_REPLY_SHFT 29 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_PUT_HAD_ERROR_REPLY_MASK 0x0000000020000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_PUT_TIMEOUT_SHFT 30 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_PUT_TIMEOUT_MASK 0x0000000040000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_SB_ACTIVATION_OVERRUN_SHFT 31 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_SB_ACTIVATION_OVERRUN_MASK 0x0000000080000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_COMPLETED_GB_ACTIVATION_HAD_ERROR_REPLY_SHFT 32 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_COMPLETED_GB_ACTIVATION_HAD_ERROR_REPLY_MASK 0x0000000100000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_COMPLETED_GB_ACTIVATION_TIMEOUT_SHFT 33 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_COMPLETED_GB_ACTIVATION_TIMEOUT_MASK 0x0000000200000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_DESCRIPTOR_BUFFER_0_PARITY_SHFT 34 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_DESCRIPTOR_BUFFER_0_PARITY_MASK 0x0000000400000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_DESCRIPTOR_BUFFER_1_PARITY_SHFT 35 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_DESCRIPTOR_BUFFER_1_PARITY_MASK 0x0000000800000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_SOCKET_DESTINATION_TABLE_PARITY_SHFT 36 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_SOCKET_DESTINATION_TABLE_PARITY_MASK 0x0000001000000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_BAU_REPLY_PAYLOAD_CORRUPTION_SHFT 37 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_BAU_REPLY_PAYLOAD_CORRUPTION_MASK 0x0000002000000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_IO_PORT_DESTINATION_TABLE_PARITY_SHFT 38 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_IO_PORT_DESTINATION_TABLE_PARITY_MASK 0x0000004000000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_INTD_SOFT_ACK_TIMEOUT_SHFT 39 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_INTD_SOFT_ACK_TIMEOUT_MASK 0x0000008000000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_INT_REP_OBESE_MSG_SHFT 40 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_INT_REP_OBESE_MSG_MASK 0x0000010000000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_INT_REP_COMMAND_ERR_SHFT 41 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_INT_REP_COMMAND_ERR_MASK 0x0000020000000000UL -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_INT_TIMEOUT_SHFT 42 -#define UVH_LB_MCAST_AOERR0_RPT_ENABLE_INT_TIMEOUT_MASK 0x0000040000000000UL - -union uvh_lb_mcast_aoerr0_rpt_enable_u { - unsigned long v; - struct uvh_lb_mcast_aoerr0_rpt_enable_s { - unsigned long mcast_obese_msg : 1; /* RW */ - unsigned long mcast_data_sb_err : 1; /* RW */ - unsigned long mcast_nack_buff_parity : 1; /* RW */ - unsigned long mcast_timeout : 1; /* RW */ - unsigned long mcast_inactive_reply : 1; /* RW */ - unsigned long mcast_upgrade_error : 1; /* RW */ - unsigned long mcast_reg_count_underflow : 1; /* RW */ - unsigned long mcast_rep_obese_msg : 1; /* RW */ - unsigned long ucache_req_runt_msg : 1; /* RW */ - unsigned long ucache_req_obese_msg : 1; /* RW */ - unsigned long ucache_req_data_sb_err : 1; /* RW */ - unsigned long ucache_rep_runt_msg : 1; /* RW */ - unsigned long ucache_rep_obese_msg : 1; /* RW */ - unsigned long ucache_rep_data_sb_err : 1; /* RW */ - unsigned long ucache_rep_command_err : 1; /* RW */ - unsigned long ucache_pend_timeout : 1; /* RW */ - unsigned long macc_req_runt_msg : 1; /* RW */ - unsigned long macc_req_obese_msg : 1; /* RW */ - unsigned long macc_req_data_sb_err : 1; /* RW */ - unsigned long macc_rep_runt_msg : 1; /* RW */ - unsigned long macc_rep_obese_msg : 1; /* RW */ - unsigned long macc_rep_data_sb_err : 1; /* RW */ - unsigned long macc_amo_timeout : 1; /* RW */ - unsigned long macc_put_timeout : 1; /* RW */ - unsigned long macc_spurious_event : 1; /* RW */ - unsigned long ioh_destination_table_parity : 1; /* RW */ - unsigned long get_had_error_reply : 1; /* RW */ - unsigned long get_timeout : 1; /* RW */ - unsigned long lock_manager_had_error_reply : 1; /* RW */ - unsigned long put_had_error_reply : 1; /* RW */ - unsigned long put_timeout : 1; /* RW */ - unsigned long sb_activation_overrun : 1; /* RW */ - unsigned long completed_gb_activation_had_error_reply : 1; /* RW */ - unsigned long completed_gb_activation_timeout : 1; /* RW */ - unsigned long descriptor_buffer_0_parity : 1; /* RW */ - unsigned long descriptor_buffer_1_parity : 1; /* RW */ - unsigned long socket_destination_table_parity : 1; /* RW */ - unsigned long bau_reply_payload_corruption : 1; /* RW */ - unsigned long io_port_destination_table_parity : 1; /* RW */ - unsigned long intd_soft_ack_timeout : 1; /* RW */ - unsigned long int_rep_obese_msg : 1; /* RW */ - unsigned long int_rep_command_err : 1; /* RW */ - unsigned long int_timeout : 1; /* RW */ - unsigned long rsvd_43_63 : 21; /* */ - } s; -}; - -/* ========================================================================= */ -/* UVH_LOCAL_INT0_CONFIG */ -/* ========================================================================= */ -#define UVH_LOCAL_INT0_CONFIG 0x61000UL - -#define UVH_LOCAL_INT0_CONFIG_VECTOR_SHFT 0 -#define UVH_LOCAL_INT0_CONFIG_VECTOR_MASK 0x00000000000000ffUL -#define UVH_LOCAL_INT0_CONFIG_DM_SHFT 8 -#define UVH_LOCAL_INT0_CONFIG_DM_MASK 0x0000000000000700UL -#define UVH_LOCAL_INT0_CONFIG_DESTMODE_SHFT 11 -#define UVH_LOCAL_INT0_CONFIG_DESTMODE_MASK 0x0000000000000800UL -#define UVH_LOCAL_INT0_CONFIG_STATUS_SHFT 12 -#define UVH_LOCAL_INT0_CONFIG_STATUS_MASK 0x0000000000001000UL -#define UVH_LOCAL_INT0_CONFIG_P_SHFT 13 -#define UVH_LOCAL_INT0_CONFIG_P_MASK 0x0000000000002000UL -#define UVH_LOCAL_INT0_CONFIG_T_SHFT 15 -#define UVH_LOCAL_INT0_CONFIG_T_MASK 0x0000000000008000UL -#define UVH_LOCAL_INT0_CONFIG_M_SHFT 16 -#define UVH_LOCAL_INT0_CONFIG_M_MASK 0x0000000000010000UL -#define UVH_LOCAL_INT0_CONFIG_APIC_ID_SHFT 32 -#define UVH_LOCAL_INT0_CONFIG_APIC_ID_MASK 0xffffffff00000000UL - -union uvh_local_int0_config_u { - unsigned long v; - struct uvh_local_int0_config_s { - unsigned long vector_ : 8; /* RW */ - unsigned long dm : 3; /* RW */ - unsigned long destmode : 1; /* RW */ - unsigned long status : 1; /* RO */ - unsigned long p : 1; /* RO */ - unsigned long rsvd_14 : 1; /* */ - unsigned long t : 1; /* RO */ - unsigned long m : 1; /* RW */ - unsigned long rsvd_17_31: 15; /* */ - unsigned long apic_id : 32; /* RW */ - } s; -}; - -/* ========================================================================= */ -/* UVH_LOCAL_INT0_ENABLE */ -/* ========================================================================= */ -#define UVH_LOCAL_INT0_ENABLE 0x65000UL - -#define UVH_LOCAL_INT0_ENABLE_LB_HCERR_SHFT 0 -#define UVH_LOCAL_INT0_ENABLE_LB_HCERR_MASK 0x0000000000000001UL -#define UVH_LOCAL_INT0_ENABLE_GR0_HCERR_SHFT 1 -#define UVH_LOCAL_INT0_ENABLE_GR0_HCERR_MASK 0x0000000000000002UL -#define UVH_LOCAL_INT0_ENABLE_GR1_HCERR_SHFT 2 -#define UVH_LOCAL_INT0_ENABLE_GR1_HCERR_MASK 0x0000000000000004UL -#define UVH_LOCAL_INT0_ENABLE_LH_HCERR_SHFT 3 -#define UVH_LOCAL_INT0_ENABLE_LH_HCERR_MASK 0x0000000000000008UL -#define UVH_LOCAL_INT0_ENABLE_RH_HCERR_SHFT 4 -#define UVH_LOCAL_INT0_ENABLE_RH_HCERR_MASK 0x0000000000000010UL -#define UVH_LOCAL_INT0_ENABLE_XN_HCERR_SHFT 5 -#define UVH_LOCAL_INT0_ENABLE_XN_HCERR_MASK 0x0000000000000020UL -#define UVH_LOCAL_INT0_ENABLE_SI_HCERR_SHFT 6 -#define UVH_LOCAL_INT0_ENABLE_SI_HCERR_MASK 0x0000000000000040UL -#define UVH_LOCAL_INT0_ENABLE_LB_AOERR0_SHFT 7 -#define UVH_LOCAL_INT0_ENABLE_LB_AOERR0_MASK 0x0000000000000080UL -#define UVH_LOCAL_INT0_ENABLE_GR0_AOERR0_SHFT 8 -#define UVH_LOCAL_INT0_ENABLE_GR0_AOERR0_MASK 0x0000000000000100UL -#define UVH_LOCAL_INT0_ENABLE_GR1_AOERR0_SHFT 9 -#define UVH_LOCAL_INT0_ENABLE_GR1_AOERR0_MASK 0x0000000000000200UL -#define UVH_LOCAL_INT0_ENABLE_LH_AOERR0_SHFT 10 -#define UVH_LOCAL_INT0_ENABLE_LH_AOERR0_MASK 0x0000000000000400UL -#define UVH_LOCAL_INT0_ENABLE_RH_AOERR0_SHFT 11 -#define UVH_LOCAL_INT0_ENABLE_RH_AOERR0_MASK 0x0000000000000800UL -#define UVH_LOCAL_INT0_ENABLE_XN_AOERR0_SHFT 12 -#define UVH_LOCAL_INT0_ENABLE_XN_AOERR0_MASK 0x0000000000001000UL -#define UVH_LOCAL_INT0_ENABLE_SI_AOERR0_SHFT 13 -#define UVH_LOCAL_INT0_ENABLE_SI_AOERR0_MASK 0x0000000000002000UL -#define UVH_LOCAL_INT0_ENABLE_LB_AOERR1_SHFT 14 -#define UVH_LOCAL_INT0_ENABLE_LB_AOERR1_MASK 0x0000000000004000UL -#define UVH_LOCAL_INT0_ENABLE_GR0_AOERR1_SHFT 15 -#define UVH_LOCAL_INT0_ENABLE_GR0_AOERR1_MASK 0x0000000000008000UL -#define UVH_LOCAL_INT0_ENABLE_GR1_AOERR1_SHFT 16 -#define UVH_LOCAL_INT0_ENABLE_GR1_AOERR1_MASK 0x0000000000010000UL -#define UVH_LOCAL_INT0_ENABLE_LH_AOERR1_SHFT 17 -#define UVH_LOCAL_INT0_ENABLE_LH_AOERR1_MASK 0x0000000000020000UL -#define UVH_LOCAL_INT0_ENABLE_RH_AOERR1_SHFT 18 -#define UVH_LOCAL_INT0_ENABLE_RH_AOERR1_MASK 0x0000000000040000UL -#define UVH_LOCAL_INT0_ENABLE_XN_AOERR1_SHFT 19 -#define UVH_LOCAL_INT0_ENABLE_XN_AOERR1_MASK 0x0000000000080000UL -#define UVH_LOCAL_INT0_ENABLE_SI_AOERR1_SHFT 20 -#define UVH_LOCAL_INT0_ENABLE_SI_AOERR1_MASK 0x0000000000100000UL -#define UVH_LOCAL_INT0_ENABLE_RH_VPI_INT_SHFT 21 -#define UVH_LOCAL_INT0_ENABLE_RH_VPI_INT_MASK 0x0000000000200000UL -#define UVH_LOCAL_INT0_ENABLE_SYSTEM_SHUTDOWN_INT_SHFT 22 -#define UVH_LOCAL_INT0_ENABLE_SYSTEM_SHUTDOWN_INT_MASK 0x0000000000400000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_0_SHFT 23 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_0_MASK 0x0000000000800000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_1_SHFT 24 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_1_MASK 0x0000000001000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_2_SHFT 25 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_2_MASK 0x0000000002000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_3_SHFT 26 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_3_MASK 0x0000000004000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_4_SHFT 27 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_4_MASK 0x0000000008000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_5_SHFT 28 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_5_MASK 0x0000000010000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_6_SHFT 29 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_6_MASK 0x0000000020000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_7_SHFT 30 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_7_MASK 0x0000000040000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_8_SHFT 31 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_8_MASK 0x0000000080000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_9_SHFT 32 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_9_MASK 0x0000000100000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_10_SHFT 33 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_10_MASK 0x0000000200000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_11_SHFT 34 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_11_MASK 0x0000000400000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_12_SHFT 35 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_12_MASK 0x0000000800000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_13_SHFT 36 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_13_MASK 0x0000001000000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_14_SHFT 37 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_14_MASK 0x0000002000000000UL -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_15_SHFT 38 -#define UVH_LOCAL_INT0_ENABLE_LB_IRQ_INT_15_MASK 0x0000004000000000UL -#define UVH_LOCAL_INT0_ENABLE_L1_NMI_INT_SHFT 39 -#define UVH_LOCAL_INT0_ENABLE_L1_NMI_INT_MASK 0x0000008000000000UL -#define UVH_LOCAL_INT0_ENABLE_STOP_CLOCK_SHFT 40 -#define UVH_LOCAL_INT0_ENABLE_STOP_CLOCK_MASK 0x0000010000000000UL -#define UVH_LOCAL_INT0_ENABLE_ASIC_TO_L1_SHFT 41 -#define UVH_LOCAL_INT0_ENABLE_ASIC_TO_L1_MASK 0x0000020000000000UL -#define UVH_LOCAL_INT0_ENABLE_L1_TO_ASIC_SHFT 42 -#define UVH_LOCAL_INT0_ENABLE_L1_TO_ASIC_MASK 0x0000040000000000UL -#define UVH_LOCAL_INT0_ENABLE_LTC_INT_SHFT 43 -#define UVH_LOCAL_INT0_ENABLE_LTC_INT_MASK 0x0000080000000000UL -#define UVH_LOCAL_INT0_ENABLE_LA_SEQ_TRIGGER_SHFT 44 -#define UVH_LOCAL_INT0_ENABLE_LA_SEQ_TRIGGER_MASK 0x0000100000000000UL - -union uvh_local_int0_enable_u { - unsigned long v; - struct uvh_local_int0_enable_s { - unsigned long lb_hcerr : 1; /* RW */ - unsigned long gr0_hcerr : 1; /* RW */ - unsigned long gr1_hcerr : 1; /* RW */ - unsigned long lh_hcerr : 1; /* RW */ - unsigned long rh_hcerr : 1; /* RW */ - unsigned long xn_hcerr : 1; /* RW */ - unsigned long si_hcerr : 1; /* RW */ - unsigned long lb_aoerr0 : 1; /* RW */ - unsigned long gr0_aoerr0 : 1; /* RW */ - unsigned long gr1_aoerr0 : 1; /* RW */ - unsigned long lh_aoerr0 : 1; /* RW */ - unsigned long rh_aoerr0 : 1; /* RW */ - unsigned long xn_aoerr0 : 1; /* RW */ - unsigned long si_aoerr0 : 1; /* RW */ - unsigned long lb_aoerr1 : 1; /* RW */ - unsigned long gr0_aoerr1 : 1; /* RW */ - unsigned long gr1_aoerr1 : 1; /* RW */ - unsigned long lh_aoerr1 : 1; /* RW */ - unsigned long rh_aoerr1 : 1; /* RW */ - unsigned long xn_aoerr1 : 1; /* RW */ - unsigned long si_aoerr1 : 1; /* RW */ - unsigned long rh_vpi_int : 1; /* RW */ - unsigned long system_shutdown_int : 1; /* RW */ - unsigned long lb_irq_int_0 : 1; /* RW */ - unsigned long lb_irq_int_1 : 1; /* RW */ - unsigned long lb_irq_int_2 : 1; /* RW */ - unsigned long lb_irq_int_3 : 1; /* RW */ - unsigned long lb_irq_int_4 : 1; /* RW */ - unsigned long lb_irq_int_5 : 1; /* RW */ - unsigned long lb_irq_int_6 : 1; /* RW */ - unsigned long lb_irq_int_7 : 1; /* RW */ - unsigned long lb_irq_int_8 : 1; /* RW */ - unsigned long lb_irq_int_9 : 1; /* RW */ - unsigned long lb_irq_int_10 : 1; /* RW */ - unsigned long lb_irq_int_11 : 1; /* RW */ - unsigned long lb_irq_int_12 : 1; /* RW */ - unsigned long lb_irq_int_13 : 1; /* RW */ - unsigned long lb_irq_int_14 : 1; /* RW */ - unsigned long lb_irq_int_15 : 1; /* RW */ - unsigned long l1_nmi_int : 1; /* RW */ - unsigned long stop_clock : 1; /* RW */ - unsigned long asic_to_l1 : 1; /* RW */ - unsigned long l1_to_asic : 1; /* RW */ - unsigned long ltc_int : 1; /* RW */ - unsigned long la_seq_trigger : 1; /* RW */ - unsigned long rsvd_45_63 : 19; /* */ - } s; -}; - -/* ========================================================================= */ /* UVH_NODE_ID */ /* ========================================================================= */ #define UVH_NODE_ID 0x0UL @@ -1112,26 +857,6 @@ union uvh_rh_gam_alias210_redirect_config_2_mmr_u { }; /* ========================================================================= */ -/* UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR */ -/* ========================================================================= */ -#define UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR 0x1600020UL - -#define UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR_BASE_SHFT 26 -#define UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR_BASE_MASK 0x00003ffffc000000UL -#define UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR_ENABLE_SHFT 63 -#define UVH_RH_GAM_CFG_OVERLAY_CONFIG_MMR_ENABLE_MASK 0x8000000000000000UL - -union uvh_rh_gam_cfg_overlay_config_mmr_u { - unsigned long v; - struct uvh_rh_gam_cfg_overlay_config_mmr_s { - unsigned long rsvd_0_25: 26; /* */ - unsigned long base : 20; /* RW */ - unsigned long rsvd_46_62: 17; /* */ - unsigned long enable : 1; /* RW */ - } s; -}; - -/* ========================================================================= */ /* UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR */ /* ========================================================================= */ #define UVH_RH_GAM_GRU_OVERLAY_CONFIG_MMR 0x1600010UL @@ -1263,101 +988,6 @@ union uvh_rtc1_int_config_u { }; /* ========================================================================= */ -/* UVH_RTC2_INT_CONFIG */ -/* ========================================================================= */ -#define UVH_RTC2_INT_CONFIG 0x61600UL - -#define UVH_RTC2_INT_CONFIG_VECTOR_SHFT 0 -#define UVH_RTC2_INT_CONFIG_VECTOR_MASK 0x00000000000000ffUL -#define UVH_RTC2_INT_CONFIG_DM_SHFT 8 -#define UVH_RTC2_INT_CONFIG_DM_MASK 0x0000000000000700UL -#define UVH_RTC2_INT_CONFIG_DESTMODE_SHFT 11 -#define UVH_RTC2_INT_CONFIG_DESTMODE_MASK 0x0000000000000800UL -#define UVH_RTC2_INT_CONFIG_STATUS_SHFT 12 -#define UVH_RTC2_INT_CONFIG_STATUS_MASK 0x0000000000001000UL -#define UVH_RTC2_INT_CONFIG_P_SHFT 13 -#define UVH_RTC2_INT_CONFIG_P_MASK 0x0000000000002000UL -#define UVH_RTC2_INT_CONFIG_T_SHFT 15 -#define UVH_RTC2_INT_CONFIG_T_MASK 0x0000000000008000UL -#define UVH_RTC2_INT_CONFIG_M_SHFT 16 -#define UVH_RTC2_INT_CONFIG_M_MASK 0x0000000000010000UL -#define UVH_RTC2_INT_CONFIG_APIC_ID_SHFT 32 -#define UVH_RTC2_INT_CONFIG_APIC_ID_MASK 0xffffffff00000000UL - -union uvh_rtc2_int_config_u { - unsigned long v; - struct uvh_rtc2_int_config_s { - unsigned long vector_ : 8; /* RW */ - unsigned long dm : 3; /* RW */ - unsigned long destmode : 1; /* RW */ - unsigned long status : 1; /* RO */ - unsigned long p : 1; /* RO */ - unsigned long rsvd_14 : 1; /* */ - unsigned long t : 1; /* RO */ - unsigned long m : 1; /* RW */ - unsigned long rsvd_17_31: 15; /* */ - unsigned long apic_id : 32; /* RW */ - } s; -}; - -/* ========================================================================= */ -/* UVH_RTC3_INT_CONFIG */ -/* ========================================================================= */ -#define UVH_RTC3_INT_CONFIG 0x61640UL - -#define UVH_RTC3_INT_CONFIG_VECTOR_SHFT 0 -#define UVH_RTC3_INT_CONFIG_VECTOR_MASK 0x00000000000000ffUL -#define UVH_RTC3_INT_CONFIG_DM_SHFT 8 -#define UVH_RTC3_INT_CONFIG_DM_MASK 0x0000000000000700UL -#define UVH_RTC3_INT_CONFIG_DESTMODE_SHFT 11 -#define UVH_RTC3_INT_CONFIG_DESTMODE_MASK 0x0000000000000800UL -#define UVH_RTC3_INT_CONFIG_STATUS_SHFT 12 -#define UVH_RTC3_INT_CONFIG_STATUS_MASK 0x0000000000001000UL -#define UVH_RTC3_INT_CONFIG_P_SHFT 13 -#define UVH_RTC3_INT_CONFIG_P_MASK 0x0000000000002000UL -#define UVH_RTC3_INT_CONFIG_T_SHFT 15 -#define UVH_RTC3_INT_CONFIG_T_MASK 0x0000000000008000UL -#define UVH_RTC3_INT_CONFIG_M_SHFT 16 -#define UVH_RTC3_INT_CONFIG_M_MASK 0x0000000000010000UL -#define UVH_RTC3_INT_CONFIG_APIC_ID_SHFT 32 -#define UVH_RTC3_INT_CONFIG_APIC_ID_MASK 0xffffffff00000000UL - -union uvh_rtc3_int_config_u { - unsigned long v; - struct uvh_rtc3_int_config_s { - unsigned long vector_ : 8; /* RW */ - unsigned long dm : 3; /* RW */ - unsigned long destmode : 1; /* RW */ - unsigned long status : 1; /* RO */ - unsigned long p : 1; /* RO */ - unsigned long rsvd_14 : 1; /* */ - unsigned long t : 1; /* RO */ - unsigned long m : 1; /* RW */ - unsigned long rsvd_17_31: 15; /* */ - unsigned long apic_id : 32; /* RW */ - } s; -}; - -/* ========================================================================= */ -/* UVH_RTC_INC_RATIO */ -/* ========================================================================= */ -#define UVH_RTC_INC_RATIO 0x350000UL - -#define UVH_RTC_INC_RATIO_FRACTION_SHFT 0 -#define UVH_RTC_INC_RATIO_FRACTION_MASK 0x00000000000fffffUL -#define UVH_RTC_INC_RATIO_RATIO_SHFT 20 -#define UVH_RTC_INC_RATIO_RATIO_MASK 0x0000000000700000UL - -union uvh_rtc_inc_ratio_u { - unsigned long v; - struct uvh_rtc_inc_ratio_s { - unsigned long fraction : 20; /* RW */ - unsigned long ratio : 3; /* RW */ - unsigned long rsvd_23_63: 41; /* */ - } s; -}; - -/* ========================================================================= */ /* UVH_SI_ADDR_MAP_CONFIG */ /* ========================================================================= */ #define UVH_SI_ADDR_MAP_CONFIG 0xc80000UL diff --git a/arch/x86/kernel/tlb_uv.c b/arch/x86/kernel/tlb_uv.c index 364d015ef..ef68ba4 100644 --- a/arch/x86/kernel/tlb_uv.c +++ b/arch/x86/kernel/tlb_uv.c @@ -20,6 +20,8 @@ #include #include +#define UV_INTD_SOFT_ACK_TIMEOUT_PERIOD 0x000000000bUL + static struct bau_control **uv_bau_table_bases __read_mostly; static int uv_bau_retry_limit __read_mostly; @@ -478,16 +480,16 @@ static void uv_enable_timeouts(void) * To program the period, the SOFT_ACK_MODE must be off. */ mmr_image &= ~((unsigned long)1 << - UV_ENABLE_INTD_SOFT_ACK_MODE_SHIFT); + UVH_LB_BAU_MISC_CONTROL_ENABLE_INTD_SOFT_ACK_MODE_SHFT); uv_write_global_mmr64 (pnode, UVH_LB_BAU_MISC_CONTROL, mmr_image); /* * Set the 4-bit period. */ mmr_image &= ~((unsigned long)0xf << - UV_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHIFT); + UVH_LB_BAU_MISC_CONTROL_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHFT); mmr_image |= (UV_INTD_SOFT_ACK_TIMEOUT_PERIOD << - UV_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHIFT); + UVH_LB_BAU_MISC_CONTROL_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHFT); uv_write_global_mmr64 (pnode, UVH_LB_BAU_MISC_CONTROL, mmr_image); /* @@ -496,7 +498,7 @@ static void uv_enable_timeouts(void) * indicated in bits 2:0 (7 causes all of them to timeout). */ mmr_image |= ((unsigned long)1 << - UV_ENABLE_INTD_SOFT_ACK_MODE_SHIFT); + UVH_LB_BAU_MISC_CONTROL_ENABLE_INTD_SOFT_ACK_MODE_SHFT); uv_write_global_mmr64 (pnode, UVH_LB_BAU_MISC_CONTROL, mmr_image); } -- cgit v0.10.2 From 41acab8851a0408c1d5ad6c21a07456f88b54d40 Mon Sep 17 00:00:00 2001 From: Lucas De Marchi Date: Wed, 10 Mar 2010 23:37:45 -0300 Subject: sched: Implement group scheduler statistics in one struct Put all statistic fields of sched_entity in one struct, sched_statistics, and embed it into sched_entity. This change allows to memset the sched_statistics to 0 when needed (for instance when forking), avoiding bugs of non initialized fields. Signed-off-by: Lucas De Marchi Signed-off-by: Peter Zijlstra LKML-Reference: <1268275065-18542-1-git-send-email-lucas.de.marchi@gmail.com> Signed-off-by: Ingo Molnar diff --git a/include/linux/sched.h b/include/linux/sched.h index 4b1753f..8cc863d 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1127,36 +1127,8 @@ struct load_weight { unsigned long weight, inv_weight; }; -/* - * CFS stats for a schedulable entity (task, task-group etc) - * - * Current field usage histogram: - * - * 4 se->block_start - * 4 se->run_node - * 4 se->sleep_start - * 6 se->load.weight - */ -struct sched_entity { - struct load_weight load; /* for load-balancing */ - struct rb_node run_node; - struct list_head group_node; - unsigned int on_rq; - - u64 exec_start; - u64 sum_exec_runtime; - u64 vruntime; - u64 prev_sum_exec_runtime; - - u64 last_wakeup; - u64 avg_overlap; - - u64 nr_migrations; - - u64 start_runtime; - u64 avg_wakeup; - #ifdef CONFIG_SCHEDSTATS +struct sched_statistics { u64 wait_start; u64 wait_max; u64 wait_count; @@ -1188,6 +1160,30 @@ struct sched_entity { u64 nr_wakeups_affine_attempts; u64 nr_wakeups_passive; u64 nr_wakeups_idle; +}; +#endif + +struct sched_entity { + struct load_weight load; /* for load-balancing */ + struct rb_node run_node; + struct list_head group_node; + unsigned int on_rq; + + u64 exec_start; + u64 sum_exec_runtime; + u64 vruntime; + u64 prev_sum_exec_runtime; + + u64 last_wakeup; + u64 avg_overlap; + + u64 nr_migrations; + + u64 start_runtime; + u64 avg_wakeup; + +#ifdef CONFIG_SCHEDSTATS + struct sched_statistics statistics; #endif #ifdef CONFIG_FAIR_GROUP_SCHED diff --git a/kernel/sched.c b/kernel/sched.c index 2c1db81..a4aa071 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2437,15 +2437,15 @@ static int try_to_wake_up(struct task_struct *p, unsigned int state, out_activate: #endif /* CONFIG_SMP */ - schedstat_inc(p, se.nr_wakeups); + schedstat_inc(p, se.statistics.nr_wakeups); if (wake_flags & WF_SYNC) - schedstat_inc(p, se.nr_wakeups_sync); + schedstat_inc(p, se.statistics.nr_wakeups_sync); if (orig_cpu != cpu) - schedstat_inc(p, se.nr_wakeups_migrate); + schedstat_inc(p, se.statistics.nr_wakeups_migrate); if (cpu == this_cpu) - schedstat_inc(p, se.nr_wakeups_local); + schedstat_inc(p, se.statistics.nr_wakeups_local); else - schedstat_inc(p, se.nr_wakeups_remote); + schedstat_inc(p, se.statistics.nr_wakeups_remote); activate_task(rq, p, 1); success = 1; @@ -2532,36 +2532,7 @@ static void __sched_fork(struct task_struct *p) p->se.avg_wakeup = sysctl_sched_wakeup_granularity; #ifdef CONFIG_SCHEDSTATS - p->se.wait_start = 0; - p->se.wait_max = 0; - p->se.wait_count = 0; - p->se.wait_sum = 0; - - p->se.sleep_start = 0; - p->se.sleep_max = 0; - p->se.sum_sleep_runtime = 0; - - p->se.block_start = 0; - p->se.block_max = 0; - p->se.exec_max = 0; - p->se.slice_max = 0; - - p->se.nr_migrations_cold = 0; - p->se.nr_failed_migrations_affine = 0; - p->se.nr_failed_migrations_running = 0; - p->se.nr_failed_migrations_hot = 0; - p->se.nr_forced_migrations = 0; - - p->se.nr_wakeups = 0; - p->se.nr_wakeups_sync = 0; - p->se.nr_wakeups_migrate = 0; - p->se.nr_wakeups_local = 0; - p->se.nr_wakeups_remote = 0; - p->se.nr_wakeups_affine = 0; - p->se.nr_wakeups_affine_attempts = 0; - p->se.nr_wakeups_passive = 0; - p->se.nr_wakeups_idle = 0; - + memset(&p->se.statistics, 0, sizeof(p->se.statistics)); #endif INIT_LIST_HEAD(&p->rt.run_list); @@ -7910,9 +7881,9 @@ void normalize_rt_tasks(void) p->se.exec_start = 0; #ifdef CONFIG_SCHEDSTATS - p->se.wait_start = 0; - p->se.sleep_start = 0; - p->se.block_start = 0; + p->se.statistics.wait_start = 0; + p->se.statistics.sleep_start = 0; + p->se.statistics.block_start = 0; #endif if (!rt_task(p)) { diff --git a/kernel/sched_debug.c b/kernel/sched_debug.c index 67f95aa..ad9df44 100644 --- a/kernel/sched_debug.c +++ b/kernel/sched_debug.c @@ -70,16 +70,16 @@ static void print_cfs_group_stats(struct seq_file *m, int cpu, PN(se->vruntime); PN(se->sum_exec_runtime); #ifdef CONFIG_SCHEDSTATS - PN(se->wait_start); - PN(se->sleep_start); - PN(se->block_start); - PN(se->sleep_max); - PN(se->block_max); - PN(se->exec_max); - PN(se->slice_max); - PN(se->wait_max); - PN(se->wait_sum); - P(se->wait_count); + PN(se->statistics.wait_start); + PN(se->statistics.sleep_start); + PN(se->statistics.block_start); + PN(se->statistics.sleep_max); + PN(se->statistics.block_max); + PN(se->statistics.exec_max); + PN(se->statistics.slice_max); + PN(se->statistics.wait_max); + PN(se->statistics.wait_sum); + P(se->statistics.wait_count); #endif P(se->load.weight); #undef PN @@ -104,7 +104,7 @@ print_task(struct seq_file *m, struct rq *rq, struct task_struct *p) SEQ_printf(m, "%9Ld.%06ld %9Ld.%06ld %9Ld.%06ld", SPLIT_NS(p->se.vruntime), SPLIT_NS(p->se.sum_exec_runtime), - SPLIT_NS(p->se.sum_sleep_runtime)); + SPLIT_NS(p->se.statistics.sum_sleep_runtime)); #else SEQ_printf(m, "%15Ld %15Ld %15Ld.%06ld %15Ld.%06ld %15Ld.%06ld", 0LL, 0LL, 0LL, 0L, 0LL, 0L, 0LL, 0L); @@ -413,34 +413,34 @@ void proc_sched_show_task(struct task_struct *p, struct seq_file *m) nr_switches = p->nvcsw + p->nivcsw; #ifdef CONFIG_SCHEDSTATS - PN(se.wait_start); - PN(se.sleep_start); - PN(se.block_start); - PN(se.sleep_max); - PN(se.block_max); - PN(se.exec_max); - PN(se.slice_max); - PN(se.wait_max); - PN(se.wait_sum); - P(se.wait_count); - PN(se.iowait_sum); - P(se.iowait_count); + PN(se.statistics.wait_start); + PN(se.statistics.sleep_start); + PN(se.statistics.block_start); + PN(se.statistics.sleep_max); + PN(se.statistics.block_max); + PN(se.statistics.exec_max); + PN(se.statistics.slice_max); + PN(se.statistics.wait_max); + PN(se.statistics.wait_sum); + P(se.statistics.wait_count); + PN(se.statistics.iowait_sum); + P(se.statistics.iowait_count); P(sched_info.bkl_count); P(se.nr_migrations); - P(se.nr_migrations_cold); - P(se.nr_failed_migrations_affine); - P(se.nr_failed_migrations_running); - P(se.nr_failed_migrations_hot); - P(se.nr_forced_migrations); - P(se.nr_wakeups); - P(se.nr_wakeups_sync); - P(se.nr_wakeups_migrate); - P(se.nr_wakeups_local); - P(se.nr_wakeups_remote); - P(se.nr_wakeups_affine); - P(se.nr_wakeups_affine_attempts); - P(se.nr_wakeups_passive); - P(se.nr_wakeups_idle); + P(se.statistics.nr_migrations_cold); + P(se.statistics.nr_failed_migrations_affine); + P(se.statistics.nr_failed_migrations_running); + P(se.statistics.nr_failed_migrations_hot); + P(se.statistics.nr_forced_migrations); + P(se.statistics.nr_wakeups); + P(se.statistics.nr_wakeups_sync); + P(se.statistics.nr_wakeups_migrate); + P(se.statistics.nr_wakeups_local); + P(se.statistics.nr_wakeups_remote); + P(se.statistics.nr_wakeups_affine); + P(se.statistics.nr_wakeups_affine_attempts); + P(se.statistics.nr_wakeups_passive); + P(se.statistics.nr_wakeups_idle); { u64 avg_atom, avg_per_cpu; @@ -491,32 +491,7 @@ void proc_sched_show_task(struct task_struct *p, struct seq_file *m) void proc_sched_set_task(struct task_struct *p) { #ifdef CONFIG_SCHEDSTATS - p->se.wait_max = 0; - p->se.wait_sum = 0; - p->se.wait_count = 0; - p->se.iowait_sum = 0; - p->se.iowait_count = 0; - p->se.sleep_max = 0; - p->se.sum_sleep_runtime = 0; - p->se.block_max = 0; - p->se.exec_max = 0; - p->se.slice_max = 0; - p->se.nr_migrations = 0; - p->se.nr_migrations_cold = 0; - p->se.nr_failed_migrations_affine = 0; - p->se.nr_failed_migrations_running = 0; - p->se.nr_failed_migrations_hot = 0; - p->se.nr_forced_migrations = 0; - p->se.nr_wakeups = 0; - p->se.nr_wakeups_sync = 0; - p->se.nr_wakeups_migrate = 0; - p->se.nr_wakeups_local = 0; - p->se.nr_wakeups_remote = 0; - p->se.nr_wakeups_affine = 0; - p->se.nr_wakeups_affine_attempts = 0; - p->se.nr_wakeups_passive = 0; - p->se.nr_wakeups_idle = 0; - p->sched_info.bkl_count = 0; + memset(&p->se.statistics, 0, sizeof(p->se.statistics)); #endif p->se.sum_exec_runtime = 0; p->se.prev_sum_exec_runtime = 0; diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 3e1fd96..8ad164b 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -505,7 +505,8 @@ __update_curr(struct cfs_rq *cfs_rq, struct sched_entity *curr, { unsigned long delta_exec_weighted; - schedstat_set(curr->exec_max, max((u64)delta_exec, curr->exec_max)); + schedstat_set(curr->statistics.exec_max, + max((u64)delta_exec, curr->statistics.exec_max)); curr->sum_exec_runtime += delta_exec; schedstat_add(cfs_rq, exec_clock, delta_exec); @@ -548,7 +549,7 @@ static void update_curr(struct cfs_rq *cfs_rq) static inline void update_stats_wait_start(struct cfs_rq *cfs_rq, struct sched_entity *se) { - schedstat_set(se->wait_start, rq_of(cfs_rq)->clock); + schedstat_set(se->statistics.wait_start, rq_of(cfs_rq)->clock); } /* @@ -567,18 +568,18 @@ static void update_stats_enqueue(struct cfs_rq *cfs_rq, struct sched_entity *se) static void update_stats_wait_end(struct cfs_rq *cfs_rq, struct sched_entity *se) { - schedstat_set(se->wait_max, max(se->wait_max, - rq_of(cfs_rq)->clock - se->wait_start)); - schedstat_set(se->wait_count, se->wait_count + 1); - schedstat_set(se->wait_sum, se->wait_sum + - rq_of(cfs_rq)->clock - se->wait_start); + schedstat_set(se->statistics.wait_max, max(se->statistics.wait_max, + rq_of(cfs_rq)->clock - se->statistics.wait_start)); + schedstat_set(se->statistics.wait_count, se->statistics.wait_count + 1); + schedstat_set(se->statistics.wait_sum, se->statistics.wait_sum + + rq_of(cfs_rq)->clock - se->statistics.wait_start); #ifdef CONFIG_SCHEDSTATS if (entity_is_task(se)) { trace_sched_stat_wait(task_of(se), - rq_of(cfs_rq)->clock - se->wait_start); + rq_of(cfs_rq)->clock - se->statistics.wait_start); } #endif - schedstat_set(se->wait_start, 0); + schedstat_set(se->statistics.wait_start, 0); } static inline void @@ -657,39 +658,39 @@ static void enqueue_sleeper(struct cfs_rq *cfs_rq, struct sched_entity *se) if (entity_is_task(se)) tsk = task_of(se); - if (se->sleep_start) { - u64 delta = rq_of(cfs_rq)->clock - se->sleep_start; + if (se->statistics.sleep_start) { + u64 delta = rq_of(cfs_rq)->clock - se->statistics.sleep_start; if ((s64)delta < 0) delta = 0; - if (unlikely(delta > se->sleep_max)) - se->sleep_max = delta; + if (unlikely(delta > se->statistics.sleep_max)) + se->statistics.sleep_max = delta; - se->sleep_start = 0; - se->sum_sleep_runtime += delta; + se->statistics.sleep_start = 0; + se->statistics.sum_sleep_runtime += delta; if (tsk) { account_scheduler_latency(tsk, delta >> 10, 1); trace_sched_stat_sleep(tsk, delta); } } - if (se->block_start) { - u64 delta = rq_of(cfs_rq)->clock - se->block_start; + if (se->statistics.block_start) { + u64 delta = rq_of(cfs_rq)->clock - se->statistics.block_start; if ((s64)delta < 0) delta = 0; - if (unlikely(delta > se->block_max)) - se->block_max = delta; + if (unlikely(delta > se->statistics.block_max)) + se->statistics.block_max = delta; - se->block_start = 0; - se->sum_sleep_runtime += delta; + se->statistics.block_start = 0; + se->statistics.sum_sleep_runtime += delta; if (tsk) { if (tsk->in_iowait) { - se->iowait_sum += delta; - se->iowait_count++; + se->statistics.iowait_sum += delta; + se->statistics.iowait_count++; trace_sched_stat_iowait(tsk, delta); } @@ -826,9 +827,9 @@ dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int sleep) struct task_struct *tsk = task_of(se); if (tsk->state & TASK_INTERRUPTIBLE) - se->sleep_start = rq_of(cfs_rq)->clock; + se->statistics.sleep_start = rq_of(cfs_rq)->clock; if (tsk->state & TASK_UNINTERRUPTIBLE) - se->block_start = rq_of(cfs_rq)->clock; + se->statistics.block_start = rq_of(cfs_rq)->clock; } #endif } @@ -912,7 +913,7 @@ set_next_entity(struct cfs_rq *cfs_rq, struct sched_entity *se) * when there are only lesser-weight tasks around): */ if (rq_of(cfs_rq)->load.weight >= 2*se->load.weight) { - se->slice_max = max(se->slice_max, + se->statistics.slice_max = max(se->statistics.slice_max, se->sum_exec_runtime - se->prev_sum_exec_runtime); } #endif @@ -1306,7 +1307,7 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync) if (sync && balanced) return 1; - schedstat_inc(p, se.nr_wakeups_affine_attempts); + schedstat_inc(p, se.statistics.nr_wakeups_affine_attempts); tl_per_task = cpu_avg_load_per_task(this_cpu); if (balanced || @@ -1318,7 +1319,7 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync) * there is no bad imbalance. */ schedstat_inc(sd, ttwu_move_affine); - schedstat_inc(p, se.nr_wakeups_affine); + schedstat_inc(p, se.statistics.nr_wakeups_affine); return 1; } @@ -1844,13 +1845,13 @@ int can_migrate_task(struct task_struct *p, struct rq *rq, int this_cpu, * 3) are cache-hot on their current CPU. */ if (!cpumask_test_cpu(this_cpu, &p->cpus_allowed)) { - schedstat_inc(p, se.nr_failed_migrations_affine); + schedstat_inc(p, se.statistics.nr_failed_migrations_affine); return 0; } *all_pinned = 0; if (task_running(rq, p)) { - schedstat_inc(p, se.nr_failed_migrations_running); + schedstat_inc(p, se.statistics.nr_failed_migrations_running); return 0; } @@ -1866,14 +1867,14 @@ int can_migrate_task(struct task_struct *p, struct rq *rq, int this_cpu, #ifdef CONFIG_SCHEDSTATS if (tsk_cache_hot) { schedstat_inc(sd, lb_hot_gained[idle]); - schedstat_inc(p, se.nr_forced_migrations); + schedstat_inc(p, se.statistics.nr_forced_migrations); } #endif return 1; } if (tsk_cache_hot) { - schedstat_inc(p, se.nr_failed_migrations_hot); + schedstat_inc(p, se.statistics.nr_failed_migrations_hot); return 0; } return 1; diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c index c4fb42a..0335e87 100644 --- a/kernel/sched_rt.c +++ b/kernel/sched_rt.c @@ -613,7 +613,7 @@ static void update_curr_rt(struct rq *rq) if (unlikely((s64)delta_exec < 0)) delta_exec = 0; - schedstat_set(curr->se.exec_max, max(curr->se.exec_max, delta_exec)); + schedstat_set(curr->se.statistics.exec_max, max(curr->se.statistics.exec_max, delta_exec)); curr->se.sum_exec_runtime += delta_exec; account_group_exec_runtime(curr, delta_exec); -- cgit v0.10.2 From 9b33fa6ba0e2f90fdf407501db801c2511121564 Mon Sep 17 00:00:00 2001 From: "eranian@google.com" Date: Wed, 10 Mar 2010 22:26:05 -0800 Subject: perf_events: Improve task_sched_in() This patch is an optimization in perf_event_task_sched_in() to avoid scheduling the events twice in a row. Without it, the perf_disable()/perf_enable() pair is invoked twice, thereby pinned events counts while scheduling flexible events and we go throuh hw_perf_enable() twice. By encapsulating, the whole sequence into perf_disable()/perf_enable() we ensure, hw_perf_enable() is going to be invoked only once because of the refcount protection. Signed-off-by: Stephane Eranian Signed-off-by: Peter Zijlstra LKML-Reference: <1268288765-5326-1-git-send-email-eranian@google.com> Signed-off-by: Ingo Molnar diff --git a/kernel/perf_event.c b/kernel/perf_event.c index 52c69a3..3853d49 100644 --- a/kernel/perf_event.c +++ b/kernel/perf_event.c @@ -1368,6 +1368,8 @@ void perf_event_task_sched_in(struct task_struct *task) if (cpuctx->task_ctx == ctx) return; + perf_disable(); + /* * We want to keep the following priority order: * cpu pinned (that don't need to move), task pinned, @@ -1380,6 +1382,8 @@ void perf_event_task_sched_in(struct task_struct *task) ctx_sched_in(ctx, cpuctx, EVENT_FLEXIBLE); cpuctx->task_ctx = ctx; + + perf_enable(); } #define MAX_INTERRUPTS (~0ULL) -- cgit v0.10.2 From cf14c2e987ba0a09a7b09be2ecd55af0bc9c17b4 Mon Sep 17 00:00:00 2001 From: Philipp Reisner Date: Tue, 2 Feb 2010 21:03:50 +0100 Subject: drbd: --dry-run option for drbdsetup net ( drbdadm -- --dry-run connect ) Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h index 2bf3a6e..1aae724 100644 --- a/drivers/block/drbd/drbd_int.h +++ b/drivers/block/drbd/drbd_int.h @@ -443,13 +443,18 @@ struct p_rs_param_89 { char csums_alg[SHARED_SECRET_MAX]; } __packed; +enum drbd_conn_flags { + CF_WANT_LOSE = 1, + CF_DRY_RUN = 2, +}; + struct p_protocol { struct p_header head; u32 protocol; u32 after_sb_0p; u32 after_sb_1p; u32 after_sb_2p; - u32 want_lose; + u32 conn_flags; u32 two_primaries; /* Since protocol version 87 and higher. */ @@ -791,6 +796,7 @@ enum { * while this is set. */ RESIZE_PENDING, /* Size change detected locally, waiting for the response from * the peer, if it changed there as well. */ + CONN_DRY_RUN, /* Expect disconnect after resync handshake. */ }; struct drbd_bitmap; /* opaque for drbd_conf */ diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c index ab871e0..b2d347d 100644 --- a/drivers/block/drbd/drbd_main.c +++ b/drivers/block/drbd/drbd_main.c @@ -1668,7 +1668,7 @@ int drbd_send_sync_param(struct drbd_conf *mdev, struct syncer_conf *sc) int drbd_send_protocol(struct drbd_conf *mdev) { struct p_protocol *p; - int size, rv; + int size, cf, rv; size = sizeof(struct p_protocol); @@ -1685,9 +1685,21 @@ int drbd_send_protocol(struct drbd_conf *mdev) p->after_sb_0p = cpu_to_be32(mdev->net_conf->after_sb_0p); p->after_sb_1p = cpu_to_be32(mdev->net_conf->after_sb_1p); p->after_sb_2p = cpu_to_be32(mdev->net_conf->after_sb_2p); - p->want_lose = cpu_to_be32(mdev->net_conf->want_lose); p->two_primaries = cpu_to_be32(mdev->net_conf->two_primaries); + cf = 0; + if (mdev->net_conf->want_lose) + cf |= CF_WANT_LOSE; + if (mdev->net_conf->dry_run) { + if (mdev->agreed_pro_version >= 92) + cf |= CF_DRY_RUN; + else { + dev_err(DEV, "--dry-run is not supported by peer"); + return 0; + } + } + p->conn_flags = cpu_to_be32(cf); + if (mdev->agreed_pro_version >= 87) strcpy(p->integrity_alg, mdev->net_conf->integrity_alg); diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c index d065c64..8bcde4a 100644 --- a/drivers/block/drbd/drbd_receiver.c +++ b/drivers/block/drbd/drbd_receiver.c @@ -2538,6 +2538,16 @@ static enum drbd_conns drbd_sync_handshake(struct drbd_conf *mdev, enum drbd_rol } } + if (mdev->net_conf->dry_run || test_bit(CONN_DRY_RUN, &mdev->flags)) { + if (hg == 0) + dev_info(DEV, "dry-run connect: No resync, would become Connected immediately.\n"); + else + dev_info(DEV, "dry-run connect: Would become %s, doing a %s resync.", + drbd_conn_str(hg > 0 ? C_SYNC_SOURCE : C_SYNC_TARGET), + abs(hg) >= 2 ? "full" : "bit-map based"); + return C_MASK; + } + if (abs(hg) >= 2) { dev_info(DEV, "Writing the whole bitmap, full sync required after drbd_sync_handshake.\n"); if (drbd_bitmap_io(mdev, &drbd_bmio_set_n_write, "set_n_write from sync_handshake")) @@ -2585,7 +2595,7 @@ static int receive_protocol(struct drbd_conf *mdev, struct p_header *h) struct p_protocol *p = (struct p_protocol *)h; int header_size, data_size; int p_proto, p_after_sb_0p, p_after_sb_1p, p_after_sb_2p; - int p_want_lose, p_two_primaries; + int p_want_lose, p_two_primaries, cf; char p_integrity_alg[SHARED_SECRET_MAX] = ""; header_size = sizeof(*p) - sizeof(*h); @@ -2598,8 +2608,14 @@ static int receive_protocol(struct drbd_conf *mdev, struct p_header *h) p_after_sb_0p = be32_to_cpu(p->after_sb_0p); p_after_sb_1p = be32_to_cpu(p->after_sb_1p); p_after_sb_2p = be32_to_cpu(p->after_sb_2p); - p_want_lose = be32_to_cpu(p->want_lose); p_two_primaries = be32_to_cpu(p->two_primaries); + cf = be32_to_cpu(p->conn_flags); + p_want_lose = cf & CF_WANT_LOSE; + + clear_bit(CONN_DRY_RUN, &mdev->flags); + + if (cf & CF_DRY_RUN) + set_bit(CONN_DRY_RUN, &mdev->flags); if (p_proto != mdev->net_conf->wire_protocol) { dev_err(DEV, "incompatible communication protocols\n"); @@ -3125,6 +3141,8 @@ static int receive_state(struct drbd_conf *mdev, struct p_header *h) dev_err(DEV, "Disk attach process on the peer node was aborted.\n"); peer_state.disk = D_DISKLESS; } else { + if (test_and_clear_bit(CONN_DRY_RUN, &mdev->flags)) + return FALSE; D_ASSERT(oconn == C_WF_REPORT_PARAMS); drbd_force_state(mdev, NS(conn, C_DISCONNECTING)); return FALSE; diff --git a/include/linux/drbd.h b/include/linux/drbd.h index 7896227..4341b1a 100644 --- a/include/linux/drbd.h +++ b/include/linux/drbd.h @@ -56,7 +56,7 @@ extern const char *drbd_buildtag(void); #define REL_VERSION "8.3.7" #define API_VERSION 88 #define PRO_VERSION_MIN 86 -#define PRO_VERSION_MAX 91 +#define PRO_VERSION_MAX 92 enum drbd_io_error_p { diff --git a/include/linux/drbd_nl.h b/include/linux/drbd_nl.h index a4d82f8..b41050e 100644 --- a/include/linux/drbd_nl.h +++ b/include/linux/drbd_nl.h @@ -63,6 +63,7 @@ NL_PACKET(net_conf, 5, NL_BIT( 41, T_MAY_IGNORE, always_asbp) NL_BIT( 61, T_MAY_IGNORE, no_cork) NL_BIT( 62, T_MANDATORY, auto_sndbuf_size) + NL_BIT( 70, T_MANDATORY, dry_run) ) NL_PACKET(disconnect, 6, ) -- cgit v0.10.2 From 4aa83b7bf122106669346eef40632289f540653f Mon Sep 17 00:00:00 2001 From: Lars Ellenberg Date: Fri, 26 Feb 2010 16:53:24 +0100 Subject: drbd: fix NULL pointer dereference on 4k hard sect size we still don't support 4k 'physical' sectors 'natively', but use a read-modify-write workaround. And we even tried to use the extra page before we allocated it :( Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c index 4df3b40..d53d36c 100644 --- a/drivers/block/drbd/drbd_nl.c +++ b/drivers/block/drbd/drbd_nl.c @@ -941,6 +941,25 @@ static int drbd_nl_disk_conf(struct drbd_conf *mdev, struct drbd_nl_cfg_req *nlp drbd_md_set_sector_offsets(mdev, nbc); + /* allocate a second IO page if logical_block_size != 512 */ + logical_block_size = bdev_logical_block_size(nbc->md_bdev); + if (logical_block_size == 0) + logical_block_size = MD_SECTOR_SIZE; + + if (logical_block_size != MD_SECTOR_SIZE) { + if (!mdev->md_io_tmpp) { + struct page *page = alloc_page(GFP_NOIO); + if (!page) + goto force_diskless_dec; + + dev_warn(DEV, "Meta data's bdev logical_block_size = %d != %d\n", + logical_block_size, MD_SECTOR_SIZE); + dev_warn(DEV, "Workaround engaged (has performance impact).\n"); + + mdev->md_io_tmpp = page; + } + } + if (!mdev->bitmap) { if (drbd_bm_init(mdev)) { retcode = ERR_NOMEM; @@ -980,25 +999,6 @@ static int drbd_nl_disk_conf(struct drbd_conf *mdev, struct drbd_nl_cfg_req *nlp goto force_diskless_dec; } - /* allocate a second IO page if logical_block_size != 512 */ - logical_block_size = bdev_logical_block_size(nbc->md_bdev); - if (logical_block_size == 0) - logical_block_size = MD_SECTOR_SIZE; - - if (logical_block_size != MD_SECTOR_SIZE) { - if (!mdev->md_io_tmpp) { - struct page *page = alloc_page(GFP_NOIO); - if (!page) - goto force_diskless_dec; - - dev_warn(DEV, "Meta data's bdev logical_block_size = %d != %d\n", - logical_block_size, MD_SECTOR_SIZE); - dev_warn(DEV, "Workaround engaged (has performance impact).\n"); - - mdev->md_io_tmpp = page; - } - } - /* Reset the "barriers don't work" bits here, then force meta data to * be written, to ensure we determine if barriers are supported. */ if (nbc->dc.no_md_flush) -- cgit v0.10.2 From 580b9767dbdf2c049c4d05330c70ea786ef01016 Mon Sep 17 00:00:00 2001 From: Lars Ellenberg Date: Fri, 26 Feb 2010 23:15:23 +0100 Subject: drbd: fix broken state change after split-brain attach while connected Situation: we have diverging data sets, i.e. we had a split brain somewhen, but currently are connected, one node diskless. Then we try to attach that disk, figure it is consistent, but has a diverging data set, we refuse to attach. This led to strange state changes: 22:18:35 bb drbd1: peer( Unknown -> Primary ) conn( WFReportParams -> Connected) pdsk( DUnknown -> UpToDate ) 22:19:30 bb drbd1: disk( Diskless -> Attaching ) 22:19:30 bb drbd1: disk( Attaching -> Negotiating ) 22:19:30 bb drbd1: drbd_sync_handshake: 22:19:30 bb drbd1: self 97BF25798B9D5222:F33D1F62ADE698DD:4269796F9D027C83:AC45D8B5C3C1BF93 bits:19449 flags:0 22:19:30 bb drbd1: peer 280DFB6E125465D3:F33D1F62ADE698DC:4269796F9D027C82:AC45D8B5C3C1BF93 bits:2575806 flags:0 22:19:30 bb drbd1: uuid_compare()=100 by rule 90 22:19:30 bb drbd1: Split-Brain detected, dropping connection! 22:19:30 bb drbd1: disk( Negotiating -> Diskless ) while the other side says: 22:19:30 aa drbd1: Split-Brain detected, dropping connection! 22:19:30 aa drbd1: Disk attach process on the peer node was aborted. 22:19:30 aa drbd1: conn( Connected -> TOO_LARGE ) pdsk( Diskless -> Consistent ) This should be fixed now. Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c index 8bcde4a..41f36a9 100644 --- a/drivers/block/drbd/drbd_receiver.c +++ b/drivers/block/drbd/drbd_receiver.c @@ -2513,6 +2513,10 @@ static enum drbd_conns drbd_sync_handshake(struct drbd_conf *mdev, enum drbd_rol } if (hg == -100) { + /* FIXME this log message is not correct if we end up here + * after an attempted attach on a diskless node. + * We just refuse to attach -- well, we drop the "connection" + * to that disk, in a way... */ dev_alert(DEV, "Split-Brain detected, dropping connection!\n"); drbd_khelper(mdev, "split-brain"); return C_MASK; @@ -3134,12 +3138,13 @@ static int receive_state(struct drbd_conf *mdev, struct p_header *h) put_ldev(mdev); if (nconn == C_MASK) { + nconn = C_CONNECTED; if (mdev->state.disk == D_NEGOTIATING) { drbd_force_state(mdev, NS(disk, D_DISKLESS)); - nconn = C_CONNECTED; } else if (peer_state.disk == D_NEGOTIATING) { dev_err(DEV, "Disk attach process on the peer node was aborted.\n"); peer_state.disk = D_DISKLESS; + real_peer_disk = D_DISKLESS; } else { if (test_and_clear_bit(CONN_DRY_RUN, &mdev->flags)) return FALSE; -- cgit v0.10.2 From 676396d545350a70d922605ec23c2ed26124334a Mon Sep 17 00:00:00 2001 From: Lars Ellenberg Date: Wed, 3 Mar 2010 02:08:22 +0100 Subject: fix unit of rs_same_csums accounting Depending on resync request size, we need to account for more than one bit. Impact: cosmetic If SyncTarget reported correctly 100% equal checksums, the SyncSource usually reported 12% equal checksums instead, because it only counted requests, we typically do 32k resync requests, and the bitmap granularity is still 4k. Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg diff --git a/drivers/block/drbd/drbd_worker.c b/drivers/block/drbd/drbd_worker.c index b453c2b..d97a811 100644 --- a/drivers/block/drbd/drbd_worker.c +++ b/drivers/block/drbd/drbd_worker.c @@ -938,7 +938,8 @@ int w_e_end_csum_rs_req(struct drbd_conf *mdev, struct drbd_work *w, int cancel) if (eq) { drbd_set_in_sync(mdev, e->sector, e->size); - mdev->rs_same_csum++; + /* rs_same_csums unit is BM_BLOCK_SIZE */ + mdev->rs_same_csum += e->size >> BM_BLOCK_SHIFT; ok = drbd_send_ack(mdev, P_RS_IS_IN_SYNC, e); } else { inc_rs_pending(mdev); -- cgit v0.10.2 From 4589d7f829951c1713ef5a4ad1a9bb563da329b5 Mon Sep 17 00:00:00 2001 From: Lars Ellenberg Date: Wed, 3 Mar 2010 02:25:33 +0100 Subject: drbd_disconnect: grab meta.socket mutex as well Fixes a race and potential kernel panic if e.g. the worker was just about to send a few P_RS_IS_IN_SYNC via the meta socket for checksum based resync, while the receiver destroys the sockets in drbd_disconnect. To make sure no-one is using the meta socket, it is not enough to stop the asender... Grab the meta socket mutex before destroying it. Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c index b2d347d..67e0fc5 100644 --- a/drivers/block/drbd/drbd_main.c +++ b/drivers/block/drbd/drbd_main.c @@ -3173,14 +3173,18 @@ void drbd_free_bc(struct drbd_backing_dev *ldev) void drbd_free_sock(struct drbd_conf *mdev) { if (mdev->data.socket) { + mutex_lock(&mdev->data.mutex); kernel_sock_shutdown(mdev->data.socket, SHUT_RDWR); sock_release(mdev->data.socket); mdev->data.socket = NULL; + mutex_unlock(&mdev->data.mutex); } if (mdev->meta.socket) { + mutex_lock(&mdev->meta.mutex); kernel_sock_shutdown(mdev->meta.socket, SHUT_RDWR); sock_release(mdev->meta.socket); mdev->meta.socket = NULL; + mutex_unlock(&mdev->meta.mutex); } } diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c index 41f36a9..d803e6c 100644 --- a/drivers/block/drbd/drbd_receiver.c +++ b/drivers/block/drbd/drbd_receiver.c @@ -3617,10 +3617,7 @@ static void drbd_disconnect(struct drbd_conf *mdev) /* asender does not clean up anything. it must not interfere, either */ drbd_thread_stop(&mdev->asender); - - mutex_lock(&mdev->data.mutex); drbd_free_sock(mdev); - mutex_unlock(&mdev->data.mutex); spin_lock_irq(&mdev->req_lock); _drbd_wait_ee_list_empty(mdev, &mdev->active_ee); -- cgit v0.10.2 From c42b6cf4b38c9726d4b46c48d04197c9ca74d773 Mon Sep 17 00:00:00 2001 From: Lars Ellenberg Date: Wed, 3 Mar 2010 02:44:11 +0100 Subject: drbd: add missing drbd command names to avoid in error messages cmdname() should map command number to its human readable representation. The string table was incomplete, though. Maybe rather do a switch() block, and let the compiler help us to keep it complete? Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h index 1aae724..844206c 100644 --- a/drivers/block/drbd/drbd_int.h +++ b/drivers/block/drbd/drbd_int.h @@ -261,6 +261,9 @@ static inline const char *cmdname(enum drbd_packets cmd) [P_OV_REQUEST] = "OVRequest", [P_OV_REPLY] = "OVReply", [P_OV_RESULT] = "OVResult", + [P_CSUM_RS_REQUEST] = "CsumRSRequest", + [P_RS_IS_IN_SYNC] = "CsumRSIsInSync", + [P_COMPRESSED_BITMAP] = "CBitmap", [P_MAX_CMD] = NULL, }; -- cgit v0.10.2 From 309d1608cce32903d67d47e7545e232c400b6aa0 Mon Sep 17 00:00:00 2001 From: Philipp Reisner Date: Tue, 2 Mar 2010 15:03:44 +0100 Subject: drbd: Reduce the time an empty resync takes usually This mitigates changes introduced with commit: http://git.drbd.org/?p=drbd-8.3.git;a=commit;h=4b6803a3276652da3737 Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h index 844206c..2d5cebb 100644 --- a/drivers/block/drbd/drbd_int.h +++ b/drivers/block/drbd/drbd_int.h @@ -800,6 +800,7 @@ enum { RESIZE_PENDING, /* Size change detected locally, waiting for the response from * the peer, if it changed there as well. */ CONN_DRY_RUN, /* Expect disconnect after resync handshake. */ + GOT_PING_ACK, /* set when we receive a ping_ack packet, misc wait gets woken */ }; struct drbd_bitmap; /* opaque for drbd_conf */ diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c index d803e6c..ed9f1de 100644 --- a/drivers/block/drbd/drbd_receiver.c +++ b/drivers/block/drbd/drbd_receiver.c @@ -4074,6 +4074,8 @@ static int got_PingAck(struct drbd_conf *mdev, struct p_header *h) { /* restore idle timeout */ mdev->meta.socket->sk->sk_rcvtimeo = mdev->net_conf->ping_int*HZ; + if (!test_and_set_bit(GOT_PING_ACK, &mdev->flags)) + wake_up(&mdev->misc_wait); return TRUE; } diff --git a/drivers/block/drbd/drbd_worker.c b/drivers/block/drbd/drbd_worker.c index d97a811..4672f2f 100644 --- a/drivers/block/drbd/drbd_worker.c +++ b/drivers/block/drbd/drbd_worker.c @@ -1289,6 +1289,14 @@ int drbd_alter_sa(struct drbd_conf *mdev, int na) return retcode; } +static void ping_peer(struct drbd_conf *mdev) +{ + clear_bit(GOT_PING_ACK, &mdev->flags); + request_ping(mdev); + wait_event(mdev->misc_wait, + test_bit(GOT_PING_ACK, &mdev->flags) || mdev->state.conn < C_CONNECTED); +} + /** * drbd_start_resync() - Start the resync process * @mdev: DRBD device. @@ -1383,9 +1391,7 @@ void drbd_start_resync(struct drbd_conf *mdev, enum drbd_conns side) if (mdev->rs_total == 0) { /* Peer still reachable? Beware of failing before-resync-target handlers! */ - request_ping(mdev); - __set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(mdev->net_conf->ping_timeo*HZ/9); /* 9 instead 10 */ + ping_peer(mdev); drbd_resync_finished(mdev); return; } -- cgit v0.10.2 From d0c3f60f3611ceac9b1e4fdffd1497337568e7cb Mon Sep 17 00:00:00 2001 From: Philipp Reisner Date: Tue, 2 Mar 2010 15:06:45 +0100 Subject: drbd: Make sure we do not send state updates during an empty resync [Bugz 271] This is a race condition that existed for ages. The previous commit reduces the window, this one closes it. Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg diff --git a/drivers/block/drbd/drbd_worker.c b/drivers/block/drbd/drbd_worker.c index 4672f2f..44bf6d1 100644 --- a/drivers/block/drbd/drbd_worker.c +++ b/drivers/block/drbd/drbd_worker.c @@ -1380,7 +1380,6 @@ void drbd_start_resync(struct drbd_conf *mdev, enum drbd_conns side) _drbd_pause_after(mdev); } write_unlock_irq(&global_state_lock); - drbd_state_unlock(mdev); put_ldev(mdev); if (r == SS_SUCCESS) { @@ -1393,7 +1392,6 @@ void drbd_start_resync(struct drbd_conf *mdev, enum drbd_conns side) /* Peer still reachable? Beware of failing before-resync-target handlers! */ ping_peer(mdev); drbd_resync_finished(mdev); - return; } /* ns.conn may already be != mdev->state.conn, @@ -1405,6 +1403,7 @@ void drbd_start_resync(struct drbd_conf *mdev, enum drbd_conns side) drbd_md_sync(mdev); } + drbd_state_unlock(mdev); } int drbd_worker(struct drbd_thread *thi) -- cgit v0.10.2 From d10a33c68b8526d95ef6ee72b371c392d48df4d3 Mon Sep 17 00:00:00 2001 From: Philipp Reisner Date: Thu, 4 Mar 2010 15:11:39 +0100 Subject: drbd: Forcing primary should also work for Consistent disks [Bugz 266] Up to now this only worked for Outdated and Inconsistent disks, that it did not worked for Consistent disks was an inconsistent omission. Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c index d53d36c..6492e32 100644 --- a/drivers/block/drbd/drbd_nl.c +++ b/drivers/block/drbd/drbd_nl.c @@ -285,8 +285,8 @@ int drbd_set_role(struct drbd_conf *mdev, enum drbd_role new_role, int force) } if (r == SS_NO_UP_TO_DATE_DISK && force && - (mdev->state.disk == D_INCONSISTENT || - mdev->state.disk == D_OUTDATED)) { + (mdev->state.disk < D_UP_TO_DATE && + mdev->state.disk >= D_INCONSISTENT)) { mask.disk = D_MASK; val.disk = D_UP_TO_DATE; forced = 1; -- cgit v0.10.2 From 1f55243024087b56aef0b1e6d9c0ea89c76f0a6b Mon Sep 17 00:00:00 2001 From: Philipp Reisner Date: Thu, 4 Mar 2010 15:51:01 +0100 Subject: drbd: Renamed overwrite_peer to primary_force Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg diff --git a/drivers/block/drbd/drbd_nl.c b/drivers/block/drbd/drbd_nl.c index 6492e32..6429d2b 100644 --- a/drivers/block/drbd/drbd_nl.c +++ b/drivers/block/drbd/drbd_nl.c @@ -407,7 +407,7 @@ static int drbd_nl_primary(struct drbd_conf *mdev, struct drbd_nl_cfg_req *nlp, } reply->ret_code = - drbd_set_role(mdev, R_PRIMARY, primary_args.overwrite_peer); + drbd_set_role(mdev, R_PRIMARY, primary_args.primary_force); return 0; } diff --git a/include/linux/drbd_nl.h b/include/linux/drbd_nl.h index b41050e..f7431a4 100644 --- a/include/linux/drbd_nl.h +++ b/include/linux/drbd_nl.h @@ -12,7 +12,7 @@ #endif NL_PACKET(primary, 1, - NL_BIT( 1, T_MAY_IGNORE, overwrite_peer) + NL_BIT( 1, T_MAY_IGNORE, primary_force) ) NL_PACKET(secondary, 2, ) -- cgit v0.10.2 From 39ad2bbb5900d1bc9ae8f06cebb4cb2529d9e42e Mon Sep 17 00:00:00 2001 From: Lars Ellenberg Date: Thu, 4 Mar 2010 15:52:30 +0100 Subject: drbd: fix al-to-on-disk-bitmap for 4k logical_block_size Up to now, applying the in-core activity-log to the on-disk bitmap did not care for logical_block_size. On logical_block_size != 512 byte, this very likely results in misalligned block access and spurious "io errors". We now simply always submit aligned whole 4k blocks, fixing this for logical block sizes of 512, 1024, 2048 and 4096. For even larger logical block sizes, this won't work. But I'm not aware of devices with such properties being available. Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg diff --git a/drivers/block/drbd/drbd_actlog.c b/drivers/block/drbd/drbd_actlog.c index 17956ff..43e57f3 100644 --- a/drivers/block/drbd/drbd_actlog.c +++ b/drivers/block/drbd/drbd_actlog.c @@ -536,7 +536,9 @@ static void atodb_endio(struct bio *bio, int error) put_ldev(mdev); } +/* sector to word */ #define S2W(s) ((s)<<(BM_EXT_SHIFT-BM_BLOCK_SHIFT-LN2_BPL)) + /* activity log to on disk bitmap -- prepare bio unless that sector * is already covered by previously prepared bios */ static int atodb_prepare_unless_covered(struct drbd_conf *mdev, @@ -546,13 +548,20 @@ static int atodb_prepare_unless_covered(struct drbd_conf *mdev, { struct bio *bio; struct page *page; - sector_t on_disk_sector = enr + mdev->ldev->md.md_offset - + mdev->ldev->md.bm_offset; + sector_t on_disk_sector; unsigned int page_offset = PAGE_SIZE; int offset; int i = 0; int err = -ENOMEM; + /* We always write aligned, full 4k blocks, + * so we can ignore the logical_block_size (for now) */ + enr &= ~7U; + on_disk_sector = enr + mdev->ldev->md.md_offset + + mdev->ldev->md.bm_offset; + + D_ASSERT(!(on_disk_sector & 7U)); + /* Check if that enr is already covered by an already created bio. * Caution, bios[] is not NULL terminated, * but only initialized to all NULL. @@ -588,7 +597,7 @@ static int atodb_prepare_unless_covered(struct drbd_conf *mdev, offset = S2W(enr); drbd_bm_get_lel(mdev, offset, - min_t(size_t, S2W(1), drbd_bm_words(mdev) - offset), + min_t(size_t, S2W(8), drbd_bm_words(mdev) - offset), kmap(page) + page_offset); kunmap(page); @@ -597,7 +606,7 @@ static int atodb_prepare_unless_covered(struct drbd_conf *mdev, bio->bi_bdev = mdev->ldev->md_bdev; bio->bi_sector = on_disk_sector; - if (bio_add_page(bio, page, MD_SECTOR_SIZE, page_offset) != MD_SECTOR_SIZE) + if (bio_add_page(bio, page, 4096, page_offset) != 4096) goto out_put_page; atomic_inc(&wc->count); -- cgit v0.10.2 From 39c0cbe2150cbd848a25ba6cdb271d1ad46818ad Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Thu, 11 Mar 2010 17:17:13 +0100 Subject: sched: Rate-limit nohz Entering nohz code on every micro-idle is costing ~10% throughput for netperf TCP_RR when scheduling cross-cpu. Rate limiting entry fixes this, but raises ticks a bit. On my Q6600, an idle box goes from ~85 interrupts/sec to 128. The higher the context switch rate, the more nohz entry costs. With this patch and some cycle recovery patches in my tree, max cross cpu context switch rate is improved by ~16%, a large portion of which of which is this ratelimiting. Signed-off-by: Mike Galbraith Signed-off-by: Peter Zijlstra LKML-Reference: <1268301003.6785.28.camel@marge.simson.net> Signed-off-by: Ingo Molnar diff --git a/include/linux/sched.h b/include/linux/sched.h index 8cc863d..13efe7d 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -271,11 +271,17 @@ extern cpumask_var_t nohz_cpu_mask; #if defined(CONFIG_SMP) && defined(CONFIG_NO_HZ) extern int select_nohz_load_balancer(int cpu); extern int get_nohz_load_balancer(void); +extern int nohz_ratelimit(int cpu); #else static inline int select_nohz_load_balancer(int cpu) { return 0; } + +static inline int nohz_ratelimit(int cpu) +{ + return 0; +} #endif /* diff --git a/kernel/sched.c b/kernel/sched.c index a4aa071..60b1bbe 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -492,6 +492,7 @@ struct rq { #define CPU_LOAD_IDX_MAX 5 unsigned long cpu_load[CPU_LOAD_IDX_MAX]; #ifdef CONFIG_NO_HZ + u64 nohz_stamp; unsigned char in_nohz_recently; #endif /* capture load from *all* tasks on this cpu: */ @@ -1228,6 +1229,17 @@ void wake_up_idle_cpu(int cpu) if (!tsk_is_polling(rq->idle)) smp_send_reschedule(cpu); } + +int nohz_ratelimit(int cpu) +{ + struct rq *rq = cpu_rq(cpu); + u64 diff = rq->clock - rq->nohz_stamp; + + rq->nohz_stamp = rq->clock; + + return diff < (NSEC_PER_SEC / HZ) >> 1; +} + #endif /* CONFIG_NO_HZ */ static u64 sched_avg_period(void) diff --git a/kernel/time/tick-sched.c b/kernel/time/tick-sched.c index f992762..f25735a 100644 --- a/kernel/time/tick-sched.c +++ b/kernel/time/tick-sched.c @@ -262,6 +262,9 @@ void tick_nohz_stop_sched_tick(int inidle) goto end; } + if (nohz_ratelimit(cpu)) + goto end; + ts->idle_calls++; /* Read jiffies and the time when jiffies were updated last */ do { -- cgit v0.10.2 From b42e0c41a422a212ddea0666d5a3a0e3c35206db Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Thu, 11 Mar 2010 17:15:38 +0100 Subject: sched: Remove avg_wakeup Testing the load which led to this heuristic (nfs4 kbuild) shows that it has outlived it's usefullness. With intervening load balancing changes, I cannot see any difference with/without, so recover there fastpath cycles. Signed-off-by: Mike Galbraith Signed-off-by: Peter Zijlstra LKML-Reference: <1268301062.6785.29.camel@marge.simson.net> Signed-off-by: Ingo Molnar diff --git a/include/linux/sched.h b/include/linux/sched.h index 13efe7d..70c560f 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1185,9 +1185,6 @@ struct sched_entity { u64 nr_migrations; - u64 start_runtime; - u64 avg_wakeup; - #ifdef CONFIG_SCHEDSTATS struct sched_statistics statistics; #endif diff --git a/kernel/sched.c b/kernel/sched.c index 60b1bbe..35a8626 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -1880,9 +1880,6 @@ static void update_avg(u64 *avg, u64 sample) static void enqueue_task(struct rq *rq, struct task_struct *p, int wakeup, bool head) { - if (wakeup) - p->se.start_runtime = p->se.sum_exec_runtime; - sched_info_queued(p); p->sched_class->enqueue_task(rq, p, wakeup, head); p->se.on_rq = 1; @@ -1890,17 +1887,11 @@ enqueue_task(struct rq *rq, struct task_struct *p, int wakeup, bool head) static void dequeue_task(struct rq *rq, struct task_struct *p, int sleep) { - if (sleep) { - if (p->se.last_wakeup) { - update_avg(&p->se.avg_overlap, - p->se.sum_exec_runtime - p->se.last_wakeup); - p->se.last_wakeup = 0; - } else { - update_avg(&p->se.avg_wakeup, - sysctl_sched_wakeup_granularity); - } + if (sleep && p->se.last_wakeup) { + update_avg(&p->se.avg_overlap, + p->se.sum_exec_runtime - p->se.last_wakeup); + p->se.last_wakeup = 0; } - sched_info_dequeued(p); p->sched_class->dequeue_task(rq, p, sleep); p->se.on_rq = 0; @@ -2466,13 +2457,6 @@ out_activate: */ if (!in_interrupt()) { struct sched_entity *se = ¤t->se; - u64 sample = se->sum_exec_runtime; - - if (se->last_wakeup) - sample -= se->last_wakeup; - else - sample -= se->start_runtime; - update_avg(&se->avg_wakeup, sample); se->last_wakeup = se->sum_exec_runtime; } @@ -2540,8 +2524,6 @@ static void __sched_fork(struct task_struct *p) p->se.nr_migrations = 0; p->se.last_wakeup = 0; p->se.avg_overlap = 0; - p->se.start_runtime = 0; - p->se.avg_wakeup = sysctl_sched_wakeup_granularity; #ifdef CONFIG_SCHEDSTATS memset(&p->se.statistics, 0, sizeof(p->se.statistics)); diff --git a/kernel/sched_debug.c b/kernel/sched_debug.c index ad9df44..20b95a4 100644 --- a/kernel/sched_debug.c +++ b/kernel/sched_debug.c @@ -408,7 +408,6 @@ void proc_sched_show_task(struct task_struct *p, struct seq_file *m) PN(se.vruntime); PN(se.sum_exec_runtime); PN(se.avg_overlap); - PN(se.avg_wakeup); nr_switches = p->nvcsw + p->nivcsw; diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 8ad164b..6fc6285 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -1592,42 +1592,11 @@ static int select_task_rq_fair(struct task_struct *p, int sd_flag, int wake_flag } #endif /* CONFIG_SMP */ -/* - * Adaptive granularity - * - * se->avg_wakeup gives the average time a task runs until it does a wakeup, - * with the limit of wakeup_gran -- when it never does a wakeup. - * - * So the smaller avg_wakeup is the faster we want this task to preempt, - * but we don't want to treat the preemptee unfairly and therefore allow it - * to run for at least the amount of time we'd like to run. - * - * NOTE: we use 2*avg_wakeup to increase the probability of actually doing one - * - * NOTE: we use *nr_running to scale with load, this nicely matches the - * degrading latency on load. - */ -static unsigned long -adaptive_gran(struct sched_entity *curr, struct sched_entity *se) -{ - u64 this_run = curr->sum_exec_runtime - curr->prev_sum_exec_runtime; - u64 expected_wakeup = 2*se->avg_wakeup * cfs_rq_of(se)->nr_running; - u64 gran = 0; - - if (this_run < expected_wakeup) - gran = expected_wakeup - this_run; - - return min_t(s64, gran, sysctl_sched_wakeup_granularity); -} - static unsigned long wakeup_gran(struct sched_entity *curr, struct sched_entity *se) { unsigned long gran = sysctl_sched_wakeup_granularity; - if (cfs_rq_of(curr)->curr && sched_feat(ADAPTIVE_GRAN)) - gran = adaptive_gran(curr, se); - /* * Since its curr running now, convert the gran from real-time * to virtual-time in his units. diff --git a/kernel/sched_features.h b/kernel/sched_features.h index d5059fd..96ef5db 100644 --- a/kernel/sched_features.h +++ b/kernel/sched_features.h @@ -31,12 +31,6 @@ SCHED_FEAT(START_DEBIT, 1) SCHED_FEAT(WAKEUP_PREEMPT, 1) /* - * Compute wakeup_gran based on task behaviour, clipped to - * [0, sched_wakeup_gran_ns] - */ -SCHED_FEAT(ADAPTIVE_GRAN, 1) - -/* * When converting the wakeup granularity to virtual time, do it such * that heavier tasks preempting a lighter task have an edge. */ -- cgit v0.10.2 From e12f31d3e5d36328c7fbd0fce40a95e70b59152c Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Thu, 11 Mar 2010 17:15:51 +0100 Subject: sched: Remove avg_overlap Both avg_overlap and avg_wakeup had an inherent problem in that their accuracy was detrimentally affected by cross-cpu wakeups, this because we are missing the necessary call to update_curr(). This can't be fixed without increasing overhead in our already too fat fastpath. Additionally, with recent load balancing changes making us prefer to place tasks in an idle cache domain (which is good for compute bound loads), communicating tasks suffer when a sync wakeup, which would enable affine placement, is turned into a non-sync wakeup by SYNC_LESS. With one task on the runqueue, wake_affine() rejects the affine wakeup request, leaving the unfortunate where placed, taking frequent cache misses. Remove it, and recover some fastpath cycles. Signed-off-by: Mike Galbraith Signed-off-by: Peter Zijlstra LKML-Reference: <1268301121.6785.30.camel@marge.simson.net> Signed-off-by: Ingo Molnar diff --git a/include/linux/sched.h b/include/linux/sched.h index 70c560f..8604884 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1180,9 +1180,6 @@ struct sched_entity { u64 vruntime; u64 prev_sum_exec_runtime; - u64 last_wakeup; - u64 avg_overlap; - u64 nr_migrations; #ifdef CONFIG_SCHEDSTATS diff --git a/kernel/sched.c b/kernel/sched.c index 35a8626..68ed6f4 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -1887,11 +1887,6 @@ enqueue_task(struct rq *rq, struct task_struct *p, int wakeup, bool head) static void dequeue_task(struct rq *rq, struct task_struct *p, int sleep) { - if (sleep && p->se.last_wakeup) { - update_avg(&p->se.avg_overlap, - p->se.sum_exec_runtime - p->se.last_wakeup); - p->se.last_wakeup = 0; - } sched_info_dequeued(p); p->sched_class->dequeue_task(rq, p, sleep); p->se.on_rq = 0; @@ -2452,15 +2447,6 @@ out_activate: activate_task(rq, p, 1); success = 1; - /* - * Only attribute actual wakeups done by this task. - */ - if (!in_interrupt()) { - struct sched_entity *se = ¤t->se; - - se->last_wakeup = se->sum_exec_runtime; - } - out_running: trace_sched_wakeup(rq, p, success); check_preempt_curr(rq, p, wake_flags); @@ -2522,8 +2508,6 @@ static void __sched_fork(struct task_struct *p) p->se.sum_exec_runtime = 0; p->se.prev_sum_exec_runtime = 0; p->se.nr_migrations = 0; - p->se.last_wakeup = 0; - p->se.avg_overlap = 0; #ifdef CONFIG_SCHEDSTATS memset(&p->se.statistics, 0, sizeof(p->se.statistics)); @@ -3594,23 +3578,6 @@ static inline void schedule_debug(struct task_struct *prev) static void put_prev_task(struct rq *rq, struct task_struct *prev) { - if (prev->state == TASK_RUNNING) { - u64 runtime = prev->se.sum_exec_runtime; - - runtime -= prev->se.prev_sum_exec_runtime; - runtime = min_t(u64, runtime, 2*sysctl_sched_migration_cost); - - /* - * In order to avoid avg_overlap growing stale when we are - * indeed overlapping and hence not getting put to sleep, grow - * the avg_overlap on preemption. - * - * We use the average preemption runtime because that - * correlates to the amount of cache footprint a task can - * build up. - */ - update_avg(&prev->se.avg_overlap, runtime); - } prev->sched_class->put_prev_task(rq, prev); } diff --git a/kernel/sched_debug.c b/kernel/sched_debug.c index 20b95a4..8a46a71 100644 --- a/kernel/sched_debug.c +++ b/kernel/sched_debug.c @@ -407,7 +407,6 @@ void proc_sched_show_task(struct task_struct *p, struct seq_file *m) PN(se.exec_start); PN(se.vruntime); PN(se.sum_exec_runtime); - PN(se.avg_overlap); nr_switches = p->nvcsw + p->nivcsw; diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 6fc6285..c3b69d4 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -1241,7 +1241,6 @@ static inline unsigned long effective_load(struct task_group *tg, int cpu, static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync) { - struct task_struct *curr = current; unsigned long this_load, load; int idx, this_cpu, prev_cpu; unsigned long tl_per_task; @@ -1256,18 +1255,6 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync) load = source_load(prev_cpu, idx); this_load = target_load(this_cpu, idx); - if (sync) { - if (sched_feat(SYNC_LESS) && - (curr->se.avg_overlap > sysctl_sched_migration_cost || - p->se.avg_overlap > sysctl_sched_migration_cost)) - sync = 0; - } else { - if (sched_feat(SYNC_MORE) && - (curr->se.avg_overlap < sysctl_sched_migration_cost && - p->se.avg_overlap < sysctl_sched_migration_cost)) - sync = 1; - } - /* * If sync wakeup then subtract the (maximum possible) * effect of the currently running task from the load @@ -1711,11 +1698,6 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_ if (sched_feat(WAKEUP_SYNC) && sync) goto preempt; - if (sched_feat(WAKEUP_OVERLAP) && - se->avg_overlap < sysctl_sched_migration_cost && - pse->avg_overlap < sysctl_sched_migration_cost) - goto preempt; - if (!sched_feat(WAKEUP_PREEMPT)) return; diff --git a/kernel/sched_features.h b/kernel/sched_features.h index 96ef5db..c545e04 100644 --- a/kernel/sched_features.h +++ b/kernel/sched_features.h @@ -42,12 +42,6 @@ SCHED_FEAT(ASYM_GRAN, 1) SCHED_FEAT(WAKEUP_SYNC, 0) /* - * Wakeup preempt based on task behaviour. Tasks that do not overlap - * don't get preempted. - */ -SCHED_FEAT(WAKEUP_OVERLAP, 0) - -/* * Use the SYNC wakeup hint, pipes and the likes use this to indicate * the remote end is likely to consume the data we just wrote, and * therefore has cache benefit from being placed on the same cpu, see @@ -64,16 +58,6 @@ SCHED_FEAT(SYNC_WAKEUPS, 1) SCHED_FEAT(AFFINE_WAKEUPS, 1) /* - * Weaken SYNC hint based on overlap - */ -SCHED_FEAT(SYNC_LESS, 1) - -/* - * Add SYNC hint based on overlap - */ -SCHED_FEAT(SYNC_MORE, 0) - -/* * Prefer to schedule the task we woke last (assuming it failed * wakeup-preemption), since its likely going to consume data we * touched, increases cache locality. -- cgit v0.10.2 From a64692a3afd85fe048551ab89142fd5ca99a0dbd Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Thu, 11 Mar 2010 17:16:20 +0100 Subject: sched: Cleanup/optimize clock updates Now that we no longer depend on the clock being updated prior to enqueueing on migratory wakeup, we can clean up a bit, placing calls to update_rq_clock() exactly where they are needed, ie on enqueue, dequeue and schedule events. In the case of a freshly enqueued task immediately preempting, we can skip the update during preemption, as the clock was just updated by the enqueue event. We also save an unneeded call during a migratory wakeup by not updating the previous runqueue, where update_curr() won't be invoked. Signed-off-by: Mike Galbraith Signed-off-by: Peter Zijlstra LKML-Reference: <1268301199.6785.32.camel@marge.simson.net> Signed-off-by: Ingo Molnar diff --git a/kernel/sched.c b/kernel/sched.c index 68ed6f4..16559de 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -495,6 +495,8 @@ struct rq { u64 nohz_stamp; unsigned char in_nohz_recently; #endif + unsigned int skip_clock_update; + /* capture load from *all* tasks on this cpu: */ struct load_weight load; unsigned long nr_load_updates; @@ -592,6 +594,13 @@ static inline void check_preempt_curr(struct rq *rq, struct task_struct *p, int flags) { rq->curr->sched_class->check_preempt_curr(rq, p, flags); + + /* + * A queue event has occurred, and we're going to schedule. In + * this case, we can save a useless back to back clock update. + */ + if (test_tsk_need_resched(p)) + rq->skip_clock_update = 1; } static inline int cpu_of(struct rq *rq) @@ -626,7 +635,8 @@ static inline int cpu_of(struct rq *rq) inline void update_rq_clock(struct rq *rq) { - rq->clock = sched_clock_cpu(cpu_of(rq)); + if (!rq->skip_clock_update) + rq->clock = sched_clock_cpu(cpu_of(rq)); } /* @@ -1782,8 +1792,6 @@ static void double_rq_lock(struct rq *rq1, struct rq *rq2) raw_spin_lock_nested(&rq1->lock, SINGLE_DEPTH_NESTING); } } - update_rq_clock(rq1); - update_rq_clock(rq2); } /* @@ -1880,6 +1888,7 @@ static void update_avg(u64 *avg, u64 sample) static void enqueue_task(struct rq *rq, struct task_struct *p, int wakeup, bool head) { + update_rq_clock(rq); sched_info_queued(p); p->sched_class->enqueue_task(rq, p, wakeup, head); p->se.on_rq = 1; @@ -1887,6 +1896,7 @@ enqueue_task(struct rq *rq, struct task_struct *p, int wakeup, bool head) static void dequeue_task(struct rq *rq, struct task_struct *p, int sleep) { + update_rq_clock(rq); sched_info_dequeued(p); p->sched_class->dequeue_task(rq, p, sleep); p->se.on_rq = 0; @@ -2366,7 +2376,6 @@ static int try_to_wake_up(struct task_struct *p, unsigned int state, smp_wmb(); rq = task_rq_lock(p, &flags); - update_rq_clock(rq); if (!(p->state & state)) goto out; @@ -2407,7 +2416,6 @@ static int try_to_wake_up(struct task_struct *p, unsigned int state, rq = cpu_rq(cpu); raw_spin_lock(&rq->lock); - update_rq_clock(rq); /* * We migrated the task without holding either rq->lock, however @@ -2624,7 +2632,6 @@ void wake_up_new_task(struct task_struct *p, unsigned long clone_flags) BUG_ON(p->state != TASK_WAKING); p->state = TASK_RUNNING; - update_rq_clock(rq); activate_task(rq, p, 0); trace_sched_wakeup_new(rq, p, 1); check_preempt_curr(rq, p, WF_FORK); @@ -3578,6 +3585,9 @@ static inline void schedule_debug(struct task_struct *prev) static void put_prev_task(struct rq *rq, struct task_struct *prev) { + if (prev->se.on_rq) + update_rq_clock(rq); + rq->skip_clock_update = 0; prev->sched_class->put_prev_task(rq, prev); } @@ -3640,7 +3650,6 @@ need_resched_nonpreemptible: hrtick_clear(rq); raw_spin_lock_irq(&rq->lock); - update_rq_clock(rq); clear_tsk_need_resched(prev); if (prev->state && !(preempt_count() & PREEMPT_ACTIVE)) { @@ -4197,7 +4206,6 @@ void rt_mutex_setprio(struct task_struct *p, int prio) BUG_ON(prio < 0 || prio > MAX_PRIO); rq = task_rq_lock(p, &flags); - update_rq_clock(rq); oldprio = p->prio; prev_class = p->sched_class; @@ -4240,7 +4248,6 @@ void set_user_nice(struct task_struct *p, long nice) * the task might be in the middle of scheduling on another CPU. */ rq = task_rq_lock(p, &flags); - update_rq_clock(rq); /* * The RT priorities are set via sched_setscheduler(), but we still * allow the 'normal' nice value to be set - but as expected @@ -4523,7 +4530,6 @@ recheck: raw_spin_unlock_irqrestore(&p->pi_lock, flags); goto recheck; } - update_rq_clock(rq); on_rq = p->se.on_rq; running = task_current(rq, p); if (on_rq) @@ -5530,7 +5536,6 @@ void sched_idle_next(void) __setscheduler(rq, p, SCHED_FIFO, MAX_RT_PRIO-1); - update_rq_clock(rq); activate_task(rq, p, 0); raw_spin_unlock_irqrestore(&rq->lock, flags); @@ -5585,7 +5590,6 @@ static void migrate_dead_tasks(unsigned int dead_cpu) for ( ; ; ) { if (!rq->nr_running) break; - update_rq_clock(rq); next = pick_next_task(rq); if (!next) break; @@ -5869,7 +5873,6 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) rq->migration_thread = NULL; /* Idle task back to normal (off runqueue, low prio) */ raw_spin_lock_irq(&rq->lock); - update_rq_clock(rq); deactivate_task(rq, rq->idle, 0); __setscheduler(rq, rq->idle, SCHED_NORMAL, 0); rq->idle->sched_class = &idle_sched_class; @@ -7815,7 +7818,6 @@ static void normalize_task(struct rq *rq, struct task_struct *p) { int on_rq; - update_rq_clock(rq); on_rq = p->se.on_rq; if (on_rq) deactivate_task(rq, p, 0); @@ -8177,8 +8179,6 @@ void sched_move_task(struct task_struct *tsk) rq = task_rq_lock(tsk, &flags); - update_rq_clock(rq); - running = task_current(rq, tsk); on_rq = tsk->se.on_rq; diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index c3b69d4..69e5820 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -3064,8 +3064,6 @@ static void active_load_balance(struct rq *busiest_rq, int busiest_cpu) /* move a task from busiest_rq to target_rq */ double_lock_balance(busiest_rq, target_rq); - update_rq_clock(busiest_rq); - update_rq_clock(target_rq); /* Search for an sd spanning us and the target CPU. */ for_each_domain(target_cpu, sd) { -- cgit v0.10.2 From 21406928afe43f1db6acab4931bb8c886f4d04ce Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Thu, 11 Mar 2010 17:17:15 +0100 Subject: sched: Tweak sched_latency and min_granularity Allow LAST_BUDDY to kick in sooner, improving cache utilization as soon as a second buddy pair arrives on scene. The cost is latency starting to climb sooner, the tbenefit for tbench 8 on my Q6600 box is ~2%. No detrimental effects noted in normal idesktop usage. Signed-off-by: Mike Galbraith Signed-off-by: Peter Zijlstra LKML-Reference: <1268301285.6785.34.camel@marge.simson.net> Signed-off-by: Ingo Molnar diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 69e5820..d19df5b 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -35,8 +35,8 @@ * (to see the precise effective timeslice length of your workload, * run vmstat and monitor the context-switches (cs) field) */ -unsigned int sysctl_sched_latency = 5000000ULL; -unsigned int normalized_sysctl_sched_latency = 5000000ULL; +unsigned int sysctl_sched_latency = 6000000ULL; +unsigned int normalized_sysctl_sched_latency = 6000000ULL; /* * The initial- and re-scaling of tunables is configurable @@ -52,15 +52,15 @@ enum sched_tunable_scaling sysctl_sched_tunable_scaling /* * Minimal preemption granularity for CPU-bound tasks: - * (default: 1 msec * (1 + ilog(ncpus)), units: nanoseconds) + * (default: 2 msec * (1 + ilog(ncpus)), units: nanoseconds) */ -unsigned int sysctl_sched_min_granularity = 1000000ULL; -unsigned int normalized_sysctl_sched_min_granularity = 1000000ULL; +unsigned int sysctl_sched_min_granularity = 2000000ULL; +unsigned int normalized_sysctl_sched_min_granularity = 2000000ULL; /* * is kept at sysctl_sched_latency / sysctl_sched_min_granularity */ -static unsigned int sched_nr_latency = 5; +static unsigned int sched_nr_latency = 3; /* * After fork, child runs first. If set to 0 (default) then -- cgit v0.10.2 From 8b911acdf08477c059d1c36c21113ab1696c612b Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Thu, 11 Mar 2010 17:17:16 +0100 Subject: sched: Fix select_idle_sibling() Don't bother with selection when the current cpu is idle. Recent load balancing changes also make it no longer necessary to check wake_affine() success before returning the selected sibling, so we now always use it. Signed-off-by: Mike Galbraith Signed-off-by: Peter Zijlstra LKML-Reference: <1268301369.6785.36.camel@marge.simson.net> Signed-off-by: Ingo Molnar diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index d19df5b..0008cc4 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -1439,7 +1439,7 @@ static int select_task_rq_fair(struct task_struct *p, int sd_flag, int wake_flag int cpu = smp_processor_id(); int prev_cpu = task_cpu(p); int new_cpu = cpu; - int want_affine = 0; + int want_affine = 0, cpu_idle = !current->pid; int want_sd = 1; int sync = wake_flags & WF_SYNC; @@ -1497,13 +1497,15 @@ static int select_task_rq_fair(struct task_struct *p, int sd_flag, int wake_flag * If there's an idle sibling in this domain, make that * the wake_affine target instead of the current cpu. */ - if (tmp->flags & SD_SHARE_PKG_RESOURCES) + if (!cpu_idle && tmp->flags & SD_SHARE_PKG_RESOURCES) target = select_idle_sibling(p, tmp, target); if (target >= 0) { if (tmp->flags & SD_WAKE_AFFINE) { affine_sd = tmp; want_affine = 0; + if (target != cpu) + cpu_idle = 1; } cpu = target; } @@ -1519,6 +1521,7 @@ static int select_task_rq_fair(struct task_struct *p, int sd_flag, int wake_flag sd = tmp; } +#ifdef CONFIG_FAIR_GROUP_SCHED if (sched_feat(LB_SHARES_UPDATE)) { /* * Pick the largest domain to update shares over @@ -1532,9 +1535,12 @@ static int select_task_rq_fair(struct task_struct *p, int sd_flag, int wake_flag if (tmp) update_shares(tmp); } +#endif - if (affine_sd && wake_affine(affine_sd, p, sync)) - return cpu; + if (affine_sd) { + if (cpu_idle || cpu == prev_cpu || wake_affine(affine_sd, p, sync)) + return cpu; + } while (sd) { int load_idx = sd->forkexec_idx; -- cgit v0.10.2 From 6bc6cf2b61336ed0c55a615eb4c0c8ed5daf3f08 Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Thu, 11 Mar 2010 17:17:17 +0100 Subject: sched: Remove NORMALIZED_SLEEPER This feature hasn't been enabled in a long time, remove effectively dead code. Signed-off-by: Mike Galbraith Signed-off-by: Peter Zijlstra LKML-Reference: <1268301447.6785.38.camel@marge.simson.net> Signed-off-by: Ingo Molnar diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 0008cc4..de98e2e 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -742,16 +742,6 @@ place_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int initial) unsigned long thresh = sysctl_sched_latency; /* - * Convert the sleeper threshold into virtual time. - * SCHED_IDLE is a special sub-class. We care about - * fairness only relative to other SCHED_IDLE tasks, - * all of which have the same weight. - */ - if (sched_feat(NORMALIZED_SLEEPER) && (!entity_is_task(se) || - task_of(se)->policy != SCHED_IDLE)) - thresh = calc_delta_fair(thresh, se); - - /* * Halve their sleep time's effect, to allow * for a gentler effect of sleepers: */ diff --git a/kernel/sched_features.h b/kernel/sched_features.h index c545e04..4042883 100644 --- a/kernel/sched_features.h +++ b/kernel/sched_features.h @@ -13,13 +13,6 @@ SCHED_FEAT(FAIR_SLEEPERS, 1) SCHED_FEAT(GENTLE_FAIR_SLEEPERS, 1) /* - * By not normalizing the sleep time, heavy tasks get an effective - * longer period, and lighter task an effective shorter period they - * are considered running. - */ -SCHED_FEAT(NORMALIZED_SLEEPER, 0) - -/* * Place new tasks ahead so that they do not starve already running * tasks */ -- cgit v0.10.2 From 5ca9880c6f4ba4c84b517bc2fed5366adf63d191 Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Thu, 11 Mar 2010 17:17:17 +0100 Subject: sched: Remove FAIR_SLEEPERS feature Our preemption model relies too heavily on sleeper fairness to disable it without dire consequences. Remove the feature, and save a branch or two. Signed-off-by: Mike Galbraith Signed-off-by: Peter Zijlstra LKML-Reference: <1268301520.6785.40.camel@marge.simson.net> Signed-off-by: Ingo Molnar diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index de98e2e..97682f9 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -738,7 +738,7 @@ place_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int initial) vruntime += sched_vslice(cfs_rq, se); /* sleeps up to a single latency don't count. */ - if (!initial && sched_feat(FAIR_SLEEPERS)) { + if (!initial) { unsigned long thresh = sysctl_sched_latency; /* diff --git a/kernel/sched_features.h b/kernel/sched_features.h index 4042883..850f980 100644 --- a/kernel/sched_features.h +++ b/kernel/sched_features.h @@ -1,11 +1,4 @@ /* - * Disregards a certain amount of sleep time (sched_latency_ns) and - * considers the task to be running during that period. This gives it - * a service deficit on wakeup, allowing it to run sooner. - */ -SCHED_FEAT(FAIR_SLEEPERS, 1) - -/* * Only give sleepers 50% of their service deficit. This allows * them to run sooner, but does not allow tons of sleepers to * rip the spread apart. -- cgit v0.10.2 From f2e74eeac03ffb779d64b66a643c5e598145a28b Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Thu, 11 Mar 2010 17:17:18 +0100 Subject: sched: Remove WAKEUP_SYNC feature This feature never earned its keep, remove it. Signed-off-by: Mike Galbraith Signed-off-by: Peter Zijlstra LKML-Reference: <1268301591.6785.42.camel@marge.simson.net> Signed-off-by: Ingo Molnar diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 97682f9..1d99535 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -1658,7 +1658,6 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_ struct task_struct *curr = rq->curr; struct sched_entity *se = &curr->se, *pse = &p->se; struct cfs_rq *cfs_rq = task_cfs_rq(curr); - int sync = wake_flags & WF_SYNC; int scale = cfs_rq->nr_running >= sched_nr_latency; if (unlikely(rt_prio(p->prio))) @@ -1691,9 +1690,6 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_ if (unlikely(curr->policy == SCHED_IDLE)) goto preempt; - if (sched_feat(WAKEUP_SYNC) && sync) - goto preempt; - if (!sched_feat(WAKEUP_PREEMPT)) return; diff --git a/kernel/sched_features.h b/kernel/sched_features.h index 850f980..1cb7c47 100644 --- a/kernel/sched_features.h +++ b/kernel/sched_features.h @@ -23,11 +23,6 @@ SCHED_FEAT(WAKEUP_PREEMPT, 1) SCHED_FEAT(ASYM_GRAN, 1) /* - * Always wakeup-preempt SYNC wakeups, see SYNC_WAKEUPS. - */ -SCHED_FEAT(WAKEUP_SYNC, 0) - -/* * Use the SYNC wakeup hint, pipes and the likes use this to indicate * the remote end is likely to consume the data we just wrote, and * therefore has cache benefit from being placed on the same cpu, see -- cgit v0.10.2 From c6ee36c423c3ed1fb86bb3eabba9fc256a300d16 Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Thu, 11 Mar 2010 17:16:43 +0100 Subject: sched: Remove SYNC_WAKEUPS feature Sync wakeups are critical functionality with a long history. Remove it, we don't need the branch or icache footprint. Signed-off-by: Mike Galbraith Signed-off-by: Peter Zijlstra LKML-Reference: <1268301817.6785.47.camel@marge.simson.net> Signed-off-by: Ingo Molnar diff --git a/kernel/sched.c b/kernel/sched.c index 16559de..cc6dc8c 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2369,9 +2369,6 @@ static int try_to_wake_up(struct task_struct *p, unsigned int state, unsigned long flags; struct rq *rq; - if (!sched_feat(SYNC_WAKEUPS)) - wake_flags &= ~WF_SYNC; - this_cpu = get_cpu(); smp_wmb(); diff --git a/kernel/sched_features.h b/kernel/sched_features.h index 1cb7c47..f54b6f9 100644 --- a/kernel/sched_features.h +++ b/kernel/sched_features.h @@ -23,14 +23,6 @@ SCHED_FEAT(WAKEUP_PREEMPT, 1) SCHED_FEAT(ASYM_GRAN, 1) /* - * Use the SYNC wakeup hint, pipes and the likes use this to indicate - * the remote end is likely to consume the data we just wrote, and - * therefore has cache benefit from being placed on the same cpu, see - * also AFFINE_WAKEUPS. - */ -SCHED_FEAT(SYNC_WAKEUPS, 1) - -/* * Based on load and program behaviour, see if it makes sense to place * a newly woken task on the same cpu as the task that woke it -- * improve cache locality. Typically used with SYNC wakeups as -- cgit v0.10.2 From 13814d42e45dfbe845a0bbe5184565d9236896ae Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Thu, 11 Mar 2010 17:17:04 +0100 Subject: sched: Remove ASYM_GRAN feature This features has been enabled for quite a while, after testing showed that easing preemption for light tasks was harmful to high priority threads. Remove the feature flag. Signed-off-by: Mike Galbraith Signed-off-by: Peter Zijlstra LKML-Reference: <1268301675.6785.44.camel@marge.simson.net> Signed-off-by: Ingo Molnar diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 1d99535..9357ecd 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -1583,24 +1583,18 @@ wakeup_gran(struct sched_entity *curr, struct sched_entity *se) /* * Since its curr running now, convert the gran from real-time * to virtual-time in his units. + * + * By using 'se' instead of 'curr' we penalize light tasks, so + * they get preempted easier. That is, if 'se' < 'curr' then + * the resulting gran will be larger, therefore penalizing the + * lighter, if otoh 'se' > 'curr' then the resulting gran will + * be smaller, again penalizing the lighter task. + * + * This is especially important for buddies when the leftmost + * task is higher priority than the buddy. */ - if (sched_feat(ASYM_GRAN)) { - /* - * By using 'se' instead of 'curr' we penalize light tasks, so - * they get preempted easier. That is, if 'se' < 'curr' then - * the resulting gran will be larger, therefore penalizing the - * lighter, if otoh 'se' > 'curr' then the resulting gran will - * be smaller, again penalizing the lighter task. - * - * This is especially important for buddies when the leftmost - * task is higher priority than the buddy. - */ - if (unlikely(se->load.weight != NICE_0_LOAD)) - gran = calc_delta_fair(gran, se); - } else { - if (unlikely(curr->load.weight != NICE_0_LOAD)) - gran = calc_delta_fair(gran, curr); - } + if (unlikely(se->load.weight != NICE_0_LOAD)) + gran = calc_delta_fair(gran, se); return gran; } diff --git a/kernel/sched_features.h b/kernel/sched_features.h index f54b6f9..83c66e8 100644 --- a/kernel/sched_features.h +++ b/kernel/sched_features.h @@ -17,12 +17,6 @@ SCHED_FEAT(START_DEBIT, 1) SCHED_FEAT(WAKEUP_PREEMPT, 1) /* - * When converting the wakeup granularity to virtual time, do it such - * that heavier tasks preempting a lighter task have an edge. - */ -SCHED_FEAT(ASYM_GRAN, 1) - -/* * Based on load and program behaviour, see if it makes sense to place * a newly woken task on the same cpu as the task that woke it -- * improve cache locality. Typically used with SYNC wakeups as -- cgit v0.10.2 From beac4c7e4a1cc6d57801f690e5e82fa2c9c245c8 Mon Sep 17 00:00:00 2001 From: Mike Galbraith Date: Thu, 11 Mar 2010 17:17:20 +0100 Subject: sched: Remove AFFINE_WAKEUPS feature Disabling affine wakeups is too horrible to contemplate. Remove the feature flag. Signed-off-by: Mike Galbraith Signed-off-by: Peter Zijlstra LKML-Reference: <1268301890.6785.50.camel@marge.simson.net> Signed-off-by: Ingo Molnar diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 9357ecd..35a5c64 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -1434,8 +1434,7 @@ static int select_task_rq_fair(struct task_struct *p, int sd_flag, int wake_flag int sync = wake_flags & WF_SYNC; if (sd_flag & SD_BALANCE_WAKE) { - if (sched_feat(AFFINE_WAKEUPS) && - cpumask_test_cpu(cpu, &p->cpus_allowed)) + if (cpumask_test_cpu(cpu, &p->cpus_allowed)) want_affine = 1; new_cpu = prev_cpu; } -- cgit v0.10.2 From a072738e04f0eb26370e39ec679e9a0d65e49aea Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Thu, 11 Mar 2010 19:54:39 +0300 Subject: perf, x86: Implement initial P4 PMU driver The netburst PMU is way different from the "architectural perfomance monitoring" specification that current CPUs use. P4 uses a tuple of ESCR+CCCR+COUNTER MSR registers to handle perfomance monitoring events. A few implementational details: 1) We need a separate x86_pmu::hw_config helper in struct x86_pmu since register bit-fields are quite different from P6, Core and later cpu series. 2) For the same reason is a x86_pmu::schedule_events helper introduced. 3) hw_perf_event::config consists of packed ESCR+CCCR values. It's allowed since in reality both registers only use a half of their size. Of course before making a real write into a particular MSR we need to unpack the value and extend it to a proper size. 4) The tuple of packed ESCR+CCCR in hw_perf_event::config doesn't describe the memory address of ESCR MSR register so that we need to keep a mapping between these tuples used and available ESCR (various P4 events may use same ESCRs but not simultaneously), for this sake every active event has a per-cpu map of hw_perf_event::idx <--> ESCR addresses. 5) Since hw_perf_event::idx is an offset to counter/control register we need to lift X86_PMC_MAX_GENERIC up, otherwise kernel strips it down to 8 registers and event armed may never be turned off (ie the bit in active_mask is set but the loop never reaches this index to check), thanks to Peter Zijlstra Restrictions: - No cascaded counters support (do we ever need them?) - No dependent events support (so PERF_COUNT_HW_INSTRUCTIONS doesn't work for now) - There are events with same counters which can't work simultaneously (need to use intersected ones due to broken counter 1) - No PERF_COUNT_HW_CACHE_ events yet Todo: - Implement dependent events - Need proper hashing for event opcodes (no linear search, good for debugging stage but not in real loads) - Some events counted during a clock cycle -- need to set threshold for them and count every clock cycle just to get summary statistics (ie to behave the same way as other PMUs do) - Need to swicth to use event_constraints - To support RAW events we need to encode a global list of P4 events into p4_templates - Cache events need to be added Event support status matrix: Event status ----------------------------- cycles works cache-references works cache-misses works branch-misses works bus-cycles partially (does not work on 64bit cpu with HT enabled) instruction doesnt work (needs dependent event [mop tagging]) branches doesnt work Signed-off-by: Cyrill Gorcunov Signed-off-by: Lin Ming Cc: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Stephane Eranian Cc: Robert Richter Cc: Frederic Weisbecker LKML-Reference: <20100311165439.GB5129@lenovo> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/perf_event.h b/arch/x86/include/asm/perf_event.h index a9038c9..124dddd 100644 --- a/arch/x86/include/asm/perf_event.h +++ b/arch/x86/include/asm/perf_event.h @@ -5,7 +5,7 @@ * Performance event hw details: */ -#define X86_PMC_MAX_GENERIC 8 +#define X86_PMC_MAX_GENERIC 32 #define X86_PMC_MAX_FIXED 3 #define X86_PMC_IDX_GENERIC 0 diff --git a/arch/x86/include/asm/perf_event_p4.h b/arch/x86/include/asm/perf_event_p4.h new file mode 100644 index 0000000..829f471 --- /dev/null +++ b/arch/x86/include/asm/perf_event_p4.h @@ -0,0 +1,707 @@ +/* + * Netburst Perfomance Events (P4, old Xeon) + */ + +#ifndef PERF_EVENT_P4_H +#define PERF_EVENT_P4_H + +#include +#include + +/* + * NetBurst has perfomance MSRs shared between + * threads if HT is turned on, ie for both logical + * processors (mem: in turn in Atom with HT support + * perf-MSRs are not shared and every thread has its + * own perf-MSRs set) + */ +#define ARCH_P4_TOTAL_ESCR (46) +#define ARCH_P4_RESERVED_ESCR (2) /* IQ_ESCR(0,1) not always present */ +#define ARCH_P4_MAX_ESCR (ARCH_P4_TOTAL_ESCR - ARCH_P4_RESERVED_ESCR) +#define ARCH_P4_MAX_CCCR (18) +#define ARCH_P4_MAX_COUNTER (ARCH_P4_MAX_CCCR / 2) + +#define P4_EVNTSEL_EVENT_MASK 0x7e000000U +#define P4_EVNTSEL_EVENT_SHIFT 25 +#define P4_EVNTSEL_EVENTMASK_MASK 0x01fffe00U +#define P4_EVNTSEL_EVENTMASK_SHIFT 9 +#define P4_EVNTSEL_TAG_MASK 0x000001e0U +#define P4_EVNTSEL_TAG_SHIFT 5 +#define P4_EVNTSEL_TAG_ENABLE 0x00000010U +#define P4_EVNTSEL_T0_OS 0x00000008U +#define P4_EVNTSEL_T0_USR 0x00000004U +#define P4_EVNTSEL_T1_OS 0x00000002U +#define P4_EVNTSEL_T1_USR 0x00000001U + +/* Non HT mask */ +#define P4_EVNTSEL_MASK \ + (P4_EVNTSEL_EVENT_MASK | \ + P4_EVNTSEL_EVENTMASK_MASK | \ + P4_EVNTSEL_TAG_MASK | \ + P4_EVNTSEL_TAG_ENABLE | \ + P4_EVNTSEL_T0_OS | \ + P4_EVNTSEL_T0_USR) + +/* HT mask */ +#define P4_EVNTSEL_MASK_HT \ + (P4_EVNTSEL_MASK | \ + P4_EVNTSEL_T1_OS | \ + P4_EVNTSEL_T1_USR) + +#define P4_CCCR_OVF 0x80000000U +#define P4_CCCR_CASCADE 0x40000000U +#define P4_CCCR_OVF_PMI_T0 0x04000000U +#define P4_CCCR_OVF_PMI_T1 0x08000000U +#define P4_CCCR_FORCE_OVF 0x02000000U +#define P4_CCCR_EDGE 0x01000000U +#define P4_CCCR_THRESHOLD_MASK 0x00f00000U +#define P4_CCCR_THRESHOLD_SHIFT 20 +#define P4_CCCR_THRESHOLD(v) ((v) << P4_CCCR_THRESHOLD_SHIFT) +#define P4_CCCR_COMPLEMENT 0x00080000U +#define P4_CCCR_COMPARE 0x00040000U +#define P4_CCCR_ESCR_SELECT_MASK 0x0000e000U +#define P4_CCCR_ESCR_SELECT_SHIFT 13 +#define P4_CCCR_ENABLE 0x00001000U +#define P4_CCCR_THREAD_SINGLE 0x00010000U +#define P4_CCCR_THREAD_BOTH 0x00020000U +#define P4_CCCR_THREAD_ANY 0x00030000U + +/* Non HT mask */ +#define P4_CCCR_MASK \ + (P4_CCCR_OVF | \ + P4_CCCR_CASCADE | \ + P4_CCCR_OVF_PMI_T0 | \ + P4_CCCR_FORCE_OVF | \ + P4_CCCR_EDGE | \ + P4_CCCR_THRESHOLD_MASK | \ + P4_CCCR_COMPLEMENT | \ + P4_CCCR_COMPARE | \ + P4_CCCR_ESCR_SELECT_MASK | \ + P4_CCCR_ENABLE) + +/* HT mask */ +#define P4_CCCR_MASK_HT \ + (P4_CCCR_MASK | \ + P4_CCCR_THREAD_ANY) + +/* + * format is 32 bit: ee ss aa aa + * where + * ee - 8 bit event + * ss - 8 bit selector + * aa aa - 16 bits reserved for tags/attributes + */ +#define P4_EVENT_PACK(event, selector) (((event) << 24) | ((selector) << 16)) +#define P4_EVENT_UNPACK_EVENT(packed) (((packed) >> 24) & 0xff) +#define P4_EVENT_UNPACK_SELECTOR(packed) (((packed) >> 16) & 0xff) +#define P4_EVENT_PACK_ATTR(attr) ((attr)) +#define P4_EVENT_UNPACK_ATTR(packed) ((packed) & 0xffff) +#define P4_MAKE_EVENT_ATTR(class, name, bit) class##_##name = (1 << bit) +#define P4_EVENT_ATTR(class, name) class##_##name +#define P4_EVENT_ATTR_STR(class, name) __stringify(class##_##name) + +/* + * config field is 64bit width and consists of + * HT << 63 | ESCR << 32 | CCCR + * where HT is HyperThreading bit (since ESCR + * has it reserved we may use it for own purpose) + * + * note that this is NOT the addresses of respective + * ESCR and CCCR but rather an only packed value should + * be unpacked and written to a proper addresses + * + * the base idea is to pack as much info as + * possible + */ +#define p4_config_pack_escr(v) (((u64)(v)) << 32) +#define p4_config_pack_cccr(v) (((u64)(v)) & 0xffffffffULL) +#define p4_config_unpack_escr(v) (((u64)(v)) >> 32) +#define p4_config_unpack_cccr(v) (((u64)(v)) & 0xffffffffULL) + +#define p4_config_unpack_emask(v) \ + ({ \ + u32 t = p4_config_unpack_escr((v)); \ + t &= P4_EVNTSEL_EVENTMASK_MASK; \ + t >>= P4_EVNTSEL_EVENTMASK_SHIFT; \ + t; \ + }) + +#define P4_CONFIG_HT_SHIFT 63 +#define P4_CONFIG_HT (1ULL << P4_CONFIG_HT_SHIFT) + +static inline u32 p4_config_unpack_opcode(u64 config) +{ + u32 e, s; + + /* + * we don't care about HT presence here since + * event opcode doesn't depend on it + */ + e = (p4_config_unpack_escr(config) & P4_EVNTSEL_EVENT_MASK) >> P4_EVNTSEL_EVENT_SHIFT; + s = (p4_config_unpack_cccr(config) & P4_CCCR_ESCR_SELECT_MASK) >> P4_CCCR_ESCR_SELECT_SHIFT; + + return P4_EVENT_PACK(e, s); +} + +static inline bool p4_is_event_cascaded(u64 config) +{ + u32 cccr = p4_config_unpack_cccr(config); + return !!(cccr & P4_CCCR_CASCADE); +} + +static inline int p4_ht_config_thread(u64 config) +{ + return !!(config & P4_CONFIG_HT); +} + +static inline u64 p4_set_ht_bit(u64 config) +{ + return config | P4_CONFIG_HT; +} + +static inline u64 p4_clear_ht_bit(u64 config) +{ + return config & ~P4_CONFIG_HT; +} + +static inline int p4_ht_active(void) +{ +#ifdef CONFIG_SMP + return smp_num_siblings > 1; +#endif + return 0; +} + +static inline int p4_ht_thread(int cpu) +{ +#ifdef CONFIG_SMP + if (smp_num_siblings == 2) + return cpu != cpumask_first(__get_cpu_var(cpu_sibling_map)); +#endif + return 0; +} + +static inline int p4_should_swap_ts(u64 config, int cpu) +{ + return p4_ht_config_thread(config) ^ p4_ht_thread(cpu); +} + +static inline u32 p4_default_cccr_conf(int cpu) +{ + /* + * Note that P4_CCCR_THREAD_ANY is "required" on + * non-HT machines (on HT machines we count TS events + * regardless the state of second logical processor + */ + u32 cccr = P4_CCCR_THREAD_ANY; + + if (!p4_ht_thread(cpu)) + cccr |= P4_CCCR_OVF_PMI_T0; + else + cccr |= P4_CCCR_OVF_PMI_T1; + + return cccr; +} + +static inline u32 p4_default_escr_conf(int cpu, int exclude_os, int exclude_usr) +{ + u32 escr = 0; + + if (!p4_ht_thread(cpu)) { + if (!exclude_os) + escr |= P4_EVNTSEL_T0_OS; + if (!exclude_usr) + escr |= P4_EVNTSEL_T0_USR; + } else { + if (!exclude_os) + escr |= P4_EVNTSEL_T1_OS; + if (!exclude_usr) + escr |= P4_EVNTSEL_T1_USR; + } + + return escr; +} + +/* + * Comments below the event represent ESCR restriction + * for this event and counter index per ESCR + * + * MSR_P4_IQ_ESCR0 and MSR_P4_IQ_ESCR1 are available only on early + * processor builds (family 0FH, models 01H-02H). These MSRs + * are not available on later versions, so that we don't use + * them completely + * + * Also note that CCCR1 do not have P4_CCCR_ENABLE bit properly + * working so that we should not use this CCCR and respective + * counter as result + */ +#define P4_TC_DELIVER_MODE P4_EVENT_PACK(0x01, 0x01) + /* + * MSR_P4_TC_ESCR0: 4, 5 + * MSR_P4_TC_ESCR1: 6, 7 + */ + +#define P4_BPU_FETCH_REQUEST P4_EVENT_PACK(0x03, 0x00) + /* + * MSR_P4_BPU_ESCR0: 0, 1 + * MSR_P4_BPU_ESCR1: 2, 3 + */ + +#define P4_ITLB_REFERENCE P4_EVENT_PACK(0x18, 0x03) + /* + * MSR_P4_ITLB_ESCR0: 0, 1 + * MSR_P4_ITLB_ESCR1: 2, 3 + */ + +#define P4_MEMORY_CANCEL P4_EVENT_PACK(0x02, 0x05) + /* + * MSR_P4_DAC_ESCR0: 8, 9 + * MSR_P4_DAC_ESCR1: 10, 11 + */ + +#define P4_MEMORY_COMPLETE P4_EVENT_PACK(0x08, 0x02) + /* + * MSR_P4_SAAT_ESCR0: 8, 9 + * MSR_P4_SAAT_ESCR1: 10, 11 + */ + +#define P4_LOAD_PORT_REPLAY P4_EVENT_PACK(0x04, 0x02) + /* + * MSR_P4_SAAT_ESCR0: 8, 9 + * MSR_P4_SAAT_ESCR1: 10, 11 + */ + +#define P4_STORE_PORT_REPLAY P4_EVENT_PACK(0x05, 0x02) + /* + * MSR_P4_SAAT_ESCR0: 8, 9 + * MSR_P4_SAAT_ESCR1: 10, 11 + */ + +#define P4_MOB_LOAD_REPLAY P4_EVENT_PACK(0x03, 0x02) + /* + * MSR_P4_MOB_ESCR0: 0, 1 + * MSR_P4_MOB_ESCR1: 2, 3 + */ + +#define P4_PAGE_WALK_TYPE P4_EVENT_PACK(0x01, 0x04) + /* + * MSR_P4_PMH_ESCR0: 0, 1 + * MSR_P4_PMH_ESCR1: 2, 3 + */ + +#define P4_BSQ_CACHE_REFERENCE P4_EVENT_PACK(0x0c, 0x07) + /* + * MSR_P4_BSU_ESCR0: 0, 1 + * MSR_P4_BSU_ESCR1: 2, 3 + */ + +#define P4_IOQ_ALLOCATION P4_EVENT_PACK(0x03, 0x06) + /* + * MSR_P4_FSB_ESCR0: 0, 1 + * MSR_P4_FSB_ESCR1: 2, 3 + */ + +#define P4_IOQ_ACTIVE_ENTRIES P4_EVENT_PACK(0x1a, 0x06) + /* + * MSR_P4_FSB_ESCR1: 2, 3 + */ + +#define P4_FSB_DATA_ACTIVITY P4_EVENT_PACK(0x17, 0x06) + /* + * MSR_P4_FSB_ESCR0: 0, 1 + * MSR_P4_FSB_ESCR1: 2, 3 + */ + +#define P4_BSQ_ALLOCATION P4_EVENT_PACK(0x05, 0x07) + /* + * MSR_P4_BSU_ESCR0: 0, 1 + */ + +#define P4_BSQ_ACTIVE_ENTRIES P4_EVENT_PACK(0x06, 0x07) + /* + * MSR_P4_BSU_ESCR1: 2, 3 + */ + +#define P4_SSE_INPUT_ASSIST P4_EVENT_PACK(0x34, 0x01) + /* + * MSR_P4_FIRM_ESCR: 8, 9 + * MSR_P4_FIRM_ESCR: 10, 11 + */ + +#define P4_PACKED_SP_UOP P4_EVENT_PACK(0x08, 0x01) + /* + * MSR_P4_FIRM_ESCR0: 8, 9 + * MSR_P4_FIRM_ESCR1: 10, 11 + */ + +#define P4_PACKED_DP_UOP P4_EVENT_PACK(0x0c, 0x01) + /* + * MSR_P4_FIRM_ESCR0: 8, 9 + * MSR_P4_FIRM_ESCR1: 10, 11 + */ + +#define P4_SCALAR_SP_UOP P4_EVENT_PACK(0x0a, 0x01) + /* + * MSR_P4_FIRM_ESCR0: 8, 9 + * MSR_P4_FIRM_ESCR1: 10, 11 + */ + +#define P4_SCALAR_DP_UOP P4_EVENT_PACK(0x0e, 0x01) + /* + * MSR_P4_FIRM_ESCR0: 8, 9 + * MSR_P4_FIRM_ESCR1: 10, 11 + */ + +#define P4_64BIT_MMX_UOP P4_EVENT_PACK(0x02, 0x01) + /* + * MSR_P4_FIRM_ESCR0: 8, 9 + * MSR_P4_FIRM_ESCR1: 10, 11 + */ + +#define P4_128BIT_MMX_UOP P4_EVENT_PACK(0x1a, 0x01) + /* + * MSR_P4_FIRM_ESCR0: 8, 9 + * MSR_P4_FIRM_ESCR1: 10, 11 + */ + +#define P4_X87_FP_UOP P4_EVENT_PACK(0x04, 0x01) + /* + * MSR_P4_FIRM_ESCR0: 8, 9 + * MSR_P4_FIRM_ESCR1: 10, 11 + */ + +#define P4_TC_MISC P4_EVENT_PACK(0x06, 0x01) + /* + * MSR_P4_TC_ESCR0: 4, 5 + * MSR_P4_TC_ESCR1: 6, 7 + */ + +#define P4_GLOBAL_POWER_EVENTS P4_EVENT_PACK(0x13, 0x06) + /* + * MSR_P4_FSB_ESCR0: 0, 1 + * MSR_P4_FSB_ESCR1: 2, 3 + */ + +#define P4_TC_MS_XFER P4_EVENT_PACK(0x05, 0x00) + /* + * MSR_P4_MS_ESCR0: 4, 5 + * MSR_P4_MS_ESCR1: 6, 7 + */ + +#define P4_UOP_QUEUE_WRITES P4_EVENT_PACK(0x09, 0x00) + /* + * MSR_P4_MS_ESCR0: 4, 5 + * MSR_P4_MS_ESCR1: 6, 7 + */ + +#define P4_RETIRED_MISPRED_BRANCH_TYPE P4_EVENT_PACK(0x05, 0x02) + /* + * MSR_P4_TBPU_ESCR0: 4, 5 + * MSR_P4_TBPU_ESCR0: 6, 7 + */ + +#define P4_RETIRED_BRANCH_TYPE P4_EVENT_PACK(0x04, 0x02) + /* + * MSR_P4_TBPU_ESCR0: 4, 5 + * MSR_P4_TBPU_ESCR0: 6, 7 + */ + +#define P4_RESOURCE_STALL P4_EVENT_PACK(0x01, 0x01) + /* + * MSR_P4_ALF_ESCR0: 12, 13, 16 + * MSR_P4_ALF_ESCR1: 14, 15, 17 + */ + +#define P4_WC_BUFFER P4_EVENT_PACK(0x05, 0x05) + /* + * MSR_P4_DAC_ESCR0: 8, 9 + * MSR_P4_DAC_ESCR1: 10, 11 + */ + +#define P4_B2B_CYCLES P4_EVENT_PACK(0x16, 0x03) + /* + * MSR_P4_FSB_ESCR0: 0, 1 + * MSR_P4_FSB_ESCR1: 2, 3 + */ + +#define P4_BNR P4_EVENT_PACK(0x08, 0x03) + /* + * MSR_P4_FSB_ESCR0: 0, 1 + * MSR_P4_FSB_ESCR1: 2, 3 + */ + +#define P4_SNOOP P4_EVENT_PACK(0x06, 0x03) + /* + * MSR_P4_FSB_ESCR0: 0, 1 + * MSR_P4_FSB_ESCR1: 2, 3 + */ + +#define P4_RESPONSE P4_EVENT_PACK(0x04, 0x03) + /* + * MSR_P4_FSB_ESCR0: 0, 1 + * MSR_P4_FSB_ESCR1: 2, 3 + */ + +#define P4_FRONT_END_EVENT P4_EVENT_PACK(0x08, 0x05) + /* + * MSR_P4_CRU_ESCR2: 12, 13, 16 + * MSR_P4_CRU_ESCR3: 14, 15, 17 + */ + +#define P4_EXECUTION_EVENT P4_EVENT_PACK(0x0c, 0x05) + /* + * MSR_P4_CRU_ESCR2: 12, 13, 16 + * MSR_P4_CRU_ESCR3: 14, 15, 17 + */ + +#define P4_REPLAY_EVENT P4_EVENT_PACK(0x09, 0x05) + /* + * MSR_P4_CRU_ESCR2: 12, 13, 16 + * MSR_P4_CRU_ESCR3: 14, 15, 17 + */ + +#define P4_INSTR_RETIRED P4_EVENT_PACK(0x02, 0x04) + /* + * MSR_P4_CRU_ESCR2: 12, 13, 16 + * MSR_P4_CRU_ESCR3: 14, 15, 17 + */ + +#define P4_UOPS_RETIRED P4_EVENT_PACK(0x01, 0x04) + /* + * MSR_P4_CRU_ESCR2: 12, 13, 16 + * MSR_P4_CRU_ESCR3: 14, 15, 17 + */ + +#define P4_UOP_TYPE P4_EVENT_PACK(0x02, 0x02) + /* + * MSR_P4_RAT_ESCR0: 12, 13, 16 + * MSR_P4_RAT_ESCR1: 14, 15, 17 + */ + +#define P4_BRANCH_RETIRED P4_EVENT_PACK(0x06, 0x05) + /* + * MSR_P4_CRU_ESCR2: 12, 13, 16 + * MSR_P4_CRU_ESCR3: 14, 15, 17 + */ + +#define P4_MISPRED_BRANCH_RETIRED P4_EVENT_PACK(0x03, 0x04) + /* + * MSR_P4_CRU_ESCR0: 12, 13, 16 + * MSR_P4_CRU_ESCR1: 14, 15, 17 + */ + +#define P4_X87_ASSIST P4_EVENT_PACK(0x03, 0x05) + /* + * MSR_P4_CRU_ESCR2: 12, 13, 16 + * MSR_P4_CRU_ESCR3: 14, 15, 17 + */ + +#define P4_MACHINE_CLEAR P4_EVENT_PACK(0x02, 0x05) + /* + * MSR_P4_CRU_ESCR2: 12, 13, 16 + * MSR_P4_CRU_ESCR3: 14, 15, 17 + */ + +#define P4_INSTR_COMPLETED P4_EVENT_PACK(0x07, 0x04) + /* + * MSR_P4_CRU_ESCR0: 12, 13, 16 + * MSR_P4_CRU_ESCR1: 14, 15, 17 + */ + +/* + * a caller should use P4_EVENT_ATTR helper to + * pick the attribute needed, for example + * + * P4_EVENT_ATTR(P4_TC_DELIVER_MODE, DD) + */ +enum P4_EVENTS_ATTR { + P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, DD, 0), + P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, DB, 1), + P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, DI, 2), + P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, BD, 3), + P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, BB, 4), + P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, BI, 5), + P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, ID, 6), + + P4_MAKE_EVENT_ATTR(P4_BPU_FETCH_REQUEST, TCMISS, 0), + + P4_MAKE_EVENT_ATTR(P4_ITLB_REFERENCE, HIT, 0), + P4_MAKE_EVENT_ATTR(P4_ITLB_REFERENCE, MISS, 1), + P4_MAKE_EVENT_ATTR(P4_ITLB_REFERENCE, HIT_UK, 2), + + P4_MAKE_EVENT_ATTR(P4_MEMORY_CANCEL, ST_RB_FULL, 2), + P4_MAKE_EVENT_ATTR(P4_MEMORY_CANCEL, 64K_CONF, 3), + + P4_MAKE_EVENT_ATTR(P4_MEMORY_COMPLETE, LSC, 0), + P4_MAKE_EVENT_ATTR(P4_MEMORY_COMPLETE, SSC, 1), + + P4_MAKE_EVENT_ATTR(P4_LOAD_PORT_REPLAY, SPLIT_LD, 1), + + P4_MAKE_EVENT_ATTR(P4_STORE_PORT_REPLAY, SPLIT_ST, 1), + + P4_MAKE_EVENT_ATTR(P4_MOB_LOAD_REPLAY, NO_STA, 1), + P4_MAKE_EVENT_ATTR(P4_MOB_LOAD_REPLAY, NO_STD, 3), + P4_MAKE_EVENT_ATTR(P4_MOB_LOAD_REPLAY, PARTIAL_DATA, 4), + P4_MAKE_EVENT_ATTR(P4_MOB_LOAD_REPLAY, UNALGN_ADDR, 5), + + P4_MAKE_EVENT_ATTR(P4_PAGE_WALK_TYPE, DTMISS, 0), + P4_MAKE_EVENT_ATTR(P4_PAGE_WALK_TYPE, ITMISS, 1), + + P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITS, 0), + P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITE, 1), + P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITM, 2), + P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_HITS, 3), + P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_HITE, 4), + P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_HITM, 5), + P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_MISS, 8), + P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_MISS, 9), + P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, WR_2ndL_MISS, 10), + + P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, DEFAULT, 0), + P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, ALL_READ, 5), + P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, ALL_WRITE, 6), + P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, MEM_UC, 7), + P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, MEM_WC, 8), + P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, MEM_WT, 9), + P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, MEM_WP, 10), + P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, MEM_WB, 11), + P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, OWN, 13), + P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, OTHER, 14), + P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, PREFETCH, 15), + + P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, DEFAULT, 0), + P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, ALL_READ, 5), + P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, ALL_WRITE, 6), + P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, MEM_UC, 7), + P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, MEM_WC, 8), + P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, MEM_WT, 9), + P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, MEM_WP, 10), + P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, MEM_WB, 11), + P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, OWN, 13), + P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, OTHER, 14), + P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, PREFETCH, 15), + + P4_MAKE_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DRDY_DRV, 0), + P4_MAKE_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DRDY_OWN, 1), + P4_MAKE_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DRDY_OTHER, 2), + P4_MAKE_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DBSY_DRV, 3), + P4_MAKE_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DBSY_OWN, 4), + P4_MAKE_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DBSY_OTHER, 5), + + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_TYPE0, 0), + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_TYPE1, 1), + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_LEN0, 2), + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_LEN1, 3), + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_IO_TYPE, 5), + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_LOCK_TYPE, 6), + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_CACHE_TYPE, 7), + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_SPLIT_TYPE, 8), + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_DEM_TYPE, 9), + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_ORD_TYPE, 10), + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, MEM_TYPE0, 11), + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, MEM_TYPE1, 12), + P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, MEM_TYPE2, 13), + + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_TYPE0, 0), + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_TYPE1, 1), + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_LEN0, 2), + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_LEN1, 3), + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_IO_TYPE, 5), + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_LOCK_TYPE, 6), + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_CACHE_TYPE, 7), + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_SPLIT_TYPE, 8), + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_DEM_TYPE, 9), + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_ORD_TYPE, 10), + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, MEM_TYPE0, 11), + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, MEM_TYPE1, 12), + P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, MEM_TYPE2, 13), + + P4_MAKE_EVENT_ATTR(P4_SSE_INPUT_ASSIST, ALL, 15), + + P4_MAKE_EVENT_ATTR(P4_PACKED_SP_UOP, ALL, 15), + + P4_MAKE_EVENT_ATTR(P4_PACKED_DP_UOP, ALL, 15), + + P4_MAKE_EVENT_ATTR(P4_SCALAR_SP_UOP, ALL, 15), + + P4_MAKE_EVENT_ATTR(P4_SCALAR_DP_UOP, ALL, 15), + + P4_MAKE_EVENT_ATTR(P4_64BIT_MMX_UOP, ALL, 15), + + P4_MAKE_EVENT_ATTR(P4_128BIT_MMX_UOP, ALL, 15), + + P4_MAKE_EVENT_ATTR(P4_X87_FP_UOP, ALL, 15), + + P4_MAKE_EVENT_ATTR(P4_TC_MISC, FLUSH, 4), + + P4_MAKE_EVENT_ATTR(P4_GLOBAL_POWER_EVENTS, RUNNING, 0), + + P4_MAKE_EVENT_ATTR(P4_TC_MS_XFER, CISC, 0), + + P4_MAKE_EVENT_ATTR(P4_UOP_QUEUE_WRITES, FROM_TC_BUILD, 0), + P4_MAKE_EVENT_ATTR(P4_UOP_QUEUE_WRITES, FROM_TC_DELIVER, 1), + P4_MAKE_EVENT_ATTR(P4_UOP_QUEUE_WRITES, FROM_ROM, 2), + + P4_MAKE_EVENT_ATTR(P4_RETIRED_MISPRED_BRANCH_TYPE, CONDITIONAL, 1), + P4_MAKE_EVENT_ATTR(P4_RETIRED_MISPRED_BRANCH_TYPE, CALL, 2), + P4_MAKE_EVENT_ATTR(P4_RETIRED_MISPRED_BRANCH_TYPE, RETURN, 3), + P4_MAKE_EVENT_ATTR(P4_RETIRED_MISPRED_BRANCH_TYPE, INDIRECT, 4), + + P4_MAKE_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, CONDITIONAL, 1), + P4_MAKE_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, CALL, 2), + P4_MAKE_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, RETURN, 3), + P4_MAKE_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, INDIRECT, 4), + + P4_MAKE_EVENT_ATTR(P4_RESOURCE_STALL, SBFULL, 5), + + P4_MAKE_EVENT_ATTR(P4_WC_BUFFER, WCB_EVICTS, 0), + P4_MAKE_EVENT_ATTR(P4_WC_BUFFER, WCB_FULL_EVICTS, 1), + + P4_MAKE_EVENT_ATTR(P4_FRONT_END_EVENT, NBOGUS, 0), + P4_MAKE_EVENT_ATTR(P4_FRONT_END_EVENT, BOGUS, 1), + + P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, NBOGUS0, 0), + P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, NBOGUS1, 1), + P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, NBOGUS2, 2), + P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, NBOGUS3, 3), + P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, BOGUS0, 4), + P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, BOGUS1, 5), + P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, BOGUS2, 6), + P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, BOGUS3, 7), + + P4_MAKE_EVENT_ATTR(P4_REPLAY_EVENT, NBOGUS, 0), + P4_MAKE_EVENT_ATTR(P4_REPLAY_EVENT, BOGUS, 1), + + P4_MAKE_EVENT_ATTR(P4_INSTR_RETIRED, NBOGUSNTAG, 0), + P4_MAKE_EVENT_ATTR(P4_INSTR_RETIRED, NBOGUSTAG, 1), + P4_MAKE_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSNTAG, 2), + P4_MAKE_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSTAG, 3), + + P4_MAKE_EVENT_ATTR(P4_UOPS_RETIRED, NBOGUS, 0), + P4_MAKE_EVENT_ATTR(P4_UOPS_RETIRED, BOGUS, 1), + + P4_MAKE_EVENT_ATTR(P4_UOP_TYPE, TAGLOADS, 1), + P4_MAKE_EVENT_ATTR(P4_UOP_TYPE, TAGSTORES, 2), + + P4_MAKE_EVENT_ATTR(P4_BRANCH_RETIRED, MMNP, 0), + P4_MAKE_EVENT_ATTR(P4_BRANCH_RETIRED, MMNM, 1), + P4_MAKE_EVENT_ATTR(P4_BRANCH_RETIRED, MMTP, 2), + P4_MAKE_EVENT_ATTR(P4_BRANCH_RETIRED, MMTM, 3), + + P4_MAKE_EVENT_ATTR(P4_MISPRED_BRANCH_RETIRED, NBOGUS, 0), + + P4_MAKE_EVENT_ATTR(P4_X87_ASSIST, FPSU, 0), + P4_MAKE_EVENT_ATTR(P4_X87_ASSIST, FPSO, 1), + P4_MAKE_EVENT_ATTR(P4_X87_ASSIST, POAO, 2), + P4_MAKE_EVENT_ATTR(P4_X87_ASSIST, POAU, 3), + P4_MAKE_EVENT_ATTR(P4_X87_ASSIST, PREA, 4), + + P4_MAKE_EVENT_ATTR(P4_MACHINE_CLEAR, CLEAR, 0), + P4_MAKE_EVENT_ATTR(P4_MACHINE_CLEAR, MOCLEAR, 1), + P4_MAKE_EVENT_ATTR(P4_MACHINE_CLEAR, SMCLEAR, 2), + + P4_MAKE_EVENT_ATTR(P4_INSTR_COMPLETED, NBOGUS, 0), + P4_MAKE_EVENT_ATTR(P4_INSTR_COMPLETED, BOGUS, 1), +}; + +#endif /* PERF_EVENT_P4_H */ diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index e24f637..e6a3f5f 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -190,6 +190,8 @@ struct x86_pmu { void (*enable_all)(void); void (*enable)(struct perf_event *); void (*disable)(struct perf_event *); + int (*hw_config)(struct perf_event_attr *attr, struct hw_perf_event *hwc); + int (*schedule_events)(struct cpu_hw_events *cpuc, int n, int *assign); unsigned eventsel; unsigned perfctr; u64 (*event_map)(int); @@ -415,6 +417,25 @@ set_ext_hw_attr(struct hw_perf_event *hwc, struct perf_event_attr *attr) return 0; } +static int x86_hw_config(struct perf_event_attr *attr, struct hw_perf_event *hwc) +{ + /* + * Generate PMC IRQs: + * (keep 'enabled' bit clear for now) + */ + hwc->config = ARCH_PERFMON_EVENTSEL_INT; + + /* + * Count user and OS events unless requested not to + */ + if (!attr->exclude_user) + hwc->config |= ARCH_PERFMON_EVENTSEL_USR; + if (!attr->exclude_kernel) + hwc->config |= ARCH_PERFMON_EVENTSEL_OS; + + return 0; +} + /* * Setup the hardware configuration for a given attr_type */ @@ -446,23 +467,13 @@ static int __hw_perf_event_init(struct perf_event *event) event->destroy = hw_perf_event_destroy; - /* - * Generate PMC IRQs: - * (keep 'enabled' bit clear for now) - */ - hwc->config = ARCH_PERFMON_EVENTSEL_INT; - hwc->idx = -1; hwc->last_cpu = -1; hwc->last_tag = ~0ULL; - /* - * Count user and OS events unless requested not to. - */ - if (!attr->exclude_user) - hwc->config |= ARCH_PERFMON_EVENTSEL_USR; - if (!attr->exclude_kernel) - hwc->config |= ARCH_PERFMON_EVENTSEL_OS; + /* Processor specifics */ + if (x86_pmu.hw_config(attr, hwc)) + return -EOPNOTSUPP; if (!hwc->sample_period) { hwc->sample_period = x86_pmu.max_period; @@ -517,7 +528,7 @@ static int __hw_perf_event_init(struct perf_event *event) return -EOPNOTSUPP; /* BTS is currently only allowed for user-mode. */ - if (hwc->config & ARCH_PERFMON_EVENTSEL_OS) + if (!attr->exclude_kernel) return -EOPNOTSUPP; } @@ -931,7 +942,7 @@ static int x86_pmu_enable(struct perf_event *event) if (n < 0) return n; - ret = x86_schedule_events(cpuc, n, assign); + ret = x86_pmu.schedule_events(cpuc, n, assign); if (ret) return ret; /* @@ -1263,7 +1274,7 @@ int hw_perf_group_sched_in(struct perf_event *leader, if (n0 < 0) return n0; - ret = x86_schedule_events(cpuc, n0, assign); + ret = x86_pmu.schedule_events(cpuc, n0, assign); if (ret) return ret; @@ -1313,6 +1324,7 @@ undo: #include "perf_event_amd.c" #include "perf_event_p6.c" +#include "perf_event_p4.c" #include "perf_event_intel_lbr.c" #include "perf_event_intel_ds.c" #include "perf_event_intel.c" @@ -1515,7 +1527,7 @@ static int validate_group(struct perf_event *event) fake_cpuc->n_events = n; - ret = x86_schedule_events(fake_cpuc, n, NULL); + ret = x86_pmu.schedule_events(fake_cpuc, n, NULL); out_free: kfree(fake_cpuc); diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c index 573458f..358a8e3 100644 --- a/arch/x86/kernel/cpu/perf_event_amd.c +++ b/arch/x86/kernel/cpu/perf_event_amd.c @@ -363,6 +363,8 @@ static __initconst struct x86_pmu amd_pmu = { .enable_all = x86_pmu_enable_all, .enable = x86_pmu_enable_event, .disable = x86_pmu_disable_event, + .hw_config = x86_hw_config, + .schedule_events = x86_schedule_events, .eventsel = MSR_K7_EVNTSEL0, .perfctr = MSR_K7_PERFCTR0, .event_map = amd_pmu_event_map, diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 971dc6e..044b843 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -749,6 +749,8 @@ static __initconst struct x86_pmu core_pmu = { .enable_all = x86_pmu_enable_all, .enable = x86_pmu_enable_event, .disable = x86_pmu_disable_event, + .hw_config = x86_hw_config, + .schedule_events = x86_schedule_events, .eventsel = MSR_ARCH_PERFMON_EVENTSEL0, .perfctr = MSR_ARCH_PERFMON_PERFCTR0, .event_map = intel_pmu_event_map, @@ -786,6 +788,8 @@ static __initconst struct x86_pmu intel_pmu = { .enable_all = intel_pmu_enable_all, .enable = intel_pmu_enable_event, .disable = intel_pmu_disable_event, + .hw_config = x86_hw_config, + .schedule_events = x86_schedule_events, .eventsel = MSR_ARCH_PERFMON_EVENTSEL0, .perfctr = MSR_ARCH_PERFMON_PERFCTR0, .event_map = intel_pmu_event_map, @@ -839,12 +843,13 @@ static __init int intel_pmu_init(void) int version; if (!cpu_has(&boot_cpu_data, X86_FEATURE_ARCH_PERFMON)) { - /* check for P6 processor family */ - if (boot_cpu_data.x86 == 6) { - return p6_pmu_init(); - } else { + switch (boot_cpu_data.x86) { + case 0x6: + return p6_pmu_init(); + case 0xf: + return p4_pmu_init(); + } return -ENODEV; - } } /* diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c new file mode 100644 index 0000000..381f593 --- /dev/null +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -0,0 +1,607 @@ +/* + * Netburst Perfomance Events (P4, old Xeon) + * + * Copyright (C) 2010 Parallels, Inc., Cyrill Gorcunov + * Copyright (C) 2010 Intel Corporation, Lin Ming + * + * For licencing details see kernel-base/COPYING + */ + +#ifdef CONFIG_CPU_SUP_INTEL + +#include + +/* + * array indices: 0,1 - HT threads, used with HT enabled cpu + */ +struct p4_event_template { + u32 opcode; /* ESCR event + CCCR selector */ + u64 config; /* packed predefined bits */ + int dep; /* upstream dependency event index */ + unsigned int emask; /* ESCR EventMask */ + unsigned int escr_msr[2]; /* ESCR MSR for this event */ + unsigned int cntr[2]; /* counter index (offset) */ +}; + +struct p4_pmu_res { + /* maps hw_conf::idx into template for ESCR sake */ + struct p4_event_template *tpl[ARCH_P4_MAX_CCCR]; +}; + +static DEFINE_PER_CPU(struct p4_pmu_res, p4_pmu_config); + +/* + * WARN: CCCR1 doesn't have a working enable bit so try to not + * use it if possible + * + * Also as only we start to support raw events we will need to + * append _all_ P4_EVENT_PACK'ed events here + */ +struct p4_event_template p4_templates[] = { + [0] = { + .opcode = P4_UOP_TYPE, + .config = 0, + .dep = -1, + .emask = + P4_EVENT_ATTR(P4_UOP_TYPE, TAGLOADS) | + P4_EVENT_ATTR(P4_UOP_TYPE, TAGSTORES), + .escr_msr = { MSR_P4_RAT_ESCR0, MSR_P4_RAT_ESCR1 }, + .cntr = { 16, 17 }, + }, + [1] = { + .opcode = P4_GLOBAL_POWER_EVENTS, + .config = 0, + .dep = -1, + .emask = + P4_EVENT_ATTR(P4_GLOBAL_POWER_EVENTS, RUNNING), + .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, + .cntr = { 0, 2 }, + }, + [2] = { + .opcode = P4_INSTR_RETIRED, + .config = 0, + .dep = 0, /* needs front-end tagging */ + .emask = + P4_EVENT_ATTR(P4_INSTR_RETIRED, NBOGUSNTAG) | + P4_EVENT_ATTR(P4_INSTR_RETIRED, NBOGUSTAG) | + P4_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSNTAG) | + P4_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSTAG), + .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR3 }, + .cntr = { 12, 14 }, + }, + [3] = { + .opcode = P4_BSQ_CACHE_REFERENCE, + .config = 0, + .dep = -1, + .emask = + P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITS) | + P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITE) | + P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITM) | + P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_HITS) | + P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_HITE) | + P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_HITM), + .escr_msr = { MSR_P4_BSU_ESCR0, MSR_P4_BSU_ESCR1 }, + .cntr = { 0, 2 }, + }, + [4] = { + .opcode = P4_BSQ_CACHE_REFERENCE, + .config = 0, + .dep = -1, + .emask = + P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_MISS) | + P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_MISS) | + P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, WR_2ndL_MISS), + .escr_msr = { MSR_P4_BSU_ESCR0, MSR_P4_BSU_ESCR1 }, + .cntr = { 0, 3 }, + }, + [5] = { + .opcode = P4_RETIRED_BRANCH_TYPE, + .config = 0, + .dep = -1, + .emask = + P4_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, CONDITIONAL) | + P4_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, CALL) | + P4_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, RETURN) | + P4_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, INDIRECT), + .escr_msr = { MSR_P4_TBPU_ESCR0, MSR_P4_TBPU_ESCR1 }, + .cntr = { 4, 6 }, + }, + [6] = { + .opcode = P4_MISPRED_BRANCH_RETIRED, + .config = 0, + .dep = -1, + .emask = + P4_EVENT_ATTR(P4_MISPRED_BRANCH_RETIRED, NBOGUS), + .escr_msr = { MSR_P4_CRU_ESCR0, MSR_P4_CRU_ESCR1 }, + .cntr = { 12, 14 }, + }, + [7] = { + .opcode = P4_FSB_DATA_ACTIVITY, + .config = p4_config_pack_cccr(P4_CCCR_EDGE | P4_CCCR_COMPARE), + .dep = -1, + .emask = + P4_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DRDY_DRV) | + P4_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DRDY_OWN), + .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, + .cntr = { 0, 2 }, + }, +}; + +static struct p4_event_template *p4_event_map[PERF_COUNT_HW_MAX] = { + /* non-halted CPU clocks */ + [PERF_COUNT_HW_CPU_CYCLES] = &p4_templates[1], + + /* retired instructions: dep on tagging the FSB */ + [PERF_COUNT_HW_INSTRUCTIONS] = &p4_templates[2], + + /* cache hits */ + [PERF_COUNT_HW_CACHE_REFERENCES] = &p4_templates[3], + + /* cache misses */ + [PERF_COUNT_HW_CACHE_MISSES] = &p4_templates[4], + + /* branch instructions retired */ + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = &p4_templates[5], + + /* mispredicted branches retired */ + [PERF_COUNT_HW_BRANCH_MISSES] = &p4_templates[6], + + /* bus ready clocks (cpu is driving #DRDY_DRV\#DRDY_OWN): */ + [PERF_COUNT_HW_BUS_CYCLES] = &p4_templates[7], +}; + +static u64 p4_pmu_event_map(int hw_event) +{ + struct p4_event_template *tpl; + u64 config; + + if (hw_event > ARRAY_SIZE(p4_event_map)) { + printk_once(KERN_ERR "PMU: Incorrect event index\n"); + return 0; + } + tpl = p4_event_map[hw_event]; + + /* + * fill config up according to + * a predefined event template + */ + config = tpl->config; + config |= p4_config_pack_escr(P4_EVENT_UNPACK_EVENT(tpl->opcode) << P4_EVNTSEL_EVENT_SHIFT); + config |= p4_config_pack_escr(tpl->emask << P4_EVNTSEL_EVENTMASK_SHIFT); + config |= p4_config_pack_cccr(P4_EVENT_UNPACK_SELECTOR(tpl->opcode) << P4_CCCR_ESCR_SELECT_SHIFT); + + /* on HT machine we need a special bit */ + if (p4_ht_active() && p4_ht_thread(raw_smp_processor_id())) + config = p4_set_ht_bit(config); + + return config; +} + +/* + * Note that we still have 5 events (from global events SDM list) + * intersected in opcode+emask bits so we will need another + * scheme there do distinguish templates. + */ +static inline int p4_pmu_emask_match(unsigned int dst, unsigned int src) +{ + return dst & src; +} + +static struct p4_event_template *p4_pmu_template_lookup(u64 config) +{ + u32 opcode = p4_config_unpack_opcode(config); + unsigned int emask = p4_config_unpack_emask(config); + unsigned int i; + + for (i = 0; i < ARRAY_SIZE(p4_templates); i++) { + if (opcode == p4_templates[i].opcode && + p4_pmu_emask_match(emask, p4_templates[i].emask)) + return &p4_templates[i]; + } + + return NULL; +} + +/* + * We don't control raw events so it's up to the caller + * to pass sane values (and we don't count the thread number + * on HT machine but allow HT-compatible specifics to be + * passed on) + */ +static u64 p4_pmu_raw_event(u64 hw_event) +{ + return hw_event & + (p4_config_pack_escr(P4_EVNTSEL_MASK_HT) | + p4_config_pack_cccr(P4_CCCR_MASK_HT)); +} + +static int p4_hw_config(struct perf_event_attr *attr, struct hw_perf_event *hwc) +{ + int cpu = raw_smp_processor_id(); + + /* + * the reason we use cpu that early is that: if we get scheduled + * first time on the same cpu -- we will not need swap thread + * specific flags in config (and will save some cpu cycles) + */ + + /* CCCR by default */ + hwc->config = p4_config_pack_cccr(p4_default_cccr_conf(cpu)); + + /* Count user and OS events unless not requested to */ + hwc->config |= p4_config_pack_escr(p4_default_escr_conf(cpu, attr->exclude_kernel, + attr->exclude_user)); + return 0; +} + +static inline void p4_pmu_clear_cccr_ovf(struct hw_perf_event *hwc) +{ + unsigned long dummy; + + rdmsrl(hwc->config_base + hwc->idx, dummy); + if (dummy & P4_CCCR_OVF) { + (void)checking_wrmsrl(hwc->config_base + hwc->idx, + ((u64)dummy) & ~P4_CCCR_OVF); + } +} + +static inline void p4_pmu_disable_event(struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + + /* + * If event gets disabled while counter is in overflowed + * state we need to clear P4_CCCR_OVF, otherwise interrupt get + * asserted again and again + */ + (void)checking_wrmsrl(hwc->config_base + hwc->idx, + (u64)(p4_config_unpack_cccr(hwc->config)) & + ~P4_CCCR_ENABLE & ~P4_CCCR_OVF); +} + +static void p4_pmu_disable_all(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + int idx; + + for (idx = 0; idx < x86_pmu.num_events; idx++) { + struct perf_event *event = cpuc->events[idx]; + if (!test_bit(idx, cpuc->active_mask)) + continue; + p4_pmu_disable_event(event); + } +} + +static void p4_pmu_enable_event(struct perf_event *event) +{ + struct hw_perf_event *hwc = &event->hw; + int thread = p4_ht_config_thread(hwc->config); + u64 escr_conf = p4_config_unpack_escr(p4_clear_ht_bit(hwc->config)); + u64 escr_base; + struct p4_event_template *tpl; + struct p4_pmu_res *c; + + /* + * some preparation work from per-cpu private fields + * since we need to find out which ESCR to use + */ + c = &__get_cpu_var(p4_pmu_config); + tpl = c->tpl[hwc->idx]; + if (!tpl) { + pr_crit("%s: Wrong index: %d\n", __func__, hwc->idx); + return; + } + escr_base = (u64)tpl->escr_msr[thread]; + + /* + * - we dont support cascaded counters yet + * - and counter 1 is broken (erratum) + */ + WARN_ON_ONCE(p4_is_event_cascaded(hwc->config)); + WARN_ON_ONCE(hwc->idx == 1); + + (void)checking_wrmsrl(escr_base, escr_conf); + (void)checking_wrmsrl(hwc->config_base + hwc->idx, + (u64)(p4_config_unpack_cccr(hwc->config)) | P4_CCCR_ENABLE); +} + +static void p4_pmu_enable_all(void) +{ + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + int idx; + + for (idx = 0; idx < x86_pmu.num_events; idx++) { + struct perf_event *event = cpuc->events[idx]; + if (!test_bit(idx, cpuc->active_mask)) + continue; + p4_pmu_enable_event(event); + } +} + +static int p4_pmu_handle_irq(struct pt_regs *regs) +{ + struct perf_sample_data data; + struct cpu_hw_events *cpuc; + struct perf_event *event; + struct hw_perf_event *hwc; + int idx, handled = 0; + u64 val; + + data.addr = 0; + data.raw = NULL; + + cpuc = &__get_cpu_var(cpu_hw_events); + + for (idx = 0; idx < x86_pmu.num_events; idx++) { + + if (!test_bit(idx, cpuc->active_mask)) + continue; + + event = cpuc->events[idx]; + hwc = &event->hw; + + WARN_ON_ONCE(hwc->idx != idx); + + /* + * FIXME: Redundant call, actually not needed + * but just to check if we're screwed + */ + p4_pmu_clear_cccr_ovf(hwc); + + val = x86_perf_event_update(event); + if (val & (1ULL << (x86_pmu.event_bits - 1))) + continue; + + /* + * event overflow + */ + handled = 1; + data.period = event->hw.last_period; + + if (!x86_perf_event_set_period(event)) + continue; + if (perf_event_overflow(event, 1, &data, regs)) + p4_pmu_disable_event(event); + } + + if (handled) { + /* p4 quirk: unmask it again */ + apic_write(APIC_LVTPC, apic_read(APIC_LVTPC) & ~APIC_LVT_MASKED); + inc_irq_stat(apic_perf_irqs); + } + + return handled; +} + +/* + * swap thread specific fields according to a thread + * we are going to run on + */ +static void p4_pmu_swap_config_ts(struct hw_perf_event *hwc, int cpu) +{ + u32 escr, cccr; + + /* + * we either lucky and continue on same cpu or no HT support + */ + if (!p4_should_swap_ts(hwc->config, cpu)) + return; + + /* + * the event is migrated from an another logical + * cpu, so we need to swap thread specific flags + */ + + escr = p4_config_unpack_escr(hwc->config); + cccr = p4_config_unpack_cccr(hwc->config); + + if (p4_ht_thread(cpu)) { + cccr &= ~P4_CCCR_OVF_PMI_T0; + cccr |= P4_CCCR_OVF_PMI_T1; + if (escr & P4_EVNTSEL_T0_OS) { + escr &= ~P4_EVNTSEL_T0_OS; + escr |= P4_EVNTSEL_T1_OS; + } + if (escr & P4_EVNTSEL_T0_USR) { + escr &= ~P4_EVNTSEL_T0_USR; + escr |= P4_EVNTSEL_T1_USR; + } + hwc->config = p4_config_pack_escr(escr); + hwc->config |= p4_config_pack_cccr(cccr); + hwc->config |= P4_CONFIG_HT; + } else { + cccr &= ~P4_CCCR_OVF_PMI_T1; + cccr |= P4_CCCR_OVF_PMI_T0; + if (escr & P4_EVNTSEL_T1_OS) { + escr &= ~P4_EVNTSEL_T1_OS; + escr |= P4_EVNTSEL_T0_OS; + } + if (escr & P4_EVNTSEL_T1_USR) { + escr &= ~P4_EVNTSEL_T1_USR; + escr |= P4_EVNTSEL_T0_USR; + } + hwc->config = p4_config_pack_escr(escr); + hwc->config |= p4_config_pack_cccr(cccr); + hwc->config &= ~P4_CONFIG_HT; + } +} + +/* ESCRs are not sequential in memory so we need a map */ +static unsigned int p4_escr_map[ARCH_P4_TOTAL_ESCR] = { + MSR_P4_ALF_ESCR0, /* 0 */ + MSR_P4_ALF_ESCR1, /* 1 */ + MSR_P4_BPU_ESCR0, /* 2 */ + MSR_P4_BPU_ESCR1, /* 3 */ + MSR_P4_BSU_ESCR0, /* 4 */ + MSR_P4_BSU_ESCR1, /* 5 */ + MSR_P4_CRU_ESCR0, /* 6 */ + MSR_P4_CRU_ESCR1, /* 7 */ + MSR_P4_CRU_ESCR2, /* 8 */ + MSR_P4_CRU_ESCR3, /* 9 */ + MSR_P4_CRU_ESCR4, /* 10 */ + MSR_P4_CRU_ESCR5, /* 11 */ + MSR_P4_DAC_ESCR0, /* 12 */ + MSR_P4_DAC_ESCR1, /* 13 */ + MSR_P4_FIRM_ESCR0, /* 14 */ + MSR_P4_FIRM_ESCR1, /* 15 */ + MSR_P4_FLAME_ESCR0, /* 16 */ + MSR_P4_FLAME_ESCR1, /* 17 */ + MSR_P4_FSB_ESCR0, /* 18 */ + MSR_P4_FSB_ESCR1, /* 19 */ + MSR_P4_IQ_ESCR0, /* 20 */ + MSR_P4_IQ_ESCR1, /* 21 */ + MSR_P4_IS_ESCR0, /* 22 */ + MSR_P4_IS_ESCR1, /* 23 */ + MSR_P4_ITLB_ESCR0, /* 24 */ + MSR_P4_ITLB_ESCR1, /* 25 */ + MSR_P4_IX_ESCR0, /* 26 */ + MSR_P4_IX_ESCR1, /* 27 */ + MSR_P4_MOB_ESCR0, /* 28 */ + MSR_P4_MOB_ESCR1, /* 29 */ + MSR_P4_MS_ESCR0, /* 30 */ + MSR_P4_MS_ESCR1, /* 31 */ + MSR_P4_PMH_ESCR0, /* 32 */ + MSR_P4_PMH_ESCR1, /* 33 */ + MSR_P4_RAT_ESCR0, /* 34 */ + MSR_P4_RAT_ESCR1, /* 35 */ + MSR_P4_SAAT_ESCR0, /* 36 */ + MSR_P4_SAAT_ESCR1, /* 37 */ + MSR_P4_SSU_ESCR0, /* 38 */ + MSR_P4_SSU_ESCR1, /* 39 */ + MSR_P4_TBPU_ESCR0, /* 40 */ + MSR_P4_TBPU_ESCR1, /* 41 */ + MSR_P4_TC_ESCR0, /* 42 */ + MSR_P4_TC_ESCR1, /* 43 */ + MSR_P4_U2L_ESCR0, /* 44 */ + MSR_P4_U2L_ESCR1, /* 45 */ +}; + +static int p4_get_escr_idx(unsigned int addr) +{ + unsigned int i; + + for (i = 0; i < ARRAY_SIZE(p4_escr_map); i++) { + if (addr == p4_escr_map[i]) + return i; + } + + return -1; +} + +static int p4_pmu_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign) +{ + unsigned long used_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; + unsigned long escr_mask[BITS_TO_LONGS(ARCH_P4_TOTAL_ESCR)]; + + struct hw_perf_event *hwc; + struct p4_event_template *tpl; + struct p4_pmu_res *c; + int cpu = raw_smp_processor_id(); + int escr_idx, thread, i, num; + + bitmap_zero(used_mask, X86_PMC_IDX_MAX); + bitmap_zero(escr_mask, ARCH_P4_TOTAL_ESCR); + + c = &__get_cpu_var(p4_pmu_config); + /* + * Firstly find out which resource events are going + * to use, if ESCR+CCCR tuple is already borrowed + * then get out of here + */ + for (i = 0, num = n; i < n; i++, num--) { + hwc = &cpuc->event_list[i]->hw; + tpl = p4_pmu_template_lookup(hwc->config); + if (!tpl) + goto done; + thread = p4_ht_thread(cpu); + escr_idx = p4_get_escr_idx(tpl->escr_msr[thread]); + if (escr_idx == -1) + goto done; + + /* already allocated and remains on the same cpu */ + if (hwc->idx != -1 && !p4_should_swap_ts(hwc->config, cpu)) { + if (assign) + assign[i] = hwc->idx; + /* upstream dependent event */ + if (unlikely(tpl->dep != -1)) + printk_once(KERN_WARNING "PMU: Dep events are " + "not implemented yet\n"); + goto reserve; + } + + /* it may be already borrowed */ + if (test_bit(tpl->cntr[thread], used_mask) || + test_bit(escr_idx, escr_mask)) + goto done; + + /* + * ESCR+CCCR+COUNTERs are available to use lets swap + * thread specific bits, push assigned bits + * back and save template into per-cpu + * area (which will allow us to find out the ESCR + * to be used at moment of "enable event via real MSR") + */ + p4_pmu_swap_config_ts(hwc, cpu); + if (assign) { + assign[i] = tpl->cntr[thread]; + c->tpl[assign[i]] = tpl; + } +reserve: + set_bit(tpl->cntr[thread], used_mask); + set_bit(escr_idx, escr_mask); + } + +done: + return num ? -ENOSPC : 0; +} + +static __initconst struct x86_pmu p4_pmu = { + .name = "Netburst P4/Xeon", + .handle_irq = p4_pmu_handle_irq, + .disable_all = p4_pmu_disable_all, + .enable_all = p4_pmu_enable_all, + .enable = p4_pmu_enable_event, + .disable = p4_pmu_disable_event, + .eventsel = MSR_P4_BPU_CCCR0, + .perfctr = MSR_P4_BPU_PERFCTR0, + .event_map = p4_pmu_event_map, + .raw_event = p4_pmu_raw_event, + .max_events = ARRAY_SIZE(p4_event_map), + .get_event_constraints = x86_get_event_constraints, + /* + * IF HT disabled we may need to use all + * ARCH_P4_MAX_CCCR counters simulaneously + * though leave it restricted at moment assuming + * HT is on + */ + .num_events = ARCH_P4_MAX_CCCR, + .apic = 1, + .event_bits = 40, + .event_mask = (1ULL << 40) - 1, + .max_period = (1ULL << 39) - 1, + .hw_config = p4_hw_config, + .schedule_events = p4_pmu_schedule_events, +}; + +static __init int p4_pmu_init(void) +{ + unsigned int low, high; + + /* If we get stripped -- indexig fails */ + BUILD_BUG_ON(ARCH_P4_MAX_CCCR > X86_PMC_MAX_GENERIC); + + rdmsr(MSR_IA32_MISC_ENABLE, low, high); + if (!(low & (1 << 7))) { + pr_cont("unsupported Netburst CPU model %d ", + boot_cpu_data.x86_model); + return -ENODEV; + } + + pr_cont("Netburst events, "); + + x86_pmu = p4_pmu; + + return 0; +} + +#endif /* CONFIG_CPU_SUP_INTEL */ diff --git a/arch/x86/kernel/cpu/perf_event_p6.c b/arch/x86/kernel/cpu/perf_event_p6.c index a330485..6ff4d01 100644 --- a/arch/x86/kernel/cpu/perf_event_p6.c +++ b/arch/x86/kernel/cpu/perf_event_p6.c @@ -109,6 +109,8 @@ static __initconst struct x86_pmu p6_pmu = { .enable_all = p6_pmu_enable_all, .enable = p6_pmu_enable_event, .disable = p6_pmu_disable_event, + .hw_config = x86_hw_config, + .schedule_events = x86_schedule_events, .eventsel = MSR_P6_EVNTSEL0, .perfctr = MSR_P6_PERFCTR0, .event_map = p6_pmu_event_map, -- cgit v0.10.2 From 6230f2c7ef01a69e2ba9370326572c287209d32a Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 11 Mar 2010 15:53:12 -0300 Subject: perf record: Mention paranoid sysctl when failing to create counter MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit [acme@mica linux-2.6-tip]$ perf record -a -f Fatal: Permission error - are you root? Consider tweaking /proc/sys/kernel/perf_event_paranoid. [acme@mica linux-2.6-tip]$ Suggested-by: Ingo Molnar Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1268333592-30872-2-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 771533c..6e4a393 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -280,7 +280,8 @@ try_again: int err = errno; if (err == EPERM || err == EACCES) - die("Permission error - are you root?\n"); + die("Permission error - are you root?\n" + "\t Consider tweaking /proc/sys/kernel/perf_event_paranoid.\n"); else if (err == ENODEV && profile_cpu != -1) die("No such device - did you specify an out-of-range profile CPU?\n"); -- cgit v0.10.2 From c12ec0a2d94001003dfb929ce14c287fca0522b0 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Thu, 11 Mar 2010 14:09:47 -0800 Subject: paride: fix off-by-one test With `while (j++ < PX_SPIN)' j reaches PX_SPIN + 1 after the loop. This is probably unlikely to produce a problem. Signed-off-by: Roel Kluin Cc: Jens Axboe Signed-off-by: Andrew Morton Signed-off-by: Jens Axboe diff --git a/drivers/block/paride/pcd.c b/drivers/block/paride/pcd.c index 8866ca3..71acf4e 100644 --- a/drivers/block/paride/pcd.c +++ b/drivers/block/paride/pcd.c @@ -341,11 +341,11 @@ static int pcd_wait(struct pcd_unit *cd, int go, int stop, char *fun, char *msg) && (j++ < PCD_SPIN)) udelay(PCD_DELAY); - if ((r & (IDE_ERR & stop)) || (j >= PCD_SPIN)) { + if ((r & (IDE_ERR & stop)) || (j > PCD_SPIN)) { s = read_reg(cd, 7); e = read_reg(cd, 1); p = read_reg(cd, 2); - if (j >= PCD_SPIN) + if (j > PCD_SPIN) e |= 0x100; if (fun) printk("%s: %s %s: alt=0x%x stat=0x%x err=0x%x" diff --git a/drivers/block/paride/pf.c b/drivers/block/paride/pf.c index ddb4f9a..c059aab 100644 --- a/drivers/block/paride/pf.c +++ b/drivers/block/paride/pf.c @@ -391,11 +391,11 @@ static int pf_wait(struct pf_unit *pf, int go, int stop, char *fun, char *msg) && (j++ < PF_SPIN)) udelay(PF_SPIN_DEL); - if ((r & (STAT_ERR & stop)) || (j >= PF_SPIN)) { + if ((r & (STAT_ERR & stop)) || (j > PF_SPIN)) { s = read_reg(pf, 7); e = read_reg(pf, 1); p = read_reg(pf, 2); - if (j >= PF_SPIN) + if (j > PF_SPIN) e |= 0x100; if (fun) printk("%s: %s %s: alt=0x%x stat=0x%x err=0x%x" diff --git a/drivers/block/paride/pt.c b/drivers/block/paride/pt.c index 1e4006e..bc5825f 100644 --- a/drivers/block/paride/pt.c +++ b/drivers/block/paride/pt.c @@ -274,11 +274,11 @@ static int pt_wait(struct pt_unit *tape, int go, int stop, char *fun, char *msg) && (j++ < PT_SPIN)) udelay(PT_SPIN_DEL); - if ((r & (STAT_ERR & stop)) || (j >= PT_SPIN)) { + if ((r & (STAT_ERR & stop)) || (j > PT_SPIN)) { s = read_reg(pi, 7); e = read_reg(pi, 1); p = read_reg(pi, 2); - if (j >= PT_SPIN) + if (j > PT_SPIN) e |= 0x100; if (fun) printk("%s: %s %s: alt=0x%x stat=0x%x err=0x%x" -- cgit v0.10.2 From f11c9c5c259cb2c3d698548dc3936f773ab1f5b9 Mon Sep 17 00:00:00 2001 From: Edward Shishkin Date: Thu, 11 Mar 2010 14:09:47 -0800 Subject: vfs: improve writeback_inodes_wb() Do not pin/unpin superblock for every inode in writeback_inodes_wb(), pin it for the whole group of inodes which belong to the same superblock and call writeback_sb_inodes() handler for them. Signed-off-by: Edward Shishkin Cc: Jens Axboe Cc: Wu Fengguang Signed-off-by: Andrew Morton Signed-off-by: Jens Axboe diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 76fc4d5..6841eff 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -553,108 +553,85 @@ select_queue: return ret; } -static void unpin_sb_for_writeback(struct super_block **psb) +static void unpin_sb_for_writeback(struct super_block *sb) { - struct super_block *sb = *psb; - - if (sb) { - up_read(&sb->s_umount); - put_super(sb); - *psb = NULL; - } + up_read(&sb->s_umount); + put_super(sb); } +enum sb_pin_state { + SB_PINNED, + SB_NOT_PINNED, + SB_PIN_FAILED +}; + /* * For WB_SYNC_NONE writeback, the caller does not have the sb pinned * before calling writeback. So make sure that we do pin it, so it doesn't * go away while we are writing inodes from it. - * - * Returns 0 if the super was successfully pinned (or pinning wasn't needed), - * 1 if we failed. */ -static int pin_sb_for_writeback(struct writeback_control *wbc, - struct inode *inode, struct super_block **psb) +static enum sb_pin_state pin_sb_for_writeback(struct writeback_control *wbc, + struct super_block *sb) { - struct super_block *sb = inode->i_sb; - - /* - * If this sb is already pinned, nothing more to do. If not and - * *psb is non-NULL, unpin the old one first - */ - if (sb == *psb) - return 0; - else if (*psb) - unpin_sb_for_writeback(psb); - /* * Caller must already hold the ref for this */ if (wbc->sync_mode == WB_SYNC_ALL) { WARN_ON(!rwsem_is_locked(&sb->s_umount)); - return 0; + return SB_NOT_PINNED; } - spin_lock(&sb_lock); sb->s_count++; if (down_read_trylock(&sb->s_umount)) { if (sb->s_root) { spin_unlock(&sb_lock); - goto pinned; + return SB_PINNED; } /* * umounted, drop rwsem again and fall through to failure */ up_read(&sb->s_umount); } - sb->s_count--; spin_unlock(&sb_lock); - return 1; -pinned: - *psb = sb; - return 0; + return SB_PIN_FAILED; } -static void writeback_inodes_wb(struct bdi_writeback *wb, - struct writeback_control *wbc) +/* + * Write a portion of b_io inodes which belong to @sb. + * If @wbc->sb != NULL, then find and write all such + * inodes. Otherwise write only ones which go sequentially + * in reverse order. + * Return 1, if the caller writeback routine should be + * interrupted. Otherwise return 0. + */ +static int writeback_sb_inodes(struct super_block *sb, + struct bdi_writeback *wb, + struct writeback_control *wbc) { - struct super_block *sb = wbc->sb, *pin_sb = NULL; - const unsigned long start = jiffies; /* livelock avoidance */ - - spin_lock(&inode_lock); - - if (!wbc->for_kupdate || list_empty(&wb->b_io)) - queue_io(wb, wbc->older_than_this); - while (!list_empty(&wb->b_io)) { - struct inode *inode = list_entry(wb->b_io.prev, - struct inode, i_list); long pages_skipped; - - /* - * super block given and doesn't match, skip this inode - */ - if (sb && sb != inode->i_sb) { + struct inode *inode = list_entry(wb->b_io.prev, + struct inode, i_list); + if (wbc->sb && sb != inode->i_sb) { + /* super block given and doesn't + match, skip this inode */ redirty_tail(inode); continue; } - + if (sb != inode->i_sb) + /* finish with this superblock */ + return 0; if (inode->i_state & (I_NEW | I_WILL_FREE)) { requeue_io(inode); continue; } - /* * Was this inode dirtied after sync_sb_inodes was called? * This keeps sync from extra jobs and livelock. */ - if (inode_dirtied_after(inode, start)) - break; - - if (pin_sb_for_writeback(wbc, inode, &pin_sb)) { - requeue_io(inode); - continue; - } + if (inode_dirtied_after(inode, wbc->wb_start)) + return 1; BUG_ON(inode->i_state & (I_FREEING | I_CLEAR)); __iget(inode); @@ -673,14 +650,50 @@ static void writeback_inodes_wb(struct bdi_writeback *wb, spin_lock(&inode_lock); if (wbc->nr_to_write <= 0) { wbc->more_io = 1; - break; + return 1; } if (!list_empty(&wb->b_more_io)) wbc->more_io = 1; } + /* b_io is empty */ + return 1; +} + +static void writeback_inodes_wb(struct bdi_writeback *wb, + struct writeback_control *wbc) +{ + int ret = 0; - unpin_sb_for_writeback(&pin_sb); + wbc->wb_start = jiffies; /* livelock avoidance */ + spin_lock(&inode_lock); + if (!wbc->for_kupdate || list_empty(&wb->b_io)) + queue_io(wb, wbc->older_than_this); + + while (!list_empty(&wb->b_io)) { + struct inode *inode = list_entry(wb->b_io.prev, + struct inode, i_list); + struct super_block *sb = inode->i_sb; + enum sb_pin_state state; + + if (wbc->sb && sb != wbc->sb) { + /* super block given and doesn't + match, skip this inode */ + redirty_tail(inode); + continue; + } + state = pin_sb_for_writeback(wbc, sb); + + if (state == SB_PIN_FAILED) { + requeue_io(inode); + continue; + } + ret = writeback_sb_inodes(sb, wb, wbc); + if (state == SB_PINNED) + unpin_sb_for_writeback(sb); + if (ret) + break; + } spin_unlock(&inode_lock); /* Leave any unwritten inodes on b_io */ } diff --git a/include/linux/writeback.h b/include/linux/writeback.h index 76e8903..36520de 100644 --- a/include/linux/writeback.h +++ b/include/linux/writeback.h @@ -34,6 +34,9 @@ struct writeback_control { enum writeback_sync_modes sync_mode; unsigned long *older_than_this; /* If !NULL, only write back inodes older than this */ + unsigned long wb_start; /* Time writeback_inodes_wb was + called. This is needed to avoid + extra jobs and livelock */ long nr_to_write; /* Write this many pages, and decrement this for each page written */ long pages_skipped; /* Pages which were not written */ -- cgit v0.10.2 From 0b861225a5890f22445f08ca9cc7a87cff276ff7 Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Fri, 12 Mar 2010 00:50:16 +0300 Subject: x86, perf: Fix NULL deref on not assigned x86_pmu In case of not assigned x86_pmu and software events NULL dereference may being hit via x86_pmu::schedule_events method. Fix it by checking if x86_pmu is initialized at all. Signed-off-by: Cyrill Gorcunov Cc: Lin Ming Cc: Arnaldo Carvalho de Melo Cc: Stephane Eranian Cc: Robert Richter Cc: Frederic Weisbecker Cc: Peter Zijlstra LKML-Reference: <20100311215016.GG25162@lenovo> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index e6a3f5f..5586a02 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -1269,6 +1269,9 @@ int hw_perf_group_sched_in(struct perf_event *leader, int assign[X86_PMC_IDX_MAX]; int n0, n1, ret; + if (!x86_pmu_initialized()) + return 0; + /* n0 = total number of events */ n0 = collect_events(cpuc, leader, true); if (n0 < 0) -- cgit v0.10.2 From fe2197b8bb2f318c1e0b0f589f24f781dd27d1f2 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 11 Mar 2010 20:12:40 -0300 Subject: perf symbols: Bump plt synthesizing warning debug level MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1268349164-5822-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 323c0ae..75cd468 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -862,8 +862,8 @@ out_close: if (err == 0) return nr; out: - pr_warning("%s: problems reading %s PLT info.\n", - __func__, self->long_name); + pr_debug("%s: problems reading %s PLT info.\n", + __func__, self->long_name); return 0; } -- cgit v0.10.2 From 895f0edc3cd04a2a3de2c608ba20821b832a8abb Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 11 Mar 2010 20:12:41 -0300 Subject: perf top: Export get_window_dimensions MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Will be used by the newt code too. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1268349164-5822-2-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index ec48223..57e232f 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -133,7 +133,7 @@ static inline struct symbol *sym_entry__symbol(struct sym_entry *self) return ((void *)self) + symbol_conf.priv_size; } -static void get_term_dimensions(struct winsize *ws) +void get_term_dimensions(struct winsize *ws) { char *s = getenv("LINES"); diff --git a/tools/perf/perf.h b/tools/perf/perf.h index 6fb379b..aa78615 100644 --- a/tools/perf/perf.h +++ b/tools/perf/perf.h @@ -1,6 +1,10 @@ #ifndef _PERF_PERF_H #define _PERF_PERF_H +struct winsize; + +void get_term_dimensions(struct winsize *ws); + #if defined(__i386__) #include "../../arch/x86/include/asm/unistd.h" #define rmb() asm volatile("lock; addl $0,0(%%esp)" ::: "memory") -- cgit v0.10.2 From b4f5296f0eec2aa7061dfd8bb8c0744f095f9bd1 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 11 Mar 2010 20:12:42 -0300 Subject: perf tools: Use eprintf for pr_{err,warning,info} too MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Just like we do for pr_debug, so that we can have a single point where to redirect to the currently used output system, be it stdio or newt. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1268349164-5822-3-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/debug.h b/tools/perf/util/debug.h index c6c24c5..58720a1 100644 --- a/tools/perf/util/debug.h +++ b/tools/perf/util/debug.h @@ -7,8 +7,6 @@ extern int verbose; extern int dump_trace; -int eprintf(int level, - const char *fmt, ...) __attribute__((format(printf, 2, 3))); int dump_printf(const char *fmt, ...) __attribute__((format(printf, 1, 2))); void trace_event(event_t *event); diff --git a/tools/perf/util/include/linux/kernel.h b/tools/perf/util/include/linux/kernel.h index f261165..388ab1b 100644 --- a/tools/perf/util/include/linux/kernel.h +++ b/tools/perf/util/include/linux/kernel.h @@ -85,16 +85,19 @@ simple_strtoul(const char *nptr, char **endptr, int base) return strtoul(nptr, endptr, base); } +int eprintf(int level, + const char *fmt, ...) __attribute__((format(printf, 2, 3))); + #ifndef pr_fmt #define pr_fmt(fmt) fmt #endif #define pr_err(fmt, ...) \ - do { fprintf(stderr, pr_fmt(fmt), ##__VA_ARGS__); } while (0) + eprintf(0, pr_fmt(fmt), ##__VA_ARGS__) #define pr_warning(fmt, ...) \ - do { fprintf(stderr, pr_fmt(fmt), ##__VA_ARGS__); } while (0) + eprintf(0, pr_fmt(fmt), ##__VA_ARGS__) #define pr_info(fmt, ...) \ - do { fprintf(stderr, pr_fmt(fmt), ##__VA_ARGS__); } while (0) + eprintf(0, pr_fmt(fmt), ##__VA_ARGS__) #define pr_debug(fmt, ...) \ eprintf(1, pr_fmt(fmt), ##__VA_ARGS__) #define pr_debugN(n, fmt, ...) \ -- cgit v0.10.2 From dd2ee78dd8e4c6d6f1a333fd60c3dd27d1b07042 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 11 Mar 2010 20:12:43 -0300 Subject: perf tools: Add missing bytes printed in hist_entry__fprintf MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We need those to properly size the browser widht in the newt TUI. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1268349164-5822-4-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index bdcfd61..d43be34 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -455,11 +455,11 @@ static size_t hist_entry_callchain__fprintf(FILE *fp, struct hist_entry *self, return ret; } -static size_t hist_entry__fprintf(struct hist_entry *self, - struct perf_session *pair_session, - bool show_displacement, - long displacement, FILE *fp, - u64 session_total) +size_t hist_entry__fprintf(struct hist_entry *self, + struct perf_session *pair_session, + bool show_displacement, + long displacement, FILE *fp, + u64 session_total) { struct sort_entry *se; u64 count, total; @@ -485,9 +485,9 @@ static size_t hist_entry__fprintf(struct hist_entry *self, if (symbol_conf.show_nr_samples) { if (sep) - fprintf(fp, "%c%lld", *sep, count); + ret += fprintf(fp, "%c%lld", *sep, count); else - fprintf(fp, "%11lld", count); + ret += fprintf(fp, "%11lld", count); } if (pair_session) { @@ -518,9 +518,9 @@ static size_t hist_entry__fprintf(struct hist_entry *self, snprintf(bf, sizeof(bf), " "); if (sep) - fprintf(fp, "%c%s", *sep, bf); + ret += fprintf(fp, "%c%s", *sep, bf); else - fprintf(fp, "%6.6s", bf); + ret += fprintf(fp, "%6.6s", bf); } } @@ -528,7 +528,7 @@ static size_t hist_entry__fprintf(struct hist_entry *self, if (se->elide) continue; - fprintf(fp, "%s", sep ?: " "); + ret += fprintf(fp, "%s", sep ?: " "); ret += se->print(fp, self, se->width ? *se->width : 0); } @@ -544,8 +544,8 @@ static size_t hist_entry__fprintf(struct hist_entry *self, left_margin -= thread__comm_len(self->thread); } - hist_entry_callchain__fprintf(fp, self, session_total, - left_margin); + ret += hist_entry_callchain__fprintf(fp, self, session_total, + left_margin); } return ret; -- cgit v0.10.2 From f9224c5c944b60cf709db4adf1f5195264b8d194 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 11 Mar 2010 20:12:44 -0300 Subject: perf report: Implement initial UI using newt MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Newt has widespread availability and provides a rather simple API as can be seen by the size of this patch. The work needed to support it will benefit other frontends too. In this initial patch it just checks if the output is a tty, if not it falls back to the previous behaviour, also if newt-devel/libnewt-dev is not installed the previous behaviour is maintaned. Pressing enter on a symbol will annotate it, ESC in the annotation window will return to the report symbol list. More work will be done to remove the special casing in color_fprintf, stop using fmemopen/FILE in the printing of hist_entries, etc. Also the annotation doesn't need to be done via spawning "perf annotate" and then browsing its output, we can do better by calling directly the builtin-annotate.c functions, that would then be moved to tools/perf/util/annotate.c and shared with perf top, etc But lets go by baby steps, this patch already improves perf usability by allowing to quickly do annotations on symbols from the report screen and provides a first experimentation with libnewt/TUI integration of tools. Tested on RHEL5 and Fedora12 X86_64 and on Debian PARISC64 to browse a perf.data file collected on a Fedora12 x86_64 box. Signed-off-by: Arnaldo Carvalho de Melo Cc: Avi Kivity Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1268349164-5822-5-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/Makefile b/tools/perf/Makefile index 8a8f52d..0abd25e 100644 --- a/tools/perf/Makefile +++ b/tools/perf/Makefile @@ -513,6 +513,14 @@ else LIB_OBJS += util/probe-finder.o endif +ifneq ($(shell sh -c "(echo '\#include '; echo 'int main(void) { newtInit(); newtCls(); return newtFinished(); }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -lnewt -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) + msg := $(warning newt not found, disables TUI support. Please install newt-devel or libnewt-dev); + BASIC_CFLAGS += -DNO_NEWT_SUPPORT +else + EXTLIBS += -lnewt + LIB_OBJS += util/newt.o +endif + ifndef NO_LIBPERL PERL_EMBED_LDOPTS = `perl -MExtUtils::Embed -e ldopts 2>/dev/null` PERL_EMBED_CCOPTS = `perl -MExtUtils::Embed -e ccopts 2>/dev/null` diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index f815de2..1f9f869 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -267,6 +267,7 @@ static int __cmd_report(void) int ret = -EINVAL; struct perf_session *session; struct rb_node *next; + const char *help = "For a higher level overview, try: perf report --sort comm,dso"; session = perf_session__new(input_name, O_RDONLY, force); if (session == NULL) @@ -301,30 +302,38 @@ static int __cmd_report(void) stats = rb_entry(next, struct event_stat_id, rb_node); perf_session__collapse_resort(&stats->hists); perf_session__output_resort(&stats->hists, stats->stats.total); - if (rb_first(&session->stats_by_id) == - rb_last(&session->stats_by_id)) - fprintf(stdout, "# Samples: %Ld\n#\n", - stats->stats.total); - else - fprintf(stdout, "# Samples: %Ld %s\n#\n", - stats->stats.total, - __event_name(stats->type, stats->config)); - perf_session__fprintf_hists(&stats->hists, NULL, false, stdout, + if (use_browser) + perf_session__browse_hists(&stats->hists, + stats->stats.total, help); + else { + if (rb_first(&session->stats_by_id) == + rb_last(&session->stats_by_id)) + fprintf(stdout, "# Samples: %Ld\n#\n", + stats->stats.total); + else + fprintf(stdout, "# Samples: %Ld %s\n#\n", + stats->stats.total, + __event_name(stats->type, stats->config)); + + perf_session__fprintf_hists(&stats->hists, NULL, false, stdout, stats->stats.total); - fprintf(stdout, "\n\n"); + fprintf(stdout, "\n\n"); + } + next = rb_next(&stats->rb_node); } - if (sort_order == default_sort_order && - parent_pattern == default_parent_pattern) - fprintf(stdout, "#\n# (For a higher level overview, try: perf report --sort comm,dso)\n#\n"); + if (!use_browser && sort_order == default_sort_order && + parent_pattern == default_parent_pattern) { + fprintf(stdout, "#\n# (%s)\n#\n", help); - if (show_threads) { - bool raw_printing_style = !strcmp(pretty_printing_style, "raw"); - perf_read_values_display(stdout, &show_threads_values, - raw_printing_style); - perf_read_values_destroy(&show_threads_values); + if (show_threads) { + bool style = !strcmp(pretty_printing_style, "raw"); + perf_read_values_display(stdout, &show_threads_values, + style); + perf_read_values_destroy(&show_threads_values); + } } out_delete: perf_session__delete(session); @@ -447,7 +456,7 @@ int cmd_report(int argc, const char **argv, const char *prefix __used) { argc = parse_options(argc, argv, options, report_usage, 0); - setup_pager(); + setup_browser(); if (symbol__init() < 0) return -1; diff --git a/tools/perf/perf.c b/tools/perf/perf.c index 57cb107..9ff186b 100644 --- a/tools/perf/perf.c +++ b/tools/perf/perf.c @@ -265,6 +265,8 @@ static int run_builtin(struct cmd_struct *p, int argc, const char **argv) if (status) return status & 0xff; + exit_browser(); + /* Somebody closed stdout? */ if (fstat(fileno(stdout), &st)) return 0; diff --git a/tools/perf/util/cache.h b/tools/perf/util/cache.h index 918eb37..47b12a3 100644 --- a/tools/perf/util/cache.h +++ b/tools/perf/util/cache.h @@ -1,6 +1,7 @@ #ifndef __PERF_CACHE_H #define __PERF_CACHE_H +#include #include "util.h" #include "strbuf.h" #include "../perf.h" @@ -69,6 +70,19 @@ extern const char *pager_program; extern int pager_in_use(void); extern int pager_use_color; +extern bool use_browser; + +#ifdef NO_NEWT_SUPPORT +static inline void setup_browser(void) +{ + setup_pager(); +} +static inline void exit_browser(void) {} +#else +void setup_browser(void); +void exit_browser(void); +#endif + extern const char *editor_program; extern const char *excludes_file; diff --git a/tools/perf/util/color.c b/tools/perf/util/color.c index e88bca5..9da0191 100644 --- a/tools/perf/util/color.c +++ b/tools/perf/util/color.c @@ -203,7 +203,10 @@ int color_fprintf(FILE *fp, const char *color, const char *fmt, ...) int r; va_start(args, fmt); - r = color_vfprintf(fp, color, fmt, args); + if (use_browser) + r = vfprintf(fp, fmt, args); + else + r = color_vfprintf(fp, color, fmt, args); va_end(args); return r; } diff --git a/tools/perf/util/debug.c b/tools/perf/util/debug.c index 0905600..033d66d 100644 --- a/tools/perf/util/debug.c +++ b/tools/perf/util/debug.c @@ -6,6 +6,7 @@ #include #include +#include "cache.h" #include "color.h" #include "event.h" #include "debug.h" @@ -21,7 +22,10 @@ int eprintf(int level, const char *fmt, ...) if (verbose >= level) { va_start(args, fmt); - ret = vfprintf(stderr, fmt, args); + if (use_browser) + ret = browser__show_help(fmt, args); + else + ret = vfprintf(stderr, fmt, args); va_end(args); } diff --git a/tools/perf/util/debug.h b/tools/perf/util/debug.h index 58720a1..03accb8 100644 --- a/tools/perf/util/debug.h +++ b/tools/perf/util/debug.h @@ -9,5 +9,6 @@ extern int dump_trace; int dump_printf(const char *fmt, ...) __attribute__((format(printf, 1, 2))); void trace_event(event_t *event); +int browser__show_help(const char *format, va_list ap); #endif /* __PERF_DEBUG_H */ diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h index 16f360c..fe366ce 100644 --- a/tools/perf/util/hist.h +++ b/tools/perf/util/hist.h @@ -18,6 +18,11 @@ struct hist_entry *__perf_session__add_hist_entry(struct rb_root *hists, u64 count, bool *hit); extern int64_t hist_entry__cmp(struct hist_entry *, struct hist_entry *); extern int64_t hist_entry__collapse(struct hist_entry *, struct hist_entry *); +size_t hist_entry__fprintf(struct hist_entry *self, + struct perf_session *pair_session, + bool show_displacement, + long displacement, FILE *fp, + u64 session_total); void hist_entry__free(struct hist_entry *); void perf_session__output_resort(struct rb_root *hists, u64 total_samples); diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c new file mode 100644 index 0000000..3d3a936 --- /dev/null +++ b/tools/perf/util/newt.c @@ -0,0 +1,194 @@ +#define _GNU_SOURCE +#include +#undef _GNU_SOURCE + +#include +#include + +#include "cache.h" +#include "hist.h" +#include "session.h" +#include "sort.h" +#include "symbol.h" + +static size_t hist_entry__append_browser(struct hist_entry *self, + newtComponent listbox, u64 total) +{ + char bf[1024]; + size_t len; + FILE *fp; + + if (symbol_conf.exclude_other && !self->parent) + return 0; + + fp = fmemopen(bf, sizeof(bf), "w"); + if (fp == NULL) + return 0; + + len = hist_entry__fprintf(self, NULL, false, 0, fp, total); + + fclose(fp); + newtListboxAppendEntry(listbox, bf, self); + return len; +} + +static void hist_entry__annotate_browser(struct hist_entry *self) +{ + FILE *fp; + struct winsize ws; + newtComponent form, listbox; + struct newtExitStruct es; + char *str; + size_t line_len, max_line_len = 0; + size_t max_usable_width; + char *line = NULL; + + if (self->sym == NULL) + return; + + if (asprintf(&str, "perf annotate %s | expand", self->sym->name) < 0) + return; + + fp = popen(str, "r"); + if (fp == NULL) + goto out_free_str; + + newtPushHelpLine("Press ESC to exit"); + get_term_dimensions(&ws); + listbox = newtListbox(0, 0, ws.ws_row - 5, NEWT_FLAG_SCROLL); + + while (!feof(fp)) { + if (getline(&line, &line_len, fp) < 0 || !line_len) + break; + while (line_len != 0 && isspace(line[line_len - 1])) + line[--line_len] = '\0'; + + if (line_len > max_line_len) + max_line_len = line_len; + newtListboxAppendEntry(listbox, line, NULL); + } + fclose(fp); + free(line); + + max_usable_width = ws.ws_col - 22; + if (max_line_len > max_usable_width) + max_line_len = max_usable_width; + + newtListboxSetWidth(listbox, max_line_len); + + newtCenteredWindow(max_line_len + 2, ws.ws_row - 5, self->sym->name); + form = newtForm(NULL, NULL, 0); + newtFormAddHotKey(form, NEWT_KEY_ESCAPE); + newtFormAddComponents(form, listbox, NULL); + + newtFormRun(form, &es); + newtFormDestroy(form); + newtPopWindow(); + newtPopHelpLine(); +out_free_str: + free(str); +} + +void perf_session__browse_hists(struct rb_root *hists, u64 session_total, + const char *helpline) +{ + struct sort_entry *se; + struct rb_node *nd; + unsigned int width; + char *col_width = symbol_conf.col_width_list_str; + struct winsize ws; + size_t max_len = 0; + char str[1024]; + newtComponent form, listbox; + struct newtExitStruct es; + + snprintf(str, sizeof(str), "Samples: %Ld", session_total); + newtDrawRootText(0, 0, str); + newtPushHelpLine(helpline); + + get_term_dimensions(&ws); + + form = newtForm(NULL, NULL, 0); + newtFormAddHotKey(form, NEWT_KEY_ESCAPE); + + listbox = newtListbox(1, 1, ws.ws_row - 2, (NEWT_FLAG_SCROLL | + NEWT_FLAG_BORDER | + NEWT_FLAG_RETURNEXIT)); + + list_for_each_entry(se, &hist_entry__sort_list, list) { + if (se->elide) + continue; + width = strlen(se->header); + if (se->width) { + if (symbol_conf.col_width_list_str) { + if (col_width) { + *se->width = atoi(col_width); + col_width = strchr(col_width, ','); + if (col_width) + ++col_width; + } + } + *se->width = max(*se->width, width); + } + } + + for (nd = rb_first(hists); nd; nd = rb_next(nd)) { + struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node); + size_t len = hist_entry__append_browser(h, listbox, session_total); + if (len > max_len) + max_len = len; + } + + newtListboxSetWidth(listbox, max_len); + newtFormAddComponents(form, listbox, NULL); + + while (1) { + struct hist_entry *selection; + + newtFormRun(form, &es); + if (es.reason == NEWT_EXIT_HOTKEY) + break; + selection = newtListboxGetCurrent(listbox); + hist_entry__annotate_browser(selection); + } + + newtFormDestroy(form); +} + +int browser__show_help(const char *format, va_list ap) +{ + int ret; + static int backlog; + static char msg[1024]; + + ret = vsnprintf(msg + backlog, sizeof(msg) - backlog, format, ap); + backlog += ret; + + if (msg[backlog - 1] == '\n') { + newtPopHelpLine(); + newtPushHelpLine(msg); + newtRefresh(); + backlog = 0; + } + + return ret; +} + +bool use_browser; + +void setup_browser(void) +{ + if (!isatty(1)) + return; + + use_browser = true; + newtInit(); + newtCls(); + newtPushHelpLine(" "); +} + +void exit_browser(void) +{ + if (use_browser) + newtFinished(); +} diff --git a/tools/perf/util/session.h b/tools/perf/util/session.h index 5c33417..34d7339 100644 --- a/tools/perf/util/session.h +++ b/tools/perf/util/session.h @@ -86,4 +86,13 @@ static inline struct map * { return map_groups__new_module(&self->kmaps, start, filename); } + +#ifdef NO_NEWT_SUPPORT +static inline void perf_session__browse_hists(struct rb_root *hists __used, + u64 session_total __used, + const char *helpline __used) {} +#else +void perf_session__browse_hists(struct rb_root *hists, u64 session_total, + const char *helpline); +#endif #endif /* __PERF_SESSION_H */ -- cgit v0.10.2 From 7081e087b90d4eb4348f7970bd6b266d837321ef Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 12 Mar 2010 10:48:12 -0300 Subject: perf newt: Add 'Q', 'q' and Ctrl+C as ways to exit from forms MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit These are keys people expect when pressed to exit the current widget, so have associate all of them to this semantic. Suggested-by: Ingo Molnar Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1268401692-9361-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index 3d3a936..2a4308a 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -4,6 +4,7 @@ #include #include +#include #include "cache.h" #include "hist.h" @@ -11,6 +12,22 @@ #include "sort.h" #include "symbol.h" +static void newt_form__set_exit_keys(newtComponent self) +{ + newtFormAddHotKey(self, NEWT_KEY_ESCAPE); + newtFormAddHotKey(self, 'Q'); + newtFormAddHotKey(self, 'q'); + newtFormAddHotKey(self, CTRL('c')); +} + +static newtComponent newt_form__new(void) +{ + newtComponent self = newtForm(NULL, NULL, 0); + if (self) + newt_form__set_exit_keys(self); + return self; +} + static size_t hist_entry__append_browser(struct hist_entry *self, newtComponent listbox, u64 total) { @@ -77,8 +94,7 @@ static void hist_entry__annotate_browser(struct hist_entry *self) newtListboxSetWidth(listbox, max_line_len); newtCenteredWindow(max_line_len + 2, ws.ws_row - 5, self->sym->name); - form = newtForm(NULL, NULL, 0); - newtFormAddHotKey(form, NEWT_KEY_ESCAPE); + form = newt_form__new(); newtFormAddComponents(form, listbox, NULL); newtFormRun(form, &es); @@ -108,8 +124,7 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, get_term_dimensions(&ws); - form = newtForm(NULL, NULL, 0); - newtFormAddHotKey(form, NEWT_KEY_ESCAPE); + form = newt_form__new(); listbox = newtListbox(1, 1, ws.ws_row - 2, (NEWT_FLAG_SCROLL | NEWT_FLAG_BORDER | -- cgit v0.10.2 From cb7afb7092bc502b890f0a897ffd67c2b078d347 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 12 Mar 2010 12:46:47 -0300 Subject: perf newt: Use newtGetScreenSize MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit For consistency, use the newt API more fully. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1268408808-13595-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index 2a4308a..f6ec6f5 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -52,7 +52,7 @@ static size_t hist_entry__append_browser(struct hist_entry *self, static void hist_entry__annotate_browser(struct hist_entry *self) { FILE *fp; - struct winsize ws; + int cols, rows; newtComponent form, listbox; struct newtExitStruct es; char *str; @@ -71,8 +71,8 @@ static void hist_entry__annotate_browser(struct hist_entry *self) goto out_free_str; newtPushHelpLine("Press ESC to exit"); - get_term_dimensions(&ws); - listbox = newtListbox(0, 0, ws.ws_row - 5, NEWT_FLAG_SCROLL); + newtGetScreenSize(&cols, &rows); + listbox = newtListbox(0, 0, rows - 5, NEWT_FLAG_SCROLL); while (!feof(fp)) { if (getline(&line, &line_len, fp) < 0 || !line_len) @@ -87,13 +87,13 @@ static void hist_entry__annotate_browser(struct hist_entry *self) fclose(fp); free(line); - max_usable_width = ws.ws_col - 22; + max_usable_width = cols - 22; if (max_line_len > max_usable_width) max_line_len = max_usable_width; newtListboxSetWidth(listbox, max_line_len); - newtCenteredWindow(max_line_len + 2, ws.ws_row - 5, self->sym->name); + newtCenteredWindow(max_line_len + 2, rows - 5, self->sym->name); form = newt_form__new(); newtFormAddComponents(form, listbox, NULL); @@ -112,7 +112,7 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, struct rb_node *nd; unsigned int width; char *col_width = symbol_conf.col_width_list_str; - struct winsize ws; + int rows; size_t max_len = 0; char str[1024]; newtComponent form, listbox; @@ -122,13 +122,13 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, newtDrawRootText(0, 0, str); newtPushHelpLine(helpline); - get_term_dimensions(&ws); + newtGetScreenSize(NULL, &rows); form = newt_form__new(); - listbox = newtListbox(1, 1, ws.ws_row - 2, (NEWT_FLAG_SCROLL | - NEWT_FLAG_BORDER | - NEWT_FLAG_RETURNEXIT)); + listbox = newtListbox(1, 1, rows - 2, (NEWT_FLAG_SCROLL | + NEWT_FLAG_BORDER | + NEWT_FLAG_RETURNEXIT)); list_for_each_entry(se, &hist_entry__sort_list, list) { if (se->elide) -- cgit v0.10.2 From 3997d3776a6e89586e76a0ef355bfbbd8a76966c Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 12 Mar 2010 12:46:48 -0300 Subject: perf hist: Don't fprintf the callgraph unconditionally MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit [root@doppio ~]# perf report -i newt.data | head -10 # Samples: 11999679868 # # Overhead Command Shared Object Symbol # ........ ....... ............................. ...... # 63.61% perf libslang.so.2.1.4 [.] SLsmg_write_chars 6.30% perf perf [.] symbols__find 2.19% perf libnewt.so.0.52.10 [.] newtListboxAppendEntry 2.08% perf libslang.so.2.1.4 [.] SLsmg_write_chars@plt 1.99% perf libc-2.10.2.so [.] _IO_vfprintf_internal [root@doppio ~]# Not good, the newt form for report works, but slang has to eat the cost of the additional callgraph lines everytime it prints a line, and the callgraph doesn't appear on the screen, so move the callgraph printing to a separate function and don't use it in newt.c. Newt tree widgets are being investigated to properly support callgraphs, but till that gets merged, lets remove this huge overhead and show at least the symbol overheads for a callgraph rich perf.data with good performance. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1268408808-13595-2-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index d43be34..1a4e837 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -532,23 +532,23 @@ size_t hist_entry__fprintf(struct hist_entry *self, ret += se->print(fp, self, se->width ? *se->width : 0); } - ret += fprintf(fp, "\n"); - - if (symbol_conf.use_callchain) { - int left_margin = 0; + return ret + fprintf(fp, "\n"); +} - if (sort__first_dimension == SORT_COMM) { - se = list_first_entry(&hist_entry__sort_list, typeof(*se), - list); - left_margin = se->width ? *se->width : 0; - left_margin -= thread__comm_len(self->thread); - } +static size_t hist_entry__fprintf_callchain(struct hist_entry *self, FILE *fp, + u64 session_total) +{ + int left_margin = 0; - ret += hist_entry_callchain__fprintf(fp, self, session_total, - left_margin); + if (sort__first_dimension == SORT_COMM) { + struct sort_entry *se = list_first_entry(&hist_entry__sort_list, + typeof(*se), list); + left_margin = se->width ? *se->width : 0; + left_margin -= thread__comm_len(self->thread); } - return ret; + return hist_entry_callchain__fprintf(fp, self, session_total, + left_margin); } size_t perf_session__fprintf_hists(struct rb_root *hists, @@ -655,6 +655,10 @@ print_entries: } ret += hist_entry__fprintf(h, pair, show_displacement, displacement, fp, session_total); + + if (symbol_conf.use_callchain) + ret += hist_entry__fprintf_callchain(h, fp, session_total); + if (h->map == NULL && verbose > 1) { __map_groups__fprintf_maps(&h->thread->mg, MAP__FUNCTION, fp); -- cgit v0.10.2 From f55db609042faecd5e518ce372b87f846659b32e Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Thu, 11 Mar 2010 14:04:37 -0800 Subject: cpu-timers: Simplify RLIMIT_CPU handling Let always set signal->cputime_expires expiration cache when setting new itimer, POSIX 1.b timer, and RLIMIT_CPU. Since we are initializing prof_exp expiration cache during fork(), this allows to remove "RLIMIT_CPU != inf" check from fastpath_timer_check() and do some other cleanups. Checked against regression using test cases from: http://marc.info/?l=linux-kernel&m=123749066504641&w=4 http://marc.info/?l=linux-kernel&m=123811277916642&w=2 Signed-off-by: Stanislaw Gruszka Cc: Ingo Molnar Cc: Oleg Nesterov Cc: Peter Zijlstra Cc: Hidetoshi Seto Cc: Balbir Singh Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index 1a22dfd..d01e0a3 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -11,19 +11,18 @@ #include /* - * Called after updating RLIMIT_CPU to set timer expiration if necessary. + * Called after updating RLIMIT_CPU to run cpu timer and update + * tsk->signal->cputime_expires expiration cache if necessary. Needs + * siglock protection since other code may update expiration cache as + * well. */ void update_rlimit_cpu(unsigned long rlim_new) { cputime_t cputime = secs_to_cputime(rlim_new); - struct signal_struct *const sig = current->signal; - if (cputime_eq(sig->it[CPUCLOCK_PROF].expires, cputime_zero) || - cputime_gt(sig->it[CPUCLOCK_PROF].expires, cputime)) { - spin_lock_irq(¤t->sighand->siglock); - set_process_cpu_timer(current, CPUCLOCK_PROF, &cputime, NULL); - spin_unlock_irq(¤t->sighand->siglock); - } + spin_lock_irq(¤t->sighand->siglock); + set_process_cpu_timer(current, CPUCLOCK_PROF, &cputime, NULL); + spin_unlock_irq(¤t->sighand->siglock); } static int check_clock(const clockid_t which_clock) @@ -564,7 +563,6 @@ static void arm_timer(struct k_itimer *timer, union cpu_time_count now) struct list_head *head, *listpos; struct cpu_timer_list *const nt = &timer->it.cpu; struct cpu_timer_list *next; - unsigned long i; head = (CPUCLOCK_PERTHREAD(timer->it_clock) ? p->cpu_timers : p->signal->cpu_timers); @@ -630,20 +628,11 @@ static void arm_timer(struct k_itimer *timer, union cpu_time_count now) default: BUG(); case CPUCLOCK_VIRT: - if (expires_le(sig->it[CPUCLOCK_VIRT].expires, - exp->cpu)) - break; - sig->cputime_expires.virt_exp = exp->cpu; - break; + if (expires_gt(sig->cputime_expires.virt_exp, exp->cpu)) + sig->cputime_expires.virt_exp = exp->cpu; case CPUCLOCK_PROF: - if (expires_le(sig->it[CPUCLOCK_PROF].expires, - exp->cpu)) - break; - i = sig->rlim[RLIMIT_CPU].rlim_cur; - if (i != RLIM_INFINITY && - i <= cputime_to_secs(exp->cpu)) - break; - sig->cputime_expires.prof_exp = exp->cpu; + if (expires_gt(sig->cputime_expires.prof_exp, exp->cpu)) + sig->cputime_expires.prof_exp = exp->cpu; break; case CPUCLOCK_SCHED: sig->cputime_expires.sched_exp = exp->sched; @@ -1386,7 +1375,7 @@ static inline int fastpath_timer_check(struct task_struct *tsk) return 1; } - return sig->rlim[RLIMIT_CPU].rlim_cur != RLIM_INFINITY; + return 0; } /* @@ -1452,21 +1441,23 @@ void run_posix_cpu_timers(struct task_struct *tsk) } /* - * Set one of the process-wide special case CPU timers. + * Set one of the process-wide special case CPU timers or RLIMIT_CPU. * The tsk->sighand->siglock must be held by the caller. - * The *newval argument is relative and we update it to be absolute, *oldval - * is absolute and we update it to be relative. */ void set_process_cpu_timer(struct task_struct *tsk, unsigned int clock_idx, cputime_t *newval, cputime_t *oldval) { union cpu_time_count now; - struct list_head *head; BUG_ON(clock_idx == CPUCLOCK_SCHED); cpu_timer_sample_group(clock_idx, tsk, &now); if (oldval) { + /* + * We are setting itimer. The *oldval is absolute and we update + * it to be relative, *newval argument is relative and we update + * it to be absolute. + */ if (!cputime_eq(*oldval, cputime_zero)) { if (cputime_le(*oldval, now.cpu)) { /* Just about to fire. */ @@ -1479,33 +1470,21 @@ void set_process_cpu_timer(struct task_struct *tsk, unsigned int clock_idx, if (cputime_eq(*newval, cputime_zero)) return; *newval = cputime_add(*newval, now.cpu); - - /* - * If the RLIMIT_CPU timer will expire before the - * ITIMER_PROF timer, we have nothing else to do. - */ - if (tsk->signal->rlim[RLIMIT_CPU].rlim_cur - < cputime_to_secs(*newval)) - return; } /* - * Check whether there are any process timers already set to fire - * before this one. If so, we don't have anything more to do. + * Update expiration cache if we are the earliest timer, or eventually + * RLIMIT_CPU limit is earlier than prof_exp cpu timer expire. */ - head = &tsk->signal->cpu_timers[clock_idx]; - if (list_empty(head) || - cputime_ge(list_first_entry(head, - struct cpu_timer_list, entry)->expires.cpu, - *newval)) { - switch (clock_idx) { - case CPUCLOCK_PROF: + switch (clock_idx) { + case CPUCLOCK_PROF: + if (expires_gt(tsk->signal->cputime_expires.prof_exp, *newval)) tsk->signal->cputime_expires.prof_exp = *newval; - break; - case CPUCLOCK_VIRT: + break; + case CPUCLOCK_VIRT: + if (expires_gt(tsk->signal->cputime_expires.virt_exp, *newval)) tsk->signal->cputime_expires.virt_exp = *newval; - break; - } + break; } } -- cgit v0.10.2 From 5eb9aa6414bdab6d075a8763bc3b647181ef3aab Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Thu, 11 Mar 2010 14:04:38 -0800 Subject: cpu-timers: Cleanup arm_timer() Signed-off-by: Stanislaw Gruszka Cc: Ingo Molnar Cc: Oleg Nesterov Cc: Peter Zijlstra Cc: Hidetoshi Seto Cc: Balbir Singh Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index d01e0a3..7c7166f 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -547,97 +547,63 @@ static inline int expires_gt(cputime_t expires, cputime_t new_exp) cputime_gt(expires, new_exp); } -static inline int expires_le(cputime_t expires, cputime_t new_exp) -{ - return !cputime_eq(expires, cputime_zero) && - cputime_le(expires, new_exp); -} /* * Insert the timer on the appropriate list before any timers that * expire later. This must be called with the tasklist_lock held * for reading, and interrupts disabled. */ -static void arm_timer(struct k_itimer *timer, union cpu_time_count now) +static void arm_timer(struct k_itimer *timer) { struct task_struct *p = timer->it.cpu.task; struct list_head *head, *listpos; + struct task_cputime *cputime_expires; struct cpu_timer_list *const nt = &timer->it.cpu; struct cpu_timer_list *next; - head = (CPUCLOCK_PERTHREAD(timer->it_clock) ? - p->cpu_timers : p->signal->cpu_timers); + if (CPUCLOCK_PERTHREAD(timer->it_clock)) { + head = p->cpu_timers; + cputime_expires = &p->cputime_expires; + } else { + head = p->signal->cpu_timers; + cputime_expires = &p->signal->cputime_expires; + } head += CPUCLOCK_WHICH(timer->it_clock); BUG_ON(!irqs_disabled()); spin_lock(&p->sighand->siglock); listpos = head; - if (CPUCLOCK_WHICH(timer->it_clock) == CPUCLOCK_SCHED) { - list_for_each_entry(next, head, entry) { - if (next->expires.sched > nt->expires.sched) - break; - listpos = &next->entry; - } - } else { - list_for_each_entry(next, head, entry) { - if (cputime_gt(next->expires.cpu, nt->expires.cpu)) - break; - listpos = &next->entry; - } + list_for_each_entry(next, head, entry) { + if (cpu_time_before(timer->it_clock, nt->expires, next->expires)) + break; + listpos = &next->entry; } list_add(&nt->entry, listpos); if (listpos == head) { + union cpu_time_count *exp = &nt->expires; + /* - * We are the new earliest-expiring timer. - * If we are a thread timer, there can always - * be a process timer telling us to stop earlier. + * We are the new earliest-expiring POSIX 1.b timer, hence + * need to update expiration cache. Take into account that + * for process timers we share expiration cache with itimers + * and RLIMIT_CPU and for thread timers with RLIMIT_RTTIME. */ - if (CPUCLOCK_PERTHREAD(timer->it_clock)) { - union cpu_time_count *exp = &nt->expires; - - switch (CPUCLOCK_WHICH(timer->it_clock)) { - default: - BUG(); - case CPUCLOCK_PROF: - if (expires_gt(p->cputime_expires.prof_exp, - exp->cpu)) - p->cputime_expires.prof_exp = exp->cpu; - break; - case CPUCLOCK_VIRT: - if (expires_gt(p->cputime_expires.virt_exp, - exp->cpu)) - p->cputime_expires.virt_exp = exp->cpu; - break; - case CPUCLOCK_SCHED: - if (p->cputime_expires.sched_exp == 0 || - p->cputime_expires.sched_exp > exp->sched) - p->cputime_expires.sched_exp = - exp->sched; - break; - } - } else { - struct signal_struct *const sig = p->signal; - union cpu_time_count *exp = &timer->it.cpu.expires; - - /* - * For a process timer, set the cached expiration time. - */ - switch (CPUCLOCK_WHICH(timer->it_clock)) { - default: - BUG(); - case CPUCLOCK_VIRT: - if (expires_gt(sig->cputime_expires.virt_exp, exp->cpu)) - sig->cputime_expires.virt_exp = exp->cpu; - case CPUCLOCK_PROF: - if (expires_gt(sig->cputime_expires.prof_exp, exp->cpu)) - sig->cputime_expires.prof_exp = exp->cpu; - break; - case CPUCLOCK_SCHED: - sig->cputime_expires.sched_exp = exp->sched; - break; - } + switch (CPUCLOCK_WHICH(timer->it_clock)) { + case CPUCLOCK_PROF: + if (expires_gt(cputime_expires->prof_exp, exp->cpu)) + cputime_expires->prof_exp = exp->cpu; + break; + case CPUCLOCK_VIRT: + if (expires_gt(cputime_expires->virt_exp, exp->cpu)) + cputime_expires->virt_exp = exp->cpu; + break; + case CPUCLOCK_SCHED: + if (cputime_expires->sched_exp == 0 || + cputime_expires->sched_exp > exp->sched) + cputime_expires->sched_exp = exp->sched; + break; } } @@ -819,7 +785,7 @@ int posix_cpu_timer_set(struct k_itimer *timer, int flags, if (new_expires.sched != 0 && (timer->it_sigev_notify & ~SIGEV_THREAD_ID) != SIGEV_NONE && cpu_time_before(timer->it_clock, val, new_expires)) { - arm_timer(timer, val); + arm_timer(timer); } read_unlock(&tasklist_lock); @@ -1283,7 +1249,7 @@ void posix_cpu_timer_schedule(struct k_itimer *timer) /* * Now re-arm for the new expiry time. */ - arm_timer(timer, now); + arm_timer(timer); out_unlock: read_unlock(&tasklist_lock); -- cgit v0.10.2 From ae1a78eecc45fe41215d9dbfd7079999455772d6 Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Thu, 11 Mar 2010 14:04:39 -0800 Subject: cpu-timers: Return correct previous timer reload value According POSIX we need to correctly set old timer it_interval value when user request that in timer_settime(). Tested using below program. #include #include #include #include #include #include #include int main(void) { struct sigaction act; struct sigevent evt = { }; timer_t tid; struct itimerspec spec, u_spec, k_spec; evt.sigev_notify = SIGEV_SIGNAL; evt.sigev_signo = SIGPROF; assert(timer_create(CLOCK_PROCESS_CPUTIME_ID, &evt, &tid) == 0); spec.it_value.tv_sec = 1; spec.it_value.tv_nsec = 2; spec.it_interval.tv_sec = 3; spec.it_interval.tv_nsec = 4; u_spec = spec; assert(timer_settime(tid, 0, &spec, NULL) == 0); spec.it_value.tv_sec = 5; spec.it_value.tv_nsec = 6; spec.it_interval.tv_sec = 7; spec.it_interval.tv_nsec = 8; assert(timer_settime(tid, 0, &spec, &k_spec) == 0); #define PRT(val) printf(#val ":\t%d/%d\n", (int) u_spec.val, (int) k_spec.val) PRT(it_value.tv_sec); PRT(it_value.tv_nsec); PRT(it_interval.tv_sec); PRT(it_interval.tv_nsec); return 0; } Signed-off-by: Stanislaw Gruszka Cc: Ingo Molnar Cc: Oleg Nesterov Cc: Peter Zijlstra Cc: Hidetoshi Seto Cc: Balbir Singh Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index 7c7166f..cce2f0b 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -676,7 +676,7 @@ int posix_cpu_timer_set(struct k_itimer *timer, int flags, struct itimerspec *new, struct itimerspec *old) { struct task_struct *p = timer->it.cpu.task; - union cpu_time_count old_expires, new_expires, val; + union cpu_time_count old_expires, new_expires, old_incr, val; int ret; if (unlikely(p == NULL)) { @@ -707,6 +707,7 @@ int posix_cpu_timer_set(struct k_itimer *timer, int flags, BUG_ON(!irqs_disabled()); ret = 0; + old_incr = timer->it.cpu.incr; spin_lock(&p->sighand->siglock); old_expires = timer->it.cpu.expires; if (unlikely(timer->it.cpu.firing)) { @@ -822,7 +823,7 @@ int posix_cpu_timer_set(struct k_itimer *timer, int flags, out: if (old) { sample_to_timespec(timer->it_clock, - timer->it.cpu.incr, &old->it_interval); + old_incr, &old->it_interval); } return ret; } -- cgit v0.10.2 From 1f169f84d25a74fb2dc67274d31d082ce30c60fb Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Thu, 11 Mar 2010 14:04:41 -0800 Subject: cpu-timers: Change SIGEV_NONE timer implementation When user sets up a timer without associated signal and process does not use any other cpu timers and does not exit, tsk->signal->cputimer is enabled and running forever. Avoid running the timer for no reason. I used below program to check patch does not break current user space visible behavior. #include #include #include #include #include #include #include #include void consume_cpu(void) { int i = 0; int count = 0; for(i=0; i<100000000; i++) count++; } int main(void) { int i; struct sigaction act; struct sigevent evt = { }; timer_t tid; struct itimerspec spec = { }; evt.sigev_notify = SIGEV_NONE; assert(timer_create(CLOCK_PROCESS_CPUTIME_ID, &evt, &tid) == 0); spec.it_value.tv_sec = 10; assert(timer_settime(tid, 0, &spec, NULL) == 0); for (i = 0; i < 30; i++) { consume_cpu(); memset(&spec, 0, sizeof(spec)); assert(timer_gettime(tid, &spec) == 0); printf("%lu.%09lu\n", (unsigned long) spec.it_value.tv_sec, (unsigned long) spec.it_value.tv_nsec); } assert(timer_delete(tid) == 0); return 0; } Signed-off-by: Stanislaw Gruszka Cc: Ingo Molnar Cc: Oleg Nesterov Cc: Peter Zijlstra Cc: Hidetoshi Seto Cc: Balbir Singh Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index cce2f0b..7d9d0fa 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -615,7 +615,12 @@ static void arm_timer(struct k_itimer *timer) */ static void cpu_timer_fire(struct k_itimer *timer) { - if (unlikely(timer->sigq == NULL)) { + if ((timer->it_sigev_notify & ~SIGEV_THREAD_ID) == SIGEV_NONE) { + /* + * User don't want any signal. + */ + timer->it.cpu.expires.sched = 0; + } else if (unlikely(timer->sigq == NULL)) { /* * This a special case for clock_nanosleep, * not a normal timer from sys_timer_create. @@ -784,7 +789,6 @@ int posix_cpu_timer_set(struct k_itimer *timer, int flags, */ timer->it.cpu.expires = new_expires; if (new_expires.sched != 0 && - (timer->it_sigev_notify & ~SIGEV_THREAD_ID) != SIGEV_NONE && cpu_time_before(timer->it_clock, val, new_expires)) { arm_timer(timer); } @@ -809,7 +813,6 @@ int posix_cpu_timer_set(struct k_itimer *timer, int flags, timer->it_overrun = -1; if (new_expires.sched != 0 && - (timer->it_sigev_notify & ~SIGEV_THREAD_ID) != SIGEV_NONE && !cpu_time_before(timer->it_clock, val, new_expires)) { /* * The designated time already passed, so we notify @@ -883,25 +886,6 @@ void posix_cpu_timer_get(struct k_itimer *timer, struct itimerspec *itp) read_unlock(&tasklist_lock); } - if ((timer->it_sigev_notify & ~SIGEV_THREAD_ID) == SIGEV_NONE) { - if (timer->it.cpu.incr.sched == 0 && - cpu_time_before(timer->it_clock, - timer->it.cpu.expires, now)) { - /* - * Do-nothing timer expired and has no reload, - * so it's as if it was never set. - */ - timer->it.cpu.expires.sched = 0; - itp->it_value.tv_sec = itp->it_value.tv_nsec = 0; - return; - } - /* - * Account for any expirations and reloads that should - * have happened. - */ - bump_cpu_timer(timer, now); - } - if (unlikely(clear_dead)) { /* * We've noticed that the thread is dead, but -- cgit v0.10.2 From c28739375bf0d6e239b4fa939ec8372aa2c707d2 Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Thu, 11 Mar 2010 14:04:42 -0800 Subject: cpu-timers: Avoid iterating over all threads in fastpath_timer_check() Spread p->sighand->siglock locking scope to make sure that fastpath_timer_check() never iterates over all threads. Without locking there is small possibility that signal->cputimer will stop running while we write values to signal->cputime_expires. Calling thread_group_cputime() from fastpath_timer_check() is not only bad because it is slow, also it is racy with __exit_signal() which can lead to invalid signal->{s,u}time values. Signed-off-by: Stanislaw Gruszka Cc: Ingo Molnar Cc: Oleg Nesterov Cc: Peter Zijlstra Cc: Hidetoshi Seto Cc: Balbir Singh Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c index 7d9d0fa..564b3b0 100644 --- a/kernel/posix-cpu-timers.c +++ b/kernel/posix-cpu-timers.c @@ -550,7 +550,7 @@ static inline int expires_gt(cputime_t expires, cputime_t new_exp) /* * Insert the timer on the appropriate list before any timers that * expire later. This must be called with the tasklist_lock held - * for reading, and interrupts disabled. + * for reading, interrupts disabled and p->sighand->siglock taken. */ static void arm_timer(struct k_itimer *timer) { @@ -569,9 +569,6 @@ static void arm_timer(struct k_itimer *timer) } head += CPUCLOCK_WHICH(timer->it_clock); - BUG_ON(!irqs_disabled()); - spin_lock(&p->sighand->siglock); - listpos = head; list_for_each_entry(next, head, entry) { if (cpu_time_before(timer->it_clock, nt->expires, next->expires)) @@ -606,8 +603,6 @@ static void arm_timer(struct k_itimer *timer) break; } } - - spin_unlock(&p->sighand->siglock); } /* @@ -720,7 +715,6 @@ int posix_cpu_timer_set(struct k_itimer *timer, int flags, ret = TIMER_RETRY; } else list_del_init(&timer->it.cpu.entry); - spin_unlock(&p->sighand->siglock); /* * We need to sample the current value to convert the new @@ -774,6 +768,7 @@ int posix_cpu_timer_set(struct k_itimer *timer, int flags, * disable this firing since we are already reporting * it as an overrun (thanks to bump_cpu_timer above). */ + spin_unlock(&p->sighand->siglock); read_unlock(&tasklist_lock); goto out; } @@ -793,6 +788,7 @@ int posix_cpu_timer_set(struct k_itimer *timer, int flags, arm_timer(timer); } + spin_unlock(&p->sighand->siglock); read_unlock(&tasklist_lock); /* @@ -1206,6 +1202,7 @@ void posix_cpu_timer_schedule(struct k_itimer *timer) goto out; } read_lock(&tasklist_lock); /* arm_timer needs it. */ + spin_lock(&p->sighand->siglock); } else { read_lock(&tasklist_lock); if (unlikely(p->signal == NULL)) { @@ -1226,6 +1223,7 @@ void posix_cpu_timer_schedule(struct k_itimer *timer) clear_dead_task(timer, now); goto out_unlock; } + spin_lock(&p->sighand->siglock); cpu_timer_sample_group(timer->it_clock, p, &now); bump_cpu_timer(timer, now); /* Leave the tasklist_lock locked for the call below. */ @@ -1234,7 +1232,9 @@ void posix_cpu_timer_schedule(struct k_itimer *timer) /* * Now re-arm for the new expiry time. */ + BUG_ON(!irqs_disabled()); arm_timer(timer); + spin_unlock(&p->sighand->siglock); out_unlock: read_unlock(&tasklist_lock); -- cgit v0.10.2 From 64ce4c2f5252f25798117fa80a027993163d6d84 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Thu, 11 Mar 2010 14:04:47 -0800 Subject: time: Clean up warp_clock() warp_clock() currently accesses timekeeping internal state directly, which is unnecessary. Convert it to use the proper timekeeping interfaces. Signed-off-by: John Stultz Cc: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner diff --git a/kernel/time.c b/kernel/time.c index 8047980..2358a36 100644 --- a/kernel/time.c +++ b/kernel/time.c @@ -133,12 +133,11 @@ SYSCALL_DEFINE2(gettimeofday, struct timeval __user *, tv, */ static inline void warp_clock(void) { - write_seqlock_irq(&xtime_lock); - wall_to_monotonic.tv_sec -= sys_tz.tz_minuteswest * 60; - xtime.tv_sec += sys_tz.tz_minuteswest * 60; - update_xtime_cache(0); - write_sequnlock_irq(&xtime_lock); - clock_was_set(); + struct timespec delta, adjust; + delta.tv_sec = sys_tz.tz_minuteswest * 60; + delta.tv_nsec = 0; + adjust = timespec_add_safe(current_kernel_time(), delta); + do_settimeofday(&adjust); } /* -- cgit v0.10.2 From 06f71b922ce5a05352acd706564ca4ae1f2add0e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Thu, 11 Mar 2010 14:04:46 -0800 Subject: timer: Print function name for timer callbacks modifying preemption count MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A function scheduled with a timer must not exit with a different preempt count than it was entered. To make helping users running into the corresponding BUG() easier also print the name of the bad function not only its address. [ tglx: Sanitized printk ] Signed-off-by: Uwe Kleine-König Cc: johnstul@us.ibm.com Cc: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner diff --git a/kernel/timer.c b/kernel/timer.c index c61a794..f82f4bf 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -1027,11 +1027,8 @@ static inline void __run_timers(struct tvec_base *base) lock_map_release(&lockdep_map); if (preempt_count != preempt_count()) { - printk(KERN_ERR "huh, entered %p " - "with preempt_count %08x, exited" - " with %08x?\n", - fn, preempt_count, - preempt_count()); + printk(KERN_ERR "timer: %pF preempt leak: %08x -> %08x\n", + fn, preempt_count, preempt_count()); BUG(); } } -- cgit v0.10.2 From 576da126a6c7364d70dfd58d0bbe43d05cf5859f Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 12 Mar 2010 21:10:29 +0100 Subject: timer: Split out timer function call The ident level is starting to be annoying. More white space than actual code. Split out the timer function call into its own function. Signed-off-by: Thomas Gleixner diff --git a/kernel/timer.c b/kernel/timer.c index f82f4bf..4522969 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -953,6 +953,41 @@ static int cascade(struct tvec_base *base, struct tvec *tv, int index) return index; } +static void call_timer_fn(struct timer_list *timer, void (*fn)(unsigned long), + unsigned long data) +{ + int preempt_count = preempt_count(); + +#ifdef CONFIG_LOCKDEP + /* + * It is permissible to free the timer from inside the + * function that is called from it, this we need to take into + * account for lockdep too. To avoid bogus "held lock freed" + * warnings as well as problems when looking into + * timer->lockdep_map, make a copy and use that here. + */ + struct lockdep_map lockdep_map = timer->lockdep_map; +#endif + /* + * Couple the lock chain with the lock chain at + * del_timer_sync() by acquiring the lock_map around the fn() + * call here and in del_timer_sync(). + */ + lock_map_acquire(&lockdep_map); + + trace_timer_expire_entry(timer); + fn(data); + trace_timer_expire_exit(timer); + + lock_map_release(&lockdep_map); + + if (preempt_count != preempt_count()) { + printk(KERN_ERR "timer: %pF preempt leak: %08x -> %08x\n", + fn, preempt_count, preempt_count()); + BUG(); + } +} + #define INDEX(N) ((base->timer_jiffies >> (TVR_BITS + (N) * TVN_BITS)) & TVN_MASK) /** @@ -996,42 +1031,7 @@ static inline void __run_timers(struct tvec_base *base) detach_timer(timer, 1); spin_unlock_irq(&base->lock); - { - int preempt_count = preempt_count(); - -#ifdef CONFIG_LOCKDEP - /* - * It is permissible to free the timer from - * inside the function that is called from - * it, this we need to take into account for - * lockdep too. To avoid bogus "held lock - * freed" warnings as well as problems when - * looking into timer->lockdep_map, make a - * copy and use that here. - */ - struct lockdep_map lockdep_map = - timer->lockdep_map; -#endif - /* - * Couple the lock chain with the lock chain at - * del_timer_sync() by acquiring the lock_map - * around the fn() call here and in - * del_timer_sync(). - */ - lock_map_acquire(&lockdep_map); - - trace_timer_expire_entry(timer); - fn(data); - trace_timer_expire_exit(timer); - - lock_map_release(&lockdep_map); - - if (preempt_count != preempt_count()) { - printk(KERN_ERR "timer: %pF preempt leak: %08x -> %08x\n", - fn, preempt_count, preempt_count()); - BUG(); - } - } + call_timer_fn(timer, fn, data); spin_lock_irq(&base->lock); } } -- cgit v0.10.2 From 802702e0c2618465b813242d4dfee6a233ba0beb Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Fri, 12 Mar 2010 20:13:23 +0100 Subject: timer: Try to survive timer callback preempt_count leak If a timer callback leaks preempt_count we currently assert a BUG(). That makes it unnecessarily hard to retrieve information about the problem especially on laptops and headless stations. There is a decent chance to survive the preempt_count leak by restoring the preempt_count to the value before the callback. That allows in many cases to get valuable information about the root cause of the problem. We carried that fixup in preempt-rt for years and were able to decode such wreckage quite a few times. Signed-off-by: Thomas Gleixner Cc: Linux Torvalds Cc: Andrew Morton Cc: Arjan van de Veen diff --git a/kernel/timer.c b/kernel/timer.c index 4522969..7e12e7b 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -982,9 +982,15 @@ static void call_timer_fn(struct timer_list *timer, void (*fn)(unsigned long), lock_map_release(&lockdep_map); if (preempt_count != preempt_count()) { - printk(KERN_ERR "timer: %pF preempt leak: %08x -> %08x\n", - fn, preempt_count, preempt_count()); - BUG(); + WARN_ONCE(1, "timer: %pF preempt leak: %08x -> %08x\n", + fn, preempt_count, preempt_count()); + /* + * Restore the preempt count. That gives us a decent + * chance to survive and extract information. If the + * callback kept a lock held, bad luck, but not worse + * than the BUG() we had. + */ + preempt_count() = preempt_count; } } -- cgit v0.10.2 From fdb19a6cb48407c59a2007063f4005c9384603c3 Mon Sep 17 00:00:00 2001 From: Jens Rottmann Date: Thu, 11 Mar 2010 14:04:44 -0800 Subject: cs5535-clockevt: Free timer in IRQ setup error path Due to a hardware limitation cs5535_mfgpt_free_timer() cannot actually release the timer hardware, but it will at least free the now unreferenced struct associated with it so calling it is the cleaner thing to do. Signed-off-by: Jens Rottmann Acked-by: Andres Salomon Cc: Jordan Crouse Cc: Ingo Molnar Cc: john stultz Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner diff --git a/drivers/clocksource/cs5535-clockevt.c b/drivers/clocksource/cs5535-clockevt.c index b314a99..d7be69f1 100644 --- a/drivers/clocksource/cs5535-clockevt.c +++ b/drivers/clocksource/cs5535-clockevt.c @@ -154,14 +154,14 @@ static int __init cs5535_mfgpt_init(void) if (cs5535_mfgpt_setup_irq(timer, MFGPT_CMP2, &timer_irq)) { printk(KERN_ERR DRV_NAME ": Could not set up IRQ %d\n", timer_irq); - return -EIO; + goto err_timer; } /* And register it with the kernel */ ret = setup_irq(timer_irq, &mfgptirq); if (ret) { printk(KERN_ERR DRV_NAME ": Unable to set up the interrupt.\n"); - goto err; + goto err_irq; } /* Set the clock scale and enable the event mode for CMP2 */ @@ -184,8 +184,10 @@ static int __init cs5535_mfgpt_init(void) return 0; -err: +err_irq: cs5535_mfgpt_release_irq(cs5535_event_clock, MFGPT_CMP2, &timer_irq); +err_timer: + cs5535_mfgpt_free_timer(cs5535_event_clock); printk(KERN_ERR DRV_NAME ": Unable to set up the MFGPT clock source\n"); return -EIO; } -- cgit v0.10.2 From 25268498c9e07870323aead10751b7c6e99a3a78 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Fri, 12 Mar 2010 14:56:00 -0800 Subject: time: Add xtime, wall_to_monotonic to feature-removal-schedule They aren't really features, but fair warning to out of tree driver folks who might be accessing these variables. Signed-off-by: John Stultz CC: Ingo Molnar LKML-Reference: <1268434560-2677-1-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index a5cc0db..198acdc 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -582,3 +582,13 @@ Why: The paravirt mmu host support is slower than non-paravirt mmu, both Who: Avi Kivity ---------------------------- + +What: xtime, wall_to_monotonic +When: 2.6.36+ +Files: kernel/time/timekeeping.c include/linux/time.h +Why: Cleaning up timekeeping internal values. Please use + existing timekeeping accessor functions to access + the equivalent functionality. +Who: John Stultz + +---------------------------- -- cgit v0.10.2 From c4507257764ae0387f0d8c2ca57808f37e4c9439 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Thu, 11 Mar 2010 14:04:47 -0800 Subject: time: Clean up direct xtime usage in xen Cleanup xen's direct use of internal timekeeping values. Signed-off-by: John Stultz Acked-by: Jeremy Fitzhardinge Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c index 0d3f07c..6365df9 100644 --- a/arch/x86/xen/time.c +++ b/arch/x86/xen/time.c @@ -475,6 +475,7 @@ void xen_timer_resume(void) __init void xen_time_init(void) { int cpu = smp_processor_id(); + struct timespec tp; clocksource_register(&xen_clocksource); @@ -486,9 +487,8 @@ __init void xen_time_init(void) } /* Set initial system time with full resolution */ - xen_read_wallclock(&xtime); - set_normalized_timespec(&wall_to_monotonic, - -xtime.tv_sec, -xtime.tv_nsec); + xen_read_wallclock(&tp); + do_settimeofday(&tp); setup_force_cpu_cap(X86_FEATURE_TSC); -- cgit v0.10.2 From 104f9c471ffb425fd081074ef31521b38d77592a Mon Sep 17 00:00:00 2001 From: John Stultz Date: Thu, 11 Mar 2010 14:11:39 -0800 Subject: xtensa: Fix unnecessary setting of xtime xtensa supports read_persisitent_clock(), so there is no reason for the arch specific code to be setting xtime. Setting xtime will be done by the generic timekeeping code. Signed-off-by: John Stultz Cc: Chris Zankel Cc: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner diff --git a/arch/xtensa/kernel/time.c b/arch/xtensa/kernel/time.c index 19f7df3..19df764 100644 --- a/arch/xtensa/kernel/time.c +++ b/arch/xtensa/kernel/time.c @@ -60,11 +60,6 @@ static struct irqaction timer_irqaction = { void __init time_init(void) { - /* FIXME: xtime&wall_to_monotonic are set in timekeeping_init. */ - read_persistent_clock(&xtime); - set_normalized_timespec(&wall_to_monotonic, - -xtime.tv_sec, -xtime.tv_nsec); - #ifdef CONFIG_XTENSA_CALIBRATE_CCOUNT printk("Calibrating CPU frequency "); platform_calibrate_ccount(); -- cgit v0.10.2 From 1e871be1aa97babb467a929d6adcb1960659928b Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 3 Mar 2010 19:57:16 -0800 Subject: alpha: Convert alpha to use read/update_persistent_clock This patch converts the alpha architecture to use the generic read_persistent_clock and update_persistent_clock interfaces, reducing the amount of arch specific code we have to maintain, and allowing for further cleanups in the future. I have not built or tested this patch, so help from arch maintainers would be appreciated. igned-off-by: John Stultz Cc: Richard Henderson Cc: Andrew Morton LKML-Reference: <1267675049-12337-2-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/arch/alpha/Kconfig b/arch/alpha/Kconfig index bd7261e..75fff7e 100644 --- a/arch/alpha/Kconfig +++ b/arch/alpha/Kconfig @@ -54,6 +54,9 @@ config ARCH_USES_GETTIMEOFFSET bool default y +config GENERIC_CMOS_UPDATE + def_bool y + config ZONE_DMA bool default y diff --git a/arch/alpha/kernel/time.c b/arch/alpha/kernel/time.c index 5d08266..5465e93 100644 --- a/arch/alpha/kernel/time.c +++ b/arch/alpha/kernel/time.c @@ -75,8 +75,6 @@ static struct { __u32 last_time; /* ticks/cycle * 2^48 */ unsigned long scaled_ticks_per_cycle; - /* last time the CMOS clock got updated */ - time_t last_rtc_update; /* partial unused tick */ unsigned long partial_tick; } state; @@ -91,6 +89,52 @@ static inline __u32 rpcc(void) return result; } +int update_persistent_clock(struct timespec now) +{ + return set_rtc_mmss(now.tv_sec); +} + +void read_persistent_clock(struct timespec *ts) +{ + unsigned int year, mon, day, hour, min, sec, epoch; + + sec = CMOS_READ(RTC_SECONDS); + min = CMOS_READ(RTC_MINUTES); + hour = CMOS_READ(RTC_HOURS); + day = CMOS_READ(RTC_DAY_OF_MONTH); + mon = CMOS_READ(RTC_MONTH); + year = CMOS_READ(RTC_YEAR); + + if (!(CMOS_READ(RTC_CONTROL) & RTC_DM_BINARY) || RTC_ALWAYS_BCD) { + sec = bcd2bin(sec); + min = bcd2bin(min); + hour = bcd2bin(hour); + day = bcd2bin(day); + mon = bcd2bin(mon); + year = bcd2bin(year); + } + + /* PC-like is standard; used for year >= 70 */ + epoch = 1900; + if (year < 20) + epoch = 2000; + else if (year >= 20 && year < 48) + /* NT epoch */ + epoch = 1980; + else if (year >= 48 && year < 70) + /* Digital UNIX epoch */ + epoch = 1952; + + printk(KERN_INFO "Using epoch = %d\n", epoch); + + if ((year += epoch) < 1970) + year += 100; + + ts->tv_sec = mktime(year, mon, day, hour, min, sec); +} + + + /* * timer_interrupt() needs to keep up the real-time clock, * as well as call the "do_timer()" routine every clocktick @@ -123,19 +167,6 @@ irqreturn_t timer_interrupt(int irq, void *dev) if (nticks) do_timer(nticks); - /* - * If we have an externally synchronized Linux clock, then update - * CMOS clock accordingly every ~11 minutes. Set_rtc_mmss() has to be - * called as close as possible to 500 ms before the new second starts. - */ - if (ntp_synced() - && xtime.tv_sec > state.last_rtc_update + 660 - && xtime.tv_nsec >= 500000 - ((unsigned) TICK_SIZE) / 2 - && xtime.tv_nsec <= 500000 + ((unsigned) TICK_SIZE) / 2) { - int tmp = set_rtc_mmss(xtime.tv_sec); - state.last_rtc_update = xtime.tv_sec - (tmp ? 600 : 0); - } - write_sequnlock(&xtime_lock); #ifndef CONFIG_SMP @@ -304,7 +335,7 @@ rpcc_after_update_in_progress(void) void __init time_init(void) { - unsigned int year, mon, day, hour, min, sec, cc1, cc2, epoch; + unsigned int cc1, cc2; unsigned long cycle_freq, tolerance; long diff; @@ -348,43 +379,6 @@ time_init(void) bogomips yet, but this is close on a 500Mhz box. */ __delay(1000000); - sec = CMOS_READ(RTC_SECONDS); - min = CMOS_READ(RTC_MINUTES); - hour = CMOS_READ(RTC_HOURS); - day = CMOS_READ(RTC_DAY_OF_MONTH); - mon = CMOS_READ(RTC_MONTH); - year = CMOS_READ(RTC_YEAR); - - if (!(CMOS_READ(RTC_CONTROL) & RTC_DM_BINARY) || RTC_ALWAYS_BCD) { - sec = bcd2bin(sec); - min = bcd2bin(min); - hour = bcd2bin(hour); - day = bcd2bin(day); - mon = bcd2bin(mon); - year = bcd2bin(year); - } - - /* PC-like is standard; used for year >= 70 */ - epoch = 1900; - if (year < 20) - epoch = 2000; - else if (year >= 20 && year < 48) - /* NT epoch */ - epoch = 1980; - else if (year >= 48 && year < 70) - /* Digital UNIX epoch */ - epoch = 1952; - - printk(KERN_INFO "Using epoch = %d\n", epoch); - - if ((year += epoch) < 1970) - year += 100; - - xtime.tv_sec = mktime(year, mon, day, hour, min, sec); - xtime.tv_nsec = 0; - - wall_to_monotonic.tv_sec -= xtime.tv_sec; - wall_to_monotonic.tv_nsec = 0; if (HZ > (1<<16)) { extern void __you_loose (void); @@ -394,7 +388,6 @@ time_init(void) state.last_time = cc1; state.scaled_ticks_per_cycle = ((unsigned long) HZ << FIX_SHIFT) / cycle_freq; - state.last_rtc_update = 0; state.partial_tick = 0L; /* Startup the timer source. */ -- cgit v0.10.2 From f7a56575fa4e62e94a644f72ef8c5d423c073982 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 3 Mar 2010 19:57:17 -0800 Subject: mn10300: Convert mn10300 to use read/update_persistent_clock This patch converts the mn10300 architecture to use the generic read_persistent_clock and update_persistent_clock interfaces, reducing the amount of arch specific code we have to maintain, and allowing for further cleanups in the future. I have not built or tested this patch, so help from arch maintainers would be appreciated. Signed-off-by: John Stultz Cc: David Howells Cc: Koichi Yasutake Cc: Andrew Morton LKML-Reference: <1267675049-12337-3-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/arch/mn10300/Kconfig b/arch/mn10300/Kconfig index 89faaca..1c4565a 100644 --- a/arch/mn10300/Kconfig +++ b/arch/mn10300/Kconfig @@ -37,6 +37,9 @@ config GENERIC_HARDIRQS_NO__DO_IRQ config GENERIC_CALIBRATE_DELAY def_bool y +config GENERIC_CMOS_UPDATE + def_bool y + config GENERIC_FIND_NEXT_BIT def_bool y diff --git a/arch/mn10300/kernel/rtc.c b/arch/mn10300/kernel/rtc.c index 7978470..815a933 100644 --- a/arch/mn10300/kernel/rtc.c +++ b/arch/mn10300/kernel/rtc.c @@ -26,17 +26,15 @@ static long last_rtc_update; /* time for RTC to update itself in ioclks */ static unsigned long mn10300_rtc_update_period; -/* - * read the current RTC time - */ -unsigned long __init get_initial_rtc_time(void) +void read_persistent_clock(struct timespec *ts) { struct rtc_time tm; get_rtc_time(&tm); - return mktime(tm.tm_year, tm.tm_mon, tm.tm_mday, + ts->tv_sec = mktime(tm.tm_year, tm.tm_mon, tm.tm_mday, tm.tm_hour, tm.tm_min, tm.tm_sec); + ts->tv_nsec = 0; } /* @@ -110,24 +108,9 @@ static int set_rtc_mmss(unsigned long nowtime) return retval; } -void check_rtc_time(void) +int update_persistent_clock(struct timespec now) { - /* the RTC clock just finished ticking over again this second - * - if we have an externally synchronized Linux clock, then update - * RTC clock accordingly every ~11 minutes. set_rtc_mmss() has to be - * called as close as possible to 500 ms before the new second starts. - */ - if ((time_status & STA_UNSYNC) == 0 && - xtime.tv_sec > last_rtc_update + 660 && - xtime.tv_nsec / 1000 >= 500000 - ((unsigned) TICK_SIZE) / 2 && - xtime.tv_nsec / 1000 <= 500000 + ((unsigned) TICK_SIZE) / 2 - ) { - if (set_rtc_mmss(xtime.tv_sec) == 0) - last_rtc_update = xtime.tv_sec; - else - /* do it again in 60s */ - last_rtc_update = xtime.tv_sec - 600; - } + return set_rtc_mms(now.tv_sec); } /* diff --git a/arch/mn10300/kernel/time.c b/arch/mn10300/kernel/time.c index 395caf0..8f7f6d2 100644 --- a/arch/mn10300/kernel/time.c +++ b/arch/mn10300/kernel/time.c @@ -111,7 +111,6 @@ static irqreturn_t timer_interrupt(int irq, void *dev_id) /* advance the kernel's time tracking system */ profile_tick(CPU_PROFILING); do_timer(1); - check_rtc_time(); } write_sequnlock(&xtime_lock); @@ -139,9 +138,6 @@ void __init time_init(void) " (calibrated against RTC)\n", MN10300_TSCCLK / 1000000, (MN10300_TSCCLK / 10000) % 100); - xtime.tv_sec = get_initial_rtc_time(); - xtime.tv_nsec = 0; - mn10300_last_tsc = TMTSCBC; /* use timer 0 & 1 cascaded to tick at as close to HZ as possible */ -- cgit v0.10.2 From 739b27116a0e6903d6c85d6769e69f43d8fa4128 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 3 Mar 2010 19:57:18 -0800 Subject: frv: Convert frv to use read/update_persistent_clock This patch converts the frv architecture to use the generic read_persistent_clock and update_persistent_clock interfaces, reducing the amount of arch specific code we have to maintain, and allowing for further cleanups in the future. I have not built or tested this patch, so help from arch maintainers would be appreciated. Signed-off-by: John Stultz Cc: David Howells Cc: Andrew Morton LKML-Reference: <1267675049-12337-4-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/arch/frv/kernel/time.c b/arch/frv/kernel/time.c index fb0ce75..0ddbbae 100644 --- a/arch/frv/kernel/time.c +++ b/arch/frv/kernel/time.c @@ -48,20 +48,12 @@ static struct irqaction timer_irq = { .name = "timer", }; -static inline int set_rtc_mmss(unsigned long nowtime) -{ - return -1; -} - /* * timer_interrupt() needs to keep up the real-time clock, * as well as call the "do_timer()" routine every clocktick */ static irqreturn_t timer_interrupt(int irq, void *dummy) { - /* last time the cmos clock got updated */ - static long last_rtc_update = 0; - profile_tick(CPU_PROFILING); /* * Here we are in the timer irq handler. We just have irqs locally @@ -74,22 +66,6 @@ static irqreturn_t timer_interrupt(int irq, void *dummy) do_timer(1); - /* - * If we have an externally synchronized Linux clock, then update - * CMOS clock accordingly every ~11 minutes. Set_rtc_mmss() has to be - * called as close as possible to 500 ms before the new second starts. - */ - if (ntp_synced() && - xtime.tv_sec > last_rtc_update + 660 && - (xtime.tv_nsec / 1000) >= 500000 - ((unsigned) TICK_SIZE) / 2 && - (xtime.tv_nsec / 1000) <= 500000 + ((unsigned) TICK_SIZE) / 2 - ) { - if (set_rtc_mmss(xtime.tv_sec) == 0) - last_rtc_update = xtime.tv_sec; - else - last_rtc_update = xtime.tv_sec - 600; /* do it again in 60 s */ - } - #ifdef CONFIG_HEARTBEAT static unsigned short n; n++; @@ -119,7 +95,8 @@ void time_divisor_init(void) __set_TCSR_DATA(0, base >> 8); } -void time_init(void) + +void read_persistent_clock(struct timespec *ts) { unsigned int year, mon, day, hour, min, sec; @@ -135,9 +112,12 @@ void time_init(void) if ((year += 1900) < 1970) year += 100; - xtime.tv_sec = mktime(year, mon, day, hour, min, sec); - xtime.tv_nsec = 0; + ts->tv_sec = mktime(year, mon, day, hour, min, sec); + ts->tv_nsec = 0; +} +void time_init(void) +{ /* install scheduling interrupt handler */ setup_irq(IRQ_CPU_TIMER0, &timer_irq); -- cgit v0.10.2 From 12833c1e373ad2b12667ca159d335fae29968473 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 3 Mar 2010 19:57:20 -0800 Subject: h8300: Convert h8300 to use read/update_persistent_clock This patch converts the h8300 architecture to use the generic read_persistent_clock and update_persistent_clock interfaces, reducing the amount of arch specific code we have to maintain, and allowing for further cleanups in the future. I have not built or tested this patch, so help from arch maintainers would be appreciated. Signed-off-by: John Stultz Cc: Yoshinori Sato Cc: Andrew Morton LKML-Reference: <1267675049-12337-6-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/arch/h8300/kernel/time.c b/arch/h8300/kernel/time.c index 7f2d6cf..165005a 100644 --- a/arch/h8300/kernel/time.c +++ b/arch/h8300/kernel/time.c @@ -41,7 +41,7 @@ void h8300_timer_tick(void) update_process_times(user_mode(get_irq_regs())); } -void __init time_init(void) +void read_persistent_clock(struct timespec *ts) { unsigned int year, mon, day, hour, min, sec; @@ -56,8 +56,12 @@ void __init time_init(void) #endif if ((year += 1900) < 1970) year += 100; - xtime.tv_sec = mktime(year, mon, day, hour, min, sec); - xtime.tv_nsec = 0; + ts->tv_sec = mktime(year, mon, day, hour, min, sec); + ts->tv_nsec = 0; +} + +void __init time_init(void) +{ h8300_timer_setup(); } -- cgit v0.10.2 From e2032a4a7237e97cd2ee36e2d806c5742ace189f Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 3 Mar 2010 19:57:21 -0800 Subject: avr32: Convert avr32 to use read/update_persistent_clock This patch converts the avr32 architecture to use the generic read_persistent_clock and update_persistent_clock interfaces, reducing the amount of arch specific code we have to maintain, and allowing for further cleanups in the future. I have not built or tested this patch, so help from arch maintainers would be appreciated. Signed-off-by: John Stultz Cc: Haavard Skinnemoen Cc: Andrew Morton LKML-Reference: <1267675049-12337-7-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/arch/avr32/kernel/time.c b/arch/avr32/kernel/time.c index f27aa3b..a047d69 100644 --- a/arch/avr32/kernel/time.c +++ b/arch/avr32/kernel/time.c @@ -110,17 +110,17 @@ static struct clock_event_device comparator = { .set_mode = comparator_mode, }; +void read_persistent_clock(struct timespec *ts) +{ + ts->tv.sec = mktime(2007, 1, 1, 0, 0, 0); + ts->tv_nsec = 0; +} + void __init time_init(void) { unsigned long counter_hz; int ret; - xtime.tv_sec = mktime(2007, 1, 1, 0, 0, 0); - xtime.tv_nsec = 0; - - set_normalized_timespec(&wall_to_monotonic, - -xtime.tv_sec, -xtime.tv_nsec); - /* figure rate for counter */ counter_hz = clk_get_rate(boot_cpu_data.clk); counter.mult = clocksource_hz2mult(counter_hz, counter.shift); -- cgit v0.10.2 From 6ffdc5774a9ef80e58db398a8307d5b2db2644ce Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 3 Mar 2010 19:57:22 -0800 Subject: ia64: Convert ia64 to use read/update_persistent_clock This patch converts the ia64 architecture to use the generic read_persistent_clock and update_persistent_clock interfaces, reducing the amount of arch specific code we have to maintain, and allowing for further cleanups in the future. I have not built or tested this patch, so help from arch maintainers would be appreciated. Signed-off-by: John Stultz Cc: Tony Luck Cc: Fenghua Yu Cc: Andrew Morton LKML-Reference: <1267675049-12337-8-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/arch/ia64/kernel/time.c b/arch/ia64/kernel/time.c index 47a1927..653b3c4 100644 --- a/arch/ia64/kernel/time.c +++ b/arch/ia64/kernel/time.c @@ -430,18 +430,16 @@ static int __init rtc_init(void) } module_init(rtc_init); +void read_persistent_clock(struct timespec *ts) +{ + efi_gettimeofday(ts); +} + void __init time_init (void) { register_percpu_irq(IA64_TIMER_VECTOR, &timer_irqaction); - efi_gettimeofday(&xtime); ia64_init_itm(); - - /* - * Initialize wall_to_monotonic such that adding it to xtime will yield zero, the - * tv_nsec field must be normalized (i.e., 0 <= nsec < NSEC_PER_SEC). - */ - set_normalized_timespec(&wall_to_monotonic, -xtime.tv_sec, -xtime.tv_nsec); } /* -- cgit v0.10.2 From cb0e996378900fa05d5c779d6bfa52d55ebc3407 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 3 Mar 2010 19:57:24 -0800 Subject: blackfin: Convert blackfin to use read/update_persistent_clock This patch converts the blackfin architecture to use the generic read_persistent_clock and update_persistent_clock interfaces, reducing the amount of arch specific code we have to maintain, and allowing for further cleanups in the future. I have not built or tested this patch, so help from arch maintainers would be appreciated. Signed-off-by: John Stultz Cc: Mike Frysinger Cc: Andrew Morton LKML-Reference: <1267675049-12337-10-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/arch/blackfin/kernel/time-ts.c b/arch/blackfin/kernel/time-ts.c index 17c38c5..08dd3e8 100644 --- a/arch/blackfin/kernel/time-ts.c +++ b/arch/blackfin/kernel/time-ts.c @@ -348,9 +348,15 @@ static int __init bfin_clockevent_init(void) return 0; } -void __init time_init(void) +void read_persistent_clock(struct timespec *ts) { time_t secs_since_1970 = (365 * 37 + 9) * 24 * 60 * 60; /* 1 Jan 2007 */ + ts->tv_sec = secs_since_1970; + ts->tv_nsec = 0; +} + +void __init time_init(void) +{ #ifdef CONFIG_RTC_DRV_BFIN /* [#2663] hack to filter junk RTC values that would cause @@ -363,11 +369,6 @@ void __init time_init(void) } #endif - /* Initialize xtime. From now on, xtime is updated with timer interrupts */ - xtime.tv_sec = secs_since_1970; - xtime.tv_nsec = 0; - set_normalized_timespec(&wall_to_monotonic, -xtime.tv_sec, -xtime.tv_nsec); - bfin_cs_cycles_init(); bfin_cs_gptimer0_init(); bfin_clockevent_init(); diff --git a/arch/blackfin/kernel/time.c b/arch/blackfin/kernel/time.c index 13c1ee3..c911361 100644 --- a/arch/blackfin/kernel/time.c +++ b/arch/blackfin/kernel/time.c @@ -112,11 +112,6 @@ u32 arch_gettimeoffset(void) } #endif -static inline int set_rtc_mmss(unsigned long nowtime) -{ - return 0; -} - /* * timer_interrupt() needs to keep up the real-time clock, * as well as call the "do_timer()" routine every clocktick @@ -126,29 +121,8 @@ __attribute__((l1_text)) #endif irqreturn_t timer_interrupt(int irq, void *dummy) { - /* last time the cmos clock got updated */ - static long last_rtc_update; - write_seqlock(&xtime_lock); do_timer(1); - - /* - * If we have an externally synchronized Linux clock, then update - * CMOS clock accordingly every ~11 minutes. Set_rtc_mmss() has to be - * called as close as possible to 500 ms before the new second starts. - */ - if (ntp_synced() && - xtime.tv_sec > last_rtc_update + 660 && - (xtime.tv_nsec / NSEC_PER_USEC) >= - 500000 - ((unsigned)TICK_SIZE) / 2 - && (xtime.tv_nsec / NSEC_PER_USEC) <= - 500000 + ((unsigned)TICK_SIZE) / 2) { - if (set_rtc_mmss(xtime.tv_sec) == 0) - last_rtc_update = xtime.tv_sec; - else - /* Do it again in 60s. */ - last_rtc_update = xtime.tv_sec - 600; - } write_sequnlock(&xtime_lock); #ifdef CONFIG_IPIPE @@ -161,10 +135,15 @@ irqreturn_t timer_interrupt(int irq, void *dummy) return IRQ_HANDLED; } -void __init time_init(void) +void read_persistent_clock(struct timespec *ts) { time_t secs_since_1970 = (365 * 37 + 9) * 24 * 60 * 60; /* 1 Jan 2007 */ + ts->tv_sec = secs_since_1970; + ts->tv_nsec = 0; +} +void __init time_init(void) +{ #ifdef CONFIG_RTC_DRV_BFIN /* [#2663] hack to filter junk RTC values that would cause * userspace to have to deal with time values greater than @@ -176,11 +155,5 @@ void __init time_init(void) } #endif - /* Initialize xtime. From now on, xtime is updated with timer interrupts */ - xtime.tv_sec = secs_since_1970; - xtime.tv_nsec = 0; - - wall_to_monotonic.tv_sec = -xtime.tv_sec; - time_sched_init(timer_interrupt); } -- cgit v0.10.2 From 944694716d6ea3c274a73c830bf33e194bad4bcd Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 3 Mar 2010 19:57:25 -0800 Subject: m32r: Convert m32r to use read/update_peristent_clock This patch converts the m32r architecture to use the generic read_persistent_clock and update_persistent_clock interfaces, reducing the amount of arch specific code we have to maintain, and allowing for further cleanups in the future. I have not built or tested this patch, so help from arch maintainers would be appreciated. Signed-off-by: John Stultz Cc: Hirokazu Takata Cc: Andrew Morton LKML-Reference: <1267675049-12337-11-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/arch/m32r/kernel/time.c b/arch/m32r/kernel/time.c index 9cedcef..bda8682 100644 --- a/arch/m32r/kernel/time.c +++ b/arch/m32r/kernel/time.c @@ -106,24 +106,6 @@ u32 arch_gettimeoffset(void) } /* - * In order to set the CMOS clock precisely, set_rtc_mmss has to be - * called 500 ms after the second nowtime has started, because when - * nowtime is written into the registers of the CMOS clock, it will - * jump to the next second precisely 500 ms later. Check the Motorola - * MC146818A or Dallas DS12887 data sheet for details. - * - * BUG: This routine does not handle hour overflow properly; it just - * sets the minutes. Usually you won't notice until after reboot! - */ -static inline int set_rtc_mmss(unsigned long nowtime) -{ - return 0; -} - -/* last time the cmos clock got updated */ -static long last_rtc_update = 0; - -/* * timer_interrupt() needs to keep up the real-time clock, * as well as call the "do_timer()" routine every clocktick */ @@ -138,23 +120,6 @@ static irqreturn_t timer_interrupt(int irq, void *dev_id) #ifndef CONFIG_SMP update_process_times(user_mode(get_irq_regs())); #endif - /* - * If we have an externally synchronized Linux clock, then update - * CMOS clock accordingly every ~11 minutes. Set_rtc_mmss() has to be - * called as close as possible to 500 ms before the new second starts. - */ - write_seqlock(&xtime_lock); - if (ntp_synced() - && xtime.tv_sec > last_rtc_update + 660 - && (xtime.tv_nsec / 1000) >= 500000 - ((unsigned)TICK_SIZE) / 2 - && (xtime.tv_nsec / 1000) <= 500000 + ((unsigned)TICK_SIZE) / 2) - { - if (set_rtc_mmss(xtime.tv_sec) == 0) - last_rtc_update = xtime.tv_sec; - else /* do it again in 60 s */ - last_rtc_update = xtime.tv_sec - 600; - } - write_sequnlock(&xtime_lock); /* As we return to user mode fire off the other CPU schedulers.. this is basically because we don't yet share IRQ's around. This message is rigged to be safe on the 386 - basically it's @@ -174,7 +139,7 @@ static struct irqaction irq0 = { .name = "MFT2", }; -void __init time_init(void) +void read_persistent_clock(struct timespec *ts) { unsigned int epoch, year, mon, day, hour, min, sec; @@ -194,11 +159,13 @@ void __init time_init(void) epoch = 1952; year += epoch; - xtime.tv_sec = mktime(year, mon, day, hour, min, sec); - xtime.tv_nsec = (INITIAL_JIFFIES % HZ) * (NSEC_PER_SEC / HZ); - set_normalized_timespec(&wall_to_monotonic, - -xtime.tv_sec, -xtime.tv_nsec); + ts->tv_sec = mktime(year, mon, day, hour, min, sec); + ts->tv_nsec = (INITIAL_JIFFIES % HZ) * (NSEC_PER_SEC / HZ); +} + +void __init time_init(void) +{ #if defined(CONFIG_CHIP_M32102) || defined(CONFIG_CHIP_XNUX2) \ || defined(CONFIG_CHIP_VDEC2) || defined(CONFIG_CHIP_M32700) \ || defined(CONFIG_CHIP_OPSP) || defined(CONFIG_CHIP_M32104) -- cgit v0.10.2 From ffbcad49e79cd82428010b44a87401446ea7f370 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 3 Mar 2010 19:57:26 -0800 Subject: m68k: Convert m68k to use read/update_persistent_clock This patch converts the m68k architecture to use the generic read_persistent_clock and update_persistent_clock interfaces, reducing the amount of arch specific code we have to maintain, and allowing for further cleanups in the future. I have not built or tested this patch, so help from arch maintainers would be appreciated. Signed-off-by: John Stultz Cc: Geert Uytterhoeven Cc: Andrew Morton LKML-Reference: <1267675049-12337-12-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/arch/m68k/kernel/time.c b/arch/m68k/kernel/time.c index 17dc2a3..4926b38 100644 --- a/arch/m68k/kernel/time.c +++ b/arch/m68k/kernel/time.c @@ -73,21 +73,24 @@ static irqreturn_t timer_interrupt(int irq, void *dummy) return IRQ_HANDLED; } -void __init time_init(void) +void read_persistent_clock(struct timespec *ts) { struct rtc_time time; + ts->tv_sec = 0; + ts->tv_nsec = 0; if (mach_hwclk) { mach_hwclk(0, &time); if ((time.tm_year += 1900) < 1970) time.tm_year += 100; - xtime.tv_sec = mktime(time.tm_year, time.tm_mon, time.tm_mday, + ts->tv_sec = mktime(time.tm_year, time.tm_mon, time.tm_mday, time.tm_hour, time.tm_min, time.tm_sec); - xtime.tv_nsec = 0; } - wall_to_monotonic.tv_sec = -xtime.tv_sec; +} +void __init time_init(void) +{ mach_sched_init(timer_interrupt); } -- cgit v0.10.2 From 8eff8a5c1d3a1de7a2d173e0effc6cc9bcbb5177 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 3 Mar 2010 19:57:28 -0800 Subject: cris: Convert cris to use read/update_persistent_clock This patch converts the cris architecture to use the generic read_persistent_clock and update_persistent_clock interfaces, reducing the amount of arch specific code we have to maintain, and allowing for further cleanups in the future. I have not built or tested this patch, so help from arch maintainers would be appreciated. Signed-off-by: John Stultz Cc: Mikael Starvik Cc: Jesper Nilsson Cc: Andrew Morton LKML-Reference: <1267675049-12337-14-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/arch/cris/Kconfig b/arch/cris/Kconfig index 059eac6..e25bf44 100644 --- a/arch/cris/Kconfig +++ b/arch/cris/Kconfig @@ -23,6 +23,9 @@ config RWSEM_XCHGADD_ALGORITHM config GENERIC_TIME def_bool y +config GENERIC_CMOS_UPDATE + def_bool y + config ARCH_USES_GETTIMEOFFSET def_bool y diff --git a/arch/cris/arch-v10/kernel/time.c b/arch/cris/arch-v10/kernel/time.c index 31ca141..30adae5 100644 --- a/arch/cris/arch-v10/kernel/time.c +++ b/arch/cris/arch-v10/kernel/time.c @@ -26,7 +26,6 @@ /* it will make jiffies at 96 hz instead of 100 hz though */ #undef USE_CASCADE_TIMERS -extern void update_xtime_from_cmos(void); extern int set_rtc_mmss(unsigned long nowtime); extern int have_rtc; @@ -188,8 +187,6 @@ stop_watchdog(void) #endif } -/* last time the cmos clock got updated */ -static long last_rtc_update = 0; /* * timer_interrupt() needs to keep up the real-time clock, @@ -232,24 +229,6 @@ timer_interrupt(int irq, void *dev_id) do_timer(1); cris_do_profile(regs); /* Save profiling information */ - - /* - * If we have an externally synchronized Linux clock, then update - * CMOS clock accordingly every ~11 minutes. Set_rtc_mmss() has to be - * called as close as possible to 500 ms before the new second starts. - * - * The division here is not time critical since it will run once in - * 11 minutes - */ - if (ntp_synced() && - xtime.tv_sec > last_rtc_update + 660 && - (xtime.tv_nsec / 1000) >= 500000 - (tick_nsec / 1000) / 2 && - (xtime.tv_nsec / 1000) <= 500000 + (tick_nsec / 1000) / 2) { - if (set_rtc_mmss(xtime.tv_sec) == 0) - last_rtc_update = xtime.tv_sec; - else - last_rtc_update = xtime.tv_sec - 600; /* do it again in 60 s */ - } return IRQ_HANDLED; } @@ -274,22 +253,10 @@ time_init(void) */ loops_per_usec = 50; - if(RTC_INIT() < 0) { - /* no RTC, start at 1980 */ - xtime.tv_sec = 0; - xtime.tv_nsec = 0; + if(RTC_INIT() < 0) have_rtc = 0; - } else { - /* get the current time */ + else have_rtc = 1; - update_xtime_from_cmos(); - } - - /* - * Initialize wall_to_monotonic such that adding it to xtime will yield zero, the - * tv_nsec field must be normalized (i.e., 0 <= nsec < NSEC_PER_SEC). - */ - set_normalized_timespec(&wall_to_monotonic, -xtime.tv_sec, -xtime.tv_nsec); /* Setup the etrax timers * Base frequency is 25000 hz, divider 250 -> 100 HZ diff --git a/arch/cris/arch-v32/kernel/time.c b/arch/cris/arch-v32/kernel/time.c index b1920d8..1ee0e10 100644 --- a/arch/cris/arch-v32/kernel/time.c +++ b/arch/cris/arch-v32/kernel/time.c @@ -44,7 +44,6 @@ unsigned long timer_regs[NR_CPUS] = #endif }; -extern void update_xtime_from_cmos(void); extern int set_rtc_mmss(unsigned long nowtime); extern int have_rtc; @@ -198,9 +197,6 @@ handle_watchdog_bite(struct pt_regs* regs) #endif } -/* Last time the cmos clock got updated. */ -static long last_rtc_update = 0; - /* * timer_interrupt() needs to keep up the real-time clock, * as well as call the "do_timer()" routine every clocktick. @@ -238,25 +234,6 @@ timer_interrupt(int irq, void *dev_id) /* Call the real timer interrupt handler */ do_timer(1); - - /* - * If we have an externally synchronized Linux clock, then update - * CMOS clock accordingly every ~11 minutes. Set_rtc_mmss() has to be - * called as close as possible to 500 ms before the new second starts. - * - * The division here is not time critical since it will run once in - * 11 minutes - */ - if ((time_status & STA_UNSYNC) == 0 && - xtime.tv_sec > last_rtc_update + 660 && - (xtime.tv_nsec / 1000) >= 500000 - (tick_nsec / 1000) / 2 && - (xtime.tv_nsec / 1000) <= 500000 + (tick_nsec / 1000) / 2) { - if (set_rtc_mmss(xtime.tv_sec) == 0) - last_rtc_update = xtime.tv_sec; - else - /* Do it again in 60 s */ - last_rtc_update = xtime.tv_sec - 600; - } return IRQ_HANDLED; } @@ -309,23 +286,10 @@ time_init(void) */ loops_per_usec = 50; - if(RTC_INIT() < 0) { - /* No RTC, start at 1980 */ - xtime.tv_sec = 0; - xtime.tv_nsec = 0; + if(RTC_INIT() < 0) have_rtc = 0; - } else { - /* Get the current time */ + else have_rtc = 1; - update_xtime_from_cmos(); - } - - /* - * Initialize wall_to_monotonic such that adding it to - * xtime will yield zero, the tv_nsec field must be normalized - * (i.e., 0 <= nsec < NSEC_PER_SEC). - */ - set_normalized_timespec(&wall_to_monotonic, -xtime.tv_sec, -xtime.tv_nsec); /* Start CPU local timer. */ cris_timer_init(); diff --git a/arch/cris/kernel/time.c b/arch/cris/kernel/time.c index a05dd31..c72730d 100644 --- a/arch/cris/kernel/time.c +++ b/arch/cris/kernel/time.c @@ -98,6 +98,8 @@ unsigned long get_cmos_time(void) { unsigned int year, mon, day, hour, min, sec; + if(!have_rtc) + return 0; sec = CMOS_READ(RTC_SECONDS); min = CMOS_READ(RTC_MINUTES); @@ -119,19 +121,19 @@ get_cmos_time(void) return mktime(year, mon, day, hour, min, sec); } -/* update xtime from the CMOS settings. used when /dev/rtc gets a SET_TIME. - * TODO: this doesn't reset the fancy NTP phase stuff as do_settimeofday does. - */ -void -update_xtime_from_cmos(void) +int update_persistent_clock(struct timespec now) { - if(have_rtc) { - xtime.tv_sec = get_cmos_time(); - xtime.tv_nsec = 0; - } + return set_rtc_mmss(now.tv_sec); } +void read_persistent_clock(struct timespec *ts) +{ + ts->tv_sec = get_cmos_time(); + ts->tv_nsec = 0; +} + + extern void cris_profile_sample(struct pt_regs* regs); void -- cgit v0.10.2 From f5c9c9be267f73eda6a668ac81ddde56ed7fa793 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 3 Mar 2010 19:57:27 -0800 Subject: sparc: Convert sparc to use read/update_persistent_clock This patch converts the sparc architecture to use the generic read_persistent_clock and update_persistent_clock interfaces, reducing the amount of arch specific code we have to maintain, and allowing for further cleanups in the future. [ davem: compile fix: Here's a version that compiles, you have to get rid of the now unused variably last_rtc_update since we build with -Werror ] Signed-off-by: John Stultz Acked-by: David S. Miller Cc: Andrew Morton Cc: sparclinux@vger.kernel.org Signed-off-by: Thomas Gleixner diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig index 4097f6a..0e8ec2a 100644 --- a/arch/sparc/Kconfig +++ b/arch/sparc/Kconfig @@ -72,7 +72,7 @@ config ARCH_USES_GETTIMEOFFSET config GENERIC_CMOS_UPDATE bool - default y if SPARC64 + default y config GENERIC_CLOCKEVENTS bool diff --git a/arch/sparc/kernel/time_32.c b/arch/sparc/kernel/time_32.c index 0d4c09b..4453003 100644 --- a/arch/sparc/kernel/time_32.c +++ b/arch/sparc/kernel/time_32.c @@ -78,6 +78,11 @@ __volatile__ unsigned int *master_l10_counter; u32 (*do_arch_gettimeoffset)(void); +int update_persistent_clock(struct timespec now) +{ + return set_rtc_mmss(now.tv_sec); +} + /* * timer_interrupt() needs to keep up the real-time clock, * as well as call the "do_timer()" routine every clocktick @@ -87,9 +92,6 @@ u32 (*do_arch_gettimeoffset)(void); static irqreturn_t timer_interrupt(int dummy, void *dev_id) { - /* last time the cmos clock got updated */ - static long last_rtc_update; - #ifndef CONFIG_SMP profile_tick(CPU_PROFILING); #endif @@ -101,16 +103,6 @@ static irqreturn_t timer_interrupt(int dummy, void *dev_id) do_timer(1); - /* Determine when to update the Mostek clock. */ - if (ntp_synced() && - xtime.tv_sec > last_rtc_update + 660 && - (xtime.tv_nsec / 1000) >= 500000 - ((unsigned) TICK_SIZE) / 2 && - (xtime.tv_nsec / 1000) <= 500000 + ((unsigned) TICK_SIZE) / 2) { - if (set_rtc_mmss(xtime.tv_sec) == 0) - last_rtc_update = xtime.tv_sec; - else - last_rtc_update = xtime.tv_sec - 600; /* do it again in 60 s */ - } write_sequnlock(&xtime_lock); #ifndef CONFIG_SMP -- cgit v0.10.2 From 567e54790e5c07152a93b6de4d0210af8b77da87 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 12 Mar 2010 21:05:10 -0300 Subject: perf tools: Fix non-newt build MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The use_browser needs to be in a file that is always built and also we need a browser__show_help stub in that case. Reported-by: Anton Blanchard Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1268438710-32697-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/perf.c b/tools/perf/perf.c index 9ff186b..0d4b9ed 100644 --- a/tools/perf/perf.c +++ b/tools/perf/perf.c @@ -16,6 +16,8 @@ #include "util/string.h" #include "util/debugfs.h" +bool use_browser; + const char perf_usage_string[] = "perf [--version] [--help] COMMAND [ARGS]"; diff --git a/tools/perf/util/debug.h b/tools/perf/util/debug.h index 03accb8..0172edf 100644 --- a/tools/perf/util/debug.h +++ b/tools/perf/util/debug.h @@ -9,6 +9,14 @@ extern int dump_trace; int dump_printf(const char *fmt, ...) __attribute__((format(printf, 1, 2))); void trace_event(event_t *event); + +#ifdef NO_NEWT_SUPPORT +static inline int browser__show_help(const char *format __used, va_list ap __used) +{ + return 0; +} +#else int browser__show_help(const char *format, va_list ap); +#endif #endif /* __PERF_DEBUG_H */ diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index f6ec6f5..2836394 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -189,8 +189,6 @@ int browser__show_help(const char *format, va_list ap) return ret; } -bool use_browser; - void setup_browser(void) { if (!isatty(1)) -- cgit v0.10.2 From 8576e1971663ffdb6139041de97cdd2e1d4791cc Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Sat, 13 Mar 2010 11:11:16 +0300 Subject: x86, perf: Unmask LVTPC only if we have APIC supported Ingo reported: | | There's a build failure on -tip with the P4 driver, on UP 32-bit, if | PERF_EVENTS is enabled but UP_APIC is disabled: | | arch/x86/built-in.o: In function `p4_pmu_handle_irq': | perf_event.c:(.text+0xa756): undefined reference to `apic' | perf_event.c:(.text+0xa76e): undefined reference to `apic' | So we have to unmask LVTPC only if we're configured to have one. Reported-by: Ingo Molnar Signed-off-by: Cyrill Gorcunov CC: Lin Ming CC: Peter Zijlstra LKML-Reference: <20100313081116.GA5179@lenovo> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index 381f593..ef861da1 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -365,8 +365,10 @@ static int p4_pmu_handle_irq(struct pt_regs *regs) } if (handled) { +#ifdef CONFIG_X86_LOCAL_APIC /* p4 quirk: unmask it again */ apic_write(APIC_LVTPC, apic_read(APIC_LVTPC) & ~APIC_LVT_MASKED); +#endif inc_irq_stat(apic_perf_irqs); } -- cgit v0.10.2 From 3af4974eb2c7867d6e160977195dfde586d0e564 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Wed, 3 Feb 2010 17:31:31 +1100 Subject: sunrpc: don't keep expired entries in the auth caches. currently expired entries remain in the auth caches as long as there is a reference. This was needed long ago when the auth_domain cache used the same cache infrastructure. But since that (being a very different sort of cache) was separated, this test is no longer needed. So remove the test on refcnt and tidy up the surrounding code. This allows the cache_dequeue call (which needed to be there to drop a potentially awkward reference) can be moved outside of the spinlock which is a better place for it. Signed-off-by: NeilBrown Signed-off-by: J. Bruce Fields diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c index 39bddba..83592e0 100644 --- a/net/sunrpc/cache.c +++ b/net/sunrpc/cache.c @@ -397,31 +397,28 @@ static int cache_clean(void) /* Ok, now to clean this strand */ cp = & current_detail->hash_table[current_index]; - ch = *cp; - for (; ch; cp= & ch->next, ch= *cp) { + for (ch = *cp ; ch ; cp = & ch->next, ch = *cp) { if (current_detail->nextcheck > ch->expiry_time) current_detail->nextcheck = ch->expiry_time+1; if (ch->expiry_time >= get_seconds() && ch->last_refresh >= current_detail->flush_time) continue; - if (test_and_clear_bit(CACHE_PENDING, &ch->flags)) - cache_dequeue(current_detail, ch); - if (atomic_read(&ch->ref.refcount) == 1) - break; - } - if (ch) { *cp = ch->next; ch->next = NULL; current_detail->entries--; rv = 1; + break; } + write_unlock(¤t_detail->hash_lock); d = current_detail; if (!ch) current_index ++; spin_unlock(&cache_list_lock); if (ch) { + if (test_and_clear_bit(CACHE_PENDING, &ch->flags)) + cache_dequeue(current_detail, ch); cache_revisit_request(ch); cache_put(ch, d); } -- cgit v0.10.2 From 2f50d8b63dd6e5320a9d223298df19df3502da29 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Wed, 3 Feb 2010 17:31:31 +1100 Subject: sunrpc/cache: factor out cache_is_expired This removes a tiny bit of code duplication, but more important prepares for following patch which will perform the expiry check in cache_lookup and the rest of the validity check in cache_check. Signed-off-by: NeilBrown Signed-off-by: J. Bruce Fields diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c index 83592e0..9826c5c 100644 --- a/net/sunrpc/cache.c +++ b/net/sunrpc/cache.c @@ -49,6 +49,12 @@ static void cache_init(struct cache_head *h) h->last_refresh = now; } +static inline int cache_is_expired(struct cache_detail *detail, struct cache_head *h) +{ + return (h->expiry_time < get_seconds()) || + (detail->flush_time > h->last_refresh); +} + struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail, struct cache_head *key, int hash) { @@ -184,9 +190,7 @@ static int cache_make_upcall(struct cache_detail *cd, struct cache_head *h) static inline int cache_is_valid(struct cache_detail *detail, struct cache_head *h) { if (!test_bit(CACHE_VALID, &h->flags) || - h->expiry_time < get_seconds()) - return -EAGAIN; - else if (detail->flush_time > h->last_refresh) + cache_is_expired(detail, h)) return -EAGAIN; else { /* entry is valid */ @@ -400,8 +404,7 @@ static int cache_clean(void) for (ch = *cp ; ch ; cp = & ch->next, ch = *cp) { if (current_detail->nextcheck > ch->expiry_time) current_detail->nextcheck = ch->expiry_time+1; - if (ch->expiry_time >= get_seconds() && - ch->last_refresh >= current_detail->flush_time) + if (!cache_is_expired(current_detail, ch)) continue; *cp = ch->next; -- cgit v0.10.2 From d202cce8963d9268ff355a386e20243e8332b308 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Wed, 3 Feb 2010 17:31:31 +1100 Subject: sunrpc: never return expired entries in sunrpc_cache_lookup If sunrpc_cache_lookup finds an expired entry, remove it from the cache and return a freshly created non-VALID entry instead. This ensures that we only ever get a usable entry, or an entry that will become usable once an update arrives. i.e. we will never need to repeat the lookup. This allows us to remove the 'is_expired' test from cache_check (i.e. from cache_is_valid). cache_check should never get an expired entry as 'lookup' will never return one. If it does happen - due to inconvenient timing - then just accept it as still valid, it won't be very much past it's use-by date. Signed-off-by: NeilBrown Signed-off-by: J. Bruce Fields diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c index 9826c5c..3e1ef8b 100644 --- a/net/sunrpc/cache.c +++ b/net/sunrpc/cache.c @@ -59,7 +59,7 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail, struct cache_head *key, int hash) { struct cache_head **head, **hp; - struct cache_head *new = NULL; + struct cache_head *new = NULL, *freeme = NULL; head = &detail->hash_table[hash]; @@ -68,6 +68,9 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail, for (hp=head; *hp != NULL ; hp = &(*hp)->next) { struct cache_head *tmp = *hp; if (detail->match(tmp, key)) { + if (cache_is_expired(detail, tmp)) + /* This entry is expired, we will discard it. */ + break; cache_get(tmp); read_unlock(&detail->hash_lock); return tmp; @@ -92,6 +95,13 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail, for (hp=head; *hp != NULL ; hp = &(*hp)->next) { struct cache_head *tmp = *hp; if (detail->match(tmp, key)) { + if (cache_is_expired(detail, tmp)) { + *hp = tmp->next; + tmp->next = NULL; + detail->entries --; + freeme = tmp; + break; + } cache_get(tmp); write_unlock(&detail->hash_lock); cache_put(new, detail); @@ -104,6 +114,8 @@ struct cache_head *sunrpc_cache_lookup(struct cache_detail *detail, cache_get(new); write_unlock(&detail->hash_lock); + if (freeme) + cache_put(freeme, detail); return new; } EXPORT_SYMBOL_GPL(sunrpc_cache_lookup); @@ -189,8 +201,7 @@ static int cache_make_upcall(struct cache_detail *cd, struct cache_head *h) static inline int cache_is_valid(struct cache_detail *detail, struct cache_head *h) { - if (!test_bit(CACHE_VALID, &h->flags) || - cache_is_expired(detail, h)) + if (!test_bit(CACHE_VALID, &h->flags)) return -EAGAIN; else { /* entry is valid */ -- cgit v0.10.2 From e4495262826d1eabca3529fa6ac22394eb348132 Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Mon, 15 Mar 2010 12:58:22 +0800 Subject: perf, x86: Enable not tagged retired instruction counting on P4s This should turn on instruction counting on P4s, which was missing in the first version of the new PMU driver. It's inaccurate for now, we still need dependant event to tag mops before we can count them precisely. The result is that the number of instruction may be lifted up. Signed-off-by: Cyrill Gorcunov Signed-off-by: Lin Ming Cc: Peter Zijlstra LKML-Reference: <1268629102.3355.11.camel@minggr.sh.intel.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/perf_event_p4.h b/arch/x86/include/asm/perf_event_p4.h index 829f471..b47b9e9 100644 --- a/arch/x86/include/asm/perf_event_p4.h +++ b/arch/x86/include/asm/perf_event_p4.h @@ -324,8 +324,8 @@ static inline u32 p4_default_escr_conf(int cpu, int exclude_os, int exclude_usr) #define P4_SSE_INPUT_ASSIST P4_EVENT_PACK(0x34, 0x01) /* - * MSR_P4_FIRM_ESCR: 8, 9 - * MSR_P4_FIRM_ESCR: 10, 11 + * MSR_P4_FIRM_ESCR0: 8, 9 + * MSR_P4_FIRM_ESCR1: 10, 11 */ #define P4_PACKED_SP_UOP P4_EVENT_PACK(0x08, 0x01) @@ -462,8 +462,8 @@ static inline u32 p4_default_escr_conf(int cpu, int exclude_os, int exclude_usr) #define P4_INSTR_RETIRED P4_EVENT_PACK(0x02, 0x04) /* - * MSR_P4_CRU_ESCR2: 12, 13, 16 - * MSR_P4_CRU_ESCR3: 14, 15, 17 + * MSR_P4_CRU_ESCR0: 12, 13, 16 + * MSR_P4_CRU_ESCR1: 14, 15, 17 */ #define P4_UOPS_RETIRED P4_EVENT_PACK(0x01, 0x04) diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index ef861da1..a11ce73 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -60,13 +60,11 @@ struct p4_event_template p4_templates[] = { [2] = { .opcode = P4_INSTR_RETIRED, .config = 0, - .dep = 0, /* needs front-end tagging */ + .dep = -1, /* needs front-end tagging */ .emask = P4_EVENT_ATTR(P4_INSTR_RETIRED, NBOGUSNTAG) | - P4_EVENT_ATTR(P4_INSTR_RETIRED, NBOGUSTAG) | - P4_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSNTAG) | - P4_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSTAG), - .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR3 }, + P4_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSNTAG), + .escr_msr = { MSR_P4_CRU_ESCR0, MSR_P4_CRU_ESCR1 }, .cntr = { 12, 14 }, }, [3] = { -- cgit v0.10.2 From 2cda2728aa1c8c006418a24f867b25e5eb7a32e2 Mon Sep 17 00:00:00 2001 From: "Martin K. Petersen" Date: Mon, 15 Mar 2010 12:46:51 +0100 Subject: block: Fix overrun in lcm() and move it to lib lcm() was defined to take integer-sized arguments. The supplied arguments are multiplied, however, causing us to overflow given sufficiently large input. That in turn led to incorrect optimal I/O size reporting in some cases (RAID over RAID). Switch lcm() over to unsigned long similar to gcd() and move the function from blk-settings.c to lib. Signed-off-by: Martin K. Petersen Signed-off-by: Jens Axboe diff --git a/block/blk-settings.c b/block/blk-settings.c index 31e7a93..4c4700d 100644 --- a/block/blk-settings.c +++ b/block/blk-settings.c @@ -8,6 +8,7 @@ #include #include /* for max_pfn/max_low_pfn */ #include +#include #include #include "blk.h" @@ -461,16 +462,6 @@ void blk_queue_stack_limits(struct request_queue *t, struct request_queue *b) } EXPORT_SYMBOL(blk_queue_stack_limits); -static unsigned int lcm(unsigned int a, unsigned int b) -{ - if (a && b) - return (a * b) / gcd(a, b); - else if (b) - return b; - - return a; -} - /** * blk_stack_limits - adjust queue_limits for stacked devices * @t: the stacking driver limits (top device) diff --git a/include/linux/lcm.h b/include/linux/lcm.h new file mode 100644 index 0000000..7bf01d7 --- /dev/null +++ b/include/linux/lcm.h @@ -0,0 +1,8 @@ +#ifndef _LCM_H +#define _LCM_H + +#include + +unsigned long lcm(unsigned long a, unsigned long b) __attribute_const__; + +#endif /* _LCM_H */ diff --git a/lib/Makefile b/lib/Makefile index 2e152ae..0d40152 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -21,7 +21,7 @@ lib-y += kobject.o kref.o klist.o obj-y += bcd.o div64.o sort.o parser.o halfmd4.o debug_locks.o random32.o \ bust_spinlocks.o hexdump.o kasprintf.o bitmap.o scatterlist.o \ - string_helpers.o gcd.o list_sort.o + string_helpers.o gcd.o lcm.o list_sort.o ifeq ($(CONFIG_DEBUG_KOBJECT),y) CFLAGS_kobject.o += -DDEBUG diff --git a/lib/lcm.c b/lib/lcm.c new file mode 100644 index 0000000..157cd88 --- /dev/null +++ b/lib/lcm.c @@ -0,0 +1,15 @@ +#include +#include +#include + +/* Lowest common multiple */ +unsigned long lcm(unsigned long a, unsigned long b) +{ + if (a && b) + return (a * b) / gcd(a, b); + else if (b) + return b; + + return a; +} +EXPORT_SYMBOL_GPL(lcm); -- cgit v0.10.2 From ee714f2dd33e726346e34f5cda12543162f4753e Mon Sep 17 00:00:00 2001 From: "Martin K. Petersen" Date: Wed, 10 Mar 2010 00:48:32 -0500 Subject: block: Finalize conversion of block limits functions Remove compatibility wrappers and update remaining drivers. Signed-off-by: Martin K. Petersen Signed-off-by: Jens Axboe diff --git a/drivers/block/DAC960.c b/drivers/block/DAC960.c index 459f1bc..c5f22bb 100644 --- a/drivers/block/DAC960.c +++ b/drivers/block/DAC960.c @@ -2533,7 +2533,6 @@ static bool DAC960_RegisterBlockDevice(DAC960_Controller_T *Controller) Controller->RequestQueue[n] = RequestQueue; blk_queue_bounce_limit(RequestQueue, Controller->BounceBufferLimit); RequestQueue->queuedata = Controller; - blk_queue_max_hw_segments(RequestQueue, Controller->DriverScatterGatherLimit); blk_queue_max_segments(RequestQueue, Controller->DriverScatterGatherLimit); blk_queue_max_hw_sectors(RequestQueue, Controller->MaxBlocksPerCommand); disk->queue = RequestQueue; diff --git a/drivers/block/virtio_blk.c b/drivers/block/virtio_blk.c index 3c64af0..653817c 100644 --- a/drivers/block/virtio_blk.c +++ b/drivers/block/virtio_blk.c @@ -347,14 +347,13 @@ static int __devinit virtblk_probe(struct virtio_device *vdev) set_capacity(vblk->disk, cap); /* We can handle whatever the host told us to handle. */ - blk_queue_max_phys_segments(q, vblk->sg_elems-2); - blk_queue_max_hw_segments(q, vblk->sg_elems-2); + blk_queue_max_segments(q, vblk->sg_elems-2); /* No need to bounce any requests */ blk_queue_bounce_limit(q, BLK_BOUNCE_ANY); /* No real sector limit. */ - blk_queue_max_sectors(q, -1U); + blk_queue_max_hw_sectors(q, -1U); /* Host can optionally specify maximum segment size and number of * segments. */ diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index ebd22db..41551c9 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -921,26 +921,7 @@ extern void blk_cleanup_queue(struct request_queue *); extern void blk_queue_make_request(struct request_queue *, make_request_fn *); extern void blk_queue_bounce_limit(struct request_queue *, u64); extern void blk_queue_max_hw_sectors(struct request_queue *, unsigned int); - -/* Temporary compatibility wrapper */ -static inline void blk_queue_max_sectors(struct request_queue *q, unsigned int max) -{ - blk_queue_max_hw_sectors(q, max); -} - extern void blk_queue_max_segments(struct request_queue *, unsigned short); - -static inline void blk_queue_max_phys_segments(struct request_queue *q, unsigned short max) -{ - blk_queue_max_segments(q, max); -} - -static inline void blk_queue_max_hw_segments(struct request_queue *q, unsigned short max) -{ - blk_queue_max_segments(q, max); -} - - extern void blk_queue_max_segment_size(struct request_queue *, unsigned int); extern void blk_queue_max_discard_sectors(struct request_queue *q, unsigned int max_discard_sectors); @@ -1030,11 +1011,6 @@ static inline int sb_issue_discard(struct super_block *sb, extern int blk_verify_command(unsigned char *cmd, fmode_t has_write_perm); -#define MAX_PHYS_SEGMENTS 128 -#define MAX_HW_SEGMENTS 128 -#define SAFE_MAX_SECTORS 255 -#define MAX_SEGMENT_SIZE 65536 - enum blk_default_limits { BLK_MAX_SEGMENTS = 128, BLK_SAFE_MAX_SECTORS = 255, -- cgit v0.10.2 From c77a5710b7e23847bfdb81fcaa10b585f65c960a Mon Sep 17 00:00:00 2001 From: "Martin K. Petersen" Date: Wed, 10 Mar 2010 00:48:33 -0500 Subject: block: Export max number of segments and max segment size in sysfs These two values are useful when debugging issues surrounding maximum I/O size. Put them in sysfs with the rest of the queue limits. Signed-off-by: Martin K. Petersen Signed-off-by: Jens Axboe diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index e854424..fad8655 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -106,6 +106,19 @@ static ssize_t queue_max_sectors_show(struct request_queue *q, char *page) return queue_var_show(max_sectors_kb, (page)); } +static ssize_t queue_max_segments_show(struct request_queue *q, char *page) +{ + return queue_var_show(queue_max_segments(q), (page)); +} + +static ssize_t queue_max_segment_size_show(struct request_queue *q, char *page) +{ + if (test_bit(QUEUE_FLAG_CLUSTER, &q->queue_flags)) + return queue_var_show(queue_max_segment_size(q), (page)); + + return queue_var_show(PAGE_CACHE_SIZE, (page)); +} + static ssize_t queue_logical_block_size_show(struct request_queue *q, char *page) { return queue_var_show(queue_logical_block_size(q), page); @@ -280,6 +293,16 @@ static struct queue_sysfs_entry queue_max_hw_sectors_entry = { .show = queue_max_hw_sectors_show, }; +static struct queue_sysfs_entry queue_max_segments_entry = { + .attr = {.name = "max_segments", .mode = S_IRUGO }, + .show = queue_max_segments_show, +}; + +static struct queue_sysfs_entry queue_max_segment_size_entry = { + .attr = {.name = "max_segment_size", .mode = S_IRUGO }, + .show = queue_max_segment_size_show, +}; + static struct queue_sysfs_entry queue_iosched_entry = { .attr = {.name = "scheduler", .mode = S_IRUGO | S_IWUSR }, .show = elv_iosched_show, @@ -355,6 +378,8 @@ static struct attribute *default_attrs[] = { &queue_ra_entry.attr, &queue_max_hw_sectors_entry.attr, &queue_max_sectors_entry.attr, + &queue_max_segments_entry.attr, + &queue_max_segment_size_entry.attr, &queue_iosched_entry.attr, &queue_hw_sector_size_entry.attr, &queue_logical_block_size_entry.attr, -- cgit v0.10.2 From bedbfdea31daf3880745001d56450c683959ee7e Mon Sep 17 00:00:00 2001 From: Eric B Munson Date: Mon, 15 Mar 2010 11:46:57 -0300 Subject: perf record: Enable the enable_on_exec flag if record forks the target When forking its target, perf record can capture data from before the target application is started. Perf stat uses the enable_on_exec flag in the event attributes to keep from displaying events from before the target program starts, this patch adds the same functionality to perf record when it is will fork the target process. Signed-off-by: Eric B Munson Signed-off-by: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Frederic Weisbecker LKML-Reference: <1268664418-28328-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index bed175d..962cdbf 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -225,7 +225,7 @@ static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int n return h_attr; } -static void create_counter(int counter, int cpu, pid_t pid) +static void create_counter(int counter, int cpu, pid_t pid, bool forks) { char *filter = filters[counter]; struct perf_event_attr *attr = attrs + counter; @@ -277,6 +277,9 @@ static void create_counter(int counter, int cpu, pid_t pid) attr->inherit = inherit; attr->disabled = 1; + if (forks) + attr->enable_on_exec = 1; + try_again: fd[nr_cpu][counter] = sys_perf_event_open(attr, pid, cpu, group_fd, 0); @@ -381,13 +384,13 @@ try_again: ioctl(fd[nr_cpu][counter], PERF_EVENT_IOC_ENABLE); } -static void open_counters(int cpu, pid_t pid) +static void open_counters(int cpu, pid_t pid, bool forks) { int counter; group_fd = -1; for (counter = 0; counter < nr_counters; counter++) - create_counter(counter, cpu, pid); + create_counter(counter, cpu, pid, forks); nr_cpu++; } @@ -547,11 +550,11 @@ static int __cmd_record(int argc, const char **argv) if ((!system_wide && !inherit) || profile_cpu != -1) { - open_counters(profile_cpu, target_pid); + open_counters(profile_cpu, target_pid, forks); } else { nr_cpus = read_cpu_map(); for (i = 0; i < nr_cpus; i++) - open_counters(cpumap[i], target_pid); + open_counters(cpumap[i], target_pid, forks); } if (file_new) { -- cgit v0.10.2 From b0a9ab62ab96e258a0ddd81d7fe2719c3db36006 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 15 Mar 2010 11:46:58 -0300 Subject: perf top: Properly notify the user that vmlinux is missing MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Before this patch this message would very briefly appear on the screen and then the screen would get updates only on the top, for number of interrupts received, etc, but no annotation would be performed: [root@doppio linux-2.6-tip]# perf top -s n_tty_write > /tmp/bla objdump: '[kernel.kallsyms]': No such file Now this is what the user gets: [root@doppio linux-2.6-tip]# perf top -s n_tty_write Can't annotate n_tty_write: No vmlinux file was found in the path: [0] vmlinux [1] /boot/vmlinux [2] /boot/vmlinux-2.6.33-rc5 [3] /lib/modules/2.6.33-rc5/build/vmlinux [4] /usr/lib/debug/lib/modules/2.6.33-rc5/vmlinux [root@doppio linux-2.6-tip]# This bug was introduced when we added automatic search for vmlinux, before that time the user had to specify a vmlinux file. Reported-by: David S. Miller Reported-by: Ingo Molnar Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Cc: LKML-Reference: <1268664418-28328-2-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index 57e232f..c968bd3 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -169,7 +169,7 @@ static void sig_winch_handler(int sig __used) update_print_entries(&winsize); } -static void parse_source(struct sym_entry *syme) +static int parse_source(struct sym_entry *syme) { struct symbol *sym; struct sym_entry_source *source; @@ -180,12 +180,21 @@ static void parse_source(struct sym_entry *syme) u64 len; if (!syme) - return; + return -1; + + sym = sym_entry__symbol(syme); + map = syme->map; + + /* + * We can't annotate with just /proc/kallsyms + */ + if (map->dso->origin == DSO__ORIG_KERNEL) + return -1; if (syme->src == NULL) { syme->src = zalloc(sizeof(*source)); if (syme->src == NULL) - return; + return -1; pthread_mutex_init(&syme->src->lock, NULL); } @@ -195,9 +204,6 @@ static void parse_source(struct sym_entry *syme) pthread_mutex_lock(&source->lock); goto out_assign; } - - sym = sym_entry__symbol(syme); - map = syme->map; path = map->dso->long_name; len = sym->end - sym->start; @@ -209,7 +215,7 @@ static void parse_source(struct sym_entry *syme) file = popen(command, "r"); if (!file) - return; + return -1; pthread_mutex_lock(&source->lock); source->lines_tail = &source->lines; @@ -245,6 +251,7 @@ static void parse_source(struct sym_entry *syme) out_assign: sym_filter_entry = syme; pthread_mutex_unlock(&source->lock); + return 0; } static void __zero_source_counters(struct sym_entry *syme) @@ -991,7 +998,17 @@ static void event__process_sample(const event_t *self, if (sym_filter_entry_sched) { sym_filter_entry = sym_filter_entry_sched; sym_filter_entry_sched = NULL; - parse_source(sym_filter_entry); + if (parse_source(sym_filter_entry) < 0) { + struct symbol *sym = sym_entry__symbol(sym_filter_entry); + + pr_err("Can't annotate %s", sym->name); + if (sym_filter_entry->map->dso->origin == DSO__ORIG_KERNEL) { + pr_err(": No vmlinux file was found in the path:\n"); + vmlinux_path__fprintf(stderr); + } else + pr_err(".\n"); + exit(1); + } } syme = symbol__priv(al.sym); diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 75cd468..292f941 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -18,18 +18,6 @@ #define NT_GNU_BUILD_ID 3 #endif -enum dso_origin { - DSO__ORIG_KERNEL = 0, - DSO__ORIG_JAVA_JIT, - DSO__ORIG_BUILD_ID_CACHE, - DSO__ORIG_FEDORA, - DSO__ORIG_UBUNTU, - DSO__ORIG_BUILDID, - DSO__ORIG_DSO, - DSO__ORIG_KMODULE, - DSO__ORIG_NOT_FOUND, -}; - static void dsos__add(struct list_head *head, struct dso *dso); static struct map *map__new2(u64 start, struct dso *dso, enum map_type type); static int dso__load_kernel_sym(struct dso *self, struct map *map, @@ -1017,7 +1005,7 @@ static int dso__load_sym(struct dso *self, struct map *map, const char *name, } curr_map->map_ip = identity__map_ip; curr_map->unmap_ip = identity__map_ip; - curr_dso->origin = DSO__ORIG_KERNEL; + curr_dso->origin = self->origin; map_groups__insert(kmap->kmaps, curr_map); dsos__add(&dsos__kernel, curr_dso); dso__set_loaded(curr_dso, map->type); @@ -1887,6 +1875,17 @@ out_fail: return -1; } +size_t vmlinux_path__fprintf(FILE *fp) +{ + int i; + size_t printed = 0; + + for (i = 0; i < vmlinux_path__nr_entries; ++i) + printed += fprintf(fp, "[%d] %s\n", i, vmlinux_path[i]); + + return printed; +} + static int setup_list(struct strlist **list, const char *list_str, const char *list_name) { diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h index 280dadd..d983bba 100644 --- a/tools/perf/util/symbol.h +++ b/tools/perf/util/symbol.h @@ -149,6 +149,19 @@ size_t dsos__fprintf_buildid(FILE *fp, bool with_hits); size_t dso__fprintf_buildid(struct dso *self, FILE *fp); size_t dso__fprintf(struct dso *self, enum map_type type, FILE *fp); + +enum dso_origin { + DSO__ORIG_KERNEL = 0, + DSO__ORIG_JAVA_JIT, + DSO__ORIG_BUILD_ID_CACHE, + DSO__ORIG_FEDORA, + DSO__ORIG_UBUNTU, + DSO__ORIG_BUILDID, + DSO__ORIG_DSO, + DSO__ORIG_KMODULE, + DSO__ORIG_NOT_FOUND, +}; + char dso__symtab_origin(const struct dso *self); void dso__set_long_name(struct dso *self, char *name); void dso__set_build_id(struct dso *self, void *build_id); @@ -168,4 +181,6 @@ int kallsyms__parse(const char *filename, void *arg, int symbol__init(void); bool symbol_type__is_a(char symbol_type, enum map_type map_type); +size_t vmlinux_path__fprintf(FILE *fp); + #endif /* __PERF_SYMBOL */ -- cgit v0.10.2 From d06d92b7c9b99ea52bdaeb13f544675529891b8a Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 15 Mar 2010 13:04:33 -0300 Subject: perf annotate: Properly notify the user that vmlinux is missing MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Before this patch we would not find a vmlinux, then try to pass objdump "[kernel.kallsyms]" as the filename, it would get confused and produce no output: [root@doppio ~]# perf annotate n_tty_write ------------------------------------------------ Percent | Source code & Disassembly of [kernel.kallsyms] ------------------------------------------------ Now we check that and emit meaningful warning: [root@doppio ~]# perf annotate n_tty_write Can't annotate n_tty_write: No vmlinux file was found in the path: [0] vmlinux [1] /boot/vmlinux [2] /boot/vmlinux-2.6.34-rc1-tip+ [3] /lib/modules/2.6.34-rc1-tip+/build/vmlinux [4] /usr/lib/debug/lib/modules/2.6.34-rc1-tip+/vmlinux [root@doppio ~]# This bug was introduced when we added automatic search for vmlinux, before that time the user had to specify a vmlinux file. v2: Print the warning just for the first symbol found when no symbol name is specified, otherwise it will spam the screen repeating the warning for each symbol. Reported-by: Ingo Molnar Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Cc: LKML-Reference: <1268669073-6856-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index 6ad7148..45d1466 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -452,6 +452,16 @@ static void annotate_sym(struct hist_entry *he) if (!filename) return; + if (dso->origin == DSO__ORIG_KERNEL) { + if (dso->annotate_warned) + return; + dso->annotate_warned = 1; + pr_err("Can't annotate %s: No vmlinux file was found in the " + "path:\n", sym->name); + vmlinux_path__fprintf(stderr); + return; + } + pr_debug("%s: filename=%s, sym=%s, start=%#Lx, end=%#Lx\n", __func__, filename, sym->name, map->unmap_ip(map, sym->start), map->unmap_ip(map, sym->end)); diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index 2836394..2d19e7a 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -63,7 +63,7 @@ static void hist_entry__annotate_browser(struct hist_entry *self) if (self->sym == NULL) return; - if (asprintf(&str, "perf annotate %s | expand", self->sym->name) < 0) + if (asprintf(&str, "perf annotate %s 2>&1 | expand", self->sym->name) < 0) return; fp = popen(str, "r"); diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h index d983bba..5bd91d1 100644 --- a/tools/perf/util/symbol.h +++ b/tools/perf/util/symbol.h @@ -106,6 +106,7 @@ struct dso { u8 has_build_id:1; u8 kernel:1; u8 hit:1; + u8 annotate_warned:1; unsigned char origin; u8 sorted_by_name; u8 loaded; -- cgit v0.10.2 From e88d556dc5f0ef437e3538277a1dd33e5038be77 Mon Sep 17 00:00:00 2001 From: Sergio Aguirre Date: Sat, 27 Feb 2010 14:13:43 -0600 Subject: OMAP3: serial: Check for zero-based physical addr This is for protecting a wrong mapping attempt of a zero-based physical address. The result is that, no serial port will be attempted to be mapped. Also add an additional protection for NULL clocks before attempting to enable them (if above condition applies) Signed-off-by: Sergio Aguirre diff --git a/arch/arm/mach-omap2/serial.c b/arch/arm/mach-omap2/serial.c index da77930..ef91fc0 100644 --- a/arch/arm/mach-omap2/serial.c +++ b/arch/arm/mach-omap2/serial.c @@ -664,6 +664,12 @@ void __init omap_serial_early_init(void) struct device *dev = &pdev->dev; struct plat_serial8250_port *p = dev->platform_data; + /* Don't map zero-based physical address */ + if (p->mapbase == 0) { + printk(KERN_WARNING "omap serial: No physical address" + " for uart#%d, so skipping early_init...\n", i); + continue; + } /* * Module 4KB + L4 interconnect 4KB * Static mapping, never released @@ -727,6 +733,13 @@ void __init omap_serial_init_port(int port) pdev = &uart->pdev; dev = &pdev->dev; + /* Don't proceed if there's no clocks available */ + if (unlikely(!uart->ick || !uart->fck)) { + WARN(1, "%s: can't init uart%d, no clocks available\n", + kobject_name(&dev->kobj), port); + return; + } + omap_uart_enable_clocks(uart); omap_uart_reset(uart); -- cgit v0.10.2 From 10c805eb4f89d44fe4e457d727b59af15c4a4a36 Mon Sep 17 00:00:00 2001 From: Sergio Aguirre Date: Tue, 9 Mar 2010 13:22:14 -0600 Subject: OMAP3: serial: Use dev_* macros instead of printk As we have a struct device populated at the time we are printing the errors, using dev_* macros makes more sense, as could give a better idea where the error/warning came from. Signed-off-by: Sergio Aguirre diff --git a/arch/arm/mach-omap2/serial.c b/arch/arm/mach-omap2/serial.c index ef91fc0..a55e6ae 100644 --- a/arch/arm/mach-omap2/serial.c +++ b/arch/arm/mach-omap2/serial.c @@ -550,7 +550,7 @@ static ssize_t sleep_timeout_store(struct device *dev, unsigned int value; if (sscanf(buf, "%u", &value) != 1) { - printk(KERN_ERR "sleep_timeout_store: Invalid value\n"); + dev_err(dev, "sleep_timeout_store: Invalid value\n"); return -EINVAL; } @@ -666,8 +666,8 @@ void __init omap_serial_early_init(void) /* Don't map zero-based physical address */ if (p->mapbase == 0) { - printk(KERN_WARNING "omap serial: No physical address" - " for uart#%d, so skipping early_init...\n", i); + dev_warn(dev, "no physical address for uart#%d," + " so skipping early_init...\n", i); continue; } /* @@ -676,21 +676,21 @@ void __init omap_serial_early_init(void) */ p->membase = ioremap(p->mapbase, SZ_8K); if (!p->membase) { - printk(KERN_ERR "ioremap failed for uart%i\n", i + 1); + dev_err(dev, "ioremap failed for uart%i\n", i + 1); continue; } sprintf(name, "uart%d_ick", i + 1); uart->ick = clk_get(NULL, name); if (IS_ERR(uart->ick)) { - printk(KERN_ERR "Could not get uart%d_ick\n", i + 1); + dev_err(dev, "Could not get uart%d_ick\n", i + 1); uart->ick = NULL; } sprintf(name, "uart%d_fck", i+1); uart->fck = clk_get(NULL, name); if (IS_ERR(uart->fck)) { - printk(KERN_ERR "Could not get uart%d_fck\n", i + 1); + dev_err(dev, "Could not get uart%d_fck\n", i + 1); uart->fck = NULL; } -- cgit v0.10.2 From 4b1bbd3fd9384ad6ca62903172ccb6a8729dd897 Mon Sep 17 00:00:00 2001 From: Sergio Aguirre Date: Fri, 26 Feb 2010 19:58:47 -0600 Subject: omap2/3/4: serial: Remove condition for getting uart4_phys This check is invalid, since we haven't filled the omap_revision var at this point. Signed-off-by: Sergio Aguirre diff --git a/arch/arm/mach-omap2/serial.c b/arch/arm/mach-omap2/serial.c index a55e6ae..3771254 100644 --- a/arch/arm/mach-omap2/serial.c +++ b/arch/arm/mach-omap2/serial.c @@ -115,7 +115,6 @@ static struct plat_serial8250_port serial_platform_data2[] = { } }; -#if defined(CONFIG_ARCH_OMAP3) || defined(CONFIG_ARCH_OMAP4) static struct plat_serial8250_port serial_platform_data3[] = { { .irq = 70, @@ -128,23 +127,12 @@ static struct plat_serial8250_port serial_platform_data3[] = { } }; -static inline void omap2_set_globals_uart4(struct omap_globals *omap2_globals) -{ - serial_platform_data3[0].mapbase = omap2_globals->uart4_phys; -} -#else -static inline void omap2_set_globals_uart4(struct omap_globals *omap2_globals) -{ -} -#endif - void __init omap2_set_globals_uart(struct omap_globals *omap2_globals) { serial_platform_data0[0].mapbase = omap2_globals->uart1_phys; serial_platform_data1[0].mapbase = omap2_globals->uart2_phys; serial_platform_data2[0].mapbase = omap2_globals->uart3_phys; - if (cpu_is_omap3630() || cpu_is_omap44xx()) - omap2_set_globals_uart4(omap2_globals); + serial_platform_data3[0].mapbase = omap2_globals->uart4_phys; } static inline unsigned int __serial_read_reg(struct uart_port *up, -- cgit v0.10.2 From 2d7f66104cc9ef508c382c8d6cb1ce6123383350 Mon Sep 17 00:00:00 2001 From: Sergio Aguirre Date: Mon, 1 Mar 2010 11:48:26 -0600 Subject: omap3: zoom2/3 / 3630sdp: Don't init always all uarts This is useless, since in Zoom2/3 boards, the ports aren't even physically accessible. They must be explicitly initted in the board-zoom2.c, board-zoom3.c and board-3630sdp.c files instead. Signed-off-by: Sergio Aguirre diff --git a/arch/arm/mach-omap2/board-zoom-peripherals.c b/arch/arm/mach-omap2/board-zoom-peripherals.c index ca95d8d..6b39849 100644 --- a/arch/arm/mach-omap2/board-zoom-peripherals.c +++ b/arch/arm/mach-omap2/board-zoom-peripherals.c @@ -280,7 +280,6 @@ static void enable_board_wakeup_source(void) void __init zoom_peripherals_init(void) { omap_i2c_init(); - omap_serial_init(); usb_musb_init(&musb_board_data); enable_board_wakeup_source(); } -- cgit v0.10.2 From 9e265aa7ac493aa0a3de80aa3092d8677b3e8546 Mon Sep 17 00:00:00 2001 From: Sergio Aguirre Date: Mon, 1 Mar 2010 11:52:36 -0600 Subject: omap3: 3630sdp: Explicitly enable all UARTs All UARTs seem physically reachable, so, enable them all. Signed-off-by: Sergio Aguirre diff --git a/arch/arm/mach-omap2/board-3630sdp.c b/arch/arm/mach-omap2/board-3630sdp.c index a0a2a11..504d2bd 100644 --- a/arch/arm/mach-omap2/board-3630sdp.c +++ b/arch/arm/mach-omap2/board-3630sdp.c @@ -96,6 +96,7 @@ static struct omap_board_mux board_mux[] __initdata = { static void __init omap_sdp_init(void) { omap3_mux_init(board_mux, OMAP_PACKAGE_CBP); + omap_serial_init(); zoom_peripherals_init(); board_smc91x_init(); enable_board_wakeup_source(); -- cgit v0.10.2 From fcbcea93abbe361e14dd458d0a5a3e9dadaa5ba9 Mon Sep 17 00:00:00 2001 From: Sergio Aguirre Date: Mon, 1 Mar 2010 12:17:15 -0600 Subject: omap3: zoom 2/3: Change debugboard serial port id This is now changed to PLAT8250_DEV_PLATFORM (= 0), because it's the only port that's going to be initialized in Zoom 2/3 boards. So, it doesn't make sense to keep the hardcoded 3 value anymore. Signed-off-by: Sergio Aguirre diff --git a/arch/arm/mach-omap2/board-zoom-debugboard.c b/arch/arm/mach-omap2/board-zoom-debugboard.c index bb4018b..e15d2e8 100644 --- a/arch/arm/mach-omap2/board-zoom-debugboard.c +++ b/arch/arm/mach-omap2/board-zoom-debugboard.c @@ -96,7 +96,7 @@ static struct plat_serial8250_port serial_platform_data[] = { static struct platform_device zoom_debugboard_serial_device = { .name = "serial8250", - .id = 3, + .id = PLAT8250_DEV_PLATFORM, .dev = { .platform_data = serial_platform_data, }, -- cgit v0.10.2 From e6e8ea63f5d745e5a3030eec1964d0429b93107d Mon Sep 17 00:00:00 2001 From: Sergio Aguirre Date: Mon, 1 Mar 2010 14:26:45 -0600 Subject: omap3: zoom2/3: Register only 1 8250 port There's no more serial ports available, so, doesn't make sense to create 4 device nodes. Signed-off-by: Sergio Aguirre diff --git a/arch/arm/configs/omap_zoom2_defconfig b/arch/arm/configs/omap_zoom2_defconfig index f5c6e11..881faea 100644 --- a/arch/arm/configs/omap_zoom2_defconfig +++ b/arch/arm/configs/omap_zoom2_defconfig @@ -661,7 +661,7 @@ CONFIG_DEVKMEM=y CONFIG_SERIAL_8250=y CONFIG_SERIAL_8250_CONSOLE=y CONFIG_SERIAL_8250_NR_UARTS=32 -CONFIG_SERIAL_8250_RUNTIME_UARTS=4 +CONFIG_SERIAL_8250_RUNTIME_UARTS=1 CONFIG_SERIAL_8250_EXTENDED=y CONFIG_SERIAL_8250_MANY_PORTS=y CONFIG_SERIAL_8250_SHARE_IRQ=y diff --git a/arch/arm/configs/omap_zoom3_defconfig b/arch/arm/configs/omap_zoom3_defconfig index ea9a501..5e55b55 100644 --- a/arch/arm/configs/omap_zoom3_defconfig +++ b/arch/arm/configs/omap_zoom3_defconfig @@ -680,7 +680,7 @@ CONFIG_DEVKMEM=y CONFIG_SERIAL_8250=y CONFIG_SERIAL_8250_CONSOLE=y CONFIG_SERIAL_8250_NR_UARTS=32 -CONFIG_SERIAL_8250_RUNTIME_UARTS=4 +CONFIG_SERIAL_8250_RUNTIME_UARTS=1 CONFIG_SERIAL_8250_EXTENDED=y CONFIG_SERIAL_8250_MANY_PORTS=y CONFIG_SERIAL_8250_SHARE_IRQ=y -- cgit v0.10.2 From 97fedbbe1046b3118f49df249840ca21041eefe4 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Tue, 16 Mar 2010 08:55:32 +0100 Subject: Remove GENHD_FL_DRIVERFS This flag is not used, so best discarded. Signed-off-by: NeilBrown -- Hi Jens, I came across this recently - these are the only two occurances of "GENHD_FL_DRIVERFS" in the kernel, so it cannot be needed. NeilBrown Signed-off-by: Jens Axboe diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c index 1dd4d84..466fae8 100644 --- a/drivers/scsi/sd.c +++ b/drivers/scsi/sd.c @@ -2185,7 +2185,7 @@ static void sd_probe_async(void *data, async_cookie_t cookie) blk_queue_prep_rq(sdp->request_queue, sd_prep_fn); gd->driverfs_dev = &sdp->sdev_gendev; - gd->flags = GENHD_FL_EXT_DEVT | GENHD_FL_DRIVERFS; + gd->flags = GENHD_FL_EXT_DEVT; if (sdp->removable) gd->flags |= GENHD_FL_REMOVABLE; diff --git a/include/linux/genhd.h b/include/linux/genhd.h index 56b5051..5f2f4c4 100644 --- a/include/linux/genhd.h +++ b/include/linux/genhd.h @@ -109,7 +109,7 @@ struct hd_struct { }; #define GENHD_FL_REMOVABLE 1 -#define GENHD_FL_DRIVERFS 2 +/* 2 is unused */ #define GENHD_FL_MEDIA_CHANGE_NOTIFY 4 #define GENHD_FL_CD 8 #define GENHD_FL_UP 16 -- cgit v0.10.2 From 910ac735bad53ce54741a72a5b19ab69794ae069 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Tue, 16 Mar 2010 08:57:15 +0100 Subject: block: make CONFIG_BLK_CGROUP visible Make the config visible, so we can choose from CONFIG_BLK_CGROUP=y and CONFIG_BLK_CGROUP=m when CONFIG_IOSCHED_CFQ=m. Signed-off-by: Li Zefan Signed-off-by: Jens Axboe diff --git a/block/Kconfig b/block/Kconfig index e20fbde..f9e89f4 100644 --- a/block/Kconfig +++ b/block/Kconfig @@ -78,8 +78,9 @@ config BLK_DEV_INTEGRITY Protection. If in doubt, say N. config BLK_CGROUP - bool + tristate "Block cgroup support" depends on CGROUPS + depends on CFQ_GROUP_IOSCHED default n ---help--- Generic block IO controller cgroup interface. This is the common -- cgit v0.10.2 From 1d199b1ad606ae8b88acebd295b101c4e1cf2a57 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Tue, 16 Mar 2010 01:05:02 +0100 Subject: perf: Fix unexported generic perf_arch_fetch_caller_regs perf_arch_fetch_caller_regs() is exported for the overriden x86 version, but not for the generic weak version. As a general rule, weak functions should not have their symbol exported in the same file they are defined. So let's export it on trace_event_perf.c as it is used by trace events only. This fixes: ERROR: ".perf_arch_fetch_caller_regs" [fs/xfs/xfs.ko] undefined! ERROR: ".perf_arch_fetch_caller_regs" [arch/powerpc/platforms/cell/spufs/spufs.ko] undefined! -v2: And also only build it if trace events are enabled. -v3: Fix changelog mistake Reported-by: Stephen Rothwell Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Xiao Guangrong Cc: Paul Mackerras LKML-Reference: <1268697902-9518-1-git-send-regression-fweisbec@gmail.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 978d297..0d3466c 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -1695,6 +1695,7 @@ struct perf_callchain_entry *perf_callchain(struct pt_regs *regs) return entry; } +#ifdef CONFIG_EVENT_TRACING void perf_arch_fetch_caller_regs(struct pt_regs *regs, unsigned long ip, int skip) { regs->ip = ip; @@ -1706,4 +1707,4 @@ void perf_arch_fetch_caller_regs(struct pt_regs *regs, unsigned long ip, int ski regs->cs = __KERNEL_CS; local_save_flags(regs->flags); } -EXPORT_SYMBOL_GPL(perf_arch_fetch_caller_regs); +#endif diff --git a/kernel/perf_event.c b/kernel/perf_event.c index 8bf6127..455393e 100644 --- a/kernel/perf_event.c +++ b/kernel/perf_event.c @@ -2790,10 +2790,12 @@ __weak struct perf_callchain_entry *perf_callchain(struct pt_regs *regs) return NULL; } +#ifdef CONFIG_EVENT_TRACING __weak void perf_arch_fetch_caller_regs(struct pt_regs *regs, unsigned long ip, int skip) { } +#endif /* * Output diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 0709e4f..7d79a10 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -12,6 +12,8 @@ DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); EXPORT_PER_CPU_SYMBOL_GPL(perf_trace_regs); +EXPORT_SYMBOL_GPL(perf_arch_fetch_caller_regs); + static char *perf_trace_buf; static char *perf_trace_buf_nmi; -- cgit v0.10.2 From 8ea7f544100844307072cae2f5fc108afdef999a Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Tue, 16 Mar 2010 10:12:36 +0800 Subject: x86, perf: Fix comments in Pentium-4 PMU definitions Reported-by: Cyrill Gorcunov Signed-off-by: Lin Ming Cc: Peter Zijlstra Cc: Mike Galbraith Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker LKML-Reference: <1268705556.3379.8.camel@minggr.sh.intel.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/perf_event_p4.h b/arch/x86/include/asm/perf_event_p4.h index b47b9e9..b842b32 100644 --- a/arch/x86/include/asm/perf_event_p4.h +++ b/arch/x86/include/asm/perf_event_p4.h @@ -319,6 +319,7 @@ static inline u32 p4_default_escr_conf(int cpu, int exclude_os, int exclude_usr) #define P4_BSQ_ACTIVE_ENTRIES P4_EVENT_PACK(0x06, 0x07) /* + * NOTE: no ESCR name in docs, it's guessed * MSR_P4_BSU_ESCR1: 2, 3 */ @@ -468,8 +469,8 @@ static inline u32 p4_default_escr_conf(int cpu, int exclude_os, int exclude_usr) #define P4_UOPS_RETIRED P4_EVENT_PACK(0x01, 0x04) /* - * MSR_P4_CRU_ESCR2: 12, 13, 16 - * MSR_P4_CRU_ESCR3: 14, 15, 17 + * MSR_P4_CRU_ESCR0: 12, 13, 16 + * MSR_P4_CRU_ESCR1: 14, 15, 17 */ #define P4_UOP_TYPE P4_EVENT_PACK(0x02, 0x02) -- cgit v0.10.2 From 6427462bfa50f50dc6c088c07037264fcc73eca1 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Mon, 15 Mar 2010 11:21:48 +0300 Subject: sched: Remove some dead code This was left over from "7c9414385e sched: Remove USER_SCHED" Signed-off-by: Dan Carpenter Acked-by: Dhaval Giani Cc: Kay Sievers Cc: Greg Kroah-Hartman LKML-Reference: <20100315082148.GD18181@bicker> Signed-off-by: Ingo Molnar diff --git a/kernel/user.c b/kernel/user.c index 766467b..ec3b222 100644 --- a/kernel/user.c +++ b/kernel/user.c @@ -178,8 +178,6 @@ struct user_struct *alloc_uid(struct user_namespace *ns, uid_t uid) return up; - put_user_ns(new->user_ns); - kmem_cache_free(uid_cachep, new); out_unlock: return NULL; } -- cgit v0.10.2 From 5cc718b9dad682329a60e73547c6e708faa5bbe4 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Mon, 15 Mar 2010 13:00:54 -0400 Subject: kprobes: Hide CONFIG_OPTPROBES and set if arch supports optimized kprobes Hide CONFIG_OPTPROBES and set if the arch supports optimized kprobes (IOW, HAVE_OPTPROBES=y), since this option doesn't change the major behavior of kprobes, and workarounds for minor changes are documented. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Dieter Ries Cc: Ananth N Mavinakayanahalli Cc: OGAWA Hirofumi Cc: Peter Zijlstra Cc: Mike Galbraith Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker LKML-Reference: <20100315170054.31593.3153.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/Documentation/kprobes.txt b/Documentation/kprobes.txt index 2f9115c..61c291c 100644 --- a/Documentation/kprobes.txt +++ b/Documentation/kprobes.txt @@ -165,8 +165,8 @@ the user entry_handler invocation is also skipped. 1.4 How Does Jump Optimization Work? -If you configured your kernel with CONFIG_OPTPROBES=y (currently -this option is supported on x86/x86-64, non-preemptive kernel) and +If your kernel is built with CONFIG_OPTPROBES=y (currently this flag +is automatically set 'y' on x86/x86-64, non-preemptive kernel) and the "debug.kprobes_optimization" kernel parameter is set to 1 (see sysctl(8)), Kprobes tries to reduce probe-hit overhead by using a jump instruction instead of a breakpoint instruction at each probepoint. @@ -271,8 +271,6 @@ tweak the kernel's execution path, you need to suppress optimization, using one of the following techniques: - Specify an empty function for the kprobe's post_handler or break_handler. or -- Config CONFIG_OPTPROBES=n. - or - Execute 'sysctl -w debug.kprobes_optimization=n' 2. Architectures Supported @@ -307,10 +305,6 @@ it useful to "Compile the kernel with debug info" (CONFIG_DEBUG_INFO), so you can use "objdump -d -l vmlinux" to see the source-to-object code mapping. -If you want to reduce probing overhead, set "Kprobes jump optimization -support" (CONFIG_OPTPROBES) to "y". You can find this option under the -"Kprobes" line. - 4. API Reference The Kprobes API includes a "register" function and an "unregister" diff --git a/arch/Kconfig b/arch/Kconfig index e5eb133..f06010f 100644 --- a/arch/Kconfig +++ b/arch/Kconfig @@ -42,15 +42,10 @@ config KPROBES If in doubt, say "N". config OPTPROBES - bool "Kprobes jump optimization support (EXPERIMENTAL)" - default y - depends on KPROBES + def_bool y + depends on KPROBES && HAVE_OPTPROBES depends on !PREEMPT - depends on HAVE_OPTPROBES select KALLSYMS_ALL - help - This option will allow kprobes to optimize breakpoint to - a jump for reducing its overhead. config HAVE_EFFICIENT_UNALIGNED_ACCESS bool -- cgit v0.10.2 From 61f8603d93fa0b0e2f73be7a4f048696417a24a3 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Wed, 3 Feb 2010 17:31:31 +1100 Subject: nfsd: factor out hash functions for export caches. Both the _lookup and the _update functions for these two caches independently calculate the hash of the key. So factor out that code for improved reuse. Signed-off-by: NeilBrown Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/export.c b/fs/nfsd/export.c index a0c4016..65ddc5b 100644 --- a/fs/nfsd/export.c +++ b/fs/nfsd/export.c @@ -258,10 +258,9 @@ static struct cache_detail svc_expkey_cache = { .alloc = expkey_alloc, }; -static struct svc_expkey * -svc_expkey_lookup(struct svc_expkey *item) +static int +svc_expkey_hash(struct svc_expkey *item) { - struct cache_head *ch; int hash = item->ek_fsidtype; char * cp = (char*)item->ek_fsid; int len = key_len(item->ek_fsidtype); @@ -269,6 +268,14 @@ svc_expkey_lookup(struct svc_expkey *item) hash ^= hash_mem(cp, len, EXPKEY_HASHBITS); hash ^= hash_ptr(item->ek_client, EXPKEY_HASHBITS); hash &= EXPKEY_HASHMASK; + return hash; +} + +static struct svc_expkey * +svc_expkey_lookup(struct svc_expkey *item) +{ + struct cache_head *ch; + int hash = svc_expkey_hash(item); ch = sunrpc_cache_lookup(&svc_expkey_cache, &item->h, hash); @@ -282,13 +289,7 @@ static struct svc_expkey * svc_expkey_update(struct svc_expkey *new, struct svc_expkey *old) { struct cache_head *ch; - int hash = new->ek_fsidtype; - char * cp = (char*)new->ek_fsid; - int len = key_len(new->ek_fsidtype); - - hash ^= hash_mem(cp, len, EXPKEY_HASHBITS); - hash ^= hash_ptr(new->ek_client, EXPKEY_HASHBITS); - hash &= EXPKEY_HASHMASK; + int hash = svc_expkey_hash(new); ch = sunrpc_cache_update(&svc_expkey_cache, &new->h, &old->h, hash); @@ -737,14 +738,22 @@ struct cache_detail svc_export_cache = { .alloc = svc_export_alloc, }; -static struct svc_export * -svc_export_lookup(struct svc_export *exp) +static int +svc_export_hash(struct svc_export *exp) { - struct cache_head *ch; int hash; + hash = hash_ptr(exp->ex_client, EXPORT_HASHBITS); hash ^= hash_ptr(exp->ex_path.dentry, EXPORT_HASHBITS); hash ^= hash_ptr(exp->ex_path.mnt, EXPORT_HASHBITS); + return hash; +} + +static struct svc_export * +svc_export_lookup(struct svc_export *exp) +{ + struct cache_head *ch; + int hash = svc_export_hash(exp); ch = sunrpc_cache_lookup(&svc_export_cache, &exp->h, hash); @@ -758,10 +767,7 @@ static struct svc_export * svc_export_update(struct svc_export *new, struct svc_export *old) { struct cache_head *ch; - int hash; - hash = hash_ptr(old->ex_client, EXPORT_HASHBITS); - hash ^= hash_ptr(old->ex_path.dentry, EXPORT_HASHBITS); - hash ^= hash_ptr(old->ex_path.mnt, EXPORT_HASHBITS); + int hash = svc_export_hash(old); ch = sunrpc_cache_update(&svc_export_cache, &new->h, &old->h, -- cgit v0.10.2 From 984763cb90d4b5444baa0c3e43feff7926bf1834 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Tue, 16 Mar 2010 17:07:33 +0100 Subject: perf, x86: Report error code that returned from x86_pmu.hw_config() If x86_pmu.hw_config() fails a fixed error code (-EOPNOTSUPP) is returned even if a different error was reported. This patch fixes this. Signed-off-by: Robert Richter Acked-by: Cyrill Gorcunov Acked-by: Lin Ming Cc: acme@redhat.com Cc: eranian@google.com Cc: gorcunov@openvz.org Cc: peterz@infradead.org Cc: fweisbec@gmail.com LKML-Reference: <20100316160733.GR1585@erda.amd.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 0d3466c..5dacf63 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -472,8 +472,9 @@ static int __hw_perf_event_init(struct perf_event *event) hwc->last_tag = ~0ULL; /* Processor specifics */ - if (x86_pmu.hw_config(attr, hwc)) - return -EOPNOTSUPP; + err = x86_pmu.hw_config(attr, hwc); + if (err) + return err; if (!hwc->sample_period) { hwc->sample_period = x86_pmu.max_period; -- cgit v0.10.2 From a6b84574eed7e4fd8cb8dac2d0926fe2cf34b941 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Tue, 16 Mar 2010 01:05:02 +0100 Subject: perf: Fix unexported generic perf_arch_fetch_caller_regs perf_arch_fetch_caller_regs() is exported for the overriden x86 version, but not for the generic weak version. As a general rule, weak functions should not have their symbol exported in the same file they are defined. So let's export it on trace_event_perf.c as it is used by trace events only. This fixes: ERROR: ".perf_arch_fetch_caller_regs" [fs/xfs/xfs.ko] undefined! ERROR: ".perf_arch_fetch_caller_regs" [arch/powerpc/platforms/cell/spufs/spufs.ko] undefined! -v2: And also only build it if trace events are enabled. -v3: Fix changelog mistake Reported-by: Stephen Rothwell Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Xiao Guangrong Cc: Paul Mackerras LKML-Reference: <1268697902-9518-1-git-send-regression-fweisbec@gmail.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 7645fae..60398a0 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -1702,6 +1702,7 @@ struct perf_callchain_entry *perf_callchain(struct pt_regs *regs) return entry; } +#ifdef CONFIG_EVENT_TRACING void perf_arch_fetch_caller_regs(struct pt_regs *regs, unsigned long ip, int skip) { regs->ip = ip; @@ -1713,4 +1714,4 @@ void perf_arch_fetch_caller_regs(struct pt_regs *regs, unsigned long ip, int ski regs->cs = __KERNEL_CS; local_save_flags(regs->flags); } -EXPORT_SYMBOL_GPL(perf_arch_fetch_caller_regs); +#endif diff --git a/kernel/perf_event.c b/kernel/perf_event.c index fb3031c..574ee58 100644 --- a/kernel/perf_event.c +++ b/kernel/perf_event.c @@ -2786,10 +2786,12 @@ __weak struct perf_callchain_entry *perf_callchain(struct pt_regs *regs) return NULL; } +#ifdef CONFIG_EVENT_TRACING __weak void perf_arch_fetch_caller_regs(struct pt_regs *regs, unsigned long ip, int skip) { } +#endif /* * Output diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 0709e4f..7d79a10 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -12,6 +12,8 @@ DEFINE_PER_CPU(struct pt_regs, perf_trace_regs); EXPORT_PER_CPU_SYMBOL_GPL(perf_trace_regs); +EXPORT_SYMBOL_GPL(perf_arch_fetch_caller_regs); + static char *perf_trace_buf; static char *perf_trace_buf_nmi; -- cgit v0.10.2 From a1d37d5285bcda07f9c0b80a2634ca20ab545297 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 16 Mar 2010 18:05:21 -0400 Subject: perf tools: Introduce xzalloc() for detecting out of memory conditions Introducing xzalloc() which wrapping zalloc() for detecting out of memory conditions. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <20100316220521.32050.85155.stgit@localhost6.localdomain6> [ -v2: small cleanups in surrounding code ] Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index 0f5b2a6..5270108 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -295,6 +295,13 @@ extern void *xmemdupz(const void *data, size_t len); extern char *xstrndup(const char *str, size_t len); extern void *xrealloc(void *ptr, size_t size) __attribute__((weak)); +static inline void *xzalloc(size_t size) +{ + void *buf = xmalloc(size); + + return memset(buf, 0, size); +} + static inline void *zalloc(size_t size) { return calloc(1, size); @@ -309,6 +316,7 @@ static inline int has_extension(const char *filename, const char *ext) { size_t len = strlen(filename); size_t extlen = strlen(ext); + return len > extlen && !memcmp(filename + len - extlen, ext, extlen); } @@ -322,6 +330,7 @@ static inline int has_extension(const char *filename, const char *ext) #undef isalnum #undef tolower #undef toupper + extern unsigned char sane_ctype[256]; #define GIT_SPACE 0x01 #define GIT_DIGIT 0x02 -- cgit v0.10.2 From 31facc5f1ac674fbcc29f212377e589396bb934c Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 16 Mar 2010 18:05:30 -0400 Subject: perf probe: Use wrapper functions Use wrapped functions as much as possible, to check out of memory conditions in perf probe. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <20100316220530.32050.53951.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c index 152d6c9..b6afe7b 100644 --- a/tools/perf/builtin-probe.c +++ b/tools/perf/builtin-probe.c @@ -87,9 +87,7 @@ static void parse_probe_event_argv(int argc, const char **argv) len = 0; for (i = 0; i < argc; i++) len += strlen(argv[i]) + 1; - buf = zalloc(len + 1); - if (!buf) - die("Failed to allocate memory for binding arguments."); + buf = xzalloc(len + 1); len = 0; for (i = 0; i < argc; i++) len += sprintf(&buf[len], "%s ", argv[i]); diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index 7c004b6..88a3b6d 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -33,6 +33,7 @@ #include #undef _GNU_SOURCE +#include "util.h" #include "event.h" #include "string.h" #include "strlist.h" @@ -90,9 +91,9 @@ void parse_line_range_desc(const char *arg, struct line_range *lr) if (*tmp != '\0') semantic_error("Tailing with invalid character '%d'.", *tmp); - tmp = strndup(arg, (ptr - arg)); + tmp = xstrndup(arg, (ptr - arg)); } else - tmp = strdup(arg); + tmp = xstrdup(arg); if (strchr(tmp, '.')) lr->file = tmp; @@ -135,7 +136,7 @@ static void parse_perf_probe_probepoint(char *arg, struct probe_point *pp) if (!check_event_name(arg)) semantic_error("%s is bad for event name -it must " "follow C symbol-naming rule.", arg); - pp->event = strdup(arg); + pp->event = xstrdup(arg); arg = tmp; } @@ -147,17 +148,16 @@ static void parse_perf_probe_probepoint(char *arg, struct probe_point *pp) /* Check arg is function or file and copy it */ if (strchr(arg, '.')) /* File */ - pp->file = strdup(arg); + pp->file = xstrdup(arg); else /* Function */ - pp->function = strdup(arg); - DIE_IF(pp->file == NULL && pp->function == NULL); + pp->function = xstrdup(arg); /* Parse other options */ while (ptr) { arg = ptr; c = nc; if (c == ';') { /* Lazy pattern must be the last part */ - pp->lazy_line = strdup(arg); + pp->lazy_line = xstrdup(arg); break; } ptr = strpbrk(arg, ";:+@%"); @@ -181,8 +181,7 @@ static void parse_perf_probe_probepoint(char *arg, struct probe_point *pp) case '@': /* File name */ if (pp->file) semantic_error("SRC@SRC is not allowed."); - pp->file = strdup(arg); - DIE_IF(pp->file == NULL); + pp->file = xstrdup(arg); break; case '%': /* Probe places */ if (strcmp(arg, "return") == 0) { @@ -247,11 +246,9 @@ void parse_perf_probe_event(const char *str, struct probe_point *pp, /* Copy arguments and ensure return probe has no C argument */ pp->nr_args = argc - 1; - pp->args = zalloc(sizeof(char *) * pp->nr_args); + pp->args = xzalloc(sizeof(char *) * pp->nr_args); for (i = 0; i < pp->nr_args; i++) { - pp->args[i] = strdup(argv[i + 1]); - if (!pp->args[i]) - die("Failed to copy argument."); + pp->args[i] = xstrdup(argv[i + 1]); if (is_c_varname(pp->args[i])) { if (pp->retprobe) semantic_error("You can't specify local" @@ -299,14 +296,12 @@ void parse_trace_kprobe_event(const char *str, struct probe_point *pp) pp->file = NULL; pp->nr_args = argc - 2; - pp->args = zalloc(sizeof(char *) * pp->nr_args); + pp->args = xzalloc(sizeof(char *) * pp->nr_args); for (i = 0; i < pp->nr_args; i++) { p = strchr(argv[i + 2], '='); if (p) /* We don't need which register is assigned. */ *p = '\0'; - pp->args[i] = strdup(argv[i + 2]); - if (!pp->args[i]) - die("Failed to copy argument."); + pp->args[i] = xstrdup(argv[i + 2]); } argv_free(argv); @@ -319,10 +314,8 @@ int synthesize_perf_probe_point(struct probe_point *pp) char offs[64] = "", line[64] = ""; int ret; - pp->probes[0] = buf = zalloc(MAX_CMDLEN); + pp->probes[0] = buf = xzalloc(MAX_CMDLEN); pp->found = 1; - if (!buf) - die("Failed to allocate memory by zalloc."); if (pp->offset) { ret = e_snprintf(offs, 64, "+%d", pp->offset); if (ret <= 0) @@ -380,9 +373,7 @@ int synthesize_trace_kprobe_event(struct probe_point *pp) char *buf; int i, len, ret; - pp->probes[0] = buf = zalloc(MAX_CMDLEN); - if (!buf) - die("Failed to allocate memory by zalloc."); + pp->probes[0] = buf = xzalloc(MAX_CMDLEN); ret = e_snprintf(buf, MAX_CMDLEN, "%s+%d", pp->function, pp->offset); if (ret <= 0) goto error; @@ -612,10 +603,9 @@ void add_trace_kprobe_events(struct probe_point *probes, int nr_probes, for (j = 0; j < nr_probes; j++) { pp = probes + j; if (!pp->event) - pp->event = strdup(pp->function); + pp->event = xstrdup(pp->function); if (!pp->group) - pp->group = strdup(PERFPROBE_GROUP); - DIE_IF(!pp->event || !pp->group); + pp->group = xstrdup(PERFPROBE_GROUP); /* If force_add is true, suffix search is allowed */ allow_suffix = force_add; for (i = 0; i < pp->found; i++) { @@ -709,9 +699,7 @@ void del_trace_kprobe_events(struct strlist *dellist) namelist = get_perf_event_names(fd, true); strlist__for_each(ent, dellist) { - str = strdup(ent->s); - if (!str) - die("Failed to copy event."); + str = xstrdup(ent->s); pr_debug("Parsing: %s\n", str); p = strchr(str, ':'); if (p) { diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index c171a24..e887bb6 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -125,8 +125,7 @@ static void line_list__add_line(struct list_head *head, unsigned int line) p = head; found: pr_debug("line list: add a line %u\n", line); - ln = zalloc(sizeof(struct line_node)); - DIE_IF(ln == NULL); + ln = xzalloc(sizeof(struct line_node)); ln->line = line; INIT_LIST_HEAD(&ln->list); list_add(&ln->list, p); @@ -416,7 +415,7 @@ static void show_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf) (unsigned long)(pf->addr - eaddr)); /* Copy the function name if possible */ if (!pp->function) { - pp->function = strdup(name); + pp->function = xstrdup(name); pp->offset = (size_t)(pf->addr - eaddr); } } else { @@ -425,7 +424,7 @@ static void show_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf) (uintmax_t)pf->addr); if (!pp->function) { /* TODO: Use _stext */ - pp->function = strdup(""); + pp->function = xstrdup(""); pp->offset = (size_t)pf->addr; } } @@ -456,7 +455,7 @@ static void show_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf) if (pp->found == MAX_PROBES) die("Too many( > %d) probe point found.\n", MAX_PROBES); - pp->probes[pp->found] = strdup(tmp); + pp->probes[pp->found] = xstrdup(tmp); pp->found++; } @@ -506,8 +505,7 @@ static int find_lazy_match_lines(struct list_head *head, if (fd < 0) die("failed to open %s", fname); DIE_IF(fstat(fd, &st) < 0); - fbuf = malloc(st.st_size + 2); - DIE_IF(fbuf == NULL); + fbuf = xmalloc(st.st_size + 2); DIE_IF(read(fd, fbuf, st.st_size) < 0); close(fd); fbuf[st.st_size] = '\n'; /* Dummy line */ @@ -727,7 +725,7 @@ static void find_line_range_by_line(Dwarf_Die *sp_die, struct line_finder *lf) /* Copy real path */ if (!lf->lr->path) - lf->lr->path = strdup(src); + lf->lr->path = xstrdup(src); line_list__add_line(&lf->lr->line_list, (unsigned int)lineno); } /* Update status */ -- cgit v0.10.2 From e0faa8d35845bb1893cf9e608a5a5d92e9390bf0 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 16 Mar 2010 18:05:37 -0400 Subject: perf probe: Move add-probe routine to util/ Move add-probe routine to util/probe_event.c. This simplifies main routine for reducing maintenance cost. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <20100316220537.32050.72214.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c index b6afe7b..2087034 100644 --- a/tools/perf/builtin-probe.c +++ b/tools/perf/builtin-probe.c @@ -36,11 +36,9 @@ #include "builtin.h" #include "util/util.h" #include "util/strlist.h" -#include "util/event.h" +#include "util/symbol.h" #include "util/debug.h" #include "util/debugfs.h" -#include "util/symbol.h" -#include "util/thread.h" #include "util/parse-options.h" #include "util/parse-events.h" /* For debugfs_path */ #include "util/probe-finder.h" @@ -57,8 +55,6 @@ static struct { int nr_probe; struct probe_point probes[MAX_PROBES]; struct strlist *dellist; - struct map_groups kmap_groups; - struct map *kmaps[MAP__NR_TYPES]; struct line_range line_range; } session; @@ -114,29 +110,7 @@ static int opt_del_probe_event(const struct option *opt __used, return 0; } -/* Currently just checking function name from symbol map */ -static void evaluate_probe_point(struct probe_point *pp) -{ - struct symbol *sym; - sym = map__find_symbol_by_name(session.kmaps[MAP__FUNCTION], - pp->function, NULL); - if (!sym) - die("Kernel symbol \'%s\' not found - probe not added.", - pp->function); -} - #ifndef NO_DWARF_SUPPORT -static int open_vmlinux(void) -{ - if (map__load(session.kmaps[MAP__FUNCTION], NULL) < 0) { - pr_debug("Failed to load kernel map.\n"); - return -EINVAL; - } - pr_debug("Try to open %s\n", - session.kmaps[MAP__FUNCTION]->dso->long_name); - return open(session.kmaps[MAP__FUNCTION]->dso->long_name, O_RDONLY); -} - static int opt_show_lines(const struct option *opt __used, const char *str, int unset __used) { @@ -204,31 +178,8 @@ static const struct option options[] = { OPT_END() }; -/* Initialize symbol maps for vmlinux */ -static void init_vmlinux(void) -{ - symbol_conf.sort_by_name = true; - if (symbol_conf.vmlinux_name == NULL) - symbol_conf.try_vmlinux_path = true; - else - pr_debug("Use vmlinux: %s\n", symbol_conf.vmlinux_name); - if (symbol__init() < 0) - die("Failed to init symbol map."); - - map_groups__init(&session.kmap_groups); - if (map_groups__create_kernel_maps(&session.kmap_groups, - session.kmaps) < 0) - die("Failed to create kernel maps."); -} - int cmd_probe(int argc, const char **argv, const char *prefix __used) { - int i, ret; -#ifndef NO_DWARF_SUPPORT - int fd; -#endif - struct probe_point *pp; - argc = parse_options(argc, argv, options, probe_usage, PARSE_OPT_STOP_AT_NON_OPTION); if (argc > 0) { @@ -267,14 +218,7 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) " --add/--del.\n"); usage_with_options(probe_usage, options); } - init_vmlinux(); - fd = open_vmlinux(); - if (fd < 0) - die("Could not open debuginfo file."); - ret = find_line_range(fd, &session.line_range); - if (ret <= 0) - die("Source line is not found.\n"); - close(fd); + show_line_range(&session.line_range); return 0; } @@ -287,72 +231,8 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) return 0; } - /* Add probes */ - init_vmlinux(); - - if (session.need_dwarf) -#ifdef NO_DWARF_SUPPORT - die("Debuginfo-analysis is not supported"); -#else /* !NO_DWARF_SUPPORT */ - pr_debug("Some probes require debuginfo.\n"); - - fd = open_vmlinux(); - if (fd < 0) { - if (session.need_dwarf) - die("Could not open debuginfo file."); - - pr_debug("Could not open vmlinux/module file." - " Try to use symbols.\n"); - goto end_dwarf; - } - - /* Searching probe points */ - for (i = 0; i < session.nr_probe; i++) { - pp = &session.probes[i]; - if (pp->found) - continue; - - lseek(fd, SEEK_SET, 0); - ret = find_probe_point(fd, pp); - if (ret > 0) - continue; - if (ret == 0) { /* No error but failed to find probe point. */ - synthesize_perf_probe_point(pp); - die("Probe point '%s' not found. - probe not added.", - pp->probes[0]); - } - /* Error path */ - if (session.need_dwarf) { - if (ret == -ENOENT) - pr_warning("No dwarf info found in the vmlinux - please rebuild with CONFIG_DEBUG_INFO=y.\n"); - die("Could not analyze debuginfo."); - } - pr_debug("An error occurred in debuginfo analysis." - " Try to use symbols.\n"); - break; - } - close(fd); - -end_dwarf: -#endif /* !NO_DWARF_SUPPORT */ - - /* Synthesize probes without dwarf */ - for (i = 0; i < session.nr_probe; i++) { - pp = &session.probes[i]; - if (pp->found) /* This probe is already found. */ - continue; - - evaluate_probe_point(pp); - ret = synthesize_trace_kprobe_event(pp); - if (ret == -E2BIG) - die("probe point definition becomes too long."); - else if (ret < 0) - die("Failed to synthesize a probe point."); - } - - /* Settng up probe points */ add_trace_kprobe_events(session.probes, session.nr_probe, - session.force_add); + session.force_add, session.need_dwarf); return 0; } diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index 88a3b6d..1e60a65 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -40,6 +40,8 @@ #include "debug.h" #include "cache.h" #include "color.h" +#include "symbol.h" +#include "thread.h" #include "parse-events.h" /* For debugfs_path */ #include "probe-event.h" @@ -65,6 +67,38 @@ static int e_snprintf(char *str, size_t size, const char *format, ...) return ret; } + +static struct map_groups kmap_groups; +static struct map *kmaps[MAP__NR_TYPES]; + +/* Initialize symbol maps for vmlinux */ +static void init_vmlinux(void) +{ + symbol_conf.sort_by_name = true; + if (symbol_conf.vmlinux_name == NULL) + symbol_conf.try_vmlinux_path = true; + else + pr_debug("Use vmlinux: %s\n", symbol_conf.vmlinux_name); + if (symbol__init() < 0) + die("Failed to init symbol map."); + + map_groups__init(&kmap_groups); + if (map_groups__create_kernel_maps(&kmap_groups, kmaps) < 0) + die("Failed to create kernel maps."); +} + +#ifndef NO_DWARF_SUPPORT +static int open_vmlinux(void) +{ + if (map__load(kmaps[MAP__FUNCTION], NULL) < 0) { + pr_debug("Failed to load kernel map.\n"); + return -EINVAL; + } + pr_debug("Try to open %s\n", kmaps[MAP__FUNCTION]->dso->long_name); + return open(kmaps[MAP__FUNCTION]->dso->long_name, O_RDONLY); +} +#endif + void parse_line_range_desc(const char *arg, struct line_range *lr) { const char *ptr; @@ -586,8 +620,8 @@ static void get_new_event_name(char *buf, size_t len, const char *base, die("Too many events are on the same function."); } -void add_trace_kprobe_events(struct probe_point *probes, int nr_probes, - bool force_add) +static void __add_trace_kprobe_events(struct probe_point *probes, + int nr_probes, bool force_add) { int i, j, fd; struct probe_point *pp; @@ -640,6 +674,92 @@ void add_trace_kprobe_events(struct probe_point *probes, int nr_probes, close(fd); } +/* Currently just checking function name from symbol map */ +static void evaluate_probe_point(struct probe_point *pp) +{ + struct symbol *sym; + sym = map__find_symbol_by_name(kmaps[MAP__FUNCTION], + pp->function, NULL); + if (!sym) + die("Kernel symbol \'%s\' not found - probe not added.", + pp->function); +} + +void add_trace_kprobe_events(struct probe_point *probes, int nr_probes, + bool force_add, bool need_dwarf) +{ + int i, ret; + struct probe_point *pp; +#ifndef NO_DWARF_SUPPORT + int fd; +#endif + /* Add probes */ + init_vmlinux(); + + if (need_dwarf) +#ifdef NO_DWARF_SUPPORT + die("Debuginfo-analysis is not supported"); +#else /* !NO_DWARF_SUPPORT */ + pr_debug("Some probes require debuginfo.\n"); + + fd = open_vmlinux(); + if (fd < 0) { + if (need_dwarf) + die("Could not open debuginfo file."); + + pr_debug("Could not open vmlinux/module file." + " Try to use symbols.\n"); + goto end_dwarf; + } + + /* Searching probe points */ + for (i = 0; i < nr_probes; i++) { + pp = &probes[i]; + if (pp->found) + continue; + + lseek(fd, SEEK_SET, 0); + ret = find_probe_point(fd, pp); + if (ret > 0) + continue; + if (ret == 0) { /* No error but failed to find probe point. */ + synthesize_perf_probe_point(pp); + die("Probe point '%s' not found. - probe not added.", + pp->probes[0]); + } + /* Error path */ + if (need_dwarf) { + if (ret == -ENOENT) + pr_warning("No dwarf info found in the vmlinux - please rebuild with CONFIG_DEBUG_INFO=y.\n"); + die("Could not analyze debuginfo."); + } + pr_debug("An error occurred in debuginfo analysis." + " Try to use symbols.\n"); + break; + } + close(fd); + +end_dwarf: +#endif /* !NO_DWARF_SUPPORT */ + + /* Synthesize probes without dwarf */ + for (i = 0; i < nr_probes; i++) { + pp = &probes[i]; + if (pp->found) /* This probe is already found. */ + continue; + + evaluate_probe_point(pp); + ret = synthesize_trace_kprobe_event(pp); + if (ret == -E2BIG) + die("probe point definition becomes too long."); + else if (ret < 0) + die("Failed to synthesize a probe point."); + } + + /* Settng up probe points */ + __add_trace_kprobe_events(probes, nr_probes, force_add); +} + static void __del_trace_kprobe_event(int fd, struct str_node *ent) { char *p; @@ -759,6 +879,17 @@ void show_line_range(struct line_range *lr) unsigned int l = 1; struct line_node *ln; FILE *fp; + int fd, ret; + + /* Search a line range */ + init_vmlinux(); + fd = open_vmlinux(); + if (fd < 0) + die("Could not open debuginfo file."); + ret = find_line_range(fd, lr); + if (ret <= 0) + die("Source line is not found.\n"); + close(fd); setup_pager(); @@ -788,3 +919,5 @@ void show_line_range(struct line_range *lr) fclose(fp); } + + diff --git a/tools/perf/util/probe-event.h b/tools/perf/util/probe-event.h index 711287d..3865e16 100644 --- a/tools/perf/util/probe-event.h +++ b/tools/perf/util/probe-event.h @@ -13,7 +13,7 @@ extern int synthesize_perf_probe_event(struct probe_point *pp); extern void parse_trace_kprobe_event(const char *str, struct probe_point *pp); extern int synthesize_trace_kprobe_event(struct probe_point *pp); extern void add_trace_kprobe_events(struct probe_point *probes, int nr_probes, - bool force_add); + bool force_add, bool need_dwarf); extern void del_trace_kprobe_events(struct strlist *dellist); extern void show_perf_probe_events(void); extern void show_line_range(struct line_range *lr); -- cgit v0.10.2 From 12a1fadb41b5a6733c36b488b881fb19a28c92d3 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 16 Mar 2010 18:05:44 -0400 Subject: perf probe: Rename session to param Since this name 'session' conflicts with 'perf_session', and this structure just holds parameters anymore. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <20100316220544.32050.8788.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c index 2087034..f577e14 100644 --- a/tools/perf/builtin-probe.c +++ b/tools/perf/builtin-probe.c @@ -56,20 +56,20 @@ static struct { struct probe_point probes[MAX_PROBES]; struct strlist *dellist; struct line_range line_range; -} session; +} params; /* Parse an event definition. Note that any error must die. */ static void parse_probe_event(const char *str) { - struct probe_point *pp = &session.probes[session.nr_probe]; + struct probe_point *pp = ¶ms.probes[params.nr_probe]; - pr_debug("probe-definition(%d): %s\n", session.nr_probe, str); - if (++session.nr_probe == MAX_PROBES) + pr_debug("probe-definition(%d): %s\n", params.nr_probe, str); + if (++params.nr_probe == MAX_PROBES) die("Too many probes (> %d) are specified.", MAX_PROBES); /* Parse perf-probe event into probe_point */ - parse_perf_probe_event(str, pp, &session.need_dwarf); + parse_perf_probe_event(str, pp, ¶ms.need_dwarf); pr_debug("%d arguments\n", pp->nr_args); } @@ -103,9 +103,9 @@ static int opt_del_probe_event(const struct option *opt __used, const char *str, int unset __used) { if (str) { - if (!session.dellist) - session.dellist = strlist__new(true, NULL); - strlist__add(session.dellist, str); + if (!params.dellist) + params.dellist = strlist__new(true, NULL); + strlist__add(params.dellist, str); } return 0; } @@ -115,9 +115,9 @@ static int opt_show_lines(const struct option *opt __used, const char *str, int unset __used) { if (str) - parse_line_range_desc(str, &session.line_range); - INIT_LIST_HEAD(&session.line_range.line_list); - session.show_lines = true; + parse_line_range_desc(str, ¶ms.line_range); + INIT_LIST_HEAD(¶ms.line_range.line_list); + params.show_lines = true; return 0; } #endif @@ -140,7 +140,7 @@ static const struct option options[] = { OPT_STRING('k', "vmlinux", &symbol_conf.vmlinux_name, "file", "vmlinux pathname"), #endif - OPT_BOOLEAN('l', "list", &session.list_events, + OPT_BOOLEAN('l', "list", ¶ms.list_events, "list up current probe events"), OPT_CALLBACK('d', "del", NULL, "[GROUP:]EVENT", "delete a probe event.", opt_del_probe_event), @@ -168,7 +168,7 @@ static const struct option options[] = { #endif "\t\t\tkprobe-tracer argument format.)\n", opt_add_probe_event), - OPT_BOOLEAN('f', "force", &session.force_add, "forcibly add events" + OPT_BOOLEAN('f', "force", ¶ms.force_add, "forcibly add events" " with existing name"), #ifndef NO_DWARF_SUPPORT OPT_CALLBACK('L', "line", NULL, @@ -190,20 +190,20 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) parse_probe_event_argv(argc, argv); } - if ((!session.nr_probe && !session.dellist && !session.list_events && - !session.show_lines)) + if ((!params.nr_probe && !params.dellist && !params.list_events && + !params.show_lines)) usage_with_options(probe_usage, options); if (debugfs_valid_mountpoint(debugfs_path) < 0) die("Failed to find debugfs path."); - if (session.list_events) { - if (session.nr_probe != 0 || session.dellist) { + if (params.list_events) { + if (params.nr_probe != 0 || params.dellist) { pr_warning(" Error: Don't use --list with" " --add/--del.\n"); usage_with_options(probe_usage, options); } - if (session.show_lines) { + if (params.show_lines) { pr_warning(" Error: Don't use --list with --line.\n"); usage_with_options(probe_usage, options); } @@ -212,27 +212,27 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) } #ifndef NO_DWARF_SUPPORT - if (session.show_lines) { - if (session.nr_probe != 0 || session.dellist) { + if (params.show_lines) { + if (params.nr_probe != 0 || params.dellist) { pr_warning(" Error: Don't use --line with" " --add/--del.\n"); usage_with_options(probe_usage, options); } - show_line_range(&session.line_range); + show_line_range(¶ms.line_range); return 0; } #endif - if (session.dellist) { - del_trace_kprobe_events(session.dellist); - strlist__delete(session.dellist); - if (session.nr_probe == 0) + if (params.dellist) { + del_trace_kprobe_events(params.dellist); + strlist__delete(params.dellist); + if (params.nr_probe == 0) return 0; } - add_trace_kprobe_events(session.probes, session.nr_probe, - session.force_add, session.need_dwarf); + add_trace_kprobe_events(params.probes, params.nr_probe, + params.force_add, params.need_dwarf); return 0; } -- cgit v0.10.2 From 95a3e4c4e21de1920a2ddb54bfc57c0af7e2561e Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 16 Mar 2010 18:05:51 -0400 Subject: perf probe: Rename some die_get_* functions Rename die_get_real_subprogram and die_get_inlinefunc to die_find_real_subprogram and die_find_inlinefunc respectively, because these functions search its children. After that, 'die_get_' means getting a property of that die, and 'die_find_' means searching DIE-tree to get an appropriate child die. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <20100316220551.32050.36181.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index e887bb6..c91a960 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -204,8 +204,8 @@ static int __die_search_func_cb(Dwarf_Die *fn_die, void *data) } /* Search a real subprogram including this line, */ -static Dwarf_Die *die_get_real_subprogram(Dwarf_Die *cu_die, Dwarf_Addr addr, - Dwarf_Die *die_mem) +static Dwarf_Die *die_find_real_subprogram(Dwarf_Die *cu_die, Dwarf_Addr addr, + Dwarf_Die *die_mem) { struct __addr_die_search_param ad; ad.addr = addr; @@ -218,8 +218,8 @@ static Dwarf_Die *die_get_real_subprogram(Dwarf_Die *cu_die, Dwarf_Addr addr, } /* Similar to dwarf_getfuncs, but returns inlined_subroutine if exists. */ -static Dwarf_Die *die_get_inlinefunc(Dwarf_Die *sp_die, Dwarf_Addr addr, - Dwarf_Die *die_mem) +static Dwarf_Die *die_find_inlinefunc(Dwarf_Die *sp_die, Dwarf_Addr addr, + Dwarf_Die *die_mem) { Dwarf_Die child_die; int ret; @@ -233,7 +233,7 @@ static Dwarf_Die *die_get_inlinefunc(Dwarf_Die *sp_die, Dwarf_Addr addr, dwarf_haspc(die_mem, addr)) return die_mem; - if (die_get_inlinefunc(die_mem, addr, &child_die)) { + if (die_find_inlinefunc(die_mem, addr, &child_die)) { memcpy(die_mem, &child_die, sizeof(Dwarf_Die)); return die_mem; } @@ -401,7 +401,7 @@ static void show_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf) /* If no real subprogram, find a real one */ if (!sp_die || dwarf_tag(sp_die) != DW_TAG_subprogram) { - sp_die = die_get_real_subprogram(&pf->cu_die, + sp_die = die_find_real_subprogram(&pf->cu_die, pf->addr, &die_mem); if (!sp_die) die("Probe point is not found in subprograms."); @@ -564,7 +564,7 @@ static void find_probe_point_lazy(Dwarf_Die *sp_die, struct probe_finder *pf) if (!dwarf_haspc(sp_die, addr)) continue; /* Address filtering 2: No child include addr? */ - if (die_get_inlinefunc(sp_die, addr, &die_mem)) + if (die_find_inlinefunc(sp_die, addr, &die_mem)) continue; } @@ -714,7 +714,7 @@ static void find_line_range_by_line(Dwarf_Die *sp_die, struct line_finder *lf) continue; /* Address filtering 2: No child include addr? */ - if (die_get_inlinefunc(sp_die, addr, &die_mem)) + if (die_find_inlinefunc(sp_die, addr, &die_mem)) continue; } -- cgit v0.10.2 From 016f262e4fb10c6ecff709317098912f94a21efa Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 16 Mar 2010 18:05:58 -0400 Subject: perf probe: Introduce die_find_child() function Introduce die_find_child() function to integrate DIE-tree searching functions. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <20100316220558.32050.7905.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index c91a960..3942e14 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -186,6 +186,62 @@ static const char *cu_find_realpath(Dwarf_Die *cu_die, const char *fname) return src; } +/* Compare diename and tname */ +static bool die_compare_name(Dwarf_Die *dw_die, const char *tname) +{ + const char *name; + name = dwarf_diename(dw_die); + DIE_IF(name == NULL); + return strcmp(tname, name); +} + +/* Get entry pc(or low pc, 1st entry of ranges) of the die */ +static Dwarf_Addr die_get_entrypc(Dwarf_Die *dw_die) +{ + Dwarf_Addr epc; + int ret; + + ret = dwarf_entrypc(dw_die, &epc); + DIE_IF(ret == -1); + return epc; +} + +/* Return values for die_find callbacks */ +enum { + DIE_FIND_CB_FOUND = 0, /* End of Search */ + DIE_FIND_CB_CHILD = 1, /* Search only children */ + DIE_FIND_CB_SIBLING = 2, /* Search only siblings */ + DIE_FIND_CB_CONTINUE = 3, /* Search children and siblings */ +}; + +/* Search a child die */ +static Dwarf_Die *die_find_child(Dwarf_Die *rt_die, + int (*callback)(Dwarf_Die *, void *), + void *data, Dwarf_Die *die_mem) +{ + Dwarf_Die child_die; + int ret; + + ret = dwarf_child(rt_die, die_mem); + if (ret != 0) + return NULL; + + do { + ret = callback(die_mem, data); + if (ret == DIE_FIND_CB_FOUND) + return die_mem; + + if ((ret & DIE_FIND_CB_CHILD) && + die_find_child(die_mem, callback, data, &child_die)) { + memcpy(die_mem, &child_die, sizeof(Dwarf_Die)); + return die_mem; + } + } while ((ret & DIE_FIND_CB_SIBLING) && + dwarf_siblingof(die_mem, die_mem) == 0); + + return NULL; +} + struct __addr_die_search_param { Dwarf_Addr addr; Dwarf_Die *die_mem; @@ -217,77 +273,45 @@ static Dwarf_Die *die_find_real_subprogram(Dwarf_Die *cu_die, Dwarf_Addr addr, return die_mem; } -/* Similar to dwarf_getfuncs, but returns inlined_subroutine if exists. */ -static Dwarf_Die *die_find_inlinefunc(Dwarf_Die *sp_die, Dwarf_Addr addr, - Dwarf_Die *die_mem) +/* die_find callback for inline function search */ +static int __die_find_inline_cb(Dwarf_Die *die_mem, void *data) { - Dwarf_Die child_die; - int ret; + Dwarf_Addr *addr = data; - ret = dwarf_child(sp_die, die_mem); - if (ret != 0) - return NULL; - - do { - if (dwarf_tag(die_mem) == DW_TAG_inlined_subroutine && - dwarf_haspc(die_mem, addr)) - return die_mem; + if (dwarf_tag(die_mem) == DW_TAG_inlined_subroutine && + dwarf_haspc(die_mem, *addr)) + return DIE_FIND_CB_FOUND; - if (die_find_inlinefunc(die_mem, addr, &child_die)) { - memcpy(die_mem, &child_die, sizeof(Dwarf_Die)); - return die_mem; - } - } while (dwarf_siblingof(die_mem, die_mem) == 0); - - return NULL; + return DIE_FIND_CB_CONTINUE; } -/* Compare diename and tname */ -static bool die_compare_name(Dwarf_Die *dw_die, const char *tname) +/* Similar to dwarf_getfuncs, but returns inlined_subroutine if exists. */ +static Dwarf_Die *die_find_inlinefunc(Dwarf_Die *sp_die, Dwarf_Addr addr, + Dwarf_Die *die_mem) { - const char *name; - name = dwarf_diename(dw_die); - DIE_IF(name == NULL); - return strcmp(tname, name); + return die_find_child(sp_die, __die_find_inline_cb, &addr, die_mem); } -/* Get entry pc(or low pc, 1st entry of ranges) of the die */ -static Dwarf_Addr die_get_entrypc(Dwarf_Die *dw_die) +static int __die_find_variable_cb(Dwarf_Die *die_mem, void *data) { - Dwarf_Addr epc; - int ret; + const char *name = data; + int tag; - ret = dwarf_entrypc(dw_die, &epc); - DIE_IF(ret == -1); - return epc; + tag = dwarf_tag(die_mem); + if ((tag == DW_TAG_formal_parameter || + tag == DW_TAG_variable) && + (die_compare_name(die_mem, name) == 0)) + return DIE_FIND_CB_FOUND; + + return DIE_FIND_CB_CONTINUE; } -/* Get a variable die */ +/* Find a variable called 'name' */ static Dwarf_Die *die_find_variable(Dwarf_Die *sp_die, const char *name, Dwarf_Die *die_mem) { - Dwarf_Die child_die; - int tag; - int ret; - - ret = dwarf_child(sp_die, die_mem); - if (ret != 0) - return NULL; - - do { - tag = dwarf_tag(die_mem); - if ((tag == DW_TAG_formal_parameter || - tag == DW_TAG_variable) && - (die_compare_name(die_mem, name) == 0)) - return die_mem; - - if (die_find_variable(die_mem, name, &child_die)) { - memcpy(die_mem, &child_die, sizeof(Dwarf_Die)); - return die_mem; - } - } while (dwarf_siblingof(die_mem, die_mem) == 0); - - return NULL; + return die_find_child(sp_die, __die_find_variable_cb, (void *)name, + die_mem); } /* -- cgit v0.10.2 From f4d7da499e4fc1fdff8f26fdeb1a058d475a7a6c Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 16 Mar 2010 18:06:05 -0400 Subject: perf probe: Add --dry-run option Add --dry-run option for debugging and testing. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <20100316220605.32050.6571.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/Documentation/perf-probe.txt b/tools/perf/Documentation/perf-probe.txt index 34202b1..0f944b3 100644 --- a/tools/perf/Documentation/perf-probe.txt +++ b/tools/perf/Documentation/perf-probe.txt @@ -57,6 +57,11 @@ OPTIONS --force:: Forcibly add events with existing name. +-n:: +--dry-run:: + Dry run. With this option, --add and --del doesn't execute actual + adding and removal operations. + PROBE SYNTAX ------------ Probe points are defined by following syntax. diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c index f577e14..a1a2891 100644 --- a/tools/perf/builtin-probe.c +++ b/tools/perf/builtin-probe.c @@ -175,6 +175,7 @@ static const struct option options[] = { "FUNC[:RLN[+NUM|:RLN2]]|SRC:ALN[+NUM|:ALN2]", "Show source code lines.", opt_show_lines), #endif + OPT__DRY_RUN(&probe_event_dry_run), OPT_END() }; diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index 1e60a65..ac41578 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -49,6 +49,8 @@ #define MAX_PROBE_ARGS 128 #define PERFPROBE_GROUP "probe" +bool probe_event_dry_run; /* Dry run flag */ + #define semantic_error(msg ...) die("Semantic error :" msg) /* If there is no space to write, returns -E2BIG. */ @@ -430,7 +432,7 @@ error: return ret; } -static int open_kprobe_events(int flags, int mode) +static int open_kprobe_events(bool readwrite) { char buf[PATH_MAX]; int ret; @@ -439,7 +441,11 @@ static int open_kprobe_events(int flags, int mode) if (ret < 0) die("Failed to make kprobe_events path."); - ret = open(buf, flags, mode); + if (readwrite && !probe_event_dry_run) + ret = open(buf, O_RDWR, O_APPEND); + else + ret = open(buf, O_RDONLY, 0); + if (ret < 0) { if (errno == ENOENT) die("kprobe_events file does not exist -" @@ -535,7 +541,7 @@ void show_perf_probe_events(void) setup_pager(); memset(&pp, 0, sizeof(pp)); - fd = open_kprobe_events(O_RDONLY, 0); + fd = open_kprobe_events(false); rawlist = get_trace_kprobe_event_rawlist(fd); close(fd); @@ -585,9 +591,11 @@ static void write_trace_kprobe_event(int fd, const char *buf) int ret; pr_debug("Writing event: %s\n", buf); - ret = write(fd, buf, strlen(buf)); - if (ret <= 0) - die("Failed to write event: %s", strerror(errno)); + if (!probe_event_dry_run) { + ret = write(fd, buf, strlen(buf)); + if (ret <= 0) + die("Failed to write event: %s", strerror(errno)); + } } static void get_new_event_name(char *buf, size_t len, const char *base, @@ -630,7 +638,7 @@ static void __add_trace_kprobe_events(struct probe_point *probes, struct strlist *namelist; bool allow_suffix; - fd = open_kprobe_events(O_RDWR, O_APPEND); + fd = open_kprobe_events(true); /* Get current event names */ namelist = get_perf_event_names(fd, false); @@ -814,7 +822,7 @@ void del_trace_kprobe_events(struct strlist *dellist) struct str_node *ent; struct strlist *namelist; - fd = open_kprobe_events(O_RDWR, O_APPEND); + fd = open_kprobe_events(true); /* Get current event names */ namelist = get_perf_event_names(fd, true); diff --git a/tools/perf/util/probe-event.h b/tools/perf/util/probe-event.h index 3865e16..703b887 100644 --- a/tools/perf/util/probe-event.h +++ b/tools/perf/util/probe-event.h @@ -5,6 +5,8 @@ #include "probe-finder.h" #include "strlist.h" +extern bool probe_event_dry_run; + extern void parse_line_range_desc(const char *arg, struct line_range *lr); extern void parse_perf_probe_event(const char *str, struct probe_point *pp, bool *need_dwarf); -- cgit v0.10.2 From 4235b0454ebeefc2295ad8417e18a8761425b19e Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 16 Mar 2010 18:06:12 -0400 Subject: perf probe: Introduce kprobe_trace_event and perf_probe_event Introduce kprobe_trace_event and perf_probe_event and replace old probe_point structure with it. probe_point structure is not enough flexible nor extensible. New data structures will help implementing further features. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <20100316220612.32050.33806.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c index a1a2891..e0dafd9 100644 --- a/tools/perf/builtin-probe.c +++ b/tools/perf/builtin-probe.c @@ -48,12 +48,11 @@ /* Session management structure */ static struct { - bool need_dwarf; bool list_events; bool force_add; bool show_lines; - int nr_probe; - struct probe_point probes[MAX_PROBES]; + int nevents; + struct perf_probe_event events[MAX_PROBES]; struct strlist *dellist; struct line_range line_range; } params; @@ -62,16 +61,16 @@ static struct { /* Parse an event definition. Note that any error must die. */ static void parse_probe_event(const char *str) { - struct probe_point *pp = ¶ms.probes[params.nr_probe]; + struct perf_probe_event *pev = ¶ms.events[params.nevents]; - pr_debug("probe-definition(%d): %s\n", params.nr_probe, str); - if (++params.nr_probe == MAX_PROBES) + pr_debug("probe-definition(%d): %s\n", params.nevents, str); + if (++params.nevents == MAX_PROBES) die("Too many probes (> %d) are specified.", MAX_PROBES); - /* Parse perf-probe event into probe_point */ - parse_perf_probe_event(str, pp, ¶ms.need_dwarf); + /* Parse a perf-probe command into event */ + parse_perf_probe_command(str, pev); - pr_debug("%d arguments\n", pp->nr_args); + pr_debug("%d arguments\n", pev->nargs); } static void parse_probe_event_argv(int argc, const char **argv) @@ -191,7 +190,7 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) parse_probe_event_argv(argc, argv); } - if ((!params.nr_probe && !params.dellist && !params.list_events && + if ((!params.nevents && !params.dellist && !params.list_events && !params.show_lines)) usage_with_options(probe_usage, options); @@ -199,7 +198,7 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) die("Failed to find debugfs path."); if (params.list_events) { - if (params.nr_probe != 0 || params.dellist) { + if (params.nevents != 0 || params.dellist) { pr_warning(" Error: Don't use --list with" " --add/--del.\n"); usage_with_options(probe_usage, options); @@ -214,7 +213,7 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) #ifndef NO_DWARF_SUPPORT if (params.show_lines) { - if (params.nr_probe != 0 || params.dellist) { + if (params.nevents != 0 || params.dellist) { pr_warning(" Error: Don't use --line with" " --add/--del.\n"); usage_with_options(probe_usage, options); @@ -226,14 +225,13 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) #endif if (params.dellist) { - del_trace_kprobe_events(params.dellist); + del_perf_probe_events(params.dellist); strlist__delete(params.dellist); - if (params.nr_probe == 0) + if (params.nevents == 0) return 0; } - add_trace_kprobe_events(params.probes, params.nr_probe, - params.force_add, params.need_dwarf); + add_perf_probe_events(params.events, params.nevents, params.force_add); return 0; } diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index ac41578..b44ddfb 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -44,6 +44,7 @@ #include "thread.h" #include "parse-events.h" /* For debugfs_path */ #include "probe-event.h" +#include "probe-finder.h" #define MAX_CMDLEN 256 #define MAX_PROBE_ARGS 128 @@ -150,8 +151,9 @@ static bool check_event_name(const char *name) } /* Parse probepoint definition. */ -static void parse_perf_probe_probepoint(char *arg, struct probe_point *pp) +static void parse_perf_probe_point(char *arg, struct perf_probe_event *pev) { + struct perf_probe_point *pp = &pev->point; char *ptr, *tmp; char c, nc = 0; /* @@ -172,7 +174,8 @@ static void parse_perf_probe_probepoint(char *arg, struct probe_point *pp) if (!check_event_name(arg)) semantic_error("%s is bad for event name -it must " "follow C symbol-naming rule.", arg); - pp->event = xstrdup(arg); + pev->event = xstrdup(arg); + pev->group = NULL; arg = tmp; } @@ -255,57 +258,65 @@ static void parse_perf_probe_probepoint(char *arg, struct probe_point *pp) semantic_error("Offset/Line/Lazy pattern can't be used with " "return probe."); - pr_debug("symbol:%s file:%s line:%d offset:%d return:%d lazy:%s\n", + pr_debug("symbol:%s file:%s line:%d offset:%lu return:%d lazy:%s\n", pp->function, pp->file, pp->line, pp->offset, pp->retprobe, pp->lazy_line); } -/* Parse perf-probe event definition */ -void parse_perf_probe_event(const char *str, struct probe_point *pp, - bool *need_dwarf) +/* Parse perf-probe event command */ +void parse_perf_probe_command(const char *cmd, struct perf_probe_event *pev) { char **argv; int argc, i; - *need_dwarf = false; - - argv = argv_split(str, &argc); + argv = argv_split(cmd, &argc); if (!argv) die("argv_split failed."); if (argc > MAX_PROBE_ARGS + 1) semantic_error("Too many arguments"); /* Parse probe point */ - parse_perf_probe_probepoint(argv[0], pp); - if (pp->file || pp->line || pp->lazy_line) - *need_dwarf = true; + parse_perf_probe_point(argv[0], pev); /* Copy arguments and ensure return probe has no C argument */ - pp->nr_args = argc - 1; - pp->args = xzalloc(sizeof(char *) * pp->nr_args); - for (i = 0; i < pp->nr_args; i++) { - pp->args[i] = xstrdup(argv[i + 1]); - if (is_c_varname(pp->args[i])) { - if (pp->retprobe) - semantic_error("You can't specify local" - " variable for kretprobe"); - *need_dwarf = true; - } + pev->nargs = argc - 1; + pev->args = xzalloc(sizeof(struct perf_probe_arg) * pev->nargs); + for (i = 0; i < pev->nargs; i++) { + pev->args[i].name = xstrdup(argv[i + 1]); + if (is_c_varname(pev->args[i].name) && pev->point.retprobe) + semantic_error("You can't specify local variable for" + " kretprobe"); } argv_free(argv); } +/* Return true if this perf_probe_event requires debuginfo */ +bool perf_probe_event_need_dwarf(struct perf_probe_event *pev) +{ + int i; + + if (pev->point.file || pev->point.line || pev->point.lazy_line) + return true; + + for (i = 0; i < pev->nargs; i++) + if (is_c_varname(pev->args[i].name)) + return true; + + return false; +} + /* Parse kprobe_events event into struct probe_point */ -void parse_trace_kprobe_event(const char *str, struct probe_point *pp) +void parse_kprobe_trace_command(const char *cmd, struct kprobe_trace_event *tev) { + struct kprobe_trace_point *tp = &tev->point; char pr; char *p; int ret, i, argc; char **argv; - pr_debug("Parsing kprobe_events: %s\n", str); - argv = argv_split(str, &argc); + pr_debug("Parsing kprobe_events: %s\n", cmd); + argv = argv_split(cmd, &argc); if (!argv) die("argv_split failed."); if (argc < 2) @@ -313,47 +324,46 @@ void parse_trace_kprobe_event(const char *str, struct probe_point *pp) /* Scan event and group name. */ ret = sscanf(argv[0], "%c:%a[^/ \t]/%a[^ \t]", - &pr, (float *)(void *)&pp->group, - (float *)(void *)&pp->event); + &pr, (float *)(void *)&tev->group, + (float *)(void *)&tev->event); if (ret != 3) semantic_error("Failed to parse event name: %s", argv[0]); - pr_debug("Group:%s Event:%s probe:%c\n", pp->group, pp->event, pr); + pr_debug("Group:%s Event:%s probe:%c\n", tev->group, tev->event, pr); - pp->retprobe = (pr == 'r'); + tp->retprobe = (pr == 'r'); /* Scan function name and offset */ - ret = sscanf(argv[1], "%a[^+]+%d", (float *)(void *)&pp->function, - &pp->offset); + ret = sscanf(argv[1], "%a[^+]+%lu", (float *)(void *)&tp->symbol, + &tp->offset); if (ret == 1) - pp->offset = 0; - - /* kprobe_events doesn't have this information */ - pp->line = 0; - pp->file = NULL; + tp->offset = 0; - pp->nr_args = argc - 2; - pp->args = xzalloc(sizeof(char *) * pp->nr_args); - for (i = 0; i < pp->nr_args; i++) { + tev->nargs = argc - 2; + tev->args = xzalloc(sizeof(struct kprobe_trace_arg) * tev->nargs); + for (i = 0; i < tev->nargs; i++) { p = strchr(argv[i + 2], '='); if (p) /* We don't need which register is assigned. */ - *p = '\0'; - pp->args[i] = xstrdup(argv[i + 2]); + *p++ = '\0'; + else + p = argv[i + 2]; + tev->args[i].name = xstrdup(argv[i + 2]); + /* TODO: parse regs and offset */ + tev->args[i].value = xstrdup(p); } argv_free(argv); } -/* Synthesize only probe point (not argument) */ -int synthesize_perf_probe_point(struct probe_point *pp) +/* Compose only probe point (not argument) */ +static char *synthesize_perf_probe_point(struct perf_probe_point *pp) { char *buf; char offs[64] = "", line[64] = ""; int ret; - pp->probes[0] = buf = xzalloc(MAX_CMDLEN); - pp->found = 1; + buf = xzalloc(MAX_CMDLEN); if (pp->offset) { - ret = e_snprintf(offs, 64, "+%d", pp->offset); + ret = e_snprintf(offs, 64, "+%lu", pp->offset); if (ret <= 0) goto error; } @@ -368,68 +378,209 @@ int synthesize_perf_probe_point(struct probe_point *pp) offs, pp->retprobe ? "%return" : "", line); else ret = e_snprintf(buf, MAX_CMDLEN, "%s%s", pp->file, line); - if (ret <= 0) { + if (ret <= 0) + goto error; + + return buf; error: - free(pp->probes[0]); - pp->probes[0] = NULL; - pp->found = 0; - } - return ret; + die("Failed to synthesize perf probe point: %s", strerror(-ret)); } -int synthesize_perf_probe_event(struct probe_point *pp) +#if 0 +char *synthesize_perf_probe_command(struct perf_probe_event *pev) { char *buf; int i, len, ret; - len = synthesize_perf_probe_point(pp); - if (len < 0) - return 0; + buf = synthesize_perf_probe_point(&pev->point); + if (!buf) + return NULL; - buf = pp->probes[0]; - for (i = 0; i < pp->nr_args; i++) { + len = strlen(buf); + for (i = 0; i < pev->nargs; i++) { ret = e_snprintf(&buf[len], MAX_CMDLEN - len, " %s", - pp->args[i]); - if (ret <= 0) - goto error; + pev->args[i].name); + if (ret <= 0) { + free(buf); + return NULL; + } len += ret; } - pp->found = 1; - return pp->found; -error: - free(pp->probes[0]); - pp->probes[0] = NULL; + return buf; +} +#endif + +static int __synthesize_kprobe_trace_arg_ref(struct kprobe_trace_arg_ref *ref, + char **buf, size_t *buflen, + int depth) +{ + int ret; + if (ref->next) { + depth = __synthesize_kprobe_trace_arg_ref(ref->next, buf, + buflen, depth + 1); + if (depth < 0) + goto out; + } + + ret = e_snprintf(*buf, *buflen, "%+ld(", ref->offset); + if (ret < 0) + depth = ret; + else { + *buf += ret; + *buflen -= ret; + } +out: + return depth; - return ret; } -int synthesize_trace_kprobe_event(struct probe_point *pp) +static int synthesize_kprobe_trace_arg(struct kprobe_trace_arg *arg, + char *buf, size_t buflen) { + int ret, depth = 0; + char *tmp = buf; + + /* Argument name or separator */ + if (arg->name) + ret = e_snprintf(buf, buflen, " %s=", arg->name); + else + ret = e_snprintf(buf, buflen, " "); + if (ret < 0) + return ret; + buf += ret; + buflen -= ret; + + /* Dereferencing arguments */ + if (arg->ref) { + depth = __synthesize_kprobe_trace_arg_ref(arg->ref, &buf, + &buflen, 1); + if (depth < 0) + return depth; + } + + /* Print argument value */ + ret = e_snprintf(buf, buflen, "%s", arg->value); + if (ret < 0) + return ret; + buf += ret; + buflen -= ret; + + /* Closing */ + while (depth--) { + ret = e_snprintf(buf, buflen, ")"); + if (ret < 0) + return ret; + buf += ret; + buflen -= ret; + } + + return buf - tmp; +} + +char *synthesize_kprobe_trace_command(struct kprobe_trace_event *tev) +{ + struct kprobe_trace_point *tp = &tev->point; char *buf; int i, len, ret; - pp->probes[0] = buf = xzalloc(MAX_CMDLEN); - ret = e_snprintf(buf, MAX_CMDLEN, "%s+%d", pp->function, pp->offset); - if (ret <= 0) + buf = xzalloc(MAX_CMDLEN); + len = e_snprintf(buf, MAX_CMDLEN, "%c:%s/%s %s+%lu", + tp->retprobe ? 'r' : 'p', + tev->group, tev->event, + tp->symbol, tp->offset); + if (len <= 0) goto error; - len = ret; - for (i = 0; i < pp->nr_args; i++) { - ret = e_snprintf(&buf[len], MAX_CMDLEN - len, " %s", - pp->args[i]); + for (i = 0; i < tev->nargs; i++) { + ret = synthesize_kprobe_trace_arg(&tev->args[i], buf + len, + MAX_CMDLEN - len); if (ret <= 0) goto error; len += ret; } - pp->found = 1; - return pp->found; + return buf; error: - free(pp->probes[0]); - pp->probes[0] = NULL; + free(buf); + return NULL; +} - return ret; +void convert_to_perf_probe_event(struct kprobe_trace_event *tev, + struct perf_probe_event *pev) +{ + char buf[64]; + int i; + + pev->event = xstrdup(tev->event); + pev->group = xstrdup(tev->group); + + /* Convert trace_point to probe_point */ + pev->point.function = xstrdup(tev->point.symbol); + pev->point.offset = tev->point.offset; + pev->point.retprobe = tev->point.retprobe; + + /* Convert trace_arg to probe_arg */ + pev->nargs = tev->nargs; + pev->args = xzalloc(sizeof(struct perf_probe_arg) * pev->nargs); + for (i = 0; i < tev->nargs; i++) + if (tev->args[i].name) + pev->args[i].name = xstrdup(tev->args[i].name); + else { + synthesize_kprobe_trace_arg(&tev->args[i], buf, 64); + pev->args[i].name = xstrdup(buf); + } +} + +void clear_perf_probe_event(struct perf_probe_event *pev) +{ + struct perf_probe_point *pp = &pev->point; + int i; + + if (pev->event) + free(pev->event); + if (pev->group) + free(pev->group); + if (pp->file) + free(pp->file); + if (pp->function) + free(pp->function); + if (pp->lazy_line) + free(pp->lazy_line); + for (i = 0; i < pev->nargs; i++) + if (pev->args[i].name) + free(pev->args[i].name); + if (pev->args) + free(pev->args); + memset(pev, 0, sizeof(*pev)); +} + +void clear_kprobe_trace_event(struct kprobe_trace_event *tev) +{ + struct kprobe_trace_arg_ref *ref, *next; + int i; + + if (tev->event) + free(tev->event); + if (tev->group) + free(tev->group); + if (tev->point.symbol) + free(tev->point.symbol); + for (i = 0; i < tev->nargs; i++) { + if (tev->args[i].name) + free(tev->args[i].name); + if (tev->args[i].value) + free(tev->args[i].value); + ref = tev->args[i].ref; + while (ref) { + next = ref->next; + free(ref); + ref = next; + } + } + if (tev->args) + free(tev->args); + memset(tev, 0, sizeof(*tev)); } static int open_kprobe_events(bool readwrite) @@ -458,7 +609,7 @@ static int open_kprobe_events(bool readwrite) } /* Get raw string list of current kprobe_events */ -static struct strlist *get_trace_kprobe_event_rawlist(int fd) +static struct strlist *get_kprobe_trace_command_rawlist(int fd) { int ret, idx; FILE *fp; @@ -486,99 +637,82 @@ static struct strlist *get_trace_kprobe_event_rawlist(int fd) return sl; } -/* Free and zero clear probe_point */ -static void clear_probe_point(struct probe_point *pp) -{ - int i; - - if (pp->event) - free(pp->event); - if (pp->group) - free(pp->group); - if (pp->function) - free(pp->function); - if (pp->file) - free(pp->file); - if (pp->lazy_line) - free(pp->lazy_line); - for (i = 0; i < pp->nr_args; i++) - free(pp->args[i]); - if (pp->args) - free(pp->args); - for (i = 0; i < pp->found; i++) - free(pp->probes[i]); - memset(pp, 0, sizeof(*pp)); -} - /* Show an event */ -static void show_perf_probe_event(const char *event, const char *place, - struct probe_point *pp) +static void show_perf_probe_event(struct perf_probe_event *pev) { int i, ret; char buf[128]; + char *place; - ret = e_snprintf(buf, 128, "%s:%s", pp->group, event); + /* Synthesize only event probe point */ + place = synthesize_perf_probe_point(&pev->point); + + ret = e_snprintf(buf, 128, "%s:%s", pev->group, pev->event); if (ret < 0) die("Failed to copy event: %s", strerror(-ret)); printf(" %-40s (on %s", buf, place); - if (pp->nr_args > 0) { + if (pev->nargs > 0) { printf(" with"); - for (i = 0; i < pp->nr_args; i++) - printf(" %s", pp->args[i]); + for (i = 0; i < pev->nargs; i++) + printf(" %s", pev->args[i].name); } printf(")\n"); + free(place); } /* List up current perf-probe events */ void show_perf_probe_events(void) { int fd; - struct probe_point pp; + struct kprobe_trace_event tev; + struct perf_probe_event pev; struct strlist *rawlist; struct str_node *ent; setup_pager(); - memset(&pp, 0, sizeof(pp)); + + memset(&tev, 0, sizeof(tev)); + memset(&pev, 0, sizeof(pev)); fd = open_kprobe_events(false); - rawlist = get_trace_kprobe_event_rawlist(fd); + rawlist = get_kprobe_trace_command_rawlist(fd); close(fd); strlist__for_each(ent, rawlist) { - parse_trace_kprobe_event(ent->s, &pp); - /* Synthesize only event probe point */ - synthesize_perf_probe_point(&pp); + parse_kprobe_trace_command(ent->s, &tev); + convert_to_perf_probe_event(&tev, &pev); /* Show an event */ - show_perf_probe_event(pp.event, pp.probes[0], &pp); - clear_probe_point(&pp); + show_perf_probe_event(&pev); + clear_perf_probe_event(&pev); + clear_kprobe_trace_event(&tev); } strlist__delete(rawlist); } /* Get current perf-probe event names */ -static struct strlist *get_perf_event_names(int fd, bool include_group) +static struct strlist *get_kprobe_trace_event_names(int fd, bool include_group) { char buf[128]; struct strlist *sl, *rawlist; struct str_node *ent; - struct probe_point pp; + struct kprobe_trace_event tev; - memset(&pp, 0, sizeof(pp)); - rawlist = get_trace_kprobe_event_rawlist(fd); + memset(&tev, 0, sizeof(tev)); + rawlist = get_kprobe_trace_command_rawlist(fd); sl = strlist__new(true, NULL); strlist__for_each(ent, rawlist) { - parse_trace_kprobe_event(ent->s, &pp); + parse_kprobe_trace_command(ent->s, &tev); if (include_group) { - if (e_snprintf(buf, 128, "%s:%s", pp.group, - pp.event) < 0) + if (e_snprintf(buf, 128, "%s:%s", tev.group, + tev.event) < 0) die("Failed to copy group:event name."); strlist__add(sl, buf); } else - strlist__add(sl, pp.event); - clear_probe_point(&pp); + strlist__add(sl, tev.event); + clear_kprobe_trace_event(&tev); } strlist__delete(rawlist); @@ -586,9 +720,10 @@ static struct strlist *get_perf_event_names(int fd, bool include_group) return sl; } -static void write_trace_kprobe_event(int fd, const char *buf) +static void write_kprobe_trace_event(int fd, struct kprobe_trace_event *tev) { int ret; + char *buf = synthesize_kprobe_trace_command(tev); pr_debug("Writing event: %s\n", buf); if (!probe_event_dry_run) { @@ -596,6 +731,7 @@ static void write_trace_kprobe_event(int fd, const char *buf) if (ret <= 0) die("Failed to write event: %s", strerror(errno)); } + free(buf); } static void get_new_event_name(char *buf, size_t len, const char *base, @@ -628,81 +764,83 @@ static void get_new_event_name(char *buf, size_t len, const char *base, die("Too many events are on the same function."); } -static void __add_trace_kprobe_events(struct probe_point *probes, - int nr_probes, bool force_add) +static void __add_kprobe_trace_events(struct perf_probe_event *pev, + struct kprobe_trace_event *tevs, + int ntevs, bool allow_suffix) { - int i, j, fd; - struct probe_point *pp; - char buf[MAX_CMDLEN]; - char event[64]; + int i, fd; + struct kprobe_trace_event *tev; + char buf[64]; + const char *event, *group; struct strlist *namelist; - bool allow_suffix; fd = open_kprobe_events(true); /* Get current event names */ - namelist = get_perf_event_names(fd, false); - - for (j = 0; j < nr_probes; j++) { - pp = probes + j; - if (!pp->event) - pp->event = xstrdup(pp->function); - if (!pp->group) - pp->group = xstrdup(PERFPROBE_GROUP); - /* If force_add is true, suffix search is allowed */ - allow_suffix = force_add; - for (i = 0; i < pp->found; i++) { - /* Get an unused new event name */ - get_new_event_name(event, 64, pp->event, namelist, - allow_suffix); - snprintf(buf, MAX_CMDLEN, "%c:%s/%s %s\n", - pp->retprobe ? 'r' : 'p', - pp->group, event, - pp->probes[i]); - write_trace_kprobe_event(fd, buf); - printf("Added new event:\n"); - /* Get the first parameter (probe-point) */ - sscanf(pp->probes[i], "%s", buf); - show_perf_probe_event(event, buf, pp); - /* Add added event name to namelist */ - strlist__add(namelist, event); - /* - * Probes after the first probe which comes from same - * user input are always allowed to add suffix, because - * there might be several addresses corresponding to - * one code line. - */ - allow_suffix = true; - } + namelist = get_kprobe_trace_event_names(fd, false); + + printf("Add new event%s\n", (ntevs > 1) ? "s:" : ":"); + for (i = 0; i < ntevs; i++) { + tev = &tevs[i]; + if (pev->event) + event = pev->event; + else + if (pev->point.function) + event = pev->point.function; + else + event = tev->point.symbol; + if (pev->group) + group = pev->group; + else + group = PERFPROBE_GROUP; + + /* Get an unused new event name */ + get_new_event_name(buf, 64, event, namelist, allow_suffix); + event = buf; + + tev->event = xstrdup(event); + tev->group = xstrdup(group); + write_kprobe_trace_event(fd, tev); + /* Add added event name to namelist */ + strlist__add(namelist, event); + + /* Trick here - save current event/group */ + event = pev->event; + group = pev->group; + pev->event = tev->event; + pev->group = tev->group; + show_perf_probe_event(pev); + /* Trick here - restore current event/group */ + pev->event = (char *)event; + pev->group = (char *)group; + + /* + * Probes after the first probe which comes from same + * user input are always allowed to add suffix, because + * there might be several addresses corresponding to + * one code line. + */ + allow_suffix = true; } /* Show how to use the event. */ printf("\nYou can now use it on all perf tools, such as:\n\n"); - printf("\tperf record -e %s:%s -a sleep 1\n\n", PERFPROBE_GROUP, event); + printf("\tperf record -e %s:%s -a sleep 1\n\n", tev->group, tev->event); strlist__delete(namelist); close(fd); } -/* Currently just checking function name from symbol map */ -static void evaluate_probe_point(struct probe_point *pp) +static int convert_to_kprobe_trace_events(struct perf_probe_event *pev, + struct kprobe_trace_event **tevs) { struct symbol *sym; - sym = map__find_symbol_by_name(kmaps[MAP__FUNCTION], - pp->function, NULL); - if (!sym) - die("Kernel symbol \'%s\' not found - probe not added.", - pp->function); -} - -void add_trace_kprobe_events(struct probe_point *probes, int nr_probes, - bool force_add, bool need_dwarf) -{ - int i, ret; - struct probe_point *pp; + bool need_dwarf; #ifndef NO_DWARF_SUPPORT int fd; #endif - /* Add probes */ - init_vmlinux(); + int ntevs = 0, i; + struct kprobe_trace_event *tev; + + need_dwarf = perf_probe_event_need_dwarf(pev); if (need_dwarf) #ifdef NO_DWARF_SUPPORT @@ -721,57 +859,90 @@ void add_trace_kprobe_events(struct probe_point *probes, int nr_probes, } /* Searching probe points */ - for (i = 0; i < nr_probes; i++) { - pp = &probes[i]; - if (pp->found) - continue; - - lseek(fd, SEEK_SET, 0); - ret = find_probe_point(fd, pp); - if (ret > 0) - continue; - if (ret == 0) { /* No error but failed to find probe point. */ - synthesize_perf_probe_point(pp); - die("Probe point '%s' not found. - probe not added.", - pp->probes[0]); - } - /* Error path */ - if (need_dwarf) { - if (ret == -ENOENT) - pr_warning("No dwarf info found in the vmlinux - please rebuild with CONFIG_DEBUG_INFO=y.\n"); - die("Could not analyze debuginfo."); - } - pr_debug("An error occurred in debuginfo analysis." - " Try to use symbols.\n"); - break; + ntevs = find_kprobe_trace_events(fd, pev, tevs); + + if (ntevs > 0) /* Found */ + goto found; + + if (ntevs == 0) /* No error but failed to find probe point. */ + die("Probe point '%s' not found. - probe not added.", + synthesize_perf_probe_point(&pev->point)); + + /* Error path */ + if (need_dwarf) { + if (ntevs == -ENOENT) + pr_warning("No dwarf info found in the vmlinux - please rebuild with CONFIG_DEBUG_INFO=y.\n"); + die("Could not analyze debuginfo."); } - close(fd); + pr_debug("An error occurred in debuginfo analysis." + " Try to use symbols.\n"); end_dwarf: #endif /* !NO_DWARF_SUPPORT */ - /* Synthesize probes without dwarf */ - for (i = 0; i < nr_probes; i++) { - pp = &probes[i]; - if (pp->found) /* This probe is already found. */ - continue; + /* Allocate trace event buffer */ + ntevs = 1; + tev = *tevs = xzalloc(sizeof(struct kprobe_trace_event)); + + /* Copy parameters */ + tev->point.symbol = xstrdup(pev->point.function); + tev->point.offset = pev->point.offset; + tev->nargs = pev->nargs; + if (tev->nargs) { + tev->args = xzalloc(sizeof(struct kprobe_trace_arg) + * tev->nargs); + for (i = 0; i < tev->nargs; i++) + tev->args[i].value = xstrdup(pev->args[i].name); + } + + /* Currently just checking function name from symbol map */ + sym = map__find_symbol_by_name(kmaps[MAP__FUNCTION], + tev->point.symbol, NULL); + if (!sym) + die("Kernel symbol \'%s\' not found - probe not added.", + tev->point.symbol); +found: + close(fd); + return ntevs; +} + +struct __event_package { + struct perf_probe_event *pev; + struct kprobe_trace_event *tevs; + int ntevs; +}; - evaluate_probe_point(pp); - ret = synthesize_trace_kprobe_event(pp); - if (ret == -E2BIG) - die("probe point definition becomes too long."); - else if (ret < 0) - die("Failed to synthesize a probe point."); +void add_perf_probe_events(struct perf_probe_event *pevs, int npevs, + bool force_add) +{ + int i; + struct __event_package *pkgs; + + pkgs = xzalloc(sizeof(struct __event_package) * npevs); + + /* Init vmlinux path */ + init_vmlinux(); + + /* Loop 1: convert all events */ + for (i = 0; i < npevs; i++) { + pkgs[i].pev = &pevs[i]; + /* Convert with or without debuginfo */ + pkgs[i].ntevs = convert_to_kprobe_trace_events(pkgs[i].pev, + &pkgs[i].tevs); } - /* Settng up probe points */ - __add_trace_kprobe_events(probes, nr_probes, force_add); + /* Loop 2: add all events */ + for (i = 0; i < npevs; i++) + __add_kprobe_trace_events(pkgs[i].pev, pkgs[i].tevs, + pkgs[i].ntevs, force_add); + /* TODO: cleanup all trace events? */ } static void __del_trace_kprobe_event(int fd, struct str_node *ent) { char *p; char buf[128]; + int ret; /* Convert from perf-probe event to trace-kprobe event */ if (e_snprintf(buf, 128, "-:%s", ent->s) < 0) @@ -781,7 +952,10 @@ static void __del_trace_kprobe_event(int fd, struct str_node *ent) die("Internal error: %s should have ':' but not.", ent->s); *p = '/'; - write_trace_kprobe_event(fd, buf); + pr_debug("Writing event: %s\n", buf); + ret = write(fd, buf, strlen(buf)); + if (ret <= 0) + die("Failed to write event: %s", strerror(errno)); printf("Remove event: %s\n", ent->s); } @@ -814,7 +988,7 @@ static void del_trace_kprobe_event(int fd, const char *group, pr_info("Info: event \"%s\" does not exist, could not remove it.\n", buf); } -void del_trace_kprobe_events(struct strlist *dellist) +void del_perf_probe_events(struct strlist *dellist) { int fd; const char *group, *event; @@ -824,7 +998,7 @@ void del_trace_kprobe_events(struct strlist *dellist) fd = open_kprobe_events(true); /* Get current event names */ - namelist = get_perf_event_names(fd, true); + namelist = get_kprobe_trace_event_names(fd, true); strlist__for_each(ent, dellist) { str = xstrdup(ent->s); diff --git a/tools/perf/util/probe-event.h b/tools/perf/util/probe-event.h index 703b887..2a2f0a2 100644 --- a/tools/perf/util/probe-event.h +++ b/tools/perf/util/probe-event.h @@ -2,24 +2,113 @@ #define _PROBE_EVENT_H #include -#include "probe-finder.h" #include "strlist.h" extern bool probe_event_dry_run; -extern void parse_line_range_desc(const char *arg, struct line_range *lr); -extern void parse_perf_probe_event(const char *str, struct probe_point *pp, - bool *need_dwarf); -extern int synthesize_perf_probe_point(struct probe_point *pp); -extern int synthesize_perf_probe_event(struct probe_point *pp); -extern void parse_trace_kprobe_event(const char *str, struct probe_point *pp); -extern int synthesize_trace_kprobe_event(struct probe_point *pp); -extern void add_trace_kprobe_events(struct probe_point *probes, int nr_probes, - bool force_add, bool need_dwarf); -extern void del_trace_kprobe_events(struct strlist *dellist); +/* kprobe-tracer tracing point */ +struct kprobe_trace_point { + char *symbol; /* Base symbol */ + unsigned long offset; /* Offset from symbol */ + bool retprobe; /* Return probe flag */ +}; + +/* kprobe-tracer tracing argument referencing offset */ +struct kprobe_trace_arg_ref { + struct kprobe_trace_arg_ref *next; /* Next reference */ + long offset; /* Offset value */ +}; + +/* kprobe-tracer tracing argument */ +struct kprobe_trace_arg { + char *name; /* Argument name */ + char *value; /* Base value */ + struct kprobe_trace_arg_ref *ref; /* Referencing offset */ +}; + +/* kprobe-tracer tracing event (point + arg) */ +struct kprobe_trace_event { + char *event; /* Event name */ + char *group; /* Group name */ + struct kprobe_trace_point point; /* Trace point */ + int nargs; /* Number of args */ + struct kprobe_trace_arg *args; /* Arguments */ +}; + +/* Perf probe probing point */ +struct perf_probe_point { + char *file; /* File path */ + char *function; /* Function name */ + int line; /* Line number */ + char *lazy_line; /* Lazy matching pattern */ + unsigned long offset; /* Offset from function entry */ + bool retprobe; /* Return probe flag */ +}; + +/* Perf probe probing argument */ +struct perf_probe_arg { + char *name; /* Argument name */ +}; + +/* Perf probe probing event (point + arg) */ +struct perf_probe_event { + char *event; /* Event name */ + char *group; /* Group name */ + struct perf_probe_point point; /* Probe point */ + int nargs; /* Number of arguments */ + struct perf_probe_arg *args; /* Arguments */ +}; + + +/* Line number container */ +struct line_node { + struct list_head list; + unsigned int line; +}; + +/* Line range */ +struct line_range { + char *file; /* File name */ + char *function; /* Function name */ + unsigned int start; /* Start line number */ + unsigned int end; /* End line number */ + int offset; /* Start line offset */ + char *path; /* Real path name */ + struct list_head line_list; /* Visible lines */ +}; + +/* Command string to events */ +extern void parse_perf_probe_command(const char *cmd, + struct perf_probe_event *pev); +extern void parse_kprobe_trace_command(const char *cmd, + struct kprobe_trace_event *tev); + +/* Events to command string */ +extern char *synthesize_perf_probe_command(struct perf_probe_event *pev); +extern char *synthesize_kprobe_trace_command(struct kprobe_trace_event *tev); + +/* Check the perf_probe_event needs debuginfo */ +extern bool perf_probe_event_need_dwarf(struct perf_probe_event *pev); + +/* Convert from kprobe_trace_event to perf_probe_event */ +extern void convert_to_perf_probe_event(struct kprobe_trace_event *tev, + struct perf_probe_event *pev); + +/* Release event contents */ +extern void clear_perf_probe_event(struct perf_probe_event *pev); +extern void clear_kprobe_trace_event(struct kprobe_trace_event *tev); + +/* Command string to line-range */ +extern void parse_line_range_desc(const char *cmd, struct line_range *lr); + + +extern void add_perf_probe_events(struct perf_probe_event *pevs, int ntevs, + bool force_add); +extern void del_perf_probe_events(struct strlist *dellist); extern void show_perf_probe_events(void); extern void show_line_range(struct line_range *lr); + /* Maximum index number of event-name postfix */ #define MAX_EVENT_INDEX 1024 diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index 3942e14..251b4c4 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -319,19 +319,20 @@ static Dwarf_Die *die_find_variable(Dwarf_Die *sp_die, const char *name, */ /* Show a location */ -static void show_location(Dwarf_Op *op, struct probe_finder *pf) +static void convert_location(Dwarf_Op *op, struct probe_finder *pf) { unsigned int regn; Dwarf_Word offs = 0; - int deref = 0, ret; + bool ref = false; const char *regs; + struct kprobe_trace_arg *tvar = pf->tvar; /* TODO: support CFA */ /* If this is based on frame buffer, set the offset */ if (op->atom == DW_OP_fbreg) { if (pf->fb_ops == NULL) die("The attribute of frame base is not supported.\n"); - deref = 1; + ref = true; offs = op->number; op = &pf->fb_ops[0]; } @@ -339,13 +340,13 @@ static void show_location(Dwarf_Op *op, struct probe_finder *pf) if (op->atom >= DW_OP_breg0 && op->atom <= DW_OP_breg31) { regn = op->atom - DW_OP_breg0; offs += op->number; - deref = 1; + ref = true; } else if (op->atom >= DW_OP_reg0 && op->atom <= DW_OP_reg31) { regn = op->atom - DW_OP_reg0; } else if (op->atom == DW_OP_bregx) { regn = op->number; offs += op->number2; - deref = 1; + ref = true; } else if (op->atom == DW_OP_regx) { regn = op->number; } else @@ -355,17 +356,15 @@ static void show_location(Dwarf_Op *op, struct probe_finder *pf) if (!regs) die("%u exceeds max register number.", regn); - if (deref) - ret = snprintf(pf->buf, pf->len, " %s=%+jd(%s)", - pf->var, (intmax_t)offs, regs); - else - ret = snprintf(pf->buf, pf->len, " %s=%s", pf->var, regs); - DIE_IF(ret < 0); - DIE_IF(ret >= pf->len); + tvar->value = xstrdup(regs); + if (ref) { + tvar->ref = xzalloc(sizeof(struct kprobe_trace_arg_ref)); + tvar->ref->offset = (long)offs; + } } /* Show a variables in kprobe event format */ -static void show_variable(Dwarf_Die *vr_die, struct probe_finder *pf) +static void convert_variable(Dwarf_Die *vr_die, struct probe_finder *pf) { Dwarf_Attribute attr; Dwarf_Op *expr; @@ -379,50 +378,51 @@ static void show_variable(Dwarf_Die *vr_die, struct probe_finder *pf) if (ret <= 0 || nexpr == 0) goto error; - show_location(expr, pf); + convert_location(expr, pf); /* *expr will be cached in libdw. Don't free it. */ return ; error: /* TODO: Support const_value */ die("Failed to find the location of %s at this address.\n" - " Perhaps, it has been optimized out.", pf->var); + " Perhaps, it has been optimized out.", pf->pvar->name); } /* Find a variable in a subprogram die */ static void find_variable(Dwarf_Die *sp_die, struct probe_finder *pf) { - int ret; Dwarf_Die vr_die; /* TODO: Support struct members and arrays */ - if (!is_c_varname(pf->var)) { - /* Output raw parameters */ - ret = snprintf(pf->buf, pf->len, " %s", pf->var); - DIE_IF(ret < 0); - DIE_IF(ret >= pf->len); - return ; + if (!is_c_varname(pf->pvar->name)) { + /* Copy raw parameters */ + pf->tvar->value = xstrdup(pf->pvar->name); + } else { + pf->tvar->name = xstrdup(pf->pvar->name); + pr_debug("Searching '%s' variable in context.\n", + pf->pvar->name); + /* Search child die for local variables and parameters. */ + if (!die_find_variable(sp_die, pf->pvar->name, &vr_die)) + die("Failed to find '%s' in this function.", + pf->pvar->name); + convert_variable(&vr_die, pf); } - - pr_debug("Searching '%s' variable in context.\n", pf->var); - /* Search child die for local variables and parameters. */ - if (!die_find_variable(sp_die, pf->var, &vr_die)) - die("Failed to find '%s' in this function.", pf->var); - - show_variable(&vr_die, pf); } /* Show a probe point to output buffer */ -static void show_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf) +static void convert_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf) { - struct probe_point *pp = pf->pp; + struct kprobe_trace_event *tev; Dwarf_Addr eaddr; Dwarf_Die die_mem; const char *name; - char tmp[MAX_PROBE_BUFFER]; - int ret, i, len; + int ret, i; Dwarf_Attribute fb_attr; size_t nops; + if (pf->ntevs == MAX_PROBES) + die("Too many( > %d) probe point found.\n", MAX_PROBES); + tev = &pf->tevs[pf->ntevs++]; + /* If no real subprogram, find a real one */ if (!sp_die || dwarf_tag(sp_die) != DW_TAG_subprogram) { sp_die = die_find_real_subprogram(&pf->cu_die, @@ -431,31 +431,18 @@ static void show_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf) die("Probe point is not found in subprograms."); } - /* Output name of probe point */ + /* Copy the name of probe point */ name = dwarf_diename(sp_die); if (name) { dwarf_entrypc(sp_die, &eaddr); - ret = snprintf(tmp, MAX_PROBE_BUFFER, "%s+%lu", name, - (unsigned long)(pf->addr - eaddr)); - /* Copy the function name if possible */ - if (!pp->function) { - pp->function = xstrdup(name); - pp->offset = (size_t)(pf->addr - eaddr); - } - } else { + tev->point.symbol = xstrdup(name); + tev->point.offset = (unsigned long)(pf->addr - eaddr); + } else /* This function has no name. */ - ret = snprintf(tmp, MAX_PROBE_BUFFER, "0x%jx", - (uintmax_t)pf->addr); - if (!pp->function) { - /* TODO: Use _stext */ - pp->function = xstrdup(""); - pp->offset = (size_t)pf->addr; - } - } - DIE_IF(ret < 0); - DIE_IF(ret >= MAX_PROBE_BUFFER); - len = ret; - pr_debug("Probe point found: %s\n", tmp); + tev->point.offset = (unsigned long)pf->addr; + + pr_debug("Probe point found: %s+%lu\n", tev->point.symbol, + tev->point.offset); /* Get the frame base attribute/ops */ dwarf_attr(sp_die, DW_AT_frame_base, &fb_attr); @@ -465,22 +452,16 @@ static void show_probe_point(Dwarf_Die *sp_die, struct probe_finder *pf) /* Find each argument */ /* TODO: use dwarf_cfi_addrframe */ - for (i = 0; i < pp->nr_args; i++) { - pf->var = pp->args[i]; - pf->buf = &tmp[len]; - pf->len = MAX_PROBE_BUFFER - len; + tev->nargs = pf->pev->nargs; + tev->args = xzalloc(sizeof(struct kprobe_trace_arg) * tev->nargs); + for (i = 0; i < pf->pev->nargs; i++) { + pf->pvar = &pf->pev->args[i]; + pf->tvar = &tev->args[i]; find_variable(sp_die, pf); - len += strlen(pf->buf); } /* *pf->fb_ops will be cached in libdw. Don't free it. */ pf->fb_ops = NULL; - - if (pp->found == MAX_PROBES) - die("Too many( > %d) probe point found.\n", MAX_PROBES); - - pp->probes[pp->found] = xstrdup(tmp); - pp->found++; } /* Find probe point from its line number */ @@ -512,7 +493,7 @@ static void find_probe_point_by_line(struct probe_finder *pf) (int)i, lineno, (uintmax_t)addr); pf->addr = addr; - show_probe_point(NULL, pf); + convert_probe_point(NULL, pf); /* Continuing, because target line might be inlined. */ } } @@ -563,7 +544,7 @@ static void find_probe_point_lazy(Dwarf_Die *sp_die, struct probe_finder *pf) if (list_empty(&pf->lcache)) { /* Matching lazy line pattern */ ret = find_lazy_match_lines(&pf->lcache, pf->fname, - pf->pp->lazy_line); + pf->pev->point.lazy_line); if (ret <= 0) die("No matched lines found in %s.", pf->fname); } @@ -596,7 +577,7 @@ static void find_probe_point_lazy(Dwarf_Die *sp_die, struct probe_finder *pf) (int)i, lineno, (unsigned long long)addr); pf->addr = addr; - show_probe_point(sp_die, pf); + convert_probe_point(sp_die, pf); /* Continuing, because target line might be inlined. */ } /* TODO: deallocate lines, but how? */ @@ -605,7 +586,7 @@ static void find_probe_point_lazy(Dwarf_Die *sp_die, struct probe_finder *pf) static int probe_point_inline_cb(Dwarf_Die *in_die, void *data) { struct probe_finder *pf = (struct probe_finder *)data; - struct probe_point *pp = pf->pp; + struct perf_probe_point *pp = &pf->pev->point; if (pp->lazy_line) find_probe_point_lazy(in_die, pf); @@ -616,7 +597,7 @@ static int probe_point_inline_cb(Dwarf_Die *in_die, void *data) pr_debug("found inline addr: 0x%jx\n", (uintmax_t)pf->addr); - show_probe_point(in_die, pf); + convert_probe_point(in_die, pf); } return DWARF_CB_OK; @@ -626,7 +607,7 @@ static int probe_point_inline_cb(Dwarf_Die *in_die, void *data) static int probe_point_search_cb(Dwarf_Die *sp_die, void *data) { struct probe_finder *pf = (struct probe_finder *)data; - struct probe_point *pp = pf->pp; + struct perf_probe_point *pp = &pf->pev->point; /* Check tag and diename */ if (dwarf_tag(sp_die) != DW_TAG_subprogram || @@ -646,7 +627,7 @@ static int probe_point_search_cb(Dwarf_Die *sp_die, void *data) pf->addr = die_get_entrypc(sp_die); pf->addr += pp->offset; /* TODO: Check the address in this function */ - show_probe_point(sp_die, pf); + convert_probe_point(sp_die, pf); } } else /* Inlined function: search instances */ @@ -660,20 +641,25 @@ static void find_probe_point_by_func(struct probe_finder *pf) dwarf_getfuncs(&pf->cu_die, probe_point_search_cb, pf, 0); } -/* Find a probe point */ -int find_probe_point(int fd, struct probe_point *pp) +/* Find kprobe_trace_events specified by perf_probe_event from debuginfo */ +int find_kprobe_trace_events(int fd, struct perf_probe_event *pev, + struct kprobe_trace_event **tevs) { - struct probe_finder pf = {.pp = pp}; + struct probe_finder pf = {.pev = pev}; + struct perf_probe_point *pp = &pev->point; Dwarf_Off off, noff; size_t cuhl; Dwarf_Die *diep; Dwarf *dbg; + pf.tevs = xzalloc(sizeof(struct kprobe_trace_event) * MAX_PROBES); + *tevs = pf.tevs; + pf.ntevs = 0; + dbg = dwarf_begin(fd, DWARF_C_READ); if (!dbg) return -ENOENT; - pp->found = 0; off = 0; line_list__init(&pf.lcache); /* Loop on CUs (Compilation Unit) */ @@ -704,7 +690,7 @@ int find_probe_point(int fd, struct probe_point *pp) line_list__free(&pf.lcache); dwarf_end(dbg); - return pp->found; + return pf.ntevs; } /* Find line range from its line number */ diff --git a/tools/perf/util/probe-finder.h b/tools/perf/util/probe-finder.h index 21f7354..4949526 100644 --- a/tools/perf/util/probe-finder.h +++ b/tools/perf/util/probe-finder.h @@ -3,6 +3,7 @@ #include #include "util.h" +#include "probe-event.h" #define MAX_PATH_LEN 256 #define MAX_PROBE_BUFFER 1024 @@ -14,67 +15,32 @@ static inline int is_c_varname(const char *name) return isalpha(name[0]) || name[0] == '_'; } -struct probe_point { - char *event; /* Event name */ - char *group; /* Event group */ - - /* Inputs */ - char *file; /* File name */ - int line; /* Line number */ - char *lazy_line; /* Lazy line pattern */ - - char *function; /* Function name */ - int offset; /* Offset bytes */ - - int nr_args; /* Number of arguments */ - char **args; /* Arguments */ - - int retprobe; /* Return probe */ - - /* Output */ - int found; /* Number of found probe points */ - char *probes[MAX_PROBES]; /* Output buffers (will be allocated)*/ -}; - -/* Line number container */ -struct line_node { - struct list_head list; - unsigned int line; -}; - -/* Line range */ -struct line_range { - char *file; /* File name */ - char *function; /* Function name */ - unsigned int start; /* Start line number */ - unsigned int end; /* End line number */ - int offset; /* Start line offset */ - char *path; /* Real path name */ - struct list_head line_list; /* Visible lines */ -}; - #ifndef NO_DWARF_SUPPORT -extern int find_probe_point(int fd, struct probe_point *pp); +/* Find kprobe_trace_events specified by perf_probe_event from debuginfo */ +extern int find_kprobe_trace_events(int fd, struct perf_probe_event *pev, + struct kprobe_trace_event **tevs); + extern int find_line_range(int fd, struct line_range *lr); #include #include struct probe_finder { - struct probe_point *pp; /* Target probe point */ + struct perf_probe_event *pev; /* Target probe event */ + int ntevs; /* number of trace events */ + struct kprobe_trace_event *tevs; /* Result trace events */ /* For function searching */ Dwarf_Addr addr; /* Address */ - const char *fname; /* File name */ + const char *fname; /* Real file name */ int lno; /* Line number */ Dwarf_Die cu_die; /* Current CU */ + struct list_head lcache; /* Line cache for lazy match */ /* For variable searching */ Dwarf_Op *fb_ops; /* Frame base attribute */ - const char *var; /* Current variable name */ - char *buf; /* Current output buffer */ - int len; /* Length of output buffer */ - struct list_head lcache; /* Line cache for lazy match */ + struct perf_probe_arg *pvar; /* Current target variable */ + struct kprobe_trace_arg *tvar; /* Current result variable */ }; struct line_finder { -- cgit v0.10.2 From fb1587d869a399554220e166d4b90b581a8ade01 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 16 Mar 2010 18:06:19 -0400 Subject: perf probe: List probes with line number and file name Improve --list to show current exist probes with line number and file name. This enables user easily to check which line is already probed. for example: ./perf probe --list probe:vfs_read (on vfs_read:8@linux-2.6-tip/fs/read_write.c) Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <20100316220619.32050.48702.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index b44ddfb..4e3c1ae 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -70,7 +70,6 @@ static int e_snprintf(char *str, size_t size, const char *format, ...) return ret; } - static struct map_groups kmap_groups; static struct map *kmaps[MAP__NR_TYPES]; @@ -357,27 +356,39 @@ void parse_kprobe_trace_command(const char *cmd, struct kprobe_trace_event *tev) /* Compose only probe point (not argument) */ static char *synthesize_perf_probe_point(struct perf_probe_point *pp) { - char *buf; - char offs[64] = "", line[64] = ""; - int ret; + char *buf, *tmp; + char offs[32] = "", line[32] = "", file[32] = ""; + int ret, len; buf = xzalloc(MAX_CMDLEN); if (pp->offset) { - ret = e_snprintf(offs, 64, "+%lu", pp->offset); + ret = e_snprintf(offs, 32, "+%lu", pp->offset); if (ret <= 0) goto error; } if (pp->line) { - ret = e_snprintf(line, 64, ":%d", pp->line); + ret = e_snprintf(line, 32, ":%d", pp->line); + if (ret <= 0) + goto error; + } + if (pp->file) { + len = strlen(pp->file) - 32; + if (len < 0) + len = 0; + tmp = strchr(pp->file + len, '/'); + if (!tmp) + tmp = pp->file + len - 1; + ret = e_snprintf(file, 32, "@%s", tmp + 1); if (ret <= 0) goto error; } if (pp->function) - ret = e_snprintf(buf, MAX_CMDLEN, "%s%s%s%s", pp->function, - offs, pp->retprobe ? "%return" : "", line); + ret = e_snprintf(buf, MAX_CMDLEN, "%s%s%s%s%s", pp->function, + offs, pp->retprobe ? "%return" : "", line, + file); else - ret = e_snprintf(buf, MAX_CMDLEN, "%s%s", pp->file, line); + ret = e_snprintf(buf, MAX_CMDLEN, "%s%s", file, line); if (ret <= 0) goto error; @@ -511,15 +522,32 @@ void convert_to_perf_probe_event(struct kprobe_trace_event *tev, { char buf[64]; int i; +#ifndef NO_DWARF_SUPPORT + struct symbol *sym; + int fd, ret = 0; - pev->event = xstrdup(tev->event); - pev->group = xstrdup(tev->group); - + sym = map__find_symbol_by_name(kmaps[MAP__FUNCTION], + tev->point.symbol, NULL); + if (sym) { + fd = open_vmlinux(); + ret = find_perf_probe_point(fd, sym->start + tev->point.offset, + &pev->point); + close(fd); + } + if (ret <= 0) { + pev->point.function = xstrdup(tev->point.symbol); + pev->point.offset = tev->point.offset; + } +#else /* Convert trace_point to probe_point */ pev->point.function = xstrdup(tev->point.symbol); pev->point.offset = tev->point.offset; +#endif pev->point.retprobe = tev->point.retprobe; + pev->event = xstrdup(tev->event); + pev->group = xstrdup(tev->group); + /* Convert trace_arg to probe_arg */ pev->nargs = tev->nargs; pev->args = xzalloc(sizeof(struct perf_probe_arg) * pev->nargs); @@ -650,7 +678,7 @@ static void show_perf_probe_event(struct perf_probe_event *pev) ret = e_snprintf(buf, 128, "%s:%s", pev->group, pev->event); if (ret < 0) die("Failed to copy event: %s", strerror(-ret)); - printf(" %-40s (on %s", buf, place); + printf(" %-20s (on %s", buf, place); if (pev->nargs > 0) { printf(" with"); @@ -671,6 +699,7 @@ void show_perf_probe_events(void) struct str_node *ent; setup_pager(); + init_vmlinux(); memset(&tev, 0, sizeof(tev)); memset(&pev, 0, sizeof(pev)); diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index 251b4c4..e02b607 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -693,6 +693,76 @@ int find_kprobe_trace_events(int fd, struct perf_probe_event *pev, return pf.ntevs; } +/* Reverse search */ +int find_perf_probe_point(int fd, unsigned long addr, + struct perf_probe_point *ppt) +{ + Dwarf_Die cudie, spdie, indie; + Dwarf *dbg; + Dwarf_Line *line; + Dwarf_Addr laddr, eaddr; + const char *tmp; + int lineno, ret = 0; + + dbg = dwarf_begin(fd, DWARF_C_READ); + if (!dbg) + return -ENOENT; + + /* Find cu die */ + if (!dwarf_addrdie(dbg, (Dwarf_Addr)addr, &cudie)) + return -EINVAL; + + /* Find a corresponding line */ + line = dwarf_getsrc_die(&cudie, (Dwarf_Addr)addr); + if (line) { + dwarf_lineaddr(line, &laddr); + if ((Dwarf_Addr)addr == laddr) { + dwarf_lineno(line, &lineno); + ppt->line = lineno; + + tmp = dwarf_linesrc(line, NULL, NULL); + DIE_IF(!tmp); + ppt->file = xstrdup(tmp); + ret = 1; + } + } + + /* Find a corresponding function */ + if (die_find_real_subprogram(&cudie, (Dwarf_Addr)addr, &spdie)) { + tmp = dwarf_diename(&spdie); + if (!tmp) + goto end; + + dwarf_entrypc(&spdie, &eaddr); + if (!lineno) { + /* We don't have a line number, let's use offset */ + ppt->function = xstrdup(tmp); + ppt->offset = addr - (unsigned long)eaddr; + ret = 1; + goto end; + } + if (die_find_inlinefunc(&spdie, (Dwarf_Addr)addr, &indie)) { + /* addr in an inline function */ + tmp = dwarf_diename(&indie); + if (!tmp) + goto end; + dwarf_decl_line(&indie, &lineno); + } else { + if (eaddr == addr) /* No offset: function entry */ + lineno = ppt->line; + else + dwarf_decl_line(&spdie, &lineno); + } + ppt->function = xstrdup(tmp); + ppt->line -= lineno; /* Make a relative line number */ + } + +end: + dwarf_end(dbg); + return ret; +} + + /* Find line range from its line number */ static void find_line_range_by_line(Dwarf_Die *sp_die, struct line_finder *lf) { diff --git a/tools/perf/util/probe-finder.h b/tools/perf/util/probe-finder.h index 4949526..2f2307d 100644 --- a/tools/perf/util/probe-finder.h +++ b/tools/perf/util/probe-finder.h @@ -20,6 +20,10 @@ static inline int is_c_varname(const char *name) extern int find_kprobe_trace_events(int fd, struct perf_probe_event *pev, struct kprobe_trace_event **tevs); +/* Find a perf_probe_point from debuginfo */ +extern int find_perf_probe_point(int fd, unsigned long addr, + struct perf_probe_point *ppt); + extern int find_line_range(int fd, struct line_range *lr); #include -- cgit v0.10.2 From 7df2f32956cf0f1a45df38cd0e0fe0c3467580e8 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Tue, 16 Mar 2010 18:06:26 -0400 Subject: perf probe: Add data structure member access support Support accessing members in the data structures. With this, perf-probe accepts data-structure members(IOW, it now accepts dot '.' and arrow '->' operators) as probe arguemnts. e.g. ./perf probe --add 'schedule:44 rq->curr' ./perf probe --add 'vfs_read file->f_op->read file->f_path.dentry' Note that '>' can be interpreted as redirection in command-line. Signed-off-by: Masami Hiramatsu Cc: systemtap Cc: DLE Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <20100316220626.32050.57552.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index 4e3c1ae..64dea6c 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -262,6 +262,49 @@ static void parse_perf_probe_point(char *arg, struct perf_probe_event *pev) pp->lazy_line); } +/* Parse perf-probe event argument */ +static void parse_perf_probe_arg(const char *str, struct perf_probe_arg *arg) +{ + const char *tmp; + struct perf_probe_arg_field **fieldp; + + pr_debug("parsing arg: %s into ", str); + + tmp = strpbrk(str, "-."); + if (!is_c_varname(str) || !tmp) { + /* A variable, register, symbol or special value */ + arg->name = xstrdup(str); + pr_debug("%s\n", arg->name); + return; + } + + /* Structure fields */ + arg->name = xstrndup(str, tmp - str); + pr_debug("%s, ", arg->name); + fieldp = &arg->field; + + do { + *fieldp = xzalloc(sizeof(struct perf_probe_arg_field)); + if (*tmp == '.') { + str = tmp + 1; + (*fieldp)->ref = false; + } else if (tmp[1] == '>') { + str = tmp + 2; + (*fieldp)->ref = true; + } else + semantic_error("Argument parse error: %s", str); + + tmp = strpbrk(str, "-."); + if (tmp) { + (*fieldp)->name = xstrndup(str, tmp - str); + pr_debug("%s(%d), ", (*fieldp)->name, (*fieldp)->ref); + fieldp = &(*fieldp)->next; + } + } while (tmp); + (*fieldp)->name = xstrdup(str); + pr_debug("%s(%d)\n", (*fieldp)->name, (*fieldp)->ref); +} + /* Parse perf-probe event command */ void parse_perf_probe_command(const char *cmd, struct perf_probe_event *pev) { @@ -281,7 +324,7 @@ void parse_perf_probe_command(const char *cmd, struct perf_probe_event *pev) pev->nargs = argc - 1; pev->args = xzalloc(sizeof(struct perf_probe_arg) * pev->nargs); for (i = 0; i < pev->nargs; i++) { - pev->args[i].name = xstrdup(argv[i + 1]); + parse_perf_probe_arg(argv[i + 1], &pev->args[i]); if (is_c_varname(pev->args[i].name) && pev->point.retprobe) semantic_error("You can't specify local variable for" " kretprobe"); @@ -353,6 +396,33 @@ void parse_kprobe_trace_command(const char *cmd, struct kprobe_trace_event *tev) argv_free(argv); } +/* Compose only probe arg */ +int synthesize_perf_probe_arg(struct perf_probe_arg *pa, char *buf, size_t len) +{ + struct perf_probe_arg_field *field = pa->field; + int ret; + char *tmp = buf; + + ret = e_snprintf(tmp, len, "%s", pa->name); + if (ret <= 0) + goto error; + tmp += ret; + len -= ret; + + while (field) { + ret = e_snprintf(tmp, len, "%s%s", field->ref ? "->" : ".", + field->name); + if (ret <= 0) + goto error; + tmp += ret; + len -= ret; + field = field->next; + } + return tmp - buf; +error: + die("Failed to synthesize perf probe argument: %s", strerror(-ret)); +} + /* Compose only probe point (not argument) */ static char *synthesize_perf_probe_point(struct perf_probe_point *pp) { @@ -563,6 +633,7 @@ void convert_to_perf_probe_event(struct kprobe_trace_event *tev, void clear_perf_probe_event(struct perf_probe_event *pev) { struct perf_probe_point *pp = &pev->point; + struct perf_probe_arg_field *field, *next; int i; if (pev->event) @@ -575,9 +646,18 @@ void clear_perf_probe_event(struct perf_probe_event *pev) free(pp->function); if (pp->lazy_line) free(pp->lazy_line); - for (i = 0; i < pev->nargs; i++) + for (i = 0; i < pev->nargs; i++) { if (pev->args[i].name) free(pev->args[i].name); + field = pev->args[i].field; + while (field) { + next = field->next; + if (field->name) + free(field->name); + free(field); + field = next; + } + } if (pev->args) free(pev->args); memset(pev, 0, sizeof(*pev)); @@ -682,8 +762,10 @@ static void show_perf_probe_event(struct perf_probe_event *pev) if (pev->nargs > 0) { printf(" with"); - for (i = 0; i < pev->nargs; i++) - printf(" %s", pev->args[i].name); + for (i = 0; i < pev->nargs; i++) { + synthesize_perf_probe_arg(&pev->args[i], buf, 128); + printf(" %s", buf); + } } printf(")\n"); free(place); diff --git a/tools/perf/util/probe-event.h b/tools/perf/util/probe-event.h index 2a2f0a2..cd308b0 100644 --- a/tools/perf/util/probe-event.h +++ b/tools/perf/util/probe-event.h @@ -45,9 +45,17 @@ struct perf_probe_point { bool retprobe; /* Return probe flag */ }; +/* Perf probe probing argument field chain */ +struct perf_probe_arg_field { + struct perf_probe_arg_field *next; /* Next field */ + char *name; /* Name of the field */ + bool ref; /* Referencing flag */ +}; + /* Perf probe probing argument */ struct perf_probe_arg { - char *name; /* Argument name */ + char *name; /* Argument name */ + struct perf_probe_arg_field *field; /* Structure fields */ }; /* Perf probe probing event (point + arg) */ @@ -86,6 +94,8 @@ extern void parse_kprobe_trace_command(const char *cmd, /* Events to command string */ extern char *synthesize_perf_probe_command(struct perf_probe_event *pev); extern char *synthesize_kprobe_trace_command(struct kprobe_trace_event *tev); +extern int synthesize_perf_probe_arg(struct perf_probe_arg *pa, char *buf, + size_t len); /* Check the perf_probe_event needs debuginfo */ extern bool perf_probe_event_need_dwarf(struct perf_probe_event *pev); diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index e02b607..db52ec2 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -206,6 +206,28 @@ static Dwarf_Addr die_get_entrypc(Dwarf_Die *dw_die) return epc; } +/* Get type die, but skip qualifiers and typedef */ +static Dwarf_Die *die_get_real_type(Dwarf_Die *vr_die, Dwarf_Die *die_mem) +{ + Dwarf_Attribute attr; + int tag; + + do { + if (dwarf_attr(vr_die, DW_AT_type, &attr) == NULL || + dwarf_formref_die(&attr, die_mem) == NULL) + return NULL; + + tag = dwarf_tag(die_mem); + vr_die = die_mem; + } while (tag == DW_TAG_const_type || + tag == DW_TAG_restrict_type || + tag == DW_TAG_volatile_type || + tag == DW_TAG_shared_type || + tag == DW_TAG_typedef); + + return die_mem; +} + /* Return values for die_find callbacks */ enum { DIE_FIND_CB_FOUND = 0, /* End of Search */ @@ -314,6 +336,25 @@ static Dwarf_Die *die_find_variable(Dwarf_Die *sp_die, const char *name, die_mem); } +static int __die_find_member_cb(Dwarf_Die *die_mem, void *data) +{ + const char *name = data; + + if ((dwarf_tag(die_mem) == DW_TAG_member) && + (die_compare_name(die_mem, name) == 0)) + return DIE_FIND_CB_FOUND; + + return DIE_FIND_CB_SIBLING; +} + +/* Find a member called 'name' */ +static Dwarf_Die *die_find_member(Dwarf_Die *st_die, const char *name, + Dwarf_Die *die_mem) +{ + return die_find_child(st_die, __die_find_member_cb, (void *)name, + die_mem); +} + /* * Probe finder related functions */ @@ -363,6 +404,62 @@ static void convert_location(Dwarf_Op *op, struct probe_finder *pf) } } +static void convert_variable_fields(Dwarf_Die *vr_die, const char *varname, + struct perf_probe_arg_field *field, + struct kprobe_trace_arg_ref **ref_ptr) +{ + struct kprobe_trace_arg_ref *ref = *ref_ptr; + Dwarf_Attribute attr; + Dwarf_Die member; + Dwarf_Die type; + Dwarf_Word offs; + + pr_debug("converting %s in %s\n", field->name, varname); + if (die_get_real_type(vr_die, &type) == NULL) + die("Failed to get a type information of %s.", varname); + + /* Check the pointer and dereference */ + if (dwarf_tag(&type) == DW_TAG_pointer_type) { + if (!field->ref) + die("Semantic error: %s must be referred by '->'", + field->name); + /* Get the type pointed by this pointer */ + if (die_get_real_type(&type, &type) == NULL) + die("Failed to get a type information of %s.", varname); + + ref = xzalloc(sizeof(struct kprobe_trace_arg_ref)); + if (*ref_ptr) + (*ref_ptr)->next = ref; + else + *ref_ptr = ref; + } else { + if (field->ref) + die("Semantic error: %s must be referred by '.'", + field->name); + if (!ref) + die("Structure on a register is not supported yet."); + } + + /* Verify it is a data structure */ + if (dwarf_tag(&type) != DW_TAG_structure_type) + die("%s is not a data structure.", varname); + + if (die_find_member(&type, field->name, &member) == NULL) + die("%s(tyep:%s) has no member %s.", varname, + dwarf_diename(&type), field->name); + + /* Get the offset of the field */ + if (dwarf_attr(&member, DW_AT_data_member_location, &attr) == NULL || + dwarf_formudata(&attr, &offs) != 0) + die("Failed to get the offset of %s.", field->name); + ref->offset += (long)offs; + + /* Converting next field */ + if (field->next) + convert_variable_fields(&member, field->name, field->next, + &ref); +} + /* Show a variables in kprobe event format */ static void convert_variable(Dwarf_Die *vr_die, struct probe_finder *pf) { @@ -379,6 +476,10 @@ static void convert_variable(Dwarf_Die *vr_die, struct probe_finder *pf) goto error; convert_location(expr, pf); + + if (pf->pvar->field) + convert_variable_fields(vr_die, pf->pvar->name, + pf->pvar->field, &pf->tvar->ref); /* *expr will be cached in libdw. Don't free it. */ return ; error: @@ -391,13 +492,15 @@ error: static void find_variable(Dwarf_Die *sp_die, struct probe_finder *pf) { Dwarf_Die vr_die; + char buf[128]; /* TODO: Support struct members and arrays */ if (!is_c_varname(pf->pvar->name)) { /* Copy raw parameters */ pf->tvar->value = xstrdup(pf->pvar->name); } else { - pf->tvar->name = xstrdup(pf->pvar->name); + synthesize_perf_probe_arg(pf->pvar, buf, 128); + pf->tvar->name = xstrdup(buf); pr_debug("Searching '%s' variable in context.\n", pf->pvar->name); /* Search child die for local variables and parameters. */ -- cgit v0.10.2 From 3b0d516463f8deb897a55cb81e9dbbe58a2490ed Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Wed, 17 Mar 2010 12:13:28 +0100 Subject: perf probe: Fix !dwarf build Fix the !drawf build. This uses the existing NO_DWARF_SUPPORT mechanism we use for that, but it's really fragile and needs a cleanup. (in a separate patch) 1) Such uses: #ifndef NO_DWARF_SUPPORT are double inverted logic a'la 'not not'. Instead the flag should be called DWARF_SUPPORT. 2) Furthermore, assymetric #ifdef polluted code flow like: if (need_dwarf) #ifdef NO_DWARF_SUPPORT die("Debuginfo-analysis is not supported"); #else /* !NO_DWARF_SUPPORT */ pr_debug("Some probes require debuginfo.\n"); fd = open_vmlinux(); is very fragile and not acceptable. Instead of that helper functions should be created and the dwarf/no-dwarf logic should be separated more cleanly. 3) Local variable #ifdefs like this: #ifndef NO_DWARF_SUPPORT int fd; #endif Are fragile as well and should be eliminated. Helper functions achieve that too. Cc: Masami Hiramatsu Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <20100316220612.32050.33806.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index 64dea6c..f333269 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -1012,8 +1012,10 @@ end_dwarf: if (!sym) die("Kernel symbol \'%s\' not found - probe not added.", tev->point.symbol); +#ifndef NO_DWARF_SUPPORT found: close(fd); +#endif return ntevs; } @@ -1172,10 +1174,13 @@ void show_line_range(struct line_range *lr) unsigned int l = 1; struct line_node *ln; FILE *fp; +#ifndef NO_DWARF_SUPPORT int fd, ret; +#endif /* Search a line range */ init_vmlinux(); +#ifndef NO_DWARF_SUPPORT fd = open_vmlinux(); if (fd < 0) die("Could not open debuginfo file."); @@ -1183,6 +1188,7 @@ void show_line_range(struct line_range *lr) if (ret <= 0) die("Source line is not found.\n"); close(fd); +#endif setup_pager(); -- cgit v0.10.2 From b27ea29c6267889be255f2217fa7a6106e6a8b04 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Wed, 17 Mar 2010 12:49:10 +0100 Subject: perf/core, x86: Reduce number of CONFIG_X86_LOCAL_APIC macros The function reserve_pmc_hardware() and release_pmc_hardware() were hard to read. This patch improves readability of the code by removing most of the CONFIG_X86_LOCAL_APIC macros. Signed-off-by: Robert Richter Cc: Stephane Eranian Cc: Peter Zijlstra LKML-Reference: <1268826553-19518-2-git-send-email-robert.richter@amd.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 5dacf63..793e63f 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -313,9 +313,10 @@ again: static atomic_t active_events; static DEFINE_MUTEX(pmc_reserve_mutex); +#ifdef CONFIG_X86_LOCAL_APIC + static bool reserve_pmc_hardware(void) { -#ifdef CONFIG_X86_LOCAL_APIC int i; if (nmi_watchdog == NMI_LOCAL_APIC) @@ -330,11 +331,9 @@ static bool reserve_pmc_hardware(void) if (!reserve_evntsel_nmi(x86_pmu.eventsel + i)) goto eventsel_fail; } -#endif return true; -#ifdef CONFIG_X86_LOCAL_APIC eventsel_fail: for (i--; i >= 0; i--) release_evntsel_nmi(x86_pmu.eventsel + i); @@ -349,12 +348,10 @@ perfctr_fail: enable_lapic_nmi_watchdog(); return false; -#endif } static void release_pmc_hardware(void) { -#ifdef CONFIG_X86_LOCAL_APIC int i; for (i = 0; i < x86_pmu.num_events; i++) { @@ -364,9 +361,15 @@ static void release_pmc_hardware(void) if (nmi_watchdog == NMI_LOCAL_APIC) enable_lapic_nmi_watchdog(); -#endif } +#else + +static bool reserve_pmc_hardware(void) { return true; } +static void release_pmc_hardware(void) {} + +#endif + static int reserve_ds_buffers(void); static void release_ds_buffers(void); -- cgit v0.10.2 From 141c4296cb630a7ed4c3730913bc3c0617ef9753 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Wed, 17 Mar 2010 12:49:11 +0100 Subject: perf/core: Correct files in MAINTAINERS entry This corrects the file entries for perf_events. The following files are caught now: $ xargs | eval ls $(cat) | sort -u kernel/perf_event*.c include/linux/perf_event.h arch/*/kernel/perf_event*.c arch/*/kernel/*/perf_event*.c arch/*/kernel/*/*/perf_event*.c arch/*/include/asm/perf_event.h arch/*/lib/perf_event*.c arch/*/kernel/perf_callchain.c arch/alpha/include/asm/perf_event.h arch/arm/include/asm/perf_event.h arch/arm/kernel/perf_event.c arch/frv/include/asm/perf_event.h arch/frv/lib/perf_event.c arch/parisc/include/asm/perf_event.h arch/powerpc/include/asm/perf_event.h arch/powerpc/kernel/perf_callchain.c arch/powerpc/kernel/perf_event.c arch/s390/include/asm/perf_event.h arch/sh/include/asm/perf_event.h arch/sh/kernel/cpu/sh4a/perf_event.c arch/sh/kernel/cpu/sh4/perf_event.c arch/sh/kernel/perf_callchain.c arch/sh/kernel/perf_event.c arch/sparc/include/asm/perf_event.h arch/sparc/kernel/perf_event.c arch/x86/include/asm/perf_event.h arch/x86/kernel/cpu/perf_event_amd.c arch/x86/kernel/cpu/perf_event.c arch/x86/kernel/cpu/perf_event_intel.c arch/x86/kernel/cpu/perf_event_p6.c include/linux/perf_event.h kernel/perf_event.c Signed-off-by: Robert Richter Cc: Stephane Eranian Cc: Peter Zijlstra LKML-Reference: <1268826553-19518-3-git-send-email-robert.richter@amd.com> Signed-off-by: Ingo Molnar diff --git a/MAINTAINERS b/MAINTAINERS index 14d5c4c..e3aa850 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -4295,13 +4295,13 @@ M: Paul Mackerras M: Ingo Molnar M: Arnaldo Carvalho de Melo S: Supported -F: kernel/perf_event.c +F: kernel/perf_event*.c F: include/linux/perf_event.h -F: arch/*/kernel/perf_event.c -F: arch/*/kernel/*/perf_event.c -F: arch/*/kernel/*/*/perf_event.c +F: arch/*/kernel/perf_event*.c +F: arch/*/kernel/*/perf_event*.c +F: arch/*/kernel/*/*/perf_event*.c F: arch/*/include/asm/perf_event.h -F: arch/*/lib/perf_event.c +F: arch/*/lib/perf_event*.c F: arch/*/kernel/perf_callchain.c F: tools/perf/ -- cgit v0.10.2 From 10f1014d86fd4fe5087080d609b51183396c5e4c Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Wed, 17 Mar 2010 12:49:12 +0100 Subject: perf/core, x86: Remove cpu_hw_events.interrupts This member in the struct is not used anymore and can be removed. Signed-off-by: Robert Richter Cc: Stephane Eranian Cc: Peter Zijlstra LKML-Reference: <1268826553-19518-4-git-send-email-robert.richter@amd.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 793e63f..104292a 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -102,7 +102,6 @@ struct cpu_hw_events { */ struct perf_event *events[X86_PMC_IDX_MAX]; /* in counter order */ unsigned long active_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; - unsigned long interrupts; int enabled; int n_events; -- cgit v0.10.2 From d6dc0b4ead6e8720096ecfa3d9e899b47ddbc8ed Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Wed, 17 Mar 2010 12:49:13 +0100 Subject: perf/core, x86: Remove duplicate perf_event_mask variable The same information is stored also in x86_pmu.intel_ctrl. This patch removes perf_event_mask and instead uses x86_pmu.intel_ctrl directly. Signed-off-by: Robert Richter Cc: Stephane Eranian Cc: Peter Zijlstra LKML-Reference: <1268826553-19518-5-git-send-email-robert.richter@amd.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 104292a..c97d5b5 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -75,8 +75,6 @@ copy_from_user_nmi(void *to, const void __user *from, unsigned long n) return len; } -static u64 perf_event_mask __read_mostly; - struct event_constraint { union { unsigned long idxmsk[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; @@ -1406,7 +1404,7 @@ void __init init_hw_perf_events(void) x86_pmu.num_events, X86_PMC_MAX_GENERIC); x86_pmu.num_events = X86_PMC_MAX_GENERIC; } - perf_event_mask = (1 << x86_pmu.num_events) - 1; + x86_pmu.intel_ctrl = (1 << x86_pmu.num_events) - 1; perf_max_events = x86_pmu.num_events; if (x86_pmu.num_events_fixed > X86_PMC_MAX_FIXED) { @@ -1415,9 +1413,8 @@ void __init init_hw_perf_events(void) x86_pmu.num_events_fixed = X86_PMC_MAX_FIXED; } - perf_event_mask |= + x86_pmu.intel_ctrl |= ((1LL << x86_pmu.num_events_fixed)-1) << X86_PMC_IDX_FIXED; - x86_pmu.intel_ctrl = perf_event_mask; perf_events_lapic_init(); register_die_notifier(&perf_event_nmi_notifier); @@ -1442,7 +1439,7 @@ void __init init_hw_perf_events(void) pr_info("... value mask: %016Lx\n", x86_pmu.event_mask); pr_info("... max period: %016Lx\n", x86_pmu.max_period); pr_info("... fixed-purpose events: %d\n", x86_pmu.num_events_fixed); - pr_info("... event mask: %016Lx\n", perf_event_mask); + pr_info("... event mask: %016Lx\n", x86_pmu.intel_ctrl); perf_cpu_notifier(x86_pmu_notifier); } -- cgit v0.10.2 From 2acebe9ecb2b77876e87a1480729cfb2db4570dd Mon Sep 17 00:00:00 2001 From: Jack Steiner Date: Wed, 17 Mar 2010 10:40:38 -0500 Subject: x86, UV: Delete unneeded boot messages SGI:UV: Delete extra boot messages that describe the system topology. These messages are no longer useful. Signed-off-by: Jack Steiner LKML-Reference: <20100317154038.GA29346@sgi.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c index 3740c8a..9a93b07 100644 --- a/arch/x86/kernel/apic/x2apic_uv_x.c +++ b/arch/x86/kernel/apic/x2apic_uv_x.c @@ -736,9 +736,6 @@ void __init uv_system_init(void) uv_node_to_blade[nid] = blade; uv_cpu_to_blade[cpu] = blade; max_pnode = max(pnode, max_pnode); - - printk(KERN_DEBUG "UV: cpu %d, apicid 0x%x, pnode %d, nid %d, lcpu %d, blade %d\n", - cpu, apicid, pnode, nid, lcpu, blade); } /* Add blade/pnode info for nodes without cpus */ -- cgit v0.10.2 From dadf28a10c3eb29421837a2e413ab869ebd9e168 Mon Sep 17 00:00:00 2001 From: Alexey Starikovskiy Date: Wed, 17 Mar 2010 13:14:13 -0400 Subject: ACPI: EC: Allow multibyte access to EC http://bugzilla.kernel.org/show_bug.cgi?id=14667 Signed-off-by: Alexey Starikovskiy Signed-off-by: Len Brown diff --git a/drivers/acpi/acpica/exprep.c b/drivers/acpi/acpica/exprep.c index edf62bf..a610ebe 100644 --- a/drivers/acpi/acpica/exprep.c +++ b/drivers/acpi/acpica/exprep.c @@ -468,6 +468,18 @@ acpi_status acpi_ex_prep_field_value(struct acpi_create_field_info *info) acpi_ut_add_reference(obj_desc->field.region_obj); + /* allow full data read from EC address space */ + if (obj_desc->field.region_obj->region.space_id == + ACPI_ADR_SPACE_EC) { + if (obj_desc->common_field.bit_length > 8) + obj_desc->common_field.access_bit_width = + ACPI_ROUND_UP(obj_desc->common_field. + bit_length, 8); + obj_desc->common_field.access_byte_width = + ACPI_DIV_8(obj_desc->common_field. + access_bit_width); + } + ACPI_DEBUG_PRINT((ACPI_DB_BFIELD, "RegionField: BitOff %X, Off %X, Gran %X, Region %p\n", obj_desc->field.start_field_bit_offset, diff --git a/drivers/acpi/ec.c b/drivers/acpi/ec.c index 1ac28c6..7208a69 100644 --- a/drivers/acpi/ec.c +++ b/drivers/acpi/ec.c @@ -628,12 +628,12 @@ static u32 acpi_ec_gpe_handler(void *data) static acpi_status acpi_ec_space_handler(u32 function, acpi_physical_address address, - u32 bits, u64 *value, + u32 bits, u64 *value64, void *handler_context, void *region_context) { struct acpi_ec *ec = handler_context; - int result = 0, i; - u8 temp = 0; + int result = 0, i, bytes = bits / 8; + u8 *value = (u8 *)value64; if ((address > 0xFF) || !value || !handler_context) return AE_BAD_PARAMETER; @@ -641,32 +641,15 @@ acpi_ec_space_handler(u32 function, acpi_physical_address address, if (function != ACPI_READ && function != ACPI_WRITE) return AE_BAD_PARAMETER; - if (bits != 8 && acpi_strict) - return AE_BAD_PARAMETER; - - if (EC_FLAGS_MSI) + if (EC_FLAGS_MSI || bits > 8) acpi_ec_burst_enable(ec); - if (function == ACPI_READ) { - result = acpi_ec_read(ec, address, &temp); - *value = temp; - } else { - temp = 0xff & (*value); - result = acpi_ec_write(ec, address, temp); - } - - for (i = 8; unlikely(bits - i > 0); i += 8) { - ++address; - if (function == ACPI_READ) { - result = acpi_ec_read(ec, address, &temp); - (*value) |= ((u64)temp) << i; - } else { - temp = 0xff & ((*value) >> i); - result = acpi_ec_write(ec, address, temp); - } - } + for (i = 0; i < bytes; ++i, ++address, ++value) + result = (function == ACPI_READ) ? + acpi_ec_read(ec, address, value) : + acpi_ec_write(ec, address, *value); - if (EC_FLAGS_MSI) + if (EC_FLAGS_MSI || bits > 8) acpi_ec_burst_disable(ec); switch (result) { -- cgit v0.10.2 From 6be2850effd6a8bae11d623c8c52e88d2fbc0e96 Mon Sep 17 00:00:00 2001 From: "Zhang, Yanmin" Date: Thu, 18 Mar 2010 11:36:03 -0300 Subject: perf stat: Enable counters when collecting process-wide or system-wide data Command 'perf stat' doesn't enable counters when collecting an existing (by -p) process or system-wide statistics. Fix the issue. Change the condition of fork/exec subcommand. If there is a subcommand parameter, perf always forks/execs it. The usage example is: # perf stat -a sleep 10 So this command could collect statistics for 10 seconds precisely. User still could stop it by CTRL+C. Without the new capability, user could only use CTRL+C to stop it without precise time clock. Another issue is 'perf stat -a' consumes 100% time of a full single logical cpu. It has a bad impact on running workload. Fix it by adding a sleep(1) in the while(!done) loop in function run_perf_stat. Signed-off-by: Zhang Yanmin Signed-off-by: Arnaldo Carvalho de Melo Cc: Avi Kivity Cc: Peter Zijlstra Cc: Sheng Yang Cc: Marcelo Tosatti Cc: Joerg Roedel Cc: Jes Sorensen Cc: Gleb Natapov Cc: Zachary Amsden Cc: LKML-Reference: <1268922965-14774-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c index 95db31c..5f41244 100644 --- a/tools/perf/builtin-stat.c +++ b/tools/perf/builtin-stat.c @@ -159,8 +159,10 @@ static void create_perf_stat_counter(int counter, int pid) } } else { attr->inherit = inherit; - attr->disabled = 1; - attr->enable_on_exec = 1; + if (target_pid == -1) { + attr->disabled = 1; + attr->enable_on_exec = 1; + } fd[0][counter] = sys_perf_event_open(attr, pid, -1, -1, 0); if (fd[0][counter] < 0 && verbose) @@ -251,9 +253,9 @@ static int run_perf_stat(int argc __used, const char **argv) unsigned long long t0, t1; int status = 0; int counter; - int pid = target_pid; + int pid; int child_ready_pipe[2], go_pipe[2]; - const bool forks = (target_pid == -1 && argc > 0); + const bool forks = (argc > 0); char buf; if (!system_wide) @@ -265,10 +267,10 @@ static int run_perf_stat(int argc __used, const char **argv) } if (forks) { - if ((pid = fork()) < 0) + if ((child_pid = fork()) < 0) perror("failed to fork"); - if (!pid) { + if (!child_pid) { close(child_ready_pipe[0]); close(go_pipe[1]); fcntl(go_pipe[0], F_SETFD, FD_CLOEXEC); @@ -297,8 +299,6 @@ static int run_perf_stat(int argc __used, const char **argv) exit(-1); } - child_pid = pid; - /* * Wait for the child to be ready to exec. */ @@ -309,6 +309,10 @@ static int run_perf_stat(int argc __used, const char **argv) close(child_ready_pipe[0]); } + if (target_pid == -1) + pid = child_pid; + else + pid = target_pid; for (counter = 0; counter < nr_counters; counter++) create_perf_stat_counter(counter, pid); @@ -321,7 +325,7 @@ static int run_perf_stat(int argc __used, const char **argv) close(go_pipe[1]); wait(&status); } else { - while(!done); + while(!done) sleep(1); } t1 = rdclock(); @@ -459,7 +463,7 @@ static volatile int signr = -1; static void skip_signal(int signo) { - if(target_pid != -1) + if(child_pid == -1) done = 1; signr = signo; -- cgit v0.10.2 From 46be604b5ba738d53e5f5314813a4e7092864baf Mon Sep 17 00:00:00 2001 From: "Zhang, Yanmin" Date: Thu, 18 Mar 2010 11:36:04 -0300 Subject: perf record: Enable counters only when kernel is execing subcommand 'perf record' starts counters before subcommand is execed, so the statistics is not precise because it includes data of some preparation steps. I fix it with the patch. In addition, change the condition to fork/exec subcommand. If there is a subcommand parameter, perf always fork/exec it. The usage example is: # perf record -f -a sleep 10 So this command could collect statistics for 10 seconds precisely. User still could stop it by CTRL+C. Without the new capability, user could only input CTRL+C to stop it without precise time clock. Signed-off-by: Zhang Yanmin Signed-off-by: Arnaldo Carvalho de Melo Cc: Avi Kivity Cc: Peter Zijlstra Cc: Sheng Yang Cc: oerg Roedel Cc: Jes Sorensen Cc: Marcelo Tosatti Cc: Gleb Natapov Cc: Cc: Zachary Amsden LKML-Reference: <1268922965-14774-2-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 962cdbf..e2b35ad 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -225,7 +225,7 @@ static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int n return h_attr; } -static void create_counter(int counter, int cpu, pid_t pid, bool forks) +static void create_counter(int counter, int cpu, pid_t pid) { char *filter = filters[counter]; struct perf_event_attr *attr = attrs + counter; @@ -275,10 +275,10 @@ static void create_counter(int counter, int cpu, pid_t pid, bool forks) attr->mmap = track; attr->comm = track; attr->inherit = inherit; - attr->disabled = 1; - - if (forks) + if (target_pid == -1 && !system_wide) { + attr->disabled = 1; attr->enable_on_exec = 1; + } try_again: fd[nr_cpu][counter] = sys_perf_event_open(attr, pid, cpu, group_fd, 0); @@ -380,17 +380,15 @@ try_again: exit(-1); } } - - ioctl(fd[nr_cpu][counter], PERF_EVENT_IOC_ENABLE); } -static void open_counters(int cpu, pid_t pid, bool forks) +static void open_counters(int cpu, pid_t pid) { int counter; group_fd = -1; for (counter = 0; counter < nr_counters; counter++) - create_counter(counter, cpu, pid, forks); + create_counter(counter, cpu, pid); nr_cpu++; } @@ -425,7 +423,7 @@ static int __cmd_record(int argc, const char **argv) int err; unsigned long waking = 0; int child_ready_pipe[2], go_pipe[2]; - const bool forks = target_pid == -1 && argc > 0; + const bool forks = argc > 0; char buf; page_size = sysconf(_SC_PAGE_SIZE); @@ -496,13 +494,13 @@ static int __cmd_record(int argc, const char **argv) atexit(atexit_header); if (forks) { - pid = fork(); + child_pid = fork(); if (pid < 0) { perror("failed to fork"); exit(-1); } - if (!pid) { + if (!child_pid) { close(child_ready_pipe[0]); close(go_pipe[1]); fcntl(go_pipe[0], F_SETFD, FD_CLOEXEC); @@ -531,11 +529,6 @@ static int __cmd_record(int argc, const char **argv) exit(-1); } - child_pid = pid; - - if (!system_wide) - target_pid = pid; - close(child_ready_pipe[1]); close(go_pipe[0]); /* @@ -548,13 +541,17 @@ static int __cmd_record(int argc, const char **argv) close(child_ready_pipe[0]); } + if (forks && target_pid == -1 && !system_wide) + pid = child_pid; + else + pid = target_pid; if ((!system_wide && !inherit) || profile_cpu != -1) { - open_counters(profile_cpu, target_pid, forks); + open_counters(profile_cpu, pid); } else { nr_cpus = read_cpu_map(); for (i = 0; i < nr_cpus; i++) - open_counters(cpumap[i], target_pid, forks); + open_counters(cpumap[i], pid); } if (file_new) { -- cgit v0.10.2 From d6d901c23a9c4c7361aa901b5b2dda69703dd5e0 Mon Sep 17 00:00:00 2001 From: "Zhang, Yanmin" Date: Thu, 18 Mar 2010 11:36:05 -0300 Subject: perf events: Change perf parameter --pid to process-wide collection instead of thread-wide Parameter --pid (or -p) of perf currently means a thread-wide collection. For exmaple, if a process whose id is 8888 has 10 threads, 'perf top -p 8888' just collects the main thread statistics. That's misleading. Users are used to attach a whole process when debugging a process by gdb. To follow normal usage style, the patch change --pid to process-wide collection and add --tid (-t) to mean a thread-wide collection. Usage example is: # perf top -p 8888 # perf record -p 8888 -f sleep 10 # perf stat -p 8888 -f sleep 10 Above commands collect the statistics of all threads of process 8888. Signed-off-by: Zhang Yanmin Signed-off-by: Arnaldo Carvalho de Melo Cc: Avi Kivity Cc: Peter Zijlstra Cc: Sheng Yang Cc: Joerg Roedel Cc: Jes Sorensen Cc: Marcelo Tosatti Cc: Gleb Natapov Cc: zhiteng.huang@intel.com Cc: Zachary Amsden LKML-Reference: <1268922965-14774-3-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index e2b35ad..bb5b23d 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -27,7 +27,7 @@ #include #include -static int fd[MAX_NR_CPUS][MAX_COUNTERS]; +static int *fd[MAX_NR_CPUS][MAX_COUNTERS]; static long default_interval = 0; @@ -43,6 +43,9 @@ static int raw_samples = 0; static int system_wide = 0; static int profile_cpu = -1; static pid_t target_pid = -1; +static pid_t target_tid = -1; +static pid_t *all_tids = NULL; +static int thread_num = 0; static pid_t child_pid = -1; static int inherit = 1; static int force = 0; @@ -60,7 +63,7 @@ static struct timeval this_read; static u64 bytes_written = 0; -static struct pollfd event_array[MAX_NR_CPUS * MAX_COUNTERS]; +static struct pollfd *event_array; static int nr_poll = 0; static int nr_cpu = 0; @@ -77,7 +80,7 @@ struct mmap_data { unsigned int prev; }; -static struct mmap_data mmap_array[MAX_NR_CPUS][MAX_COUNTERS]; +static struct mmap_data *mmap_array[MAX_NR_CPUS][MAX_COUNTERS]; static unsigned long mmap_read_head(struct mmap_data *md) { @@ -225,12 +228,13 @@ static struct perf_header_attr *get_header_attr(struct perf_event_attr *a, int n return h_attr; } -static void create_counter(int counter, int cpu, pid_t pid) +static void create_counter(int counter, int cpu) { char *filter = filters[counter]; struct perf_event_attr *attr = attrs + counter; struct perf_header_attr *h_attr; int track = !counter; /* only the first counter needs these */ + int thread_index; int ret; struct { u64 count; @@ -280,115 +284,124 @@ static void create_counter(int counter, int cpu, pid_t pid) attr->enable_on_exec = 1; } + for (thread_index = 0; thread_index < thread_num; thread_index++) { try_again: - fd[nr_cpu][counter] = sys_perf_event_open(attr, pid, cpu, group_fd, 0); - - if (fd[nr_cpu][counter] < 0) { - int err = errno; - - if (err == EPERM || err == EACCES) - die("Permission error - are you root?\n" - "\t Consider tweaking /proc/sys/kernel/perf_event_paranoid.\n"); - else if (err == ENODEV && profile_cpu != -1) - die("No such device - did you specify an out-of-range profile CPU?\n"); + fd[nr_cpu][counter][thread_index] = sys_perf_event_open(attr, + all_tids[thread_index], cpu, group_fd, 0); + + if (fd[nr_cpu][counter][thread_index] < 0) { + int err = errno; + + if (err == EPERM || err == EACCES) + die("Permission error - are you root?\n" + "\t Consider tweaking" + " /proc/sys/kernel/perf_event_paranoid.\n"); + else if (err == ENODEV && profile_cpu != -1) { + die("No such device - did you specify" + " an out-of-range profile CPU?\n"); + } - /* - * If it's cycles then fall back to hrtimer - * based cpu-clock-tick sw counter, which - * is always available even if no PMU support: - */ - if (attr->type == PERF_TYPE_HARDWARE - && attr->config == PERF_COUNT_HW_CPU_CYCLES) { - - if (verbose) - warning(" ... trying to fall back to cpu-clock-ticks\n"); - attr->type = PERF_TYPE_SOFTWARE; - attr->config = PERF_COUNT_SW_CPU_CLOCK; - goto try_again; - } - printf("\n"); - error("perfcounter syscall returned with %d (%s)\n", - fd[nr_cpu][counter], strerror(err)); + /* + * If it's cycles then fall back to hrtimer + * based cpu-clock-tick sw counter, which + * is always available even if no PMU support: + */ + if (attr->type == PERF_TYPE_HARDWARE + && attr->config == PERF_COUNT_HW_CPU_CYCLES) { + + if (verbose) + warning(" ... trying to fall back to cpu-clock-ticks\n"); + attr->type = PERF_TYPE_SOFTWARE; + attr->config = PERF_COUNT_SW_CPU_CLOCK; + goto try_again; + } + printf("\n"); + error("perfcounter syscall returned with %d (%s)\n", + fd[nr_cpu][counter][thread_index], strerror(err)); #if defined(__i386__) || defined(__x86_64__) - if (attr->type == PERF_TYPE_HARDWARE && err == EOPNOTSUPP) - die("No hardware sampling interrupt available. No APIC? If so then you can boot the kernel with the \"lapic\" boot parameter to force-enable it.\n"); + if (attr->type == PERF_TYPE_HARDWARE && err == EOPNOTSUPP) + die("No hardware sampling interrupt available." + " No APIC? If so then you can boot the kernel" + " with the \"lapic\" boot parameter to" + " force-enable it.\n"); #endif - die("No CONFIG_PERF_EVENTS=y kernel support configured?\n"); - exit(-1); - } + die("No CONFIG_PERF_EVENTS=y kernel support configured?\n"); + exit(-1); + } - h_attr = get_header_attr(attr, counter); - if (h_attr == NULL) - die("nomem\n"); + h_attr = get_header_attr(attr, counter); + if (h_attr == NULL) + die("nomem\n"); - if (!file_new) { - if (memcmp(&h_attr->attr, attr, sizeof(*attr))) { - fprintf(stderr, "incompatible append\n"); - exit(-1); + if (!file_new) { + if (memcmp(&h_attr->attr, attr, sizeof(*attr))) { + fprintf(stderr, "incompatible append\n"); + exit(-1); + } } - } - if (read(fd[nr_cpu][counter], &read_data, sizeof(read_data)) == -1) { - perror("Unable to read perf file descriptor\n"); - exit(-1); - } + if (read(fd[nr_cpu][counter][thread_index], &read_data, sizeof(read_data)) == -1) { + perror("Unable to read perf file descriptor\n"); + exit(-1); + } - if (perf_header_attr__add_id(h_attr, read_data.id) < 0) { - pr_warning("Not enough memory to add id\n"); - exit(-1); - } + if (perf_header_attr__add_id(h_attr, read_data.id) < 0) { + pr_warning("Not enough memory to add id\n"); + exit(-1); + } - assert(fd[nr_cpu][counter] >= 0); - fcntl(fd[nr_cpu][counter], F_SETFL, O_NONBLOCK); + assert(fd[nr_cpu][counter][thread_index] >= 0); + fcntl(fd[nr_cpu][counter][thread_index], F_SETFL, O_NONBLOCK); - /* - * First counter acts as the group leader: - */ - if (group && group_fd == -1) - group_fd = fd[nr_cpu][counter]; - if (multiplex && multiplex_fd == -1) - multiplex_fd = fd[nr_cpu][counter]; + /* + * First counter acts as the group leader: + */ + if (group && group_fd == -1) + group_fd = fd[nr_cpu][counter][thread_index]; + if (multiplex && multiplex_fd == -1) + multiplex_fd = fd[nr_cpu][counter][thread_index]; - if (multiplex && fd[nr_cpu][counter] != multiplex_fd) { + if (multiplex && fd[nr_cpu][counter][thread_index] != multiplex_fd) { - ret = ioctl(fd[nr_cpu][counter], PERF_EVENT_IOC_SET_OUTPUT, multiplex_fd); - assert(ret != -1); - } else { - event_array[nr_poll].fd = fd[nr_cpu][counter]; - event_array[nr_poll].events = POLLIN; - nr_poll++; - - mmap_array[nr_cpu][counter].counter = counter; - mmap_array[nr_cpu][counter].prev = 0; - mmap_array[nr_cpu][counter].mask = mmap_pages*page_size - 1; - mmap_array[nr_cpu][counter].base = mmap(NULL, (mmap_pages+1)*page_size, - PROT_READ|PROT_WRITE, MAP_SHARED, fd[nr_cpu][counter], 0); - if (mmap_array[nr_cpu][counter].base == MAP_FAILED) { - error("failed to mmap with %d (%s)\n", errno, strerror(errno)); - exit(-1); + ret = ioctl(fd[nr_cpu][counter][thread_index], PERF_EVENT_IOC_SET_OUTPUT, multiplex_fd); + assert(ret != -1); + } else { + event_array[nr_poll].fd = fd[nr_cpu][counter][thread_index]; + event_array[nr_poll].events = POLLIN; + nr_poll++; + + mmap_array[nr_cpu][counter][thread_index].counter = counter; + mmap_array[nr_cpu][counter][thread_index].prev = 0; + mmap_array[nr_cpu][counter][thread_index].mask = mmap_pages*page_size - 1; + mmap_array[nr_cpu][counter][thread_index].base = mmap(NULL, (mmap_pages+1)*page_size, + PROT_READ|PROT_WRITE, MAP_SHARED, fd[nr_cpu][counter][thread_index], 0); + if (mmap_array[nr_cpu][counter][thread_index].base == MAP_FAILED) { + error("failed to mmap with %d (%s)\n", errno, strerror(errno)); + exit(-1); + } } - } - if (filter != NULL) { - ret = ioctl(fd[nr_cpu][counter], - PERF_EVENT_IOC_SET_FILTER, filter); - if (ret) { - error("failed to set filter with %d (%s)\n", errno, - strerror(errno)); - exit(-1); + if (filter != NULL) { + ret = ioctl(fd[nr_cpu][counter][thread_index], + PERF_EVENT_IOC_SET_FILTER, filter); + if (ret) { + error("failed to set filter with %d (%s)\n", errno, + strerror(errno)); + exit(-1); + } } } } -static void open_counters(int cpu, pid_t pid) +static void open_counters(int cpu) { int counter; group_fd = -1; for (counter = 0; counter < nr_counters; counter++) - create_counter(counter, cpu, pid); + create_counter(counter, cpu); nr_cpu++; } @@ -529,6 +542,9 @@ static int __cmd_record(int argc, const char **argv) exit(-1); } + if (!system_wide && target_tid == -1 && target_pid == -1) + all_tids[0] = child_pid; + close(child_ready_pipe[1]); close(go_pipe[0]); /* @@ -541,17 +557,12 @@ static int __cmd_record(int argc, const char **argv) close(child_ready_pipe[0]); } - if (forks && target_pid == -1 && !system_wide) - pid = child_pid; - else - pid = target_pid; - if ((!system_wide && !inherit) || profile_cpu != -1) { - open_counters(profile_cpu, pid); + open_counters(profile_cpu); } else { nr_cpus = read_cpu_map(); for (i = 0; i < nr_cpus; i++) - open_counters(cpumap[i], pid); + open_counters(cpumap[i]); } if (file_new) { @@ -576,7 +587,7 @@ static int __cmd_record(int argc, const char **argv) } if (!system_wide && profile_cpu == -1) - event__synthesize_thread(target_pid, process_synthesized_event, + event__synthesize_thread(target_tid, process_synthesized_event, session); else event__synthesize_threads(process_synthesized_event, session); @@ -599,11 +610,16 @@ static int __cmd_record(int argc, const char **argv) for (;;) { int hits = samples; + int thread; for (i = 0; i < nr_cpu; i++) { for (counter = 0; counter < nr_counters; counter++) { - if (mmap_array[i][counter].base) - mmap_read(&mmap_array[i][counter]); + for (thread = 0; + thread < thread_num; thread++) { + if (mmap_array[i][counter][thread].base) + mmap_read(&mmap_array[i][counter][thread]); + } + } } @@ -616,8 +632,15 @@ static int __cmd_record(int argc, const char **argv) if (done) { for (i = 0; i < nr_cpu; i++) { - for (counter = 0; counter < nr_counters; counter++) - ioctl(fd[i][counter], PERF_EVENT_IOC_DISABLE); + for (counter = 0; + counter < nr_counters; + counter++) { + for (thread = 0; + thread < thread_num; + thread++) + ioctl(fd[i][counter][thread], + PERF_EVENT_IOC_DISABLE); + } } } } @@ -649,7 +672,9 @@ static const struct option options[] = { OPT_CALLBACK(0, "filter", NULL, "filter", "event filter", parse_filter), OPT_INTEGER('p', "pid", &target_pid, - "record events on existing pid"), + "record events on existing process id"), + OPT_INTEGER('t', "tid", &target_tid, + "record events on existing thread id"), OPT_INTEGER('r', "realtime", &realtime_prio, "collect data with this RT SCHED_FIFO priority"), OPT_BOOLEAN('R', "raw-samples", &raw_samples, @@ -690,10 +715,12 @@ static const struct option options[] = { int cmd_record(int argc, const char **argv, const char *prefix __used) { int counter; + int i,j; argc = parse_options(argc, argv, options, record_usage, PARSE_OPT_STOP_AT_NON_OPTION); - if (!argc && target_pid == -1 && !system_wide && profile_cpu == -1) + if (!argc && target_pid == -1 && target_tid == -1 && + !system_wide && profile_cpu == -1) usage_with_options(record_usage, options); symbol__init(); @@ -704,6 +731,37 @@ int cmd_record(int argc, const char **argv, const char *prefix __used) attrs[0].config = PERF_COUNT_HW_CPU_CYCLES; } + if (target_pid != -1) { + target_tid = target_pid; + thread_num = find_all_tid(target_pid, &all_tids); + if (thread_num <= 0) { + fprintf(stderr, "Can't find all threads of pid %d\n", + target_pid); + usage_with_options(record_usage, options); + } + } else { + all_tids=malloc(sizeof(pid_t)); + if (!all_tids) + return -ENOMEM; + + all_tids[0] = target_tid; + thread_num = 1; + } + + for (i = 0; i < MAX_NR_CPUS; i++) { + for (j = 0; j < MAX_COUNTERS; j++) { + fd[i][j] = malloc(sizeof(int)*thread_num); + mmap_array[i][j] = malloc( + sizeof(struct mmap_data)*thread_num); + if (!fd[i][j] || !mmap_array[i][j]) + return -ENOMEM; + } + } + event_array = malloc( + sizeof(struct pollfd)*MAX_NR_CPUS*MAX_COUNTERS*thread_num); + if (!event_array) + return -ENOMEM; + /* * User specified count overrides default frequency. */ diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c index 5f41244..c92f90f 100644 --- a/tools/perf/builtin-stat.c +++ b/tools/perf/builtin-stat.c @@ -46,6 +46,7 @@ #include "util/debug.h" #include "util/header.h" #include "util/cpumap.h" +#include "util/thread.h" #include #include @@ -74,10 +75,13 @@ static int run_count = 1; static int inherit = 1; static int scale = 1; static pid_t target_pid = -1; +static pid_t target_tid = -1; +static pid_t *all_tids = NULL; +static int thread_num = 0; static pid_t child_pid = -1; static int null_run = 0; -static int fd[MAX_NR_CPUS][MAX_COUNTERS]; +static int *fd[MAX_NR_CPUS][MAX_COUNTERS]; static int event_scaled[MAX_COUNTERS]; @@ -140,9 +144,10 @@ struct stats runtime_branches_stats; #define ERR_PERF_OPEN \ "Error: counter %d, sys_perf_event_open() syscall returned with %d (%s)\n" -static void create_perf_stat_counter(int counter, int pid) +static void create_perf_stat_counter(int counter) { struct perf_event_attr *attr = attrs + counter; + int thread; if (scale) attr->read_format = PERF_FORMAT_TOTAL_TIME_ENABLED | @@ -152,10 +157,11 @@ static void create_perf_stat_counter(int counter, int pid) unsigned int cpu; for (cpu = 0; cpu < nr_cpus; cpu++) { - fd[cpu][counter] = sys_perf_event_open(attr, -1, cpumap[cpu], -1, 0); - if (fd[cpu][counter] < 0 && verbose) + fd[cpu][counter][0] = sys_perf_event_open(attr, + -1, cpumap[cpu], -1, 0); + if (fd[cpu][counter][0] < 0 && verbose) fprintf(stderr, ERR_PERF_OPEN, counter, - fd[cpu][counter], strerror(errno)); + fd[cpu][counter][0], strerror(errno)); } } else { attr->inherit = inherit; @@ -163,11 +169,14 @@ static void create_perf_stat_counter(int counter, int pid) attr->disabled = 1; attr->enable_on_exec = 1; } - - fd[0][counter] = sys_perf_event_open(attr, pid, -1, -1, 0); - if (fd[0][counter] < 0 && verbose) - fprintf(stderr, ERR_PERF_OPEN, counter, - fd[0][counter], strerror(errno)); + for (thread = 0; thread < thread_num; thread++) { + fd[0][counter][thread] = sys_perf_event_open(attr, + all_tids[thread], -1, -1, 0); + if (fd[0][counter][thread] < 0 && verbose) + fprintf(stderr, ERR_PERF_OPEN, counter, + fd[0][counter][thread], + strerror(errno)); + } } } @@ -192,25 +201,28 @@ static void read_counter(int counter) unsigned int cpu; size_t res, nv; int scaled; - int i; + int i, thread; count[0] = count[1] = count[2] = 0; nv = scale ? 3 : 1; for (cpu = 0; cpu < nr_cpus; cpu++) { - if (fd[cpu][counter] < 0) - continue; - - res = read(fd[cpu][counter], single_count, nv * sizeof(u64)); - assert(res == nv * sizeof(u64)); - - close(fd[cpu][counter]); - fd[cpu][counter] = -1; - - count[0] += single_count[0]; - if (scale) { - count[1] += single_count[1]; - count[2] += single_count[2]; + for (thread = 0; thread < thread_num; thread++) { + if (fd[cpu][counter][thread] < 0) + continue; + + res = read(fd[cpu][counter][thread], + single_count, nv * sizeof(u64)); + assert(res == nv * sizeof(u64)); + + close(fd[cpu][counter][thread]); + fd[cpu][counter][thread] = -1; + + count[0] += single_count[0]; + if (scale) { + count[1] += single_count[1]; + count[2] += single_count[2]; + } } } @@ -253,7 +265,6 @@ static int run_perf_stat(int argc __used, const char **argv) unsigned long long t0, t1; int status = 0; int counter; - int pid; int child_ready_pipe[2], go_pipe[2]; const bool forks = (argc > 0); char buf; @@ -299,6 +310,9 @@ static int run_perf_stat(int argc __used, const char **argv) exit(-1); } + if (target_tid == -1 && target_pid == -1 && !system_wide) + all_tids[0] = child_pid; + /* * Wait for the child to be ready to exec. */ @@ -309,12 +323,8 @@ static int run_perf_stat(int argc __used, const char **argv) close(child_ready_pipe[0]); } - if (target_pid == -1) - pid = child_pid; - else - pid = target_pid; for (counter = 0; counter < nr_counters; counter++) - create_perf_stat_counter(counter, pid); + create_perf_stat_counter(counter); /* * Enable counters and exec the command: @@ -433,12 +443,14 @@ static void print_stat(int argc, const char **argv) fprintf(stderr, "\n"); fprintf(stderr, " Performance counter stats for "); - if(target_pid == -1) { + if(target_pid == -1 && target_tid == -1) { fprintf(stderr, "\'%s", argv[0]); for (i = 1; i < argc; i++) fprintf(stderr, " %s", argv[i]); - }else - fprintf(stderr, "task pid \'%d", target_pid); + } else if (target_pid != -1) + fprintf(stderr, "process id \'%d", target_pid); + else + fprintf(stderr, "thread id \'%d", target_tid); fprintf(stderr, "\'"); if (run_count > 1) @@ -493,7 +505,9 @@ static const struct option options[] = { OPT_BOOLEAN('i', "inherit", &inherit, "child tasks inherit counters"), OPT_INTEGER('p', "pid", &target_pid, - "stat events on existing pid"), + "stat events on existing process id"), + OPT_INTEGER('t', "tid", &target_tid, + "stat events on existing thread id"), OPT_BOOLEAN('a', "all-cpus", &system_wide, "system-wide collection from all CPUs"), OPT_BOOLEAN('c', "scale", &scale, @@ -510,10 +524,11 @@ static const struct option options[] = { int cmd_stat(int argc, const char **argv, const char *prefix __used) { int status; + int i,j; argc = parse_options(argc, argv, options, stat_usage, PARSE_OPT_STOP_AT_NON_OPTION); - if (!argc && target_pid == -1) + if (!argc && target_pid == -1 && target_tid == -1) usage_with_options(stat_usage, options); if (run_count <= 0) usage_with_options(stat_usage, options); @@ -529,6 +544,31 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used) else nr_cpus = 1; + if (target_pid != -1) { + target_tid = target_pid; + thread_num = find_all_tid(target_pid, &all_tids); + if (thread_num <= 0) { + fprintf(stderr, "Can't find all threads of pid %d\n", + target_pid); + usage_with_options(stat_usage, options); + } + } else { + all_tids=malloc(sizeof(pid_t)); + if (!all_tids) + return -ENOMEM; + + all_tids[0] = target_tid; + thread_num = 1; + } + + for (i = 0; i < MAX_NR_CPUS; i++) { + for (j = 0; j < MAX_COUNTERS; j++) { + fd[i][j] = malloc(sizeof(int)*thread_num); + if (!fd[i][j]) + return -ENOMEM; + } + } + /* * We dont want to block the signals - that would cause * child tasks to inherit that and Ctrl-C would not work. diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index 887ebbf..5f3ac9f 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -55,7 +55,7 @@ #include #include -static int fd[MAX_NR_CPUS][MAX_COUNTERS]; +static int *fd[MAX_NR_CPUS][MAX_COUNTERS]; static int system_wide = 0; @@ -65,6 +65,9 @@ static int count_filter = 5; static int print_entries; static int target_pid = -1; +static int target_tid = -1; +static pid_t *all_tids = NULL; +static int thread_num = 0; static int inherit = 0; static int profile_cpu = -1; static int nr_cpus = 0; @@ -524,13 +527,15 @@ static void print_sym_table(void) if (target_pid != -1) printf(" (target_pid: %d", target_pid); + else if (target_tid != -1) + printf(" (target_tid: %d", target_tid); else printf(" (all"); if (profile_cpu != -1) printf(", cpu: %d)\n", profile_cpu); else { - if (target_pid != -1) + if (target_tid != -1) printf(")\n"); else printf(", %d CPUs)\n", nr_cpus); @@ -1129,16 +1134,21 @@ static void perf_session__mmap_read_counter(struct perf_session *self, md->prev = old; } -static struct pollfd event_array[MAX_NR_CPUS * MAX_COUNTERS]; -static struct mmap_data mmap_array[MAX_NR_CPUS][MAX_COUNTERS]; +static struct pollfd *event_array; +static struct mmap_data *mmap_array[MAX_NR_CPUS][MAX_COUNTERS]; static void perf_session__mmap_read(struct perf_session *self) { - int i, counter; + int i, counter, thread_index; for (i = 0; i < nr_cpus; i++) { for (counter = 0; counter < nr_counters; counter++) - perf_session__mmap_read_counter(self, &mmap_array[i][counter]); + for (thread_index = 0; + thread_index < thread_num; + thread_index++) { + perf_session__mmap_read_counter(self, + &mmap_array[i][counter][thread_index]); + } } } @@ -1149,9 +1159,10 @@ static void start_counter(int i, int counter) { struct perf_event_attr *attr; int cpu; + int thread_index; cpu = profile_cpu; - if (target_pid == -1 && profile_cpu == -1) + if (target_tid == -1 && profile_cpu == -1) cpu = cpumap[i]; attr = attrs + counter; @@ -1167,55 +1178,58 @@ static void start_counter(int i, int counter) attr->inherit = (cpu < 0) && inherit; attr->mmap = 1; + for (thread_index = 0; thread_index < thread_num; thread_index++) { try_again: - fd[i][counter] = sys_perf_event_open(attr, target_pid, cpu, group_fd, 0); - - if (fd[i][counter] < 0) { - int err = errno; + fd[i][counter][thread_index] = sys_perf_event_open(attr, + all_tids[thread_index], cpu, group_fd, 0); + + if (fd[i][counter][thread_index] < 0) { + int err = errno; + + if (err == EPERM || err == EACCES) + die("No permission - are you root?\n"); + /* + * If it's cycles then fall back to hrtimer + * based cpu-clock-tick sw counter, which + * is always available even if no PMU support: + */ + if (attr->type == PERF_TYPE_HARDWARE + && attr->config == PERF_COUNT_HW_CPU_CYCLES) { + + if (verbose) + warning(" ... trying to fall back to cpu-clock-ticks\n"); + + attr->type = PERF_TYPE_SOFTWARE; + attr->config = PERF_COUNT_SW_CPU_CLOCK; + goto try_again; + } + printf("\n"); + error("perfcounter syscall returned with %d (%s)\n", + fd[i][counter][thread_index], strerror(err)); + die("No CONFIG_PERF_EVENTS=y kernel support configured?\n"); + exit(-1); + } + assert(fd[i][counter][thread_index] >= 0); + fcntl(fd[i][counter][thread_index], F_SETFL, O_NONBLOCK); - if (err == EPERM || err == EACCES) - die("No permission - are you root?\n"); /* - * If it's cycles then fall back to hrtimer - * based cpu-clock-tick sw counter, which - * is always available even if no PMU support: + * First counter acts as the group leader: */ - if (attr->type == PERF_TYPE_HARDWARE - && attr->config == PERF_COUNT_HW_CPU_CYCLES) { - - if (verbose) - warning(" ... trying to fall back to cpu-clock-ticks\n"); - - attr->type = PERF_TYPE_SOFTWARE; - attr->config = PERF_COUNT_SW_CPU_CLOCK; - goto try_again; - } - printf("\n"); - error("perfcounter syscall returned with %d (%s)\n", - fd[i][counter], strerror(err)); - die("No CONFIG_PERF_EVENTS=y kernel support configured?\n"); - exit(-1); + if (group && group_fd == -1) + group_fd = fd[i][counter][thread_index]; + + event_array[nr_poll].fd = fd[i][counter][thread_index]; + event_array[nr_poll].events = POLLIN; + nr_poll++; + + mmap_array[i][counter][thread_index].counter = counter; + mmap_array[i][counter][thread_index].prev = 0; + mmap_array[i][counter][thread_index].mask = mmap_pages*page_size - 1; + mmap_array[i][counter][thread_index].base = mmap(NULL, (mmap_pages+1)*page_size, + PROT_READ, MAP_SHARED, fd[i][counter][thread_index], 0); + if (mmap_array[i][counter][thread_index].base == MAP_FAILED) + die("failed to mmap with %d (%s)\n", errno, strerror(errno)); } - assert(fd[i][counter] >= 0); - fcntl(fd[i][counter], F_SETFL, O_NONBLOCK); - - /* - * First counter acts as the group leader: - */ - if (group && group_fd == -1) - group_fd = fd[i][counter]; - - event_array[nr_poll].fd = fd[i][counter]; - event_array[nr_poll].events = POLLIN; - nr_poll++; - - mmap_array[i][counter].counter = counter; - mmap_array[i][counter].prev = 0; - mmap_array[i][counter].mask = mmap_pages*page_size - 1; - mmap_array[i][counter].base = mmap(NULL, (mmap_pages+1)*page_size, - PROT_READ, MAP_SHARED, fd[i][counter], 0); - if (mmap_array[i][counter].base == MAP_FAILED) - die("failed to mmap with %d (%s)\n", errno, strerror(errno)); } static int __cmd_top(void) @@ -1231,8 +1245,8 @@ static int __cmd_top(void) if (session == NULL) return -ENOMEM; - if (target_pid != -1) - event__synthesize_thread(target_pid, event__process, session); + if (target_tid != -1) + event__synthesize_thread(target_tid, event__process, session); else event__synthesize_threads(event__process, session); @@ -1243,7 +1257,7 @@ static int __cmd_top(void) } /* Wait for a minimal set of events before starting the snapshot */ - poll(event_array, nr_poll, 100); + poll(&event_array[0], nr_poll, 100); perf_session__mmap_read(session); @@ -1286,7 +1300,9 @@ static const struct option options[] = { OPT_INTEGER('c', "count", &default_interval, "event period to sample"), OPT_INTEGER('p', "pid", &target_pid, - "profile events on existing pid"), + "profile events on existing process id"), + OPT_INTEGER('t', "tid", &target_tid, + "profile events on existing thread id"), OPT_BOOLEAN('a', "all-cpus", &system_wide, "system-wide collection from all CPUs"), OPT_INTEGER('C', "CPU", &profile_cpu, @@ -1327,6 +1343,7 @@ static const struct option options[] = { int cmd_top(int argc, const char **argv, const char *prefix __used) { int counter; + int i,j; page_size = sysconf(_SC_PAGE_SIZE); @@ -1334,8 +1351,39 @@ int cmd_top(int argc, const char **argv, const char *prefix __used) if (argc) usage_with_options(top_usage, options); + if (target_pid != -1) { + target_tid = target_pid; + thread_num = find_all_tid(target_pid, &all_tids); + if (thread_num <= 0) { + fprintf(stderr, "Can't find all threads of pid %d\n", + target_pid); + usage_with_options(top_usage, options); + } + } else { + all_tids=malloc(sizeof(pid_t)); + if (!all_tids) + return -ENOMEM; + + all_tids[0] = target_tid; + thread_num = 1; + } + + for (i = 0; i < MAX_NR_CPUS; i++) { + for (j = 0; j < MAX_COUNTERS; j++) { + fd[i][j] = malloc(sizeof(int)*thread_num); + mmap_array[i][j] = malloc( + sizeof(struct mmap_data)*thread_num); + if (!fd[i][j] || !mmap_array[i][j]) + return -ENOMEM; + } + } + event_array = malloc( + sizeof(struct pollfd)*MAX_NR_CPUS*MAX_COUNTERS*thread_num); + if (!event_array) + return -ENOMEM; + /* CPU and PID are mutually exclusive */ - if (target_pid != -1 && profile_cpu != -1) { + if (target_tid > 0 && profile_cpu != -1) { printf("WARNING: PID switch overriding CPU\n"); sleep(1); profile_cpu = -1; @@ -1376,7 +1424,7 @@ int cmd_top(int argc, const char **argv, const char *prefix __used) attrs[counter].sample_period = default_interval; } - if (target_pid != -1 || profile_cpu != -1) + if (target_tid != -1 || profile_cpu != -1) nr_cpus = 1; else nr_cpus = read_cpu_map(); diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c index fa96831..ea65062 100644 --- a/tools/perf/util/thread.c +++ b/tools/perf/util/thread.c @@ -7,6 +7,37 @@ #include "util.h" #include "debug.h" +int find_all_tid(int pid, pid_t ** all_tid) +{ + char name[256]; + int items; + struct dirent **namelist = NULL; + int ret = 0; + int i; + + sprintf(name, "/proc/%d/task", pid); + items = scandir(name, &namelist, NULL, NULL); + if (items <= 0) + return -ENOENT; + *all_tid = malloc(sizeof(pid_t) * items); + if (!*all_tid) { + ret = -ENOMEM; + goto failure; + } + + for (i = 0; i < items; i++) + (*all_tid)[i] = atoi(namelist[i]->d_name); + + ret = items; + +failure: + for (i=0; i Date: Wed, 17 Mar 2010 13:37:00 +0300 Subject: x86, apic: Allow to use certain functions without APIC built-in support In case even if the kernel is configured so that no APIC support is built-in we still may allow to use certain apic functions as dummy calls. In particular we start using it in perf-events code. Note that this is not that same as NOOP apic driver (which is used if APIC support is present but no physical APIC is available), this is for the case when we don't have apic code compiled in at all. Signed-off-by: Cyrill Gorcunov Cc: H. Peter Anvin Cc: Yinghai Lu Cc: Yinghai Lu LKML-Reference: <20100317104356.011052632@openvz.org> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h index b4ac2cd..1fa03e0 100644 --- a/arch/x86/include/asm/apic.h +++ b/arch/x86/include/asm/apic.h @@ -373,6 +373,7 @@ extern atomic_t init_deasserted; extern int wakeup_secondary_cpu_via_nmi(int apicid, unsigned long start_eip); #endif +#ifdef CONFIG_X86_LOCAL_APIC static inline u32 apic_read(u32 reg) { return apic->read(reg); @@ -403,10 +404,19 @@ static inline u32 safe_apic_wait_icr_idle(void) return apic->safe_wait_icr_idle(); } +#else /* CONFIG_X86_LOCAL_APIC */ + +static inline u32 apic_read(u32 reg) { return 0; } +static inline void apic_write(u32 reg, u32 val) { } +static inline u64 apic_icr_read(void) { return 0; } +static inline void apic_icr_write(u32 low, u32 high) { } +static inline void apic_wait_icr_idle(void) { } +static inline u32 safe_apic_wait_icr_idle(void) { return 0; } + +#endif /* CONFIG_X86_LOCAL_APIC */ static inline void ack_APIC_irq(void) { -#ifdef CONFIG_X86_LOCAL_APIC /* * ack_APIC_irq() actually gets compiled as a single instruction * ... yummie. @@ -414,7 +424,6 @@ static inline void ack_APIC_irq(void) /* Docs say use 0 for future compatibility */ apic_write(APIC_EOI, 0); -#endif } static inline unsigned default_get_apic_id(unsigned long x) -- cgit v0.10.2 From 7335f75e9ca166044e38a96abad422d8e6e364b5 Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Wed, 17 Mar 2010 13:37:01 +0300 Subject: x86, perf: Use apic_write unconditionally Since apic_write() maps to a plain noop in the !CONFIG_X86_LOCAL_APIC case we're safe to remove this conditional compilation and clean up the code a bit. Signed-off-by: Cyrill Gorcunov Cc: fweisbec@gmail.com Cc: acme@redhat.com Cc: eranian@google.com Cc: peterz@infradead.org LKML-Reference: <20100317104356.232371479@openvz.org> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index c97d5b5..14eca80 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -1136,7 +1136,6 @@ void set_perf_event_pending(void) void perf_events_lapic_init(void) { -#ifdef CONFIG_X86_LOCAL_APIC if (!x86_pmu.apic || !x86_pmu_initialized()) return; @@ -1144,7 +1143,6 @@ void perf_events_lapic_init(void) * Always use NMI for PMU */ apic_write(APIC_LVTPC, APIC_DM_NMI); -#endif } static int __kprobes @@ -1168,9 +1166,7 @@ perf_event_nmi_handler(struct notifier_block *self, regs = args->regs; -#ifdef CONFIG_X86_LOCAL_APIC apic_write(APIC_LVTPC, APIC_DM_NMI); -#endif /* * Can't rely on the handled return value to say it was our NMI, two * events could trigger 'simultaneously' raising two back-to-back NMIs. diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index a11ce73..0367889 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -363,10 +363,8 @@ static int p4_pmu_handle_irq(struct pt_regs *regs) } if (handled) { -#ifdef CONFIG_X86_LOCAL_APIC /* p4 quirk: unmask it again */ apic_write(APIC_LVTPC, apic_read(APIC_LVTPC) & ~APIC_LVT_MASKED); -#endif inc_irq_stat(apic_perf_irqs); } -- cgit v0.10.2 From 55632770d7298835645489828af87f854c47749c Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Thu, 18 Mar 2010 16:51:16 +0100 Subject: perf events: Fix false positive build warning with older GCC's gcc 4.2.1 produces: util/probe-event.c: In function 'add_perf_probe_events': util/probe-event.c:883: warning: 'tev' may be used uninitialized in this function make: *** [util/probe-event.o] Error 1 Newer GCCs get this right. To work it around, initialize the variable to NULL so that older GCCs see it as initialized too. Cc: Masami Hiramatsu Cc: Frederic Weisbecker Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <20100316220612.32050.33806.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index f333269..c6603f3 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -880,7 +880,7 @@ static void __add_kprobe_trace_events(struct perf_probe_event *pev, int ntevs, bool allow_suffix) { int i, fd; - struct kprobe_trace_event *tev; + struct kprobe_trace_event *tev = NULL; char buf[64]; const char *event, *group; struct strlist *namelist; -- cgit v0.10.2 From f34edbc1cdb0f8f83d94e1d668dd6e41abf0defb Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Thu, 18 Mar 2010 18:33:07 +0800 Subject: perf, x86: Add a key to simplify template lookup in Pentium-4 PMU Currently, we use opcode(Event and Event-Selector) + emask to look up template in p4_templates. But cache events (L1-dcache-load-misses, LLC-load-misses, etc) use the same event(P4_REPLAY_EVENT) to do the counting, ie, they have the same opcode and emask. So we can not use current lookup mechanism to find the template for cache events. This patch introduces a "key", which is the index into p4_templates. The low 12 bits of CCCR are reserved, so we can hide the "key" in the low 12 bits of hwc->config. We extract the key from hwc->config and then quickly find the template. Signed-off-by: Lin Ming Reviewed-by: Cyrill Gorcunov Cc: Peter Zijlstra LKML-Reference: <1268908387.13901.127.camel@minggr.sh.intel.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/perf_event_p4.h b/arch/x86/include/asm/perf_event_p4.h index b842b32..7d3406a 100644 --- a/arch/x86/include/asm/perf_event_p4.h +++ b/arch/x86/include/asm/perf_event_p4.h @@ -65,6 +65,7 @@ #define P4_CCCR_THREAD_SINGLE 0x00010000U #define P4_CCCR_THREAD_BOTH 0x00020000U #define P4_CCCR_THREAD_ANY 0x00030000U +#define P4_CCCR_RESERVED 0x00000fffU /* Non HT mask */ #define P4_CCCR_MASK \ @@ -116,7 +117,7 @@ #define p4_config_pack_escr(v) (((u64)(v)) << 32) #define p4_config_pack_cccr(v) (((u64)(v)) & 0xffffffffULL) #define p4_config_unpack_escr(v) (((u64)(v)) >> 32) -#define p4_config_unpack_cccr(v) (((u64)(v)) & 0xffffffffULL) +#define p4_config_unpack_cccr(v) (((u64)(v)) & 0xfffff000ULL) #define p4_config_unpack_emask(v) \ ({ \ @@ -126,6 +127,8 @@ t; \ }) +#define p4_config_unpack_key(v) (((u64)(v)) & P4_CCCR_RESERVED) + #define P4_CONFIG_HT_SHIFT 63 #define P4_CONFIG_HT (1ULL << P4_CONFIG_HT_SHIFT) diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index 0367889..3e97ed3 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -18,6 +18,7 @@ struct p4_event_template { u32 opcode; /* ESCR event + CCCR selector */ u64 config; /* packed predefined bits */ int dep; /* upstream dependency event index */ + int key; /* index into p4_templates */ unsigned int emask; /* ESCR EventMask */ unsigned int escr_msr[2]; /* ESCR MSR for this event */ unsigned int cntr[2]; /* counter index (offset) */ @@ -39,38 +40,31 @@ static DEFINE_PER_CPU(struct p4_pmu_res, p4_pmu_config); */ struct p4_event_template p4_templates[] = { [0] = { - .opcode = P4_UOP_TYPE, - .config = 0, - .dep = -1, - .emask = - P4_EVENT_ATTR(P4_UOP_TYPE, TAGLOADS) | - P4_EVENT_ATTR(P4_UOP_TYPE, TAGSTORES), - .escr_msr = { MSR_P4_RAT_ESCR0, MSR_P4_RAT_ESCR1 }, - .cntr = { 16, 17 }, - }, - [1] = { .opcode = P4_GLOBAL_POWER_EVENTS, .config = 0, .dep = -1, + .key = 0, .emask = P4_EVENT_ATTR(P4_GLOBAL_POWER_EVENTS, RUNNING), .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, .cntr = { 0, 2 }, }, - [2] = { + [1] = { .opcode = P4_INSTR_RETIRED, .config = 0, .dep = -1, /* needs front-end tagging */ + .key = 1, .emask = P4_EVENT_ATTR(P4_INSTR_RETIRED, NBOGUSNTAG) | P4_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSNTAG), .escr_msr = { MSR_P4_CRU_ESCR0, MSR_P4_CRU_ESCR1 }, .cntr = { 12, 14 }, }, - [3] = { + [2] = { .opcode = P4_BSQ_CACHE_REFERENCE, .config = 0, .dep = -1, + .key = 2, .emask = P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITS) | P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITE) | @@ -81,10 +75,11 @@ struct p4_event_template p4_templates[] = { .escr_msr = { MSR_P4_BSU_ESCR0, MSR_P4_BSU_ESCR1 }, .cntr = { 0, 2 }, }, - [4] = { + [3] = { .opcode = P4_BSQ_CACHE_REFERENCE, .config = 0, .dep = -1, + .key = 3, .emask = P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_MISS) | P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_MISS) | @@ -92,10 +87,11 @@ struct p4_event_template p4_templates[] = { .escr_msr = { MSR_P4_BSU_ESCR0, MSR_P4_BSU_ESCR1 }, .cntr = { 0, 3 }, }, - [5] = { + [4] = { .opcode = P4_RETIRED_BRANCH_TYPE, .config = 0, .dep = -1, + .key = 4, .emask = P4_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, CONDITIONAL) | P4_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, CALL) | @@ -104,48 +100,38 @@ struct p4_event_template p4_templates[] = { .escr_msr = { MSR_P4_TBPU_ESCR0, MSR_P4_TBPU_ESCR1 }, .cntr = { 4, 6 }, }, - [6] = { + [5] = { .opcode = P4_MISPRED_BRANCH_RETIRED, .config = 0, .dep = -1, + .key = 5, .emask = P4_EVENT_ATTR(P4_MISPRED_BRANCH_RETIRED, NBOGUS), .escr_msr = { MSR_P4_CRU_ESCR0, MSR_P4_CRU_ESCR1 }, .cntr = { 12, 14 }, }, - [7] = { + [6] = { .opcode = P4_FSB_DATA_ACTIVITY, .config = p4_config_pack_cccr(P4_CCCR_EDGE | P4_CCCR_COMPARE), .dep = -1, + .key = 6, .emask = P4_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DRDY_DRV) | P4_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DRDY_OWN), .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, .cntr = { 0, 2 }, }, -}; - -static struct p4_event_template *p4_event_map[PERF_COUNT_HW_MAX] = { - /* non-halted CPU clocks */ - [PERF_COUNT_HW_CPU_CYCLES] = &p4_templates[1], - - /* retired instructions: dep on tagging the FSB */ - [PERF_COUNT_HW_INSTRUCTIONS] = &p4_templates[2], - - /* cache hits */ - [PERF_COUNT_HW_CACHE_REFERENCES] = &p4_templates[3], - - /* cache misses */ - [PERF_COUNT_HW_CACHE_MISSES] = &p4_templates[4], - - /* branch instructions retired */ - [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = &p4_templates[5], - - /* mispredicted branches retired */ - [PERF_COUNT_HW_BRANCH_MISSES] = &p4_templates[6], - - /* bus ready clocks (cpu is driving #DRDY_DRV\#DRDY_OWN): */ - [PERF_COUNT_HW_BUS_CYCLES] = &p4_templates[7], + [7] = { + .opcode = P4_UOP_TYPE, + .config = 0, + .dep = -1, + .key = 7, + .emask = + P4_EVENT_ATTR(P4_UOP_TYPE, TAGLOADS) | + P4_EVENT_ATTR(P4_UOP_TYPE, TAGSTORES), + .escr_msr = { MSR_P4_RAT_ESCR0, MSR_P4_RAT_ESCR1 }, + .cntr = { 16, 17 }, + }, }; static u64 p4_pmu_event_map(int hw_event) @@ -153,11 +139,11 @@ static u64 p4_pmu_event_map(int hw_event) struct p4_event_template *tpl; u64 config; - if (hw_event > ARRAY_SIZE(p4_event_map)) { + if (hw_event > ARRAY_SIZE(p4_templates)) { printk_once(KERN_ERR "PMU: Incorrect event index\n"); return 0; } - tpl = p4_event_map[hw_event]; + tpl = &p4_templates[hw_event]; /* * fill config up according to @@ -167,6 +153,7 @@ static u64 p4_pmu_event_map(int hw_event) config |= p4_config_pack_escr(P4_EVENT_UNPACK_EVENT(tpl->opcode) << P4_EVNTSEL_EVENT_SHIFT); config |= p4_config_pack_escr(tpl->emask << P4_EVNTSEL_EVENTMASK_SHIFT); config |= p4_config_pack_cccr(P4_EVENT_UNPACK_SELECTOR(tpl->opcode) << P4_CCCR_ESCR_SELECT_SHIFT); + config |= p4_config_pack_cccr(hw_event & P4_CCCR_RESERVED); /* on HT machine we need a special bit */ if (p4_ht_active() && p4_ht_thread(raw_smp_processor_id())) @@ -187,17 +174,12 @@ static inline int p4_pmu_emask_match(unsigned int dst, unsigned int src) static struct p4_event_template *p4_pmu_template_lookup(u64 config) { - u32 opcode = p4_config_unpack_opcode(config); - unsigned int emask = p4_config_unpack_emask(config); - unsigned int i; - - for (i = 0; i < ARRAY_SIZE(p4_templates); i++) { - if (opcode == p4_templates[i].opcode && - p4_pmu_emask_match(emask, p4_templates[i].emask)) - return &p4_templates[i]; - } + int key = p4_config_unpack_key(config); - return NULL; + if (key < ARRAY_SIZE(p4_templates)) + return &p4_templates[key]; + else + return NULL; } /* @@ -564,7 +546,7 @@ static __initconst struct x86_pmu p4_pmu = { .perfctr = MSR_P4_BPU_PERFCTR0, .event_map = p4_pmu_event_map, .raw_event = p4_pmu_raw_event, - .max_events = ARRAY_SIZE(p4_event_map), + .max_events = ARRAY_SIZE(p4_templates), .get_event_constraints = x86_get_event_constraints, /* * IF HT disabled we may need to use all -- cgit v0.10.2 From cb7d6b5053e86598735d9af19930f5929f007b7f Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Thu, 18 Mar 2010 18:33:12 +0800 Subject: perf, x86: Add cache events for the Pentium-4 PMU Move the HT bit setting code from p4_pmu_event_map to p4_hw_config. So the cache events can get HT bit set correctly. Tested on my P4 desktop, below 6 cache events work: L1-dcache-load-misses LLC-load-misses dTLB-load-misses dTLB-store-misses iTLB-loads iTLB-load-misses Signed-off-by: Lin Ming Reviewed-by: Cyrill Gorcunov Cc: Peter Zijlstra LKML-Reference: <1268908392.13901.128.camel@minggr.sh.intel.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/msr-index.h b/arch/x86/include/asm/msr-index.h index 1cd58cd..aef562c 100644 --- a/arch/x86/include/asm/msr-index.h +++ b/arch/x86/include/asm/msr-index.h @@ -357,6 +357,8 @@ #define MSR_P4_U2L_ESCR0 0x000003b0 #define MSR_P4_U2L_ESCR1 0x000003b1 +#define MSR_P4_PEBS_MATRIX_VERT 0x000003f2 + /* Intel Core-based CPU performance counters */ #define MSR_CORE_PERF_FIXED_CTR0 0x00000309 #define MSR_CORE_PERF_FIXED_CTR1 0x0000030a diff --git a/arch/x86/include/asm/perf_event_p4.h b/arch/x86/include/asm/perf_event_p4.h index 7d3406a..871249c 100644 --- a/arch/x86/include/asm/perf_event_p4.h +++ b/arch/x86/include/asm/perf_event_p4.h @@ -708,4 +708,14 @@ enum P4_EVENTS_ATTR { P4_MAKE_EVENT_ATTR(P4_INSTR_COMPLETED, BOGUS, 1), }; +enum { + KEY_P4_L1D_OP_READ_RESULT_MISS, + KEY_P4_LL_OP_READ_RESULT_MISS, + KEY_P4_DTLB_OP_READ_RESULT_MISS, + KEY_P4_DTLB_OP_WRITE_RESULT_MISS, + KEY_P4_ITLB_OP_READ_RESULT_ACCESS, + KEY_P4_ITLB_OP_READ_RESULT_MISS, + KEY_P4_UOP_TYPE, +}; + #endif /* PERF_EVENT_P4_H */ diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index 3e97ed3..b7bf991 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -19,6 +19,11 @@ struct p4_event_template { u64 config; /* packed predefined bits */ int dep; /* upstream dependency event index */ int key; /* index into p4_templates */ + u64 msr; /* + * the high 32 bits set into MSR_IA32_PEBS_ENABLE and + * the low 32 bits set into MSR_P4_PEBS_MATRIX_VERT + * for cache events + */ unsigned int emask; /* ESCR EventMask */ unsigned int escr_msr[2]; /* ESCR MSR for this event */ unsigned int cntr[2]; /* counter index (offset) */ @@ -31,6 +36,67 @@ struct p4_pmu_res { static DEFINE_PER_CPU(struct p4_pmu_res, p4_pmu_config); +#define P4_CACHE_EVENT_CONFIG(event, bit) \ + p4_config_pack_escr(P4_EVENT_UNPACK_EVENT(event) << P4_EVNTSEL_EVENT_SHIFT) | \ + p4_config_pack_escr((event##_##bit) << P4_EVNTSEL_EVENTMASK_SHIFT) | \ + p4_config_pack_cccr(P4_EVENT_UNPACK_SELECTOR(event) << P4_CCCR_ESCR_SELECT_SHIFT) + +static __initconst u64 p4_hw_cache_event_ids + [PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX] = +{ + [ C(L1D ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0, + /* 1stL_cache_load_miss_retired */ + [ C(RESULT_MISS) ] = P4_CACHE_EVENT_CONFIG(P4_REPLAY_EVENT, NBOGUS) + | KEY_P4_L1D_OP_READ_RESULT_MISS, + }, + }, + [ C(LL ) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0, + /* 2ndL_cache_load_miss_retired */ + [ C(RESULT_MISS) ] = P4_CACHE_EVENT_CONFIG(P4_REPLAY_EVENT, NBOGUS) + | KEY_P4_LL_OP_READ_RESULT_MISS, + }, + }, + [ C(DTLB) ] = { + [ C(OP_READ) ] = { + [ C(RESULT_ACCESS) ] = 0x0, + /* DTLB_load_miss_retired */ + [ C(RESULT_MISS) ] = P4_CACHE_EVENT_CONFIG(P4_REPLAY_EVENT, NBOGUS) + | KEY_P4_DTLB_OP_READ_RESULT_MISS, + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = 0x0, + /* DTLB_store_miss_retired */ + [ C(RESULT_MISS) ] = P4_CACHE_EVENT_CONFIG(P4_REPLAY_EVENT, NBOGUS) + | KEY_P4_DTLB_OP_WRITE_RESULT_MISS, + }, + }, + [ C(ITLB) ] = { + [ C(OP_READ) ] = { + /* ITLB_reference.HIT */ + [ C(RESULT_ACCESS) ] = P4_CACHE_EVENT_CONFIG(P4_ITLB_REFERENCE, HIT) + | KEY_P4_ITLB_OP_READ_RESULT_ACCESS, + + /* ITLB_reference.MISS */ + [ C(RESULT_MISS) ] = P4_CACHE_EVENT_CONFIG(P4_ITLB_REFERENCE, MISS) + | KEY_P4_ITLB_OP_READ_RESULT_MISS, + }, + [ C(OP_WRITE) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + [ C(OP_PREFETCH) ] = { + [ C(RESULT_ACCESS) ] = -1, + [ C(RESULT_MISS) ] = -1, + }, + }, +}; + /* * WARN: CCCR1 doesn't have a working enable bit so try to not * use it if possible @@ -121,11 +187,77 @@ struct p4_event_template p4_templates[] = { .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, .cntr = { 0, 2 }, }, - [7] = { + [KEY_P4_L1D_OP_READ_RESULT_MISS] = { + .opcode = P4_REPLAY_EVENT, + .config = 0, + .dep = -1, + .msr = (u64)(1 << 0 | 1 << 24) << 32 | (1 << 0), + .key = KEY_P4_L1D_OP_READ_RESULT_MISS, + .emask = + P4_EVENT_ATTR(P4_REPLAY_EVENT, NBOGUS), + .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR2 }, + .cntr = { 16, 17 }, + }, + [KEY_P4_LL_OP_READ_RESULT_MISS] = { + .opcode = P4_REPLAY_EVENT, + .config = 0, + .dep = -1, + .msr = (u64)(1 << 1 | 1 << 24) << 32 | (1 << 0), + .key = KEY_P4_LL_OP_READ_RESULT_MISS, + .emask = + P4_EVENT_ATTR(P4_REPLAY_EVENT, NBOGUS), + .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR2 }, + .cntr = { 16, 17 }, + }, + [KEY_P4_DTLB_OP_READ_RESULT_MISS] = { + .opcode = P4_REPLAY_EVENT, + .config = 0, + .dep = -1, + .msr = (u64)(1 << 2 | 1 << 24) << 32 | (1 << 0), + .key = KEY_P4_DTLB_OP_READ_RESULT_MISS, + .emask = + P4_EVENT_ATTR(P4_REPLAY_EVENT, NBOGUS), + .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR2 }, + .cntr = { 16, 17 }, + }, + [KEY_P4_DTLB_OP_WRITE_RESULT_MISS] = { + .opcode = P4_REPLAY_EVENT, + .config = 0, + .dep = -1, + .msr = (u64)(1 << 2 | 1 << 24) << 32 | (1 << 1), + .key = KEY_P4_DTLB_OP_WRITE_RESULT_MISS, + .emask = + P4_EVENT_ATTR(P4_REPLAY_EVENT, NBOGUS), + .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR2 }, + .cntr = { 16, 17 }, + }, + [KEY_P4_ITLB_OP_READ_RESULT_ACCESS] = { + .opcode = P4_ITLB_REFERENCE, + .config = 0, + .dep = -1, + .msr = 0, + .key = KEY_P4_ITLB_OP_READ_RESULT_ACCESS, + .emask = + P4_EVENT_ATTR(P4_ITLB_REFERENCE, HIT), + .escr_msr = { MSR_P4_ITLB_ESCR0, MSR_P4_ITLB_ESCR1 }, + .cntr = { 0, 2 }, + }, + [KEY_P4_ITLB_OP_READ_RESULT_MISS] = { + .opcode = P4_ITLB_REFERENCE, + .config = 0, + .dep = -1, + .msr = 0, + .key = KEY_P4_ITLB_OP_READ_RESULT_MISS, + .emask = + P4_EVENT_ATTR(P4_ITLB_REFERENCE, MISS), + .escr_msr = { MSR_P4_ITLB_ESCR0, MSR_P4_ITLB_ESCR1 }, + .cntr = { 0, 2 }, + }, + [KEY_P4_UOP_TYPE] = { .opcode = P4_UOP_TYPE, .config = 0, .dep = -1, - .key = 7, + .key = KEY_P4_UOP_TYPE, .emask = P4_EVENT_ATTR(P4_UOP_TYPE, TAGLOADS) | P4_EVENT_ATTR(P4_UOP_TYPE, TAGSTORES), @@ -155,10 +287,6 @@ static u64 p4_pmu_event_map(int hw_event) config |= p4_config_pack_cccr(P4_EVENT_UNPACK_SELECTOR(tpl->opcode) << P4_CCCR_ESCR_SELECT_SHIFT); config |= p4_config_pack_cccr(hw_event & P4_CCCR_RESERVED); - /* on HT machine we need a special bit */ - if (p4_ht_active() && p4_ht_thread(raw_smp_processor_id())) - config = p4_set_ht_bit(config); - return config; } @@ -211,6 +339,10 @@ static int p4_hw_config(struct perf_event_attr *attr, struct hw_perf_event *hwc) /* Count user and OS events unless not requested to */ hwc->config |= p4_config_pack_escr(p4_default_escr_conf(cpu, attr->exclude_kernel, attr->exclude_user)); + /* on HT machine we need a special bit */ + if (p4_ht_active() && p4_ht_thread(cpu)) + hwc->config = p4_set_ht_bit(hwc->config); + return 0; } @@ -271,6 +403,12 @@ static void p4_pmu_enable_event(struct perf_event *event) pr_crit("%s: Wrong index: %d\n", __func__, hwc->idx); return; } + + if (tpl->msr) { + (void)checking_wrmsrl(MSR_IA32_PEBS_ENABLE, tpl->msr >> 32); + (void)checking_wrmsrl(MSR_P4_PEBS_MATRIX_VERT, tpl->msr & 0xffffffff); + } + escr_base = (u64)tpl->escr_msr[thread]; /* @@ -577,6 +715,9 @@ static __init int p4_pmu_init(void) return -ENODEV; } + memcpy(hw_cache_event_ids, p4_hw_cache_event_ids, + sizeof(hw_cache_event_ids)); + pr_cont("Netburst events, "); x86_pmu = p4_pmu; -- cgit v0.10.2 From 4b24a88b35e15e04bd8f2c5dda65b5dc8ebca05f Mon Sep 17 00:00:00 2001 From: Stephane Eranian Date: Wed, 17 Mar 2010 23:21:01 +0200 Subject: perf_events: Fix resource leak in x86 __hw_perf_event_init() If reserve_pmc_hardware() succeeds but reserve_ds_buffers() fails, then we need to release_pmc_hardware. It won't be done by the destroy() callback because we return before setting it in case of error. Signed-off-by: Stephane Eranian Cc: Cc: peterz@infradead.org Cc: paulus@samba.org Cc: davem@davemloft.net Cc: fweisbec@gmail.com Cc: robert.richter@amd.com Cc: perfmon2-devel@lists.sf.net LKML-Reference: <4ba1568b.15185e0a.182a.7802@mx.google.com> Signed-off-by: Ingo Molnar -- arch/x86/kernel/cpu/perf_event.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 14eca80..f571f51 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -455,8 +455,11 @@ static int __hw_perf_event_init(struct perf_event *event) if (atomic_read(&active_events) == 0) { if (!reserve_pmc_hardware()) err = -EBUSY; - else + else { err = reserve_ds_buffers(); + if (err) + release_pmc_hardware(); + } } if (!err) atomic_inc(&active_events); -- cgit v0.10.2 From 9c8c6bad3137112d2c7bf3d215b736ee4215fa74 Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Fri, 19 Mar 2010 00:12:56 +0300 Subject: x86, perf: Fix few cosmetic dabs for P4 pmu (comments and constantify) - A few ESCR have escaped fixing at previous attempt. - p4_escr_map is read only, make it const. Nothing serious. Signed-off-by: Cyrill Gorcunov Cc: Lin Ming LKML-Reference: <20100318211256.GH5062@lenovo> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/perf_event_p4.h b/arch/x86/include/asm/perf_event_p4.h index 871249c..2a1a57f 100644 --- a/arch/x86/include/asm/perf_event_p4.h +++ b/arch/x86/include/asm/perf_event_p4.h @@ -401,13 +401,13 @@ static inline u32 p4_default_escr_conf(int cpu, int exclude_os, int exclude_usr) #define P4_RETIRED_MISPRED_BRANCH_TYPE P4_EVENT_PACK(0x05, 0x02) /* * MSR_P4_TBPU_ESCR0: 4, 5 - * MSR_P4_TBPU_ESCR0: 6, 7 + * MSR_P4_TBPU_ESCR1: 6, 7 */ #define P4_RETIRED_BRANCH_TYPE P4_EVENT_PACK(0x04, 0x02) /* * MSR_P4_TBPU_ESCR0: 4, 5 - * MSR_P4_TBPU_ESCR0: 6, 7 + * MSR_P4_TBPU_ESCR1: 6, 7 */ #define P4_RESOURCE_STALL P4_EVENT_PACK(0x01, 0x01) diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index b7bf991..b8a811a 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -545,7 +545,7 @@ static void p4_pmu_swap_config_ts(struct hw_perf_event *hwc, int cpu) } /* ESCRs are not sequential in memory so we need a map */ -static unsigned int p4_escr_map[ARCH_P4_TOTAL_ESCR] = { +static const unsigned int p4_escr_map[ARCH_P4_TOTAL_ESCR] = { MSR_P4_ALF_ESCR0, /* 0 */ MSR_P4_ALF_ESCR1, /* 1 */ MSR_P4_BPU_ESCR0, /* 2 */ -- cgit v0.10.2 From e9ce335df51ff782035a15c261a3c0c9892a1767 Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Fri, 19 Mar 2010 08:03:04 +0100 Subject: cfq-iosched: fix a kbuild regression Alex Shi reported a kbuild regression which is about 10% performance lost. He bisected to this commit: 3dde36ddea3e07dd025c4c1ba47edec91606fec0. The reason is cfqq_close() can't find close cooperator. Restoring cfq_rq_close()'s threshold to original value makes the regression go away. Since for_preempt parameter isn't used anymore, this patch deletes it. Reported-by: Alex Shi Signed-off-by: Shaohua Li Acked-by: Corrado Zoccolo Signed-off-by: Jens Axboe diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index dee9d93..8d5a2f2 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -47,6 +47,7 @@ static const int cfq_hist_divisor = 4; #define CFQ_SERVICE_SHIFT 12 #define CFQQ_SEEK_THR (sector_t)(8 * 100) +#define CFQQ_CLOSE_THR (sector_t)(8 * 1024) #define CFQQ_SECT_THR_NONROT (sector_t)(2 * 32) #define CFQQ_SEEKY(cfqq) (hweight32(cfqq->seek_history) > 32/8) @@ -1660,9 +1661,9 @@ static inline sector_t cfq_dist_from_last(struct cfq_data *cfqd, } static inline int cfq_rq_close(struct cfq_data *cfqd, struct cfq_queue *cfqq, - struct request *rq, bool for_preempt) + struct request *rq) { - return cfq_dist_from_last(cfqd, rq) <= CFQQ_SEEK_THR; + return cfq_dist_from_last(cfqd, rq) <= CFQQ_CLOSE_THR; } static struct cfq_queue *cfqq_close(struct cfq_data *cfqd, @@ -1689,7 +1690,7 @@ static struct cfq_queue *cfqq_close(struct cfq_data *cfqd, * will contain the closest sector. */ __cfqq = rb_entry(parent, struct cfq_queue, p_node); - if (cfq_rq_close(cfqd, cur_cfqq, __cfqq->next_rq, false)) + if (cfq_rq_close(cfqd, cur_cfqq, __cfqq->next_rq)) return __cfqq; if (blk_rq_pos(__cfqq->next_rq) < sector) @@ -1700,7 +1701,7 @@ static struct cfq_queue *cfqq_close(struct cfq_data *cfqd, return NULL; __cfqq = rb_entry(node, struct cfq_queue, p_node); - if (cfq_rq_close(cfqd, cur_cfqq, __cfqq->next_rq, false)) + if (cfq_rq_close(cfqd, cur_cfqq, __cfqq->next_rq)) return __cfqq; return NULL; @@ -3103,7 +3104,7 @@ cfq_should_preempt(struct cfq_data *cfqd, struct cfq_queue *new_cfqq, * if this request is as-good as one we would expect from the * current cfqq, let it preempt */ - if (cfq_rq_close(cfqd, cfqq, rq, true)) + if (cfq_rq_close(cfqd, cfqq, rq)) return true; return false; -- cgit v0.10.2 From 181fdde3b4268cb7b4af76ba6337e7ec8accbb36 Mon Sep 17 00:00:00 2001 From: Richard Kennedy Date: Fri, 19 Mar 2010 08:58:16 +0100 Subject: block: remove 16 bytes of padding from struct request on 64bits Remove alignment padding to shrink struct request from 336 to 320 bytes so needing one fewer cacheline and therefore removing 48 bytes from struct request_queue. Signed-off-by: Richard Kennedy Signed-off-by: Jens Axboe diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 41551c9..6690e8b 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -158,7 +158,6 @@ enum rq_flag_bits { struct request { struct list_head queuelist; struct call_single_data csd; - int cpu; struct request_queue *q; @@ -166,9 +165,11 @@ struct request { enum rq_cmd_type_bits cmd_type; unsigned long atomic_flags; + int cpu; + /* the following two fields are internal, NEVER access directly */ - sector_t __sector; /* sector cursor */ unsigned int __data_len; /* total data len */ + sector_t __sector; /* sector cursor */ struct bio *bio; struct bio *biotail; @@ -201,20 +202,20 @@ struct request { unsigned short ioprio; + int ref_count; + void *special; /* opaque pointer available for LLD use */ char *buffer; /* kaddr of the current segment if available */ int tag; int errors; - int ref_count; - /* * when request is used as a packet command carrier */ - unsigned short cmd_len; unsigned char __cmd[BLK_MAX_CDB]; unsigned char *cmd; + unsigned short cmd_len; unsigned int extra_len; /* length of alignment and padding */ unsigned int sense_len; -- cgit v0.10.2 From 40b7e05e17eef31ff30fe08dfc2424ef653a792c Mon Sep 17 00:00:00 2001 From: Lin Ming Date: Fri, 19 Mar 2010 15:28:58 +0800 Subject: perf, x86: Fix key indexing in Pentium-4 PMU Index 0-6 in p4_templates are reserved for common hardware events. So p4_templates is arranged as below: 0 - 6: common hardware events 7 - N: cache events N+1 - ...: other raw events Reported-by: Cyrill Gorcunov Signed-off-by: Lin Ming Acked-by: Cyrill Gorcunov Cc: Peter Zijlstra LKML-Reference: <1268983738.13901.142.camel@minggr.sh.intel.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/perf_event_p4.h b/arch/x86/include/asm/perf_event_p4.h index 2a1a57f..facf961 100644 --- a/arch/x86/include/asm/perf_event_p4.h +++ b/arch/x86/include/asm/perf_event_p4.h @@ -709,7 +709,7 @@ enum P4_EVENTS_ATTR { }; enum { - KEY_P4_L1D_OP_READ_RESULT_MISS, + KEY_P4_L1D_OP_READ_RESULT_MISS = PERF_COUNT_HW_MAX, KEY_P4_LL_OP_READ_RESULT_MISS, KEY_P4_DTLB_OP_READ_RESULT_MISS, KEY_P4_DTLB_OP_WRITE_RESULT_MISS, -- cgit v0.10.2 From 48fae657f04924c06fe8c9afdb149efdd99d68dd Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Thu, 18 Mar 2010 16:55:45 +0100 Subject: i.MX51 Babbage: Add uncompress output Signed-off-by: Sascha Hauer diff --git a/arch/arm/plat-mxc/include/mach/uncompress.h b/arch/arm/plat-mxc/include/mach/uncompress.h index 52e476a..b6d3d0f 100644 --- a/arch/arm/plat-mxc/include/mach/uncompress.h +++ b/arch/arm/plat-mxc/include/mach/uncompress.h @@ -66,6 +66,7 @@ static inline void flush(void) #define MX2X_UART1_BASE_ADDR 0x1000a000 #define MX3X_UART1_BASE_ADDR 0x43F90000 #define MX3X_UART2_BASE_ADDR 0x43F94000 +#define MX51_UART1_BASE_ADDR 0x73fbc000 static __inline__ void __arch_decomp_setup(unsigned long arch_id) { @@ -101,6 +102,9 @@ static __inline__ void __arch_decomp_setup(unsigned long arch_id) case MACH_TYPE_MAGX_ZN5: uart_base = MX3X_UART2_BASE_ADDR; break; + case MACH_TYPE_MX51_BABBAGE: + uart_base = MX51_UART1_BASE_ADDR; + break; default: break; } -- cgit v0.10.2 From 1b6a2b2d0ff2ced5fe608e0b2e13ccd2b7a283e5 Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Thu, 18 Mar 2010 16:56:11 +0100 Subject: i.MX51: Use correct clock for gpt The gpt uses the ipg clock, not ipg_perclk Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx5/clock-mx51.c b/arch/arm/mach-mx5/clock-mx51.c index be90c03..8f85f73 100644 --- a/arch/arm/mach-mx5/clock-mx51.c +++ b/arch/arm/mach-mx5/clock-mx51.c @@ -757,7 +757,7 @@ DEFINE_CLOCK(uart3_ipg_clk, 2, MXC_CCM_CCGR1, MXC_CCM_CCGRx_CG7_OFFSET, /* GPT */ DEFINE_CLOCK(gpt_clk, 0, MXC_CCM_CCGR2, MXC_CCM_CCGRx_CG9_OFFSET, - NULL, NULL, &ipg_perclk, NULL); + NULL, NULL, &ipg_clk, NULL); DEFINE_CLOCK(gpt_ipg_clk, 0, MXC_CCM_CCGR2, MXC_CCM_CCGRx_CG10_OFFSET, NULL, NULL, &ipg_clk, NULL); -- cgit v0.10.2 From 3d1bc8626c7b17facfcb7fb5dee4686f47a1e75d Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Thu, 18 Mar 2010 16:56:30 +0100 Subject: i.MX51: map TZIC dynamically This looks cleaner and allows us to call mx51_revision later when we can use ioremap to determine the silicon revision dynamically. Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx5/mm.c b/arch/arm/mach-mx5/mm.c index c21e18b..9fe7beb 100644 --- a/arch/arm/mach-mx5/mm.c +++ b/arch/arm/mach-mx5/mm.c @@ -35,11 +35,6 @@ static struct map_desc mxc_io_desc[] __initdata = { .length = MX51_DEBUG_SIZE, .type = MT_DEVICE }, { - .virtual = MX51_TZIC_BASE_ADDR_VIRT, - .pfn = __phys_to_pfn(MX51_TZIC_BASE_ADDR), - .length = MX51_TZIC_SIZE, - .type = MT_DEVICE - }, { .virtual = MX51_AIPS1_BASE_ADDR_VIRT, .pfn = __phys_to_pfn(MX51_AIPS1_BASE_ADDR), .length = MX51_AIPS1_SIZE, @@ -69,14 +64,6 @@ static struct map_desc mxc_io_desc[] __initdata = { */ void __init mx51_map_io(void) { - u32 tzic_addr; - - if (mx51_revision() < MX51_CHIP_REV_2_0) - tzic_addr = 0x8FFFC000; - else - tzic_addr = 0xE0003000; - mxc_io_desc[2].pfn = __phys_to_pfn(tzic_addr); - mxc_set_cpu_type(MXC_CPU_MX51); mxc_iomux_v3_init(MX51_IO_ADDRESS(MX51_IOMUXC_BASE_ADDR)); mxc_arch_reset_init(MX51_IO_ADDRESS(MX51_WDOG_BASE_ADDR)); @@ -85,5 +72,17 @@ void __init mx51_map_io(void) void __init mx51_init_irq(void) { - tzic_init_irq(MX51_IO_ADDRESS(MX51_TZIC_BASE_ADDR)); + unsigned long tzic_addr; + void __iomem *tzic_virt; + + if (mx51_revision() < MX51_CHIP_REV_2_0) + tzic_addr = MX51_TZIC_BASE_ADDR_TO1; + else + tzic_addr = MX51_TZIC_BASE_ADDR; + + tzic_virt = ioremap(tzic_addr, SZ_16K); + if (!tzic_virt) + panic("unable to map TZIC interrupt controller\n"); + + tzic_init_irq(tzic_virt); } diff --git a/arch/arm/plat-mxc/include/mach/mx51.h b/arch/arm/plat-mxc/include/mach/mx51.h index 771532b..f1396bd 100644 --- a/arch/arm/plat-mxc/include/mach/mx51.h +++ b/arch/arm/plat-mxc/include/mach/mx51.h @@ -14,7 +14,7 @@ * FB100000 70000000 1M SPBA 0 * FB000000 73F00000 1M AIPS 1 * FB200000 83F00000 1M AIPS 2 - * FA100000 8FFFC000 16K TZIC (interrupt controller) + * 8FFFC000 16K TZIC (interrupt controller) * 90000000 256M CSD0 SDRAM/DDR * A0000000 256M CSD1 SDRAM/DDR * B0000000 128M CS0 Flash @@ -49,9 +49,8 @@ #define MX51_GPU_BASE_ADDR 0x20000000 #define MX51_GPU2D_BASE_ADDR 0xD0000000 -#define MX51_TZIC_BASE_ADDR 0x8FFFC000 -#define MX51_TZIC_BASE_ADDR_VIRT 0xFA100000 -#define MX51_TZIC_SIZE SZ_16K +#define MX51_TZIC_BASE_ADDR_TO1 0x8FFFC000 +#define MX51_TZIC_BASE_ADDR 0xE0000000 #define MX51_DEBUG_BASE_ADDR 0x60000000 #define MX51_DEBUG_BASE_ADDR_VIRT 0xFA200000 @@ -232,7 +231,6 @@ #define MX51_IO_ADDRESS(x) \ (void __iomem *) \ (MX51_IS_MODULE(x, IRAM) ? MX51_IRAM_IO_ADDRESS(x) : \ - MX51_IS_MODULE(x, TZIC) ? MX51_TZIC_IO_ADDRESS(x) : \ MX51_IS_MODULE(x, DEBUG) ? MX51_DEBUG_IO_ADDRESS(x) : \ MX51_IS_MODULE(x, SPBA0) ? MX51_SPBA0_IO_ADDRESS(x) : \ MX51_IS_MODULE(x, AIPS1) ? MX51_AIPS1_IO_ADDRESS(x) : \ @@ -246,9 +244,6 @@ #define MX51_IRAM_IO_ADDRESS(x) \ (((x) - MX51_IRAM_BASE_ADDR) + MX51_IRAM_BASE_ADDR_VIRT) -#define MX51_TZIC_IO_ADDRESS(x) \ - (((x) - MX51_TZIC_BASE_ADDR) + MX51_TZIC_BASE_ADDR_VIRT) - #define MX51_DEBUG_IO_ADDRESS(x) \ (((x) - MX51_DEBUG_BASE_ADDR) + MX51_DEBUG_BASE_ADDR_VIRT) -- cgit v0.10.2 From 5443856cadac7faaaeefeed9d769f497a8c6fa4b Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Fri, 19 Mar 2010 10:50:55 +0100 Subject: i.MX51: determine silicon revision dynamically Freescale redboot passes the silicon revision via ATAG_REVISION. Remove this bootloader dependency by doing the same as redboot does in the Kernel. Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx5/cpu.c b/arch/arm/mach-mx5/cpu.c index 41c769f..2d37785 100644 --- a/arch/arm/mach-mx5/cpu.c +++ b/arch/arm/mach-mx5/cpu.c @@ -14,9 +14,62 @@ #include #include #include +#include #include #include +static int cpu_silicon_rev = -1; + +#define SI_REV 0x48 + +static void query_silicon_parameter(void) +{ + void __iomem *rom = ioremap(MX51_IROM_BASE_ADDR, MX51_IROM_SIZE); + u32 rev; + + if (!rom) { + cpu_silicon_rev = -EINVAL; + return; + } + + rev = readl(rom + SI_REV); + switch (rev) { + case 0x1: + cpu_silicon_rev = MX51_CHIP_REV_1_0; + break; + case 0x2: + cpu_silicon_rev = MX51_CHIP_REV_1_1; + break; + case 0x10: + cpu_silicon_rev = MX51_CHIP_REV_2_0; + break; + case 0x20: + cpu_silicon_rev = MX51_CHIP_REV_3_0; + break; + default: + cpu_silicon_rev = 0; + } + + iounmap(rom); +} + +/* + * Returns: + * the silicon revision of the cpu + * -EINVAL - not a mx51 + */ +int mx51_revision(void) +{ + if (!cpu_is_mx51()) + return -EINVAL; + + if (cpu_silicon_rev == -1) + query_silicon_parameter(); + + return cpu_silicon_rev; +} +EXPORT_SYMBOL(mx51_revision); + static int __init post_cpu_init(void) { unsigned int reg; diff --git a/arch/arm/plat-mxc/include/mach/mx51.h b/arch/arm/plat-mxc/include/mach/mx51.h index f1396bd..fd255a9 100644 --- a/arch/arm/plat-mxc/include/mach/mx51.h +++ b/arch/arm/plat-mxc/include/mach/mx51.h @@ -28,6 +28,12 @@ */ /* + * IROM + */ +#define MX51_IROM_BASE_ADDR 0x0 +#define MX51_IROM_SIZE SZ_64K + +/* * IRAM */ #define MX51_IRAM_BASE_ADDR 0x1FFE0000 /* internal ram */ @@ -438,12 +444,7 @@ #if !defined(__ASSEMBLY__) && !defined(__MXC_BOOT_UNCOMPRESS) -extern unsigned int system_rev; - -static inline unsigned int mx51_revision(void) -{ - return system_rev; -} +extern int mx51_revision(void); #endif #endif /* __ASM_ARCH_MXC_MX51_H__ */ -- cgit v0.10.2 From fc300206ad07e771ed003d35b1dc179eaf0c508f Mon Sep 17 00:00:00 2001 From: Sascha Hauer Date: Fri, 19 Mar 2010 11:12:02 +0100 Subject: i.MX51: remove NFC AXI static mapping This area contains the Nand Flash controller registers. There is no need to map them statically as the Nand driver uses ioremap(). Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx5/mm.c b/arch/arm/mach-mx5/mm.c index 9fe7beb..b7677ef 100644 --- a/arch/arm/mach-mx5/mm.c +++ b/arch/arm/mach-mx5/mm.c @@ -49,11 +49,6 @@ static struct map_desc mxc_io_desc[] __initdata = { .pfn = __phys_to_pfn(MX51_AIPS2_BASE_ADDR), .length = MX51_AIPS2_SIZE, .type = MT_DEVICE - }, { - .virtual = MX51_NFC_AXI_BASE_ADDR_VIRT, - .pfn = __phys_to_pfn(MX51_NFC_AXI_BASE_ADDR), - .length = MX51_NFC_AXI_SIZE, - .type = MT_DEVICE }, }; diff --git a/arch/arm/plat-mxc/include/mach/mx51.h b/arch/arm/plat-mxc/include/mach/mx51.h index fd255a9..5aad344 100644 --- a/arch/arm/plat-mxc/include/mach/mx51.h +++ b/arch/arm/plat-mxc/include/mach/mx51.h @@ -23,7 +23,7 @@ * C8000000 64M CS3 Flash * CC000000 32M CS4 SRAM * CE000000 32M CS5 SRAM - * F9000000 CFFF0000 64K NFC (NAND Flash AXI) + * CFFF0000 64K NFC (NAND Flash AXI) * */ @@ -46,7 +46,6 @@ * NFC */ #define MX51_NFC_AXI_BASE_ADDR 0xCFFF0000 /* NAND flash AXI */ -#define MX51_NFC_AXI_BASE_ADDR_VIRT 0xF9000000 #define MX51_NFC_AXI_SIZE SZ_64K /* @@ -240,8 +239,7 @@ MX51_IS_MODULE(x, DEBUG) ? MX51_DEBUG_IO_ADDRESS(x) : \ MX51_IS_MODULE(x, SPBA0) ? MX51_SPBA0_IO_ADDRESS(x) : \ MX51_IS_MODULE(x, AIPS1) ? MX51_AIPS1_IO_ADDRESS(x) : \ - MX51_IS_MODULE(x, AIPS2) ? MX51_AIPS2_IO_ADDRESS(x) : \ - MX51_IS_MODULE(x, NFC_AXI) ? MX51_NFC_AXI_IO_ADDRESS(x) : \ + MX51_IS_MODULE(x, AIPS2) ? MX51_AIPS2_IO_ADDRESS(x) : \ 0xDEADBEEF) /* @@ -262,9 +260,6 @@ #define MX51_AIPS2_IO_ADDRESS(x) \ (((x) - MX51_AIPS2_BASE_ADDR) + MX51_AIPS2_BASE_ADDR_VIRT) -#define MX51_NFC_AXI_IO_ADDRESS(x) \ - (((x) - MX51_NFC_AXI_BASE_ADDR) + MX51_NFC_AXI_BASE_ADDR_VIRT) - #define MX51_IS_MEM_DEVICE_NONSHARED(x) 0 /* -- cgit v0.10.2 From 0deac0bf66110fe70507c5dcec661c9e85fa07ca Mon Sep 17 00:00:00 2001 From: Russell King Date: Mon, 15 Mar 2010 20:36:22 +0000 Subject: ARM: Arrange for COMMON_CLKDEV to select HAVE_CLK If support for COMMON_CLKDEV is enabled, we have CLK support. Signed-off-by: Russell King diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index c5408bf..bcaa8c8 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -238,7 +238,6 @@ config ARCH_INTEGRATOR bool "ARM Ltd. Integrator family" select ARM_AMBA select ARCH_HAS_CPUFREQ - select HAVE_CLK select COMMON_CLKDEV select ICST525 help @@ -247,7 +246,6 @@ config ARCH_INTEGRATOR config ARCH_REALVIEW bool "ARM Ltd. RealView family" select ARM_AMBA - select HAVE_CLK select COMMON_CLKDEV select ICST307 select GENERIC_TIME @@ -260,7 +258,6 @@ config ARCH_VERSATILE bool "ARM Ltd. Versatile family" select ARM_AMBA select ARM_VIC - select HAVE_CLK select COMMON_CLKDEV select ICST307 select GENERIC_TIME @@ -321,7 +318,6 @@ config ARCH_EP93XX select ARM_AMBA select ARM_VIC select GENERIC_GPIO - select HAVE_CLK select COMMON_CLKDEV select ARCH_REQUIRE_GPIOLIB select ARCH_HAS_HOLES_MEMORYMODEL @@ -341,7 +337,6 @@ config ARCH_MXC select GENERIC_TIME select GENERIC_CLOCKEVENTS select ARCH_REQUIRE_GPIOLIB - select HAVE_CLK select COMMON_CLKDEV help Support for Freescale MXC/iMX-based family of processors @@ -349,7 +344,6 @@ config ARCH_MXC config ARCH_STMP3XXX bool "Freescale STMP3xxx" select CPU_ARM926T - select HAVE_CLK select COMMON_CLKDEV select ARCH_REQUIRE_GPIOLIB select GENERIC_TIME @@ -516,7 +510,6 @@ config ARCH_MMP depends on MMU select GENERIC_GPIO select ARCH_REQUIRE_GPIOLIB - select HAVE_CLK select COMMON_CLKDEV select GENERIC_TIME select GENERIC_CLOCKEVENTS @@ -552,7 +545,6 @@ config ARCH_W90X900 select CPU_ARM926T select ARCH_REQUIRE_GPIOLIB select GENERIC_GPIO - select HAVE_CLK select COMMON_CLKDEV select GENERIC_TIME select GENERIC_CLOCKEVENTS @@ -568,7 +560,6 @@ config ARCH_W90X900 config ARCH_NUC93X bool "Nuvoton NUC93X CPU" select CPU_ARM926T - select HAVE_CLK select COMMON_CLKDEV help Support for Nuvoton (Winbond logic dept.) NUC93X MCU,The NUC93X is a @@ -577,7 +568,6 @@ config ARCH_NUC93X config ARCH_PNX4008 bool "Philips Nexperia PNX4008 Mobile" select CPU_ARM926T - select HAVE_CLK select COMMON_CLKDEV help This enables support for Philips PNX4008 mobile platform. @@ -588,7 +578,6 @@ config ARCH_PXA select ARCH_MTD_XIP select ARCH_HAS_CPUFREQ select GENERIC_GPIO - select HAVE_CLK select COMMON_CLKDEV select ARCH_REQUIRE_GPIOLIB select GENERIC_TIME @@ -743,7 +732,6 @@ config ARCH_U300 select ARM_VIC select GENERIC_TIME select GENERIC_CLOCKEVENTS - select HAVE_CLK select COMMON_CLKDEV select GENERIC_GPIO help @@ -764,7 +752,6 @@ config ARCH_NOMADIK select ARM_AMBA select ARM_VIC select CPU_ARM926T - select HAVE_CLK select COMMON_CLKDEV select GENERIC_TIME select GENERIC_CLOCKEVENTS @@ -780,7 +767,6 @@ config ARCH_DAVINCI select GENERIC_CLOCKEVENTS select GENERIC_GPIO select ARCH_REQUIRE_GPIOLIB - select HAVE_CLK select ZONE_DMA select HAVE_IDE select COMMON_CLKDEV diff --git a/arch/arm/common/Kconfig b/arch/arm/common/Kconfig index 4efbb9d..b64bb78 100644 --- a/arch/arm/common/Kconfig +++ b/arch/arm/common/Kconfig @@ -40,3 +40,4 @@ config SHARP_SCOOP config COMMON_CLKDEV bool + select HAVE_CLK -- cgit v0.10.2 From e4bf5becccf4685754c4d8e4485bb2ff7d28147f Mon Sep 17 00:00:00 2001 From: viresh kumar Date: Tue, 9 Mar 2010 11:54:30 +0100 Subject: ARM: 5979/1: CLKDEV: return from clk_find as soon as we get maching clk clk_find function looks for clock with matching dev_id and con_id. We should return from clk_find as soon as we get a clock with matching dev_id and con_id. This can be done by comparing variable "match" with 3. Acked-by: Uwe Kleine-Konig Signed-off-by: Viresh Kumar Signed-off-by: Russell King diff --git a/arch/arm/common/clkdev.c b/arch/arm/common/clkdev.c index 6416d5b..4f8390d 100644 --- a/arch/arm/common/clkdev.c +++ b/arch/arm/common/clkdev.c @@ -52,12 +52,13 @@ static struct clk *clk_find(const char *dev_id, const char *con_id) continue; match += 1; } - if (match == 0) - continue; if (match > best) { clk = p->clk; - best = match; + if (match != 3) + best = match; + else + break; } } return clk; -- cgit v0.10.2 From 3a5e854e09b77c49c8030c7dfd3a3f506571a89a Mon Sep 17 00:00:00 2001 From: Rabin Vincent Date: Wed, 3 Mar 2010 04:46:58 +0100 Subject: ARM: 5969/1: nomadik: move gpio to plat-nomadik Move the Nomadik GPIO driver to plat-nomadik so that it can also be used on the Ux500 platform. The mach-nomadik include/mach/gpio.h is replaced by one that includes the one now in plat-nomadik, so that code doesn't need to include the one in plat specifically, and can instead use as usual. Acked-by: Alessandro Rubini Acked-by: Linus Walleij Signed-off-by: Rabin Vincent Signed-off-by: Russell King diff --git a/arch/arm/mach-nomadik/Kconfig b/arch/arm/mach-nomadik/Kconfig index 3c5e0f5..71f3ea6 100644 --- a/arch/arm/mach-nomadik/Kconfig +++ b/arch/arm/mach-nomadik/Kconfig @@ -6,6 +6,7 @@ config MACH_NOMADIK_8815NHK bool "ST 8815 Nomadik Hardware Kit (evaluation board)" select NOMADIK_8815 select HAS_MTU + select NOMADIK_GPIO endmenu diff --git a/arch/arm/mach-nomadik/Makefile b/arch/arm/mach-nomadik/Makefile index 36f67fb..a6bbd1a 100644 --- a/arch/arm/mach-nomadik/Makefile +++ b/arch/arm/mach-nomadik/Makefile @@ -7,7 +7,7 @@ # Object file lists. -obj-y += clock.o gpio.o +obj-y += clock.o # Cpu revision obj-$(CONFIG_NOMADIK_8815) += cpu-8815.o diff --git a/arch/arm/mach-nomadik/gpio.c b/arch/arm/mach-nomadik/gpio.c deleted file mode 100644 index 9a09b27..0000000 --- a/arch/arm/mach-nomadik/gpio.c +++ /dev/null @@ -1,396 +0,0 @@ -/* - * Generic GPIO driver for logic cells found in the Nomadik SoC - * - * Copyright (C) 2008,2009 STMicroelectronics - * Copyright (C) 2009 Alessandro Rubini - * Rewritten based on work by Prafulla WADASKAR - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - */ -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#include -#include - -/* - * The GPIO module in the Nomadik family of Systems-on-Chip is an - * AMBA device, managing 32 pins and alternate functions. The logic block - * is currently only used in the Nomadik. - * - * Symbols in this file are called "nmk_gpio" for "nomadik gpio" - */ - -#define NMK_GPIO_PER_CHIP 32 -struct nmk_gpio_chip { - struct gpio_chip chip; - void __iomem *addr; - unsigned int parent_irq; - spinlock_t *lock; - /* Keep track of configured edges */ - u32 edge_rising; - u32 edge_falling; -}; - -/* Mode functions */ -int nmk_gpio_set_mode(int gpio, int gpio_mode) -{ - struct nmk_gpio_chip *nmk_chip; - unsigned long flags; - u32 afunc, bfunc, bit; - - nmk_chip = get_irq_chip_data(NOMADIK_GPIO_TO_IRQ(gpio)); - if (!nmk_chip) - return -EINVAL; - - bit = 1 << (gpio - nmk_chip->chip.base); - - spin_lock_irqsave(&nmk_chip->lock, flags); - afunc = readl(nmk_chip->addr + NMK_GPIO_AFSLA) & ~bit; - bfunc = readl(nmk_chip->addr + NMK_GPIO_AFSLB) & ~bit; - if (gpio_mode & NMK_GPIO_ALT_A) - afunc |= bit; - if (gpio_mode & NMK_GPIO_ALT_B) - bfunc |= bit; - writel(afunc, nmk_chip->addr + NMK_GPIO_AFSLA); - writel(bfunc, nmk_chip->addr + NMK_GPIO_AFSLB); - spin_unlock_irqrestore(&nmk_chip->lock, flags); - - return 0; -} -EXPORT_SYMBOL(nmk_gpio_set_mode); - -int nmk_gpio_get_mode(int gpio) -{ - struct nmk_gpio_chip *nmk_chip; - u32 afunc, bfunc, bit; - - nmk_chip = get_irq_chip_data(NOMADIK_GPIO_TO_IRQ(gpio)); - if (!nmk_chip) - return -EINVAL; - - bit = 1 << (gpio - nmk_chip->chip.base); - - afunc = readl(nmk_chip->addr + NMK_GPIO_AFSLA) & bit; - bfunc = readl(nmk_chip->addr + NMK_GPIO_AFSLB) & bit; - - return (afunc ? NMK_GPIO_ALT_A : 0) | (bfunc ? NMK_GPIO_ALT_B : 0); -} -EXPORT_SYMBOL(nmk_gpio_get_mode); - - -/* IRQ functions */ -static inline int nmk_gpio_get_bitmask(int gpio) -{ - return 1 << (gpio % 32); -} - -static void nmk_gpio_irq_ack(unsigned int irq) -{ - int gpio; - struct nmk_gpio_chip *nmk_chip; - - gpio = NOMADIK_IRQ_TO_GPIO(irq); - nmk_chip = get_irq_chip_data(irq); - if (!nmk_chip) - return; - writel(nmk_gpio_get_bitmask(gpio), nmk_chip->addr + NMK_GPIO_IC); -} - -static void nmk_gpio_irq_mask(unsigned int irq) -{ - int gpio; - struct nmk_gpio_chip *nmk_chip; - unsigned long flags; - u32 bitmask, reg; - - gpio = NOMADIK_IRQ_TO_GPIO(irq); - nmk_chip = get_irq_chip_data(irq); - bitmask = nmk_gpio_get_bitmask(gpio); - if (!nmk_chip) - return; - - /* we must individually clear the two edges */ - spin_lock_irqsave(&nmk_chip->lock, flags); - if (nmk_chip->edge_rising & bitmask) { - reg = readl(nmk_chip->addr + NMK_GPIO_RWIMSC); - reg &= ~bitmask; - writel(reg, nmk_chip->addr + NMK_GPIO_RWIMSC); - } - if (nmk_chip->edge_falling & bitmask) { - reg = readl(nmk_chip->addr + NMK_GPIO_FWIMSC); - reg &= ~bitmask; - writel(reg, nmk_chip->addr + NMK_GPIO_FWIMSC); - } - spin_unlock_irqrestore(&nmk_chip->lock, flags); -}; - -static void nmk_gpio_irq_unmask(unsigned int irq) -{ - int gpio; - struct nmk_gpio_chip *nmk_chip; - unsigned long flags; - u32 bitmask, reg; - - gpio = NOMADIK_IRQ_TO_GPIO(irq); - nmk_chip = get_irq_chip_data(irq); - bitmask = nmk_gpio_get_bitmask(gpio); - if (!nmk_chip) - return; - - /* we must individually set the two edges */ - spin_lock_irqsave(&nmk_chip->lock, flags); - if (nmk_chip->edge_rising & bitmask) { - reg = readl(nmk_chip->addr + NMK_GPIO_RWIMSC); - reg |= bitmask; - writel(reg, nmk_chip->addr + NMK_GPIO_RWIMSC); - } - if (nmk_chip->edge_falling & bitmask) { - reg = readl(nmk_chip->addr + NMK_GPIO_FWIMSC); - reg |= bitmask; - writel(reg, nmk_chip->addr + NMK_GPIO_FWIMSC); - } - spin_unlock_irqrestore(&nmk_chip->lock, flags); -} - -static int nmk_gpio_irq_set_type(unsigned int irq, unsigned int type) -{ - int gpio; - struct nmk_gpio_chip *nmk_chip; - unsigned long flags; - u32 bitmask; - - gpio = NOMADIK_IRQ_TO_GPIO(irq); - nmk_chip = get_irq_chip_data(irq); - bitmask = nmk_gpio_get_bitmask(gpio); - if (!nmk_chip) - return -EINVAL; - - if (type & IRQ_TYPE_LEVEL_HIGH) - return -EINVAL; - if (type & IRQ_TYPE_LEVEL_LOW) - return -EINVAL; - - spin_lock_irqsave(&nmk_chip->lock, flags); - - nmk_chip->edge_rising &= ~bitmask; - if (type & IRQ_TYPE_EDGE_RISING) - nmk_chip->edge_rising |= bitmask; - writel(nmk_chip->edge_rising, nmk_chip->addr + NMK_GPIO_RIMSC); - - nmk_chip->edge_falling &= ~bitmask; - if (type & IRQ_TYPE_EDGE_FALLING) - nmk_chip->edge_falling |= bitmask; - writel(nmk_chip->edge_falling, nmk_chip->addr + NMK_GPIO_FIMSC); - - spin_unlock_irqrestore(&nmk_chip->lock, flags); - - nmk_gpio_irq_unmask(irq); - - return 0; -} - -static struct irq_chip nmk_gpio_irq_chip = { - .name = "Nomadik-GPIO", - .ack = nmk_gpio_irq_ack, - .mask = nmk_gpio_irq_mask, - .unmask = nmk_gpio_irq_unmask, - .set_type = nmk_gpio_irq_set_type, -}; - -static void nmk_gpio_irq_handler(unsigned int irq, struct irq_desc *desc) -{ - struct nmk_gpio_chip *nmk_chip; - struct irq_chip *host_chip; - unsigned int gpio_irq; - u32 pending; - unsigned int first_irq; - - nmk_chip = get_irq_data(irq); - first_irq = NOMADIK_GPIO_TO_IRQ(nmk_chip->chip.base); - while ( (pending = readl(nmk_chip->addr + NMK_GPIO_IS)) ) { - gpio_irq = first_irq + __ffs(pending); - generic_handle_irq(gpio_irq); - } - if (0) {/* don't ack parent irq, as ack == disable */ - host_chip = get_irq_chip(irq); - host_chip->ack(irq); - } -} - -static int nmk_gpio_init_irq(struct nmk_gpio_chip *nmk_chip) -{ - unsigned int first_irq; - int i; - - first_irq = NOMADIK_GPIO_TO_IRQ(nmk_chip->chip.base); - for (i = first_irq; i < first_irq + NMK_GPIO_PER_CHIP; i++) { - set_irq_chip(i, &nmk_gpio_irq_chip); - set_irq_handler(i, handle_edge_irq); - set_irq_flags(i, IRQF_VALID); - set_irq_chip_data(i, nmk_chip); - } - set_irq_chained_handler(nmk_chip->parent_irq, nmk_gpio_irq_handler); - set_irq_data(nmk_chip->parent_irq, nmk_chip); - return 0; -} - -/* I/O Functions */ -static int nmk_gpio_make_input(struct gpio_chip *chip, unsigned offset) -{ - struct nmk_gpio_chip *nmk_chip = - container_of(chip, struct nmk_gpio_chip, chip); - - writel(1 << offset, nmk_chip->addr + NMK_GPIO_DIRC); - return 0; -} - -static int nmk_gpio_make_output(struct gpio_chip *chip, unsigned offset, - int val) -{ - struct nmk_gpio_chip *nmk_chip = - container_of(chip, struct nmk_gpio_chip, chip); - - writel(1 << offset, nmk_chip->addr + NMK_GPIO_DIRS); - return 0; -} - -static int nmk_gpio_get_input(struct gpio_chip *chip, unsigned offset) -{ - struct nmk_gpio_chip *nmk_chip = - container_of(chip, struct nmk_gpio_chip, chip); - u32 bit = 1 << offset; - - return (readl(nmk_chip->addr + NMK_GPIO_DAT) & bit) != 0; -} - -static void nmk_gpio_set_output(struct gpio_chip *chip, unsigned offset, - int val) -{ - struct nmk_gpio_chip *nmk_chip = - container_of(chip, struct nmk_gpio_chip, chip); - u32 bit = 1 << offset; - - if (val) - writel(bit, nmk_chip->addr + NMK_GPIO_DATS); - else - writel(bit, nmk_chip->addr + NMK_GPIO_DATC); -} - -/* This structure is replicated for each GPIO block allocated at probe time */ -static struct gpio_chip nmk_gpio_template = { - .direction_input = nmk_gpio_make_input, - .get = nmk_gpio_get_input, - .direction_output = nmk_gpio_make_output, - .set = nmk_gpio_set_output, - .ngpio = NMK_GPIO_PER_CHIP, - .can_sleep = 0, -}; - -static int __init nmk_gpio_probe(struct amba_device *dev, struct amba_id *id) -{ - struct nmk_gpio_platform_data *pdata; - struct nmk_gpio_chip *nmk_chip; - struct gpio_chip *chip; - int ret; - - pdata = dev->dev.platform_data; - ret = amba_request_regions(dev, pdata->name); - if (ret) - return ret; - - nmk_chip = kzalloc(sizeof(*nmk_chip), GFP_KERNEL); - if (!nmk_chip) { - ret = -ENOMEM; - goto out_amba; - } - /* - * The virt address in nmk_chip->addr is in the nomadik register space, - * so we can simply convert the resource address, without remapping - */ - nmk_chip->addr = io_p2v(dev->res.start); - nmk_chip->chip = nmk_gpio_template; - nmk_chip->parent_irq = pdata->parent_irq; - - chip = &nmk_chip->chip; - chip->base = pdata->first_gpio; - chip->label = pdata->name; - chip->dev = &dev->dev; - chip->owner = THIS_MODULE; - - ret = gpiochip_add(&nmk_chip->chip); - if (ret) - goto out_free; - - amba_set_drvdata(dev, nmk_chip); - - nmk_gpio_init_irq(nmk_chip); - - dev_info(&dev->dev, "Bits %i-%i at address %p\n", - nmk_chip->chip.base, nmk_chip->chip.base+31, nmk_chip->addr); - return 0; - - out_free: - kfree(nmk_chip); - out_amba: - amba_release_regions(dev); - dev_err(&dev->dev, "Failure %i for GPIO %i-%i\n", ret, - pdata->first_gpio, pdata->first_gpio+31); - return ret; -} - -static int nmk_gpio_remove(struct amba_device *dev) -{ - struct nmk_gpio_chip *nmk_chip; - - nmk_chip = amba_get_drvdata(dev); - gpiochip_remove(&nmk_chip->chip); - kfree(nmk_chip); - amba_release_regions(dev); - return 0; -} - - -/* We have 0x1f080060 and 0x1f180060, accept both using the mask */ -static struct amba_id nmk_gpio_ids[] = { - { - .id = 0x1f080060, - .mask = 0xffefffff, - }, - {0, 0}, -}; - -static struct amba_driver nmk_gpio_driver = { - .drv = { - .owner = THIS_MODULE, - .name = "gpio", - }, - .probe = nmk_gpio_probe, - .remove = nmk_gpio_remove, - .suspend = NULL, /* to be done */ - .resume = NULL, - .id_table = nmk_gpio_ids, -}; - -static int __init nmk_gpio_init(void) -{ - return amba_driver_register(&nmk_gpio_driver); -} - -arch_initcall(nmk_gpio_init); - -MODULE_AUTHOR("Prafulla WADASKAR and Alessandro Rubini"); -MODULE_DESCRIPTION("Nomadik GPIO Driver"); -MODULE_LICENSE("GPL"); - - diff --git a/arch/arm/mach-nomadik/include/mach/gpio.h b/arch/arm/mach-nomadik/include/mach/gpio.h index 61577c9..7a81a04 100644 --- a/arch/arm/mach-nomadik/include/mach/gpio.h +++ b/arch/arm/mach-nomadik/include/mach/gpio.h @@ -1,71 +1,6 @@ -/* - * Structures and registers for GPIO access in the Nomadik SoC - * - * Copyright (C) 2008 STMicroelectronics - * Author: Prafulla WADASKAR - * Copyright (C) 2009 Alessandro Rubini - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - */ #ifndef __ASM_ARCH_GPIO_H #define __ASM_ARCH_GPIO_H -#include - -/* - * These currently cause a function call to happen, they may be optimized - * if needed by adding cpu-specific defines to identify blocks - * (see mach-pxa/include/mach/gpio.h as an example using GPLR etc) - */ -#define gpio_get_value __gpio_get_value -#define gpio_set_value __gpio_set_value -#define gpio_cansleep __gpio_cansleep -#define gpio_to_irq __gpio_to_irq - -/* - * "nmk_gpio" and "NMK_GPIO" stand for "Nomadik GPIO", leaving - * the "gpio" namespace for generic and cross-machine functions - */ - -/* Register in the logic block */ -#define NMK_GPIO_DAT 0x00 -#define NMK_GPIO_DATS 0x04 -#define NMK_GPIO_DATC 0x08 -#define NMK_GPIO_PDIS 0x0c -#define NMK_GPIO_DIR 0x10 -#define NMK_GPIO_DIRS 0x14 -#define NMK_GPIO_DIRC 0x18 -#define NMK_GPIO_SLPC 0x1c -#define NMK_GPIO_AFSLA 0x20 -#define NMK_GPIO_AFSLB 0x24 - -#define NMK_GPIO_RIMSC 0x40 -#define NMK_GPIO_FIMSC 0x44 -#define NMK_GPIO_IS 0x48 -#define NMK_GPIO_IC 0x4c -#define NMK_GPIO_RWIMSC 0x50 -#define NMK_GPIO_FWIMSC 0x54 -#define NMK_GPIO_WKS 0x58 - -/* Alternate functions: function C is set in hw by setting both A and B */ -#define NMK_GPIO_ALT_GPIO 0 -#define NMK_GPIO_ALT_A 1 -#define NMK_GPIO_ALT_B 2 -#define NMK_GPIO_ALT_C (NMK_GPIO_ALT_A | NMK_GPIO_ALT_B) - -extern int nmk_gpio_set_mode(int gpio, int gpio_mode); -extern int nmk_gpio_get_mode(int gpio); - -/* - * Platform data to register a block: only the initial gpio/irq number. - */ -struct nmk_gpio_platform_data { - char *name; - int first_gpio; - int first_irq; - int parent_irq; -}; +#include #endif /* __ASM_ARCH_GPIO_H */ diff --git a/arch/arm/plat-nomadik/Kconfig b/arch/arm/plat-nomadik/Kconfig index 159daf5..5da3f97 100644 --- a/arch/arm/plat-nomadik/Kconfig +++ b/arch/arm/plat-nomadik/Kconfig @@ -19,4 +19,9 @@ config HAS_MTU to multiple interrupt generating programmable 32-bit free running decrementing counters. +config NOMADIK_GPIO + bool + help + Support for the Nomadik GPIO controller. + endif diff --git a/arch/arm/plat-nomadik/Makefile b/arch/arm/plat-nomadik/Makefile index 37c7cdd..c335473 100644 --- a/arch/arm/plat-nomadik/Makefile +++ b/arch/arm/plat-nomadik/Makefile @@ -3,3 +3,4 @@ # Licensed under GPLv2 obj-$(CONFIG_HAS_MTU) += timer.o +obj-$(CONFIG_NOMADIK_GPIO) += gpio.o diff --git a/arch/arm/plat-nomadik/gpio.c b/arch/arm/plat-nomadik/gpio.c new file mode 100644 index 0000000..9a09b27 --- /dev/null +++ b/arch/arm/plat-nomadik/gpio.c @@ -0,0 +1,396 @@ +/* + * Generic GPIO driver for logic cells found in the Nomadik SoC + * + * Copyright (C) 2008,2009 STMicroelectronics + * Copyright (C) 2009 Alessandro Rubini + * Rewritten based on work by Prafulla WADASKAR + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include + +/* + * The GPIO module in the Nomadik family of Systems-on-Chip is an + * AMBA device, managing 32 pins and alternate functions. The logic block + * is currently only used in the Nomadik. + * + * Symbols in this file are called "nmk_gpio" for "nomadik gpio" + */ + +#define NMK_GPIO_PER_CHIP 32 +struct nmk_gpio_chip { + struct gpio_chip chip; + void __iomem *addr; + unsigned int parent_irq; + spinlock_t *lock; + /* Keep track of configured edges */ + u32 edge_rising; + u32 edge_falling; +}; + +/* Mode functions */ +int nmk_gpio_set_mode(int gpio, int gpio_mode) +{ + struct nmk_gpio_chip *nmk_chip; + unsigned long flags; + u32 afunc, bfunc, bit; + + nmk_chip = get_irq_chip_data(NOMADIK_GPIO_TO_IRQ(gpio)); + if (!nmk_chip) + return -EINVAL; + + bit = 1 << (gpio - nmk_chip->chip.base); + + spin_lock_irqsave(&nmk_chip->lock, flags); + afunc = readl(nmk_chip->addr + NMK_GPIO_AFSLA) & ~bit; + bfunc = readl(nmk_chip->addr + NMK_GPIO_AFSLB) & ~bit; + if (gpio_mode & NMK_GPIO_ALT_A) + afunc |= bit; + if (gpio_mode & NMK_GPIO_ALT_B) + bfunc |= bit; + writel(afunc, nmk_chip->addr + NMK_GPIO_AFSLA); + writel(bfunc, nmk_chip->addr + NMK_GPIO_AFSLB); + spin_unlock_irqrestore(&nmk_chip->lock, flags); + + return 0; +} +EXPORT_SYMBOL(nmk_gpio_set_mode); + +int nmk_gpio_get_mode(int gpio) +{ + struct nmk_gpio_chip *nmk_chip; + u32 afunc, bfunc, bit; + + nmk_chip = get_irq_chip_data(NOMADIK_GPIO_TO_IRQ(gpio)); + if (!nmk_chip) + return -EINVAL; + + bit = 1 << (gpio - nmk_chip->chip.base); + + afunc = readl(nmk_chip->addr + NMK_GPIO_AFSLA) & bit; + bfunc = readl(nmk_chip->addr + NMK_GPIO_AFSLB) & bit; + + return (afunc ? NMK_GPIO_ALT_A : 0) | (bfunc ? NMK_GPIO_ALT_B : 0); +} +EXPORT_SYMBOL(nmk_gpio_get_mode); + + +/* IRQ functions */ +static inline int nmk_gpio_get_bitmask(int gpio) +{ + return 1 << (gpio % 32); +} + +static void nmk_gpio_irq_ack(unsigned int irq) +{ + int gpio; + struct nmk_gpio_chip *nmk_chip; + + gpio = NOMADIK_IRQ_TO_GPIO(irq); + nmk_chip = get_irq_chip_data(irq); + if (!nmk_chip) + return; + writel(nmk_gpio_get_bitmask(gpio), nmk_chip->addr + NMK_GPIO_IC); +} + +static void nmk_gpio_irq_mask(unsigned int irq) +{ + int gpio; + struct nmk_gpio_chip *nmk_chip; + unsigned long flags; + u32 bitmask, reg; + + gpio = NOMADIK_IRQ_TO_GPIO(irq); + nmk_chip = get_irq_chip_data(irq); + bitmask = nmk_gpio_get_bitmask(gpio); + if (!nmk_chip) + return; + + /* we must individually clear the two edges */ + spin_lock_irqsave(&nmk_chip->lock, flags); + if (nmk_chip->edge_rising & bitmask) { + reg = readl(nmk_chip->addr + NMK_GPIO_RWIMSC); + reg &= ~bitmask; + writel(reg, nmk_chip->addr + NMK_GPIO_RWIMSC); + } + if (nmk_chip->edge_falling & bitmask) { + reg = readl(nmk_chip->addr + NMK_GPIO_FWIMSC); + reg &= ~bitmask; + writel(reg, nmk_chip->addr + NMK_GPIO_FWIMSC); + } + spin_unlock_irqrestore(&nmk_chip->lock, flags); +}; + +static void nmk_gpio_irq_unmask(unsigned int irq) +{ + int gpio; + struct nmk_gpio_chip *nmk_chip; + unsigned long flags; + u32 bitmask, reg; + + gpio = NOMADIK_IRQ_TO_GPIO(irq); + nmk_chip = get_irq_chip_data(irq); + bitmask = nmk_gpio_get_bitmask(gpio); + if (!nmk_chip) + return; + + /* we must individually set the two edges */ + spin_lock_irqsave(&nmk_chip->lock, flags); + if (nmk_chip->edge_rising & bitmask) { + reg = readl(nmk_chip->addr + NMK_GPIO_RWIMSC); + reg |= bitmask; + writel(reg, nmk_chip->addr + NMK_GPIO_RWIMSC); + } + if (nmk_chip->edge_falling & bitmask) { + reg = readl(nmk_chip->addr + NMK_GPIO_FWIMSC); + reg |= bitmask; + writel(reg, nmk_chip->addr + NMK_GPIO_FWIMSC); + } + spin_unlock_irqrestore(&nmk_chip->lock, flags); +} + +static int nmk_gpio_irq_set_type(unsigned int irq, unsigned int type) +{ + int gpio; + struct nmk_gpio_chip *nmk_chip; + unsigned long flags; + u32 bitmask; + + gpio = NOMADIK_IRQ_TO_GPIO(irq); + nmk_chip = get_irq_chip_data(irq); + bitmask = nmk_gpio_get_bitmask(gpio); + if (!nmk_chip) + return -EINVAL; + + if (type & IRQ_TYPE_LEVEL_HIGH) + return -EINVAL; + if (type & IRQ_TYPE_LEVEL_LOW) + return -EINVAL; + + spin_lock_irqsave(&nmk_chip->lock, flags); + + nmk_chip->edge_rising &= ~bitmask; + if (type & IRQ_TYPE_EDGE_RISING) + nmk_chip->edge_rising |= bitmask; + writel(nmk_chip->edge_rising, nmk_chip->addr + NMK_GPIO_RIMSC); + + nmk_chip->edge_falling &= ~bitmask; + if (type & IRQ_TYPE_EDGE_FALLING) + nmk_chip->edge_falling |= bitmask; + writel(nmk_chip->edge_falling, nmk_chip->addr + NMK_GPIO_FIMSC); + + spin_unlock_irqrestore(&nmk_chip->lock, flags); + + nmk_gpio_irq_unmask(irq); + + return 0; +} + +static struct irq_chip nmk_gpio_irq_chip = { + .name = "Nomadik-GPIO", + .ack = nmk_gpio_irq_ack, + .mask = nmk_gpio_irq_mask, + .unmask = nmk_gpio_irq_unmask, + .set_type = nmk_gpio_irq_set_type, +}; + +static void nmk_gpio_irq_handler(unsigned int irq, struct irq_desc *desc) +{ + struct nmk_gpio_chip *nmk_chip; + struct irq_chip *host_chip; + unsigned int gpio_irq; + u32 pending; + unsigned int first_irq; + + nmk_chip = get_irq_data(irq); + first_irq = NOMADIK_GPIO_TO_IRQ(nmk_chip->chip.base); + while ( (pending = readl(nmk_chip->addr + NMK_GPIO_IS)) ) { + gpio_irq = first_irq + __ffs(pending); + generic_handle_irq(gpio_irq); + } + if (0) {/* don't ack parent irq, as ack == disable */ + host_chip = get_irq_chip(irq); + host_chip->ack(irq); + } +} + +static int nmk_gpio_init_irq(struct nmk_gpio_chip *nmk_chip) +{ + unsigned int first_irq; + int i; + + first_irq = NOMADIK_GPIO_TO_IRQ(nmk_chip->chip.base); + for (i = first_irq; i < first_irq + NMK_GPIO_PER_CHIP; i++) { + set_irq_chip(i, &nmk_gpio_irq_chip); + set_irq_handler(i, handle_edge_irq); + set_irq_flags(i, IRQF_VALID); + set_irq_chip_data(i, nmk_chip); + } + set_irq_chained_handler(nmk_chip->parent_irq, nmk_gpio_irq_handler); + set_irq_data(nmk_chip->parent_irq, nmk_chip); + return 0; +} + +/* I/O Functions */ +static int nmk_gpio_make_input(struct gpio_chip *chip, unsigned offset) +{ + struct nmk_gpio_chip *nmk_chip = + container_of(chip, struct nmk_gpio_chip, chip); + + writel(1 << offset, nmk_chip->addr + NMK_GPIO_DIRC); + return 0; +} + +static int nmk_gpio_make_output(struct gpio_chip *chip, unsigned offset, + int val) +{ + struct nmk_gpio_chip *nmk_chip = + container_of(chip, struct nmk_gpio_chip, chip); + + writel(1 << offset, nmk_chip->addr + NMK_GPIO_DIRS); + return 0; +} + +static int nmk_gpio_get_input(struct gpio_chip *chip, unsigned offset) +{ + struct nmk_gpio_chip *nmk_chip = + container_of(chip, struct nmk_gpio_chip, chip); + u32 bit = 1 << offset; + + return (readl(nmk_chip->addr + NMK_GPIO_DAT) & bit) != 0; +} + +static void nmk_gpio_set_output(struct gpio_chip *chip, unsigned offset, + int val) +{ + struct nmk_gpio_chip *nmk_chip = + container_of(chip, struct nmk_gpio_chip, chip); + u32 bit = 1 << offset; + + if (val) + writel(bit, nmk_chip->addr + NMK_GPIO_DATS); + else + writel(bit, nmk_chip->addr + NMK_GPIO_DATC); +} + +/* This structure is replicated for each GPIO block allocated at probe time */ +static struct gpio_chip nmk_gpio_template = { + .direction_input = nmk_gpio_make_input, + .get = nmk_gpio_get_input, + .direction_output = nmk_gpio_make_output, + .set = nmk_gpio_set_output, + .ngpio = NMK_GPIO_PER_CHIP, + .can_sleep = 0, +}; + +static int __init nmk_gpio_probe(struct amba_device *dev, struct amba_id *id) +{ + struct nmk_gpio_platform_data *pdata; + struct nmk_gpio_chip *nmk_chip; + struct gpio_chip *chip; + int ret; + + pdata = dev->dev.platform_data; + ret = amba_request_regions(dev, pdata->name); + if (ret) + return ret; + + nmk_chip = kzalloc(sizeof(*nmk_chip), GFP_KERNEL); + if (!nmk_chip) { + ret = -ENOMEM; + goto out_amba; + } + /* + * The virt address in nmk_chip->addr is in the nomadik register space, + * so we can simply convert the resource address, without remapping + */ + nmk_chip->addr = io_p2v(dev->res.start); + nmk_chip->chip = nmk_gpio_template; + nmk_chip->parent_irq = pdata->parent_irq; + + chip = &nmk_chip->chip; + chip->base = pdata->first_gpio; + chip->label = pdata->name; + chip->dev = &dev->dev; + chip->owner = THIS_MODULE; + + ret = gpiochip_add(&nmk_chip->chip); + if (ret) + goto out_free; + + amba_set_drvdata(dev, nmk_chip); + + nmk_gpio_init_irq(nmk_chip); + + dev_info(&dev->dev, "Bits %i-%i at address %p\n", + nmk_chip->chip.base, nmk_chip->chip.base+31, nmk_chip->addr); + return 0; + + out_free: + kfree(nmk_chip); + out_amba: + amba_release_regions(dev); + dev_err(&dev->dev, "Failure %i for GPIO %i-%i\n", ret, + pdata->first_gpio, pdata->first_gpio+31); + return ret; +} + +static int nmk_gpio_remove(struct amba_device *dev) +{ + struct nmk_gpio_chip *nmk_chip; + + nmk_chip = amba_get_drvdata(dev); + gpiochip_remove(&nmk_chip->chip); + kfree(nmk_chip); + amba_release_regions(dev); + return 0; +} + + +/* We have 0x1f080060 and 0x1f180060, accept both using the mask */ +static struct amba_id nmk_gpio_ids[] = { + { + .id = 0x1f080060, + .mask = 0xffefffff, + }, + {0, 0}, +}; + +static struct amba_driver nmk_gpio_driver = { + .drv = { + .owner = THIS_MODULE, + .name = "gpio", + }, + .probe = nmk_gpio_probe, + .remove = nmk_gpio_remove, + .suspend = NULL, /* to be done */ + .resume = NULL, + .id_table = nmk_gpio_ids, +}; + +static int __init nmk_gpio_init(void) +{ + return amba_driver_register(&nmk_gpio_driver); +} + +arch_initcall(nmk_gpio_init); + +MODULE_AUTHOR("Prafulla WADASKAR and Alessandro Rubini"); +MODULE_DESCRIPTION("Nomadik GPIO Driver"); +MODULE_LICENSE("GPL"); + + diff --git a/arch/arm/plat-nomadik/include/plat/gpio.h b/arch/arm/plat-nomadik/include/plat/gpio.h new file mode 100644 index 0000000..1d665a0a --- /dev/null +++ b/arch/arm/plat-nomadik/include/plat/gpio.h @@ -0,0 +1,71 @@ +/* + * Structures and registers for GPIO access in the Nomadik SoC + * + * Copyright (C) 2008 STMicroelectronics + * Author: Prafulla WADASKAR + * Copyright (C) 2009 Alessandro Rubini + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#ifndef __ASM_PLAT_GPIO_H +#define __ASM_PLAT_GPIO_H + +#include + +/* + * These currently cause a function call to happen, they may be optimized + * if needed by adding cpu-specific defines to identify blocks + * (see mach-pxa/include/mach/gpio.h as an example using GPLR etc) + */ +#define gpio_get_value __gpio_get_value +#define gpio_set_value __gpio_set_value +#define gpio_cansleep __gpio_cansleep +#define gpio_to_irq __gpio_to_irq + +/* + * "nmk_gpio" and "NMK_GPIO" stand for "Nomadik GPIO", leaving + * the "gpio" namespace for generic and cross-machine functions + */ + +/* Register in the logic block */ +#define NMK_GPIO_DAT 0x00 +#define NMK_GPIO_DATS 0x04 +#define NMK_GPIO_DATC 0x08 +#define NMK_GPIO_PDIS 0x0c +#define NMK_GPIO_DIR 0x10 +#define NMK_GPIO_DIRS 0x14 +#define NMK_GPIO_DIRC 0x18 +#define NMK_GPIO_SLPC 0x1c +#define NMK_GPIO_AFSLA 0x20 +#define NMK_GPIO_AFSLB 0x24 + +#define NMK_GPIO_RIMSC 0x40 +#define NMK_GPIO_FIMSC 0x44 +#define NMK_GPIO_IS 0x48 +#define NMK_GPIO_IC 0x4c +#define NMK_GPIO_RWIMSC 0x50 +#define NMK_GPIO_FWIMSC 0x54 +#define NMK_GPIO_WKS 0x58 + +/* Alternate functions: function C is set in hw by setting both A and B */ +#define NMK_GPIO_ALT_GPIO 0 +#define NMK_GPIO_ALT_A 1 +#define NMK_GPIO_ALT_B 2 +#define NMK_GPIO_ALT_C (NMK_GPIO_ALT_A | NMK_GPIO_ALT_B) + +extern int nmk_gpio_set_mode(int gpio, int gpio_mode); +extern int nmk_gpio_get_mode(int gpio); + +/* + * Platform data to register a block: only the initial gpio/irq number. + */ +struct nmk_gpio_platform_data { + char *name; + int first_gpio; + int first_irq; + int parent_irq; +}; + +#endif /* __ASM_PLAT_GPIO_H */ -- cgit v0.10.2 From c0fcb8dba829421fe20652a376a2bedaf168238f Mon Sep 17 00:00:00 2001 From: Rabin Vincent Date: Wed, 3 Mar 2010 04:48:54 +0100 Subject: ARM: 5970/1: nomadik-gpio: fix spinlock usage Fix the incorrect spinlock usage which leads to these compiler warnings: arch/arm/plat-nomadik/gpio.c:60: warning: passing argument 1 of `spinlock_check' from incompatible pointer type arch/arm/plat-nomadik/gpio.c:69: warning: passing argument 1 of `spin_unlock_irqrestore' from incompatible pointer type Acked-by: Alessandro Rubini Acked-by: Linus Walleij Signed-off-by: Rabin Vincent Signed-off-by: Russell King diff --git a/arch/arm/plat-nomadik/gpio.c b/arch/arm/plat-nomadik/gpio.c index 9a09b27..a9ee988 100644 --- a/arch/arm/plat-nomadik/gpio.c +++ b/arch/arm/plat-nomadik/gpio.c @@ -36,7 +36,7 @@ struct nmk_gpio_chip { struct gpio_chip chip; void __iomem *addr; unsigned int parent_irq; - spinlock_t *lock; + spinlock_t lock; /* Keep track of configured edges */ u32 edge_rising; u32 edge_falling; @@ -321,6 +321,7 @@ static int __init nmk_gpio_probe(struct amba_device *dev, struct amba_id *id) nmk_chip->addr = io_p2v(dev->res.start); nmk_chip->chip = nmk_gpio_template; nmk_chip->parent_irq = pdata->parent_irq; + spin_lock_init(&nmk_chip->lock); chip = &nmk_chip->chip; chip->base = pdata->first_gpio; -- cgit v0.10.2 From aaedaa2b5c610ae97f863078075d8d3c6ef91575 Mon Sep 17 00:00:00 2001 From: Rabin Vincent Date: Wed, 3 Mar 2010 04:50:27 +0100 Subject: ARM: 5971/1: nomadik-gpio: mask/ack/unmask the parent irq Since we register as a chained handler, we need to take care of acking the parent irq. Current code had the acking disabled because just acking the interrupt disables it on some interrupt controllers (like the VIC used on Nomadik). However, not acking at all is incorrect because interrupt controllers like the GIC need an EOI indication, which is done in the ack routine. Solve this by maskacking and unmasking it appropriately, similar to how handle_level_irq operates. Acked-by: Alessandro Rubini Acked-by: Linus Walleij Signed-off-by: Rabin Vincent Signed-off-by: Russell King diff --git a/arch/arm/plat-nomadik/gpio.c b/arch/arm/plat-nomadik/gpio.c index a9ee988..4c3ea1a 100644 --- a/arch/arm/plat-nomadik/gpio.c +++ b/arch/arm/plat-nomadik/gpio.c @@ -211,21 +211,27 @@ static struct irq_chip nmk_gpio_irq_chip = { static void nmk_gpio_irq_handler(unsigned int irq, struct irq_desc *desc) { struct nmk_gpio_chip *nmk_chip; - struct irq_chip *host_chip; + struct irq_chip *host_chip = get_irq_chip(irq); unsigned int gpio_irq; u32 pending; unsigned int first_irq; + if (host_chip->mask_ack) + host_chip->mask_ack(irq); + else { + host_chip->mask(irq); + if (host_chip->ack) + host_chip->ack(irq); + } + nmk_chip = get_irq_data(irq); first_irq = NOMADIK_GPIO_TO_IRQ(nmk_chip->chip.base); while ( (pending = readl(nmk_chip->addr + NMK_GPIO_IS)) ) { gpio_irq = first_irq + __ffs(pending); generic_handle_irq(gpio_irq); } - if (0) {/* don't ack parent irq, as ack == disable */ - host_chip = get_irq_chip(irq); - host_chip->ack(irq); - } + + host_chip->unmask(irq); } static int nmk_gpio_init_irq(struct nmk_gpio_chip *nmk_chip) -- cgit v0.10.2 From 3e3c62ca5395df48319b808379bc9fd487ff3c29 Mon Sep 17 00:00:00 2001 From: Rabin Vincent Date: Wed, 3 Mar 2010 04:52:34 +0100 Subject: ARM: 5972/1: nomadik-gpio: convert to platform driver On the U8500 platform there are four GPIO blocks, each with a 4K address space, including the peripheral identification. However, each of these blocks have a varying number of banks, each of which have 32 GPIOs and an interrupt line. The current nomadik-gpio driver implementation can handle each of these sub-banks easily with one instance each, but cannot as-is be hooked up to them because it is an AMBA driver and it expects to see a peripheral with the appropriate peripheral ids but having only one bank and only one interrupt. Solve this by converting the driver to a platform driver. Acked-by: Alessandro Rubini Acked-by: Linus Walleij Signed-off-by: Rabin Vincent Signed-off-by: Russell King diff --git a/arch/arm/mach-nomadik/cpu-8815.c b/arch/arm/mach-nomadik/cpu-8815.c index 9bf33b3..91c3c90 100644 --- a/arch/arm/mach-nomadik/cpu-8815.c +++ b/arch/arm/mach-nomadik/cpu-8815.c @@ -20,6 +20,7 @@ #include #include #include +#include #include #include @@ -30,60 +31,66 @@ #include #include +#define __MEM_4K_RESOURCE(x) \ + .res = {.start = (x), .end = (x) + SZ_4K - 1, .flags = IORESOURCE_MEM} + /* The 8815 has 4 GPIO blocks, let's register them immediately */ + +#define GPIO_RESOURCE(block) \ + { \ + .start = NOMADIK_GPIO##block##_BASE, \ + .end = NOMADIK_GPIO##block##_BASE + SZ_4K - 1, \ + .flags = IORESOURCE_MEM, \ + }, \ + { \ + .start = IRQ_GPIO##block, \ + .end = IRQ_GPIO##block, \ + .flags = IORESOURCE_IRQ, \ + } + +#define GPIO_DEVICE(block) \ + { \ + .name = "gpio", \ + .id = block, \ + .num_resources = 2, \ + .resource = &cpu8815_gpio_resources[block * 2], \ + .dev = { \ + .platform_data = &cpu8815_gpio[block], \ + }, \ + } + static struct nmk_gpio_platform_data cpu8815_gpio[] = { { .name = "GPIO-0-31", .first_gpio = 0, .first_irq = NOMADIK_GPIO_TO_IRQ(0), - .parent_irq = IRQ_GPIO0, }, { .name = "GPIO-32-63", .first_gpio = 32, .first_irq = NOMADIK_GPIO_TO_IRQ(32), - .parent_irq = IRQ_GPIO1, }, { .name = "GPIO-64-95", .first_gpio = 64, .first_irq = NOMADIK_GPIO_TO_IRQ(64), - .parent_irq = IRQ_GPIO2, }, { .name = "GPIO-96-127", /* 124..127 not routed to pin */ .first_gpio = 96, .first_irq = NOMADIK_GPIO_TO_IRQ(96), - .parent_irq = IRQ_GPIO3, } }; -#define __MEM_4K_RESOURCE(x) \ - .res = {.start = (x), .end = (x) + SZ_4K - 1, .flags = IORESOURCE_MEM} +static struct resource cpu8815_gpio_resources[] = { + GPIO_RESOURCE(0), + GPIO_RESOURCE(1), + GPIO_RESOURCE(2), + GPIO_RESOURCE(3), +}; -static struct amba_device cpu8815_amba_gpio[] = { - { - .dev = { - .init_name = "gpio0", - .platform_data = cpu8815_gpio + 0, - }, - __MEM_4K_RESOURCE(NOMADIK_GPIO0_BASE), - }, { - .dev = { - .init_name = "gpio1", - .platform_data = cpu8815_gpio + 1, - }, - __MEM_4K_RESOURCE(NOMADIK_GPIO1_BASE), - }, { - .dev = { - .init_name = "gpio2", - .platform_data = cpu8815_gpio + 2, - }, - __MEM_4K_RESOURCE(NOMADIK_GPIO2_BASE), - }, { - .dev = { - .init_name = "gpio3", - .platform_data = cpu8815_gpio + 3, - }, - __MEM_4K_RESOURCE(NOMADIK_GPIO3_BASE), - }, +static struct platform_device cpu8815_platform_gpio[] = { + GPIO_DEVICE(0), + GPIO_DEVICE(1), + GPIO_DEVICE(2), + GPIO_DEVICE(3), }; static struct amba_device cpu8815_amba_rng = { @@ -93,11 +100,14 @@ static struct amba_device cpu8815_amba_rng = { __MEM_4K_RESOURCE(NOMADIK_RNG_BASE), }; +static struct platform_device *platform_devs[] __initdata = { + cpu8815_platform_gpio + 0, + cpu8815_platform_gpio + 1, + cpu8815_platform_gpio + 2, + cpu8815_platform_gpio + 3, +}; + static struct amba_device *amba_devs[] __initdata = { - cpu8815_amba_gpio + 0, - cpu8815_amba_gpio + 1, - cpu8815_amba_gpio + 2, - cpu8815_amba_gpio + 3, &cpu8815_amba_rng }; @@ -105,6 +115,7 @@ static int __init cpu8815_init(void) { int i; + platform_add_devices(platform_devs, ARRAY_SIZE(platform_devs)); for (i = 0; i < ARRAY_SIZE(amba_devs); i++) amba_device_register(amba_devs[i], &iomem_resource); return 0; diff --git a/arch/arm/plat-nomadik/gpio.c b/arch/arm/plat-nomadik/gpio.c index 4c3ea1a..092f380 100644 --- a/arch/arm/plat-nomadik/gpio.c +++ b/arch/arm/plat-nomadik/gpio.c @@ -13,7 +13,7 @@ #include #include #include -#include +#include #include #include #include @@ -303,30 +303,48 @@ static struct gpio_chip nmk_gpio_template = { .can_sleep = 0, }; -static int __init nmk_gpio_probe(struct amba_device *dev, struct amba_id *id) +static int __init nmk_gpio_probe(struct platform_device *dev) { - struct nmk_gpio_platform_data *pdata; + struct nmk_gpio_platform_data *pdata = dev->dev.platform_data; struct nmk_gpio_chip *nmk_chip; struct gpio_chip *chip; + struct resource *res; + int irq; int ret; - pdata = dev->dev.platform_data; - ret = amba_request_regions(dev, pdata->name); - if (ret) - return ret; + if (!pdata) + return -ENODEV; + + res = platform_get_resource(dev, IORESOURCE_MEM, 0); + if (!res) { + ret = -ENOENT; + goto out; + } + + irq = platform_get_irq(dev, 0); + if (irq < 0) { + ret = irq; + goto out; + } + + if (request_mem_region(res->start, resource_size(res), + dev_name(&dev->dev)) == NULL) { + ret = -EBUSY; + goto out; + } nmk_chip = kzalloc(sizeof(*nmk_chip), GFP_KERNEL); if (!nmk_chip) { ret = -ENOMEM; - goto out_amba; + goto out_release; } /* * The virt address in nmk_chip->addr is in the nomadik register space, * so we can simply convert the resource address, without remapping */ - nmk_chip->addr = io_p2v(dev->res.start); + nmk_chip->addr = io_p2v(res->start); nmk_chip->chip = nmk_gpio_template; - nmk_chip->parent_irq = pdata->parent_irq; + nmk_chip->parent_irq = irq; spin_lock_init(&nmk_chip->lock); chip = &nmk_chip->chip; @@ -339,7 +357,7 @@ static int __init nmk_gpio_probe(struct amba_device *dev, struct amba_id *id) if (ret) goto out_free; - amba_set_drvdata(dev, nmk_chip); + platform_set_drvdata(dev, nmk_chip); nmk_gpio_init_irq(nmk_chip); @@ -347,51 +365,45 @@ static int __init nmk_gpio_probe(struct amba_device *dev, struct amba_id *id) nmk_chip->chip.base, nmk_chip->chip.base+31, nmk_chip->addr); return 0; - out_free: +out_free: kfree(nmk_chip); - out_amba: - amba_release_regions(dev); +out_release: + release_mem_region(res->start, resource_size(res)); +out: dev_err(&dev->dev, "Failure %i for GPIO %i-%i\n", ret, pdata->first_gpio, pdata->first_gpio+31); return ret; } -static int nmk_gpio_remove(struct amba_device *dev) +static int __exit nmk_gpio_remove(struct platform_device *dev) { struct nmk_gpio_chip *nmk_chip; + struct resource *res; + + res = platform_get_resource(dev, IORESOURCE_MEM, 0); - nmk_chip = amba_get_drvdata(dev); + nmk_chip = platform_get_drvdata(dev); gpiochip_remove(&nmk_chip->chip); kfree(nmk_chip); - amba_release_regions(dev); + release_mem_region(res->start, resource_size(res)); return 0; } -/* We have 0x1f080060 and 0x1f180060, accept both using the mask */ -static struct amba_id nmk_gpio_ids[] = { - { - .id = 0x1f080060, - .mask = 0xffefffff, - }, - {0, 0}, -}; - -static struct amba_driver nmk_gpio_driver = { - .drv = { +static struct platform_driver nmk_gpio_driver = { + .driver = { .owner = THIS_MODULE, .name = "gpio", }, .probe = nmk_gpio_probe, - .remove = nmk_gpio_remove, + .remove = __exit_p(nmk_gpio_remove), .suspend = NULL, /* to be done */ .resume = NULL, - .id_table = nmk_gpio_ids, }; static int __init nmk_gpio_init(void) { - return amba_driver_register(&nmk_gpio_driver); + return platform_driver_register(&nmk_gpio_driver); } arch_initcall(nmk_gpio_init); diff --git a/arch/arm/plat-nomadik/include/plat/gpio.h b/arch/arm/plat-nomadik/include/plat/gpio.h index 1d665a0a..4200811 100644 --- a/arch/arm/plat-nomadik/include/plat/gpio.h +++ b/arch/arm/plat-nomadik/include/plat/gpio.h @@ -65,7 +65,6 @@ struct nmk_gpio_platform_data { char *name; int first_gpio; int first_irq; - int parent_irq; }; #endif /* __ASM_PLAT_GPIO_H */ -- cgit v0.10.2 From cc2c13342508d52cf08ea81f639b5de69607fd5d Mon Sep 17 00:00:00 2001 From: Rabin Vincent Date: Mon, 1 Mar 2010 05:03:31 +0100 Subject: ARM: 5962/1: ux500: move system timer to cpu file There is nothing board-specific about the system timer, so move it to the CPU file. Acked-by: Linus Walleij Acked-by: Srinidhi Kasagar Signed-off-by: Rabin Vincent Signed-off-by: Russell King diff --git a/arch/arm/mach-ux500/board-mop500.c b/arch/arm/mach-ux500/board-mop500.c index 803aec1..c2b4a35 100644 --- a/arch/arm/mach-ux500/board-mop500.c +++ b/arch/arm/mach-ux500/board-mop500.c @@ -17,11 +17,9 @@ #include #include -#include #include #include -#include #include #include @@ -201,22 +199,6 @@ static struct platform_device *platform_devs[] __initdata = { &i2c_controller3, }; -static void __init u8500_timer_init(void) -{ -#ifdef CONFIG_LOCAL_TIMERS - /* Setup the local timer base */ - twd_base = __io_address(U8500_TWD_BASE); -#endif - /* Setup the MTU base */ - mtu_base = __io_address(U8500_MTU0_BASE); - - nmdk_timer_init(); -} - -static struct sys_timer u8500_timer = { - .init = u8500_timer_init, -}; - static void __init u8500_init_machine(void) { int i; diff --git a/arch/arm/mach-ux500/cpu-u8500.c b/arch/arm/mach-ux500/cpu-u8500.c index 397bc1f..588b059 100644 --- a/arch/arm/mach-ux500/cpu-u8500.c +++ b/arch/arm/mach-ux500/cpu-u8500.c @@ -14,10 +14,14 @@ #include #include #include +#include +#include #include #include +#include #include +#include /* add any platform devices here - TODO */ static struct platform_device *platform_devs[] __initdata = { @@ -63,3 +67,19 @@ void __init u8500_init_devices(void) return ; } + +static void __init u8500_timer_init(void) +{ +#ifdef CONFIG_LOCAL_TIMERS + /* Setup the local timer base */ + twd_base = __io_address(U8500_TWD_BASE); +#endif + /* Setup the MTU base */ + mtu_base = __io_address(U8500_MTU0_BASE); + + nmdk_timer_init(); +} + +struct sys_timer u8500_timer = { + .init = u8500_timer_init, +}; diff --git a/arch/arm/mach-ux500/include/mach/setup.h b/arch/arm/mach-ux500/include/mach/setup.h index cf0ce16..65112bf 100644 --- a/arch/arm/mach-ux500/include/mach/setup.h +++ b/arch/arm/mach-ux500/include/mach/setup.h @@ -20,4 +20,7 @@ extern void u8500_init_irq(void); /* We re-use nomadik_timer for this platform */ extern void nmdk_timer_init(void); +struct sys_timer; +extern struct sys_timer u8500_timer; + #endif /* __ASM_ARCH_SETUP_H */ -- cgit v0.10.2 From 75a36ee01e860be693220f1c40e2c04fb6cc69ba Mon Sep 17 00:00:00 2001 From: Rabin Vincent Date: Mon, 1 Mar 2010 05:05:56 +0100 Subject: ARM: 5963/1: ux500: add support for u8500 v1 revision Add cpu_is_u8500{ed/v1}() functions to determine the variant based on the CPU id, add the changed peripheral addresses, and fixup the MTU address. Acked-by: Linus Walleij Acked-by: Srinidhi Kasagar Signed-off-by: Rabin Vincent Signed-off-by: Russell King diff --git a/arch/arm/mach-ux500/cpu-u8500.c b/arch/arm/mach-ux500/cpu-u8500.c index 588b059..f368504 100644 --- a/arch/arm/mach-ux500/cpu-u8500.c +++ b/arch/arm/mach-ux500/cpu-u8500.c @@ -40,15 +40,27 @@ static struct map_desc u8500_io_desc[] __initdata = { __IO_DEV_DESC(U8500_UART2_BASE, SZ_4K), __IO_DEV_DESC(U8500_GIC_CPU_BASE, SZ_4K), __IO_DEV_DESC(U8500_GIC_DIST_BASE, SZ_4K), - __IO_DEV_DESC(U8500_MTU0_BASE, SZ_4K), __IO_DEV_DESC(U8500_TWD_BASE, SZ_4K), __IO_DEV_DESC(U8500_SCU_BASE, SZ_4K), __IO_DEV_DESC(U8500_BACKUPRAM0_BASE, SZ_8K), }; +static struct map_desc u8500ed_io_desc[] __initdata = { + __IO_DEV_DESC(U8500_MTU0_BASE_ED, SZ_4K), +}; + +static struct map_desc u8500v1_io_desc[] __initdata = { + __IO_DEV_DESC(U8500_MTU0_BASE_V1, SZ_4K), +}; + void __init u8500_map_io(void) { iotable_init(u8500_io_desc, ARRAY_SIZE(u8500_io_desc)); + + if (cpu_is_u8500ed()) + iotable_init(u8500ed_io_desc, ARRAY_SIZE(u8500ed_io_desc)); + else + iotable_init(u8500v1_io_desc, ARRAY_SIZE(u8500v1_io_desc)); } void __init u8500_init_irq(void) @@ -75,7 +87,10 @@ static void __init u8500_timer_init(void) twd_base = __io_address(U8500_TWD_BASE); #endif /* Setup the MTU base */ - mtu_base = __io_address(U8500_MTU0_BASE); + if (cpu_is_u8500ed()) + mtu_base = __io_address(U8500_MTU0_BASE_ED); + else + mtu_base = __io_address(U8500_MTU0_BASE_V1); nmdk_timer_init(); } diff --git a/arch/arm/mach-ux500/include/mach/hardware.h b/arch/arm/mach-ux500/include/mach/hardware.h index 04ea836..f29a43d 100644 --- a/arch/arm/mach-ux500/include/mach/hardware.h +++ b/arch/arm/mach-ux500/include/mach/hardware.h @@ -56,16 +56,19 @@ #define U8500_TWD_SIZE 0x100 /* per7 base addressess */ -#define U8500_CR_BASE (U8500_PER7_BASE + 0x8000) -#define U8500_MTU0_BASE (U8500_PER7_BASE + 0xa000) -#define U8500_MTU1_BASE (U8500_PER7_BASE + 0xb000) -#define U8500_TZPC0_BASE (U8500_PER7_BASE + 0xc000) -#define U8500_CLKRST7_BASE (U8500_PER7_BASE + 0xf000) +#define U8500_CR_BASE_ED (U8500_PER7_BASE + 0x8000) +#define U8500_MTU0_BASE_ED (U8500_PER7_BASE + 0xa000) +#define U8500_MTU1_BASE_ED (U8500_PER7_BASE + 0xb000) +#define U8500_TZPC0_BASE_ED (U8500_PER7_BASE + 0xc000) +#define U8500_CLKRST7_BASE_ED (U8500_PER7_BASE + 0xf000) /* per6 base addressess */ #define U8500_RNG_BASE (U8500_PER6_BASE + 0x0000) #define U8500_PKA_BASE (U8500_PER6_BASE + 0x1000) #define U8500_PKAM_BASE (U8500_PER6_BASE + 0x2000) +#define U8500_MTU0_BASE_V1 (U8500_PER6_BASE + 0x6000) +#define U8500_MTU1_BASE_V1 (U8500_PER6_BASE + 0x7000) +#define U8500_CR_BASE_V1 (U8500_PER6_BASE + 0x8000) #define U8500_CRYPTO0_BASE (U8500_PER6_BASE + 0xa000) #define U8500_CRYPTO1_BASE (U8500_PER6_BASE + 0xb000) #define U8500_CLKRST6_BASE (U8500_PER6_BASE + 0xf000) @@ -128,4 +131,20 @@ /* ST-Ericsson modified pl022 id */ #define SSP_PER_ID 0x01080022 +#ifndef __ASSEMBLY__ + +#include + +static inline bool cpu_is_u8500ed(void) +{ + return (read_cpuid_id() & 15) == 0; +} + +static inline bool cpu_is_u8500v1(void) +{ + return (read_cpuid_id() & 15) == 1; +} + +#endif + #endif /* __MACH_HARDWARE_H */ -- cgit v0.10.2 From 1df20afce51e7af2a587425d1aafca2608fe0066 Mon Sep 17 00:00:00 2001 From: Rabin Vincent Date: Mon, 1 Mar 2010 05:07:47 +0100 Subject: ARM: 5964/1: ux500: support clock gating Implement clock gating support for the u8500 clocks. Acked-by: Linus Walleij Acked-by: Srinidhi Kasagar Signed-off-by: Rabin Vincent Signed-off-by: Russell King diff --git a/arch/arm/mach-ux500/board-mop500.c b/arch/arm/mach-ux500/board-mop500.c index c2b4a35..4335186 100644 --- a/arch/arm/mach-ux500/board-mop500.c +++ b/arch/arm/mach-ux500/board-mop500.c @@ -94,7 +94,7 @@ static struct pl022_ssp_controller ssp0_platform_data = { static struct amba_device pl022_device = { .dev = { .coherent_dma_mask = ~0, - .init_name = "pl022", + .init_name = "ssp0", .platform_data = &ssp0_platform_data, }, .res = { diff --git a/arch/arm/mach-ux500/clock.c b/arch/arm/mach-ux500/clock.c index 8359a73..d16b8b5 100644 --- a/arch/arm/mach-ux500/clock.c +++ b/arch/arm/mach-ux500/clock.c @@ -1,6 +1,6 @@ /* * Copyright (C) 2009 ST-Ericsson - * heavily based on realview platform + * Copyright (C) 2009 STMicroelectronics * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as @@ -12,33 +12,130 @@ #include #include #include -#include +#include #include -/* currently the clk structure - * just supports rate. This would - * be extended as and when new devices are - * added - TODO - */ -struct clk { - unsigned long rate; -}; +#include +#include "clock.h" + +#define PRCC_PCKEN 0x00 +#define PRCC_PCKDIS 0x04 +#define PRCC_KCKEN 0x08 +#define PRCC_KCKDIS 0x0C + +#define PRCM_YYCLKEN0_MGT_SET 0x510 +#define PRCM_YYCLKEN1_MGT_SET 0x514 +#define PRCM_YYCLKEN0_MGT_CLR 0x518 +#define PRCM_YYCLKEN1_MGT_CLR 0x51C +#define PRCM_YYCLKEN0_MGT_VAL 0x520 +#define PRCM_YYCLKEN1_MGT_VAL 0x524 + +#define PRCM_SVAMMDSPCLK_MGT 0x008 +#define PRCM_SIAMMDSPCLK_MGT 0x00C +#define PRCM_SGACLK_MGT 0x014 +#define PRCM_UARTCLK_MGT 0x018 +#define PRCM_MSP02CLK_MGT 0x01C +#define PRCM_MSP1CLK_MGT 0x288 +#define PRCM_I2CCLK_MGT 0x020 +#define PRCM_SDMMCCLK_MGT 0x024 +#define PRCM_SLIMCLK_MGT 0x028 +#define PRCM_PER1CLK_MGT 0x02C +#define PRCM_PER2CLK_MGT 0x030 +#define PRCM_PER3CLK_MGT 0x034 +#define PRCM_PER5CLK_MGT 0x038 +#define PRCM_PER6CLK_MGT 0x03C +#define PRCM_PER7CLK_MGT 0x040 +#define PRCM_LCDCLK_MGT 0x044 +#define PRCM_BMLCLK_MGT 0x04C +#define PRCM_HSITXCLK_MGT 0x050 +#define PRCM_HSIRXCLK_MGT 0x054 +#define PRCM_HDMICLK_MGT 0x058 +#define PRCM_APEATCLK_MGT 0x05C +#define PRCM_APETRACECLK_MGT 0x060 +#define PRCM_MCDECLK_MGT 0x064 +#define PRCM_IPI2CCLK_MGT 0x068 +#define PRCM_DSIALTCLK_MGT 0x06C +#define PRCM_DMACLK_MGT 0x074 +#define PRCM_B2R2CLK_MGT 0x078 +#define PRCM_TVCLK_MGT 0x07C +#define PRCM_UNIPROCLK_MGT 0x278 +#define PRCM_SSPCLK_MGT 0x280 +#define PRCM_RNGCLK_MGT 0x284 +#define PRCM_UICCCLK_MGT 0x27C + +#define PRCM_MGT_ENABLE (1 << 8) + +static DEFINE_SPINLOCK(clocks_lock); + +static void __clk_enable(struct clk *clk) +{ + if (clk->enabled++ == 0) { + if (clk->parent_cluster) + __clk_enable(clk->parent_cluster); + + if (clk->parent_periph) + __clk_enable(clk->parent_periph); + + if (clk->ops && clk->ops->enable) + clk->ops->enable(clk); + } +} int clk_enable(struct clk *clk) { + unsigned long flags; + + spin_lock_irqsave(&clocks_lock, flags); + __clk_enable(clk); + spin_unlock_irqrestore(&clocks_lock, flags); + return 0; } EXPORT_SYMBOL(clk_enable); +static void __clk_disable(struct clk *clk) +{ + if (--clk->enabled == 0) { + if (clk->ops && clk->ops->disable) + clk->ops->disable(clk); + + if (clk->parent_periph) + __clk_disable(clk->parent_periph); + + if (clk->parent_cluster) + __clk_disable(clk->parent_cluster); + } +} + void clk_disable(struct clk *clk) { + unsigned long flags; + + WARN_ON(!clk->enabled); + + spin_lock_irqsave(&clocks_lock, flags); + __clk_disable(clk); + spin_unlock_irqrestore(&clocks_lock, flags); } EXPORT_SYMBOL(clk_disable); unsigned long clk_get_rate(struct clk *clk) { - return clk->rate; + unsigned long rate; + + if (clk->ops && clk->ops->get_rate) + return clk->ops->get_rate(clk); + + rate = clk->rate; + if (!rate) { + if (clk->parent_periph) + rate = clk_get_rate(clk->parent_periph); + else if (clk->parent_cluster) + rate = clk_get_rate(clk->parent_cluster); + } + + return rate; } EXPORT_SYMBOL(clk_get_rate); @@ -56,37 +153,363 @@ int clk_set_rate(struct clk *clk, unsigned long rate) } EXPORT_SYMBOL(clk_set_rate); -/* ssp clock */ -static struct clk ssp_clk = { - .rate = 48000000, +static void clk_prcmu_enable(struct clk *clk) +{ + void __iomem *cg_set_reg = __io_address(U8500_PRCMU_BASE) + + PRCM_YYCLKEN0_MGT_SET + clk->prcmu_cg_off; + + writel(1 << clk->prcmu_cg_bit, cg_set_reg); +} + +static void clk_prcmu_disable(struct clk *clk) +{ + void __iomem *cg_clr_reg = __io_address(U8500_PRCMU_BASE) + + PRCM_YYCLKEN0_MGT_CLR + clk->prcmu_cg_off; + + writel(1 << clk->prcmu_cg_bit, cg_clr_reg); +} + +/* ED doesn't have the combined set/clr registers */ +static void clk_prcmu_ed_enable(struct clk *clk) +{ + void __iomem *addr = __io_address(U8500_PRCMU_BASE) + + clk->prcmu_cg_mgt; + + writel(readl(addr) | PRCM_MGT_ENABLE, addr); +} + +static void clk_prcmu_ed_disable(struct clk *clk) +{ + void __iomem *addr = __io_address(U8500_PRCMU_BASE) + + clk->prcmu_cg_mgt; + + writel(readl(addr) & ~PRCM_MGT_ENABLE, addr); +} + +static struct clkops clk_prcmu_ops = { + .enable = clk_prcmu_enable, + .disable = clk_prcmu_disable, }; -/* fixed clock */ -static struct clk f38_clk = { - .rate = 38400000, +static unsigned int clkrst_base[] = { + [1] = U8500_CLKRST1_BASE, + [2] = U8500_CLKRST2_BASE, + [3] = U8500_CLKRST3_BASE, + [5] = U8500_CLKRST5_BASE, + [6] = U8500_CLKRST6_BASE, + [7] = U8500_CLKRST7_BASE_ED, }; -static struct clk_lookup lookups[] = { - { - /* UART0 */ - .dev_id = "uart0", - .clk = &f38_clk, - }, { /* UART1 */ - .dev_id = "uart1", - .clk = &f38_clk, - }, { /* UART2 */ - .dev_id = "uart2", - .clk = &f38_clk, - }, { /* SSP */ - .dev_id = "pl022", - .clk = &ssp_clk, - } +static void clk_prcc_enable(struct clk *clk) +{ + void __iomem *addr = __io_address(clkrst_base[clk->cluster]); + + if (clk->prcc_kernel != -1) + writel(1 << clk->prcc_kernel, addr + PRCC_KCKEN); + + if (clk->prcc_bus != -1) + writel(1 << clk->prcc_bus, addr + PRCC_PCKEN); +} + +static void clk_prcc_disable(struct clk *clk) +{ + void __iomem *addr = __io_address(clkrst_base[clk->cluster]); + + if (clk->prcc_bus != -1) + writel(1 << clk->prcc_bus, addr + PRCC_PCKDIS); + + if (clk->prcc_kernel != -1) + writel(1 << clk->prcc_kernel, addr + PRCC_KCKDIS); +} + +static struct clkops clk_prcc_ops = { + .enable = clk_prcc_enable, + .disable = clk_prcc_disable, +}; + +static struct clk clk_32khz = { + .rate = 32000, +}; + +/* + * PRCMU level clock gating + */ + +/* Bank 0 */ +static DEFINE_PRCMU_CLK(svaclk, 0x0, 2, SVAMMDSPCLK); +static DEFINE_PRCMU_CLK(siaclk, 0x0, 3, SIAMMDSPCLK); +static DEFINE_PRCMU_CLK(sgaclk, 0x0, 4, SGACLK); +static DEFINE_PRCMU_CLK_RATE(uartclk, 0x0, 5, UARTCLK, 38400000); +static DEFINE_PRCMU_CLK(msp02clk, 0x0, 6, MSP02CLK); +static DEFINE_PRCMU_CLK(msp1clk, 0x0, 7, MSP1CLK); /* v1 */ +static DEFINE_PRCMU_CLK_RATE(i2cclk, 0x0, 8, I2CCLK, 48000000); +static DEFINE_PRCMU_CLK_RATE(sdmmcclk, 0x0, 9, SDMMCCLK, 50000000); +static DEFINE_PRCMU_CLK(slimclk, 0x0, 10, SLIMCLK); +static DEFINE_PRCMU_CLK(per1clk, 0x0, 11, PER1CLK); +static DEFINE_PRCMU_CLK(per2clk, 0x0, 12, PER2CLK); +static DEFINE_PRCMU_CLK(per3clk, 0x0, 13, PER3CLK); +static DEFINE_PRCMU_CLK(per5clk, 0x0, 14, PER5CLK); +static DEFINE_PRCMU_CLK_RATE(per6clk, 0x0, 15, PER6CLK, 133330000); +static DEFINE_PRCMU_CLK_RATE(per7clk, 0x0, 16, PER7CLK, 100000000); +static DEFINE_PRCMU_CLK(lcdclk, 0x0, 17, LCDCLK); +static DEFINE_PRCMU_CLK(bmlclk, 0x0, 18, BMLCLK); +static DEFINE_PRCMU_CLK(hsitxclk, 0x0, 19, HSITXCLK); +static DEFINE_PRCMU_CLK(hsirxclk, 0x0, 20, HSIRXCLK); +static DEFINE_PRCMU_CLK(hdmiclk, 0x0, 21, HDMICLK); +static DEFINE_PRCMU_CLK(apeatclk, 0x0, 22, APEATCLK); +static DEFINE_PRCMU_CLK(apetraceclk, 0x0, 23, APETRACECLK); +static DEFINE_PRCMU_CLK(mcdeclk, 0x0, 24, MCDECLK); +static DEFINE_PRCMU_CLK(ipi2clk, 0x0, 25, IPI2CCLK); +static DEFINE_PRCMU_CLK(dsialtclk, 0x0, 26, DSIALTCLK); /* v1 */ +static DEFINE_PRCMU_CLK(dmaclk, 0x0, 27, DMACLK); +static DEFINE_PRCMU_CLK(b2r2clk, 0x0, 28, B2R2CLK); +static DEFINE_PRCMU_CLK(tvclk, 0x0, 29, TVCLK); +static DEFINE_PRCMU_CLK(uniproclk, 0x0, 30, UNIPROCLK); /* v1 */ +static DEFINE_PRCMU_CLK_RATE(sspclk, 0x0, 31, SSPCLK, 48000000); /* v1 */ + +/* Bank 1 */ +static DEFINE_PRCMU_CLK(rngclk, 0x4, 0, RNGCLK); /* v1 */ +static DEFINE_PRCMU_CLK(uiccclk, 0x4, 1, UICCCLK); /* v1 */ + +/* + * PRCC level clock gating + * Format: per#, clk, PCKEN bit, KCKEN bit, parent + */ + +/* Peripheral Cluster #1 */ +static DEFINE_PRCC_CLK(1, i2c4, 10, 9, &clk_i2cclk); +static DEFINE_PRCC_CLK(1, gpio0, 9, -1, NULL); +static DEFINE_PRCC_CLK(1, slimbus0, 8, 8, &clk_slimclk); +static DEFINE_PRCC_CLK(1, spi3_ed, 7, 7, NULL); +static DEFINE_PRCC_CLK(1, spi3_v1, 7, -1, NULL); +static DEFINE_PRCC_CLK(1, i2c2, 6, 6, &clk_i2cclk); +static DEFINE_PRCC_CLK(1, sdi0, 5, 5, &clk_sdmmcclk); +static DEFINE_PRCC_CLK(1, msp1_ed, 4, 4, &clk_msp02clk); +static DEFINE_PRCC_CLK(1, msp1_v1, 4, 4, &clk_msp1clk); +static DEFINE_PRCC_CLK(1, msp0, 3, 3, &clk_msp02clk); +static DEFINE_PRCC_CLK(1, i2c1, 2, 2, &clk_i2cclk); +static DEFINE_PRCC_CLK(1, uart1, 1, 1, &clk_uartclk); +static DEFINE_PRCC_CLK(1, uart0, 0, 0, &clk_uartclk); + +/* Peripheral Cluster #2 */ + +static DEFINE_PRCC_CLK(2, gpio1_ed, 12, -1, NULL); +static DEFINE_PRCC_CLK(2, ssitx_ed, 11, -1, NULL); +static DEFINE_PRCC_CLK(2, ssirx_ed, 10, -1, NULL); +static DEFINE_PRCC_CLK(2, spi0_ed, 9, -1, NULL); +static DEFINE_PRCC_CLK(2, sdi3_ed, 8, 6, &clk_sdmmcclk); +static DEFINE_PRCC_CLK(2, sdi1_ed, 7, 5, &clk_sdmmcclk); +static DEFINE_PRCC_CLK(2, msp2_ed, 6, 4, &clk_msp02clk); +static DEFINE_PRCC_CLK(2, sdi4_ed, 4, 2, &clk_sdmmcclk); +static DEFINE_PRCC_CLK(2, pwl_ed, 3, 1, NULL); +static DEFINE_PRCC_CLK(2, spi1_ed, 2, -1, NULL); +static DEFINE_PRCC_CLK(2, spi2_ed, 1, -1, NULL); +static DEFINE_PRCC_CLK(2, i2c3_ed, 0, 0, &clk_i2cclk); + +static DEFINE_PRCC_CLK(2, gpio1_v1, 11, -1, NULL); +static DEFINE_PRCC_CLK(2, ssitx_v1, 10, 7, NULL); +static DEFINE_PRCC_CLK(2, ssirx_v1, 9, 6, NULL); +static DEFINE_PRCC_CLK(2, spi0_v1, 8, -1, NULL); +static DEFINE_PRCC_CLK(2, sdi3_v1, 7, 5, &clk_sdmmcclk); +static DEFINE_PRCC_CLK(2, sdi1_v1, 6, 4, &clk_sdmmcclk); +static DEFINE_PRCC_CLK(2, msp2_v1, 5, 3, &clk_msp02clk); +static DEFINE_PRCC_CLK(2, sdi4_v1, 4, 2, &clk_sdmmcclk); +static DEFINE_PRCC_CLK(2, pwl_v1, 3, 1, NULL); +static DEFINE_PRCC_CLK(2, spi1_v1, 2, -1, NULL); +static DEFINE_PRCC_CLK(2, spi2_v1, 1, -1, NULL); +static DEFINE_PRCC_CLK(2, i2c3_v1, 0, 0, &clk_i2cclk); + +/* Peripheral Cluster #3 */ +static DEFINE_PRCC_CLK(3, gpio2, 8, -1, NULL); +static DEFINE_PRCC_CLK(3, sdi5, 7, 7, &clk_sdmmcclk); +static DEFINE_PRCC_CLK(3, uart2, 6, 6, &clk_uartclk); +static DEFINE_PRCC_CLK(3, ske, 5, 5, &clk_32khz); +static DEFINE_PRCC_CLK(3, sdi2, 4, 4, &clk_sdmmcclk); +static DEFINE_PRCC_CLK(3, i2c0, 3, 3, &clk_i2cclk); +static DEFINE_PRCC_CLK(3, ssp1_ed, 2, 2, &clk_i2cclk); +static DEFINE_PRCC_CLK(3, ssp0_ed, 1, 1, &clk_i2cclk); +static DEFINE_PRCC_CLK(3, ssp1_v1, 2, 2, &clk_sspclk); +static DEFINE_PRCC_CLK(3, ssp0_v1, 1, 1, &clk_sspclk); +static DEFINE_PRCC_CLK(3, fsmc, 0, -1, NULL); + +/* Peripheral Cluster #4 is in the always on domain */ + +/* Peripheral Cluster #5 */ +static DEFINE_PRCC_CLK(5, gpio3, 1, -1, NULL); +static DEFINE_PRCC_CLK(5, usb_ed, 0, 0, &clk_i2cclk); +static DEFINE_PRCC_CLK(5, usb_v1, 0, 0, NULL); + +/* Peripheral Cluster #6 */ + +static DEFINE_PRCC_CLK(6, mtu1_v1, 8, -1, NULL); +static DEFINE_PRCC_CLK(6, mtu0_v1, 7, -1, NULL); +static DEFINE_PRCC_CLK(6, cfgreg_v1, 6, 6, NULL); +static DEFINE_PRCC_CLK(6, dmc_ed, 6, 6, NULL); +static DEFINE_PRCC_CLK(6, hash1, 5, -1, NULL); +static DEFINE_PRCC_CLK(6, unipro_v1, 4, 1, &clk_uniproclk); +static DEFINE_PRCC_CLK(6, cryp1_ed, 4, -1, NULL); +static DEFINE_PRCC_CLK(6, pka, 3, -1, NULL); +static DEFINE_PRCC_CLK(6, hash0, 2, -1, NULL); +static DEFINE_PRCC_CLK(6, cryp0, 1, -1, NULL); +static DEFINE_PRCC_CLK(6, rng_ed, 0, 0, &clk_i2cclk); +static DEFINE_PRCC_CLK(6, rng_v1, 0, 0, &clk_rngclk); + +/* Peripheral Cluster #7 */ + +static DEFINE_PRCC_CLK(7, tzpc0_ed, 4, -1, NULL); +static DEFINE_PRCC_CLK(7, mtu1_ed, 3, -1, NULL); +static DEFINE_PRCC_CLK(7, mtu0_ed, 2, -1, NULL); +static DEFINE_PRCC_CLK(7, wdg_ed, 1, -1, NULL); +static DEFINE_PRCC_CLK(7, cfgreg_ed, 0, -1, NULL); + +static struct clk_lookup u8500_common_clks[] = { + /* Peripheral Cluster #1 */ + CLK(gpio0, "gpioblock0", NULL), + CLK(slimbus0, "slimbus0", NULL), + CLK(i2c2, "nmk-i2c.2", NULL), + CLK(sdi0, "sdi0", NULL), + CLK(msp0, "msp0", NULL), + CLK(i2c1, "nmk-i2c.1", NULL), + CLK(uart1, "uart1", NULL), + CLK(uart0, "uart0", NULL), + + /* Peripheral Cluster #3 */ + CLK(gpio2, "gpioblock2", NULL), + CLK(sdi5, "sdi5", NULL), + CLK(uart2, "uart2", NULL), + CLK(ske, "ske", NULL), + CLK(sdi2, "sdi2", NULL), + CLK(i2c0, "nmk-i2c.0", NULL), + CLK(fsmc, "fsmc", NULL), + + /* Peripheral Cluster #5 */ + CLK(gpio3, "gpioblock3", NULL), + + /* Peripheral Cluster #6 */ + CLK(hash1, "hash1", NULL), + CLK(pka, "pka", NULL), + CLK(hash0, "hash0", NULL), + CLK(cryp0, "cryp0", NULL), + + /* PRCMU level clock gating */ + + /* Bank 0 */ + CLK(svaclk, "sva", NULL), + CLK(siaclk, "sia", NULL), + CLK(sgaclk, "sga", NULL), + CLK(slimclk, "slim", NULL), + CLK(lcdclk, "lcd", NULL), + CLK(bmlclk, "bml", NULL), + CLK(hsitxclk, "stm-hsi.0", NULL), + CLK(hsirxclk, "stm-hsi.1", NULL), + CLK(hdmiclk, "hdmi", NULL), + CLK(apeatclk, "apeat", NULL), + CLK(apetraceclk, "apetrace", NULL), + CLK(mcdeclk, "mcde", NULL), + CLK(ipi2clk, "ipi2", NULL), + CLK(dmaclk, "dma40", NULL), + CLK(b2r2clk, "b2r2", NULL), + CLK(tvclk, "tv", NULL), +}; + +static struct clk_lookup u8500_ed_clks[] = { + /* Peripheral Cluster #1 */ + CLK(spi3_ed, "spi3", NULL), + CLK(msp1_ed, "msp1", NULL), + + /* Peripheral Cluster #2 */ + CLK(gpio1_ed, "gpioblock1", NULL), + CLK(ssitx_ed, "ssitx", NULL), + CLK(ssirx_ed, "ssirx", NULL), + CLK(spi0_ed, "spi0", NULL), + CLK(sdi3_ed, "sdi3", NULL), + CLK(sdi1_ed, "sdi1", NULL), + CLK(msp2_ed, "msp2", NULL), + CLK(sdi4_ed, "sdi4", NULL), + CLK(pwl_ed, "pwl", NULL), + CLK(spi1_ed, "spi1", NULL), + CLK(spi2_ed, "spi2", NULL), + CLK(i2c3_ed, "nmk-i2c.3", NULL), + + /* Peripheral Cluster #3 */ + CLK(ssp1_ed, "ssp1", NULL), + CLK(ssp0_ed, "ssp0", NULL), + + /* Peripheral Cluster #5 */ + CLK(usb_ed, "musb_hdrc.0", "usb"), + + /* Peripheral Cluster #6 */ + CLK(dmc_ed, "dmc", NULL), + CLK(cryp1_ed, "cryp1", NULL), + CLK(rng_ed, "rng", NULL), + + /* Peripheral Cluster #7 */ + CLK(tzpc0_ed, "tzpc0", NULL), + CLK(mtu1_ed, "mtu1", NULL), + CLK(mtu0_ed, "mtu0", NULL), + CLK(wdg_ed, "wdg", NULL), + CLK(cfgreg_ed, "cfgreg", NULL), +}; + +static struct clk_lookup u8500_v1_clks[] = { + /* Peripheral Cluster #1 */ + CLK(i2c4, "nmk-i2c.4", NULL), + CLK(spi3_v1, "spi3", NULL), + CLK(msp1_v1, "msp1", NULL), + + /* Peripheral Cluster #2 */ + CLK(gpio1_v1, "gpioblock1", NULL), + CLK(ssitx_v1, "ssitx", NULL), + CLK(ssirx_v1, "ssirx", NULL), + CLK(spi0_v1, "spi0", NULL), + CLK(sdi3_v1, "sdi3", NULL), + CLK(sdi1_v1, "sdi1", NULL), + CLK(msp2_v1, "msp2", NULL), + CLK(sdi4_v1, "sdi4", NULL), + CLK(pwl_v1, "pwl", NULL), + CLK(spi1_v1, "spi1", NULL), + CLK(spi2_v1, "spi2", NULL), + CLK(i2c3_v1, "nmk-i2c.3", NULL), + + /* Peripheral Cluster #3 */ + CLK(ssp1_v1, "ssp1", NULL), + CLK(ssp0_v1, "ssp0", NULL), + + /* Peripheral Cluster #5 */ + CLK(usb_v1, "musb_hdrc.0", "usb"), + + /* Peripheral Cluster #6 */ + CLK(mtu1_v1, "mtu1", NULL), + CLK(mtu0_v1, "mtu0", NULL), + CLK(cfgreg_v1, "cfgreg", NULL), + CLK(hash1, "hash1", NULL), + CLK(unipro_v1, "unipro", NULL), + CLK(rng_v1, "rng", NULL), + + /* PRCMU level clock gating */ + + /* Bank 0 */ + CLK(uniproclk, "uniproclk", NULL), + CLK(dsialtclk, "dsialt", NULL), + + /* Bank 1 */ + CLK(rngclk, "rng", NULL), + CLK(uiccclk, "uicc", NULL), }; static int __init clk_init(void) { - /* register the clock lookups */ - clkdev_add_table(lookups, ARRAY_SIZE(lookups)); + if (cpu_is_u8500ed()) { + clk_prcmu_ops.enable = clk_prcmu_ed_enable; + clk_prcmu_ops.disable = clk_prcmu_ed_disable; + } + + clkdev_add_table(u8500_common_clks, ARRAY_SIZE(u8500_common_clks)); + if (cpu_is_u8500ed()) + clkdev_add_table(u8500_ed_clks, ARRAY_SIZE(u8500_ed_clks)); + else + clkdev_add_table(u8500_v1_clks, ARRAY_SIZE(u8500_v1_clks)); + return 0; } arch_initcall(clk_init); diff --git a/arch/arm/mach-ux500/clock.h b/arch/arm/mach-ux500/clock.h new file mode 100644 index 0000000..e4f99b6 --- /dev/null +++ b/arch/arm/mach-ux500/clock.h @@ -0,0 +1,125 @@ +/* + * Copyright (C) 2010 ST-Ericsson + * Copyright (C) 2009 STMicroelectronics + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +/** + * struct clkops - ux500 clock operations + * @enable: function to enable the clock + * @disable: function to disable the clock + * @get_rate: function to get the current clock rate + * + * This structure contains function pointers to functions that will be used to + * control the clock. All of these functions are optional. If get_rate is + * NULL, the rate in the struct clk will be used. + */ +struct clkops { + void (*enable) (struct clk *); + void (*disable) (struct clk *); + unsigned long (*get_rate) (struct clk *); +}; + +/** + * struct clk - ux500 clock structure + * @ops: pointer to clkops struct used to control this clock + * @name: name, for debugging + * @enabled: refcount. positive if enabled, zero if disabled + * @rate: fixed rate for clocks which don't implement + * ops->getrate + * @prcmu_cg_off: address offset of the combined enable/disable register + * (used on u8500v1) + * @prcmu_cg_bit: bit in the combined enable/disable register (used on + * u8500v1) + * @prcmu_cg_mgt: address of the enable/disable register (used on + * u8500ed) + * @cluster: peripheral cluster number + * @prcc_bus: bit for the bus clock in the peripheral's CLKRST + * @prcc_kernel: bit for the kernel clock in the peripheral's CLKRST. + * -1 if no kernel clock exists. + * @parent_cluster: pointer to parent's cluster clk struct + * @parent_periph: pointer to parent's peripheral clk struct + * + * Peripherals are organised into clusters, and each cluster has an associated + * bus clock. Some peripherals also have a parent peripheral clock. + * + * In order to enable a clock for a peripheral, we need to enable: + * (1) the parent cluster (bus) clock at the PRCMU level + * (2) the parent peripheral clock (if any) at the PRCMU level + * (3) the peripheral's bus & kernel clock at the PRCC level + * + * (1) and (2) are handled by defining clk structs (DEFINE_PRCMU_CLK) for each + * of the cluster and peripheral clocks, and hooking these as the parents of + * the individual peripheral clocks. + * + * (3) is handled by specifying the bits in the PRCC control registers required + * to enable these clocks and modifying them in the ->enable and + * ->disable callbacks of the peripheral clocks (DEFINE_PRCC_CLK). + * + * This structure describes both the PRCMU-level clocks and PRCC-level clocks. + * The prcmu_* fields are only used for the PRCMU clocks, and the cluster, + * prcc, and parent pointers are only used for the PRCC-level clocks. + */ +struct clk { + const struct clkops *ops; + const char *name; + unsigned int enabled; + + unsigned long rate; + struct list_head list; + + /* These three are only for PRCMU clks */ + + unsigned int prcmu_cg_off; + unsigned int prcmu_cg_bit; + unsigned int prcmu_cg_mgt; + + /* The rest are only for PRCC clks */ + + int cluster; + unsigned int prcc_bus; + unsigned int prcc_kernel; + + struct clk *parent_cluster; + struct clk *parent_periph; +}; + +#define DEFINE_PRCMU_CLK(_name, _cg_off, _cg_bit, _reg) \ +struct clk clk_##_name = { \ + .name = #_name, \ + .ops = &clk_prcmu_ops, \ + .prcmu_cg_off = _cg_off, \ + .prcmu_cg_bit = _cg_bit, \ + .prcmu_cg_mgt = PRCM_##_reg##_MGT \ + } + +#define DEFINE_PRCMU_CLK_RATE(_name, _cg_off, _cg_bit, _reg, _rate) \ +struct clk clk_##_name = { \ + .name = #_name, \ + .ops = &clk_prcmu_ops, \ + .prcmu_cg_off = _cg_off, \ + .prcmu_cg_bit = _cg_bit, \ + .rate = _rate, \ + .prcmu_cg_mgt = PRCM_##_reg##_MGT \ + } + +#define DEFINE_PRCC_CLK(_pclust, _name, _bus_en, _kernel_en, _kernclk) \ +struct clk clk_##_name = { \ + .name = #_name, \ + .ops = &clk_prcc_ops, \ + .cluster = _pclust, \ + .prcc_bus = _bus_en, \ + .prcc_kernel = _kernel_en, \ + .parent_cluster = &clk_per##_pclust##clk, \ + .parent_periph = _kernclk \ + } + +#define CLK(_clk, _devname, _conname) \ + { \ + .clk = &clk_##_clk, \ + .dev_id = _devname, \ + .con_id = _conname, \ + } diff --git a/arch/arm/mach-ux500/cpu-u8500.c b/arch/arm/mach-ux500/cpu-u8500.c index f368504..09bcba1 100644 --- a/arch/arm/mach-ux500/cpu-u8500.c +++ b/arch/arm/mach-ux500/cpu-u8500.c @@ -43,10 +43,17 @@ static struct map_desc u8500_io_desc[] __initdata = { __IO_DEV_DESC(U8500_TWD_BASE, SZ_4K), __IO_DEV_DESC(U8500_SCU_BASE, SZ_4K), __IO_DEV_DESC(U8500_BACKUPRAM0_BASE, SZ_8K), + __IO_DEV_DESC(U8500_PRCMU_BASE, SZ_4K), + __IO_DEV_DESC(U8500_CLKRST1_BASE, SZ_4K), + __IO_DEV_DESC(U8500_CLKRST2_BASE, SZ_4K), + __IO_DEV_DESC(U8500_CLKRST3_BASE, SZ_4K), + __IO_DEV_DESC(U8500_CLKRST5_BASE, SZ_4K), + __IO_DEV_DESC(U8500_CLKRST6_BASE, SZ_4K), }; static struct map_desc u8500ed_io_desc[] __initdata = { __IO_DEV_DESC(U8500_MTU0_BASE_ED, SZ_4K), + __IO_DEV_DESC(U8500_CLKRST7_BASE_ED, SZ_8K), }; static struct map_desc u8500v1_io_desc[] __initdata = { -- cgit v0.10.2 From 94bdc0e2d76c5f2467346bf14e7e16d6d8e0395d Mon Sep 17 00:00:00 2001 From: Rabin Vincent Date: Wed, 3 Mar 2010 04:54:37 +0100 Subject: ARM: 5973/1: ux500: add gpio support Add support for the GPIOs on the U8500, using the plat-nomadik GPIO driver. Acked-by: Linus Walleij Acked-by: Srinidhi Kasagar Signed-off-by: Rabin Vincent Signed-off-by: Russell King diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index c5408bf..540f859 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -756,6 +756,7 @@ config ARCH_U8500 select GENERIC_TIME select GENERIC_CLOCKEVENTS select COMMON_CLKDEV + select ARCH_REQUIRE_GPIOLIB help Support for ST-Ericsson's Ux500 architecture diff --git a/arch/arm/mach-ux500/Kconfig b/arch/arm/mach-ux500/Kconfig index 03625d7..cd412a3 100644 --- a/arch/arm/mach-ux500/Kconfig +++ b/arch/arm/mach-ux500/Kconfig @@ -8,6 +8,7 @@ config MACH_U8500_MOP default y select ARM_GIC select HAS_MTU + select NOMADIK_GPIO help Include support for mop500 development platform based on U8500 architecture. The platform is based diff --git a/arch/arm/mach-ux500/cpu-u8500.c b/arch/arm/mach-ux500/cpu-u8500.c index 09bcba1..5fb4466 100644 --- a/arch/arm/mach-ux500/cpu-u8500.c +++ b/arch/arm/mach-ux500/cpu-u8500.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include @@ -23,9 +24,82 @@ #include #include -/* add any platform devices here - TODO */ +#define GPIO_RESOURCE(block) \ + { \ + .start = U8500_GPIOBANK##block##_BASE, \ + .end = U8500_GPIOBANK##block##_BASE + 127, \ + .flags = IORESOURCE_MEM, \ + }, \ + { \ + .start = IRQ_GPIO##block, \ + .end = IRQ_GPIO##block, \ + .flags = IORESOURCE_IRQ, \ + } + +#define GPIO_DEVICE(block) \ + { \ + .name = "gpio", \ + .id = block, \ + .num_resources = 2, \ + .resource = &u8500_gpio_resources[block * 2], \ + .dev = { \ + .platform_data = &u8500_gpio_data[block], \ + }, \ + } + +#define GPIO_DATA(_name, first) \ + { \ + .name = _name, \ + .first_gpio = first, \ + .first_irq = NOMADIK_GPIO_TO_IRQ(first), \ + } + +static struct nmk_gpio_platform_data u8500_gpio_data[] = { + GPIO_DATA("GPIO-0-31", 0), + GPIO_DATA("GPIO-32-63", 32), /* 37..63 not routed to pin */ + GPIO_DATA("GPIO-64-95", 64), + GPIO_DATA("GPIO-96-127", 96), /* 97..127 not routed to pin */ + GPIO_DATA("GPIO-128-159", 128), + GPIO_DATA("GPIO-160-191", 160), /* 172..191 not routed to pin */ + GPIO_DATA("GPIO-192-223", 192), + GPIO_DATA("GPIO-224-255", 224), /* 231..255 not routed to pin */ + GPIO_DATA("GPIO-256-288", 256), /* 258..288 not routed to pin */ +}; + +static struct resource u8500_gpio_resources[] = { + GPIO_RESOURCE(0), + GPIO_RESOURCE(1), + GPIO_RESOURCE(2), + GPIO_RESOURCE(3), + GPIO_RESOURCE(4), + GPIO_RESOURCE(5), + GPIO_RESOURCE(6), + GPIO_RESOURCE(7), + GPIO_RESOURCE(8), +}; + +static struct platform_device u8500_gpio_devs[] = { + GPIO_DEVICE(0), + GPIO_DEVICE(1), + GPIO_DEVICE(2), + GPIO_DEVICE(3), + GPIO_DEVICE(4), + GPIO_DEVICE(5), + GPIO_DEVICE(6), + GPIO_DEVICE(7), + GPIO_DEVICE(8), +}; + static struct platform_device *platform_devs[] __initdata = { - /* yet to be added, add i2c0, gpio.. */ + &u8500_gpio_devs[0], + &u8500_gpio_devs[1], + &u8500_gpio_devs[2], + &u8500_gpio_devs[3], + &u8500_gpio_devs[4], + &u8500_gpio_devs[5], + &u8500_gpio_devs[6], + &u8500_gpio_devs[7], + &u8500_gpio_devs[8], }; #define __IO_DEV_DESC(x, sz) { \ @@ -49,6 +123,10 @@ static struct map_desc u8500_io_desc[] __initdata = { __IO_DEV_DESC(U8500_CLKRST3_BASE, SZ_4K), __IO_DEV_DESC(U8500_CLKRST5_BASE, SZ_4K), __IO_DEV_DESC(U8500_CLKRST6_BASE, SZ_4K), + __IO_DEV_DESC(U8500_GPIO1_BASE, SZ_4K), + __IO_DEV_DESC(U8500_GPIO2_BASE, SZ_4K), + __IO_DEV_DESC(U8500_GPIO3_BASE, SZ_4K), + __IO_DEV_DESC(U8500_GPIO5_BASE, SZ_4K), }; static struct map_desc u8500ed_io_desc[] __initdata = { diff --git a/arch/arm/mach-ux500/include/mach/gpio.h b/arch/arm/mach-ux500/include/mach/gpio.h new file mode 100644 index 0000000..3c4cd31 --- /dev/null +++ b/arch/arm/mach-ux500/include/mach/gpio.h @@ -0,0 +1,12 @@ +#ifndef __ASM_ARCH_GPIO_H +#define __ASM_ARCH_GPIO_H + +/* + * 288 (#267 is the highest one actually hooked up) onchip GPIOs, plus enough + * room for a couple of GPIO expanders. + */ +#define ARCH_NR_GPIOS 350 + +#include + +#endif /* __ASM_ARCH_GPIO_H */ diff --git a/arch/arm/mach-ux500/include/mach/hardware.h b/arch/arm/mach-ux500/include/mach/hardware.h index f29a43d..99ca89f 100644 --- a/arch/arm/mach-ux500/include/mach/hardware.h +++ b/arch/arm/mach-ux500/include/mach/hardware.h @@ -23,6 +23,8 @@ /* typesafe io address */ #define __io_address(n) __io(IO_ADDRESS(n)) +/* used by some plat-nomadik code */ +#define io_p2v(n) __io_address(n) /* * Base address definitions for U8500 Onchip IPs. All the @@ -128,6 +130,16 @@ #define U8500_GPIO1_BASE (U8500_PER1_BASE + 0xe000) #define U8500_CLKRST1_BASE (U8500_PER1_BASE + 0xf000) +#define U8500_GPIOBANK0_BASE U8500_GPIO1_BASE +#define U8500_GPIOBANK1_BASE (U8500_GPIO1_BASE + 0x80) +#define U8500_GPIOBANK2_BASE U8500_GPIO3_BASE +#define U8500_GPIOBANK3_BASE (U8500_GPIO3_BASE + 0x80) +#define U8500_GPIOBANK4_BASE (U8500_GPIO3_BASE + 0x100) +#define U8500_GPIOBANK5_BASE (U8500_GPIO3_BASE + 0x180) +#define U8500_GPIOBANK6_BASE U8500_GPIO2_BASE +#define U8500_GPIOBANK7_BASE (U8500_GPIO2_BASE + 0x80) +#define U8500_GPIOBANK8_BASE U8500_GPIO5_BASE + /* ST-Ericsson modified pl022 id */ #define SSP_PER_ID 0x01080022 diff --git a/arch/arm/mach-ux500/include/mach/irqs.h b/arch/arm/mach-ux500/include/mach/irqs.h index 394b5dd..8a1f697 100644 --- a/arch/arm/mach-ux500/include/mach/irqs.h +++ b/arch/arm/mach-ux500/include/mach/irqs.h @@ -66,6 +66,12 @@ /* There are 128 shared peripheral interrupts assigned to * INTID[160:32]. The first 32 interrupts are reserved. */ -#define NR_IRQS 161 +#define U8500_SOC_NR_IRQS 161 + +/* After chip-specific IRQ numbers we have the GPIO ones */ +#define NOMADIK_NR_GPIO 288 +#define NOMADIK_GPIO_TO_IRQ(gpio) ((gpio) + U8500_SOC_NR_IRQS) +#define NOMADIK_IRQ_TO_GPIO(irq) ((irq) - U8500_SOC_NR_IRQS) +#define NR_IRQS NOMADIK_GPIO_TO_IRQ(NOMADIK_NR_GPIO) #endif /*ASM_ARCH_IRQS_H*/ -- cgit v0.10.2 From b102c01faed5e0083a4e6d29a2d61f6b57716e94 Mon Sep 17 00:00:00 2001 From: Alessandro Rubini Date: Fri, 5 Mar 2010 12:38:51 +0100 Subject: ARM: 5978/1: plat-nomadik: use one-shot clock events This is a complete rewrite of the MTU driver, using one-shot for events and a free-running timer for stamping. It allows CONFIG_NO_HZ and CONFIG_HIGH_RES_TIMERS to work on Nomadik and Ux500. Signed-off-by: Alessandro Rubini Acked-by: Linus Walleij Acked-by: Andrea Gallo Signed-off-by: Russell King diff --git a/arch/arm/plat-nomadik/timer.c b/arch/arm/plat-nomadik/timer.c index fa7cb3a..db67402 100644 --- a/arch/arm/plat-nomadik/timer.c +++ b/arch/arm/plat-nomadik/timer.c @@ -2,7 +2,7 @@ * linux/arch/arm/mach-nomadik/timer.c * * Copyright (C) 2008 STMicroelectronics - * Copyright (C) 2009 Alessandro Rubini, somewhat based on at91sam926x + * Copyright (C) 2010 Alessandro Rubini * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2, as @@ -18,123 +18,124 @@ #include -static u32 nmdk_count; /* accumulated count */ -static u32 nmdk_cycle; /* write-once */ +void __iomem *mtu_base; /* ssigned by machine code */ -/* setup by the platform code */ -void __iomem *mtu_base; - -/* - * clocksource: the MTU device is a decrementing counters, so we negate - * the value being read. - */ +/* clocksource: MTU decrements, so we negate the value being read. */ static cycle_t nmdk_read_timer(struct clocksource *cs) { - u32 count = readl(mtu_base + MTU_VAL(0)); - return nmdk_count + nmdk_cycle - count; - + return -readl(mtu_base + MTU_VAL(0)); } static struct clocksource nmdk_clksrc = { .name = "mtu_0", - .rating = 120, + .rating = 200, .read = nmdk_read_timer, + .mask = CLOCKSOURCE_MASK(32), .shift = 20, .flags = CLOCK_SOURCE_IS_CONTINUOUS, }; -/* - * Clockevent device: currently only periodic mode is supported - */ +/* Clockevent device: use one-shot mode */ static void nmdk_clkevt_mode(enum clock_event_mode mode, struct clock_event_device *dev) { + u32 cr; + switch (mode) { case CLOCK_EVT_MODE_PERIODIC: - /* count current value? */ - writel(readl(mtu_base + MTU_IMSC) | 1, mtu_base + MTU_IMSC); + pr_err("%s: periodic mode not supported\n", __func__); break; case CLOCK_EVT_MODE_ONESHOT: - BUG(); /* Not supported, yet */ - /* FALLTHROUGH */ + /* Load highest value, enable device, enable interrupts */ + cr = readl(mtu_base + MTU_CR(1)); + writel(0, mtu_base + MTU_LR(1)); + writel(cr | MTU_CRn_ENA, mtu_base + MTU_CR(1)); + writel(0x2, mtu_base + MTU_IMSC); + break; case CLOCK_EVT_MODE_SHUTDOWN: case CLOCK_EVT_MODE_UNUSED: - writel(readl(mtu_base + MTU_IMSC) & ~1, mtu_base + MTU_IMSC); + /* disable irq */ + writel(0, mtu_base + MTU_IMSC); break; case CLOCK_EVT_MODE_RESUME: break; } } +static int nmdk_clkevt_next(unsigned long evt, struct clock_event_device *ev) +{ + /* writing the value has immediate effect */ + writel(evt, mtu_base + MTU_LR(1)); + return 0; +} + static struct clock_event_device nmdk_clkevt = { - .name = "mtu_0", - .features = CLOCK_EVT_FEAT_PERIODIC, + .name = "mtu_1", + .features = CLOCK_EVT_FEAT_ONESHOT, .shift = 32, - .rating = 100, + .rating = 200, .set_mode = nmdk_clkevt_mode, + .set_next_event = nmdk_clkevt_next, }; /* - * IRQ Handler for the timer 0 of the MTU block. The irq is not shared - * as we are the only users of mtu0 by now. + * IRQ Handler for timer 1 of the MTU block. */ static irqreturn_t nmdk_timer_interrupt(int irq, void *dev_id) { - /* ack: "interrupt clear register" */ - writel(1 << 0, mtu_base + MTU_ICR); - - /* we can't count lost ticks, unfortunately */ - nmdk_count += nmdk_cycle; - nmdk_clkevt.event_handler(&nmdk_clkevt); + struct clock_event_device *evdev = dev_id; + writel(1 << 1, mtu_base + MTU_ICR); /* Interrupt clear reg */ + evdev->event_handler(evdev); return IRQ_HANDLED; } -/* - * Set up timer interrupt, and return the current time in seconds. - */ static struct irqaction nmdk_timer_irq = { .name = "Nomadik Timer Tick", .flags = IRQF_DISABLED | IRQF_TIMER, .handler = nmdk_timer_interrupt, + .dev_id = &nmdk_clkevt, }; -static void nmdk_timer_reset(void) -{ - u32 cr; - - writel(0, mtu_base + MTU_CR(0)); /* off */ - - /* configure load and background-load, and fire it up */ - writel(nmdk_cycle, mtu_base + MTU_LR(0)); - writel(nmdk_cycle, mtu_base + MTU_BGLR(0)); - cr = MTU_CRn_PERIODIC | MTU_CRn_PRESCALE_1 | MTU_CRn_32BITS; - writel(cr, mtu_base + MTU_CR(0)); - writel(cr | MTU_CRn_ENA, mtu_base + MTU_CR(0)); -} - void __init nmdk_timer_init(void) { unsigned long rate; - int bits; - - rate = CLOCK_TICK_RATE; /* 2.4MHz */ - nmdk_cycle = (rate + HZ/2) / HZ; + u32 cr = MTU_CRn_32BITS;; + + /* + * Tick rate is 2.4MHz for Nomadik and 110MHz for ux500: + * use a divide-by-16 counter if it's more than 16MHz + */ + rate = CLOCK_TICK_RATE; + if (rate > 16 << 20) { + rate /= 16; + cr |= MTU_CRn_PRESCALE_16; + } else { + cr |= MTU_CRn_PRESCALE_1; + } - /* Init the timer and register clocksource */ - nmdk_timer_reset(); + /* Timer 0 is the free running clocksource */ + writel(cr, mtu_base + MTU_CR(0)); + writel(0, mtu_base + MTU_LR(0)); + writel(0, mtu_base + MTU_BGLR(0)); + writel(cr | MTU_CRn_ENA, mtu_base + MTU_CR(0)); nmdk_clksrc.mult = clocksource_hz2mult(rate, nmdk_clksrc.shift); - bits = 8*sizeof(nmdk_count); - nmdk_clksrc.mask = CLOCKSOURCE_MASK(bits); if (clocksource_register(&nmdk_clksrc)) - printk(KERN_ERR "timer: failed to initialize clock " - "source %s\n", nmdk_clksrc.name); + pr_err("timer: failed to initialize clock source %s\n", + nmdk_clksrc.name); + + /* Timer 1 is used for events, fix according to rate */ + writel(cr | MTU_CRn_ONESHOT, mtu_base + MTU_CR(1)); /* off, currently */ + nmdk_clkevt.mult = div_sc(rate, NSEC_PER_SEC, nmdk_clkevt.shift); + nmdk_clkevt.max_delta_ns = + clockevent_delta2ns(0xffffffff, &nmdk_clkevt); + nmdk_clkevt.min_delta_ns = + clockevent_delta2ns(0x00000002, &nmdk_clkevt); + nmdk_clkevt.cpumask = cpumask_of(0); /* Register irq and clockevents */ setup_irq(IRQ_MTU0, &nmdk_timer_irq); - nmdk_clkevt.mult = div_sc(rate, NSEC_PER_SEC, nmdk_clkevt.shift); - nmdk_clkevt.cpumask = cpumask_of(0); clockevents_register_device(&nmdk_clkevt); } -- cgit v0.10.2 From 4bd96a7a8185755b091233b16034c7436cbf57af Mon Sep 17 00:00:00 2001 From: Shane Wang Date: Wed, 10 Mar 2010 14:36:10 +0800 Subject: x86, tboot: Add support for S3 memory integrity protection This patch adds support for S3 memory integrity protection within an Intel(R) TXT launched kernel, for all kernel and userspace memory. All RAM used by the kernel and userspace, as indicated by memory ranges of type E820_RAM and E820_RESERVED_KERN in the e820 table, will be integrity protected. The MAINTAINERS file is also updated to reflect the maintainers of the TXT-related code. All MACing is done in tboot, based on a complexity analysis and tradeoff. v3: Compared with v2, this patch adds a check of array size in tboot.c, and a note to specify which c/s of tboot supports this kind of MACing in intel_txt.txt. Signed-off-by: Shane Wang LKML-Reference: <4B973DDA.6050902@intel.com> Signed-off-by: Joseph Cihula Acked-by: Pavel Machek Acked-by: Rafael J. Wysocki Signed-off-by: H. Peter Anvin diff --git a/Documentation/intel_txt.txt b/Documentation/intel_txt.txt index f40a1f0..87c8990 100644 --- a/Documentation/intel_txt.txt +++ b/Documentation/intel_txt.txt @@ -161,13 +161,15 @@ o In order to put a system into any of the sleep states after a TXT has been restored, it will restore the TPM PCRs and then transfer control back to the kernel's S3 resume vector. In order to preserve system integrity across S3, the kernel - provides tboot with a set of memory ranges (kernel - code/data/bss, S3 resume code, and AP trampoline) that tboot - will calculate a MAC (message authentication code) over and then - seal with the TPM. On resume and once the measured environment - has been re-established, tboot will re-calculate the MAC and - verify it against the sealed value. Tboot's policy determines - what happens if the verification fails. + provides tboot with a set of memory ranges (RAM and RESERVED_KERN + in the e820 table, but not any memory that BIOS might alter over + the S3 transition) that tboot will calculate a MAC (message + authentication code) over and then seal with the TPM. On resume + and once the measured environment has been re-established, tboot + will re-calculate the MAC and verify it against the sealed value. + Tboot's policy determines what happens if the verification fails. + Note that the c/s 194 of tboot which has the new MAC code supports + this. That's pretty much it for TXT support. diff --git a/MAINTAINERS b/MAINTAINERS index 47cc449..d3072cb 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2940,6 +2940,17 @@ S: Odd Fixes F: Documentation/networking/README.ipw2200 F: drivers/net/wireless/ipw2x00/ipw2200.* +INTEL(R) TRUSTED EXECUTION TECHNOLOGY (TXT) +M: Joseph Cihula +M: Shane Wang +L: tboot-devel@lists.sourceforge.net +W: http://tboot.sourceforge.net +T: Mercurial http://www.bughost.org/repos.hg/tboot.hg +S: Supported +F: Documentation/intel_txt.txt +F: include/linux/tboot.h +F: arch/x86/kernel/tboot.c + INTEL WIRELESS WIMAX CONNECTION 2400 M: Inaky Perez-Gonzalez M: linux-wimax@intel.com diff --git a/arch/x86/include/asm/e820.h b/arch/x86/include/asm/e820.h index 0e22296..ec8a52d 100644 --- a/arch/x86/include/asm/e820.h +++ b/arch/x86/include/asm/e820.h @@ -45,7 +45,12 @@ #define E820_NVS 4 #define E820_UNUSABLE 5 -/* reserved RAM used by kernel itself */ +/* + * reserved RAM used by kernel itself + * if CONFIG_INTEL_TXT is enabled, memory of this type will be + * included in the S3 integrity calculation and so should not include + * any memory that BIOS might alter over the S3 transition + */ #define E820_RESERVED_KERN 128 #ifndef __ASSEMBLY__ diff --git a/arch/x86/kernel/tboot.c b/arch/x86/kernel/tboot.c index 86c9f91..cc2c604 100644 --- a/arch/x86/kernel/tboot.c +++ b/arch/x86/kernel/tboot.c @@ -175,6 +175,9 @@ static void add_mac_region(phys_addr_t start, unsigned long size) struct tboot_mac_region *mr; phys_addr_t end = start + size; + if (tboot->num_mac_regions >= MAX_TB_MAC_REGIONS) + panic("tboot: Too many MAC regions\n"); + if (start && size) { mr = &tboot->mac_regions[tboot->num_mac_regions++]; mr->start = round_down(start, PAGE_SIZE); @@ -184,18 +187,17 @@ static void add_mac_region(phys_addr_t start, unsigned long size) static int tboot_setup_sleep(void) { + int i; + tboot->num_mac_regions = 0; - /* S3 resume code */ - add_mac_region(acpi_wakeup_address, WAKEUP_SIZE); + for (i = 0; i < e820.nr_map; i++) { + if ((e820.map[i].type != E820_RAM) + && (e820.map[i].type != E820_RESERVED_KERN)) + continue; -#ifdef CONFIG_X86_TRAMPOLINE - /* AP trampoline code */ - add_mac_region(virt_to_phys(trampoline_base), TRAMPOLINE_SIZE); -#endif - - /* kernel code + data + bss */ - add_mac_region(virt_to_phys(_text), _end - _text); + add_mac_region(e820.map[i].addr, e820.map[i].size); + } tboot->acpi_sinfo.kernel_s3_resume_vector = acpi_wakeup_address; -- cgit v0.10.2 From 88fcf710c13bd41f2b98c5411e4f21ab98da4fb4 Mon Sep 17 00:00:00 2001 From: Yong Wang Date: Fri, 19 Mar 2010 23:02:16 -0700 Subject: Input: sparse-keymap - free the right keymap on error 'map' is allocated in sparse_keymap_setup() and it it the one that should be freed on error instead of 'keymap'. Signed-off-by: Yong Wang Cc: stable@kernel.org Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/sparse-keymap.c b/drivers/input/sparse-keymap.c index e6bde55..f64e004 100644 --- a/drivers/input/sparse-keymap.c +++ b/drivers/input/sparse-keymap.c @@ -163,7 +163,7 @@ int sparse_keymap_setup(struct input_dev *dev, return 0; err_out: - kfree(keymap); + kfree(map); return error; } -- cgit v0.10.2 From d90e6f6aad0ffdc674bc3a05c85c40dcc18e604c Mon Sep 17 00:00:00 2001 From: Henrik Rydberg Date: Sat, 20 Mar 2010 00:06:40 -0700 Subject: Input: bcm5974 - retract efi-broken suspend_resume With the recent system-wide improvements on suspend/resume and EFI booting the suspend_resume method of the bcm5974 has broken. When waking up from the S3 state on the MacBookAir, the trackpad is found in a yet unknown state, unable to switch to the proper multitouch mode. The result is a frozen touchpad, and a flood of errors of the kind bcm5974: bad trackpad package, length: 8. This patch retracts the reset_resume method altogether, falling back on the generic unbind/rebind functionality of the usb layer until further investigations can be made as how to reset the device when booting from efi. Signed-off-by: Henrik Rydberg Signed-off-by: Andrew Morton Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/mouse/bcm5974.c b/drivers/input/mouse/bcm5974.c index 4f8fe08..b89879b 100644 --- a/drivers/input/mouse/bcm5974.c +++ b/drivers/input/mouse/bcm5974.c @@ -803,7 +803,6 @@ static struct usb_driver bcm5974_driver = { .disconnect = bcm5974_disconnect, .suspend = bcm5974_suspend, .resume = bcm5974_resume, - .reset_resume = bcm5974_resume, .id_table = bcm5974_table, .supports_autosuspend = 1, }; -- cgit v0.10.2 From 97065e033ec5fd824d0133ecd0c614fe1e5c20bd Mon Sep 17 00:00:00 2001 From: Henrik Rydberg Date: Sun, 21 Mar 2010 22:31:26 -0700 Subject: Input: clarify the no-finger event in multitouch protocol The current documentation does not explicitly specify how to report zero fingers, leaving a potential problem in the driver implementations and giving no parsing directive to userland. This patch defines two equally valid ways. Signed-off-by: Henrik Rydberg Signed-off-by: Dmitry Torokhov diff --git a/Documentation/input/multi-touch-protocol.txt b/Documentation/input/multi-touch-protocol.txt index 8490480..eacb226 100644 --- a/Documentation/input/multi-touch-protocol.txt +++ b/Documentation/input/multi-touch-protocol.txt @@ -68,6 +68,22 @@ like: SYN_MT_REPORT SYN_REPORT +Here is the sequence after lifting one of the fingers: + + ABS_MT_POSITION_X + ABS_MT_POSITION_Y + SYN_MT_REPORT + SYN_REPORT + +And here is the sequence after lifting the remaining finger: + + SYN_MT_REPORT + SYN_REPORT + +If the driver reports one of BTN_TOUCH or ABS_PRESSURE in addition to the +ABS_MT events, the last SYN_MT_REPORT event may be omitted. Otherwise, the +last SYN_REPORT will be dropped by the input core, resulting in no +zero-finger event reaching userland. Event Semantics --------------- -- cgit v0.10.2 From 13bad37b04c779d98983307a27f97e9caa36f9b1 Mon Sep 17 00:00:00 2001 From: Henrik Rydberg Date: Sun, 21 Mar 2010 22:31:26 -0700 Subject: Input: update the status of the Multitouch X driver project The Multitouch X driver project has moved to alpha status. This patch updates the documentation accordingly. Signed-off-by: Henrik Rydberg Signed-off-by: Dmitry Torokhov diff --git a/Documentation/input/multi-touch-protocol.txt b/Documentation/input/multi-touch-protocol.txt index eacb226..c0fc1c7 100644 --- a/Documentation/input/multi-touch-protocol.txt +++ b/Documentation/input/multi-touch-protocol.txt @@ -233,11 +233,6 @@ where examples can be found. difference between the contact position and the approaching tool position could be used to derive tilt. [2] The list can of course be extended. -[3] The multi-touch X driver is currently in the prototyping stage. At the -time of writing (April 2009), the MT protocol is not yet merged, and the -prototype implements finger matching, basic mouse support and two-finger -scrolling. The project aims at improving the quality of current multi-touch -functionality available in the Synaptics X driver, and in addition -implement more advanced gestures. +[3] Multitouch X driver project: http://bitmath.org/code/multitouch/. [4] See the section on event computation. [5] See the section on finger tracking. -- cgit v0.10.2 From 2e2e3b96d98d5c17e9c09bc6088df3e182a71814 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Sun, 21 Mar 2010 22:56:15 -0700 Subject: Input: sparse-keymap - implement safer freeing of the keymap Allow calling sparse_keymap_free() before unregistering input device whithout risk of racing with EVIOCGETKEYCODE and EVIOCSETKEYCODE. This makes life of drivers writers easier. Acked-by: Yong Wang Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/input.c b/drivers/input/input.c index e2aad0a..be18fa9 100644 --- a/drivers/input/input.c +++ b/drivers/input/input.c @@ -659,7 +659,14 @@ static int input_default_setkeycode(struct input_dev *dev, int input_get_keycode(struct input_dev *dev, unsigned int scancode, unsigned int *keycode) { - return dev->getkeycode(dev, scancode, keycode); + unsigned long flags; + int retval; + + spin_lock_irqsave(&dev->event_lock, flags); + retval = dev->getkeycode(dev, scancode, keycode); + spin_unlock_irqrestore(&dev->event_lock, flags); + + return retval; } EXPORT_SYMBOL(input_get_keycode); diff --git a/drivers/input/sparse-keymap.c b/drivers/input/sparse-keymap.c index f64e004..2434ac5 100644 --- a/drivers/input/sparse-keymap.c +++ b/drivers/input/sparse-keymap.c @@ -67,12 +67,14 @@ static int sparse_keymap_getkeycode(struct input_dev *dev, unsigned int scancode, unsigned int *keycode) { - const struct key_entry *key = - sparse_keymap_entry_from_scancode(dev, scancode); + const struct key_entry *key; - if (key && key->type == KE_KEY) { - *keycode = key->keycode; - return 0; + if (dev->keycode) { + key = sparse_keymap_entry_from_scancode(dev, scancode); + if (key && key->type == KE_KEY) { + *keycode = key->keycode; + return 0; + } } return -EINVAL; @@ -85,17 +87,16 @@ static int sparse_keymap_setkeycode(struct input_dev *dev, struct key_entry *key; int old_keycode; - if (keycode < 0 || keycode > KEY_MAX) - return -EINVAL; - - key = sparse_keymap_entry_from_scancode(dev, scancode); - if (key && key->type == KE_KEY) { - old_keycode = key->keycode; - key->keycode = keycode; - set_bit(keycode, dev->keybit); - if (!sparse_keymap_entry_from_keycode(dev, old_keycode)) - clear_bit(old_keycode, dev->keybit); - return 0; + if (dev->keycode) { + key = sparse_keymap_entry_from_scancode(dev, scancode); + if (key && key->type == KE_KEY) { + old_keycode = key->keycode; + key->keycode = keycode; + set_bit(keycode, dev->keybit); + if (!sparse_keymap_entry_from_keycode(dev, old_keycode)) + clear_bit(old_keycode, dev->keybit); + return 0; + } } return -EINVAL; @@ -175,14 +176,27 @@ EXPORT_SYMBOL(sparse_keymap_setup); * * This function is used to free memory allocated by sparse keymap * in an input device that was set up by sparse_keymap_setup(). + * NOTE: It is safe to cal this function while input device is + * still registered (however the drivers should care not to try to + * use freed keymap and thus have to shut off interrups/polling + * before freeing the keymap). */ void sparse_keymap_free(struct input_dev *dev) { + unsigned long flags; + + /* + * Take event lock to prevent racing with input_get_keycode() + * and input_set_keycode() if we are called while input device + * is still registered. + */ + spin_lock_irqsave(&dev->event_lock, flags); + kfree(dev->keycode); dev->keycode = NULL; dev->keycodemax = 0; - dev->getkeycode = NULL; - dev->setkeycode = NULL; + + spin_unlock_irqrestore(&dev->event_lock, flags); } EXPORT_SYMBOL(sparse_keymap_free); -- cgit v0.10.2 From 301fde27c7fcd0380b02b175d547e894ff65d78a Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Mon, 22 Mar 2010 13:09:33 -0300 Subject: perf: Fix orphan callchain branches Callchains have markers inside their capture to tell we enter a context (kernel, user, ...). Those are not displayed in the callchains but they are incidentally an active part of the radix tree where callchains are stored, just like any other address. If we have the two following callchains: addr1 -> addr2 -> user context -> addr3 addr1 -> addr2 -> user context -> addr4 addr1 -> addr2 -> addr 5 This is pretty common if addr1 and addr2 are part of an interrupt path, addr3 and addr4 are user addresses and addr5 is a kernel non interrupt path. This will be stored as follows in the tree: addr1 addr2 / \ / addr5 user context / \ addr3 addr4 But we ignore the context markers in the report, hence the addr3 and addr4 will appear as orphan branches: |--28.30%-- hrtimer_interrupt | smp_apic_timer_interrupt | apic_timer_interrupt | | <------------- here, no parent! | | | | | |--11.11%-- 0x7fae7bccb875 | | | | | |--11.11%-- 0xffffffffff60013b | | | | | |--11.11%-- __pthread_mutex_lock_internal | | | | | |--11.11%-- __errno_location Fix this by removing the context markers when we process the callchains to the tree. Reported-by: Arnaldo Carvalho de Melo Signed-off-by: Frederic Weisbecker Cc: Paul Mackerras Cc: Peter Zijlstra Signed-off-by: Arnaldo Carvalho de Melo LKML-Reference: <1269274173-20328-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 1f9f869..d609afb 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -83,6 +83,7 @@ static int perf_session__add_hist_entry(struct perf_session *self, { struct symbol **syms = NULL, *parent = NULL; bool hit; + int err; struct hist_entry *he; struct event_stat_id *stats; struct perf_event_attr *attr; @@ -109,8 +110,11 @@ static int perf_session__add_hist_entry(struct perf_session *self, if (symbol_conf.use_callchain) { if (!hit) callchain_init(&he->callchain); - append_chain(&he->callchain, data->callchain, syms); + err = append_chain(&he->callchain, data->callchain, syms); free(syms); + + if (err) + return err; } return 0; diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c index b3b7125..883844e 100644 --- a/tools/perf/util/callchain.c +++ b/tools/perf/util/callchain.c @@ -1,5 +1,5 @@ /* - * Copyright (C) 2009, Frederic Weisbecker + * Copyright (C) 2009-2010, Frederic Weisbecker * * Handle the callchains from the stream in an ad-hoc radix tree and then * sort them in an rbtree. @@ -183,12 +183,23 @@ create_child(struct callchain_node *parent, bool inherit_children) return new; } + +struct resolved_ip { + u64 ip; + struct symbol *sym; +}; + +struct resolved_chain { + u64 nr; + struct resolved_ip ips[0]; +}; + + /* * Fill the node with callchain values */ static void -fill_node(struct callchain_node *node, struct ip_callchain *chain, - int start, struct symbol **syms) +fill_node(struct callchain_node *node, struct resolved_chain *chain, int start) { unsigned int i; @@ -200,8 +211,8 @@ fill_node(struct callchain_node *node, struct ip_callchain *chain, perror("not enough memory for the code path tree"); return; } - call->ip = chain->ips[i]; - call->sym = syms[i]; + call->ip = chain->ips[i].ip; + call->sym = chain->ips[i].sym; list_add_tail(&call->list, &node->val); } node->val_nr = chain->nr - start; @@ -210,13 +221,13 @@ fill_node(struct callchain_node *node, struct ip_callchain *chain, } static void -add_child(struct callchain_node *parent, struct ip_callchain *chain, - int start, struct symbol **syms) +add_child(struct callchain_node *parent, struct resolved_chain *chain, + int start) { struct callchain_node *new; new = create_child(parent, false); - fill_node(new, chain, start, syms); + fill_node(new, chain, start); new->children_hit = 0; new->hit = 1; @@ -228,9 +239,8 @@ add_child(struct callchain_node *parent, struct ip_callchain *chain, * Then create another child to host the given callchain of new branch */ static void -split_add_child(struct callchain_node *parent, struct ip_callchain *chain, - struct callchain_list *to_split, int idx_parents, int idx_local, - struct symbol **syms) +split_add_child(struct callchain_node *parent, struct resolved_chain *chain, + struct callchain_list *to_split, int idx_parents, int idx_local) { struct callchain_node *new; struct list_head *old_tail; @@ -257,7 +267,7 @@ split_add_child(struct callchain_node *parent, struct ip_callchain *chain, /* create a new child for the new branch if any */ if (idx_total < chain->nr) { parent->hit = 0; - add_child(parent, chain, idx_total, syms); + add_child(parent, chain, idx_total); parent->children_hit++; } else { parent->hit = 1; @@ -265,32 +275,33 @@ split_add_child(struct callchain_node *parent, struct ip_callchain *chain, } static int -__append_chain(struct callchain_node *root, struct ip_callchain *chain, - unsigned int start, struct symbol **syms); +__append_chain(struct callchain_node *root, struct resolved_chain *chain, + unsigned int start); static void -__append_chain_children(struct callchain_node *root, struct ip_callchain *chain, - struct symbol **syms, unsigned int start) +__append_chain_children(struct callchain_node *root, + struct resolved_chain *chain, + unsigned int start) { struct callchain_node *rnode; /* lookup in childrens */ chain_for_each_child(rnode, root) { - unsigned int ret = __append_chain(rnode, chain, start, syms); + unsigned int ret = __append_chain(rnode, chain, start); if (!ret) goto inc_children_hit; } /* nothing in children, add to the current node */ - add_child(root, chain, start, syms); + add_child(root, chain, start); inc_children_hit: root->children_hit++; } static int -__append_chain(struct callchain_node *root, struct ip_callchain *chain, - unsigned int start, struct symbol **syms) +__append_chain(struct callchain_node *root, struct resolved_chain *chain, + unsigned int start) { struct callchain_list *cnode; unsigned int i = start; @@ -302,13 +313,19 @@ __append_chain(struct callchain_node *root, struct ip_callchain *chain, * anywhere inside a function. */ list_for_each_entry(cnode, &root->val, list) { + struct symbol *sym; + if (i == chain->nr) break; - if (cnode->sym && syms[i]) { - if (cnode->sym->start != syms[i]->start) + + sym = chain->ips[i].sym; + + if (cnode->sym && sym) { + if (cnode->sym->start != sym->start) break; - } else if (cnode->ip != chain->ips[i]) + } else if (cnode->ip != chain->ips[i].ip) break; + if (!found) found = true; i++; @@ -320,7 +337,7 @@ __append_chain(struct callchain_node *root, struct ip_callchain *chain, /* we match only a part of the node. Split it and add the new chain */ if (i - start < root->val_nr) { - split_add_child(root, chain, cnode, start, i - start, syms); + split_add_child(root, chain, cnode, start, i - start); return 0; } @@ -331,15 +348,51 @@ __append_chain(struct callchain_node *root, struct ip_callchain *chain, } /* We match the node and still have a part remaining */ - __append_chain_children(root, chain, syms, i); + __append_chain_children(root, chain, i); return 0; } -void append_chain(struct callchain_node *root, struct ip_callchain *chain, +static void +filter_context(struct ip_callchain *old, struct resolved_chain *new, + struct symbol **syms) +{ + int i, j = 0; + + for (i = 0; i < (int)old->nr; i++) { + if (old->ips[i] >= PERF_CONTEXT_MAX) + continue; + + new->ips[j].ip = old->ips[i]; + new->ips[j].sym = syms[i]; + j++; + } + + new->nr = j; +} + + +int append_chain(struct callchain_node *root, struct ip_callchain *chain, struct symbol **syms) { + struct resolved_chain *filtered; + if (!chain->nr) - return; - __append_chain_children(root, chain, syms, 0); + return 0; + + filtered = malloc(sizeof(*filtered) + + chain->nr * sizeof(struct resolved_ip)); + if (!filtered) + return -ENOMEM; + + filter_context(chain, filtered, syms); + + if (!filtered->nr) + goto end; + + __append_chain_children(root, filtered, 0); +end: + free(filtered); + + return 0; } diff --git a/tools/perf/util/callchain.h b/tools/perf/util/callchain.h index ad4626d..bbd76da 100644 --- a/tools/perf/util/callchain.h +++ b/tools/perf/util/callchain.h @@ -56,6 +56,6 @@ static inline u64 cumul_hits(struct callchain_node *node) } int register_callchain_param(struct callchain_param *param); -void append_chain(struct callchain_node *root, struct ip_callchain *chain, - struct symbol **syms); +int append_chain(struct callchain_node *root, struct ip_callchain *chain, + struct symbol **syms); #endif /* __PERF_CALLCHAIN_H */ diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index c37da8b..5843a9c 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -328,8 +328,6 @@ static size_t __callchain__fprintf_graph(FILE *fp, struct callchain_node *self, left_margin); i = 0; list_for_each_entry(chain, &child->val, list) { - if (chain->ip >= PERF_CONTEXT_MAX) - continue; ret += ipchain__fprintf_graph(fp, chain, depth, new_depth_mask, i++, new_total, @@ -368,9 +366,6 @@ static size_t callchain__fprintf_graph(FILE *fp, struct callchain_node *self, int ret = 0; list_for_each_entry(chain, &self->val, list) { - if (chain->ip >= PERF_CONTEXT_MAX) - continue; - if (!i++ && sort__first_dimension == SORT_SYM) continue; -- cgit v0.10.2 From f3a1f0ea9432ec395cd112f42201e8e523c07bc5 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 22 Mar 2010 13:10:25 -0300 Subject: perf newt: Properly restore the screen when error exiting MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Show an OK message box with the last message sent via pr_err, etc. Reported-by: Ingo Molnar Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269274229-20442-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/perf.c b/tools/perf/perf.c index d2de839..2826e70 100644 --- a/tools/perf/perf.c +++ b/tools/perf/perf.c @@ -264,11 +264,11 @@ static int run_builtin(struct cmd_struct *p, int argc, const char **argv) set_debugfs_path(); status = p->fn(argc, argv, prefix); + exit_browser(status); + if (status) return status & 0xff; - exit_browser(); - /* Somebody closed stdout? */ if (fstat(fileno(stdout), &st)) return 0; diff --git a/tools/perf/util/cache.h b/tools/perf/util/cache.h index 47b12a3..4b9aab7 100644 --- a/tools/perf/util/cache.h +++ b/tools/perf/util/cache.h @@ -77,10 +77,10 @@ static inline void setup_browser(void) { setup_pager(); } -static inline void exit_browser(void) {} +static inline void exit_browser(bool wait_for_ok __used) {} #else void setup_browser(void); -void exit_browser(void); +void exit_browser(bool wait_for_ok); #endif extern const char *editor_program; diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index 2d19e7a..3c2ef95 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -170,18 +170,20 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, newtFormDestroy(form); } +static char browser__last_msg[1024]; + int browser__show_help(const char *format, va_list ap) { int ret; static int backlog; - static char msg[1024]; - ret = vsnprintf(msg + backlog, sizeof(msg) - backlog, format, ap); + ret = vsnprintf(browser__last_msg + backlog, + sizeof(browser__last_msg) - backlog, format, ap); backlog += ret; - if (msg[backlog - 1] == '\n') { + if (browser__last_msg[backlog - 1] == '\n') { newtPopHelpLine(); - newtPushHelpLine(msg); + newtPushHelpLine(browser__last_msg); newtRefresh(); backlog = 0; } @@ -200,8 +202,13 @@ void setup_browser(void) newtPushHelpLine(" "); } -void exit_browser(void) +void exit_browser(bool wait_for_ok) { - if (use_browser) + if (use_browser) { + if (wait_for_ok) { + char title[] = "Fatal Error", ok[] = "Ok"; + newtWinMessage(title, ok, browser__last_msg); + } newtFinished(); + } } -- cgit v0.10.2 From 4b4da7f76660ea8b5aa45615165c48f62167ffa8 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Mon, 22 Mar 2010 13:10:26 -0300 Subject: perf probe: Cleanup debuginfo related code Cleanup debuginfo related code to eliminate fragile code which pointed by Ingo (Thanks!). 1) Invert logic of NO_DWARF_SUPPORT to DWARF_SUPPORT. 2) For removing assymetric/local variable ifdefs, introduce more helper functions. 3) Change options order to reduce the number of ifdefs. Reported-by: Ingo Molnar Signed-off-by: Masami Hiramatsu Signed-off-by: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <1269274229-20442-2-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/Makefile b/tools/perf/Makefile index 0abd25e..c991818 100644 --- a/tools/perf/Makefile +++ b/tools/perf/Makefile @@ -506,9 +506,8 @@ endif ifneq ($(shell sh -c "(echo '\#include '; echo '\#include '; echo 'int main(void) { Dwarf *dbg; dbg = dwarf_begin(0, DWARF_C_READ); return (long)dbg; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/elfutils -ldw -lelf -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) msg := $(warning No libdw.h found or old libdw.h found, disables dwarf support. Please install elfutils-devel/elfutils-dev); - BASIC_CFLAGS += -DNO_DWARF_SUPPORT else - BASIC_CFLAGS += -I/usr/include/elfutils + BASIC_CFLAGS += -I/usr/include/elfutils -DDWARF_SUPPORT EXTLIBS += -lelf -ldw LIB_OBJS += util/probe-finder.o endif diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c index e0dafd9..cf2ffa5 100644 --- a/tools/perf/builtin-probe.c +++ b/tools/perf/builtin-probe.c @@ -109,7 +109,7 @@ static int opt_del_probe_event(const struct option *opt __used, return 0; } -#ifndef NO_DWARF_SUPPORT +#ifdef DWARF_SUPPORT static int opt_show_lines(const struct option *opt __used, const char *str, int unset __used) { @@ -126,7 +126,7 @@ static const char * const probe_usage[] = { "perf probe [] --add 'PROBEDEF' [--add 'PROBEDEF' ...]", "perf probe [] --del '[GROUP:]EVENT' ...", "perf probe --list", -#ifndef NO_DWARF_SUPPORT +#ifdef DWARF_SUPPORT "perf probe --line 'LINEDESC'", #endif NULL @@ -135,20 +135,16 @@ static const char * const probe_usage[] = { static const struct option options[] = { OPT_BOOLEAN('v', "verbose", &verbose, "be more verbose (show parsed arguments, etc)"), -#ifndef NO_DWARF_SUPPORT - OPT_STRING('k', "vmlinux", &symbol_conf.vmlinux_name, - "file", "vmlinux pathname"), -#endif OPT_BOOLEAN('l', "list", ¶ms.list_events, "list up current probe events"), OPT_CALLBACK('d', "del", NULL, "[GROUP:]EVENT", "delete a probe event.", opt_del_probe_event), OPT_CALLBACK('a', "add", NULL, -#ifdef NO_DWARF_SUPPORT - "[EVENT=]FUNC[+OFF|%return] [ARG ...]", -#else +#ifdef DWARF_SUPPORT "[EVENT=]FUNC[@SRC][+OFF|%return|:RL|;PT]|SRC:AL|SRC;PT" " [ARG ...]", +#else + "[EVENT=]FUNC[+OFF|%return] [ARG ...]", #endif "probe point definition, where\n" "\t\tGROUP:\tGroup name (optional)\n" @@ -156,23 +152,25 @@ static const struct option options[] = { "\t\tFUNC:\tFunction name\n" "\t\tOFF:\tOffset from function entry (in byte)\n" "\t\t%return:\tPut the probe at function return\n" -#ifdef NO_DWARF_SUPPORT - "\t\tARG:\tProbe argument (only \n" -#else +#ifdef DWARF_SUPPORT "\t\tSRC:\tSource code path\n" "\t\tRL:\tRelative line number from function entry.\n" "\t\tAL:\tAbsolute line number in file.\n" "\t\tPT:\tLazy expression of line code.\n" "\t\tARG:\tProbe argument (local variable name or\n" -#endif "\t\t\tkprobe-tracer argument format.)\n", +#else + "\t\tARG:\tProbe argument (kprobe-tracer argument format.)\n", +#endif opt_add_probe_event), OPT_BOOLEAN('f', "force", ¶ms.force_add, "forcibly add events" " with existing name"), -#ifndef NO_DWARF_SUPPORT +#ifdef DWARF_SUPPORT OPT_CALLBACK('L', "line", NULL, "FUNC[:RLN[+NUM|:RLN2]]|SRC:ALN[+NUM|:ALN2]", "Show source code lines.", opt_show_lines), + OPT_STRING('k', "vmlinux", &symbol_conf.vmlinux_name, + "file", "vmlinux pathname"), #endif OPT__DRY_RUN(&probe_event_dry_run), OPT_END() @@ -211,7 +209,7 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used) return 0; } -#ifndef NO_DWARF_SUPPORT +#ifdef DWARF_SUPPORT if (params.show_lines) { if (params.nevents != 0 || params.dellist) { pr_warning(" Error: Don't use --line with" diff --git a/tools/perf/util/probe-event.c b/tools/perf/util/probe-event.c index c6603f3..3fc0be7 100644 --- a/tools/perf/util/probe-event.c +++ b/tools/perf/util/probe-event.c @@ -42,7 +42,8 @@ #include "color.h" #include "symbol.h" #include "thread.h" -#include "parse-events.h" /* For debugfs_path */ +#include "trace-event.h" /* For __unused */ +#include "parse-events.h" /* For debugfs_path */ #include "probe-event.h" #include "probe-finder.h" @@ -70,10 +71,11 @@ static int e_snprintf(char *str, size_t size, const char *format, ...) return ret; } +static char *synthesize_perf_probe_point(struct perf_probe_point *pp); static struct map_groups kmap_groups; static struct map *kmaps[MAP__NR_TYPES]; -/* Initialize symbol maps for vmlinux */ +/* Initialize symbol maps and path of vmlinux */ static void init_vmlinux(void) { symbol_conf.sort_by_name = true; @@ -89,7 +91,7 @@ static void init_vmlinux(void) die("Failed to create kernel maps."); } -#ifndef NO_DWARF_SUPPORT +#ifdef DWARF_SUPPORT static int open_vmlinux(void) { if (map__load(kmaps[MAP__FUNCTION], NULL) < 0) { @@ -99,6 +101,176 @@ static int open_vmlinux(void) pr_debug("Try to open %s\n", kmaps[MAP__FUNCTION]->dso->long_name); return open(kmaps[MAP__FUNCTION]->dso->long_name, O_RDONLY); } + +static void convert_to_perf_probe_point(struct kprobe_trace_point *tp, + struct perf_probe_point *pp) +{ + struct symbol *sym; + int fd, ret = 0; + + sym = map__find_symbol_by_name(kmaps[MAP__FUNCTION], + tp->symbol, NULL); + if (sym) { + fd = open_vmlinux(); + ret = find_perf_probe_point(fd, sym->start + tp->offset, pp); + close(fd); + } + if (ret <= 0) { + pp->function = xstrdup(tp->symbol); + pp->offset = tp->offset; + } + pp->retprobe = tp->retprobe; +} + +/* Try to find perf_probe_event with debuginfo */ +static int try_to_find_kprobe_trace_events(struct perf_probe_event *pev, + struct kprobe_trace_event **tevs) +{ + bool need_dwarf = perf_probe_event_need_dwarf(pev); + int fd, ntevs; + + fd = open_vmlinux(); + if (fd < 0) { + if (need_dwarf) + die("Could not open debuginfo file."); + + pr_debug("Could not open vmlinux. Try to use symbols.\n"); + return 0; + } + + /* Searching trace events corresponding to probe event */ + ntevs = find_kprobe_trace_events(fd, pev, tevs); + close(fd); + + if (ntevs > 0) /* Succeeded to find trace events */ + return ntevs; + + if (ntevs == 0) /* No error but failed to find probe point. */ + die("Probe point '%s' not found. - probe not added.", + synthesize_perf_probe_point(&pev->point)); + + /* Error path */ + if (need_dwarf) { + if (ntevs == -ENOENT) + pr_warning("No dwarf info found in the vmlinux - " + "please rebuild with CONFIG_DEBUG_INFO=y.\n"); + die("Could not analyze debuginfo."); + } + pr_debug("An error occurred in debuginfo analysis." + " Try to use symbols.\n"); + return 0; + +} + +#define LINEBUF_SIZE 256 +#define NR_ADDITIONAL_LINES 2 + +static void show_one_line(FILE *fp, unsigned int l, bool skip, bool show_num) +{ + char buf[LINEBUF_SIZE]; + const char *color = PERF_COLOR_BLUE; + + if (fgets(buf, LINEBUF_SIZE, fp) == NULL) + goto error; + if (!skip) { + if (show_num) + fprintf(stdout, "%7u %s", l, buf); + else + color_fprintf(stdout, color, " %s", buf); + } + + while (strlen(buf) == LINEBUF_SIZE - 1 && + buf[LINEBUF_SIZE - 2] != '\n') { + if (fgets(buf, LINEBUF_SIZE, fp) == NULL) + goto error; + if (!skip) { + if (show_num) + fprintf(stdout, "%s", buf); + else + color_fprintf(stdout, color, "%s", buf); + } + } + return; +error: + if (feof(fp)) + die("Source file is shorter than expected."); + else + die("File read error: %s", strerror(errno)); +} + +/* + * Show line-range always requires debuginfo to find source file and + * line number. + */ +void show_line_range(struct line_range *lr) +{ + unsigned int l = 1; + struct line_node *ln; + FILE *fp; + int fd, ret; + + /* Search a line range */ + init_vmlinux(); + fd = open_vmlinux(); + if (fd < 0) + die("Could not open debuginfo file."); + ret = find_line_range(fd, lr); + if (ret <= 0) + die("Source line is not found.\n"); + close(fd); + + setup_pager(); + + if (lr->function) + fprintf(stdout, "<%s:%d>\n", lr->function, + lr->start - lr->offset); + else + fprintf(stdout, "<%s:%d>\n", lr->file, lr->start); + + fp = fopen(lr->path, "r"); + if (fp == NULL) + die("Failed to open %s: %s", lr->path, strerror(errno)); + /* Skip to starting line number */ + while (l < lr->start) + show_one_line(fp, l++, true, false); + + list_for_each_entry(ln, &lr->line_list, list) { + while (ln->line > l) + show_one_line(fp, (l++) - lr->offset, false, false); + show_one_line(fp, (l++) - lr->offset, false, true); + } + + if (lr->end == INT_MAX) + lr->end = l + NR_ADDITIONAL_LINES; + while (l < lr->end && !feof(fp)) + show_one_line(fp, (l++) - lr->offset, false, false); + + fclose(fp); +} + +#else /* !DWARF_SUPPORT */ + +static void convert_to_perf_probe_point(struct kprobe_trace_point *tp, + struct perf_probe_point *pp) +{ + pp->function = xstrdup(tp->symbol); + pp->offset = tp->offset; + pp->retprobe = tp->retprobe; +} + +static int try_to_find_kprobe_trace_events(struct perf_probe_event *pev, + struct kprobe_trace_event **tevs __unused) +{ + if (perf_probe_event_need_dwarf(pev)) + die("Debuginfo-analysis is not supported"); + return 0; +} + +void show_line_range(struct line_range *lr __unused) +{ + die("Debuginfo-analysis is not supported"); +} + #endif void parse_line_range_desc(const char *arg, struct line_range *lr) @@ -592,32 +764,14 @@ void convert_to_perf_probe_event(struct kprobe_trace_event *tev, { char buf[64]; int i; -#ifndef NO_DWARF_SUPPORT - struct symbol *sym; - int fd, ret = 0; - - sym = map__find_symbol_by_name(kmaps[MAP__FUNCTION], - tev->point.symbol, NULL); - if (sym) { - fd = open_vmlinux(); - ret = find_perf_probe_point(fd, sym->start + tev->point.offset, - &pev->point); - close(fd); - } - if (ret <= 0) { - pev->point.function = xstrdup(tev->point.symbol); - pev->point.offset = tev->point.offset; - } -#else - /* Convert trace_point to probe_point */ - pev->point.function = xstrdup(tev->point.symbol); - pev->point.offset = tev->point.offset; -#endif - pev->point.retprobe = tev->point.retprobe; + /* Convert event/group name */ pev->event = xstrdup(tev->event); pev->group = xstrdup(tev->group); + /* Convert trace_point to probe_point */ + convert_to_perf_probe_point(&tev->point, &pev->point); + /* Convert trace_arg to probe_arg */ pev->nargs = tev->nargs; pev->args = xzalloc(sizeof(struct perf_probe_arg) * pev->nargs); @@ -944,52 +1098,13 @@ static int convert_to_kprobe_trace_events(struct perf_probe_event *pev, struct kprobe_trace_event **tevs) { struct symbol *sym; - bool need_dwarf; -#ifndef NO_DWARF_SUPPORT - int fd; -#endif int ntevs = 0, i; struct kprobe_trace_event *tev; - need_dwarf = perf_probe_event_need_dwarf(pev); - - if (need_dwarf) -#ifdef NO_DWARF_SUPPORT - die("Debuginfo-analysis is not supported"); -#else /* !NO_DWARF_SUPPORT */ - pr_debug("Some probes require debuginfo.\n"); - - fd = open_vmlinux(); - if (fd < 0) { - if (need_dwarf) - die("Could not open debuginfo file."); - - pr_debug("Could not open vmlinux/module file." - " Try to use symbols.\n"); - goto end_dwarf; - } - - /* Searching probe points */ - ntevs = find_kprobe_trace_events(fd, pev, tevs); - - if (ntevs > 0) /* Found */ - goto found; - - if (ntevs == 0) /* No error but failed to find probe point. */ - die("Probe point '%s' not found. - probe not added.", - synthesize_perf_probe_point(&pev->point)); - - /* Error path */ - if (need_dwarf) { - if (ntevs == -ENOENT) - pr_warning("No dwarf info found in the vmlinux - please rebuild with CONFIG_DEBUG_INFO=y.\n"); - die("Could not analyze debuginfo."); - } - pr_debug("An error occurred in debuginfo analysis." - " Try to use symbols.\n"); - -end_dwarf: -#endif /* !NO_DWARF_SUPPORT */ + /* Convert perf_probe_event with debuginfo */ + ntevs = try_to_find_kprobe_trace_events(pev, tevs); + if (ntevs > 0) + return ntevs; /* Allocate trace event buffer */ ntevs = 1; @@ -1012,10 +1127,7 @@ end_dwarf: if (!sym) die("Kernel symbol \'%s\' not found - probe not added.", tev->point.symbol); -#ifndef NO_DWARF_SUPPORT -found: - close(fd); -#endif + return ntevs; } @@ -1133,90 +1245,3 @@ void del_perf_probe_events(struct strlist *dellist) close(fd); } -#define LINEBUF_SIZE 256 -#define NR_ADDITIONAL_LINES 2 - -static void show_one_line(FILE *fp, unsigned int l, bool skip, bool show_num) -{ - char buf[LINEBUF_SIZE]; - const char *color = PERF_COLOR_BLUE; - - if (fgets(buf, LINEBUF_SIZE, fp) == NULL) - goto error; - if (!skip) { - if (show_num) - fprintf(stdout, "%7u %s", l, buf); - else - color_fprintf(stdout, color, " %s", buf); - } - - while (strlen(buf) == LINEBUF_SIZE - 1 && - buf[LINEBUF_SIZE - 2] != '\n') { - if (fgets(buf, LINEBUF_SIZE, fp) == NULL) - goto error; - if (!skip) { - if (show_num) - fprintf(stdout, "%s", buf); - else - color_fprintf(stdout, color, "%s", buf); - } - } - return; -error: - if (feof(fp)) - die("Source file is shorter than expected."); - else - die("File read error: %s", strerror(errno)); -} - -void show_line_range(struct line_range *lr) -{ - unsigned int l = 1; - struct line_node *ln; - FILE *fp; -#ifndef NO_DWARF_SUPPORT - int fd, ret; -#endif - - /* Search a line range */ - init_vmlinux(); -#ifndef NO_DWARF_SUPPORT - fd = open_vmlinux(); - if (fd < 0) - die("Could not open debuginfo file."); - ret = find_line_range(fd, lr); - if (ret <= 0) - die("Source line is not found.\n"); - close(fd); -#endif - - setup_pager(); - - if (lr->function) - fprintf(stdout, "<%s:%d>\n", lr->function, - lr->start - lr->offset); - else - fprintf(stdout, "<%s:%d>\n", lr->file, lr->start); - - fp = fopen(lr->path, "r"); - if (fp == NULL) - die("Failed to open %s: %s", lr->path, strerror(errno)); - /* Skip to starting line number */ - while (l < lr->start) - show_one_line(fp, l++, true, false); - - list_for_each_entry(ln, &lr->line_list, list) { - while (ln->line > l) - show_one_line(fp, (l++) - lr->offset, false, false); - show_one_line(fp, (l++) - lr->offset, false, true); - } - - if (lr->end == INT_MAX) - lr->end = l + NR_ADDITIONAL_LINES; - while (l < lr->end && !feof(fp)) - show_one_line(fp, (l++) - lr->offset, false, false); - - fclose(fp); -} - - diff --git a/tools/perf/util/probe-finder.h b/tools/perf/util/probe-finder.h index 2f2307d..3564f22 100644 --- a/tools/perf/util/probe-finder.h +++ b/tools/perf/util/probe-finder.h @@ -15,7 +15,7 @@ static inline int is_c_varname(const char *name) return isalpha(name[0]) || name[0] == '_'; } -#ifndef NO_DWARF_SUPPORT +#ifdef DWARF_SUPPORT /* Find kprobe_trace_events specified by perf_probe_event from debuginfo */ extern int find_kprobe_trace_events(int fd, struct perf_probe_event *pev, struct kprobe_trace_event **tevs); @@ -57,6 +57,6 @@ struct line_finder { int found; }; -#endif /* NO_DWARF_SUPPORT */ +#endif /* DWARF_SUPPORT */ #endif /*_PROBE_FINDER_H */ -- cgit v0.10.2 From ca721e45b39209415d2288dbac3667b26d9d1def Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Mon, 22 Mar 2010 13:10:27 -0300 Subject: perf probe: Add NO_DWARF make option Add NO_DWARF make option for testing build without libdw. Signed-off-by: Masami Hiramatsu Signed-off-by: Arnaldo Carvalho de Melo Cc: Frederic Weisbecker Cc: Paul Mackerras Cc: Mike Galbraith Cc: Peter Zijlstra LKML-Reference: <1269274229-20442-3-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/Makefile b/tools/perf/Makefile index c991818..6903645 100644 --- a/tools/perf/Makefile +++ b/tools/perf/Makefile @@ -150,6 +150,8 @@ all:: # Define LDFLAGS=-static to build a static binary. # # Define EXTRA_CFLAGS=-m64 or EXTRA_CFLAGS=-m32 as appropriate for cross-builds. +# +# Define NO_DWARF if you do not want debug-info analysis feature at all. PERF-VERSION-FILE: .FORCE-PERF-VERSION-FILE @$(SHELL_PATH) util/PERF-VERSION-GEN @@ -507,10 +509,12 @@ endif ifneq ($(shell sh -c "(echo '\#include '; echo '\#include '; echo 'int main(void) { Dwarf *dbg; dbg = dwarf_begin(0, DWARF_C_READ); return (long)dbg; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/elfutils -ldw -lelf -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) msg := $(warning No libdw.h found or old libdw.h found, disables dwarf support. Please install elfutils-devel/elfutils-dev); else +ifndef NO_DWARF BASIC_CFLAGS += -I/usr/include/elfutils -DDWARF_SUPPORT EXTLIBS += -lelf -ldw LIB_OBJS += util/probe-finder.o endif +endif ifneq ($(shell sh -c "(echo '\#include '; echo 'int main(void) { newtInit(); newtCls(); return newtFinished(); }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -lnewt -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) msg := $(warning newt not found, disables TUI support. Please install newt-devel or libnewt-dev); -- cgit v0.10.2 From 084ab9f862416b2ddb4bb9804884de19bf09774d Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 22 Mar 2010 13:10:28 -0300 Subject: perf stat: Better report failure to collect system wide stats MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Before: [acme@doppio linux-2.6-tip]$ perf stat -a sleep 1s Performance counter stats for 'sleep 1s': task-clock-msecs context-switches CPU-migrations page-faults cycles instructions branches branch-misses cache-references cache-misses 1.016998463 seconds time elapsed [acme@doppio linux-2.6-tip]$ Now: [acme@doppio linux-2.6-tip]$ perf stat -a sleep 1s No permission to collect system-wide stats. Consider tweaking /proc/sys/kernel/perf_event_paranoid. [acme@doppio linux-2.6-tip]$ Reported-by: Ingo Molnar Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269274229-20442-4-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c index c92f90f..1036ca7 100644 --- a/tools/perf/builtin-stat.c +++ b/tools/perf/builtin-stat.c @@ -144,10 +144,11 @@ struct stats runtime_branches_stats; #define ERR_PERF_OPEN \ "Error: counter %d, sys_perf_event_open() syscall returned with %d (%s)\n" -static void create_perf_stat_counter(int counter) +static int create_perf_stat_counter(int counter) { struct perf_event_attr *attr = attrs + counter; int thread; + int ncreated = 0; if (scale) attr->read_format = PERF_FORMAT_TOTAL_TIME_ENABLED | @@ -159,9 +160,11 @@ static void create_perf_stat_counter(int counter) for (cpu = 0; cpu < nr_cpus; cpu++) { fd[cpu][counter][0] = sys_perf_event_open(attr, -1, cpumap[cpu], -1, 0); - if (fd[cpu][counter][0] < 0 && verbose) - fprintf(stderr, ERR_PERF_OPEN, counter, - fd[cpu][counter][0], strerror(errno)); + if (fd[cpu][counter][0] < 0) + pr_debug(ERR_PERF_OPEN, counter, + fd[cpu][counter][0], strerror(errno)); + else + ++ncreated; } } else { attr->inherit = inherit; @@ -172,12 +175,16 @@ static void create_perf_stat_counter(int counter) for (thread = 0; thread < thread_num; thread++) { fd[0][counter][thread] = sys_perf_event_open(attr, all_tids[thread], -1, -1, 0); - if (fd[0][counter][thread] < 0 && verbose) - fprintf(stderr, ERR_PERF_OPEN, counter, - fd[0][counter][thread], - strerror(errno)); + if (fd[0][counter][thread] < 0) + pr_debug(ERR_PERF_OPEN, counter, + fd[0][counter][thread], + strerror(errno)); + else + ++ncreated; } } + + return ncreated; } /* @@ -264,7 +271,7 @@ static int run_perf_stat(int argc __used, const char **argv) { unsigned long long t0, t1; int status = 0; - int counter; + int counter, ncreated = 0; int child_ready_pipe[2], go_pipe[2]; const bool forks = (argc > 0); char buf; @@ -324,7 +331,16 @@ static int run_perf_stat(int argc __used, const char **argv) } for (counter = 0; counter < nr_counters; counter++) - create_perf_stat_counter(counter); + ncreated += create_perf_stat_counter(counter); + + if (ncreated == 0) { + pr_err("No permission to collect %sstats.\n" + "Consider tweaking /proc/sys/kernel/perf_event_paranoid.\n", + system_wide ? "system-wide " : ""); + if (child_pid != -1) + kill(child_pid, SIGTERM); + return -1; + } /* * Enable counters and exec the command: @@ -587,7 +603,8 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used) status = run_perf_stat(argc, argv); } - print_stat(argc, argv); + if (status != -1) + print_stat(argc, argv); return status; } -- cgit v0.10.2 From 478b0973bf8c90db3677fbb8d812e2bdefc43d9b Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 22 Mar 2010 13:10:29 -0300 Subject: perf tools: Exit browser before printing usage when unkown option passed MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If not the screen will get garbled when using newt. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269274229-20442-5-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/parse-options.c b/tools/perf/util/parse-options.c index efebd5b..79dfa0c 100644 --- a/tools/perf/util/parse-options.c +++ b/tools/perf/util/parse-options.c @@ -500,6 +500,7 @@ int usage_with_options_internal(const char * const *usagestr, void usage_with_options(const char * const *usagestr, const struct option *opts) { + exit_browser(false); usage_with_options_internal(usagestr, opts, 0); exit(129); } -- cgit v0.10.2 From 91885258e8343bb65c08f668d7e6c16563eb4284 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 19 Mar 2010 08:06:28 -0400 Subject: nfsd: don't break lease while servicing a COMMIT This is the second attempt to fix the problem whereby a COMMIT call causes a lease break and triggers a possible deadlock. The problem is that nfsd attempts to break a lease on a COMMIT call. This triggers a delegation recall if the lease is held for a delegation. If the client is the one holding the delegation and it's the same one on which it's issuing the COMMIT, then it can't return that delegation until the COMMIT is complete. But, nfsd won't complete the COMMIT until the delegation is returned. The client and server are essentially deadlocked until the state is marked bad (due to the client not responding on the callback channel). The first patch attempted to deal with this by eliminating the open of the file altogether and simply had nfsd_commit pass a NULL file pointer to the vfs_fsync_range. That would conflict with some work in progress by Christoph Hellwig to clean up the fsync interface, so this patch takes a different approach. This declares a new NFSD_MAY_NOT_BREAK_LEASE access flag that indicates to nfsd_open that it should not break any leases when opening the file, and has nfsd_commit set that flag on the nfsd_open call. For now, this patch leaves nfsd_commit opening the file with write access since I'm not clear on what sort of access would be more appropriate. Signed-off-by: Jeff Layton Cc: stable@kernel.org Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index a11b0e867..c2dcb4c 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -723,7 +723,7 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, struct inode *inode; int flags = O_RDONLY|O_LARGEFILE; __be32 err; - int host_err; + int host_err = 0; validate_process_creds(); @@ -760,7 +760,8 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, * Check to see if there are any leases on this file. * This may block while leases are broken. */ - host_err = break_lease(inode, O_NONBLOCK | ((access & NFSD_MAY_WRITE) ? O_WRONLY : 0)); + if (!(access & NFSD_MAY_NOT_BREAK_LEASE)) + host_err = break_lease(inode, O_NONBLOCK | ((access & NFSD_MAY_WRITE) ? O_WRONLY : 0)); if (host_err == -EWOULDBLOCK) host_err = -ETIMEDOUT; if (host_err) /* NOMEM or WOULDBLOCK */ @@ -1168,7 +1169,8 @@ nfsd_commit(struct svc_rqst *rqstp, struct svc_fh *fhp, goto out; } - err = nfsd_open(rqstp, fhp, S_IFREG, NFSD_MAY_WRITE, &file); + err = nfsd_open(rqstp, fhp, S_IFREG, + NFSD_MAY_WRITE|NFSD_MAY_NOT_BREAK_LEASE, &file); if (err) goto out; if (EX_ISSYNC(fhp->fh_export)) { diff --git a/fs/nfsd/vfs.h b/fs/nfsd/vfs.h index 4b1de0a..217a62c 100644 --- a/fs/nfsd/vfs.h +++ b/fs/nfsd/vfs.h @@ -20,6 +20,7 @@ #define NFSD_MAY_OWNER_OVERRIDE 64 #define NFSD_MAY_LOCAL_ACCESS 128 /* IRIX doing local access check on device special file*/ #define NFSD_MAY_BYPASS_GSS_ON_ROOT 256 +#define NFSD_MAY_NOT_BREAK_LEASE 512 #define NFSD_MAY_CREATE (NFSD_MAY_EXEC|NFSD_MAY_WRITE) #define NFSD_MAY_REMOVE (NFSD_MAY_EXEC|NFSD_MAY_WRITE|NFSD_MAY_TRUNC) -- cgit v0.10.2 From 4ded2b250f1fbba4e414d17dc55ee513485c0aa1 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Mon, 22 Mar 2010 17:52:49 -0300 Subject: perf report: Implement Newt callgraphs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Starts collapsed, allows annotating by pressing 'A' or 'a' on the symbol, be it the top level one or any of the symbols in the chains. It (ab)uses the only tree widget in newt, that is actually a checkbox tree that we use with just one option ('.'), end result is usable but we really need to create a custom widget tree so that we can use the data structures we have (hist_entry rb_tree + callchain rb_tree + lists), so that we reduce the memory footprint by not creating a mirror set of data structures in the newtCheckboxTree widget. Thanks to Frédéric Weisbacker for fixing the orphanage problem in 301fde2, without that we were tripping a newt bug (fix already sent to newt's maintainer). Signed-off-by: Arnaldo Carvalho de Melo Cc: Avi Kivity Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269291169-29820-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index 3c2ef95..12b229b 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -28,11 +28,197 @@ static newtComponent newt_form__new(void) return self; } +/* + * When debugging newt problems it was useful to be able to "unroll" + * the calls to newtCheckBoxTreeAdd{Array,Item}, so that we can generate + * a source file with the sequence of calls to these methods, to then + * tweak the arrays to get the intended results, so I'm keeping this code + * here, may be useful again in the future. + */ +#undef NEWT_DEBUG + +static void newt_checkbox_tree__add(newtComponent tree, const char *str, + void *priv, int *indexes) +{ +#ifdef NEWT_DEBUG + /* Print the newtCheckboxTreeAddArray to tinker with its index arrays */ + int i = 0, len = 40 - strlen(str); + + fprintf(stderr, + "\tnewtCheckboxTreeAddItem(tree, %*.*s\"%s\", (void *)%p, 0, ", + len, len, " ", str, priv); + while (indexes[i] != NEWT_ARG_LAST) { + if (indexes[i] != NEWT_ARG_APPEND) + fprintf(stderr, " %d,", indexes[i]); + else + fprintf(stderr, " %s,", "NEWT_ARG_APPEND"); + ++i; + } + fprintf(stderr, " %s", " NEWT_ARG_LAST);\n"); + fflush(stderr); +#endif + newtCheckboxTreeAddArray(tree, str, priv, 0, indexes); +} + +static char *callchain_list__sym_name(struct callchain_list *self, + char *bf, size_t bfsize) +{ + if (self->sym) + return self->sym->name; + + snprintf(bf, bfsize, "%#Lx", self->ip); + return bf; +} + +static void __callchain__append_graph_browser(struct callchain_node *self, + newtComponent tree, u64 total, + int *indexes, int depth) +{ + struct rb_node *node; + u64 new_total, remaining; + int idx = 0; + + if (callchain_param.mode == CHAIN_GRAPH_REL) + new_total = self->children_hit; + else + new_total = total; + + remaining = new_total; + node = rb_first(&self->rb_root); + while (node) { + struct callchain_node *child = rb_entry(node, struct callchain_node, rb_node); + struct rb_node *next = rb_next(node); + u64 cumul = cumul_hits(child); + struct callchain_list *chain; + int first = true, printed = 0; + int chain_idx = -1; + remaining -= cumul; + + indexes[depth] = NEWT_ARG_APPEND; + indexes[depth + 1] = NEWT_ARG_LAST; + + list_for_each_entry(chain, &child->val, list) { + char ipstr[BITS_PER_LONG / 4 + 1], + *alloc_str = NULL; + const char *str = callchain_list__sym_name(chain, ipstr, sizeof(ipstr)); + + if (first) { + double percent = cumul * 100.0 / new_total; + + first = false; + if (asprintf(&alloc_str, "%2.2f%% %s", percent, str) < 0) + str = "Not enough memory!"; + else + str = alloc_str; + } else { + indexes[depth] = idx; + indexes[depth + 1] = NEWT_ARG_APPEND; + indexes[depth + 2] = NEWT_ARG_LAST; + ++chain_idx; + } + newt_checkbox_tree__add(tree, str, chain->sym, indexes); + free(alloc_str); + ++printed; + } + + indexes[depth] = idx; + if (chain_idx != -1) + indexes[depth + 1] = chain_idx; + if (printed != 0) + ++idx; + __callchain__append_graph_browser(child, tree, new_total, indexes, + depth + (chain_idx != -1 ? 2 : 1)); + node = next; + } +} + +static void callchain__append_graph_browser(struct callchain_node *self, + newtComponent tree, u64 total, + int *indexes, int parent_idx) +{ + struct callchain_list *chain; + int i = 0; + + indexes[1] = NEWT_ARG_APPEND; + indexes[2] = NEWT_ARG_LAST; + + list_for_each_entry(chain, &self->val, list) { + char ipstr[BITS_PER_LONG / 4 + 1], *str; + + if (chain->ip >= PERF_CONTEXT_MAX) + continue; + + if (!i++ && sort__first_dimension == SORT_SYM) + continue; + + str = callchain_list__sym_name(chain, ipstr, sizeof(ipstr)); + newt_checkbox_tree__add(tree, str, chain->sym, indexes); + } + + indexes[1] = parent_idx; + indexes[2] = NEWT_ARG_APPEND; + indexes[3] = NEWT_ARG_LAST; + __callchain__append_graph_browser(self, tree, total, indexes, 2); +} + +static void hist_entry__append_callchain_browser(struct hist_entry *self, + newtComponent tree, u64 total, int parent_idx) +{ + struct rb_node *rb_node; + int indexes[1024] = { [0] = parent_idx, }; + int idx = 0; + struct callchain_node *chain; + + rb_node = rb_first(&self->sorted_chain); + while (rb_node) { + chain = rb_entry(rb_node, struct callchain_node, rb_node); + switch (callchain_param.mode) { + case CHAIN_FLAT: + break; + case CHAIN_GRAPH_ABS: /* falldown */ + case CHAIN_GRAPH_REL: + callchain__append_graph_browser(chain, tree, total, indexes, idx++); + break; + case CHAIN_NONE: + default: + break; + } + rb_node = rb_next(rb_node); + } +} + +/* + * FIXME: get lib/string.c linked with perf somehow + */ +static char *skip_spaces(const char *str) +{ + while (isspace(*str)) + ++str; + return (char *)str; +} + +static char *strim(char *s) +{ + size_t size; + char *end; + + s = skip_spaces(s); + size = strlen(s); + if (!size) + return s; + + end = s + size - 1; + while (end >= s && isspace(*end)) + end--; + *(end + 1) = '\0'; + + return s; +} + static size_t hist_entry__append_browser(struct hist_entry *self, - newtComponent listbox, u64 total) + newtComponent tree, u64 total) { - char bf[1024]; - size_t len; + char bf[1024], *s; FILE *fp; if (symbol_conf.exclude_other && !self->parent) @@ -42,28 +228,46 @@ static size_t hist_entry__append_browser(struct hist_entry *self, if (fp == NULL) return 0; - len = hist_entry__fprintf(self, NULL, false, 0, fp, total); - + hist_entry__fprintf(self, NULL, false, 0, fp, total); fclose(fp); - newtListboxAppendEntry(listbox, bf, self); - return len; + + /* + * FIXME: We shouldn't need to trim, as the printing routines shouldn't + * add spaces it in the first place, the stdio output routines should + * call a __snprintf method instead of the current __print (that + * actually is a __fprintf) one, but get the raw string and _then_ add + * the newline, as this is a detail of stdio printing, not needed in + * other UIs, e.g. newt. + */ + s = strim(bf); + + if (symbol_conf.use_callchain) { + int indexes[2]; + + indexes[0] = NEWT_ARG_APPEND; + indexes[1] = NEWT_ARG_LAST; + newt_checkbox_tree__add(tree, s, self->sym, indexes); + } else + newtListboxAppendEntry(tree, s, self->sym); + + return strlen(s); } -static void hist_entry__annotate_browser(struct hist_entry *self) +static void symbol__annotate_browser(const struct symbol *self) { FILE *fp; int cols, rows; - newtComponent form, listbox; + newtComponent form, tree; struct newtExitStruct es; char *str; size_t line_len, max_line_len = 0; size_t max_usable_width; char *line = NULL; - if (self->sym == NULL) + if (self == NULL) return; - if (asprintf(&str, "perf annotate %s 2>&1 | expand", self->sym->name) < 0) + if (asprintf(&str, "perf annotate %s 2>&1 | expand", self->name) < 0) return; fp = popen(str, "r"); @@ -72,7 +276,7 @@ static void hist_entry__annotate_browser(struct hist_entry *self) newtPushHelpLine("Press ESC to exit"); newtGetScreenSize(&cols, &rows); - listbox = newtListbox(0, 0, rows - 5, NEWT_FLAG_SCROLL); + tree = newtListbox(0, 0, rows - 5, NEWT_FLAG_SCROLL); while (!feof(fp)) { if (getline(&line, &line_len, fp) < 0 || !line_len) @@ -82,7 +286,7 @@ static void hist_entry__annotate_browser(struct hist_entry *self) if (line_len > max_line_len) max_line_len = line_len; - newtListboxAppendEntry(listbox, line, NULL); + newtListboxAppendEntry(tree, line, NULL); } fclose(fp); free(line); @@ -91,11 +295,11 @@ static void hist_entry__annotate_browser(struct hist_entry *self) if (max_line_len > max_usable_width) max_line_len = max_usable_width; - newtListboxSetWidth(listbox, max_line_len); + newtListboxSetWidth(tree, max_line_len); - newtCenteredWindow(max_line_len + 2, rows - 5, self->sym->name); + newtCenteredWindow(max_line_len + 2, rows - 5, self->name); form = newt_form__new(); - newtFormAddComponents(form, listbox, NULL); + newtFormAddComponents(form, tree, NULL); newtFormRun(form, &es); newtFormDestroy(form); @@ -110,25 +314,27 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, { struct sort_entry *se; struct rb_node *nd; + char seq[] = "."; unsigned int width; char *col_width = symbol_conf.col_width_list_str; - int rows; - size_t max_len = 0; + int rows, cols, idx; + int max_len = 0; char str[1024]; - newtComponent form, listbox; + newtComponent form, tree; struct newtExitStruct es; snprintf(str, sizeof(str), "Samples: %Ld", session_total); newtDrawRootText(0, 0, str); newtPushHelpLine(helpline); - newtGetScreenSize(NULL, &rows); - - form = newt_form__new(); + newtGetScreenSize(&cols, &rows); - listbox = newtListbox(1, 1, rows - 2, (NEWT_FLAG_SCROLL | - NEWT_FLAG_BORDER | - NEWT_FLAG_RETURNEXIT)); + if (symbol_conf.use_callchain) + tree = newtCheckboxTreeMulti(0, 0, rows - 5, seq, + NEWT_FLAG_SCROLL); + else + tree = newtListbox(0, 0, rows - 5, (NEWT_FLAG_SCROLL | + NEWT_FLAG_RETURNEXIT)); list_for_each_entry(se, &hist_entry__sort_list, list) { if (se->elide) @@ -147,27 +353,48 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, } } + idx = 0; for (nd = rb_first(hists); nd; nd = rb_next(nd)) { struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node); - size_t len = hist_entry__append_browser(h, listbox, session_total); + int len = hist_entry__append_browser(h, tree, session_total); if (len > max_len) max_len = len; + if (symbol_conf.use_callchain) { + hist_entry__append_callchain_browser(h, tree, session_total, idx++); + if (idx > 3300) + break; + } } - newtListboxSetWidth(listbox, max_len); - newtFormAddComponents(form, listbox, NULL); + if (max_len > cols) + max_len = cols - 3; + + if (!symbol_conf.use_callchain) + newtListboxSetWidth(tree, max_len); + + newtCenteredWindow(max_len + (symbol_conf.use_callchain ? 5 : 0), + rows - 5, "Report"); + form = newt_form__new(); + newtFormAddHotKey(form, 'A'); + newtFormAddHotKey(form, 'a'); + newtFormAddComponents(form, tree, NULL); while (1) { - struct hist_entry *selection; + const struct symbol *selection; newtFormRun(form, &es); - if (es.reason == NEWT_EXIT_HOTKEY) + if (es.reason == NEWT_EXIT_HOTKEY && + toupper(es.u.key) != 'A') break; - selection = newtListboxGetCurrent(listbox); - hist_entry__annotate_browser(selection); + if (!symbol_conf.use_callchain) + selection = newtListboxGetCurrent(tree); + else + selection = newtCheckboxTreeGetCurrent(tree); + symbol__annotate_browser(selection); } newtFormDestroy(form); + newtPopWindow(); } static char browser__last_msg[1024]; -- cgit v0.10.2 From 1ee4d61fd9822fb89e63b88a66848477087cd82e Mon Sep 17 00:00:00 2001 From: Zhang Rui Date: Mon, 22 Mar 2010 15:46:49 +0800 Subject: ACPI dock: support multiple ACPI dock devices There may be multiple ACPI dock devices exist in ACPI namespace and we should probe all of them. http://bugzilla.kernel.org/show_bug.cgi?id=15521 CC: Li Shaohua Signed-off-by: Zhang Rui Signed-off-by: Len Brown diff --git a/drivers/acpi/dock.c b/drivers/acpi/dock.c index d9a85f1..9d67bc6 100644 --- a/drivers/acpi/dock.c +++ b/drivers/acpi/dock.c @@ -1025,13 +1025,10 @@ static int dock_remove(struct dock_station *ds) static acpi_status find_dock(acpi_handle handle, u32 lvl, void *context, void **rv) { - acpi_status status = AE_OK; - if (is_dock(handle)) - if (dock_add(handle) >= 0) - status = AE_CTRL_TERMINATE; + dock_add(handle); - return status; + return AE_OK; } static acpi_status -- cgit v0.10.2 From bc73675b99fd9850dd914be01d71af99c5d2a1ae Mon Sep 17 00:00:00 2001 From: Zhang Rui Date: Mon, 22 Mar 2010 15:48:54 +0800 Subject: ACPI: fixes a false alarm from lockdep fixes a false alarm from lockdep, as acpi hotplug workqueue waits other workqueues. http://bugzilla.kernel.org/show_bug.cgi?id=14553 https://bugzilla.kernel.org/show_bug.cgi?id=15521 Original-patch-from: Andrew Morton Signed-off-by: Shaohua Li Signed-off-by: Zhang Rui Signed-off-by: Len Brown diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c index 8e6d866..900da68 100644 --- a/drivers/acpi/osl.c +++ b/drivers/acpi/osl.c @@ -758,7 +758,14 @@ static acpi_status __acpi_os_execute(acpi_execute_type type, queue = hp ? kacpi_hotplug_wq : (type == OSL_NOTIFY_HANDLER ? kacpi_notify_wq : kacpid_wq); dpc->wait = hp ? 1 : 0; - INIT_WORK(&dpc->work, acpi_os_execute_deferred); + + if (queue == kacpi_hotplug_wq) + INIT_WORK(&dpc->work, acpi_os_execute_deferred); + else if (queue == kacpi_notify_wq) + INIT_WORK(&dpc->work, acpi_os_execute_deferred); + else + INIT_WORK(&dpc->work, acpi_os_execute_deferred); + ret = queue_work(queue, &dpc->work); if (!ret) { -- cgit v0.10.2 From bf02bd2590eb78d79ba1033d6df80c778b2f5ddf Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Tue, 16 Mar 2010 23:21:55 +0100 Subject: ACPI / ACPICA: Do not check reference counters in acpi_ev_enable_gpe() acpi_ev_enable_gpe() should enable the GPE at the hardware level regardless of the value of the GPE's runtime reference counter. There are only two callers of acpi_ev_enable_gpe(), acpi_enable_gpe() and acpi_set_gpe(). The first one checks the GPE's runtime reference counter itself and only calls acpi_ev_enable_gpe() if it's equal to one, and the other one is supposed to enable the GPE unconditionally (if called with ACPI_GPE_ENABLE). This change fixes the problem in acpi_enable_wakeup_device() where the GPE will not be enabled for wakeup if it's runtime reference counter is zero, which is a regression from 2.6.33. Signed-off-by: Rafael J. Wysocki Reported-by: Robert Moore Signed-off-by: Len Brown diff --git a/drivers/acpi/acpica/evgpe.c b/drivers/acpi/acpica/evgpe.c index 837de66..78c5550 100644 --- a/drivers/acpi/acpica/evgpe.c +++ b/drivers/acpi/acpica/evgpe.c @@ -117,19 +117,14 @@ acpi_status acpi_ev_enable_gpe(struct acpi_gpe_event_info *gpe_event_info) if (ACPI_FAILURE(status)) return_ACPI_STATUS(status); - /* Mark wake-enabled or HW enable, or both */ - - if (gpe_event_info->runtime_count) { - /* Clear the GPE (of stale events), then enable it */ - status = acpi_hw_clear_gpe(gpe_event_info); - if (ACPI_FAILURE(status)) - return_ACPI_STATUS(status); - - /* Enable the requested runtime GPE */ - status = acpi_hw_write_gpe_enable_reg(gpe_event_info); - } + /* Clear the GPE (of stale events), then enable it */ + status = acpi_hw_clear_gpe(gpe_event_info); + if (ACPI_FAILURE(status)) + return_ACPI_STATUS(status); - return_ACPI_STATUS(AE_OK); + /* Enable the requested GPE */ + status = acpi_hw_write_gpe_enable_reg(gpe_event_info); + return_ACPI_STATUS(status); } /******************************************************************************* -- cgit v0.10.2 From 3089f381fbaf53560dcbcb4aef6ef17fe44e347c Mon Sep 17 00:00:00 2001 From: Guennadi Liakhovetski Date: Fri, 19 Mar 2010 13:53:04 +0000 Subject: SH: extend SCI DMA support to work on SCIFA ports SCIFA ports have additional bits to control DMA requests and they must have respective interrupt sources enabled, as the datasheet suggests, the only way to avoid actually taking interrupts in addition to DMA events is by masking the IRQ on the CPU. Signed-off-by: Guennadi Liakhovetski Signed-off-by: Paul Mundt diff --git a/drivers/serial/sh-sci.c b/drivers/serial/sh-sci.c index cd52dd6..304a877 100644 --- a/drivers/serial/sh-sci.c +++ b/drivers/serial/sh-sci.c @@ -106,6 +106,7 @@ struct sci_port { struct work_struct work_tx; struct work_struct work_rx; struct timer_list rx_timer; + unsigned int rx_timeout; #endif }; @@ -673,22 +674,22 @@ static irqreturn_t sci_rx_interrupt(int irq, void *ptr) struct sci_port *s = to_sci_port(port); if (s->chan_rx) { - unsigned long tout; u16 scr = sci_in(port, SCSCR); u16 ssr = sci_in(port, SCxSR); /* Disable future Rx interrupts */ - sci_out(port, SCSCR, scr & ~SCI_CTRL_FLAGS_RIE); + if (port->type == PORT_SCIFA) { + disable_irq_nosync(irq); + scr |= 0x4000; + } else { + scr &= ~SCI_CTRL_FLAGS_RIE; + } + sci_out(port, SCSCR, scr); /* Clear current interrupt */ sci_out(port, SCxSR, ssr & ~(1 | SCxSR_RDxF(port))); - /* Calculate delay for 1.5 DMA buffers */ - tout = (port->timeout - HZ / 50) * s->buf_len_rx * 3 / - port->fifosize / 2; - dev_dbg(port->dev, "Rx IRQ: setup timeout in %lu ms\n", - tout * 1000 / HZ); - if (tout < 2) - tout = 2; - mod_timer(&s->rx_timer, jiffies + tout); + dev_dbg(port->dev, "Rx IRQ %lu: setup t-out in %u jiffies\n", + jiffies, s->rx_timeout); + mod_timer(&s->rx_timer, jiffies + s->rx_timeout); return IRQ_HANDLED; } @@ -925,13 +926,17 @@ static void sci_dma_tx_complete(void *arg) s->cookie_tx = -EINVAL; s->desc_tx = NULL; - spin_unlock_irqrestore(&port->lock, flags); - if (uart_circ_chars_pending(xmit) < WAKEUP_CHARS) uart_write_wakeup(port); - if (uart_circ_chars_pending(xmit)) + if (!uart_circ_empty(xmit)) { schedule_work(&s->work_tx); + } else if (port->type == PORT_SCIFA) { + u16 ctrl = sci_in(port, SCSCR); + sci_out(port, SCSCR, ctrl & ~SCI_CTRL_FLAGS_TIE); + } + + spin_unlock_irqrestore(&port->lock, flags); } /* Locking: called with port lock held */ @@ -975,13 +980,13 @@ static void sci_dma_rx_complete(void *arg) unsigned long flags; int count; - dev_dbg(port->dev, "%s(%d)\n", __func__, port->line); + dev_dbg(port->dev, "%s(%d) active #%d\n", __func__, port->line, s->active_rx); spin_lock_irqsave(&port->lock, flags); count = sci_dma_rx_push(s, tty, s->buf_len_rx); - mod_timer(&s->rx_timer, jiffies + msecs_to_jiffies(5)); + mod_timer(&s->rx_timer, jiffies + s->rx_timeout); spin_unlock_irqrestore(&port->lock, flags); @@ -1053,6 +1058,8 @@ static void sci_submit_rx(struct sci_port *s) sci_rx_dma_release(s, true); return; } + dev_dbg(s->port.dev, "%s(): cookie %d to #%d\n", __func__, + s->cookie_rx[i], i); } s->active_rx = s->cookie_rx[0]; @@ -1110,10 +1117,10 @@ static void work_fn_rx(struct work_struct *work) return; } - dev_dbg(port->dev, "%s: cookie %d #%d\n", __func__, - s->cookie_rx[new], new); - s->active_rx = s->cookie_rx[!new]; + + dev_dbg(port->dev, "%s: cookie %d #%d, new active #%d\n", __func__, + s->cookie_rx[new], new, s->active_rx); } static void work_fn_tx(struct work_struct *work) @@ -1175,23 +1182,28 @@ static void work_fn_tx(struct work_struct *work) static void sci_start_tx(struct uart_port *port) { + struct sci_port *s = to_sci_port(port); unsigned short ctrl; #ifdef CONFIG_SERIAL_SH_SCI_DMA - struct sci_port *s = to_sci_port(port); - - if (s->chan_tx) { - if (!uart_circ_empty(&s->port.state->xmit) && s->cookie_tx < 0) - schedule_work(&s->work_tx); - - return; + if (port->type == PORT_SCIFA) { + u16 new, scr = sci_in(port, SCSCR); + if (s->chan_tx) + new = scr | 0x8000; + else + new = scr & ~0x8000; + if (new != scr) + sci_out(port, SCSCR, new); } + if (s->chan_tx && !uart_circ_empty(&s->port.state->xmit) && + s->cookie_tx < 0) + schedule_work(&s->work_tx); #endif - - /* Set TIE (Transmit Interrupt Enable) bit in SCSCR */ - ctrl = sci_in(port, SCSCR); - ctrl |= SCI_CTRL_FLAGS_TIE; - sci_out(port, SCSCR, ctrl); + if (!s->chan_tx || port->type == PORT_SCIFA) { + /* Set TIE (Transmit Interrupt Enable) bit in SCSCR */ + ctrl = sci_in(port, SCSCR); + sci_out(port, SCSCR, ctrl | SCI_CTRL_FLAGS_TIE); + } } static void sci_stop_tx(struct uart_port *port) @@ -1200,6 +1212,8 @@ static void sci_stop_tx(struct uart_port *port) /* Clear TIE (Transmit Interrupt Enable) bit in SCSCR */ ctrl = sci_in(port, SCSCR); + if (port->type == PORT_SCIFA) + ctrl &= ~0x8000; ctrl &= ~SCI_CTRL_FLAGS_TIE; sci_out(port, SCSCR, ctrl); } @@ -1210,6 +1224,8 @@ static void sci_start_rx(struct uart_port *port) /* Set RIE (Receive Interrupt Enable) bit in SCSCR */ ctrl |= sci_in(port, SCSCR); + if (port->type == PORT_SCIFA) + ctrl &= ~0x4000; sci_out(port, SCSCR, ctrl); } @@ -1219,6 +1235,8 @@ static void sci_stop_rx(struct uart_port *port) /* Clear RIE (Receive Interrupt Enable) bit in SCSCR */ ctrl = sci_in(port, SCSCR); + if (port->type == PORT_SCIFA) + ctrl &= ~0x4000; ctrl &= ~(SCI_CTRL_FLAGS_RIE | SCI_CTRL_FLAGS_REIE); sci_out(port, SCSCR, ctrl); } @@ -1253,8 +1271,12 @@ static void rx_timer_fn(unsigned long arg) { struct sci_port *s = (struct sci_port *)arg; struct uart_port *port = &s->port; - u16 scr = sci_in(port, SCSCR); + + if (port->type == PORT_SCIFA) { + scr &= ~0x4000; + enable_irq(s->irqs[1]); + } sci_out(port, SCSCR, scr | SCI_CTRL_FLAGS_RIE); dev_dbg(port->dev, "DMA Rx timed out\n"); schedule_work(&s->work_rx); @@ -1404,8 +1426,12 @@ static void sci_shutdown(struct uart_port *port) static void sci_set_termios(struct uart_port *port, struct ktermios *termios, struct ktermios *old) { +#ifdef CONFIG_SERIAL_SH_SCI_DMA + struct sci_port *s = to_sci_port(port); +#endif unsigned int status, baud, smr_val, max_baud; int t = -1; + u16 scfcr = 0; /* * earlyprintk comes here early on with port->uartclk set to zero. @@ -1428,7 +1454,7 @@ static void sci_set_termios(struct uart_port *port, struct ktermios *termios, sci_out(port, SCSCR, 0x00); /* TE=0, RE=0, CKE1=0 */ if (port->type != PORT_SCI) - sci_out(port, SCFCR, SCFCR_RFRST | SCFCR_TFRST); + sci_out(port, SCFCR, scfcr | SCFCR_RFRST | SCFCR_TFRST); smr_val = sci_in(port, SCSMR) & 3; if ((termios->c_cflag & CSIZE) == CS7) @@ -1459,10 +1485,32 @@ static void sci_set_termios(struct uart_port *port, struct ktermios *termios, } sci_init_pins(port, termios->c_cflag); - sci_out(port, SCFCR, (termios->c_cflag & CRTSCTS) ? SCFCR_MCE : 0); + sci_out(port, SCFCR, scfcr | ((termios->c_cflag & CRTSCTS) ? SCFCR_MCE : 0)); sci_out(port, SCSCR, SCSCR_INIT(port)); +#ifdef CONFIG_SERIAL_SH_SCI_DMA + /* + * Calculate delay for 1.5 DMA buffers: see + * drivers/serial/serial_core.c::uart_update_timeout(). With 10 bits + * (CS8), 250Hz, 115200 baud and 64 bytes FIFO, the above function + * calculates 1 jiffie for the data plus 5 jiffies for the "slop(e)." + * Then below we calculate 3 jiffies (12ms) for 1.5 DMA buffers (3 FIFO + * sizes), but it has been found out experimentally, that this is not + * enough: the driver too often needlessly runs on a DMA timeout. 20ms + * as a minimum seem to work perfectly. + */ + if (s->chan_rx) { + s->rx_timeout = (port->timeout - HZ / 50) * s->buf_len_rx * 3 / + port->fifosize / 2; + dev_dbg(port->dev, + "DMA Rx t-out %ums, tty t-out %u jiffies\n", + s->rx_timeout * 1000 / HZ, port->timeout); + if (s->rx_timeout < msecs_to_jiffies(20)) + s->rx_timeout = msecs_to_jiffies(20); + } +#endif + if ((termios->c_cflag & CREAD) != 0) sci_start_rx(port); } -- cgit v0.10.2 From e1292ba164742e3a236e407148e00300b7196906 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Thu, 18 Mar 2010 20:19:27 -0700 Subject: ntp: Make time_adjust static Now that no arches are accessing time_adjust directly, make it static. Signed-off-by: John Stultz LKML-Reference: <1268968769-19209-1-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/include/linux/timex.h b/include/linux/timex.h index 7a082b3..e2de51e 100644 --- a/include/linux/timex.h +++ b/include/linux/timex.h @@ -238,7 +238,6 @@ extern int tickadj; /* amount of adjustment per tick */ * phase-lock loop variables */ extern int time_status; /* clock synchronization status bits */ -extern long time_adjust; /* The amount of adjtime left */ extern void ntp_init(void); extern void ntp_clear(void); diff --git a/kernel/time/ntp.c b/kernel/time/ntp.c index 7c0f180..c631168 100644 --- a/kernel/time/ntp.c +++ b/kernel/time/ntp.c @@ -69,7 +69,7 @@ static s64 time_freq; /* time at last adjustment (secs): */ static long time_reftime; -long time_adjust; +static long time_adjust; /* constant (boot-param configurable) NTP tick adjustment (upscaled) */ static s64 ntp_tick_adj; -- cgit v0.10.2 From 3d0205bd1383aa3cac93c209b7c7d03b27930195 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Thu, 18 Mar 2010 20:19:28 -0700 Subject: ntp: Remove tickadj There are zero users of tickadj. So remove it. Signed-off-by: John Stultz LKML-Reference: <1268968769-19209-2-git-send-email-johnstul@us.ibm.com> Signed-off-by: Thomas Gleixner diff --git a/include/linux/timex.h b/include/linux/timex.h index e2de51e..32d852f 100644 --- a/include/linux/timex.h +++ b/include/linux/timex.h @@ -232,7 +232,6 @@ struct timex { */ extern unsigned long tick_usec; /* USER_HZ period (usec) */ extern unsigned long tick_nsec; /* ACTHZ period (nsec) */ -extern int tickadj; /* amount of adjustment per tick */ /* * phase-lock loop variables @@ -270,9 +269,6 @@ extern void second_overflow(void); extern void update_ntp_one_tick(void); extern int do_adjtimex(struct timex *); -/* Don't use! Compatibility define for existing users. */ -#define tickadj (500/HZ ? : 1) - int read_current_timer(unsigned long *timer_val); /* The clock frequency of the i8253/i8254 PIT */ -- cgit v0.10.2 From 157f1071354db1aed885816094888e0e257c9d0a Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Thu, 11 Feb 2010 07:10:38 -0600 Subject: eCryptfs: Fix metadata in xattr feature regression Fixes regression in 8faece5f906725c10e7a1f6caf84452abadbdc7b When using the ecryptfs_xattr_metadata mount option, eCryptfs stores the metadata (normally stored at the front of the file) in the user.ecryptfs xattr. This causes ecryptfs_crypt_stat.num_header_bytes_at_front to be 0, since there is no header data at the front of the file. This results in too much memory being requested and ENOMEM being returned from ecryptfs_write_metadata(). This patch fixes the problem by using the num_header_bytes_at_front variable for specifying the max size of the metadata, despite whether it is stored in the header or xattr. Reviewed-by: Eric Sandeen Signed-off-by: Tyler Hicks diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index 7cb0a59..c907f6f 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -381,8 +381,8 @@ out: static void ecryptfs_lower_offset_for_extent(loff_t *offset, loff_t extent_num, struct ecryptfs_crypt_stat *crypt_stat) { - (*offset) = (crypt_stat->num_header_bytes_at_front - + (crypt_stat->extent_size * extent_num)); + (*offset) = ecryptfs_lower_header_size(crypt_stat) + + (crypt_stat->extent_size * extent_num); } /** @@ -834,7 +834,8 @@ void ecryptfs_set_default_sizes(struct ecryptfs_crypt_stat *crypt_stat) set_extent_mask_and_shift(crypt_stat); crypt_stat->iv_bytes = ECRYPTFS_DEFAULT_IV_BYTES; if (crypt_stat->flags & ECRYPTFS_METADATA_IN_XATTR) - crypt_stat->num_header_bytes_at_front = 0; + crypt_stat->num_header_bytes_at_front = + ECRYPTFS_MINIMUM_HEADER_EXTENT_SIZE; else { if (PAGE_CACHE_SIZE <= ECRYPTFS_MINIMUM_HEADER_EXTENT_SIZE) crypt_stat->num_header_bytes_at_front = diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index 542f625..8456f70 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -464,6 +464,14 @@ struct ecryptfs_daemon { extern struct mutex ecryptfs_daemon_hash_mux; +static inline size_t +ecryptfs_lower_header_size(struct ecryptfs_crypt_stat *crypt_stat) +{ + if (crypt_stat->flags & ECRYPTFS_METADATA_IN_XATTR) + return 0; + return crypt_stat->num_header_bytes_at_front; +} + static inline struct ecryptfs_file_info * ecryptfs_file_to_private(struct file *file) { diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index 4a430ab..1a73953 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -768,7 +768,7 @@ upper_size_to_lower_size(struct ecryptfs_crypt_stat *crypt_stat, { loff_t lower_size; - lower_size = crypt_stat->num_header_bytes_at_front; + lower_size = ecryptfs_lower_header_size(crypt_stat); if (upper_size != 0) { loff_t num_extents; diff --git a/fs/ecryptfs/mmap.c b/fs/ecryptfs/mmap.c index df4ce99..5a30e01 100644 --- a/fs/ecryptfs/mmap.c +++ b/fs/ecryptfs/mmap.c @@ -97,19 +97,6 @@ out: * (big-endian) * Octet 26: Begin RFC 2440 authentication token packet set */ -static void set_header_info(char *page_virt, - struct ecryptfs_crypt_stat *crypt_stat) -{ - size_t written; - size_t save_num_header_bytes_at_front = - crypt_stat->num_header_bytes_at_front; - - crypt_stat->num_header_bytes_at_front = - ECRYPTFS_MINIMUM_HEADER_EXTENT_SIZE; - ecryptfs_write_header_metadata(page_virt + 20, crypt_stat, &written); - crypt_stat->num_header_bytes_at_front = - save_num_header_bytes_at_front; -} /** * ecryptfs_copy_up_encrypted_with_header @@ -146,9 +133,13 @@ ecryptfs_copy_up_encrypted_with_header(struct page *page, memset(page_virt, 0, PAGE_CACHE_SIZE); /* TODO: Support more than one header extent */ if (view_extent_num == 0) { + size_t written; + rc = ecryptfs_read_xattr_region( page_virt, page->mapping->host); - set_header_info(page_virt, crypt_stat); + ecryptfs_write_header_metadata(page_virt + 20, + crypt_stat, + &written); } kunmap_atomic(page_virt, KM_USER0); flush_dcache_page(page); -- cgit v0.10.2 From fa3ef1cb4e6e9958a9bfaa977c107c515907f102 Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Thu, 11 Feb 2010 05:09:14 -0600 Subject: eCryptfs: Rename ecryptfs_crypt_stat.num_header_bytes_at_front This patch renames the num_header_bytes_at_front variable to metadata_size since it now contains the max size of the metadata. Reviewed-by: Eric Sandeen Signed-off-by: Tyler Hicks diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index c907f6f..391f558 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -834,14 +834,13 @@ void ecryptfs_set_default_sizes(struct ecryptfs_crypt_stat *crypt_stat) set_extent_mask_and_shift(crypt_stat); crypt_stat->iv_bytes = ECRYPTFS_DEFAULT_IV_BYTES; if (crypt_stat->flags & ECRYPTFS_METADATA_IN_XATTR) - crypt_stat->num_header_bytes_at_front = - ECRYPTFS_MINIMUM_HEADER_EXTENT_SIZE; + crypt_stat->metadata_size = ECRYPTFS_MINIMUM_HEADER_EXTENT_SIZE; else { if (PAGE_CACHE_SIZE <= ECRYPTFS_MINIMUM_HEADER_EXTENT_SIZE) - crypt_stat->num_header_bytes_at_front = + crypt_stat->metadata_size = ECRYPTFS_MINIMUM_HEADER_EXTENT_SIZE; else - crypt_stat->num_header_bytes_at_front = PAGE_CACHE_SIZE; + crypt_stat->metadata_size = PAGE_CACHE_SIZE; } } @@ -1238,8 +1237,7 @@ ecryptfs_write_header_metadata(char *virt, header_extent_size = (u32)crypt_stat->extent_size; num_header_extents_at_front = - (u16)(crypt_stat->num_header_bytes_at_front - / crypt_stat->extent_size); + (u16)(crypt_stat->metadata_size / crypt_stat->extent_size); put_unaligned_be32(header_extent_size, virt); virt += 4; put_unaligned_be16(num_header_extents_at_front, virt); @@ -1382,7 +1380,7 @@ int ecryptfs_write_metadata(struct dentry *ecryptfs_dentry) rc = -EINVAL; goto out; } - virt_len = crypt_stat->num_header_bytes_at_front; + virt_len = crypt_stat->metadata_size; order = get_order(virt_len); /* Released in this function */ virt = (char *)ecryptfs_get_zeroed_pages(GFP_KERNEL, order); @@ -1428,16 +1426,15 @@ static int parse_header_metadata(struct ecryptfs_crypt_stat *crypt_stat, header_extent_size = get_unaligned_be32(virt); virt += sizeof(__be32); num_header_extents_at_front = get_unaligned_be16(virt); - crypt_stat->num_header_bytes_at_front = - (((size_t)num_header_extents_at_front - * (size_t)header_extent_size)); + crypt_stat->metadata_size = (((size_t)num_header_extents_at_front + * (size_t)header_extent_size)); (*bytes_read) = (sizeof(__be32) + sizeof(__be16)); if ((validate_header_size == ECRYPTFS_VALIDATE_HEADER_SIZE) - && (crypt_stat->num_header_bytes_at_front + && (crypt_stat->metadata_size < ECRYPTFS_MINIMUM_HEADER_EXTENT_SIZE)) { rc = -EINVAL; printk(KERN_WARNING "Invalid header size: [%zd]\n", - crypt_stat->num_header_bytes_at_front); + crypt_stat->metadata_size); } return rc; } @@ -1452,8 +1449,7 @@ static int parse_header_metadata(struct ecryptfs_crypt_stat *crypt_stat, */ static void set_default_header_data(struct ecryptfs_crypt_stat *crypt_stat) { - crypt_stat->num_header_bytes_at_front = - ECRYPTFS_MINIMUM_HEADER_EXTENT_SIZE; + crypt_stat->metadata_size = ECRYPTFS_MINIMUM_HEADER_EXTENT_SIZE; } /** diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index 8456f70..d031efd 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -273,7 +273,7 @@ struct ecryptfs_crypt_stat { u32 flags; unsigned int file_version; size_t iv_bytes; - size_t num_header_bytes_at_front; + size_t metadata_size; size_t extent_size; /* Data extent size; default is 4096 */ size_t key_size; size_t extent_shift; @@ -469,7 +469,7 @@ ecryptfs_lower_header_size(struct ecryptfs_crypt_stat *crypt_stat) { if (crypt_stat->flags & ECRYPTFS_METADATA_IN_XATTR) return 0; - return crypt_stat->num_header_bytes_at_front; + return crypt_stat->metadata_size; } static inline struct ecryptfs_file_info * diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index 1a73953..a50efb1 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -335,7 +335,7 @@ int ecryptfs_lookup_and_interpose_lower(struct dentry *ecryptfs_dentry, ecryptfs_dentry->d_sb)->mount_crypt_stat; if (mount_crypt_stat->flags & ECRYPTFS_ENCRYPTED_VIEW_ENABLED) { if (crypt_stat->flags & ECRYPTFS_METADATA_IN_XATTR) - file_size = (crypt_stat->num_header_bytes_at_front + file_size = (crypt_stat->metadata_size + i_size_read(lower_dentry->d_inode)); else file_size = i_size_read(lower_dentry->d_inode); diff --git a/fs/ecryptfs/mmap.c b/fs/ecryptfs/mmap.c index 5a30e01..270f42a 100644 --- a/fs/ecryptfs/mmap.c +++ b/fs/ecryptfs/mmap.c @@ -122,8 +122,7 @@ ecryptfs_copy_up_encrypted_with_header(struct page *page, * num_extents_per_page) + extent_num_in_page); size_t num_header_extents_at_front = - (crypt_stat->num_header_bytes_at_front - / crypt_stat->extent_size); + (crypt_stat->metadata_size / crypt_stat->extent_size); if (view_extent_num < num_header_extents_at_front) { /* This is a header extent */ @@ -152,7 +151,7 @@ ecryptfs_copy_up_encrypted_with_header(struct page *page, /* This is an encrypted data extent */ loff_t lower_offset = ((view_extent_num * crypt_stat->extent_size) - - crypt_stat->num_header_bytes_at_front); + - crypt_stat->metadata_size); rc = ecryptfs_read_lower_page_segment( page, (lower_offset >> PAGE_CACHE_SHIFT), -- cgit v0.10.2 From 1984c23f9e0cdb432d90a85ecf88b424d36878fc Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Wed, 10 Feb 2010 23:17:44 -0600 Subject: eCryptfs: Clear buffer before reading in metadata xattr We initially read in the first PAGE_CACHE_SIZE of a file to if the eCryptfs header marker can be found. If it isn't found and ecryptfs_xattr_metadata was given as a mount option, then the user.ecryptfs xattr is read into the same buffer. Since the data from the first page of the file wasn't cleared, it is possible that we think we've found a second tag 3 or tag 1 packet and then error out after the packet contents aren't as expected. This patch clears the buffer before filling it with metadata from the user.ecryptfs xattr. Reviewed-by: Eric Sandeen Signed-off-by: Tyler Hicks diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index 391f558..4d9db0e 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -1603,6 +1603,7 @@ int ecryptfs_read_metadata(struct dentry *ecryptfs_dentry) ecryptfs_dentry, ECRYPTFS_VALIDATE_HEADER_SIZE); if (rc) { + memset(page_virt, 0, PAGE_CACHE_SIZE); rc = ecryptfs_read_xattr_region(page_virt, ecryptfs_inode); if (rc) { printk(KERN_DEBUG "Valid eCryptfs headers not found in " diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index a50efb1..ddbd096 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -323,6 +323,7 @@ int ecryptfs_lookup_and_interpose_lower(struct dentry *ecryptfs_dentry, rc = ecryptfs_read_and_validate_header_region(page_virt, ecryptfs_dentry->d_inode); if (rc) { + memset(page_virt, 0, PAGE_CACHE_SIZE); rc = ecryptfs_read_and_validate_xattr_region(page_virt, ecryptfs_dentry); if (rc) { -- cgit v0.10.2 From f4e60e6b303bc46cdc477d3174dbf9cb5dd013aa Mon Sep 17 00:00:00 2001 From: Tyler Hicks Date: Thu, 11 Feb 2010 00:02:32 -0600 Subject: eCryptfs: Strip metadata in xattr flag in encrypted view The ecryptfs_encrypted_view mount option provides a unified way of viewing encrypted eCryptfs files. If the metadata is stored in a xattr, the metadata is moved to the file header when the file is read inside the eCryptfs mount. Because of this, we should strip the ECRYPTFS_METADATA_IN_XATTR flag from the header's flag section. This allows eCryptfs to treat the file as an eCryptfs file with a header at the front. Reviewed-by: Eric Sandeen Signed-off-by: Tyler Hicks diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index 4d9db0e..fad5bf6 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -1107,9 +1107,9 @@ static void write_ecryptfs_marker(char *page_virt, size_t *written) (*written) = MAGIC_ECRYPTFS_MARKER_SIZE_BYTES; } -static void -write_ecryptfs_flags(char *page_virt, struct ecryptfs_crypt_stat *crypt_stat, - size_t *written) +void ecryptfs_write_crypt_stat_flags(char *page_virt, + struct ecryptfs_crypt_stat *crypt_stat, + size_t *written) { u32 flags = 0; int i; @@ -1290,7 +1290,8 @@ static int ecryptfs_write_headers_virt(char *page_virt, size_t max, offset = ECRYPTFS_FILE_SIZE_BYTES; write_ecryptfs_marker((page_virt + offset), &written); offset += written; - write_ecryptfs_flags((page_virt + offset), crypt_stat, &written); + ecryptfs_write_crypt_stat_flags((page_virt + offset), crypt_stat, + &written); offset += written; ecryptfs_write_header_metadata((page_virt + offset), crypt_stat, &written); diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index d031efd..bc71154 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -659,6 +659,9 @@ int ecryptfs_decrypt_page(struct page *page); int ecryptfs_write_metadata(struct dentry *ecryptfs_dentry); int ecryptfs_read_metadata(struct dentry *ecryptfs_dentry); int ecryptfs_new_file_context(struct dentry *ecryptfs_dentry); +void ecryptfs_write_crypt_stat_flags(char *page_virt, + struct ecryptfs_crypt_stat *crypt_stat, + size_t *written); int ecryptfs_read_and_validate_header_region(char *data, struct inode *ecryptfs_inode); int ecryptfs_read_and_validate_xattr_region(char *page_virt, diff --git a/fs/ecryptfs/mmap.c b/fs/ecryptfs/mmap.c index 270f42a..bea998a 100644 --- a/fs/ecryptfs/mmap.c +++ b/fs/ecryptfs/mmap.c @@ -82,6 +82,19 @@ out: return rc; } +static void strip_xattr_flag(char *page_virt, + struct ecryptfs_crypt_stat *crypt_stat) +{ + if (crypt_stat->flags & ECRYPTFS_METADATA_IN_XATTR) { + size_t written; + + crypt_stat->flags &= ~ECRYPTFS_METADATA_IN_XATTR; + ecryptfs_write_crypt_stat_flags(page_virt, crypt_stat, + &written); + crypt_stat->flags |= ECRYPTFS_METADATA_IN_XATTR; + } +} + /** * Header Extent: * Octets 0-7: Unencrypted file size (big-endian) @@ -136,6 +149,7 @@ ecryptfs_copy_up_encrypted_with_header(struct page *page, rc = ecryptfs_read_xattr_region( page_virt, page->mapping->host); + strip_xattr_flag(page_virt + 16, crypt_stat); ecryptfs_write_header_metadata(page_virt + 20, crypt_stat, &written); -- cgit v0.10.2 From 88978e562302c836c1c4597700c79d971e93abc0 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 23 Mar 2010 14:33:58 -0300 Subject: perf archive: Explain how to use the generated tarball MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit [root@doppio ~]# perf archive Now please run: $ tar xvf perf.data.tar.bz2 -C ~/.debug wherever you need to run 'perf report' on. [root@doppio ~]# Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269365638-10223-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/perf-archive.sh b/tools/perf/perf-archive.sh index 910468e..2e7a4f4 100644 --- a/tools/perf/perf-archive.sh +++ b/tools/perf/perf-archive.sh @@ -30,4 +30,7 @@ done tar cfj $PERF_DATA.tar.bz2 -C $DEBUGDIR -T $MANIFEST rm -f $MANIFEST $BUILDIDS +echo -e "Now please run:\n" +echo -e "$ tar xvf $PERF_DATA.tar.bz2 -C ~/.debug\n" +echo "wherever you need to run 'perf report' on." exit 0 -- cgit v0.10.2 From 8a6d9b149f105f8bdfa8e42dd9753e45a1887a16 Mon Sep 17 00:00:00 2001 From: Ferenc Wagner Date: Wed, 24 Mar 2010 08:20:03 +0100 Subject: i2o: Remove the dangerous kobj_to_i2o_device macro This macro worked only when applied to variables named 'kobj'. While this could have been fixed by simply renaming the macro argument, a more type-safe replacement by an inline function would be preferred. However, nobody uses this macro, so it's simpler to just remove it. Signed-off-by: Ferenc Wagner Signed-off-by: Jens Axboe diff --git a/include/linux/i2o.h b/include/linux/i2o.h index 87018dc..9e7a12d 100644 --- a/include/linux/i2o.h +++ b/include/linux/i2o.h @@ -782,7 +782,6 @@ extern int i2o_exec_lct_get(struct i2o_controller *); #define to_i2o_driver(drv) container_of(drv,struct i2o_driver, driver) #define to_i2o_device(dev) container_of(dev, struct i2o_device, device) #define to_i2o_controller(dev) container_of(dev, struct i2o_controller, device) -#define kobj_to_i2o_device(kobj) to_i2o_device(container_of(kobj, struct device, kobj)) /** * i2o_out_to_virt - Turn an I2O message to a virtual address -- cgit v0.10.2 From 5cfc8ee0bb5134e1204e1ec5843e164f24315021 Mon Sep 17 00:00:00 2001 From: John Stultz Date: Wed, 24 Mar 2010 00:22:36 +0000 Subject: ARM: convert arm to arch_gettimeoffset() Convert arm to use GENERIC_TIME via the arch_getoffset() infrastructure, reducing the amount of arch specific code we need to maintain. The arm architecture is the last arch that need to be converted. Signed-off-by: John Stultz Signed-off-by: Russell King Cc: Andrew Morton diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index bcaa8c8..bf856f4 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -42,6 +42,11 @@ config GENERIC_GPIO config GENERIC_TIME bool + default y + +config ARCH_USES_GETTIMEOFFSET + bool + default n config GENERIC_CLOCKEVENTS bool @@ -231,6 +236,7 @@ config ARCH_AAEC2000 select CPU_ARM920T select ARM_AMBA select HAVE_CLK + select ARCH_USES_GETTIMEOFFSET help This enables support for systems based on the Agilent AAEC-2000 @@ -271,6 +277,7 @@ config ARCH_AT91 select GENERIC_GPIO select ARCH_REQUIRE_GPIOLIB select HAVE_CLK + select ARCH_USES_GETTIMEOFFSET help This enables support for systems based on the Atmel AT91RM9200, AT91SAM9 and AT91CAP9 processors. @@ -290,6 +297,7 @@ config ARCH_BCMRING config ARCH_CLPS711X bool "Cirrus Logic CLPS711x/EP721x-based" select CPU_ARM720T + select ARCH_USES_GETTIMEOFFSET help Support for Cirrus Logic 711x/721x based boards. @@ -298,6 +306,7 @@ config ARCH_GEMINI select CPU_FA526 select GENERIC_GPIO select ARCH_REQUIRE_GPIOLIB + select ARCH_USES_GETTIMEOFFSET help Support for the Cortina Systems Gemini family SoCs @@ -306,6 +315,7 @@ config ARCH_EBSA110 select CPU_SA110 select ISA select NO_IOPORT + select ARCH_USES_GETTIMEOFFSET help This is an evaluation board for the StrongARM processor available from Digital. It has limited hardware on-board, including an @@ -321,6 +331,7 @@ config ARCH_EP93XX select COMMON_CLKDEV select ARCH_REQUIRE_GPIOLIB select ARCH_HAS_HOLES_MEMORYMODEL + select ARCH_USES_GETTIMEOFFSET help This enables support for the Cirrus EP93xx series of CPUs. @@ -328,6 +339,7 @@ config ARCH_FOOTBRIDGE bool "FootBridge" select CPU_SA110 select FOOTBRIDGE + select ARCH_USES_GETTIMEOFFSET help Support for systems based on the DC21285 companion chip ("FootBridge"), such as the Simtec CATS and the Rebel NetWinder. @@ -366,6 +378,7 @@ config ARCH_H720X bool "Hynix HMS720x-based" select CPU_ARM720T select ISA_DMA_API + select ARCH_USES_GETTIMEOFFSET help This enables support for systems based on the Hynix HMS720x @@ -408,6 +421,7 @@ config ARCH_IXP23XX depends on MMU select CPU_XSC3 select PCI + select ARCH_USES_GETTIMEOFFSET help Support for Intel's IXP23xx (XScale) family of processors. @@ -416,6 +430,7 @@ config ARCH_IXP2000 depends on MMU select CPU_XSCALE select PCI + select ARCH_USES_GETTIMEOFFSET help Support for Intel's IXP2400/2800 (XScale) family of processors. @@ -434,6 +449,7 @@ config ARCH_L7200 bool "LinkUp-L7200" select CPU_ARM720T select FIQ + select ARCH_USES_GETTIMEOFFSET help Say Y here if you intend to run this kernel on a LinkUp Systems L7200 Software Development Board which uses an ARM720T processor. @@ -523,6 +539,7 @@ config ARCH_KS8695 select CPU_ARM922T select GENERIC_GPIO select ARCH_REQUIRE_GPIOLIB + select ARCH_USES_GETTIMEOFFSET help Support for Micrel/Kendin KS8695 "Centaur" (ARM922T) based System-on-Chip devices. @@ -569,6 +586,7 @@ config ARCH_PNX4008 bool "Philips Nexperia PNX4008 Mobile" select CPU_ARM926T select COMMON_CLKDEV + select ARCH_USES_GETTIMEOFFSET help This enables support for Philips PNX4008 mobile platform. @@ -613,6 +631,7 @@ config ARCH_RPC select ISA_DMA_API select NO_IOPORT select ARCH_SPARSEMEM_ENABLE + select ARCH_USES_GETTIMEOFFSET help On the Acorn Risc-PC, Linux can support the internal IDE disk and CD-ROM interface, serial and parallel port, and the floppy drive. @@ -639,6 +658,7 @@ config ARCH_S3C2410 select GENERIC_GPIO select ARCH_HAS_CPUFREQ select HAVE_CLK + select ARCH_USES_GETTIMEOFFSET help Samsung S3C2410X CPU based systems, such as the Simtec Electronics BAST (), the IPAQ 1940 or @@ -652,6 +672,7 @@ config ARCH_S3C64XX select ARM_VIC select HAVE_CLK select NO_IOPORT + select ARCH_USES_GETTIMEOFFSET select ARCH_HAS_CPUFREQ select ARCH_REQUIRE_GPIOLIB select SAMSUNG_CLKSRC @@ -708,6 +729,7 @@ config ARCH_SHARK select ISA_DMA select ZONE_DMA select PCI + select ARCH_USES_GETTIMEOFFSET help Support for the StrongARM based Digital DNARD machine, also known as "Shark" (). @@ -717,6 +739,7 @@ config ARCH_LH7A40X select CPU_ARM922T select ARCH_DISCONTIGMEM_ENABLE if !LH7A40X_CONTIGMEM select ARCH_SPARSEMEM_ENABLE if !LH7A40X_CONTIGMEM + select ARCH_USES_GETTIMEOFFSET help Say Y here for systems based on one of the Sharp LH7A40X System on a Chip processors. These CPUs include an ARM922T diff --git a/arch/arm/include/asm/mach/time.h b/arch/arm/include/asm/mach/time.h index 8bffc3f..35d408f 100644 --- a/arch/arm/include/asm/mach/time.h +++ b/arch/arm/include/asm/mach/time.h @@ -38,7 +38,7 @@ struct sys_timer { void (*init)(void); void (*suspend)(void); void (*resume)(void); -#ifndef CONFIG_GENERIC_TIME +#ifdef CONFIG_ARCH_USES_GETTIMEOFFSET unsigned long (*offset)(void); #endif }; diff --git a/arch/arm/kernel/time.c b/arch/arm/kernel/time.c index 2875380..38c261f 100644 --- a/arch/arm/kernel/time.c +++ b/arch/arm/kernel/time.c @@ -72,12 +72,15 @@ unsigned long profile_pc(struct pt_regs *regs) EXPORT_SYMBOL(profile_pc); #endif -#ifndef CONFIG_GENERIC_TIME -static unsigned long dummy_gettimeoffset(void) +#ifdef CONFIG_ARCH_USES_GETTIMEOFFSET +u32 arch_gettimeoffset(void) { + if (system_timer->offset != NULL) + return system_timer->offset() * 1000; + return 0; } -#endif +#endif /* CONFIG_ARCH_USES_GETTIMEOFFSET */ #ifdef CONFIG_LEDS_TIMER static inline void do_leds(void) @@ -93,63 +96,6 @@ static inline void do_leds(void) #define do_leds() #endif -#ifndef CONFIG_GENERIC_TIME -void do_gettimeofday(struct timeval *tv) -{ - unsigned long flags; - unsigned long seq; - unsigned long usec, sec; - - do { - seq = read_seqbegin_irqsave(&xtime_lock, flags); - usec = system_timer->offset(); - sec = xtime.tv_sec; - usec += xtime.tv_nsec / 1000; - } while (read_seqretry_irqrestore(&xtime_lock, seq, flags)); - - /* usec may have gone up a lot: be safe */ - while (usec >= 1000000) { - usec -= 1000000; - sec++; - } - - tv->tv_sec = sec; - tv->tv_usec = usec; -} - -EXPORT_SYMBOL(do_gettimeofday); - -int do_settimeofday(struct timespec *tv) -{ - time_t wtm_sec, sec = tv->tv_sec; - long wtm_nsec, nsec = tv->tv_nsec; - - if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC) - return -EINVAL; - - write_seqlock_irq(&xtime_lock); - /* - * This is revolting. We need to set "xtime" correctly. However, the - * value in this location is the value at the most recent update of - * wall time. Discover what correction gettimeofday() would have - * done, and then undo it! - */ - nsec -= system_timer->offset() * NSEC_PER_USEC; - - wtm_sec = wall_to_monotonic.tv_sec + (xtime.tv_sec - sec); - wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - nsec); - - set_normalized_timespec(&xtime, sec, nsec); - set_normalized_timespec(&wall_to_monotonic, wtm_sec, wtm_nsec); - - ntp_clear(); - write_sequnlock_irq(&xtime_lock); - clock_was_set(); - return 0; -} - -EXPORT_SYMBOL(do_settimeofday); -#endif /* !CONFIG_GENERIC_TIME */ #ifndef CONFIG_GENERIC_CLOCKEVENTS /* @@ -214,10 +160,6 @@ device_initcall(timer_init_sysfs); void __init time_init(void) { -#ifndef CONFIG_GENERIC_TIME - if (system_timer->offset == NULL) - system_timer->offset = dummy_gettimeoffset; -#endif system_timer->init(); } diff --git a/arch/arm/mach-at91/Kconfig b/arch/arm/mach-at91/Kconfig index 2db43a5..8a5e8e9 100644 --- a/arch/arm/mach-at91/Kconfig +++ b/arch/arm/mach-at91/Kconfig @@ -97,6 +97,7 @@ config ARCH_AT572D940HF config ARCH_AT91X40 bool "AT91x40" + select ARCH_USES_GETTIMEOFFSET endchoice -- cgit v0.10.2 From 92d6b71ab906be706f3679353b30a8d2c3831144 Mon Sep 17 00:00:00 2001 From: Dimitri Sivanich Date: Thu, 11 Mar 2010 14:08:56 -0800 Subject: genirq: Expose irq_desc->node in proc/irq Expose irq_desc->node as /proc/irq/*/node. This file provides device hardware locality information for apps desiring to include hardware locality in irq mapping decisions. Signed-off-by: Dimitri Sivanich Cc: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner diff --git a/Documentation/filesystems/proc.txt b/Documentation/filesystems/proc.txt index a4f30fa..6507d2a 100644 --- a/Documentation/filesystems/proc.txt +++ b/Documentation/filesystems/proc.txt @@ -566,6 +566,10 @@ The default_smp_affinity mask applies to all non-active IRQs, which are the IRQs which have not yet been allocated/activated, and hence which lack a /proc/irq/[0-9]* directory. +The node file on an SMP system shows the node to which the device using the IRQ +reports itself as being attached. This hardware locality information does not +include information about any possible driver locality preference. + prof_cpu_mask specifies which CPUs are to be profiled by the system wide profiler. Default value is ffffffff (all cpus). diff --git a/kernel/irq/proc.c b/kernel/irq/proc.c index 6f50ecc..e346e08 100644 --- a/kernel/irq/proc.c +++ b/kernel/irq/proc.c @@ -146,6 +146,26 @@ static const struct file_operations default_affinity_proc_fops = { .release = single_release, .write = default_affinity_write, }; + +static int irq_node_proc_show(struct seq_file *m, void *v) +{ + struct irq_desc *desc = irq_to_desc((long) m->private); + + seq_printf(m, "%d\n", desc->node); + return 0; +} + +static int irq_node_proc_open(struct inode *inode, struct file *file) +{ + return single_open(file, irq_node_proc_show, PDE(inode)->data); +} + +static const struct file_operations irq_node_proc_fops = { + .open = irq_node_proc_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; #endif static int irq_spurious_proc_show(struct seq_file *m, void *v) @@ -230,6 +250,9 @@ void register_irq_proc(unsigned int irq, struct irq_desc *desc) /* create /proc/irq//smp_affinity */ proc_create_data("smp_affinity", 0600, desc->dir, &irq_affinity_proc_fops, (void *)(long)irq); + + proc_create_data("node", 0444, desc->dir, + &irq_node_proc_fops, (void *)(long)irq); #endif proc_create_data("spurious", 0444, desc->dir, -- cgit v0.10.2 From a8462bde78fdb77c8ede61e1af99617905a78ccf Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Wed, 24 Mar 2010 14:58:34 +0300 Subject: ASoC: wm8994: playback => capture Sparse caught that initialize "playback" two times instead of initializing "capture". Signed-off-by: Dan Carpenter Acked-by: Liam Girdwood Signed-off-by: Mark Brown diff --git a/sound/soc/codecs/wm8994.c b/sound/soc/codecs/wm8994.c index 29f3771..d10d651 100644 --- a/sound/soc/codecs/wm8994.c +++ b/sound/soc/codecs/wm8994.c @@ -3401,7 +3401,7 @@ struct snd_soc_dai wm8994_dai[] = { .rates = WM8994_RATES, .formats = WM8994_FORMATS, }, - .playback = { + .capture = { .stream_name = "AIF3 Capture", .channels_min = 2, .channels_max = 2, -- cgit v0.10.2 From a5990ea1254cd186b38744507aeec3136a0c1c95 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Thu, 11 Mar 2010 14:08:10 -0800 Subject: sunrpc/cache: fix module refcnt leak in a failure path Don't forget to release the module refcnt if seq_open() returns failure. Signed-off-by: Li Zefan Cc: J. Bruce Fields Cc: Neil Brown Cc: Trond Myklebust Signed-off-by: Andrew Morton Signed-off-by: J. Bruce Fields diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c index 3e1ef8b..a3f340c 100644 --- a/net/sunrpc/cache.c +++ b/net/sunrpc/cache.c @@ -1244,8 +1244,10 @@ static int content_open(struct inode *inode, struct file *file, if (!cd || !try_module_get(cd->owner)) return -EACCES; han = __seq_open_private(file, &cache_content_op, sizeof(*han)); - if (han == NULL) + if (han == NULL) { + module_put(cd->owner); return -ENOMEM; + } han->cd = cd; return 0; -- cgit v0.10.2 From 6dc2d6bf5148a9ed027af21285b0e17b83b3a8f4 Mon Sep 17 00:00:00 2001 From: Vladimir Zapolskiy Date: Wed, 24 Mar 2010 14:32:16 +0300 Subject: imx31: fix parent clock for rtc According to imx31 reference manual the signal from external low frequency clock is sent to RTC clock. The patch makes redundant the previously defined mxc_rtc clock. Signed-off-by: Vladimir Zapolskiy Cc: Sascha Hauer Cc: Daniel Mack Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/clock-imx31.c b/arch/arm/mach-mx3/clock-imx31.c index 80dba99..9a9eb6d 100644 --- a/arch/arm/mach-mx3/clock-imx31.c +++ b/arch/arm/mach-mx3/clock-imx31.c @@ -468,6 +468,7 @@ static struct clk ahb_clk = { } DEFINE_CLOCK(perclk_clk, 0, NULL, 0, NULL, NULL, &ipg_clk); +DEFINE_CLOCK(ckil_clk, 0, NULL, 0, clk_ckil_get_rate, NULL, NULL); DEFINE_CLOCK(sdhc1_clk, 0, MXC_CCM_CGR0, 0, NULL, NULL, &perclk_clk); DEFINE_CLOCK(sdhc2_clk, 1, MXC_CCM_CGR0, 2, NULL, NULL, &perclk_clk); @@ -490,7 +491,7 @@ DEFINE_CLOCK(mpeg4_clk, 0, MXC_CCM_CGR1, 0, NULL, NULL, &ahb_clk); DEFINE_CLOCK(mstick1_clk, 0, MXC_CCM_CGR1, 2, mstick1_get_rate, NULL, &usb_pll_clk); DEFINE_CLOCK(mstick2_clk, 1, MXC_CCM_CGR1, 4, mstick2_get_rate, NULL, &usb_pll_clk); DEFINE_CLOCK1(csi_clk, 0, MXC_CCM_CGR1, 6, csi, NULL, &serial_pll_clk); -DEFINE_CLOCK(rtc_clk, 0, MXC_CCM_CGR1, 8, NULL, NULL, &ipg_clk); +DEFINE_CLOCK(rtc_clk, 0, MXC_CCM_CGR1, 8, NULL, NULL, &ckil_clk); DEFINE_CLOCK(wdog_clk, 0, MXC_CCM_CGR1, 10, NULL, NULL, &ipg_clk); DEFINE_CLOCK(pwm_clk, 0, MXC_CCM_CGR1, 12, NULL, NULL, &perclk_clk); DEFINE_CLOCK(usb_clk2, 0, MXC_CCM_CGR1, 18, usb_get_rate, NULL, &ahb_clk); @@ -514,7 +515,6 @@ DEFINE_CLOCK(usb_clk1, 0, NULL, 0, usb_get_rate, NULL, &usb_pll_clk) DEFINE_CLOCK(nfc_clk, 0, NULL, 0, nfc_get_rate, NULL, &ahb_clk); DEFINE_CLOCK(scc_clk, 0, NULL, 0, NULL, NULL, &ipg_clk); DEFINE_CLOCK(ipg_clk, 0, NULL, 0, ipg_get_rate, NULL, &ahb_clk); -DEFINE_CLOCK(ckil_clk, 0, NULL, 0, clk_ckil_get_rate, NULL, NULL); #define _REGISTER_CLOCK(d, n, c) \ { \ @@ -572,7 +572,6 @@ static struct clk_lookup lookups[] = { _REGISTER_CLOCK(NULL, "iim", iim_clk) _REGISTER_CLOCK(NULL, "mpeg4", mpeg4_clk) _REGISTER_CLOCK(NULL, "mbx", mbx_clk) - _REGISTER_CLOCK("mxc_rtc", NULL, ckil_clk) }; int __init mx31_clocks_init(unsigned long fref) -- cgit v0.10.2 From 11a332adfeacee2634f73ad21b55aad5cc7fd54a Mon Sep 17 00:00:00 2001 From: Alberto Panizzo Date: Tue, 23 Mar 2010 19:46:57 +0100 Subject: MXC: mach-mx31_3ds: Update variable names over recent mach name modification. Signed-off-by: Alberto Panizzo Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/mach-mx31_3ds.c b/arch/arm/mach-mx3/mach-mx31_3ds.c index b88c18a..4a94c3c 100644 --- a/arch/arm/mach-mx3/mach-mx31_3ds.c +++ b/arch/arm/mach-mx3/mach-mx31_3ds.c @@ -31,20 +31,20 @@ #include #include #include -#include +#include #include #include #include "devices.h" /*! - * @file mx31pdk.c + * @file mx31_3ds.c * * @brief This file contains the board-specific initialization routines. * * @ingroup System */ -static int mx31pdk_pins[] = { +static int mx31_3ds_pins[] = { /* UART1 */ MX31_PIN_CTS1__CTS1, MX31_PIN_RTS1__RTS1, @@ -95,7 +95,7 @@ static struct platform_device smsc911x_device = { * LEDs, switches, interrupts for Ethernet. */ -static void mx31pdk_expio_irq_handler(uint32_t irq, struct irq_desc *desc) +static void mx31_3ds_expio_irq_handler(uint32_t irq, struct irq_desc *desc) { uint32_t imr_val; uint32_t int_valid; @@ -163,7 +163,7 @@ static struct irq_chip expio_irq_chip = { .unmask = expio_unmask_irq, }; -static int __init mx31pdk_init_expio(void) +static int __init mx31_3ds_init_expio(void) { int i; int ret; @@ -176,7 +176,7 @@ static int __init mx31pdk_init_expio(void) return -ENODEV; } - pr_info("i.MX31PDK Debug board detected, rev = 0x%04X\n", + pr_info("i.MX31 3DS Debug board detected, rev = 0x%04X\n", __raw_readw(CPLD_CODE_VER_REG)); /* @@ -201,7 +201,7 @@ static int __init mx31pdk_init_expio(void) set_irq_flags(i, IRQF_VALID); } set_irq_type(EXPIO_PARENT_INT, IRQ_TYPE_LEVEL_LOW); - set_irq_chained_handler(EXPIO_PARENT_INT, mx31pdk_expio_irq_handler); + set_irq_chained_handler(EXPIO_PARENT_INT, mx31_3ds_expio_irq_handler); return 0; } @@ -209,7 +209,7 @@ static int __init mx31pdk_init_expio(void) /* * This structure defines the MX31 memory map. */ -static struct map_desc mx31pdk_io_desc[] __initdata = { +static struct map_desc mx31_3ds_io_desc[] __initdata = { { .virtual = MX31_CS5_BASE_ADDR_VIRT, .pfn = __phys_to_pfn(MX31_CS5_BASE_ADDR), @@ -221,10 +221,10 @@ static struct map_desc mx31pdk_io_desc[] __initdata = { /* * Set up static virtual mappings. */ -static void __init mx31pdk_map_io(void) +static void __init mx31_3ds_map_io(void) { mx31_map_io(); - iotable_init(mx31pdk_io_desc, ARRAY_SIZE(mx31pdk_io_desc)); + iotable_init(mx31_3ds_io_desc, ARRAY_SIZE(mx31_3ds_io_desc)); } /*! @@ -232,35 +232,35 @@ static void __init mx31pdk_map_io(void) */ static void __init mxc_board_init(void) { - mxc_iomux_setup_multiple_pins(mx31pdk_pins, ARRAY_SIZE(mx31pdk_pins), - "mx31pdk"); + mxc_iomux_setup_multiple_pins(mx31_3ds_pins, ARRAY_SIZE(mx31_3ds_pins), + "mx31_3ds"); mxc_register_device(&mxc_uart_device0, &uart_pdata); - if (!mx31pdk_init_expio()) + if (!mx31_3ds_init_expio()) platform_device_register(&smsc911x_device); } -static void __init mx31pdk_timer_init(void) +static void __init mx31_3ds_timer_init(void) { mx31_clocks_init(26000000); } -static struct sys_timer mx31pdk_timer = { - .init = mx31pdk_timer_init, +static struct sys_timer mx31_3ds_timer = { + .init = mx31_3ds_timer_init, }; /* * The following uses standard kernel macros defined in arch.h in order to - * initialize __mach_desc_MX31PDK data structure. + * initialize __mach_desc_MX31_3DS data structure. */ MACHINE_START(MX31_3DS, "Freescale MX31PDK (3DS)") /* Maintainer: Freescale Semiconductor, Inc. */ .phys_io = MX31_AIPS1_BASE_ADDR, .io_pg_offst = (MX31_AIPS1_BASE_ADDR_VIRT >> 18) & 0xfffc, .boot_params = MX3x_PHYS_OFFSET + 0x100, - .map_io = mx31pdk_map_io, + .map_io = mx31_3ds_map_io, .init_irq = mx31_init_irq, .init_machine = mxc_board_init, - .timer = &mx31pdk_timer, + .timer = &mx31_3ds_timer, MACHINE_END diff --git a/arch/arm/plat-mxc/include/mach/board-mx31_3ds.h b/arch/arm/plat-mxc/include/mach/board-mx31_3ds.h new file mode 100644 index 0000000..da92933 --- /dev/null +++ b/arch/arm/plat-mxc/include/mach/board-mx31_3ds.h @@ -0,0 +1,59 @@ +/* + * Copyright 2008 Freescale Semiconductor, Inc. All Rights Reserved. + */ + +/* + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef __ASM_ARCH_MXC_BOARD_MX31_3DS_H__ +#define __ASM_ARCH_MXC_BOARD_MX31_3DS_H__ + +/* Definitions for components on the Debug board */ + +/* Base address of CPLD controller on the Debug board */ +#define DEBUG_BASE_ADDRESS CS5_IO_ADDRESS(CS5_BASE_ADDR) + +/* LAN9217 ethernet base address */ +#define LAN9217_BASE_ADDR CS5_BASE_ADDR + +/* CPLD config and interrupt base address */ +#define CPLD_ADDR (DEBUG_BASE_ADDRESS + 0x20000) + +/* LED switchs */ +#define CPLD_LED_REG (CPLD_ADDR + 0x00) +/* buttons */ +#define CPLD_SWITCH_BUTTONS_REG (EXPIO_ADDR + 0x08) +/* status, interrupt */ +#define CPLD_INT_STATUS_REG (CPLD_ADDR + 0x10) +#define CPLD_INT_MASK_REG (CPLD_ADDR + 0x38) +#define CPLD_INT_RESET_REG (CPLD_ADDR + 0x20) +/* magic word for debug CPLD */ +#define CPLD_MAGIC_NUMBER1_REG (CPLD_ADDR + 0x40) +#define CPLD_MAGIC_NUMBER2_REG (CPLD_ADDR + 0x48) +/* CPLD code version */ +#define CPLD_CODE_VER_REG (CPLD_ADDR + 0x50) +/* magic word for debug CPLD */ +#define CPLD_MAGIC_NUMBER3_REG (CPLD_ADDR + 0x58) +/* module reset register */ +#define CPLD_MODULE_RESET_REG (CPLD_ADDR + 0x60) +/* CPU ID and Personality ID */ +#define CPLD_MCU_BOARD_ID_REG (CPLD_ADDR + 0x68) + +/* CPLD IRQ line for external uart, external ethernet etc */ +#define EXPIO_PARENT_INT IOMUX_TO_IRQ(MX31_PIN_GPIO1_1) + +#define MXC_EXP_IO_BASE (MXC_BOARD_IRQ_START) +#define MXC_IRQ_TO_EXPIO(irq) ((irq) - MXC_EXP_IO_BASE) + +#define EXPIO_INT_ENET (MXC_EXP_IO_BASE + 0) +#define EXPIO_INT_XUART_A (MXC_EXP_IO_BASE + 1) +#define EXPIO_INT_XUART_B (MXC_EXP_IO_BASE + 2) +#define EXPIO_INT_BUTTON_A (MXC_EXP_IO_BASE + 3) +#define EXPIO_INT_BUTTON_B (MXC_EXP_IO_BASE + 4) + +#define MXC_MAX_EXP_IO_LINES 16 + +#endif /* __ASM_ARCH_MXC_BOARD_MX31_3DS_H__ */ diff --git a/arch/arm/plat-mxc/include/mach/board-mx31pdk.h b/arch/arm/plat-mxc/include/mach/board-mx31pdk.h deleted file mode 100644 index 2bbd6ed..0000000 --- a/arch/arm/plat-mxc/include/mach/board-mx31pdk.h +++ /dev/null @@ -1,59 +0,0 @@ -/* - * Copyright 2008 Freescale Semiconductor, Inc. All Rights Reserved. - */ - -/* - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - */ - -#ifndef __ASM_ARCH_MXC_BOARD_MX31PDK_H__ -#define __ASM_ARCH_MXC_BOARD_MX31PDK_H__ - -/* Definitions for components on the Debug board */ - -/* Base address of CPLD controller on the Debug board */ -#define DEBUG_BASE_ADDRESS CS5_IO_ADDRESS(CS5_BASE_ADDR) - -/* LAN9217 ethernet base address */ -#define LAN9217_BASE_ADDR CS5_BASE_ADDR - -/* CPLD config and interrupt base address */ -#define CPLD_ADDR (DEBUG_BASE_ADDRESS + 0x20000) - -/* LED switchs */ -#define CPLD_LED_REG (CPLD_ADDR + 0x00) -/* buttons */ -#define CPLD_SWITCH_BUTTONS_REG (EXPIO_ADDR + 0x08) -/* status, interrupt */ -#define CPLD_INT_STATUS_REG (CPLD_ADDR + 0x10) -#define CPLD_INT_MASK_REG (CPLD_ADDR + 0x38) -#define CPLD_INT_RESET_REG (CPLD_ADDR + 0x20) -/* magic word for debug CPLD */ -#define CPLD_MAGIC_NUMBER1_REG (CPLD_ADDR + 0x40) -#define CPLD_MAGIC_NUMBER2_REG (CPLD_ADDR + 0x48) -/* CPLD code version */ -#define CPLD_CODE_VER_REG (CPLD_ADDR + 0x50) -/* magic word for debug CPLD */ -#define CPLD_MAGIC_NUMBER3_REG (CPLD_ADDR + 0x58) -/* module reset register */ -#define CPLD_MODULE_RESET_REG (CPLD_ADDR + 0x60) -/* CPU ID and Personality ID */ -#define CPLD_MCU_BOARD_ID_REG (CPLD_ADDR + 0x68) - -/* CPLD IRQ line for external uart, external ethernet etc */ -#define EXPIO_PARENT_INT IOMUX_TO_IRQ(MX31_PIN_GPIO1_1) - -#define MXC_EXP_IO_BASE (MXC_BOARD_IRQ_START) -#define MXC_IRQ_TO_EXPIO(irq) ((irq) - MXC_EXP_IO_BASE) - -#define EXPIO_INT_ENET (MXC_EXP_IO_BASE + 0) -#define EXPIO_INT_XUART_A (MXC_EXP_IO_BASE + 1) -#define EXPIO_INT_XUART_B (MXC_EXP_IO_BASE + 2) -#define EXPIO_INT_BUTTON_A (MXC_EXP_IO_BASE + 3) -#define EXPIO_INT_BUTTON_B (MXC_EXP_IO_BASE + 4) - -#define MXC_MAX_EXP_IO_LINES 16 - -#endif /* __ASM_ARCH_MXC_BOARD_MX31PDK_H__ */ -- cgit v0.10.2 From a1b67b957ea3b9138ed8239796cb48d8b1bf66b0 Mon Sep 17 00:00:00 2001 From: Alberto Panizzo Date: Tue, 23 Mar 2010 19:49:35 +0100 Subject: MXC: mach-mx31_3ds: Add support for on board NAND Flash. Since the using of Bad Block Table is not constantly a good behave I had made it configurable. Signed-off-by: Alberto Panizzo Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/Kconfig b/arch/arm/mach-mx3/Kconfig index 3872af1..9a2911e 100644 --- a/arch/arm/mach-mx3/Kconfig +++ b/arch/arm/mach-mx3/Kconfig @@ -62,6 +62,15 @@ config MACH_MX31_3DS Include support for MX31PDK (3DS) platform. This includes specific configurations for the board and its peripherals. +config MACH_MX31_3DS_MXC_NAND_USE_BBT + bool "Make the MXC NAND driver use the in flash Bad Block Table" + depends on MACH_MX31_3DS + depends on MTD_NAND_MXC + help + Enable this if you want that the MXC NAND driver uses the in flash + Bad Block Table to know what blocks are bad instead of scanning the + entire flash looking for bad block markers. + config MACH_MX31MOBOARD bool "Support mx31moboard platforms (EPFL Mobots group)" select ARCH_MX31 diff --git a/arch/arm/mach-mx3/mach-mx31_3ds.c b/arch/arm/mach-mx3/mach-mx31_3ds.c index 4a94c3c..bf1f54a 100644 --- a/arch/arm/mach-mx3/mach-mx31_3ds.c +++ b/arch/arm/mach-mx3/mach-mx31_3ds.c @@ -34,6 +34,7 @@ #include #include #include +#include #include "devices.h" /*! @@ -53,6 +54,17 @@ static int mx31_3ds_pins[] = { IOMUX_MODE(MX31_PIN_GPIO1_1, IOMUX_CONFIG_GPIO), }; +/* + * NAND Flash + */ +static struct mxc_nand_platform_data imx31_3ds_nand_flash_pdata = { + .width = 1, + .hw_ecc = 1, +#ifdef MACH_MX31_3DS_MXC_NAND_USE_BBT + .flash_bbt = 1, +#endif +}; + static struct imxuart_platform_data uart_pdata = { .flags = IMXUART_HAVE_RTSCTS, }; @@ -236,6 +248,7 @@ static void __init mxc_board_init(void) "mx31_3ds"); mxc_register_device(&mxc_uart_device0, &uart_pdata); + mxc_register_device(&mxc_nand_device, &imx31_3ds_nand_flash_pdata); if (!mx31_3ds_init_expio()) platform_device_register(&smsc911x_device); -- cgit v0.10.2 From a1ac442443342f778d5230b16efadc4c32d96298 Mon Sep 17 00:00:00 2001 From: Alberto Panizzo Date: Tue, 23 Mar 2010 19:50:28 +0100 Subject: MXC: mach-mx31_3ds: Add SPI1 device support. Signed-off-by: Alberto Panizzo Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/mach-mx31_3ds.c b/arch/arm/mach-mx3/mach-mx31_3ds.c index bf1f54a..80179cf 100644 --- a/arch/arm/mach-mx3/mach-mx31_3ds.c +++ b/arch/arm/mach-mx3/mach-mx31_3ds.c @@ -35,6 +35,7 @@ #include #include #include +#include #include "devices.h" /*! @@ -52,6 +53,24 @@ static int mx31_3ds_pins[] = { MX31_PIN_TXD1__TXD1, MX31_PIN_RXD1__RXD1, IOMUX_MODE(MX31_PIN_GPIO1_1, IOMUX_CONFIG_GPIO), + /* SPI 1 */ + MX31_PIN_CSPI2_SCLK__SCLK, + MX31_PIN_CSPI2_MOSI__MOSI, + MX31_PIN_CSPI2_MISO__MISO, + MX31_PIN_CSPI2_SPI_RDY__SPI_RDY, + MX31_PIN_CSPI2_SS0__SS0, + MX31_PIN_CSPI2_SS2__SS2, /*CS for MC13783 */ +}; + +/* SPI */ +static int spi1_internal_chipselect[] = { + MXC_SPI_CS(0), + MXC_SPI_CS(2), +}; + +static struct spi_imx_master spi1_pdata = { + .chipselect = spi1_internal_chipselect, + .num_chipselect = ARRAY_SIZE(spi1_internal_chipselect), }; /* @@ -249,6 +268,7 @@ static void __init mxc_board_init(void) mxc_register_device(&mxc_uart_device0, &uart_pdata); mxc_register_device(&mxc_nand_device, &imx31_3ds_nand_flash_pdata); + mxc_register_device(&mxc_spi_device1, &spi1_pdata); if (!mx31_3ds_init_expio()) platform_device_register(&smsc911x_device); -- cgit v0.10.2 From ae7a3f13ab59784d9d5041d8ecb08477a049e2c6 Mon Sep 17 00:00:00 2001 From: Alberto Panizzo Date: Tue, 23 Mar 2010 19:51:45 +0100 Subject: MXC: mach-mx31_3ds: add support for freescale mc13783 power management device. Power Gates must to be always enabled. Signed-off-by: Alberto Panizzo Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/mach-mx31_3ds.c b/arch/arm/mach-mx3/mach-mx31_3ds.c index 80179cf..f54af1e 100644 --- a/arch/arm/mach-mx3/mach-mx31_3ds.c +++ b/arch/arm/mach-mx3/mach-mx31_3ds.c @@ -23,6 +23,9 @@ #include #include #include +#include +#include +#include #include #include @@ -60,6 +63,33 @@ static int mx31_3ds_pins[] = { MX31_PIN_CSPI2_SPI_RDY__SPI_RDY, MX31_PIN_CSPI2_SS0__SS0, MX31_PIN_CSPI2_SS2__SS2, /*CS for MC13783 */ + /* MC13783 IRQ */ + IOMUX_MODE(MX31_PIN_GPIO1_3, IOMUX_CONFIG_GPIO), +}; + +/* Regulators */ +static struct regulator_init_data pwgtx_init = { + .constraints = { + .boot_on = 1, + .always_on = 1, + }, +}; + +static struct mc13783_regulator_init_data mx31_3ds_regulators[] = { + { + .id = MC13783_REGU_PWGT1SPI, /* Power Gate for ARM core. */ + .init_data = &pwgtx_init, + }, { + .id = MC13783_REGU_PWGT2SPI, /* Power Gate for L2 Cache. */ + .init_data = &pwgtx_init, + }, +}; + +/* MC13783 */ +static struct mc13783_platform_data mc13783_pdata __initdata = { + .regulators = mx31_3ds_regulators, + .num_regulators = ARRAY_SIZE(mx31_3ds_regulators), + .flags = MC13783_USE_REGULATOR, }; /* SPI */ @@ -73,6 +103,18 @@ static struct spi_imx_master spi1_pdata = { .num_chipselect = ARRAY_SIZE(spi1_internal_chipselect), }; +static struct spi_board_info mx31_3ds_spi_devs[] __initdata = { + { + .modalias = "mc13783", + .max_speed_hz = 1000000, + .bus_num = 1, + .chip_select = 1, /* SS2 */ + .platform_data = &mc13783_pdata, + .irq = IOMUX_TO_IRQ(MX31_PIN_GPIO1_3), + .mode = SPI_CS_HIGH, + }, +}; + /* * NAND Flash */ @@ -268,7 +310,10 @@ static void __init mxc_board_init(void) mxc_register_device(&mxc_uart_device0, &uart_pdata); mxc_register_device(&mxc_nand_device, &imx31_3ds_nand_flash_pdata); + mxc_register_device(&mxc_spi_device1, &spi1_pdata); + spi_register_board_info(mx31_3ds_spi_devs, + ARRAY_SIZE(mx31_3ds_spi_devs)); if (!mx31_3ds_init_expio()) platform_device_register(&smsc911x_device); -- cgit v0.10.2 From b1ffe737f5b743115ee46ffb59e338e580c54902 Mon Sep 17 00:00:00 2001 From: Divyesh Shah Date: Thu, 25 Mar 2010 15:45:03 +0100 Subject: cfq-iosched: Add additional blktrace log messages in CFQ for easier debugging These have helped us debug some issues we've noticed in earlier IO controller versions and should be useful now as well. The extra logging covers: - idling behavior. Since there are so many conditions based on which we decide to idle or not, this patch adds a log message for some conditions that we've found useful. - workload slices and current prio and workload type Changelog from v1: o moved log message from cfq_set_active_queue() to __cfq_set_active_queue() o changed queue_count to st->count Signed-off-by: Divyesh Shah Signed-off-by: Jens Axboe diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index 8d5a2f2..2f91c53 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -1518,7 +1518,8 @@ static void __cfq_set_active_queue(struct cfq_data *cfqd, struct cfq_queue *cfqq) { if (cfqq) { - cfq_log_cfqq(cfqd, cfqq, "set_active"); + cfq_log_cfqq(cfqd, cfqq, "set_active wl_prio:%d wl_type:%d", + cfqd->serving_prio, cfqd->serving_type); cfqq->slice_start = 0; cfqq->dispatch_start = jiffies; cfqq->allocated_slice = 0; @@ -1788,7 +1789,11 @@ static bool cfq_should_idle(struct cfq_data *cfqd, struct cfq_queue *cfqq) * Otherwise, we do only if they are the last ones * in their service tree. */ - return service_tree->count == 1 && cfq_cfqq_sync(cfqq); + if (service_tree->count == 1 && cfq_cfqq_sync(cfqq)) + return 1; + cfq_log_cfqq(cfqd, cfqq, "Not idling. st->count:%d", + service_tree->count); + return 0; } static void cfq_arm_slice_timer(struct cfq_data *cfqd) @@ -1833,8 +1838,11 @@ static void cfq_arm_slice_timer(struct cfq_data *cfqd) * time slice. */ if (sample_valid(cic->ttime_samples) && - (cfqq->slice_end - jiffies < cic->ttime_mean)) + (cfqq->slice_end - jiffies < cic->ttime_mean)) { + cfq_log_cfqq(cfqd, cfqq, "Not idling. think_time:%d", + cic->ttime_mean); return; + } cfq_mark_cfqq_wait_request(cfqq); @@ -2042,6 +2050,7 @@ static void choose_service_tree(struct cfq_data *cfqd, struct cfq_group *cfqg) slice = max(slice, 2 * cfqd->cfq_slice_idle); slice = max_t(unsigned, slice, CFQ_MIN_TT); + cfq_log(cfqd, "workload slice:%d", slice); cfqd->workload_expires = jiffies + slice; cfqd->noidle_tree_requires_idle = false; } @@ -3308,6 +3317,7 @@ static void cfq_completed_request(struct request_queue *q, struct request *rq) if (cfq_should_wait_busy(cfqd, cfqq)) { cfqq->slice_end = jiffies + cfqd->cfq_slice_idle; cfq_mark_cfqq_wait_busy(cfqq); + cfq_log_cfqq(cfqd, cfqq, "will busy wait"); } /* -- cgit v0.10.2 From 39c01b219fd30c74869b6fc8749f7900f04e9ef6 Mon Sep 17 00:00:00 2001 From: Divyesh Shah Date: Thu, 25 Mar 2010 15:45:57 +0100 Subject: cfq-iosched: Do not merge queues of BE and IDLE classes Even if they are found to be co-operating. The prio_trees do not have any IDLE cfqqs on them. cfq_close_cooperator() is called from cfq_select_queue() and cfq_completed_request(). The latter ensures that the close cooperator code does not get invoked if the current cfqq is of class IDLE but the former doesn't seem to have any such checks. So an IDLE cfqq may get merged with a BE cfqq from the same group which should be avoided. Signed-off-by: Divyesh Shah Acked-by: Vivek Goyal Signed-off-by: Jens Axboe diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index 2f91c53..2c7a0f4 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -1723,6 +1723,8 @@ static struct cfq_queue *cfq_close_cooperator(struct cfq_data *cfqd, { struct cfq_queue *cfqq; + if (cfq_class_idle(cur_cfqq)) + return NULL; if (!cfq_cfqq_sync(cur_cfqq)) return NULL; if (CFQQ_SEEKY(cur_cfqq)) -- cgit v0.10.2 From 8d06a1e1e9c69244f08beb7d17146483f9dcd120 Mon Sep 17 00:00:00 2001 From: Robert Hooker Date: Fri, 19 Mar 2010 15:13:27 -0400 Subject: drm/i915: Disable FBC on 915GM and 945GM. It is causing hangs after a suspend/resume cycle with the default powersave=1 module option on these chipsets since 2.6.32-rc. BugLink: http://bugs.launchpad.net/bugs/492392 Signed-off-by: Robert Hooker Acked-by: Jesse Barnes Signed-off-by: Eric Anholt diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 4b26919..1a39ec7 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -80,14 +80,14 @@ const static struct intel_device_info intel_i915g_info = { .is_i915g = 1, .is_i9xx = 1, .cursor_needs_physical = 1, }; const static struct intel_device_info intel_i915gm_info = { - .is_i9xx = 1, .is_mobile = 1, .has_fbc = 1, + .is_i9xx = 1, .is_mobile = 1, .cursor_needs_physical = 1, }; const static struct intel_device_info intel_i945g_info = { .is_i9xx = 1, .has_hotplug = 1, .cursor_needs_physical = 1, }; const static struct intel_device_info intel_i945gm_info = { - .is_i945gm = 1, .is_i9xx = 1, .is_mobile = 1, .has_fbc = 1, + .is_i945gm = 1, .is_i9xx = 1, .is_mobile = 1, .has_hotplug = 1, .cursor_needs_physical = 1, }; diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 58fc7fa..f021490 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -4814,7 +4814,7 @@ static void intel_init_display(struct drm_device *dev) dev_priv->display.fbc_enabled = g4x_fbc_enabled; dev_priv->display.enable_fbc = g4x_enable_fbc; dev_priv->display.disable_fbc = g4x_disable_fbc; - } else if (IS_I965GM(dev) || IS_I945GM(dev) || IS_I915GM(dev)) { + } else if (IS_I965GM(dev)) { dev_priv->display.fbc_enabled = i8xx_fbc_enabled; dev_priv->display.enable_fbc = i8xx_enable_fbc; dev_priv->display.disable_fbc = i8xx_disable_fbc; -- cgit v0.10.2 From 23010e43b353c2cdc9725cbedc7e364708039bf7 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Mon, 8 Mar 2010 13:35:02 +0100 Subject: drm/i915: introduce to_intel_bo helper This is a purely cosmetic change to make changes in this area easier. And hey, it's not only clearer and typechecked, but actually shorter, too! [anholt: To clarify, this is a change to let us later make drm_i915_gem_object subclass drm_gem_object, instead of having drm_gem_object have a pointer to i915's private data] Signed-off-by: Daniel Vetter Acked-by: Dave Airlie Signed-off-by: Eric Anholt diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index 1376dfe..bb3a4a8 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -225,7 +225,7 @@ static int i915_gem_fence_regs_info(struct seq_file *m, void *data) } else { struct drm_i915_gem_object *obj_priv; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); seq_printf(m, "Fenced object[%2d] = %p: %s " "%08x %08zx %08x %s %08x %08x %d", i, obj, get_pin_flag(obj_priv), diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 1a39ec7..0af3dcc 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -361,7 +361,7 @@ int i965_reset(struct drm_device *dev, u8 flags) !dev_priv->mm.suspended) { drm_i915_ring_buffer_t *ring = &dev_priv->ring; struct drm_gem_object *obj = ring->ring_obj; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); dev_priv->mm.suspended = 0; /* Stop the ring if it's running. */ diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index aba8260..b7cb4aa 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -731,6 +731,8 @@ struct drm_i915_gem_object { atomic_t pending_flip; }; +#define to_intel_bo(x) ((struct drm_i915_gem_object *) (x)->driver_private) + /** * Request queue structure. * diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 933e865..b85727c 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -162,7 +162,7 @@ fast_shmem_read(struct page **pages, static int i915_gem_object_needs_bit17_swizzle(struct drm_gem_object *obj) { drm_i915_private_t *dev_priv = obj->dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); return dev_priv->mm.bit_6_swizzle_x == I915_BIT_6_SWIZZLE_9_10_17 && obj_priv->tiling_mode != I915_TILING_NONE; @@ -263,7 +263,7 @@ i915_gem_shmem_pread_fast(struct drm_device *dev, struct drm_gem_object *obj, struct drm_i915_gem_pread *args, struct drm_file *file_priv) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); ssize_t remain; loff_t offset, page_base; char __user *user_data; @@ -284,7 +284,7 @@ i915_gem_shmem_pread_fast(struct drm_device *dev, struct drm_gem_object *obj, if (ret != 0) goto fail_put_pages; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); offset = args->offset; while (remain > 0) { @@ -353,7 +353,7 @@ i915_gem_shmem_pread_slow(struct drm_device *dev, struct drm_gem_object *obj, struct drm_i915_gem_pread *args, struct drm_file *file_priv) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); struct mm_struct *mm = current->mm; struct page **user_pages; ssize_t remain; @@ -402,7 +402,7 @@ i915_gem_shmem_pread_slow(struct drm_device *dev, struct drm_gem_object *obj, if (ret != 0) goto fail_put_pages; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); offset = args->offset; while (remain > 0) { @@ -478,7 +478,7 @@ i915_gem_pread_ioctl(struct drm_device *dev, void *data, obj = drm_gem_object_lookup(dev, file_priv, args->handle); if (obj == NULL) return -EBADF; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); /* Bounds check source. * @@ -580,7 +580,7 @@ i915_gem_gtt_pwrite_fast(struct drm_device *dev, struct drm_gem_object *obj, struct drm_i915_gem_pwrite *args, struct drm_file *file_priv) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); drm_i915_private_t *dev_priv = dev->dev_private; ssize_t remain; loff_t offset, page_base; @@ -604,7 +604,7 @@ i915_gem_gtt_pwrite_fast(struct drm_device *dev, struct drm_gem_object *obj, if (ret) goto fail; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); offset = obj_priv->gtt_offset + args->offset; while (remain > 0) { @@ -654,7 +654,7 @@ i915_gem_gtt_pwrite_slow(struct drm_device *dev, struct drm_gem_object *obj, struct drm_i915_gem_pwrite *args, struct drm_file *file_priv) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); drm_i915_private_t *dev_priv = dev->dev_private; ssize_t remain; loff_t gtt_page_base, offset; @@ -698,7 +698,7 @@ i915_gem_gtt_pwrite_slow(struct drm_device *dev, struct drm_gem_object *obj, if (ret) goto out_unpin_object; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); offset = obj_priv->gtt_offset + args->offset; while (remain > 0) { @@ -760,7 +760,7 @@ i915_gem_shmem_pwrite_fast(struct drm_device *dev, struct drm_gem_object *obj, struct drm_i915_gem_pwrite *args, struct drm_file *file_priv) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); ssize_t remain; loff_t offset, page_base; char __user *user_data; @@ -780,7 +780,7 @@ i915_gem_shmem_pwrite_fast(struct drm_device *dev, struct drm_gem_object *obj, if (ret != 0) goto fail_put_pages; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); offset = args->offset; obj_priv->dirty = 1; @@ -828,7 +828,7 @@ i915_gem_shmem_pwrite_slow(struct drm_device *dev, struct drm_gem_object *obj, struct drm_i915_gem_pwrite *args, struct drm_file *file_priv) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); struct mm_struct *mm = current->mm; struct page **user_pages; ssize_t remain; @@ -876,7 +876,7 @@ i915_gem_shmem_pwrite_slow(struct drm_device *dev, struct drm_gem_object *obj, if (ret != 0) goto fail_put_pages; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); offset = args->offset; obj_priv->dirty = 1; @@ -951,7 +951,7 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data, obj = drm_gem_object_lookup(dev, file_priv, args->handle); if (obj == NULL) return -EBADF; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); /* Bounds check destination. * @@ -1033,7 +1033,7 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data, obj = drm_gem_object_lookup(dev, file_priv, args->handle); if (obj == NULL) return -EBADF; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); mutex_lock(&dev->struct_mutex); @@ -1095,7 +1095,7 @@ i915_gem_sw_finish_ioctl(struct drm_device *dev, void *data, DRM_INFO("%s: sw_finish %d (%p %zd)\n", __func__, args->handle, obj, obj->size); #endif - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); /* Pinned buffers may be scanout, so flush the cache */ if (obj_priv->pin_count) @@ -1166,7 +1166,7 @@ int i915_gem_fault(struct vm_area_struct *vma, struct vm_fault *vmf) struct drm_gem_object *obj = vma->vm_private_data; struct drm_device *dev = obj->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); pgoff_t page_offset; unsigned long pfn; int ret = 0; @@ -1233,7 +1233,7 @@ i915_gem_create_mmap_offset(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; struct drm_gem_mm *mm = dev->mm_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); struct drm_map_list *list; struct drm_local_map *map; int ret = 0; @@ -1304,7 +1304,7 @@ void i915_gem_release_mmap(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); if (dev->dev_mapping) unmap_mapping_range(dev->dev_mapping, @@ -1315,7 +1315,7 @@ static void i915_gem_free_mmap_offset(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); struct drm_gem_mm *mm = dev->mm_private; struct drm_map_list *list; @@ -1346,7 +1346,7 @@ static uint32_t i915_gem_get_gtt_alignment(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int start, i; /* @@ -1405,7 +1405,7 @@ i915_gem_mmap_gtt_ioctl(struct drm_device *dev, void *data, mutex_lock(&dev->struct_mutex); - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); if (obj_priv->madv != I915_MADV_WILLNEED) { DRM_ERROR("Attempting to mmap a purgeable buffer\n"); @@ -1449,7 +1449,7 @@ i915_gem_mmap_gtt_ioctl(struct drm_device *dev, void *data, void i915_gem_object_put_pages(struct drm_gem_object *obj) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int page_count = obj->size / PAGE_SIZE; int i; @@ -1485,7 +1485,7 @@ i915_gem_object_move_to_active(struct drm_gem_object *obj, uint32_t seqno) { struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); /* Add a reference if we're newly entering the active list. */ if (!obj_priv->active) { @@ -1505,7 +1505,7 @@ i915_gem_object_move_to_flushing(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); BUG_ON(!obj_priv->active); list_move_tail(&obj_priv->list, &dev_priv->mm.flushing_list); @@ -1516,7 +1516,7 @@ i915_gem_object_move_to_flushing(struct drm_gem_object *obj) static void i915_gem_object_truncate(struct drm_gem_object *obj) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); struct inode *inode; inode = obj->filp->f_path.dentry->d_inode; @@ -1537,7 +1537,7 @@ i915_gem_object_move_to_inactive(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); i915_verify_inactive(dev, __FILE__, __LINE__); if (obj_priv->pin_count != 0) @@ -1964,7 +1964,7 @@ static int i915_gem_object_wait_rendering(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int ret; /* This function only exists to support waiting for existing rendering, @@ -1996,7 +1996,7 @@ i915_gem_object_unbind(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int ret = 0; #if WATCH_BUF @@ -2172,7 +2172,7 @@ i915_gem_evict_something(struct drm_device *dev, int min_size) #if WATCH_LRU DRM_INFO("%s: evicting %p\n", __func__, obj); #endif - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); BUG_ON(obj_priv->pin_count != 0); BUG_ON(obj_priv->active); @@ -2243,7 +2243,7 @@ int i915_gem_object_get_pages(struct drm_gem_object *obj, gfp_t gfpmask) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int page_count, i; struct address_space *mapping; struct inode *inode; @@ -2296,7 +2296,7 @@ static void sandybridge_write_fence_reg(struct drm_i915_fence_reg *reg) struct drm_gem_object *obj = reg->obj; struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int regnum = obj_priv->fence_reg; uint64_t val; @@ -2318,7 +2318,7 @@ static void i965_write_fence_reg(struct drm_i915_fence_reg *reg) struct drm_gem_object *obj = reg->obj; struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int regnum = obj_priv->fence_reg; uint64_t val; @@ -2338,7 +2338,7 @@ static void i915_write_fence_reg(struct drm_i915_fence_reg *reg) struct drm_gem_object *obj = reg->obj; struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int regnum = obj_priv->fence_reg; int tile_width; uint32_t fence_reg, val; @@ -2380,7 +2380,7 @@ static void i830_write_fence_reg(struct drm_i915_fence_reg *reg) struct drm_gem_object *obj = reg->obj; struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int regnum = obj_priv->fence_reg; uint32_t val; uint32_t pitch_val; @@ -2424,7 +2424,7 @@ static int i915_find_fence_reg(struct drm_device *dev) if (!reg->obj) return i; - obj_priv = reg->obj->driver_private; + obj_priv = to_intel_bo(reg->obj); if (!obj_priv->pin_count) avail++; } @@ -2479,7 +2479,7 @@ i915_gem_object_get_fence_reg(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); struct drm_i915_fence_reg *reg = NULL; int ret; @@ -2546,7 +2546,7 @@ i915_gem_clear_fence_reg(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); if (IS_GEN6(dev)) { I915_WRITE64(FENCE_REG_SANDYBRIDGE_0 + @@ -2582,7 +2582,7 @@ int i915_gem_object_put_fence_reg(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); if (obj_priv->fence_reg == I915_FENCE_REG_NONE) return 0; @@ -2620,7 +2620,7 @@ i915_gem_object_bind_to_gtt(struct drm_gem_object *obj, unsigned alignment) { struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); struct drm_mm_node *free_space; gfp_t gfpmask = __GFP_NORETRY | __GFP_NOWARN; int ret; @@ -2727,7 +2727,7 @@ i915_gem_object_bind_to_gtt(struct drm_gem_object *obj, unsigned alignment) void i915_gem_clflush_object(struct drm_gem_object *obj) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); /* If we don't have a page list set up, then we're not pinned * to GPU, and we can ignore the cache flush because it'll happen @@ -2828,7 +2828,7 @@ i915_gem_object_flush_write_domain(struct drm_gem_object *obj) int i915_gem_object_set_to_gtt_domain(struct drm_gem_object *obj, int write) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); uint32_t old_write_domain, old_read_domains; int ret; @@ -2878,7 +2878,7 @@ int i915_gem_object_set_to_display_plane(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); uint32_t old_write_domain, old_read_domains; int ret; @@ -3091,7 +3091,7 @@ static void i915_gem_object_set_to_gpu_domain(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); uint32_t invalidate_domains = 0; uint32_t flush_domains = 0; uint32_t old_read_domains; @@ -3176,7 +3176,7 @@ i915_gem_object_set_to_gpu_domain(struct drm_gem_object *obj) static void i915_gem_object_set_to_full_cpu_read_domain(struct drm_gem_object *obj) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); if (!obj_priv->page_cpu_valid) return; @@ -3216,7 +3216,7 @@ static int i915_gem_object_set_cpu_read_domain_range(struct drm_gem_object *obj, uint64_t offset, uint64_t size) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); uint32_t old_read_domains; int i, ret; @@ -3285,7 +3285,7 @@ i915_gem_object_pin_and_relocate(struct drm_gem_object *obj, { struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int i, ret; void __iomem *reloc_page; bool need_fence; @@ -3336,7 +3336,7 @@ i915_gem_object_pin_and_relocate(struct drm_gem_object *obj, i915_gem_object_unpin(obj); return -EBADF; } - target_obj_priv = target_obj->driver_private; + target_obj_priv = to_intel_bo(target_obj); #if WATCH_RELOC DRM_INFO("%s: obj %p offset %08x target %d " @@ -3688,7 +3688,7 @@ i915_gem_wait_for_pending_flip(struct drm_device *dev, prepare_to_wait(&dev_priv->pending_flip_queue, &wait, TASK_INTERRUPTIBLE); for (i = 0; i < count; i++) { - obj_priv = object_list[i]->driver_private; + obj_priv = to_intel_bo(object_list[i]); if (atomic_read(&obj_priv->pending_flip) > 0) break; } @@ -3797,7 +3797,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data, goto err; } - obj_priv = object_list[i]->driver_private; + obj_priv = to_intel_bo(object_list[i]); if (obj_priv->in_execbuffer) { DRM_ERROR("Object %p appears more than once in object list\n", object_list[i]); @@ -3923,7 +3923,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data, for (i = 0; i < args->buffer_count; i++) { struct drm_gem_object *obj = object_list[i]; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); uint32_t old_write_domain = obj->write_domain; obj->write_domain = obj->pending_write_domain; @@ -3998,7 +3998,7 @@ err: for (i = 0; i < args->buffer_count; i++) { if (object_list[i]) { - obj_priv = object_list[i]->driver_private; + obj_priv = to_intel_bo(object_list[i]); obj_priv->in_execbuffer = false; } drm_gem_object_unreference(object_list[i]); @@ -4176,7 +4176,7 @@ int i915_gem_object_pin(struct drm_gem_object *obj, uint32_t alignment) { struct drm_device *dev = obj->dev; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int ret; i915_verify_inactive(dev, __FILE__, __LINE__); @@ -4209,7 +4209,7 @@ i915_gem_object_unpin(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); i915_verify_inactive(dev, __FILE__, __LINE__); obj_priv->pin_count--; @@ -4249,7 +4249,7 @@ i915_gem_pin_ioctl(struct drm_device *dev, void *data, mutex_unlock(&dev->struct_mutex); return -EBADF; } - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); if (obj_priv->madv != I915_MADV_WILLNEED) { DRM_ERROR("Attempting to pin a purgeable buffer\n"); @@ -4306,7 +4306,7 @@ i915_gem_unpin_ioctl(struct drm_device *dev, void *data, return -EBADF; } - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); if (obj_priv->pin_filp != file_priv) { DRM_ERROR("Not pinned by caller in i915_gem_pin_ioctl(): %d\n", args->handle); @@ -4348,7 +4348,7 @@ i915_gem_busy_ioctl(struct drm_device *dev, void *data, */ i915_gem_retire_requests(dev); - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); /* Don't count being on the flushing list against the object being * done. Otherwise, a buffer left on the flushing list but not getting * flushed (because nobody's flushing that domain) won't ever return @@ -4394,7 +4394,7 @@ i915_gem_madvise_ioctl(struct drm_device *dev, void *data, } mutex_lock(&dev->struct_mutex); - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); if (obj_priv->pin_count) { drm_gem_object_unreference(obj); @@ -4455,7 +4455,7 @@ int i915_gem_init_object(struct drm_gem_object *obj) void i915_gem_free_object(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); trace_i915_gem_object_destroy(obj); @@ -4564,7 +4564,7 @@ i915_gem_init_hws(struct drm_device *dev) DRM_ERROR("Failed to allocate status page\n"); return -ENOMEM; } - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); obj_priv->agp_type = AGP_USER_CACHED_MEMORY; ret = i915_gem_object_pin(obj, 4096); @@ -4608,7 +4608,7 @@ i915_gem_cleanup_hws(struct drm_device *dev) return; obj = dev_priv->hws_obj; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); kunmap(obj_priv->pages[0]); i915_gem_object_unpin(obj); @@ -4642,7 +4642,7 @@ i915_gem_init_ringbuffer(struct drm_device *dev) i915_gem_cleanup_hws(dev); return -ENOMEM; } - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); ret = i915_gem_object_pin(obj, 4096); if (ret != 0) { @@ -4935,7 +4935,7 @@ void i915_gem_detach_phys_object(struct drm_device *dev, int ret; int page_count; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); if (!obj_priv->phys_obj) return; @@ -4974,7 +4974,7 @@ i915_gem_attach_phys_object(struct drm_device *dev, if (id > I915_MAX_PHYS_OBJECT) return -EINVAL; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); if (obj_priv->phys_obj) { if (obj_priv->phys_obj->id == id) @@ -5025,7 +5025,7 @@ i915_gem_phys_pwrite(struct drm_device *dev, struct drm_gem_object *obj, struct drm_i915_gem_pwrite *args, struct drm_file *file_priv) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); void *obj_addr; int ret; char __user *user_data; diff --git a/drivers/gpu/drm/i915/i915_gem_debug.c b/drivers/gpu/drm/i915/i915_gem_debug.c index e602614..35507cf 100644 --- a/drivers/gpu/drm/i915/i915_gem_debug.c +++ b/drivers/gpu/drm/i915/i915_gem_debug.c @@ -72,7 +72,7 @@ void i915_gem_dump_object(struct drm_gem_object *obj, int len, const char *where, uint32_t mark) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int page; DRM_INFO("%s: object at offset %08x\n", where, obj_priv->gtt_offset); @@ -137,7 +137,7 @@ void i915_gem_object_check_coherency(struct drm_gem_object *obj, int handle) { struct drm_device *dev = obj->dev; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int page; uint32_t *gtt_mapping; uint32_t *backing_map = NULL; diff --git a/drivers/gpu/drm/i915/i915_gem_tiling.c b/drivers/gpu/drm/i915/i915_gem_tiling.c index c01c878..449157f 100644 --- a/drivers/gpu/drm/i915/i915_gem_tiling.c +++ b/drivers/gpu/drm/i915/i915_gem_tiling.c @@ -240,7 +240,7 @@ bool i915_gem_object_fence_offset_ok(struct drm_gem_object *obj, int tiling_mode) { struct drm_device *dev = obj->dev; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); if (obj_priv->gtt_space == NULL) return true; @@ -280,7 +280,7 @@ i915_gem_set_tiling(struct drm_device *dev, void *data, obj = drm_gem_object_lookup(dev, file_priv, args->handle); if (obj == NULL) return -EINVAL; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); if (!i915_tiling_ok(dev, args->stride, obj->size, args->tiling_mode)) { drm_gem_object_unreference_unlocked(obj); @@ -364,7 +364,7 @@ i915_gem_get_tiling(struct drm_device *dev, void *data, obj = drm_gem_object_lookup(dev, file_priv, args->handle); if (obj == NULL) return -EINVAL; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); mutex_lock(&dev->struct_mutex); @@ -427,7 +427,7 @@ i915_gem_object_do_bit_17_swizzle(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int page_count = obj->size >> PAGE_SHIFT; int i; @@ -456,7 +456,7 @@ i915_gem_object_save_bit_17_swizzle(struct drm_gem_object *obj) { struct drm_device *dev = obj->dev; drm_i915_private_t *dev_priv = dev->dev_private; - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); int page_count = obj->size >> PAGE_SHIFT; int i; diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 5388354..bfbdad9 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -443,7 +443,7 @@ i915_error_object_create(struct drm_device *dev, if (src == NULL) return NULL; - src_priv = src->driver_private; + src_priv = to_intel_bo(src); if (src_priv->pages == NULL) return NULL; diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index f021490..7adb3a5 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -1002,7 +1002,7 @@ static void i8xx_enable_fbc(struct drm_crtc *crtc, unsigned long interval) struct drm_i915_private *dev_priv = dev->dev_private; struct drm_framebuffer *fb = crtc->fb; struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb); - struct drm_i915_gem_object *obj_priv = intel_fb->obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(intel_fb->obj); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); int plane, i; u32 fbc_ctl, fbc_ctl2; @@ -1079,7 +1079,7 @@ static void g4x_enable_fbc(struct drm_crtc *crtc, unsigned long interval) struct drm_i915_private *dev_priv = dev->dev_private; struct drm_framebuffer *fb = crtc->fb; struct intel_framebuffer *intel_fb = to_intel_framebuffer(fb); - struct drm_i915_gem_object *obj_priv = intel_fb->obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(intel_fb->obj); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); int plane = (intel_crtc->plane == 0 ? DPFC_CTL_PLANEA : DPFC_CTL_PLANEB); @@ -1175,7 +1175,7 @@ static void intel_update_fbc(struct drm_crtc *crtc, return; intel_fb = to_intel_framebuffer(fb); - obj_priv = intel_fb->obj->driver_private; + obj_priv = to_intel_bo(intel_fb->obj); /* * If FBC is already on, we just have to verify that we can @@ -1242,7 +1242,7 @@ out_disable: static int intel_pin_and_fence_fb_obj(struct drm_device *dev, struct drm_gem_object *obj) { - struct drm_i915_gem_object *obj_priv = obj->driver_private; + struct drm_i915_gem_object *obj_priv = to_intel_bo(obj); u32 alignment; int ret; @@ -1322,7 +1322,7 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y, intel_fb = to_intel_framebuffer(crtc->fb); obj = intel_fb->obj; - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); mutex_lock(&dev->struct_mutex); ret = intel_pin_and_fence_fb_obj(dev, obj); @@ -1400,7 +1400,7 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y, if (old_fb) { intel_fb = to_intel_framebuffer(old_fb); - obj_priv = intel_fb->obj->driver_private; + obj_priv = to_intel_bo(intel_fb->obj); i915_gem_object_unpin(intel_fb->obj); } intel_increase_pllclock(crtc, true); @@ -3510,7 +3510,7 @@ static int intel_crtc_cursor_set(struct drm_crtc *crtc, if (!bo) return -ENOENT; - obj_priv = bo->driver_private; + obj_priv = to_intel_bo(bo); if (bo->size < width * height * 4) { DRM_ERROR("buffer is to small\n"); @@ -4155,7 +4155,7 @@ void intel_finish_page_flip(struct drm_device *dev, int pipe) work = intel_crtc->unpin_work; if (work == NULL || !work->pending) { if (work && !work->pending) { - obj_priv = work->pending_flip_obj->driver_private; + obj_priv = to_intel_bo(work->pending_flip_obj); DRM_DEBUG_DRIVER("flip finish: %p (%d) not pending?\n", obj_priv, atomic_read(&obj_priv->pending_flip)); @@ -4180,7 +4180,7 @@ void intel_finish_page_flip(struct drm_device *dev, int pipe) spin_unlock_irqrestore(&dev->event_lock, flags); - obj_priv = work->pending_flip_obj->driver_private; + obj_priv = to_intel_bo(work->pending_flip_obj); /* Initial scanout buffer will have a 0 pending flip count */ if ((atomic_read(&obj_priv->pending_flip) == 0) || @@ -4251,7 +4251,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, ret = intel_pin_and_fence_fb_obj(dev, obj); if (ret != 0) { DRM_DEBUG_DRIVER("flip queue: %p pin & fence failed\n", - obj->driver_private); + to_intel_bo(obj)); kfree(work); intel_crtc->unpin_work = NULL; mutex_unlock(&dev->struct_mutex); @@ -4265,7 +4265,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, crtc->fb = fb; i915_gem_object_flush_write_domain(obj); drm_vblank_get(dev, intel_crtc->pipe); - obj_priv = obj->driver_private; + obj_priv = to_intel_bo(obj); atomic_inc(&obj_priv->pending_flip); work->pending_flip_obj = obj; @@ -4778,14 +4778,14 @@ void intel_init_clock_gating(struct drm_device *dev) struct drm_i915_gem_object *obj_priv = NULL; if (dev_priv->pwrctx) { - obj_priv = dev_priv->pwrctx->driver_private; + obj_priv = to_intel_bo(dev_priv->pwrctx); } else { struct drm_gem_object *pwrctx; pwrctx = intel_alloc_power_context(dev); if (pwrctx) { dev_priv->pwrctx = pwrctx; - obj_priv = pwrctx->driver_private; + obj_priv = to_intel_bo(pwrctx); } } @@ -4956,7 +4956,7 @@ void intel_modeset_cleanup(struct drm_device *dev) if (dev_priv->pwrctx) { struct drm_i915_gem_object *obj_priv; - obj_priv = dev_priv->pwrctx->driver_private; + obj_priv = to_intel_bo(dev_priv->pwrctx); I915_WRITE(PWRCTXA, obj_priv->gtt_offset &~ PWRCTX_EN); I915_READ(PWRCTXA); i915_gem_object_unpin(dev_priv->pwrctx); diff --git a/drivers/gpu/drm/i915/intel_fb.c b/drivers/gpu/drm/i915/intel_fb.c index 8cd791d..c9fbdfa 100644 --- a/drivers/gpu/drm/i915/intel_fb.c +++ b/drivers/gpu/drm/i915/intel_fb.c @@ -145,7 +145,7 @@ static int intelfb_create(struct drm_device *dev, uint32_t fb_width, ret = -ENOMEM; goto out; } - obj_priv = fbo->driver_private; + obj_priv = to_intel_bo(fbo); mutex_lock(&dev->struct_mutex); diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c index 60595fc..6d524a1 100644 --- a/drivers/gpu/drm/i915/intel_overlay.c +++ b/drivers/gpu/drm/i915/intel_overlay.c @@ -724,7 +724,7 @@ int intel_overlay_do_put_image(struct intel_overlay *overlay, int ret, tmp_width; struct overlay_registers *regs; bool scale_changed = false; - struct drm_i915_gem_object *bo_priv = new_bo->driver_private; + struct drm_i915_gem_object *bo_priv = to_intel_bo(new_bo); struct drm_device *dev = overlay->dev; BUG_ON(!mutex_is_locked(&dev->struct_mutex)); @@ -809,7 +809,7 @@ int intel_overlay_do_put_image(struct intel_overlay *overlay, intel_overlay_continue(overlay, scale_changed); overlay->old_vid_bo = overlay->vid_bo; - overlay->vid_bo = new_bo->driver_private; + overlay->vid_bo = to_intel_bo(new_bo); return 0; @@ -1344,7 +1344,7 @@ void intel_setup_overlay(struct drm_device *dev) reg_bo = drm_gem_object_alloc(dev, PAGE_SIZE); if (!reg_bo) goto out_free; - overlay->reg_bo = reg_bo->driver_private; + overlay->reg_bo = to_intel_bo(reg_bo); if (OVERLAY_NONPHYSICAL(dev)) { ret = i915_gem_object_pin(reg_bo, PAGE_SIZE); -- cgit v0.10.2 From 5e64499f3d39c633de49320e399479642c2b1743 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Fri, 19 Mar 2010 21:46:23 +0100 Subject: agp/intel: intel_845_driver is an agp driver! ... not a GTT driver. So the additional chipset flush introduced in commit 2162e6a2b0cd5acbb9bd8a3c94e1c1269b078295 Author: Dave Airlie Date: Wed Nov 21 16:36:31 2007 +1000 agp/intel: Add chipset flushing support for i8xx chipsets. to fix a GTT problem makes absolutely no sense. If this would really be needed for AGP chipsets, too, we should add it to all i8xx agp drivers, not just one. Signed-off-by: Daniel Vetter Signed-off-by: Eric Anholt diff --git a/drivers/char/agp/intel-agp.c b/drivers/char/agp/intel-agp.c index b78d5c3..a34fc9f 100644 --- a/drivers/char/agp/intel-agp.c +++ b/drivers/char/agp/intel-agp.c @@ -1816,8 +1816,6 @@ static int intel_845_configure(void) pci_write_config_byte(agp_bridge->dev, INTEL_I845_AGPM, temp2 | (1 << 1)); /* clear any possible error conditions */ pci_write_config_word(agp_bridge->dev, INTEL_I845_ERRSTS, 0x001c); - - intel_i830_setup_flush(); return 0; } @@ -2187,7 +2185,6 @@ static const struct agp_bridge_driver intel_845_driver = { .agp_destroy_page = agp_generic_destroy_page, .agp_destroy_pages = agp_generic_destroy_pages, .agp_type_to_mask_type = agp_generic_type_to_mask_type, - .chipset_flush = intel_i830_chipset_flush, }; static const struct agp_bridge_driver intel_850_driver = { -- cgit v0.10.2 From 21d40d37eca86872f2bf0af995809ebdef25c9d9 Mon Sep 17 00:00:00 2001 From: Eric Anholt Date: Thu, 25 Mar 2010 11:11:14 -0700 Subject: drm/i915: Rename intel_output to intel_encoder. The intel_output naming is inherited from the UMS code, which had a structure of screen -> CRTC -> output. The DRM code has an additional notion of encoder/connector, so the structure is screen -> CRTC -> encoder -> connector. This is a useful structure for SDVO encoders which can support multiple connectors (each of which requires different programming in the one encoder and could be connected to different CRTCs), or for DVI-I, where multiple encoders feed into the connector for whether it's used for digital or analog. Most of our code is encoder-related, so transition it to talking about encoders before we start trying to distinguish connectors. This patch is produced by sed s/intel_output/intel_encoder/ over the driver. Signed-off-by: Eric Anholt diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index bfbdad9..5b53adf 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -259,10 +259,10 @@ static void i915_hotplug_work_func(struct work_struct *work) if (mode_config->num_connector) { list_for_each_entry(connector, &mode_config->connector_list, head) { - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); - if (intel_output->hot_plug) - (*intel_output->hot_plug) (intel_output); + if (intel_encoder->hot_plug) + (*intel_encoder->hot_plug) (intel_encoder); } } /* Just fire off a uevent and let userspace tell us what to do */ diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c index fccf074..36c4ad7 100644 --- a/drivers/gpu/drm/i915/intel_crt.c +++ b/drivers/gpu/drm/i915/intel_crt.c @@ -246,19 +246,19 @@ static bool intel_crt_detect_hotplug(struct drm_connector *connector) static bool intel_crt_detect_ddc(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); /* CRT should always be at 0, but check anyway */ - if (intel_output->type != INTEL_OUTPUT_ANALOG) + if (intel_encoder->type != INTEL_OUTPUT_ANALOG) return false; - return intel_ddc_probe(intel_output); + return intel_ddc_probe(intel_encoder); } static enum drm_connector_status -intel_crt_load_detect(struct drm_crtc *crtc, struct intel_output *intel_output) +intel_crt_load_detect(struct drm_crtc *crtc, struct intel_encoder *intel_encoder) { - struct drm_encoder *encoder = &intel_output->enc; + struct drm_encoder *encoder = &intel_encoder->enc; struct drm_device *dev = encoder->dev; struct drm_i915_private *dev_priv = dev->dev_private; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); @@ -386,8 +386,8 @@ intel_crt_load_detect(struct drm_crtc *crtc, struct intel_output *intel_output) static enum drm_connector_status intel_crt_detect(struct drm_connector *connector) { struct drm_device *dev = connector->dev; - struct intel_output *intel_output = to_intel_output(connector); - struct drm_encoder *encoder = &intel_output->enc; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct drm_encoder *encoder = &intel_encoder->enc; struct drm_crtc *crtc; int dpms_mode; enum drm_connector_status status; @@ -404,13 +404,13 @@ static enum drm_connector_status intel_crt_detect(struct drm_connector *connecto /* for pre-945g platforms use load detect */ if (encoder->crtc && encoder->crtc->enabled) { - status = intel_crt_load_detect(encoder->crtc, intel_output); + status = intel_crt_load_detect(encoder->crtc, intel_encoder); } else { - crtc = intel_get_load_detect_pipe(intel_output, + crtc = intel_get_load_detect_pipe(intel_encoder, NULL, &dpms_mode); if (crtc) { - status = intel_crt_load_detect(crtc, intel_output); - intel_release_load_detect_pipe(intel_output, dpms_mode); + status = intel_crt_load_detect(crtc, intel_encoder); + intel_release_load_detect_pipe(intel_encoder, dpms_mode); } else status = connector_status_unknown; } @@ -420,9 +420,9 @@ static enum drm_connector_status intel_crt_detect(struct drm_connector *connecto static void intel_crt_destroy(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); - intel_i2c_destroy(intel_output->ddc_bus); + intel_i2c_destroy(intel_encoder->ddc_bus); drm_sysfs_connector_remove(connector); drm_connector_cleanup(connector); kfree(connector); @@ -431,28 +431,28 @@ static void intel_crt_destroy(struct drm_connector *connector) static int intel_crt_get_modes(struct drm_connector *connector) { int ret; - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); struct i2c_adapter *ddcbus; struct drm_device *dev = connector->dev; - ret = intel_ddc_get_modes(intel_output); + ret = intel_ddc_get_modes(intel_encoder); if (ret || !IS_G4X(dev)) goto end; - ddcbus = intel_output->ddc_bus; + ddcbus = intel_encoder->ddc_bus; /* Try to probe digital port for output in DVI-I -> VGA mode. */ - intel_output->ddc_bus = + intel_encoder->ddc_bus = intel_i2c_create(connector->dev, GPIOD, "CRTDDC_D"); - if (!intel_output->ddc_bus) { - intel_output->ddc_bus = ddcbus; + if (!intel_encoder->ddc_bus) { + intel_encoder->ddc_bus = ddcbus; dev_printk(KERN_ERR, &connector->dev->pdev->dev, "DDC bus registration failed for CRTDDC_D.\n"); goto end; } /* Try to get modes by GPIOD port */ - ret = intel_ddc_get_modes(intel_output); + ret = intel_ddc_get_modes(intel_encoder); intel_i2c_destroy(ddcbus); end: @@ -505,23 +505,23 @@ static const struct drm_encoder_funcs intel_crt_enc_funcs = { void intel_crt_init(struct drm_device *dev) { struct drm_connector *connector; - struct intel_output *intel_output; + struct intel_encoder *intel_encoder; struct drm_i915_private *dev_priv = dev->dev_private; u32 i2c_reg; - intel_output = kzalloc(sizeof(struct intel_output), GFP_KERNEL); - if (!intel_output) + intel_encoder = kzalloc(sizeof(struct intel_encoder), GFP_KERNEL); + if (!intel_encoder) return; - connector = &intel_output->base; - drm_connector_init(dev, &intel_output->base, + connector = &intel_encoder->base; + drm_connector_init(dev, &intel_encoder->base, &intel_crt_connector_funcs, DRM_MODE_CONNECTOR_VGA); - drm_encoder_init(dev, &intel_output->enc, &intel_crt_enc_funcs, + drm_encoder_init(dev, &intel_encoder->enc, &intel_crt_enc_funcs, DRM_MODE_ENCODER_DAC); - drm_mode_connector_attach_encoder(&intel_output->base, - &intel_output->enc); + drm_mode_connector_attach_encoder(&intel_encoder->base, + &intel_encoder->enc); /* Set up the DDC bus. */ if (HAS_PCH_SPLIT(dev)) @@ -532,22 +532,22 @@ void intel_crt_init(struct drm_device *dev) if (dev_priv->crt_ddc_bus != 0) i2c_reg = dev_priv->crt_ddc_bus; } - intel_output->ddc_bus = intel_i2c_create(dev, i2c_reg, "CRTDDC_A"); - if (!intel_output->ddc_bus) { + intel_encoder->ddc_bus = intel_i2c_create(dev, i2c_reg, "CRTDDC_A"); + if (!intel_encoder->ddc_bus) { dev_printk(KERN_ERR, &dev->pdev->dev, "DDC bus registration " "failed.\n"); return; } - intel_output->type = INTEL_OUTPUT_ANALOG; - intel_output->clone_mask = (1 << INTEL_SDVO_NON_TV_CLONE_BIT) | + intel_encoder->type = INTEL_OUTPUT_ANALOG; + intel_encoder->clone_mask = (1 << INTEL_SDVO_NON_TV_CLONE_BIT) | (1 << INTEL_ANALOG_CLONE_BIT) | (1 << INTEL_SDVO_LVDS_CLONE_BIT); - intel_output->crtc_mask = (1 << 0) | (1 << 1); + intel_encoder->crtc_mask = (1 << 0) | (1 << 1); connector->interlace_allowed = 0; connector->doublescan_allowed = 0; - drm_encoder_helper_add(&intel_output->enc, &intel_crt_helper_funcs); + drm_encoder_helper_add(&intel_encoder->enc, &intel_crt_helper_funcs); drm_connector_helper_add(connector, &intel_crt_connector_helper_funcs); drm_sysfs_connector_add(connector); diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 7adb3a5..2595c4c 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -746,8 +746,8 @@ bool intel_pipe_has_type (struct drm_crtc *crtc, int type) list_for_each_entry(l_entry, &mode_config->connector_list, head) { if (l_entry->encoder && l_entry->encoder->crtc == crtc) { - struct intel_output *intel_output = to_intel_output(l_entry); - if (intel_output->type == type) + struct intel_encoder *intel_encoder = to_intel_encoder(l_entry); + if (intel_encoder->type == type) return true; } } @@ -2942,19 +2942,19 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc, drm_vblank_pre_modeset(dev, pipe); list_for_each_entry(connector, &mode_config->connector_list, head) { - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); if (!connector->encoder || connector->encoder->crtc != crtc) continue; - switch (intel_output->type) { + switch (intel_encoder->type) { case INTEL_OUTPUT_LVDS: is_lvds = true; break; case INTEL_OUTPUT_SDVO: case INTEL_OUTPUT_HDMI: is_sdvo = true; - if (intel_output->needs_tv_clock) + if (intel_encoder->needs_tv_clock) is_tv = true; break; case INTEL_OUTPUT_DVO: @@ -3049,7 +3049,7 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc, struct drm_connector *edp; target_clock = mode->clock; edp = intel_pipe_get_output(crtc); - intel_edp_link_config(to_intel_output(edp), + intel_edp_link_config(to_intel_encoder(edp), &lane, &link_bw); } else { /* DP over FDI requires target mode clock @@ -3669,14 +3669,14 @@ static struct drm_display_mode load_detect_mode = { 704, 832, 0, 480, 489, 491, 520, 0, DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC), }; -struct drm_crtc *intel_get_load_detect_pipe(struct intel_output *intel_output, +struct drm_crtc *intel_get_load_detect_pipe(struct intel_encoder *intel_encoder, struct drm_display_mode *mode, int *dpms_mode) { struct intel_crtc *intel_crtc; struct drm_crtc *possible_crtc; struct drm_crtc *supported_crtc =NULL; - struct drm_encoder *encoder = &intel_output->enc; + struct drm_encoder *encoder = &intel_encoder->enc; struct drm_crtc *crtc = NULL; struct drm_device *dev = encoder->dev; struct drm_encoder_helper_funcs *encoder_funcs = encoder->helper_private; @@ -3728,8 +3728,8 @@ struct drm_crtc *intel_get_load_detect_pipe(struct intel_output *intel_output, } encoder->crtc = crtc; - intel_output->base.encoder = encoder; - intel_output->load_detect_temp = true; + intel_encoder->base.encoder = encoder; + intel_encoder->load_detect_temp = true; intel_crtc = to_intel_crtc(crtc); *dpms_mode = intel_crtc->dpms_mode; @@ -3754,18 +3754,18 @@ struct drm_crtc *intel_get_load_detect_pipe(struct intel_output *intel_output, return crtc; } -void intel_release_load_detect_pipe(struct intel_output *intel_output, int dpms_mode) +void intel_release_load_detect_pipe(struct intel_encoder *intel_encoder, int dpms_mode) { - struct drm_encoder *encoder = &intel_output->enc; + struct drm_encoder *encoder = &intel_encoder->enc; struct drm_device *dev = encoder->dev; struct drm_crtc *crtc = encoder->crtc; struct drm_encoder_helper_funcs *encoder_funcs = encoder->helper_private; struct drm_crtc_helper_funcs *crtc_funcs = crtc->helper_private; - if (intel_output->load_detect_temp) { + if (intel_encoder->load_detect_temp) { encoder->crtc = NULL; - intel_output->base.encoder = NULL; - intel_output->load_detect_temp = false; + intel_encoder->base.encoder = NULL; + intel_encoder->load_detect_temp = false; crtc->enabled = drm_helper_crtc_in_use(crtc); drm_helper_disable_unused_functions(dev); } @@ -4398,8 +4398,8 @@ static int intel_connector_clones(struct drm_device *dev, int type_mask) int entry = 0; list_for_each_entry(connector, &dev->mode_config.connector_list, head) { - struct intel_output *intel_output = to_intel_output(connector); - if (type_mask & intel_output->clone_mask) + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + if (type_mask & intel_encoder->clone_mask) index_mask |= (1 << entry); entry++; } @@ -4494,12 +4494,12 @@ static void intel_setup_outputs(struct drm_device *dev) intel_tv_init(dev); list_for_each_entry(connector, &dev->mode_config.connector_list, head) { - struct intel_output *intel_output = to_intel_output(connector); - struct drm_encoder *encoder = &intel_output->enc; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct drm_encoder *encoder = &intel_encoder->enc; - encoder->possible_crtcs = intel_output->crtc_mask; + encoder->possible_crtcs = intel_encoder->crtc_mask; encoder->possible_clones = intel_connector_clones(dev, - intel_output->clone_mask); + intel_encoder->clone_mask); } } @@ -4977,9 +4977,9 @@ void intel_modeset_cleanup(struct drm_device *dev) */ struct drm_encoder *intel_best_encoder(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); - return &intel_output->enc; + return &intel_encoder->enc; } /* diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 3ef3a0d..0a7e326 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -54,23 +54,23 @@ struct intel_dp_priv { uint8_t link_bw; uint8_t lane_count; uint8_t dpcd[4]; - struct intel_output *intel_output; + struct intel_encoder *intel_encoder; struct i2c_adapter adapter; struct i2c_algo_dp_aux_data algo; }; static void -intel_dp_link_train(struct intel_output *intel_output, uint32_t DP, +intel_dp_link_train(struct intel_encoder *intel_encoder, uint32_t DP, uint8_t link_configuration[DP_LINK_CONFIGURATION_SIZE]); static void -intel_dp_link_down(struct intel_output *intel_output, uint32_t DP); +intel_dp_link_down(struct intel_encoder *intel_encoder, uint32_t DP); void -intel_edp_link_config (struct intel_output *intel_output, +intel_edp_link_config (struct intel_encoder *intel_encoder, int *lane_num, int *link_bw) { - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; *lane_num = dp_priv->lane_count; if (dp_priv->link_bw == DP_LINK_BW_1_62) @@ -80,9 +80,9 @@ intel_edp_link_config (struct intel_output *intel_output, } static int -intel_dp_max_lane_count(struct intel_output *intel_output) +intel_dp_max_lane_count(struct intel_encoder *intel_encoder) { - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; int max_lane_count = 4; if (dp_priv->dpcd[0] >= 0x11) { @@ -98,9 +98,9 @@ intel_dp_max_lane_count(struct intel_output *intel_output) } static int -intel_dp_max_link_bw(struct intel_output *intel_output) +intel_dp_max_link_bw(struct intel_encoder *intel_encoder) { - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; int max_link_bw = dp_priv->dpcd[1]; switch (max_link_bw) { @@ -126,11 +126,11 @@ intel_dp_link_clock(uint8_t link_bw) /* I think this is a fiction */ static int intel_dp_link_required(struct drm_device *dev, - struct intel_output *intel_output, int pixel_clock) + struct intel_encoder *intel_encoder, int pixel_clock) { struct drm_i915_private *dev_priv = dev->dev_private; - if (IS_eDP(intel_output)) + if (IS_eDP(intel_encoder)) return (pixel_clock * dev_priv->edp_bpp) / 8; else return pixel_clock * 3; @@ -140,11 +140,11 @@ static int intel_dp_mode_valid(struct drm_connector *connector, struct drm_display_mode *mode) { - struct intel_output *intel_output = to_intel_output(connector); - int max_link_clock = intel_dp_link_clock(intel_dp_max_link_bw(intel_output)); - int max_lanes = intel_dp_max_lane_count(intel_output); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + int max_link_clock = intel_dp_link_clock(intel_dp_max_link_bw(intel_encoder)); + int max_lanes = intel_dp_max_lane_count(intel_encoder); - if (intel_dp_link_required(connector->dev, intel_output, mode->clock) + if (intel_dp_link_required(connector->dev, intel_encoder, mode->clock) > max_link_clock * max_lanes) return MODE_CLOCK_HIGH; @@ -208,13 +208,13 @@ intel_hrawclk(struct drm_device *dev) } static int -intel_dp_aux_ch(struct intel_output *intel_output, +intel_dp_aux_ch(struct intel_encoder *intel_encoder, uint8_t *send, int send_bytes, uint8_t *recv, int recv_size) { - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; uint32_t output_reg = dp_priv->output_reg; - struct drm_device *dev = intel_output->base.dev; + struct drm_device *dev = intel_encoder->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; uint32_t ch_ctl = output_reg + 0x10; uint32_t ch_data = ch_ctl + 4; @@ -229,7 +229,7 @@ intel_dp_aux_ch(struct intel_output *intel_output, * and would like to run at 2MHz. So, take the * hrawclk value and divide by 2 and use that */ - if (IS_eDP(intel_output)) + if (IS_eDP(intel_encoder)) aux_clock_divider = 225; /* eDP input clock at 450Mhz */ else if (HAS_PCH_SPLIT(dev)) aux_clock_divider = 62; /* IRL input clock fixed at 125Mhz */ @@ -312,7 +312,7 @@ intel_dp_aux_ch(struct intel_output *intel_output, /* Write data to the aux channel in native mode */ static int -intel_dp_aux_native_write(struct intel_output *intel_output, +intel_dp_aux_native_write(struct intel_encoder *intel_encoder, uint16_t address, uint8_t *send, int send_bytes) { int ret; @@ -329,7 +329,7 @@ intel_dp_aux_native_write(struct intel_output *intel_output, memcpy(&msg[4], send, send_bytes); msg_bytes = send_bytes + 4; for (;;) { - ret = intel_dp_aux_ch(intel_output, msg, msg_bytes, &ack, 1); + ret = intel_dp_aux_ch(intel_encoder, msg, msg_bytes, &ack, 1); if (ret < 0) return ret; if ((ack & AUX_NATIVE_REPLY_MASK) == AUX_NATIVE_REPLY_ACK) @@ -344,15 +344,15 @@ intel_dp_aux_native_write(struct intel_output *intel_output, /* Write a single byte to the aux channel in native mode */ static int -intel_dp_aux_native_write_1(struct intel_output *intel_output, +intel_dp_aux_native_write_1(struct intel_encoder *intel_encoder, uint16_t address, uint8_t byte) { - return intel_dp_aux_native_write(intel_output, address, &byte, 1); + return intel_dp_aux_native_write(intel_encoder, address, &byte, 1); } /* read bytes from a native aux channel */ static int -intel_dp_aux_native_read(struct intel_output *intel_output, +intel_dp_aux_native_read(struct intel_encoder *intel_encoder, uint16_t address, uint8_t *recv, int recv_bytes) { uint8_t msg[4]; @@ -371,7 +371,7 @@ intel_dp_aux_native_read(struct intel_output *intel_output, reply_bytes = recv_bytes + 1; for (;;) { - ret = intel_dp_aux_ch(intel_output, msg, msg_bytes, + ret = intel_dp_aux_ch(intel_encoder, msg, msg_bytes, reply, reply_bytes); if (ret == 0) return -EPROTO; @@ -397,7 +397,7 @@ intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode, struct intel_dp_priv *dp_priv = container_of(adapter, struct intel_dp_priv, adapter); - struct intel_output *intel_output = dp_priv->intel_output; + struct intel_encoder *intel_encoder = dp_priv->intel_encoder; uint16_t address = algo_data->address; uint8_t msg[5]; uint8_t reply[2]; @@ -436,7 +436,7 @@ intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode, } for (;;) { - ret = intel_dp_aux_ch(intel_output, + ret = intel_dp_aux_ch(intel_encoder, msg, msg_bytes, reply, reply_bytes); if (ret < 0) { @@ -464,9 +464,9 @@ intel_dp_i2c_aux_ch(struct i2c_adapter *adapter, int mode, } static int -intel_dp_i2c_init(struct intel_output *intel_output, const char *name) +intel_dp_i2c_init(struct intel_encoder *intel_encoder, const char *name) { - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; DRM_DEBUG_KMS("i2c_init %s\n", name); dp_priv->algo.running = false; @@ -479,7 +479,7 @@ intel_dp_i2c_init(struct intel_output *intel_output, const char *name) strncpy (dp_priv->adapter.name, name, sizeof(dp_priv->adapter.name) - 1); dp_priv->adapter.name[sizeof(dp_priv->adapter.name) - 1] = '\0'; dp_priv->adapter.algo_data = &dp_priv->algo; - dp_priv->adapter.dev.parent = &intel_output->base.kdev; + dp_priv->adapter.dev.parent = &intel_encoder->base.kdev; return i2c_dp_aux_add_bus(&dp_priv->adapter); } @@ -488,18 +488,18 @@ static bool intel_dp_mode_fixup(struct drm_encoder *encoder, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode) { - struct intel_output *intel_output = enc_to_intel_output(encoder); - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; int lane_count, clock; - int max_lane_count = intel_dp_max_lane_count(intel_output); - int max_clock = intel_dp_max_link_bw(intel_output) == DP_LINK_BW_2_7 ? 1 : 0; + int max_lane_count = intel_dp_max_lane_count(intel_encoder); + int max_clock = intel_dp_max_link_bw(intel_encoder) == DP_LINK_BW_2_7 ? 1 : 0; static int bws[2] = { DP_LINK_BW_1_62, DP_LINK_BW_2_7 }; for (lane_count = 1; lane_count <= max_lane_count; lane_count <<= 1) { for (clock = 0; clock <= max_clock; clock++) { int link_avail = intel_dp_link_clock(bws[clock]) * lane_count; - if (intel_dp_link_required(encoder->dev, intel_output, mode->clock) + if (intel_dp_link_required(encoder->dev, intel_encoder, mode->clock) <= link_avail) { dp_priv->link_bw = bws[clock]; dp_priv->lane_count = lane_count; @@ -561,16 +561,16 @@ intel_dp_set_m_n(struct drm_crtc *crtc, struct drm_display_mode *mode, struct intel_dp_m_n m_n; /* - * Find the lane count in the intel_output private + * Find the lane count in the intel_encoder private */ list_for_each_entry(connector, &mode_config->connector_list, head) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; if (!connector->encoder || connector->encoder->crtc != crtc) continue; - if (intel_output->type == INTEL_OUTPUT_DISPLAYPORT) { + if (intel_encoder->type == INTEL_OUTPUT_DISPLAYPORT) { lane_count = dp_priv->lane_count; break; } @@ -625,9 +625,9 @@ static void intel_dp_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode) { - struct intel_output *intel_output = enc_to_intel_output(encoder); - struct intel_dp_priv *dp_priv = intel_output->dev_priv; - struct drm_crtc *crtc = intel_output->enc.crtc; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; + struct drm_crtc *crtc = intel_encoder->enc.crtc; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); dp_priv->DP = (DP_LINK_TRAIN_OFF | @@ -666,7 +666,7 @@ intel_dp_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode, if (intel_crtc->pipe == 1) dp_priv->DP |= DP_PIPEB_SELECT; - if (IS_eDP(intel_output)) { + if (IS_eDP(intel_encoder)) { /* don't miss out required setting for eDP */ dp_priv->DP |= DP_PLL_ENABLE; if (adjusted_mode->clock < 200000) @@ -701,22 +701,22 @@ static void ironlake_edp_backlight_off (struct drm_device *dev) static void intel_dp_dpms(struct drm_encoder *encoder, int mode) { - struct intel_output *intel_output = enc_to_intel_output(encoder); - struct intel_dp_priv *dp_priv = intel_output->dev_priv; - struct drm_device *dev = intel_output->base.dev; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; + struct drm_device *dev = intel_encoder->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; uint32_t dp_reg = I915_READ(dp_priv->output_reg); if (mode != DRM_MODE_DPMS_ON) { if (dp_reg & DP_PORT_EN) { - intel_dp_link_down(intel_output, dp_priv->DP); - if (IS_eDP(intel_output)) + intel_dp_link_down(intel_encoder, dp_priv->DP); + if (IS_eDP(intel_encoder)) ironlake_edp_backlight_off(dev); } } else { if (!(dp_reg & DP_PORT_EN)) { - intel_dp_link_train(intel_output, dp_priv->DP, dp_priv->link_configuration); - if (IS_eDP(intel_output)) + intel_dp_link_train(intel_encoder, dp_priv->DP, dp_priv->link_configuration); + if (IS_eDP(intel_encoder)) ironlake_edp_backlight_on(dev); } } @@ -728,12 +728,12 @@ intel_dp_dpms(struct drm_encoder *encoder, int mode) * link status information */ static bool -intel_dp_get_link_status(struct intel_output *intel_output, +intel_dp_get_link_status(struct intel_encoder *intel_encoder, uint8_t link_status[DP_LINK_STATUS_SIZE]) { int ret; - ret = intel_dp_aux_native_read(intel_output, + ret = intel_dp_aux_native_read(intel_encoder, DP_LANE0_1_STATUS, link_status, DP_LINK_STATUS_SIZE); if (ret != DP_LINK_STATUS_SIZE) @@ -751,13 +751,13 @@ intel_dp_link_status(uint8_t link_status[DP_LINK_STATUS_SIZE], static void intel_dp_save(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct drm_device *dev = intel_output->base.dev; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct drm_device *dev = intel_encoder->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; dp_priv->save_DP = I915_READ(dp_priv->output_reg); - intel_dp_aux_native_read(intel_output, DP_LINK_BW_SET, + intel_dp_aux_native_read(intel_encoder, DP_LINK_BW_SET, dp_priv->save_link_configuration, sizeof (dp_priv->save_link_configuration)); } @@ -824,7 +824,7 @@ intel_dp_pre_emphasis_max(uint8_t voltage_swing) } static void -intel_get_adjust_train(struct intel_output *intel_output, +intel_get_adjust_train(struct intel_encoder *intel_encoder, uint8_t link_status[DP_LINK_STATUS_SIZE], int lane_count, uint8_t train_set[4]) @@ -941,15 +941,15 @@ intel_channel_eq_ok(uint8_t link_status[DP_LINK_STATUS_SIZE], int lane_count) } static bool -intel_dp_set_link_train(struct intel_output *intel_output, +intel_dp_set_link_train(struct intel_encoder *intel_encoder, uint32_t dp_reg_value, uint8_t dp_train_pat, uint8_t train_set[4], bool first) { - struct drm_device *dev = intel_output->base.dev; + struct drm_device *dev = intel_encoder->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; int ret; I915_WRITE(dp_priv->output_reg, dp_reg_value); @@ -957,11 +957,11 @@ intel_dp_set_link_train(struct intel_output *intel_output, if (first) intel_wait_for_vblank(dev); - intel_dp_aux_native_write_1(intel_output, + intel_dp_aux_native_write_1(intel_encoder, DP_TRAINING_PATTERN_SET, dp_train_pat); - ret = intel_dp_aux_native_write(intel_output, + ret = intel_dp_aux_native_write(intel_encoder, DP_TRAINING_LANE0_SET, train_set, 4); if (ret != 4) return false; @@ -970,12 +970,12 @@ intel_dp_set_link_train(struct intel_output *intel_output, } static void -intel_dp_link_train(struct intel_output *intel_output, uint32_t DP, +intel_dp_link_train(struct intel_encoder *intel_encoder, uint32_t DP, uint8_t link_configuration[DP_LINK_CONFIGURATION_SIZE]) { - struct drm_device *dev = intel_output->base.dev; + struct drm_device *dev = intel_encoder->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; uint8_t train_set[4]; uint8_t link_status[DP_LINK_STATUS_SIZE]; int i; @@ -986,7 +986,7 @@ intel_dp_link_train(struct intel_output *intel_output, uint32_t DP, int tries; /* Write the link configuration data */ - intel_dp_aux_native_write(intel_output, 0x100, + intel_dp_aux_native_write(intel_encoder, 0x100, link_configuration, DP_LINK_CONFIGURATION_SIZE); DP |= DP_PORT_EN; @@ -1000,14 +1000,14 @@ intel_dp_link_train(struct intel_output *intel_output, uint32_t DP, uint32_t signal_levels = intel_dp_signal_levels(train_set[0], dp_priv->lane_count); DP = (DP & ~(DP_VOLTAGE_MASK|DP_PRE_EMPHASIS_MASK)) | signal_levels; - if (!intel_dp_set_link_train(intel_output, DP | DP_LINK_TRAIN_PAT_1, + if (!intel_dp_set_link_train(intel_encoder, DP | DP_LINK_TRAIN_PAT_1, DP_TRAINING_PATTERN_1, train_set, first)) break; first = false; /* Set training pattern 1 */ udelay(100); - if (!intel_dp_get_link_status(intel_output, link_status)) + if (!intel_dp_get_link_status(intel_encoder, link_status)) break; if (intel_clock_recovery_ok(link_status, dp_priv->lane_count)) { @@ -1032,7 +1032,7 @@ intel_dp_link_train(struct intel_output *intel_output, uint32_t DP, voltage = train_set[0] & DP_TRAIN_VOLTAGE_SWING_MASK; /* Compute new train_set as requested by target */ - intel_get_adjust_train(intel_output, link_status, dp_priv->lane_count, train_set); + intel_get_adjust_train(intel_encoder, link_status, dp_priv->lane_count, train_set); } /* channel equalization */ @@ -1044,13 +1044,13 @@ intel_dp_link_train(struct intel_output *intel_output, uint32_t DP, DP = (DP & ~(DP_VOLTAGE_MASK|DP_PRE_EMPHASIS_MASK)) | signal_levels; /* channel eq pattern */ - if (!intel_dp_set_link_train(intel_output, DP | DP_LINK_TRAIN_PAT_2, + if (!intel_dp_set_link_train(intel_encoder, DP | DP_LINK_TRAIN_PAT_2, DP_TRAINING_PATTERN_2, train_set, false)) break; udelay(400); - if (!intel_dp_get_link_status(intel_output, link_status)) + if (!intel_dp_get_link_status(intel_encoder, link_status)) break; if (intel_channel_eq_ok(link_status, dp_priv->lane_count)) { @@ -1063,26 +1063,26 @@ intel_dp_link_train(struct intel_output *intel_output, uint32_t DP, break; /* Compute new train_set as requested by target */ - intel_get_adjust_train(intel_output, link_status, dp_priv->lane_count, train_set); + intel_get_adjust_train(intel_encoder, link_status, dp_priv->lane_count, train_set); ++tries; } I915_WRITE(dp_priv->output_reg, DP | DP_LINK_TRAIN_OFF); POSTING_READ(dp_priv->output_reg); - intel_dp_aux_native_write_1(intel_output, + intel_dp_aux_native_write_1(intel_encoder, DP_TRAINING_PATTERN_SET, DP_TRAINING_PATTERN_DISABLE); } static void -intel_dp_link_down(struct intel_output *intel_output, uint32_t DP) +intel_dp_link_down(struct intel_encoder *intel_encoder, uint32_t DP) { - struct drm_device *dev = intel_output->base.dev; + struct drm_device *dev = intel_encoder->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; DRM_DEBUG_KMS("\n"); - if (IS_eDP(intel_output)) { + if (IS_eDP(intel_encoder)) { DP &= ~DP_PLL_ENABLE; I915_WRITE(dp_priv->output_reg, DP); POSTING_READ(dp_priv->output_reg); @@ -1095,7 +1095,7 @@ intel_dp_link_down(struct intel_output *intel_output, uint32_t DP) udelay(17000); - if (IS_eDP(intel_output)) + if (IS_eDP(intel_encoder)) DP |= DP_LINK_TRAIN_OFF; I915_WRITE(dp_priv->output_reg, DP & ~DP_PORT_EN); POSTING_READ(dp_priv->output_reg); @@ -1104,13 +1104,13 @@ intel_dp_link_down(struct intel_output *intel_output, uint32_t DP) static void intel_dp_restore(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; if (dp_priv->save_DP & DP_PORT_EN) - intel_dp_link_train(intel_output, dp_priv->save_DP, dp_priv->save_link_configuration); + intel_dp_link_train(intel_encoder, dp_priv->save_DP, dp_priv->save_link_configuration); else - intel_dp_link_down(intel_output, dp_priv->save_DP); + intel_dp_link_down(intel_encoder, dp_priv->save_DP); } /* @@ -1123,32 +1123,32 @@ intel_dp_restore(struct drm_connector *connector) */ static void -intel_dp_check_link_status(struct intel_output *intel_output) +intel_dp_check_link_status(struct intel_encoder *intel_encoder) { - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; uint8_t link_status[DP_LINK_STATUS_SIZE]; - if (!intel_output->enc.crtc) + if (!intel_encoder->enc.crtc) return; - if (!intel_dp_get_link_status(intel_output, link_status)) { - intel_dp_link_down(intel_output, dp_priv->DP); + if (!intel_dp_get_link_status(intel_encoder, link_status)) { + intel_dp_link_down(intel_encoder, dp_priv->DP); return; } if (!intel_channel_eq_ok(link_status, dp_priv->lane_count)) - intel_dp_link_train(intel_output, dp_priv->DP, dp_priv->link_configuration); + intel_dp_link_train(intel_encoder, dp_priv->DP, dp_priv->link_configuration); } static enum drm_connector_status ironlake_dp_detect(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; enum drm_connector_status status; status = connector_status_disconnected; - if (intel_dp_aux_native_read(intel_output, + if (intel_dp_aux_native_read(intel_encoder, 0x000, dp_priv->dpcd, sizeof (dp_priv->dpcd)) == sizeof (dp_priv->dpcd)) { @@ -1167,10 +1167,10 @@ ironlake_dp_detect(struct drm_connector *connector) static enum drm_connector_status intel_dp_detect(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct drm_device *dev = intel_output->base.dev; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct drm_device *dev = intel_encoder->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; uint32_t temp, bit; enum drm_connector_status status; @@ -1209,7 +1209,7 @@ intel_dp_detect(struct drm_connector *connector) return connector_status_disconnected; status = connector_status_disconnected; - if (intel_dp_aux_native_read(intel_output, + if (intel_dp_aux_native_read(intel_encoder, 0x000, dp_priv->dpcd, sizeof (dp_priv->dpcd)) == sizeof (dp_priv->dpcd)) { @@ -1221,20 +1221,20 @@ intel_dp_detect(struct drm_connector *connector) static int intel_dp_get_modes(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct drm_device *dev = intel_output->base.dev; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct drm_device *dev = intel_encoder->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; int ret; /* We should parse the EDID data and find out if it has an audio sink */ - ret = intel_ddc_get_modes(intel_output); + ret = intel_ddc_get_modes(intel_encoder); if (ret) return ret; /* if eDP has no EDID, try to use fixed panel mode from VBT */ - if (IS_eDP(intel_output)) { + if (IS_eDP(intel_encoder)) { if (dev_priv->panel_fixed_mode != NULL) { struct drm_display_mode *mode; mode = drm_mode_duplicate(dev, dev_priv->panel_fixed_mode); @@ -1248,13 +1248,13 @@ static int intel_dp_get_modes(struct drm_connector *connector) static void intel_dp_destroy (struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); - if (intel_output->i2c_bus) - intel_i2c_destroy(intel_output->i2c_bus); + if (intel_encoder->i2c_bus) + intel_i2c_destroy(intel_encoder->i2c_bus); drm_sysfs_connector_remove(connector); drm_connector_cleanup(connector); - kfree(intel_output); + kfree(intel_encoder); } static const struct drm_encoder_helper_funcs intel_dp_helper_funcs = { @@ -1290,12 +1290,12 @@ static const struct drm_encoder_funcs intel_dp_enc_funcs = { }; void -intel_dp_hot_plug(struct intel_output *intel_output) +intel_dp_hot_plug(struct intel_encoder *intel_encoder) { - struct intel_dp_priv *dp_priv = intel_output->dev_priv; + struct intel_dp_priv *dp_priv = intel_encoder->dev_priv; if (dp_priv->dpms_mode == DRM_MODE_DPMS_ON) - intel_dp_check_link_status(intel_output); + intel_dp_check_link_status(intel_encoder); } void @@ -1303,53 +1303,53 @@ intel_dp_init(struct drm_device *dev, int output_reg) { struct drm_i915_private *dev_priv = dev->dev_private; struct drm_connector *connector; - struct intel_output *intel_output; + struct intel_encoder *intel_encoder; struct intel_dp_priv *dp_priv; const char *name = NULL; - intel_output = kcalloc(sizeof(struct intel_output) + + intel_encoder = kcalloc(sizeof(struct intel_encoder) + sizeof(struct intel_dp_priv), 1, GFP_KERNEL); - if (!intel_output) + if (!intel_encoder) return; - dp_priv = (struct intel_dp_priv *)(intel_output + 1); + dp_priv = (struct intel_dp_priv *)(intel_encoder + 1); - connector = &intel_output->base; + connector = &intel_encoder->base; drm_connector_init(dev, connector, &intel_dp_connector_funcs, DRM_MODE_CONNECTOR_DisplayPort); drm_connector_helper_add(connector, &intel_dp_connector_helper_funcs); if (output_reg == DP_A) - intel_output->type = INTEL_OUTPUT_EDP; + intel_encoder->type = INTEL_OUTPUT_EDP; else - intel_output->type = INTEL_OUTPUT_DISPLAYPORT; + intel_encoder->type = INTEL_OUTPUT_DISPLAYPORT; if (output_reg == DP_B || output_reg == PCH_DP_B) - intel_output->clone_mask = (1 << INTEL_DP_B_CLONE_BIT); + intel_encoder->clone_mask = (1 << INTEL_DP_B_CLONE_BIT); else if (output_reg == DP_C || output_reg == PCH_DP_C) - intel_output->clone_mask = (1 << INTEL_DP_C_CLONE_BIT); + intel_encoder->clone_mask = (1 << INTEL_DP_C_CLONE_BIT); else if (output_reg == DP_D || output_reg == PCH_DP_D) - intel_output->clone_mask = (1 << INTEL_DP_D_CLONE_BIT); + intel_encoder->clone_mask = (1 << INTEL_DP_D_CLONE_BIT); - if (IS_eDP(intel_output)) - intel_output->clone_mask = (1 << INTEL_EDP_CLONE_BIT); + if (IS_eDP(intel_encoder)) + intel_encoder->clone_mask = (1 << INTEL_EDP_CLONE_BIT); - intel_output->crtc_mask = (1 << 0) | (1 << 1); + intel_encoder->crtc_mask = (1 << 0) | (1 << 1); connector->interlace_allowed = true; connector->doublescan_allowed = 0; - dp_priv->intel_output = intel_output; + dp_priv->intel_encoder = intel_encoder; dp_priv->output_reg = output_reg; dp_priv->has_audio = false; dp_priv->dpms_mode = DRM_MODE_DPMS_ON; - intel_output->dev_priv = dp_priv; + intel_encoder->dev_priv = dp_priv; - drm_encoder_init(dev, &intel_output->enc, &intel_dp_enc_funcs, + drm_encoder_init(dev, &intel_encoder->enc, &intel_dp_enc_funcs, DRM_MODE_ENCODER_TMDS); - drm_encoder_helper_add(&intel_output->enc, &intel_dp_helper_funcs); + drm_encoder_helper_add(&intel_encoder->enc, &intel_dp_helper_funcs); - drm_mode_connector_attach_encoder(&intel_output->base, - &intel_output->enc); + drm_mode_connector_attach_encoder(&intel_encoder->base, + &intel_encoder->enc); drm_sysfs_connector_add(connector); /* Set up the DDC bus. */ @@ -1377,10 +1377,10 @@ intel_dp_init(struct drm_device *dev, int output_reg) break; } - intel_dp_i2c_init(intel_output, name); + intel_dp_i2c_init(intel_encoder, name); - intel_output->ddc_bus = &dp_priv->adapter; - intel_output->hot_plug = intel_dp_hot_plug; + intel_encoder->ddc_bus = &dp_priv->adapter; + intel_encoder->hot_plug = intel_dp_hot_plug; if (output_reg == DP_A) { /* initialize panel mode from VBT if available for eDP */ diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 3a467ca..e302537 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -95,7 +95,7 @@ struct intel_framebuffer { }; -struct intel_output { +struct intel_encoder { struct drm_connector base; struct drm_encoder enc; @@ -105,7 +105,7 @@ struct intel_output { bool load_detect_temp; bool needs_tv_clock; void *dev_priv; - void (*hot_plug)(struct intel_output *); + void (*hot_plug)(struct intel_encoder *); int crtc_mask; int clone_mask; }; @@ -152,15 +152,15 @@ struct intel_crtc { }; #define to_intel_crtc(x) container_of(x, struct intel_crtc, base) -#define to_intel_output(x) container_of(x, struct intel_output, base) -#define enc_to_intel_output(x) container_of(x, struct intel_output, enc) +#define to_intel_encoder(x) container_of(x, struct intel_encoder, base) +#define enc_to_intel_encoder(x) container_of(x, struct intel_encoder, enc) #define to_intel_framebuffer(x) container_of(x, struct intel_framebuffer, base) struct i2c_adapter *intel_i2c_create(struct drm_device *dev, const u32 reg, const char *name); void intel_i2c_destroy(struct i2c_adapter *adapter); -int intel_ddc_get_modes(struct intel_output *intel_output); -extern bool intel_ddc_probe(struct intel_output *intel_output); +int intel_ddc_get_modes(struct intel_encoder *intel_encoder); +extern bool intel_ddc_probe(struct intel_encoder *intel_encoder); void intel_i2c_quirk_set(struct drm_device *dev, bool enable); void intel_i2c_reset_gmbus(struct drm_device *dev); @@ -175,7 +175,7 @@ extern void intel_dp_init(struct drm_device *dev, int dp_reg); void intel_dp_set_m_n(struct drm_crtc *crtc, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode); -extern void intel_edp_link_config (struct intel_output *, int *, int *); +extern void intel_edp_link_config (struct intel_encoder *, int *, int *); extern int intel_panel_fitter_pipe (struct drm_device *dev); @@ -191,10 +191,10 @@ int intel_get_pipe_from_crtc_id(struct drm_device *dev, void *data, struct drm_file *file_priv); extern void intel_wait_for_vblank(struct drm_device *dev); extern struct drm_crtc *intel_get_crtc_from_pipe(struct drm_device *dev, int pipe); -extern struct drm_crtc *intel_get_load_detect_pipe(struct intel_output *intel_output, +extern struct drm_crtc *intel_get_load_detect_pipe(struct intel_encoder *intel_encoder, struct drm_display_mode *mode, int *dpms_mode); -extern void intel_release_load_detect_pipe(struct intel_output *intel_output, +extern void intel_release_load_detect_pipe(struct intel_encoder *intel_encoder, int dpms_mode); extern struct drm_connector* intel_sdvo_find(struct drm_device *dev, int sdvoB); diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c index a4d2606..62282f3 100644 --- a/drivers/gpu/drm/i915/intel_dvo.c +++ b/drivers/gpu/drm/i915/intel_dvo.c @@ -79,8 +79,8 @@ static struct intel_dvo_device intel_dvo_devices[] = { static void intel_dvo_dpms(struct drm_encoder *encoder, int mode) { struct drm_i915_private *dev_priv = encoder->dev->dev_private; - struct intel_output *intel_output = enc_to_intel_output(encoder); - struct intel_dvo_device *dvo = intel_output->dev_priv; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_dvo_device *dvo = intel_encoder->dev_priv; u32 dvo_reg = dvo->dvo_reg; u32 temp = I915_READ(dvo_reg); @@ -98,8 +98,8 @@ static void intel_dvo_dpms(struct drm_encoder *encoder, int mode) static void intel_dvo_save(struct drm_connector *connector) { struct drm_i915_private *dev_priv = connector->dev->dev_private; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_dvo_device *dvo = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_dvo_device *dvo = intel_encoder->dev_priv; /* Each output should probably just save the registers it touches, * but for now, use more overkill. @@ -114,8 +114,8 @@ static void intel_dvo_save(struct drm_connector *connector) static void intel_dvo_restore(struct drm_connector *connector) { struct drm_i915_private *dev_priv = connector->dev->dev_private; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_dvo_device *dvo = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_dvo_device *dvo = intel_encoder->dev_priv; dvo->dev_ops->restore(dvo); @@ -127,8 +127,8 @@ static void intel_dvo_restore(struct drm_connector *connector) static int intel_dvo_mode_valid(struct drm_connector *connector, struct drm_display_mode *mode) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_dvo_device *dvo = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_dvo_device *dvo = intel_encoder->dev_priv; if (mode->flags & DRM_MODE_FLAG_DBLSCAN) return MODE_NO_DBLESCAN; @@ -149,8 +149,8 @@ static bool intel_dvo_mode_fixup(struct drm_encoder *encoder, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode) { - struct intel_output *intel_output = enc_to_intel_output(encoder); - struct intel_dvo_device *dvo = intel_output->dev_priv; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_dvo_device *dvo = intel_encoder->dev_priv; /* If we have timings from the BIOS for the panel, put them in * to the adjusted mode. The CRTC will be set up for this mode, @@ -185,8 +185,8 @@ static void intel_dvo_mode_set(struct drm_encoder *encoder, struct drm_device *dev = encoder->dev; struct drm_i915_private *dev_priv = dev->dev_private; struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc); - struct intel_output *intel_output = enc_to_intel_output(encoder); - struct intel_dvo_device *dvo = intel_output->dev_priv; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_dvo_device *dvo = intel_encoder->dev_priv; int pipe = intel_crtc->pipe; u32 dvo_val; u32 dvo_reg = dvo->dvo_reg, dvo_srcdim_reg; @@ -240,23 +240,23 @@ static void intel_dvo_mode_set(struct drm_encoder *encoder, */ static enum drm_connector_status intel_dvo_detect(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_dvo_device *dvo = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_dvo_device *dvo = intel_encoder->dev_priv; return dvo->dev_ops->detect(dvo); } static int intel_dvo_get_modes(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_dvo_device *dvo = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_dvo_device *dvo = intel_encoder->dev_priv; /* We should probably have an i2c driver get_modes function for those * devices which will have a fixed set of modes determined by the chip * (TV-out, for example), but for now with just TMDS and LVDS, * that's not the case. */ - intel_ddc_get_modes(intel_output); + intel_ddc_get_modes(intel_encoder); if (!list_empty(&connector->probed_modes)) return 1; @@ -274,8 +274,8 @@ static int intel_dvo_get_modes(struct drm_connector *connector) static void intel_dvo_destroy (struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_dvo_device *dvo = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_dvo_device *dvo = intel_encoder->dev_priv; if (dvo) { if (dvo->dev_ops->destroy) @@ -285,13 +285,13 @@ static void intel_dvo_destroy (struct drm_connector *connector) /* no need, in i830_dvoices[] now */ //kfree(dvo); } - if (intel_output->i2c_bus) - intel_i2c_destroy(intel_output->i2c_bus); - if (intel_output->ddc_bus) - intel_i2c_destroy(intel_output->ddc_bus); + if (intel_encoder->i2c_bus) + intel_i2c_destroy(intel_encoder->i2c_bus); + if (intel_encoder->ddc_bus) + intel_i2c_destroy(intel_encoder->ddc_bus); drm_sysfs_connector_remove(connector); drm_connector_cleanup(connector); - kfree(intel_output); + kfree(intel_encoder); } #ifdef RANDR_GET_CRTC_INTERFACE @@ -299,8 +299,8 @@ static struct drm_crtc *intel_dvo_get_crtc(struct drm_connector *connector) { struct drm_device *dev = connector->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_dvo_device *dvo = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_dvo_device *dvo = intel_encoder->dev_priv; int pipe = !!(I915_READ(dvo->dvo_reg) & SDVO_PIPE_B_SELECT); return intel_pipe_to_crtc(pScrn, pipe); @@ -351,8 +351,8 @@ intel_dvo_get_current_mode (struct drm_connector *connector) { struct drm_device *dev = connector->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_dvo_device *dvo = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_dvo_device *dvo = intel_encoder->dev_priv; uint32_t dvo_reg = dvo->dvo_reg; uint32_t dvo_val = I915_READ(dvo_reg); struct drm_display_mode *mode = NULL; @@ -382,24 +382,24 @@ intel_dvo_get_current_mode (struct drm_connector *connector) void intel_dvo_init(struct drm_device *dev) { - struct intel_output *intel_output; + struct intel_encoder *intel_encoder; struct intel_dvo_device *dvo; struct i2c_adapter *i2cbus = NULL; int ret = 0; int i; int encoder_type = DRM_MODE_ENCODER_NONE; - intel_output = kzalloc (sizeof(struct intel_output), GFP_KERNEL); - if (!intel_output) + intel_encoder = kzalloc (sizeof(struct intel_encoder), GFP_KERNEL); + if (!intel_encoder) return; /* Set up the DDC bus */ - intel_output->ddc_bus = intel_i2c_create(dev, GPIOD, "DVODDC_D"); - if (!intel_output->ddc_bus) + intel_encoder->ddc_bus = intel_i2c_create(dev, GPIOD, "DVODDC_D"); + if (!intel_encoder->ddc_bus) goto free_intel; /* Now, try to find a controller */ for (i = 0; i < ARRAY_SIZE(intel_dvo_devices); i++) { - struct drm_connector *connector = &intel_output->base; + struct drm_connector *connector = &intel_encoder->base; int gpio; dvo = &intel_dvo_devices[i]; @@ -434,11 +434,11 @@ void intel_dvo_init(struct drm_device *dev) if (!ret) continue; - intel_output->type = INTEL_OUTPUT_DVO; - intel_output->crtc_mask = (1 << 0) | (1 << 1); + intel_encoder->type = INTEL_OUTPUT_DVO; + intel_encoder->crtc_mask = (1 << 0) | (1 << 1); switch (dvo->type) { case INTEL_DVO_CHIP_TMDS: - intel_output->clone_mask = + intel_encoder->clone_mask = (1 << INTEL_DVO_TMDS_CLONE_BIT) | (1 << INTEL_ANALOG_CLONE_BIT); drm_connector_init(dev, connector, @@ -447,7 +447,7 @@ void intel_dvo_init(struct drm_device *dev) encoder_type = DRM_MODE_ENCODER_TMDS; break; case INTEL_DVO_CHIP_LVDS: - intel_output->clone_mask = + intel_encoder->clone_mask = (1 << INTEL_DVO_LVDS_CLONE_BIT); drm_connector_init(dev, connector, &intel_dvo_connector_funcs, @@ -462,16 +462,16 @@ void intel_dvo_init(struct drm_device *dev) connector->interlace_allowed = false; connector->doublescan_allowed = false; - intel_output->dev_priv = dvo; - intel_output->i2c_bus = i2cbus; + intel_encoder->dev_priv = dvo; + intel_encoder->i2c_bus = i2cbus; - drm_encoder_init(dev, &intel_output->enc, + drm_encoder_init(dev, &intel_encoder->enc, &intel_dvo_enc_funcs, encoder_type); - drm_encoder_helper_add(&intel_output->enc, + drm_encoder_helper_add(&intel_encoder->enc, &intel_dvo_helper_funcs); - drm_mode_connector_attach_encoder(&intel_output->base, - &intel_output->enc); + drm_mode_connector_attach_encoder(&intel_encoder->base, + &intel_encoder->enc); if (dvo->type == INTEL_DVO_CHIP_LVDS) { /* For our LVDS chipsets, we should hopefully be able * to dig the fixed panel mode out of the BIOS data. @@ -489,10 +489,10 @@ void intel_dvo_init(struct drm_device *dev) return; } - intel_i2c_destroy(intel_output->ddc_bus); + intel_i2c_destroy(intel_encoder->ddc_bus); /* Didn't find a chip, so tear down. */ if (i2cbus != NULL) intel_i2c_destroy(i2cbus); free_intel: - kfree(intel_output); + kfree(intel_encoder); } diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index a30f8bf..9777504 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -50,8 +50,8 @@ static void intel_hdmi_mode_set(struct drm_encoder *encoder, struct drm_i915_private *dev_priv = dev->dev_private; struct drm_crtc *crtc = encoder->crtc; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_output *intel_output = enc_to_intel_output(encoder); - struct intel_hdmi_priv *hdmi_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_hdmi_priv *hdmi_priv = intel_encoder->dev_priv; u32 sdvox; sdvox = SDVO_ENCODING_HDMI | @@ -73,8 +73,8 @@ static void intel_hdmi_dpms(struct drm_encoder *encoder, int mode) { struct drm_device *dev = encoder->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_output *intel_output = enc_to_intel_output(encoder); - struct intel_hdmi_priv *hdmi_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_hdmi_priv *hdmi_priv = intel_encoder->dev_priv; u32 temp; temp = I915_READ(hdmi_priv->sdvox_reg); @@ -109,8 +109,8 @@ static void intel_hdmi_save(struct drm_connector *connector) { struct drm_device *dev = connector->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_hdmi_priv *hdmi_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_hdmi_priv *hdmi_priv = intel_encoder->dev_priv; hdmi_priv->save_SDVOX = I915_READ(hdmi_priv->sdvox_reg); } @@ -119,8 +119,8 @@ static void intel_hdmi_restore(struct drm_connector *connector) { struct drm_device *dev = connector->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_hdmi_priv *hdmi_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_hdmi_priv *hdmi_priv = intel_encoder->dev_priv; I915_WRITE(hdmi_priv->sdvox_reg, hdmi_priv->save_SDVOX); POSTING_READ(hdmi_priv->sdvox_reg); @@ -150,21 +150,21 @@ static bool intel_hdmi_mode_fixup(struct drm_encoder *encoder, static enum drm_connector_status intel_hdmi_detect(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_hdmi_priv *hdmi_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_hdmi_priv *hdmi_priv = intel_encoder->dev_priv; struct edid *edid = NULL; enum drm_connector_status status = connector_status_disconnected; hdmi_priv->has_hdmi_sink = false; - edid = drm_get_edid(&intel_output->base, - intel_output->ddc_bus); + edid = drm_get_edid(&intel_encoder->base, + intel_encoder->ddc_bus); if (edid) { if (edid->input & DRM_EDID_INPUT_DIGITAL) { status = connector_status_connected; hdmi_priv->has_hdmi_sink = drm_detect_hdmi_monitor(edid); } - intel_output->base.display_info.raw_edid = NULL; + intel_encoder->base.display_info.raw_edid = NULL; kfree(edid); } @@ -173,24 +173,24 @@ intel_hdmi_detect(struct drm_connector *connector) static int intel_hdmi_get_modes(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); /* We should parse the EDID data and find out if it's an HDMI sink so * we can send audio to it. */ - return intel_ddc_get_modes(intel_output); + return intel_ddc_get_modes(intel_encoder); } static void intel_hdmi_destroy(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); - if (intel_output->i2c_bus) - intel_i2c_destroy(intel_output->i2c_bus); + if (intel_encoder->i2c_bus) + intel_i2c_destroy(intel_encoder->i2c_bus); drm_sysfs_connector_remove(connector); drm_connector_cleanup(connector); - kfree(intel_output); + kfree(intel_encoder); } static const struct drm_encoder_helper_funcs intel_hdmi_helper_funcs = { @@ -229,63 +229,63 @@ void intel_hdmi_init(struct drm_device *dev, int sdvox_reg) { struct drm_i915_private *dev_priv = dev->dev_private; struct drm_connector *connector; - struct intel_output *intel_output; + struct intel_encoder *intel_encoder; struct intel_hdmi_priv *hdmi_priv; - intel_output = kcalloc(sizeof(struct intel_output) + + intel_encoder = kcalloc(sizeof(struct intel_encoder) + sizeof(struct intel_hdmi_priv), 1, GFP_KERNEL); - if (!intel_output) + if (!intel_encoder) return; - hdmi_priv = (struct intel_hdmi_priv *)(intel_output + 1); + hdmi_priv = (struct intel_hdmi_priv *)(intel_encoder + 1); - connector = &intel_output->base; + connector = &intel_encoder->base; drm_connector_init(dev, connector, &intel_hdmi_connector_funcs, DRM_MODE_CONNECTOR_HDMIA); drm_connector_helper_add(connector, &intel_hdmi_connector_helper_funcs); - intel_output->type = INTEL_OUTPUT_HDMI; + intel_encoder->type = INTEL_OUTPUT_HDMI; connector->interlace_allowed = 0; connector->doublescan_allowed = 0; - intel_output->crtc_mask = (1 << 0) | (1 << 1); + intel_encoder->crtc_mask = (1 << 0) | (1 << 1); /* Set up the DDC bus. */ if (sdvox_reg == SDVOB) { - intel_output->clone_mask = (1 << INTEL_HDMIB_CLONE_BIT); - intel_output->ddc_bus = intel_i2c_create(dev, GPIOE, "HDMIB"); + intel_encoder->clone_mask = (1 << INTEL_HDMIB_CLONE_BIT); + intel_encoder->ddc_bus = intel_i2c_create(dev, GPIOE, "HDMIB"); dev_priv->hotplug_supported_mask |= HDMIB_HOTPLUG_INT_STATUS; } else if (sdvox_reg == SDVOC) { - intel_output->clone_mask = (1 << INTEL_HDMIC_CLONE_BIT); - intel_output->ddc_bus = intel_i2c_create(dev, GPIOD, "HDMIC"); + intel_encoder->clone_mask = (1 << INTEL_HDMIC_CLONE_BIT); + intel_encoder->ddc_bus = intel_i2c_create(dev, GPIOD, "HDMIC"); dev_priv->hotplug_supported_mask |= HDMIC_HOTPLUG_INT_STATUS; } else if (sdvox_reg == HDMIB) { - intel_output->clone_mask = (1 << INTEL_HDMID_CLONE_BIT); - intel_output->ddc_bus = intel_i2c_create(dev, PCH_GPIOE, + intel_encoder->clone_mask = (1 << INTEL_HDMID_CLONE_BIT); + intel_encoder->ddc_bus = intel_i2c_create(dev, PCH_GPIOE, "HDMIB"); dev_priv->hotplug_supported_mask |= HDMIB_HOTPLUG_INT_STATUS; } else if (sdvox_reg == HDMIC) { - intel_output->clone_mask = (1 << INTEL_HDMIE_CLONE_BIT); - intel_output->ddc_bus = intel_i2c_create(dev, PCH_GPIOD, + intel_encoder->clone_mask = (1 << INTEL_HDMIE_CLONE_BIT); + intel_encoder->ddc_bus = intel_i2c_create(dev, PCH_GPIOD, "HDMIC"); dev_priv->hotplug_supported_mask |= HDMIC_HOTPLUG_INT_STATUS; } else if (sdvox_reg == HDMID) { - intel_output->clone_mask = (1 << INTEL_HDMIF_CLONE_BIT); - intel_output->ddc_bus = intel_i2c_create(dev, PCH_GPIOF, + intel_encoder->clone_mask = (1 << INTEL_HDMIF_CLONE_BIT); + intel_encoder->ddc_bus = intel_i2c_create(dev, PCH_GPIOF, "HDMID"); dev_priv->hotplug_supported_mask |= HDMID_HOTPLUG_INT_STATUS; } - if (!intel_output->ddc_bus) + if (!intel_encoder->ddc_bus) goto err_connector; hdmi_priv->sdvox_reg = sdvox_reg; - intel_output->dev_priv = hdmi_priv; + intel_encoder->dev_priv = hdmi_priv; - drm_encoder_init(dev, &intel_output->enc, &intel_hdmi_enc_funcs, + drm_encoder_init(dev, &intel_encoder->enc, &intel_hdmi_enc_funcs, DRM_MODE_ENCODER_TMDS); - drm_encoder_helper_add(&intel_output->enc, &intel_hdmi_helper_funcs); + drm_encoder_helper_add(&intel_encoder->enc, &intel_hdmi_helper_funcs); - drm_mode_connector_attach_encoder(&intel_output->base, - &intel_output->enc); + drm_mode_connector_attach_encoder(&intel_encoder->base, + &intel_encoder->enc); drm_sysfs_connector_add(connector); /* For G4X desktop chip, PEG_BAND_GAP_DATA 3:0 must first be written @@ -301,7 +301,7 @@ void intel_hdmi_init(struct drm_device *dev, int sdvox_reg) err_connector: drm_connector_cleanup(connector); - kfree(intel_output); + kfree(intel_encoder); return; } diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c index 2b3fa7a..9d99ddc 100644 --- a/drivers/gpu/drm/i915/intel_lvds.c +++ b/drivers/gpu/drm/i915/intel_lvds.c @@ -238,8 +238,8 @@ static bool intel_lvds_mode_fixup(struct drm_encoder *encoder, struct drm_i915_private *dev_priv = dev->dev_private; struct intel_crtc *intel_crtc = to_intel_crtc(encoder->crtc); struct drm_encoder *tmp_encoder; - struct intel_output *intel_output = enc_to_intel_output(encoder); - struct intel_lvds_priv *lvds_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_lvds_priv *lvds_priv = intel_encoder->dev_priv; u32 pfit_control = 0, pfit_pgm_ratios = 0; int left_border = 0, right_border = 0, top_border = 0; int bottom_border = 0; @@ -586,8 +586,8 @@ static void intel_lvds_mode_set(struct drm_encoder *encoder, { struct drm_device *dev = encoder->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_output *intel_output = enc_to_intel_output(encoder); - struct intel_lvds_priv *lvds_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_lvds_priv *lvds_priv = intel_encoder->dev_priv; /* * The LVDS pin pair will already have been turned on in the @@ -634,11 +634,11 @@ static enum drm_connector_status intel_lvds_detect(struct drm_connector *connect static int intel_lvds_get_modes(struct drm_connector *connector) { struct drm_device *dev = connector->dev; - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); struct drm_i915_private *dev_priv = dev->dev_private; int ret = 0; - ret = intel_ddc_get_modes(intel_output); + ret = intel_ddc_get_modes(intel_encoder); if (ret) return ret; @@ -714,11 +714,11 @@ static int intel_lid_notify(struct notifier_block *nb, unsigned long val, static void intel_lvds_destroy(struct drm_connector *connector) { struct drm_device *dev = connector->dev; - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); struct drm_i915_private *dev_priv = dev->dev_private; - if (intel_output->ddc_bus) - intel_i2c_destroy(intel_output->ddc_bus); + if (intel_encoder->ddc_bus) + intel_i2c_destroy(intel_encoder->ddc_bus); if (dev_priv->lid_notifier.notifier_call) acpi_lid_notifier_unregister(&dev_priv->lid_notifier); drm_sysfs_connector_remove(connector); @@ -731,13 +731,13 @@ static int intel_lvds_set_property(struct drm_connector *connector, uint64_t value) { struct drm_device *dev = connector->dev; - struct intel_output *intel_output = - to_intel_output(connector); + struct intel_encoder *intel_encoder = + to_intel_encoder(connector); if (property == dev->mode_config.scaling_mode_property && connector->encoder) { struct drm_crtc *crtc = connector->encoder->crtc; - struct intel_lvds_priv *lvds_priv = intel_output->dev_priv; + struct intel_lvds_priv *lvds_priv = intel_encoder->dev_priv; if (value == DRM_MODE_SCALE_NONE) { DRM_DEBUG_KMS("no scaling not supported\n"); return 0; @@ -967,7 +967,7 @@ static int lvds_is_present_in_vbt(struct drm_device *dev) void intel_lvds_init(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_output *intel_output; + struct intel_encoder *intel_encoder; struct drm_connector *connector; struct drm_encoder *encoder; struct drm_display_mode *scan; /* *modes, *bios_mode; */ @@ -995,40 +995,40 @@ void intel_lvds_init(struct drm_device *dev) gpio = PCH_GPIOC; } - intel_output = kzalloc(sizeof(struct intel_output) + + intel_encoder = kzalloc(sizeof(struct intel_encoder) + sizeof(struct intel_lvds_priv), GFP_KERNEL); - if (!intel_output) { + if (!intel_encoder) { return; } - connector = &intel_output->base; - encoder = &intel_output->enc; - drm_connector_init(dev, &intel_output->base, &intel_lvds_connector_funcs, + connector = &intel_encoder->base; + encoder = &intel_encoder->enc; + drm_connector_init(dev, &intel_encoder->base, &intel_lvds_connector_funcs, DRM_MODE_CONNECTOR_LVDS); - drm_encoder_init(dev, &intel_output->enc, &intel_lvds_enc_funcs, + drm_encoder_init(dev, &intel_encoder->enc, &intel_lvds_enc_funcs, DRM_MODE_ENCODER_LVDS); - drm_mode_connector_attach_encoder(&intel_output->base, &intel_output->enc); - intel_output->type = INTEL_OUTPUT_LVDS; + drm_mode_connector_attach_encoder(&intel_encoder->base, &intel_encoder->enc); + intel_encoder->type = INTEL_OUTPUT_LVDS; - intel_output->clone_mask = (1 << INTEL_LVDS_CLONE_BIT); - intel_output->crtc_mask = (1 << 1); + intel_encoder->clone_mask = (1 << INTEL_LVDS_CLONE_BIT); + intel_encoder->crtc_mask = (1 << 1); drm_encoder_helper_add(encoder, &intel_lvds_helper_funcs); drm_connector_helper_add(connector, &intel_lvds_connector_helper_funcs); connector->display_info.subpixel_order = SubPixelHorizontalRGB; connector->interlace_allowed = false; connector->doublescan_allowed = false; - lvds_priv = (struct intel_lvds_priv *)(intel_output + 1); - intel_output->dev_priv = lvds_priv; + lvds_priv = (struct intel_lvds_priv *)(intel_encoder + 1); + intel_encoder->dev_priv = lvds_priv; /* create the scaling mode property */ drm_mode_create_scaling_mode_property(dev); /* * the initial panel fitting mode will be FULL_SCREEN. */ - drm_connector_attach_property(&intel_output->base, + drm_connector_attach_property(&intel_encoder->base, dev->mode_config.scaling_mode_property, DRM_MODE_SCALE_FULLSCREEN); lvds_priv->fitting_mode = DRM_MODE_SCALE_FULLSCREEN; @@ -1043,8 +1043,8 @@ void intel_lvds_init(struct drm_device *dev) */ /* Set up the DDC bus. */ - intel_output->ddc_bus = intel_i2c_create(dev, gpio, "LVDSDDC_C"); - if (!intel_output->ddc_bus) { + intel_encoder->ddc_bus = intel_i2c_create(dev, gpio, "LVDSDDC_C"); + if (!intel_encoder->ddc_bus) { dev_printk(KERN_ERR, &dev->pdev->dev, "DDC bus registration " "failed.\n"); goto failed; @@ -1054,7 +1054,7 @@ void intel_lvds_init(struct drm_device *dev) * Attempt to get the fixed panel mode from DDC. Assume that the * preferred mode is the right one. */ - intel_ddc_get_modes(intel_output); + intel_ddc_get_modes(intel_encoder); list_for_each_entry(scan, &connector->probed_modes, head) { mutex_lock(&dev->mode_config.mutex); @@ -1132,9 +1132,9 @@ out: failed: DRM_DEBUG_KMS("No LVDS modes found, disabling.\n"); - if (intel_output->ddc_bus) - intel_i2c_destroy(intel_output->ddc_bus); + if (intel_encoder->ddc_bus) + intel_i2c_destroy(intel_encoder->ddc_bus); drm_connector_cleanup(connector); drm_encoder_cleanup(encoder); - kfree(intel_output); + kfree(intel_encoder); } diff --git a/drivers/gpu/drm/i915/intel_modes.c b/drivers/gpu/drm/i915/intel_modes.c index 67e2f46..3111a1c 100644 --- a/drivers/gpu/drm/i915/intel_modes.c +++ b/drivers/gpu/drm/i915/intel_modes.c @@ -33,7 +33,7 @@ * intel_ddc_probe * */ -bool intel_ddc_probe(struct intel_output *intel_output) +bool intel_ddc_probe(struct intel_encoder *intel_encoder) { u8 out_buf[] = { 0x0, 0x0}; u8 buf[2]; @@ -53,9 +53,9 @@ bool intel_ddc_probe(struct intel_output *intel_output) } }; - intel_i2c_quirk_set(intel_output->base.dev, true); - ret = i2c_transfer(intel_output->ddc_bus, msgs, 2); - intel_i2c_quirk_set(intel_output->base.dev, false); + intel_i2c_quirk_set(intel_encoder->base.dev, true); + ret = i2c_transfer(intel_encoder->ddc_bus, msgs, 2); + intel_i2c_quirk_set(intel_encoder->base.dev, false); if (ret == 2) return true; @@ -68,19 +68,19 @@ bool intel_ddc_probe(struct intel_output *intel_output) * * Fetch the EDID information from @connector using the DDC bus. */ -int intel_ddc_get_modes(struct intel_output *intel_output) +int intel_ddc_get_modes(struct intel_encoder *intel_encoder) { struct edid *edid; int ret = 0; - intel_i2c_quirk_set(intel_output->base.dev, true); - edid = drm_get_edid(&intel_output->base, intel_output->ddc_bus); - intel_i2c_quirk_set(intel_output->base.dev, false); + intel_i2c_quirk_set(intel_encoder->base.dev, true); + edid = drm_get_edid(&intel_encoder->base, intel_encoder->ddc_bus); + intel_i2c_quirk_set(intel_encoder->base.dev, false); if (edid) { - drm_mode_connector_update_edid_property(&intel_output->base, + drm_mode_connector_update_edid_property(&intel_encoder->base, edid); - ret = drm_add_edid_modes(&intel_output->base, edid); - intel_output->base.display_info.raw_edid = NULL; + ret = drm_add_edid_modes(&intel_encoder->base, edid); + intel_encoder->base.display_info.raw_edid = NULL; kfree(edid); } diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c index 48daee5..ea6de3b 100644 --- a/drivers/gpu/drm/i915/intel_sdvo.c +++ b/drivers/gpu/drm/i915/intel_sdvo.c @@ -161,18 +161,18 @@ struct intel_sdvo_priv { }; static bool -intel_sdvo_output_setup(struct intel_output *intel_output, uint16_t flags); +intel_sdvo_output_setup(struct intel_encoder *intel_encoder, uint16_t flags); /** * Writes the SDVOB or SDVOC with the given value, but always writes both * SDVOB and SDVOC to work around apparent hardware issues (according to * comments in the BIOS). */ -static void intel_sdvo_write_sdvox(struct intel_output *intel_output, u32 val) +static void intel_sdvo_write_sdvox(struct intel_encoder *intel_encoder, u32 val) { - struct drm_device *dev = intel_output->base.dev; + struct drm_device *dev = intel_encoder->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; u32 bval = val, cval = val; int i; @@ -195,10 +195,10 @@ static void intel_sdvo_write_sdvox(struct intel_output *intel_output, u32 val) } } -static bool intel_sdvo_read_byte(struct intel_output *intel_output, u8 addr, +static bool intel_sdvo_read_byte(struct intel_encoder *intel_encoder, u8 addr, u8 *ch) { - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; u8 out_buf[2]; u8 buf[2]; int ret; @@ -221,7 +221,7 @@ static bool intel_sdvo_read_byte(struct intel_output *intel_output, u8 addr, out_buf[0] = addr; out_buf[1] = 0; - if ((ret = i2c_transfer(intel_output->i2c_bus, msgs, 2)) == 2) + if ((ret = i2c_transfer(intel_encoder->i2c_bus, msgs, 2)) == 2) { *ch = buf[0]; return true; @@ -231,10 +231,10 @@ static bool intel_sdvo_read_byte(struct intel_output *intel_output, u8 addr, return false; } -static bool intel_sdvo_write_byte(struct intel_output *intel_output, int addr, +static bool intel_sdvo_write_byte(struct intel_encoder *intel_encoder, int addr, u8 ch) { - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; u8 out_buf[2]; struct i2c_msg msgs[] = { { @@ -248,7 +248,7 @@ static bool intel_sdvo_write_byte(struct intel_output *intel_output, int addr, out_buf[0] = addr; out_buf[1] = ch; - if (i2c_transfer(intel_output->i2c_bus, msgs, 1) == 1) + if (i2c_transfer(intel_encoder->i2c_bus, msgs, 1) == 1) { return true; } @@ -355,10 +355,10 @@ static const struct _sdvo_cmd_name { #define SDVO_NAME(dev_priv) ((dev_priv)->output_device == SDVOB ? "SDVOB" : "SDVOC") #define SDVO_PRIV(output) ((struct intel_sdvo_priv *) (output)->dev_priv) -static void intel_sdvo_debug_write(struct intel_output *intel_output, u8 cmd, +static void intel_sdvo_debug_write(struct intel_encoder *intel_encoder, u8 cmd, void *args, int args_len) { - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; int i; DRM_DEBUG_KMS("%s: W: %02X ", @@ -378,19 +378,19 @@ static void intel_sdvo_debug_write(struct intel_output *intel_output, u8 cmd, DRM_LOG_KMS("\n"); } -static void intel_sdvo_write_cmd(struct intel_output *intel_output, u8 cmd, +static void intel_sdvo_write_cmd(struct intel_encoder *intel_encoder, u8 cmd, void *args, int args_len) { int i; - intel_sdvo_debug_write(intel_output, cmd, args, args_len); + intel_sdvo_debug_write(intel_encoder, cmd, args, args_len); for (i = 0; i < args_len; i++) { - intel_sdvo_write_byte(intel_output, SDVO_I2C_ARG_0 - i, + intel_sdvo_write_byte(intel_encoder, SDVO_I2C_ARG_0 - i, ((u8*)args)[i]); } - intel_sdvo_write_byte(intel_output, SDVO_I2C_OPCODE, cmd); + intel_sdvo_write_byte(intel_encoder, SDVO_I2C_OPCODE, cmd); } static const char *cmd_status_names[] = { @@ -403,11 +403,11 @@ static const char *cmd_status_names[] = { "Scaling not supported" }; -static void intel_sdvo_debug_response(struct intel_output *intel_output, +static void intel_sdvo_debug_response(struct intel_encoder *intel_encoder, void *response, int response_len, u8 status) { - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; int i; DRM_DEBUG_KMS("%s: R: ", SDVO_NAME(sdvo_priv)); @@ -422,7 +422,7 @@ static void intel_sdvo_debug_response(struct intel_output *intel_output, DRM_LOG_KMS("\n"); } -static u8 intel_sdvo_read_response(struct intel_output *intel_output, +static u8 intel_sdvo_read_response(struct intel_encoder *intel_encoder, void *response, int response_len) { int i; @@ -432,16 +432,16 @@ static u8 intel_sdvo_read_response(struct intel_output *intel_output, while (retry--) { /* Read the command response */ for (i = 0; i < response_len; i++) { - intel_sdvo_read_byte(intel_output, + intel_sdvo_read_byte(intel_encoder, SDVO_I2C_RETURN_0 + i, &((u8 *)response)[i]); } /* read the return status */ - intel_sdvo_read_byte(intel_output, SDVO_I2C_CMD_STATUS, + intel_sdvo_read_byte(intel_encoder, SDVO_I2C_CMD_STATUS, &status); - intel_sdvo_debug_response(intel_output, response, response_len, + intel_sdvo_debug_response(intel_encoder, response, response_len, status); if (status != SDVO_CMD_STATUS_PENDING) return status; @@ -469,10 +469,10 @@ static int intel_sdvo_get_pixel_multiplier(struct drm_display_mode *mode) * another I2C transaction after issuing the DDC bus switch, it will be * switched to the internal SDVO register. */ -static void intel_sdvo_set_control_bus_switch(struct intel_output *intel_output, +static void intel_sdvo_set_control_bus_switch(struct intel_encoder *intel_encoder, u8 target) { - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; u8 out_buf[2], cmd_buf[2], ret_value[2], ret; struct i2c_msg msgs[] = { { @@ -496,10 +496,10 @@ static void intel_sdvo_set_control_bus_switch(struct intel_output *intel_output, }, }; - intel_sdvo_debug_write(intel_output, SDVO_CMD_SET_CONTROL_BUS_SWITCH, + intel_sdvo_debug_write(intel_encoder, SDVO_CMD_SET_CONTROL_BUS_SWITCH, &target, 1); /* write the DDC switch command argument */ - intel_sdvo_write_byte(intel_output, SDVO_I2C_ARG_0, target); + intel_sdvo_write_byte(intel_encoder, SDVO_I2C_ARG_0, target); out_buf[0] = SDVO_I2C_OPCODE; out_buf[1] = SDVO_CMD_SET_CONTROL_BUS_SWITCH; @@ -508,7 +508,7 @@ static void intel_sdvo_set_control_bus_switch(struct intel_output *intel_output, ret_value[0] = 0; ret_value[1] = 0; - ret = i2c_transfer(intel_output->i2c_bus, msgs, 3); + ret = i2c_transfer(intel_encoder->i2c_bus, msgs, 3); if (ret != 3) { /* failure in I2C transfer */ DRM_DEBUG_KMS("I2c transfer returned %d\n", ret); @@ -522,7 +522,7 @@ static void intel_sdvo_set_control_bus_switch(struct intel_output *intel_output, return; } -static bool intel_sdvo_set_target_input(struct intel_output *intel_output, bool target_0, bool target_1) +static bool intel_sdvo_set_target_input(struct intel_encoder *intel_encoder, bool target_0, bool target_1) { struct intel_sdvo_set_target_input_args targets = {0}; u8 status; @@ -533,10 +533,10 @@ static bool intel_sdvo_set_target_input(struct intel_output *intel_output, bool if (target_1) targets.target_1 = 1; - intel_sdvo_write_cmd(intel_output, SDVO_CMD_SET_TARGET_INPUT, &targets, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_TARGET_INPUT, &targets, sizeof(targets)); - status = intel_sdvo_read_response(intel_output, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); return (status == SDVO_CMD_STATUS_SUCCESS); } @@ -547,13 +547,13 @@ static bool intel_sdvo_set_target_input(struct intel_output *intel_output, bool * This function is making an assumption about the layout of the response, * which should be checked against the docs. */ -static bool intel_sdvo_get_trained_inputs(struct intel_output *intel_output, bool *input_1, bool *input_2) +static bool intel_sdvo_get_trained_inputs(struct intel_encoder *intel_encoder, bool *input_1, bool *input_2) { struct intel_sdvo_get_trained_inputs_response response; u8 status; - intel_sdvo_write_cmd(intel_output, SDVO_CMD_GET_TRAINED_INPUTS, NULL, 0); - status = intel_sdvo_read_response(intel_output, &response, sizeof(response)); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_TRAINED_INPUTS, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, &response, sizeof(response)); if (status != SDVO_CMD_STATUS_SUCCESS) return false; @@ -562,29 +562,29 @@ static bool intel_sdvo_get_trained_inputs(struct intel_output *intel_output, boo return true; } -static bool intel_sdvo_get_active_outputs(struct intel_output *intel_output, +static bool intel_sdvo_get_active_outputs(struct intel_encoder *intel_encoder, u16 *outputs) { u8 status; - intel_sdvo_write_cmd(intel_output, SDVO_CMD_GET_ACTIVE_OUTPUTS, NULL, 0); - status = intel_sdvo_read_response(intel_output, outputs, sizeof(*outputs)); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_ACTIVE_OUTPUTS, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, outputs, sizeof(*outputs)); return (status == SDVO_CMD_STATUS_SUCCESS); } -static bool intel_sdvo_set_active_outputs(struct intel_output *intel_output, +static bool intel_sdvo_set_active_outputs(struct intel_encoder *intel_encoder, u16 outputs) { u8 status; - intel_sdvo_write_cmd(intel_output, SDVO_CMD_SET_ACTIVE_OUTPUTS, &outputs, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_ACTIVE_OUTPUTS, &outputs, sizeof(outputs)); - status = intel_sdvo_read_response(intel_output, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); return (status == SDVO_CMD_STATUS_SUCCESS); } -static bool intel_sdvo_set_encoder_power_state(struct intel_output *intel_output, +static bool intel_sdvo_set_encoder_power_state(struct intel_encoder *intel_encoder, int mode) { u8 status, state = SDVO_ENCODER_STATE_ON; @@ -604,24 +604,24 @@ static bool intel_sdvo_set_encoder_power_state(struct intel_output *intel_output break; } - intel_sdvo_write_cmd(intel_output, SDVO_CMD_SET_ENCODER_POWER_STATE, &state, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_ENCODER_POWER_STATE, &state, sizeof(state)); - status = intel_sdvo_read_response(intel_output, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); return (status == SDVO_CMD_STATUS_SUCCESS); } -static bool intel_sdvo_get_input_pixel_clock_range(struct intel_output *intel_output, +static bool intel_sdvo_get_input_pixel_clock_range(struct intel_encoder *intel_encoder, int *clock_min, int *clock_max) { struct intel_sdvo_pixel_clock_range clocks; u8 status; - intel_sdvo_write_cmd(intel_output, SDVO_CMD_GET_INPUT_PIXEL_CLOCK_RANGE, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_INPUT_PIXEL_CLOCK_RANGE, NULL, 0); - status = intel_sdvo_read_response(intel_output, &clocks, sizeof(clocks)); + status = intel_sdvo_read_response(intel_encoder, &clocks, sizeof(clocks)); if (status != SDVO_CMD_STATUS_SUCCESS) return false; @@ -633,31 +633,31 @@ static bool intel_sdvo_get_input_pixel_clock_range(struct intel_output *intel_ou return true; } -static bool intel_sdvo_set_target_output(struct intel_output *intel_output, +static bool intel_sdvo_set_target_output(struct intel_encoder *intel_encoder, u16 outputs) { u8 status; - intel_sdvo_write_cmd(intel_output, SDVO_CMD_SET_TARGET_OUTPUT, &outputs, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_TARGET_OUTPUT, &outputs, sizeof(outputs)); - status = intel_sdvo_read_response(intel_output, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); return (status == SDVO_CMD_STATUS_SUCCESS); } -static bool intel_sdvo_get_timing(struct intel_output *intel_output, u8 cmd, +static bool intel_sdvo_get_timing(struct intel_encoder *intel_encoder, u8 cmd, struct intel_sdvo_dtd *dtd) { u8 status; - intel_sdvo_write_cmd(intel_output, cmd, NULL, 0); - status = intel_sdvo_read_response(intel_output, &dtd->part1, + intel_sdvo_write_cmd(intel_encoder, cmd, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, &dtd->part1, sizeof(dtd->part1)); if (status != SDVO_CMD_STATUS_SUCCESS) return false; - intel_sdvo_write_cmd(intel_output, cmd + 1, NULL, 0); - status = intel_sdvo_read_response(intel_output, &dtd->part2, + intel_sdvo_write_cmd(intel_encoder, cmd + 1, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, &dtd->part2, sizeof(dtd->part2)); if (status != SDVO_CMD_STATUS_SUCCESS) return false; @@ -665,54 +665,54 @@ static bool intel_sdvo_get_timing(struct intel_output *intel_output, u8 cmd, return true; } -static bool intel_sdvo_get_input_timing(struct intel_output *intel_output, +static bool intel_sdvo_get_input_timing(struct intel_encoder *intel_encoder, struct intel_sdvo_dtd *dtd) { - return intel_sdvo_get_timing(intel_output, + return intel_sdvo_get_timing(intel_encoder, SDVO_CMD_GET_INPUT_TIMINGS_PART1, dtd); } -static bool intel_sdvo_get_output_timing(struct intel_output *intel_output, +static bool intel_sdvo_get_output_timing(struct intel_encoder *intel_encoder, struct intel_sdvo_dtd *dtd) { - return intel_sdvo_get_timing(intel_output, + return intel_sdvo_get_timing(intel_encoder, SDVO_CMD_GET_OUTPUT_TIMINGS_PART1, dtd); } -static bool intel_sdvo_set_timing(struct intel_output *intel_output, u8 cmd, +static bool intel_sdvo_set_timing(struct intel_encoder *intel_encoder, u8 cmd, struct intel_sdvo_dtd *dtd) { u8 status; - intel_sdvo_write_cmd(intel_output, cmd, &dtd->part1, sizeof(dtd->part1)); - status = intel_sdvo_read_response(intel_output, NULL, 0); + intel_sdvo_write_cmd(intel_encoder, cmd, &dtd->part1, sizeof(dtd->part1)); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); if (status != SDVO_CMD_STATUS_SUCCESS) return false; - intel_sdvo_write_cmd(intel_output, cmd + 1, &dtd->part2, sizeof(dtd->part2)); - status = intel_sdvo_read_response(intel_output, NULL, 0); + intel_sdvo_write_cmd(intel_encoder, cmd + 1, &dtd->part2, sizeof(dtd->part2)); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); if (status != SDVO_CMD_STATUS_SUCCESS) return false; return true; } -static bool intel_sdvo_set_input_timing(struct intel_output *intel_output, +static bool intel_sdvo_set_input_timing(struct intel_encoder *intel_encoder, struct intel_sdvo_dtd *dtd) { - return intel_sdvo_set_timing(intel_output, + return intel_sdvo_set_timing(intel_encoder, SDVO_CMD_SET_INPUT_TIMINGS_PART1, dtd); } -static bool intel_sdvo_set_output_timing(struct intel_output *intel_output, +static bool intel_sdvo_set_output_timing(struct intel_encoder *intel_encoder, struct intel_sdvo_dtd *dtd) { - return intel_sdvo_set_timing(intel_output, + return intel_sdvo_set_timing(intel_encoder, SDVO_CMD_SET_OUTPUT_TIMINGS_PART1, dtd); } static bool -intel_sdvo_create_preferred_input_timing(struct intel_output *output, +intel_sdvo_create_preferred_input_timing(struct intel_encoder *output, uint16_t clock, uint16_t width, uint16_t height) @@ -741,7 +741,7 @@ intel_sdvo_create_preferred_input_timing(struct intel_output *output, return true; } -static bool intel_sdvo_get_preferred_input_timing(struct intel_output *output, +static bool intel_sdvo_get_preferred_input_timing(struct intel_encoder *output, struct intel_sdvo_dtd *dtd) { bool status; @@ -765,12 +765,12 @@ static bool intel_sdvo_get_preferred_input_timing(struct intel_output *output, return false; } -static int intel_sdvo_get_clock_rate_mult(struct intel_output *intel_output) +static int intel_sdvo_get_clock_rate_mult(struct intel_encoder *intel_encoder) { u8 response, status; - intel_sdvo_write_cmd(intel_output, SDVO_CMD_GET_CLOCK_RATE_MULT, NULL, 0); - status = intel_sdvo_read_response(intel_output, &response, 1); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_CLOCK_RATE_MULT, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, &response, 1); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Couldn't get SDVO clock rate multiplier\n"); @@ -782,12 +782,12 @@ static int intel_sdvo_get_clock_rate_mult(struct intel_output *intel_output) return response; } -static bool intel_sdvo_set_clock_rate_mult(struct intel_output *intel_output, u8 val) +static bool intel_sdvo_set_clock_rate_mult(struct intel_encoder *intel_encoder, u8 val) { u8 status; - intel_sdvo_write_cmd(intel_output, SDVO_CMD_SET_CLOCK_RATE_MULT, &val, 1); - status = intel_sdvo_read_response(intel_output, NULL, 0); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_CLOCK_RATE_MULT, &val, 1); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); if (status != SDVO_CMD_STATUS_SUCCESS) return false; @@ -876,7 +876,7 @@ static void intel_sdvo_get_mode_from_dtd(struct drm_display_mode * mode, mode->flags |= DRM_MODE_FLAG_PVSYNC; } -static bool intel_sdvo_get_supp_encode(struct intel_output *output, +static bool intel_sdvo_get_supp_encode(struct intel_encoder *output, struct intel_sdvo_encode *encode) { uint8_t status; @@ -891,7 +891,7 @@ static bool intel_sdvo_get_supp_encode(struct intel_output *output, return true; } -static bool intel_sdvo_set_encode(struct intel_output *output, uint8_t mode) +static bool intel_sdvo_set_encode(struct intel_encoder *output, uint8_t mode) { uint8_t status; @@ -901,7 +901,7 @@ static bool intel_sdvo_set_encode(struct intel_output *output, uint8_t mode) return (status == SDVO_CMD_STATUS_SUCCESS); } -static bool intel_sdvo_set_colorimetry(struct intel_output *output, +static bool intel_sdvo_set_colorimetry(struct intel_encoder *output, uint8_t mode) { uint8_t status; @@ -913,7 +913,7 @@ static bool intel_sdvo_set_colorimetry(struct intel_output *output, } #if 0 -static void intel_sdvo_dump_hdmi_buf(struct intel_output *output) +static void intel_sdvo_dump_hdmi_buf(struct intel_encoder *output) { int i, j; uint8_t set_buf_index[2]; @@ -943,7 +943,7 @@ static void intel_sdvo_dump_hdmi_buf(struct intel_output *output) } #endif -static void intel_sdvo_set_hdmi_buf(struct intel_output *output, int index, +static void intel_sdvo_set_hdmi_buf(struct intel_encoder *output, int index, uint8_t *data, int8_t size, uint8_t tx_rate) { uint8_t set_buf_index[2]; @@ -1033,7 +1033,7 @@ struct dip_infoframe { } __attribute__ ((packed)) u; } __attribute__((packed)); -static void intel_sdvo_set_avi_infoframe(struct intel_output *output, +static void intel_sdvo_set_avi_infoframe(struct intel_encoder *output, struct drm_display_mode * mode) { struct dip_infoframe avi_if = { @@ -1048,7 +1048,7 @@ static void intel_sdvo_set_avi_infoframe(struct intel_output *output, SDVO_HBUF_TX_VSYNC); } -static void intel_sdvo_set_tv_format(struct intel_output *output) +static void intel_sdvo_set_tv_format(struct intel_encoder *output) { struct intel_sdvo_tv_format format; @@ -1078,7 +1078,7 @@ static bool intel_sdvo_mode_fixup(struct drm_encoder *encoder, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode) { - struct intel_output *output = enc_to_intel_output(encoder); + struct intel_encoder *output = enc_to_intel_encoder(encoder); struct intel_sdvo_priv *dev_priv = output->dev_priv; if (dev_priv->is_tv) { @@ -1181,7 +1181,7 @@ static void intel_sdvo_mode_set(struct drm_encoder *encoder, struct drm_i915_private *dev_priv = dev->dev_private; struct drm_crtc *crtc = encoder->crtc; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_output *output = enc_to_intel_output(encoder); + struct intel_encoder *output = enc_to_intel_encoder(encoder); struct intel_sdvo_priv *sdvo_priv = output->dev_priv; u32 sdvox = 0; int sdvo_pixel_multiply; @@ -1305,19 +1305,19 @@ static void intel_sdvo_dpms(struct drm_encoder *encoder, int mode) { struct drm_device *dev = encoder->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_output *intel_output = enc_to_intel_output(encoder); - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; u32 temp; if (mode != DRM_MODE_DPMS_ON) { - intel_sdvo_set_active_outputs(intel_output, 0); + intel_sdvo_set_active_outputs(intel_encoder, 0); if (0) - intel_sdvo_set_encoder_power_state(intel_output, mode); + intel_sdvo_set_encoder_power_state(intel_encoder, mode); if (mode == DRM_MODE_DPMS_OFF) { temp = I915_READ(sdvo_priv->output_device); if ((temp & SDVO_ENABLE) != 0) { - intel_sdvo_write_sdvox(intel_output, temp & ~SDVO_ENABLE); + intel_sdvo_write_sdvox(intel_encoder, temp & ~SDVO_ENABLE); } } } else { @@ -1327,11 +1327,11 @@ static void intel_sdvo_dpms(struct drm_encoder *encoder, int mode) temp = I915_READ(sdvo_priv->output_device); if ((temp & SDVO_ENABLE) == 0) - intel_sdvo_write_sdvox(intel_output, temp | SDVO_ENABLE); + intel_sdvo_write_sdvox(intel_encoder, temp | SDVO_ENABLE); for (i = 0; i < 2; i++) intel_wait_for_vblank(dev); - status = intel_sdvo_get_trained_inputs(intel_output, &input1, + status = intel_sdvo_get_trained_inputs(intel_encoder, &input1, &input2); @@ -1345,8 +1345,8 @@ static void intel_sdvo_dpms(struct drm_encoder *encoder, int mode) } if (0) - intel_sdvo_set_encoder_power_state(intel_output, mode); - intel_sdvo_set_active_outputs(intel_output, sdvo_priv->controlled_output); + intel_sdvo_set_encoder_power_state(intel_encoder, mode); + intel_sdvo_set_active_outputs(intel_encoder, sdvo_priv->controlled_output); } return; } @@ -1355,22 +1355,22 @@ static void intel_sdvo_save(struct drm_connector *connector) { struct drm_device *dev = connector->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; int o; - sdvo_priv->save_sdvo_mult = intel_sdvo_get_clock_rate_mult(intel_output); - intel_sdvo_get_active_outputs(intel_output, &sdvo_priv->save_active_outputs); + sdvo_priv->save_sdvo_mult = intel_sdvo_get_clock_rate_mult(intel_encoder); + intel_sdvo_get_active_outputs(intel_encoder, &sdvo_priv->save_active_outputs); if (sdvo_priv->caps.sdvo_inputs_mask & 0x1) { - intel_sdvo_set_target_input(intel_output, true, false); - intel_sdvo_get_input_timing(intel_output, + intel_sdvo_set_target_input(intel_encoder, true, false); + intel_sdvo_get_input_timing(intel_encoder, &sdvo_priv->save_input_dtd_1); } if (sdvo_priv->caps.sdvo_inputs_mask & 0x2) { - intel_sdvo_set_target_input(intel_output, false, true); - intel_sdvo_get_input_timing(intel_output, + intel_sdvo_set_target_input(intel_encoder, false, true); + intel_sdvo_get_input_timing(intel_encoder, &sdvo_priv->save_input_dtd_2); } @@ -1379,8 +1379,8 @@ static void intel_sdvo_save(struct drm_connector *connector) u16 this_output = (1 << o); if (sdvo_priv->caps.output_flags & this_output) { - intel_sdvo_set_target_output(intel_output, this_output); - intel_sdvo_get_output_timing(intel_output, + intel_sdvo_set_target_output(intel_encoder, this_output); + intel_sdvo_get_output_timing(intel_encoder, &sdvo_priv->save_output_dtd[o]); } } @@ -1394,60 +1394,60 @@ static void intel_sdvo_save(struct drm_connector *connector) static void intel_sdvo_restore(struct drm_connector *connector) { struct drm_device *dev = connector->dev; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; int o; int i; bool input1, input2; u8 status; - intel_sdvo_set_active_outputs(intel_output, 0); + intel_sdvo_set_active_outputs(intel_encoder, 0); for (o = SDVO_OUTPUT_FIRST; o <= SDVO_OUTPUT_LAST; o++) { u16 this_output = (1 << o); if (sdvo_priv->caps.output_flags & this_output) { - intel_sdvo_set_target_output(intel_output, this_output); - intel_sdvo_set_output_timing(intel_output, &sdvo_priv->save_output_dtd[o]); + intel_sdvo_set_target_output(intel_encoder, this_output); + intel_sdvo_set_output_timing(intel_encoder, &sdvo_priv->save_output_dtd[o]); } } if (sdvo_priv->caps.sdvo_inputs_mask & 0x1) { - intel_sdvo_set_target_input(intel_output, true, false); - intel_sdvo_set_input_timing(intel_output, &sdvo_priv->save_input_dtd_1); + intel_sdvo_set_target_input(intel_encoder, true, false); + intel_sdvo_set_input_timing(intel_encoder, &sdvo_priv->save_input_dtd_1); } if (sdvo_priv->caps.sdvo_inputs_mask & 0x2) { - intel_sdvo_set_target_input(intel_output, false, true); - intel_sdvo_set_input_timing(intel_output, &sdvo_priv->save_input_dtd_2); + intel_sdvo_set_target_input(intel_encoder, false, true); + intel_sdvo_set_input_timing(intel_encoder, &sdvo_priv->save_input_dtd_2); } - intel_sdvo_set_clock_rate_mult(intel_output, sdvo_priv->save_sdvo_mult); + intel_sdvo_set_clock_rate_mult(intel_encoder, sdvo_priv->save_sdvo_mult); if (sdvo_priv->is_tv) { /* XXX: Restore TV format/enhancements. */ } - intel_sdvo_write_sdvox(intel_output, sdvo_priv->save_SDVOX); + intel_sdvo_write_sdvox(intel_encoder, sdvo_priv->save_SDVOX); if (sdvo_priv->save_SDVOX & SDVO_ENABLE) { for (i = 0; i < 2; i++) intel_wait_for_vblank(dev); - status = intel_sdvo_get_trained_inputs(intel_output, &input1, &input2); + status = intel_sdvo_get_trained_inputs(intel_encoder, &input1, &input2); if (status == SDVO_CMD_STATUS_SUCCESS && !input1) DRM_DEBUG_KMS("First %s output reported failure to " "sync\n", SDVO_NAME(sdvo_priv)); } - intel_sdvo_set_active_outputs(intel_output, sdvo_priv->save_active_outputs); + intel_sdvo_set_active_outputs(intel_encoder, sdvo_priv->save_active_outputs); } static int intel_sdvo_mode_valid(struct drm_connector *connector, struct drm_display_mode *mode) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; if (mode->flags & DRM_MODE_FLAG_DBLSCAN) return MODE_NO_DBLESCAN; @@ -1472,12 +1472,12 @@ static int intel_sdvo_mode_valid(struct drm_connector *connector, return MODE_OK; } -static bool intel_sdvo_get_capabilities(struct intel_output *intel_output, struct intel_sdvo_caps *caps) +static bool intel_sdvo_get_capabilities(struct intel_encoder *intel_encoder, struct intel_sdvo_caps *caps) { u8 status; - intel_sdvo_write_cmd(intel_output, SDVO_CMD_GET_DEVICE_CAPS, NULL, 0); - status = intel_sdvo_read_response(intel_output, caps, sizeof(*caps)); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_DEVICE_CAPS, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, caps, sizeof(*caps)); if (status != SDVO_CMD_STATUS_SUCCESS) return false; @@ -1487,12 +1487,12 @@ static bool intel_sdvo_get_capabilities(struct intel_output *intel_output, struc struct drm_connector* intel_sdvo_find(struct drm_device *dev, int sdvoB) { struct drm_connector *connector = NULL; - struct intel_output *iout = NULL; + struct intel_encoder *iout = NULL; struct intel_sdvo_priv *sdvo; /* find the sdvo connector */ list_for_each_entry(connector, &dev->mode_config.connector_list, head) { - iout = to_intel_output(connector); + iout = to_intel_encoder(connector); if (iout->type != INTEL_OUTPUT_SDVO) continue; @@ -1514,16 +1514,16 @@ int intel_sdvo_supports_hotplug(struct drm_connector *connector) { u8 response[2]; u8 status; - struct intel_output *intel_output; + struct intel_encoder *intel_encoder; DRM_DEBUG_KMS("\n"); if (!connector) return 0; - intel_output = to_intel_output(connector); + intel_encoder = to_intel_encoder(connector); - intel_sdvo_write_cmd(intel_output, SDVO_CMD_GET_HOT_PLUG_SUPPORT, NULL, 0); - status = intel_sdvo_read_response(intel_output, &response, 2); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_HOT_PLUG_SUPPORT, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, &response, 2); if (response[0] !=0) return 1; @@ -1535,30 +1535,30 @@ void intel_sdvo_set_hotplug(struct drm_connector *connector, int on) { u8 response[2]; u8 status; - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); - intel_sdvo_write_cmd(intel_output, SDVO_CMD_GET_ACTIVE_HOT_PLUG, NULL, 0); - intel_sdvo_read_response(intel_output, &response, 2); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_ACTIVE_HOT_PLUG, NULL, 0); + intel_sdvo_read_response(intel_encoder, &response, 2); if (on) { - intel_sdvo_write_cmd(intel_output, SDVO_CMD_GET_HOT_PLUG_SUPPORT, NULL, 0); - status = intel_sdvo_read_response(intel_output, &response, 2); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_HOT_PLUG_SUPPORT, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, &response, 2); - intel_sdvo_write_cmd(intel_output, SDVO_CMD_SET_ACTIVE_HOT_PLUG, &response, 2); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_ACTIVE_HOT_PLUG, &response, 2); } else { response[0] = 0; response[1] = 0; - intel_sdvo_write_cmd(intel_output, SDVO_CMD_SET_ACTIVE_HOT_PLUG, &response, 2); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_ACTIVE_HOT_PLUG, &response, 2); } - intel_sdvo_write_cmd(intel_output, SDVO_CMD_GET_ACTIVE_HOT_PLUG, NULL, 0); - intel_sdvo_read_response(intel_output, &response, 2); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_ACTIVE_HOT_PLUG, NULL, 0); + intel_sdvo_read_response(intel_encoder, &response, 2); } static bool -intel_sdvo_multifunc_encoder(struct intel_output *intel_output) +intel_sdvo_multifunc_encoder(struct intel_encoder *intel_encoder) { - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; int caps = 0; if (sdvo_priv->caps.output_flags & @@ -1592,11 +1592,11 @@ static struct drm_connector * intel_find_analog_connector(struct drm_device *dev) { struct drm_connector *connector; - struct intel_output *intel_output; + struct intel_encoder *intel_encoder; list_for_each_entry(connector, &dev->mode_config.connector_list, head) { - intel_output = to_intel_output(connector); - if (intel_output->type == INTEL_OUTPUT_ANALOG) + intel_encoder = to_intel_encoder(connector); + if (intel_encoder->type == INTEL_OUTPUT_ANALOG) return connector; } return NULL; @@ -1621,16 +1621,16 @@ intel_analog_is_connected(struct drm_device *dev) enum drm_connector_status intel_sdvo_hdmi_sink_detect(struct drm_connector *connector, u16 response) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; enum drm_connector_status status = connector_status_connected; struct edid *edid = NULL; - edid = drm_get_edid(&intel_output->base, - intel_output->ddc_bus); + edid = drm_get_edid(&intel_encoder->base, + intel_encoder->ddc_bus); /* This is only applied to SDVO cards with multiple outputs */ - if (edid == NULL && intel_sdvo_multifunc_encoder(intel_output)) { + if (edid == NULL && intel_sdvo_multifunc_encoder(intel_encoder)) { uint8_t saved_ddc, temp_ddc; saved_ddc = sdvo_priv->ddc_bus; temp_ddc = sdvo_priv->ddc_bus >> 1; @@ -1640,8 +1640,8 @@ intel_sdvo_hdmi_sink_detect(struct drm_connector *connector, u16 response) */ while(temp_ddc > 1) { sdvo_priv->ddc_bus = temp_ddc; - edid = drm_get_edid(&intel_output->base, - intel_output->ddc_bus); + edid = drm_get_edid(&intel_encoder->base, + intel_encoder->ddc_bus); if (edid) { /* * When we can get the EDID, maybe it is the @@ -1660,8 +1660,8 @@ intel_sdvo_hdmi_sink_detect(struct drm_connector *connector, u16 response) */ if (edid == NULL && sdvo_priv->analog_ddc_bus && - !intel_analog_is_connected(intel_output->base.dev)) - edid = drm_get_edid(&intel_output->base, + !intel_analog_is_connected(intel_encoder->base.dev)) + edid = drm_get_edid(&intel_encoder->base, sdvo_priv->analog_ddc_bus); if (edid != NULL) { /* Don't report the output as connected if it's a DVI-I @@ -1676,7 +1676,7 @@ intel_sdvo_hdmi_sink_detect(struct drm_connector *connector, u16 response) } kfree(edid); - intel_output->base.display_info.raw_edid = NULL; + intel_encoder->base.display_info.raw_edid = NULL; } else if (response & (SDVO_OUTPUT_TMDS0 | SDVO_OUTPUT_TMDS1)) status = connector_status_disconnected; @@ -1688,16 +1688,16 @@ static enum drm_connector_status intel_sdvo_detect(struct drm_connector *connect { uint16_t response; u8 status; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_ATTACHED_DISPLAYS, NULL, 0); if (sdvo_priv->is_tv) { /* add 30ms delay when the output type is SDVO-TV */ mdelay(30); } - status = intel_sdvo_read_response(intel_output, &response, 2); + status = intel_sdvo_read_response(intel_encoder, &response, 2); DRM_DEBUG_KMS("SDVO response %d %d\n", response & 0xff, response >> 8); @@ -1707,10 +1707,10 @@ static enum drm_connector_status intel_sdvo_detect(struct drm_connector *connect if (response == 0) return connector_status_disconnected; - if (intel_sdvo_multifunc_encoder(intel_output) && + if (intel_sdvo_multifunc_encoder(intel_encoder) && sdvo_priv->attached_output != response) { if (sdvo_priv->controlled_output != response && - intel_sdvo_output_setup(intel_output, response) != true) + intel_sdvo_output_setup(intel_encoder, response) != true) return connector_status_unknown; sdvo_priv->attached_output = response; } @@ -1719,12 +1719,12 @@ static enum drm_connector_status intel_sdvo_detect(struct drm_connector *connect static void intel_sdvo_get_ddc_modes(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; int num_modes; /* set the bus switch and get the modes */ - num_modes = intel_ddc_get_modes(intel_output); + num_modes = intel_ddc_get_modes(intel_encoder); /* * Mac mini hack. On this device, the DVI-I connector shares one DDC @@ -1734,17 +1734,17 @@ static void intel_sdvo_get_ddc_modes(struct drm_connector *connector) */ if (num_modes == 0 && sdvo_priv->analog_ddc_bus && - !intel_analog_is_connected(intel_output->base.dev)) { + !intel_analog_is_connected(intel_encoder->base.dev)) { struct i2c_adapter *digital_ddc_bus; /* Switch to the analog ddc bus and try that */ - digital_ddc_bus = intel_output->ddc_bus; - intel_output->ddc_bus = sdvo_priv->analog_ddc_bus; + digital_ddc_bus = intel_encoder->ddc_bus; + intel_encoder->ddc_bus = sdvo_priv->analog_ddc_bus; - (void) intel_ddc_get_modes(intel_output); + (void) intel_ddc_get_modes(intel_encoder); - intel_output->ddc_bus = digital_ddc_bus; + intel_encoder->ddc_bus = digital_ddc_bus; } } @@ -1815,7 +1815,7 @@ struct drm_display_mode sdvo_tv_modes[] = { static void intel_sdvo_get_tv_modes(struct drm_connector *connector) { - struct intel_output *output = to_intel_output(connector); + struct intel_encoder *output = to_intel_encoder(connector); struct intel_sdvo_priv *sdvo_priv = output->dev_priv; struct intel_sdvo_sdtv_resolution_request tv_res; uint32_t reply = 0, format_map = 0; @@ -1857,9 +1857,9 @@ static void intel_sdvo_get_tv_modes(struct drm_connector *connector) static void intel_sdvo_get_lvds_modes(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); struct drm_i915_private *dev_priv = connector->dev->dev_private; - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; struct drm_display_mode *newmode; /* @@ -1867,7 +1867,7 @@ static void intel_sdvo_get_lvds_modes(struct drm_connector *connector) * Assume that the preferred modes are * arranged in priority order. */ - intel_ddc_get_modes(intel_output); + intel_ddc_get_modes(intel_encoder); if (list_empty(&connector->probed_modes) == false) goto end; @@ -1896,7 +1896,7 @@ end: static int intel_sdvo_get_modes(struct drm_connector *connector) { - struct intel_output *output = to_intel_output(connector); + struct intel_encoder *output = to_intel_encoder(connector); struct intel_sdvo_priv *sdvo_priv = output->dev_priv; if (sdvo_priv->is_tv) @@ -1914,8 +1914,8 @@ static int intel_sdvo_get_modes(struct drm_connector *connector) static void intel_sdvo_destroy_enhance_property(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; struct drm_device *dev = connector->dev; if (sdvo_priv->is_tv) { @@ -1952,13 +1952,13 @@ void intel_sdvo_destroy_enhance_property(struct drm_connector *connector) static void intel_sdvo_destroy(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; - if (intel_output->i2c_bus) - intel_i2c_destroy(intel_output->i2c_bus); - if (intel_output->ddc_bus) - intel_i2c_destroy(intel_output->ddc_bus); + if (intel_encoder->i2c_bus) + intel_i2c_destroy(intel_encoder->i2c_bus); + if (intel_encoder->ddc_bus) + intel_i2c_destroy(intel_encoder->ddc_bus); if (sdvo_priv->analog_ddc_bus) intel_i2c_destroy(sdvo_priv->analog_ddc_bus); @@ -1976,7 +1976,7 @@ static void intel_sdvo_destroy(struct drm_connector *connector) drm_sysfs_connector_remove(connector); drm_connector_cleanup(connector); - kfree(intel_output); + kfree(intel_encoder); } static int @@ -1984,9 +1984,9 @@ intel_sdvo_set_property(struct drm_connector *connector, struct drm_property *property, uint64_t val) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; - struct drm_encoder *encoder = &intel_output->enc; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; + struct drm_encoder *encoder = &intel_encoder->enc; struct drm_crtc *crtc = encoder->crtc; int ret = 0; bool changed = false; @@ -2094,8 +2094,8 @@ intel_sdvo_set_property(struct drm_connector *connector, sdvo_priv->cur_brightness = temp_value; } if (cmd) { - intel_sdvo_write_cmd(intel_output, cmd, &temp_value, 2); - status = intel_sdvo_read_response(intel_output, + intel_sdvo_write_cmd(intel_encoder, cmd, &temp_value, 2); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO command \n"); @@ -2190,7 +2190,7 @@ intel_sdvo_select_ddc_bus(struct intel_sdvo_priv *dev_priv) } static bool -intel_sdvo_get_digital_encoding_mode(struct intel_output *output) +intel_sdvo_get_digital_encoding_mode(struct intel_encoder *output) { struct intel_sdvo_priv *sdvo_priv = output->dev_priv; uint8_t status; @@ -2204,42 +2204,42 @@ intel_sdvo_get_digital_encoding_mode(struct intel_output *output) return true; } -static struct intel_output * -intel_sdvo_chan_to_intel_output(struct intel_i2c_chan *chan) +static struct intel_encoder * +intel_sdvo_chan_to_intel_encoder(struct intel_i2c_chan *chan) { struct drm_device *dev = chan->drm_dev; struct drm_connector *connector; - struct intel_output *intel_output = NULL; + struct intel_encoder *intel_encoder = NULL; list_for_each_entry(connector, &dev->mode_config.connector_list, head) { - if (to_intel_output(connector)->ddc_bus == &chan->adapter) { - intel_output = to_intel_output(connector); + if (to_intel_encoder(connector)->ddc_bus == &chan->adapter) { + intel_encoder = to_intel_encoder(connector); break; } } - return intel_output; + return intel_encoder; } static int intel_sdvo_master_xfer(struct i2c_adapter *i2c_adap, struct i2c_msg msgs[], int num) { - struct intel_output *intel_output; + struct intel_encoder *intel_encoder; struct intel_sdvo_priv *sdvo_priv; struct i2c_algo_bit_data *algo_data; const struct i2c_algorithm *algo; algo_data = (struct i2c_algo_bit_data *)i2c_adap->algo_data; - intel_output = - intel_sdvo_chan_to_intel_output( + intel_encoder = + intel_sdvo_chan_to_intel_encoder( (struct intel_i2c_chan *)(algo_data->data)); - if (intel_output == NULL) + if (intel_encoder == NULL) return -EINVAL; - sdvo_priv = intel_output->dev_priv; - algo = intel_output->i2c_bus->algo; + sdvo_priv = intel_encoder->dev_priv; + algo = intel_encoder->i2c_bus->algo; - intel_sdvo_set_control_bus_switch(intel_output, sdvo_priv->ddc_bus); + intel_sdvo_set_control_bus_switch(intel_encoder, sdvo_priv->ddc_bus); return algo->master_xfer(i2c_adap, msgs, num); } @@ -2304,15 +2304,15 @@ static struct dmi_system_id intel_sdvo_bad_tv[] = { }; static bool -intel_sdvo_output_setup(struct intel_output *intel_output, uint16_t flags) +intel_sdvo_output_setup(struct intel_encoder *intel_encoder, uint16_t flags) { - struct drm_connector *connector = &intel_output->base; - struct drm_encoder *encoder = &intel_output->enc; - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct drm_connector *connector = &intel_encoder->base; + struct drm_encoder *encoder = &intel_encoder->enc; + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; bool ret = true, registered = false; sdvo_priv->is_tv = false; - intel_output->needs_tv_clock = false; + intel_encoder->needs_tv_clock = false; sdvo_priv->is_lvds = false; if (device_is_registered(&connector->kdev)) { @@ -2330,16 +2330,16 @@ intel_sdvo_output_setup(struct intel_output *intel_output, uint16_t flags) encoder->encoder_type = DRM_MODE_ENCODER_TMDS; connector->connector_type = DRM_MODE_CONNECTOR_DVID; - if (intel_sdvo_get_supp_encode(intel_output, + if (intel_sdvo_get_supp_encode(intel_encoder, &sdvo_priv->encode) && - intel_sdvo_get_digital_encoding_mode(intel_output) && + intel_sdvo_get_digital_encoding_mode(intel_encoder) && sdvo_priv->is_hdmi) { /* enable hdmi encoding mode if supported */ - intel_sdvo_set_encode(intel_output, SDVO_ENCODE_HDMI); - intel_sdvo_set_colorimetry(intel_output, + intel_sdvo_set_encode(intel_encoder, SDVO_ENCODE_HDMI); + intel_sdvo_set_colorimetry(intel_encoder, SDVO_COLORIMETRY_RGB256); connector->connector_type = DRM_MODE_CONNECTOR_HDMIA; - intel_output->clone_mask = + intel_encoder->clone_mask = (1 << INTEL_SDVO_NON_TV_CLONE_BIT) | (1 << INTEL_ANALOG_CLONE_BIT); } @@ -2350,21 +2350,21 @@ intel_sdvo_output_setup(struct intel_output *intel_output, uint16_t flags) encoder->encoder_type = DRM_MODE_ENCODER_TVDAC; connector->connector_type = DRM_MODE_CONNECTOR_SVIDEO; sdvo_priv->is_tv = true; - intel_output->needs_tv_clock = true; - intel_output->clone_mask = 1 << INTEL_SDVO_TV_CLONE_BIT; + intel_encoder->needs_tv_clock = true; + intel_encoder->clone_mask = 1 << INTEL_SDVO_TV_CLONE_BIT; } else if (flags & SDVO_OUTPUT_RGB0) { sdvo_priv->controlled_output = SDVO_OUTPUT_RGB0; encoder->encoder_type = DRM_MODE_ENCODER_DAC; connector->connector_type = DRM_MODE_CONNECTOR_VGA; - intel_output->clone_mask = (1 << INTEL_SDVO_NON_TV_CLONE_BIT) | + intel_encoder->clone_mask = (1 << INTEL_SDVO_NON_TV_CLONE_BIT) | (1 << INTEL_ANALOG_CLONE_BIT); } else if (flags & SDVO_OUTPUT_RGB1) { sdvo_priv->controlled_output = SDVO_OUTPUT_RGB1; encoder->encoder_type = DRM_MODE_ENCODER_DAC; connector->connector_type = DRM_MODE_CONNECTOR_VGA; - intel_output->clone_mask = (1 << INTEL_SDVO_NON_TV_CLONE_BIT) | + intel_encoder->clone_mask = (1 << INTEL_SDVO_NON_TV_CLONE_BIT) | (1 << INTEL_ANALOG_CLONE_BIT); } else if (flags & SDVO_OUTPUT_CVBS0) { @@ -2372,15 +2372,15 @@ intel_sdvo_output_setup(struct intel_output *intel_output, uint16_t flags) encoder->encoder_type = DRM_MODE_ENCODER_TVDAC; connector->connector_type = DRM_MODE_CONNECTOR_SVIDEO; sdvo_priv->is_tv = true; - intel_output->needs_tv_clock = true; - intel_output->clone_mask = 1 << INTEL_SDVO_TV_CLONE_BIT; + intel_encoder->needs_tv_clock = true; + intel_encoder->clone_mask = 1 << INTEL_SDVO_TV_CLONE_BIT; } else if (flags & SDVO_OUTPUT_LVDS0) { sdvo_priv->controlled_output = SDVO_OUTPUT_LVDS0; encoder->encoder_type = DRM_MODE_ENCODER_LVDS; connector->connector_type = DRM_MODE_CONNECTOR_LVDS; sdvo_priv->is_lvds = true; - intel_output->clone_mask = (1 << INTEL_ANALOG_CLONE_BIT) | + intel_encoder->clone_mask = (1 << INTEL_ANALOG_CLONE_BIT) | (1 << INTEL_SDVO_LVDS_CLONE_BIT); } else if (flags & SDVO_OUTPUT_LVDS1) { @@ -2388,7 +2388,7 @@ intel_sdvo_output_setup(struct intel_output *intel_output, uint16_t flags) encoder->encoder_type = DRM_MODE_ENCODER_LVDS; connector->connector_type = DRM_MODE_CONNECTOR_LVDS; sdvo_priv->is_lvds = true; - intel_output->clone_mask = (1 << INTEL_ANALOG_CLONE_BIT) | + intel_encoder->clone_mask = (1 << INTEL_ANALOG_CLONE_BIT) | (1 << INTEL_SDVO_LVDS_CLONE_BIT); } else { @@ -2401,7 +2401,7 @@ intel_sdvo_output_setup(struct intel_output *intel_output, uint16_t flags) bytes[0], bytes[1]); ret = false; } - intel_output->crtc_mask = (1 << 0) | (1 << 1); + intel_encoder->crtc_mask = (1 << 0) | (1 << 1); if (ret && registered) ret = drm_sysfs_connector_add(connector) == 0 ? true : false; @@ -2413,18 +2413,18 @@ intel_sdvo_output_setup(struct intel_output *intel_output, uint16_t flags) static void intel_sdvo_tv_create_property(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; struct intel_sdvo_tv_format format; uint32_t format_map, i; uint8_t status; - intel_sdvo_set_target_output(intel_output, + intel_sdvo_set_target_output(intel_encoder, sdvo_priv->controlled_output); - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_SUPPORTED_TV_FORMATS, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &format, sizeof(format)); if (status != SDVO_CMD_STATUS_SUCCESS) return; @@ -2462,16 +2462,16 @@ static void intel_sdvo_tv_create_property(struct drm_connector *connector) static void intel_sdvo_create_enhance_property(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; struct intel_sdvo_enhancements_reply sdvo_data; struct drm_device *dev = connector->dev; uint8_t status; uint16_t response, data_value[2]; - intel_sdvo_write_cmd(intel_output, SDVO_CMD_GET_SUPPORTED_ENHANCEMENTS, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_SUPPORTED_ENHANCEMENTS, NULL, 0); - status = intel_sdvo_read_response(intel_output, &sdvo_data, + status = intel_sdvo_read_response(intel_encoder, &sdvo_data, sizeof(sdvo_data)); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS(" incorrect response is returned\n"); @@ -2487,18 +2487,18 @@ static void intel_sdvo_create_enhance_property(struct drm_connector *connector) * property */ if (sdvo_data.overscan_h) { - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_MAX_OVERSCAN_H, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &data_value, 4); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO max " "h_overscan\n"); return; } - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_OVERSCAN_H, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &response, 2); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO h_overscan\n"); @@ -2528,18 +2528,18 @@ static void intel_sdvo_create_enhance_property(struct drm_connector *connector) data_value[0], data_value[1], response); } if (sdvo_data.overscan_v) { - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_MAX_OVERSCAN_V, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &data_value, 4); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO max " "v_overscan\n"); return; } - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_OVERSCAN_V, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &response, 2); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO v_overscan\n"); @@ -2569,17 +2569,17 @@ static void intel_sdvo_create_enhance_property(struct drm_connector *connector) data_value[0], data_value[1], response); } if (sdvo_data.position_h) { - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_MAX_POSITION_H, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &data_value, 4); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO Max h_pos\n"); return; } - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_POSITION_H, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &response, 2); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO get h_postion\n"); @@ -2600,17 +2600,17 @@ static void intel_sdvo_create_enhance_property(struct drm_connector *connector) data_value[0], data_value[1], response); } if (sdvo_data.position_v) { - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_MAX_POSITION_V, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &data_value, 4); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO Max v_pos\n"); return; } - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_POSITION_V, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &response, 2); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO get v_postion\n"); @@ -2633,17 +2633,17 @@ static void intel_sdvo_create_enhance_property(struct drm_connector *connector) } if (sdvo_priv->is_tv) { if (sdvo_data.saturation) { - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_MAX_SATURATION, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &data_value, 4); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO Max sat\n"); return; } - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_SATURATION, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &response, 2); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO get sat\n"); @@ -2665,17 +2665,17 @@ static void intel_sdvo_create_enhance_property(struct drm_connector *connector) data_value[0], data_value[1], response); } if (sdvo_data.contrast) { - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_MAX_CONTRAST, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &data_value, 4); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO Max contrast\n"); return; } - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_CONTRAST, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &response, 2); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO get contrast\n"); @@ -2696,17 +2696,17 @@ static void intel_sdvo_create_enhance_property(struct drm_connector *connector) data_value[0], data_value[1], response); } if (sdvo_data.hue) { - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_MAX_HUE, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &data_value, 4); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO Max hue\n"); return; } - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_HUE, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &response, 2); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO get hue\n"); @@ -2729,17 +2729,17 @@ static void intel_sdvo_create_enhance_property(struct drm_connector *connector) } if (sdvo_priv->is_tv || sdvo_priv->is_lvds) { if (sdvo_data.brightness) { - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_MAX_BRIGHTNESS, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &data_value, 4); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO Max bright\n"); return; } - intel_sdvo_write_cmd(intel_output, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_BRIGHTNESS, NULL, 0); - status = intel_sdvo_read_response(intel_output, + status = intel_sdvo_read_response(intel_encoder, &response, 2); if (status != SDVO_CMD_STATUS_SUCCESS) { DRM_DEBUG_KMS("Incorrect SDVO get brigh\n"); @@ -2768,40 +2768,40 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) { struct drm_i915_private *dev_priv = dev->dev_private; struct drm_connector *connector; - struct intel_output *intel_output; + struct intel_encoder *intel_encoder; struct intel_sdvo_priv *sdvo_priv; u8 ch[0x40]; int i; - intel_output = kcalloc(sizeof(struct intel_output)+sizeof(struct intel_sdvo_priv), 1, GFP_KERNEL); - if (!intel_output) { + intel_encoder = kcalloc(sizeof(struct intel_encoder)+sizeof(struct intel_sdvo_priv), 1, GFP_KERNEL); + if (!intel_encoder) { return false; } - sdvo_priv = (struct intel_sdvo_priv *)(intel_output + 1); + sdvo_priv = (struct intel_sdvo_priv *)(intel_encoder + 1); sdvo_priv->output_device = output_device; - intel_output->dev_priv = sdvo_priv; - intel_output->type = INTEL_OUTPUT_SDVO; + intel_encoder->dev_priv = sdvo_priv; + intel_encoder->type = INTEL_OUTPUT_SDVO; /* setup the DDC bus. */ if (output_device == SDVOB) - intel_output->i2c_bus = intel_i2c_create(dev, GPIOE, "SDVOCTRL_E for SDVOB"); + intel_encoder->i2c_bus = intel_i2c_create(dev, GPIOE, "SDVOCTRL_E for SDVOB"); else - intel_output->i2c_bus = intel_i2c_create(dev, GPIOE, "SDVOCTRL_E for SDVOC"); + intel_encoder->i2c_bus = intel_i2c_create(dev, GPIOE, "SDVOCTRL_E for SDVOC"); - if (!intel_output->i2c_bus) + if (!intel_encoder->i2c_bus) goto err_inteloutput; sdvo_priv->slave_addr = intel_sdvo_get_slave_addr(dev, output_device); /* Save the bit-banging i2c functionality for use by the DDC wrapper */ - intel_sdvo_i2c_bit_algo.functionality = intel_output->i2c_bus->algo->functionality; + intel_sdvo_i2c_bit_algo.functionality = intel_encoder->i2c_bus->algo->functionality; /* Read the regs to test if we can talk to the device */ for (i = 0; i < 0x40; i++) { - if (!intel_sdvo_read_byte(intel_output, i, &ch[i])) { + if (!intel_sdvo_read_byte(intel_encoder, i, &ch[i])) { DRM_DEBUG_KMS("No SDVO device found on SDVO%c\n", output_device == SDVOB ? 'B' : 'C'); goto err_i2c; @@ -2810,27 +2810,27 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) /* setup the DDC bus. */ if (output_device == SDVOB) { - intel_output->ddc_bus = intel_i2c_create(dev, GPIOE, "SDVOB DDC BUS"); + intel_encoder->ddc_bus = intel_i2c_create(dev, GPIOE, "SDVOB DDC BUS"); sdvo_priv->analog_ddc_bus = intel_i2c_create(dev, GPIOA, "SDVOB/VGA DDC BUS"); dev_priv->hotplug_supported_mask |= SDVOB_HOTPLUG_INT_STATUS; } else { - intel_output->ddc_bus = intel_i2c_create(dev, GPIOE, "SDVOC DDC BUS"); + intel_encoder->ddc_bus = intel_i2c_create(dev, GPIOE, "SDVOC DDC BUS"); sdvo_priv->analog_ddc_bus = intel_i2c_create(dev, GPIOA, "SDVOC/VGA DDC BUS"); dev_priv->hotplug_supported_mask |= SDVOC_HOTPLUG_INT_STATUS; } - if (intel_output->ddc_bus == NULL) + if (intel_encoder->ddc_bus == NULL) goto err_i2c; /* Wrap with our custom algo which switches to DDC mode */ - intel_output->ddc_bus->algo = &intel_sdvo_i2c_bit_algo; + intel_encoder->ddc_bus->algo = &intel_sdvo_i2c_bit_algo; /* In default case sdvo lvds is false */ - intel_sdvo_get_capabilities(intel_output, &sdvo_priv->caps); + intel_sdvo_get_capabilities(intel_encoder, &sdvo_priv->caps); - if (intel_sdvo_output_setup(intel_output, + if (intel_sdvo_output_setup(intel_encoder, sdvo_priv->caps.output_flags) != true) { DRM_DEBUG_KMS("SDVO output failed to setup on SDVO%c\n", output_device == SDVOB ? 'B' : 'C'); @@ -2838,7 +2838,7 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) } - connector = &intel_output->base; + connector = &intel_encoder->base; drm_connector_init(dev, connector, &intel_sdvo_connector_funcs, connector->connector_type); @@ -2847,12 +2847,12 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) connector->doublescan_allowed = 0; connector->display_info.subpixel_order = SubPixelHorizontalRGB; - drm_encoder_init(dev, &intel_output->enc, - &intel_sdvo_enc_funcs, intel_output->enc.encoder_type); + drm_encoder_init(dev, &intel_encoder->enc, + &intel_sdvo_enc_funcs, intel_encoder->enc.encoder_type); - drm_encoder_helper_add(&intel_output->enc, &intel_sdvo_helper_funcs); + drm_encoder_helper_add(&intel_encoder->enc, &intel_sdvo_helper_funcs); - drm_mode_connector_attach_encoder(&intel_output->base, &intel_output->enc); + drm_mode_connector_attach_encoder(&intel_encoder->base, &intel_encoder->enc); if (sdvo_priv->is_tv) intel_sdvo_tv_create_property(connector); @@ -2864,9 +2864,9 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) intel_sdvo_select_ddc_bus(sdvo_priv); /* Set the input timing to the screen. Assume always input 0. */ - intel_sdvo_set_target_input(intel_output, true, false); + intel_sdvo_set_target_input(intel_encoder, true, false); - intel_sdvo_get_input_pixel_clock_range(intel_output, + intel_sdvo_get_input_pixel_clock_range(intel_encoder, &sdvo_priv->pixel_clock_min, &sdvo_priv->pixel_clock_max); @@ -2893,12 +2893,12 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) err_i2c: if (sdvo_priv->analog_ddc_bus != NULL) intel_i2c_destroy(sdvo_priv->analog_ddc_bus); - if (intel_output->ddc_bus != NULL) - intel_i2c_destroy(intel_output->ddc_bus); - if (intel_output->i2c_bus != NULL) - intel_i2c_destroy(intel_output->i2c_bus); + if (intel_encoder->ddc_bus != NULL) + intel_i2c_destroy(intel_encoder->ddc_bus); + if (intel_encoder->i2c_bus != NULL) + intel_i2c_destroy(intel_encoder->i2c_bus); err_inteloutput: - kfree(intel_output); + kfree(intel_encoder); return false; } diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c index 552ec11..d7d39b2 100644 --- a/drivers/gpu/drm/i915/intel_tv.c +++ b/drivers/gpu/drm/i915/intel_tv.c @@ -921,8 +921,8 @@ intel_tv_save(struct drm_connector *connector) { struct drm_device *dev = connector->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_tv_priv *tv_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_tv_priv *tv_priv = intel_encoder->dev_priv; int i; tv_priv->save_TV_H_CTL_1 = I915_READ(TV_H_CTL_1); @@ -971,8 +971,8 @@ intel_tv_restore(struct drm_connector *connector) { struct drm_device *dev = connector->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_tv_priv *tv_priv = intel_output->dev_priv; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_tv_priv *tv_priv = intel_encoder->dev_priv; struct drm_crtc *crtc = connector->encoder->crtc; struct intel_crtc *intel_crtc; int i; @@ -1068,9 +1068,9 @@ intel_tv_mode_lookup (char *tv_format) } static const struct tv_mode * -intel_tv_mode_find (struct intel_output *intel_output) +intel_tv_mode_find (struct intel_encoder *intel_encoder) { - struct intel_tv_priv *tv_priv = intel_output->dev_priv; + struct intel_tv_priv *tv_priv = intel_encoder->dev_priv; return intel_tv_mode_lookup(tv_priv->tv_format); } @@ -1078,8 +1078,8 @@ intel_tv_mode_find (struct intel_output *intel_output) static enum drm_mode_status intel_tv_mode_valid(struct drm_connector *connector, struct drm_display_mode *mode) { - struct intel_output *intel_output = to_intel_output(connector); - const struct tv_mode *tv_mode = intel_tv_mode_find(intel_output); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + const struct tv_mode *tv_mode = intel_tv_mode_find(intel_encoder); /* Ensure TV refresh is close to desired refresh */ if (tv_mode && abs(tv_mode->refresh - drm_mode_vrefresh(mode) * 1000) @@ -1095,8 +1095,8 @@ intel_tv_mode_fixup(struct drm_encoder *encoder, struct drm_display_mode *mode, { struct drm_device *dev = encoder->dev; struct drm_mode_config *drm_config = &dev->mode_config; - struct intel_output *intel_output = enc_to_intel_output(encoder); - const struct tv_mode *tv_mode = intel_tv_mode_find (intel_output); + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + const struct tv_mode *tv_mode = intel_tv_mode_find (intel_encoder); struct drm_encoder *other_encoder; if (!tv_mode) @@ -1121,9 +1121,9 @@ intel_tv_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode, struct drm_i915_private *dev_priv = dev->dev_private; struct drm_crtc *crtc = encoder->crtc; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_output *intel_output = enc_to_intel_output(encoder); - struct intel_tv_priv *tv_priv = intel_output->dev_priv; - const struct tv_mode *tv_mode = intel_tv_mode_find(intel_output); + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_tv_priv *tv_priv = intel_encoder->dev_priv; + const struct tv_mode *tv_mode = intel_tv_mode_find(intel_encoder); u32 tv_ctl; u32 hctl1, hctl2, hctl3; u32 vctl1, vctl2, vctl3, vctl4, vctl5, vctl6, vctl7; @@ -1360,9 +1360,9 @@ static const struct drm_display_mode reported_modes[] = { * \return false if TV is disconnected. */ static int -intel_tv_detect_type (struct drm_crtc *crtc, struct intel_output *intel_output) +intel_tv_detect_type (struct drm_crtc *crtc, struct intel_encoder *intel_encoder) { - struct drm_encoder *encoder = &intel_output->enc; + struct drm_encoder *encoder = &intel_encoder->enc; struct drm_device *dev = encoder->dev; struct drm_i915_private *dev_priv = dev->dev_private; unsigned long irqflags; @@ -1441,9 +1441,9 @@ intel_tv_detect_type (struct drm_crtc *crtc, struct intel_output *intel_output) */ static void intel_tv_find_better_format(struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); - struct intel_tv_priv *tv_priv = intel_output->dev_priv; - const struct tv_mode *tv_mode = intel_tv_mode_find(intel_output); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_tv_priv *tv_priv = intel_encoder->dev_priv; + const struct tv_mode *tv_mode = intel_tv_mode_find(intel_encoder); int i; if ((tv_priv->type == DRM_MODE_CONNECTOR_Component) == @@ -1475,9 +1475,9 @@ intel_tv_detect(struct drm_connector *connector) { struct drm_crtc *crtc; struct drm_display_mode mode; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_tv_priv *tv_priv = intel_output->dev_priv; - struct drm_encoder *encoder = &intel_output->enc; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_tv_priv *tv_priv = intel_encoder->dev_priv; + struct drm_encoder *encoder = &intel_encoder->enc; int dpms_mode; int type = tv_priv->type; @@ -1485,12 +1485,12 @@ intel_tv_detect(struct drm_connector *connector) drm_mode_set_crtcinfo(&mode, CRTC_INTERLACE_HALVE_V); if (encoder->crtc && encoder->crtc->enabled) { - type = intel_tv_detect_type(encoder->crtc, intel_output); + type = intel_tv_detect_type(encoder->crtc, intel_encoder); } else { - crtc = intel_get_load_detect_pipe(intel_output, &mode, &dpms_mode); + crtc = intel_get_load_detect_pipe(intel_encoder, &mode, &dpms_mode); if (crtc) { - type = intel_tv_detect_type(crtc, intel_output); - intel_release_load_detect_pipe(intel_output, dpms_mode); + type = intel_tv_detect_type(crtc, intel_encoder); + intel_release_load_detect_pipe(intel_encoder, dpms_mode); } else type = -1; } @@ -1525,8 +1525,8 @@ static void intel_tv_chose_preferred_modes(struct drm_connector *connector, struct drm_display_mode *mode_ptr) { - struct intel_output *intel_output = to_intel_output(connector); - const struct tv_mode *tv_mode = intel_tv_mode_find(intel_output); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + const struct tv_mode *tv_mode = intel_tv_mode_find(intel_encoder); if (tv_mode->nbr_end < 480 && mode_ptr->vdisplay == 480) mode_ptr->type |= DRM_MODE_TYPE_PREFERRED; @@ -1550,8 +1550,8 @@ static int intel_tv_get_modes(struct drm_connector *connector) { struct drm_display_mode *mode_ptr; - struct intel_output *intel_output = to_intel_output(connector); - const struct tv_mode *tv_mode = intel_tv_mode_find(intel_output); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + const struct tv_mode *tv_mode = intel_tv_mode_find(intel_encoder); int j, count = 0; u64 tmp; @@ -1604,11 +1604,11 @@ intel_tv_get_modes(struct drm_connector *connector) static void intel_tv_destroy (struct drm_connector *connector) { - struct intel_output *intel_output = to_intel_output(connector); + struct intel_encoder *intel_encoder = to_intel_encoder(connector); drm_sysfs_connector_remove(connector); drm_connector_cleanup(connector); - kfree(intel_output); + kfree(intel_encoder); } @@ -1617,9 +1617,9 @@ intel_tv_set_property(struct drm_connector *connector, struct drm_property *prop uint64_t val) { struct drm_device *dev = connector->dev; - struct intel_output *intel_output = to_intel_output(connector); - struct intel_tv_priv *tv_priv = intel_output->dev_priv; - struct drm_encoder *encoder = &intel_output->enc; + struct intel_encoder *intel_encoder = to_intel_encoder(connector); + struct intel_tv_priv *tv_priv = intel_encoder->dev_priv; + struct drm_encoder *encoder = &intel_encoder->enc; struct drm_crtc *crtc = encoder->crtc; int ret = 0; bool changed = false; @@ -1740,7 +1740,7 @@ intel_tv_init(struct drm_device *dev) { struct drm_i915_private *dev_priv = dev->dev_private; struct drm_connector *connector; - struct intel_output *intel_output; + struct intel_encoder *intel_encoder; struct intel_tv_priv *tv_priv; u32 tv_dac_on, tv_dac_off, save_tv_dac; char **tv_format_names; @@ -1780,28 +1780,28 @@ intel_tv_init(struct drm_device *dev) (tv_dac_off & TVDAC_STATE_CHG_EN) != 0) return; - intel_output = kzalloc(sizeof(struct intel_output) + + intel_encoder = kzalloc(sizeof(struct intel_encoder) + sizeof(struct intel_tv_priv), GFP_KERNEL); - if (!intel_output) { + if (!intel_encoder) { return; } - connector = &intel_output->base; + connector = &intel_encoder->base; drm_connector_init(dev, connector, &intel_tv_connector_funcs, DRM_MODE_CONNECTOR_SVIDEO); - drm_encoder_init(dev, &intel_output->enc, &intel_tv_enc_funcs, + drm_encoder_init(dev, &intel_encoder->enc, &intel_tv_enc_funcs, DRM_MODE_ENCODER_TVDAC); - drm_mode_connector_attach_encoder(&intel_output->base, &intel_output->enc); - tv_priv = (struct intel_tv_priv *)(intel_output + 1); - intel_output->type = INTEL_OUTPUT_TVOUT; - intel_output->crtc_mask = (1 << 0) | (1 << 1); - intel_output->clone_mask = (1 << INTEL_TV_CLONE_BIT); - intel_output->enc.possible_crtcs = ((1 << 0) | (1 << 1)); - intel_output->enc.possible_clones = (1 << INTEL_OUTPUT_TVOUT); - intel_output->dev_priv = tv_priv; + drm_mode_connector_attach_encoder(&intel_encoder->base, &intel_encoder->enc); + tv_priv = (struct intel_tv_priv *)(intel_encoder + 1); + intel_encoder->type = INTEL_OUTPUT_TVOUT; + intel_encoder->crtc_mask = (1 << 0) | (1 << 1); + intel_encoder->clone_mask = (1 << INTEL_TV_CLONE_BIT); + intel_encoder->enc.possible_crtcs = ((1 << 0) | (1 << 1)); + intel_encoder->enc.possible_clones = (1 << INTEL_OUTPUT_TVOUT); + intel_encoder->dev_priv = tv_priv; tv_priv->type = DRM_MODE_CONNECTOR_Unknown; /* BIOS margin values */ @@ -1812,7 +1812,7 @@ intel_tv_init(struct drm_device *dev) tv_priv->tv_format = kstrdup(tv_modes[initial_mode].name, GFP_KERNEL); - drm_encoder_helper_add(&intel_output->enc, &intel_tv_helper_funcs); + drm_encoder_helper_add(&intel_encoder->enc, &intel_tv_helper_funcs); drm_connector_helper_add(connector, &intel_tv_connector_helper_funcs); connector->interlace_allowed = false; connector->doublescan_allowed = false; -- cgit v0.10.2 From c751ce4f52b11ea93764a7cd44e6ae9c098d361b Mon Sep 17 00:00:00 2001 From: Eric Anholt Date: Thu, 25 Mar 2010 11:48:48 -0700 Subject: drm/i915: Rename many remaining uses of "output" to encoder or connector. Signed-off-by: Eric Anholt diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 2595c4c..34d2652 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -754,8 +754,8 @@ bool intel_pipe_has_type (struct drm_crtc *crtc, int type) return false; } -struct drm_connector * -intel_pipe_get_output (struct drm_crtc *crtc) +static struct drm_connector * +intel_pipe_get_connector (struct drm_crtc *crtc) { struct drm_device *dev = crtc->dev; struct drm_mode_config *mode_config = &dev->mode_config; @@ -2916,7 +2916,7 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc, int dspsize_reg = (plane == 0) ? DSPASIZE : DSPBSIZE; int dsppos_reg = (plane == 0) ? DSPAPOS : DSPBPOS; int pipesrc_reg = (pipe == 0) ? PIPEASRC : PIPEBSRC; - int refclk, num_outputs = 0; + int refclk, num_connectors = 0; intel_clock_t clock, reduced_clock; u32 dpll = 0, fp = 0, fp2 = 0, dspcntr, pipeconf; bool ok, has_reduced_clock = false, is_sdvo = false, is_dvo = false; @@ -2974,10 +2974,10 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc, break; } - num_outputs++; + num_connectors++; } - if (is_lvds && dev_priv->lvds_use_ssc && num_outputs < 2) { + if (is_lvds && dev_priv->lvds_use_ssc && num_connectors < 2) { refclk = dev_priv->lvds_ssc_freq * 1000; DRM_DEBUG_KMS("using SSC reference clock of %d MHz\n", refclk / 1000); @@ -3048,7 +3048,7 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc, if (is_edp) { struct drm_connector *edp; target_clock = mode->clock; - edp = intel_pipe_get_output(crtc); + edp = intel_pipe_get_connector(crtc); intel_edp_link_config(to_intel_encoder(edp), &lane, &link_bw); } else { @@ -3230,7 +3230,7 @@ static int intel_crtc_mode_set(struct drm_crtc *crtc, /* XXX: just matching BIOS for now */ /* dpll |= PLL_REF_INPUT_TVCLKINBC; */ dpll |= 3; - else if (is_lvds && dev_priv->lvds_use_ssc && num_outputs < 2) + else if (is_lvds && dev_priv->lvds_use_ssc && num_connectors < 2) dpll |= PLLB_REF_INPUT_SPREADSPECTRUMIN; else dpll |= PLL_REF_INPUT_DREFCLK; @@ -3654,9 +3654,9 @@ static void intel_crtc_gamma_set(struct drm_crtc *crtc, u16 *red, u16 *green, * detection. * * It will be up to the load-detect code to adjust the pipe as appropriate for - * its requirements. The pipe will be connected to no other outputs. + * its requirements. The pipe will be connected to no other encoders. * - * Currently this code will only succeed if there is a pipe with no outputs + * Currently this code will only succeed if there is a pipe with no encoders * configured for it. In the future, it could choose to temporarily disable * some outputs to free up a pipe for its use. * @@ -3770,7 +3770,7 @@ void intel_release_load_detect_pipe(struct intel_encoder *intel_encoder, int dpm drm_helper_disable_unused_functions(dev); } - /* Switch crtc and output back off if necessary */ + /* Switch crtc and encoder back off if necessary */ if (crtc->enabled && dpms_mode != DRM_MODE_DPMS_ON) { if (encoder->crtc == crtc) encoder_funcs->dpms(encoder, dpms_mode); diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c index ea6de3b..a5b049f 100644 --- a/drivers/gpu/drm/i915/intel_sdvo.c +++ b/drivers/gpu/drm/i915/intel_sdvo.c @@ -53,7 +53,7 @@ struct intel_sdvo_priv { u8 slave_addr; /* Register for the SDVO device: SDVOB or SDVOC */ - int output_device; + int sdvo_reg; /* Active outputs controlled by this SDVO output */ uint16_t controlled_output; @@ -123,7 +123,7 @@ struct intel_sdvo_priv { */ struct intel_sdvo_encode encode; - /* DDC bus used by this SDVO output */ + /* DDC bus used by this SDVO encoder */ uint8_t ddc_bus; /* Mac mini hack -- use the same DDC as the analog connector */ @@ -176,7 +176,7 @@ static void intel_sdvo_write_sdvox(struct intel_encoder *intel_encoder, u32 val) u32 bval = val, cval = val; int i; - if (sdvo_priv->output_device == SDVOB) { + if (sdvo_priv->sdvo_reg == SDVOB) { cval = I915_READ(SDVOC); } else { bval = I915_READ(SDVOB); @@ -352,8 +352,8 @@ static const struct _sdvo_cmd_name { SDVO_CMD_NAME_ENTRY(SDVO_CMD_GET_HBUF_DATA), }; -#define SDVO_NAME(dev_priv) ((dev_priv)->output_device == SDVOB ? "SDVOB" : "SDVOC") -#define SDVO_PRIV(output) ((struct intel_sdvo_priv *) (output)->dev_priv) +#define SDVO_NAME(dev_priv) ((dev_priv)->sdvo_reg == SDVOB ? "SDVOB" : "SDVOC") +#define SDVO_PRIV(encoder) ((struct intel_sdvo_priv *) (encoder)->dev_priv) static void intel_sdvo_debug_write(struct intel_encoder *intel_encoder, u8 cmd, void *args, int args_len) @@ -712,13 +712,13 @@ static bool intel_sdvo_set_output_timing(struct intel_encoder *intel_encoder, } static bool -intel_sdvo_create_preferred_input_timing(struct intel_encoder *output, +intel_sdvo_create_preferred_input_timing(struct intel_encoder *intel_encoder, uint16_t clock, uint16_t width, uint16_t height) { struct intel_sdvo_preferred_input_timing_args args; - struct intel_sdvo_priv *sdvo_priv = output->dev_priv; + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; uint8_t status; memset(&args, 0, sizeof(args)); @@ -732,32 +732,33 @@ intel_sdvo_create_preferred_input_timing(struct intel_encoder *output, sdvo_priv->sdvo_lvds_fixed_mode->vdisplay != height)) args.scaled = 1; - intel_sdvo_write_cmd(output, SDVO_CMD_CREATE_PREFERRED_INPUT_TIMING, + intel_sdvo_write_cmd(intel_encoder, + SDVO_CMD_CREATE_PREFERRED_INPUT_TIMING, &args, sizeof(args)); - status = intel_sdvo_read_response(output, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); if (status != SDVO_CMD_STATUS_SUCCESS) return false; return true; } -static bool intel_sdvo_get_preferred_input_timing(struct intel_encoder *output, +static bool intel_sdvo_get_preferred_input_timing(struct intel_encoder *intel_encoder, struct intel_sdvo_dtd *dtd) { bool status; - intel_sdvo_write_cmd(output, SDVO_CMD_GET_PREFERRED_INPUT_TIMING_PART1, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_PREFERRED_INPUT_TIMING_PART1, NULL, 0); - status = intel_sdvo_read_response(output, &dtd->part1, + status = intel_sdvo_read_response(intel_encoder, &dtd->part1, sizeof(dtd->part1)); if (status != SDVO_CMD_STATUS_SUCCESS) return false; - intel_sdvo_write_cmd(output, SDVO_CMD_GET_PREFERRED_INPUT_TIMING_PART2, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_PREFERRED_INPUT_TIMING_PART2, NULL, 0); - status = intel_sdvo_read_response(output, &dtd->part2, + status = intel_sdvo_read_response(intel_encoder, &dtd->part2, sizeof(dtd->part2)); if (status != SDVO_CMD_STATUS_SUCCESS) return false; @@ -876,13 +877,13 @@ static void intel_sdvo_get_mode_from_dtd(struct drm_display_mode * mode, mode->flags |= DRM_MODE_FLAG_PVSYNC; } -static bool intel_sdvo_get_supp_encode(struct intel_encoder *output, +static bool intel_sdvo_get_supp_encode(struct intel_encoder *intel_encoder, struct intel_sdvo_encode *encode) { uint8_t status; - intel_sdvo_write_cmd(output, SDVO_CMD_GET_SUPP_ENCODE, NULL, 0); - status = intel_sdvo_read_response(output, encode, sizeof(*encode)); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_GET_SUPP_ENCODE, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, encode, sizeof(*encode)); if (status != SDVO_CMD_STATUS_SUCCESS) { /* non-support means DVI */ memset(encode, 0, sizeof(*encode)); return false; @@ -891,29 +892,30 @@ static bool intel_sdvo_get_supp_encode(struct intel_encoder *output, return true; } -static bool intel_sdvo_set_encode(struct intel_encoder *output, uint8_t mode) +static bool intel_sdvo_set_encode(struct intel_encoder *intel_encoder, + uint8_t mode) { uint8_t status; - intel_sdvo_write_cmd(output, SDVO_CMD_SET_ENCODE, &mode, 1); - status = intel_sdvo_read_response(output, NULL, 0); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_ENCODE, &mode, 1); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); return (status == SDVO_CMD_STATUS_SUCCESS); } -static bool intel_sdvo_set_colorimetry(struct intel_encoder *output, +static bool intel_sdvo_set_colorimetry(struct intel_encoder *intel_encoder, uint8_t mode) { uint8_t status; - intel_sdvo_write_cmd(output, SDVO_CMD_SET_COLORIMETRY, &mode, 1); - status = intel_sdvo_read_response(output, NULL, 0); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_COLORIMETRY, &mode, 1); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); return (status == SDVO_CMD_STATUS_SUCCESS); } #if 0 -static void intel_sdvo_dump_hdmi_buf(struct intel_encoder *output) +static void intel_sdvo_dump_hdmi_buf(struct intel_encoder *intel_encoder) { int i, j; uint8_t set_buf_index[2]; @@ -922,43 +924,45 @@ static void intel_sdvo_dump_hdmi_buf(struct intel_encoder *output) uint8_t buf[48]; uint8_t *pos; - intel_sdvo_write_cmd(output, SDVO_CMD_GET_HBUF_AV_SPLIT, NULL, 0); - intel_sdvo_read_response(output, &av_split, 1); + intel_sdvo_write_cmd(encoder, SDVO_CMD_GET_HBUF_AV_SPLIT, NULL, 0); + intel_sdvo_read_response(encoder, &av_split, 1); for (i = 0; i <= av_split; i++) { set_buf_index[0] = i; set_buf_index[1] = 0; - intel_sdvo_write_cmd(output, SDVO_CMD_SET_HBUF_INDEX, + intel_sdvo_write_cmd(encoder, SDVO_CMD_SET_HBUF_INDEX, set_buf_index, 2); - intel_sdvo_write_cmd(output, SDVO_CMD_GET_HBUF_INFO, NULL, 0); - intel_sdvo_read_response(output, &buf_size, 1); + intel_sdvo_write_cmd(encoder, SDVO_CMD_GET_HBUF_INFO, NULL, 0); + intel_sdvo_read_response(encoder, &buf_size, 1); pos = buf; for (j = 0; j <= buf_size; j += 8) { - intel_sdvo_write_cmd(output, SDVO_CMD_GET_HBUF_DATA, + intel_sdvo_write_cmd(encoder, SDVO_CMD_GET_HBUF_DATA, NULL, 0); - intel_sdvo_read_response(output, pos, 8); + intel_sdvo_read_response(encoder, pos, 8); pos += 8; } } } #endif -static void intel_sdvo_set_hdmi_buf(struct intel_encoder *output, int index, - uint8_t *data, int8_t size, uint8_t tx_rate) +static void intel_sdvo_set_hdmi_buf(struct intel_encoder *intel_encoder, + int index, + uint8_t *data, int8_t size, uint8_t tx_rate) { uint8_t set_buf_index[2]; set_buf_index[0] = index; set_buf_index[1] = 0; - intel_sdvo_write_cmd(output, SDVO_CMD_SET_HBUF_INDEX, set_buf_index, 2); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_HBUF_INDEX, + set_buf_index, 2); for (; size > 0; size -= 8) { - intel_sdvo_write_cmd(output, SDVO_CMD_SET_HBUF_DATA, data, 8); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_HBUF_DATA, data, 8); data += 8; } - intel_sdvo_write_cmd(output, SDVO_CMD_SET_HBUF_TXRATE, &tx_rate, 1); + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_HBUF_TXRATE, &tx_rate, 1); } static uint8_t intel_sdvo_calc_hbuf_csum(uint8_t *data, uint8_t size) @@ -1033,7 +1037,7 @@ struct dip_infoframe { } __attribute__ ((packed)) u; } __attribute__((packed)); -static void intel_sdvo_set_avi_infoframe(struct intel_encoder *output, +static void intel_sdvo_set_avi_infoframe(struct intel_encoder *intel_encoder, struct drm_display_mode * mode) { struct dip_infoframe avi_if = { @@ -1044,15 +1048,16 @@ static void intel_sdvo_set_avi_infoframe(struct intel_encoder *output, avi_if.checksum = intel_sdvo_calc_hbuf_csum((uint8_t *)&avi_if, 4 + avi_if.len); - intel_sdvo_set_hdmi_buf(output, 1, (uint8_t *)&avi_if, 4 + avi_if.len, + intel_sdvo_set_hdmi_buf(intel_encoder, 1, (uint8_t *)&avi_if, + 4 + avi_if.len, SDVO_HBUF_TX_VSYNC); } -static void intel_sdvo_set_tv_format(struct intel_encoder *output) +static void intel_sdvo_set_tv_format(struct intel_encoder *intel_encoder) { struct intel_sdvo_tv_format format; - struct intel_sdvo_priv *sdvo_priv = output->dev_priv; + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; uint32_t format_map, i; uint8_t status; @@ -1065,10 +1070,10 @@ static void intel_sdvo_set_tv_format(struct intel_encoder *output) memcpy(&format, &format_map, sizeof(format_map) > sizeof(format) ? sizeof(format) : sizeof(format_map)); - intel_sdvo_write_cmd(output, SDVO_CMD_SET_TV_FORMAT, &format_map, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_TV_FORMAT, &format_map, sizeof(format)); - status = intel_sdvo_read_response(output, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); if (status != SDVO_CMD_STATUS_SUCCESS) DRM_DEBUG_KMS("%s: Failed to set TV format\n", SDVO_NAME(sdvo_priv)); @@ -1078,8 +1083,8 @@ static bool intel_sdvo_mode_fixup(struct drm_encoder *encoder, struct drm_display_mode *mode, struct drm_display_mode *adjusted_mode) { - struct intel_encoder *output = enc_to_intel_encoder(encoder); - struct intel_sdvo_priv *dev_priv = output->dev_priv; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_sdvo_priv *dev_priv = intel_encoder->dev_priv; if (dev_priv->is_tv) { struct intel_sdvo_dtd output_dtd; @@ -1094,22 +1099,22 @@ static bool intel_sdvo_mode_fixup(struct drm_encoder *encoder, /* Set output timings */ intel_sdvo_get_dtd_from_mode(&output_dtd, mode); - intel_sdvo_set_target_output(output, + intel_sdvo_set_target_output(intel_encoder, dev_priv->controlled_output); - intel_sdvo_set_output_timing(output, &output_dtd); + intel_sdvo_set_output_timing(intel_encoder, &output_dtd); /* Set the input timing to the screen. Assume always input 0. */ - intel_sdvo_set_target_input(output, true, false); + intel_sdvo_set_target_input(intel_encoder, true, false); - success = intel_sdvo_create_preferred_input_timing(output, + success = intel_sdvo_create_preferred_input_timing(intel_encoder, mode->clock / 10, mode->hdisplay, mode->vdisplay); if (success) { struct intel_sdvo_dtd input_dtd; - intel_sdvo_get_preferred_input_timing(output, + intel_sdvo_get_preferred_input_timing(intel_encoder, &input_dtd); intel_sdvo_get_mode_from_dtd(adjusted_mode, &input_dtd); dev_priv->sdvo_flags = input_dtd.part2.sdvo_flags; @@ -1132,16 +1137,16 @@ static bool intel_sdvo_mode_fixup(struct drm_encoder *encoder, intel_sdvo_get_dtd_from_mode(&output_dtd, dev_priv->sdvo_lvds_fixed_mode); - intel_sdvo_set_target_output(output, + intel_sdvo_set_target_output(intel_encoder, dev_priv->controlled_output); - intel_sdvo_set_output_timing(output, &output_dtd); + intel_sdvo_set_output_timing(intel_encoder, &output_dtd); /* Set the input timing to the screen. Assume always input 0. */ - intel_sdvo_set_target_input(output, true, false); + intel_sdvo_set_target_input(intel_encoder, true, false); success = intel_sdvo_create_preferred_input_timing( - output, + intel_encoder, mode->clock / 10, mode->hdisplay, mode->vdisplay); @@ -1149,7 +1154,7 @@ static bool intel_sdvo_mode_fixup(struct drm_encoder *encoder, if (success) { struct intel_sdvo_dtd input_dtd; - intel_sdvo_get_preferred_input_timing(output, + intel_sdvo_get_preferred_input_timing(intel_encoder, &input_dtd); intel_sdvo_get_mode_from_dtd(adjusted_mode, &input_dtd); dev_priv->sdvo_flags = input_dtd.part2.sdvo_flags; @@ -1181,8 +1186,8 @@ static void intel_sdvo_mode_set(struct drm_encoder *encoder, struct drm_i915_private *dev_priv = dev->dev_private; struct drm_crtc *crtc = encoder->crtc; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - struct intel_encoder *output = enc_to_intel_encoder(encoder); - struct intel_sdvo_priv *sdvo_priv = output->dev_priv; + struct intel_encoder *intel_encoder = enc_to_intel_encoder(encoder); + struct intel_sdvo_priv *sdvo_priv = intel_encoder->dev_priv; u32 sdvox = 0; int sdvo_pixel_multiply; struct intel_sdvo_in_out_map in_out; @@ -1201,12 +1206,12 @@ static void intel_sdvo_mode_set(struct drm_encoder *encoder, in_out.in0 = sdvo_priv->controlled_output; in_out.in1 = 0; - intel_sdvo_write_cmd(output, SDVO_CMD_SET_IN_OUT_MAP, + intel_sdvo_write_cmd(intel_encoder, SDVO_CMD_SET_IN_OUT_MAP, &in_out, sizeof(in_out)); - status = intel_sdvo_read_response(output, NULL, 0); + status = intel_sdvo_read_response(intel_encoder, NULL, 0); if (sdvo_priv->is_hdmi) { - intel_sdvo_set_avi_infoframe(output, mode); + intel_sdvo_set_avi_infoframe(intel_encoder, mode); sdvox |= SDVO_AUDIO_ENABLE; } @@ -1223,16 +1228,16 @@ static void intel_sdvo_mode_set(struct drm_encoder *encoder, */ if (!sdvo_priv->is_tv && !sdvo_priv->is_lvds) { /* Set the output timing to the screen */ - intel_sdvo_set_target_output(output, + intel_sdvo_set_target_output(intel_encoder, sdvo_priv->controlled_output); - intel_sdvo_set_output_timing(output, &input_dtd); + intel_sdvo_set_output_timing(intel_encoder, &input_dtd); } /* Set the input timing to the screen. Assume always input 0. */ - intel_sdvo_set_target_input(output, true, false); + intel_sdvo_set_target_input(intel_encoder, true, false); if (sdvo_priv->is_tv) - intel_sdvo_set_tv_format(output); + intel_sdvo_set_tv_format(intel_encoder); /* We would like to use intel_sdvo_create_preferred_input_timing() to * provide the device with a timing it can support, if it supports that @@ -1240,29 +1245,29 @@ static void intel_sdvo_mode_set(struct drm_encoder *encoder, * output the preferred timing, and we don't support that currently. */ #if 0 - success = intel_sdvo_create_preferred_input_timing(output, clock, + success = intel_sdvo_create_preferred_input_timing(encoder, clock, width, height); if (success) { struct intel_sdvo_dtd *input_dtd; - intel_sdvo_get_preferred_input_timing(output, &input_dtd); - intel_sdvo_set_input_timing(output, &input_dtd); + intel_sdvo_get_preferred_input_timing(encoder, &input_dtd); + intel_sdvo_set_input_timing(encoder, &input_dtd); } #else - intel_sdvo_set_input_timing(output, &input_dtd); + intel_sdvo_set_input_timing(intel_encoder, &input_dtd); #endif switch (intel_sdvo_get_pixel_multiplier(mode)) { case 1: - intel_sdvo_set_clock_rate_mult(output, + intel_sdvo_set_clock_rate_mult(intel_encoder, SDVO_CLOCK_RATE_MULT_1X); break; case 2: - intel_sdvo_set_clock_rate_mult(output, + intel_sdvo_set_clock_rate_mult(intel_encoder, SDVO_CLOCK_RATE_MULT_2X); break; case 4: - intel_sdvo_set_clock_rate_mult(output, + intel_sdvo_set_clock_rate_mult(intel_encoder, SDVO_CLOCK_RATE_MULT_4X); break; } @@ -1273,8 +1278,8 @@ static void intel_sdvo_mode_set(struct drm_encoder *encoder, SDVO_VSYNC_ACTIVE_HIGH | SDVO_HSYNC_ACTIVE_HIGH; } else { - sdvox |= I915_READ(sdvo_priv->output_device); - switch (sdvo_priv->output_device) { + sdvox |= I915_READ(sdvo_priv->sdvo_reg); + switch (sdvo_priv->sdvo_reg) { case SDVOB: sdvox &= SDVOB_PRESERVE_MASK; break; @@ -1298,7 +1303,7 @@ static void intel_sdvo_mode_set(struct drm_encoder *encoder, if (sdvo_priv->sdvo_flags & SDVO_NEED_TO_STALL) sdvox |= SDVO_STALL_SELECT; - intel_sdvo_write_sdvox(output, sdvox); + intel_sdvo_write_sdvox(intel_encoder, sdvox); } static void intel_sdvo_dpms(struct drm_encoder *encoder, int mode) @@ -1315,7 +1320,7 @@ static void intel_sdvo_dpms(struct drm_encoder *encoder, int mode) intel_sdvo_set_encoder_power_state(intel_encoder, mode); if (mode == DRM_MODE_DPMS_OFF) { - temp = I915_READ(sdvo_priv->output_device); + temp = I915_READ(sdvo_priv->sdvo_reg); if ((temp & SDVO_ENABLE) != 0) { intel_sdvo_write_sdvox(intel_encoder, temp & ~SDVO_ENABLE); } @@ -1325,7 +1330,7 @@ static void intel_sdvo_dpms(struct drm_encoder *encoder, int mode) int i; u8 status; - temp = I915_READ(sdvo_priv->output_device); + temp = I915_READ(sdvo_priv->sdvo_reg); if ((temp & SDVO_ENABLE) == 0) intel_sdvo_write_sdvox(intel_encoder, temp | SDVO_ENABLE); for (i = 0; i < 2; i++) @@ -1388,7 +1393,7 @@ static void intel_sdvo_save(struct drm_connector *connector) /* XXX: Save TV format/enhancements. */ } - sdvo_priv->save_SDVOX = I915_READ(sdvo_priv->output_device); + sdvo_priv->save_SDVOX = I915_READ(sdvo_priv->sdvo_reg); } static void intel_sdvo_restore(struct drm_connector *connector) @@ -1499,10 +1504,10 @@ struct drm_connector* intel_sdvo_find(struct drm_device *dev, int sdvoB) sdvo = iout->dev_priv; - if (sdvo->output_device == SDVOB && sdvoB) + if (sdvo->sdvo_reg == SDVOB && sdvoB) return connector; - if (sdvo->output_device == SDVOC && !sdvoB) + if (sdvo->sdvo_reg == SDVOC && !sdvoB) return connector; } @@ -2248,12 +2253,12 @@ static struct i2c_algorithm intel_sdvo_i2c_bit_algo = { }; static u8 -intel_sdvo_get_slave_addr(struct drm_device *dev, int output_device) +intel_sdvo_get_slave_addr(struct drm_device *dev, int sdvo_reg) { struct drm_i915_private *dev_priv = dev->dev_private; struct sdvo_device_mapping *my_mapping, *other_mapping; - if (output_device == SDVOB) { + if (sdvo_reg == SDVOB) { my_mapping = &dev_priv->sdvo_mappings[0]; other_mapping = &dev_priv->sdvo_mappings[1]; } else { @@ -2278,7 +2283,7 @@ intel_sdvo_get_slave_addr(struct drm_device *dev, int output_device) /* No SDVO device info is found for another DVO port, * so use mapping assumption we had before BIOS parsing. */ - if (output_device == SDVOB) + if (sdvo_reg == SDVOB) return 0x70; else return 0x72; @@ -2764,7 +2769,7 @@ static void intel_sdvo_create_enhance_property(struct drm_connector *connector) return; } -bool intel_sdvo_init(struct drm_device *dev, int output_device) +bool intel_sdvo_init(struct drm_device *dev, int sdvo_reg) { struct drm_i915_private *dev_priv = dev->dev_private; struct drm_connector *connector; @@ -2780,13 +2785,13 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) } sdvo_priv = (struct intel_sdvo_priv *)(intel_encoder + 1); - sdvo_priv->output_device = output_device; + sdvo_priv->sdvo_reg = sdvo_reg; intel_encoder->dev_priv = sdvo_priv; intel_encoder->type = INTEL_OUTPUT_SDVO; /* setup the DDC bus. */ - if (output_device == SDVOB) + if (sdvo_reg == SDVOB) intel_encoder->i2c_bus = intel_i2c_create(dev, GPIOE, "SDVOCTRL_E for SDVOB"); else intel_encoder->i2c_bus = intel_i2c_create(dev, GPIOE, "SDVOCTRL_E for SDVOC"); @@ -2794,7 +2799,7 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) if (!intel_encoder->i2c_bus) goto err_inteloutput; - sdvo_priv->slave_addr = intel_sdvo_get_slave_addr(dev, output_device); + sdvo_priv->slave_addr = intel_sdvo_get_slave_addr(dev, sdvo_reg); /* Save the bit-banging i2c functionality for use by the DDC wrapper */ intel_sdvo_i2c_bit_algo.functionality = intel_encoder->i2c_bus->algo->functionality; @@ -2803,13 +2808,13 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) for (i = 0; i < 0x40; i++) { if (!intel_sdvo_read_byte(intel_encoder, i, &ch[i])) { DRM_DEBUG_KMS("No SDVO device found on SDVO%c\n", - output_device == SDVOB ? 'B' : 'C'); + sdvo_reg == SDVOB ? 'B' : 'C'); goto err_i2c; } } /* setup the DDC bus. */ - if (output_device == SDVOB) { + if (sdvo_reg == SDVOB) { intel_encoder->ddc_bus = intel_i2c_create(dev, GPIOE, "SDVOB DDC BUS"); sdvo_priv->analog_ddc_bus = intel_i2c_create(dev, GPIOA, "SDVOB/VGA DDC BUS"); @@ -2833,7 +2838,7 @@ bool intel_sdvo_init(struct drm_device *dev, int output_device) if (intel_sdvo_output_setup(intel_encoder, sdvo_priv->caps.output_flags) != true) { DRM_DEBUG_KMS("SDVO output failed to setup on SDVO%c\n", - output_device == SDVOB ? 'B' : 'C'); + sdvo_reg == SDVOB ? 'B' : 'C'); goto err_i2c; } -- cgit v0.10.2 From d814f30105798b6677ecb73ed61d691ff96dada9 Mon Sep 17 00:00:00 2001 From: Cyrill Gorcunov Date: Wed, 24 Mar 2010 12:09:26 +0800 Subject: x86, perf: Add raw events support for the P4 PMU The adding of raw event support lead to complete code refactoring. I hope is became more readable then it was. The list of changes: 1) The 64bit config field is enough to hold all information we need to track event details. To achieve it we used *own* enum for events selection in ESCR register and map this key into proper value at moment of event enabling. For the same reason we use 12LSB bits in CCCR register -- to track which exactly cache trace event was requested. And we cear this bits at real 'write' moment. 2) There is no per-cpu area reserved for P4 PMU anymore. We don't need it. All is held by config. 3) Now we may use any available counter, ie we try to grab any possible counter. v2: - Lin Ming reported the lack of ESCR selector in CCCR for cache events v3: - Don't loose cache event codes at config unpacking procedure, we may need it one day so no obscure hack behind our back, better to clear reserved bits explicitly when needed (thanks Ming for pointing out) - Lin Ming fixed misplaced opcodes in cache events Signed-off-by: Cyrill Gorcunov Tested-by: Lin Ming Signed-off-by: Lin Ming Cc: Arnaldo Carvalho de Melo Cc: Stephane Eranian Cc: Robert Richter Cc: Frederic Weisbecker Cc: Cyrill Gorcunov Cc: Peter Zijlstra LKML-Reference: <1269403766.3409.6.camel@minggr.sh.intel.com> [ v4: did a few whitespace fixlets ] Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/perf_event_p4.h b/arch/x86/include/asm/perf_event_p4.h index facf961..b05400a 100644 --- a/arch/x86/include/asm/perf_event_p4.h +++ b/arch/x86/include/asm/perf_event_p4.h @@ -15,38 +15,40 @@ * perf-MSRs are not shared and every thread has its * own perf-MSRs set) */ -#define ARCH_P4_TOTAL_ESCR (46) -#define ARCH_P4_RESERVED_ESCR (2) /* IQ_ESCR(0,1) not always present */ -#define ARCH_P4_MAX_ESCR (ARCH_P4_TOTAL_ESCR - ARCH_P4_RESERVED_ESCR) -#define ARCH_P4_MAX_CCCR (18) -#define ARCH_P4_MAX_COUNTER (ARCH_P4_MAX_CCCR / 2) - -#define P4_EVNTSEL_EVENT_MASK 0x7e000000U -#define P4_EVNTSEL_EVENT_SHIFT 25 -#define P4_EVNTSEL_EVENTMASK_MASK 0x01fffe00U -#define P4_EVNTSEL_EVENTMASK_SHIFT 9 -#define P4_EVNTSEL_TAG_MASK 0x000001e0U -#define P4_EVNTSEL_TAG_SHIFT 5 -#define P4_EVNTSEL_TAG_ENABLE 0x00000010U -#define P4_EVNTSEL_T0_OS 0x00000008U -#define P4_EVNTSEL_T0_USR 0x00000004U -#define P4_EVNTSEL_T1_OS 0x00000002U -#define P4_EVNTSEL_T1_USR 0x00000001U +#define ARCH_P4_TOTAL_ESCR (46) +#define ARCH_P4_RESERVED_ESCR (2) /* IQ_ESCR(0,1) not always present */ +#define ARCH_P4_MAX_ESCR (ARCH_P4_TOTAL_ESCR - ARCH_P4_RESERVED_ESCR) +#define ARCH_P4_MAX_CCCR (18) +#define ARCH_P4_MAX_COUNTER (ARCH_P4_MAX_CCCR / 2) + +#define P4_ESCR_EVENT_MASK 0x7e000000U +#define P4_ESCR_EVENT_SHIFT 25 +#define P4_ESCR_EVENTMASK_MASK 0x01fffe00U +#define P4_ESCR_EVENTMASK_SHIFT 9 +#define P4_ESCR_TAG_MASK 0x000001e0U +#define P4_ESCR_TAG_SHIFT 5 +#define P4_ESCR_TAG_ENABLE 0x00000010U +#define P4_ESCR_T0_OS 0x00000008U +#define P4_ESCR_T0_USR 0x00000004U +#define P4_ESCR_T1_OS 0x00000002U +#define P4_ESCR_T1_USR 0x00000001U + +#define P4_ESCR_EVENT(v) ((v) << P4_ESCR_EVENT_SHIFT) +#define P4_ESCR_EMASK(v) ((v) << P4_ESCR_EVENTMASK_SHIFT) +#define P4_ESCR_TAG(v) ((v) << P4_ESCR_TAG_SHIFT) /* Non HT mask */ -#define P4_EVNTSEL_MASK \ - (P4_EVNTSEL_EVENT_MASK | \ - P4_EVNTSEL_EVENTMASK_MASK | \ - P4_EVNTSEL_TAG_MASK | \ - P4_EVNTSEL_TAG_ENABLE | \ - P4_EVNTSEL_T0_OS | \ - P4_EVNTSEL_T0_USR) +#define P4_ESCR_MASK \ + (P4_ESCR_EVENT_MASK | \ + P4_ESCR_EVENTMASK_MASK | \ + P4_ESCR_TAG_MASK | \ + P4_ESCR_TAG_ENABLE | \ + P4_ESCR_T0_OS | \ + P4_ESCR_T0_USR) /* HT mask */ -#define P4_EVNTSEL_MASK_HT \ - (P4_EVNTSEL_MASK | \ - P4_EVNTSEL_T1_OS | \ - P4_EVNTSEL_T1_USR) +#define P4_ESCR_MASK_HT \ + (P4_ESCR_MASK | P4_ESCR_T1_OS | P4_ESCR_T1_USR) #define P4_CCCR_OVF 0x80000000U #define P4_CCCR_CASCADE 0x40000000U @@ -56,7 +58,6 @@ #define P4_CCCR_EDGE 0x01000000U #define P4_CCCR_THRESHOLD_MASK 0x00f00000U #define P4_CCCR_THRESHOLD_SHIFT 20 -#define P4_CCCR_THRESHOLD(v) ((v) << P4_CCCR_THRESHOLD_SHIFT) #define P4_CCCR_COMPLEMENT 0x00080000U #define P4_CCCR_COMPARE 0x00040000U #define P4_CCCR_ESCR_SELECT_MASK 0x0000e000U @@ -67,6 +68,13 @@ #define P4_CCCR_THREAD_ANY 0x00030000U #define P4_CCCR_RESERVED 0x00000fffU +#define P4_CCCR_THRESHOLD(v) ((v) << P4_CCCR_THRESHOLD_SHIFT) +#define P4_CCCR_ESEL(v) ((v) << P4_CCCR_ESCR_SELECT_SHIFT) + +/* Custom bits in reerved CCCR area */ +#define P4_CCCR_CACHE_OPS_MASK 0x0000003fU + + /* Non HT mask */ #define P4_CCCR_MASK \ (P4_CCCR_OVF | \ @@ -81,25 +89,11 @@ P4_CCCR_ENABLE) /* HT mask */ -#define P4_CCCR_MASK_HT \ - (P4_CCCR_MASK | \ - P4_CCCR_THREAD_ANY) +#define P4_CCCR_MASK_HT (P4_CCCR_MASK | P4_CCCR_THREAD_ANY) -/* - * format is 32 bit: ee ss aa aa - * where - * ee - 8 bit event - * ss - 8 bit selector - * aa aa - 16 bits reserved for tags/attributes - */ -#define P4_EVENT_PACK(event, selector) (((event) << 24) | ((selector) << 16)) -#define P4_EVENT_UNPACK_EVENT(packed) (((packed) >> 24) & 0xff) -#define P4_EVENT_UNPACK_SELECTOR(packed) (((packed) >> 16) & 0xff) -#define P4_EVENT_PACK_ATTR(attr) ((attr)) -#define P4_EVENT_UNPACK_ATTR(packed) ((packed) & 0xffff) -#define P4_MAKE_EVENT_ATTR(class, name, bit) class##_##name = (1 << bit) -#define P4_EVENT_ATTR(class, name) class##_##name -#define P4_EVENT_ATTR_STR(class, name) __stringify(class##_##name) +#define P4_GEN_ESCR_EMASK(class, name, bit) \ + class##__##name = ((1 << bit) << P4_ESCR_EVENTMASK_SHIFT) +#define P4_ESCR_EMASK_BIT(class, name) class##__##name /* * config field is 64bit width and consists of @@ -117,35 +111,29 @@ #define p4_config_pack_escr(v) (((u64)(v)) << 32) #define p4_config_pack_cccr(v) (((u64)(v)) & 0xffffffffULL) #define p4_config_unpack_escr(v) (((u64)(v)) >> 32) -#define p4_config_unpack_cccr(v) (((u64)(v)) & 0xfffff000ULL) +#define p4_config_unpack_cccr(v) (((u64)(v)) & 0xffffffffULL) #define p4_config_unpack_emask(v) \ ({ \ u32 t = p4_config_unpack_escr((v)); \ - t &= P4_EVNTSEL_EVENTMASK_MASK; \ - t >>= P4_EVNTSEL_EVENTMASK_SHIFT; \ + t = t & P4_ESCR_EVENTMASK_MASK; \ + t = t >> P4_ESCR_EVENTMASK_SHIFT; \ + t; \ + }) + +#define p4_config_unpack_event(v) \ + ({ \ + u32 t = p4_config_unpack_escr((v)); \ + t = t & P4_ESCR_EVENT_MASK; \ + t = t >> P4_ESCR_EVENT_SHIFT; \ t; \ }) -#define p4_config_unpack_key(v) (((u64)(v)) & P4_CCCR_RESERVED) +#define p4_config_unpack_cache_event(v) (((u64)(v)) & P4_CCCR_CACHE_OPS_MASK) #define P4_CONFIG_HT_SHIFT 63 #define P4_CONFIG_HT (1ULL << P4_CONFIG_HT_SHIFT) -static inline u32 p4_config_unpack_opcode(u64 config) -{ - u32 e, s; - - /* - * we don't care about HT presence here since - * event opcode doesn't depend on it - */ - e = (p4_config_unpack_escr(config) & P4_EVNTSEL_EVENT_MASK) >> P4_EVNTSEL_EVENT_SHIFT; - s = (p4_config_unpack_cccr(config) & P4_CCCR_ESCR_SELECT_MASK) >> P4_CCCR_ESCR_SELECT_SHIFT; - - return P4_EVENT_PACK(e, s); -} - static inline bool p4_is_event_cascaded(u64 config) { u32 cccr = p4_config_unpack_cccr(config); @@ -212,19 +200,73 @@ static inline u32 p4_default_escr_conf(int cpu, int exclude_os, int exclude_usr) if (!p4_ht_thread(cpu)) { if (!exclude_os) - escr |= P4_EVNTSEL_T0_OS; + escr |= P4_ESCR_T0_OS; if (!exclude_usr) - escr |= P4_EVNTSEL_T0_USR; + escr |= P4_ESCR_T0_USR; } else { if (!exclude_os) - escr |= P4_EVNTSEL_T1_OS; + escr |= P4_ESCR_T1_OS; if (!exclude_usr) - escr |= P4_EVNTSEL_T1_USR; + escr |= P4_ESCR_T1_USR; } return escr; } +enum P4_EVENTS { + P4_EVENT_TC_DELIVER_MODE, + P4_EVENT_BPU_FETCH_REQUEST, + P4_EVENT_ITLB_REFERENCE, + P4_EVENT_MEMORY_CANCEL, + P4_EVENT_MEMORY_COMPLETE, + P4_EVENT_LOAD_PORT_REPLAY, + P4_EVENT_STORE_PORT_REPLAY, + P4_EVENT_MOB_LOAD_REPLAY, + P4_EVENT_PAGE_WALK_TYPE, + P4_EVENT_BSQ_CACHE_REFERENCE, + P4_EVENT_IOQ_ALLOCATION, + P4_EVENT_IOQ_ACTIVE_ENTRIES, + P4_EVENT_FSB_DATA_ACTIVITY, + P4_EVENT_BSQ_ALLOCATION, + P4_EVENT_BSQ_ACTIVE_ENTRIES, + P4_EVENT_SSE_INPUT_ASSIST, + P4_EVENT_PACKED_SP_UOP, + P4_EVENT_PACKED_DP_UOP, + P4_EVENT_SCALAR_SP_UOP, + P4_EVENT_SCALAR_DP_UOP, + P4_EVENT_64BIT_MMX_UOP, + P4_EVENT_128BIT_MMX_UOP, + P4_EVENT_X87_FP_UOP, + P4_EVENT_TC_MISC, + P4_EVENT_GLOBAL_POWER_EVENTS, + P4_EVENT_TC_MS_XFER, + P4_EVENT_UOP_QUEUE_WRITES, + P4_EVENT_RETIRED_MISPRED_BRANCH_TYPE, + P4_EVENT_RETIRED_BRANCH_TYPE, + P4_EVENT_RESOURCE_STALL, + P4_EVENT_WC_BUFFER, + P4_EVENT_B2B_CYCLES, + P4_EVENT_BNR, + P4_EVENT_SNOOP, + P4_EVENT_RESPONSE, + P4_EVENT_FRONT_END_EVENT, + P4_EVENT_EXECUTION_EVENT, + P4_EVENT_REPLAY_EVENT, + P4_EVENT_INSTR_RETIRED, + P4_EVENT_UOPS_RETIRED, + P4_EVENT_UOP_TYPE, + P4_EVENT_BRANCH_RETIRED, + P4_EVENT_MISPRED_BRANCH_RETIRED, + P4_EVENT_X87_ASSIST, + P4_EVENT_MACHINE_CLEAR, + P4_EVENT_INSTR_COMPLETED, +}; + +#define P4_OPCODE(event) event##_OPCODE +#define P4_OPCODE_ESEL(opcode) ((opcode & 0x00ff) >> 0) +#define P4_OPCODE_EVNT(opcode) ((opcode & 0xff00) >> 8) +#define P4_OPCODE_PACK(event, sel) (((event) << 8) | sel) + /* * Comments below the event represent ESCR restriction * for this event and counter index per ESCR @@ -238,484 +280,515 @@ static inline u32 p4_default_escr_conf(int cpu, int exclude_os, int exclude_usr) * working so that we should not use this CCCR and respective * counter as result */ -#define P4_TC_DELIVER_MODE P4_EVENT_PACK(0x01, 0x01) +enum P4_EVENT_OPCODES { + P4_OPCODE(P4_EVENT_TC_DELIVER_MODE) = P4_OPCODE_PACK(0x01, 0x01), /* * MSR_P4_TC_ESCR0: 4, 5 * MSR_P4_TC_ESCR1: 6, 7 */ -#define P4_BPU_FETCH_REQUEST P4_EVENT_PACK(0x03, 0x00) + P4_OPCODE(P4_EVENT_BPU_FETCH_REQUEST) = P4_OPCODE_PACK(0x03, 0x00), /* * MSR_P4_BPU_ESCR0: 0, 1 * MSR_P4_BPU_ESCR1: 2, 3 */ -#define P4_ITLB_REFERENCE P4_EVENT_PACK(0x18, 0x03) + P4_OPCODE(P4_EVENT_ITLB_REFERENCE) = P4_OPCODE_PACK(0x18, 0x03), /* * MSR_P4_ITLB_ESCR0: 0, 1 * MSR_P4_ITLB_ESCR1: 2, 3 */ -#define P4_MEMORY_CANCEL P4_EVENT_PACK(0x02, 0x05) + P4_OPCODE(P4_EVENT_MEMORY_CANCEL) = P4_OPCODE_PACK(0x02, 0x05), /* * MSR_P4_DAC_ESCR0: 8, 9 * MSR_P4_DAC_ESCR1: 10, 11 */ -#define P4_MEMORY_COMPLETE P4_EVENT_PACK(0x08, 0x02) + P4_OPCODE(P4_EVENT_MEMORY_COMPLETE) = P4_OPCODE_PACK(0x08, 0x02), /* * MSR_P4_SAAT_ESCR0: 8, 9 * MSR_P4_SAAT_ESCR1: 10, 11 */ -#define P4_LOAD_PORT_REPLAY P4_EVENT_PACK(0x04, 0x02) + P4_OPCODE(P4_EVENT_LOAD_PORT_REPLAY) = P4_OPCODE_PACK(0x04, 0x02), /* * MSR_P4_SAAT_ESCR0: 8, 9 * MSR_P4_SAAT_ESCR1: 10, 11 */ -#define P4_STORE_PORT_REPLAY P4_EVENT_PACK(0x05, 0x02) + P4_OPCODE(P4_EVENT_STORE_PORT_REPLAY) = P4_OPCODE_PACK(0x05, 0x02), /* * MSR_P4_SAAT_ESCR0: 8, 9 * MSR_P4_SAAT_ESCR1: 10, 11 */ -#define P4_MOB_LOAD_REPLAY P4_EVENT_PACK(0x03, 0x02) + P4_OPCODE(P4_EVENT_MOB_LOAD_REPLAY) = P4_OPCODE_PACK(0x03, 0x02), /* * MSR_P4_MOB_ESCR0: 0, 1 * MSR_P4_MOB_ESCR1: 2, 3 */ -#define P4_PAGE_WALK_TYPE P4_EVENT_PACK(0x01, 0x04) + P4_OPCODE(P4_EVENT_PAGE_WALK_TYPE) = P4_OPCODE_PACK(0x01, 0x04), /* * MSR_P4_PMH_ESCR0: 0, 1 * MSR_P4_PMH_ESCR1: 2, 3 */ -#define P4_BSQ_CACHE_REFERENCE P4_EVENT_PACK(0x0c, 0x07) + P4_OPCODE(P4_EVENT_BSQ_CACHE_REFERENCE) = P4_OPCODE_PACK(0x0c, 0x07), /* * MSR_P4_BSU_ESCR0: 0, 1 * MSR_P4_BSU_ESCR1: 2, 3 */ -#define P4_IOQ_ALLOCATION P4_EVENT_PACK(0x03, 0x06) + P4_OPCODE(P4_EVENT_IOQ_ALLOCATION) = P4_OPCODE_PACK(0x03, 0x06), /* * MSR_P4_FSB_ESCR0: 0, 1 * MSR_P4_FSB_ESCR1: 2, 3 */ -#define P4_IOQ_ACTIVE_ENTRIES P4_EVENT_PACK(0x1a, 0x06) + P4_OPCODE(P4_EVENT_IOQ_ACTIVE_ENTRIES) = P4_OPCODE_PACK(0x1a, 0x06), /* * MSR_P4_FSB_ESCR1: 2, 3 */ -#define P4_FSB_DATA_ACTIVITY P4_EVENT_PACK(0x17, 0x06) + P4_OPCODE(P4_EVENT_FSB_DATA_ACTIVITY) = P4_OPCODE_PACK(0x17, 0x06), /* * MSR_P4_FSB_ESCR0: 0, 1 * MSR_P4_FSB_ESCR1: 2, 3 */ -#define P4_BSQ_ALLOCATION P4_EVENT_PACK(0x05, 0x07) + P4_OPCODE(P4_EVENT_BSQ_ALLOCATION) = P4_OPCODE_PACK(0x05, 0x07), /* * MSR_P4_BSU_ESCR0: 0, 1 */ -#define P4_BSQ_ACTIVE_ENTRIES P4_EVENT_PACK(0x06, 0x07) + P4_OPCODE(P4_EVENT_BSQ_ACTIVE_ENTRIES) = P4_OPCODE_PACK(0x06, 0x07), /* * NOTE: no ESCR name in docs, it's guessed * MSR_P4_BSU_ESCR1: 2, 3 */ -#define P4_SSE_INPUT_ASSIST P4_EVENT_PACK(0x34, 0x01) + P4_OPCODE(P4_EVENT_SSE_INPUT_ASSIST) = P4_OPCODE_PACK(0x34, 0x01), /* * MSR_P4_FIRM_ESCR0: 8, 9 * MSR_P4_FIRM_ESCR1: 10, 11 */ -#define P4_PACKED_SP_UOP P4_EVENT_PACK(0x08, 0x01) + P4_OPCODE(P4_EVENT_PACKED_SP_UOP) = P4_OPCODE_PACK(0x08, 0x01), /* * MSR_P4_FIRM_ESCR0: 8, 9 * MSR_P4_FIRM_ESCR1: 10, 11 */ -#define P4_PACKED_DP_UOP P4_EVENT_PACK(0x0c, 0x01) + P4_OPCODE(P4_EVENT_PACKED_DP_UOP) = P4_OPCODE_PACK(0x0c, 0x01), /* * MSR_P4_FIRM_ESCR0: 8, 9 * MSR_P4_FIRM_ESCR1: 10, 11 */ -#define P4_SCALAR_SP_UOP P4_EVENT_PACK(0x0a, 0x01) + P4_OPCODE(P4_EVENT_SCALAR_SP_UOP) = P4_OPCODE_PACK(0x0a, 0x01), /* * MSR_P4_FIRM_ESCR0: 8, 9 * MSR_P4_FIRM_ESCR1: 10, 11 */ -#define P4_SCALAR_DP_UOP P4_EVENT_PACK(0x0e, 0x01) + P4_OPCODE(P4_EVENT_SCALAR_DP_UOP) = P4_OPCODE_PACK(0x0e, 0x01), /* * MSR_P4_FIRM_ESCR0: 8, 9 * MSR_P4_FIRM_ESCR1: 10, 11 */ -#define P4_64BIT_MMX_UOP P4_EVENT_PACK(0x02, 0x01) + P4_OPCODE(P4_EVENT_64BIT_MMX_UOP) = P4_OPCODE_PACK(0x02, 0x01), /* * MSR_P4_FIRM_ESCR0: 8, 9 * MSR_P4_FIRM_ESCR1: 10, 11 */ -#define P4_128BIT_MMX_UOP P4_EVENT_PACK(0x1a, 0x01) + P4_OPCODE(P4_EVENT_128BIT_MMX_UOP) = P4_OPCODE_PACK(0x1a, 0x01), /* * MSR_P4_FIRM_ESCR0: 8, 9 * MSR_P4_FIRM_ESCR1: 10, 11 */ -#define P4_X87_FP_UOP P4_EVENT_PACK(0x04, 0x01) + P4_OPCODE(P4_EVENT_X87_FP_UOP) = P4_OPCODE_PACK(0x04, 0x01), /* * MSR_P4_FIRM_ESCR0: 8, 9 * MSR_P4_FIRM_ESCR1: 10, 11 */ -#define P4_TC_MISC P4_EVENT_PACK(0x06, 0x01) + P4_OPCODE(P4_EVENT_TC_MISC) = P4_OPCODE_PACK(0x06, 0x01), /* * MSR_P4_TC_ESCR0: 4, 5 * MSR_P4_TC_ESCR1: 6, 7 */ -#define P4_GLOBAL_POWER_EVENTS P4_EVENT_PACK(0x13, 0x06) + P4_OPCODE(P4_EVENT_GLOBAL_POWER_EVENTS) = P4_OPCODE_PACK(0x13, 0x06), /* * MSR_P4_FSB_ESCR0: 0, 1 * MSR_P4_FSB_ESCR1: 2, 3 */ -#define P4_TC_MS_XFER P4_EVENT_PACK(0x05, 0x00) + P4_OPCODE(P4_EVENT_TC_MS_XFER) = P4_OPCODE_PACK(0x05, 0x00), /* * MSR_P4_MS_ESCR0: 4, 5 * MSR_P4_MS_ESCR1: 6, 7 */ -#define P4_UOP_QUEUE_WRITES P4_EVENT_PACK(0x09, 0x00) + P4_OPCODE(P4_EVENT_UOP_QUEUE_WRITES) = P4_OPCODE_PACK(0x09, 0x00), /* * MSR_P4_MS_ESCR0: 4, 5 * MSR_P4_MS_ESCR1: 6, 7 */ -#define P4_RETIRED_MISPRED_BRANCH_TYPE P4_EVENT_PACK(0x05, 0x02) + P4_OPCODE(P4_EVENT_RETIRED_MISPRED_BRANCH_TYPE) = P4_OPCODE_PACK(0x05, 0x02), /* * MSR_P4_TBPU_ESCR0: 4, 5 * MSR_P4_TBPU_ESCR1: 6, 7 */ -#define P4_RETIRED_BRANCH_TYPE P4_EVENT_PACK(0x04, 0x02) + P4_OPCODE(P4_EVENT_RETIRED_BRANCH_TYPE) = P4_OPCODE_PACK(0x04, 0x02), /* * MSR_P4_TBPU_ESCR0: 4, 5 * MSR_P4_TBPU_ESCR1: 6, 7 */ -#define P4_RESOURCE_STALL P4_EVENT_PACK(0x01, 0x01) + P4_OPCODE(P4_EVENT_RESOURCE_STALL) = P4_OPCODE_PACK(0x01, 0x01), /* * MSR_P4_ALF_ESCR0: 12, 13, 16 * MSR_P4_ALF_ESCR1: 14, 15, 17 */ -#define P4_WC_BUFFER P4_EVENT_PACK(0x05, 0x05) + P4_OPCODE(P4_EVENT_WC_BUFFER) = P4_OPCODE_PACK(0x05, 0x05), /* * MSR_P4_DAC_ESCR0: 8, 9 * MSR_P4_DAC_ESCR1: 10, 11 */ -#define P4_B2B_CYCLES P4_EVENT_PACK(0x16, 0x03) + P4_OPCODE(P4_EVENT_B2B_CYCLES) = P4_OPCODE_PACK(0x16, 0x03), /* * MSR_P4_FSB_ESCR0: 0, 1 * MSR_P4_FSB_ESCR1: 2, 3 */ -#define P4_BNR P4_EVENT_PACK(0x08, 0x03) + P4_OPCODE(P4_EVENT_BNR) = P4_OPCODE_PACK(0x08, 0x03), /* * MSR_P4_FSB_ESCR0: 0, 1 * MSR_P4_FSB_ESCR1: 2, 3 */ -#define P4_SNOOP P4_EVENT_PACK(0x06, 0x03) + P4_OPCODE(P4_EVENT_SNOOP) = P4_OPCODE_PACK(0x06, 0x03), /* * MSR_P4_FSB_ESCR0: 0, 1 * MSR_P4_FSB_ESCR1: 2, 3 */ -#define P4_RESPONSE P4_EVENT_PACK(0x04, 0x03) + P4_OPCODE(P4_EVENT_RESPONSE) = P4_OPCODE_PACK(0x04, 0x03), /* * MSR_P4_FSB_ESCR0: 0, 1 * MSR_P4_FSB_ESCR1: 2, 3 */ -#define P4_FRONT_END_EVENT P4_EVENT_PACK(0x08, 0x05) + P4_OPCODE(P4_EVENT_FRONT_END_EVENT) = P4_OPCODE_PACK(0x08, 0x05), /* * MSR_P4_CRU_ESCR2: 12, 13, 16 * MSR_P4_CRU_ESCR3: 14, 15, 17 */ -#define P4_EXECUTION_EVENT P4_EVENT_PACK(0x0c, 0x05) + P4_OPCODE(P4_EVENT_EXECUTION_EVENT) = P4_OPCODE_PACK(0x0c, 0x05), /* * MSR_P4_CRU_ESCR2: 12, 13, 16 * MSR_P4_CRU_ESCR3: 14, 15, 17 */ -#define P4_REPLAY_EVENT P4_EVENT_PACK(0x09, 0x05) + P4_OPCODE(P4_EVENT_REPLAY_EVENT) = P4_OPCODE_PACK(0x09, 0x05), /* * MSR_P4_CRU_ESCR2: 12, 13, 16 * MSR_P4_CRU_ESCR3: 14, 15, 17 */ -#define P4_INSTR_RETIRED P4_EVENT_PACK(0x02, 0x04) + P4_OPCODE(P4_EVENT_INSTR_RETIRED) = P4_OPCODE_PACK(0x02, 0x04), /* * MSR_P4_CRU_ESCR0: 12, 13, 16 * MSR_P4_CRU_ESCR1: 14, 15, 17 */ -#define P4_UOPS_RETIRED P4_EVENT_PACK(0x01, 0x04) + P4_OPCODE(P4_EVENT_UOPS_RETIRED) = P4_OPCODE_PACK(0x01, 0x04), /* * MSR_P4_CRU_ESCR0: 12, 13, 16 * MSR_P4_CRU_ESCR1: 14, 15, 17 */ -#define P4_UOP_TYPE P4_EVENT_PACK(0x02, 0x02) + P4_OPCODE(P4_EVENT_UOP_TYPE) = P4_OPCODE_PACK(0x02, 0x02), /* * MSR_P4_RAT_ESCR0: 12, 13, 16 * MSR_P4_RAT_ESCR1: 14, 15, 17 */ -#define P4_BRANCH_RETIRED P4_EVENT_PACK(0x06, 0x05) + P4_OPCODE(P4_EVENT_BRANCH_RETIRED) = P4_OPCODE_PACK(0x06, 0x05), /* * MSR_P4_CRU_ESCR2: 12, 13, 16 * MSR_P4_CRU_ESCR3: 14, 15, 17 */ -#define P4_MISPRED_BRANCH_RETIRED P4_EVENT_PACK(0x03, 0x04) + P4_OPCODE(P4_EVENT_MISPRED_BRANCH_RETIRED) = P4_OPCODE_PACK(0x03, 0x04), /* * MSR_P4_CRU_ESCR0: 12, 13, 16 * MSR_P4_CRU_ESCR1: 14, 15, 17 */ -#define P4_X87_ASSIST P4_EVENT_PACK(0x03, 0x05) + P4_OPCODE(P4_EVENT_X87_ASSIST) = P4_OPCODE_PACK(0x03, 0x05), /* * MSR_P4_CRU_ESCR2: 12, 13, 16 * MSR_P4_CRU_ESCR3: 14, 15, 17 */ -#define P4_MACHINE_CLEAR P4_EVENT_PACK(0x02, 0x05) + P4_OPCODE(P4_EVENT_MACHINE_CLEAR) = P4_OPCODE_PACK(0x02, 0x05), /* * MSR_P4_CRU_ESCR2: 12, 13, 16 * MSR_P4_CRU_ESCR3: 14, 15, 17 */ -#define P4_INSTR_COMPLETED P4_EVENT_PACK(0x07, 0x04) + P4_OPCODE(P4_EVENT_INSTR_COMPLETED) = P4_OPCODE_PACK(0x07, 0x04), /* * MSR_P4_CRU_ESCR0: 12, 13, 16 * MSR_P4_CRU_ESCR1: 14, 15, 17 */ +}; /* - * a caller should use P4_EVENT_ATTR helper to - * pick the attribute needed, for example + * a caller should use P4_ESCR_EMASK_NAME helper to + * pick the EventMask needed, for example * - * P4_EVENT_ATTR(P4_TC_DELIVER_MODE, DD) + * P4_ESCR_EMASK_NAME(P4_EVENT_TC_DELIVER_MODE, DD) */ -enum P4_EVENTS_ATTR { - P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, DD, 0), - P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, DB, 1), - P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, DI, 2), - P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, BD, 3), - P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, BB, 4), - P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, BI, 5), - P4_MAKE_EVENT_ATTR(P4_TC_DELIVER_MODE, ID, 6), - - P4_MAKE_EVENT_ATTR(P4_BPU_FETCH_REQUEST, TCMISS, 0), - - P4_MAKE_EVENT_ATTR(P4_ITLB_REFERENCE, HIT, 0), - P4_MAKE_EVENT_ATTR(P4_ITLB_REFERENCE, MISS, 1), - P4_MAKE_EVENT_ATTR(P4_ITLB_REFERENCE, HIT_UK, 2), - - P4_MAKE_EVENT_ATTR(P4_MEMORY_CANCEL, ST_RB_FULL, 2), - P4_MAKE_EVENT_ATTR(P4_MEMORY_CANCEL, 64K_CONF, 3), - - P4_MAKE_EVENT_ATTR(P4_MEMORY_COMPLETE, LSC, 0), - P4_MAKE_EVENT_ATTR(P4_MEMORY_COMPLETE, SSC, 1), - - P4_MAKE_EVENT_ATTR(P4_LOAD_PORT_REPLAY, SPLIT_LD, 1), - - P4_MAKE_EVENT_ATTR(P4_STORE_PORT_REPLAY, SPLIT_ST, 1), - - P4_MAKE_EVENT_ATTR(P4_MOB_LOAD_REPLAY, NO_STA, 1), - P4_MAKE_EVENT_ATTR(P4_MOB_LOAD_REPLAY, NO_STD, 3), - P4_MAKE_EVENT_ATTR(P4_MOB_LOAD_REPLAY, PARTIAL_DATA, 4), - P4_MAKE_EVENT_ATTR(P4_MOB_LOAD_REPLAY, UNALGN_ADDR, 5), - - P4_MAKE_EVENT_ATTR(P4_PAGE_WALK_TYPE, DTMISS, 0), - P4_MAKE_EVENT_ATTR(P4_PAGE_WALK_TYPE, ITMISS, 1), - - P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITS, 0), - P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITE, 1), - P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITM, 2), - P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_HITS, 3), - P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_HITE, 4), - P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_HITM, 5), - P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_MISS, 8), - P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_MISS, 9), - P4_MAKE_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, WR_2ndL_MISS, 10), - - P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, DEFAULT, 0), - P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, ALL_READ, 5), - P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, ALL_WRITE, 6), - P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, MEM_UC, 7), - P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, MEM_WC, 8), - P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, MEM_WT, 9), - P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, MEM_WP, 10), - P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, MEM_WB, 11), - P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, OWN, 13), - P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, OTHER, 14), - P4_MAKE_EVENT_ATTR(P4_IOQ_ALLOCATION, PREFETCH, 15), - - P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, DEFAULT, 0), - P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, ALL_READ, 5), - P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, ALL_WRITE, 6), - P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, MEM_UC, 7), - P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, MEM_WC, 8), - P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, MEM_WT, 9), - P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, MEM_WP, 10), - P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, MEM_WB, 11), - P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, OWN, 13), - P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, OTHER, 14), - P4_MAKE_EVENT_ATTR(P4_IOQ_ACTIVE_ENTRIES, PREFETCH, 15), - - P4_MAKE_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DRDY_DRV, 0), - P4_MAKE_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DRDY_OWN, 1), - P4_MAKE_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DRDY_OTHER, 2), - P4_MAKE_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DBSY_DRV, 3), - P4_MAKE_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DBSY_OWN, 4), - P4_MAKE_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DBSY_OTHER, 5), - - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_TYPE0, 0), - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_TYPE1, 1), - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_LEN0, 2), - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_LEN1, 3), - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_IO_TYPE, 5), - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_LOCK_TYPE, 6), - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_CACHE_TYPE, 7), - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_SPLIT_TYPE, 8), - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_DEM_TYPE, 9), - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, REQ_ORD_TYPE, 10), - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, MEM_TYPE0, 11), - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, MEM_TYPE1, 12), - P4_MAKE_EVENT_ATTR(P4_BSQ_ALLOCATION, MEM_TYPE2, 13), - - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_TYPE0, 0), - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_TYPE1, 1), - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_LEN0, 2), - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_LEN1, 3), - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_IO_TYPE, 5), - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_LOCK_TYPE, 6), - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_CACHE_TYPE, 7), - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_SPLIT_TYPE, 8), - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_DEM_TYPE, 9), - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, REQ_ORD_TYPE, 10), - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, MEM_TYPE0, 11), - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, MEM_TYPE1, 12), - P4_MAKE_EVENT_ATTR(P4_BSQ_ACTIVE_ENTRIES, MEM_TYPE2, 13), - - P4_MAKE_EVENT_ATTR(P4_SSE_INPUT_ASSIST, ALL, 15), - - P4_MAKE_EVENT_ATTR(P4_PACKED_SP_UOP, ALL, 15), - - P4_MAKE_EVENT_ATTR(P4_PACKED_DP_UOP, ALL, 15), - - P4_MAKE_EVENT_ATTR(P4_SCALAR_SP_UOP, ALL, 15), - - P4_MAKE_EVENT_ATTR(P4_SCALAR_DP_UOP, ALL, 15), - - P4_MAKE_EVENT_ATTR(P4_64BIT_MMX_UOP, ALL, 15), - - P4_MAKE_EVENT_ATTR(P4_128BIT_MMX_UOP, ALL, 15), - - P4_MAKE_EVENT_ATTR(P4_X87_FP_UOP, ALL, 15), - - P4_MAKE_EVENT_ATTR(P4_TC_MISC, FLUSH, 4), - - P4_MAKE_EVENT_ATTR(P4_GLOBAL_POWER_EVENTS, RUNNING, 0), - - P4_MAKE_EVENT_ATTR(P4_TC_MS_XFER, CISC, 0), - - P4_MAKE_EVENT_ATTR(P4_UOP_QUEUE_WRITES, FROM_TC_BUILD, 0), - P4_MAKE_EVENT_ATTR(P4_UOP_QUEUE_WRITES, FROM_TC_DELIVER, 1), - P4_MAKE_EVENT_ATTR(P4_UOP_QUEUE_WRITES, FROM_ROM, 2), - - P4_MAKE_EVENT_ATTR(P4_RETIRED_MISPRED_BRANCH_TYPE, CONDITIONAL, 1), - P4_MAKE_EVENT_ATTR(P4_RETIRED_MISPRED_BRANCH_TYPE, CALL, 2), - P4_MAKE_EVENT_ATTR(P4_RETIRED_MISPRED_BRANCH_TYPE, RETURN, 3), - P4_MAKE_EVENT_ATTR(P4_RETIRED_MISPRED_BRANCH_TYPE, INDIRECT, 4), - - P4_MAKE_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, CONDITIONAL, 1), - P4_MAKE_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, CALL, 2), - P4_MAKE_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, RETURN, 3), - P4_MAKE_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, INDIRECT, 4), - - P4_MAKE_EVENT_ATTR(P4_RESOURCE_STALL, SBFULL, 5), - - P4_MAKE_EVENT_ATTR(P4_WC_BUFFER, WCB_EVICTS, 0), - P4_MAKE_EVENT_ATTR(P4_WC_BUFFER, WCB_FULL_EVICTS, 1), - - P4_MAKE_EVENT_ATTR(P4_FRONT_END_EVENT, NBOGUS, 0), - P4_MAKE_EVENT_ATTR(P4_FRONT_END_EVENT, BOGUS, 1), - - P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, NBOGUS0, 0), - P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, NBOGUS1, 1), - P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, NBOGUS2, 2), - P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, NBOGUS3, 3), - P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, BOGUS0, 4), - P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, BOGUS1, 5), - P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, BOGUS2, 6), - P4_MAKE_EVENT_ATTR(P4_EXECUTION_EVENT, BOGUS3, 7), - - P4_MAKE_EVENT_ATTR(P4_REPLAY_EVENT, NBOGUS, 0), - P4_MAKE_EVENT_ATTR(P4_REPLAY_EVENT, BOGUS, 1), - - P4_MAKE_EVENT_ATTR(P4_INSTR_RETIRED, NBOGUSNTAG, 0), - P4_MAKE_EVENT_ATTR(P4_INSTR_RETIRED, NBOGUSTAG, 1), - P4_MAKE_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSNTAG, 2), - P4_MAKE_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSTAG, 3), - - P4_MAKE_EVENT_ATTR(P4_UOPS_RETIRED, NBOGUS, 0), - P4_MAKE_EVENT_ATTR(P4_UOPS_RETIRED, BOGUS, 1), +enum P4_ESCR_EMASKS { + P4_GEN_ESCR_EMASK(P4_EVENT_TC_DELIVER_MODE, DD, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_TC_DELIVER_MODE, DB, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_TC_DELIVER_MODE, DI, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_TC_DELIVER_MODE, BD, 3), + P4_GEN_ESCR_EMASK(P4_EVENT_TC_DELIVER_MODE, BB, 4), + P4_GEN_ESCR_EMASK(P4_EVENT_TC_DELIVER_MODE, BI, 5), + P4_GEN_ESCR_EMASK(P4_EVENT_TC_DELIVER_MODE, ID, 6), + + P4_GEN_ESCR_EMASK(P4_EVENT_BPU_FETCH_REQUEST, TCMISS, 0), + + P4_GEN_ESCR_EMASK(P4_EVENT_ITLB_REFERENCE, HIT, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_ITLB_REFERENCE, MISS, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_ITLB_REFERENCE, HIT_UK, 2), + + P4_GEN_ESCR_EMASK(P4_EVENT_MEMORY_CANCEL, ST_RB_FULL, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_MEMORY_CANCEL, 64K_CONF, 3), + + P4_GEN_ESCR_EMASK(P4_EVENT_MEMORY_COMPLETE, LSC, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_MEMORY_COMPLETE, SSC, 1), + + P4_GEN_ESCR_EMASK(P4_EVENT_LOAD_PORT_REPLAY, SPLIT_LD, 1), + + P4_GEN_ESCR_EMASK(P4_EVENT_STORE_PORT_REPLAY, SPLIT_ST, 1), + + P4_GEN_ESCR_EMASK(P4_EVENT_MOB_LOAD_REPLAY, NO_STA, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_MOB_LOAD_REPLAY, NO_STD, 3), + P4_GEN_ESCR_EMASK(P4_EVENT_MOB_LOAD_REPLAY, PARTIAL_DATA, 4), + P4_GEN_ESCR_EMASK(P4_EVENT_MOB_LOAD_REPLAY, UNALGN_ADDR, 5), + + P4_GEN_ESCR_EMASK(P4_EVENT_PAGE_WALK_TYPE, DTMISS, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_PAGE_WALK_TYPE, ITMISS, 1), + + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_CACHE_REFERENCE, RD_2ndL_HITS, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_CACHE_REFERENCE, RD_2ndL_HITE, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_CACHE_REFERENCE, RD_2ndL_HITM, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_CACHE_REFERENCE, RD_3rdL_HITS, 3), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_CACHE_REFERENCE, RD_3rdL_HITE, 4), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_CACHE_REFERENCE, RD_3rdL_HITM, 5), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_CACHE_REFERENCE, RD_2ndL_MISS, 8), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_CACHE_REFERENCE, RD_3rdL_MISS, 9), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_CACHE_REFERENCE, WR_2ndL_MISS, 10), + + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ALLOCATION, DEFAULT, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ALLOCATION, ALL_READ, 5), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ALLOCATION, ALL_WRITE, 6), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ALLOCATION, MEM_UC, 7), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ALLOCATION, MEM_WC, 8), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ALLOCATION, MEM_WT, 9), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ALLOCATION, MEM_WP, 10), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ALLOCATION, MEM_WB, 11), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ALLOCATION, OWN, 13), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ALLOCATION, OTHER, 14), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ALLOCATION, PREFETCH, 15), + + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ACTIVE_ENTRIES, DEFAULT, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ACTIVE_ENTRIES, ALL_READ, 5), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ACTIVE_ENTRIES, ALL_WRITE, 6), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ACTIVE_ENTRIES, MEM_UC, 7), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ACTIVE_ENTRIES, MEM_WC, 8), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ACTIVE_ENTRIES, MEM_WT, 9), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ACTIVE_ENTRIES, MEM_WP, 10), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ACTIVE_ENTRIES, MEM_WB, 11), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ACTIVE_ENTRIES, OWN, 13), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ACTIVE_ENTRIES, OTHER, 14), + P4_GEN_ESCR_EMASK(P4_EVENT_IOQ_ACTIVE_ENTRIES, PREFETCH, 15), + + P4_GEN_ESCR_EMASK(P4_EVENT_FSB_DATA_ACTIVITY, DRDY_DRV, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_FSB_DATA_ACTIVITY, DRDY_OWN, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_FSB_DATA_ACTIVITY, DRDY_OTHER, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_FSB_DATA_ACTIVITY, DBSY_DRV, 3), + P4_GEN_ESCR_EMASK(P4_EVENT_FSB_DATA_ACTIVITY, DBSY_OWN, 4), + P4_GEN_ESCR_EMASK(P4_EVENT_FSB_DATA_ACTIVITY, DBSY_OTHER, 5), + + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, REQ_TYPE0, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, REQ_TYPE1, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, REQ_LEN0, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, REQ_LEN1, 3), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, REQ_IO_TYPE, 5), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, REQ_LOCK_TYPE, 6), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, REQ_CACHE_TYPE, 7), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, REQ_SPLIT_TYPE, 8), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, REQ_DEM_TYPE, 9), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, REQ_ORD_TYPE, 10), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, MEM_TYPE0, 11), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, MEM_TYPE1, 12), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ALLOCATION, MEM_TYPE2, 13), + + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, REQ_TYPE0, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, REQ_TYPE1, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, REQ_LEN0, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, REQ_LEN1, 3), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, REQ_IO_TYPE, 5), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, REQ_LOCK_TYPE, 6), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, REQ_CACHE_TYPE, 7), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, REQ_SPLIT_TYPE, 8), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, REQ_DEM_TYPE, 9), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, REQ_ORD_TYPE, 10), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, MEM_TYPE0, 11), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, MEM_TYPE1, 12), + P4_GEN_ESCR_EMASK(P4_EVENT_BSQ_ACTIVE_ENTRIES, MEM_TYPE2, 13), + + P4_GEN_ESCR_EMASK(P4_EVENT_SSE_INPUT_ASSIST, ALL, 15), + + P4_GEN_ESCR_EMASK(P4_EVENT_PACKED_SP_UOP, ALL, 15), + + P4_GEN_ESCR_EMASK(P4_EVENT_PACKED_DP_UOP, ALL, 15), + + P4_GEN_ESCR_EMASK(P4_EVENT_SCALAR_SP_UOP, ALL, 15), + + P4_GEN_ESCR_EMASK(P4_EVENT_SCALAR_DP_UOP, ALL, 15), + + P4_GEN_ESCR_EMASK(P4_EVENT_64BIT_MMX_UOP, ALL, 15), + + P4_GEN_ESCR_EMASK(P4_EVENT_128BIT_MMX_UOP, ALL, 15), + + P4_GEN_ESCR_EMASK(P4_EVENT_X87_FP_UOP, ALL, 15), + + P4_GEN_ESCR_EMASK(P4_EVENT_TC_MISC, FLUSH, 4), + + P4_GEN_ESCR_EMASK(P4_EVENT_GLOBAL_POWER_EVENTS, RUNNING, 0), + + P4_GEN_ESCR_EMASK(P4_EVENT_TC_MS_XFER, CISC, 0), + + P4_GEN_ESCR_EMASK(P4_EVENT_UOP_QUEUE_WRITES, FROM_TC_BUILD, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_UOP_QUEUE_WRITES, FROM_TC_DELIVER, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_UOP_QUEUE_WRITES, FROM_ROM, 2), + + P4_GEN_ESCR_EMASK(P4_EVENT_RETIRED_MISPRED_BRANCH_TYPE, CONDITIONAL, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_RETIRED_MISPRED_BRANCH_TYPE, CALL, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_RETIRED_MISPRED_BRANCH_TYPE, RETURN, 3), + P4_GEN_ESCR_EMASK(P4_EVENT_RETIRED_MISPRED_BRANCH_TYPE, INDIRECT, 4), + + P4_GEN_ESCR_EMASK(P4_EVENT_RETIRED_BRANCH_TYPE, CONDITIONAL, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_RETIRED_BRANCH_TYPE, CALL, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_RETIRED_BRANCH_TYPE, RETURN, 3), + P4_GEN_ESCR_EMASK(P4_EVENT_RETIRED_BRANCH_TYPE, INDIRECT, 4), + + P4_GEN_ESCR_EMASK(P4_EVENT_RESOURCE_STALL, SBFULL, 5), + + P4_GEN_ESCR_EMASK(P4_EVENT_WC_BUFFER, WCB_EVICTS, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_WC_BUFFER, WCB_FULL_EVICTS, 1), + + P4_GEN_ESCR_EMASK(P4_EVENT_FRONT_END_EVENT, NBOGUS, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_FRONT_END_EVENT, BOGUS, 1), + + P4_GEN_ESCR_EMASK(P4_EVENT_EXECUTION_EVENT, NBOGUS0, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_EXECUTION_EVENT, NBOGUS1, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_EXECUTION_EVENT, NBOGUS2, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_EXECUTION_EVENT, NBOGUS3, 3), + P4_GEN_ESCR_EMASK(P4_EVENT_EXECUTION_EVENT, BOGUS0, 4), + P4_GEN_ESCR_EMASK(P4_EVENT_EXECUTION_EVENT, BOGUS1, 5), + P4_GEN_ESCR_EMASK(P4_EVENT_EXECUTION_EVENT, BOGUS2, 6), + P4_GEN_ESCR_EMASK(P4_EVENT_EXECUTION_EVENT, BOGUS3, 7), + + P4_GEN_ESCR_EMASK(P4_EVENT_REPLAY_EVENT, NBOGUS, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_REPLAY_EVENT, BOGUS, 1), + + P4_GEN_ESCR_EMASK(P4_EVENT_INSTR_RETIRED, NBOGUSNTAG, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_INSTR_RETIRED, NBOGUSTAG, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_INSTR_RETIRED, BOGUSNTAG, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_INSTR_RETIRED, BOGUSTAG, 3), + + P4_GEN_ESCR_EMASK(P4_EVENT_UOPS_RETIRED, NBOGUS, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_UOPS_RETIRED, BOGUS, 1), - P4_MAKE_EVENT_ATTR(P4_UOP_TYPE, TAGLOADS, 1), - P4_MAKE_EVENT_ATTR(P4_UOP_TYPE, TAGSTORES, 2), - - P4_MAKE_EVENT_ATTR(P4_BRANCH_RETIRED, MMNP, 0), - P4_MAKE_EVENT_ATTR(P4_BRANCH_RETIRED, MMNM, 1), - P4_MAKE_EVENT_ATTR(P4_BRANCH_RETIRED, MMTP, 2), - P4_MAKE_EVENT_ATTR(P4_BRANCH_RETIRED, MMTM, 3), + P4_GEN_ESCR_EMASK(P4_EVENT_UOP_TYPE, TAGLOADS, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_UOP_TYPE, TAGSTORES, 2), + + P4_GEN_ESCR_EMASK(P4_EVENT_BRANCH_RETIRED, MMNP, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_BRANCH_RETIRED, MMNM, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_BRANCH_RETIRED, MMTP, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_BRANCH_RETIRED, MMTM, 3), - P4_MAKE_EVENT_ATTR(P4_MISPRED_BRANCH_RETIRED, NBOGUS, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_MISPRED_BRANCH_RETIRED, NBOGUS, 0), - P4_MAKE_EVENT_ATTR(P4_X87_ASSIST, FPSU, 0), - P4_MAKE_EVENT_ATTR(P4_X87_ASSIST, FPSO, 1), - P4_MAKE_EVENT_ATTR(P4_X87_ASSIST, POAO, 2), - P4_MAKE_EVENT_ATTR(P4_X87_ASSIST, POAU, 3), - P4_MAKE_EVENT_ATTR(P4_X87_ASSIST, PREA, 4), + P4_GEN_ESCR_EMASK(P4_EVENT_X87_ASSIST, FPSU, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_X87_ASSIST, FPSO, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_X87_ASSIST, POAO, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_X87_ASSIST, POAU, 3), + P4_GEN_ESCR_EMASK(P4_EVENT_X87_ASSIST, PREA, 4), - P4_MAKE_EVENT_ATTR(P4_MACHINE_CLEAR, CLEAR, 0), - P4_MAKE_EVENT_ATTR(P4_MACHINE_CLEAR, MOCLEAR, 1), - P4_MAKE_EVENT_ATTR(P4_MACHINE_CLEAR, SMCLEAR, 2), + P4_GEN_ESCR_EMASK(P4_EVENT_MACHINE_CLEAR, CLEAR, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_MACHINE_CLEAR, MOCLEAR, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_MACHINE_CLEAR, SMCLEAR, 2), - P4_MAKE_EVENT_ATTR(P4_INSTR_COMPLETED, NBOGUS, 0), - P4_MAKE_EVENT_ATTR(P4_INSTR_COMPLETED, BOGUS, 1), + P4_GEN_ESCR_EMASK(P4_EVENT_INSTR_COMPLETED, NBOGUS, 0), + P4_GEN_ESCR_EMASK(P4_EVENT_INSTR_COMPLETED, BOGUS, 1), }; -enum { - KEY_P4_L1D_OP_READ_RESULT_MISS = PERF_COUNT_HW_MAX, - KEY_P4_LL_OP_READ_RESULT_MISS, - KEY_P4_DTLB_OP_READ_RESULT_MISS, - KEY_P4_DTLB_OP_WRITE_RESULT_MISS, - KEY_P4_ITLB_OP_READ_RESULT_ACCESS, - KEY_P4_ITLB_OP_READ_RESULT_MISS, - KEY_P4_UOP_TYPE, +/* P4 PEBS: stale for a while */ +#define P4_PEBS_METRIC_MASK 0x00001fffU +#define P4_PEBS_UOB_TAG 0x01000000U +#define P4_PEBS_ENABLE 0x02000000U + +/* Replay metrics for MSR_IA32_PEBS_ENABLE and MSR_P4_PEBS_MATRIX_VERT */ +#define P4_PEBS__1stl_cache_load_miss_retired 0x3000001 +#define P4_PEBS__2ndl_cache_load_miss_retired 0x3000002 +#define P4_PEBS__dtlb_load_miss_retired 0x3000004 +#define P4_PEBS__dtlb_store_miss_retired 0x3000004 +#define P4_PEBS__dtlb_all_miss_retired 0x3000004 +#define P4_PEBS__tagged_mispred_branch 0x3018000 +#define P4_PEBS__mob_load_replay_retired 0x3000200 +#define P4_PEBS__split_load_retired 0x3000400 +#define P4_PEBS__split_store_retired 0x3000400 + +#define P4_VERT__1stl_cache_load_miss_retired 0x0000001 +#define P4_VERT__2ndl_cache_load_miss_retired 0x0000001 +#define P4_VERT__dtlb_load_miss_retired 0x0000001 +#define P4_VERT__dtlb_store_miss_retired 0x0000002 +#define P4_VERT__dtlb_all_miss_retired 0x0000003 +#define P4_VERT__tagged_mispred_branch 0x0000010 +#define P4_VERT__mob_load_replay_retired 0x0000001 +#define P4_VERT__split_load_retired 0x0000001 +#define P4_VERT__split_store_retired 0x0000002 + +enum P4_CACHE_EVENTS { + P4_CACHE__NONE, + + P4_CACHE__1stl_cache_load_miss_retired, + P4_CACHE__2ndl_cache_load_miss_retired, + P4_CACHE__dtlb_load_miss_retired, + P4_CACHE__dtlb_store_miss_retired, + P4_CACHE__itlb_reference_hit, + P4_CACHE__itlb_reference_miss, + + P4_CACHE__MAX }; #endif /* PERF_EVENT_P4_H */ diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index b8a811a..f8fe069 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -11,35 +11,281 @@ #include +#define P4_CNTR_LIMIT 3 /* * array indices: 0,1 - HT threads, used with HT enabled cpu */ -struct p4_event_template { - u32 opcode; /* ESCR event + CCCR selector */ - u64 config; /* packed predefined bits */ - int dep; /* upstream dependency event index */ - int key; /* index into p4_templates */ - u64 msr; /* - * the high 32 bits set into MSR_IA32_PEBS_ENABLE and - * the low 32 bits set into MSR_P4_PEBS_MATRIX_VERT - * for cache events - */ - unsigned int emask; /* ESCR EventMask */ - unsigned int escr_msr[2]; /* ESCR MSR for this event */ - unsigned int cntr[2]; /* counter index (offset) */ +struct p4_event_bind { + unsigned int opcode; /* Event code and ESCR selector */ + unsigned int escr_msr[2]; /* ESCR MSR for this event */ + unsigned char cntr[2][P4_CNTR_LIMIT]; /* counter index (offset), -1 on abscence */ }; -struct p4_pmu_res { - /* maps hw_conf::idx into template for ESCR sake */ - struct p4_event_template *tpl[ARCH_P4_MAX_CCCR]; +struct p4_cache_event_bind { + unsigned int metric_pebs; + unsigned int metric_vert; }; -static DEFINE_PER_CPU(struct p4_pmu_res, p4_pmu_config); +#define P4_GEN_CACHE_EVENT_BIND(name) \ + [P4_CACHE__##name] = { \ + .metric_pebs = P4_PEBS__##name, \ + .metric_vert = P4_VERT__##name, \ + } + +static struct p4_cache_event_bind p4_cache_event_bind_map[] = { + P4_GEN_CACHE_EVENT_BIND(1stl_cache_load_miss_retired), + P4_GEN_CACHE_EVENT_BIND(2ndl_cache_load_miss_retired), + P4_GEN_CACHE_EVENT_BIND(dtlb_load_miss_retired), + P4_GEN_CACHE_EVENT_BIND(dtlb_store_miss_retired), +}; + +/* + * Note that we don't use CCCR1 here, there is an + * exception for P4_BSQ_ALLOCATION but we just have + * no workaround + * + * consider this binding as resources which particular + * event may borrow, it doesn't contain EventMask, + * Tags and friends -- they are left to a caller + */ +static struct p4_event_bind p4_event_bind_map[] = { + [P4_EVENT_TC_DELIVER_MODE] = { + .opcode = P4_OPCODE(P4_EVENT_TC_DELIVER_MODE), + .escr_msr = { MSR_P4_TC_ESCR0, MSR_P4_TC_ESCR1 }, + .cntr = { {4, 5, -1}, {6, 7, -1} }, + }, + [P4_EVENT_BPU_FETCH_REQUEST] = { + .opcode = P4_OPCODE(P4_EVENT_BPU_FETCH_REQUEST), + .escr_msr = { MSR_P4_BPU_ESCR0, MSR_P4_BPU_ESCR1 }, + .cntr = { {0, -1, -1}, {2, -1, -1} }, + }, + [P4_EVENT_ITLB_REFERENCE] = { + .opcode = P4_OPCODE(P4_EVENT_ITLB_REFERENCE), + .escr_msr = { MSR_P4_ITLB_ESCR0, MSR_P4_ITLB_ESCR1 }, + .cntr = { {0, -1, -1}, {2, -1, -1} }, + }, + [P4_EVENT_MEMORY_CANCEL] = { + .opcode = P4_OPCODE(P4_EVENT_MEMORY_CANCEL), + .escr_msr = { MSR_P4_DAC_ESCR0, MSR_P4_DAC_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_MEMORY_COMPLETE] = { + .opcode = P4_OPCODE(P4_EVENT_MEMORY_COMPLETE), + .escr_msr = { MSR_P4_SAAT_ESCR0 , MSR_P4_SAAT_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_LOAD_PORT_REPLAY] = { + .opcode = P4_OPCODE(P4_EVENT_LOAD_PORT_REPLAY), + .escr_msr = { MSR_P4_SAAT_ESCR0, MSR_P4_SAAT_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_STORE_PORT_REPLAY] = { + .opcode = P4_OPCODE(P4_EVENT_STORE_PORT_REPLAY), + .escr_msr = { MSR_P4_SAAT_ESCR0 , MSR_P4_SAAT_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_MOB_LOAD_REPLAY] = { + .opcode = P4_OPCODE(P4_EVENT_MOB_LOAD_REPLAY), + .escr_msr = { MSR_P4_MOB_ESCR0, MSR_P4_MOB_ESCR1 }, + .cntr = { {0, -1, -1}, {2, -1, -1} }, + }, + [P4_EVENT_PAGE_WALK_TYPE] = { + .opcode = P4_OPCODE(P4_EVENT_PAGE_WALK_TYPE), + .escr_msr = { MSR_P4_PMH_ESCR0, MSR_P4_PMH_ESCR1 }, + .cntr = { {0, -1, -1}, {2, -1, -1} }, + }, + [P4_EVENT_BSQ_CACHE_REFERENCE] = { + .opcode = P4_OPCODE(P4_EVENT_BSQ_CACHE_REFERENCE), + .escr_msr = { MSR_P4_BSU_ESCR0, MSR_P4_BSU_ESCR1 }, + .cntr = { {0, -1, -1}, {2, -1, -1} }, + }, + [P4_EVENT_IOQ_ALLOCATION] = { + .opcode = P4_OPCODE(P4_EVENT_IOQ_ALLOCATION), + .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, + .cntr = { {0, -1, -1}, {2, -1, -1} }, + }, + [P4_EVENT_IOQ_ACTIVE_ENTRIES] = { /* shared ESCR */ + .opcode = P4_OPCODE(P4_EVENT_IOQ_ACTIVE_ENTRIES), + .escr_msr = { MSR_P4_FSB_ESCR1, MSR_P4_FSB_ESCR1 }, + .cntr = { {2, -1, -1}, {3, -1, -1} }, + }, + [P4_EVENT_FSB_DATA_ACTIVITY] = { + .opcode = P4_OPCODE(P4_EVENT_FSB_DATA_ACTIVITY), + .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, + .cntr = { {0, -1, -1}, {2, -1, -1} }, + }, + [P4_EVENT_BSQ_ALLOCATION] = { /* shared ESCR, broken CCCR1 */ + .opcode = P4_OPCODE(P4_EVENT_BSQ_ALLOCATION), + .escr_msr = { MSR_P4_BSU_ESCR0, MSR_P4_BSU_ESCR0 }, + .cntr = { {0, -1, -1}, {1, -1, -1} }, + }, + [P4_EVENT_BSQ_ACTIVE_ENTRIES] = { /* shared ESCR */ + .opcode = P4_OPCODE(P4_EVENT_BSQ_ACTIVE_ENTRIES), + .escr_msr = { MSR_P4_BSU_ESCR1 , MSR_P4_BSU_ESCR1 }, + .cntr = { {2, -1, -1}, {3, -1, -1} }, + }, + [P4_EVENT_SSE_INPUT_ASSIST] = { + .opcode = P4_OPCODE(P4_EVENT_SSE_INPUT_ASSIST), + .escr_msr = { MSR_P4_FIRM_ESCR0, MSR_P4_FIRM_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_PACKED_SP_UOP] = { + .opcode = P4_OPCODE(P4_EVENT_PACKED_SP_UOP), + .escr_msr = { MSR_P4_FIRM_ESCR0, MSR_P4_FIRM_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_PACKED_DP_UOP] = { + .opcode = P4_OPCODE(P4_EVENT_PACKED_DP_UOP), + .escr_msr = { MSR_P4_FIRM_ESCR0, MSR_P4_FIRM_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_SCALAR_SP_UOP] = { + .opcode = P4_OPCODE(P4_EVENT_SCALAR_SP_UOP), + .escr_msr = { MSR_P4_FIRM_ESCR0, MSR_P4_FIRM_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_SCALAR_DP_UOP] = { + .opcode = P4_OPCODE(P4_EVENT_SCALAR_DP_UOP), + .escr_msr = { MSR_P4_FIRM_ESCR0, MSR_P4_FIRM_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_64BIT_MMX_UOP] = { + .opcode = P4_OPCODE(P4_EVENT_64BIT_MMX_UOP), + .escr_msr = { MSR_P4_FIRM_ESCR0, MSR_P4_FIRM_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_128BIT_MMX_UOP] = { + .opcode = P4_OPCODE(P4_EVENT_128BIT_MMX_UOP), + .escr_msr = { MSR_P4_FIRM_ESCR0, MSR_P4_FIRM_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_X87_FP_UOP] = { + .opcode = P4_OPCODE(P4_EVENT_X87_FP_UOP), + .escr_msr = { MSR_P4_FIRM_ESCR0, MSR_P4_FIRM_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_TC_MISC] = { + .opcode = P4_OPCODE(P4_EVENT_TC_MISC), + .escr_msr = { MSR_P4_TC_ESCR0, MSR_P4_TC_ESCR1 }, + .cntr = { {4, 5, -1}, {6, 7, -1} }, + }, + [P4_EVENT_GLOBAL_POWER_EVENTS] = { + .opcode = P4_OPCODE(P4_EVENT_GLOBAL_POWER_EVENTS), + .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, + .cntr = { {0, -1, -1}, {2, -1, -1} }, + }, + [P4_EVENT_TC_MS_XFER] = { + .opcode = P4_OPCODE(P4_EVENT_TC_MS_XFER), + .escr_msr = { MSR_P4_MS_ESCR0, MSR_P4_MS_ESCR1 }, + .cntr = { {4, 5, -1}, {6, 7, -1} }, + }, + [P4_EVENT_UOP_QUEUE_WRITES] = { + .opcode = P4_OPCODE(P4_EVENT_UOP_QUEUE_WRITES), + .escr_msr = { MSR_P4_MS_ESCR0, MSR_P4_MS_ESCR1 }, + .cntr = { {4, 5, -1}, {6, 7, -1} }, + }, + [P4_EVENT_RETIRED_MISPRED_BRANCH_TYPE] = { + .opcode = P4_OPCODE(P4_EVENT_RETIRED_MISPRED_BRANCH_TYPE), + .escr_msr = { MSR_P4_TBPU_ESCR0 , MSR_P4_TBPU_ESCR0 }, + .cntr = { {4, 5, -1}, {6, 7, -1} }, + }, + [P4_EVENT_RETIRED_BRANCH_TYPE] = { + .opcode = P4_OPCODE(P4_EVENT_RETIRED_BRANCH_TYPE), + .escr_msr = { MSR_P4_TBPU_ESCR0 , MSR_P4_TBPU_ESCR1 }, + .cntr = { {4, 5, -1}, {6, 7, -1} }, + }, + [P4_EVENT_RESOURCE_STALL] = { + .opcode = P4_OPCODE(P4_EVENT_RESOURCE_STALL), + .escr_msr = { MSR_P4_ALF_ESCR0, MSR_P4_ALF_ESCR1 }, + .cntr = { {12, 13, 16}, {14, 15, 17} }, + }, + [P4_EVENT_WC_BUFFER] = { + .opcode = P4_OPCODE(P4_EVENT_WC_BUFFER), + .escr_msr = { MSR_P4_DAC_ESCR0, MSR_P4_DAC_ESCR1 }, + .cntr = { {8, 9, -1}, {10, 11, -1} }, + }, + [P4_EVENT_B2B_CYCLES] = { + .opcode = P4_OPCODE(P4_EVENT_B2B_CYCLES), + .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, + .cntr = { {0, -1, -1}, {2, -1, -1} }, + }, + [P4_EVENT_BNR] = { + .opcode = P4_OPCODE(P4_EVENT_BNR), + .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, + .cntr = { {0, -1, -1}, {2, -1, -1} }, + }, + [P4_EVENT_SNOOP] = { + .opcode = P4_OPCODE(P4_EVENT_SNOOP), + .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, + .cntr = { {0, -1, -1}, {2, -1, -1} }, + }, + [P4_EVENT_RESPONSE] = { + .opcode = P4_OPCODE(P4_EVENT_RESPONSE), + .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, + .cntr = { {0, -1, -1}, {2, -1, -1} }, + }, + [P4_EVENT_FRONT_END_EVENT] = { + .opcode = P4_OPCODE(P4_EVENT_FRONT_END_EVENT), + .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR3 }, + .cntr = { {12, 13, 16}, {14, 15, 17} }, + }, + [P4_EVENT_EXECUTION_EVENT] = { + .opcode = P4_OPCODE(P4_EVENT_EXECUTION_EVENT), + .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR3 }, + .cntr = { {12, 13, 16}, {14, 15, 17} }, + }, + [P4_EVENT_REPLAY_EVENT] = { + .opcode = P4_OPCODE(P4_EVENT_REPLAY_EVENT), + .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR3 }, + .cntr = { {12, 13, 16}, {14, 15, 17} }, + }, + [P4_EVENT_INSTR_RETIRED] = { + .opcode = P4_OPCODE(P4_EVENT_INSTR_RETIRED), + .escr_msr = { MSR_P4_CRU_ESCR0, MSR_P4_CRU_ESCR1 }, + .cntr = { {12, 13, 16}, {14, 15, 17} }, + }, + [P4_EVENT_UOPS_RETIRED] = { + .opcode = P4_OPCODE(P4_EVENT_UOPS_RETIRED), + .escr_msr = { MSR_P4_CRU_ESCR0, MSR_P4_CRU_ESCR1 }, + .cntr = { {12, 13, 16}, {14, 15, 17} }, + }, + [P4_EVENT_UOP_TYPE] = { + .opcode = P4_OPCODE(P4_EVENT_UOP_TYPE), + .escr_msr = { MSR_P4_RAT_ESCR0, MSR_P4_RAT_ESCR1 }, + .cntr = { {12, 13, 16}, {14, 15, 17} }, + }, + [P4_EVENT_BRANCH_RETIRED] = { + .opcode = P4_OPCODE(P4_EVENT_BRANCH_RETIRED), + .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR3 }, + .cntr = { {12, 13, 16}, {14, 15, 17} }, + }, + [P4_EVENT_MISPRED_BRANCH_RETIRED] = { + .opcode = P4_OPCODE(P4_EVENT_MISPRED_BRANCH_RETIRED), + .escr_msr = { MSR_P4_CRU_ESCR0, MSR_P4_CRU_ESCR1 }, + .cntr = { {12, 13, 16}, {14, 15, 17} }, + }, + [P4_EVENT_X87_ASSIST] = { + .opcode = P4_OPCODE(P4_EVENT_X87_ASSIST), + .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR3 }, + .cntr = { {12, 13, 16}, {14, 15, 17} }, + }, + [P4_EVENT_MACHINE_CLEAR] = { + .opcode = P4_OPCODE(P4_EVENT_MACHINE_CLEAR), + .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR3 }, + .cntr = { {12, 13, 16}, {14, 15, 17} }, + }, + [P4_EVENT_INSTR_COMPLETED] = { + .opcode = P4_OPCODE(P4_EVENT_INSTR_COMPLETED), + .escr_msr = { MSR_P4_CRU_ESCR0, MSR_P4_CRU_ESCR1 }, + .cntr = { {12, 13, 16}, {14, 15, 17} }, + }, +}; -#define P4_CACHE_EVENT_CONFIG(event, bit) \ - p4_config_pack_escr(P4_EVENT_UNPACK_EVENT(event) << P4_EVNTSEL_EVENT_SHIFT) | \ - p4_config_pack_escr((event##_##bit) << P4_EVNTSEL_EVENTMASK_SHIFT) | \ - p4_config_pack_cccr(P4_EVENT_UNPACK_SELECTOR(event) << P4_CCCR_ESCR_SELECT_SHIFT) +#define P4_GEN_CACHE_EVENT(event, bit, cache_event) \ + p4_config_pack_escr(P4_ESCR_EVENT(event) | \ + P4_ESCR_EMASK_BIT(event, bit)) | \ + p4_config_pack_cccr(cache_event | \ + P4_CCCR_ESEL(P4_OPCODE_ESEL(P4_OPCODE(event)))) static __initconst u64 p4_hw_cache_event_ids [PERF_COUNT_HW_CACHE_MAX] @@ -49,42 +295,35 @@ static __initconst u64 p4_hw_cache_event_ids [ C(L1D ) ] = { [ C(OP_READ) ] = { [ C(RESULT_ACCESS) ] = 0x0, - /* 1stL_cache_load_miss_retired */ - [ C(RESULT_MISS) ] = P4_CACHE_EVENT_CONFIG(P4_REPLAY_EVENT, NBOGUS) - | KEY_P4_L1D_OP_READ_RESULT_MISS, + [ C(RESULT_MISS) ] = P4_GEN_CACHE_EVENT(P4_EVENT_REPLAY_EVENT, NBOGUS, + P4_CACHE__1stl_cache_load_miss_retired), }, }, [ C(LL ) ] = { [ C(OP_READ) ] = { [ C(RESULT_ACCESS) ] = 0x0, - /* 2ndL_cache_load_miss_retired */ - [ C(RESULT_MISS) ] = P4_CACHE_EVENT_CONFIG(P4_REPLAY_EVENT, NBOGUS) - | KEY_P4_LL_OP_READ_RESULT_MISS, + [ C(RESULT_MISS) ] = P4_GEN_CACHE_EVENT(P4_EVENT_REPLAY_EVENT, NBOGUS, + P4_CACHE__2ndl_cache_load_miss_retired), }, - }, +}, [ C(DTLB) ] = { [ C(OP_READ) ] = { [ C(RESULT_ACCESS) ] = 0x0, - /* DTLB_load_miss_retired */ - [ C(RESULT_MISS) ] = P4_CACHE_EVENT_CONFIG(P4_REPLAY_EVENT, NBOGUS) - | KEY_P4_DTLB_OP_READ_RESULT_MISS, + [ C(RESULT_MISS) ] = P4_GEN_CACHE_EVENT(P4_EVENT_REPLAY_EVENT, NBOGUS, + P4_CACHE__dtlb_load_miss_retired), }, [ C(OP_WRITE) ] = { [ C(RESULT_ACCESS) ] = 0x0, - /* DTLB_store_miss_retired */ - [ C(RESULT_MISS) ] = P4_CACHE_EVENT_CONFIG(P4_REPLAY_EVENT, NBOGUS) - | KEY_P4_DTLB_OP_WRITE_RESULT_MISS, + [ C(RESULT_MISS) ] = P4_GEN_CACHE_EVENT(P4_EVENT_REPLAY_EVENT, NBOGUS, + P4_CACHE__dtlb_store_miss_retired), }, }, [ C(ITLB) ] = { [ C(OP_READ) ] = { - /* ITLB_reference.HIT */ - [ C(RESULT_ACCESS) ] = P4_CACHE_EVENT_CONFIG(P4_ITLB_REFERENCE, HIT) - | KEY_P4_ITLB_OP_READ_RESULT_ACCESS, - - /* ITLB_reference.MISS */ - [ C(RESULT_MISS) ] = P4_CACHE_EVENT_CONFIG(P4_ITLB_REFERENCE, MISS) - | KEY_P4_ITLB_OP_READ_RESULT_MISS, + [ C(RESULT_ACCESS) ] = P4_GEN_CACHE_EVENT(P4_EVENT_ITLB_REFERENCE, HIT, + P4_CACHE__itlb_reference_hit), + [ C(RESULT_MISS) ] = P4_GEN_CACHE_EVENT(P4_EVENT_ITLB_REFERENCE, MISS, + P4_CACHE__itlb_reference_miss), }, [ C(OP_WRITE) ] = { [ C(RESULT_ACCESS) ] = -1, @@ -97,220 +336,90 @@ static __initconst u64 p4_hw_cache_event_ids }, }; -/* - * WARN: CCCR1 doesn't have a working enable bit so try to not - * use it if possible - * - * Also as only we start to support raw events we will need to - * append _all_ P4_EVENT_PACK'ed events here - */ -struct p4_event_template p4_templates[] = { - [0] = { - .opcode = P4_GLOBAL_POWER_EVENTS, - .config = 0, - .dep = -1, - .key = 0, - .emask = - P4_EVENT_ATTR(P4_GLOBAL_POWER_EVENTS, RUNNING), - .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, - .cntr = { 0, 2 }, - }, - [1] = { - .opcode = P4_INSTR_RETIRED, - .config = 0, - .dep = -1, /* needs front-end tagging */ - .key = 1, - .emask = - P4_EVENT_ATTR(P4_INSTR_RETIRED, NBOGUSNTAG) | - P4_EVENT_ATTR(P4_INSTR_RETIRED, BOGUSNTAG), - .escr_msr = { MSR_P4_CRU_ESCR0, MSR_P4_CRU_ESCR1 }, - .cntr = { 12, 14 }, - }, - [2] = { - .opcode = P4_BSQ_CACHE_REFERENCE, - .config = 0, - .dep = -1, - .key = 2, - .emask = - P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITS) | - P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITE) | - P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_HITM) | - P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_HITS) | - P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_HITE) | - P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_HITM), - .escr_msr = { MSR_P4_BSU_ESCR0, MSR_P4_BSU_ESCR1 }, - .cntr = { 0, 2 }, - }, - [3] = { - .opcode = P4_BSQ_CACHE_REFERENCE, - .config = 0, - .dep = -1, - .key = 3, - .emask = - P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_2ndL_MISS) | - P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, RD_3rdL_MISS) | - P4_EVENT_ATTR(P4_BSQ_CACHE_REFERENCE, WR_2ndL_MISS), - .escr_msr = { MSR_P4_BSU_ESCR0, MSR_P4_BSU_ESCR1 }, - .cntr = { 0, 3 }, - }, - [4] = { - .opcode = P4_RETIRED_BRANCH_TYPE, - .config = 0, - .dep = -1, - .key = 4, - .emask = - P4_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, CONDITIONAL) | - P4_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, CALL) | - P4_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, RETURN) | - P4_EVENT_ATTR(P4_RETIRED_BRANCH_TYPE, INDIRECT), - .escr_msr = { MSR_P4_TBPU_ESCR0, MSR_P4_TBPU_ESCR1 }, - .cntr = { 4, 6 }, - }, - [5] = { - .opcode = P4_MISPRED_BRANCH_RETIRED, - .config = 0, - .dep = -1, - .key = 5, - .emask = - P4_EVENT_ATTR(P4_MISPRED_BRANCH_RETIRED, NBOGUS), - .escr_msr = { MSR_P4_CRU_ESCR0, MSR_P4_CRU_ESCR1 }, - .cntr = { 12, 14 }, - }, - [6] = { - .opcode = P4_FSB_DATA_ACTIVITY, - .config = p4_config_pack_cccr(P4_CCCR_EDGE | P4_CCCR_COMPARE), - .dep = -1, - .key = 6, - .emask = - P4_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DRDY_DRV) | - P4_EVENT_ATTR(P4_FSB_DATA_ACTIVITY, DRDY_OWN), - .escr_msr = { MSR_P4_FSB_ESCR0, MSR_P4_FSB_ESCR1 }, - .cntr = { 0, 2 }, - }, - [KEY_P4_L1D_OP_READ_RESULT_MISS] = { - .opcode = P4_REPLAY_EVENT, - .config = 0, - .dep = -1, - .msr = (u64)(1 << 0 | 1 << 24) << 32 | (1 << 0), - .key = KEY_P4_L1D_OP_READ_RESULT_MISS, - .emask = - P4_EVENT_ATTR(P4_REPLAY_EVENT, NBOGUS), - .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR2 }, - .cntr = { 16, 17 }, - }, - [KEY_P4_LL_OP_READ_RESULT_MISS] = { - .opcode = P4_REPLAY_EVENT, - .config = 0, - .dep = -1, - .msr = (u64)(1 << 1 | 1 << 24) << 32 | (1 << 0), - .key = KEY_P4_LL_OP_READ_RESULT_MISS, - .emask = - P4_EVENT_ATTR(P4_REPLAY_EVENT, NBOGUS), - .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR2 }, - .cntr = { 16, 17 }, - }, - [KEY_P4_DTLB_OP_READ_RESULT_MISS] = { - .opcode = P4_REPLAY_EVENT, - .config = 0, - .dep = -1, - .msr = (u64)(1 << 2 | 1 << 24) << 32 | (1 << 0), - .key = KEY_P4_DTLB_OP_READ_RESULT_MISS, - .emask = - P4_EVENT_ATTR(P4_REPLAY_EVENT, NBOGUS), - .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR2 }, - .cntr = { 16, 17 }, - }, - [KEY_P4_DTLB_OP_WRITE_RESULT_MISS] = { - .opcode = P4_REPLAY_EVENT, - .config = 0, - .dep = -1, - .msr = (u64)(1 << 2 | 1 << 24) << 32 | (1 << 1), - .key = KEY_P4_DTLB_OP_WRITE_RESULT_MISS, - .emask = - P4_EVENT_ATTR(P4_REPLAY_EVENT, NBOGUS), - .escr_msr = { MSR_P4_CRU_ESCR2, MSR_P4_CRU_ESCR2 }, - .cntr = { 16, 17 }, - }, - [KEY_P4_ITLB_OP_READ_RESULT_ACCESS] = { - .opcode = P4_ITLB_REFERENCE, - .config = 0, - .dep = -1, - .msr = 0, - .key = KEY_P4_ITLB_OP_READ_RESULT_ACCESS, - .emask = - P4_EVENT_ATTR(P4_ITLB_REFERENCE, HIT), - .escr_msr = { MSR_P4_ITLB_ESCR0, MSR_P4_ITLB_ESCR1 }, - .cntr = { 0, 2 }, - }, - [KEY_P4_ITLB_OP_READ_RESULT_MISS] = { - .opcode = P4_ITLB_REFERENCE, - .config = 0, - .dep = -1, - .msr = 0, - .key = KEY_P4_ITLB_OP_READ_RESULT_MISS, - .emask = - P4_EVENT_ATTR(P4_ITLB_REFERENCE, MISS), - .escr_msr = { MSR_P4_ITLB_ESCR0, MSR_P4_ITLB_ESCR1 }, - .cntr = { 0, 2 }, - }, - [KEY_P4_UOP_TYPE] = { - .opcode = P4_UOP_TYPE, - .config = 0, - .dep = -1, - .key = KEY_P4_UOP_TYPE, - .emask = - P4_EVENT_ATTR(P4_UOP_TYPE, TAGLOADS) | - P4_EVENT_ATTR(P4_UOP_TYPE, TAGSTORES), - .escr_msr = { MSR_P4_RAT_ESCR0, MSR_P4_RAT_ESCR1 }, - .cntr = { 16, 17 }, - }, +static u64 p4_general_events[PERF_COUNT_HW_MAX] = { + /* non-halted CPU clocks */ + [PERF_COUNT_HW_CPU_CYCLES] = + p4_config_pack_escr(P4_ESCR_EVENT(P4_EVENT_GLOBAL_POWER_EVENTS) | + P4_ESCR_EMASK_BIT(P4_EVENT_GLOBAL_POWER_EVENTS, RUNNING)), + + /* + * retired instructions + * in a sake of simplicity we don't use the FSB tagging + */ + [PERF_COUNT_HW_INSTRUCTIONS] = + p4_config_pack_escr(P4_ESCR_EVENT(P4_EVENT_INSTR_RETIRED) | + P4_ESCR_EMASK_BIT(P4_EVENT_INSTR_RETIRED, NBOGUSNTAG) | + P4_ESCR_EMASK_BIT(P4_EVENT_INSTR_RETIRED, BOGUSNTAG)), + + /* cache hits */ + [PERF_COUNT_HW_CACHE_REFERENCES] = + p4_config_pack_escr(P4_ESCR_EVENT(P4_EVENT_BSQ_CACHE_REFERENCE) | + P4_ESCR_EMASK_BIT(P4_EVENT_BSQ_CACHE_REFERENCE, RD_2ndL_HITS) | + P4_ESCR_EMASK_BIT(P4_EVENT_BSQ_CACHE_REFERENCE, RD_2ndL_HITE) | + P4_ESCR_EMASK_BIT(P4_EVENT_BSQ_CACHE_REFERENCE, RD_2ndL_HITM) | + P4_ESCR_EMASK_BIT(P4_EVENT_BSQ_CACHE_REFERENCE, RD_3rdL_HITS) | + P4_ESCR_EMASK_BIT(P4_EVENT_BSQ_CACHE_REFERENCE, RD_3rdL_HITE) | + P4_ESCR_EMASK_BIT(P4_EVENT_BSQ_CACHE_REFERENCE, RD_3rdL_HITM)), + + /* cache misses */ + [PERF_COUNT_HW_CACHE_MISSES] = + p4_config_pack_escr(P4_ESCR_EVENT(P4_EVENT_BSQ_CACHE_REFERENCE) | + P4_ESCR_EMASK_BIT(P4_EVENT_BSQ_CACHE_REFERENCE, RD_2ndL_MISS) | + P4_ESCR_EMASK_BIT(P4_EVENT_BSQ_CACHE_REFERENCE, RD_3rdL_MISS) | + P4_ESCR_EMASK_BIT(P4_EVENT_BSQ_CACHE_REFERENCE, WR_2ndL_MISS)), + + /* branch instructions retired */ + [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = + p4_config_pack_escr(P4_ESCR_EVENT(P4_EVENT_RETIRED_BRANCH_TYPE) | + P4_ESCR_EMASK_BIT(P4_EVENT_RETIRED_BRANCH_TYPE, CONDITIONAL) | + P4_ESCR_EMASK_BIT(P4_EVENT_RETIRED_BRANCH_TYPE, CALL) | + P4_ESCR_EMASK_BIT(P4_EVENT_RETIRED_BRANCH_TYPE, RETURN) | + P4_ESCR_EMASK_BIT(P4_EVENT_RETIRED_BRANCH_TYPE, INDIRECT)), + + /* mispredicted branches retired */ + [PERF_COUNT_HW_BRANCH_MISSES] = + p4_config_pack_escr(P4_ESCR_EVENT(P4_EVENT_MISPRED_BRANCH_RETIRED) | + P4_ESCR_EMASK_BIT(P4_EVENT_MISPRED_BRANCH_RETIRED, NBOGUS)), + + /* bus ready clocks (cpu is driving #DRDY_DRV\#DRDY_OWN): */ + [PERF_COUNT_HW_BUS_CYCLES] = + p4_config_pack_escr(P4_ESCR_EVENT(P4_EVENT_FSB_DATA_ACTIVITY) | + P4_ESCR_EMASK_BIT(P4_EVENT_FSB_DATA_ACTIVITY, DRDY_DRV) | + P4_ESCR_EMASK_BIT(P4_EVENT_FSB_DATA_ACTIVITY, DRDY_OWN)) | + p4_config_pack_cccr(P4_CCCR_EDGE | P4_CCCR_COMPARE), }; +static struct p4_event_bind *p4_config_get_bind(u64 config) +{ + unsigned int evnt = p4_config_unpack_event(config); + struct p4_event_bind *bind = NULL; + + if (evnt < ARRAY_SIZE(p4_event_bind_map)) + bind = &p4_event_bind_map[evnt]; + + return bind; +} + static u64 p4_pmu_event_map(int hw_event) { - struct p4_event_template *tpl; + struct p4_event_bind *bind; + unsigned int esel; u64 config; - if (hw_event > ARRAY_SIZE(p4_templates)) { - printk_once(KERN_ERR "PMU: Incorrect event index\n"); + if (hw_event > ARRAY_SIZE(p4_general_events)) { + printk_once(KERN_ERR "P4 PMU: Bad index: %i\n", hw_event); return 0; } - tpl = &p4_templates[hw_event]; - /* - * fill config up according to - * a predefined event template - */ - config = tpl->config; - config |= p4_config_pack_escr(P4_EVENT_UNPACK_EVENT(tpl->opcode) << P4_EVNTSEL_EVENT_SHIFT); - config |= p4_config_pack_escr(tpl->emask << P4_EVNTSEL_EVENTMASK_SHIFT); - config |= p4_config_pack_cccr(P4_EVENT_UNPACK_SELECTOR(tpl->opcode) << P4_CCCR_ESCR_SELECT_SHIFT); - config |= p4_config_pack_cccr(hw_event & P4_CCCR_RESERVED); + config = p4_general_events[hw_event]; + bind = p4_config_get_bind(config); + esel = P4_OPCODE_ESEL(bind->opcode); + config |= p4_config_pack_cccr(P4_CCCR_ESEL(esel)); return config; } /* - * Note that we still have 5 events (from global events SDM list) - * intersected in opcode+emask bits so we will need another - * scheme there do distinguish templates. - */ -static inline int p4_pmu_emask_match(unsigned int dst, unsigned int src) -{ - return dst & src; -} - -static struct p4_event_template *p4_pmu_template_lookup(u64 config) -{ - int key = p4_config_unpack_key(config); - - if (key < ARRAY_SIZE(p4_templates)) - return &p4_templates[key]; - else - return NULL; -} - -/* * We don't control raw events so it's up to the caller * to pass sane values (and we don't count the thread number * on HT machine but allow HT-compatible specifics to be @@ -319,13 +428,14 @@ static struct p4_event_template *p4_pmu_template_lookup(u64 config) static u64 p4_pmu_raw_event(u64 hw_event) { return hw_event & - (p4_config_pack_escr(P4_EVNTSEL_MASK_HT) | + (p4_config_pack_escr(P4_ESCR_MASK_HT) | p4_config_pack_cccr(P4_CCCR_MASK_HT)); } static int p4_hw_config(struct perf_event_attr *attr, struct hw_perf_event *hwc) { int cpu = raw_smp_processor_id(); + u32 escr, cccr; /* * the reason we use cpu that early is that: if we get scheduled @@ -333,13 +443,10 @@ static int p4_hw_config(struct perf_event_attr *attr, struct hw_perf_event *hwc) * specific flags in config (and will save some cpu cycles) */ - /* CCCR by default */ - hwc->config = p4_config_pack_cccr(p4_default_cccr_conf(cpu)); + cccr = p4_default_cccr_conf(cpu); + escr = p4_default_escr_conf(cpu, attr->exclude_kernel, attr->exclude_user); + hwc->config = p4_config_pack_escr(escr) | p4_config_pack_cccr(cccr); - /* Count user and OS events unless not requested to */ - hwc->config |= p4_config_pack_escr(p4_default_escr_conf(cpu, attr->exclude_kernel, - attr->exclude_user)); - /* on HT machine we need a special bit */ if (p4_ht_active() && p4_ht_thread(cpu)) hwc->config = p4_set_ht_bit(hwc->config); @@ -368,7 +475,7 @@ static inline void p4_pmu_disable_event(struct perf_event *event) */ (void)checking_wrmsrl(hwc->config_base + hwc->idx, (u64)(p4_config_unpack_cccr(hwc->config)) & - ~P4_CCCR_ENABLE & ~P4_CCCR_OVF); + ~P4_CCCR_ENABLE & ~P4_CCCR_OVF & ~P4_CCCR_RESERVED); } static void p4_pmu_disable_all(void) @@ -389,27 +496,14 @@ static void p4_pmu_enable_event(struct perf_event *event) struct hw_perf_event *hwc = &event->hw; int thread = p4_ht_config_thread(hwc->config); u64 escr_conf = p4_config_unpack_escr(p4_clear_ht_bit(hwc->config)); - u64 escr_base; - struct p4_event_template *tpl; - struct p4_pmu_res *c; + unsigned int idx = p4_config_unpack_event(hwc->config); + unsigned int idx_cache = p4_config_unpack_cache_event(hwc->config); + struct p4_event_bind *bind; + struct p4_cache_event_bind *bind_cache; + u64 escr_addr, cccr; - /* - * some preparation work from per-cpu private fields - * since we need to find out which ESCR to use - */ - c = &__get_cpu_var(p4_pmu_config); - tpl = c->tpl[hwc->idx]; - if (!tpl) { - pr_crit("%s: Wrong index: %d\n", __func__, hwc->idx); - return; - } - - if (tpl->msr) { - (void)checking_wrmsrl(MSR_IA32_PEBS_ENABLE, tpl->msr >> 32); - (void)checking_wrmsrl(MSR_P4_PEBS_MATRIX_VERT, tpl->msr & 0xffffffff); - } - - escr_base = (u64)tpl->escr_msr[thread]; + bind = &p4_event_bind_map[idx]; + escr_addr = (u64)bind->escr_msr[thread]; /* * - we dont support cascaded counters yet @@ -418,9 +512,27 @@ static void p4_pmu_enable_event(struct perf_event *event) WARN_ON_ONCE(p4_is_event_cascaded(hwc->config)); WARN_ON_ONCE(hwc->idx == 1); - (void)checking_wrmsrl(escr_base, escr_conf); + /* we need a real Event value */ + escr_conf &= ~P4_ESCR_EVENT_MASK; + escr_conf |= P4_ESCR_EVENT(P4_OPCODE_EVNT(bind->opcode)); + + cccr = p4_config_unpack_cccr(hwc->config); + + /* + * it could be Cache event so that we need to + * set metrics into additional MSRs + */ + BUILD_BUG_ON(P4_CACHE__MAX > P4_CCCR_CACHE_OPS_MASK); + if (idx_cache > P4_CACHE__NONE && + idx_cache < ARRAY_SIZE(p4_cache_event_bind_map)) { + bind_cache = &p4_cache_event_bind_map[idx_cache]; + (void)checking_wrmsrl(MSR_IA32_PEBS_ENABLE, (u64)bind_cache->metric_pebs); + (void)checking_wrmsrl(MSR_P4_PEBS_MATRIX_VERT, (u64)bind_cache->metric_vert); + } + + (void)checking_wrmsrl(escr_addr, escr_conf); (void)checking_wrmsrl(hwc->config_base + hwc->idx, - (u64)(p4_config_unpack_cccr(hwc->config)) | P4_CCCR_ENABLE); + (cccr & ~P4_CCCR_RESERVED) | P4_CCCR_ENABLE); } static void p4_pmu_enable_all(void) @@ -516,13 +628,13 @@ static void p4_pmu_swap_config_ts(struct hw_perf_event *hwc, int cpu) if (p4_ht_thread(cpu)) { cccr &= ~P4_CCCR_OVF_PMI_T0; cccr |= P4_CCCR_OVF_PMI_T1; - if (escr & P4_EVNTSEL_T0_OS) { - escr &= ~P4_EVNTSEL_T0_OS; - escr |= P4_EVNTSEL_T1_OS; + if (escr & P4_ESCR_T0_OS) { + escr &= ~P4_ESCR_T0_OS; + escr |= P4_ESCR_T1_OS; } - if (escr & P4_EVNTSEL_T0_USR) { - escr &= ~P4_EVNTSEL_T0_USR; - escr |= P4_EVNTSEL_T1_USR; + if (escr & P4_ESCR_T0_USR) { + escr &= ~P4_ESCR_T0_USR; + escr |= P4_ESCR_T1_USR; } hwc->config = p4_config_pack_escr(escr); hwc->config |= p4_config_pack_cccr(cccr); @@ -530,13 +642,13 @@ static void p4_pmu_swap_config_ts(struct hw_perf_event *hwc, int cpu) } else { cccr &= ~P4_CCCR_OVF_PMI_T1; cccr |= P4_CCCR_OVF_PMI_T0; - if (escr & P4_EVNTSEL_T1_OS) { - escr &= ~P4_EVNTSEL_T1_OS; - escr |= P4_EVNTSEL_T0_OS; + if (escr & P4_ESCR_T1_OS) { + escr &= ~P4_ESCR_T1_OS; + escr |= P4_ESCR_T0_OS; } - if (escr & P4_EVNTSEL_T1_USR) { - escr &= ~P4_EVNTSEL_T1_USR; - escr |= P4_EVNTSEL_T0_USR; + if (escr & P4_ESCR_T1_USR) { + escr &= ~P4_ESCR_T1_USR; + escr |= P4_ESCR_T0_USR; } hwc->config = p4_config_pack_escr(escr); hwc->config |= p4_config_pack_cccr(cccr); @@ -606,66 +718,56 @@ static int p4_get_escr_idx(unsigned int addr) return -1; } +static int p4_next_cntr(int thread, unsigned long *used_mask, + struct p4_event_bind *bind) +{ + int i = 0, j; + + for (i = 0; i < P4_CNTR_LIMIT; i++) { + j = bind->cntr[thread][i++]; + if (j == -1 || !test_bit(j, used_mask)) + return j; + } + + return -1; +} + static int p4_pmu_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign) { unsigned long used_mask[BITS_TO_LONGS(X86_PMC_IDX_MAX)]; unsigned long escr_mask[BITS_TO_LONGS(ARCH_P4_TOTAL_ESCR)]; - - struct hw_perf_event *hwc; - struct p4_event_template *tpl; - struct p4_pmu_res *c; int cpu = raw_smp_processor_id(); - int escr_idx, thread, i, num; + struct hw_perf_event *hwc; + struct p4_event_bind *bind; + unsigned int i, thread, num; + int cntr_idx, escr_idx; bitmap_zero(used_mask, X86_PMC_IDX_MAX); bitmap_zero(escr_mask, ARCH_P4_TOTAL_ESCR); - c = &__get_cpu_var(p4_pmu_config); - /* - * Firstly find out which resource events are going - * to use, if ESCR+CCCR tuple is already borrowed - * then get out of here - */ for (i = 0, num = n; i < n; i++, num--) { + hwc = &cpuc->event_list[i]->hw; - tpl = p4_pmu_template_lookup(hwc->config); - if (!tpl) - goto done; thread = p4_ht_thread(cpu); - escr_idx = p4_get_escr_idx(tpl->escr_msr[thread]); - if (escr_idx == -1) - goto done; + bind = p4_config_get_bind(hwc->config); + escr_idx = p4_get_escr_idx(bind->escr_msr[thread]); - /* already allocated and remains on the same cpu */ if (hwc->idx != -1 && !p4_should_swap_ts(hwc->config, cpu)) { + cntr_idx = hwc->idx; if (assign) assign[i] = hwc->idx; - /* upstream dependent event */ - if (unlikely(tpl->dep != -1)) - printk_once(KERN_WARNING "PMU: Dep events are " - "not implemented yet\n"); goto reserve; } - /* it may be already borrowed */ - if (test_bit(tpl->cntr[thread], used_mask) || - test_bit(escr_idx, escr_mask)) + cntr_idx = p4_next_cntr(thread, used_mask, bind); + if (cntr_idx == -1 || test_bit(escr_idx, escr_mask)) goto done; - /* - * ESCR+CCCR+COUNTERs are available to use lets swap - * thread specific bits, push assigned bits - * back and save template into per-cpu - * area (which will allow us to find out the ESCR - * to be used at moment of "enable event via real MSR") - */ p4_pmu_swap_config_ts(hwc, cpu); - if (assign) { - assign[i] = tpl->cntr[thread]; - c->tpl[assign[i]] = tpl; - } + if (assign) + assign[i] = cntr_idx; reserve: - set_bit(tpl->cntr[thread], used_mask); + set_bit(cntr_idx, used_mask); set_bit(escr_idx, escr_mask); } @@ -684,7 +786,7 @@ static __initconst struct x86_pmu p4_pmu = { .perfctr = MSR_P4_BPU_PERFCTR0, .event_map = p4_pmu_event_map, .raw_event = p4_pmu_raw_event, - .max_events = ARRAY_SIZE(p4_templates), + .max_events = ARRAY_SIZE(p4_general_events), .get_event_constraints = x86_get_event_constraints, /* * IF HT disabled we may need to use all @@ -716,7 +818,7 @@ static __init int p4_pmu_init(void) } memcpy(hw_cache_event_ids, p4_hw_cache_event_ids, - sizeof(hw_cache_event_ids)); + sizeof(hw_cache_event_ids)); pr_cont("Netburst events, "); -- cgit v0.10.2 From 53c540195724b52422da067a31ef6916d2c70202 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 24 Mar 2010 16:40:14 -0300 Subject: perf report: Add a popup menu to ask what operation is to be performed MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Right now it presents a menu with these options: +------------------------------+ | Annotate CURRENT_SYMBOL_NAME | | Exit | +------------------------------+ If the highlighted (current) symbol is not annotatable only the "Exit" option will appear. Also add a confirmation dialog when ESC is pressed on the top level to avoid exiting the application by pressing one too many ESC key. To get to the menu just press the -> (Right navigation key), to exit just press ESC. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269459619-982-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index 12b229b..a3465a0 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -28,6 +28,47 @@ static newtComponent newt_form__new(void) return self; } +static int popup_menu(int argc, const char *argv[]) +{ + struct newtExitStruct es; + int i, rc = -1, max_len = 5; + newtComponent listbox, form = newt_form__new(); + + if (form == NULL) + return -1; + + listbox = newtListbox(0, 0, argc, NEWT_FLAG_RETURNEXIT); + if (listbox == NULL) + goto out_destroy_form; + + newtFormAddComponents(form, listbox, NULL); + + for (i = 0; i < argc; ++i) { + int len = strlen(argv[i]); + if (len > max_len) + max_len = len; + if (newtListboxAddEntry(listbox, argv[i], (void *)(long)i)) + goto out_destroy_form; + } + + newtCenteredWindow(max_len, argc, NULL); + newtFormRun(form, &es); + rc = newtListboxGetCurrent(listbox) - NULL; + if (es.reason == NEWT_EXIT_HOTKEY) + rc = -1; + newtPopWindow(); +out_destroy_form: + newtFormDestroy(form); + return rc; +} + +static bool dialog_yesno(const char *msg) +{ + /* newtWinChoice should really be accepting const char pointers... */ + char yes[] = "Yes", no[] = "No"; + return newtWinChoice(NULL, no, yes, (char *)msg) == 2; +} + /* * When debugging newt problems it was useful to be able to "unroll" * the calls to newtCheckBoxTreeAdd{Array,Item}, so that we can generate @@ -309,6 +350,19 @@ out_free_str: free(str); } +static const void *newt__symbol_tree_get_current(newtComponent self) +{ + if (symbol_conf.use_callchain) + return newtCheckboxTreeGetCurrent(self); + return newtListboxGetCurrent(self); +} + +static void perf_session__selection(newtComponent self, void *data) +{ + const struct symbol **symbol_ptr = data; + *symbol_ptr = newt__symbol_tree_get_current(self); +} + void perf_session__browse_hists(struct rb_root *hists, u64 session_total, const char *helpline) { @@ -322,6 +376,7 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, char str[1024]; newtComponent form, tree; struct newtExitStruct es; + const struct symbol *selection; snprintf(str, sizeof(str), "Samples: %Ld", session_total); newtDrawRootText(0, 0, str); @@ -336,6 +391,8 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, tree = newtListbox(0, 0, rows - 5, (NEWT_FLAG_SCROLL | NEWT_FLAG_RETURNEXIT)); + newtComponentAddCallback(tree, perf_session__selection, &selection); + list_for_each_entry(se, &hist_entry__sort_list, list) { if (se->elide) continue; @@ -377,20 +434,43 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, form = newt_form__new(); newtFormAddHotKey(form, 'A'); newtFormAddHotKey(form, 'a'); + newtFormAddHotKey(form, NEWT_KEY_RIGHT); newtFormAddComponents(form, tree, NULL); + selection = newt__symbol_tree_get_current(tree); while (1) { - const struct symbol *selection; + char annotate[512]; + const char *options[2]; + int nr_options = 0, choice; newtFormRun(form, &es); - if (es.reason == NEWT_EXIT_HOTKEY && - toupper(es.u.key) != 'A') + if (es.reason == NEWT_EXIT_HOTKEY) { + if (toupper(es.u.key) == 'A') { + symbol__annotate_browser(selection); + continue; + } + if (es.u.key == NEWT_KEY_ESCAPE || + toupper(es.u.key) == 'Q' || + es.u.key == CTRL('c')) { + if (dialog_yesno("Do you really want to exit?")) + break; + else + continue; + } + } + + if (selection != NULL) { + snprintf(annotate, sizeof(annotate), + "Annotate %s", selection->name); + options[nr_options++] = annotate; + } + + options[nr_options++] = "Exit"; + choice = popup_menu(nr_options, options); + if (choice == nr_options - 1) break; - if (!symbol_conf.use_callchain) - selection = newtListboxGetCurrent(tree); - else - selection = newtCheckboxTreeGetCurrent(tree); - symbol__annotate_browser(selection); + else if (selection != NULL && choice >= 0) + symbol__annotate_browser(selection); } newtFormDestroy(form); -- cgit v0.10.2 From 96415e4d3f5fdf9cdb12eedfcbc58152b1e1458c Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 24 Mar 2010 16:40:15 -0300 Subject: perf symbols: Avoid unnecessary symbol loading when dso list is specified MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We were performing the full thread__find_addr_location operation, i.e. resolving to a map/dso _and_ loading its symbols when we can optimize it by first calling thread__find_addr_map to find just the map/dso, check if it is one that we are interested in (passed via --dsos/-d in 'perf annotate', 'perf report', etc) and if not avoid loading the symtab. Nice speedup when we know which DSO we're interested in. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269459619-982-2-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c index 705ec63..c2808ad 100644 --- a/tools/perf/util/event.c +++ b/tools/perf/util/event.c @@ -513,24 +513,32 @@ int event__preprocess_sample(const event_t *self, struct perf_session *session, dump_printf(" ... thread: %s:%d\n", thread->comm, thread->pid); - thread__find_addr_location(thread, session, cpumode, MAP__FUNCTION, - self->ip.ip, al, filter); + thread__find_addr_map(thread, session, cpumode, MAP__FUNCTION, + self->ip.ip, al); dump_printf(" ...... dso: %s\n", al->map ? al->map->dso->long_name : al->level == 'H' ? "[hypervisor]" : ""); - /* - * We have to do this here as we may have a dso with no symbol hit that - * has a name longer than the ones with symbols sampled. - */ - if (al->map && !sort_dso.elide && !al->map->dso->slen_calculated) - dso__calc_col_width(al->map->dso); - - if (symbol_conf.dso_list && - (!al->map || !al->map->dso || - !(strlist__has_entry(symbol_conf.dso_list, al->map->dso->short_name) || - (al->map->dso->short_name != al->map->dso->long_name && - strlist__has_entry(symbol_conf.dso_list, al->map->dso->long_name))))) - goto out_filtered; + al->sym = NULL; + + if (al->map) { + if (symbol_conf.dso_list && + (!al->map || !al->map->dso || + !(strlist__has_entry(symbol_conf.dso_list, + al->map->dso->short_name) || + (al->map->dso->short_name != al->map->dso->long_name && + strlist__has_entry(symbol_conf.dso_list, + al->map->dso->long_name))))) + goto out_filtered; + /* + * We have to do this here as we may have a dso with no symbol + * hit that has a name longer than the ones with symbols + * sampled. + */ + if (!sort_dso.elide && !al->map->dso->slen_calculated) + dso__calc_col_width(al->map->dso); + + al->sym = map__find_symbol(al->map, al->addr, filter); + } if (symbol_conf.sym_list && al->sym && !strlist__has_entry(symbol_conf.sym_list, al->sym->name)) -- cgit v0.10.2 From ac73c5a9c1767b2771e6d2b5accafdef89db04c2 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 24 Mar 2010 16:40:16 -0300 Subject: perf annotate: Allow specifying DSOs where to look for symbol MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Using the same parameter as in 'perf report', allowing to specify just one and disambiguate between DSOs that have the symbol of interest. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269459619-982-3-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index 45d1466..ce9b1ef 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -591,6 +591,8 @@ static const char * const annotate_usage[] = { static const struct option options[] = { OPT_STRING('i', "input", &input_name, "file", "input file name"), + OPT_STRING('d', "dsos", &symbol_conf.dso_list_str, "dso[,dso...]", + "only consider symbols in these dsos"), OPT_STRING('s', "symbol", &sym_hist_filter, "symbol", "symbol to annotate"), OPT_BOOLEAN('f', "force", &force, "don't complain, do it"), -- cgit v0.10.2 From 59fd53062f71011a68d03f4cd0ba93d822ac3249 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 24 Mar 2010 16:40:17 -0300 Subject: perf tools: Introduce struct map_symbol MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit That will be in both struct hist_entry and struct callchain_list, so that the TUI can store a pointer to the pair (map, symbol) in the trees where hist_entries and callchain_lists are present, to allow precise annotation instead of looking for the first symbol with the selected name. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269459619-982-4-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index ce9b1ef..887e8e0 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -69,13 +69,13 @@ static int sym__alloc_hist(struct symbol *self) static int annotate__hist_hit(struct hist_entry *he, u64 ip) { unsigned int sym_size, offset; - struct symbol *sym = he->sym; + struct symbol *sym = he->ms.sym; struct sym_priv *priv; struct sym_hist *h; he->count++; - if (!sym || !he->map) + if (!sym || !he->ms.map) return 0; priv = symbol__priv(sym); @@ -85,7 +85,7 @@ static int annotate__hist_hit(struct hist_entry *he, u64 ip) sym_size = sym->end - sym->start; offset = ip - sym->start; - pr_debug3("%s: ip=%#Lx\n", __func__, he->map->unmap_ip(he->map, ip)); + pr_debug3("%s: ip=%#Lx\n", __func__, he->ms.map->unmap_ip(he->ms.map, ip)); if (offset >= sym_size) return 0; @@ -94,8 +94,8 @@ static int annotate__hist_hit(struct hist_entry *he, u64 ip) h->sum++; h->ip[offset]++; - pr_debug3("%#Lx %s: count++ [ip: %#Lx, %#Lx] => %Ld\n", he->sym->start, - he->sym->name, ip, ip - he->sym->start, h->ip[offset]); + pr_debug3("%#Lx %s: count++ [ip: %#Lx, %#Lx] => %Ld\n", he->ms.sym->start, + he->ms.sym->name, ip, ip - he->ms.sym->start, h->ip[offset]); return 0; } @@ -187,7 +187,7 @@ static struct objdump_line *objdump__get_next_ip_line(struct list_head *head, static int parse_line(FILE *file, struct hist_entry *he, struct list_head *head) { - struct symbol *sym = he->sym; + struct symbol *sym = he->ms.sym; struct objdump_line *objdump_line; char *line = NULL, *tmp, *tmp2; size_t line_len; @@ -226,7 +226,7 @@ static int parse_line(FILE *file, struct hist_entry *he, } if (line_ip != -1) { - u64 start = map__rip_2objdump(he->map, sym->start); + u64 start = map__rip_2objdump(he->ms.map, sym->start); offset = line_ip - start; } @@ -244,7 +244,7 @@ static int objdump_line__print(struct objdump_line *self, struct list_head *head, struct hist_entry *he, u64 len) { - struct symbol *sym = he->sym; + struct symbol *sym = he->ms.sym; static const char *prev_line; static const char *prev_color; @@ -327,7 +327,7 @@ static void insert_source_line(struct sym_ext *sym_ext) static void free_source_line(struct hist_entry *he, int len) { - struct sym_priv *priv = symbol__priv(he->sym); + struct sym_priv *priv = symbol__priv(he->ms.sym); struct sym_ext *sym_ext = priv->ext; int i; @@ -346,7 +346,7 @@ static void free_source_line(struct hist_entry *he, int len) static void get_source_line(struct hist_entry *he, int len, const char *filename) { - struct symbol *sym = he->sym; + struct symbol *sym = he->ms.sym; u64 start; int i; char cmd[PATH_MAX * 2]; @@ -361,7 +361,7 @@ get_source_line(struct hist_entry *he, int len, const char *filename) if (!priv->ext) return; - start = he->map->unmap_ip(he->map, sym->start); + start = he->ms.map->unmap_ip(he->ms.map, sym->start); for (i = 0; i < len; i++) { char *path = NULL; @@ -425,7 +425,7 @@ static void print_summary(const char *filename) static void hist_entry__print_hits(struct hist_entry *self) { - struct symbol *sym = self->sym; + struct symbol *sym = self->ms.sym; struct sym_priv *priv = symbol__priv(sym); struct sym_hist *h = priv->hist; u64 len = sym->end - sym->start, offset; @@ -439,9 +439,9 @@ static void hist_entry__print_hits(struct hist_entry *self) static void annotate_sym(struct hist_entry *he) { - struct map *map = he->map; + struct map *map = he->ms.map; struct dso *dso = map->dso; - struct symbol *sym = he->sym; + struct symbol *sym = he->ms.sym; const char *filename = dso->long_name, *d_filename; u64 len; char command[PATH_MAX*2]; @@ -526,17 +526,17 @@ static void perf_session__find_annotations(struct perf_session *self) struct hist_entry *he = rb_entry(nd, struct hist_entry, rb_node); struct sym_priv *priv; - if (he->sym == NULL) + if (he->ms.sym == NULL) continue; - priv = symbol__priv(he->sym); + priv = symbol__priv(he->ms.sym); if (priv->hist == NULL) continue; annotate_sym(he); /* * Since we have a hist_entry per IP for the same symbol, free - * he->sym->hist to signal we already processed this symbol. + * he->ms.sym->hist to signal we already processed this symbol. */ free(priv->hist); priv->hist = NULL; diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index 5843a9c..4eefb52 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -22,8 +22,10 @@ struct hist_entry *__perf_session__add_hist_entry(struct rb_root *hists, struct hist_entry *he; struct hist_entry entry = { .thread = al->thread, - .map = al->map, - .sym = al->sym, + .ms = { + .map = al->map, + .sym = al->sym, + }, .ip = al->addr, .level = al->level, .count = count, @@ -654,7 +656,7 @@ print_entries: if (symbol_conf.use_callchain) ret += hist_entry__fprintf_callchain(h, fp, session_total); - if (h->map == NULL && verbose > 1) { + if (h->ms.map == NULL && verbose > 1) { __map_groups__fprintf_maps(&h->thread->mg, MAP__FUNCTION, fp); fprintf(fp, "%.10s end\n", graph_dotted_line); diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index a3465a0..25cd2e1 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -287,9 +287,9 @@ static size_t hist_entry__append_browser(struct hist_entry *self, indexes[0] = NEWT_ARG_APPEND; indexes[1] = NEWT_ARG_LAST; - newt_checkbox_tree__add(tree, s, self->sym, indexes); + newt_checkbox_tree__add(tree, s, self->ms.sym, indexes); } else - newtListboxAppendEntry(tree, s, self->sym); + newtListboxAppendEntry(tree, s, self->ms.sym); return strlen(s); } diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c index cb0f327..9b80c13 100644 --- a/tools/perf/util/sort.c +++ b/tools/perf/util/sort.c @@ -131,8 +131,8 @@ sort__comm_print(FILE *fp, struct hist_entry *self, unsigned int width) int64_t sort__dso_cmp(struct hist_entry *left, struct hist_entry *right) { - struct dso *dso_l = left->map ? left->map->dso : NULL; - struct dso *dso_r = right->map ? right->map->dso : NULL; + struct dso *dso_l = left->ms.map ? left->ms.map->dso : NULL; + struct dso *dso_r = right->ms.map ? right->ms.map->dso : NULL; const char *dso_name_l, *dso_name_r; if (!dso_l || !dso_r) @@ -152,9 +152,9 @@ sort__dso_cmp(struct hist_entry *left, struct hist_entry *right) size_t sort__dso_print(FILE *fp, struct hist_entry *self, unsigned int width) { - if (self->map && self->map->dso) { - const char *dso_name = !verbose ? self->map->dso->short_name : - self->map->dso->long_name; + if (self->ms.map && self->ms.map->dso) { + const char *dso_name = !verbose ? self->ms.map->dso->short_name : + self->ms.map->dso->long_name; return repsep_fprintf(fp, "%-*s", width, dso_name); } @@ -168,11 +168,11 @@ sort__sym_cmp(struct hist_entry *left, struct hist_entry *right) { u64 ip_l, ip_r; - if (left->sym == right->sym) + if (left->ms.sym == right->ms.sym) return 0; - ip_l = left->sym ? left->sym->start : left->ip; - ip_r = right->sym ? right->sym->start : right->ip; + ip_l = left->ms.sym ? left->ms.sym->start : left->ip; + ip_r = right->ms.sym ? right->ms.sym->start : right->ip; return (int64_t)(ip_r - ip_l); } @@ -184,13 +184,13 @@ sort__sym_print(FILE *fp, struct hist_entry *self, unsigned int width __used) size_t ret = 0; if (verbose) { - char o = self->map ? dso__symtab_origin(self->map->dso) : '!'; + char o = self->ms.map ? dso__symtab_origin(self->ms.map->dso) : '!'; ret += repsep_fprintf(fp, "%#018llx %c ", (u64)self->ip, o); } ret += repsep_fprintf(fp, "[%c] ", self->level); - if (self->sym) - ret += repsep_fprintf(fp, "%s", self->sym->name); + if (self->ms.sym) + ret += repsep_fprintf(fp, "%s", self->ms.sym->name); else ret += repsep_fprintf(fp, "%#016llx", (u64)self->ip); diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h index 753f9ea..5985686 100644 --- a/tools/perf/util/sort.h +++ b/tools/perf/util/sort.h @@ -45,8 +45,7 @@ struct hist_entry { struct rb_node rb_node; u64 count; struct thread *thread; - struct map *map; - struct symbol *sym; + struct map_symbol ms; u64 ip; char level; struct symbol *parent; diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h index 0da2455..a4a894b 100644 --- a/tools/perf/util/symbol.h +++ b/tools/perf/util/symbol.h @@ -88,6 +88,11 @@ struct ref_reloc_sym { u64 unrelocated_addr; }; +struct map_symbol { + struct map *map; + struct symbol *sym; +}; + struct addr_location { struct thread *thread; struct map *map; -- cgit v0.10.2 From b3c9ac0846c654dea4df095999ee202e8b4cb253 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 24 Mar 2010 16:40:18 -0300 Subject: perf callchains: Store the map together with the symbol MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We need this to know where a symbol in a callchain came from, for various reasons, among them precise annotation from a TUI/GUI tool. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269459619-982-5-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index d609afb..6ab1698 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -81,7 +81,8 @@ static int perf_session__add_hist_entry(struct perf_session *self, struct addr_location *al, struct sample_data *data) { - struct symbol **syms = NULL, *parent = NULL; + struct map_symbol *syms = NULL; + struct symbol *parent = NULL; bool hit; int err; struct hist_entry *he; diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c index 883844e..db628af 100644 --- a/tools/perf/util/callchain.c +++ b/tools/perf/util/callchain.c @@ -185,8 +185,8 @@ create_child(struct callchain_node *parent, bool inherit_children) struct resolved_ip { - u64 ip; - struct symbol *sym; + u64 ip; + struct map_symbol ms; }; struct resolved_chain { @@ -212,7 +212,7 @@ fill_node(struct callchain_node *node, struct resolved_chain *chain, int start) return; } call->ip = chain->ips[i].ip; - call->sym = chain->ips[i].sym; + call->ms = chain->ips[i].ms; list_add_tail(&call->list, &node->val); } node->val_nr = chain->nr - start; @@ -318,10 +318,10 @@ __append_chain(struct callchain_node *root, struct resolved_chain *chain, if (i == chain->nr) break; - sym = chain->ips[i].sym; + sym = chain->ips[i].ms.sym; - if (cnode->sym && sym) { - if (cnode->sym->start != sym->start) + if (cnode->ms.sym && sym) { + if (cnode->ms.sym->start != sym->start) break; } else if (cnode->ip != chain->ips[i].ip) break; @@ -353,9 +353,8 @@ __append_chain(struct callchain_node *root, struct resolved_chain *chain, return 0; } -static void -filter_context(struct ip_callchain *old, struct resolved_chain *new, - struct symbol **syms) +static void filter_context(struct ip_callchain *old, struct resolved_chain *new, + struct map_symbol *syms) { int i, j = 0; @@ -364,7 +363,7 @@ filter_context(struct ip_callchain *old, struct resolved_chain *new, continue; new->ips[j].ip = old->ips[i]; - new->ips[j].sym = syms[i]; + new->ips[j].ms = syms[i]; j++; } @@ -373,7 +372,7 @@ filter_context(struct ip_callchain *old, struct resolved_chain *new, int append_chain(struct callchain_node *root, struct ip_callchain *chain, - struct symbol **syms) + struct map_symbol *syms) { struct resolved_chain *filtered; diff --git a/tools/perf/util/callchain.h b/tools/perf/util/callchain.h index bbd76da..8a7e8bb 100644 --- a/tools/perf/util/callchain.h +++ b/tools/perf/util/callchain.h @@ -39,7 +39,7 @@ struct callchain_param { struct callchain_list { u64 ip; - struct symbol *sym; + struct map_symbol ms; struct list_head list; }; @@ -57,5 +57,5 @@ static inline u64 cumul_hits(struct callchain_node *node) int register_callchain_param(struct callchain_param *param); int append_chain(struct callchain_node *root, struct ip_callchain *chain, - struct symbol **syms); + struct map_symbol *syms); #endif /* __PERF_CALLCHAIN_H */ diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index 4eefb52..09e09e7 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -260,8 +260,8 @@ static size_t ipchain__fprintf_graph(FILE *fp, struct callchain_list *chain, } else ret += fprintf(fp, "%s", " "); } - if (chain->sym) - ret += fprintf(fp, "%s\n", chain->sym->name); + if (chain->ms.sym) + ret += fprintf(fp, "%s\n", chain->ms.sym->name); else ret += fprintf(fp, "%p\n", (void *)(long)chain->ip); @@ -280,7 +280,7 @@ static void init_rem_hits(void) } strcpy(rem_sq_bracket->name, "[...]"); - rem_hits.sym = rem_sq_bracket; + rem_hits.ms.sym = rem_sq_bracket; } static size_t __callchain__fprintf_graph(FILE *fp, struct callchain_node *self, @@ -382,8 +382,8 @@ static size_t callchain__fprintf_graph(FILE *fp, struct callchain_node *self, } else ret += callchain__fprintf_left_margin(fp, left_margin); - if (chain->sym) - ret += fprintf(fp, " %s\n", chain->sym->name); + if (chain->ms.sym) + ret += fprintf(fp, " %s\n", chain->ms.sym->name); else ret += fprintf(fp, " %p\n", (void *)(long)chain->ip); } @@ -408,8 +408,8 @@ static size_t callchain__fprintf_flat(FILE *fp, struct callchain_node *self, list_for_each_entry(chain, &self->val, list) { if (chain->ip >= PERF_CONTEXT_MAX) continue; - if (chain->sym) - ret += fprintf(fp, " %s\n", chain->sym->name); + if (chain->ms.sym) + ret += fprintf(fp, " %s\n", chain->ms.sym->name); else ret += fprintf(fp, " %p\n", (void *)(long)chain->ip); diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index 25cd2e1..f6953ca 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -104,8 +104,8 @@ static void newt_checkbox_tree__add(newtComponent tree, const char *str, static char *callchain_list__sym_name(struct callchain_list *self, char *bf, size_t bfsize) { - if (self->sym) - return self->sym->name; + if (self->ms.sym) + return self->ms.sym->name; snprintf(bf, bfsize, "%#Lx", self->ip); return bf; @@ -157,7 +157,7 @@ static void __callchain__append_graph_browser(struct callchain_node *self, indexes[depth + 2] = NEWT_ARG_LAST; ++chain_idx; } - newt_checkbox_tree__add(tree, str, chain->sym, indexes); + newt_checkbox_tree__add(tree, str, chain->ms.sym, indexes); free(alloc_str); ++printed; } @@ -193,7 +193,7 @@ static void callchain__append_graph_browser(struct callchain_node *self, continue; str = callchain_list__sym_name(chain, ipstr, sizeof(ipstr)); - newt_checkbox_tree__add(tree, str, chain->sym, indexes); + newt_checkbox_tree__add(tree, str, chain->ms.sym, indexes); } indexes[1] = parent_idx; diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index eed1cb8..2cef373 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -117,13 +117,13 @@ static bool symbol__match_parent_regex(struct symbol *sym) return 0; } -struct symbol **perf_session__resolve_callchain(struct perf_session *self, - struct thread *thread, - struct ip_callchain *chain, - struct symbol **parent) +struct map_symbol *perf_session__resolve_callchain(struct perf_session *self, + struct thread *thread, + struct ip_callchain *chain, + struct symbol **parent) { u8 cpumode = PERF_RECORD_MISC_USER; - struct symbol **syms = NULL; + struct map_symbol *syms = NULL; unsigned int i; if (symbol_conf.use_callchain) { @@ -160,7 +160,8 @@ struct symbol **perf_session__resolve_callchain(struct perf_session *self, *parent = al.sym; if (!symbol_conf.use_callchain) break; - syms[i] = al.sym; + syms[i].map = al.map; + syms[i].sym = al.sym; } } diff --git a/tools/perf/util/session.h b/tools/perf/util/session.h index 34d7339..631f815 100644 --- a/tools/perf/util/session.h +++ b/tools/perf/util/session.h @@ -57,10 +57,10 @@ int __perf_session__process_events(struct perf_session *self, int perf_session__process_events(struct perf_session *self, struct perf_event_ops *event_ops); -struct symbol **perf_session__resolve_callchain(struct perf_session *self, - struct thread *thread, - struct ip_callchain *chain, - struct symbol **parent); +struct map_symbol *perf_session__resolve_callchain(struct perf_session *self, + struct thread *thread, + struct ip_callchain *chain, + struct symbol **parent); bool perf_session__has_traces(struct perf_session *self, const char *msg); -- cgit v0.10.2 From d5679ae4d2d4369477dc3b60027cca222aef33e9 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 24 Mar 2010 16:40:19 -0300 Subject: perf report: Pass the DSO to 'perf annotate' MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit So that we ensure that the symbol asked for annotation really is in the DSO we are interested in. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269459619-982-6-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index f6953ca..e99bcc8 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -157,7 +157,7 @@ static void __callchain__append_graph_browser(struct callchain_node *self, indexes[depth + 2] = NEWT_ARG_LAST; ++chain_idx; } - newt_checkbox_tree__add(tree, str, chain->ms.sym, indexes); + newt_checkbox_tree__add(tree, str, &chain->ms, indexes); free(alloc_str); ++printed; } @@ -193,7 +193,7 @@ static void callchain__append_graph_browser(struct callchain_node *self, continue; str = callchain_list__sym_name(chain, ipstr, sizeof(ipstr)); - newt_checkbox_tree__add(tree, str, chain->ms.sym, indexes); + newt_checkbox_tree__add(tree, str, &chain->ms, indexes); } indexes[1] = parent_idx; @@ -287,14 +287,14 @@ static size_t hist_entry__append_browser(struct hist_entry *self, indexes[0] = NEWT_ARG_APPEND; indexes[1] = NEWT_ARG_LAST; - newt_checkbox_tree__add(tree, s, self->ms.sym, indexes); + newt_checkbox_tree__add(tree, s, &self->ms, indexes); } else - newtListboxAppendEntry(tree, s, self->ms.sym); + newtListboxAppendEntry(tree, s, &self->ms); return strlen(s); } -static void symbol__annotate_browser(const struct symbol *self) +static void map_symbol__annotate_browser(const struct map_symbol *self) { FILE *fp; int cols, rows; @@ -305,10 +305,11 @@ static void symbol__annotate_browser(const struct symbol *self) size_t max_usable_width; char *line = NULL; - if (self == NULL) + if (self->sym == NULL) return; - if (asprintf(&str, "perf annotate %s 2>&1 | expand", self->name) < 0) + if (asprintf(&str, "perf annotate -d \"%s\" %s 2>&1 | expand", + self->map->dso->name, self->sym->name) < 0) return; fp = popen(str, "r"); @@ -338,7 +339,7 @@ static void symbol__annotate_browser(const struct symbol *self) newtListboxSetWidth(tree, max_line_len); - newtCenteredWindow(max_line_len + 2, rows - 5, self->name); + newtCenteredWindow(max_line_len + 2, rows - 5, self->sym->name); form = newt_form__new(); newtFormAddComponents(form, tree, NULL); @@ -359,7 +360,7 @@ static const void *newt__symbol_tree_get_current(newtComponent self) static void perf_session__selection(newtComponent self, void *data) { - const struct symbol **symbol_ptr = data; + const struct map_symbol **symbol_ptr = data; *symbol_ptr = newt__symbol_tree_get_current(self); } @@ -376,7 +377,7 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, char str[1024]; newtComponent form, tree; struct newtExitStruct es; - const struct symbol *selection; + const struct map_symbol *selection; snprintf(str, sizeof(str), "Samples: %Ld", session_total); newtDrawRootText(0, 0, str); @@ -416,11 +417,8 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, int len = hist_entry__append_browser(h, tree, session_total); if (len > max_len) max_len = len; - if (symbol_conf.use_callchain) { + if (symbol_conf.use_callchain) hist_entry__append_callchain_browser(h, tree, session_total, idx++); - if (idx > 3300) - break; - } } if (max_len > cols) @@ -441,14 +439,12 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, while (1) { char annotate[512]; const char *options[2]; - int nr_options = 0, choice; + int nr_options = 0, choice = 0; newtFormRun(form, &es); if (es.reason == NEWT_EXIT_HOTKEY) { - if (toupper(es.u.key) == 'A') { - symbol__annotate_browser(selection); - continue; - } + if (toupper(es.u.key) == 'A') + goto do_annotate; if (es.u.key == NEWT_KEY_ESCAPE || toupper(es.u.key) == 'Q' || es.u.key == CTRL('c')) { @@ -459,9 +455,9 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, } } - if (selection != NULL) { + if (selection->sym != NULL) { snprintf(annotate, sizeof(annotate), - "Annotate %s", selection->name); + "Annotate %s", selection->sym->name); options[nr_options++] = annotate; } @@ -469,8 +465,17 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, choice = popup_menu(nr_options, options); if (choice == nr_options - 1) break; - else if (selection != NULL && choice >= 0) - symbol__annotate_browser(selection); +do_annotate: + if (selection->sym != NULL && choice >= 0) { + if (selection->map->dso->origin == DSO__ORIG_KERNEL) { + newtPopHelpLine(); + newtPushHelpLine("No vmlinux file found, can't " + "annotate with just a " + "kallsyms file"); + continue; + } + map_symbol__annotate_browser(selection); + } } newtFormDestroy(form); -- cgit v0.10.2 From b177f63f5226e75280855bbcd106e677250778bd Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 25 Mar 2010 19:58:57 -0300 Subject: perf symbols: Pass the mmap parameters instead of using mmap_event MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit To reduce the coupling of the symbol system with the rest of perf. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269557941-15617-2-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c index c2808ad..052eaecc 100644 --- a/tools/perf/util/event.c +++ b/tools/perf/util/event.c @@ -393,7 +393,8 @@ int event__process_mmap(event_t *self, struct perf_session *session) } thread = perf_session__findnew(session, self->mmap.pid); - map = map__new(&self->mmap, MAP__FUNCTION, + map = map__new(self->mmap.start, self->mmap.len, self->mmap.pgoff, + self->mmap.pid, self->mmap.filename, MAP__FUNCTION, session->cwd, session->cwdlen); if (thread == NULL || map == NULL) diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index e509cd5..a9b4227 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -1,4 +1,3 @@ -#include "event.h" #include "symbol.h" #include #include @@ -38,13 +37,12 @@ void map__init(struct map *self, enum map_type type, RB_CLEAR_NODE(&self->rb_node); } -struct map *map__new(struct mmap_event *event, enum map_type type, - char *cwd, int cwdlen) +struct map *map__new(u64 start, u64 len, u64 pgoff, u32 pid, char *filename, + enum map_type type, char *cwd, int cwdlen) { struct map *self = malloc(sizeof(*self)); if (self != NULL) { - const char *filename = event->filename; char newfilename[PATH_MAX]; struct dso *dso; int anon; @@ -62,7 +60,7 @@ struct map *map__new(struct mmap_event *event, enum map_type type, anon = is_anon_memory(filename); if (anon) { - snprintf(newfilename, sizeof(newfilename), "/tmp/perf-%d.map", event->pid); + snprintf(newfilename, sizeof(newfilename), "/tmp/perf-%d.map", pid); filename = newfilename; } @@ -70,8 +68,7 @@ struct map *map__new(struct mmap_event *event, enum map_type type, if (dso == NULL) goto out_delete; - map__init(self, type, event->start, event->start + event->len, - event->pgoff, dso); + map__init(self, type, start, start + len, pgoff, dso); if (anon) { set_identity: diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h index b756368..a4a5bc4 100644 --- a/tools/perf/util/map.h +++ b/tools/perf/util/map.h @@ -68,14 +68,13 @@ u64 map__rip_2objdump(struct map *map, u64 rip); u64 map__objdump_2ip(struct map *map, u64 addr); struct symbol; -struct mmap_event; typedef int (*symbol_filter_t)(struct map *map, struct symbol *sym); void map__init(struct map *self, enum map_type type, u64 start, u64 end, u64 pgoff, struct dso *dso); -struct map *map__new(struct mmap_event *event, enum map_type, - char *cwd, int cwdlen); +struct map *map__new(u64 start, u64 len, u64 pgoff, u32 pid, char *filename, + enum map_type type, char *cwd, int cwdlen); void map__delete(struct map *self); struct map *map__clone(struct map *self); int map__overlap(struct map *l, struct map *r); -- cgit v0.10.2 From 4b8cf84624e9a58a21aaac3d064222092ae234e0 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 25 Mar 2010 19:58:58 -0300 Subject: perf symbols: Move map related routines to map.c MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Thru series of refactorings functions were being renamed but not moved to map.c to reduce patch noise, now lets have them in the same place so that use of the symbol system by tools can be constrained to building and linking fewer source files: symbol.c, map.c and rbtree.c. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269557941-15617-3-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index a9b4227..9f2963f 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -1,8 +1,9 @@ #include "symbol.h" +#include #include #include #include -#include "debug.h" +#include "map.h" const char *map_type__name[MAP__NR_TYPES] = { [MAP__FUNCTION] = "Functions", @@ -232,3 +233,84 @@ u64 map__objdump_2ip(struct map *map, u64 addr) map->unmap_ip(map, addr); /* RIP -> IP */ return ip; } + +struct symbol *map_groups__find_symbol(struct map_groups *self, + enum map_type type, u64 addr, + symbol_filter_t filter) +{ + struct map *map = map_groups__find(self, type, addr); + + if (map != NULL) + return map__find_symbol(map, map->map_ip(map, addr), filter); + + return NULL; +} + +static u64 map__reloc_map_ip(struct map *map, u64 ip) +{ + return ip + (s64)map->pgoff; +} + +static u64 map__reloc_unmap_ip(struct map *map, u64 ip) +{ + return ip - (s64)map->pgoff; +} + +void map__reloc_vmlinux(struct map *self) +{ + struct kmap *kmap = map__kmap(self); + s64 reloc; + + if (!kmap->ref_reloc_sym || !kmap->ref_reloc_sym->unrelocated_addr) + return; + + reloc = (kmap->ref_reloc_sym->unrelocated_addr - + kmap->ref_reloc_sym->addr); + + if (!reloc) + return; + + self->map_ip = map__reloc_map_ip; + self->unmap_ip = map__reloc_unmap_ip; + self->pgoff = reloc; +} + +void maps__insert(struct rb_root *maps, struct map *map) +{ + struct rb_node **p = &maps->rb_node; + struct rb_node *parent = NULL; + const u64 ip = map->start; + struct map *m; + + while (*p != NULL) { + parent = *p; + m = rb_entry(parent, struct map, rb_node); + if (ip < m->start) + p = &(*p)->rb_left; + else + p = &(*p)->rb_right; + } + + rb_link_node(&map->rb_node, parent, p); + rb_insert_color(&map->rb_node, maps); +} + +struct map *maps__find(struct rb_root *maps, u64 ip) +{ + struct rb_node **p = &maps->rb_node; + struct rb_node *parent = NULL; + struct map *m; + + while (*p != NULL) { + parent = *p; + m = rb_entry(parent, struct map, rb_node); + if (ip < m->start) + p = &(*p)->rb_left; + else if (ip > m->end) + p = &(*p)->rb_right; + else + return m; + } + + return NULL; +} diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h index a4a5bc4..6a703fa 100644 --- a/tools/perf/util/map.h +++ b/tools/perf/util/map.h @@ -4,7 +4,8 @@ #include #include #include -#include +#include +#include "types.h" enum map_type { MAP__FUNCTION = 0, @@ -90,4 +91,48 @@ void map__fixup_end(struct map *self); void map__reloc_vmlinux(struct map *self); +struct map_groups { + struct rb_root maps[MAP__NR_TYPES]; + struct list_head removed_maps[MAP__NR_TYPES]; +}; + +size_t __map_groups__fprintf_maps(struct map_groups *self, + enum map_type type, FILE *fp); +void maps__insert(struct rb_root *maps, struct map *map); +struct map *maps__find(struct rb_root *maps, u64 addr); +void map_groups__init(struct map_groups *self); +size_t map_groups__fprintf_maps(struct map_groups *self, FILE *fp); + +static inline void map_groups__insert(struct map_groups *self, struct map *map) +{ + maps__insert(&self->maps[map->type], map); +} + +static inline struct map *map_groups__find(struct map_groups *self, + enum map_type type, u64 addr) +{ + return maps__find(&self->maps[type], addr); +} + +struct symbol *map_groups__find_symbol(struct map_groups *self, + enum map_type type, u64 addr, + symbol_filter_t filter); + +static inline struct symbol *map_groups__find_function(struct map_groups *self, + u64 addr, + symbol_filter_t filter) +{ + return map_groups__find_symbol(self, MAP__FUNCTION, addr, filter); +} + +struct map *map_groups__find_by_name(struct map_groups *self, + enum map_type type, const char *name); +int __map_groups__create_kernel_maps(struct map_groups *self, + struct map *vmlinux_maps[MAP__NR_TYPES], + struct dso *kernel); +int map_groups__create_kernel_maps(struct map_groups *self, + struct map *vmlinux_maps[MAP__NR_TYPES]); +struct map *map_groups__new_module(struct map_groups *self, u64 start, + const char *filename); + #endif /* __PERF_MAP_H */ diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index 2cef373..76b4ac6 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -544,32 +544,3 @@ int perf_session__set_kallsyms_ref_reloc_sym(struct perf_session *self, return 0; } - -static u64 map__reloc_map_ip(struct map *map, u64 ip) -{ - return ip + (s64)map->pgoff; -} - -static u64 map__reloc_unmap_ip(struct map *map, u64 ip) -{ - return ip - (s64)map->pgoff; -} - -void map__reloc_vmlinux(struct map *self) -{ - struct kmap *kmap = map__kmap(self); - s64 reloc; - - if (!kmap->ref_reloc_sym || !kmap->ref_reloc_sym->unrelocated_addr) - return; - - reloc = (kmap->ref_reloc_sym->unrelocated_addr - - kmap->ref_reloc_sym->addr); - - if (!reloc) - return; - - self->map_ip = map__reloc_map_ip; - self->unmap_ip = map__reloc_unmap_ip; - self->pgoff = reloc; -} diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c index ea65062..9bbe27d 100644 --- a/tools/perf/util/thread.c +++ b/tools/perf/util/thread.c @@ -272,46 +272,6 @@ static int map_groups__fixup_overlappings(struct map_groups *self, return 0; } -void maps__insert(struct rb_root *maps, struct map *map) -{ - struct rb_node **p = &maps->rb_node; - struct rb_node *parent = NULL; - const u64 ip = map->start; - struct map *m; - - while (*p != NULL) { - parent = *p; - m = rb_entry(parent, struct map, rb_node); - if (ip < m->start) - p = &(*p)->rb_left; - else - p = &(*p)->rb_right; - } - - rb_link_node(&map->rb_node, parent, p); - rb_insert_color(&map->rb_node, maps); -} - -struct map *maps__find(struct rb_root *maps, u64 ip) -{ - struct rb_node **p = &maps->rb_node; - struct rb_node *parent = NULL; - struct map *m; - - while (*p != NULL) { - parent = *p; - m = rb_entry(parent, struct map, rb_node); - if (ip < m->start) - p = &(*p)->rb_left; - else if (ip > m->end) - p = &(*p)->rb_right; - else - return m; - } - - return NULL; -} - void thread__insert_map(struct thread *self, struct map *map) { map_groups__fixup_overlappings(&self->mg, map); @@ -367,16 +327,3 @@ size_t perf_session__fprintf(struct perf_session *self, FILE *fp) return ret; } - -struct symbol *map_groups__find_symbol(struct map_groups *self, - enum map_type type, u64 addr, - symbol_filter_t filter) -{ - struct map *map = map_groups__find(self, type, addr); - - if (map != NULL) - return map__find_symbol(map, map->map_ip(map, addr), filter); - - return NULL; -} - diff --git a/tools/perf/util/thread.h b/tools/perf/util/thread.h index a81426a..9c488fc 100644 --- a/tools/perf/util/thread.h +++ b/tools/perf/util/thread.h @@ -5,14 +5,6 @@ #include #include "symbol.h" -struct map_groups { - struct rb_root maps[MAP__NR_TYPES]; - struct list_head removed_maps[MAP__NR_TYPES]; -}; - -size_t __map_groups__fprintf_maps(struct map_groups *self, - enum map_type type, FILE *fp); - struct thread { struct rb_node rb_node; struct map_groups mg; @@ -23,30 +15,16 @@ struct thread { int comm_len; }; +struct perf_session; + int find_all_tid(int pid, pid_t ** all_tid); -void map_groups__init(struct map_groups *self); int thread__set_comm(struct thread *self, const char *comm); int thread__comm_len(struct thread *self); struct thread *perf_session__findnew(struct perf_session *self, pid_t pid); void thread__insert_map(struct thread *self, struct map *map); int thread__fork(struct thread *self, struct thread *parent); -size_t map_groups__fprintf_maps(struct map_groups *self, FILE *fp); size_t perf_session__fprintf(struct perf_session *self, FILE *fp); -void maps__insert(struct rb_root *maps, struct map *map); -struct map *maps__find(struct rb_root *maps, u64 addr); - -static inline void map_groups__insert(struct map_groups *self, struct map *map) -{ - maps__insert(&self->maps[map->type], map); -} - -static inline struct map *map_groups__find(struct map_groups *self, - enum map_type type, u64 addr) -{ - return maps__find(&self->maps[type], addr); -} - static inline struct map *thread__find_map(struct thread *self, enum map_type type, u64 addr) { @@ -63,26 +41,4 @@ void thread__find_addr_location(struct thread *self, enum map_type type, u64 addr, struct addr_location *al, symbol_filter_t filter); -struct symbol *map_groups__find_symbol(struct map_groups *self, - enum map_type type, u64 addr, - symbol_filter_t filter); - -static inline struct symbol *map_groups__find_function(struct map_groups *self, - u64 addr, - symbol_filter_t filter) -{ - return map_groups__find_symbol(self, MAP__FUNCTION, addr, filter); -} - -struct map *map_groups__find_by_name(struct map_groups *self, - enum map_type type, const char *name); - -int __map_groups__create_kernel_maps(struct map_groups *self, - struct map *vmlinux_maps[MAP__NR_TYPES], - struct dso *kernel); -int map_groups__create_kernel_maps(struct map_groups *self, - struct map *vmlinux_maps[MAP__NR_TYPES]); - -struct map *map_groups__new_module(struct map_groups *self, u64 start, - const char *filename); #endif /* __PERF_THREAD_H */ -- cgit v0.10.2 From 618038df3588fdfcaccfd40057f36ce792bee252 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 25 Mar 2010 19:58:59 -0300 Subject: perf tools: Move __used from perf.h to linux/compiler.h MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Just like in the kernel and also to remove the need to include perf.h in the symbol subsystem. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269557941-15617-4-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/perf.h b/tools/perf/perf.h index aa78615..ec21274 100644 --- a/tools/perf/perf.h +++ b/tools/perf/perf.h @@ -106,8 +106,6 @@ static inline unsigned long long rdclock(void) #define __user #define asmlinkage -#define __used __attribute__((__unused__)) - #define unlikely(x) __builtin_expect(!!(x), 0) #define min(x, y) ({ \ typeof(x) _min1 = (x); \ diff --git a/tools/perf/util/include/linux/compiler.h b/tools/perf/util/include/linux/compiler.h index dfb0713..791f9dd 100644 --- a/tools/perf/util/include/linux/compiler.h +++ b/tools/perf/util/include/linux/compiler.h @@ -7,4 +7,6 @@ #define __user #define __attribute_const__ +#define __used __attribute__((__unused__)) + #endif -- cgit v0.10.2 From 5aab621b7bf024608f0c089e21656e7fe875a150 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 25 Mar 2010 19:59:00 -0300 Subject: perf symbols: Move hex2u64 and strxfrchar to symbol.c MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Mostly used in symbol.c so move them there to reduce the number of files needed to use the symbol system. Also do some header adjustments with the same intent. Signed-off-by: Arnaldo Carvalho de Melo Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <1269557941-15617-5-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c index a201445..435781e 100644 --- a/tools/perf/util/parse-events.c +++ b/tools/perf/util/parse-events.c @@ -5,6 +5,7 @@ #include "parse-events.h" #include "exec_cmd.h" #include "string.h" +#include "symbol.h" #include "cache.h" #include "header.h" #include "debugfs.h" diff --git a/tools/perf/util/string.c b/tools/perf/util/string.c index a175949..d438924 100644 --- a/tools/perf/util/string.c +++ b/tools/perf/util/string.c @@ -1,49 +1,6 @@ #include "string.h" #include "util.h" -static int hex(char ch) -{ - if ((ch >= '0') && (ch <= '9')) - return ch - '0'; - if ((ch >= 'a') && (ch <= 'f')) - return ch - 'a' + 10; - if ((ch >= 'A') && (ch <= 'F')) - return ch - 'A' + 10; - return -1; -} - -/* - * While we find nice hex chars, build a long_val. - * Return number of chars processed. - */ -int hex2u64(const char *ptr, u64 *long_val) -{ - const char *p = ptr; - *long_val = 0; - - while (*p) { - const int hex_val = hex(*p); - - if (hex_val < 0) - break; - - *long_val = (*long_val << 4) | hex_val; - p++; - } - - return p - ptr; -} - -char *strxfrchar(char *s, char from, char to) -{ - char *p = s; - - while ((p = strchr(p, from)) != NULL) - *p++ = to; - - return s; -} - #define K 1024LL /* * perf_atoll() diff --git a/tools/perf/util/string.h b/tools/perf/util/string.h index 542e44d..7005824 100644 --- a/tools/perf/util/string.h +++ b/tools/perf/util/string.h @@ -4,8 +4,6 @@ #include #include "types.h" -int hex2u64(const char *ptr, u64 *val); -char *strxfrchar(char *s, char from, char to); s64 perf_atoll(const char *str); char **argv_split(const char *str, int *argcp); void argv_free(char **argv); diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 3eb9de4..f3d4151 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -1,13 +1,19 @@ -#include "util.h" -#include "../perf.h" -#include "sort.h" -#include "string.h" +#define _GNU_SOURCE +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include #include "symbol.h" -#include "thread.h" +#include "strlist.h" -#include "debug.h" - -#include #include #include #include @@ -114,8 +120,8 @@ static void map_groups__fixup_end(struct map_groups *self) static struct symbol *symbol__new(u64 start, u64 len, const char *name) { size_t namelen = strlen(name) + 1; - struct symbol *self = zalloc(symbol_conf.priv_size + - sizeof(*self) + namelen); + struct symbol *self = calloc(1, (symbol_conf.priv_size + + sizeof(*self) + namelen)); if (self == NULL) return NULL; @@ -166,7 +172,7 @@ static void dso__set_basename(struct dso *self) struct dso *dso__new(const char *name) { - struct dso *self = zalloc(sizeof(*self) + strlen(name) + 1); + struct dso *self = calloc(1, sizeof(*self) + strlen(name) + 1); if (self != NULL) { int i; @@ -1382,13 +1388,13 @@ static int dso__kernel_module_get_build_id(struct dso *self) return 0; } -static int map_groups__set_modules_path_dir(struct map_groups *self, char *dirname) +static int map_groups__set_modules_path_dir(struct map_groups *self, char *dir_name) { struct dirent *dent; - DIR *dir = opendir(dirname); + DIR *dir = opendir(dir_name); if (!dir) { - pr_debug("%s: cannot open %s dir\n", __func__, dirname); + pr_debug("%s: cannot open %s dir\n", __func__, dir_name); return -1; } @@ -1401,7 +1407,7 @@ static int map_groups__set_modules_path_dir(struct map_groups *self, char *dirna continue; snprintf(path, sizeof(path), "%s/%s", - dirname, dent->d_name); + dir_name, dent->d_name); if (map_groups__set_modules_path_dir(self, path) < 0) goto failure; } else { @@ -1421,7 +1427,7 @@ static int map_groups__set_modules_path_dir(struct map_groups *self, char *dirna continue; snprintf(path, sizeof(path), "%s/%s", - dirname, dent->d_name); + dir_name, dent->d_name); long_name = strdup(path); if (long_name == NULL) @@ -1458,8 +1464,8 @@ static int map_groups__set_modules_path(struct map_groups *self) */ static struct map *map__new2(u64 start, struct dso *dso, enum map_type type) { - struct map *self = zalloc(sizeof(*self) + - (dso->kernel ? sizeof(struct kmap) : 0)); + struct map *self = calloc(1, (sizeof(*self) + + (dso->kernel ? sizeof(struct kmap) : 0))); if (self != NULL) { /* * ->end will be filled after we load all the symbols @@ -1963,3 +1969,46 @@ int map_groups__create_kernel_maps(struct map_groups *self, map_groups__fixup_end(self); return 0; } + +static int hex(char ch) +{ + if ((ch >= '0') && (ch <= '9')) + return ch - '0'; + if ((ch >= 'a') && (ch <= 'f')) + return ch - 'a' + 10; + if ((ch >= 'A') && (ch <= 'F')) + return ch - 'A' + 10; + return -1; +} + +/* + * While we find nice hex chars, build a long_val. + * Return number of chars processed. + */ +int hex2u64(const char *ptr, u64 *long_val) +{ + const char *p = ptr; + *long_val = 0; + + while (*p) { + const int hex_val = hex(*p); + + if (hex_val < 0) + break; + + *long_val = (*long_val << 4) | hex_val; + p++; + } + + return p - ptr; +} + +char *strxfrchar(char *s, char from, char to) +{ + char *p = s; + + while ((p = strchr(p, from)) != NULL) + *p++ = to; + + return s; +} diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h index a4a894b..757fae3 100644 --- a/tools/perf/util/symbol.h +++ b/tools/perf/util/symbol.h @@ -3,10 +3,11 @@ #include #include -#include "types.h" +#include +#include "map.h" #include #include -#include "event.h" +#include #define DEBUG_CACHE_DIR ".debug" @@ -29,6 +30,9 @@ static inline char *bfd_demangle(void __used *v, const char __used *c, #endif #endif +int hex2u64(const char *ptr, u64 *val); +char *strxfrchar(char *s, char from, char to); + /* * libelf 0.8.x and earlier do not support ELF_C_READ_MMAP; * for newer versions we can use mmap to reduce memory usage: @@ -44,6 +48,8 @@ static inline char *bfd_demangle(void __used *v, const char __used *c, #define DMGL_ANSI (1 << 1) /* Include const, volatile, etc */ #endif +#define BUILD_ID_SIZE 20 + struct symbol { struct rb_node rb_node; u64 start; -- cgit v0.10.2 From 5a10317483f606106395814ee2fdaa2f1256a3b3 Mon Sep 17 00:00:00 2001 From: "Zhang, Yanmin" Date: Thu, 25 Mar 2010 19:59:01 -0300 Subject: perf record: Zero out mmap_array to fix segfault Reported-by: Li Zefan Tested-by: Li Zefan Signed-off-by: Zhang Yanmin Signed-off-by: Arnaldo Carvalho de Melo LKML-Reference: <1269557941-15617-6-git-send-email-acme@infradead.org> Cc: Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index bb5b23d..60ecdd3 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -751,7 +751,7 @@ int cmd_record(int argc, const char **argv, const char *prefix __used) for (i = 0; i < MAX_NR_CPUS; i++) { for (j = 0; j < MAX_COUNTERS; j++) { fd[i][j] = malloc(sizeof(int)*thread_num); - mmap_array[i][j] = malloc( + mmap_array[i][j] = zalloc( sizeof(struct mmap_data)*thread_num); if (!fd[i][j] || !mmap_array[i][j]) return -ENOMEM; diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index 5f3ac9f..4abdd9b 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -1371,7 +1371,7 @@ int cmd_top(int argc, const char **argv, const char *prefix __used) for (i = 0; i < MAX_NR_CPUS; i++) { for (j = 0; j < MAX_COUNTERS; j++) { fd[i][j] = malloc(sizeof(int)*thread_num); - mmap_array[i][j] = malloc( + mmap_array[i][j] = zalloc( sizeof(struct mmap_data)*thread_num); if (!fd[i][j] || !mmap_array[i][j]) return -ENOMEM; -- cgit v0.10.2 From 7c5ecaf7666617889f337296c610815b519abfa9 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 25 Mar 2010 14:51:49 +0100 Subject: perf, x86: Clean up debugctlmsr bit definitions Move all debugctlmsr thingies into msr-index.h Signed-off-by: Peter Zijlstra LKML-Reference: <20100325135413.861425293@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/msr-index.h b/arch/x86/include/asm/msr-index.h index aef562c..06e4cf0 100644 --- a/arch/x86/include/asm/msr-index.h +++ b/arch/x86/include/asm/msr-index.h @@ -71,11 +71,14 @@ #define MSR_IA32_LASTINTTOIP 0x000001de /* DEBUGCTLMSR bits (others vary by model): */ -#define _DEBUGCTLMSR_LBR 0 /* last branch recording */ -#define _DEBUGCTLMSR_BTF 1 /* single-step on branches */ - -#define DEBUGCTLMSR_LBR (1UL << _DEBUGCTLMSR_LBR) -#define DEBUGCTLMSR_BTF (1UL << _DEBUGCTLMSR_BTF) +#define DEBUGCTLMSR_LBR (1UL << 0) /* last branch recording */ +#define DEBUGCTLMSR_BTF (1UL << 1) /* single-step on branches */ +#define DEBUGCTLMSR_TR (1UL << 6) +#define DEBUGCTLMSR_BTS (1UL << 7) +#define DEBUGCTLMSR_BTINT (1UL << 8) +#define DEBUGCTLMSR_BTS_OFF_OS (1UL << 9) +#define DEBUGCTLMSR_BTS_OFF_USR (1UL << 10) +#define DEBUGCTLMSR_FREEZE_LBRS_ON_PMI (1UL << 11) #define MSR_IA32_MC0_CTL 0x00000400 #define MSR_IA32_MC0_STATUS 0x00000401 diff --git a/arch/x86/kernel/cpu/perf_event_intel_ds.c b/arch/x86/kernel/cpu/perf_event_intel_ds.c index c59678a..2fea362 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_ds.c +++ b/arch/x86/kernel/cpu/perf_event_intel_ds.c @@ -38,15 +38,6 @@ struct pebs_record_nhm { }; /* - * Bits in the debugctlmsr controlling branch tracing. - */ -#define X86_DEBUGCTL_TR (1 << 6) -#define X86_DEBUGCTL_BTS (1 << 7) -#define X86_DEBUGCTL_BTINT (1 << 8) -#define X86_DEBUGCTL_BTS_OFF_OS (1 << 9) -#define X86_DEBUGCTL_BTS_OFF_USR (1 << 10) - -/* * A debug store configuration. * * We only support architectures that use 64bit fields. @@ -193,15 +184,15 @@ static void intel_pmu_enable_bts(u64 config) debugctlmsr = get_debugctlmsr(); - debugctlmsr |= X86_DEBUGCTL_TR; - debugctlmsr |= X86_DEBUGCTL_BTS; - debugctlmsr |= X86_DEBUGCTL_BTINT; + debugctlmsr |= DEBUGCTLMSR_TR; + debugctlmsr |= DEBUGCTLMSR_BTS; + debugctlmsr |= DEBUGCTLMSR_BTINT; if (!(config & ARCH_PERFMON_EVENTSEL_OS)) - debugctlmsr |= X86_DEBUGCTL_BTS_OFF_OS; + debugctlmsr |= DEBUGCTLMSR_BTS_OFF_OS; if (!(config & ARCH_PERFMON_EVENTSEL_USR)) - debugctlmsr |= X86_DEBUGCTL_BTS_OFF_USR; + debugctlmsr |= DEBUGCTLMSR_BTS_OFF_USR; update_debugctlmsr(debugctlmsr); } @@ -217,8 +208,8 @@ static void intel_pmu_disable_bts(void) debugctlmsr = get_debugctlmsr(); debugctlmsr &= - ~(X86_DEBUGCTL_TR | X86_DEBUGCTL_BTS | X86_DEBUGCTL_BTINT | - X86_DEBUGCTL_BTS_OFF_OS | X86_DEBUGCTL_BTS_OFF_USR); + ~(DEBUGCTLMSR_TR | DEBUGCTLMSR_BTS | DEBUGCTLMSR_BTINT | + DEBUGCTLMSR_BTS_OFF_OS | DEBUGCTLMSR_BTS_OFF_USR); update_debugctlmsr(debugctlmsr); } diff --git a/arch/x86/kernel/cpu/perf_event_intel_lbr.c b/arch/x86/kernel/cpu/perf_event_intel_lbr.c index df4c98e..d202c1b 100644 --- a/arch/x86/kernel/cpu/perf_event_intel_lbr.c +++ b/arch/x86/kernel/cpu/perf_event_intel_lbr.c @@ -12,15 +12,12 @@ enum { * otherwise it becomes near impossible to get a reliable stack. */ -#define X86_DEBUGCTL_LBR (1 << 0) -#define X86_DEBUGCTL_FREEZE_LBRS_ON_PMI (1 << 11) - static void __intel_pmu_lbr_enable(void) { u64 debugctl; rdmsrl(MSR_IA32_DEBUGCTLMSR, debugctl); - debugctl |= (X86_DEBUGCTL_LBR | X86_DEBUGCTL_FREEZE_LBRS_ON_PMI); + debugctl |= (DEBUGCTLMSR_LBR | DEBUGCTLMSR_FREEZE_LBRS_ON_PMI); wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctl); } @@ -29,7 +26,7 @@ static void __intel_pmu_lbr_disable(void) u64 debugctl; rdmsrl(MSR_IA32_DEBUGCTLMSR, debugctl); - debugctl &= ~(X86_DEBUGCTL_LBR | X86_DEBUGCTL_FREEZE_LBRS_ON_PMI); + debugctl &= ~(DEBUGCTLMSR_LBR | DEBUGCTLMSR_FREEZE_LBRS_ON_PMI); wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctl); } -- cgit v0.10.2 From faa4602e47690fb11221e00f9b9697c8dc0d4b19 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 25 Mar 2010 14:51:50 +0100 Subject: x86, perf, bts, mm: Delete the never used BTS-ptrace code Support for the PMU's BTS features has been upstreamed in v2.6.32, but we still have the old and disabled ptrace-BTS, as Linus noticed it not so long ago. It's buggy: TIF_DEBUGCTLMSR is trampling all over that MSR without regard for other uses (perf) and doesn't provide the flexibility needed for perf either. Its users are ptrace-block-step and ptrace-bts, since ptrace-bts was never used and ptrace-block-step can be implemented using a much simpler approach. So axe all 3000 lines of it. That includes the *locked_memory*() APIs in mm/mlock.c as well. Reported-by: Linus Torvalds Signed-off-by: Peter Zijlstra Cc: Roland McGrath Cc: Oleg Nesterov Cc: Markus Metzger Cc: Steven Rostedt Cc: Andrew Morton LKML-Reference: <20100325135413.938004390@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/Kconfig.cpu b/arch/x86/Kconfig.cpu index a198293..918fbb1 100644 --- a/arch/x86/Kconfig.cpu +++ b/arch/x86/Kconfig.cpu @@ -502,23 +502,3 @@ config CPU_SUP_UMC_32 CPU might render the kernel unbootable. If unsure, say N. - -config X86_DS - def_bool X86_PTRACE_BTS - depends on X86_DEBUGCTLMSR - select HAVE_HW_BRANCH_TRACER - -config X86_PTRACE_BTS - bool "Branch Trace Store" - default y - depends on X86_DEBUGCTLMSR - depends on BROKEN - ---help--- - This adds a ptrace interface to the hardware's branch trace store. - - Debuggers may use it to collect an execution trace of the debugged - application in order to answer the question 'how did I get here?'. - Debuggers may trace user mode as well as kernel mode. - - Say Y unless there is no application development on this machine - and you want to save a small amount of code size. diff --git a/arch/x86/Kconfig.debug b/arch/x86/Kconfig.debug index bc01e3e..bd58c8a 100644 --- a/arch/x86/Kconfig.debug +++ b/arch/x86/Kconfig.debug @@ -174,15 +174,6 @@ config IOMMU_LEAK Add a simple leak tracer to the IOMMU code. This is useful when you are debugging a buggy device driver that leaks IOMMU mappings. -config X86_DS_SELFTEST - bool "DS selftest" - default y - depends on DEBUG_KERNEL - depends on X86_DS - ---help--- - Perform Debug Store selftests at boot time. - If in doubt, say "N". - config HAVE_MMIOTRACE_SUPPORT def_bool y diff --git a/arch/x86/include/asm/ds.h b/arch/x86/include/asm/ds.h deleted file mode 100644 index 70dac19..0000000 --- a/arch/x86/include/asm/ds.h +++ /dev/null @@ -1,302 +0,0 @@ -/* - * Debug Store (DS) support - * - * This provides a low-level interface to the hardware's Debug Store - * feature that is used for branch trace store (BTS) and - * precise-event based sampling (PEBS). - * - * It manages: - * - DS and BTS hardware configuration - * - buffer overflow handling (to be done) - * - buffer access - * - * It does not do: - * - security checking (is the caller allowed to trace the task) - * - buffer allocation (memory accounting) - * - * - * Copyright (C) 2007-2009 Intel Corporation. - * Markus Metzger , 2007-2009 - */ - -#ifndef _ASM_X86_DS_H -#define _ASM_X86_DS_H - - -#include -#include -#include - - -#ifdef CONFIG_X86_DS - -struct task_struct; -struct ds_context; -struct ds_tracer; -struct bts_tracer; -struct pebs_tracer; - -typedef void (*bts_ovfl_callback_t)(struct bts_tracer *); -typedef void (*pebs_ovfl_callback_t)(struct pebs_tracer *); - - -/* - * A list of features plus corresponding macros to talk about them in - * the ds_request function's flags parameter. - * - * We use the enum to index an array of corresponding control bits; - * we use the macro to index a flags bit-vector. - */ -enum ds_feature { - dsf_bts = 0, - dsf_bts_kernel, -#define BTS_KERNEL (1 << dsf_bts_kernel) - /* trace kernel-mode branches */ - - dsf_bts_user, -#define BTS_USER (1 << dsf_bts_user) - /* trace user-mode branches */ - - dsf_bts_overflow, - dsf_bts_max, - dsf_pebs = dsf_bts_max, - - dsf_pebs_max, - dsf_ctl_max = dsf_pebs_max, - dsf_bts_timestamps = dsf_ctl_max, -#define BTS_TIMESTAMPS (1 << dsf_bts_timestamps) - /* add timestamps into BTS trace */ - -#define BTS_USER_FLAGS (BTS_KERNEL | BTS_USER | BTS_TIMESTAMPS) -}; - - -/* - * Request BTS or PEBS - * - * Due to alignement constraints, the actual buffer may be slightly - * smaller than the requested or provided buffer. - * - * Returns a pointer to a tracer structure on success, or - * ERR_PTR(errcode) on failure. - * - * The interrupt threshold is independent from the overflow callback - * to allow users to use their own overflow interrupt handling mechanism. - * - * The function might sleep. - * - * task: the task to request recording for - * cpu: the cpu to request recording for - * base: the base pointer for the (non-pageable) buffer; - * size: the size of the provided buffer in bytes - * ovfl: pointer to a function to be called on buffer overflow; - * NULL if cyclic buffer requested - * th: the interrupt threshold in records from the end of the buffer; - * -1 if no interrupt threshold is requested. - * flags: a bit-mask of the above flags - */ -extern struct bts_tracer *ds_request_bts_task(struct task_struct *task, - void *base, size_t size, - bts_ovfl_callback_t ovfl, - size_t th, unsigned int flags); -extern struct bts_tracer *ds_request_bts_cpu(int cpu, void *base, size_t size, - bts_ovfl_callback_t ovfl, - size_t th, unsigned int flags); -extern struct pebs_tracer *ds_request_pebs_task(struct task_struct *task, - void *base, size_t size, - pebs_ovfl_callback_t ovfl, - size_t th, unsigned int flags); -extern struct pebs_tracer *ds_request_pebs_cpu(int cpu, - void *base, size_t size, - pebs_ovfl_callback_t ovfl, - size_t th, unsigned int flags); - -/* - * Release BTS or PEBS resources - * Suspend and resume BTS or PEBS tracing - * - * Must be called with irq's enabled. - * - * tracer: the tracer handle returned from ds_request_~() - */ -extern void ds_release_bts(struct bts_tracer *tracer); -extern void ds_suspend_bts(struct bts_tracer *tracer); -extern void ds_resume_bts(struct bts_tracer *tracer); -extern void ds_release_pebs(struct pebs_tracer *tracer); -extern void ds_suspend_pebs(struct pebs_tracer *tracer); -extern void ds_resume_pebs(struct pebs_tracer *tracer); - -/* - * Release BTS or PEBS resources - * Suspend and resume BTS or PEBS tracing - * - * Cpu tracers must call this on the traced cpu. - * Task tracers must call ds_release_~_noirq() for themselves. - * - * May be called with irq's disabled. - * - * Returns 0 if successful; - * -EPERM if the cpu tracer does not trace the current cpu. - * -EPERM if the task tracer does not trace itself. - * - * tracer: the tracer handle returned from ds_request_~() - */ -extern int ds_release_bts_noirq(struct bts_tracer *tracer); -extern int ds_suspend_bts_noirq(struct bts_tracer *tracer); -extern int ds_resume_bts_noirq(struct bts_tracer *tracer); -extern int ds_release_pebs_noirq(struct pebs_tracer *tracer); -extern int ds_suspend_pebs_noirq(struct pebs_tracer *tracer); -extern int ds_resume_pebs_noirq(struct pebs_tracer *tracer); - - -/* - * The raw DS buffer state as it is used for BTS and PEBS recording. - * - * This is the low-level, arch-dependent interface for working - * directly on the raw trace data. - */ -struct ds_trace { - /* the number of bts/pebs records */ - size_t n; - /* the size of a bts/pebs record in bytes */ - size_t size; - /* pointers into the raw buffer: - - to the first entry */ - void *begin; - /* - one beyond the last entry */ - void *end; - /* - one beyond the newest entry */ - void *top; - /* - the interrupt threshold */ - void *ith; - /* flags given on ds_request() */ - unsigned int flags; -}; - -/* - * An arch-independent view on branch trace data. - */ -enum bts_qualifier { - bts_invalid, -#define BTS_INVALID bts_invalid - - bts_branch, -#define BTS_BRANCH bts_branch - - bts_task_arrives, -#define BTS_TASK_ARRIVES bts_task_arrives - - bts_task_departs, -#define BTS_TASK_DEPARTS bts_task_departs - - bts_qual_bit_size = 4, - bts_qual_max = (1 << bts_qual_bit_size), -}; - -struct bts_struct { - __u64 qualifier; - union { - /* BTS_BRANCH */ - struct { - __u64 from; - __u64 to; - } lbr; - /* BTS_TASK_ARRIVES or BTS_TASK_DEPARTS */ - struct { - __u64 clock; - pid_t pid; - } event; - } variant; -}; - - -/* - * The BTS state. - * - * This gives access to the raw DS state and adds functions to provide - * an arch-independent view of the BTS data. - */ -struct bts_trace { - struct ds_trace ds; - - int (*read)(struct bts_tracer *tracer, const void *at, - struct bts_struct *out); - int (*write)(struct bts_tracer *tracer, const struct bts_struct *in); -}; - - -/* - * The PEBS state. - * - * This gives access to the raw DS state and the PEBS-specific counter - * reset value. - */ -struct pebs_trace { - struct ds_trace ds; - - /* the number of valid counters in the below array */ - unsigned int counters; - -#define MAX_PEBS_COUNTERS 4 - /* the counter reset value */ - unsigned long long counter_reset[MAX_PEBS_COUNTERS]; -}; - - -/* - * Read the BTS or PEBS trace. - * - * Returns a view on the trace collected for the parameter tracer. - * - * The view remains valid as long as the traced task is not running or - * the tracer is suspended. - * Writes into the trace buffer are not reflected. - * - * tracer: the tracer handle returned from ds_request_~() - */ -extern const struct bts_trace *ds_read_bts(struct bts_tracer *tracer); -extern const struct pebs_trace *ds_read_pebs(struct pebs_tracer *tracer); - - -/* - * Reset the write pointer of the BTS/PEBS buffer. - * - * Returns 0 on success; -Eerrno on error - * - * tracer: the tracer handle returned from ds_request_~() - */ -extern int ds_reset_bts(struct bts_tracer *tracer); -extern int ds_reset_pebs(struct pebs_tracer *tracer); - -/* - * Set the PEBS counter reset value. - * - * Returns 0 on success; -Eerrno on error - * - * tracer: the tracer handle returned from ds_request_pebs() - * counter: the index of the counter - * value: the new counter reset value - */ -extern int ds_set_pebs_reset(struct pebs_tracer *tracer, - unsigned int counter, u64 value); - -/* - * Initialization - */ -struct cpuinfo_x86; -extern void __cpuinit ds_init_intel(struct cpuinfo_x86 *); - -/* - * Context switch work - */ -extern void ds_switch_to(struct task_struct *prev, struct task_struct *next); - -#else /* CONFIG_X86_DS */ - -struct cpuinfo_x86; -static inline void __cpuinit ds_init_intel(struct cpuinfo_x86 *ignored) {} -static inline void ds_switch_to(struct task_struct *prev, - struct task_struct *next) {} - -#endif /* CONFIG_X86_DS */ -#endif /* _ASM_X86_DS_H */ diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h index b753ea5..5bec21a 100644 --- a/arch/x86/include/asm/processor.h +++ b/arch/x86/include/asm/processor.h @@ -21,7 +21,6 @@ struct mm_struct; #include #include #include -#include #include #include @@ -29,6 +28,7 @@ struct mm_struct; #include #include #include +#include #define HBP_NUM 4 /* @@ -473,10 +473,6 @@ struct thread_struct { unsigned long iopl; /* Max allowed port in the bitmap, in bytes: */ unsigned io_bitmap_max; -/* MSR_IA32_DEBUGCTLMSR value to switch in if TIF_DEBUGCTLMSR is set. */ - unsigned long debugctlmsr; - /* Debug Store context; see asm/ds.h */ - struct ds_context *ds_ctx; }; static inline unsigned long native_get_debugreg(int regno) @@ -814,21 +810,6 @@ static inline unsigned long get_debugctlmsr(void) return debugctlmsr; } -static inline unsigned long get_debugctlmsr_on_cpu(int cpu) -{ - u64 debugctlmsr = 0; - u32 val1, val2; - -#ifndef CONFIG_X86_DEBUGCTLMSR - if (boot_cpu_data.x86 < 6) - return 0; -#endif - rdmsr_on_cpu(cpu, MSR_IA32_DEBUGCTLMSR, &val1, &val2); - debugctlmsr = val1 | ((u64)val2 << 32); - - return debugctlmsr; -} - static inline void update_debugctlmsr(unsigned long debugctlmsr) { #ifndef CONFIG_X86_DEBUGCTLMSR @@ -838,18 +819,6 @@ static inline void update_debugctlmsr(unsigned long debugctlmsr) wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctlmsr); } -static inline void update_debugctlmsr_on_cpu(int cpu, - unsigned long debugctlmsr) -{ -#ifndef CONFIG_X86_DEBUGCTLMSR - if (boot_cpu_data.x86 < 6) - return; -#endif - wrmsr_on_cpu(cpu, MSR_IA32_DEBUGCTLMSR, - (u32)((u64)debugctlmsr), - (u32)((u64)debugctlmsr >> 32)); -} - /* * from system description table in BIOS. Mostly for MCA use, but * others may find it useful: diff --git a/arch/x86/include/asm/ptrace-abi.h b/arch/x86/include/asm/ptrace-abi.h index 8672303..52b098a 100644 --- a/arch/x86/include/asm/ptrace-abi.h +++ b/arch/x86/include/asm/ptrace-abi.h @@ -82,61 +82,6 @@ #ifndef __ASSEMBLY__ #include - -/* configuration/status structure used in PTRACE_BTS_CONFIG and - PTRACE_BTS_STATUS commands. -*/ -struct ptrace_bts_config { - /* requested or actual size of BTS buffer in bytes */ - __u32 size; - /* bitmask of below flags */ - __u32 flags; - /* buffer overflow signal */ - __u32 signal; - /* actual size of bts_struct in bytes */ - __u32 bts_size; -}; -#endif /* __ASSEMBLY__ */ - -#define PTRACE_BTS_O_TRACE 0x1 /* branch trace */ -#define PTRACE_BTS_O_SCHED 0x2 /* scheduling events w/ jiffies */ -#define PTRACE_BTS_O_SIGNAL 0x4 /* send SIG on buffer overflow - instead of wrapping around */ -#define PTRACE_BTS_O_ALLOC 0x8 /* (re)allocate buffer */ - -#define PTRACE_BTS_CONFIG 40 -/* Configure branch trace recording. - ADDR points to a struct ptrace_bts_config. - DATA gives the size of that buffer. - A new buffer is allocated, if requested in the flags. - An overflow signal may only be requested for new buffers. - Returns the number of bytes read. -*/ -#define PTRACE_BTS_STATUS 41 -/* Return the current configuration in a struct ptrace_bts_config - pointed to by ADDR; DATA gives the size of that buffer. - Returns the number of bytes written. -*/ -#define PTRACE_BTS_SIZE 42 -/* Return the number of available BTS records for draining. - DATA and ADDR are ignored. -*/ -#define PTRACE_BTS_GET 43 -/* Get a single BTS record. - DATA defines the index into the BTS array, where 0 is the newest - entry, and higher indices refer to older entries. - ADDR is pointing to struct bts_struct (see asm/ds.h). -*/ -#define PTRACE_BTS_CLEAR 44 -/* Clear the BTS buffer. - DATA and ADDR are ignored. -*/ -#define PTRACE_BTS_DRAIN 45 -/* Read all available BTS records and clear the buffer. - ADDR points to an array of struct bts_struct. - DATA gives the size of that buffer. - BTS records are read from oldest to newest. - Returns number of BTS records drained. -*/ +#endif #endif /* _ASM_X86_PTRACE_ABI_H */ diff --git a/arch/x86/include/asm/ptrace.h b/arch/x86/include/asm/ptrace.h index 69a686a..78cd1ea 100644 --- a/arch/x86/include/asm/ptrace.h +++ b/arch/x86/include/asm/ptrace.h @@ -289,12 +289,6 @@ extern int do_get_thread_area(struct task_struct *p, int idx, extern int do_set_thread_area(struct task_struct *p, int idx, struct user_desc __user *info, int can_allocate); -#ifdef CONFIG_X86_PTRACE_BTS -extern void ptrace_bts_untrace(struct task_struct *tsk); - -#define arch_ptrace_untrace(tsk) ptrace_bts_untrace(tsk) -#endif /* CONFIG_X86_PTRACE_BTS */ - #endif /* __KERNEL__ */ #endif /* !__ASSEMBLY__ */ diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h index e0d2890..dc85e12 100644 --- a/arch/x86/include/asm/thread_info.h +++ b/arch/x86/include/asm/thread_info.h @@ -92,8 +92,6 @@ struct thread_info { #define TIF_IO_BITMAP 22 /* uses I/O bitmap */ #define TIF_FREEZE 23 /* is freezing for suspend */ #define TIF_FORCED_TF 24 /* true if TF in eflags artificially */ -#define TIF_DEBUGCTLMSR 25 /* uses thread_struct.debugctlmsr */ -#define TIF_DS_AREA_MSR 26 /* uses thread_struct.ds_area_msr */ #define TIF_LAZY_MMU_UPDATES 27 /* task is updating the mmu lazily */ #define TIF_SYSCALL_TRACEPOINT 28 /* syscall tracepoint instrumentation */ @@ -115,8 +113,6 @@ struct thread_info { #define _TIF_IO_BITMAP (1 << TIF_IO_BITMAP) #define _TIF_FREEZE (1 << TIF_FREEZE) #define _TIF_FORCED_TF (1 << TIF_FORCED_TF) -#define _TIF_DEBUGCTLMSR (1 << TIF_DEBUGCTLMSR) -#define _TIF_DS_AREA_MSR (1 << TIF_DS_AREA_MSR) #define _TIF_LAZY_MMU_UPDATES (1 << TIF_LAZY_MMU_UPDATES) #define _TIF_SYSCALL_TRACEPOINT (1 << TIF_SYSCALL_TRACEPOINT) @@ -147,7 +143,7 @@ struct thread_info { /* flags to check in __switch_to() */ #define _TIF_WORK_CTXSW \ - (_TIF_IO_BITMAP|_TIF_DEBUGCTLMSR|_TIF_DS_AREA_MSR|_TIF_NOTSC) + (_TIF_IO_BITMAP|_TIF_NOTSC) #define _TIF_WORK_CTXSW_PREV (_TIF_WORK_CTXSW|_TIF_USER_RETURN_NOTIFY) #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW|_TIF_DEBUG) diff --git a/arch/x86/kernel/Makefile b/arch/x86/kernel/Makefile index 4c58352..e77b220 100644 --- a/arch/x86/kernel/Makefile +++ b/arch/x86/kernel/Makefile @@ -47,8 +47,6 @@ obj-$(CONFIG_X86_TRAMPOLINE) += trampoline.o obj-y += process.o obj-y += i387.o xsave.o obj-y += ptrace.o -obj-$(CONFIG_X86_DS) += ds.o -obj-$(CONFIG_X86_DS_SELFTEST) += ds_selftest.o obj-$(CONFIG_X86_32) += tls.o obj-$(CONFIG_IA32_EMULATION) += tls.o obj-y += step.o diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index 7e1cca1..d72377c 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -12,7 +12,6 @@ #include #include #include -#include #include #include @@ -367,7 +366,6 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c) set_cpu_cap(c, X86_FEATURE_BTS); if (!(l1 & (1<<12))) set_cpu_cap(c, X86_FEATURE_PEBS); - ds_init_intel(c); } if (c->x86 == 6 && c->x86_model == 29 && cpu_has_clflush) diff --git a/arch/x86/kernel/ds.c b/arch/x86/kernel/ds.c deleted file mode 100644 index 1c47390..0000000 --- a/arch/x86/kernel/ds.c +++ /dev/null @@ -1,1437 +0,0 @@ -/* - * Debug Store support - * - * This provides a low-level interface to the hardware's Debug Store - * feature that is used for branch trace store (BTS) and - * precise-event based sampling (PEBS). - * - * It manages: - * - DS and BTS hardware configuration - * - buffer overflow handling (to be done) - * - buffer access - * - * It does not do: - * - security checking (is the caller allowed to trace the task) - * - buffer allocation (memory accounting) - * - * - * Copyright (C) 2007-2009 Intel Corporation. - * Markus Metzger , 2007-2009 - */ - -#include -#include -#include -#include -#include -#include -#include - -#include - -#include "ds_selftest.h" - -/* - * The configuration for a particular DS hardware implementation: - */ -struct ds_configuration { - /* The name of the configuration: */ - const char *name; - - /* The size of pointer-typed fields in DS, BTS, and PEBS: */ - unsigned char sizeof_ptr_field; - - /* The size of a BTS/PEBS record in bytes: */ - unsigned char sizeof_rec[2]; - - /* The number of pebs counter reset values in the DS structure. */ - unsigned char nr_counter_reset; - - /* Control bit-masks indexed by enum ds_feature: */ - unsigned long ctl[dsf_ctl_max]; -}; -static struct ds_configuration ds_cfg __read_mostly; - - -/* Maximal size of a DS configuration: */ -#define MAX_SIZEOF_DS 0x80 - -/* Maximal size of a BTS record: */ -#define MAX_SIZEOF_BTS (3 * 8) - -/* BTS and PEBS buffer alignment: */ -#define DS_ALIGNMENT (1 << 3) - -/* Number of buffer pointers in DS: */ -#define NUM_DS_PTR_FIELDS 8 - -/* Size of a pebs reset value in DS: */ -#define PEBS_RESET_FIELD_SIZE 8 - -/* Mask of control bits in the DS MSR register: */ -#define BTS_CONTROL \ - ( ds_cfg.ctl[dsf_bts] | \ - ds_cfg.ctl[dsf_bts_kernel] | \ - ds_cfg.ctl[dsf_bts_user] | \ - ds_cfg.ctl[dsf_bts_overflow] ) - -/* - * A BTS or PEBS tracer. - * - * This holds the configuration of the tracer and serves as a handle - * to identify tracers. - */ -struct ds_tracer { - /* The DS context (partially) owned by this tracer. */ - struct ds_context *context; - /* The buffer provided on ds_request() and its size in bytes. */ - void *buffer; - size_t size; -}; - -struct bts_tracer { - /* The common DS part: */ - struct ds_tracer ds; - - /* The trace including the DS configuration: */ - struct bts_trace trace; - - /* Buffer overflow notification function: */ - bts_ovfl_callback_t ovfl; - - /* Active flags affecting trace collection. */ - unsigned int flags; -}; - -struct pebs_tracer { - /* The common DS part: */ - struct ds_tracer ds; - - /* The trace including the DS configuration: */ - struct pebs_trace trace; - - /* Buffer overflow notification function: */ - pebs_ovfl_callback_t ovfl; -}; - -/* - * Debug Store (DS) save area configuration (see Intel64 and IA32 - * Architectures Software Developer's Manual, section 18.5) - * - * The DS configuration consists of the following fields; different - * architetures vary in the size of those fields. - * - * - double-word aligned base linear address of the BTS buffer - * - write pointer into the BTS buffer - * - end linear address of the BTS buffer (one byte beyond the end of - * the buffer) - * - interrupt pointer into BTS buffer - * (interrupt occurs when write pointer passes interrupt pointer) - * - double-word aligned base linear address of the PEBS buffer - * - write pointer into the PEBS buffer - * - end linear address of the PEBS buffer (one byte beyond the end of - * the buffer) - * - interrupt pointer into PEBS buffer - * (interrupt occurs when write pointer passes interrupt pointer) - * - value to which counter is reset following counter overflow - * - * Later architectures use 64bit pointers throughout, whereas earlier - * architectures use 32bit pointers in 32bit mode. - * - * - * We compute the base address for the first 8 fields based on: - * - the field size stored in the DS configuration - * - the relative field position - * - an offset giving the start of the respective region - * - * This offset is further used to index various arrays holding - * information for BTS and PEBS at the respective index. - * - * On later 32bit processors, we only access the lower 32bit of the - * 64bit pointer fields. The upper halves will be zeroed out. - */ - -enum ds_field { - ds_buffer_base = 0, - ds_index, - ds_absolute_maximum, - ds_interrupt_threshold, -}; - -enum ds_qualifier { - ds_bts = 0, - ds_pebs -}; - -static inline unsigned long -ds_get(const unsigned char *base, enum ds_qualifier qual, enum ds_field field) -{ - base += (ds_cfg.sizeof_ptr_field * (field + (4 * qual))); - return *(unsigned long *)base; -} - -static inline void -ds_set(unsigned char *base, enum ds_qualifier qual, enum ds_field field, - unsigned long value) -{ - base += (ds_cfg.sizeof_ptr_field * (field + (4 * qual))); - (*(unsigned long *)base) = value; -} - - -/* - * Locking is done only for allocating BTS or PEBS resources. - */ -static DEFINE_SPINLOCK(ds_lock); - -/* - * We either support (system-wide) per-cpu or per-thread allocation. - * We distinguish the two based on the task_struct pointer, where a - * NULL pointer indicates per-cpu allocation for the current cpu. - * - * Allocations are use-counted. As soon as resources are allocated, - * further allocations must be of the same type (per-cpu or - * per-thread). We model this by counting allocations (i.e. the number - * of tracers of a certain type) for one type negatively: - * =0 no tracers - * >0 number of per-thread tracers - * <0 number of per-cpu tracers - * - * Tracers essentially gives the number of ds contexts for a certain - * type of allocation. - */ -static atomic_t tracers = ATOMIC_INIT(0); - -static inline int get_tracer(struct task_struct *task) -{ - int error; - - spin_lock_irq(&ds_lock); - - if (task) { - error = -EPERM; - if (atomic_read(&tracers) < 0) - goto out; - atomic_inc(&tracers); - } else { - error = -EPERM; - if (atomic_read(&tracers) > 0) - goto out; - atomic_dec(&tracers); - } - - error = 0; -out: - spin_unlock_irq(&ds_lock); - return error; -} - -static inline void put_tracer(struct task_struct *task) -{ - if (task) - atomic_dec(&tracers); - else - atomic_inc(&tracers); -} - -/* - * The DS context is either attached to a thread or to a cpu: - * - in the former case, the thread_struct contains a pointer to the - * attached context. - * - in the latter case, we use a static array of per-cpu context - * pointers. - * - * Contexts are use-counted. They are allocated on first access and - * deallocated when the last user puts the context. - */ -struct ds_context { - /* The DS configuration; goes into MSR_IA32_DS_AREA: */ - unsigned char ds[MAX_SIZEOF_DS]; - - /* The owner of the BTS and PEBS configuration, respectively: */ - struct bts_tracer *bts_master; - struct pebs_tracer *pebs_master; - - /* Use count: */ - unsigned long count; - - /* Pointer to the context pointer field: */ - struct ds_context **this; - - /* The traced task; NULL for cpu tracing: */ - struct task_struct *task; - - /* The traced cpu; only valid if task is NULL: */ - int cpu; -}; - -static DEFINE_PER_CPU(struct ds_context *, cpu_ds_context); - - -static struct ds_context *ds_get_context(struct task_struct *task, int cpu) -{ - struct ds_context **p_context = - (task ? &task->thread.ds_ctx : &per_cpu(cpu_ds_context, cpu)); - struct ds_context *context = NULL; - struct ds_context *new_context = NULL; - - /* Chances are small that we already have a context. */ - new_context = kzalloc(sizeof(*new_context), GFP_KERNEL); - if (!new_context) - return NULL; - - spin_lock_irq(&ds_lock); - - context = *p_context; - if (likely(!context)) { - context = new_context; - - context->this = p_context; - context->task = task; - context->cpu = cpu; - context->count = 0; - - *p_context = context; - } - - context->count++; - - spin_unlock_irq(&ds_lock); - - if (context != new_context) - kfree(new_context); - - return context; -} - -static void ds_put_context(struct ds_context *context) -{ - struct task_struct *task; - unsigned long irq; - - if (!context) - return; - - spin_lock_irqsave(&ds_lock, irq); - - if (--context->count) { - spin_unlock_irqrestore(&ds_lock, irq); - return; - } - - *(context->this) = NULL; - - task = context->task; - - if (task) - clear_tsk_thread_flag(task, TIF_DS_AREA_MSR); - - /* - * We leave the (now dangling) pointer to the DS configuration in - * the DS_AREA msr. This is as good or as bad as replacing it with - * NULL - the hardware would crash if we enabled tracing. - * - * This saves us some problems with having to write an msr on a - * different cpu while preventing others from doing the same for the - * next context for that same cpu. - */ - - spin_unlock_irqrestore(&ds_lock, irq); - - /* The context might still be in use for context switching. */ - if (task && (task != current)) - wait_task_context_switch(task); - - kfree(context); -} - -static void ds_install_ds_area(struct ds_context *context) -{ - unsigned long ds; - - ds = (unsigned long)context->ds; - - /* - * There is a race between the bts master and the pebs master. - * - * The thread/cpu access is synchronized via get/put_cpu() for - * task tracing and via wrmsr_on_cpu for cpu tracing. - * - * If bts and pebs are collected for the same task or same cpu, - * the same confiuration is written twice. - */ - if (context->task) { - get_cpu(); - if (context->task == current) - wrmsrl(MSR_IA32_DS_AREA, ds); - set_tsk_thread_flag(context->task, TIF_DS_AREA_MSR); - put_cpu(); - } else - wrmsr_on_cpu(context->cpu, MSR_IA32_DS_AREA, - (u32)((u64)ds), (u32)((u64)ds >> 32)); -} - -/* - * Call the tracer's callback on a buffer overflow. - * - * context: the ds context - * qual: the buffer type - */ -static void ds_overflow(struct ds_context *context, enum ds_qualifier qual) -{ - switch (qual) { - case ds_bts: - if (context->bts_master && - context->bts_master->ovfl) - context->bts_master->ovfl(context->bts_master); - break; - case ds_pebs: - if (context->pebs_master && - context->pebs_master->ovfl) - context->pebs_master->ovfl(context->pebs_master); - break; - } -} - - -/* - * Write raw data into the BTS or PEBS buffer. - * - * The remainder of any partially written record is zeroed out. - * - * context: the DS context - * qual: the buffer type - * record: the data to write - * size: the size of the data - */ -static int ds_write(struct ds_context *context, enum ds_qualifier qual, - const void *record, size_t size) -{ - int bytes_written = 0; - - if (!record) - return -EINVAL; - - while (size) { - unsigned long base, index, end, write_end, int_th; - unsigned long write_size, adj_write_size; - - /* - * Write as much as possible without producing an - * overflow interrupt. - * - * Interrupt_threshold must either be - * - bigger than absolute_maximum or - * - point to a record between buffer_base and absolute_maximum - * - * Index points to a valid record. - */ - base = ds_get(context->ds, qual, ds_buffer_base); - index = ds_get(context->ds, qual, ds_index); - end = ds_get(context->ds, qual, ds_absolute_maximum); - int_th = ds_get(context->ds, qual, ds_interrupt_threshold); - - write_end = min(end, int_th); - - /* - * If we are already beyond the interrupt threshold, - * we fill the entire buffer. - */ - if (write_end <= index) - write_end = end; - - if (write_end <= index) - break; - - write_size = min((unsigned long) size, write_end - index); - memcpy((void *)index, record, write_size); - - record = (const char *)record + write_size; - size -= write_size; - bytes_written += write_size; - - adj_write_size = write_size / ds_cfg.sizeof_rec[qual]; - adj_write_size *= ds_cfg.sizeof_rec[qual]; - - /* Zero out trailing bytes. */ - memset((char *)index + write_size, 0, - adj_write_size - write_size); - index += adj_write_size; - - if (index >= end) - index = base; - ds_set(context->ds, qual, ds_index, index); - - if (index >= int_th) - ds_overflow(context, qual); - } - - return bytes_written; -} - - -/* - * Branch Trace Store (BTS) uses the following format. Different - * architectures vary in the size of those fields. - * - source linear address - * - destination linear address - * - flags - * - * Later architectures use 64bit pointers throughout, whereas earlier - * architectures use 32bit pointers in 32bit mode. - * - * We compute the base address for the fields based on: - * - the field size stored in the DS configuration - * - the relative field position - * - * In order to store additional information in the BTS buffer, we use - * a special source address to indicate that the record requires - * special interpretation. - * - * Netburst indicated via a bit in the flags field whether the branch - * was predicted; this is ignored. - * - * We use two levels of abstraction: - * - the raw data level defined here - * - an arch-independent level defined in ds.h - */ - -enum bts_field { - bts_from, - bts_to, - bts_flags, - - bts_qual = bts_from, - bts_clock = bts_to, - bts_pid = bts_flags, - - bts_qual_mask = (bts_qual_max - 1), - bts_escape = ((unsigned long)-1 & ~bts_qual_mask) -}; - -static inline unsigned long bts_get(const char *base, unsigned long field) -{ - base += (ds_cfg.sizeof_ptr_field * field); - return *(unsigned long *)base; -} - -static inline void bts_set(char *base, unsigned long field, unsigned long val) -{ - base += (ds_cfg.sizeof_ptr_field * field); - (*(unsigned long *)base) = val; -} - - -/* - * The raw BTS data is architecture dependent. - * - * For higher-level users, we give an arch-independent view. - * - ds.h defines struct bts_struct - * - bts_read translates one raw bts record into a bts_struct - * - bts_write translates one bts_struct into the raw format and - * writes it into the top of the parameter tracer's buffer. - * - * return: bytes read/written on success; -Eerrno, otherwise - */ -static int -bts_read(struct bts_tracer *tracer, const void *at, struct bts_struct *out) -{ - if (!tracer) - return -EINVAL; - - if (at < tracer->trace.ds.begin) - return -EINVAL; - - if (tracer->trace.ds.end < (at + tracer->trace.ds.size)) - return -EINVAL; - - memset(out, 0, sizeof(*out)); - if ((bts_get(at, bts_qual) & ~bts_qual_mask) == bts_escape) { - out->qualifier = (bts_get(at, bts_qual) & bts_qual_mask); - out->variant.event.clock = bts_get(at, bts_clock); - out->variant.event.pid = bts_get(at, bts_pid); - } else { - out->qualifier = bts_branch; - out->variant.lbr.from = bts_get(at, bts_from); - out->variant.lbr.to = bts_get(at, bts_to); - - if (!out->variant.lbr.from && !out->variant.lbr.to) - out->qualifier = bts_invalid; - } - - return ds_cfg.sizeof_rec[ds_bts]; -} - -static int bts_write(struct bts_tracer *tracer, const struct bts_struct *in) -{ - unsigned char raw[MAX_SIZEOF_BTS]; - - if (!tracer) - return -EINVAL; - - if (MAX_SIZEOF_BTS < ds_cfg.sizeof_rec[ds_bts]) - return -EOVERFLOW; - - switch (in->qualifier) { - case bts_invalid: - bts_set(raw, bts_from, 0); - bts_set(raw, bts_to, 0); - bts_set(raw, bts_flags, 0); - break; - case bts_branch: - bts_set(raw, bts_from, in->variant.lbr.from); - bts_set(raw, bts_to, in->variant.lbr.to); - bts_set(raw, bts_flags, 0); - break; - case bts_task_arrives: - case bts_task_departs: - bts_set(raw, bts_qual, (bts_escape | in->qualifier)); - bts_set(raw, bts_clock, in->variant.event.clock); - bts_set(raw, bts_pid, in->variant.event.pid); - break; - default: - return -EINVAL; - } - - return ds_write(tracer->ds.context, ds_bts, raw, - ds_cfg.sizeof_rec[ds_bts]); -} - - -static void ds_write_config(struct ds_context *context, - struct ds_trace *cfg, enum ds_qualifier qual) -{ - unsigned char *ds = context->ds; - - ds_set(ds, qual, ds_buffer_base, (unsigned long)cfg->begin); - ds_set(ds, qual, ds_index, (unsigned long)cfg->top); - ds_set(ds, qual, ds_absolute_maximum, (unsigned long)cfg->end); - ds_set(ds, qual, ds_interrupt_threshold, (unsigned long)cfg->ith); -} - -static void ds_read_config(struct ds_context *context, - struct ds_trace *cfg, enum ds_qualifier qual) -{ - unsigned char *ds = context->ds; - - cfg->begin = (void *)ds_get(ds, qual, ds_buffer_base); - cfg->top = (void *)ds_get(ds, qual, ds_index); - cfg->end = (void *)ds_get(ds, qual, ds_absolute_maximum); - cfg->ith = (void *)ds_get(ds, qual, ds_interrupt_threshold); -} - -static void ds_init_ds_trace(struct ds_trace *trace, enum ds_qualifier qual, - void *base, size_t size, size_t ith, - unsigned int flags) { - unsigned long buffer, adj; - - /* - * Adjust the buffer address and size to meet alignment - * constraints: - * - buffer is double-word aligned - * - size is multiple of record size - * - * We checked the size at the very beginning; we have enough - * space to do the adjustment. - */ - buffer = (unsigned long)base; - - adj = ALIGN(buffer, DS_ALIGNMENT) - buffer; - buffer += adj; - size -= adj; - - trace->n = size / ds_cfg.sizeof_rec[qual]; - trace->size = ds_cfg.sizeof_rec[qual]; - - size = (trace->n * trace->size); - - trace->begin = (void *)buffer; - trace->top = trace->begin; - trace->end = (void *)(buffer + size); - /* - * The value for 'no threshold' is -1, which will set the - * threshold outside of the buffer, just like we want it. - */ - ith *= ds_cfg.sizeof_rec[qual]; - trace->ith = (void *)(buffer + size - ith); - - trace->flags = flags; -} - - -static int ds_request(struct ds_tracer *tracer, struct ds_trace *trace, - enum ds_qualifier qual, struct task_struct *task, - int cpu, void *base, size_t size, size_t th) -{ - struct ds_context *context; - int error; - size_t req_size; - - error = -EOPNOTSUPP; - if (!ds_cfg.sizeof_rec[qual]) - goto out; - - error = -EINVAL; - if (!base) - goto out; - - req_size = ds_cfg.sizeof_rec[qual]; - /* We might need space for alignment adjustments. */ - if (!IS_ALIGNED((unsigned long)base, DS_ALIGNMENT)) - req_size += DS_ALIGNMENT; - - error = -EINVAL; - if (size < req_size) - goto out; - - if (th != (size_t)-1) { - th *= ds_cfg.sizeof_rec[qual]; - - error = -EINVAL; - if (size <= th) - goto out; - } - - tracer->buffer = base; - tracer->size = size; - - error = -ENOMEM; - context = ds_get_context(task, cpu); - if (!context) - goto out; - tracer->context = context; - - /* - * Defer any tracer-specific initialization work for the context until - * context ownership has been clarified. - */ - - error = 0; - out: - return error; -} - -static struct bts_tracer *ds_request_bts(struct task_struct *task, int cpu, - void *base, size_t size, - bts_ovfl_callback_t ovfl, size_t th, - unsigned int flags) -{ - struct bts_tracer *tracer; - int error; - - /* Buffer overflow notification is not yet implemented. */ - error = -EOPNOTSUPP; - if (ovfl) - goto out; - - error = get_tracer(task); - if (error < 0) - goto out; - - error = -ENOMEM; - tracer = kzalloc(sizeof(*tracer), GFP_KERNEL); - if (!tracer) - goto out_put_tracer; - tracer->ovfl = ovfl; - - /* Do some more error checking and acquire a tracing context. */ - error = ds_request(&tracer->ds, &tracer->trace.ds, - ds_bts, task, cpu, base, size, th); - if (error < 0) - goto out_tracer; - - /* Claim the bts part of the tracing context we acquired above. */ - spin_lock_irq(&ds_lock); - - error = -EPERM; - if (tracer->ds.context->bts_master) - goto out_unlock; - tracer->ds.context->bts_master = tracer; - - spin_unlock_irq(&ds_lock); - - /* - * Now that we own the bts part of the context, let's complete the - * initialization for that part. - */ - ds_init_ds_trace(&tracer->trace.ds, ds_bts, base, size, th, flags); - ds_write_config(tracer->ds.context, &tracer->trace.ds, ds_bts); - ds_install_ds_area(tracer->ds.context); - - tracer->trace.read = bts_read; - tracer->trace.write = bts_write; - - /* Start tracing. */ - ds_resume_bts(tracer); - - return tracer; - - out_unlock: - spin_unlock_irq(&ds_lock); - ds_put_context(tracer->ds.context); - out_tracer: - kfree(tracer); - out_put_tracer: - put_tracer(task); - out: - return ERR_PTR(error); -} - -struct bts_tracer *ds_request_bts_task(struct task_struct *task, - void *base, size_t size, - bts_ovfl_callback_t ovfl, - size_t th, unsigned int flags) -{ - return ds_request_bts(task, 0, base, size, ovfl, th, flags); -} - -struct bts_tracer *ds_request_bts_cpu(int cpu, void *base, size_t size, - bts_ovfl_callback_t ovfl, - size_t th, unsigned int flags) -{ - return ds_request_bts(NULL, cpu, base, size, ovfl, th, flags); -} - -static struct pebs_tracer *ds_request_pebs(struct task_struct *task, int cpu, - void *base, size_t size, - pebs_ovfl_callback_t ovfl, size_t th, - unsigned int flags) -{ - struct pebs_tracer *tracer; - int error; - - /* Buffer overflow notification is not yet implemented. */ - error = -EOPNOTSUPP; - if (ovfl) - goto out; - - error = get_tracer(task); - if (error < 0) - goto out; - - error = -ENOMEM; - tracer = kzalloc(sizeof(*tracer), GFP_KERNEL); - if (!tracer) - goto out_put_tracer; - tracer->ovfl = ovfl; - - /* Do some more error checking and acquire a tracing context. */ - error = ds_request(&tracer->ds, &tracer->trace.ds, - ds_pebs, task, cpu, base, size, th); - if (error < 0) - goto out_tracer; - - /* Claim the pebs part of the tracing context we acquired above. */ - spin_lock_irq(&ds_lock); - - error = -EPERM; - if (tracer->ds.context->pebs_master) - goto out_unlock; - tracer->ds.context->pebs_master = tracer; - - spin_unlock_irq(&ds_lock); - - /* - * Now that we own the pebs part of the context, let's complete the - * initialization for that part. - */ - ds_init_ds_trace(&tracer->trace.ds, ds_pebs, base, size, th, flags); - ds_write_config(tracer->ds.context, &tracer->trace.ds, ds_pebs); - ds_install_ds_area(tracer->ds.context); - - /* Start tracing. */ - ds_resume_pebs(tracer); - - return tracer; - - out_unlock: - spin_unlock_irq(&ds_lock); - ds_put_context(tracer->ds.context); - out_tracer: - kfree(tracer); - out_put_tracer: - put_tracer(task); - out: - return ERR_PTR(error); -} - -struct pebs_tracer *ds_request_pebs_task(struct task_struct *task, - void *base, size_t size, - pebs_ovfl_callback_t ovfl, - size_t th, unsigned int flags) -{ - return ds_request_pebs(task, 0, base, size, ovfl, th, flags); -} - -struct pebs_tracer *ds_request_pebs_cpu(int cpu, void *base, size_t size, - pebs_ovfl_callback_t ovfl, - size_t th, unsigned int flags) -{ - return ds_request_pebs(NULL, cpu, base, size, ovfl, th, flags); -} - -static void ds_free_bts(struct bts_tracer *tracer) -{ - struct task_struct *task; - - task = tracer->ds.context->task; - - WARN_ON_ONCE(tracer->ds.context->bts_master != tracer); - tracer->ds.context->bts_master = NULL; - - /* Make sure tracing stopped and the tracer is not in use. */ - if (task && (task != current)) - wait_task_context_switch(task); - - ds_put_context(tracer->ds.context); - put_tracer(task); - - kfree(tracer); -} - -void ds_release_bts(struct bts_tracer *tracer) -{ - might_sleep(); - - if (!tracer) - return; - - ds_suspend_bts(tracer); - ds_free_bts(tracer); -} - -int ds_release_bts_noirq(struct bts_tracer *tracer) -{ - struct task_struct *task; - unsigned long irq; - int error; - - if (!tracer) - return 0; - - task = tracer->ds.context->task; - - local_irq_save(irq); - - error = -EPERM; - if (!task && - (tracer->ds.context->cpu != smp_processor_id())) - goto out; - - error = -EPERM; - if (task && (task != current)) - goto out; - - ds_suspend_bts_noirq(tracer); - ds_free_bts(tracer); - - error = 0; - out: - local_irq_restore(irq); - return error; -} - -static void update_task_debugctlmsr(struct task_struct *task, - unsigned long debugctlmsr) -{ - task->thread.debugctlmsr = debugctlmsr; - - get_cpu(); - if (task == current) - update_debugctlmsr(debugctlmsr); - put_cpu(); -} - -void ds_suspend_bts(struct bts_tracer *tracer) -{ - struct task_struct *task; - unsigned long debugctlmsr; - int cpu; - - if (!tracer) - return; - - tracer->flags = 0; - - task = tracer->ds.context->task; - cpu = tracer->ds.context->cpu; - - WARN_ON(!task && irqs_disabled()); - - debugctlmsr = (task ? - task->thread.debugctlmsr : - get_debugctlmsr_on_cpu(cpu)); - debugctlmsr &= ~BTS_CONTROL; - - if (task) - update_task_debugctlmsr(task, debugctlmsr); - else - update_debugctlmsr_on_cpu(cpu, debugctlmsr); -} - -int ds_suspend_bts_noirq(struct bts_tracer *tracer) -{ - struct task_struct *task; - unsigned long debugctlmsr, irq; - int cpu, error = 0; - - if (!tracer) - return 0; - - tracer->flags = 0; - - task = tracer->ds.context->task; - cpu = tracer->ds.context->cpu; - - local_irq_save(irq); - - error = -EPERM; - if (!task && (cpu != smp_processor_id())) - goto out; - - debugctlmsr = (task ? - task->thread.debugctlmsr : - get_debugctlmsr()); - debugctlmsr &= ~BTS_CONTROL; - - if (task) - update_task_debugctlmsr(task, debugctlmsr); - else - update_debugctlmsr(debugctlmsr); - - error = 0; - out: - local_irq_restore(irq); - return error; -} - -static unsigned long ds_bts_control(struct bts_tracer *tracer) -{ - unsigned long control; - - control = ds_cfg.ctl[dsf_bts]; - if (!(tracer->trace.ds.flags & BTS_KERNEL)) - control |= ds_cfg.ctl[dsf_bts_kernel]; - if (!(tracer->trace.ds.flags & BTS_USER)) - control |= ds_cfg.ctl[dsf_bts_user]; - - return control; -} - -void ds_resume_bts(struct bts_tracer *tracer) -{ - struct task_struct *task; - unsigned long debugctlmsr; - int cpu; - - if (!tracer) - return; - - tracer->flags = tracer->trace.ds.flags; - - task = tracer->ds.context->task; - cpu = tracer->ds.context->cpu; - - WARN_ON(!task && irqs_disabled()); - - debugctlmsr = (task ? - task->thread.debugctlmsr : - get_debugctlmsr_on_cpu(cpu)); - debugctlmsr |= ds_bts_control(tracer); - - if (task) - update_task_debugctlmsr(task, debugctlmsr); - else - update_debugctlmsr_on_cpu(cpu, debugctlmsr); -} - -int ds_resume_bts_noirq(struct bts_tracer *tracer) -{ - struct task_struct *task; - unsigned long debugctlmsr, irq; - int cpu, error = 0; - - if (!tracer) - return 0; - - tracer->flags = tracer->trace.ds.flags; - - task = tracer->ds.context->task; - cpu = tracer->ds.context->cpu; - - local_irq_save(irq); - - error = -EPERM; - if (!task && (cpu != smp_processor_id())) - goto out; - - debugctlmsr = (task ? - task->thread.debugctlmsr : - get_debugctlmsr()); - debugctlmsr |= ds_bts_control(tracer); - - if (task) - update_task_debugctlmsr(task, debugctlmsr); - else - update_debugctlmsr(debugctlmsr); - - error = 0; - out: - local_irq_restore(irq); - return error; -} - -static void ds_free_pebs(struct pebs_tracer *tracer) -{ - struct task_struct *task; - - task = tracer->ds.context->task; - - WARN_ON_ONCE(tracer->ds.context->pebs_master != tracer); - tracer->ds.context->pebs_master = NULL; - - ds_put_context(tracer->ds.context); - put_tracer(task); - - kfree(tracer); -} - -void ds_release_pebs(struct pebs_tracer *tracer) -{ - might_sleep(); - - if (!tracer) - return; - - ds_suspend_pebs(tracer); - ds_free_pebs(tracer); -} - -int ds_release_pebs_noirq(struct pebs_tracer *tracer) -{ - struct task_struct *task; - unsigned long irq; - int error; - - if (!tracer) - return 0; - - task = tracer->ds.context->task; - - local_irq_save(irq); - - error = -EPERM; - if (!task && - (tracer->ds.context->cpu != smp_processor_id())) - goto out; - - error = -EPERM; - if (task && (task != current)) - goto out; - - ds_suspend_pebs_noirq(tracer); - ds_free_pebs(tracer); - - error = 0; - out: - local_irq_restore(irq); - return error; -} - -void ds_suspend_pebs(struct pebs_tracer *tracer) -{ - -} - -int ds_suspend_pebs_noirq(struct pebs_tracer *tracer) -{ - return 0; -} - -void ds_resume_pebs(struct pebs_tracer *tracer) -{ - -} - -int ds_resume_pebs_noirq(struct pebs_tracer *tracer) -{ - return 0; -} - -const struct bts_trace *ds_read_bts(struct bts_tracer *tracer) -{ - if (!tracer) - return NULL; - - ds_read_config(tracer->ds.context, &tracer->trace.ds, ds_bts); - return &tracer->trace; -} - -const struct pebs_trace *ds_read_pebs(struct pebs_tracer *tracer) -{ - if (!tracer) - return NULL; - - ds_read_config(tracer->ds.context, &tracer->trace.ds, ds_pebs); - - tracer->trace.counters = ds_cfg.nr_counter_reset; - memcpy(tracer->trace.counter_reset, - tracer->ds.context->ds + - (NUM_DS_PTR_FIELDS * ds_cfg.sizeof_ptr_field), - ds_cfg.nr_counter_reset * PEBS_RESET_FIELD_SIZE); - - return &tracer->trace; -} - -int ds_reset_bts(struct bts_tracer *tracer) -{ - if (!tracer) - return -EINVAL; - - tracer->trace.ds.top = tracer->trace.ds.begin; - - ds_set(tracer->ds.context->ds, ds_bts, ds_index, - (unsigned long)tracer->trace.ds.top); - - return 0; -} - -int ds_reset_pebs(struct pebs_tracer *tracer) -{ - if (!tracer) - return -EINVAL; - - tracer->trace.ds.top = tracer->trace.ds.begin; - - ds_set(tracer->ds.context->ds, ds_pebs, ds_index, - (unsigned long)tracer->trace.ds.top); - - return 0; -} - -int ds_set_pebs_reset(struct pebs_tracer *tracer, - unsigned int counter, u64 value) -{ - if (!tracer) - return -EINVAL; - - if (ds_cfg.nr_counter_reset < counter) - return -EINVAL; - - *(u64 *)(tracer->ds.context->ds + - (NUM_DS_PTR_FIELDS * ds_cfg.sizeof_ptr_field) + - (counter * PEBS_RESET_FIELD_SIZE)) = value; - - return 0; -} - -static const struct ds_configuration ds_cfg_netburst = { - .name = "Netburst", - .ctl[dsf_bts] = (1 << 2) | (1 << 3), - .ctl[dsf_bts_kernel] = (1 << 5), - .ctl[dsf_bts_user] = (1 << 6), - .nr_counter_reset = 1, -}; -static const struct ds_configuration ds_cfg_pentium_m = { - .name = "Pentium M", - .ctl[dsf_bts] = (1 << 6) | (1 << 7), - .nr_counter_reset = 1, -}; -static const struct ds_configuration ds_cfg_core2_atom = { - .name = "Core 2/Atom", - .ctl[dsf_bts] = (1 << 6) | (1 << 7), - .ctl[dsf_bts_kernel] = (1 << 9), - .ctl[dsf_bts_user] = (1 << 10), - .nr_counter_reset = 1, -}; -static const struct ds_configuration ds_cfg_core_i7 = { - .name = "Core i7", - .ctl[dsf_bts] = (1 << 6) | (1 << 7), - .ctl[dsf_bts_kernel] = (1 << 9), - .ctl[dsf_bts_user] = (1 << 10), - .nr_counter_reset = 4, -}; - -static void -ds_configure(const struct ds_configuration *cfg, - struct cpuinfo_x86 *cpu) -{ - unsigned long nr_pebs_fields = 0; - - printk(KERN_INFO "[ds] using %s configuration\n", cfg->name); - -#ifdef __i386__ - nr_pebs_fields = 10; -#else - nr_pebs_fields = 18; -#endif - - /* - * Starting with version 2, architectural performance - * monitoring supports a format specifier. - */ - if ((cpuid_eax(0xa) & 0xff) > 1) { - unsigned long perf_capabilities, format; - - rdmsrl(MSR_IA32_PERF_CAPABILITIES, perf_capabilities); - - format = (perf_capabilities >> 8) & 0xf; - - switch (format) { - case 0: - nr_pebs_fields = 18; - break; - case 1: - nr_pebs_fields = 22; - break; - default: - printk(KERN_INFO - "[ds] unknown PEBS format: %lu\n", format); - nr_pebs_fields = 0; - break; - } - } - - memset(&ds_cfg, 0, sizeof(ds_cfg)); - ds_cfg = *cfg; - - ds_cfg.sizeof_ptr_field = - (cpu_has(cpu, X86_FEATURE_DTES64) ? 8 : 4); - - ds_cfg.sizeof_rec[ds_bts] = ds_cfg.sizeof_ptr_field * 3; - ds_cfg.sizeof_rec[ds_pebs] = ds_cfg.sizeof_ptr_field * nr_pebs_fields; - - if (!cpu_has(cpu, X86_FEATURE_BTS)) { - ds_cfg.sizeof_rec[ds_bts] = 0; - printk(KERN_INFO "[ds] bts not available\n"); - } - if (!cpu_has(cpu, X86_FEATURE_PEBS)) { - ds_cfg.sizeof_rec[ds_pebs] = 0; - printk(KERN_INFO "[ds] pebs not available\n"); - } - - printk(KERN_INFO "[ds] sizes: address: %u bit, ", - 8 * ds_cfg.sizeof_ptr_field); - printk("bts/pebs record: %u/%u bytes\n", - ds_cfg.sizeof_rec[ds_bts], ds_cfg.sizeof_rec[ds_pebs]); - - WARN_ON_ONCE(MAX_PEBS_COUNTERS < ds_cfg.nr_counter_reset); -} - -void __cpuinit ds_init_intel(struct cpuinfo_x86 *c) -{ - /* Only configure the first cpu. Others are identical. */ - if (ds_cfg.name) - return; - - switch (c->x86) { - case 0x6: - switch (c->x86_model) { - case 0x9: - case 0xd: /* Pentium M */ - ds_configure(&ds_cfg_pentium_m, c); - break; - case 0xf: - case 0x17: /* Core2 */ - case 0x1c: /* Atom */ - ds_configure(&ds_cfg_core2_atom, c); - break; - case 0x1a: /* Core i7 */ - ds_configure(&ds_cfg_core_i7, c); - break; - default: - /* Sorry, don't know about them. */ - break; - } - break; - case 0xf: - switch (c->x86_model) { - case 0x0: - case 0x1: - case 0x2: /* Netburst */ - ds_configure(&ds_cfg_netburst, c); - break; - default: - /* Sorry, don't know about them. */ - break; - } - break; - default: - /* Sorry, don't know about them. */ - break; - } -} - -static inline void ds_take_timestamp(struct ds_context *context, - enum bts_qualifier qualifier, - struct task_struct *task) -{ - struct bts_tracer *tracer = context->bts_master; - struct bts_struct ts; - - /* Prevent compilers from reading the tracer pointer twice. */ - barrier(); - - if (!tracer || !(tracer->flags & BTS_TIMESTAMPS)) - return; - - memset(&ts, 0, sizeof(ts)); - ts.qualifier = qualifier; - ts.variant.event.clock = trace_clock_global(); - ts.variant.event.pid = task->pid; - - bts_write(tracer, &ts); -} - -/* - * Change the DS configuration from tracing prev to tracing next. - */ -void ds_switch_to(struct task_struct *prev, struct task_struct *next) -{ - struct ds_context *prev_ctx = prev->thread.ds_ctx; - struct ds_context *next_ctx = next->thread.ds_ctx; - unsigned long debugctlmsr = next->thread.debugctlmsr; - - /* Make sure all data is read before we start. */ - barrier(); - - if (prev_ctx) { - update_debugctlmsr(0); - - ds_take_timestamp(prev_ctx, bts_task_departs, prev); - } - - if (next_ctx) { - ds_take_timestamp(next_ctx, bts_task_arrives, next); - - wrmsrl(MSR_IA32_DS_AREA, (unsigned long)next_ctx->ds); - } - - update_debugctlmsr(debugctlmsr); -} - -static __init int ds_selftest(void) -{ - if (ds_cfg.sizeof_rec[ds_bts]) { - int error; - - error = ds_selftest_bts(); - if (error) { - WARN(1, "[ds] selftest failed. disabling bts.\n"); - ds_cfg.sizeof_rec[ds_bts] = 0; - } - } - - if (ds_cfg.sizeof_rec[ds_pebs]) { - int error; - - error = ds_selftest_pebs(); - if (error) { - WARN(1, "[ds] selftest failed. disabling pebs.\n"); - ds_cfg.sizeof_rec[ds_pebs] = 0; - } - } - - return 0; -} -device_initcall(ds_selftest); diff --git a/arch/x86/kernel/ds_selftest.c b/arch/x86/kernel/ds_selftest.c deleted file mode 100644 index 6bc7c19..0000000 --- a/arch/x86/kernel/ds_selftest.c +++ /dev/null @@ -1,408 +0,0 @@ -/* - * Debug Store support - selftest - * - * - * Copyright (C) 2009 Intel Corporation. - * Markus Metzger , 2009 - */ - -#include "ds_selftest.h" - -#include -#include -#include -#include - -#include - - -#define BUFFER_SIZE 521 /* Intentionally chose an odd size. */ -#define SMALL_BUFFER_SIZE 24 /* A single bts entry. */ - -struct ds_selftest_bts_conf { - struct bts_tracer *tracer; - int error; - int (*suspend)(struct bts_tracer *); - int (*resume)(struct bts_tracer *); -}; - -static int ds_selftest_bts_consistency(const struct bts_trace *trace) -{ - int error = 0; - - if (!trace) { - printk(KERN_CONT "failed to access trace..."); - /* Bail out. Other tests are pointless. */ - return -1; - } - - if (!trace->read) { - printk(KERN_CONT "bts read not available..."); - error = -1; - } - - /* Do some sanity checks on the trace configuration. */ - if (!trace->ds.n) { - printk(KERN_CONT "empty bts buffer..."); - error = -1; - } - if (!trace->ds.size) { - printk(KERN_CONT "bad bts trace setup..."); - error = -1; - } - if (trace->ds.end != - (char *)trace->ds.begin + (trace->ds.n * trace->ds.size)) { - printk(KERN_CONT "bad bts buffer setup..."); - error = -1; - } - /* - * We allow top in [begin; end], since its not clear when the - * overflow adjustment happens: after the increment or before the - * write. - */ - if ((trace->ds.top < trace->ds.begin) || - (trace->ds.end < trace->ds.top)) { - printk(KERN_CONT "bts top out of bounds..."); - error = -1; - } - - return error; -} - -static int ds_selftest_bts_read(struct bts_tracer *tracer, - const struct bts_trace *trace, - const void *from, const void *to) -{ - const unsigned char *at; - - /* - * Check a few things which do not belong to this test. - * They should be covered by other tests. - */ - if (!trace) - return -1; - - if (!trace->read) - return -1; - - if (to < from) - return -1; - - if (from < trace->ds.begin) - return -1; - - if (trace->ds.end < to) - return -1; - - if (!trace->ds.size) - return -1; - - /* Now to the test itself. */ - for (at = from; (void *)at < to; at += trace->ds.size) { - struct bts_struct bts; - unsigned long index; - int error; - - if (((void *)at - trace->ds.begin) % trace->ds.size) { - printk(KERN_CONT - "read from non-integer index..."); - return -1; - } - index = ((void *)at - trace->ds.begin) / trace->ds.size; - - memset(&bts, 0, sizeof(bts)); - error = trace->read(tracer, at, &bts); - if (error < 0) { - printk(KERN_CONT - "error reading bts trace at [%lu] (0x%p)...", - index, at); - return error; - } - - switch (bts.qualifier) { - case BTS_BRANCH: - break; - default: - printk(KERN_CONT - "unexpected bts entry %llu at [%lu] (0x%p)...", - bts.qualifier, index, at); - return -1; - } - } - - return 0; -} - -static void ds_selftest_bts_cpu(void *arg) -{ - struct ds_selftest_bts_conf *conf = arg; - const struct bts_trace *trace; - void *top; - - if (IS_ERR(conf->tracer)) { - conf->error = PTR_ERR(conf->tracer); - conf->tracer = NULL; - - printk(KERN_CONT - "initialization failed (err: %d)...", conf->error); - return; - } - - /* We should meanwhile have enough trace. */ - conf->error = conf->suspend(conf->tracer); - if (conf->error < 0) - return; - - /* Let's see if we can access the trace. */ - trace = ds_read_bts(conf->tracer); - - conf->error = ds_selftest_bts_consistency(trace); - if (conf->error < 0) - return; - - /* If everything went well, we should have a few trace entries. */ - if (trace->ds.top == trace->ds.begin) { - /* - * It is possible but highly unlikely that we got a - * buffer overflow and end up at exactly the same - * position we started from. - * Let's issue a warning, but continue. - */ - printk(KERN_CONT "no trace/overflow..."); - } - - /* Let's try to read the trace we collected. */ - conf->error = - ds_selftest_bts_read(conf->tracer, trace, - trace->ds.begin, trace->ds.top); - if (conf->error < 0) - return; - - /* - * Let's read the trace again. - * Since we suspended tracing, we should get the same result. - */ - top = trace->ds.top; - - trace = ds_read_bts(conf->tracer); - conf->error = ds_selftest_bts_consistency(trace); - if (conf->error < 0) - return; - - if (top != trace->ds.top) { - printk(KERN_CONT "suspend not working..."); - conf->error = -1; - return; - } - - /* Let's collect some more trace - see if resume is working. */ - conf->error = conf->resume(conf->tracer); - if (conf->error < 0) - return; - - conf->error = conf->suspend(conf->tracer); - if (conf->error < 0) - return; - - trace = ds_read_bts(conf->tracer); - - conf->error = ds_selftest_bts_consistency(trace); - if (conf->error < 0) - return; - - if (trace->ds.top == top) { - /* - * It is possible but highly unlikely that we got a - * buffer overflow and end up at exactly the same - * position we started from. - * Let's issue a warning and check the full trace. - */ - printk(KERN_CONT - "no resume progress/overflow..."); - - conf->error = - ds_selftest_bts_read(conf->tracer, trace, - trace->ds.begin, trace->ds.end); - } else if (trace->ds.top < top) { - /* - * We had a buffer overflow - the entire buffer should - * contain trace records. - */ - conf->error = - ds_selftest_bts_read(conf->tracer, trace, - trace->ds.begin, trace->ds.end); - } else { - /* - * It is quite likely that the buffer did not overflow. - * Let's just check the delta trace. - */ - conf->error = - ds_selftest_bts_read(conf->tracer, trace, top, - trace->ds.top); - } - if (conf->error < 0) - return; - - conf->error = 0; -} - -static int ds_suspend_bts_wrap(struct bts_tracer *tracer) -{ - ds_suspend_bts(tracer); - return 0; -} - -static int ds_resume_bts_wrap(struct bts_tracer *tracer) -{ - ds_resume_bts(tracer); - return 0; -} - -static void ds_release_bts_noirq_wrap(void *tracer) -{ - (void)ds_release_bts_noirq(tracer); -} - -static int ds_selftest_bts_bad_release_noirq(int cpu, - struct bts_tracer *tracer) -{ - int error = -EPERM; - - /* Try to release the tracer on the wrong cpu. */ - get_cpu(); - if (cpu != smp_processor_id()) { - error = ds_release_bts_noirq(tracer); - if (error != -EPERM) - printk(KERN_CONT "release on wrong cpu..."); - } - put_cpu(); - - return error ? 0 : -1; -} - -static int ds_selftest_bts_bad_request_cpu(int cpu, void *buffer) -{ - struct bts_tracer *tracer; - int error; - - /* Try to request cpu tracing while task tracing is active. */ - tracer = ds_request_bts_cpu(cpu, buffer, BUFFER_SIZE, NULL, - (size_t)-1, BTS_KERNEL); - error = PTR_ERR(tracer); - if (!IS_ERR(tracer)) { - ds_release_bts(tracer); - error = 0; - } - - if (error != -EPERM) - printk(KERN_CONT "cpu/task tracing overlap..."); - - return error ? 0 : -1; -} - -static int ds_selftest_bts_bad_request_task(void *buffer) -{ - struct bts_tracer *tracer; - int error; - - /* Try to request cpu tracing while task tracing is active. */ - tracer = ds_request_bts_task(current, buffer, BUFFER_SIZE, NULL, - (size_t)-1, BTS_KERNEL); - error = PTR_ERR(tracer); - if (!IS_ERR(tracer)) { - error = 0; - ds_release_bts(tracer); - } - - if (error != -EPERM) - printk(KERN_CONT "task/cpu tracing overlap..."); - - return error ? 0 : -1; -} - -int ds_selftest_bts(void) -{ - struct ds_selftest_bts_conf conf; - unsigned char buffer[BUFFER_SIZE], *small_buffer; - unsigned long irq; - int cpu; - - printk(KERN_INFO "[ds] bts selftest..."); - conf.error = 0; - - small_buffer = (unsigned char *)ALIGN((unsigned long)buffer, 8) + 8; - - get_online_cpus(); - for_each_online_cpu(cpu) { - conf.suspend = ds_suspend_bts_wrap; - conf.resume = ds_resume_bts_wrap; - conf.tracer = - ds_request_bts_cpu(cpu, buffer, BUFFER_SIZE, - NULL, (size_t)-1, BTS_KERNEL); - ds_selftest_bts_cpu(&conf); - if (conf.error >= 0) - conf.error = ds_selftest_bts_bad_request_task(buffer); - ds_release_bts(conf.tracer); - if (conf.error < 0) - goto out; - - conf.suspend = ds_suspend_bts_noirq; - conf.resume = ds_resume_bts_noirq; - conf.tracer = - ds_request_bts_cpu(cpu, buffer, BUFFER_SIZE, - NULL, (size_t)-1, BTS_KERNEL); - smp_call_function_single(cpu, ds_selftest_bts_cpu, &conf, 1); - if (conf.error >= 0) { - conf.error = - ds_selftest_bts_bad_release_noirq(cpu, - conf.tracer); - /* We must not release the tracer twice. */ - if (conf.error < 0) - conf.tracer = NULL; - } - if (conf.error >= 0) - conf.error = ds_selftest_bts_bad_request_task(buffer); - smp_call_function_single(cpu, ds_release_bts_noirq_wrap, - conf.tracer, 1); - if (conf.error < 0) - goto out; - } - - conf.suspend = ds_suspend_bts_wrap; - conf.resume = ds_resume_bts_wrap; - conf.tracer = - ds_request_bts_task(current, buffer, BUFFER_SIZE, - NULL, (size_t)-1, BTS_KERNEL); - ds_selftest_bts_cpu(&conf); - if (conf.error >= 0) - conf.error = ds_selftest_bts_bad_request_cpu(0, buffer); - ds_release_bts(conf.tracer); - if (conf.error < 0) - goto out; - - conf.suspend = ds_suspend_bts_noirq; - conf.resume = ds_resume_bts_noirq; - conf.tracer = - ds_request_bts_task(current, small_buffer, SMALL_BUFFER_SIZE, - NULL, (size_t)-1, BTS_KERNEL); - local_irq_save(irq); - ds_selftest_bts_cpu(&conf); - if (conf.error >= 0) - conf.error = ds_selftest_bts_bad_request_cpu(0, buffer); - ds_release_bts_noirq(conf.tracer); - local_irq_restore(irq); - if (conf.error < 0) - goto out; - - conf.error = 0; - out: - put_online_cpus(); - printk(KERN_CONT "%s.\n", (conf.error ? "failed" : "passed")); - - return conf.error; -} - -int ds_selftest_pebs(void) -{ - return 0; -} diff --git a/arch/x86/kernel/ds_selftest.h b/arch/x86/kernel/ds_selftest.h deleted file mode 100644 index 2ba8745..0000000 --- a/arch/x86/kernel/ds_selftest.h +++ /dev/null @@ -1,15 +0,0 @@ -/* - * Debug Store support - selftest - * - * - * Copyright (C) 2009 Intel Corporation. - * Markus Metzger , 2009 - */ - -#ifdef CONFIG_X86_DS_SELFTEST -extern int ds_selftest_bts(void); -extern int ds_selftest_pebs(void); -#else -static inline int ds_selftest_bts(void) { return 0; } -static inline int ds_selftest_pebs(void) { return 0; } -#endif diff --git a/arch/x86/kernel/dumpstack.c b/arch/x86/kernel/dumpstack.c index 6d81755..c89a386 100644 --- a/arch/x86/kernel/dumpstack.c +++ b/arch/x86/kernel/dumpstack.c @@ -224,11 +224,6 @@ unsigned __kprobes long oops_begin(void) int cpu; unsigned long flags; - /* notify the hw-branch tracer so it may disable tracing and - add the last trace to the trace buffer - - the earlier this happens, the more useful the trace. */ - trace_hw_branch_oops(); - oops_enter(); /* racy, but better than risking deadlock. */ diff --git a/arch/x86/kernel/kprobes.c b/arch/x86/kernel/kprobes.c index b43bbae..7a880ad 100644 --- a/arch/x86/kernel/kprobes.c +++ b/arch/x86/kernel/kprobes.c @@ -422,14 +422,12 @@ static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs, static void __kprobes clear_btf(void) { - if (test_thread_flag(TIF_DEBUGCTLMSR)) - update_debugctlmsr(0); + /* XXX */ } static void __kprobes restore_btf(void) { - if (test_thread_flag(TIF_DEBUGCTLMSR)) - update_debugctlmsr(current->thread.debugctlmsr); + /* XXX */ } void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri, diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index ad95406..1a60beb 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -20,7 +20,6 @@ #include #include #include -#include #include unsigned long idle_halt; @@ -50,8 +49,6 @@ void free_thread_xstate(struct task_struct *tsk) kmem_cache_free(task_xstate_cachep, tsk->thread.xstate); tsk->thread.xstate = NULL; } - - WARN(tsk->thread.ds_ctx, "leaking DS context\n"); } void free_thread_info(struct thread_info *ti) @@ -198,12 +195,6 @@ void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p, prev = &prev_p->thread; next = &next_p->thread; - if (test_tsk_thread_flag(next_p, TIF_DS_AREA_MSR) || - test_tsk_thread_flag(prev_p, TIF_DS_AREA_MSR)) - ds_switch_to(prev_p, next_p); - else if (next->debugctlmsr != prev->debugctlmsr) - update_debugctlmsr(next->debugctlmsr); - if (test_tsk_thread_flag(prev_p, TIF_NOTSC) ^ test_tsk_thread_flag(next_p, TIF_NOTSC)) { /* prev and next are different */ diff --git a/arch/x86/kernel/process_32.c b/arch/x86/kernel/process_32.c index f6c6266..75090c5 100644 --- a/arch/x86/kernel/process_32.c +++ b/arch/x86/kernel/process_32.c @@ -55,7 +55,6 @@ #include #include #include -#include #include asmlinkage void ret_from_fork(void) __asm__("ret_from_fork"); @@ -238,13 +237,6 @@ int copy_thread(unsigned long clone_flags, unsigned long sp, kfree(p->thread.io_bitmap_ptr); p->thread.io_bitmap_max = 0; } - - clear_tsk_thread_flag(p, TIF_DS_AREA_MSR); - p->thread.ds_ctx = NULL; - - clear_tsk_thread_flag(p, TIF_DEBUGCTLMSR); - p->thread.debugctlmsr = 0; - return err; } diff --git a/arch/x86/kernel/process_64.c b/arch/x86/kernel/process_64.c index dc9690b..cc4258f 100644 --- a/arch/x86/kernel/process_64.c +++ b/arch/x86/kernel/process_64.c @@ -49,7 +49,6 @@ #include #include #include -#include #include asmlinkage extern void ret_from_fork(void); @@ -313,13 +312,6 @@ int copy_thread(unsigned long clone_flags, unsigned long sp, if (err) goto out; } - - clear_tsk_thread_flag(p, TIF_DS_AREA_MSR); - p->thread.ds_ctx = NULL; - - clear_tsk_thread_flag(p, TIF_DEBUGCTLMSR); - p->thread.debugctlmsr = 0; - err = 0; out: if (err && p->thread.io_bitmap_ptr) { diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index a503b1f..f2fd3b8 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -2,9 +2,6 @@ /* * Pentium III FXSR, SSE support * Gareth Hughes , May 2000 - * - * BTS tracing - * Markus Metzger , Dec 2007 */ #include @@ -21,7 +18,6 @@ #include #include #include -#include #include #include @@ -35,7 +31,6 @@ #include #include #include -#include #include #include "tls.h" @@ -788,342 +783,6 @@ static int ioperm_get(struct task_struct *target, 0, IO_BITMAP_BYTES); } -#ifdef CONFIG_X86_PTRACE_BTS -/* - * A branch trace store context. - * - * Contexts may only be installed by ptrace_bts_config() and only for - * ptraced tasks. - * - * Contexts are destroyed when the tracee is detached from the tracer. - * The actual destruction work requires interrupts enabled, so the - * work is deferred and will be scheduled during __ptrace_unlink(). - * - * Contexts hold an additional task_struct reference on the traced - * task, as well as a reference on the tracer's mm. - * - * Ptrace already holds a task_struct for the duration of ptrace operations, - * but since destruction is deferred, it may be executed after both - * tracer and tracee exited. - */ -struct bts_context { - /* The branch trace handle. */ - struct bts_tracer *tracer; - - /* The buffer used to store the branch trace and its size. */ - void *buffer; - unsigned int size; - - /* The mm that paid for the above buffer. */ - struct mm_struct *mm; - - /* The task this context belongs to. */ - struct task_struct *task; - - /* The signal to send on a bts buffer overflow. */ - unsigned int bts_ovfl_signal; - - /* The work struct to destroy a context. */ - struct work_struct work; -}; - -static int alloc_bts_buffer(struct bts_context *context, unsigned int size) -{ - void *buffer = NULL; - int err = -ENOMEM; - - err = account_locked_memory(current->mm, current->signal->rlim, size); - if (err < 0) - return err; - - buffer = kzalloc(size, GFP_KERNEL); - if (!buffer) - goto out_refund; - - context->buffer = buffer; - context->size = size; - context->mm = get_task_mm(current); - - return 0; - - out_refund: - refund_locked_memory(current->mm, size); - return err; -} - -static inline void free_bts_buffer(struct bts_context *context) -{ - if (!context->buffer) - return; - - kfree(context->buffer); - context->buffer = NULL; - - refund_locked_memory(context->mm, context->size); - context->size = 0; - - mmput(context->mm); - context->mm = NULL; -} - -static void free_bts_context_work(struct work_struct *w) -{ - struct bts_context *context; - - context = container_of(w, struct bts_context, work); - - ds_release_bts(context->tracer); - put_task_struct(context->task); - free_bts_buffer(context); - kfree(context); -} - -static inline void free_bts_context(struct bts_context *context) -{ - INIT_WORK(&context->work, free_bts_context_work); - schedule_work(&context->work); -} - -static inline struct bts_context *alloc_bts_context(struct task_struct *task) -{ - struct bts_context *context = kzalloc(sizeof(*context), GFP_KERNEL); - if (context) { - context->task = task; - task->bts = context; - - get_task_struct(task); - } - - return context; -} - -static int ptrace_bts_read_record(struct task_struct *child, size_t index, - struct bts_struct __user *out) -{ - struct bts_context *context; - const struct bts_trace *trace; - struct bts_struct bts; - const unsigned char *at; - int error; - - context = child->bts; - if (!context) - return -ESRCH; - - trace = ds_read_bts(context->tracer); - if (!trace) - return -ESRCH; - - at = trace->ds.top - ((index + 1) * trace->ds.size); - if ((void *)at < trace->ds.begin) - at += (trace->ds.n * trace->ds.size); - - if (!trace->read) - return -EOPNOTSUPP; - - error = trace->read(context->tracer, at, &bts); - if (error < 0) - return error; - - if (copy_to_user(out, &bts, sizeof(bts))) - return -EFAULT; - - return sizeof(bts); -} - -static int ptrace_bts_drain(struct task_struct *child, - long size, - struct bts_struct __user *out) -{ - struct bts_context *context; - const struct bts_trace *trace; - const unsigned char *at; - int error, drained = 0; - - context = child->bts; - if (!context) - return -ESRCH; - - trace = ds_read_bts(context->tracer); - if (!trace) - return -ESRCH; - - if (!trace->read) - return -EOPNOTSUPP; - - if (size < (trace->ds.top - trace->ds.begin)) - return -EIO; - - for (at = trace->ds.begin; (void *)at < trace->ds.top; - out++, drained++, at += trace->ds.size) { - struct bts_struct bts; - - error = trace->read(context->tracer, at, &bts); - if (error < 0) - return error; - - if (copy_to_user(out, &bts, sizeof(bts))) - return -EFAULT; - } - - memset(trace->ds.begin, 0, trace->ds.n * trace->ds.size); - - error = ds_reset_bts(context->tracer); - if (error < 0) - return error; - - return drained; -} - -static int ptrace_bts_config(struct task_struct *child, - long cfg_size, - const struct ptrace_bts_config __user *ucfg) -{ - struct bts_context *context; - struct ptrace_bts_config cfg; - unsigned int flags = 0; - - if (cfg_size < sizeof(cfg)) - return -EIO; - - if (copy_from_user(&cfg, ucfg, sizeof(cfg))) - return -EFAULT; - - context = child->bts; - if (!context) - context = alloc_bts_context(child); - if (!context) - return -ENOMEM; - - if (cfg.flags & PTRACE_BTS_O_SIGNAL) { - if (!cfg.signal) - return -EINVAL; - - return -EOPNOTSUPP; - context->bts_ovfl_signal = cfg.signal; - } - - ds_release_bts(context->tracer); - context->tracer = NULL; - - if ((cfg.flags & PTRACE_BTS_O_ALLOC) && (cfg.size != context->size)) { - int err; - - free_bts_buffer(context); - if (!cfg.size) - return 0; - - err = alloc_bts_buffer(context, cfg.size); - if (err < 0) - return err; - } - - if (cfg.flags & PTRACE_BTS_O_TRACE) - flags |= BTS_USER; - - if (cfg.flags & PTRACE_BTS_O_SCHED) - flags |= BTS_TIMESTAMPS; - - context->tracer = - ds_request_bts_task(child, context->buffer, context->size, - NULL, (size_t)-1, flags); - if (unlikely(IS_ERR(context->tracer))) { - int error = PTR_ERR(context->tracer); - - free_bts_buffer(context); - context->tracer = NULL; - return error; - } - - return sizeof(cfg); -} - -static int ptrace_bts_status(struct task_struct *child, - long cfg_size, - struct ptrace_bts_config __user *ucfg) -{ - struct bts_context *context; - const struct bts_trace *trace; - struct ptrace_bts_config cfg; - - context = child->bts; - if (!context) - return -ESRCH; - - if (cfg_size < sizeof(cfg)) - return -EIO; - - trace = ds_read_bts(context->tracer); - if (!trace) - return -ESRCH; - - memset(&cfg, 0, sizeof(cfg)); - cfg.size = trace->ds.end - trace->ds.begin; - cfg.signal = context->bts_ovfl_signal; - cfg.bts_size = sizeof(struct bts_struct); - - if (cfg.signal) - cfg.flags |= PTRACE_BTS_O_SIGNAL; - - if (trace->ds.flags & BTS_USER) - cfg.flags |= PTRACE_BTS_O_TRACE; - - if (trace->ds.flags & BTS_TIMESTAMPS) - cfg.flags |= PTRACE_BTS_O_SCHED; - - if (copy_to_user(ucfg, &cfg, sizeof(cfg))) - return -EFAULT; - - return sizeof(cfg); -} - -static int ptrace_bts_clear(struct task_struct *child) -{ - struct bts_context *context; - const struct bts_trace *trace; - - context = child->bts; - if (!context) - return -ESRCH; - - trace = ds_read_bts(context->tracer); - if (!trace) - return -ESRCH; - - memset(trace->ds.begin, 0, trace->ds.n * trace->ds.size); - - return ds_reset_bts(context->tracer); -} - -static int ptrace_bts_size(struct task_struct *child) -{ - struct bts_context *context; - const struct bts_trace *trace; - - context = child->bts; - if (!context) - return -ESRCH; - - trace = ds_read_bts(context->tracer); - if (!trace) - return -ESRCH; - - return (trace->ds.top - trace->ds.begin) / trace->ds.size; -} - -/* - * Called from __ptrace_unlink() after the child has been moved back - * to its original parent. - */ -void ptrace_bts_untrace(struct task_struct *child) -{ - if (unlikely(child->bts)) { - free_bts_context(child->bts); - child->bts = NULL; - } -} -#endif /* CONFIG_X86_PTRACE_BTS */ - /* * Called by kernel/ptrace.c when detaching.. * @@ -1251,39 +910,6 @@ long arch_ptrace(struct task_struct *child, long request, long addr, long data) break; #endif - /* - * These bits need more cooking - not enabled yet: - */ -#ifdef CONFIG_X86_PTRACE_BTS - case PTRACE_BTS_CONFIG: - ret = ptrace_bts_config - (child, data, (struct ptrace_bts_config __user *)addr); - break; - - case PTRACE_BTS_STATUS: - ret = ptrace_bts_status - (child, data, (struct ptrace_bts_config __user *)addr); - break; - - case PTRACE_BTS_SIZE: - ret = ptrace_bts_size(child); - break; - - case PTRACE_BTS_GET: - ret = ptrace_bts_read_record - (child, data, (struct bts_struct __user *) addr); - break; - - case PTRACE_BTS_CLEAR: - ret = ptrace_bts_clear(child); - break; - - case PTRACE_BTS_DRAIN: - ret = ptrace_bts_drain - (child, data, (struct bts_struct __user *) addr); - break; -#endif /* CONFIG_X86_PTRACE_BTS */ - default: ret = ptrace_request(child, request, addr, data); break; @@ -1543,14 +1169,6 @@ long compat_arch_ptrace(struct task_struct *child, compat_long_t request, case PTRACE_GET_THREAD_AREA: case PTRACE_SET_THREAD_AREA: -#ifdef CONFIG_X86_PTRACE_BTS - case PTRACE_BTS_CONFIG: - case PTRACE_BTS_STATUS: - case PTRACE_BTS_SIZE: - case PTRACE_BTS_GET: - case PTRACE_BTS_CLEAR: - case PTRACE_BTS_DRAIN: -#endif /* CONFIG_X86_PTRACE_BTS */ return arch_ptrace(child, request, addr, data); default: diff --git a/arch/x86/kernel/step.c b/arch/x86/kernel/step.c index 3149032..7beba07 100644 --- a/arch/x86/kernel/step.c +++ b/arch/x86/kernel/step.c @@ -158,22 +158,6 @@ static int enable_single_step(struct task_struct *child) } /* - * Install this value in MSR_IA32_DEBUGCTLMSR whenever child is running. - */ -static void write_debugctlmsr(struct task_struct *child, unsigned long val) -{ - if (child->thread.debugctlmsr == val) - return; - - child->thread.debugctlmsr = val; - - if (child != current) - return; - - update_debugctlmsr(val); -} - -/* * Enable single or block step. */ static void enable_step(struct task_struct *child, bool block) @@ -185,17 +169,9 @@ static void enable_step(struct task_struct *child, bool block) * So noone should try to use debugger block stepping in a program * that uses user-mode single stepping itself. */ - if (enable_single_step(child) && block) { - set_tsk_thread_flag(child, TIF_DEBUGCTLMSR); - write_debugctlmsr(child, - child->thread.debugctlmsr | DEBUGCTLMSR_BTF); - } else { - write_debugctlmsr(child, - child->thread.debugctlmsr & ~DEBUGCTLMSR_BTF); - - if (!child->thread.debugctlmsr) - clear_tsk_thread_flag(child, TIF_DEBUGCTLMSR); - } + if (!enable_single_step(child)) + return; + /* XXX */ } void user_enable_single_step(struct task_struct *child) @@ -213,11 +189,7 @@ void user_disable_single_step(struct task_struct *child) /* * Make sure block stepping (BTF) is disabled. */ - write_debugctlmsr(child, - child->thread.debugctlmsr & ~DEBUGCTLMSR_BTF); - - if (!child->thread.debugctlmsr) - clear_tsk_thread_flag(child, TIF_DEBUGCTLMSR); + /* XXX */ /* Always clear TIF_SINGLESTEP... */ clear_tsk_thread_flag(child, TIF_SINGLESTEP); diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 1168e44..e3da5d7 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -543,11 +543,6 @@ dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code) /* DR6 may or may not be cleared by the CPU */ set_debugreg(0, 6); - /* - * The processor cleared BTF, so don't mark that we need it set. - */ - clear_tsk_thread_flag(tsk, TIF_DEBUGCTLMSR); - tsk->thread.debugctlmsr = 0; /* Store the virtualized DR6 value */ tsk->thread.debugreg6 = dr6; diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h index 01e6ade..cc12b3c 100644 --- a/include/linux/ftrace.h +++ b/include/linux/ftrace.h @@ -504,18 +504,6 @@ extern int ftrace_dump_on_oops; #define INIT_TRACE_RECURSION #endif -#ifdef CONFIG_HW_BRANCH_TRACER - -void trace_hw_branch(u64 from, u64 to); -void trace_hw_branch_oops(void); - -#else /* CONFIG_HW_BRANCH_TRACER */ - -static inline void trace_hw_branch(u64 from, u64 to) {} -static inline void trace_hw_branch_oops(void) {} - -#endif /* CONFIG_HW_BRANCH_TRACER */ - #ifdef CONFIG_FTRACE_SYSCALLS unsigned long arch_syscall_addr(int nr); diff --git a/include/linux/mm.h b/include/linux/mm.h index e70f21b..c8442b6 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -19,7 +19,6 @@ struct anon_vma; struct file_ra_state; struct user_struct; struct writeback_control; -struct rlimit; #ifndef CONFIG_DISCONTIGMEM /* Don't use mapnrs, do it properly */ extern unsigned long max_mapnr; @@ -1449,9 +1448,6 @@ int vmemmap_populate_basepages(struct page *start_page, int vmemmap_populate(struct page *start_page, unsigned long pages, int node); void vmemmap_populate_print_last(void); -extern int account_locked_memory(struct mm_struct *mm, struct rlimit *rlim, - size_t size); -extern void refund_locked_memory(struct mm_struct *mm, size_t size); enum mf_flags { MF_COUNT_INCREASED = 1 << 0, diff --git a/include/linux/ptrace.h b/include/linux/ptrace.h index e1fb607..4272521 100644 --- a/include/linux/ptrace.h +++ b/include/linux/ptrace.h @@ -345,18 +345,6 @@ static inline void user_single_step_siginfo(struct task_struct *tsk, #define arch_ptrace_stop(code, info) do { } while (0) #endif -#ifndef arch_ptrace_untrace -/* - * Do machine-specific work before untracing child. - * - * This is called for a normal detach as well as from ptrace_exit() - * when the tracing task dies. - * - * Called with write_lock(&tasklist_lock) held. - */ -#define arch_ptrace_untrace(task) do { } while (0) -#endif - extern int task_current_syscall(struct task_struct *target, long *callno, unsigned long args[6], unsigned int maxargs, unsigned long *sp, unsigned long *pc); diff --git a/include/linux/sched.h b/include/linux/sched.h index dad7f66..e0447c6 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -99,7 +99,6 @@ struct futex_pi_state; struct robust_list_head; struct bio_list; struct fs_struct; -struct bts_context; struct perf_event_context; /* @@ -1272,12 +1271,6 @@ struct task_struct { struct list_head ptraced; struct list_head ptrace_entry; - /* - * This is the tracer handle for the ptrace BTS extension. - * This field actually belongs to the ptracer task. - */ - struct bts_context *bts; - /* PID/PID hash table linkage. */ struct pid_link pids[PIDTYPE_MAX]; struct list_head thread_group; @@ -2123,10 +2116,8 @@ extern void set_task_comm(struct task_struct *tsk, char *from); extern char *get_task_comm(char *to, struct task_struct *tsk); #ifdef CONFIG_SMP -extern void wait_task_context_switch(struct task_struct *p); extern unsigned long wait_task_inactive(struct task_struct *, long match_state); #else -static inline void wait_task_context_switch(struct task_struct *p) {} static inline unsigned long wait_task_inactive(struct task_struct *p, long match_state) { diff --git a/kernel/fork.c b/kernel/fork.c index 4799c5f..d67f1db 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1108,9 +1108,6 @@ static struct task_struct *copy_process(unsigned long clone_flags, p->memcg_batch.do_batch = 0; p->memcg_batch.memcg = NULL; #endif - - p->bts = NULL; - p->stack_start = stack_start; /* Perform scheduler related setup. Assign this task to a CPU. */ diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 42ad8ae..9fb5123 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -76,7 +76,6 @@ void __ptrace_unlink(struct task_struct *child) child->parent = child->real_parent; list_del_init(&child->ptrace_entry); - arch_ptrace_untrace(child); if (task_is_traced(child)) ptrace_untrace(child); } diff --git a/kernel/sched.c b/kernel/sched.c index 9ab3cd7..117b7ca 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2077,49 +2077,6 @@ migrate_task(struct task_struct *p, int dest_cpu, struct migration_req *req) } /* - * wait_task_context_switch - wait for a thread to complete at least one - * context switch. - * - * @p must not be current. - */ -void wait_task_context_switch(struct task_struct *p) -{ - unsigned long nvcsw, nivcsw, flags; - int running; - struct rq *rq; - - nvcsw = p->nvcsw; - nivcsw = p->nivcsw; - for (;;) { - /* - * The runqueue is assigned before the actual context - * switch. We need to take the runqueue lock. - * - * We could check initially without the lock but it is - * very likely that we need to take the lock in every - * iteration. - */ - rq = task_rq_lock(p, &flags); - running = task_running(rq, p); - task_rq_unlock(rq, &flags); - - if (likely(!running)) - break; - /* - * The switch count is incremented before the actual - * context switch. We thus wait for two switches to be - * sure at least one completed. - */ - if ((p->nvcsw - nvcsw) > 1) - break; - if ((p->nivcsw - nivcsw) > 1) - break; - - cpu_relax(); - } -} - -/* * wait_task_inactive - wait for a thread to unschedule. * * If @match_state is nonzero, it's the @p->state value just checked and diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig index 13e13d4..8b1797c 100644 --- a/kernel/trace/Kconfig +++ b/kernel/trace/Kconfig @@ -44,9 +44,6 @@ config HAVE_FTRACE_MCOUNT_RECORD help See Documentation/trace/ftrace-design.txt -config HAVE_HW_BRANCH_TRACER - bool - config HAVE_SYSCALL_TRACEPOINTS bool help @@ -374,14 +371,6 @@ config STACK_TRACER Say N if unsure. -config HW_BRANCH_TRACER - depends on HAVE_HW_BRANCH_TRACER - bool "Trace hw branches" - select GENERIC_TRACER - help - This tracer records all branches on the system in a circular - buffer, giving access to the last N branches for each cpu. - config KMEMTRACE bool "Trace SLAB allocations" select GENERIC_TRACER diff --git a/kernel/trace/Makefile b/kernel/trace/Makefile index 78edc64..ffb1a5b 100644 --- a/kernel/trace/Makefile +++ b/kernel/trace/Makefile @@ -41,7 +41,6 @@ obj-$(CONFIG_MMIOTRACE) += trace_mmiotrace.o obj-$(CONFIG_BOOT_TRACER) += trace_boot.o obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += trace_functions_graph.o obj-$(CONFIG_TRACE_BRANCH_PROFILING) += trace_branch.o -obj-$(CONFIG_HW_BRANCH_TRACER) += trace_hw_branches.o obj-$(CONFIG_KMEMTRACE) += kmemtrace.o obj-$(CONFIG_WORKQUEUE_TRACER) += trace_workqueue.o obj-$(CONFIG_BLK_DEV_IO_TRACE) += blktrace.o diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 2825ef2..bec2c97 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -34,7 +34,6 @@ enum trace_type { TRACE_GRAPH_RET, TRACE_GRAPH_ENT, TRACE_USER_STACK, - TRACE_HW_BRANCHES, TRACE_KMEM_ALLOC, TRACE_KMEM_FREE, TRACE_BLK, @@ -229,7 +228,6 @@ extern void __ftrace_bad_type(void); TRACE_GRAPH_ENT); \ IF_ASSIGN(var, ent, struct ftrace_graph_ret_entry, \ TRACE_GRAPH_RET); \ - IF_ASSIGN(var, ent, struct hw_branch_entry, TRACE_HW_BRANCHES);\ IF_ASSIGN(var, ent, struct kmemtrace_alloc_entry, \ TRACE_KMEM_ALLOC); \ IF_ASSIGN(var, ent, struct kmemtrace_free_entry, \ @@ -467,8 +465,6 @@ extern int trace_selftest_startup_sysprof(struct tracer *trace, struct trace_array *tr); extern int trace_selftest_startup_branch(struct tracer *trace, struct trace_array *tr); -extern int trace_selftest_startup_hw_branches(struct tracer *trace, - struct trace_array *tr); extern int trace_selftest_startup_ksym(struct tracer *trace, struct trace_array *tr); #endif /* CONFIG_FTRACE_STARTUP_TEST */ diff --git a/kernel/trace/trace_entries.h b/kernel/trace/trace_entries.h index c16a08f..dc008c1 100644 --- a/kernel/trace/trace_entries.h +++ b/kernel/trace/trace_entries.h @@ -318,18 +318,6 @@ FTRACE_ENTRY(branch, trace_branch, __entry->func, __entry->file, __entry->correct) ); -FTRACE_ENTRY(hw_branch, hw_branch_entry, - - TRACE_HW_BRANCHES, - - F_STRUCT( - __field( u64, from ) - __field( u64, to ) - ), - - F_printk("from: %llx to: %llx", __entry->from, __entry->to) -); - FTRACE_ENTRY(kmem_alloc, kmemtrace_alloc_entry, TRACE_KMEM_ALLOC, diff --git a/kernel/trace/trace_hw_branches.c b/kernel/trace/trace_hw_branches.c deleted file mode 100644 index 7b97000..0000000 --- a/kernel/trace/trace_hw_branches.c +++ /dev/null @@ -1,312 +0,0 @@ -/* - * h/w branch tracer for x86 based on BTS - * - * Copyright (C) 2008-2009 Intel Corporation. - * Markus Metzger , 2008-2009 - */ -#include -#include -#include -#include -#include -#include -#include - -#include - -#include "trace_output.h" -#include "trace.h" - - -#define BTS_BUFFER_SIZE (1 << 13) - -static DEFINE_PER_CPU(struct bts_tracer *, hwb_tracer); -static DEFINE_PER_CPU(unsigned char[BTS_BUFFER_SIZE], hwb_buffer); - -#define this_tracer per_cpu(hwb_tracer, smp_processor_id()) - -static int trace_hw_branches_enabled __read_mostly; -static int trace_hw_branches_suspended __read_mostly; -static struct trace_array *hw_branch_trace __read_mostly; - - -static void bts_trace_init_cpu(int cpu) -{ - per_cpu(hwb_tracer, cpu) = - ds_request_bts_cpu(cpu, per_cpu(hwb_buffer, cpu), - BTS_BUFFER_SIZE, NULL, (size_t)-1, - BTS_KERNEL); - - if (IS_ERR(per_cpu(hwb_tracer, cpu))) - per_cpu(hwb_tracer, cpu) = NULL; -} - -static int bts_trace_init(struct trace_array *tr) -{ - int cpu; - - hw_branch_trace = tr; - trace_hw_branches_enabled = 0; - - get_online_cpus(); - for_each_online_cpu(cpu) { - bts_trace_init_cpu(cpu); - - if (likely(per_cpu(hwb_tracer, cpu))) - trace_hw_branches_enabled = 1; - } - trace_hw_branches_suspended = 0; - put_online_cpus(); - - /* If we could not enable tracing on a single cpu, we fail. */ - return trace_hw_branches_enabled ? 0 : -EOPNOTSUPP; -} - -static void bts_trace_reset(struct trace_array *tr) -{ - int cpu; - - get_online_cpus(); - for_each_online_cpu(cpu) { - if (likely(per_cpu(hwb_tracer, cpu))) { - ds_release_bts(per_cpu(hwb_tracer, cpu)); - per_cpu(hwb_tracer, cpu) = NULL; - } - } - trace_hw_branches_enabled = 0; - trace_hw_branches_suspended = 0; - put_online_cpus(); -} - -static void bts_trace_start(struct trace_array *tr) -{ - int cpu; - - get_online_cpus(); - for_each_online_cpu(cpu) - if (likely(per_cpu(hwb_tracer, cpu))) - ds_resume_bts(per_cpu(hwb_tracer, cpu)); - trace_hw_branches_suspended = 0; - put_online_cpus(); -} - -static void bts_trace_stop(struct trace_array *tr) -{ - int cpu; - - get_online_cpus(); - for_each_online_cpu(cpu) - if (likely(per_cpu(hwb_tracer, cpu))) - ds_suspend_bts(per_cpu(hwb_tracer, cpu)); - trace_hw_branches_suspended = 1; - put_online_cpus(); -} - -static int __cpuinit bts_hotcpu_handler(struct notifier_block *nfb, - unsigned long action, void *hcpu) -{ - int cpu = (long)hcpu; - - switch (action) { - case CPU_ONLINE: - case CPU_DOWN_FAILED: - /* The notification is sent with interrupts enabled. */ - if (trace_hw_branches_enabled) { - bts_trace_init_cpu(cpu); - - if (trace_hw_branches_suspended && - likely(per_cpu(hwb_tracer, cpu))) - ds_suspend_bts(per_cpu(hwb_tracer, cpu)); - } - break; - - case CPU_DOWN_PREPARE: - /* The notification is sent with interrupts enabled. */ - if (likely(per_cpu(hwb_tracer, cpu))) { - ds_release_bts(per_cpu(hwb_tracer, cpu)); - per_cpu(hwb_tracer, cpu) = NULL; - } - } - - return NOTIFY_DONE; -} - -static struct notifier_block bts_hotcpu_notifier __cpuinitdata = { - .notifier_call = bts_hotcpu_handler -}; - -static void bts_trace_print_header(struct seq_file *m) -{ - seq_puts(m, "# CPU# TO <- FROM\n"); -} - -static enum print_line_t bts_trace_print_line(struct trace_iterator *iter) -{ - unsigned long symflags = TRACE_ITER_SYM_OFFSET; - struct trace_entry *entry = iter->ent; - struct trace_seq *seq = &iter->seq; - struct hw_branch_entry *it; - - trace_assign_type(it, entry); - - if (entry->type == TRACE_HW_BRANCHES) { - if (trace_seq_printf(seq, "%4d ", iter->cpu) && - seq_print_ip_sym(seq, it->to, symflags) && - trace_seq_printf(seq, "\t <- ") && - seq_print_ip_sym(seq, it->from, symflags) && - trace_seq_printf(seq, "\n")) - return TRACE_TYPE_HANDLED; - return TRACE_TYPE_PARTIAL_LINE; - } - return TRACE_TYPE_UNHANDLED; -} - -void trace_hw_branch(u64 from, u64 to) -{ - struct ftrace_event_call *call = &event_hw_branch; - struct trace_array *tr = hw_branch_trace; - struct ring_buffer_event *event; - struct ring_buffer *buf; - struct hw_branch_entry *entry; - unsigned long irq1; - int cpu; - - if (unlikely(!tr)) - return; - - if (unlikely(!trace_hw_branches_enabled)) - return; - - local_irq_save(irq1); - cpu = raw_smp_processor_id(); - if (atomic_inc_return(&tr->data[cpu]->disabled) != 1) - goto out; - - buf = tr->buffer; - event = trace_buffer_lock_reserve(buf, TRACE_HW_BRANCHES, - sizeof(*entry), 0, 0); - if (!event) - goto out; - entry = ring_buffer_event_data(event); - tracing_generic_entry_update(&entry->ent, 0, from); - entry->ent.type = TRACE_HW_BRANCHES; - entry->from = from; - entry->to = to; - if (!filter_check_discard(call, entry, buf, event)) - trace_buffer_unlock_commit(buf, event, 0, 0); - - out: - atomic_dec(&tr->data[cpu]->disabled); - local_irq_restore(irq1); -} - -static void trace_bts_at(const struct bts_trace *trace, void *at) -{ - struct bts_struct bts; - int err = 0; - - WARN_ON_ONCE(!trace->read); - if (!trace->read) - return; - - err = trace->read(this_tracer, at, &bts); - if (err < 0) - return; - - switch (bts.qualifier) { - case BTS_BRANCH: - trace_hw_branch(bts.variant.lbr.from, bts.variant.lbr.to); - break; - } -} - -/* - * Collect the trace on the current cpu and write it into the ftrace buffer. - * - * pre: tracing must be suspended on the current cpu - */ -static void trace_bts_cpu(void *arg) -{ - struct trace_array *tr = (struct trace_array *)arg; - const struct bts_trace *trace; - unsigned char *at; - - if (unlikely(!tr)) - return; - - if (unlikely(atomic_read(&tr->data[raw_smp_processor_id()]->disabled))) - return; - - if (unlikely(!this_tracer)) - return; - - trace = ds_read_bts(this_tracer); - if (!trace) - return; - - for (at = trace->ds.top; (void *)at < trace->ds.end; - at += trace->ds.size) - trace_bts_at(trace, at); - - for (at = trace->ds.begin; (void *)at < trace->ds.top; - at += trace->ds.size) - trace_bts_at(trace, at); -} - -static void trace_bts_prepare(struct trace_iterator *iter) -{ - int cpu; - - get_online_cpus(); - for_each_online_cpu(cpu) - if (likely(per_cpu(hwb_tracer, cpu))) - ds_suspend_bts(per_cpu(hwb_tracer, cpu)); - /* - * We need to collect the trace on the respective cpu since ftrace - * implicitly adds the record for the current cpu. - * Once that is more flexible, we could collect the data from any cpu. - */ - on_each_cpu(trace_bts_cpu, iter->tr, 1); - - for_each_online_cpu(cpu) - if (likely(per_cpu(hwb_tracer, cpu))) - ds_resume_bts(per_cpu(hwb_tracer, cpu)); - put_online_cpus(); -} - -static void trace_bts_close(struct trace_iterator *iter) -{ - tracing_reset_online_cpus(iter->tr); -} - -void trace_hw_branch_oops(void) -{ - if (this_tracer) { - ds_suspend_bts_noirq(this_tracer); - trace_bts_cpu(hw_branch_trace); - ds_resume_bts_noirq(this_tracer); - } -} - -struct tracer bts_tracer __read_mostly = -{ - .name = "hw-branch-tracer", - .init = bts_trace_init, - .reset = bts_trace_reset, - .print_header = bts_trace_print_header, - .print_line = bts_trace_print_line, - .start = bts_trace_start, - .stop = bts_trace_stop, - .open = trace_bts_prepare, - .close = trace_bts_close, -#ifdef CONFIG_FTRACE_SELFTEST - .selftest = trace_selftest_startup_hw_branches, -#endif /* CONFIG_FTRACE_SELFTEST */ -}; - -__init static int init_bts_trace(void) -{ - register_hotcpu_notifier(&bts_hotcpu_notifier); - return register_tracer(&bts_tracer); -} -device_initcall(init_bts_trace); diff --git a/kernel/trace/trace_selftest.c b/kernel/trace/trace_selftest.c index 280fea4..a7084e7 100644 --- a/kernel/trace/trace_selftest.c +++ b/kernel/trace/trace_selftest.c @@ -16,7 +16,6 @@ static inline int trace_valid_entry(struct trace_entry *entry) case TRACE_BRANCH: case TRACE_GRAPH_ENT: case TRACE_GRAPH_RET: - case TRACE_HW_BRANCHES: case TRACE_KSYM: return 1; } @@ -754,62 +753,6 @@ trace_selftest_startup_branch(struct tracer *trace, struct trace_array *tr) } #endif /* CONFIG_BRANCH_TRACER */ -#ifdef CONFIG_HW_BRANCH_TRACER -int -trace_selftest_startup_hw_branches(struct tracer *trace, - struct trace_array *tr) -{ - struct trace_iterator *iter; - struct tracer tracer; - unsigned long count; - int ret; - - if (!trace->open) { - printk(KERN_CONT "missing open function..."); - return -1; - } - - ret = tracer_init(trace, tr); - if (ret) { - warn_failed_init_tracer(trace, ret); - return ret; - } - - /* - * The hw-branch tracer needs to collect the trace from the various - * cpu trace buffers - before tracing is stopped. - */ - iter = kzalloc(sizeof(*iter), GFP_KERNEL); - if (!iter) - return -ENOMEM; - - memcpy(&tracer, trace, sizeof(tracer)); - - iter->trace = &tracer; - iter->tr = tr; - iter->pos = -1; - mutex_init(&iter->mutex); - - trace->open(iter); - - mutex_destroy(&iter->mutex); - kfree(iter); - - tracing_stop(); - - ret = trace_test_buffer(tr, &count); - trace->reset(tr); - tracing_start(); - - if (!ret && !count) { - printk(KERN_CONT "no entries found.."); - ret = -1; - } - - return ret; -} -#endif /* CONFIG_HW_BRANCH_TRACER */ - #ifdef CONFIG_KSYM_TRACER static int ksym_selftest_dummy; diff --git a/mm/mlock.c b/mm/mlock.c index 8f4e2df..3f82720 100644 --- a/mm/mlock.c +++ b/mm/mlock.c @@ -607,44 +607,3 @@ void user_shm_unlock(size_t size, struct user_struct *user) spin_unlock(&shmlock_user_lock); free_uid(user); } - -int account_locked_memory(struct mm_struct *mm, struct rlimit *rlim, - size_t size) -{ - unsigned long lim, vm, pgsz; - int error = -ENOMEM; - - pgsz = PAGE_ALIGN(size) >> PAGE_SHIFT; - - down_write(&mm->mmap_sem); - - lim = ACCESS_ONCE(rlim[RLIMIT_AS].rlim_cur) >> PAGE_SHIFT; - vm = mm->total_vm + pgsz; - if (lim < vm) - goto out; - - lim = ACCESS_ONCE(rlim[RLIMIT_MEMLOCK].rlim_cur) >> PAGE_SHIFT; - vm = mm->locked_vm + pgsz; - if (lim < vm) - goto out; - - mm->total_vm += pgsz; - mm->locked_vm += pgsz; - - error = 0; - out: - up_write(&mm->mmap_sem); - return error; -} - -void refund_locked_memory(struct mm_struct *mm, size_t size) -{ - unsigned long pgsz = PAGE_ALIGN(size) >> PAGE_SHIFT; - - down_write(&mm->mmap_sem); - - mm->total_vm -= pgsz; - mm->locked_vm -= pgsz; - - up_write(&mm->mmap_sem); -} -- cgit v0.10.2 From ea8e61b7bbc4a2faef77db34eb2db2a2c2372ff6 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 25 Mar 2010 14:51:51 +0100 Subject: x86, ptrace: Fix block-step Implement ptrace-block-step using TIF_BLOCKSTEP which will set DEBUGCTLMSR_BTF when set for a task while preserving any other DEBUGCTLMSR bits. Signed-off-by: Peter Zijlstra LKML-Reference: <20100325135414.017536066@chello.nl> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/processor.h b/arch/x86/include/asm/processor.h index 5bec21a..32428b4 100644 --- a/arch/x86/include/asm/processor.h +++ b/arch/x86/include/asm/processor.h @@ -799,7 +799,7 @@ extern void cpu_init(void); static inline unsigned long get_debugctlmsr(void) { - unsigned long debugctlmsr = 0; + unsigned long debugctlmsr = 0; #ifndef CONFIG_X86_DEBUGCTLMSR if (boot_cpu_data.x86 < 6) @@ -807,7 +807,7 @@ static inline unsigned long get_debugctlmsr(void) #endif rdmsrl(MSR_IA32_DEBUGCTLMSR, debugctlmsr); - return debugctlmsr; + return debugctlmsr; } static inline void update_debugctlmsr(unsigned long debugctlmsr) diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h index dc85e12..d017ed55 100644 --- a/arch/x86/include/asm/thread_info.h +++ b/arch/x86/include/asm/thread_info.h @@ -92,6 +92,7 @@ struct thread_info { #define TIF_IO_BITMAP 22 /* uses I/O bitmap */ #define TIF_FREEZE 23 /* is freezing for suspend */ #define TIF_FORCED_TF 24 /* true if TF in eflags artificially */ +#define TIF_BLOCKSTEP 25 /* set when we want DEBUGCTLMSR_BTF */ #define TIF_LAZY_MMU_UPDATES 27 /* task is updating the mmu lazily */ #define TIF_SYSCALL_TRACEPOINT 28 /* syscall tracepoint instrumentation */ @@ -113,6 +114,7 @@ struct thread_info { #define _TIF_IO_BITMAP (1 << TIF_IO_BITMAP) #define _TIF_FREEZE (1 << TIF_FREEZE) #define _TIF_FORCED_TF (1 << TIF_FORCED_TF) +#define _TIF_BLOCKSTEP (1 << TIF_BLOCKSTEP) #define _TIF_LAZY_MMU_UPDATES (1 << TIF_LAZY_MMU_UPDATES) #define _TIF_SYSCALL_TRACEPOINT (1 << TIF_SYSCALL_TRACEPOINT) @@ -143,7 +145,7 @@ struct thread_info { /* flags to check in __switch_to() */ #define _TIF_WORK_CTXSW \ - (_TIF_IO_BITMAP|_TIF_NOTSC) + (_TIF_IO_BITMAP|_TIF_NOTSC|_TIF_BLOCKSTEP) #define _TIF_WORK_CTXSW_PREV (_TIF_WORK_CTXSW|_TIF_USER_RETURN_NOTIFY) #define _TIF_WORK_CTXSW_NEXT (_TIF_WORK_CTXSW|_TIF_DEBUG) diff --git a/arch/x86/kernel/kprobes.c b/arch/x86/kernel/kprobes.c index 7a880ad..f2f56c0 100644 --- a/arch/x86/kernel/kprobes.c +++ b/arch/x86/kernel/kprobes.c @@ -422,12 +422,22 @@ static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs, static void __kprobes clear_btf(void) { - /* XXX */ + if (test_thread_flag(TIF_BLOCKSTEP)) { + unsigned long debugctl = get_debugctlmsr(); + + debugctl &= ~DEBUGCTLMSR_BTF; + update_debugctlmsr(debugctl); + } } static void __kprobes restore_btf(void) { - /* XXX */ + if (test_thread_flag(TIF_BLOCKSTEP)) { + unsigned long debugctl = get_debugctlmsr(); + + debugctl |= DEBUGCTLMSR_BTF; + update_debugctlmsr(debugctl); + } } void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri, diff --git a/arch/x86/kernel/process.c b/arch/x86/kernel/process.c index 1a60beb..8328009 100644 --- a/arch/x86/kernel/process.c +++ b/arch/x86/kernel/process.c @@ -195,6 +195,17 @@ void __switch_to_xtra(struct task_struct *prev_p, struct task_struct *next_p, prev = &prev_p->thread; next = &next_p->thread; + if (test_tsk_thread_flag(prev_p, TIF_BLOCKSTEP) ^ + test_tsk_thread_flag(next_p, TIF_BLOCKSTEP)) { + unsigned long debugctl = get_debugctlmsr(); + + debugctl &= ~DEBUGCTLMSR_BTF; + if (test_tsk_thread_flag(next_p, TIF_BLOCKSTEP)) + debugctl |= DEBUGCTLMSR_BTF; + + update_debugctlmsr(debugctl); + } + if (test_tsk_thread_flag(prev_p, TIF_NOTSC) ^ test_tsk_thread_flag(next_p, TIF_NOTSC)) { /* prev and next are different */ diff --git a/arch/x86/kernel/step.c b/arch/x86/kernel/step.c index 7beba07..58de45e 100644 --- a/arch/x86/kernel/step.c +++ b/arch/x86/kernel/step.c @@ -169,9 +169,19 @@ static void enable_step(struct task_struct *child, bool block) * So noone should try to use debugger block stepping in a program * that uses user-mode single stepping itself. */ - if (!enable_single_step(child)) - return; - /* XXX */ + if (enable_single_step(child) && block) { + unsigned long debugctl = get_debugctlmsr(); + + debugctl |= DEBUGCTLMSR_BTF; + update_debugctlmsr(debugctl); + set_tsk_thread_flag(child, TIF_BLOCKSTEP); + } else if (test_tsk_thread_flag(child, TIF_BLOCKSTEP)) { + unsigned long debugctl = get_debugctlmsr(); + + debugctl &= ~DEBUGCTLMSR_BTF; + update_debugctlmsr(debugctl); + clear_tsk_thread_flag(child, TIF_BLOCKSTEP); + } } void user_enable_single_step(struct task_struct *child) @@ -189,7 +199,13 @@ void user_disable_single_step(struct task_struct *child) /* * Make sure block stepping (BTF) is disabled. */ - /* XXX */ + if (test_tsk_thread_flag(child, TIF_BLOCKSTEP)) { + unsigned long debugctl = get_debugctlmsr(); + + debugctl &= ~DEBUGCTLMSR_BTF; + update_debugctlmsr(debugctl); + clear_tsk_thread_flag(child, TIF_BLOCKSTEP); + } /* Always clear TIF_SINGLESTEP... */ clear_tsk_thread_flag(child, TIF_SINGLESTEP); diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index e3da5d7..36f1bd9 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -544,6 +544,11 @@ dotraplinkage void __kprobes do_debug(struct pt_regs *regs, long error_code) /* DR6 may or may not be cleared by the CPU */ set_debugreg(0, 6); + /* + * The processor cleared BTF, so don't mark that we need it set. + */ + clear_tsk_thread_flag(tsk, TIF_BLOCKSTEP); + /* Store the virtualized DR6 value */ tsk->thread.debugreg6 = dr6; -- cgit v0.10.2 From 11164cd4f6dab326a88bdf27f2f8f7c11977e91a Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 26 Mar 2010 14:08:44 +0100 Subject: perf, x86: Add Nehelem PMU programming errata workaround Implement the workaround for Intel Errata AAK100 and AAP53. Also, remove the Core-i7 name for Nehalem events since there are also Westmere based i7 chips. Signed-off-by: Peter Zijlstra Cc: Stephane Eranian LKML-Reference: <1269608924.12097.147.camel@laptop> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index f571f51..6f66d4a 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -184,7 +184,7 @@ struct x86_pmu { int version; int (*handle_irq)(struct pt_regs *); void (*disable_all)(void); - void (*enable_all)(void); + void (*enable_all)(int added); void (*enable)(struct perf_event *); void (*disable)(struct perf_event *); int (*hw_config)(struct perf_event_attr *attr, struct hw_perf_event *hwc); @@ -576,7 +576,7 @@ void hw_perf_disable(void) x86_pmu.disable_all(); } -static void x86_pmu_enable_all(void) +static void x86_pmu_enable_all(int added) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); int idx; @@ -784,7 +784,7 @@ void hw_perf_enable(void) struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); struct perf_event *event; struct hw_perf_event *hwc; - int i; + int i, added = cpuc->n_added; if (!x86_pmu_initialized()) return; @@ -836,7 +836,7 @@ void hw_perf_enable(void) cpuc->enabled = 1; barrier(); - x86_pmu.enable_all(); + x86_pmu.enable_all(added); } static inline void __x86_pmu_enable_event(struct hw_perf_event *hwc) diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 044b843..676aac2 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -483,7 +483,7 @@ static void intel_pmu_disable_all(void) intel_pmu_lbr_disable_all(); } -static void intel_pmu_enable_all(void) +static void intel_pmu_enable_all(int added) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); @@ -502,6 +502,40 @@ static void intel_pmu_enable_all(void) } } +/* + * Workaround for: + * Intel Errata AAK100 (model 26) + * Intel Errata AAP53 (model 30) + * + * These chips need to be 'reset' when adding counters by programming + * the magic three (non counting) events 0x4300D2, 0x4300B1 and 0x4300B5 + * either in sequence on the same PMC or on different PMCs. + */ +static void intel_pmu_nhm_enable_all(int added) +{ + if (added) { + struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); + int i; + + wrmsrl(MSR_ARCH_PERFMON_EVENTSEL0 + 0, 0x4300D2); + wrmsrl(MSR_ARCH_PERFMON_EVENTSEL0 + 1, 0x4300B1); + wrmsrl(MSR_ARCH_PERFMON_EVENTSEL0 + 2, 0x4300B5); + + wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0x3); + wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0x0); + + for (i = 0; i < 3; i++) { + struct perf_event *event = cpuc->events[i]; + + if (!event) + continue; + + __x86_pmu_enable_event(&event->hw); + } + } + intel_pmu_enable_all(added); +} + static inline u64 intel_pmu_get_status(void) { u64 status; @@ -658,7 +692,7 @@ static int intel_pmu_handle_irq(struct pt_regs *regs) intel_pmu_drain_bts_buffer(); status = intel_pmu_get_status(); if (!status) { - intel_pmu_enable_all(); + intel_pmu_enable_all(0); return 0; } @@ -707,7 +741,7 @@ again: goto again; done: - intel_pmu_enable_all(); + intel_pmu_enable_all(0); return 1; } @@ -920,7 +954,8 @@ static __init int intel_pmu_init(void) intel_pmu_lbr_init_nhm(); x86_pmu.event_constraints = intel_nehalem_event_constraints; - pr_cont("Nehalem/Corei7 events, "); + x86_pmu.enable_all = intel_pmu_nhm_enable_all; + pr_cont("Nehalem events, "); break; case 28: /* Atom */ diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index f8fe069..0d1be36 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -535,7 +535,7 @@ static void p4_pmu_enable_event(struct perf_event *event) (cccr & ~P4_CCCR_RESERVED) | P4_CCCR_ENABLE); } -static void p4_pmu_enable_all(void) +static void p4_pmu_enable_all(int added) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); int idx; diff --git a/arch/x86/kernel/cpu/perf_event_p6.c b/arch/x86/kernel/cpu/perf_event_p6.c index 6ff4d01..877182c 100644 --- a/arch/x86/kernel/cpu/perf_event_p6.c +++ b/arch/x86/kernel/cpu/perf_event_p6.c @@ -66,7 +66,7 @@ static void p6_pmu_disable_all(void) wrmsrl(MSR_P6_EVNTSEL0, val); } -static void p6_pmu_enable_all(void) +static void p6_pmu_enable_all(int added) { unsigned long val; -- cgit v0.10.2 From 1fb2f77c037624601fd214fb7c29faa84cd7bdd7 Mon Sep 17 00:00:00 2001 From: Henrik Kretzschmar Date: Fri, 26 Mar 2010 20:38:35 +0100 Subject: debugobjects: Section mismatch cleanup This patch marks two functions, which only get called at initialization, as __init. Here is also interesting, that modpost doesn't catch here the right function name. WARNING: lib/built-in.o(.text+0x585f): Section mismatch in reference from the function T.506() to the variable .init.data:obj The function T.506() references the variable __initdata obj. This is often because T.506 lacks a __initdata annotation or the annotation of obj is wrong. Signed-off-by: Henrik Kretzschmar LKML-Reference: <1269632315-19403-1-git-send-email-henne@nachtwindheim.de> Signed-off-by: Thomas Gleixner diff --git a/lib/debugobjects.c b/lib/debugobjects.c index a9a8996..c4ecd3c 100644 --- a/lib/debugobjects.c +++ b/lib/debugobjects.c @@ -773,7 +773,7 @@ static int __init fixup_free(void *addr, enum debug_obj_state state) } } -static int +static int __init check_results(void *addr, enum debug_obj_state state, int fixups, int warnings) { struct debug_bucket *db; @@ -916,7 +916,7 @@ void __init debug_objects_early_init(void) /* * Convert the statically allocated objects to dynamic ones: */ -static int debug_objects_replace_static_objects(void) +static int __init debug_objects_replace_static_objects(void) { struct debug_bucket *db = obj_hash; struct hlist_node *node, *tmp; -- cgit v0.10.2 From 810627a013163cd294762d57c0ea2ec055ffe4f6 Mon Sep 17 00:00:00 2001 From: Pavel Shilovsky Date: Sat, 27 Mar 2010 02:00:49 +0000 Subject: [CIFS] Add mmap for direct, nobrl cifs mount types without mmap functions in file_ops OpenOffice can't save changes in existing document. The same situation you can see with gedit. Also, a.out format of files can't be executed without mmap. Signed-off-by: Pavel Shilovsky Reviewed-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 5183bc2..ded66be 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -808,6 +808,7 @@ const struct file_operations cifs_file_direct_nobrl_ops = { .release = cifs_close, .fsync = cifs_fsync, .flush = cifs_flush, + .mmap = cifs_file_mmap, .splice_read = generic_file_splice_read, #ifdef CONFIG_CIFS_POSIX .unlocked_ioctl = cifs_ioctl, -- cgit v0.10.2 From 6ae6650232ddcf2f50e8817acd63cde37cf1d093 Mon Sep 17 00:00:00 2001 From: Matt Fleming Date: Mon, 29 Mar 2010 15:24:54 +0900 Subject: sh: tlb debugfs support. Export the status of the utlb and itlb entries through debugfs. Signed-off-by: Matt Fleming Signed-off-by: Paul Mundt diff --git a/arch/sh/include/cpu-sh4/cpu/mmu_context.h b/arch/sh/include/cpu-sh4/cpu/mmu_context.h index 5963124..2941be6 100644 --- a/arch/sh/include/cpu-sh4/cpu/mmu_context.h +++ b/arch/sh/include/cpu-sh4/cpu/mmu_context.h @@ -19,10 +19,17 @@ #define MMUCR 0xFF000010 /* MMU Control Register */ +#define MMU_TLB_ENTRY_SHIFT 8 + #define MMU_ITLB_ADDRESS_ARRAY 0xF2000000 #define MMU_ITLB_ADDRESS_ARRAY2 0xF2800000 +#define MMU_ITLB_DATA_ARRAY 0xF3000000 +#define MMU_ITLB_DATA_ARRAY2 0xF3800000 + #define MMU_UTLB_ADDRESS_ARRAY 0xF6000000 #define MMU_UTLB_ADDRESS_ARRAY2 0xF6800000 +#define MMU_UTLB_DATA_ARRAY 0xF7000000 +#define MMU_UTLB_DATA_ARRAY2 0xF7800000 #define MMU_PAGE_ASSOC_BIT 0x80 #define MMUCR_TI (1<<2) diff --git a/arch/sh/mm/Makefile b/arch/sh/mm/Makefile index 3dc8a8a..c73018a 100644 --- a/arch/sh/mm/Makefile +++ b/arch/sh/mm/Makefile @@ -18,13 +18,14 @@ mmu-$(CONFIG_MMU) := extable_$(BITS).o fault_$(BITS).o \ ioremap.o kmap.o pgtable.o tlbflush_$(BITS).o obj-y += $(mmu-y) -obj-$(CONFIG_DEBUG_FS) += asids-debugfs.o -ifdef CONFIG_DEBUG_FS -obj-$(CONFIG_CPU_SH4) += cache-debugfs.o +debugfs-y := asids-debugfs.o +ifndef CONFIG_CACHE_OFF +debugfs-$(CONFIG_CPU_SH4) += cache-debugfs.o endif ifdef CONFIG_MMU +debugfs-$(CONFIG_CPU_SH4) += tlb-debugfs.o tlb-$(CONFIG_CPU_SH3) := tlb-sh3.o tlb-$(CONFIG_CPU_SH4) := tlb-sh4.o tlb-urb.o tlb-$(CONFIG_CPU_SH5) := tlb-sh5.o @@ -32,6 +33,7 @@ tlb-$(CONFIG_CPU_HAS_PTEAEX) := tlb-pteaex.o tlb-urb.o obj-y += $(tlb-y) endif +obj-$(CONFIG_DEBUG_FS) += $(debugfs-y) obj-$(CONFIG_HUGETLB_PAGE) += hugetlbpage.o obj-$(CONFIG_PMB) += pmb.o obj-$(CONFIG_NUMA) += numa.o diff --git a/arch/sh/mm/tlb-debugfs.c b/arch/sh/mm/tlb-debugfs.c new file mode 100644 index 0000000..229bf75 --- /dev/null +++ b/arch/sh/mm/tlb-debugfs.c @@ -0,0 +1,179 @@ +/* + * arch/sh/mm/tlb-debugfs.c + * + * debugfs ops for SH-4 ITLB/UTLBs. + * + * Copyright (C) 2010 Matt Fleming + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + */ +#include +#include +#include +#include +#include +#include +#include + +enum tlb_type { + TLB_TYPE_ITLB, + TLB_TYPE_UTLB, +}; + +static struct { + int bits; + const char *size; +} tlb_sizes[] = { + { 0x0, " 1KB" }, + { 0x1, " 4KB" }, + { 0x2, " 8KB" }, + { 0x4, " 64KB" }, + { 0x5, "256KB" }, + { 0x7, " 1MB" }, + { 0x8, " 4MB" }, + { 0xc, " 64MB" }, +}; + +static int tlb_seq_show(struct seq_file *file, void *iter) +{ + unsigned int tlb_type = (unsigned int)file->private; + unsigned long addr1, addr2, data1, data2; + unsigned long flags; + unsigned long mmucr; + unsigned int nentries, entry; + unsigned int urb; + + mmucr = __raw_readl(MMUCR); + if ((mmucr & 0x1) == 0) { + seq_printf(file, "address translation disabled\n"); + return 0; + } + + if (tlb_type == TLB_TYPE_ITLB) { + addr1 = MMU_ITLB_ADDRESS_ARRAY; + addr2 = MMU_ITLB_ADDRESS_ARRAY2; + data1 = MMU_ITLB_DATA_ARRAY; + data2 = MMU_ITLB_DATA_ARRAY2; + nentries = 4; + } else { + addr1 = MMU_UTLB_ADDRESS_ARRAY; + addr2 = MMU_UTLB_ADDRESS_ARRAY2; + data1 = MMU_UTLB_DATA_ARRAY; + data2 = MMU_UTLB_DATA_ARRAY2; + nentries = 64; + } + + local_irq_save(flags); + jump_to_uncached(); + + urb = (mmucr & MMUCR_URB) >> MMUCR_URB_SHIFT; + + /* Make the "entry >= urb" test fail. */ + if (urb == 0) + urb = MMUCR_URB_NENTRIES + 1; + + if (tlb_type == TLB_TYPE_ITLB) { + addr1 = MMU_ITLB_ADDRESS_ARRAY; + addr2 = MMU_ITLB_ADDRESS_ARRAY2; + data1 = MMU_ITLB_DATA_ARRAY; + data2 = MMU_ITLB_DATA_ARRAY2; + nentries = 4; + } else { + addr1 = MMU_UTLB_ADDRESS_ARRAY; + addr2 = MMU_UTLB_ADDRESS_ARRAY2; + data1 = MMU_UTLB_DATA_ARRAY; + data2 = MMU_UTLB_DATA_ARRAY2; + nentries = 64; + } + + seq_printf(file, "entry: vpn ppn asid size valid wired\n"); + + for (entry = 0; entry < nentries; entry++) { + unsigned long vpn, ppn, asid, size; + unsigned long valid; + unsigned long val; + const char *sz = " ?"; + int i; + + val = __raw_readl(addr1 | (entry << MMU_TLB_ENTRY_SHIFT)); + ctrl_barrier(); + vpn = val & 0xfffffc00; + valid = val & 0x100; + + val = __raw_readl(addr2 | (entry << MMU_TLB_ENTRY_SHIFT)); + ctrl_barrier(); + asid = val & MMU_CONTEXT_ASID_MASK; + + val = __raw_readl(data1 | (entry << MMU_TLB_ENTRY_SHIFT)); + ctrl_barrier(); + ppn = (val & 0x0ffffc00) << 4; + + val = __raw_readl(data2 | (entry << MMU_TLB_ENTRY_SHIFT)); + ctrl_barrier(); + size = (val & 0xf0) >> 4; + + for (i = 0; i < ARRAY_SIZE(tlb_sizes); i++) { + if (tlb_sizes[i].bits == size) + break; + } + + if (i != ARRAY_SIZE(tlb_sizes)) + sz = tlb_sizes[i].size; + + seq_printf(file, "%2d: 0x%08lx 0x%08lx %5lu %s %s %s\n", + entry, vpn, ppn, asid, + sz, valid ? "V" : "-", + (urb <= entry) ? "W" : "-"); + } + + back_to_cached(); + local_irq_restore(flags); + + return 0; +} + +static int tlb_debugfs_open(struct inode *inode, struct file *file) +{ + return single_open(file, tlb_seq_show, inode->i_private); +} + +static const struct file_operations tlb_debugfs_fops = { + .owner = THIS_MODULE, + .open = tlb_debugfs_open, + .read = seq_read, + .llseek = seq_lseek, + .release = single_release, +}; + +static int __init tlb_debugfs_init(void) +{ + struct dentry *itlb, *utlb; + + itlb = debugfs_create_file("itlb", S_IRUSR, sh_debugfs_root, + (unsigned int *)TLB_TYPE_ITLB, + &tlb_debugfs_fops); + if (unlikely(!itlb)) + return -ENOMEM; + if (IS_ERR(itlb)) + return PTR_ERR(itlb); + + utlb = debugfs_create_file("utlb", S_IRUSR, sh_debugfs_root, + (unsigned int *)TLB_TYPE_UTLB, + &tlb_debugfs_fops); + if (unlikely(!utlb)) { + debugfs_remove(itlb); + return -ENOMEM; + } + + if (IS_ERR(utlb)) { + debugfs_remove(itlb); + return PTR_ERR(utlb); + } + + return 0; +} +module_init(tlb_debugfs_init); + +MODULE_LICENSE("GPL v2"); -- cgit v0.10.2 From bd05086bbe3f241cd552068f9ceba9e19c6ce427 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 29 Mar 2010 15:51:35 +0900 Subject: driver core: Convert to kasprintf() for early dev_name(). This is just a simple refactoring patch on top of the early dev_name() support, converting from kstrdup() to kasprintf() as suggested by Kay. Signed-off-by: Paul Mundt diff --git a/drivers/base/platform.c b/drivers/base/platform.c index d2d4926..f2377f3 100644 --- a/drivers/base/platform.c +++ b/drivers/base/platform.c @@ -1245,19 +1245,20 @@ static int __init early_platform_driver_probe_id(char *class_str, * rest of the driver core is initialized. */ if (!match->dev.init_name) { - char buf[32]; - if (match->id != -1) - snprintf(buf, sizeof(buf), "%s.%d", - match->name, match->id); + match->dev.init_name = + kasprintf(GFP_KERNEL, "%s.%d", + match->name, + match->id); else - snprintf(buf, sizeof(buf), "%s", - match->name); + match->dev.init_name = + kasprintf(GFP_KERNEL, "%s", + match->name); - match->dev.init_name = kstrdup(buf, GFP_KERNEL); if (!match->dev.init_name) return -ENOMEM; } + if (epdrv->pdrv->probe(match)) pr_warning("%s: unable to probe %s early.\n", class_str, match->name); -- cgit v0.10.2 From 5dbd5ec6e1cf2e49128025d80813a275744a7ac5 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 29 Mar 2010 09:16:24 +0200 Subject: ALSA: hda - Fix invalid bit values passed to snd_hda_codec_amp_stereo() The mask and value parameters passed to snd_hda_codec_amp_stereo() should be 8-bit values for mute and volume. Passing AMP_IN_MUTE() is wrong, which is found in many places in patch_realtek.c as a left-over from the conversion to snd_hda_codec_amp_stereo(). Reported-by: Dan Carpenter Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 9a23444..bc55c1e 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -12459,11 +12459,11 @@ static void alc268_aspire_one_speaker_automute(struct hda_codec *codec) unsigned char bits; present = snd_hda_jack_detect(codec, 0x15); - bits = present ? AMP_IN_MUTE(0) : 0; + bits = present ? HDA_AMP_MUTE : 0; snd_hda_codec_amp_stereo(codec, 0x0f, HDA_INPUT, 0, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_amp_stereo(codec, 0x0f, HDA_INPUT, 1, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); } static void alc268_acer_lc_unsol_event(struct hda_codec *codec, @@ -13482,11 +13482,11 @@ static void alc269_quanta_fl1_speaker_automute(struct hda_codec *codec) unsigned char bits; present = snd_hda_jack_detect(codec, 0x15); - bits = present ? AMP_IN_MUTE(0) : 0; + bits = present ? HDA_AMP_MUTE : 0; snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 0, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 1, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_write(codec, 0x20, 0, AC_VERB_SET_COEF_INDEX, 0x0c); @@ -13511,11 +13511,11 @@ static void alc269_lifebook_speaker_automute(struct hda_codec *codec) /* Check port replicator headphone socket */ present |= snd_hda_jack_detect(codec, 0x1a); - bits = present ? AMP_IN_MUTE(0) : 0; + bits = present ? HDA_AMP_MUTE : 0; snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 0, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 1, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_write(codec, 0x20, 0, AC_VERB_SET_COEF_INDEX, 0x0c); @@ -13646,11 +13646,11 @@ static void alc269_speaker_automute(struct hda_codec *codec) unsigned char bits; present = snd_hda_jack_detect(codec, nid); - bits = present ? AMP_IN_MUTE(0) : 0; + bits = present ? HDA_AMP_MUTE : 0; snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 0, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 1, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); } /* unsolicited event for HP jack sensing */ @@ -17115,9 +17115,9 @@ static void alc663_m51va_speaker_automute(struct hda_codec *codec) present = snd_hda_jack_detect(codec, 0x21); bits = present ? HDA_AMP_MUTE : 0; snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 0, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 1, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); } static void alc663_21jd_two_speaker_automute(struct hda_codec *codec) @@ -17128,13 +17128,13 @@ static void alc663_21jd_two_speaker_automute(struct hda_codec *codec) present = snd_hda_jack_detect(codec, 0x21); bits = present ? HDA_AMP_MUTE : 0; snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 0, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 1, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_amp_stereo(codec, 0x0e, HDA_INPUT, 0, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_amp_stereo(codec, 0x0e, HDA_INPUT, 1, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); } static void alc663_15jd_two_speaker_automute(struct hda_codec *codec) @@ -17145,13 +17145,13 @@ static void alc663_15jd_two_speaker_automute(struct hda_codec *codec) present = snd_hda_jack_detect(codec, 0x15); bits = present ? HDA_AMP_MUTE : 0; snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 0, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 1, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_amp_stereo(codec, 0x0e, HDA_INPUT, 0, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); snd_hda_codec_amp_stereo(codec, 0x0e, HDA_INPUT, 1, - AMP_IN_MUTE(0), bits); + HDA_AMP_MUTE, bits); } static void alc662_f5z_speaker_automute(struct hda_codec *codec) @@ -17190,14 +17190,14 @@ static void alc663_two_hp_m2_speaker_automute(struct hda_codec *codec) if (present1 || present2) { snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 0, - AMP_IN_MUTE(0), AMP_IN_MUTE(0)); + HDA_AMP_MUTE, HDA_AMP_MUTE); snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 1, - AMP_IN_MUTE(0), AMP_IN_MUTE(0)); + HDA_AMP_MUTE, HDA_AMP_MUTE); } else { snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 0, - AMP_IN_MUTE(0), 0); + HDA_AMP_MUTE, 0); snd_hda_codec_amp_stereo(codec, 0x0c, HDA_INPUT, 1, - AMP_IN_MUTE(0), 0); + HDA_AMP_MUTE, 0); } } -- cgit v0.10.2 From c2a25e819717ea34e8f682d67a3fa76f4893395b Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 29 Mar 2010 16:55:43 +0900 Subject: clocksource: Deprecate clock string across the SH drivers. We want to get rid of the clock string from platform data entirely, depending on the clkdev-based clock lookup to do the right thing for us instead. This converts all of the SH drivers to request their associated function clocks directly, and if there is no match for that then we fall back on the legacy lookup while warning about it. After all of the outstanding CPUs have been converted to clkdev lookups the clock string will be killed off completely. Signed-off-by: Paul Mundt diff --git a/drivers/clocksource/sh_cmt.c b/drivers/clocksource/sh_cmt.c index a1d3811..c5f6617 100644 --- a/drivers/clocksource/sh_cmt.c +++ b/drivers/clocksource/sh_cmt.c @@ -608,11 +608,15 @@ static int sh_cmt_setup(struct sh_cmt_priv *p, struct platform_device *pdev) p->irqaction.flags = IRQF_DISABLED | IRQF_TIMER | IRQF_IRQPOLL; /* get hold of clock */ - p->clk = clk_get(&p->pdev->dev, cfg->clk); + p->clk = clk_get(&p->pdev->dev, "cmt_fck"); if (IS_ERR(p->clk)) { - dev_err(&p->pdev->dev, "cannot get clock\n"); - ret = PTR_ERR(p->clk); - goto err1; + dev_warn(&p->pdev->dev, "using deprecated clock lookup\n"); + p->clk = clk_get(&p->pdev->dev, cfg->clk); + if (IS_ERR(p->clk)) { + dev_err(&p->pdev->dev, "cannot get clock\n"); + ret = PTR_ERR(p->clk); + goto err1; + } } if (resource_size(res) == 6) { diff --git a/drivers/clocksource/sh_mtu2.c b/drivers/clocksource/sh_mtu2.c index 8ce3555..b11882e 100644 --- a/drivers/clocksource/sh_mtu2.c +++ b/drivers/clocksource/sh_mtu2.c @@ -283,11 +283,15 @@ static int sh_mtu2_setup(struct sh_mtu2_priv *p, struct platform_device *pdev) p->irqaction.flags = IRQF_DISABLED | IRQF_TIMER | IRQF_IRQPOLL; /* get hold of clock */ - p->clk = clk_get(&p->pdev->dev, cfg->clk); + p->clk = clk_get(&p->pdev->dev, "mtu2_fck"); if (IS_ERR(p->clk)) { - dev_err(&p->pdev->dev, "cannot get clock\n"); - ret = PTR_ERR(p->clk); - goto err1; + dev_warn(&p->pdev->dev, "using deprecated clock lookup\n"); + p->clk = clk_get(&p->pdev->dev, cfg->clk); + if (IS_ERR(p->clk)) { + dev_err(&p->pdev->dev, "cannot get clock\n"); + ret = PTR_ERR(p->clk); + goto err1; + } } return sh_mtu2_register(p, (char *)dev_name(&p->pdev->dev), diff --git a/drivers/clocksource/sh_tmu.c b/drivers/clocksource/sh_tmu.c index 3423945..6b62283 100644 --- a/drivers/clocksource/sh_tmu.c +++ b/drivers/clocksource/sh_tmu.c @@ -387,11 +387,15 @@ static int sh_tmu_setup(struct sh_tmu_priv *p, struct platform_device *pdev) p->irqaction.flags = IRQF_DISABLED | IRQF_TIMER | IRQF_IRQPOLL; /* get hold of clock */ - p->clk = clk_get(&p->pdev->dev, cfg->clk); + p->clk = clk_get(&p->pdev->dev, "tmu_fck"); if (IS_ERR(p->clk)) { - dev_err(&p->pdev->dev, "cannot get clock\n"); - ret = PTR_ERR(p->clk); - goto err1; + dev_warn(&p->pdev->dev, "using deprecated clock lookup\n"); + p->clk = clk_get(&p->pdev->dev, cfg->clk); + if (IS_ERR(p->clk)) { + dev_err(&p->pdev->dev, "cannot get clock\n"); + ret = PTR_ERR(p->clk); + goto err1; + } } return sh_tmu_register(p, (char *)dev_name(&p->pdev->dev), -- cgit v0.10.2 From e21d2aa76e1fc91cbf18c4b593d32112d954fab5 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 29 Mar 2010 17:03:15 +0900 Subject: sh: sh7786 clkdev lookups. This plugs in clkdev lookups for the SH7786 clocks and migrates off of clock string usage for the TMU clocks. Previously the TMU clocks were aliased to the peripheral clock, so this gets the driver actually toggling the proper clocks now as well. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7786.c b/arch/sh/kernel/cpu/sh4a/clock-sh7786.c index 85e8d57..c4a84bb 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7786.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7786.c @@ -13,6 +13,8 @@ #include #include #include +#include +#include #include #include @@ -120,12 +122,78 @@ static struct clk mstp_clks[] = { SH_CLK_MSTP32("ether_fck", -1, NULL, MSTPCR1, 2, 0), }; +static struct clk_lookup lookups[] = { + { + /* TMU0 */ + .dev_id = "sh_tmu.0", + .con_id = "tmu_fck", + .clk = &mstp_clks[17], /* tmu012_fck */ + }, { + /* TMU1 */ + .dev_id = "sh_tmu.1", + .con_id = "tmu_fck", + .clk = &mstp_clks[17], + }, { + /* TMU2 */ + .dev_id = "sh_tmu.2", + .con_id = "tmu_fck", + .clk = &mstp_clks[17], + }, { + /* TMU3 */ + .dev_id = "sh_tmu.3", + .con_id = "tmu_fck", + .clk = &mstp_clks[16], /* tmu345_fck */ + }, { + /* TMU4 */ + .dev_id = "sh_tmu.4", + .con_id = "tmu_fck", + .clk = &mstp_clks[16], + }, { + /* TMU5 */ + .dev_id = "sh_tmu.5", + .con_id = "tmu_fck", + .clk = &mstp_clks[16], + }, { + /* TMU6 */ + .dev_id = "sh_tmu.6", + .con_id = "tmu_fck", + .clk = &mstp_clks[15], /* tmu678_fck */ + }, { + /* TMU7 */ + .dev_id = "sh_tmu.7", + .con_id = "tmu_fck", + .clk = &mstp_clks[15], + }, { + /* TMU8 */ + .dev_id = "sh_tmu.8", + .con_id = "tmu_fck", + .clk = &mstp_clks[15], + }, { + /* TMU9 */ + .dev_id = "sh_tmu.9", + .con_id = "tmu_fck", + .clk = &mstp_clks[14], /* tmu9_11_fck */ + }, { + /* TMU10 */ + .dev_id = "sh_tmu.10", + .con_id = "tmu_fck", + .clk = &mstp_clks[14], + }, { + /* TMU11 */ + .dev_id = "sh_tmu.11", + .con_id = "tmu_fck", + .clk = &mstp_clks[14], + } +}; + int __init arch_clk_init(void) { int i, ret = 0; for (i = 0; i < ARRAY_SIZE(clks); i++) ret |= clk_register(clks[i]); + for (i = 0; i < ARRAY_SIZE(lookups); i++) + clkdev_add(&lookups[i]); if (!ret) ret = sh_clk_div4_register(div4_clks, ARRAY_SIZE(div4_clks), diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7786.c b/arch/sh/kernel/cpu/sh4a/setup-sh7786.c index 542ced8..f559990 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7786.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7786.c @@ -119,7 +119,6 @@ static struct platform_device scif5_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -148,7 +147,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -177,7 +175,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { @@ -205,7 +202,6 @@ static struct platform_device tmu2_device = { static struct sh_timer_config tmu3_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", }; static struct resource tmu3_resources[] = { @@ -233,7 +229,6 @@ static struct platform_device tmu3_device = { static struct sh_timer_config tmu4_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", }; static struct resource tmu4_resources[] = { @@ -261,7 +256,6 @@ static struct platform_device tmu4_device = { static struct sh_timer_config tmu5_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu5_resources[] = { @@ -289,7 +283,6 @@ static struct platform_device tmu5_device = { static struct sh_timer_config tmu6_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", }; static struct resource tmu6_resources[] = { @@ -317,7 +310,6 @@ static struct platform_device tmu6_device = { static struct sh_timer_config tmu7_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", }; static struct resource tmu7_resources[] = { @@ -345,7 +337,6 @@ static struct platform_device tmu7_device = { static struct sh_timer_config tmu8_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu8_resources[] = { @@ -373,7 +364,6 @@ static struct platform_device tmu8_device = { static struct sh_timer_config tmu9_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", }; static struct resource tmu9_resources[] = { @@ -401,7 +391,6 @@ static struct platform_device tmu9_device = { static struct sh_timer_config tmu10_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", }; static struct resource tmu10_resources[] = { @@ -429,7 +418,6 @@ static struct platform_device tmu10_device = { static struct sh_timer_config tmu11_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu11_resources[] = { -- cgit v0.10.2 From c55fbdd3f97a468d51cbdb03bb7c9c7287fbda2f Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 29 Mar 2010 17:10:06 +0900 Subject: sh: sh7785 clkdev lookups. Convert to TMU clock lookups for SH7785. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7785.c b/arch/sh/kernel/cpu/sh4a/clock-sh7785.c index 6061c73..28de049 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7785.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7785.c @@ -3,7 +3,7 @@ * * SH7785 support for the clock framework * - * Copyright (C) 2007 - 2009 Paul Mundt + * Copyright (C) 2007 - 2010 Paul Mundt * * This file is subject to the terms and conditions of the GNU General Public * License. See the file "COPYING" in the main directory of this archive @@ -14,6 +14,7 @@ #include #include #include +#include #include #include #include @@ -113,12 +114,48 @@ static struct clk mstp_clks[] = { SH_CLK_MSTP32("gdta_fck", -1, NULL, MSTPCR1, 0, 0), }; +static struct clk_lookup lookups[] = { + { + /* TMU0 */ + .dev_id = "sh_tmu.0", + .con_id = "tmu_fck", + .clk = &mstp_clks[13], /* tmu012_fck */ + }, { + /* TMU1 */ + .dev_id = "sh_tmu.1", + .con_id = "tmu_fck", + .clk = &mstp_clks[13], + }, { + /* TMU2 */ + .dev_id = "sh_tmu.2", + .con_id = "tmu_fck", + .clk = &mstp_clks[13], + }, { + /* TMU3 */ + .dev_id = "sh_tmu.3", + .con_id = "tmu_fck", + .clk = &mstp_clks[12], /* tmu345_fck */ + }, { + /* TMU4 */ + .dev_id = "sh_tmu.4", + .con_id = "tmu_fck", + .clk = &mstp_clks[12], + }, { + /* TMU5 */ + .dev_id = "sh_tmu.5", + .con_id = "tmu_fck", + .clk = &mstp_clks[12], + }, +}; + int __init arch_clk_init(void) { int i, ret = 0; for (i = 0; i < ARRAY_SIZE(clks); i++) ret |= clk_register(clks[i]); + for (i = 0; i < ARRAY_SIZE(lookups); i++) + clkdev_add(&lookups[i]); if (!ret) ret = sh_clk_div4_register(div4_clks, ARRAY_SIZE(div4_clks), diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7785.c b/arch/sh/kernel/cpu/sh4a/setup-sh7785.c index 76afe57..07bb2d4 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7785.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7785.c @@ -113,7 +113,6 @@ static struct platform_device scif5_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "tmu012_fck", .clockevent_rating = 200, }; @@ -142,7 +141,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "tmu012_fck", .clocksource_rating = 200, }; @@ -171,7 +169,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "tmu012_fck", }; static struct resource tmu2_resources[] = { @@ -199,7 +196,6 @@ static struct platform_device tmu2_device = { static struct sh_timer_config tmu3_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "tmu345_fck", }; static struct resource tmu3_resources[] = { @@ -227,7 +223,6 @@ static struct platform_device tmu3_device = { static struct sh_timer_config tmu4_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "tmu345_fck", }; static struct resource tmu4_resources[] = { @@ -255,7 +250,6 @@ static struct platform_device tmu4_device = { static struct sh_timer_config tmu5_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "tmu345_fck", }; static struct resource tmu5_resources[] = { -- cgit v0.10.2 From 0776d602abe61871d27accd8e3bcb225a96f4a9f Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 29 Mar 2010 17:22:50 +0900 Subject: sh: provide some clock lookup aliases for legacy CPG parts. For the CPUs that have not yet been converted off of legacy CPG we provide some pclk aliases in order to make driver migration easier. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/clock-cpg.c b/arch/sh/kernel/cpu/clock-cpg.c index eed5eaf..17a73ad 100644 --- a/arch/sh/kernel/cpu/clock-cpg.c +++ b/arch/sh/kernel/cpu/clock-cpg.c @@ -338,6 +338,11 @@ int __init __deprecated cpg_clk_init(void) ret |= clk_register(clk); } + clk_add_alias("tmu_fck", NULL, "peripheral_clk", NULL); + clk_add_alias("mtu2_fck", NULL, "peripheral_clk", NULL); + clk_add_alias("cmt_fck", NULL, "peripheral_clk", NULL); + clk_add_alias("sci_ick", NULL, "peripheral_clk", NULL); + return ret; } -- cgit v0.10.2 From 4dc50a068ad8fbbd2dd6dd5ac8ac2652589ba046 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 29 Mar 2010 17:25:03 +0900 Subject: sh: Kill off timer clock strings for legacy CPG parts. Now with the lookup aliases in place there is no longer any need to provide the clock string, kill it off for all legacy CPG CPUs. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/sh2/setup-sh7619.c b/arch/sh/kernel/cpu/sh2/setup-sh7619.c index 9b96457..c363851 100644 --- a/arch/sh/kernel/cpu/sh2/setup-sh7619.c +++ b/arch/sh/kernel/cpu/sh2/setup-sh7619.c @@ -130,7 +130,6 @@ static struct platform_device eth_device = { static struct sh_timer_config cmt0_platform_data = { .channel_offset = 0x02, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 125, .clocksource_rating = 0, /* disabled due to code generation issues */ }; @@ -160,7 +159,6 @@ static struct platform_device cmt0_device = { static struct sh_timer_config cmt1_platform_data = { .channel_offset = 0x08, .timer_bit = 1, - .clk = "peripheral_clk", .clockevent_rating = 125, .clocksource_rating = 0, /* disabled due to code generation issues */ }; diff --git a/arch/sh/kernel/cpu/sh2a/setup-mxg.c b/arch/sh/kernel/cpu/sh2a/setup-mxg.c index 5301450..6c96ea0 100644 --- a/arch/sh/kernel/cpu/sh2a/setup-mxg.c +++ b/arch/sh/kernel/cpu/sh2a/setup-mxg.c @@ -117,7 +117,6 @@ static DECLARE_INTC_DESC(intc_desc, "mxg", vectors, groups, static struct sh_timer_config mtu2_0_platform_data = { .channel_offset = -0x80, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -146,7 +145,6 @@ static struct platform_device mtu2_0_device = { static struct sh_timer_config mtu2_1_platform_data = { .channel_offset = -0x100, .timer_bit = 1, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -175,7 +173,6 @@ static struct platform_device mtu2_1_device = { static struct sh_timer_config mtu2_2_platform_data = { .channel_offset = 0x80, .timer_bit = 2, - .clk = "peripheral_clk", .clockevent_rating = 200, }; diff --git a/arch/sh/kernel/cpu/sh2a/setup-sh7201.c b/arch/sh/kernel/cpu/sh2a/setup-sh7201.c index 5b07ab5..d08bf4c 100644 --- a/arch/sh/kernel/cpu/sh2a/setup-sh7201.c +++ b/arch/sh/kernel/cpu/sh2a/setup-sh7201.c @@ -320,7 +320,6 @@ static struct platform_device rtc_device = { static struct sh_timer_config mtu2_0_platform_data = { .channel_offset = -0x80, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -349,7 +348,6 @@ static struct platform_device mtu2_0_device = { static struct sh_timer_config mtu2_1_platform_data = { .channel_offset = -0x100, .timer_bit = 1, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -378,7 +376,6 @@ static struct platform_device mtu2_1_device = { static struct sh_timer_config mtu2_2_platform_data = { .channel_offset = 0x80, .timer_bit = 2, - .clk = "peripheral_clk", .clockevent_rating = 200, }; diff --git a/arch/sh/kernel/cpu/sh2a/setup-sh7203.c b/arch/sh/kernel/cpu/sh2a/setup-sh7203.c index 433ac53..832f401 100644 --- a/arch/sh/kernel/cpu/sh2a/setup-sh7203.c +++ b/arch/sh/kernel/cpu/sh2a/setup-sh7203.c @@ -236,7 +236,6 @@ static struct platform_device scif3_device = { static struct sh_timer_config cmt0_platform_data = { .channel_offset = 0x02, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 125, .clocksource_rating = 0, /* disabled due to code generation issues */ }; @@ -266,7 +265,6 @@ static struct platform_device cmt0_device = { static struct sh_timer_config cmt1_platform_data = { .channel_offset = 0x08, .timer_bit = 1, - .clk = "peripheral_clk", .clockevent_rating = 125, .clocksource_rating = 0, /* disabled due to code generation issues */ }; @@ -296,7 +294,6 @@ static struct platform_device cmt1_device = { static struct sh_timer_config mtu2_0_platform_data = { .channel_offset = -0x80, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -325,7 +322,6 @@ static struct platform_device mtu2_0_device = { static struct sh_timer_config mtu2_1_platform_data = { .channel_offset = -0x100, .timer_bit = 1, - .clk = "peripheral_clk", .clockevent_rating = 200, }; diff --git a/arch/sh/kernel/cpu/sh2a/setup-sh7206.c b/arch/sh/kernel/cpu/sh2a/setup-sh7206.c index 8caecb4..dc47b04 100644 --- a/arch/sh/kernel/cpu/sh2a/setup-sh7206.c +++ b/arch/sh/kernel/cpu/sh2a/setup-sh7206.c @@ -196,7 +196,6 @@ static struct platform_device scif3_device = { static struct sh_timer_config cmt0_platform_data = { .channel_offset = 0x02, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 125, .clocksource_rating = 0, /* disabled due to code generation issues */ }; @@ -226,7 +225,6 @@ static struct platform_device cmt0_device = { static struct sh_timer_config cmt1_platform_data = { .channel_offset = 0x08, .timer_bit = 1, - .clk = "peripheral_clk", .clockevent_rating = 125, .clocksource_rating = 0, /* disabled due to code generation issues */ }; @@ -256,7 +254,6 @@ static struct platform_device cmt1_device = { static struct sh_timer_config mtu2_0_platform_data = { .channel_offset = -0x80, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -285,7 +282,6 @@ static struct platform_device mtu2_0_device = { static struct sh_timer_config mtu2_1_platform_data = { .channel_offset = -0x100, .timer_bit = 1, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -314,7 +310,6 @@ static struct platform_device mtu2_1_device = { static struct sh_timer_config mtu2_2_platform_data = { .channel_offset = 0x80, .timer_bit = 2, - .clk = "peripheral_clk", .clockevent_rating = 200, }; diff --git a/arch/sh/kernel/cpu/sh3/setup-sh7705.c b/arch/sh/kernel/cpu/sh3/setup-sh7705.c index 5e45fb7..baadd7f 100644 --- a/arch/sh/kernel/cpu/sh3/setup-sh7705.c +++ b/arch/sh/kernel/cpu/sh3/setup-sh7705.c @@ -126,7 +126,6 @@ static struct platform_device rtc_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x02, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -155,7 +154,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0xe, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -184,7 +182,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1a, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { diff --git a/arch/sh/kernel/cpu/sh3/setup-sh770x.c b/arch/sh/kernel/cpu/sh3/setup-sh770x.c index 9adb853..3cf8c8e 100644 --- a/arch/sh/kernel/cpu/sh3/setup-sh770x.c +++ b/arch/sh/kernel/cpu/sh3/setup-sh770x.c @@ -159,7 +159,6 @@ static struct platform_device scif2_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x02, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -188,7 +187,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0xe, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -217,7 +215,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1a, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { diff --git a/arch/sh/kernel/cpu/sh3/setup-sh7710.c b/arch/sh/kernel/cpu/sh3/setup-sh7710.c index 2e581e3..b0c2fb4 100644 --- a/arch/sh/kernel/cpu/sh3/setup-sh7710.c +++ b/arch/sh/kernel/cpu/sh3/setup-sh7710.c @@ -129,7 +129,6 @@ static struct platform_device scif1_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x02, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -158,7 +157,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0xe, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -187,7 +185,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1a, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { diff --git a/arch/sh/kernel/cpu/sh3/setup-sh7720.c b/arch/sh/kernel/cpu/sh3/setup-sh7720.c index 9a006e1..24b1713 100644 --- a/arch/sh/kernel/cpu/sh3/setup-sh7720.c +++ b/arch/sh/kernel/cpu/sh3/setup-sh7720.c @@ -132,7 +132,6 @@ static struct platform_device usbf_device = { static struct sh_timer_config cmt0_platform_data = { .channel_offset = 0x10, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 125, .clocksource_rating = 125, }; @@ -162,7 +161,6 @@ static struct platform_device cmt0_device = { static struct sh_timer_config cmt1_platform_data = { .channel_offset = 0x20, .timer_bit = 1, - .clk = "peripheral_clk", }; static struct resource cmt1_resources[] = { @@ -190,7 +188,6 @@ static struct platform_device cmt1_device = { static struct sh_timer_config cmt2_platform_data = { .channel_offset = 0x30, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource cmt2_resources[] = { @@ -218,7 +215,6 @@ static struct platform_device cmt2_device = { static struct sh_timer_config cmt3_platform_data = { .channel_offset = 0x40, .timer_bit = 3, - .clk = "peripheral_clk", }; static struct resource cmt3_resources[] = { @@ -246,7 +242,6 @@ static struct platform_device cmt3_device = { static struct sh_timer_config cmt4_platform_data = { .channel_offset = 0x50, .timer_bit = 4, - .clk = "peripheral_clk", }; static struct resource cmt4_resources[] = { @@ -274,7 +269,6 @@ static struct platform_device cmt4_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x02, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -303,7 +297,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0xe, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -332,7 +325,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1a, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { diff --git a/arch/sh/kernel/cpu/sh4/setup-sh4-202.c b/arch/sh/kernel/cpu/sh4/setup-sh4-202.c index 223d066..e916b18 100644 --- a/arch/sh/kernel/cpu/sh4/setup-sh4-202.c +++ b/arch/sh/kernel/cpu/sh4/setup-sh4-202.c @@ -33,7 +33,6 @@ static struct platform_device scif0_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -62,7 +61,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -91,7 +89,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { diff --git a/arch/sh/kernel/cpu/sh4/setup-sh7750.c b/arch/sh/kernel/cpu/sh4/setup-sh7750.c index 2f337b1..911d196 100644 --- a/arch/sh/kernel/cpu/sh4/setup-sh7750.c +++ b/arch/sh/kernel/cpu/sh4/setup-sh7750.c @@ -68,7 +68,6 @@ static struct platform_device scif1_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -97,7 +96,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -126,7 +124,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { @@ -159,7 +156,6 @@ static struct platform_device tmu2_device = { static struct sh_timer_config tmu3_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", }; static struct resource tmu3_resources[] = { @@ -187,7 +183,6 @@ static struct platform_device tmu3_device = { static struct sh_timer_config tmu4_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", }; static struct resource tmu4_resources[] = { diff --git a/arch/sh/kernel/cpu/sh4/setup-sh7760.c b/arch/sh/kernel/cpu/sh4/setup-sh7760.c index fa43da0..48ea8fe 100644 --- a/arch/sh/kernel/cpu/sh4/setup-sh7760.c +++ b/arch/sh/kernel/cpu/sh4/setup-sh7760.c @@ -189,7 +189,6 @@ static struct platform_device scif3_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -218,7 +217,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -247,7 +245,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7757.c b/arch/sh/kernel/cpu/sh4a/setup-sh7757.c index 25eddf2..444aca9 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7757.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7757.c @@ -65,7 +65,6 @@ static struct platform_device scif4_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -94,7 +93,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7763.c b/arch/sh/kernel/cpu/sh4a/setup-sh7763.c index a91263e..5b5f6b0 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7763.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7763.c @@ -133,7 +133,6 @@ static struct platform_device usbf_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -162,7 +161,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -191,7 +189,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { @@ -219,7 +216,6 @@ static struct platform_device tmu2_device = { static struct sh_timer_config tmu3_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", }; static struct resource tmu3_resources[] = { @@ -247,7 +243,6 @@ static struct platform_device tmu3_device = { static struct sh_timer_config tmu4_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", }; static struct resource tmu4_resources[] = { @@ -275,7 +270,6 @@ static struct platform_device tmu4_device = { static struct sh_timer_config tmu5_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu5_resources[] = { diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7770.c b/arch/sh/kernel/cpu/sh4a/setup-sh7770.c index 83a27b0..7270d7f 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7770.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7770.c @@ -167,7 +167,6 @@ static struct platform_device scif9_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -196,7 +195,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -225,7 +223,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { @@ -253,7 +250,6 @@ static struct platform_device tmu2_device = { static struct sh_timer_config tmu3_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", }; static struct resource tmu3_resources[] = { @@ -281,7 +277,6 @@ static struct platform_device tmu3_device = { static struct sh_timer_config tmu4_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", }; static struct resource tmu4_resources[] = { @@ -309,7 +304,6 @@ static struct platform_device tmu4_device = { static struct sh_timer_config tmu5_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu5_resources[] = { @@ -337,7 +331,6 @@ static struct platform_device tmu5_device = { static struct sh_timer_config tmu6_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", }; static struct resource tmu6_resources[] = { @@ -365,7 +358,6 @@ static struct platform_device tmu6_device = { static struct sh_timer_config tmu7_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", }; static struct resource tmu7_resources[] = { @@ -393,7 +385,6 @@ static struct platform_device tmu7_device = { static struct sh_timer_config tmu8_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu8_resources[] = { diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7780.c b/arch/sh/kernel/cpu/sh4a/setup-sh7780.c index f7c48e9..05fc38d 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7780.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7780.c @@ -51,7 +51,6 @@ static struct platform_device scif1_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -80,7 +79,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -109,7 +107,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { @@ -137,7 +134,6 @@ static struct platform_device tmu2_device = { static struct sh_timer_config tmu3_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", }; static struct resource tmu3_resources[] = { @@ -165,7 +161,6 @@ static struct platform_device tmu3_device = { static struct sh_timer_config tmu4_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", }; static struct resource tmu4_resources[] = { @@ -193,7 +188,6 @@ static struct platform_device tmu4_device = { static struct sh_timer_config tmu5_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu5_resources[] = { diff --git a/arch/sh/kernel/cpu/sh4a/setup-shx3.c b/arch/sh/kernel/cpu/sh4a/setup-shx3.c index ba545b5..9158bc5 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-shx3.c +++ b/arch/sh/kernel/cpu/sh4a/setup-shx3.c @@ -72,7 +72,6 @@ static struct platform_device scif2_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -101,7 +100,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -130,7 +128,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { @@ -158,7 +155,6 @@ static struct platform_device tmu2_device = { static struct sh_timer_config tmu3_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", }; static struct resource tmu3_resources[] = { @@ -186,7 +182,6 @@ static struct platform_device tmu3_device = { static struct sh_timer_config tmu4_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", }; static struct resource tmu4_resources[] = { @@ -214,7 +209,6 @@ static struct platform_device tmu4_device = { static struct sh_timer_config tmu5_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu5_resources[] = { diff --git a/arch/sh/kernel/cpu/sh5/setup-sh5.c b/arch/sh/kernel/cpu/sh5/setup-sh5.c index c84ca20..d910666 100644 --- a/arch/sh/kernel/cpu/sh5/setup-sh5.c +++ b/arch/sh/kernel/cpu/sh5/setup-sh5.c @@ -70,7 +70,6 @@ static struct platform_device rtc_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "peripheral_clk", .clockevent_rating = 200, }; @@ -99,7 +98,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "peripheral_clk", .clocksource_rating = 200, }; @@ -128,7 +126,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "peripheral_clk", }; static struct resource tmu2_resources[] = { -- cgit v0.10.2 From 5afdd0d93a32938752384f195fdbde06163b3c2e Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 29 Mar 2010 17:35:04 +0900 Subject: sh: sh7343 clock string death. Kills off clock string usage, using the new function clock definition instead. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7343.c b/arch/sh/kernel/cpu/sh4a/clock-sh7343.c index c541301..a63cdca 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7343.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7343.c @@ -154,8 +154,8 @@ static struct clk mstp_clks[] = { MSTP("sh0", &div4_clks[DIV4_P], MSTPCR0, 20, 0), MSTP("hudi0", &div4_clks[DIV4_P], MSTPCR0, 19, 0), MSTP("ubc0", &div4_clks[DIV4_P], MSTPCR0, 17, 0), - MSTP("tmu0", &div4_clks[DIV4_P], MSTPCR0, 15, 0), - MSTP("cmt0", &r_clk, MSTPCR0, 14, 0), + MSTP("tmu_fck", &div4_clks[DIV4_P], MSTPCR0, 15, 0), + MSTP("cmt_fck", &r_clk, MSTPCR0, 14, 0), MSTP("rwdt0", &r_clk, MSTPCR0, 13, 0), MSTP("mfi0", &div4_clks[DIV4_P], MSTPCR0, 11, 0), MSTP("flctl0", &div4_clks[DIV4_P], MSTPCR0, 10, 0), diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7343.c b/arch/sh/kernel/cpu/sh4a/setup-sh7343.c index 6f83949..3681cafd 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7343.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7343.c @@ -205,7 +205,6 @@ static struct platform_device jpu_device = { static struct sh_timer_config cmt_platform_data = { .channel_offset = 0x60, .timer_bit = 5, - .clk = "cmt0", .clockevent_rating = 125, .clocksource_rating = 200, }; @@ -235,7 +234,6 @@ static struct platform_device cmt_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "tmu0", .clockevent_rating = 200, }; @@ -264,7 +262,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "tmu0", .clocksource_rating = 200, }; @@ -293,7 +290,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "tmu0", }; static struct resource tmu2_resources[] = { -- cgit v0.10.2 From 2169bc1bae5d6d996276ddbd017f9d9b582abcae Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 29 Mar 2010 17:36:41 +0900 Subject: sh: sh7366 clock string death. Kills off clock string usage, using the new function clock definition instead. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7366.c b/arch/sh/kernel/cpu/sh4a/clock-sh7366.c index e10c64f..f99db94 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7366.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7366.c @@ -158,8 +158,8 @@ static struct clk mstp_clks[] = { MSTP("sh0", &div4_clks[DIV4_P], MSTPCR0, 20, 0), MSTP("hudi0", &div4_clks[DIV4_P], MSTPCR0, 19, 0), MSTP("ubc0", &div4_clks[DIV4_P], MSTPCR0, 17, 0), - MSTP("tmu0", &div4_clks[DIV4_P], MSTPCR0, 15, 0), - MSTP("cmt0", &r_clk, MSTPCR0, 14, 0), + MSTP("tmu_fck", &div4_clks[DIV4_P], MSTPCR0, 15, 0), + MSTP("cmt_fck", &r_clk, MSTPCR0, 14, 0), MSTP("rwdt0", &r_clk, MSTPCR0, 13, 0), MSTP("mfi0", &div4_clks[DIV4_P], MSTPCR0, 11, 0), MSTP("flctl0", &div4_clks[DIV4_P], MSTPCR0, 10, 0), diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7366.c b/arch/sh/kernel/cpu/sh4a/setup-sh7366.c index 1d897ed..8dab9e1 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7366.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7366.c @@ -170,7 +170,6 @@ static struct platform_device veu1_device = { static struct sh_timer_config cmt_platform_data = { .channel_offset = 0x60, .timer_bit = 5, - .clk = "cmt0", .clockevent_rating = 125, .clocksource_rating = 200, }; @@ -200,7 +199,6 @@ static struct platform_device cmt_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "tmu0", .clockevent_rating = 200, }; @@ -229,7 +227,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "tmu0", .clocksource_rating = 200, }; @@ -258,7 +255,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "tmu0", }; static struct resource tmu2_resources[] = { -- cgit v0.10.2 From 04b1731766b22cd41a0acd480e3206e93f2d5931 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 29 Mar 2010 17:37:39 +0900 Subject: sh: sh7722 clock string death. Kills off clock string usage, using the new function clock definition instead. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7722.c b/arch/sh/kernel/cpu/sh4a/clock-sh7722.c index 0732c8b..107b200 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7722.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7722.c @@ -160,8 +160,8 @@ struct clk div6_clks[] = { static struct clk mstp_clks[] = { SH_HWBLK_CLK("uram0", -1, U_CLK, HWBLK_URAM, CLK_ENABLE_ON_INIT), SH_HWBLK_CLK("xymem0", -1, B_CLK, HWBLK_XYMEM, CLK_ENABLE_ON_INIT), - SH_HWBLK_CLK("tmu0", -1, P_CLK, HWBLK_TMU, 0), - SH_HWBLK_CLK("cmt0", -1, R_CLK, HWBLK_CMT, 0), + SH_HWBLK_CLK("tmu_fck", -1, P_CLK, HWBLK_TMU, 0), + SH_HWBLK_CLK("cmt_fck", -1, R_CLK, HWBLK_CMT, 0), SH_HWBLK_CLK("rwdt0", -1, R_CLK, HWBLK_RWDT, 0), SH_HWBLK_CLK("flctl0", -1, P_CLK, HWBLK_FLCTL, 0), SH_HWBLK_CLK("sci_fck", 0, P_CLK, HWBLK_SCIF0, 0), diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7722.c b/arch/sh/kernel/cpu/sh4a/setup-sh7722.c index 847a276..dc9b30d 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7722.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7722.c @@ -400,7 +400,6 @@ static struct platform_device jpu_device = { static struct sh_timer_config cmt_platform_data = { .channel_offset = 0x60, .timer_bit = 5, - .clk = "cmt0", .clockevent_rating = 125, .clocksource_rating = 125, }; @@ -433,7 +432,6 @@ static struct platform_device cmt_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "tmu0", .clockevent_rating = 200, }; @@ -465,7 +463,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "tmu0", .clocksource_rating = 200, }; @@ -497,7 +494,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "tmu0", }; static struct resource tmu2_resources[] = { -- cgit v0.10.2 From f4221802ea3a28f7c16d157ac6f697d58267cb58 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Mon, 29 Mar 2010 17:45:12 +0900 Subject: sh: sh7723 clkdev lookups. Convert to TMU clock lookups for SH7723. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7723.c b/arch/sh/kernel/cpu/sh4a/clock-sh7723.c index 4d4c123..fc86c88 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7723.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7723.c @@ -21,6 +21,8 @@ #include #include #include +#include +#include #include #include #include @@ -171,11 +173,11 @@ static struct clk mstp_clks[] = { SH_HWBLK_CLK("sh0", -1, SH_CLK, HWBLK_SHYWAY, CLK_ENABLE_ON_INIT), SH_HWBLK_CLK("hudi0", -1, P_CLK, HWBLK_HUDI, 0), SH_HWBLK_CLK("ubc0", -1, I_CLK, HWBLK_UBC, 0), - SH_HWBLK_CLK("tmu0", -1, P_CLK, HWBLK_TMU0, 0), - SH_HWBLK_CLK("cmt0", -1, R_CLK, HWBLK_CMT, 0), + SH_HWBLK_CLK("tmu012_fck", -1, P_CLK, HWBLK_TMU0, 0), + SH_HWBLK_CLK("cmt_fck", -1, R_CLK, HWBLK_CMT, 0), SH_HWBLK_CLK("rwdt0", -1, R_CLK, HWBLK_RWDT, 0), SH_HWBLK_CLK("dmac1", -1, B_CLK, HWBLK_DMAC1, 0), - SH_HWBLK_CLK("tmu1", -1, P_CLK, HWBLK_TMU1, 0), + SH_HWBLK_CLK("tmu345_fck", -1, P_CLK, HWBLK_TMU1, 0), SH_HWBLK_CLK("flctl0", -1, P_CLK, HWBLK_FLCTL, 0), SH_HWBLK_CLK("sci_fck", 0, P_CLK, HWBLK_SCIF0, 0), SH_HWBLK_CLK("sci_fck", 1, P_CLK, HWBLK_SCIF1, 0), @@ -211,6 +213,40 @@ static struct clk mstp_clks[] = { SH_HWBLK_CLK("lcdc0", -1, B_CLK, HWBLK_LCDC, 0), }; +static struct clk_lookup lookups[] = { + { + /* TMU0 */ + .dev_id = "sh_tmu.0", + .con_id = "tmu_fck", + .clk = &mstp_clks[11], /* tmu012_fck */ + }, { + /* TMU1 */ + .dev_id = "sh_tmu.1", + .con_id = "tmu_fck", + .clk = &mstp_clks[11], + }, { + /* TMU2 */ + .dev_id = "sh_tmu.2", + .con_id = "tmu_fck", + .clk = &mstp_clks[11], + }, { + /* TMU3 */ + .dev_id = "sh_tmu.3", + .con_id = "tmu_fck", + .clk = &mstp_clks[15], /* tmu345_fck */ + }, { + /* TMU4 */ + .dev_id = "sh_tmu.4", + .con_id = "tmu_fck", + .clk = &mstp_clks[15], + }, { + /* TMU5 */ + .dev_id = "sh_tmu.5", + .con_id = "tmu_fck", + .clk = &mstp_clks[15], + }, +}; + int __init arch_clk_init(void) { int k, ret = 0; @@ -222,7 +258,9 @@ int __init arch_clk_init(void) pll_clk.parent = &extal_clk; for (k = 0; !ret && (k < ARRAY_SIZE(main_clks)); k++) - ret = clk_register(main_clks[k]); + ret |= clk_register(main_clks[k]); + + clkdev_add_table(lookups, ARRAY_SIZE(lookups)); if (!ret) ret = sh_clk_div4_register(div4_clks, DIV4_NR, &div4_table); diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7723.c b/arch/sh/kernel/cpu/sh4a/setup-sh7723.c index 187462b..0eadefd 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7723.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7723.c @@ -207,7 +207,6 @@ static struct platform_device veu1_device = { static struct sh_timer_config cmt_platform_data = { .channel_offset = 0x60, .timer_bit = 5, - .clk = "cmt0", .clockevent_rating = 125, .clocksource_rating = 125, }; @@ -240,7 +239,6 @@ static struct platform_device cmt_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "tmu0", .clockevent_rating = 200, }; @@ -272,7 +270,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "tmu0", .clocksource_rating = 200, }; @@ -304,7 +301,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "tmu0", }; static struct resource tmu2_resources[] = { @@ -335,7 +331,6 @@ static struct platform_device tmu2_device = { static struct sh_timer_config tmu3_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "tmu1", }; static struct resource tmu3_resources[] = { @@ -366,7 +361,6 @@ static struct platform_device tmu3_device = { static struct sh_timer_config tmu4_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "tmu1", }; static struct resource tmu4_resources[] = { @@ -397,7 +391,6 @@ static struct platform_device tmu4_device = { static struct sh_timer_config tmu5_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "tmu1", }; static struct resource tmu5_resources[] = { -- cgit v0.10.2 From 6694635d3ae1b038d7a0e38b80637db867c7c8e2 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 29 Mar 2010 17:21:45 +0200 Subject: ALSA: hda - Fix ADC/MUX assignment of ALC269 codec ALC269 codec has a few different variants, and each of them may have different ADC and MUX widgets. For example, one model has ADC 0x08 with MUX 0x23 while others has ADC 0x09 or ADC 0x07 with MUX 022 or 0x24. The difference of ADC appears usually as the capability of the digital mic pin (0x12), and the current driver sometimes misses the internal mic pin due to the mismatching ADC. This patch adds a bit more clever way to find the matching ADC instead of the static list. Now the driver checks all active input pins and fills only the ADC/MUX's that contain all of them. Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index bc55c1e..22aea7b 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -4984,6 +4984,69 @@ static void set_capture_mixer(struct hda_codec *codec) } } +/* fill adc_nids (and capsrc_nids) containing all active input pins */ +static void fillup_priv_adc_nids(struct hda_codec *codec, hda_nid_t *nids, + int num_nids) +{ + struct alc_spec *spec = codec->spec; + int n; + hda_nid_t fallback_adc = 0, fallback_cap = 0; + + for (n = 0; n < num_nids; n++) { + hda_nid_t adc, cap; + hda_nid_t conn[HDA_MAX_NUM_INPUTS]; + int nconns, i, j; + + adc = nids[n]; + if (get_wcaps_type(get_wcaps(codec, adc)) != AC_WID_AUD_IN) + continue; + cap = adc; + nconns = snd_hda_get_connections(codec, cap, conn, + ARRAY_SIZE(conn)); + if (nconns == 1) { + cap = conn[0]; + nconns = snd_hda_get_connections(codec, cap, conn, + ARRAY_SIZE(conn)); + } + if (nconns <= 0) + continue; + if (!fallback_adc) { + fallback_adc = adc; + fallback_cap = cap; + } + for (i = 0; i < AUTO_PIN_LAST; i++) { + hda_nid_t nid = spec->autocfg.input_pins[i]; + if (!nid) + continue; + for (j = 0; j < nconns; j++) { + if (conn[j] == nid) + break; + } + if (j >= nconns) + break; + } + if (i >= AUTO_PIN_LAST) { + int num_adcs = spec->num_adc_nids; + spec->private_adc_nids[num_adcs] = adc; + spec->private_capsrc_nids[num_adcs] = cap; + spec->num_adc_nids++; + spec->adc_nids = spec->private_adc_nids; + if (adc != cap) + spec->capsrc_nids = spec->private_capsrc_nids; + } + } + if (!spec->num_adc_nids) { + printk(KERN_WARNING "hda_codec: %s: no valid ADC found;" + " using fallback 0x%x\n", fallback_adc); + spec->private_adc_nids[0] = fallback_adc; + spec->adc_nids = spec->private_adc_nids; + if (fallback_adc != fallback_cap) { + spec->private_capsrc_nids[0] = fallback_cap; + spec->capsrc_nids = spec->private_adc_nids; + } + } +} + #ifdef CONFIG_SND_HDA_INPUT_BEEP #define set_beep_amp(spec, nid, idx, dir) \ ((spec)->beep_amp = HDA_COMPOSE_AMP_VAL(nid, 3, idx, dir)) @@ -13333,9 +13396,9 @@ static hda_nid_t alc269vb_capsrc_nids[1] = { 0x22, }; -/* NOTE: ADC2 (0x07) is connected from a recording *MIXER* (0x24), - * not a mux! - */ +static hda_nid_t alc269_adc_candidates[] = { + 0x08, 0x09, 0x07, +}; #define alc269_modes alc260_modes #define alc269_capture_source alc880_lg_lw_capture_source @@ -13842,7 +13905,6 @@ static int alc269_parse_auto_config(struct hda_codec *codec) struct alc_spec *spec = codec->spec; int err; static hda_nid_t alc269_ignore[] = { 0x1d, 0 }; - hda_nid_t real_capsrc_nids; err = snd_hda_parse_pin_def_config(codec, &spec->autocfg, alc269_ignore); @@ -13866,18 +13928,19 @@ static int alc269_parse_auto_config(struct hda_codec *codec) if ((alc_read_coef_idx(codec, 0) & 0x00f0) == 0x0010) { add_verb(spec, alc269vb_init_verbs); - real_capsrc_nids = alc269vb_capsrc_nids[0]; alc_ssid_check(codec, 0, 0x1b, 0x14, 0x21); } else { add_verb(spec, alc269_init_verbs); - real_capsrc_nids = alc269_capsrc_nids[0]; alc_ssid_check(codec, 0x15, 0x1b, 0x14, 0); } spec->num_mux_defs = 1; spec->input_mux = &spec->private_imux[0]; + fillup_priv_adc_nids(codec, alc269_adc_candidates, + sizeof(alc269_adc_candidates)); + /* set default input source */ - snd_hda_codec_write_cache(codec, real_capsrc_nids, + snd_hda_codec_write_cache(codec, spec->capsrc_nids[0], 0, AC_VERB_SET_CONNECT_SEL, spec->input_mux->items[0].index); @@ -14156,14 +14219,16 @@ static int patch_alc269(struct hda_codec *codec) spec->stream_digital_playback = &alc269_pcm_digital_playback; spec->stream_digital_capture = &alc269_pcm_digital_capture; - if (!is_alc269vb) { - spec->adc_nids = alc269_adc_nids; - spec->num_adc_nids = ARRAY_SIZE(alc269_adc_nids); - spec->capsrc_nids = alc269_capsrc_nids; - } else { - spec->adc_nids = alc269vb_adc_nids; - spec->num_adc_nids = ARRAY_SIZE(alc269vb_adc_nids); - spec->capsrc_nids = alc269vb_capsrc_nids; + if (!spec->adc_nids) { /* wasn't filled automatically? use default */ + if (!is_alc269vb) { + spec->adc_nids = alc269_adc_nids; + spec->num_adc_nids = ARRAY_SIZE(alc269_adc_nids); + spec->capsrc_nids = alc269_capsrc_nids; + } else { + spec->adc_nids = alc269vb_adc_nids; + spec->num_adc_nids = ARRAY_SIZE(alc269vb_adc_nids); + spec->capsrc_nids = alc269vb_capsrc_nids; + } } if (!spec->cap_mixer) -- cgit v0.10.2 From fb48e3c6a4d8888aff61fbf567aadac7d206e973 Mon Sep 17 00:00:00 2001 From: Graham Gower Date: Thu, 25 Mar 2010 10:52:12 +1030 Subject: ASoC: Fix passing platform_data to ac97 bus users and fix a leak [The issue is an attempt to write the pdata without the AC97 device allocated when using ac97.c - also added a comment in soc-core.c for the special case for ac97. -- broonie] Signed-off-by: Graham Gower Acked-by: Liam Girdwood Signed-off-by: Mark Brown diff --git a/sound/soc/codecs/ac97.c b/sound/soc/codecs/ac97.c index a1bbe16..bcfa532 100644 --- a/sound/soc/codecs/ac97.c +++ b/sound/soc/codecs/ac97.c @@ -80,9 +80,11 @@ static int ac97_write(struct snd_soc_codec *codec, unsigned int reg, static int ac97_soc_probe(struct platform_device *pdev) { struct snd_soc_device *socdev = platform_get_drvdata(pdev); + struct snd_soc_card *card = socdev->card; struct snd_soc_codec *codec; struct snd_ac97_bus *ac97_bus; struct snd_ac97_template ac97_template; + int i; int ret = 0; printk(KERN_INFO "AC97 SoC Audio Codec %s\n", AC97_VERSION); @@ -102,12 +104,6 @@ static int ac97_soc_probe(struct platform_device *pdev) INIT_LIST_HEAD(&codec->dapm_widgets); INIT_LIST_HEAD(&codec->dapm_paths); - ret = snd_soc_new_ac97_codec(codec, &soc_ac97_ops, 0); - if (ret < 0) { - printk(KERN_ERR "ASoC: failed to init gen ac97 glue\n"); - goto err; - } - /* register pcms */ ret = snd_soc_new_pcms(socdev, SNDRV_DEFAULT_IDX1, SNDRV_DEFAULT_STR1); if (ret < 0) @@ -123,6 +119,13 @@ static int ac97_soc_probe(struct platform_device *pdev) if (ret < 0) goto bus_err; + for (i = 0; i < card->num_links; i++) { + if (card->dai_link[i].codec_dai->ac97_control) { + snd_ac97_dev_add_pdata(codec->ac97, + card->dai_link[i].cpu_dai->ac97_pdata); + } + } + return 0; bus_err: diff --git a/sound/soc/soc-core.c b/sound/soc/soc-core.c index c8b0556..d0efd5e 100644 --- a/sound/soc/soc-core.c +++ b/sound/soc/soc-core.c @@ -1548,7 +1548,8 @@ int snd_soc_new_pcms(struct snd_soc_device *socdev, int idx, const char *xid) mutex_unlock(&codec->mutex); return ret; } - if (card->dai_link[i].codec_dai->ac97_control) { + /* Check for codec->ac97 to handle the ac97.c fun */ + if (card->dai_link[i].codec_dai->ac97_control && codec->ac97) { snd_ac97_dev_add_pdata(codec->ac97, card->dai_link[i].cpu_dai->ac97_pdata); } -- cgit v0.10.2 From 788e69e548cc8d127b90f0de1f7b7e983d1d587a Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 29 Mar 2010 21:02:31 -0400 Subject: svcrpc: don't hold sv_lock over svc_xprt_put() svc_xprt_put() can call tcp_close(), which can sleep, so we shouldn't be holding this lock. In fact, only the xpt_list removal and the sv_tmpcnt decrement should need the sv_lock here. Reported-by: Mi Jinlong Signed-off-by: J. Bruce Fields diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c index 8f0f1fb..c334f54 100644 --- a/net/sunrpc/svc_xprt.c +++ b/net/sunrpc/svc_xprt.c @@ -892,12 +892,12 @@ void svc_delete_xprt(struct svc_xprt *xprt) */ if (test_bit(XPT_TEMP, &xprt->xpt_flags)) serv->sv_tmpcnt--; + spin_unlock_bh(&serv->sv_lock); while ((dr = svc_deferred_dequeue(xprt)) != NULL) kfree(dr); svc_xprt_put(xprt); - spin_unlock_bh(&serv->sv_lock); } void svc_close_xprt(struct svc_xprt *xprt) -- cgit v0.10.2 From f4cff0d0ffc0162636357d0475409ecf1304f854 Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Tue, 30 Mar 2010 11:20:35 +0900 Subject: sh: sh7724 clkdev lookups. Conver to TMU clock lookups for SH7724. Signed-off-by: Paul Mundt diff --git a/arch/sh/kernel/cpu/sh4a/clock-sh7724.c b/arch/sh/kernel/cpu/sh4a/clock-sh7724.c index bcf2b97..f1583a2 100644 --- a/arch/sh/kernel/cpu/sh4a/clock-sh7724.c +++ b/arch/sh/kernel/cpu/sh4a/clock-sh7724.c @@ -21,6 +21,8 @@ #include #include #include +#include +#include #include #include #include @@ -189,11 +191,11 @@ static struct clk mstp_clks[] = { SH_HWBLK_CLK("sh0", -1, SH_CLK, HWBLK_SHYWAY, CLK_ENABLE_ON_INIT), SH_HWBLK_CLK("hudi0", -1, P_CLK, HWBLK_HUDI, 0), SH_HWBLK_CLK("ubc0", -1, I_CLK, HWBLK_UBC, 0), - SH_HWBLK_CLK("tmu0", -1, P_CLK, HWBLK_TMU0, 0), - SH_HWBLK_CLK("cmt0", -1, R_CLK, HWBLK_CMT, 0), + SH_HWBLK_CLK("tmu012_fck", -1, P_CLK, HWBLK_TMU0, 0), + SH_HWBLK_CLK("cmt_fck", -1, R_CLK, HWBLK_CMT, 0), SH_HWBLK_CLK("rwdt0", -1, R_CLK, HWBLK_RWDT, 0), SH_HWBLK_CLK("dmac1", -1, B_CLK, HWBLK_DMAC1, 0), - SH_HWBLK_CLK("tmu1", -1, P_CLK, HWBLK_TMU1, 0), + SH_HWBLK_CLK("tmu345_fck", -1, P_CLK, HWBLK_TMU1, 0), SH_HWBLK_CLK("sci_fck", 0, P_CLK, HWBLK_SCIF0, 0), SH_HWBLK_CLK("sci_fck", 1, P_CLK, HWBLK_SCIF1, 0), SH_HWBLK_CLK("sci_fck", 2, P_CLK, HWBLK_SCIF2, 0), @@ -233,6 +235,40 @@ static struct clk mstp_clks[] = { SH_HWBLK_CLK("lcdc0", -1, B_CLK, HWBLK_LCDC, 0), }; +static struct clk_lookup lookups[] = { + { + /* TMU0 */ + .dev_id = "sh_tmu.0", + .con_id = "tmu_fck", + .clk = &mstp_clks[12], /* tmu012_fck */ + }, { + /* TMU1 */ + .dev_id = "sh_tmu.1", + .con_id = "tmu_fck", + .clk = &mstp_clks[12], + }, { + /* TMU2 */ + .dev_id = "sh_tmu.2", + .con_id = "tmu_fck", + .clk = &mstp_clks[12], + }, { + /* TMU3 */ + .dev_id = "sh_tmu.3", + .con_id = "tmu_fck", + .clk = &mstp_clks[16], /* tmu345_fck */ + }, { + /* TMU4 */ + .dev_id = "sh_tmu.4", + .con_id = "tmu_fck", + .clk = &mstp_clks[16], + }, { + /* TMU5 */ + .dev_id = "sh_tmu.5", + .con_id = "tmu_fck", + .clk = &mstp_clks[16], + }, +}; + int __init arch_clk_init(void) { int k, ret = 0; @@ -246,6 +282,8 @@ int __init arch_clk_init(void) for (k = 0; !ret && (k < ARRAY_SIZE(main_clks)); k++) ret = clk_register(main_clks[k]); + clkdev_add_table(lookups, ARRAY_SIZE(lookups)); + if (!ret) ret = sh_clk_div4_register(div4_clks, DIV4_NR, &div4_table); diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7724.c b/arch/sh/kernel/cpu/sh4a/setup-sh7724.c index ee97890..8a0a4a9 100644 --- a/arch/sh/kernel/cpu/sh4a/setup-sh7724.c +++ b/arch/sh/kernel/cpu/sh4a/setup-sh7724.c @@ -481,7 +481,6 @@ static struct platform_device veu1_device = { static struct sh_timer_config cmt_platform_data = { .channel_offset = 0x60, .timer_bit = 5, - .clk = "cmt0", .clockevent_rating = 125, .clocksource_rating = 200, }; @@ -514,7 +513,6 @@ static struct platform_device cmt_device = { static struct sh_timer_config tmu0_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "tmu0", .clockevent_rating = 200, }; @@ -546,7 +544,6 @@ static struct platform_device tmu0_device = { static struct sh_timer_config tmu1_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "tmu0", .clocksource_rating = 200, }; @@ -578,7 +575,6 @@ static struct platform_device tmu1_device = { static struct sh_timer_config tmu2_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "tmu0", }; static struct resource tmu2_resources[] = { @@ -610,7 +606,6 @@ static struct platform_device tmu2_device = { static struct sh_timer_config tmu3_platform_data = { .channel_offset = 0x04, .timer_bit = 0, - .clk = "tmu1", }; static struct resource tmu3_resources[] = { @@ -641,7 +636,6 @@ static struct platform_device tmu3_device = { static struct sh_timer_config tmu4_platform_data = { .channel_offset = 0x10, .timer_bit = 1, - .clk = "tmu1", }; static struct resource tmu4_resources[] = { @@ -672,7 +666,6 @@ static struct platform_device tmu4_device = { static struct sh_timer_config tmu5_platform_data = { .channel_offset = 0x1c, .timer_bit = 2, - .clk = "tmu1", }; static struct resource tmu5_resources[] = { -- cgit v0.10.2 From 1f85d72d2c9c9a1d6d32cf325936bc224ad5d591 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 30 Mar 2010 07:48:05 +0200 Subject: ALSA: hda - Add missing printk argument in previous patch Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 22aea7b..ca93c4c 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -5037,7 +5037,8 @@ static void fillup_priv_adc_nids(struct hda_codec *codec, hda_nid_t *nids, } if (!spec->num_adc_nids) { printk(KERN_WARNING "hda_codec: %s: no valid ADC found;" - " using fallback 0x%x\n", fallback_adc); + " using fallback 0x%x\n", + codec->chip_name, fallback_adc); spec->private_adc_nids[0] = fallback_adc; spec->adc_nids = spec->private_adc_nids; if (fallback_adc != fallback_cap) { -- cgit v0.10.2 From a7dc12ba756021a95866730b482c6e34b13f84ee Mon Sep 17 00:00:00 2001 From: Vladimir Zapolskiy Date: Fri, 19 Mar 2010 18:16:37 +0300 Subject: imx3: Add watchdog platform device support MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This patch adds support for build-in watchdog device found on Freescale imx31 and imx35 SoCs. Signed-off-by: Vladimir Zapolskiy Cc: Sascha Hauer Cc: Uwe Kleine-König Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/devices.c b/arch/arm/mach-mx3/devices.c index 6adb586..f891115 100644 --- a/arch/arm/mach-mx3/devices.c +++ b/arch/arm/mach-mx3/devices.c @@ -575,11 +575,26 @@ struct platform_device imx_ssi_device1 = { .resource = imx_ssi_resources1, }; -static int mx3_devices_init(void) +static struct resource imx_wdt_resources[] = { + { + .flags = IORESOURCE_MEM, + }, +}; + +struct platform_device imx_wdt_device0 = { + .name = "imx-wdt", + .id = 0, + .num_resources = ARRAY_SIZE(imx_wdt_resources), + .resource = imx_wdt_resources, +}; + +static int __init mx3_devices_init(void) { if (cpu_is_mx31()) { mxc_nand_resources[0].start = MX31_NFC_BASE_ADDR; mxc_nand_resources[0].end = MX31_NFC_BASE_ADDR + 0xfff; + imx_wdt_resources[0].start = MX31_WDOG_BASE_ADDR; + imx_wdt_resources[0].end = MX31_WDOG_BASE_ADDR + 0x3fff; mxc_register_device(&mxc_rnga_device, NULL); } if (cpu_is_mx35()) { @@ -597,6 +612,8 @@ static int mx3_devices_init(void) imx_ssi_resources0[1].end = MX35_INT_SSI1; imx_ssi_resources1[1].start = MX35_INT_SSI2; imx_ssi_resources1[1].end = MX35_INT_SSI2; + imx_wdt_resources[0].start = MX35_WDOG_BASE_ADDR; + imx_wdt_resources[0].end = MX35_WDOG_BASE_ADDR + 0x3fff; } return 0; diff --git a/arch/arm/mach-mx3/devices.h b/arch/arm/mach-mx3/devices.h index 42cf175..4f77eb5 100644 --- a/arch/arm/mach-mx3/devices.h +++ b/arch/arm/mach-mx3/devices.h @@ -25,4 +25,5 @@ extern struct platform_device mxc_spi_device1; extern struct platform_device mxc_spi_device2; extern struct platform_device imx_ssi_device0; extern struct platform_device imx_ssi_device1; - +extern struct platform_device imx_ssi_device1; +extern struct platform_device imx_wdt_device0; -- cgit v0.10.2 From 7d7a77e56744bb32b8a97a9903fd12433af1c4b2 Mon Sep 17 00:00:00 2001 From: Vladimir Zapolskiy Date: Fri, 19 Mar 2010 18:16:59 +0300 Subject: imx31: add watchdog device on litekit board. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This patch adds support for SoC build-in watchdog device on litekit board. Signed-off-by: Vladimir Zapolskiy Cc: Uwe Kleine-König Cc: Sascha Hauer Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/mx31lite-db.c b/arch/arm/mach-mx3/mx31lite-db.c index ccd8742..0201683 100644 --- a/arch/arm/mach-mx3/mx31lite-db.c +++ b/arch/arm/mach-mx3/mx31lite-db.c @@ -206,5 +206,6 @@ void __init mx31lite_db_init(void) mxc_register_device(&mxcsdhc_device0, &mmc_pdata); mxc_register_device(&mxc_spi_device0, &spi0_pdata); platform_device_register(&litekit_led_device); + mxc_register_device(&imx_wdt_device0, NULL); } -- cgit v0.10.2 From 034cf2362af85259718c4930bc7b548fa4165c1c Mon Sep 17 00:00:00 2001 From: Andrea Gelmini Date: Thu, 25 Mar 2010 18:22:33 +0100 Subject: ARM mach-mx3: duplicated include arch/arm/mach-mx3/mach-pcm037.c: linux/fsl_devices.h is included more than once. Signed-off-by: Andrea Gelmini Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/mach-pcm037.c b/arch/arm/mach-mx3/mach-pcm037.c index 11f5315..a550ae9 100644 --- a/arch/arm/mach-mx3/mach-pcm037.c +++ b/arch/arm/mach-mx3/mach-pcm037.c @@ -35,7 +35,6 @@ #include #include #include -#include #include -- cgit v0.10.2 From 17669fd60e6b245945cdd4bbe87d1186a07a03b6 Mon Sep 17 00:00:00 2001 From: Andrea Gelmini Date: Thu, 25 Mar 2010 18:22:34 +0100 Subject: ARM mach-mx3: duplicated include arch/arm/mach-mx3/mx31lite-db.c: linux/platform_device.h is included more than once. Signed-off-by: Andrea Gelmini Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/mx31lite-db.c b/arch/arm/mach-mx3/mx31lite-db.c index 0201683..093c595 100644 --- a/arch/arm/mach-mx3/mx31lite-db.c +++ b/arch/arm/mach-mx3/mx31lite-db.c @@ -28,7 +28,6 @@ #include #include #include -#include #include #include -- cgit v0.10.2 From b3aa111f7ac1ad07009081a824b5ce7a71462c4c Mon Sep 17 00:00:00 2001 From: Alberto Panizzo Date: Fri, 26 Feb 2010 18:36:32 +0100 Subject: MXC: mach_armadillo5x0: Add USB Host support. This add USB Host capability. The Armadillo 500 board is supplied with two USB Host connectors driven by the USB OTG and USB Host 2 ports, through two NXP isp 1504 transceivers. Signed-off-by: Alberto Panizzo Signed-off-by: Sascha Hauer diff --git a/arch/arm/mach-mx3/Kconfig b/arch/arm/mach-mx3/Kconfig index 9a2911e..170f68e 100644 --- a/arch/arm/mach-mx3/Kconfig +++ b/arch/arm/mach-mx3/Kconfig @@ -104,6 +104,7 @@ config MACH_PCM043 config MACH_ARMADILLO5X0 bool "Support Atmark Armadillo-500 Development Base Board" select ARCH_MX31 + select MXC_ULPI if USB_ULPI help Include support for Atmark Armadillo-500 platform. This includes specific configurations for the board and its peripherals. diff --git a/arch/arm/mach-mx3/mach-armadillo5x0.c b/arch/arm/mach-mx3/mach-armadillo5x0.c index 3d72b0b..5f72ec9 100644 --- a/arch/arm/mach-mx3/mach-armadillo5x0.c +++ b/arch/arm/mach-mx3/mach-armadillo5x0.c @@ -36,6 +36,9 @@ #include #include #include +#include +#include +#include #include #include @@ -52,6 +55,8 @@ #include #include #include +#include +#include #include "devices.h" #include "crm_regs.h" @@ -103,8 +108,158 @@ static int armadillo5x0_pins[] = { /* I2C2 */ MX31_PIN_CSPI2_MOSI__SCL, MX31_PIN_CSPI2_MISO__SDA, + /* OTG */ + MX31_PIN_USBOTG_DATA0__USBOTG_DATA0, + MX31_PIN_USBOTG_DATA1__USBOTG_DATA1, + MX31_PIN_USBOTG_DATA2__USBOTG_DATA2, + MX31_PIN_USBOTG_DATA3__USBOTG_DATA3, + MX31_PIN_USBOTG_DATA4__USBOTG_DATA4, + MX31_PIN_USBOTG_DATA5__USBOTG_DATA5, + MX31_PIN_USBOTG_DATA6__USBOTG_DATA6, + MX31_PIN_USBOTG_DATA7__USBOTG_DATA7, + MX31_PIN_USBOTG_CLK__USBOTG_CLK, + MX31_PIN_USBOTG_DIR__USBOTG_DIR, + MX31_PIN_USBOTG_NXT__USBOTG_NXT, + MX31_PIN_USBOTG_STP__USBOTG_STP, + /* USB host 2 */ + IOMUX_MODE(MX31_PIN_USBH2_CLK, IOMUX_CONFIG_FUNC), + IOMUX_MODE(MX31_PIN_USBH2_DIR, IOMUX_CONFIG_FUNC), + IOMUX_MODE(MX31_PIN_USBH2_NXT, IOMUX_CONFIG_FUNC), + IOMUX_MODE(MX31_PIN_USBH2_STP, IOMUX_CONFIG_FUNC), + IOMUX_MODE(MX31_PIN_USBH2_DATA0, IOMUX_CONFIG_FUNC), + IOMUX_MODE(MX31_PIN_USBH2_DATA1, IOMUX_CONFIG_FUNC), + IOMUX_MODE(MX31_PIN_STXD3, IOMUX_CONFIG_FUNC), + IOMUX_MODE(MX31_PIN_SRXD3, IOMUX_CONFIG_FUNC), + IOMUX_MODE(MX31_PIN_SCK3, IOMUX_CONFIG_FUNC), + IOMUX_MODE(MX31_PIN_SFS3, IOMUX_CONFIG_FUNC), + IOMUX_MODE(MX31_PIN_STXD6, IOMUX_CONFIG_FUNC), + IOMUX_MODE(MX31_PIN_SRXD6, IOMUX_CONFIG_FUNC), }; +/* USB */ +#if defined(CONFIG_USB_ULPI) + +#define OTG_RESET IOMUX_TO_GPIO(MX31_PIN_STXD4) +#define USBH2_RESET IOMUX_TO_GPIO(MX31_PIN_SCK6) +#define USBH2_CS IOMUX_TO_GPIO(MX31_PIN_GPIO1_3) + +#define USB_PAD_CFG (PAD_CTL_DRV_MAX | PAD_CTL_SRE_FAST | PAD_CTL_HYS_CMOS | \ + PAD_CTL_ODE_CMOS | PAD_CTL_100K_PU) + +static int usbotg_init(struct platform_device *pdev) +{ + int err; + + mxc_iomux_set_pad(MX31_PIN_USBOTG_DATA0, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBOTG_DATA1, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBOTG_DATA2, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBOTG_DATA3, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBOTG_DATA4, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBOTG_DATA5, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBOTG_DATA6, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBOTG_DATA7, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBOTG_CLK, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBOTG_DIR, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBOTG_NXT, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBOTG_STP, USB_PAD_CFG); + + /* Chip already enabled by hardware */ + /* OTG phy reset*/ + err = gpio_request(OTG_RESET, "USB-OTG-RESET"); + if (err) { + pr_err("Failed to request the usb otg reset gpio\n"); + return err; + } + + err = gpio_direction_output(OTG_RESET, 1/*HIGH*/); + if (err) { + pr_err("Failed to reset the usb otg phy\n"); + goto otg_free_reset; + } + + gpio_set_value(OTG_RESET, 0/*LOW*/); + mdelay(5); + gpio_set_value(OTG_RESET, 1/*HIGH*/); + + return 0; + +otg_free_reset: + gpio_free(OTG_RESET); + return err; +} + +static int usbh2_init(struct platform_device *pdev) +{ + int err; + + mxc_iomux_set_pad(MX31_PIN_USBH2_CLK, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBH2_DIR, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBH2_NXT, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBH2_STP, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBH2_DATA0, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_USBH2_DATA1, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_SRXD6, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_STXD6, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_SFS3, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_SCK3, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_SRXD3, USB_PAD_CFG); + mxc_iomux_set_pad(MX31_PIN_STXD3, USB_PAD_CFG); + + mxc_iomux_set_gpr(MUX_PGP_UH2, true); + + + /* Enable the chip */ + err = gpio_request(USBH2_CS, "USB-H2-CS"); + if (err) { + pr_err("Failed to request the usb host 2 CS gpio\n"); + return err; + } + + err = gpio_direction_output(USBH2_CS, 0/*Enabled*/); + if (err) { + pr_err("Failed to drive the usb host 2 CS gpio\n"); + goto h2_free_cs; + } + + /* H2 phy reset*/ + err = gpio_request(USBH2_RESET, "USB-H2-RESET"); + if (err) { + pr_err("Failed to request the usb host 2 reset gpio\n"); + goto h2_free_cs; + } + + err = gpio_direction_output(USBH2_RESET, 1/*HIGH*/); + if (err) { + pr_err("Failed to reset the usb host 2 phy\n"); + goto h2_free_reset; + } + + gpio_set_value(USBH2_RESET, 0/*LOW*/); + mdelay(5); + gpio_set_value(USBH2_RESET, 1/*HIGH*/); + + return 0; + +h2_free_reset: + gpio_free(USBH2_RESET); +h2_free_cs: + gpio_free(USBH2_CS); + return err; +} + +static struct mxc_usbh_platform_data usbotg_pdata = { + .init = usbotg_init, + .portsc = MXC_EHCI_MODE_ULPI | MXC_EHCI_UTMI_8BIT, + .flags = MXC_EHCI_POWER_PINS_ENABLED | MXC_EHCI_INTERFACE_DIFF_UNI, +}; + +static struct mxc_usbh_platform_data usbh2_pdata = { + .init = usbh2_init, + .portsc = MXC_EHCI_MODE_ULPI | MXC_EHCI_UTMI_8BIT, + .flags = MXC_EHCI_POWER_PINS_ENABLED | MXC_EHCI_INTERFACE_DIFF_UNI, +}; +#endif /* CONFIG_USB_ULPI */ + /* RTC over I2C*/ #define ARMADILLO5X0_RTC_GPIO IOMUX_TO_GPIO(MX31_PIN_SRXD4) @@ -393,6 +548,17 @@ static void __init armadillo5x0_init(void) if (armadillo5x0_i2c_rtc.irq == 0) pr_warning("armadillo5x0_init: failed to get RTC IRQ\n"); i2c_register_board_info(1, &armadillo5x0_i2c_rtc, 1); + + /* USB */ +#if defined(CONFIG_USB_ULPI) + usbotg_pdata.otg = otg_ulpi_create(&mxc_ulpi_access_ops, + USB_OTG_DRV_VBUS | USB_OTG_DRV_VBUS_EXT); + usbh2_pdata.otg = otg_ulpi_create(&mxc_ulpi_access_ops, + USB_OTG_DRV_VBUS | USB_OTG_DRV_VBUS_EXT); + + mxc_register_device(&mxc_otg_host, &usbotg_pdata); + mxc_register_device(&mxc_usbh2, &usbh2_pdata); +#endif } static void __init armadillo5x0_timer_init(void) -- cgit v0.10.2 From e05c378f4973674a16d5b9636f2310cf88aca5f2 Mon Sep 17 00:00:00 2001 From: Joern Engel Date: Tue, 30 Mar 2010 18:25:17 +0200 Subject: [LogFS] Remove unused method All callers are long gone. Signed-off-by: Joern Engel diff --git a/fs/logfs/logfs.h b/fs/logfs/logfs.h index b84b0ee..97195b9 100644 --- a/fs/logfs/logfs.h +++ b/fs/logfs/logfs.h @@ -305,7 +305,6 @@ typedef int write_alias_t(struct super_block *sb, u64 ino, u64 bix, level_t level, int child_no, __be64 val); struct logfs_block_ops { void (*write_block)(struct logfs_block *block); - gc_level_t (*block_level)(struct logfs_block *block); void (*free_block)(struct super_block *sb, struct logfs_block*block); int (*write_alias)(struct super_block *sb, struct logfs_block *block, diff --git a/fs/logfs/readwrite.c b/fs/logfs/readwrite.c index c3a3a68..3659c37 100644 --- a/fs/logfs/readwrite.c +++ b/fs/logfs/readwrite.c @@ -429,25 +429,6 @@ static void inode_write_block(struct logfs_block *block) } } -static gc_level_t inode_block_level(struct logfs_block *block) -{ - BUG_ON(block->inode->i_ino == LOGFS_INO_MASTER); - return GC_LEVEL(LOGFS_MAX_LEVELS); -} - -static gc_level_t indirect_block_level(struct logfs_block *block) -{ - struct page *page; - struct inode *inode; - u64 bix; - level_t level; - - page = block->page; - inode = page->mapping->host; - logfs_unpack_index(page->index, &bix, &level); - return expand_level(inode->i_ino, level); -} - /* * This silences a false, yet annoying gcc warning. I hate it when my editor * jumps into bitops.h each time I recompile this file. @@ -586,14 +567,12 @@ static void indirect_free_block(struct super_block *sb, static struct logfs_block_ops inode_block_ops = { .write_block = inode_write_block, - .block_level = inode_block_level, .free_block = inode_free_block, .write_alias = inode_write_alias, }; struct logfs_block_ops indirect_block_ops = { .write_block = indirect_write_block, - .block_level = indirect_block_level, .free_block = indirect_free_block, .write_alias = indirect_write_alias, }; diff --git a/fs/logfs/segment.c b/fs/logfs/segment.c index 0ecd8f0..02db22e 100644 --- a/fs/logfs/segment.c +++ b/fs/logfs/segment.c @@ -182,14 +182,8 @@ static int btree_write_alias(struct super_block *sb, struct logfs_block *block, return 0; } -static gc_level_t btree_block_level(struct logfs_block *block) -{ - return expand_level(block->ino, block->level); -} - static struct logfs_block_ops btree_block_ops = { .write_block = btree_write_block, - .block_level = btree_block_level, .free_block = __free_block, .write_alias = btree_write_alias, }; -- cgit v0.10.2 From 8ae06d223f8203c72104e5c0c4ee49a000aedb42 Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Fri, 5 Mar 2010 08:59:32 +0800 Subject: x86-32, resume: do a global tlb flush in S4 resume Colin King reported a strange oops in S4 resume code path (see below). The test system has i5/i7 CPU. The kernel doesn't open PAE, so 4M page table is used. The oops always happen a virtual address 0xc03ff000, which is mapped to the last 4k of first 4M memory. Doing a global tlb flush fixes the issue. EIP: 0060:[] EFLAGS: 00010086 CPU: 0 EIP is at copy_loop+0xe/0x15 EAX: 36aeb000 EBX: 00000000 ECX: 00000400 EDX: f55ad46c ESI: 0f800000 EDI: c03ff000 EBP: f67fbec4 ESP: f67fbea8 DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 ... ... CR2: 00000000c03ff000 Tested-by: Colin Ian King Signed-off-by: Shaohua Li LKML-Reference: <20100305005932.GA22675@sli10-desk.sh.intel.com> Acked-by: Rafael J. Wysocki Signed-off-by: H. Peter Anvin Cc: diff --git a/arch/x86/power/hibernate_asm_32.S b/arch/x86/power/hibernate_asm_32.S index b641388..ad47dae 100644 --- a/arch/x86/power/hibernate_asm_32.S +++ b/arch/x86/power/hibernate_asm_32.S @@ -27,10 +27,17 @@ ENTRY(swsusp_arch_suspend) ret ENTRY(restore_image) + movl mmu_cr4_features, %ecx movl resume_pg_dir, %eax subl $__PAGE_OFFSET, %eax movl %eax, %cr3 + jecxz 1f # cr4 Pentium and higher, skip if zero + andl $~(X86_CR4_PGE), %ecx + movl %ecx, %cr4; # turn off PGE + movl %cr3, %eax; # flush TLB + movl %eax, %cr3 +1: movl restore_pblist, %edx .p2align 4,,7 @@ -54,16 +61,8 @@ done: movl $swapper_pg_dir, %eax subl $__PAGE_OFFSET, %eax movl %eax, %cr3 - /* Flush TLB, including "global" things (vmalloc) */ movl mmu_cr4_features, %ecx jecxz 1f # cr4 Pentium and higher, skip if zero - movl %ecx, %edx - andl $~(X86_CR4_PGE), %edx - movl %edx, %cr4; # turn off PGE -1: - movl %cr3, %eax; # flush TLB - movl %eax, %cr3 - jecxz 1f # cr4 Pentium and higher, skip if zero movl %ecx, %cr4; # turn PGE back on 1: -- cgit v0.10.2 From efd647f744f3cf504ed83580274bd4b6918139fe Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Tue, 30 Mar 2010 13:21:31 +0800 Subject: ocfs2_dlmfs: User DLM_* when decoding file open flags. In commit 0016eedc4185a3cd7e578b027a6e69001b85d6c4, we have changed dlmfs to use stackglue. So when use DLM* when we decode dlm flags from open level. Signed-off-by: Tao Ma Signed-off-by: Joel Becker diff --git a/fs/ocfs2/dlmfs/dlmfs.c b/fs/ocfs2/dlmfs/dlmfs.c index 1b0de15..a99d1ea 100644 --- a/fs/ocfs2/dlmfs/dlmfs.c +++ b/fs/ocfs2/dlmfs/dlmfs.c @@ -112,20 +112,20 @@ MODULE_PARM_DESC(capabilities, DLMFS_CAPABILITIES); * O_RDONLY -> PRMODE level * O_WRONLY -> EXMODE level * - * O_NONBLOCK -> LKM_NOQUEUE + * O_NONBLOCK -> NOQUEUE */ static int dlmfs_decode_open_flags(int open_flags, int *level, int *flags) { if (open_flags & (O_WRONLY|O_RDWR)) - *level = LKM_EXMODE; + *level = DLM_LOCK_EX; else - *level = LKM_PRMODE; + *level = DLM_LOCK_PR; *flags = 0; if (open_flags & O_NONBLOCK) - *flags |= LKM_NOQUEUE; + *flags |= DLM_LKF_NOQUEUE; return 0; } @@ -166,7 +166,7 @@ static int dlmfs_file_open(struct inode *inode, * to be able userspace to be able to distinguish a * valid lock request from one that simply couldn't be * granted. */ - if (flags & LKM_NOQUEUE && status == -EAGAIN) + if (flags & DLM_LKF_NOQUEUE && status == -EAGAIN) status = -ETXTBSY; kfree(fp); goto bail; @@ -193,7 +193,7 @@ static int dlmfs_file_release(struct inode *inode, status = 0; if (fp) { level = fp->fp_lock_level; - if (level != LKM_IVMODE) + if (level != DLM_LOCK_IV) user_dlm_cluster_unlock(&ip->ip_lockres, level); kfree(fp); -- cgit v0.10.2 From a03ab788d070c256eff1ac24cf0e3bc2ca148096 Mon Sep 17 00:00:00 2001 From: Coly Li Date: Fri, 26 Mar 2010 05:15:12 +0800 Subject: ocfs2: one more warning fix in ocfs2_file_aio_write(), v2 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This patch fixes another compiling warning in ocfs2_file_aio_write() like this, fs/ocfs2/file.c: In function ‘ocfs2_file_aio_write’: fs/ocfs2/file.c:2026: warning: suggest parentheses around ‘&&’ within ‘||’ As Joel suggested, '!ret' is unary, this version removes the wrap from '!ret'. Signed-off-by: Coly Li Signed-off-by: Joel Becker diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 17947dc..2b4235c 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -2021,9 +2021,9 @@ out_dio: if (ret < 0) written = ret; - if (!ret && (old_size != i_size_read(inode) || - old_clusters != OCFS2_I(inode)->ip_clusters || - has_refcount)) { + if (!ret && ((old_size != i_size_read(inode)) || + (old_clusters != OCFS2_I(inode)->ip_clusters) || + has_refcount)) { ret = jbd2_journal_force_commit(osb->journal->j_journal); if (ret < 0) written = ret; -- cgit v0.10.2 From 428257f8870f0e72e85ce782d091fa1f366de7df Mon Sep 17 00:00:00 2001 From: Wengang Wang Date: Wed, 24 Mar 2010 22:40:44 +0800 Subject: ocfs2: Check the owner of a lockres inside the spinlock The checking of lockres owner in dlm_update_lvb() is not inside spinlock protection. I don't see problem in current call path of dlm_update_lvb(). But just for code robustness. Signed-off-by: Wengang Wang Signed-off-by: Joel Becker diff --git a/fs/ocfs2/dlm/dlmast.c b/fs/ocfs2/dlm/dlmast.c index dccc439..b7a25ef 100644 --- a/fs/ocfs2/dlm/dlmast.c +++ b/fs/ocfs2/dlm/dlmast.c @@ -185,9 +185,8 @@ static void dlm_update_lvb(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, BUG_ON(!lksb); /* only updates if this node masters the lockres */ + spin_lock(&res->spinlock); if (res->owner == dlm->node_num) { - - spin_lock(&res->spinlock); /* check the lksb flags for the direction */ if (lksb->flags & DLM_LKSB_GET_LVB) { mlog(0, "getting lvb from lockres for %s node\n", @@ -202,8 +201,8 @@ static void dlm_update_lvb(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, * here. In the future we might want to clear it at the time * the put is actually done. */ - spin_unlock(&res->spinlock); } + spin_unlock(&res->spinlock); /* reset any lvb flags on the lksb */ lksb->flags &= ~(DLM_LKSB_PUT_LVB|DLM_LKSB_GET_LVB); -- cgit v0.10.2 From 9358c6d4c0264b1572554c49c4b92673ea9a5c72 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Tue, 30 Mar 2010 13:54:41 -0700 Subject: ceph: fix dentry rehashing on virtual .snap dir If a lookup fails on the magic .snap directory, we bind it to a magic snap directory inode in ceph_lookup_finish(). That code assumes the dentry is unhashed, but a recent server-side change started returning NULL leases on lookup failure, causing the .snap dentry to be hashed and NULL by ceph_fill_trace(). This causes dentry hash chain corruption, or a dies when d_rehash() includes BUG_ON(!d_unhashed(entry)); So, avoid processing the NULL dentry lease if it the dentry matches the snapdir name in ceph_fill_trace(). That allows the lookup completion to properly bind it to the snapdir inode. BUG there if dentry is hashed to be sure. Signed-off-by: Sage Weil diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c index 8a9116e..aed8fda 100644 --- a/fs/ceph/dir.c +++ b/fs/ceph/dir.c @@ -488,6 +488,7 @@ struct dentry *ceph_finish_lookup(struct ceph_mds_request *req, struct inode *inode = ceph_get_snapdir(parent); dout("ENOENT on snapdir %p '%.*s', linking to snapdir %p\n", dentry, dentry->d_name.len, dentry->d_name.name, inode); + BUG_ON(!d_unhashed(dentry)); d_add(dentry, inode); err = 0; } diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index aca82d5..26f883c 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -886,6 +886,7 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req, struct inode *in = NULL; struct ceph_mds_reply_inode *ininfo; struct ceph_vino vino; + struct ceph_client *client = ceph_sb_to_client(sb); int i = 0; int err = 0; @@ -949,7 +950,14 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req, return err; } - if (rinfo->head->is_dentry && !req->r_aborted) { + /* + * ignore null lease/binding on snapdir ENOENT, or else we + * will have trouble splicing in the virtual snapdir later + */ + if (rinfo->head->is_dentry && !req->r_aborted && + (rinfo->head->is_target || strncmp(req->r_dentry->d_name.name, + client->mount_args->snapdir_name, + req->r_dentry->d_name.len))) { /* * lookup link rename : null -> possibly existing inode * mknod symlink mkdir : null -> new inode -- cgit v0.10.2 From 753234007f4ac2c96921cfb19ec1ba535ac29790 Mon Sep 17 00:00:00 2001 From: Li Hong Date: Wed, 31 Mar 2010 15:41:00 +0800 Subject: nilfs2: fix a wrong type conversion in nilfs_ioctl() (void * __user *) should be (void __user *) Signed-off-by: Li Hong Signed-off-by: Ryusuke Konishi diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c index 313d0a2..c446017 100644 --- a/fs/nilfs2/ioctl.c +++ b/fs/nilfs2/ioctl.c @@ -648,7 +648,7 @@ static int nilfs_ioctl_get_info(struct inode *inode, struct file *filp, long nilfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { struct inode *inode = filp->f_dentry->d_inode; - void __user *argp = (void * __user *)arg; + void __user *argp = (void __user *)arg; switch (cmd) { case NILFS_IOCTL_CHANGE_CPMODE: -- cgit v0.10.2 From b8e80cf386419453678b01bef830f53445ebb15d Mon Sep 17 00:00:00 2001 From: Daniel T Chen Date: Tue, 30 Mar 2010 13:29:28 -0400 Subject: ALSA: hda: Fix 0 dB offset for Lenovo Thinkpad models using AD1981 BugLink: https://launchpad.net/bugs/551606 The OR's hardware distorts at PCM 100% because it does not correspond to 0 dB. Fix this in patch_ad1981() for all models using the Thinkpad quirk. Reported-by: Jane Silber Cc: Signed-off-by: Daniel T Chen Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_analog.c b/sound/pci/hda/patch_analog.c index e6d1bdf..af34606 100644 --- a/sound/pci/hda/patch_analog.c +++ b/sound/pci/hda/patch_analog.c @@ -1896,6 +1896,14 @@ static int patch_ad1981(struct hda_codec *codec) case AD1981_THINKPAD: spec->mixers[0] = ad1981_thinkpad_mixers; spec->input_mux = &ad1981_thinkpad_capture_source; + /* set the upper-limit for mixer amp to 0dB for avoiding the + * possible damage by overloading + */ + snd_hda_override_amp_caps(codec, 0x11, HDA_INPUT, + (0x17 << AC_AMPCAP_OFFSET_SHIFT) | + (0x17 << AC_AMPCAP_NUM_STEPS_SHIFT) | + (0x05 << AC_AMPCAP_STEP_SIZE_SHIFT) | + (1 << AC_AMPCAP_MUTE_SHIFT)); break; case AD1981_TOSHIBA: spec->mixers[0] = ad1981_hp_mixers; -- cgit v0.10.2 From d5647c18eac9a313a89a3049a9fff51b20021a44 Mon Sep 17 00:00:00 2001 From: Thara Gopinath Date: Wed, 31 Mar 2010 04:16:29 -0600 Subject: OMAP: HWMOD: Adding clockdomain check This patch adds check for presence of clockdomain structure in the API omap_hwmod_get_pwrdm before trying to access the powerdomain structure. This will prevent unnecessary crashing of the system in case of a clock node with out an associated clockdomain. Signed-off-by: Thara Gopinath Signed-off-by: Paul Walmsley diff --git a/arch/arm/mach-omap2/omap_hwmod.c b/arch/arm/mach-omap2/omap_hwmod.c index c664947..e436dcb 100644 --- a/arch/arm/mach-omap2/omap_hwmod.c +++ b/arch/arm/mach-omap2/omap_hwmod.c @@ -1511,6 +1511,9 @@ struct powerdomain *omap_hwmod_get_pwrdm(struct omap_hwmod *oh) c = oh->slaves[oh->_mpu_port_index]->_clk; } + if (!c->clkdm) + return NULL; + return c->clkdm->pwrdm.ptr; } -- cgit v0.10.2 From 56dc79aba14bad1b73508df9e210e93bbd81a577 Mon Sep 17 00:00:00 2001 From: Thara Gopinath Date: Wed, 31 Mar 2010 04:16:29 -0600 Subject: OMAP: HWMOD: Bug fixes in hwmod structure definitions This patch corrects the width of sysc_flags in hwmod sysconfig structure where the values to be stored to this variable exceed the current field width. Signed-off-by: Thara Gopinath [paul@pwsan.com: edited to apply; rearranged structure members to pack] Signed-off-by: Paul Walmsley diff --git a/arch/arm/plat-omap/include/plat/omap_hwmod.h b/arch/arm/plat-omap/include/plat/omap_hwmod.h index 440b416..36d6ea5 100644 --- a/arch/arm/plat-omap/include/plat/omap_hwmod.h +++ b/arch/arm/plat-omap/include/plat/omap_hwmod.h @@ -294,8 +294,8 @@ struct omap_hwmod_class_sysconfig { u16 rev_offs; u16 sysc_offs; u16 syss_offs; + u16 sysc_flags; u8 idlemodes; - u8 sysc_flags; u8 clockact; struct omap_hwmod_sysc_fields *sysc_fields; }; -- cgit v0.10.2 From 766d305fead341889e7b9611fdc97236075a29fb Mon Sep 17 00:00:00 2001 From: Rajendra Nayak Date: Wed, 31 Mar 2010 04:16:30 -0600 Subject: OMAP4: prcm: Use logical OR instead of bitwise OR This patch fixes usage of bitwise OR in if conditions, and instead uses logical OR. Signed-off-by: Rajendra Nayak Cc: Abhijit Pagare Cc: Tony Lindgren Signed-off-by: Paul Walmsley diff --git a/arch/arm/mach-omap2/clockdomain.c b/arch/arm/mach-omap2/clockdomain.c index b87ad66..6e568ec 100644 --- a/arch/arm/mach-omap2/clockdomain.c +++ b/arch/arm/mach-omap2/clockdomain.c @@ -240,7 +240,7 @@ static void _omap2_clkdm_set_hwsup(struct clockdomain *clkdm, int enable) bits = OMAP24XX_CLKSTCTRL_ENABLE_AUTO; else bits = OMAP24XX_CLKSTCTRL_DISABLE_AUTO; - } else if (cpu_is_omap34xx() | cpu_is_omap44xx()) { + } else if (cpu_is_omap34xx() || cpu_is_omap44xx()) { if (enable) bits = OMAP34XX_CLKSTCTRL_ENABLE_AUTO; else @@ -812,7 +812,7 @@ int omap2_clkdm_sleep(struct clockdomain *clkdm) cm_set_mod_reg_bits(OMAP24XX_FORCESTATE, clkdm->pwrdm.ptr->prcm_offs, OMAP2_PM_PWSTCTRL); - } else if (cpu_is_omap34xx() | cpu_is_omap44xx()) { + } else if (cpu_is_omap34xx() || cpu_is_omap44xx()) { u32 bits = (OMAP34XX_CLKSTCTRL_FORCE_SLEEP << __ffs(clkdm->clktrctrl_mask)); @@ -856,7 +856,7 @@ int omap2_clkdm_wakeup(struct clockdomain *clkdm) cm_clear_mod_reg_bits(OMAP24XX_FORCESTATE, clkdm->pwrdm.ptr->prcm_offs, OMAP2_PM_PWSTCTRL); - } else if (cpu_is_omap34xx() | cpu_is_omap44xx()) { + } else if (cpu_is_omap34xx() || cpu_is_omap44xx()) { u32 bits = (OMAP34XX_CLKSTCTRL_FORCE_WAKEUP << __ffs(clkdm->clktrctrl_mask)); diff --git a/arch/arm/mach-omap2/powerdomain.c b/arch/arm/mach-omap2/powerdomain.c index 9a0fb38..ebfce7d 100644 --- a/arch/arm/mach-omap2/powerdomain.c +++ b/arch/arm/mach-omap2/powerdomain.c @@ -222,7 +222,7 @@ void pwrdm_init(struct powerdomain **pwrdm_list) { struct powerdomain **p = NULL; - if (cpu_is_omap24xx() | cpu_is_omap34xx()) { + if (cpu_is_omap24xx() || cpu_is_omap34xx()) { pwrstctrl_reg_offs = OMAP2_PM_PWSTCTRL; pwrstst_reg_offs = OMAP2_PM_PWSTST; } else if (cpu_is_omap44xx()) { diff --git a/arch/arm/mach-omap2/prcm.c b/arch/arm/mach-omap2/prcm.c index 9537f6f..07a60f1 100644 --- a/arch/arm/mach-omap2/prcm.c +++ b/arch/arm/mach-omap2/prcm.c @@ -123,7 +123,7 @@ struct omap3_prcm_regs prcm_context; u32 omap_prcm_get_reset_sources(void) { /* XXX This presumably needs modification for 34XX */ - if (cpu_is_omap24xx() | cpu_is_omap34xx()) + if (cpu_is_omap24xx() || cpu_is_omap34xx()) return prm_read_mod_reg(WKUP_MOD, OMAP2_RM_RSTST) & 0x7f; if (cpu_is_omap44xx()) return prm_read_mod_reg(WKUP_MOD, OMAP4_RM_RSTST) & 0x7f; @@ -157,7 +157,7 @@ void omap_prcm_arch_reset(char mode, const char *cmd) else WARN_ON(1); - if (cpu_is_omap24xx() | cpu_is_omap34xx()) + if (cpu_is_omap24xx() || cpu_is_omap34xx()) prm_set_mod_reg_bits(OMAP_RST_DPLL3, prcm_offs, OMAP2_RM_RSTCTRL); if (cpu_is_omap44xx()) -- cgit v0.10.2 From d54a45e2533ef33678dc340298b022a289d2b3e3 Mon Sep 17 00:00:00 2001 From: Ranjith Lohithakshan Date: Wed, 31 Mar 2010 04:16:30 -0600 Subject: OMAP3: clock: fix enable bit used for dpll4_m4x2 clock The enable bit for dpll4_m4x2 clock should be OMAP3430_PWRDN_DSS1_SHIFT. The code erroneously uses OMAP3430_PWRDN_CAM_SHIFT which is meant for dpll4_m5x2 clock. This came into notice during a recent review of the clock tree. Signed-off-by: Ranjith Lohithakshan Signed-off-by: Paul Walmsley diff --git a/arch/arm/mach-omap2/clock3xxx_data.c b/arch/arm/mach-omap2/clock3xxx_data.c index d5153b6..9cba556 100644 --- a/arch/arm/mach-omap2/clock3xxx_data.c +++ b/arch/arm/mach-omap2/clock3xxx_data.c @@ -895,7 +895,7 @@ static struct clk dpll4_m4x2_ck = { .ops = &clkops_omap2_dflt_wait, .parent = &dpll4_m4_ck, .enable_reg = OMAP_CM_REGADDR(PLL_MOD, CM_CLKEN), - .enable_bit = OMAP3430_PWRDN_CAM_SHIFT, + .enable_bit = OMAP3430_PWRDN_DSS1_SHIFT, .flags = INVERT_ENABLE, .clkdm_name = "dpll4_clkdm", .recalc = &omap3_clkoutx2_recalc, -- cgit v0.10.2 From 499bca9b6d3243f9278a1f5a22d00e67acdd844d Mon Sep 17 00:00:00 2001 From: Amerigo Wang Date: Thu, 4 Mar 2010 03:23:46 -0500 Subject: [CPUFREQ] fix a lockdep warning There is no need to do sysfs_remove_link() or kobject_put() etc. when policy_rwsem_write is held, move them after releasing the lock. This fixes the lockdep warning: halt/4071 is trying to acquire lock: (s_active){++++.+}, at: [] .sysfs_addrm_finish+0x58/0xc0 but task is already holding lock: (&per_cpu(cpu_policy_rwsem, cpu)){+.+.+.}, at: [] .lock_policy_rwsem_write+0x84/0xf4 Reported-by: Benjamin Herrenschmidt Signed-off-by: WANG Cong Cc: Johannes Berg Cc: Venkatesh Pallipadi Signed-off-by: Dave Jones diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 2d5d575..75d293e 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -1113,6 +1113,8 @@ static int __cpufreq_remove_dev(struct sys_device *sys_dev) unsigned int cpu = sys_dev->id; unsigned long flags; struct cpufreq_policy *data; + struct kobject *kobj; + struct completion *cmp; #ifdef CONFIG_SMP struct sys_device *cpu_sys_dev; unsigned int j; @@ -1141,10 +1143,11 @@ static int __cpufreq_remove_dev(struct sys_device *sys_dev) dprintk("removing link\n"); cpumask_clear_cpu(cpu, data->cpus); spin_unlock_irqrestore(&cpufreq_driver_lock, flags); - sysfs_remove_link(&sys_dev->kobj, "cpufreq"); + kobj = &sys_dev->kobj; cpufreq_cpu_put(data); cpufreq_debug_enable_ratelimit(); unlock_policy_rwsem_write(cpu); + sysfs_remove_link(kobj, "cpufreq"); return 0; } #endif @@ -1181,7 +1184,10 @@ static int __cpufreq_remove_dev(struct sys_device *sys_dev) data->governor->name, CPUFREQ_NAME_LEN); #endif cpu_sys_dev = get_cpu_sysdev(j); - sysfs_remove_link(&cpu_sys_dev->kobj, "cpufreq"); + kobj = &cpu_sys_dev->kobj; + unlock_policy_rwsem_write(cpu); + sysfs_remove_link(kobj, "cpufreq"); + lock_policy_rwsem_write(cpu); cpufreq_cpu_put(data); } } @@ -1192,19 +1198,22 @@ static int __cpufreq_remove_dev(struct sys_device *sys_dev) if (cpufreq_driver->target) __cpufreq_governor(data, CPUFREQ_GOV_STOP); - kobject_put(&data->kobj); + kobj = &data->kobj; + cmp = &data->kobj_unregister; + unlock_policy_rwsem_write(cpu); + kobject_put(kobj); /* we need to make sure that the underlying kobj is actually * not referenced anymore by anybody before we proceed with * unloading. */ dprintk("waiting for dropping of refcount\n"); - wait_for_completion(&data->kobj_unregister); + wait_for_completion(cmp); dprintk("wait complete\n"); + lock_policy_rwsem_write(cpu); if (cpufreq_driver->exit) cpufreq_driver->exit(data); - unlock_policy_rwsem_write(cpu); free_cpumask_var(data->related_cpus); -- cgit v0.10.2 From fd187aaf980c45f1d16a94a846faa68e24de03c8 Mon Sep 17 00:00:00 2001 From: Dominik Brodowski Date: Fri, 26 Mar 2010 10:01:34 +0100 Subject: [CPUFREQ] use max load in conservative governor Instead of using the load of the last CPU in a package, use the maximum load of all CPUs in a package. Reported-by: Jean-Christian Goussard Signed-off-by: Dominik Brodowski Signed-off-by: Dave Jones diff --git a/drivers/cpufreq/cpufreq_conservative.c b/drivers/cpufreq/cpufreq_conservative.c index 599a40b..3a14787 100644 --- a/drivers/cpufreq/cpufreq_conservative.c +++ b/drivers/cpufreq/cpufreq_conservative.c @@ -444,6 +444,7 @@ static struct attribute_group dbs_attr_group_old = { static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info) { unsigned int load = 0; + unsigned int max_load = 0; unsigned int freq_target; struct cpufreq_policy *policy; @@ -501,6 +502,9 @@ static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info) continue; load = 100 * (wall_time - idle_time) / wall_time; + + if (load > max_load) + max_load = load; } /* @@ -511,7 +515,7 @@ static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info) return; /* Check for frequency increase */ - if (load > dbs_tuners_ins.up_threshold) { + if (max_load > dbs_tuners_ins.up_threshold) { this_dbs_info->down_skip = 0; /* if we are already at full speed then break out early */ @@ -538,7 +542,7 @@ static void dbs_check_cpu(struct cpu_dbs_info_s *this_dbs_info) * can support the current CPU usage without triggering the up * policy. To be safe, we focus 10 points under the threshold. */ - if (load < (dbs_tuners_ins.down_threshold - 10)) { + if (max_load < (dbs_tuners_ins.down_threshold - 10)) { freq_target = (dbs_tuners_ins.freq_step * policy->max) / 100; this_dbs_info->requested_freq -= freq_target; -- cgit v0.10.2 From 3e340c05c0def3bb68db6751299b7821c2ba0621 Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Thu, 11 Mar 2010 16:17:28 +0000 Subject: IB/cm: Fix device_create() return value check Use IS_ERR() instead of comparing to NULL. Signed-off-by: Jani Nikula Signed-off-by: Roland Dreier diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c index 764787e..c9730cb 100644 --- a/drivers/infiniband/core/cm.c +++ b/drivers/infiniband/core/cm.c @@ -3693,7 +3693,7 @@ static void cm_add_one(struct ib_device *ib_device) cm_dev->device = device_create(&cm_class, &ib_device->dev, MKDEV(0, 0), NULL, "%s", ib_device->name); - if (!cm_dev->device) { + if (IS_ERR(cm_dev->device)) { kfree(cm_dev); return; } -- cgit v0.10.2 From 9f3a5f52aa63d3aa4c64a7245153549bb66bad8c Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Mon, 29 Mar 2010 22:38:29 -0700 Subject: x86: Make e820_remove_range to handle all covered case Rusty found on lguest with trim_bios_range, max_pfn is not right anymore, and looks e820_remove_range does not work right. [ 0.000000] BIOS-provided physical RAM map: [ 0.000000] LGUEST: 0000000000000000 - 0000000004000000 (usable) [ 0.000000] Notice: NX (Execute Disable) protection missing in CPU or disabled in BIOS! [ 0.000000] DMI not present or invalid. [ 0.000000] last_pfn = 0x3fa0 max_arch_pfn = 0x100000 [ 0.000000] init_memory_mapping: 0000000000000000-0000000003fa0000 root cause is: the e820_remove_range doesn't handle the all covered case. e820_remove_range(BIOS_START, BIOS_END - BIOS_START, ...) produces a bogus range as a result. Make it match e820_update_range() by handling that case too. Reported-by: Rusty Russell Signed-off-by: Yinghai Lu Tested-by: Rusty Russell LKML-Reference: <4BB18E55.6090903@kernel.org> Signed-off-by: H. Peter Anvin diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 740b440..7bca3c6 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -519,29 +519,45 @@ u64 __init e820_remove_range(u64 start, u64 size, unsigned old_type, printk(KERN_DEBUG "e820 remove range: %016Lx - %016Lx ", (unsigned long long) start, (unsigned long long) end); - e820_print_type(old_type); + if (checktype) + e820_print_type(old_type); printk(KERN_CONT "\n"); for (i = 0; i < e820.nr_map; i++) { struct e820entry *ei = &e820.map[i]; u64 final_start, final_end; + u64 ei_end; if (checktype && ei->type != old_type) continue; + + ei_end = ei->addr + ei->size; /* totally covered? */ - if (ei->addr >= start && - (ei->addr + ei->size) <= (start + size)) { + if (ei->addr >= start && ei_end <= end) { real_removed_size += ei->size; memset(ei, 0, sizeof(struct e820entry)); continue; } + + /* new range is totally covered? */ + if (ei->addr < start && ei_end > end) { + e820_add_region(end, ei_end - end, ei->type); + ei->size = start - ei->addr; + real_removed_size += size; + continue; + } + /* partially covered */ final_start = max(start, ei->addr); - final_end = min(start + size, ei->addr + ei->size); + final_end = min(end, ei_end); if (final_start >= final_end) continue; real_removed_size += final_end - final_start; + /* + * left range could be head or tail, so need to update + * size at first. + */ ei->size -= final_end - final_start; if (ei->addr < final_start) continue; -- cgit v0.10.2 From a42ab8e1a37257da37e0f018e707bf365ac24531 Mon Sep 17 00:00:00 2001 From: Joel Becker Date: Wed, 31 Mar 2010 18:25:44 -0700 Subject: ocfs2: Compute metaecc for superblocks during online resize. Online resize writes out the new superblock and its backups directly. The metaecc data wasn't being recomputed. Let's do that directly. Signed-off-by: Joel Becker Acked-by: Mark Fasheh [ Cc: stable@kernel.org diff --git a/fs/ocfs2/buffer_head_io.c b/fs/ocfs2/buffer_head_io.c index 21c808f..b18c6d6 100644 --- a/fs/ocfs2/buffer_head_io.c +++ b/fs/ocfs2/buffer_head_io.c @@ -407,6 +407,7 @@ int ocfs2_write_super_or_backup(struct ocfs2_super *osb, struct buffer_head *bh) { int ret = 0; + struct ocfs2_dinode *di = (struct ocfs2_dinode *)bh->b_data; mlog_entry_void(); @@ -426,6 +427,7 @@ int ocfs2_write_super_or_backup(struct ocfs2_super *osb, get_bh(bh); /* for end_buffer_write_sync() */ bh->b_end_io = end_buffer_write_sync; + ocfs2_compute_meta_ecc(osb->sb, bh->b_data, &di->i_check); submit_bh(WRITE, bh); wait_on_buffer(bh); -- cgit v0.10.2 From 4bdde044dc36ac7b01f7502394d52619af9d1927 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 24 Mar 2010 10:58:05 +0800 Subject: tracing: Convert some signal events to DEFINE_TRACE Use DECLARE_EVENT_CLASS to remove duplicate code: text data bss dec hex filename 23639 6084 8 29731 7423 kernel/signal.o.orig 22727 6084 8 28819 7093 kernel/signal.o 2 events are converted: signal_queue_overflow: signal_overflow_fail, signal_lose_info No functional change. Acked-by: Masami Hiramatsu Signed-off-by: Li Zefan LKML-Reference: <4BA97FBD.8070703@cn.fujitsu.com> Signed-off-by: Steven Rostedt diff --git a/include/trace/events/signal.h b/include/trace/events/signal.h index a510b75..814566c 100644 --- a/include/trace/events/signal.h +++ b/include/trace/events/signal.h @@ -100,18 +100,7 @@ TRACE_EVENT(signal_deliver, __entry->sa_handler, __entry->sa_flags) ); -/** - * signal_overflow_fail - called when signal queue is overflow - * @sig: signal number - * @group: signal to process group or not (bool) - * @info: pointer to struct siginfo - * - * Kernel fails to generate 'sig' signal with 'info' siginfo, because - * siginfo queue is overflow, and the signal is dropped. - * 'group' is not 0 if the signal will be sent to a process group. - * 'sig' is always one of RT signals. - */ -TRACE_EVENT(signal_overflow_fail, +DECLARE_EVENT_CLASS(signal_queue_overflow, TP_PROTO(int sig, int group, struct siginfo *info), @@ -135,6 +124,24 @@ TRACE_EVENT(signal_overflow_fail, ); /** + * signal_overflow_fail - called when signal queue is overflow + * @sig: signal number + * @group: signal to process group or not (bool) + * @info: pointer to struct siginfo + * + * Kernel fails to generate 'sig' signal with 'info' siginfo, because + * siginfo queue is overflow, and the signal is dropped. + * 'group' is not 0 if the signal will be sent to a process group. + * 'sig' is always one of RT signals. + */ +DEFINE_EVENT(signal_queue_overflow, signal_overflow_fail, + + TP_PROTO(int sig, int group, struct siginfo *info), + + TP_ARGS(sig, group, info) +); + +/** * signal_lose_info - called when siginfo is lost * @sig: signal number * @group: signal to process group or not (bool) @@ -145,28 +152,13 @@ TRACE_EVENT(signal_overflow_fail, * 'group' is not 0 if the signal will be sent to a process group. * 'sig' is always one of non-RT signals. */ -TRACE_EVENT(signal_lose_info, +DEFINE_EVENT(signal_queue_overflow, signal_lose_info, TP_PROTO(int sig, int group, struct siginfo *info), - TP_ARGS(sig, group, info), - - TP_STRUCT__entry( - __field( int, sig ) - __field( int, group ) - __field( int, errno ) - __field( int, code ) - ), - - TP_fast_assign( - __entry->sig = sig; - __entry->group = group; - TP_STORE_SIGINFO(__entry, info); - ), - - TP_printk("sig=%d group=%d errno=%d code=%d", - __entry->sig, __entry->group, __entry->errno, __entry->code) + TP_ARGS(sig, group, info) ); + #endif /* _TRACE_SIGNAL_H */ /* This part must be outside protection */ -- cgit v0.10.2 From 50354a8a28d0c91695a2d6d25b5a821bfe557a07 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 24 Mar 2010 10:58:24 +0800 Subject: tracing: Update comments Make some comments consistent with the code. Signed-off-by: Li Zefan LKML-Reference: <4BA97FD0.7090202@cn.fujitsu.com> Signed-off-by: Steven Rostedt diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index ea6f9d4..75dd778 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -154,9 +154,11 @@ * * field = (typeof(field))entry; * - * p = get_cpu_var(ftrace_event_seq); + * p = &get_cpu_var(ftrace_event_seq); * trace_seq_init(p); - * ret = trace_seq_printf(s, "\n"); + * ret = trace_seq_printf(s, "%s: ", ); + * if (ret) + * ret = trace_seq_printf(s, "\n"); * put_cpu(); * if (!ret) * return TRACE_TYPE_PARTIAL_LINE; @@ -450,38 +452,38 @@ perf_trace_disable_##name(struct ftrace_event_call *unused) \ * * static void ftrace_raw_event_(proto) * { + * struct ftrace_data_offsets_ __maybe_unused __data_offsets; * struct ring_buffer_event *event; * struct ftrace_raw_ *entry; <-- defined in stage 1 * struct ring_buffer *buffer; * unsigned long irq_flags; + * int __data_size; * int pc; * * local_save_flags(irq_flags); * pc = preempt_count(); * + * __data_size = ftrace_get_offsets_(&__data_offsets, args); + * * event = trace_current_buffer_lock_reserve(&buffer, * event_.id, - * sizeof(struct ftrace_raw_), + * sizeof(*entry) + __data_size, * irq_flags, pc); * if (!event) * return; * entry = ring_buffer_event_data(event); * - * ; <-- Here we assign the entries by the __field and - * __array macros. + * { ; } <-- Here we assign the entries by the __field and + * __array macros. * - * trace_current_buffer_unlock_commit(buffer, event, irq_flags, pc); + * if (!filter_current_check_discard(buffer, event_call, entry, event)) + * trace_current_buffer_unlock_commit(buffer, + * event, irq_flags, pc); * } * * static int ftrace_raw_reg_event_(struct ftrace_event_call *unused) * { - * int ret; - * - * ret = register_trace_(ftrace_raw_event_); - * if (!ret) - * pr_info("event trace: Could not activate trace point " - * "probe to "); - * return ret; + * return register_trace_(ftrace_raw_event_); * } * * static void ftrace_unreg_event_(struct ftrace_event_call *unused) @@ -493,6 +495,8 @@ perf_trace_disable_##name(struct ftrace_event_call *unused) \ * .trace = ftrace_raw_output_, <-- stage 2 * }; * + * static const char print_fmt_[] = ; + * * static struct ftrace_event_call __used * __attribute__((__aligned__(4))) * __attribute__((section("_ftrace_events"))) event_ = { @@ -501,6 +505,8 @@ perf_trace_disable_##name(struct ftrace_event_call *unused) \ * .raw_init = trace_event_raw_init, * .regfunc = ftrace_reg_event_, * .unregfunc = ftrace_unreg_event_, + * .print_fmt = print_fmt_, + * .define_fields = ftrace_define_fields_, * } * */ @@ -569,7 +575,6 @@ ftrace_raw_event_id_##call(struct ftrace_event_call *event_call, \ return; \ entry = ring_buffer_event_data(event); \ \ - \ tstruct \ \ { assign; } \ -- cgit v0.10.2 From ae832d1e03ac9bf09fb8a07fb37908ab40c7cd0e Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 24 Mar 2010 10:57:43 +0800 Subject: tracing: Remove side effect from module tracepoints that caused a GPF Remove the @refcnt argument, because it has side-effects, and arguments with side-effects are not skipped by the jump over disabled instrumentation and are executed even when the tracepoint is disabled. This was also causing a GPF as found by Randy Dunlap: Subject: 2.6.33 GP fault only when built with tracing LKML-Reference: <4BA2B69D.3000309@oracle.com> Note, the current 2.6.34-rc has a fix for the actual cause of the GPF, but this fixes one of its triggers. Tested-by: Randy Dunlap Acked-by: Mathieu Desnoyers Signed-off-by: Li Zefan LKML-Reference: <4BA97FA7.6040406@cn.fujitsu.com> Signed-off-by: Steven Rostedt diff --git a/include/linux/module.h b/include/linux/module.h index 5e869ff..393ec39 100644 --- a/include/linux/module.h +++ b/include/linux/module.h @@ -460,8 +460,7 @@ static inline void __module_get(struct module *module) if (module) { preempt_disable(); __this_cpu_inc(module->refptr->count); - trace_module_get(module, _THIS_IP_, - __this_cpu_read(module->refptr->count)); + trace_module_get(module, _THIS_IP_); preempt_enable(); } } @@ -475,8 +474,7 @@ static inline int try_module_get(struct module *module) if (likely(module_is_live(module))) { __this_cpu_inc(module->refptr->count); - trace_module_get(module, _THIS_IP_, - __this_cpu_read(module->refptr->count)); + trace_module_get(module, _THIS_IP_); } else ret = 0; diff --git a/include/trace/events/module.h b/include/trace/events/module.h index 4b0f48b..a585f81 100644 --- a/include/trace/events/module.h +++ b/include/trace/events/module.h @@ -53,9 +53,9 @@ TRACE_EVENT(module_free, DECLARE_EVENT_CLASS(module_refcnt, - TP_PROTO(struct module *mod, unsigned long ip, int refcnt), + TP_PROTO(struct module *mod, unsigned long ip), - TP_ARGS(mod, ip, refcnt), + TP_ARGS(mod, ip), TP_STRUCT__entry( __field( unsigned long, ip ) @@ -65,7 +65,7 @@ DECLARE_EVENT_CLASS(module_refcnt, TP_fast_assign( __entry->ip = ip; - __entry->refcnt = refcnt; + __entry->refcnt = __this_cpu_read(mod->refptr->count); __assign_str(name, mod->name); ), @@ -75,16 +75,16 @@ DECLARE_EVENT_CLASS(module_refcnt, DEFINE_EVENT(module_refcnt, module_get, - TP_PROTO(struct module *mod, unsigned long ip, int refcnt), + TP_PROTO(struct module *mod, unsigned long ip), - TP_ARGS(mod, ip, refcnt) + TP_ARGS(mod, ip) ); DEFINE_EVENT(module_refcnt, module_put, - TP_PROTO(struct module *mod, unsigned long ip, int refcnt), + TP_PROTO(struct module *mod, unsigned long ip), - TP_ARGS(mod, ip, refcnt) + TP_ARGS(mod, ip) ); TRACE_EVENT(module_request, diff --git a/kernel/module.c b/kernel/module.c index c968d36..21591ad 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -800,8 +800,7 @@ void module_put(struct module *module) preempt_disable(); __this_cpu_dec(module->refptr->count); - trace_module_put(module, _RET_IP_, - __this_cpu_read(module->refptr->count)); + trace_module_put(module, _RET_IP_); /* Maybe they're waiting for us to drop reference? */ if (unlikely(!module_is_live(module))) wake_up_process(module->waiter); -- cgit v0.10.2 From eb0c53771fb2f5f66b0edb3ebce33be4bbf1c285 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Mon, 29 Mar 2010 14:25:18 -0400 Subject: tracing: Fix compile error in module tracepoints when MODULE_UNLOAD not set If modules are configured in the build but unloading of modules is not, then the refcnt is not defined. Place the get/put module tracepoints under CONFIG_MODULE_UNLOAD since it references this field in the module structure. As a side-effect, this patch also reduces the code when MODULE_UNLOAD is not set, because these unused tracepoints are not created. Signed-off-by: Steven Rostedt diff --git a/include/trace/events/module.h b/include/trace/events/module.h index a585f81..f07b44a 100644 --- a/include/trace/events/module.h +++ b/include/trace/events/module.h @@ -51,6 +51,9 @@ TRACE_EVENT(module_free, TP_printk("%s", __get_str(name)) ); +#ifdef CONFIG_MODULE_UNLOAD +/* trace_module_get/put are only used if CONFIG_MODULE_UNLOAD is defined */ + DECLARE_EVENT_CLASS(module_refcnt, TP_PROTO(struct module *mod, unsigned long ip), @@ -86,6 +89,7 @@ DEFINE_EVENT(module_refcnt, module_put, TP_ARGS(mod, ip) ); +#endif /* CONFIG_MODULE_UNLOAD */ TRACE_EVENT(module_request, diff --git a/kernel/module.c b/kernel/module.c index 21591ad..d9e2379 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -59,8 +59,6 @@ #define CREATE_TRACE_POINTS #include -EXPORT_TRACEPOINT_SYMBOL(module_get); - #if 0 #define DEBUGP printk #else @@ -467,6 +465,9 @@ MODINFO_ATTR(srcversion); static char last_unloaded_module[MODULE_NAME_LEN+1]; #ifdef CONFIG_MODULE_UNLOAD + +EXPORT_TRACEPOINT_SYMBOL(module_get); + /* Init the unload section of the module. */ static void module_unload_init(struct module *mod) { -- cgit v0.10.2 From 66a8cb95ed04025664d1db4e952155ee1dccd048 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 31 Mar 2010 13:21:56 -0400 Subject: ring-buffer: Add place holder recording of dropped events Currently, when the ring buffer drops events, it does not record the fact that it did so. It does inform the writer that the event was dropped by returning a NULL event, but it does not put in any place holder where the event was dropped. This is not a trivial thing to add because the ring buffer mostly runs in overwrite (flight recorder) mode. That is, when the ring buffer is full, new data will overwrite old data. In a produce/consumer mode, where new data is simply dropped when the ring buffer is full, it is trivial to add the placeholder for dropped events. When there's more room to write new data, then a special event can be added to notify the reader about the dropped events. But in overwrite mode, any new write can overwrite events. A place holder can not be inserted into the ring buffer since there never may be room. A reader could also come in at anytime and miss the placeholder. Luckily, the way the ring buffer works, the read side can find out if events were lost or not, and how many events. Everytime a write takes place, if it overwrites the header page (the next read) it updates a "overrun" variable that keeps track of the number of lost events. When a reader swaps out a page from the ring buffer, it can record this number, perfom the swap, and then check to see if the number changed, and take the diff if it has, which would be the number of events dropped. This can be stored by the reader and returned to callers of the reader. Since the reader page swap will fail if the writer moved the head page since the time the reader page set up the swap, this gives room to record the overruns without worrying about races. If the reader sets up the pages, records the overrun, than performs the swap, if the swap succeeds, then the overrun variable has not been updated since the setup before the swap. For binary readers of the ring buffer, a flag is set in the header of each sub page (sub buffer) of the ring buffer. This flag is embedded in the size field of the data on the sub buffer, in the 31st bit (the size can be 32 or 64 bits depending on the architecture), but only 27 bits needs to be used for the actual size (less actually). We could add a new field in the sub buffer header to also record the number of events dropped since the last read, but this will change the format of the binary ring buffer a bit too much. Perhaps this change can be made if the information on the number of events dropped is considered important enough. Note, the notification of dropped events is only used by consuming reads or peeking at the ring buffer. Iterating over the ring buffer does not keep this information because the necessary data is only available when a page swap is made, and the iterator does not swap out pages. Cc: Robert Richter Cc: Andi Kleen Cc: Li Zefan Cc: Arnaldo Carvalho de Melo Cc: "Luis Claudio R. Goncalves" Cc: Frederic Weisbecker Signed-off-by: Steven Rostedt diff --git a/drivers/oprofile/cpu_buffer.c b/drivers/oprofile/cpu_buffer.c index 166b67e..7581dbe 100644 --- a/drivers/oprofile/cpu_buffer.c +++ b/drivers/oprofile/cpu_buffer.c @@ -186,14 +186,14 @@ int op_cpu_buffer_write_commit(struct op_entry *entry) struct op_sample *op_cpu_buffer_read_entry(struct op_entry *entry, int cpu) { struct ring_buffer_event *e; - e = ring_buffer_consume(op_ring_buffer_read, cpu, NULL); + e = ring_buffer_consume(op_ring_buffer_read, cpu, NULL, NULL); if (e) goto event; if (ring_buffer_swap_cpu(op_ring_buffer_read, op_ring_buffer_write, cpu)) return NULL; - e = ring_buffer_consume(op_ring_buffer_read, cpu, NULL); + e = ring_buffer_consume(op_ring_buffer_read, cpu, NULL, NULL); if (e) goto event; return NULL; diff --git a/include/linux/ring_buffer.h b/include/linux/ring_buffer.h index 5fcc31e..c829776 100644 --- a/include/linux/ring_buffer.h +++ b/include/linux/ring_buffer.h @@ -120,9 +120,11 @@ int ring_buffer_write(struct ring_buffer *buffer, unsigned long length, void *data); struct ring_buffer_event * -ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts); +ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts, + unsigned long *lost_events); struct ring_buffer_event * -ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts); +ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts, + unsigned long *lost_events); struct ring_buffer_iter * ring_buffer_read_start(struct ring_buffer *buffer, int cpu); diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index d1187ef..8295650 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -318,6 +318,9 @@ EXPORT_SYMBOL_GPL(ring_buffer_event_data); #define TS_MASK ((1ULL << TS_SHIFT) - 1) #define TS_DELTA_TEST (~TS_MASK) +/* Flag when events were overwritten */ +#define RB_MISSED_EVENTS (1 << 31) + struct buffer_data_page { u64 time_stamp; /* page time stamp */ local_t commit; /* write committed index */ @@ -416,6 +419,12 @@ int ring_buffer_print_page_header(struct trace_seq *s) (unsigned int)sizeof(field.commit), (unsigned int)is_signed_type(long)); + ret = trace_seq_printf(s, "\tfield: int overwrite;\t" + "offset:%u;\tsize:%u;\tsigned:%u;\n", + (unsigned int)offsetof(typeof(field), commit), + 1, + (unsigned int)is_signed_type(long)); + ret = trace_seq_printf(s, "\tfield: char data;\t" "offset:%u;\tsize:%u;\tsigned:%u;\n", (unsigned int)offsetof(typeof(field), data), @@ -439,6 +448,8 @@ struct ring_buffer_per_cpu { struct buffer_page *tail_page; /* write to tail */ struct buffer_page *commit_page; /* committed pages */ struct buffer_page *reader_page; + unsigned long lost_events; + unsigned long last_overrun; local_t commit_overrun; local_t overrun; local_t entries; @@ -2835,6 +2846,7 @@ static struct buffer_page * rb_get_reader_page(struct ring_buffer_per_cpu *cpu_buffer) { struct buffer_page *reader = NULL; + unsigned long overwrite; unsigned long flags; int nr_loops = 0; int ret; @@ -2896,6 +2908,18 @@ rb_get_reader_page(struct ring_buffer_per_cpu *cpu_buffer) rb_set_list_to_head(cpu_buffer, &cpu_buffer->reader_page->list); /* + * We want to make sure we read the overruns after we set up our + * pointers to the next object. The writer side does a + * cmpxchg to cross pages which acts as the mb on the writer + * side. Note, the reader will constantly fail the swap + * while the writer is updating the pointers, so this + * guarantees that the overwrite recorded here is the one we + * want to compare with the last_overrun. + */ + smp_mb(); + overwrite = local_read(&(cpu_buffer->overrun)); + + /* * Here's the tricky part. * * We need to move the pointer past the header page. @@ -2926,6 +2950,11 @@ rb_get_reader_page(struct ring_buffer_per_cpu *cpu_buffer) cpu_buffer->reader_page = reader; rb_reset_reader_page(cpu_buffer); + if (overwrite != cpu_buffer->last_overrun) { + cpu_buffer->lost_events = overwrite - cpu_buffer->last_overrun; + cpu_buffer->last_overrun = overwrite; + } + goto again; out: @@ -3002,8 +3031,14 @@ static void rb_advance_iter(struct ring_buffer_iter *iter) rb_advance_iter(iter); } +static int rb_lost_events(struct ring_buffer_per_cpu *cpu_buffer) +{ + return cpu_buffer->lost_events; +} + static struct ring_buffer_event * -rb_buffer_peek(struct ring_buffer_per_cpu *cpu_buffer, u64 *ts) +rb_buffer_peek(struct ring_buffer_per_cpu *cpu_buffer, u64 *ts, + unsigned long *lost_events) { struct ring_buffer_event *event; struct buffer_page *reader; @@ -3055,6 +3090,8 @@ rb_buffer_peek(struct ring_buffer_per_cpu *cpu_buffer, u64 *ts) ring_buffer_normalize_time_stamp(cpu_buffer->buffer, cpu_buffer->cpu, ts); } + if (lost_events) + *lost_events = rb_lost_events(cpu_buffer); return event; default: @@ -3165,12 +3202,14 @@ static inline int rb_ok_to_lock(void) * @buffer: The ring buffer to read * @cpu: The cpu to peak at * @ts: The timestamp counter of this event. + * @lost_events: a variable to store if events were lost (may be NULL) * * This will return the event that will be read next, but does * not consume the data. */ struct ring_buffer_event * -ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts) +ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts, + unsigned long *lost_events) { struct ring_buffer_per_cpu *cpu_buffer = buffer->buffers[cpu]; struct ring_buffer_event *event; @@ -3185,7 +3224,7 @@ ring_buffer_peek(struct ring_buffer *buffer, int cpu, u64 *ts) local_irq_save(flags); if (dolock) spin_lock(&cpu_buffer->reader_lock); - event = rb_buffer_peek(cpu_buffer, ts); + event = rb_buffer_peek(cpu_buffer, ts, lost_events); if (event && event->type_len == RINGBUF_TYPE_PADDING) rb_advance_reader(cpu_buffer); if (dolock) @@ -3227,13 +3266,17 @@ ring_buffer_iter_peek(struct ring_buffer_iter *iter, u64 *ts) /** * ring_buffer_consume - return an event and consume it * @buffer: The ring buffer to get the next event from + * @cpu: the cpu to read the buffer from + * @ts: a variable to store the timestamp (may be NULL) + * @lost_events: a variable to store if events were lost (may be NULL) * * Returns the next event in the ring buffer, and that event is consumed. * Meaning, that sequential reads will keep returning a different event, * and eventually empty the ring buffer if the producer is slower. */ struct ring_buffer_event * -ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts) +ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts, + unsigned long *lost_events) { struct ring_buffer_per_cpu *cpu_buffer; struct ring_buffer_event *event = NULL; @@ -3254,9 +3297,11 @@ ring_buffer_consume(struct ring_buffer *buffer, int cpu, u64 *ts) if (dolock) spin_lock(&cpu_buffer->reader_lock); - event = rb_buffer_peek(cpu_buffer, ts); - if (event) + event = rb_buffer_peek(cpu_buffer, ts, lost_events); + if (event) { + cpu_buffer->lost_events = 0; rb_advance_reader(cpu_buffer); + } if (dolock) spin_unlock(&cpu_buffer->reader_lock); @@ -3405,6 +3450,9 @@ rb_reset_cpu(struct ring_buffer_per_cpu *cpu_buffer) cpu_buffer->write_stamp = 0; cpu_buffer->read_stamp = 0; + cpu_buffer->lost_events = 0; + cpu_buffer->last_overrun = 0; + rb_head_page_activate(cpu_buffer); } @@ -3684,6 +3732,7 @@ int ring_buffer_read_page(struct ring_buffer *buffer, unsigned int commit; unsigned int read; u64 save_timestamp; + int missed_events = 0; int ret = -1; if (!cpumask_test_cpu(cpu, buffer->cpumask)) @@ -3716,6 +3765,10 @@ int ring_buffer_read_page(struct ring_buffer *buffer, read = reader->read; commit = rb_page_commit(reader); + /* Check if any events were dropped */ + if (cpu_buffer->lost_events) + missed_events = 1; + /* * If this page has been partially read or * if len is not big enough to read the rest of the page or @@ -3779,6 +3832,13 @@ int ring_buffer_read_page(struct ring_buffer *buffer, } ret = read; + cpu_buffer->lost_events = 0; + /* + * Set a flag in the commit field if we lost events + */ + if (missed_events) + local_add(RB_MISSED_EVENTS, &bpage->commit); + out_unlock: spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags); diff --git a/kernel/trace/ring_buffer_benchmark.c b/kernel/trace/ring_buffer_benchmark.c index df74c79..dc56556 100644 --- a/kernel/trace/ring_buffer_benchmark.c +++ b/kernel/trace/ring_buffer_benchmark.c @@ -81,7 +81,7 @@ static enum event_status read_event(int cpu) int *entry; u64 ts; - event = ring_buffer_consume(buffer, cpu, &ts); + event = ring_buffer_consume(buffer, cpu, &ts, NULL); if (!event) return EVENT_DROPPED; diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 3ec2ee6..fabb003 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1556,7 +1556,7 @@ peek_next_entry(struct trace_iterator *iter, int cpu, u64 *ts) if (buf_iter) event = ring_buffer_iter_peek(buf_iter, ts); else - event = ring_buffer_peek(iter->tr->buffer, cpu, ts); + event = ring_buffer_peek(iter->tr->buffer, cpu, ts, NULL); ftrace_enable_cpu(); @@ -1635,7 +1635,7 @@ static void trace_consume(struct trace_iterator *iter) { /* Don't allow ftrace to trace into the ring buffers */ ftrace_disable_cpu(); - ring_buffer_consume(iter->tr->buffer, iter->cpu, &iter->ts); + ring_buffer_consume(iter->tr->buffer, iter->cpu, &iter->ts, NULL); ftrace_enable_cpu(); } diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index e6989d9..a7f75fb 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -489,9 +489,10 @@ get_return_for_leaf(struct trace_iterator *iter, * We need to consume the current entry to see * the next one. */ - ring_buffer_consume(iter->tr->buffer, iter->cpu, NULL); + ring_buffer_consume(iter->tr->buffer, iter->cpu, + NULL, NULL); event = ring_buffer_peek(iter->tr->buffer, iter->cpu, - NULL); + NULL, NULL); } if (!event) diff --git a/kernel/trace/trace_selftest.c b/kernel/trace/trace_selftest.c index 280fea4..e501808 100644 --- a/kernel/trace/trace_selftest.c +++ b/kernel/trace/trace_selftest.c @@ -29,7 +29,7 @@ static int trace_test_buffer_cpu(struct trace_array *tr, int cpu) struct trace_entry *entry; unsigned int loops = 0; - while ((event = ring_buffer_consume(tr->buffer, cpu, NULL))) { + while ((event = ring_buffer_consume(tr->buffer, cpu, NULL, NULL))) { entry = ring_buffer_event_data(event); /* -- cgit v0.10.2 From bc21b478425ac73f66a5ec0b375a5e0d12d609ce Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 31 Mar 2010 19:49:26 -0400 Subject: tracing: Show the lost events in the trace_pipe output Now that the ring buffer can keep track of where events are lost. Use this information to the output of trace_pipe: hackbench-3588 [001] 1326.701660: lock_acquire: ffffffff816591e0 read rcu_read_lock hackbench-3588 [001] 1326.701661: lock_acquire: ffff88003f4091f0 &(&dentry->d_lock)->rlock hackbench-3588 [001] 1326.701664: lock_release: ffff88003f4091f0 &(&dentry->d_lock)->rlock CPU:1 [LOST 673 EVENTS] hackbench-3588 [001] 1326.702711: kmem_cache_free: call_site=ffffffff81102b85 ptr=ffff880026d96738 hackbench-3588 [001] 1326.702712: lock_release: ffff88003e1480a8 &mm->mmap_sem hackbench-3588 [001] 1326.702713: lock_acquire: ffff88003e1480a8 &mm->mmap_sem Even works with the function graph tracer: 2) ! 170.098 us | } 2) 4.036 us | rcu_irq_exit(); 2) 3.657 us | idle_cpu(); 2) ! 190.301 us | } CPU:2 [LOST 2196 EVENTS] 2) 0.853 us | } /* cancel_dirty_page */ 2) | remove_from_page_cache() { 2) 1.578 us | _raw_spin_lock_irq(); 2) | __remove_from_page_cache() { Note, it does not work with the iterator "trace" file, since it requires the use of consuming the page from the ring buffer to determine how many events were lost, which the iterator does not do. Signed-off-by: Steven Rostedt diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index c0f4b36..39e71b0 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -58,6 +58,7 @@ struct trace_iterator { /* The below is zeroed out in pipe_read */ struct trace_seq seq; struct trace_entry *ent; + unsigned long lost_events; int leftover; int cpu; u64 ts; diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index fabb003..0498beb 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1545,7 +1545,8 @@ static void trace_iterator_increment(struct trace_iterator *iter) } static struct trace_entry * -peek_next_entry(struct trace_iterator *iter, int cpu, u64 *ts) +peek_next_entry(struct trace_iterator *iter, int cpu, u64 *ts, + unsigned long *lost_events) { struct ring_buffer_event *event; struct ring_buffer_iter *buf_iter = iter->buffer_iter[cpu]; @@ -1556,7 +1557,8 @@ peek_next_entry(struct trace_iterator *iter, int cpu, u64 *ts) if (buf_iter) event = ring_buffer_iter_peek(buf_iter, ts); else - event = ring_buffer_peek(iter->tr->buffer, cpu, ts, NULL); + event = ring_buffer_peek(iter->tr->buffer, cpu, ts, + lost_events); ftrace_enable_cpu(); @@ -1564,10 +1566,12 @@ peek_next_entry(struct trace_iterator *iter, int cpu, u64 *ts) } static struct trace_entry * -__find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) +__find_next_entry(struct trace_iterator *iter, int *ent_cpu, + unsigned long *missing_events, u64 *ent_ts) { struct ring_buffer *buffer = iter->tr->buffer; struct trace_entry *ent, *next = NULL; + unsigned long lost_events, next_lost = 0; int cpu_file = iter->cpu_file; u64 next_ts = 0, ts; int next_cpu = -1; @@ -1580,7 +1584,7 @@ __find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) if (cpu_file > TRACE_PIPE_ALL_CPU) { if (ring_buffer_empty_cpu(buffer, cpu_file)) return NULL; - ent = peek_next_entry(iter, cpu_file, ent_ts); + ent = peek_next_entry(iter, cpu_file, ent_ts, missing_events); if (ent_cpu) *ent_cpu = cpu_file; @@ -1592,7 +1596,7 @@ __find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) if (ring_buffer_empty_cpu(buffer, cpu)) continue; - ent = peek_next_entry(iter, cpu, &ts); + ent = peek_next_entry(iter, cpu, &ts, &lost_events); /* * Pick the entry with the smallest timestamp: @@ -1601,6 +1605,7 @@ __find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) next = ent; next_cpu = cpu; next_ts = ts; + next_lost = lost_events; } } @@ -1610,6 +1615,9 @@ __find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) if (ent_ts) *ent_ts = next_ts; + if (missing_events) + *missing_events = next_lost; + return next; } @@ -1617,13 +1625,14 @@ __find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) struct trace_entry *trace_find_next_entry(struct trace_iterator *iter, int *ent_cpu, u64 *ent_ts) { - return __find_next_entry(iter, ent_cpu, ent_ts); + return __find_next_entry(iter, ent_cpu, NULL, ent_ts); } /* Find the next real entry, and increment the iterator to the next entry */ static void *find_next_entry_inc(struct trace_iterator *iter) { - iter->ent = __find_next_entry(iter, &iter->cpu, &iter->ts); + iter->ent = __find_next_entry(iter, &iter->cpu, + &iter->lost_events, &iter->ts); if (iter->ent) trace_iterator_increment(iter); @@ -1635,7 +1644,8 @@ static void trace_consume(struct trace_iterator *iter) { /* Don't allow ftrace to trace into the ring buffers */ ftrace_disable_cpu(); - ring_buffer_consume(iter->tr->buffer, iter->cpu, &iter->ts, NULL); + ring_buffer_consume(iter->tr->buffer, iter->cpu, &iter->ts, + &iter->lost_events); ftrace_enable_cpu(); } @@ -2030,6 +2040,10 @@ static enum print_line_t print_trace_line(struct trace_iterator *iter) { enum print_line_t ret; + if (iter->lost_events) + trace_seq_printf(&iter->seq, "CPU:%d [LOST %lu EVENTS]\n", + iter->cpu, iter->lost_events); + if (iter->trace && iter->trace->print_line) { ret = iter->trace->print_line(iter); if (ret != TRACE_TYPE_UNHANDLED) -- cgit v0.10.2 From ff0ff84a0767df48d728c36510365344a7e7d582 Mon Sep 17 00:00:00 2001 From: Steven Rostedt Date: Wed, 31 Mar 2010 22:11:42 -0400 Subject: ring-buffer: Add lost event count to end of sub buffer Currently, binary readers of the ring buffer only know where events were lost, but not how many events were lost at that location. This information is available, but it would require adding another field to the sub buffer header to include it. But when a event can not fit at the end of a sub buffer, it is written to the next sub buffer. This means there is a good chance that the buffer may have room to hold this counter. If it does, write the counter at the end of the sub buffer and set another flag in the data size field that states that this information exists. Signed-off-by: Steven Rostedt diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index 8295650..dc6d563 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -320,6 +320,8 @@ EXPORT_SYMBOL_GPL(ring_buffer_event_data); /* Flag when events were overwritten */ #define RB_MISSED_EVENTS (1 << 31) +/* Missed count stored at end */ +#define RB_MISSED_STORED (1 << 30) struct buffer_data_page { u64 time_stamp; /* page time stamp */ @@ -340,6 +342,7 @@ struct buffer_page { local_t write; /* index for next write */ unsigned read; /* index for next read */ local_t entries; /* entries on this page */ + unsigned long real_end; /* real end of data */ struct buffer_data_page *page; /* Actual data page */ }; @@ -1770,6 +1773,13 @@ rb_reset_tail(struct ring_buffer_per_cpu *cpu_buffer, kmemcheck_annotate_bitfield(event, bitfield); /* + * Save the original length to the meta data. + * This will be used by the reader to add lost event + * counter. + */ + tail_page->real_end = tail; + + /* * If this event is bigger than the minimum size, then * we need to be careful that we don't subtract the * write counter enough to allow another writer to slip @@ -2888,6 +2898,7 @@ rb_get_reader_page(struct ring_buffer_per_cpu *cpu_buffer) local_set(&cpu_buffer->reader_page->write, 0); local_set(&cpu_buffer->reader_page->entries, 0); local_set(&cpu_buffer->reader_page->page->commit, 0); + cpu_buffer->reader_page->real_end = 0; spin: /* @@ -3728,11 +3739,11 @@ int ring_buffer_read_page(struct ring_buffer *buffer, struct ring_buffer_event *event; struct buffer_data_page *bpage; struct buffer_page *reader; + unsigned long missed_events; unsigned long flags; unsigned int commit; unsigned int read; u64 save_timestamp; - int missed_events = 0; int ret = -1; if (!cpumask_test_cpu(cpu, buffer->cpumask)) @@ -3766,8 +3777,7 @@ int ring_buffer_read_page(struct ring_buffer *buffer, commit = rb_page_commit(reader); /* Check if any events were dropped */ - if (cpu_buffer->lost_events) - missed_events = 1; + missed_events = cpu_buffer->lost_events; /* * If this page has been partially read or @@ -3829,6 +3839,14 @@ int ring_buffer_read_page(struct ring_buffer *buffer, local_set(&reader->entries, 0); reader->read = 0; *data_page = bpage; + + /* + * Use the real_end for the data size, + * This gives us a chance to store the lost events + * on the page. + */ + if (reader->real_end) + local_set(&bpage->commit, reader->real_end); } ret = read; @@ -3836,8 +3854,19 @@ int ring_buffer_read_page(struct ring_buffer *buffer, /* * Set a flag in the commit field if we lost events */ - if (missed_events) + if (missed_events) { + commit = local_read(&bpage->commit); + + /* If there is room at the end of the page to save the + * missed events, then record it there. + */ + if (BUF_PAGE_SIZE - commit >= sizeof(missed_events)) { + memcpy(&bpage->data[commit], &missed_events, + sizeof(missed_events)); + local_add(RB_MISSED_STORED, &bpage->commit); + } local_add(RB_MISSED_EVENTS, &bpage->commit); + } out_unlock: spin_unlock_irqrestore(&cpu_buffer->reader_lock, flags); -- cgit v0.10.2 From 6072f7491f5ef391a575e18a1165e72a3eef1601 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 31 Mar 2010 20:11:59 +0000 Subject: ide: Requeue request after DMA timeout I noticed that my KVM virtual machines were experiencing IDE issues resulting in processes stuck on waiting for buffers to complete. The root cause is of course race conditions in the ancient qemu backend that I'm using. However, the fact that the guest isn't recovering is a bug. I've tracked it down to the change made last year to dequeue requests at the start rather than at the end in the IDE layer. commit 8f6205cd572fece673da0255d74843680f67f879 Author: Tejun Heo Date: Fri May 8 11:53:59 2009 +0900 ide: dequeue in-flight request The problem is that the function ide_dma_timeout_retry does not requeue the current request, causing one request to be lost for each DMA timeout. This patch fixes this by requeueing the request. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller diff --git a/drivers/ide/ide-dma.c b/drivers/ide/ide-dma.c index ee58c88..fd40a81 100644 --- a/drivers/ide/ide-dma.c +++ b/drivers/ide/ide-dma.c @@ -492,6 +492,7 @@ ide_startstop_t ide_dma_timeout_retry(ide_drive_t *drive, int error) if (rq) { hwif->rq = NULL; rq->errors = 0; + ide_requeue_and_plug(drive, rq); } return ret; } diff --git a/drivers/ide/ide-io.c b/drivers/ide/ide-io.c index db96138..172ac92 100644 --- a/drivers/ide/ide-io.c +++ b/drivers/ide/ide-io.c @@ -566,7 +566,7 @@ plug_device_2: blk_plug_device(q); } -static void ide_requeue_and_plug(ide_drive_t *drive, struct request *rq) +void ide_requeue_and_plug(ide_drive_t *drive, struct request *rq) { struct request_queue *q = drive->queue; unsigned long flags; diff --git a/include/linux/ide.h b/include/linux/ide.h index 97e6ab4..3239d1c 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -1169,6 +1169,7 @@ extern void ide_stall_queue(ide_drive_t *drive, unsigned long timeout); extern void ide_timer_expiry(unsigned long); extern irqreturn_t ide_intr(int irq, void *dev_id); extern void do_ide_request(struct request_queue *); +extern void ide_requeue_and_plug(ide_drive_t *drive, struct request *rq); void ide_init_disk(struct gendisk *, ide_drive_t *); -- cgit v0.10.2 From 1af185034662b75279f90e1c7cb958271d4121e2 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 31 Mar 2010 20:13:39 +0000 Subject: ide: Must hold queue lock when requeueing ide-atapi requeues requests without holding the queue lock. This patch fixes it by using ide_requeue_and_plug. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller diff --git a/drivers/ide/ide-atapi.c b/drivers/ide/ide-atapi.c index eb2181a..9aedb9a 100644 --- a/drivers/ide/ide-atapi.c +++ b/drivers/ide/ide-atapi.c @@ -263,8 +263,8 @@ void ide_retry_pc(ide_drive_t *drive) * of it. The failed command will be retried after sense data * is acquired. */ - blk_requeue_request(failed_rq->q, failed_rq); drive->hwif->rq = NULL; + ide_requeue_and_plug(drive, failed_rq); if (ide_queue_sense_rq(drive, pc)) { blk_start_request(failed_rq); ide_complete_rq(drive, -EIO, blk_rq_bytes(failed_rq)); -- cgit v0.10.2 From 91cb17314e74d0e5ab572b4b84b9398c61b71abb Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Thu, 1 Apr 2010 18:08:29 +0200 Subject: ALSA: hda - Update document about MSI and interrupts Signed-off-by: Takashi Iwai diff --git a/Documentation/sound/alsa/HD-Audio.txt b/Documentation/sound/alsa/HD-Audio.txt index f4dd3bf..98d14cb 100644 --- a/Documentation/sound/alsa/HD-Audio.txt +++ b/Documentation/sound/alsa/HD-Audio.txt @@ -119,10 +119,18 @@ the codec slots 0 and 1 no matter what the hardware reports. Interrupt Handling ~~~~~~~~~~~~~~~~~~ -In rare but some cases, the interrupt isn't properly handled as -default. You would notice this by the DMA transfer error reported by -ALSA PCM core, for example. Using MSI might help in such a case. -Pass `enable_msi=1` option for enabling MSI. +HD-audio driver uses MSI as default (if available) since 2.6.33 +kernel as MSI works better on some machines, and in general, it's +better for performance. However, Nvidia controllers showed bad +regressions with MSI (especially in a combination with AMD chipset), +thus we disabled MSI for them. + +There seem also still other devices that don't work with MSI. If you +see a regression wrt the sound quality (stuttering, etc) or a lock-up +in the recent kernel, try to pass `enable_msi=0` option to disable +MSI. If it works, you can add the known bad device to the blacklist +defined in hda_intel.c. In such a case, please report and give the +patch back to the upstream developer. HD-AUDIO CODEC -- cgit v0.10.2 From 80e755fedebc8de0599a79efad2c656503df2e62 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Wed, 31 Mar 2010 21:52:10 -0700 Subject: ceph: allow writeback of snapped pages older than 'oldest' snapc On snap deletion, we don't regenerate ceph_cap_snaps for inodes with dirty pages because deletion does not affect metadata writeback. However, we did run into problems when we went to write back the pages because the 'oldest' snapc is determined by the oldest cap_snap, and that may be the newer snapc that reflects the deletion. This caused confusion and an infinite loop in ceph_update_writeable_page(). Change the snapc checks to allow writeback of any snapc that is equal to OR older than the 'oldest' snapc. When there are no cap_snaps, we were also using the realm's latest snapc for writeback, which complicates ceph_put_wrbufffer_cap_refs(). Instead, use i_head_snapc, the most snapc used for the most recent ('head') data. This makes the writeback snapc (ceph_osd_request.r_snapc) _always_ match a capsnap or i_head_snapc. Also, in writepags_finish(), drop the snapc referenced by the _page_ and do not assume it matches the request snapc (it may not anymore). Signed-off-by: Sage Weil diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index ce8ef61..a313e9b 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -356,8 +356,8 @@ static struct ceph_snap_context *__get_oldest_context(struct inode *inode, break; } } - if (!snapc && ci->i_snap_realm) { - snapc = ceph_get_snap_context(ci->i_snap_realm->cached_context); + if (!snapc && ci->i_head_snapc) { + snapc = ceph_get_snap_context(ci->i_head_snapc); dout(" head snapc %p has %d dirty pages\n", snapc, ci->i_wrbuffer_ref_head); } @@ -412,7 +412,7 @@ static int writepage_nounlock(struct page *page, struct writeback_control *wbc) dout("writepage %p page %p not dirty?\n", inode, page); goto out; } - if (snapc != get_oldest_context(inode, &snap_size)) { + if (snapc->seq > get_oldest_context(inode, &snap_size)->seq) { dout("writepage %p page %p snapc %p not writeable - noop\n", inode, page, (void *)page->private); /* we should only noop if called by kswapd */ @@ -557,9 +557,9 @@ static void writepages_finish(struct ceph_osd_request *req, dout("inode %p skipping page %p\n", inode, page); wbc->pages_skipped++; } + ceph_put_snap_context((void *)page->private); page->private = 0; ClearPagePrivate(page); - ceph_put_snap_context(snapc); dout("unlocking %d %p\n", i, page); end_page_writeback(page); @@ -617,7 +617,7 @@ static int ceph_writepages_start(struct address_space *mapping, int range_whole = 0; int should_loop = 1; pgoff_t max_pages = 0, max_pages_ever = 0; - struct ceph_snap_context *snapc = NULL, *last_snapc = NULL; + struct ceph_snap_context *snapc = NULL, *last_snapc = NULL, *pgsnapc; struct pagevec pvec; int done = 0; int rc = 0; @@ -769,9 +769,10 @@ get_more_pages: } /* only if matching snap context */ - if (snapc != (void *)page->private) { - dout("page snapc %p != oldest %p\n", - (void *)page->private, snapc); + pgsnapc = (void *)page->private; + if (pgsnapc->seq > snapc->seq) { + dout("page snapc %p %lld > oldest %p %lld\n", + pgsnapc, pgsnapc->seq, snapc, snapc->seq); unlock_page(page); if (!locked_pages) continue; /* keep looking for snap */ @@ -935,8 +936,8 @@ static int ceph_update_writeable_page(struct file *file, int pos_in_page = pos & ~PAGE_CACHE_MASK; int end_in_page = pos_in_page + len; loff_t i_size; - struct ceph_snap_context *snapc; int r; + struct ceph_snap_context *snapc, *oldest; retry_locked: /* writepages currently holds page lock, but if we change that later, */ @@ -946,16 +947,16 @@ retry_locked: BUG_ON(!ci->i_snap_realm); down_read(&mdsc->snap_rwsem); BUG_ON(!ci->i_snap_realm->cached_context); - if (page->private && - (void *)page->private != ci->i_snap_realm->cached_context) { + snapc = (void *)page->private; + if (snapc && snapc != ci->i_head_snapc) { /* * this page is already dirty in another (older) snap * context! is it writeable now? */ - snapc = get_oldest_context(inode, NULL); + oldest = get_oldest_context(inode, NULL); up_read(&mdsc->snap_rwsem); - if (snapc != (void *)page->private) { + if (snapc->seq > oldest->seq) { dout(" page %p snapc %p not current or oldest\n", page, (void *)page->private); /* -- cgit v0.10.2 From 6298a33757ba7361bb8f506c106daad77e5ac8cf Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Wed, 31 Mar 2010 22:01:38 -0700 Subject: ceph: fix snap context reference leaks The get_oldest_context() helper takes a reference to the returned snap context, but most callers weren't dropping that reference. Fix them. Also drop the unused locked __get_oldest_context() variant. Signed-off-by: Sage Weil diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index a313e9b..41f1f71 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -336,16 +336,15 @@ out: /* * Get ref for the oldest snapc for an inode with dirty data... that is, the * only snap context we are allowed to write back. - * - * Caller holds i_lock. */ -static struct ceph_snap_context *__get_oldest_context(struct inode *inode, - u64 *snap_size) +static struct ceph_snap_context *get_oldest_context(struct inode *inode, + u64 *snap_size) { struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_snap_context *snapc = NULL; struct ceph_cap_snap *capsnap = NULL; + spin_lock(&inode->i_lock); list_for_each_entry(capsnap, &ci->i_cap_snaps, ci_item) { dout(" cap_snap %p snapc %p has %d dirty pages\n", capsnap, capsnap->context, capsnap->dirty_pages); @@ -361,16 +360,6 @@ static struct ceph_snap_context *__get_oldest_context(struct inode *inode, dout(" head snapc %p has %d dirty pages\n", snapc, ci->i_wrbuffer_ref_head); } - return snapc; -} - -static struct ceph_snap_context *get_oldest_context(struct inode *inode, - u64 *snap_size) -{ - struct ceph_snap_context *snapc = NULL; - - spin_lock(&inode->i_lock); - snapc = __get_oldest_context(inode, snap_size); spin_unlock(&inode->i_lock); return snapc; } @@ -391,7 +380,7 @@ static int writepage_nounlock(struct page *page, struct writeback_control *wbc) int len = PAGE_CACHE_SIZE; loff_t i_size; int err = 0; - struct ceph_snap_context *snapc; + struct ceph_snap_context *snapc, *oldest; u64 snap_size = 0; long writeback_stat; @@ -412,13 +401,16 @@ static int writepage_nounlock(struct page *page, struct writeback_control *wbc) dout("writepage %p page %p not dirty?\n", inode, page); goto out; } - if (snapc->seq > get_oldest_context(inode, &snap_size)->seq) { + oldest = get_oldest_context(inode, &snap_size); + if (snapc->seq > oldest->seq) { dout("writepage %p page %p snapc %p not writeable - noop\n", inode, page, (void *)page->private); /* we should only noop if called by kswapd */ WARN_ON((current->flags & PF_MEMALLOC) == 0); + ceph_put_snap_context(oldest); goto out; } + ceph_put_snap_context(oldest); /* is this a partial page at end of file? */ if (snap_size) @@ -457,7 +449,7 @@ static int writepage_nounlock(struct page *page, struct writeback_control *wbc) ClearPagePrivate(page); end_page_writeback(page); ceph_put_wrbuffer_cap_refs(ci, 1, snapc); - ceph_put_snap_context(snapc); + ceph_put_snap_context(snapc); /* page's reference */ out: return err; } @@ -914,7 +906,10 @@ static int context_is_writeable_or_written(struct inode *inode, struct ceph_snap_context *snapc) { struct ceph_snap_context *oldest = get_oldest_context(inode, NULL); - return !oldest || snapc->seq <= oldest->seq; + int ret = !oldest || snapc->seq <= oldest->seq; + + ceph_put_snap_context(oldest); + return ret; } /* @@ -957,13 +952,14 @@ retry_locked: up_read(&mdsc->snap_rwsem); if (snapc->seq > oldest->seq) { + ceph_put_snap_context(oldest); dout(" page %p snapc %p not current or oldest\n", - page, (void *)page->private); + page, snapc); /* * queue for writeback, and wait for snapc to * be writeable or written */ - snapc = ceph_get_snap_context((void *)page->private); + snapc = ceph_get_snap_context(snapc); unlock_page(page); ceph_queue_writeback(inode); r = wait_event_interruptible(ci->i_cap_wq, @@ -973,6 +969,7 @@ retry_locked: return r; return -EAGAIN; } + ceph_put_snap_context(oldest); /* yay, writeable, do it now (without dropping page lock) */ dout(" page %p snapc %p not current, but oldest\n", -- cgit v0.10.2 From 819ccbfa448403992ceafc05d6d7097aaa74d4c3 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Thu, 1 Apr 2010 09:33:46 -0700 Subject: ceph: fix leaked inode ref due to snap metadata writeback race We create a ceph_cap_snap if there is dirty cap metadata (for writeback to mds) OR dirty pages (for writeback to osd). It is thus possible that the metadata has been written back to the MDS but the OSD data has not when the cap_snap is created. This results in a cap_snap with dirty(caps) == 0. The problem is that cap writeback to the MDS isn't necessary, and a FLUSHSNAP cap op gets no ack from the MDS. This leaves the cap_snap attached to the inode along with its inode reference. Fix the problem by dropping the cap_snap if it becomes 'complete' (all pages written out) and dirty(caps) == 0 in ceph_put_wrbuffer_cap_refs(). Also, BUG() in __ceph_flush_snaps() if we encounter a cap_snap with dirty(caps) == 0. Signed-off-by: Sage Weil diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 7d0a0d0..b6fdf01 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -1204,6 +1204,12 @@ retry: if (capsnap->dirty_pages || capsnap->writing) continue; + /* + * if cap writeback already occurred, we should have dropped + * the capsnap in ceph_put_wrbuffer_cap_refs. + */ + BUG_ON(capsnap->dirty == 0); + /* pick mds, take s_mutex */ mds = __ceph_get_cap_mds(ci, &mseq); if (session && session->s_mds != mds) { @@ -2117,8 +2123,8 @@ void ceph_put_cap_refs(struct ceph_inode_info *ci, int had) } spin_unlock(&inode->i_lock); - dout("put_cap_refs %p had %s %s\n", inode, ceph_cap_string(had), - last ? "last" : ""); + dout("put_cap_refs %p had %s%s%s\n", inode, ceph_cap_string(had), + last ? " last" : "", put ? " put" : ""); if (last && !flushsnaps) ceph_check_caps(ci, 0, NULL); @@ -2142,7 +2148,8 @@ void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr, { struct inode *inode = &ci->vfs_inode; int last = 0; - int last_snap = 0; + int complete_capsnap = 0; + int drop_capsnap = 0; int found = 0; struct ceph_cap_snap *capsnap = NULL; @@ -2165,19 +2172,32 @@ void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr, list_for_each_entry(capsnap, &ci->i_cap_snaps, ci_item) { if (capsnap->context == snapc) { found = 1; - capsnap->dirty_pages -= nr; - last_snap = !capsnap->dirty_pages; break; } } BUG_ON(!found); + capsnap->dirty_pages -= nr; + if (capsnap->dirty_pages == 0) { + complete_capsnap = 1; + if (capsnap->dirty == 0) + /* cap writeback completed before we created + * the cap_snap; no FLUSHSNAP is needed */ + drop_capsnap = 1; + } dout("put_wrbuffer_cap_refs on %p cap_snap %p " - " snap %lld %d/%d -> %d/%d %s%s\n", + " snap %lld %d/%d -> %d/%d %s%s%s\n", inode, capsnap, capsnap->context->seq, ci->i_wrbuffer_ref+nr, capsnap->dirty_pages + nr, ci->i_wrbuffer_ref, capsnap->dirty_pages, last ? " (wrbuffer last)" : "", - last_snap ? " (capsnap last)" : ""); + complete_capsnap ? " (complete capsnap)" : "", + drop_capsnap ? " (drop capsnap)" : ""); + if (drop_capsnap) { + ceph_put_snap_context(capsnap->context); + list_del(&capsnap->ci_item); + list_del(&capsnap->flushing_item); + ceph_put_cap_snap(capsnap); + } } spin_unlock(&inode->i_lock); @@ -2185,10 +2205,12 @@ void ceph_put_wrbuffer_cap_refs(struct ceph_inode_info *ci, int nr, if (last) { ceph_check_caps(ci, CHECK_CAPS_AUTHONLY, NULL); iput(inode); - } else if (last_snap) { + } else if (complete_capsnap) { ceph_flush_snaps(ci); wake_up(&ci->i_cap_wq); } + if (drop_capsnap) + iput(inode); } /* @@ -2464,8 +2486,8 @@ static void handle_cap_flushsnap_ack(struct inode *inode, u64 flush_tid, break; } WARN_ON(capsnap->dirty_pages || capsnap->writing); - dout(" removing cap_snap %p follows %lld\n", - capsnap, follows); + dout(" removing %p cap_snap %p follows %lld\n", + inode, capsnap, follows); ceph_put_snap_context(capsnap->context); list_del(&capsnap->ci_item); list_del(&capsnap->flushing_item); diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c index df04e21..7e3e5f9 100644 --- a/fs/ceph/snap.c +++ b/fs/ceph/snap.c @@ -521,15 +521,17 @@ int __ceph_finish_cap_snap(struct ceph_inode_info *ci, capsnap->ctime = inode->i_ctime; capsnap->time_warp_seq = ci->i_time_warp_seq; if (capsnap->dirty_pages) { - dout("finish_cap_snap %p cap_snap %p snapc %p %llu s=%llu " + dout("finish_cap_snap %p cap_snap %p snapc %p %llu %s s=%llu " "still has %d dirty pages\n", inode, capsnap, capsnap->context, capsnap->context->seq, - capsnap->size, capsnap->dirty_pages); + ceph_cap_string(capsnap->dirty), capsnap->size, + capsnap->dirty_pages); return 0; } - dout("finish_cap_snap %p cap_snap %p snapc %p %llu s=%llu clean\n", + dout("finish_cap_snap %p cap_snap %p snapc %p %llu %s s=%llu\n", inode, capsnap, capsnap->context, - capsnap->context->seq, capsnap->size); + capsnap->context->seq, ceph_cap_string(capsnap->dirty), + capsnap->size); spin_lock(&mdsc->snap_flush_lock); list_add_tail(&ci->i_snap_flush_item, &mdsc->snap_flush_list); -- cgit v0.10.2 From 909fc87b32b3b9e3f0b87dcc5d98319c41900c58 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Mon, 29 Mar 2010 09:41:11 +0200 Subject: x86: Handle overlapping mptables We found a system where the MP table MPC and MPF structures overlap. That doesn't really matter because the mptable is not used anyways with ACPI, but it leads to a panic in the early allocator due to the overlapping reservations in 2.6.33. Earlier kernels handled this without problems. Simply change these reservations to reserve_early_overlap_ok to avoid the panic. Reported-by: Thomas Renninger Tested-by: Thomas Renninger Signed-off-by: Andi Kleen LKML-Reference: <20100329074111.GA22821@basil.fritz.box> Signed-off-by: H. Peter Anvin Cc: diff --git a/arch/x86/kernel/mpparse.c b/arch/x86/kernel/mpparse.c index a2c1edd..e81030f 100644 --- a/arch/x86/kernel/mpparse.c +++ b/arch/x86/kernel/mpparse.c @@ -664,7 +664,7 @@ static void __init smp_reserve_memory(struct mpf_intel *mpf) { unsigned long size = get_mpc_size(mpf->physptr); - reserve_early(mpf->physptr, mpf->physptr+size, "MP-table mpc"); + reserve_early_overlap_ok(mpf->physptr, mpf->physptr+size, "MP-table mpc"); } static int __init smp_scan_config(unsigned long base, unsigned long length) @@ -693,7 +693,7 @@ static int __init smp_scan_config(unsigned long base, unsigned long length) mpf, (u64)virt_to_phys(mpf)); mem = virt_to_phys(mpf); - reserve_early(mem, mem + sizeof(*mpf), "MP-table mpf"); + reserve_early_overlap_ok(mem, mem + sizeof(*mpf), "MP-table mpf"); if (mpf->physptr) smp_reserve_memory(mpf); -- cgit v0.10.2 From 337998587f802535896e9ed16d19f97915ccd368 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Wed, 31 Mar 2010 20:44:09 -0700 Subject: nobootmem, x86: Fix 32bit numa system without RAM on node 0 On one system without RAM on node0, got following boot dump with a 32 bit NUMA kernel: early_node_map[4] active PFN ranges 1: 0x00000010 -> 0x00000099 1: 0x00000100 -> 0x0007da00 1: 0x0007e800 -> 0x0007ffa0 1: 0x0007ffae -> 0x0007ffb0 ... Subtract (29 early reservations) #000 [0000001000 - 0000002000] #001 [0000089000 - 000008f000] #002 [0000091000 - 0000093500] ... #027 [007cbfef40 - 007e800000] #028 [007e9ca000 - 007ff95000] (0 free memory ranges) Initializing HighMem for node 0 (00000000:00000000) Initializing HighMem for node 1 (00000000:00000000) Memory: 0k/2096832k available (6662k kernel code, 2096300k reserved, 4829k data, 484k init, 0k highmem) ... Checking if this processor honours the WP bit even in supervisor mode...Ok. swapper: page allocation failure. order:0, mode:0x0 Pid: 0, comm: swapper Not tainted 2.6.34-rc3-tip-03818-g4b1ea6c-dirty #35 Call Trace: [<4087a5dc>] ? printk+0xf/0x11 [<40286728>] __alloc_pages_nodemask+0x417/0x487 [<402a9ce1>] new_slab+0xe2/0x1fe [<402aa5b2>] kmem_cache_open+0x185/0x358 [<402abbc0>] T.954+0x1c/0x60 [<40d52a29>] kmem_cache_init+0x24/0x113 [<40d39738>] start_kernel+0x166/0x2e4 [<40d3940e>] ? unknown_bootoption+0x0/0x18e [<40d390ce>] i386_start_kernel+0xce/0xd5 Mem-Info: Node 1 DMA per-cpu: CPU 0: hi: 0, btch: 1 usd: 0 Node 1 Normal per-cpu: CPU 0: hi: 0, btch: 1 usd: 0 active_anon:0 inactive_anon:0 isolated_anon:0 active_file:0 inactive_file:0 isolated_file:0 unevictable:0 dirty:0 writeback:0 unstable:0 free:0 slab_reclaimable:0 slab_unreclaimable:0 mapped:0 shmem:0 pagetables:0 bounce:0 When 32bit NUMA is used, free_all_bootmem() will still only go over with node id 0. If node 0 doesn't have RAM installed, We need to go with node1 because early_node_map still use 1 for all ranges, and ram from node1 become low ram. Use MAX_NUMNODES like 64-bit NUMA does. Note: BOOTMEM path has the same problem. this bug exist before We have NO_BOOTMEM support. -v3: add more comments, and fix bootmem path too. -v4: seperate bootmem path fix Signed-off-by: Yinghai Lu LKML-Reference: <4BB41689.9090502@kernel.org> Signed-off-by: H. Peter Anvin diff --git a/mm/bootmem.c b/mm/bootmem.c index 9b13446..2058cb7 100644 --- a/mm/bootmem.c +++ b/mm/bootmem.c @@ -303,7 +303,14 @@ unsigned long __init free_all_bootmem_node(pg_data_t *pgdat) unsigned long __init free_all_bootmem(void) { #ifdef CONFIG_NO_BOOTMEM - return free_all_memory_core_early(NODE_DATA(0)->node_id); + /* + * We need to use MAX_NUMNODES instead of NODE_DATA(0)->node_id + * because in some case like Node0 doesnt have RAM installed + * low ram will be on Node1 + * Use MAX_NUMNODES will make sure all ranges in early_node_map[] + * will be used instead of only Node0 related + */ + return free_all_memory_core_early(MAX_NUMNODES); #else return free_all_bootmem_core(NODE_DATA(0)->bdata); #endif -- cgit v0.10.2 From aa235fc712f379d4194cff9217f07026c452c141 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Wed, 31 Mar 2010 20:45:27 -0700 Subject: bootmem, x86: Fix 32bit numa system without RAM on node 0 When 32bit numa is used, free_all_bootmem() will still only go over with node id 0. If node 0 doesn't have RAM installed, the lowest populated node becomes low RAM. This one fixes BOOTMEM path by iterating over the bdata_list. -v3: add more comments, and fix bootmem path too. -v4: seperate from one big patch Signed-off-by: Yinghai Lu LKML-Reference: <4BB416D7.6090203@kernel.org> Signed-off-by: H. Peter Anvin diff --git a/mm/bootmem.c b/mm/bootmem.c index 2058cb7..ba37d62 100644 --- a/mm/bootmem.c +++ b/mm/bootmem.c @@ -312,7 +312,13 @@ unsigned long __init free_all_bootmem(void) */ return free_all_memory_core_early(MAX_NUMNODES); #else - return free_all_bootmem_core(NODE_DATA(0)->bdata); + unsigned long total_pages = 0; + bootmem_data_t *bdata; + + list_for_each_entry(bdata, &bdata_list, list) + total_pages += free_all_bootmem_core(bdata); + + return total_pages; #endif } -- cgit v0.10.2 From 8da854cb02156c90028233ae1e85ce46a1d3f82c Mon Sep 17 00:00:00 2001 From: "Pallipadi, Venkatesh" Date: Thu, 25 Feb 2010 10:53:48 -0800 Subject: x86, hpet: Erratum workaround for read after write of HPET comparator On Wed, Feb 24, 2010 at 03:37:04PM -0800, Justin Piszcz wrote: > Hello, > > Again, on the Intel DP55KG board: > > # uname -a > Linux host 2.6.33 #1 SMP Wed Feb 24 18:31:00 EST 2010 x86_64 GNU/Linux > > [ 1.237600] ------------[ cut here ]------------ > [ 1.237890] WARNING: at arch/x86/kernel/hpet.c:404 hpet_next_event+0x70/0x80() > [ 1.238221] Hardware name: > [ 1.238504] hpet: compare register read back failed. > [ 1.238793] Modules linked in: > [ 1.239315] Pid: 0, comm: swapper Not tainted 2.6.33 #1 > [ 1.239605] Call Trace: > [ 1.239886] [] ? warn_slowpath_common+0x73/0xb0 > [ 1.240409] [] ? tick_dev_program_event+0x38/0xc0 > [ 1.240699] [] ? warn_slowpath_fmt+0x40/0x50 > [ 1.240992] [] ? tick_dev_program_event+0x38/0xc0 > [ 1.241281] [] ? hpet_next_event+0x70/0x80 > [ 1.241573] [] ? tick_dev_program_event+0x38/0xc0 > [ 1.241859] [] ? tick_handle_oneshot_broadcast+0xe2/0x100 > [ 1.246533] [] ? timer_interrupt+0x1a/0x30 > [ 1.246826] [] ? handle_IRQ_event+0x39/0xd0 > [ 1.247118] [] ? handle_edge_irq+0xb8/0x160 > [ 1.247407] [] ? handle_irq+0x15/0x20 > [ 1.247689] [] ? do_IRQ+0x62/0xe0 > [ 1.247976] [] ? ret_from_intr+0x0/0xa > [ 1.248262] [] ? mwait_idle+0x57/0x80 > [ 1.248796] [] ? cpu_idle+0x5c/0xb0 > [ 1.249080] ---[ end trace db7f668fb6fef4e1 ]--- > > Is this something Intel has to fix or is it a bug in the kernel? This is a chipset erratum. Thomas: You mentioned we can retain this check only for known-buggy and hpet debug kind of options. But here is the simple workaround patch for this particular erratum. Some chipsets have a erratum due to which read immediately following a write of HPET comparator returns old comparator value instead of most recently written value. Erratum 15 in "Intel I/O Controller Hub 9 (ICH9) Family Specification Update" (http://www.intel.com/assets/pdf/specupdate/316973.pdf) Workaround for the errata is to read the comparator twice if the first one fails. Signed-off-by: Venkatesh Pallipadi LKML-Reference: <20100225185348.GA9674@linux-os.sc.intel.com> Signed-off-by: H. Peter Anvin Cc: Venkatesh Pallipadi Cc: diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index ee4fa1b..3d422da 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -399,9 +399,15 @@ static int hpet_next_event(unsigned long delta, * then we might have a real hardware problem. We can not do * much about it here, but at least alert the user/admin with * a prominent warning. + * An erratum on some chipsets (ICH9,..), results in comparator read + * immediately following a write returning old value. Workaround + * for this is to read this value second time, when first + * read returns old value. */ - WARN_ONCE(hpet_readl(HPET_Tn_CMP(timer)) != cnt, + if (unlikely((u32)hpet_readl(HPET_Tn_CMP(timer)) != cnt)) { + WARN_ONCE(hpet_readl(HPET_Tn_CMP(timer)) != cnt, KERN_WARNING "hpet: compare register read back failed.\n"); + } return (s32)(hpet_readl(HPET_COUNTER) - cnt) >= 0 ? -ETIME : 0; } -- cgit v0.10.2 From b4a5e8a1deca7e61ebaffb37344766b0f0e9f327 Mon Sep 17 00:00:00 2001 From: Alok Kataria Date: Thu, 11 Mar 2010 14:00:16 -0800 Subject: x86, hpet: Fix bug in RTC emulation We think there exists a bug in the HPET code that emulates the RTC. In the normal case, when the RTC frequency is set, the rtc driver tells the hpet code about it here: int hpet_set_periodic_freq(unsigned long freq) { uint64_t clc; if (!is_hpet_enabled()) return 0; if (freq <= DEFAULT_RTC_INT_FREQ) hpet_pie_limit = DEFAULT_RTC_INT_FREQ / freq; else { clc = (uint64_t) hpet_clockevent.mult * NSEC_PER_SEC; do_div(clc, freq); clc >>= hpet_clockevent.shift; hpet_pie_delta = (unsigned long) clc; } return 1; } If freq is set to 64Hz (DEFAULT_RTC_INT_FREQ) or lower, then hpet_pie_limit (a static) is set to non-zero. Then, on every one-shot HPET interrupt, hpet_rtc_timer_reinit is called to compute the next timeout. Well, that function has this logic: if (!(hpet_rtc_flags & RTC_PIE) || hpet_pie_limit) delta = hpet_default_delta; else delta = hpet_pie_delta; Since hpet_pie_limit is not 0, hpet_default_delta is used. That corresponds to 64Hz. Now, if you set a different rtc frequency, you'll take the else path through hpet_set_periodic_freq, but unfortunately no one resets hpet_pie_limit back to 0. Boom....now you are stuck with 64Hz RTC interrupts forever. The patch below just resets the hpet_pie_limit value when requested freq is greater than DEFAULT_RTC_INT_FREQ, which we think fixes this problem. Signed-off-by: Alok N Kataria LKML-Reference: <201003112200.o2BM0Hre012875@imap1.linux-foundation.org> Signed-off-by: Daniel Hecht Cc: Venkatesh Pallipadi Cc: Thomas Gleixner Cc: "H. Peter Anvin" Cc: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: H. Peter Anvin diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c index 3d422da..2bda5f00 100644 --- a/arch/x86/kernel/hpet.c +++ b/arch/x86/kernel/hpet.c @@ -1149,6 +1149,7 @@ int hpet_set_periodic_freq(unsigned long freq) do_div(clc, freq); clc >>= hpet_clockevent.shift; hpet_pie_delta = clc; + hpet_pie_limit = 0; } return 1; } -- cgit v0.10.2 From 042be38e6106ed70b42d096ab4a1ed4187e510e6 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Thu, 1 Apr 2010 14:32:43 -0700 Subject: ibft, x86: Change reserve_ibft_region() to find_ibft_region() This allows arch code could decide the way to reserve the ibft. And we should reserve ibft as early as possible, instead of BOOTMEM stage, in case the table is in RAM range and is not reserved by BIOS (this will often be the case.) Move to just after find_smp_config(). Also when CONFIG_NO_BOOTMEM=y, We will not have reserve_bootmem() anymore. -v2: fix typo about ibft pointed by Konrad Rzeszutek Wilk Signed-off-by: Yinghai Lu LKML-Reference: <4BB510FB.80601@kernel.org> Cc: Pekka Enberg Cc: Peter Jones Cc: Konrad Rzeszutek Wilk CC: Jan Beulich Signed-off-by: H. Peter Anvin diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index d76e185..580e6b3 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -608,6 +608,16 @@ static int __init setup_elfcorehdr(char *arg) early_param("elfcorehdr", setup_elfcorehdr); #endif +static __init void reserve_ibft_region(void) +{ + unsigned long addr, size = 0; + + addr = find_ibft_region(&size); + + if (size) + reserve_early_overlap_ok(addr, addr + size, "ibft"); +} + #ifdef CONFIG_X86_RESERVE_LOW_64K static int __init dmi_low_memory_corruption(const struct dmi_system_id *d) { @@ -910,6 +920,8 @@ void __init setup_arch(char **cmdline_p) */ find_smp_config(); + reserve_ibft_region(); + reserve_trampoline_memory(); #ifdef CONFIG_ACPI_SLEEP @@ -977,8 +989,6 @@ void __init setup_arch(char **cmdline_p) dma32_reserve_bootmem(); - reserve_ibft_region(); - #ifdef CONFIG_KVM_CLOCK kvmclock_init(); #endif diff --git a/drivers/firmware/iscsi_ibft_find.c b/drivers/firmware/iscsi_ibft_find.c index dfb15c0..8f5d9e2 100644 --- a/drivers/firmware/iscsi_ibft_find.c +++ b/drivers/firmware/iscsi_ibft_find.c @@ -52,7 +52,7 @@ EXPORT_SYMBOL_GPL(ibft_addr); * Routine used to find the iSCSI Boot Format Table. The logical * kernel address is set in the ibft_addr global variable. */ -void __init reserve_ibft_region(void) +unsigned long __init find_ibft_region(unsigned long *sizep) { unsigned long pos; unsigned int len = 0; @@ -78,6 +78,11 @@ void __init reserve_ibft_region(void) } } } - if (ibft_addr) - reserve_bootmem(pos, PAGE_ALIGN(len), BOOTMEM_DEFAULT); + if (ibft_addr) { + *sizep = PAGE_ALIGN(len); + return pos; + } + + *sizep = 0; + return 0; } diff --git a/include/linux/iscsi_ibft.h b/include/linux/iscsi_ibft.h index 6092487..d2e4042 100644 --- a/include/linux/iscsi_ibft.h +++ b/include/linux/iscsi_ibft.h @@ -42,9 +42,13 @@ extern struct ibft_table_header *ibft_addr; * mapped address is set in the ibft_addr variable. */ #ifdef CONFIG_ISCSI_IBFT_FIND -extern void __init reserve_ibft_region(void); +unsigned long find_ibft_region(unsigned long *sizep); #else -static inline void reserve_ibft_region(void) { } +static inline unsigned long find_ibft_region(unsigned long *sizep) +{ + *sizep = 0; + return 0; +} #endif #endif /* ISCSI_IBFT_H */ -- cgit v0.10.2 From 61917bdaaf6bea4b885525cf63f65272914f6be2 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Fri, 2 Apr 2010 08:39:40 +0200 Subject: cciss: unlock on error path We take the spin_lock again in fail_all_cmds() so we need to unlock here. Signed-off-by: Dan Carpenter Signed-off-by: Jens Axboe diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index 9e3af30..eb5ff05 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -3341,6 +3341,7 @@ static irqreturn_t do_cciss_intr(int irq, void *dev_id) printk(KERN_WARNING "cciss: controller cciss%d failed, stopping.\n", h->ctlr); + spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); fail_all_cmds(h->ctlr); return IRQ_HANDLED; } -- cgit v0.10.2 From b2b163dd47024e445410b72d0c5df6d819c14dfd Mon Sep 17 00:00:00 2001 From: Philipp Reisner Date: Fri, 2 Apr 2010 08:40:33 +0200 Subject: drbd: lc_element_by_index() never returns NULL Signed-off-by: Philipp Reisner Signed-off-by: Lars Ellenberg Signed-off-by: Jens Axboe diff --git a/drivers/block/drbd/drbd_actlog.c b/drivers/block/drbd/drbd_actlog.c index 43e57f3..df01899 100644 --- a/drivers/block/drbd/drbd_actlog.c +++ b/drivers/block/drbd/drbd_actlog.c @@ -1336,7 +1336,7 @@ int drbd_rs_del_all(struct drbd_conf *mdev) /* ok, ->resync is there. */ for (i = 0; i < mdev->resync->nr_elements; i++) { e = lc_element_by_index(mdev->resync, i); - bm_ext = e ? lc_entry(e, struct bm_extent, lce) : NULL; + bm_ext = lc_entry(e, struct bm_extent, lce); if (bm_ext->lce.lc_number == LC_FREE) continue; if (bm_ext->lce.lc_number == mdev->resync_wenr) { -- cgit v0.10.2 From a506aedc51093544ff0f9610af6066d18cb6abbe Mon Sep 17 00:00:00 2001 From: "wzt.wzt@gmail.com" Date: Fri, 2 Apr 2010 08:41:14 +0200 Subject: Block: Fix block/elevator.c elevator_get() off-by-one error elevator_get() not check the name length, if the name length > sizeof(elv), elv will miss the '\0'. And elv buffer will be replace "-iosched" as something like aaaaaaaaa, then call request_module() can load an not trust module. Signed-off-by: Zhitong Wang Signed-off-by: Jens Axboe diff --git a/block/elevator.c b/block/elevator.c index df75676..76e3702 100644 --- a/block/elevator.c +++ b/block/elevator.c @@ -154,7 +154,7 @@ static struct elevator_type *elevator_get(const char *name) spin_unlock(&elv_list_lock); - sprintf(elv, "%s-iosched", name); + snprintf(elv, sizeof(elv), "%s-iosched", name); request_module("%s", elv); spin_lock(&elv_list_lock); -- cgit v0.10.2 From 144214537370b4f133a735446ebe86e90cfb2501 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Fri, 2 Apr 2010 09:46:55 +0200 Subject: backing-dev: Handle class_create() failure I hit this when we had a bug in IDR for a few days. Basically sysfs would fail to create new inodes since it uses an IDR and therefore class_create would fail. While we are unlikely to see this fail we may as well handle it instead of oopsing. Signed-off-by: Anton Blanchard Signed-off-by: Jens Axboe diff --git a/mm/backing-dev.c b/mm/backing-dev.c index 0e8ca03..f13e067 100644 --- a/mm/backing-dev.c +++ b/mm/backing-dev.c @@ -227,6 +227,9 @@ static struct device_attribute bdi_dev_attrs[] = { static __init int bdi_class_init(void) { bdi_class = class_create(THIS_MODULE, "bdi"); + if (IS_ERR(bdi_class)) + return PTR_ERR(bdi_class); + bdi_class->dev_attrs = bdi_dev_attrs; bdi_debug_init(); return 0; -- cgit v0.10.2 From 308f44193f796b1c522b3b87760e43d8d8e316d2 Mon Sep 17 00:00:00 2001 From: Li Hong Date: Fri, 2 Apr 2010 18:40:39 +0800 Subject: nilfs2: Remove an uninitialization warning in nilfs_btree_propagate_v() `make CONFIG_NILFS2_FS=m M=fs/nilfs2/` will give the following warnings: fs/nilfs2/btree.c: In function 'nilfs_btree_propagate': fs/nilfs2/btree.c:1882: warning: 'maxlevel' may be used uninitialized in this function fs/nilfs2/btree.c:1882: note: 'maxlevel' was declared here Set maxlevel = 0 to fix it. Signed-off-by: Li Hong Signed-off-by: Ryusuke Konishi diff --git a/fs/nilfs2/btree.c b/fs/nilfs2/btree.c index 7cdd98b..76c38e3 100644 --- a/fs/nilfs2/btree.c +++ b/fs/nilfs2/btree.c @@ -1879,7 +1879,7 @@ static int nilfs_btree_propagate_v(struct nilfs_btree *btree, struct nilfs_btree_path *path, int level, struct buffer_head *bh) { - int maxlevel, ret; + int maxlevel = 0, ret; struct nilfs_btree_node *parent; struct inode *dat = nilfs_bmap_get_dat(&btree->bt_bmap); __u64 ptr; -- cgit v0.10.2 From b5442a75deee293d10c2ab8f4a77013973c4c9e0 Mon Sep 17 00:00:00 2001 From: Janusz Krzysztofik Date: Sun, 28 Mar 2010 22:29:29 +0200 Subject: ASoC: OMAP: Fix capture pointer handling for OMAP1510 to work correctly with recent ALSA PCM code With recent (2.6.34) chnages in PCM handling, capture stopped working on my OMAP1510 based Amstrad Delta videophone. Using 2.6.34-rc2, I was able to correct the problem in 3 different ways: 1. reverting commit 7b3a177b0d4f92b3431b8dca777313a07533a710, 2. enabling additional jiffies check with echo 4 >/proc/asound/card0/pcm0c0/xrun_debug 3. applying the patch below. Since I wasn't able to reproduce the problem on my i686 PC, I guess the problem is probably machine specific. The patch reuses the method for software emulation of missing hardware pointer, already implemented for playback on OMAP1510. It's possible that event if a hardware pointer is available for capture on this machine, its behaviour may be not compatible with what upper layer expects. If you think the problem may be more general and should be solved differently, on a higher level, I can try to work more on it if you give me a hint. If the patch gets accepted, I suggest it goes as a fix in the current release cycle. Created and tested against linux-2.6.34-rc2. Signed-off-by: Janusz Krzysztofik Acked-by: Jarkko Nikula Acked-by: Liam Girdwood Signed-off-by: Mark Brown diff --git a/sound/soc/omap/omap-pcm.c b/sound/soc/omap/omap-pcm.c index 825db38..bdd1097 100644 --- a/sound/soc/omap/omap-pcm.c +++ b/sound/soc/omap/omap-pcm.c @@ -60,12 +60,11 @@ static void omap_pcm_dma_irq(int ch, u16 stat, void *data) struct omap_runtime_data *prtd = runtime->private_data; unsigned long flags; - if ((cpu_is_omap1510()) && - (substream->stream == SNDRV_PCM_STREAM_PLAYBACK)) { + if ((cpu_is_omap1510())) { /* * OMAP1510 doesn't fully support DMA progress counter * and there is no software emulation implemented yet, - * so have to maintain our own playback progress counter + * so have to maintain our own progress counters * that can be used by omap_pcm_pointer() instead. */ spin_lock_irqsave(&prtd->lock, flags); @@ -189,8 +188,7 @@ static int omap_pcm_prepare(struct snd_pcm_substream *substream) dma_params.frame_count = runtime->periods; omap_set_dma_params(prtd->dma_ch, &dma_params); - if ((cpu_is_omap1510()) && - (substream->stream == SNDRV_PCM_STREAM_PLAYBACK)) + if ((cpu_is_omap1510())) omap_enable_dma_irq(prtd->dma_ch, OMAP_DMA_FRAME_IRQ | OMAP_DMA_LAST_IRQ | OMAP_DMA_BLOCK_IRQ); else @@ -248,14 +246,15 @@ static snd_pcm_uframes_t omap_pcm_pointer(struct snd_pcm_substream *substream) dma_addr_t ptr; snd_pcm_uframes_t offset; - if (substream->stream == SNDRV_PCM_STREAM_CAPTURE) { + if (cpu_is_omap1510()) { + offset = prtd->period_index * runtime->period_size; + } else if (substream->stream == SNDRV_PCM_STREAM_CAPTURE) { ptr = omap_get_dma_dst_pos(prtd->dma_ch); offset = bytes_to_frames(runtime, ptr - runtime->dma_addr); - } else if (!(cpu_is_omap1510())) { + } else { ptr = omap_get_dma_src_pos(prtd->dma_ch); offset = bytes_to_frames(runtime, ptr - runtime->dma_addr); - } else - offset = prtd->period_index * runtime->period_size; + } if (offset >= runtime->buffer_size) offset = 0; -- cgit v0.10.2 From 51591e31dcb3716f03f962e26ec36a029aa46340 Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Thu, 25 Mar 2010 15:39:27 -0700 Subject: x86: Increase CONFIG_NODES_SHIFT max to 10 Some larger systems require more than 512 nodes, so increase the maximum CONFIG_NODES_SHIFT to 10 for a new max of 1024 nodes. This was tested with numa=fake=64M on systems with more than 64GB of RAM. A total of 1022 nodes were initialized. Successfully builds with no additional warnings on x86_64 allyesconfig. ( No effect on any existing config. Newly enabled CONFIG_MAXSMP=y will see the new default. ) Signed-off-by: David Rientjes LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 0eacb1f..9458685 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -1216,8 +1216,8 @@ config NUMA_EMU config NODES_SHIFT int "Maximum NUMA Nodes (as a power of 2)" if !MAXSMP - range 1 9 - default "9" if MAXSMP + range 1 10 + default "10" if MAXSMP default "6" if X86_64 default "4" if X86_NUMAQ default "3" -- cgit v0.10.2 From 085ea739adf107b5a5d131f3625e517ff4a5181e Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 2 Apr 2010 12:50:39 -0400 Subject: perf probe: Fix --line syntax help and document Just fix typos. --line option accepts ':START-END' syntax, not ':START:END'. Signed-off-by: Masami Hiramatsu Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo Cc: Peter Zijlstra Cc: Mike Galbraith Cc: Frederic Weisbecker Cc: systemtap Cc: DLE LKML-Reference: <20100402165038.23551.62590.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/Documentation/perf-probe.txt b/tools/perf/Documentation/perf-probe.txt index 0f944b3..bb671b3 100644 --- a/tools/perf/Documentation/perf-probe.txt +++ b/tools/perf/Documentation/perf-probe.txt @@ -85,7 +85,7 @@ LINE SYNTAX ----------- Line range is descripted by following syntax. - "FUNC[:RLN[+NUM|:RLN2]]|SRC:ALN[+NUM|:ALN2]" + "FUNC[:RLN[+NUM|-RLN2]]|SRC:ALN[+NUM|-ALN2]" FUNC specifies the function name of showing lines. 'RLN' is the start line number from function entry line, and 'RLN2' is the end line number. As same as diff --git a/tools/perf/builtin-probe.c b/tools/perf/builtin-probe.c index cf2ffa5..b3ba25a 100644 --- a/tools/perf/builtin-probe.c +++ b/tools/perf/builtin-probe.c @@ -167,7 +167,7 @@ static const struct option options[] = { " with existing name"), #ifdef DWARF_SUPPORT OPT_CALLBACK('L', "line", NULL, - "FUNC[:RLN[+NUM|:RLN2]]|SRC:ALN[+NUM|:ALN2]", + "FUNC[:RLN[+NUM|-RLN2]]|SRC:ALN[+NUM|-ALN2]", "Show source code lines.", opt_show_lines), OPT_STRING('k', "vmlinux", &symbol_conf.vmlinux_name, "file", "vmlinux pathname"), -- cgit v0.10.2 From c9e385826d4f1ca5a72005ab8503598f791a8dc0 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 2 Apr 2010 12:50:45 -0400 Subject: perf probe: Fix not to return non-matched file Fix cu_find_realpath() not to return the last file path if that is not matched to input pattern. Signed-off-by: Masami Hiramatsu Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo Cc: Peter Zijlstra Cc: Mike Galbraith Cc: Frederic Weisbecker Cc: systemtap Cc: DLE LKML-Reference: <20100402165045.23551.47780.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index db52ec2..b44132e 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -183,6 +183,8 @@ static const char *cu_find_realpath(Dwarf_Die *cu_die, const char *fname) if (strtailcmp(src, fname) == 0) break; } + if (i == nfiles) + return NULL; return src; } -- cgit v0.10.2 From 12e5a7ae475ccb2733d740ffb95d9ca0a18392da Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 2 Apr 2010 12:50:53 -0400 Subject: perf probe: Correct error message for non-structure type perf probe outputs incorrect error message when it is called with non-existent field on a non-data structure local variable. # perf probe vfs_read 'count.hoge' Fatal: Structure on a register is not supported yet. # perf probe vfs_read 'count->hoge' Fatal: Semantic error: hoge must be referred by '.' This corrects the messsage. # perf probe vfs_read 'count.hoge' Fatal: count is not a data structure. # perf probe vfs_read 'count->hoge' Fatal: count is not a data structure. Signed-off-by: Masami Hiramatsu Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo Cc: Peter Zijlstra Cc: Mike Galbraith Cc: Frederic Weisbecker Cc: systemtap Cc: DLE LKML-Reference: <20100402165052.23551.75866.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index b44132e..59b0115 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -429,12 +429,20 @@ static void convert_variable_fields(Dwarf_Die *vr_die, const char *varname, if (die_get_real_type(&type, &type) == NULL) die("Failed to get a type information of %s.", varname); + /* Verify it is a data structure */ + if (dwarf_tag(&type) != DW_TAG_structure_type) + die("%s is not a data structure.", varname); + ref = xzalloc(sizeof(struct kprobe_trace_arg_ref)); if (*ref_ptr) (*ref_ptr)->next = ref; else *ref_ptr = ref; } else { + /* Verify it is a data structure */ + if (dwarf_tag(&type) != DW_TAG_structure_type) + die("%s is not a data structure.", varname); + if (field->ref) die("Semantic error: %s must be referred by '.'", field->name); @@ -442,10 +450,6 @@ static void convert_variable_fields(Dwarf_Die *vr_die, const char *varname, die("Structure on a register is not supported yet."); } - /* Verify it is a data structure */ - if (dwarf_tag(&type) != DW_TAG_structure_type) - die("%s is not a data structure.", varname); - if (die_find_member(&type, field->name, &member) == NULL) die("%s(tyep:%s) has no member %s.", varname, dwarf_diename(&type), field->name); -- cgit v0.10.2 From 75ec5a245c7763c397f31ec8964d0a46c54a7386 Mon Sep 17 00:00:00 2001 From: Masami Hiramatsu Date: Fri, 2 Apr 2010 12:50:59 -0400 Subject: perf probe: Fix to close dwarf when failing to analyze it Fix to close libdw routine when failing to analyze it in find_perf_probe_point(). Signed-off-by: Masami Hiramatsu Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo Cc: Peter Zijlstra Cc: Mike Galbraith Cc: Frederic Weisbecker Cc: systemtap Cc: DLE LKML-Reference: <20100402165059.23551.95587.stgit@localhost6.localdomain6> Signed-off-by: Ingo Molnar diff --git a/tools/perf/util/probe-finder.c b/tools/perf/util/probe-finder.c index 59b0115..a851377 100644 --- a/tools/perf/util/probe-finder.c +++ b/tools/perf/util/probe-finder.c @@ -818,8 +818,10 @@ int find_perf_probe_point(int fd, unsigned long addr, return -ENOENT; /* Find cu die */ - if (!dwarf_addrdie(dbg, (Dwarf_Addr)addr, &cudie)) - return -EINVAL; + if (!dwarf_addrdie(dbg, (Dwarf_Addr)addr, &cudie)) { + ret = -EINVAL; + goto end; + } /* Find a corresponding line */ line = dwarf_getsrc_die(&cudie, (Dwarf_Addr)addr); -- cgit v0.10.2 From 948b1bb89a44561560531394c18da4a99215f772 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Mon, 29 Mar 2010 18:36:50 +0200 Subject: perf, x86: Undo some some *_counter* -> *_event* renames The big rename: cdd6c48 perf: Do the big rename: Performance Counters -> Performance Events accidentally renamed some members of stucts that were named after registers in the spec. To avoid confusion this patch reverts some changes. The related specs are MSR descriptions in AMD's BKDGs and the ARCHITECTURAL PERFORMANCE MONITORING section in the Intel 64 and IA-32 Architectures Software Developer's Manuals. This patch does: $ sed -i -e 's:num_events:num_counters:g' \ arch/x86/include/asm/perf_event.h \ arch/x86/kernel/cpu/perf_event_amd.c \ arch/x86/kernel/cpu/perf_event.c \ arch/x86/kernel/cpu/perf_event_intel.c \ arch/x86/kernel/cpu/perf_event_p6.c \ arch/x86/kernel/cpu/perf_event_p4.c \ arch/x86/oprofile/op_model_ppro.c $ sed -i -e 's:event_bits:cntval_bits:g' -e 's:event_mask:cntval_mask:g' \ arch/x86/kernel/cpu/perf_event_amd.c \ arch/x86/kernel/cpu/perf_event.c \ arch/x86/kernel/cpu/perf_event_intel.c \ arch/x86/kernel/cpu/perf_event_p6.c \ arch/x86/kernel/cpu/perf_event_p4.c Signed-off-by: Robert Richter Signed-off-by: Peter Zijlstra LKML-Reference: <1269880612-25800-2-git-send-email-robert.richter@amd.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/perf_event.h b/arch/x86/include/asm/perf_event.h index 124dddd..987bf67 100644 --- a/arch/x86/include/asm/perf_event.h +++ b/arch/x86/include/asm/perf_event.h @@ -67,7 +67,7 @@ union cpuid10_eax { struct { unsigned int version_id:8; - unsigned int num_events:8; + unsigned int num_counters:8; unsigned int bit_width:8; unsigned int mask_length:8; } split; @@ -76,7 +76,7 @@ union cpuid10_eax { union cpuid10_edx { struct { - unsigned int num_events_fixed:4; + unsigned int num_counters_fixed:4; unsigned int reserved:28; } split; unsigned int full; diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index b534356..9daaa1e 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -195,10 +195,10 @@ struct x86_pmu { u64 (*event_map)(int); u64 (*raw_event)(u64); int max_events; - int num_events; - int num_events_fixed; - int event_bits; - u64 event_mask; + int num_counters; + int num_counters_fixed; + int cntval_bits; + u64 cntval_mask; int apic; u64 max_period; struct event_constraint * @@ -268,7 +268,7 @@ static u64 x86_perf_event_update(struct perf_event *event) { struct hw_perf_event *hwc = &event->hw; - int shift = 64 - x86_pmu.event_bits; + int shift = 64 - x86_pmu.cntval_bits; u64 prev_raw_count, new_raw_count; int idx = hwc->idx; s64 delta; @@ -320,12 +320,12 @@ static bool reserve_pmc_hardware(void) if (nmi_watchdog == NMI_LOCAL_APIC) disable_lapic_nmi_watchdog(); - for (i = 0; i < x86_pmu.num_events; i++) { + for (i = 0; i < x86_pmu.num_counters; i++) { if (!reserve_perfctr_nmi(x86_pmu.perfctr + i)) goto perfctr_fail; } - for (i = 0; i < x86_pmu.num_events; i++) { + for (i = 0; i < x86_pmu.num_counters; i++) { if (!reserve_evntsel_nmi(x86_pmu.eventsel + i)) goto eventsel_fail; } @@ -336,7 +336,7 @@ eventsel_fail: for (i--; i >= 0; i--) release_evntsel_nmi(x86_pmu.eventsel + i); - i = x86_pmu.num_events; + i = x86_pmu.num_counters; perfctr_fail: for (i--; i >= 0; i--) @@ -352,7 +352,7 @@ static void release_pmc_hardware(void) { int i; - for (i = 0; i < x86_pmu.num_events; i++) { + for (i = 0; i < x86_pmu.num_counters; i++) { release_perfctr_nmi(x86_pmu.perfctr + i); release_evntsel_nmi(x86_pmu.eventsel + i); } @@ -547,7 +547,7 @@ static void x86_pmu_disable_all(void) struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); int idx; - for (idx = 0; idx < x86_pmu.num_events; idx++) { + for (idx = 0; idx < x86_pmu.num_counters; idx++) { u64 val; if (!test_bit(idx, cpuc->active_mask)) @@ -582,7 +582,7 @@ static void x86_pmu_enable_all(int added) struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); int idx; - for (idx = 0; idx < x86_pmu.num_events; idx++) { + for (idx = 0; idx < x86_pmu.num_counters; idx++) { struct perf_event *event = cpuc->events[idx]; u64 val; @@ -657,14 +657,14 @@ static int x86_schedule_events(struct cpu_hw_events *cpuc, int n, int *assign) * assign events to counters starting with most * constrained events. */ - wmax = x86_pmu.num_events; + wmax = x86_pmu.num_counters; /* * when fixed event counters are present, * wmax is incremented by 1 to account * for one more choice */ - if (x86_pmu.num_events_fixed) + if (x86_pmu.num_counters_fixed) wmax++; for (w = 1, num = n; num && w <= wmax; w++) { @@ -714,7 +714,7 @@ static int collect_events(struct cpu_hw_events *cpuc, struct perf_event *leader, struct perf_event *event; int n, max_count; - max_count = x86_pmu.num_events + x86_pmu.num_events_fixed; + max_count = x86_pmu.num_counters + x86_pmu.num_counters_fixed; /* current number of events already accepted */ n = cpuc->n_events; @@ -904,7 +904,7 @@ x86_perf_event_set_period(struct perf_event *event) atomic64_set(&hwc->prev_count, (u64)-left); wrmsrl(hwc->event_base + idx, - (u64)(-left) & x86_pmu.event_mask); + (u64)(-left) & x86_pmu.cntval_mask); perf_event_update_userpage(event); @@ -987,7 +987,7 @@ void perf_event_print_debug(void) unsigned long flags; int cpu, idx; - if (!x86_pmu.num_events) + if (!x86_pmu.num_counters) return; local_irq_save(flags); @@ -1011,7 +1011,7 @@ void perf_event_print_debug(void) } pr_info("CPU#%d: active: %016llx\n", cpu, *(u64 *)cpuc->active_mask); - for (idx = 0; idx < x86_pmu.num_events; idx++) { + for (idx = 0; idx < x86_pmu.num_counters; idx++) { rdmsrl(x86_pmu.eventsel + idx, pmc_ctrl); rdmsrl(x86_pmu.perfctr + idx, pmc_count); @@ -1024,7 +1024,7 @@ void perf_event_print_debug(void) pr_info("CPU#%d: gen-PMC%d left: %016llx\n", cpu, idx, prev_left); } - for (idx = 0; idx < x86_pmu.num_events_fixed; idx++) { + for (idx = 0; idx < x86_pmu.num_counters_fixed; idx++) { rdmsrl(MSR_ARCH_PERFMON_FIXED_CTR0 + idx, pmc_count); pr_info("CPU#%d: fixed-PMC%d count: %016llx\n", @@ -1089,7 +1089,7 @@ static int x86_pmu_handle_irq(struct pt_regs *regs) cpuc = &__get_cpu_var(cpu_hw_events); - for (idx = 0; idx < x86_pmu.num_events; idx++) { + for (idx = 0; idx < x86_pmu.num_counters; idx++) { if (!test_bit(idx, cpuc->active_mask)) continue; @@ -1097,7 +1097,7 @@ static int x86_pmu_handle_irq(struct pt_regs *regs) hwc = &event->hw; val = x86_perf_event_update(event); - if (val & (1ULL << (x86_pmu.event_bits - 1))) + if (val & (1ULL << (x86_pmu.cntval_bits - 1))) continue; /* @@ -1401,46 +1401,46 @@ void __init init_hw_perf_events(void) if (x86_pmu.quirks) x86_pmu.quirks(); - if (x86_pmu.num_events > X86_PMC_MAX_GENERIC) { + if (x86_pmu.num_counters > X86_PMC_MAX_GENERIC) { WARN(1, KERN_ERR "hw perf events %d > max(%d), clipping!", - x86_pmu.num_events, X86_PMC_MAX_GENERIC); - x86_pmu.num_events = X86_PMC_MAX_GENERIC; + x86_pmu.num_counters, X86_PMC_MAX_GENERIC); + x86_pmu.num_counters = X86_PMC_MAX_GENERIC; } - x86_pmu.intel_ctrl = (1 << x86_pmu.num_events) - 1; - perf_max_events = x86_pmu.num_events; + x86_pmu.intel_ctrl = (1 << x86_pmu.num_counters) - 1; + perf_max_events = x86_pmu.num_counters; - if (x86_pmu.num_events_fixed > X86_PMC_MAX_FIXED) { + if (x86_pmu.num_counters_fixed > X86_PMC_MAX_FIXED) { WARN(1, KERN_ERR "hw perf events fixed %d > max(%d), clipping!", - x86_pmu.num_events_fixed, X86_PMC_MAX_FIXED); - x86_pmu.num_events_fixed = X86_PMC_MAX_FIXED; + x86_pmu.num_counters_fixed, X86_PMC_MAX_FIXED); + x86_pmu.num_counters_fixed = X86_PMC_MAX_FIXED; } x86_pmu.intel_ctrl |= - ((1LL << x86_pmu.num_events_fixed)-1) << X86_PMC_IDX_FIXED; + ((1LL << x86_pmu.num_counters_fixed)-1) << X86_PMC_IDX_FIXED; perf_events_lapic_init(); register_die_notifier(&perf_event_nmi_notifier); unconstrained = (struct event_constraint) - __EVENT_CONSTRAINT(0, (1ULL << x86_pmu.num_events) - 1, - 0, x86_pmu.num_events); + __EVENT_CONSTRAINT(0, (1ULL << x86_pmu.num_counters) - 1, + 0, x86_pmu.num_counters); if (x86_pmu.event_constraints) { for_each_event_constraint(c, x86_pmu.event_constraints) { if (c->cmask != INTEL_ARCH_FIXED_MASK) continue; - c->idxmsk64 |= (1ULL << x86_pmu.num_events) - 1; - c->weight += x86_pmu.num_events; + c->idxmsk64 |= (1ULL << x86_pmu.num_counters) - 1; + c->weight += x86_pmu.num_counters; } } pr_info("... version: %d\n", x86_pmu.version); - pr_info("... bit width: %d\n", x86_pmu.event_bits); - pr_info("... generic registers: %d\n", x86_pmu.num_events); - pr_info("... value mask: %016Lx\n", x86_pmu.event_mask); + pr_info("... bit width: %d\n", x86_pmu.cntval_bits); + pr_info("... generic registers: %d\n", x86_pmu.num_counters); + pr_info("... value mask: %016Lx\n", x86_pmu.cntval_mask); pr_info("... max period: %016Lx\n", x86_pmu.max_period); - pr_info("... fixed-purpose events: %d\n", x86_pmu.num_events_fixed); + pr_info("... fixed-purpose events: %d\n", x86_pmu.num_counters_fixed); pr_info("... event mask: %016Lx\n", x86_pmu.intel_ctrl); perf_cpu_notifier(x86_pmu_notifier); diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c index 285623b..7753a5c 100644 --- a/arch/x86/kernel/cpu/perf_event_amd.c +++ b/arch/x86/kernel/cpu/perf_event_amd.c @@ -165,7 +165,7 @@ static void amd_put_event_constraints(struct cpu_hw_events *cpuc, * be removed on one CPU at a time AND PMU is disabled * when we come here */ - for (i = 0; i < x86_pmu.num_events; i++) { + for (i = 0; i < x86_pmu.num_counters; i++) { if (nb->owners[i] == event) { cmpxchg(nb->owners+i, event, NULL); break; @@ -215,7 +215,7 @@ amd_get_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event) struct hw_perf_event *hwc = &event->hw; struct amd_nb *nb = cpuc->amd_nb; struct perf_event *old = NULL; - int max = x86_pmu.num_events; + int max = x86_pmu.num_counters; int i, j, k = -1; /* @@ -293,7 +293,7 @@ static struct amd_nb *amd_alloc_nb(int cpu, int nb_id) /* * initialize all possible NB constraints */ - for (i = 0; i < x86_pmu.num_events; i++) { + for (i = 0; i < x86_pmu.num_counters; i++) { __set_bit(i, nb->event_constraints[i].idxmsk); nb->event_constraints[i].weight = 1; } @@ -385,9 +385,9 @@ static __initconst struct x86_pmu amd_pmu = { .event_map = amd_pmu_event_map, .raw_event = amd_pmu_raw_event, .max_events = ARRAY_SIZE(amd_perfmon_event_map), - .num_events = 4, - .event_bits = 48, - .event_mask = (1ULL << 48) - 1, + .num_counters = 4, + .cntval_bits = 48, + .cntval_mask = (1ULL << 48) - 1, .apic = 1, /* use highest bit to detect overflow */ .max_period = (1ULL << 47) - 1, diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 676aac2..cc4d90a 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -653,20 +653,20 @@ static void intel_pmu_reset(void) unsigned long flags; int idx; - if (!x86_pmu.num_events) + if (!x86_pmu.num_counters) return; local_irq_save(flags); printk("clearing PMU state on CPU#%d\n", smp_processor_id()); - for (idx = 0; idx < x86_pmu.num_events; idx++) { + for (idx = 0; idx < x86_pmu.num_counters; idx++) { checking_wrmsrl(x86_pmu.eventsel + idx, 0ull); checking_wrmsrl(x86_pmu.perfctr + idx, 0ull); } - for (idx = 0; idx < x86_pmu.num_events_fixed; idx++) { + for (idx = 0; idx < x86_pmu.num_counters_fixed; idx++) checking_wrmsrl(MSR_ARCH_PERFMON_FIXED_CTR0 + idx, 0ull); - } + if (ds) ds->bts_index = ds->bts_buffer_base; @@ -901,16 +901,16 @@ static __init int intel_pmu_init(void) x86_pmu = intel_pmu; x86_pmu.version = version; - x86_pmu.num_events = eax.split.num_events; - x86_pmu.event_bits = eax.split.bit_width; - x86_pmu.event_mask = (1ULL << eax.split.bit_width) - 1; + x86_pmu.num_counters = eax.split.num_counters; + x86_pmu.cntval_bits = eax.split.bit_width; + x86_pmu.cntval_mask = (1ULL << eax.split.bit_width) - 1; /* * Quirk: v2 perfmon does not report fixed-purpose events, so * assume at least 3 events: */ if (version > 1) - x86_pmu.num_events_fixed = max((int)edx.split.num_events_fixed, 3); + x86_pmu.num_counters_fixed = max((int)edx.split.num_counters_fixed, 3); /* * v2 and above have a perf capabilities MSR diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index 0d1be36..4139100 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -483,7 +483,7 @@ static void p4_pmu_disable_all(void) struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); int idx; - for (idx = 0; idx < x86_pmu.num_events; idx++) { + for (idx = 0; idx < x86_pmu.num_counters; idx++) { struct perf_event *event = cpuc->events[idx]; if (!test_bit(idx, cpuc->active_mask)) continue; @@ -540,7 +540,7 @@ static void p4_pmu_enable_all(int added) struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); int idx; - for (idx = 0; idx < x86_pmu.num_events; idx++) { + for (idx = 0; idx < x86_pmu.num_counters; idx++) { struct perf_event *event = cpuc->events[idx]; if (!test_bit(idx, cpuc->active_mask)) continue; @@ -562,7 +562,7 @@ static int p4_pmu_handle_irq(struct pt_regs *regs) cpuc = &__get_cpu_var(cpu_hw_events); - for (idx = 0; idx < x86_pmu.num_events; idx++) { + for (idx = 0; idx < x86_pmu.num_counters; idx++) { if (!test_bit(idx, cpuc->active_mask)) continue; @@ -579,7 +579,7 @@ static int p4_pmu_handle_irq(struct pt_regs *regs) p4_pmu_clear_cccr_ovf(hwc); val = x86_perf_event_update(event); - if (val & (1ULL << (x86_pmu.event_bits - 1))) + if (val & (1ULL << (x86_pmu.cntval_bits - 1))) continue; /* @@ -794,10 +794,10 @@ static __initconst struct x86_pmu p4_pmu = { * though leave it restricted at moment assuming * HT is on */ - .num_events = ARCH_P4_MAX_CCCR, + .num_counters = ARCH_P4_MAX_CCCR, .apic = 1, - .event_bits = 40, - .event_mask = (1ULL << 40) - 1, + .cntval_bits = 40, + .cntval_mask = (1ULL << 40) - 1, .max_period = (1ULL << 39) - 1, .hw_config = p4_hw_config, .schedule_events = p4_pmu_schedule_events, diff --git a/arch/x86/kernel/cpu/perf_event_p6.c b/arch/x86/kernel/cpu/perf_event_p6.c index 877182c..b26fbc7e 100644 --- a/arch/x86/kernel/cpu/perf_event_p6.c +++ b/arch/x86/kernel/cpu/perf_event_p6.c @@ -119,7 +119,7 @@ static __initconst struct x86_pmu p6_pmu = { .apic = 1, .max_period = (1ULL << 31) - 1, .version = 0, - .num_events = 2, + .num_counters = 2, /* * Events have 40 bits implemented. However they are designed such * that bits [32-39] are sign extensions of bit 31. As such the @@ -127,8 +127,8 @@ static __initconst struct x86_pmu p6_pmu = { * * See IA-32 Intel Architecture Software developer manual Vol 3B */ - .event_bits = 32, - .event_mask = (1ULL << 32) - 1, + .cntval_bits = 32, + .cntval_mask = (1ULL << 32) - 1, .get_event_constraints = x86_get_event_constraints, .event_constraints = p6_event_constraints, }; diff --git a/arch/x86/oprofile/op_model_ppro.c b/arch/x86/oprofile/op_model_ppro.c index 2bf90fa..c8abc4d 100644 --- a/arch/x86/oprofile/op_model_ppro.c +++ b/arch/x86/oprofile/op_model_ppro.c @@ -239,11 +239,11 @@ static void arch_perfmon_setup_counters(void) if (eax.split.version_id == 0 && current_cpu_data.x86 == 6 && current_cpu_data.x86_model == 15) { eax.split.version_id = 2; - eax.split.num_events = 2; + eax.split.num_counters = 2; eax.split.bit_width = 40; } - num_counters = eax.split.num_events; + num_counters = eax.split.num_counters; op_arch_perfmon_spec.num_counters = num_counters; op_arch_perfmon_spec.num_controls = num_counters; -- cgit v0.10.2 From a098f4484bc7dae23f5b62360954007b99b64600 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Tue, 30 Mar 2010 11:28:21 +0200 Subject: perf, x86: implement ARCH_PERFMON_EVENTSEL bit masks ARCH_PERFMON_EVENTSEL bit masks are often used in the kernel. This patch adds macros for the bit masks and removes local defines. The function intel_pmu_raw_event() becomes x86_pmu_raw_event() which is generic for x86 models and same also for p6. Duplicate code is removed. Signed-off-by: Robert Richter Signed-off-by: Peter Zijlstra LKML-Reference: <20100330092821.GH11907@erda.amd.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/perf_event.h b/arch/x86/include/asm/perf_event.h index 987bf67..f6d43db 100644 --- a/arch/x86/include/asm/perf_event.h +++ b/arch/x86/include/asm/perf_event.h @@ -18,39 +18,31 @@ #define MSR_ARCH_PERFMON_EVENTSEL0 0x186 #define MSR_ARCH_PERFMON_EVENTSEL1 0x187 -#define ARCH_PERFMON_EVENTSEL_ENABLE (1 << 22) -#define ARCH_PERFMON_EVENTSEL_ANY (1 << 21) -#define ARCH_PERFMON_EVENTSEL_INT (1 << 20) -#define ARCH_PERFMON_EVENTSEL_OS (1 << 17) -#define ARCH_PERFMON_EVENTSEL_USR (1 << 16) - -/* - * Includes eventsel and unit mask as well: - */ - - -#define INTEL_ARCH_EVTSEL_MASK 0x000000FFULL -#define INTEL_ARCH_UNIT_MASK 0x0000FF00ULL -#define INTEL_ARCH_EDGE_MASK 0x00040000ULL -#define INTEL_ARCH_INV_MASK 0x00800000ULL -#define INTEL_ARCH_CNT_MASK 0xFF000000ULL -#define INTEL_ARCH_EVENT_MASK (INTEL_ARCH_UNIT_MASK|INTEL_ARCH_EVTSEL_MASK) - -/* - * filter mask to validate fixed counter events. - * the following filters disqualify for fixed counters: - * - inv - * - edge - * - cnt-mask - * The other filters are supported by fixed counters. - * The any-thread option is supported starting with v3. - */ -#define INTEL_ARCH_FIXED_MASK \ - (INTEL_ARCH_CNT_MASK| \ - INTEL_ARCH_INV_MASK| \ - INTEL_ARCH_EDGE_MASK|\ - INTEL_ARCH_UNIT_MASK|\ - INTEL_ARCH_EVENT_MASK) +#define ARCH_PERFMON_EVENTSEL_EVENT 0x000000FFULL +#define ARCH_PERFMON_EVENTSEL_UMASK 0x0000FF00ULL +#define ARCH_PERFMON_EVENTSEL_USR (1ULL << 16) +#define ARCH_PERFMON_EVENTSEL_OS (1ULL << 17) +#define ARCH_PERFMON_EVENTSEL_EDGE (1ULL << 18) +#define ARCH_PERFMON_EVENTSEL_INT (1ULL << 20) +#define ARCH_PERFMON_EVENTSEL_ANY (1ULL << 21) +#define ARCH_PERFMON_EVENTSEL_ENABLE (1ULL << 22) +#define ARCH_PERFMON_EVENTSEL_INV (1ULL << 23) +#define ARCH_PERFMON_EVENTSEL_CMASK 0xFF000000ULL + +#define AMD64_EVENTSEL_EVENT \ + (ARCH_PERFMON_EVENTSEL_EVENT | (0x0FULL << 32)) +#define INTEL_ARCH_EVENT_MASK \ + (ARCH_PERFMON_EVENTSEL_UMASK | ARCH_PERFMON_EVENTSEL_EVENT) + +#define X86_RAW_EVENT_MASK \ + (ARCH_PERFMON_EVENTSEL_EVENT | \ + ARCH_PERFMON_EVENTSEL_UMASK | \ + ARCH_PERFMON_EVENTSEL_EDGE | \ + ARCH_PERFMON_EVENTSEL_INV | \ + ARCH_PERFMON_EVENTSEL_CMASK) +#define AMD64_RAW_EVENT_MASK \ + (X86_RAW_EVENT_MASK | \ + AMD64_EVENTSEL_EVENT) #define ARCH_PERFMON_UNHALTED_CORE_CYCLES_SEL 0x3c #define ARCH_PERFMON_UNHALTED_CORE_CYCLES_UMASK (0x00 << 8) diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 9daaa1e..1dd42c1 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -143,13 +143,21 @@ struct cpu_hw_events { * Constraint on the Event code. */ #define INTEL_EVENT_CONSTRAINT(c, n) \ - EVENT_CONSTRAINT(c, n, INTEL_ARCH_EVTSEL_MASK) + EVENT_CONSTRAINT(c, n, ARCH_PERFMON_EVENTSEL_EVENT) /* * Constraint on the Event code + UMask + fixed-mask + * + * filter mask to validate fixed counter events. + * the following filters disqualify for fixed counters: + * - inv + * - edge + * - cnt-mask + * The other filters are supported by fixed counters. + * The any-thread option is supported starting with v3. */ #define FIXED_EVENT_CONSTRAINT(c, n) \ - EVENT_CONSTRAINT(c, (1ULL << (32+n)), INTEL_ARCH_FIXED_MASK) + EVENT_CONSTRAINT(c, (1ULL << (32+n)), X86_RAW_EVENT_MASK) /* * Constraint on the Event code + UMask @@ -437,6 +445,11 @@ static int x86_hw_config(struct perf_event_attr *attr, struct hw_perf_event *hwc return 0; } +static u64 x86_pmu_raw_event(u64 hw_event) +{ + return hw_event & X86_RAW_EVENT_MASK; +} + /* * Setup the hardware configuration for a given attr_type */ @@ -1427,7 +1440,7 @@ void __init init_hw_perf_events(void) if (x86_pmu.event_constraints) { for_each_event_constraint(c, x86_pmu.event_constraints) { - if (c->cmask != INTEL_ARCH_FIXED_MASK) + if (c->cmask != X86_RAW_EVENT_MASK) continue; c->idxmsk64 |= (1ULL << x86_pmu.num_counters) - 1; diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c index 7753a5c..37e9517 100644 --- a/arch/x86/kernel/cpu/perf_event_amd.c +++ b/arch/x86/kernel/cpu/perf_event_amd.c @@ -113,20 +113,7 @@ static u64 amd_pmu_event_map(int hw_event) static u64 amd_pmu_raw_event(u64 hw_event) { -#define K7_EVNTSEL_EVENT_MASK 0xF000000FFULL -#define K7_EVNTSEL_UNIT_MASK 0x00000FF00ULL -#define K7_EVNTSEL_EDGE_MASK 0x000040000ULL -#define K7_EVNTSEL_INV_MASK 0x000800000ULL -#define K7_EVNTSEL_REG_MASK 0x0FF000000ULL - -#define K7_EVNTSEL_MASK \ - (K7_EVNTSEL_EVENT_MASK | \ - K7_EVNTSEL_UNIT_MASK | \ - K7_EVNTSEL_EDGE_MASK | \ - K7_EVNTSEL_INV_MASK | \ - K7_EVNTSEL_REG_MASK) - - return hw_event & K7_EVNTSEL_MASK; + return hw_event & AMD64_RAW_EVENT_MASK; } /* diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index cc4d90a..dfdd6f9 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -452,24 +452,6 @@ static __initconst u64 atom_hw_cache_event_ids }, }; -static u64 intel_pmu_raw_event(u64 hw_event) -{ -#define CORE_EVNTSEL_EVENT_MASK 0x000000FFULL -#define CORE_EVNTSEL_UNIT_MASK 0x0000FF00ULL -#define CORE_EVNTSEL_EDGE_MASK 0x00040000ULL -#define CORE_EVNTSEL_INV_MASK 0x00800000ULL -#define CORE_EVNTSEL_REG_MASK 0xFF000000ULL - -#define CORE_EVNTSEL_MASK \ - (INTEL_ARCH_EVTSEL_MASK | \ - INTEL_ARCH_UNIT_MASK | \ - INTEL_ARCH_EDGE_MASK | \ - INTEL_ARCH_INV_MASK | \ - INTEL_ARCH_CNT_MASK) - - return hw_event & CORE_EVNTSEL_MASK; -} - static void intel_pmu_disable_all(void) { struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events); @@ -788,7 +770,7 @@ static __initconst struct x86_pmu core_pmu = { .eventsel = MSR_ARCH_PERFMON_EVENTSEL0, .perfctr = MSR_ARCH_PERFMON_PERFCTR0, .event_map = intel_pmu_event_map, - .raw_event = intel_pmu_raw_event, + .raw_event = x86_pmu_raw_event, .max_events = ARRAY_SIZE(intel_perfmon_event_map), .apic = 1, /* @@ -827,7 +809,7 @@ static __initconst struct x86_pmu intel_pmu = { .eventsel = MSR_ARCH_PERFMON_EVENTSEL0, .perfctr = MSR_ARCH_PERFMON_PERFCTR0, .event_map = intel_pmu_event_map, - .raw_event = intel_pmu_raw_event, + .raw_event = x86_pmu_raw_event, .max_events = ARRAY_SIZE(intel_perfmon_event_map), .apic = 1, /* diff --git a/arch/x86/kernel/cpu/perf_event_p6.c b/arch/x86/kernel/cpu/perf_event_p6.c index b26fbc7e..03c139a 100644 --- a/arch/x86/kernel/cpu/perf_event_p6.c +++ b/arch/x86/kernel/cpu/perf_event_p6.c @@ -27,24 +27,6 @@ static u64 p6_pmu_event_map(int hw_event) */ #define P6_NOP_EVENT 0x0000002EULL -static u64 p6_pmu_raw_event(u64 hw_event) -{ -#define P6_EVNTSEL_EVENT_MASK 0x000000FFULL -#define P6_EVNTSEL_UNIT_MASK 0x0000FF00ULL -#define P6_EVNTSEL_EDGE_MASK 0x00040000ULL -#define P6_EVNTSEL_INV_MASK 0x00800000ULL -#define P6_EVNTSEL_REG_MASK 0xFF000000ULL - -#define P6_EVNTSEL_MASK \ - (P6_EVNTSEL_EVENT_MASK | \ - P6_EVNTSEL_UNIT_MASK | \ - P6_EVNTSEL_EDGE_MASK | \ - P6_EVNTSEL_INV_MASK | \ - P6_EVNTSEL_REG_MASK) - - return hw_event & P6_EVNTSEL_MASK; -} - static struct event_constraint p6_event_constraints[] = { INTEL_EVENT_CONSTRAINT(0xc1, 0x1), /* FLOPS */ @@ -114,7 +96,7 @@ static __initconst struct x86_pmu p6_pmu = { .eventsel = MSR_P6_EVNTSEL0, .perfctr = MSR_P6_PERFCTR0, .event_map = p6_pmu_event_map, - .raw_event = p6_pmu_raw_event, + .raw_event = x86_pmu_raw_event, .max_events = ARRAY_SIZE(p6_perfmon_event_map), .apic = 1, .max_period = (1ULL << 31) - 1, -- cgit v0.10.2 From b4cdc5c264b35c67007800dec3928e9547a9d70b Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 30 Mar 2010 17:00:06 +0200 Subject: perf, x86: Fix up the ANY flag stuff Stephane noticed that the ANY flag was in generic arch code, and Cyrill reported that it broke the P4 code. Solve this by merging x86_pmu::raw_event into x86_pmu::hw_config and provide intel_pmu and amd_pmu specific versions of this callback. The intel_pmu one deals with the ANY flag, the amd_pmu adds the few extra event bits AMD64 has. Reported-by: Stephane Eranian Reported-by: Cyrill Gorcunov Acked-by: Robert Richter Acked-by: Cyrill Gorcunov Acked-by: Stephane Eranian Signed-off-by: Peter Zijlstra LKML-Reference: <1269968113.5258.442.camel@laptop> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 1dd42c1..65e9c5e 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -196,12 +196,11 @@ struct x86_pmu { void (*enable_all)(int added); void (*enable)(struct perf_event *); void (*disable)(struct perf_event *); - int (*hw_config)(struct perf_event_attr *attr, struct hw_perf_event *hwc); + int (*hw_config)(struct perf_event *event); int (*schedule_events)(struct cpu_hw_events *cpuc, int n, int *assign); unsigned eventsel; unsigned perfctr; u64 (*event_map)(int); - u64 (*raw_event)(u64); int max_events; int num_counters; int num_counters_fixed; @@ -426,28 +425,26 @@ set_ext_hw_attr(struct hw_perf_event *hwc, struct perf_event_attr *attr) return 0; } -static int x86_hw_config(struct perf_event_attr *attr, struct hw_perf_event *hwc) +static int x86_pmu_hw_config(struct perf_event *event) { /* * Generate PMC IRQs: * (keep 'enabled' bit clear for now) */ - hwc->config = ARCH_PERFMON_EVENTSEL_INT; + event->hw.config = ARCH_PERFMON_EVENTSEL_INT; /* * Count user and OS events unless requested not to */ - if (!attr->exclude_user) - hwc->config |= ARCH_PERFMON_EVENTSEL_USR; - if (!attr->exclude_kernel) - hwc->config |= ARCH_PERFMON_EVENTSEL_OS; + if (!event->attr.exclude_user) + event->hw.config |= ARCH_PERFMON_EVENTSEL_USR; + if (!event->attr.exclude_kernel) + event->hw.config |= ARCH_PERFMON_EVENTSEL_OS; - return 0; -} + if (event->attr.type == PERF_TYPE_RAW) + event->hw.config |= event->attr.config & X86_RAW_EVENT_MASK; -static u64 x86_pmu_raw_event(u64 hw_event) -{ - return hw_event & X86_RAW_EVENT_MASK; + return 0; } /* @@ -489,7 +486,7 @@ static int __hw_perf_event_init(struct perf_event *event) hwc->last_tag = ~0ULL; /* Processor specifics */ - err = x86_pmu.hw_config(attr, hwc); + err = x86_pmu.hw_config(event); if (err) return err; @@ -508,16 +505,8 @@ static int __hw_perf_event_init(struct perf_event *event) return -EOPNOTSUPP; } - /* - * Raw hw_event type provide the config in the hw_event structure - */ - if (attr->type == PERF_TYPE_RAW) { - hwc->config |= x86_pmu.raw_event(attr->config); - if ((hwc->config & ARCH_PERFMON_EVENTSEL_ANY) && - perf_paranoid_cpu() && !capable(CAP_SYS_ADMIN)) - return -EACCES; + if (attr->type == PERF_TYPE_RAW) return 0; - } if (attr->type == PERF_TYPE_HW_CACHE) return set_ext_hw_attr(hwc, attr); diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c index 37e9517..bbd7339 100644 --- a/arch/x86/kernel/cpu/perf_event_amd.c +++ b/arch/x86/kernel/cpu/perf_event_amd.c @@ -111,9 +111,19 @@ static u64 amd_pmu_event_map(int hw_event) return amd_perfmon_event_map[hw_event]; } -static u64 amd_pmu_raw_event(u64 hw_event) +static int amd_pmu_hw_config(struct perf_event *event) { - return hw_event & AMD64_RAW_EVENT_MASK; + int ret = x86_pmu_hw_config(event); + + if (ret) + return ret; + + if (event->attr.type != PERF_TYPE_RAW) + return 0; + + event->hw.config |= event->attr.config & AMD64_RAW_EVENT_MASK; + + return 0; } /* @@ -365,12 +375,11 @@ static __initconst struct x86_pmu amd_pmu = { .enable_all = x86_pmu_enable_all, .enable = x86_pmu_enable_event, .disable = x86_pmu_disable_event, - .hw_config = x86_hw_config, + .hw_config = amd_pmu_hw_config, .schedule_events = x86_schedule_events, .eventsel = MSR_K7_EVNTSEL0, .perfctr = MSR_K7_PERFCTR0, .event_map = amd_pmu_event_map, - .raw_event = amd_pmu_raw_event, .max_events = ARRAY_SIZE(amd_perfmon_event_map), .num_counters = 4, .cntval_bits = 48, diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index dfdd6f9..30bf10c 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -758,6 +758,30 @@ intel_get_event_constraints(struct cpu_hw_events *cpuc, struct perf_event *event return x86_get_event_constraints(cpuc, event); } +static int intel_pmu_hw_config(struct perf_event *event) +{ + int ret = x86_pmu_hw_config(event); + + if (ret) + return ret; + + if (event->attr.type != PERF_TYPE_RAW) + return 0; + + if (!(event->attr.config & ARCH_PERFMON_EVENTSEL_ANY)) + return 0; + + if (x86_pmu.version < 3) + return -EINVAL; + + if (perf_paranoid_cpu() && !capable(CAP_SYS_ADMIN)) + return -EACCES; + + event->hw.config |= ARCH_PERFMON_EVENTSEL_ANY; + + return 0; +} + static __initconst struct x86_pmu core_pmu = { .name = "core", .handle_irq = x86_pmu_handle_irq, @@ -765,12 +789,11 @@ static __initconst struct x86_pmu core_pmu = { .enable_all = x86_pmu_enable_all, .enable = x86_pmu_enable_event, .disable = x86_pmu_disable_event, - .hw_config = x86_hw_config, + .hw_config = x86_pmu_hw_config, .schedule_events = x86_schedule_events, .eventsel = MSR_ARCH_PERFMON_EVENTSEL0, .perfctr = MSR_ARCH_PERFMON_PERFCTR0, .event_map = intel_pmu_event_map, - .raw_event = x86_pmu_raw_event, .max_events = ARRAY_SIZE(intel_perfmon_event_map), .apic = 1, /* @@ -804,12 +827,11 @@ static __initconst struct x86_pmu intel_pmu = { .enable_all = intel_pmu_enable_all, .enable = intel_pmu_enable_event, .disable = intel_pmu_disable_event, - .hw_config = x86_hw_config, + .hw_config = intel_pmu_hw_config, .schedule_events = x86_schedule_events, .eventsel = MSR_ARCH_PERFMON_EVENTSEL0, .perfctr = MSR_ARCH_PERFMON_PERFCTR0, .event_map = intel_pmu_event_map, - .raw_event = x86_pmu_raw_event, .max_events = ARRAY_SIZE(intel_perfmon_event_map), .apic = 1, /* diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index 4139100..acd237d 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -419,20 +419,7 @@ static u64 p4_pmu_event_map(int hw_event) return config; } -/* - * We don't control raw events so it's up to the caller - * to pass sane values (and we don't count the thread number - * on HT machine but allow HT-compatible specifics to be - * passed on) - */ -static u64 p4_pmu_raw_event(u64 hw_event) -{ - return hw_event & - (p4_config_pack_escr(P4_ESCR_MASK_HT) | - p4_config_pack_cccr(P4_CCCR_MASK_HT)); -} - -static int p4_hw_config(struct perf_event_attr *attr, struct hw_perf_event *hwc) +static int p4_hw_config(struct perf_event *event) { int cpu = raw_smp_processor_id(); u32 escr, cccr; @@ -444,11 +431,29 @@ static int p4_hw_config(struct perf_event_attr *attr, struct hw_perf_event *hwc) */ cccr = p4_default_cccr_conf(cpu); - escr = p4_default_escr_conf(cpu, attr->exclude_kernel, attr->exclude_user); - hwc->config = p4_config_pack_escr(escr) | p4_config_pack_cccr(cccr); + escr = p4_default_escr_conf(cpu, event->attr.exclude_kernel, + event->attr.exclude_user); + event->hw.config = p4_config_pack_escr(escr) | + p4_config_pack_cccr(cccr); if (p4_ht_active() && p4_ht_thread(cpu)) - hwc->config = p4_set_ht_bit(hwc->config); + event->hw.config = p4_set_ht_bit(event->hw.config); + + if (event->attr.type != PERF_TYPE_RAW) + return 0; + + /* + * We don't control raw events so it's up to the caller + * to pass sane values (and we don't count the thread number + * on HT machine but allow HT-compatible specifics to be + * passed on) + * + * XXX: HT wide things should check perf_paranoid_cpu() && + * CAP_SYS_ADMIN + */ + event->hw.config |= event->attr.config & + (p4_config_pack_escr(P4_ESCR_MASK_HT) | + p4_config_pack_cccr(P4_CCCR_MASK_HT)); return 0; } @@ -785,7 +790,6 @@ static __initconst struct x86_pmu p4_pmu = { .eventsel = MSR_P4_BPU_CCCR0, .perfctr = MSR_P4_BPU_PERFCTR0, .event_map = p4_pmu_event_map, - .raw_event = p4_pmu_raw_event, .max_events = ARRAY_SIZE(p4_general_events), .get_event_constraints = x86_get_event_constraints, /* diff --git a/arch/x86/kernel/cpu/perf_event_p6.c b/arch/x86/kernel/cpu/perf_event_p6.c index 03c139a..9123e8e 100644 --- a/arch/x86/kernel/cpu/perf_event_p6.c +++ b/arch/x86/kernel/cpu/perf_event_p6.c @@ -91,12 +91,11 @@ static __initconst struct x86_pmu p6_pmu = { .enable_all = p6_pmu_enable_all, .enable = p6_pmu_enable_event, .disable = p6_pmu_disable_event, - .hw_config = x86_hw_config, + .hw_config = x86_pmu_hw_config, .schedule_events = x86_schedule_events, .eventsel = MSR_P6_EVNTSEL0, .perfctr = MSR_P6_PERFCTR0, .event_map = p6_pmu_event_map, - .raw_event = x86_pmu_raw_event, .max_events = ARRAY_SIZE(p6_perfmon_event_map), .apic = 1, .max_period = (1ULL << 31) - 1, -- cgit v0.10.2 From caaa8be3b6707cb9664e573a28b00f845ce9f32e Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 29 Mar 2010 13:09:53 +0200 Subject: perf, x86: Fix __initconst vs const All variables that have __initconst should also be const. Suggested-by: Stephen Rothwell Signed-off-by: Peter Zijlstra LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c index bbd7339..611df11 100644 --- a/arch/x86/kernel/cpu/perf_event_amd.c +++ b/arch/x86/kernel/cpu/perf_event_amd.c @@ -2,7 +2,7 @@ static DEFINE_RAW_SPINLOCK(amd_nb_lock); -static __initconst u64 amd_hw_cache_event_ids +static __initconst const u64 amd_hw_cache_event_ids [PERF_COUNT_HW_CACHE_MAX] [PERF_COUNT_HW_CACHE_OP_MAX] [PERF_COUNT_HW_CACHE_RESULT_MAX] = @@ -368,7 +368,7 @@ static void amd_pmu_cpu_dead(int cpu) raw_spin_unlock(&amd_nb_lock); } -static __initconst struct x86_pmu amd_pmu = { +static __initconst const struct x86_pmu amd_pmu = { .name = "AMD", .handle_irq = x86_pmu_handle_irq, .disable_all = x86_pmu_disable_all, diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 30bf10c..1957e3f 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -88,7 +88,7 @@ static u64 intel_pmu_event_map(int hw_event) return intel_perfmon_event_map[hw_event]; } -static __initconst u64 westmere_hw_cache_event_ids +static __initconst const u64 westmere_hw_cache_event_ids [PERF_COUNT_HW_CACHE_MAX] [PERF_COUNT_HW_CACHE_OP_MAX] [PERF_COUNT_HW_CACHE_RESULT_MAX] = @@ -179,7 +179,7 @@ static __initconst u64 westmere_hw_cache_event_ids }, }; -static __initconst u64 nehalem_hw_cache_event_ids +static __initconst const u64 nehalem_hw_cache_event_ids [PERF_COUNT_HW_CACHE_MAX] [PERF_COUNT_HW_CACHE_OP_MAX] [PERF_COUNT_HW_CACHE_RESULT_MAX] = @@ -270,7 +270,7 @@ static __initconst u64 nehalem_hw_cache_event_ids }, }; -static __initconst u64 core2_hw_cache_event_ids +static __initconst const u64 core2_hw_cache_event_ids [PERF_COUNT_HW_CACHE_MAX] [PERF_COUNT_HW_CACHE_OP_MAX] [PERF_COUNT_HW_CACHE_RESULT_MAX] = @@ -361,7 +361,7 @@ static __initconst u64 core2_hw_cache_event_ids }, }; -static __initconst u64 atom_hw_cache_event_ids +static __initconst const u64 atom_hw_cache_event_ids [PERF_COUNT_HW_CACHE_MAX] [PERF_COUNT_HW_CACHE_OP_MAX] [PERF_COUNT_HW_CACHE_RESULT_MAX] = @@ -782,7 +782,7 @@ static int intel_pmu_hw_config(struct perf_event *event) return 0; } -static __initconst struct x86_pmu core_pmu = { +static __initconst const struct x86_pmu core_pmu = { .name = "core", .handle_irq = x86_pmu_handle_irq, .disable_all = x86_pmu_disable_all, @@ -820,7 +820,7 @@ static void intel_pmu_cpu_dying(int cpu) fini_debug_store_on_cpu(cpu); } -static __initconst struct x86_pmu intel_pmu = { +static __initconst const struct x86_pmu intel_pmu = { .name = "Intel", .handle_irq = intel_pmu_handle_irq, .disable_all = intel_pmu_disable_all, diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index acd237d..15367cc 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ b/arch/x86/kernel/cpu/perf_event_p4.c @@ -287,7 +287,7 @@ static struct p4_event_bind p4_event_bind_map[] = { p4_config_pack_cccr(cache_event | \ P4_CCCR_ESEL(P4_OPCODE_ESEL(P4_OPCODE(event)))) -static __initconst u64 p4_hw_cache_event_ids +static __initconst const u64 p4_hw_cache_event_ids [PERF_COUNT_HW_CACHE_MAX] [PERF_COUNT_HW_CACHE_OP_MAX] [PERF_COUNT_HW_CACHE_RESULT_MAX] = @@ -780,7 +780,7 @@ done: return num ? -ENOSPC : 0; } -static __initconst struct x86_pmu p4_pmu = { +static __initconst const struct x86_pmu p4_pmu = { .name = "Netburst P4/Xeon", .handle_irq = p4_pmu_handle_irq, .disable_all = p4_pmu_disable_all, diff --git a/arch/x86/kernel/cpu/perf_event_p6.c b/arch/x86/kernel/cpu/perf_event_p6.c index 9123e8e..34ba07b 100644 --- a/arch/x86/kernel/cpu/perf_event_p6.c +++ b/arch/x86/kernel/cpu/perf_event_p6.c @@ -84,7 +84,7 @@ static void p6_pmu_enable_event(struct perf_event *event) (void)checking_wrmsrl(hwc->config_base + hwc->idx, val); } -static __initconst struct x86_pmu p6_pmu = { +static __initconst const struct x86_pmu p6_pmu = { .name = "p6", .handle_irq = x86_pmu_handle_irq, .disable_all = p6_pmu_disable_all, -- cgit v0.10.2 From 40b91cd10f000b4c4934e48e2e5c0bec66def144 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 29 Mar 2010 16:37:17 +0200 Subject: perf, x86: Add Nehalem programming quirk to Westmere According to the Xeon-5600 errata the Westmere suffers the same PMU programming bug as the original Nehalem did. Signed-off-by: Peter Zijlstra LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 1957e3f..f168b40 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -488,6 +488,7 @@ static void intel_pmu_enable_all(int added) * Workaround for: * Intel Errata AAK100 (model 26) * Intel Errata AAP53 (model 30) + * Intel Errata BD53 (model 44) * * These chips need to be 'reset' when adding counters by programming * the magic three (non counting) events 0x4300D2, 0x4300B1 and 0x4300B5 @@ -980,6 +981,7 @@ static __init int intel_pmu_init(void) intel_pmu_lbr_init_nhm(); x86_pmu.event_constraints = intel_westmere_event_constraints; + x86_pmu.enable_all = intel_pmu_nhm_enable_all; pr_cont("Westmere events, "); break; -- cgit v0.10.2 From 32bd7eb5a7f4596c8440dd9440322fe9e686634d Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 24 Mar 2010 13:17:19 +0800 Subject: sched: Remove remaining USER_SCHED code This is left over from commit 7c9414385e ("sched: Remove USER_SCHED"") Signed-off-by: Li Zefan Acked-by: Dhaval Giani Signed-off-by: Peter Zijlstra Cc: David Howells LKML-Reference: <4BA9A05F.7010407@cn.fujitsu.com> Signed-off-by: Ingo Molnar diff --git a/init/Kconfig b/init/Kconfig index eb77e8c..5fe94b8 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -604,8 +604,7 @@ config RT_GROUP_SCHED default n help This feature lets you explicitly allocate real CPU bandwidth - to users or control groups (depending on the "Basis for grouping tasks" - setting below. If enabled, it will also make it impossible to + to task groups. If enabled, it will also make it impossible to schedule realtime tasks for non-root users until you allocate realtime bandwidth for them. See Documentation/scheduler/sched-rt-group.txt for more information. diff --git a/kernel/capability.c b/kernel/capability.c index 9e4697e..2f05303 100644 --- a/kernel/capability.c +++ b/kernel/capability.c @@ -15,7 +15,6 @@ #include #include #include -#include "cred-internals.h" /* * Leveraged for setting/resetting capabilities diff --git a/kernel/cred-internals.h b/kernel/cred-internals.h deleted file mode 100644 index 2dc4fc2..0000000 --- a/kernel/cred-internals.h +++ /dev/null @@ -1,21 +0,0 @@ -/* Internal credentials stuff - * - * Copyright (C) 2008 Red Hat, Inc. All Rights Reserved. - * Written by David Howells (dhowells@redhat.com) - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public Licence - * as published by the Free Software Foundation; either version - * 2 of the Licence, or (at your option) any later version. - */ - -/* - * user.c - */ -static inline void sched_switch_user(struct task_struct *p) -{ -#ifdef CONFIG_USER_SCHED - sched_move_task(p); -#endif /* CONFIG_USER_SCHED */ -} - diff --git a/kernel/cred.c b/kernel/cred.c index 1b1129d..19d3ccc 100644 --- a/kernel/cred.c +++ b/kernel/cred.c @@ -16,7 +16,6 @@ #include #include #include -#include "cred-internals.h" #if 0 #define kdebug(FMT, ...) \ @@ -557,8 +556,6 @@ int commit_creds(struct cred *new) atomic_dec(&old->user->processes); alter_cred_subscribers(old, -2); - sched_switch_user(task); - /* send notifications */ if (new->uid != old->uid || new->euid != old->euid || diff --git a/kernel/exit.c b/kernel/exit.c index cce59cb..84dc4b2 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -55,7 +55,6 @@ #include #include #include -#include "cred-internals.h" static void exit_mm(struct task_struct * tsk); diff --git a/kernel/sched_debug.c b/kernel/sched_debug.c index 8a46a71..0932c5c 100644 --- a/kernel/sched_debug.c +++ b/kernel/sched_debug.c @@ -173,11 +173,6 @@ void print_cfs_rq(struct seq_file *m, int cpu, struct cfs_rq *cfs_rq) task_group_path(tg, path, sizeof(path)); SEQ_printf(m, "\ncfs_rq[%d]:%s\n", cpu, path); -#elif defined(CONFIG_USER_SCHED) && defined(CONFIG_FAIR_GROUP_SCHED) - { - uid_t uid = cfs_rq->tg->uid; - SEQ_printf(m, "\ncfs_rq[%d] for UID: %u\n", cpu, uid); - } #else SEQ_printf(m, "\ncfs_rq[%d]:\n", cpu); #endif diff --git a/kernel/user.c b/kernel/user.c index ec3b222..8e1c8c0 100644 --- a/kernel/user.c +++ b/kernel/user.c @@ -16,7 +16,6 @@ #include #include #include -#include "cred-internals.h" struct user_namespace init_user_ns = { .kref = { @@ -137,9 +136,7 @@ struct user_struct *alloc_uid(struct user_namespace *ns, uid_t uid) struct hlist_head *hashent = uidhashentry(ns, uid); struct user_struct *up, *new; - /* Make uid_hash_find() + uids_user_create() + uid_hash_insert() - * atomic. - */ + /* Make uid_hash_find() + uid_hash_insert() atomic. */ spin_lock_irq(&uidhash_lock); up = uid_hash_find(uid, hashent); spin_unlock_irq(&uidhash_lock); @@ -161,11 +158,6 @@ struct user_struct *alloc_uid(struct user_namespace *ns, uid_t uid) spin_lock_irq(&uidhash_lock); up = uid_hash_find(uid, hashent); if (up) { - /* This case is not possible when CONFIG_USER_SCHED - * is defined, since we serialize alloc_uid() using - * uids_mutex. Hence no need to call - * sched_destroy_user() or remove_user_sysfs_dir(). - */ key_put(new->uid_keyring); key_put(new->session_keyring); kmem_cache_free(uid_cachep, new); -- cgit v0.10.2 From 25c2d55c00c6097e6792ebf21e31342f23b9b768 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 24 Mar 2010 13:17:50 +0800 Subject: sched: Remove USER_SCHED from documentation USER_SCHED has been removed, so update the documentation accordingly. Signed-off-by: Li Zefan Signed-off-by: Peter Zijlstra Acked-by: Serge E. Hallyn LKML-Reference: <4BA9A07E.8070508@cn.fujitsu.com> Signed-off-by: Ingo Molnar diff --git a/Documentation/scheduler/sched-design-CFS.txt b/Documentation/scheduler/sched-design-CFS.txt index 6f33593..8239ebb 100644 --- a/Documentation/scheduler/sched-design-CFS.txt +++ b/Documentation/scheduler/sched-design-CFS.txt @@ -211,7 +211,7 @@ provide fair CPU time to each such task group. For example, it may be desirable to first provide fair CPU time to each user on the system and then to each task belonging to a user. -CONFIG_GROUP_SCHED strives to achieve exactly that. It lets tasks to be +CONFIG_CGROUP_SCHED strives to achieve exactly that. It lets tasks to be grouped and divides CPU time fairly among such groups. CONFIG_RT_GROUP_SCHED permits to group real-time (i.e., SCHED_FIFO and @@ -220,38 +220,11 @@ SCHED_RR) tasks. CONFIG_FAIR_GROUP_SCHED permits to group CFS (i.e., SCHED_NORMAL and SCHED_BATCH) tasks. -At present, there are two (mutually exclusive) mechanisms to group tasks for -CPU bandwidth control purposes: - - - Based on user id (CONFIG_USER_SCHED) - - With this option, tasks are grouped according to their user id. - - - Based on "cgroup" pseudo filesystem (CONFIG_CGROUP_SCHED) - - This options needs CONFIG_CGROUPS to be defined, and lets the administrator + These options need CONFIG_CGROUPS to be defined, and let the administrator create arbitrary groups of tasks, using the "cgroup" pseudo filesystem. See Documentation/cgroups/cgroups.txt for more information about this filesystem. -Only one of these options to group tasks can be chosen and not both. - -When CONFIG_USER_SCHED is defined, a directory is created in sysfs for each new -user and a "cpu_share" file is added in that directory. - - # cd /sys/kernel/uids - # cat 512/cpu_share # Display user 512's CPU share - 1024 - # echo 2048 > 512/cpu_share # Modify user 512's CPU share - # cat 512/cpu_share # Display user 512's CPU share - 2048 - # - -CPU bandwidth between two users is divided in the ratio of their CPU shares. -For example: if you would like user "root" to get twice the bandwidth of user -"guest," then set the cpu_share for both the users such that "root"'s cpu_share -is twice "guest"'s cpu_share. - -When CONFIG_CGROUP_SCHED is defined, a "cpu.shares" file is created for each +When CONFIG_FAIR_GROUP_SCHED is defined, a "cpu.shares" file is created for each group created using the pseudo filesystem. See example steps below to create task groups and modify their CPU share using the "cgroups" pseudo filesystem. @@ -273,24 +246,3 @@ task groups and modify their CPU share using the "cgroups" pseudo filesystem. # #Launch gmplayer (or your favourite movie player) # echo > multimedia/tasks - -8. Implementation note: user namespaces - -User namespaces are intended to be hierarchical. But they are currently -only partially implemented. Each of those has ramifications for CFS. - -First, since user namespaces are hierarchical, the /sys/kernel/uids -presentation is inadequate. Eventually we will likely want to use sysfs -tagging to provide private views of /sys/kernel/uids within each user -namespace. - -Second, the hierarchical nature is intended to support completely -unprivileged use of user namespaces. So if using user groups, then -we want the users in a user namespace to be children of the user -who created it. - -That is currently unimplemented. So instead, every user in a new -user namespace will receive 1024 shares just like any user in the -initial user namespace. Note that at the moment creation of a new -user namespace requires each of CAP_SYS_ADMIN, CAP_SETUID, and -CAP_SETGID. diff --git a/Documentation/scheduler/sched-rt-group.txt b/Documentation/scheduler/sched-rt-group.txt index 86eabe6..605b0d4 100644 --- a/Documentation/scheduler/sched-rt-group.txt +++ b/Documentation/scheduler/sched-rt-group.txt @@ -126,23 +126,12 @@ priority! 2.3 Basis for grouping tasks ---------------------------- -There are two compile-time settings for allocating CPU bandwidth. These are -configured using the "Basis for grouping tasks" multiple choice menu under -General setup > Group CPU Scheduler: - -a. CONFIG_USER_SCHED (aka "Basis for grouping tasks" = "user id") - -This lets you use the virtual files under -"/sys/kernel/uids//cpu_rt_runtime_us" to control he CPU time reserved for -each user . - -The other option is: - -.o CONFIG_CGROUP_SCHED (aka "Basis for grouping tasks" = "Control groups") +Enabling CONFIG_RT_GROUP_SCHED lets you explicitly allocate real +CPU bandwidth to task groups. This uses the /cgroup virtual file system and "/cgroup//cpu.rt_runtime_us" to control the CPU time reserved for each -control group instead. +control group. For more information on working with control groups, you should read Documentation/cgroups/cgroups.txt as well. @@ -161,8 +150,7 @@ For now, this can be simplified to just the following (but see Future plans): =============== There is work in progress to make the scheduling period for each group -("/sys/kernel/uids//cpu_rt_period_us" or -"/cgroup//cpu.rt_period_us" respectively) configurable as well. +("/cgroup//cpu.rt_period_us") configurable as well. The constraint on the period is that a subgroup must have a smaller or equal period to its parent. But realistically its not very useful _yet_ -- cgit v0.10.2 From 897f0b3c3ff40b443c84e271bef19bd6ae885195 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Mon, 15 Mar 2010 10:10:03 +0100 Subject: sched: Kill the broken and deadlockable cpuset_lock/cpuset_cpus_allowed_locked code This patch just states the fact the cpusets/cpuhotplug interaction is broken and removes the deadlockable code which only pretends to work. - cpuset_lock() doesn't really work. It is needed for cpuset_cpus_allowed_locked() but we can't take this lock in try_to_wake_up()->select_fallback_rq() path. - cpuset_lock() is deadlockable. Suppose that a task T bound to CPU takes callback_mutex. If cpu_down(CPU) happens before T drops callback_mutex stop_machine() preempts T, then migration_call(CPU_DEAD) tries to take cpuset_lock() and hangs forever because CPU is already dead and thus T can't be scheduled. - cpuset_cpus_allowed_locked() is deadlockable too. It takes task_lock() which is not irq-safe, but try_to_wake_up() can be called from irq. Kill them, and change select_fallback_rq() to use cpu_possible_mask, like we currently do without CONFIG_CPUSETS. Also, with or without this patch, with or without CONFIG_CPUSETS, the callers of select_fallback_rq() can race with each other or with set_cpus_allowed() pathes. The subsequent patches try to to fix these problems. Signed-off-by: Oleg Nesterov Signed-off-by: Peter Zijlstra LKML-Reference: <20100315091003.GA9123@redhat.com> Signed-off-by: Ingo Molnar diff --git a/include/linux/cpuset.h b/include/linux/cpuset.h index a5740fc..eeaaee7 100644 --- a/include/linux/cpuset.h +++ b/include/linux/cpuset.h @@ -21,8 +21,6 @@ extern int number_of_cpusets; /* How many cpusets are defined in system? */ extern int cpuset_init(void); extern void cpuset_init_smp(void); extern void cpuset_cpus_allowed(struct task_struct *p, struct cpumask *mask); -extern void cpuset_cpus_allowed_locked(struct task_struct *p, - struct cpumask *mask); extern nodemask_t cpuset_mems_allowed(struct task_struct *p); #define cpuset_current_mems_allowed (current->mems_allowed) void cpuset_init_current_mems_allowed(void); @@ -69,9 +67,6 @@ struct seq_file; extern void cpuset_task_status_allowed(struct seq_file *m, struct task_struct *task); -extern void cpuset_lock(void); -extern void cpuset_unlock(void); - extern int cpuset_mem_spread_node(void); static inline int cpuset_do_page_mem_spread(void) @@ -105,11 +100,6 @@ static inline void cpuset_cpus_allowed(struct task_struct *p, { cpumask_copy(mask, cpu_possible_mask); } -static inline void cpuset_cpus_allowed_locked(struct task_struct *p, - struct cpumask *mask) -{ - cpumask_copy(mask, cpu_possible_mask); -} static inline nodemask_t cpuset_mems_allowed(struct task_struct *p) { @@ -157,9 +147,6 @@ static inline void cpuset_task_status_allowed(struct seq_file *m, { } -static inline void cpuset_lock(void) {} -static inline void cpuset_unlock(void) {} - static inline int cpuset_mem_spread_node(void) { return 0; diff --git a/kernel/cpuset.c b/kernel/cpuset.c index d109467..9a747f5 100644 --- a/kernel/cpuset.c +++ b/kernel/cpuset.c @@ -2182,19 +2182,10 @@ void __init cpuset_init_smp(void) void cpuset_cpus_allowed(struct task_struct *tsk, struct cpumask *pmask) { mutex_lock(&callback_mutex); - cpuset_cpus_allowed_locked(tsk, pmask); - mutex_unlock(&callback_mutex); -} - -/** - * cpuset_cpus_allowed_locked - return cpus_allowed mask from a tasks cpuset. - * Must be called with callback_mutex held. - **/ -void cpuset_cpus_allowed_locked(struct task_struct *tsk, struct cpumask *pmask) -{ task_lock(tsk); guarantee_online_cpus(task_cs(tsk), pmask); task_unlock(tsk); + mutex_unlock(&callback_mutex); } void cpuset_init_current_mems_allowed(void) @@ -2383,22 +2374,6 @@ int __cpuset_node_allowed_hardwall(int node, gfp_t gfp_mask) } /** - * cpuset_lock - lock out any changes to cpuset structures - * - * The out of memory (oom) code needs to mutex_lock cpusets - * from being changed while it scans the tasklist looking for a - * task in an overlapping cpuset. Expose callback_mutex via this - * cpuset_lock() routine, so the oom code can lock it, before - * locking the task list. The tasklist_lock is a spinlock, so - * must be taken inside callback_mutex. - */ - -void cpuset_lock(void) -{ - mutex_lock(&callback_mutex); -} - -/** * cpuset_unlock - release lock on cpuset changes * * Undo the lock taken in a previous cpuset_lock() call. diff --git a/kernel/sched.c b/kernel/sched.c index 52b7efd2..c0b3ebc 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2296,11 +2296,9 @@ static int select_fallback_rq(int cpu, struct task_struct *p) return dest_cpu; /* No more Mr. Nice Guy. */ - if (dest_cpu >= nr_cpu_ids) { - rcu_read_lock(); - cpuset_cpus_allowed_locked(p, &p->cpus_allowed); - rcu_read_unlock(); - dest_cpu = cpumask_any_and(cpu_active_mask, &p->cpus_allowed); + if (unlikely(dest_cpu >= nr_cpu_ids)) { + cpumask_copy(&p->cpus_allowed, cpu_possible_mask); + dest_cpu = cpumask_any(cpu_active_mask); /* * Don't tell them about moving exiting tasks or @@ -5866,7 +5864,6 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) case CPU_DEAD: case CPU_DEAD_FROZEN: - cpuset_lock(); /* around calls to cpuset_cpus_allowed_lock() */ migrate_live_tasks(cpu); rq = cpu_rq(cpu); kthread_stop(rq->migration_thread); @@ -5879,7 +5876,6 @@ migration_call(struct notifier_block *nfb, unsigned long action, void *hcpu) rq->idle->sched_class = &idle_sched_class; migrate_dead_tasks(cpu); raw_spin_unlock_irq(&rq->lock); - cpuset_unlock(); migrate_nr_uninterruptible(rq); BUG_ON(rq->nr_running != 0); calc_global_load_remove(rq); -- cgit v0.10.2 From 1445c08d06c5594895b4fae952ef8a457e89c390 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Mon, 15 Mar 2010 10:10:10 +0100 Subject: sched: move_task_off_dead_cpu(): Take rq->lock around select_fallback_rq() move_task_off_dead_cpu()->select_fallback_rq() reads/updates ->cpus_allowed lockless. We can race with set_cpus_allowed() running in parallel. Change it to take rq->lock around select_fallback_rq(). Note that it is not trivial to move this spin_lock() into select_fallback_rq(), we must recheck the task was not migrated after we take the lock and other callers do not need this lock. To avoid the races with other callers of select_fallback_rq() which rely on TASK_WAKING, we also check p->state != TASK_WAKING and do nothing otherwise. The owner of TASK_WAKING must update ->cpus_allowed and choose the correct CPU anyway, and the subsequent __migrate_task() is just meaningless because p->se.on_rq must be false. Alternatively, we could change select_task_rq() to take rq->lock right after it calls sched_class->select_task_rq(), but this looks a bit ugly. Also, change it to not assume irqs are disabled and absorb __migrate_task_irq(). Signed-off-by: Oleg Nesterov Signed-off-by: Peter Zijlstra LKML-Reference: <20100315091010.GA9131@redhat.com> Signed-off-by: Ingo Molnar diff --git a/kernel/sched.c b/kernel/sched.c index c0b3ebc..27774b5 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -5448,29 +5448,29 @@ static int migration_thread(void *data) } #ifdef CONFIG_HOTPLUG_CPU - -static int __migrate_task_irq(struct task_struct *p, int src_cpu, int dest_cpu) -{ - int ret; - - local_irq_disable(); - ret = __migrate_task(p, src_cpu, dest_cpu); - local_irq_enable(); - return ret; -} - /* * Figure out where task on dead CPU should go, use force if necessary. */ static void move_task_off_dead_cpu(int dead_cpu, struct task_struct *p) { - int dest_cpu; - + struct rq *rq = cpu_rq(dead_cpu); + int needs_cpu, uninitialized_var(dest_cpu); + unsigned long flags; again: - dest_cpu = select_fallback_rq(dead_cpu, p); + local_irq_save(flags); + + raw_spin_lock(&rq->lock); + needs_cpu = (task_cpu(p) == dead_cpu) && (p->state != TASK_WAKING); + if (needs_cpu) + dest_cpu = select_fallback_rq(dead_cpu, p); + raw_spin_unlock(&rq->lock); /* It can have affinity changed while we were choosing. */ - if (unlikely(!__migrate_task_irq(p, dead_cpu, dest_cpu))) + if (needs_cpu) + needs_cpu = !__migrate_task(p, dead_cpu, dest_cpu); + local_irq_restore(flags); + + if (unlikely(needs_cpu)) goto again; } -- cgit v0.10.2 From c1804d547dc098363443667609c272d1e4d15ee8 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Mon, 15 Mar 2010 10:10:14 +0100 Subject: sched: move_task_off_dead_cpu(): Remove retry logic The previous patch preserved the retry logic, but it looks unneeded. __migrate_task() can only fail if we raced with migration after we dropped the lock, but in this case the caller of set_cpus_allowed/etc must initiate migration itself if ->on_rq == T. We already fixed p->cpus_allowed, the changes in active/online masks must be visible to racer, it should migrate the task to online cpu correctly. Signed-off-by: Oleg Nesterov Signed-off-by: Peter Zijlstra LKML-Reference: <20100315091014.GA9138@redhat.com> Signed-off-by: Ingo Molnar diff --git a/kernel/sched.c b/kernel/sched.c index 27774b5..f475c60 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -5456,7 +5456,7 @@ static void move_task_off_dead_cpu(int dead_cpu, struct task_struct *p) struct rq *rq = cpu_rq(dead_cpu); int needs_cpu, uninitialized_var(dest_cpu); unsigned long flags; -again: + local_irq_save(flags); raw_spin_lock(&rq->lock); @@ -5464,14 +5464,13 @@ again: if (needs_cpu) dest_cpu = select_fallback_rq(dead_cpu, p); raw_spin_unlock(&rq->lock); - - /* It can have affinity changed while we were choosing. */ + /* + * It can only fail if we race with set_cpus_allowed(), + * in the racer should migrate the task anyway. + */ if (needs_cpu) - needs_cpu = !__migrate_task(p, dead_cpu, dest_cpu); + __migrate_task(p, dead_cpu, dest_cpu); local_irq_restore(flags); - - if (unlikely(needs_cpu)) - goto again; } /* -- cgit v0.10.2 From 30da688ef6b76e01969b00608202fff1eed2accc Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Mon, 15 Mar 2010 10:10:19 +0100 Subject: sched: sched_exec(): Remove the select_fallback_rq() logic sched_exec()->select_task_rq() reads/updates ->cpus_allowed lockless. This can race with other CPUs updating our ->cpus_allowed, and this looks meaningless to me. The task is current and running, it must have online cpus in ->cpus_allowed, the fallback mode is bogus. And, if ->sched_class returns the "wrong" cpu, this likely means we raced with set_cpus_allowed() which was called for reason, why should sched_exec() retry and call ->select_task_rq() again? Change the code to call sched_class->select_task_rq() directly and do nothing if the returned cpu is wrong after re-checking under rq->lock. From now task_struct->cpus_allowed is always stable under TASK_WAKING, select_fallback_rq() is always called under rq-lock or the caller or the caller owns TASK_WAKING (select_task_rq). Signed-off-by: Oleg Nesterov Signed-off-by: Peter Zijlstra LKML-Reference: <20100315091019.GA9141@redhat.com> Signed-off-by: Ingo Molnar diff --git a/kernel/sched.c b/kernel/sched.c index f475c60..165b532 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2280,6 +2280,9 @@ void task_oncpu_function_call(struct task_struct *p, } #ifdef CONFIG_SMP +/* + * ->cpus_allowed is protected by either TASK_WAKING or rq->lock held. + */ static int select_fallback_rq(int cpu, struct task_struct *p) { int dest_cpu; @@ -2316,12 +2319,7 @@ static int select_fallback_rq(int cpu, struct task_struct *p) } /* - * Gets called from 3 sites (exec, fork, wakeup), since it is called without - * holding rq->lock we need to ensure ->cpus_allowed is stable, this is done - * by: - * - * exec: is unstable, retry loop - * fork & wake-up: serialize ->cpus_allowed against TASK_WAKING + * The caller (fork, wakeup) owns TASK_WAKING, ->cpus_allowed is stable. */ static inline int select_task_rq(struct task_struct *p, int sd_flags, int wake_flags) @@ -3076,9 +3074,8 @@ void sched_exec(void) unsigned long flags; struct rq *rq; -again: this_cpu = get_cpu(); - dest_cpu = select_task_rq(p, SD_BALANCE_EXEC, 0); + dest_cpu = p->sched_class->select_task_rq(p, SD_BALANCE_EXEC, 0); if (dest_cpu == this_cpu) { put_cpu(); return; @@ -3086,18 +3083,12 @@ again: rq = task_rq_lock(p, &flags); put_cpu(); - /* * select_task_rq() can race against ->cpus_allowed */ - if (!cpumask_test_cpu(dest_cpu, &p->cpus_allowed) - || unlikely(!cpu_active(dest_cpu))) { - task_rq_unlock(rq, &flags); - goto again; - } - - /* force the process onto the specified CPU */ - if (migrate_task(p, dest_cpu, &req)) { + if (cpumask_test_cpu(dest_cpu, &p->cpus_allowed) && + likely(cpu_active(dest_cpu)) && + migrate_task(p, dest_cpu, &req)) { /* Need to wait for migration thread (might exit: take ref). */ struct task_struct *mt = rq->migration_thread; -- cgit v0.10.2 From 6a1bdc1b577ebcb65f6603c57f8347309bc4ab13 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Mon, 15 Mar 2010 10:10:23 +0100 Subject: sched: _cpu_down(): Don't play with current->cpus_allowed _cpu_down() changes the current task's affinity and then recovers it at the end. The problems are well known: we can't restore old_allowed if it was bound to the now-dead-cpu, and we can race with the userspace which can change cpu-affinity during unplug. _cpu_down() should not play with current->cpus_allowed at all. Instead, take_cpu_down() can migrate the caller of _cpu_down() after __cpu_disable() removes the dying cpu from cpu_online_mask. Signed-off-by: Oleg Nesterov Acked-by: Rafael J. Wysocki Signed-off-by: Peter Zijlstra LKML-Reference: <20100315091023.GA9148@redhat.com> Signed-off-by: Ingo Molnar diff --git a/include/linux/sched.h b/include/linux/sched.h index 43c9451..8bea407 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1843,6 +1843,7 @@ extern void sched_clock_idle_sleep_event(void); extern void sched_clock_idle_wakeup_event(u64 delta_ns); #ifdef CONFIG_HOTPLUG_CPU +extern void move_task_off_dead_cpu(int dead_cpu, struct task_struct *p); extern void idle_task_exit(void); #else static inline void idle_task_exit(void) {} diff --git a/kernel/cpu.c b/kernel/cpu.c index f8cced2..8d340fa 100644 --- a/kernel/cpu.c +++ b/kernel/cpu.c @@ -163,6 +163,7 @@ static inline void check_for_tasks(int cpu) } struct take_cpu_down_param { + struct task_struct *caller; unsigned long mod; void *hcpu; }; @@ -171,6 +172,7 @@ struct take_cpu_down_param { static int __ref take_cpu_down(void *_param) { struct take_cpu_down_param *param = _param; + unsigned int cpu = (unsigned long)param->hcpu; int err; /* Ensure this CPU doesn't handle any more interrupts. */ @@ -181,6 +183,8 @@ static int __ref take_cpu_down(void *_param) raw_notifier_call_chain(&cpu_chain, CPU_DYING | param->mod, param->hcpu); + if (task_cpu(param->caller) == cpu) + move_task_off_dead_cpu(cpu, param->caller); /* Force idle task to run as soon as we yield: it should immediately notice cpu is offline and die quickly. */ sched_idle_next(); @@ -191,10 +195,10 @@ static int __ref take_cpu_down(void *_param) static int __ref _cpu_down(unsigned int cpu, int tasks_frozen) { int err, nr_calls = 0; - cpumask_var_t old_allowed; void *hcpu = (void *)(long)cpu; unsigned long mod = tasks_frozen ? CPU_TASKS_FROZEN : 0; struct take_cpu_down_param tcd_param = { + .caller = current, .mod = mod, .hcpu = hcpu, }; @@ -205,9 +209,6 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen) if (!cpu_online(cpu)) return -EINVAL; - if (!alloc_cpumask_var(&old_allowed, GFP_KERNEL)) - return -ENOMEM; - cpu_hotplug_begin(); set_cpu_active(cpu, false); err = __raw_notifier_call_chain(&cpu_chain, CPU_DOWN_PREPARE | mod, @@ -224,10 +225,6 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen) goto out_release; } - /* Ensure that we are not runnable on dying cpu */ - cpumask_copy(old_allowed, ¤t->cpus_allowed); - set_cpus_allowed_ptr(current, cpu_active_mask); - err = __stop_machine(take_cpu_down, &tcd_param, cpumask_of(cpu)); if (err) { set_cpu_active(cpu, true); @@ -236,7 +233,7 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen) hcpu) == NOTIFY_BAD) BUG(); - goto out_allowed; + goto out_release; } BUG_ON(cpu_online(cpu)); @@ -254,8 +251,6 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen) check_for_tasks(cpu); -out_allowed: - set_cpus_allowed_ptr(current, old_allowed); out_release: cpu_hotplug_done(); if (!err) { @@ -263,7 +258,6 @@ out_release: hcpu) == NOTIFY_BAD) BUG(); } - free_cpumask_var(old_allowed); return err; } diff --git a/kernel/sched.c b/kernel/sched.c index 165b532..11119de 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -5442,7 +5442,7 @@ static int migration_thread(void *data) /* * Figure out where task on dead CPU should go, use force if necessary. */ -static void move_task_off_dead_cpu(int dead_cpu, struct task_struct *p) +void move_task_off_dead_cpu(int dead_cpu, struct task_struct *p) { struct rq *rq = cpu_rq(dead_cpu); int needs_cpu, uninitialized_var(dest_cpu); -- cgit v0.10.2 From 9084bb8246ea935b98320554229e2f371f7f52fa Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Mon, 15 Mar 2010 10:10:27 +0100 Subject: sched: Make select_fallback_rq() cpuset friendly Introduce cpuset_cpus_allowed_fallback() helper to fix the cpuset problems with select_fallback_rq(). It can be called from any context and can't use any cpuset locks including task_lock(). It is called when the task doesn't have online cpus in ->cpus_allowed but ttwu/etc must be able to find a suitable cpu. I am not proud of this patch. Everything which needs such a fat comment can't be good even if correct. But I'd prefer to not change the locking rules in the code I hardly understand, and in any case I believe this simple change make the code much more correct compared to deadlocks we currently have. Signed-off-by: Oleg Nesterov Signed-off-by: Peter Zijlstra LKML-Reference: <20100315091027.GA9155@redhat.com> Signed-off-by: Ingo Molnar diff --git a/include/linux/cpuset.h b/include/linux/cpuset.h index eeaaee7..a73454a 100644 --- a/include/linux/cpuset.h +++ b/include/linux/cpuset.h @@ -21,6 +21,7 @@ extern int number_of_cpusets; /* How many cpusets are defined in system? */ extern int cpuset_init(void); extern void cpuset_init_smp(void); extern void cpuset_cpus_allowed(struct task_struct *p, struct cpumask *mask); +extern int cpuset_cpus_allowed_fallback(struct task_struct *p); extern nodemask_t cpuset_mems_allowed(struct task_struct *p); #define cpuset_current_mems_allowed (current->mems_allowed) void cpuset_init_current_mems_allowed(void); @@ -101,6 +102,12 @@ static inline void cpuset_cpus_allowed(struct task_struct *p, cpumask_copy(mask, cpu_possible_mask); } +static inline int cpuset_cpus_allowed_fallback(struct task_struct *p) +{ + cpumask_copy(&p->cpus_allowed, cpu_possible_mask); + return cpumask_any(cpu_active_mask); +} + static inline nodemask_t cpuset_mems_allowed(struct task_struct *p) { return node_possible_map; diff --git a/kernel/cpuset.c b/kernel/cpuset.c index 9a747f5..9a50c5f 100644 --- a/kernel/cpuset.c +++ b/kernel/cpuset.c @@ -2188,6 +2188,48 @@ void cpuset_cpus_allowed(struct task_struct *tsk, struct cpumask *pmask) mutex_unlock(&callback_mutex); } +int cpuset_cpus_allowed_fallback(struct task_struct *tsk) +{ + const struct cpuset *cs; + int cpu; + + rcu_read_lock(); + cs = task_cs(tsk); + if (cs) + cpumask_copy(&tsk->cpus_allowed, cs->cpus_allowed); + rcu_read_unlock(); + + /* + * We own tsk->cpus_allowed, nobody can change it under us. + * + * But we used cs && cs->cpus_allowed lockless and thus can + * race with cgroup_attach_task() or update_cpumask() and get + * the wrong tsk->cpus_allowed. However, both cases imply the + * subsequent cpuset_change_cpumask()->set_cpus_allowed_ptr() + * which takes task_rq_lock(). + * + * If we are called after it dropped the lock we must see all + * changes in tsk_cs()->cpus_allowed. Otherwise we can temporary + * set any mask even if it is not right from task_cs() pov, + * the pending set_cpus_allowed_ptr() will fix things. + */ + + cpu = cpumask_any_and(&tsk->cpus_allowed, cpu_active_mask); + if (cpu >= nr_cpu_ids) { + /* + * Either tsk->cpus_allowed is wrong (see above) or it + * is actually empty. The latter case is only possible + * if we are racing with remove_tasks_in_empty_cpuset(). + * Like above we can temporary set any mask and rely on + * set_cpus_allowed_ptr() as synchronization point. + */ + cpumask_copy(&tsk->cpus_allowed, cpu_possible_mask); + cpu = cpumask_any(cpu_active_mask); + } + + return cpu; +} + void cpuset_init_current_mems_allowed(void) { nodes_setall(current->mems_allowed); diff --git a/kernel/sched.c b/kernel/sched.c index 11119de..9a38c7a 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2300,9 +2300,7 @@ static int select_fallback_rq(int cpu, struct task_struct *p) /* No more Mr. Nice Guy. */ if (unlikely(dest_cpu >= nr_cpu_ids)) { - cpumask_copy(&p->cpus_allowed, cpu_possible_mask); - dest_cpu = cpumask_any(cpu_active_mask); - + dest_cpu = cpuset_cpus_allowed_fallback(p); /* * Don't tell them about moving exiting tasks or * kernel threads (both mm NULL), since they never -- cgit v0.10.2 From 0017d735092844118bef006696a750a0e4ef6ebd Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 24 Mar 2010 18:34:10 +0100 Subject: sched: Fix TASK_WAKING vs fork deadlock Oleg noticed a few races with the TASK_WAKING usage on fork. - since TASK_WAKING is basically a spinlock, it should be IRQ safe - since we set TASK_WAKING (*) without holding rq->lock it could be there still is a rq->lock holder, thereby not actually providing full serialization. (*) in fact we clear PF_STARTING, which in effect enables TASK_WAKING. Cure the second issue by not setting TASK_WAKING in sched_fork(), but only temporarily in wake_up_new_task() while calling select_task_rq(). Cure the first by holding rq->lock around the select_task_rq() call, this will disable IRQs, this however requires that we push down the rq->lock release into select_task_rq_fair()'s cgroup stuff. Because select_task_rq_fair() still needs to drop the rq->lock we cannot fully get rid of TASK_WAKING. Reported-by: Oleg Nesterov Signed-off-by: Peter Zijlstra LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/include/linux/sched.h b/include/linux/sched.h index 8bea407..fb6c188 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1046,7 +1046,8 @@ struct sched_class { void (*put_prev_task) (struct rq *rq, struct task_struct *p); #ifdef CONFIG_SMP - int (*select_task_rq)(struct task_struct *p, int sd_flag, int flags); + int (*select_task_rq)(struct rq *rq, struct task_struct *p, + int sd_flag, int flags); void (*pre_schedule) (struct rq *this_rq, struct task_struct *task); void (*post_schedule) (struct rq *this_rq); diff --git a/kernel/sched.c b/kernel/sched.c index 9a38c7a..dcd1773 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -916,14 +916,10 @@ static inline void finish_lock_switch(struct rq *rq, struct task_struct *prev) /* * Check whether the task is waking, we use this to synchronize against * ttwu() so that task_cpu() reports a stable number. - * - * We need to make an exception for PF_STARTING tasks because the fork - * path might require task_rq_lock() to work, eg. it can call - * set_cpus_allowed_ptr() from the cpuset clone_ns code. */ static inline int task_is_waking(struct task_struct *p) { - return unlikely((p->state == TASK_WAKING) && !(p->flags & PF_STARTING)); + return unlikely(p->state == TASK_WAKING); } /* @@ -2320,9 +2316,9 @@ static int select_fallback_rq(int cpu, struct task_struct *p) * The caller (fork, wakeup) owns TASK_WAKING, ->cpus_allowed is stable. */ static inline -int select_task_rq(struct task_struct *p, int sd_flags, int wake_flags) +int select_task_rq(struct rq *rq, struct task_struct *p, int sd_flags, int wake_flags) { - int cpu = p->sched_class->select_task_rq(p, sd_flags, wake_flags); + int cpu = p->sched_class->select_task_rq(rq, p, sd_flags, wake_flags); /* * In order not to call set_task_cpu() on a blocking task we need @@ -2393,17 +2389,10 @@ static int try_to_wake_up(struct task_struct *p, unsigned int state, if (p->sched_class->task_waking) p->sched_class->task_waking(rq, p); - __task_rq_unlock(rq); - - cpu = select_task_rq(p, SD_BALANCE_WAKE, wake_flags); - if (cpu != orig_cpu) { - /* - * Since we migrate the task without holding any rq->lock, - * we need to be careful with task_rq_lock(), since that - * might end up locking an invalid rq. - */ + cpu = select_task_rq(rq, p, SD_BALANCE_WAKE, wake_flags); + if (cpu != orig_cpu) set_task_cpu(p, cpu); - } + __task_rq_unlock(rq); rq = cpu_rq(cpu); raw_spin_lock(&rq->lock); @@ -2530,11 +2519,11 @@ void sched_fork(struct task_struct *p, int clone_flags) __sched_fork(p); /* - * We mark the process as waking here. This guarantees that + * We mark the process as running here. This guarantees that * nobody will actually run it, and a signal or other external * event cannot wake it up and insert it on the runqueue either. */ - p->state = TASK_WAKING; + p->state = TASK_RUNNING; /* * Revert to default priority/policy on fork if requested. @@ -2601,28 +2590,25 @@ void wake_up_new_task(struct task_struct *p, unsigned long clone_flags) int cpu __maybe_unused = get_cpu(); #ifdef CONFIG_SMP + rq = task_rq_lock(p, &flags); + p->state = TASK_WAKING; + /* * Fork balancing, do it here and not earlier because: * - cpus_allowed can change in the fork path * - any previously selected cpu might disappear through hotplug * - * We still have TASK_WAKING but PF_STARTING is gone now, meaning - * ->cpus_allowed is stable, we have preemption disabled, meaning - * cpu_online_mask is stable. + * We set TASK_WAKING so that select_task_rq() can drop rq->lock + * without people poking at ->cpus_allowed. */ - cpu = select_task_rq(p, SD_BALANCE_FORK, 0); + cpu = select_task_rq(rq, p, SD_BALANCE_FORK, 0); set_task_cpu(p, cpu); -#endif - - /* - * Since the task is not on the rq and we still have TASK_WAKING set - * nobody else will migrate this task. - */ - rq = cpu_rq(cpu); - raw_spin_lock_irqsave(&rq->lock, flags); - BUG_ON(p->state != TASK_WAKING); p->state = TASK_RUNNING; + task_rq_unlock(rq, &flags); +#endif + + rq = task_rq_lock(p, &flags); activate_task(rq, p, 0); trace_sched_wakeup_new(rq, p, 1); check_preempt_curr(rq, p, WF_FORK); @@ -3068,19 +3054,15 @@ void sched_exec(void) { struct task_struct *p = current; struct migration_req req; - int dest_cpu, this_cpu; unsigned long flags; struct rq *rq; - - this_cpu = get_cpu(); - dest_cpu = p->sched_class->select_task_rq(p, SD_BALANCE_EXEC, 0); - if (dest_cpu == this_cpu) { - put_cpu(); - return; - } + int dest_cpu; rq = task_rq_lock(p, &flags); - put_cpu(); + dest_cpu = p->sched_class->select_task_rq(rq, p, SD_BALANCE_EXEC, 0); + if (dest_cpu == smp_processor_id()) + goto unlock; + /* * select_task_rq() can race against ->cpus_allowed */ @@ -3098,6 +3080,7 @@ void sched_exec(void) return; } +unlock: task_rq_unlock(rq, &flags); } diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 49ad993..8a5e763 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -1423,7 +1423,8 @@ select_idle_sibling(struct task_struct *p, struct sched_domain *sd, int target) * * preempt must be disabled. */ -static int select_task_rq_fair(struct task_struct *p, int sd_flag, int wake_flags) +static int +select_task_rq_fair(struct rq *rq, struct task_struct *p, int sd_flag, int wake_flags) { struct sched_domain *tmp, *affine_sd = NULL, *sd = NULL; int cpu = smp_processor_id(); @@ -1521,8 +1522,11 @@ static int select_task_rq_fair(struct task_struct *p, int sd_flag, int wake_flag cpumask_weight(sched_domain_span(sd)))) tmp = affine_sd; - if (tmp) + if (tmp) { + raw_spin_unlock(&rq->lock); update_shares(tmp); + raw_spin_lock(&rq->lock); + } } #endif diff --git a/kernel/sched_idletask.c b/kernel/sched_idletask.c index a8a6d8a..5af709f 100644 --- a/kernel/sched_idletask.c +++ b/kernel/sched_idletask.c @@ -6,7 +6,8 @@ */ #ifdef CONFIG_SMP -static int select_task_rq_idle(struct task_struct *p, int sd_flag, int flags) +static int +select_task_rq_idle(struct rq *rq, struct task_struct *p, int sd_flag, int flags) { return task_cpu(p); /* IDLE tasks as never migrated */ } diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c index 012d69b..fde895f 100644 --- a/kernel/sched_rt.c +++ b/kernel/sched_rt.c @@ -948,10 +948,9 @@ static void yield_task_rt(struct rq *rq) #ifdef CONFIG_SMP static int find_lowest_rq(struct task_struct *task); -static int select_task_rq_rt(struct task_struct *p, int sd_flag, int flags) +static int +select_task_rq_rt(struct rq *rq, struct task_struct *p, int sd_flag, int flags) { - struct rq *rq = task_rq(p); - if (sd_flag != SD_BALANCE_WAKE) return smp_processor_id(); -- cgit v0.10.2 From 65cc8e4859ff29a9ddc989c88557d6059834c2a2 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 25 Mar 2010 21:05:16 +0100 Subject: sched: Optimize task_rq_lock() Now that we hold the rq->lock over set_task_cpu() again, we can do away with most of the TASK_WAKING checks and reduce them again to set_cpus_allowed_ptr(). Removes some conditionals from scheduling hot-paths. Signed-off-by: Peter Zijlstra Cc: Oleg Nesterov LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/kernel/sched.c b/kernel/sched.c index dcd1773..51d336e 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -914,8 +914,8 @@ static inline void finish_lock_switch(struct rq *rq, struct task_struct *prev) #endif /* __ARCH_WANT_UNLOCKED_CTXSW */ /* - * Check whether the task is waking, we use this to synchronize against - * ttwu() so that task_cpu() reports a stable number. + * Check whether the task is waking, we use this to synchronize ->cpus_allowed + * against ttwu(). */ static inline int task_is_waking(struct task_struct *p) { @@ -932,11 +932,9 @@ static inline struct rq *__task_rq_lock(struct task_struct *p) struct rq *rq; for (;;) { - while (task_is_waking(p)) - cpu_relax(); rq = task_rq(p); raw_spin_lock(&rq->lock); - if (likely(rq == task_rq(p) && !task_is_waking(p))) + if (likely(rq == task_rq(p))) return rq; raw_spin_unlock(&rq->lock); } @@ -953,12 +951,10 @@ static struct rq *task_rq_lock(struct task_struct *p, unsigned long *flags) struct rq *rq; for (;;) { - while (task_is_waking(p)) - cpu_relax(); local_irq_save(*flags); rq = task_rq(p); raw_spin_lock(&rq->lock); - if (likely(rq == task_rq(p) && !task_is_waking(p))) + if (likely(rq == task_rq(p))) return rq; raw_spin_unlock_irqrestore(&rq->lock, *flags); } @@ -5262,7 +5258,18 @@ int set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask) struct rq *rq; int ret = 0; + /* + * Serialize against TASK_WAKING so that ttwu() and wunt() can + * drop the rq->lock and still rely on ->cpus_allowed. + */ +again: + while (task_is_waking(p)) + cpu_relax(); rq = task_rq_lock(p, &flags); + if (task_is_waking(p)) { + task_rq_unlock(rq, &flags); + goto again; + } if (!cpumask_intersects(new_mask, cpu_active_mask)) { ret = -EINVAL; -- cgit v0.10.2 From cc87f76a601d2d256118f7bab15e35254356ae21 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 26 Mar 2010 12:22:14 +0100 Subject: sched: Fix nr_uninterruptible count The cpuload calculation in calc_load_account_active() assumes rq->nr_uninterruptible will not change on an offline cpu after migrate_nr_uninterruptible(). However the recent migrate on wakeup changes broke that and would result in decrementing the offline cpu's rq->nr_uninterruptible. Fix this by accounting the nr_uninterruptible on the waking cpu. Signed-off-by: Peter Zijlstra LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/kernel/sched.c b/kernel/sched.c index 51d336e..14c8d2a 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -2378,8 +2378,12 @@ static int try_to_wake_up(struct task_struct *p, unsigned int state, * * First fix up the nr_uninterruptible count: */ - if (task_contributes_to_load(p)) - rq->nr_uninterruptible--; + if (task_contributes_to_load(p)) { + if (likely(cpu_online(orig_cpu))) + rq->nr_uninterruptible--; + else + this_rq()->nr_uninterruptible--; + } p->state = TASK_WAKING; if (p->sched_class->task_waking) -- cgit v0.10.2 From 371fd7e7a56a5c136d31aa980011bd2f131c3ef5 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Wed, 24 Mar 2010 16:38:48 +0100 Subject: sched: Add enqueue/dequeue flags In order to reduce the dependency on TASK_WAKING rework the enqueue interface to support a proper flags field. Replace the int wakeup, bool head arguments with an int flags argument and create the following flags: ENQUEUE_WAKEUP - the enqueue is a wakeup of a sleeping task, ENQUEUE_WAKING - the enqueue has relative vruntime due to having sched_class::task_waking() called, ENQUEUE_HEAD - the waking task should be places on the head of the priority queue (where appropriate). For symmetry also convert sched_class::dequeue() to a flags scheme. Signed-off-by: Peter Zijlstra LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/include/linux/sched.h b/include/linux/sched.h index fb6c188..e3e900f 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1032,12 +1032,17 @@ struct sched_domain; #define WF_SYNC 0x01 /* waker goes to sleep after wakup */ #define WF_FORK 0x02 /* child wakeup after fork */ +#define ENQUEUE_WAKEUP 1 +#define ENQUEUE_WAKING 2 +#define ENQUEUE_HEAD 4 + +#define DEQUEUE_SLEEP 1 + struct sched_class { const struct sched_class *next; - void (*enqueue_task) (struct rq *rq, struct task_struct *p, int wakeup, - bool head); - void (*dequeue_task) (struct rq *rq, struct task_struct *p, int sleep); + void (*enqueue_task) (struct rq *rq, struct task_struct *p, int flags); + void (*dequeue_task) (struct rq *rq, struct task_struct *p, int flags); void (*yield_task) (struct rq *rq); void (*check_preempt_curr) (struct rq *rq, struct task_struct *p, int flags); diff --git a/kernel/sched.c b/kernel/sched.c index 14c8d2a..4a57e96 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -1877,44 +1877,43 @@ static void update_avg(u64 *avg, u64 sample) *avg += diff >> 3; } -static void -enqueue_task(struct rq *rq, struct task_struct *p, int wakeup, bool head) +static void enqueue_task(struct rq *rq, struct task_struct *p, int flags) { update_rq_clock(rq); sched_info_queued(p); - p->sched_class->enqueue_task(rq, p, wakeup, head); + p->sched_class->enqueue_task(rq, p, flags); p->se.on_rq = 1; } -static void dequeue_task(struct rq *rq, struct task_struct *p, int sleep) +static void dequeue_task(struct rq *rq, struct task_struct *p, int flags) { update_rq_clock(rq); sched_info_dequeued(p); - p->sched_class->dequeue_task(rq, p, sleep); + p->sched_class->dequeue_task(rq, p, flags); p->se.on_rq = 0; } /* * activate_task - move a task to the runqueue. */ -static void activate_task(struct rq *rq, struct task_struct *p, int wakeup) +static void activate_task(struct rq *rq, struct task_struct *p, int flags) { if (task_contributes_to_load(p)) rq->nr_uninterruptible--; - enqueue_task(rq, p, wakeup, false); + enqueue_task(rq, p, flags); inc_nr_running(rq); } /* * deactivate_task - remove a task from the runqueue. */ -static void deactivate_task(struct rq *rq, struct task_struct *p, int sleep) +static void deactivate_task(struct rq *rq, struct task_struct *p, int flags) { if (task_contributes_to_load(p)) rq->nr_uninterruptible++; - dequeue_task(rq, p, sleep); + dequeue_task(rq, p, flags); dec_nr_running(rq); } @@ -2353,6 +2352,7 @@ static int try_to_wake_up(struct task_struct *p, unsigned int state, { int cpu, orig_cpu, this_cpu, success = 0; unsigned long flags; + unsigned long en_flags = ENQUEUE_WAKEUP; struct rq *rq; this_cpu = get_cpu(); @@ -2386,8 +2386,10 @@ static int try_to_wake_up(struct task_struct *p, unsigned int state, } p->state = TASK_WAKING; - if (p->sched_class->task_waking) + if (p->sched_class->task_waking) { p->sched_class->task_waking(rq, p); + en_flags |= ENQUEUE_WAKING; + } cpu = select_task_rq(rq, p, SD_BALANCE_WAKE, wake_flags); if (cpu != orig_cpu) @@ -2432,7 +2434,7 @@ out_activate: schedstat_inc(p, se.statistics.nr_wakeups_local); else schedstat_inc(p, se.statistics.nr_wakeups_remote); - activate_task(rq, p, 1); + activate_task(rq, p, en_flags); success = 1; out_running: @@ -3623,7 +3625,7 @@ need_resched_nonpreemptible: if (unlikely(signal_pending_state(prev->state, prev))) prev->state = TASK_RUNNING; else - deactivate_task(rq, prev, 1); + deactivate_task(rq, prev, DEQUEUE_SLEEP); switch_count = &prev->nvcsw; } @@ -4193,7 +4195,7 @@ void rt_mutex_setprio(struct task_struct *p, int prio) if (running) p->sched_class->set_curr_task(rq); if (on_rq) { - enqueue_task(rq, p, 0, oldprio < prio); + enqueue_task(rq, p, oldprio < prio ? ENQUEUE_HEAD : 0); check_class_changed(rq, p, prev_class, oldprio, running); } @@ -4236,7 +4238,7 @@ void set_user_nice(struct task_struct *p, long nice) delta = p->prio - old_prio; if (on_rq) { - enqueue_task(rq, p, 0, false); + enqueue_task(rq, p, 0); /* * If the task increased its priority or is running and * lowered its priority, then reschedule its CPU: @@ -8180,7 +8182,7 @@ void sched_move_task(struct task_struct *tsk) if (unlikely(running)) tsk->sched_class->set_curr_task(rq); if (on_rq) - enqueue_task(rq, tsk, 0, false); + enqueue_task(rq, tsk, 0); task_rq_unlock(rq, &flags); } diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 8a5e763..88d3053 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -757,9 +757,6 @@ place_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int initial) se->vruntime = vruntime; } -#define ENQUEUE_WAKEUP 1 -#define ENQUEUE_MIGRATE 2 - static void enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags) { @@ -767,7 +764,7 @@ enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags) * Update the normalized vruntime before updating min_vruntime * through callig update_curr(). */ - if (!(flags & ENQUEUE_WAKEUP) || (flags & ENQUEUE_MIGRATE)) + if (!(flags & ENQUEUE_WAKEUP) || (flags & ENQUEUE_WAKING)) se->vruntime += cfs_rq->min_vruntime; /* @@ -803,7 +800,7 @@ static void clear_buddies(struct cfs_rq *cfs_rq, struct sched_entity *se) } static void -dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int sleep) +dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int flags) { /* * Update run-time statistics of the 'current'. @@ -811,7 +808,7 @@ dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int sleep) update_curr(cfs_rq); update_stats_dequeue(cfs_rq, se); - if (sleep) { + if (flags & DEQUEUE_SLEEP) { #ifdef CONFIG_SCHEDSTATS if (entity_is_task(se)) { struct task_struct *tsk = task_of(se); @@ -836,7 +833,7 @@ dequeue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int sleep) * update can refer to the ->curr item and we need to reflect this * movement in our normalized position. */ - if (!sleep) + if (!(flags & DEQUEUE_SLEEP)) se->vruntime -= cfs_rq->min_vruntime; } @@ -1045,16 +1042,10 @@ static inline void hrtick_update(struct rq *rq) * then put the task into the rbtree: */ static void -enqueue_task_fair(struct rq *rq, struct task_struct *p, int wakeup, bool head) +enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags) { struct cfs_rq *cfs_rq; struct sched_entity *se = &p->se; - int flags = 0; - - if (wakeup) - flags |= ENQUEUE_WAKEUP; - if (p->state == TASK_WAKING) - flags |= ENQUEUE_MIGRATE; for_each_sched_entity(se) { if (se->on_rq) @@ -1072,18 +1063,18 @@ enqueue_task_fair(struct rq *rq, struct task_struct *p, int wakeup, bool head) * decreased. We remove the task from the rbtree and * update the fair scheduling stats: */ -static void dequeue_task_fair(struct rq *rq, struct task_struct *p, int sleep) +static void dequeue_task_fair(struct rq *rq, struct task_struct *p, int flags) { struct cfs_rq *cfs_rq; struct sched_entity *se = &p->se; for_each_sched_entity(se) { cfs_rq = cfs_rq_of(se); - dequeue_entity(cfs_rq, se, sleep); + dequeue_entity(cfs_rq, se, flags); /* Don't dequeue parent if it has other entities besides us */ if (cfs_rq->load.weight) break; - sleep = 1; + flags |= DEQUEUE_SLEEP; } hrtick_update(rq); diff --git a/kernel/sched_idletask.c b/kernel/sched_idletask.c index 5af709f..bea2b8f 100644 --- a/kernel/sched_idletask.c +++ b/kernel/sched_idletask.c @@ -33,7 +33,7 @@ static struct task_struct *pick_next_task_idle(struct rq *rq) * message if some code attempts to do it: */ static void -dequeue_task_idle(struct rq *rq, struct task_struct *p, int sleep) +dequeue_task_idle(struct rq *rq, struct task_struct *p, int flags) { raw_spin_unlock_irq(&rq->lock); printk(KERN_ERR "bad: scheduling from the idle thread!\n"); diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c index fde895f..8afb953 100644 --- a/kernel/sched_rt.c +++ b/kernel/sched_rt.c @@ -888,20 +888,20 @@ static void dequeue_rt_entity(struct sched_rt_entity *rt_se) * Adding/removing a task to/from a priority array: */ static void -enqueue_task_rt(struct rq *rq, struct task_struct *p, int wakeup, bool head) +enqueue_task_rt(struct rq *rq, struct task_struct *p, int flags) { struct sched_rt_entity *rt_se = &p->rt; - if (wakeup) + if (flags & ENQUEUE_WAKEUP) rt_se->timeout = 0; - enqueue_rt_entity(rt_se, head); + enqueue_rt_entity(rt_se, flags & ENQUEUE_HEAD); if (!task_current(rq, p) && p->rt.nr_cpus_allowed > 1) enqueue_pushable_task(rq, p); } -static void dequeue_task_rt(struct rq *rq, struct task_struct *p, int sleep) +static void dequeue_task_rt(struct rq *rq, struct task_struct *p, int flags) { struct sched_rt_entity *rt_se = &p->rt; -- cgit v0.10.2 From dd48744964296b5713032ea1d66eb9e3d990e287 Mon Sep 17 00:00:00 2001 From: Zhu Yi Date: Mon, 22 Mar 2010 02:28:41 -0700 Subject: iwlwifi: fix DMA allocation warnings Below warning is triggered sometimes at module removal time when CONFIG_DMA_API_DEBUG is enabled. This should be caused by we didn't unmap pending commands (enqueued, but no complete notification received) for the Tx command queue. [ 1583.107469] ------------[ cut here ]------------ [ 1583.107539] WARNING: at lib/dma-debug.c:688 dma_debug_device_change+0x13c/0x180() [ 1583.107617] Hardware name: ... [ 1583.107664] pci 0000:04:00.0: DMA-API: device driver has pending DMA allocations while released from device [count=1] [ 1583.107713] Modules linked in: ... [ 1583.111661] Pid: 16970, comm: modprobe Tainted: G W 2.6.34-rc1-wl #33 [ 1583.111727] Call Trace: [ 1583.111779] [] ? dma_debug_device_change+0x13c/0x180 [ 1583.111833] [] ? dma_debug_device_change+0x13c/0x180 [ 1583.111908] [] warn_slowpath_common+0x71/0xd0 [ 1583.111963] [] ? dma_debug_device_change+0x13c/0x180 [ 1583.112016] [] warn_slowpath_fmt+0x2b/0x30 [ 1583.112086] [] dma_debug_device_change+0x13c/0x180 [ 1583.112142] [] notifier_call_chain+0x53/0x90 [ 1583.112198] [] ? down_read+0x6e/0x90 [ 1583.112271] [] __blocking_notifier_call_chain+0x49/0x70 [ 1583.112326] [] blocking_notifier_call_chain+0x1f/0x30 [ 1583.112380] [] __device_release_driver+0x8c/0xa0 [ 1583.112451] [] driver_detach+0x8f/0xa0 [ 1583.112538] [] bus_remove_driver+0x82/0x100 [ 1583.112595] [] driver_unregister+0x49/0x80 [ 1583.112671] [] ? sysfs_remove_file+0x12/0x20 [ 1583.112727] [] pci_unregister_driver+0x32/0x80 [ 1583.112791] [] iwl_exit+0x12/0x19 [iwlagn] [ 1583.112848] [] sys_delete_module+0x15a/0x210 [ 1583.112870] [] ? up_read+0x1b/0x30 [ 1583.112893] [] ? trace_hardirqs_off_thunk+0xc/0x10 [ 1583.112924] [] ? trace_hardirqs_on_thunk+0xc/0x10 [ 1583.112947] [] ? do_page_fault+0x1ff/0x3c0 [ 1583.112978] [] ? restore_all_notrace+0x0/0x18 [ 1583.113002] [] ? trace_hardirqs_on_caller+0x20/0x190 [ 1583.113025] [] sysenter_do_call+0x12/0x38 [ 1583.113054] ---[ end trace fc23e059cc4c2ced ]--- Signed-off-by: Zhu Yi Signed-off-by: Reinette Chatre diff --git a/drivers/net/wireless/iwlwifi/iwl-tx.c b/drivers/net/wireless/iwlwifi/iwl-tx.c index 8c12311..6aa3090 100644 --- a/drivers/net/wireless/iwlwifi/iwl-tx.c +++ b/drivers/net/wireless/iwlwifi/iwl-tx.c @@ -193,10 +193,34 @@ void iwl_cmd_queue_free(struct iwl_priv *priv) struct iwl_queue *q = &txq->q; struct device *dev = &priv->pci_dev->dev; int i; + bool huge = false; if (q->n_bd == 0) return; + for (; q->read_ptr != q->write_ptr; + q->read_ptr = iwl_queue_inc_wrap(q->read_ptr, q->n_bd)) { + /* we have no way to tell if it is a huge cmd ATM */ + i = get_cmd_index(q, q->read_ptr, 0); + + if (txq->meta[i].flags & CMD_SIZE_HUGE) { + huge = true; + continue; + } + + pci_unmap_single(priv->pci_dev, + pci_unmap_addr(&txq->meta[i], mapping), + pci_unmap_len(&txq->meta[i], len), + PCI_DMA_BIDIRECTIONAL); + } + if (huge) { + i = q->n_window; + pci_unmap_single(priv->pci_dev, + pci_unmap_addr(&txq->meta[i], mapping), + pci_unmap_len(&txq->meta[i], len), + PCI_DMA_BIDIRECTIONAL); + } + /* De-alloc array of command/tx buffers */ for (i = 0; i <= TFD_CMD_SLOTS; i++) kfree(txq->cmd[i]); @@ -1049,6 +1073,14 @@ int iwl_enqueue_hcmd(struct iwl_priv *priv, struct iwl_host_cmd *cmd) spin_lock_irqsave(&priv->hcmd_lock, flags); + /* If this is a huge cmd, mark the huge flag also on the meta.flags + * of the _original_ cmd. This is used for DMA mapping clean up. + */ + if (cmd->flags & CMD_SIZE_HUGE) { + idx = get_cmd_index(q, q->write_ptr, 0); + txq->meta[idx].flags = CMD_SIZE_HUGE; + } + idx = get_cmd_index(q, q->write_ptr, cmd->flags & CMD_SIZE_HUGE); out_cmd = txq->cmd[idx]; out_meta = &txq->meta[idx]; @@ -1226,6 +1258,7 @@ void iwl_tx_cmd_complete(struct iwl_priv *priv, struct iwl_rx_mem_buffer *rxb) bool huge = !!(pkt->hdr.sequence & SEQ_HUGE_FRAME); struct iwl_device_cmd *cmd; struct iwl_cmd_meta *meta; + struct iwl_tx_queue *txq = &priv->txq[IWL_CMD_QUEUE_NUM]; /* If a Tx command is being handled and it isn't in the actual * command queue then there a command routing bug has been introduced @@ -1239,9 +1272,17 @@ void iwl_tx_cmd_complete(struct iwl_priv *priv, struct iwl_rx_mem_buffer *rxb) return; } - cmd_index = get_cmd_index(&priv->txq[IWL_CMD_QUEUE_NUM].q, index, huge); - cmd = priv->txq[IWL_CMD_QUEUE_NUM].cmd[cmd_index]; - meta = &priv->txq[IWL_CMD_QUEUE_NUM].meta[cmd_index]; + /* If this is a huge cmd, clear the huge flag on the meta.flags + * of the _original_ cmd. So that iwl_cmd_queue_free won't unmap + * the DMA buffer for the scan (huge) command. + */ + if (huge) { + cmd_index = get_cmd_index(&txq->q, index, 0); + txq->meta[cmd_index].flags = 0; + } + cmd_index = get_cmd_index(&txq->q, index, huge); + cmd = txq->cmd[cmd_index]; + meta = &txq->meta[cmd_index]; pci_unmap_single(priv->pci_dev, pci_unmap_addr(meta, mapping), @@ -1263,6 +1304,7 @@ void iwl_tx_cmd_complete(struct iwl_priv *priv, struct iwl_rx_mem_buffer *rxb) get_cmd_string(cmd->hdr.cmd)); wake_up_interruptible(&priv->wait_command_queue); } + meta->flags = 0; } EXPORT_SYMBOL(iwl_tx_cmd_complete); -- cgit v0.10.2 From 04f2dec1c3d375c4072613880f28f43b66524876 Mon Sep 17 00:00:00 2001 From: Shanyu Zhao Date: Fri, 19 Mar 2010 13:34:45 -0700 Subject: iwlwifi: use consistent table for tx data collect When collecting tx data for non-aggregation packets in rate scaling, if the tx data matches "other table", it still uses current table to update the stats and calculate average throughput in function rs_collect_tx_data(). This can mess up the rate scaling data structure and cause a kernel panic in a BUG_ON statement in rs_rate_scale_perform(). To fix this bug, we pass table pointer instead of window pointer (pointed to by table pointer) to function rs_collect_tx_data() so that the table being used is consistent. Signed-off-by: Shanyu Zhao Signed-off-by: Henry Zhang Signed-off-by: Reinette Chatre diff --git a/drivers/net/wireless/iwlwifi/iwl-agn-rs.c b/drivers/net/wireless/iwlwifi/iwl-agn-rs.c index 8bf7c20..be00cb3 100644 --- a/drivers/net/wireless/iwlwifi/iwl-agn-rs.c +++ b/drivers/net/wireless/iwlwifi/iwl-agn-rs.c @@ -345,6 +345,17 @@ static inline int get_num_of_ant_from_rate(u32 rate_n_flags) !!(rate_n_flags & RATE_MCS_ANT_C_MSK); } +/* + * Static function to get the expected throughput from an iwl_scale_tbl_info + * that wraps a NULL pointer check + */ +static s32 get_expected_tpt(struct iwl_scale_tbl_info *tbl, int rs_index) +{ + if (tbl->expected_tpt) + return tbl->expected_tpt[rs_index]; + return 0; +} + /** * rs_collect_tx_data - Update the success/failure sliding window * @@ -352,19 +363,21 @@ static inline int get_num_of_ant_from_rate(u32 rate_n_flags) * at this rate. window->data contains the bitmask of successful * packets. */ -static int rs_collect_tx_data(struct iwl_rate_scale_data *windows, - int scale_index, s32 tpt, int attempts, - int successes) +static int rs_collect_tx_data(struct iwl_scale_tbl_info *tbl, + int scale_index, int attempts, int successes) { struct iwl_rate_scale_data *window = NULL; static const u64 mask = (((u64)1) << (IWL_RATE_MAX_WINDOW - 1)); - s32 fail_count; + s32 fail_count, tpt; if (scale_index < 0 || scale_index >= IWL_RATE_COUNT) return -EINVAL; /* Select window for current tx bit rate */ - window = &(windows[scale_index]); + window = &(tbl->win[scale_index]); + + /* Get expected throughput */ + tpt = get_expected_tpt(tbl, scale_index); /* * Keep track of only the latest 62 tx frame attempts in this rate's @@ -738,16 +751,6 @@ static bool table_type_matches(struct iwl_scale_tbl_info *a, return (a->lq_type == b->lq_type) && (a->ant_type == b->ant_type) && (a->is_SGI == b->is_SGI); } -/* - * Static function to get the expected throughput from an iwl_scale_tbl_info - * that wraps a NULL pointer check - */ -static s32 get_expected_tpt(struct iwl_scale_tbl_info *tbl, int rs_index) -{ - if (tbl->expected_tpt) - return tbl->expected_tpt[rs_index]; - return 0; -} /* * mac80211 sends us Tx status @@ -764,12 +767,10 @@ static void rs_tx_status(void *priv_r, struct ieee80211_supported_band *sband, struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data; struct iwl_priv *priv = (struct iwl_priv *)priv_r; struct ieee80211_tx_info *info = IEEE80211_SKB_CB(skb); - struct iwl_rate_scale_data *window = NULL; enum mac80211_rate_control_flags mac_flags; u32 tx_rate; struct iwl_scale_tbl_info tbl_type; - struct iwl_scale_tbl_info *curr_tbl, *other_tbl; - s32 tpt = 0; + struct iwl_scale_tbl_info *curr_tbl, *other_tbl, *tmp_tbl; IWL_DEBUG_RATE_LIMIT(priv, "get frame ack response, update rate scale window\n"); @@ -852,7 +853,6 @@ static void rs_tx_status(void *priv_r, struct ieee80211_supported_band *sband, IWL_DEBUG_RATE(priv, "Neither active nor search matches tx rate\n"); return; } - window = (struct iwl_rate_scale_data *)&(curr_tbl->win[0]); /* * Updating the frame history depends on whether packets were @@ -865,8 +865,7 @@ static void rs_tx_status(void *priv_r, struct ieee80211_supported_band *sband, tx_rate = le32_to_cpu(table->rs_table[0].rate_n_flags); rs_get_tbl_info_from_mcs(tx_rate, priv->band, &tbl_type, &rs_index); - tpt = get_expected_tpt(curr_tbl, rs_index); - rs_collect_tx_data(window, rs_index, tpt, + rs_collect_tx_data(curr_tbl, rs_index, info->status.ampdu_ack_len, info->status.ampdu_ack_map); @@ -896,19 +895,13 @@ static void rs_tx_status(void *priv_r, struct ieee80211_supported_band *sband, * table as active/search. */ if (table_type_matches(&tbl_type, curr_tbl)) - tpt = get_expected_tpt(curr_tbl, rs_index); + tmp_tbl = curr_tbl; else if (table_type_matches(&tbl_type, other_tbl)) - tpt = get_expected_tpt(other_tbl, rs_index); + tmp_tbl = other_tbl; else continue; - - /* Constants mean 1 transmission, 0 successes */ - if (i < retries) - rs_collect_tx_data(window, rs_index, tpt, 1, - 0); - else - rs_collect_tx_data(window, rs_index, tpt, 1, - legacy_success); + rs_collect_tx_data(tmp_tbl, rs_index, 1, + i < retries ? 0 : legacy_success); } /* Update success/fail counts if not searching for new mode */ -- cgit v0.10.2 From 472a474c6630efd195d3738339fd1bdc8aa3b1aa Mon Sep 17 00:00:00 2001 From: Suresh Siddha Date: Wed, 31 Mar 2010 18:04:47 -0700 Subject: x86: Fix double enable_IR_x2apic() call on SMP kernel on !SMP boards Jan Grossmann reported kernel boot panic while booting SMP kernel on his system with a single core cpu. SMP kernels call enable_IR_x2apic() from native_smp_prepare_cpus() and on platforms where the kernel doesn't find SMP configuration we ended up again calling enable_IR_x2apic() from the APIC_init_uniprocessor() call in the smp_sanity_check(). Thus leading to kernel panic. Don't call enable_IR_x2apic() and default_setup_apic_routing() from APIC_init_uniprocessor() in CONFIG_SMP case. NOTE: this kind of non-idempotent and assymetric initialization sequence is rather fragile and unclean, we'll clean that up in v2.6.35. This is the minimal fix for v2.6.34. Reported-by: Jan.Grossmann@kielnet.net Signed-off-by: Suresh Siddha Cc: Cc: Cc: Cc: Cc: Cc: # [v2.6.32.x, v2.6.33.x] LKML-Reference: <1270083887.7835.78.camel@sbs-t61.sc.intel.com> Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index 00187f1..e5a4a1e 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -1640,8 +1640,10 @@ int __init APIC_init_uniprocessor(void) } #endif +#ifndef CONFIG_SMP enable_IR_x2apic(); default_setup_apic_routing(); +#endif verify_local_APIC(); connect_bsp_APIC(); -- cgit v0.10.2 From de0f60ea94e132c858caa64a44b2012e1e8580b0 Mon Sep 17 00:00:00 2001 From: Zhu Yi Date: Tue, 23 Mar 2010 00:45:03 -0700 Subject: iwlwifi: avoid Tx queue memory allocation in interface down We used to free all the Tx queues memory when interface is brought down and reallocate them again in interface up. This requires order-4 allocation for txq->cmd[]. In situations like s2ram, this usually leads to allocation failure in the memory subsystem. The patch fixed this problem by allocating the Tx queues memory only at the first time. Later iwl_down/iwl_up only initialize but don't free and reallocate them. The memory is freed at the device removal time. BTW, we have already done this for the Rx queue. This fixed bug https://bugzilla.kernel.org/show_bug.cgi?id=15551 Signed-off-by: Zhu Yi Acked-by: Wey-Yi Guy Signed-off-by: Reinette Chatre diff --git a/drivers/net/wireless/iwlwifi/iwl-core.c b/drivers/net/wireless/iwlwifi/iwl-core.c index 112149e..894bcb8 100644 --- a/drivers/net/wireless/iwlwifi/iwl-core.c +++ b/drivers/net/wireless/iwlwifi/iwl-core.c @@ -307,10 +307,13 @@ int iwl_hw_nic_init(struct iwl_priv *priv) spin_unlock_irqrestore(&priv->lock, flags); - /* Allocate and init all Tx and Command queues */ - ret = iwl_txq_ctx_reset(priv); - if (ret) - return ret; + /* Allocate or reset and init all Tx and Command queues */ + if (!priv->txq) { + ret = iwl_txq_ctx_alloc(priv); + if (ret) + return ret; + } else + iwl_txq_ctx_reset(priv); set_bit(STATUS_INIT, &priv->status); diff --git a/drivers/net/wireless/iwlwifi/iwl-core.h b/drivers/net/wireless/iwlwifi/iwl-core.h index 4ef7739..732590f 100644 --- a/drivers/net/wireless/iwlwifi/iwl-core.h +++ b/drivers/net/wireless/iwlwifi/iwl-core.h @@ -442,7 +442,8 @@ void iwl_rx_csa(struct iwl_priv *priv, struct iwl_rx_mem_buffer *rxb); /***************************************************** * TX ******************************************************/ -int iwl_txq_ctx_reset(struct iwl_priv *priv); +int iwl_txq_ctx_alloc(struct iwl_priv *priv); +void iwl_txq_ctx_reset(struct iwl_priv *priv); void iwl_hw_txq_free_tfd(struct iwl_priv *priv, struct iwl_tx_queue *txq); int iwl_hw_txq_attach_buf_to_tfd(struct iwl_priv *priv, struct iwl_tx_queue *txq, @@ -456,6 +457,8 @@ void iwl_free_tfds_in_queue(struct iwl_priv *priv, void iwl_txq_update_write_ptr(struct iwl_priv *priv, struct iwl_tx_queue *txq); int iwl_tx_queue_init(struct iwl_priv *priv, struct iwl_tx_queue *txq, int slots_num, u32 txq_id); +void iwl_tx_queue_reset(struct iwl_priv *priv, struct iwl_tx_queue *txq, + int slots_num, u32 txq_id); void iwl_tx_queue_free(struct iwl_priv *priv, int txq_id); int iwl_tx_agg_start(struct iwl_priv *priv, const u8 *ra, u16 tid, u16 *ssn); int iwl_tx_agg_stop(struct iwl_priv *priv , const u8 *ra, u16 tid); diff --git a/drivers/net/wireless/iwlwifi/iwl-tx.c b/drivers/net/wireless/iwlwifi/iwl-tx.c index 6aa3090..343d81a 100644 --- a/drivers/net/wireless/iwlwifi/iwl-tx.c +++ b/drivers/net/wireless/iwlwifi/iwl-tx.c @@ -433,6 +433,26 @@ out_free_arrays: } EXPORT_SYMBOL(iwl_tx_queue_init); +void iwl_tx_queue_reset(struct iwl_priv *priv, struct iwl_tx_queue *txq, + int slots_num, u32 txq_id) +{ + int actual_slots = slots_num; + + if (txq_id == IWL_CMD_QUEUE_NUM) + actual_slots++; + + memset(txq->meta, 0, sizeof(struct iwl_cmd_meta) * actual_slots); + + txq->need_update = 0; + + /* Initialize queue's high/low-water marks, and head/tail indexes */ + iwl_queue_init(priv, &txq->q, TFD_QUEUE_SIZE_MAX, slots_num, txq_id); + + /* Tell device where to find queue */ + priv->cfg->ops->lib->txq_init(priv, txq); +} +EXPORT_SYMBOL(iwl_tx_queue_reset); + /** * iwl_hw_txq_ctx_free - Free TXQ Context * @@ -444,8 +464,7 @@ void iwl_hw_txq_ctx_free(struct iwl_priv *priv) /* Tx queues */ if (priv->txq) { - for (txq_id = 0; txq_id < priv->hw_params.max_txq_num; - txq_id++) + for (txq_id = 0; txq_id < priv->hw_params.max_txq_num; txq_id++) if (txq_id == IWL_CMD_QUEUE_NUM) iwl_cmd_queue_free(priv); else @@ -461,15 +480,15 @@ void iwl_hw_txq_ctx_free(struct iwl_priv *priv) EXPORT_SYMBOL(iwl_hw_txq_ctx_free); /** - * iwl_txq_ctx_reset - Reset TX queue context - * Destroys all DMA structures and initialize them again + * iwl_txq_ctx_alloc - allocate TX queue context + * Allocate all Tx DMA structures and initialize them * * @param priv * @return error code */ -int iwl_txq_ctx_reset(struct iwl_priv *priv) +int iwl_txq_ctx_alloc(struct iwl_priv *priv) { - int ret = 0; + int ret; int txq_id, slots_num; unsigned long flags; @@ -527,8 +546,31 @@ int iwl_txq_ctx_reset(struct iwl_priv *priv) return ret; } +void iwl_txq_ctx_reset(struct iwl_priv *priv) +{ + int txq_id, slots_num; + unsigned long flags; + + spin_lock_irqsave(&priv->lock, flags); + + /* Turn off all Tx DMA fifos */ + priv->cfg->ops->lib->txq_set_sched(priv, 0); + + /* Tell NIC where to find the "keep warm" buffer */ + iwl_write_direct32(priv, FH_KW_MEM_ADDR_REG, priv->kw.dma >> 4); + + spin_unlock_irqrestore(&priv->lock, flags); + + /* Alloc and init all Tx queues, including the command queue (#4) */ + for (txq_id = 0; txq_id < priv->hw_params.max_txq_num; txq_id++) { + slots_num = txq_id == IWL_CMD_QUEUE_NUM ? + TFD_CMD_SLOTS : TFD_TX_CMD_SLOTS; + iwl_tx_queue_reset(priv, &priv->txq[txq_id], slots_num, txq_id); + } +} + /** - * iwl_txq_ctx_stop - Stop all Tx DMA channels, free Tx queue memory + * iwl_txq_ctx_stop - Stop all Tx DMA channels */ void iwl_txq_ctx_stop(struct iwl_priv *priv) { @@ -548,9 +590,6 @@ void iwl_txq_ctx_stop(struct iwl_priv *priv) 1000); } spin_unlock_irqrestore(&priv->lock, flags); - - /* Deallocate memory for all Tx queues */ - iwl_hw_txq_ctx_free(priv); } EXPORT_SYMBOL(iwl_txq_ctx_stop); -- cgit v0.10.2 From c6e718ff8cdcf5e7855077687720b37c4a07650a Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 26 Mar 2010 12:11:06 -0300 Subject: perf symbols: Move more map_groups methods to map.c MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit While writing a standalone test app that uses the symbol system to find kernel space symbols I noticed these also need to be moved. Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index 09e09e7..de31901 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -658,7 +658,7 @@ print_entries: if (h->ms.map == NULL && verbose > 1) { __map_groups__fprintf_maps(&h->thread->mg, - MAP__FUNCTION, fp); + MAP__FUNCTION, verbose, fp); fprintf(fp, "%.10s end\n", graph_dotted_line); } } diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index 9f2963f..e21f980 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -1,4 +1,5 @@ #include "symbol.h" +#include #include #include #include @@ -234,6 +235,37 @@ u64 map__objdump_2ip(struct map *map, u64 addr) return ip; } +void map_groups__init(struct map_groups *self) +{ + int i; + for (i = 0; i < MAP__NR_TYPES; ++i) { + self->maps[i] = RB_ROOT; + INIT_LIST_HEAD(&self->removed_maps[i]); + } +} + +void map_groups__flush(struct map_groups *self) +{ + int type; + + for (type = 0; type < MAP__NR_TYPES; type++) { + struct rb_root *root = &self->maps[type]; + struct rb_node *next = rb_first(root); + + while (next) { + struct map *pos = rb_entry(next, struct map, rb_node); + next = rb_next(&pos->rb_node); + rb_erase(&pos->rb_node, root); + /* + * We may have references to this map, for + * instance in some hist_entry instances, so + * just move them to a separate list. + */ + list_add_tail(&pos->node, &self->removed_maps[pos->type]); + } + } +} + struct symbol *map_groups__find_symbol(struct map_groups *self, enum map_type type, u64 addr, symbol_filter_t filter) @@ -246,6 +278,142 @@ struct symbol *map_groups__find_symbol(struct map_groups *self, return NULL; } +size_t __map_groups__fprintf_maps(struct map_groups *self, + enum map_type type, int verbose, FILE *fp) +{ + size_t printed = fprintf(fp, "%s:\n", map_type__name[type]); + struct rb_node *nd; + + for (nd = rb_first(&self->maps[type]); nd; nd = rb_next(nd)) { + struct map *pos = rb_entry(nd, struct map, rb_node); + printed += fprintf(fp, "Map:"); + printed += map__fprintf(pos, fp); + if (verbose > 2) { + printed += dso__fprintf(pos->dso, type, fp); + printed += fprintf(fp, "--\n"); + } + } + + return printed; +} + +size_t map_groups__fprintf_maps(struct map_groups *self, int verbose, FILE *fp) +{ + size_t printed = 0, i; + for (i = 0; i < MAP__NR_TYPES; ++i) + printed += __map_groups__fprintf_maps(self, i, verbose, fp); + return printed; +} + +static size_t __map_groups__fprintf_removed_maps(struct map_groups *self, + enum map_type type, + int verbose, FILE *fp) +{ + struct map *pos; + size_t printed = 0; + + list_for_each_entry(pos, &self->removed_maps[type], node) { + printed += fprintf(fp, "Map:"); + printed += map__fprintf(pos, fp); + if (verbose > 1) { + printed += dso__fprintf(pos->dso, type, fp); + printed += fprintf(fp, "--\n"); + } + } + return printed; +} + +static size_t map_groups__fprintf_removed_maps(struct map_groups *self, + int verbose, FILE *fp) +{ + size_t printed = 0, i; + for (i = 0; i < MAP__NR_TYPES; ++i) + printed += __map_groups__fprintf_removed_maps(self, i, verbose, fp); + return printed; +} + +size_t map_groups__fprintf(struct map_groups *self, int verbose, FILE *fp) +{ + size_t printed = map_groups__fprintf_maps(self, verbose, fp); + printed += fprintf(fp, "Removed maps:\n"); + return printed + map_groups__fprintf_removed_maps(self, verbose, fp); +} + +int map_groups__fixup_overlappings(struct map_groups *self, struct map *map, + int verbose, FILE *fp) +{ + struct rb_root *root = &self->maps[map->type]; + struct rb_node *next = rb_first(root); + + while (next) { + struct map *pos = rb_entry(next, struct map, rb_node); + next = rb_next(&pos->rb_node); + + if (!map__overlap(pos, map)) + continue; + + if (verbose >= 2) { + fputs("overlapping maps:\n", fp); + map__fprintf(map, fp); + map__fprintf(pos, fp); + } + + rb_erase(&pos->rb_node, root); + /* + * We may have references to this map, for instance in some + * hist_entry instances, so just move them to a separate + * list. + */ + list_add_tail(&pos->node, &self->removed_maps[map->type]); + /* + * Now check if we need to create new maps for areas not + * overlapped by the new map: + */ + if (map->start > pos->start) { + struct map *before = map__clone(pos); + + if (before == NULL) + return -ENOMEM; + + before->end = map->start - 1; + map_groups__insert(self, before); + if (verbose >= 2) + map__fprintf(before, fp); + } + + if (map->end < pos->end) { + struct map *after = map__clone(pos); + + if (after == NULL) + return -ENOMEM; + + after->start = map->end + 1; + map_groups__insert(self, after); + if (verbose >= 2) + map__fprintf(after, fp); + } + } + + return 0; +} + +/* + * XXX This should not really _copy_ te maps, but refcount them. + */ +int map_groups__clone(struct map_groups *self, + struct map_groups *parent, enum map_type type) +{ + struct rb_node *nd; + for (nd = rb_first(&parent->maps[type]); nd; nd = rb_next(nd)) { + struct map *map = rb_entry(nd, struct map, rb_node); + struct map *new = map__clone(map); + if (new == NULL) + return -ENOMEM; + map_groups__insert(self, new); + } + return 0; +} + static u64 map__reloc_map_ip(struct map *map, u64 ip) { return ip + (s64)map->pgoff; diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h index 6a703fa..4e7a11d 100644 --- a/tools/perf/util/map.h +++ b/tools/perf/util/map.h @@ -97,11 +97,14 @@ struct map_groups { }; size_t __map_groups__fprintf_maps(struct map_groups *self, - enum map_type type, FILE *fp); + enum map_type type, int verbose, FILE *fp); void maps__insert(struct rb_root *maps, struct map *map); struct map *maps__find(struct rb_root *maps, u64 addr); void map_groups__init(struct map_groups *self); -size_t map_groups__fprintf_maps(struct map_groups *self, FILE *fp); +int map_groups__clone(struct map_groups *self, + struct map_groups *parent, enum map_type type); +size_t map_groups__fprintf(struct map_groups *self, int verbose, FILE *fp); +size_t map_groups__fprintf_maps(struct map_groups *self, int verbose, FILE *fp); static inline void map_groups__insert(struct map_groups *self, struct map *map) { @@ -125,6 +128,9 @@ static inline struct symbol *map_groups__find_function(struct map_groups *self, return map_groups__find_symbol(self, MAP__FUNCTION, addr, filter); } +int map_groups__fixup_overlappings(struct map_groups *self, struct map *map, + int verbose, FILE *fp); + struct map *map_groups__find_by_name(struct map_groups *self, enum map_type type, const char *name); int __map_groups__create_kernel_maps(struct map_groups *self, @@ -134,5 +140,6 @@ int map_groups__create_kernel_maps(struct map_groups *self, struct map *vmlinux_maps[MAP__NR_TYPES]); struct map *map_groups__new_module(struct map_groups *self, u64 start, const char *filename); +void map_groups__flush(struct map_groups *self); #endif /* __PERF_MAP_H */ diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c index 9bbe27d..1f7ecd4 100644 --- a/tools/perf/util/thread.c +++ b/tools/perf/util/thread.c @@ -38,15 +38,6 @@ failure: return ret; } -void map_groups__init(struct map_groups *self) -{ - int i; - for (i = 0; i < MAP__NR_TYPES; ++i) { - self->maps[i] = RB_ROOT; - INIT_LIST_HEAD(&self->removed_maps[i]); - } -} - static struct thread *thread__new(pid_t pid) { struct thread *self = zalloc(sizeof(*self)); @@ -62,28 +53,6 @@ static struct thread *thread__new(pid_t pid) return self; } -static void map_groups__flush(struct map_groups *self) -{ - int type; - - for (type = 0; type < MAP__NR_TYPES; type++) { - struct rb_root *root = &self->maps[type]; - struct rb_node *next = rb_first(root); - - while (next) { - struct map *pos = rb_entry(next, struct map, rb_node); - next = rb_next(&pos->rb_node); - rb_erase(&pos->rb_node, root); - /* - * We may have references to this map, for - * instance in some hist_entry instances, so - * just move them to a separate list. - */ - list_add_tail(&pos->node, &self->removed_maps[pos->type]); - } - } -} - int thread__set_comm(struct thread *self, const char *comm) { int err; @@ -110,69 +79,10 @@ int thread__comm_len(struct thread *self) return self->comm_len; } -size_t __map_groups__fprintf_maps(struct map_groups *self, - enum map_type type, FILE *fp) -{ - size_t printed = fprintf(fp, "%s:\n", map_type__name[type]); - struct rb_node *nd; - - for (nd = rb_first(&self->maps[type]); nd; nd = rb_next(nd)) { - struct map *pos = rb_entry(nd, struct map, rb_node); - printed += fprintf(fp, "Map:"); - printed += map__fprintf(pos, fp); - if (verbose > 2) { - printed += dso__fprintf(pos->dso, type, fp); - printed += fprintf(fp, "--\n"); - } - } - - return printed; -} - -size_t map_groups__fprintf_maps(struct map_groups *self, FILE *fp) -{ - size_t printed = 0, i; - for (i = 0; i < MAP__NR_TYPES; ++i) - printed += __map_groups__fprintf_maps(self, i, fp); - return printed; -} - -static size_t __map_groups__fprintf_removed_maps(struct map_groups *self, - enum map_type type, FILE *fp) -{ - struct map *pos; - size_t printed = 0; - - list_for_each_entry(pos, &self->removed_maps[type], node) { - printed += fprintf(fp, "Map:"); - printed += map__fprintf(pos, fp); - if (verbose > 1) { - printed += dso__fprintf(pos->dso, type, fp); - printed += fprintf(fp, "--\n"); - } - } - return printed; -} - -static size_t map_groups__fprintf_removed_maps(struct map_groups *self, FILE *fp) -{ - size_t printed = 0, i; - for (i = 0; i < MAP__NR_TYPES; ++i) - printed += __map_groups__fprintf_removed_maps(self, i, fp); - return printed; -} - -static size_t map_groups__fprintf(struct map_groups *self, FILE *fp) -{ - size_t printed = map_groups__fprintf_maps(self, fp); - printed += fprintf(fp, "Removed maps:\n"); - return printed + map_groups__fprintf_removed_maps(self, fp); -} - static size_t thread__fprintf(struct thread *self, FILE *fp) { return fprintf(fp, "Thread %d %s\n", self->pid, self->comm) + - map_groups__fprintf(&self->mg, fp); + map_groups__fprintf(&self->mg, verbose, fp); } struct thread *perf_session__findnew(struct perf_session *self, pid_t pid) @@ -214,87 +124,12 @@ struct thread *perf_session__findnew(struct perf_session *self, pid_t pid) return th; } -static int map_groups__fixup_overlappings(struct map_groups *self, - struct map *map) -{ - struct rb_root *root = &self->maps[map->type]; - struct rb_node *next = rb_first(root); - - while (next) { - struct map *pos = rb_entry(next, struct map, rb_node); - next = rb_next(&pos->rb_node); - - if (!map__overlap(pos, map)) - continue; - - if (verbose >= 2) { - fputs("overlapping maps:\n", stderr); - map__fprintf(map, stderr); - map__fprintf(pos, stderr); - } - - rb_erase(&pos->rb_node, root); - /* - * We may have references to this map, for instance in some - * hist_entry instances, so just move them to a separate - * list. - */ - list_add_tail(&pos->node, &self->removed_maps[map->type]); - /* - * Now check if we need to create new maps for areas not - * overlapped by the new map: - */ - if (map->start > pos->start) { - struct map *before = map__clone(pos); - - if (before == NULL) - return -ENOMEM; - - before->end = map->start - 1; - map_groups__insert(self, before); - if (verbose >= 2) - map__fprintf(before, stderr); - } - - if (map->end < pos->end) { - struct map *after = map__clone(pos); - - if (after == NULL) - return -ENOMEM; - - after->start = map->end + 1; - map_groups__insert(self, after); - if (verbose >= 2) - map__fprintf(after, stderr); - } - } - - return 0; -} - void thread__insert_map(struct thread *self, struct map *map) { - map_groups__fixup_overlappings(&self->mg, map); + map_groups__fixup_overlappings(&self->mg, map, verbose, stderr); map_groups__insert(&self->mg, map); } -/* - * XXX This should not really _copy_ te maps, but refcount them. - */ -static int map_groups__clone(struct map_groups *self, - struct map_groups *parent, enum map_type type) -{ - struct rb_node *nd; - for (nd = rb_first(&parent->maps[type]); nd; nd = rb_next(nd)) { - struct map *map = rb_entry(nd, struct map, rb_node); - struct map *new = map__clone(map); - if (new == NULL) - return -ENOMEM; - map_groups__insert(self, new); - } - return 0; -} - int thread__fork(struct thread *self, struct thread *parent) { int i; -- cgit v0.10.2 From 7e5e1b1404c30db5f6bc3f5203b6c21c1d244f99 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 26 Mar 2010 12:30:40 -0300 Subject: perf symbols: map_groups__find_symbol must return the map too MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Tools need to know from which map in the map_group a symbol was resolved to, so that, for isntance, we can annotate kernel modules symbols by getting its precise name, etc. Also add the _by_name variants for completeness. Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/builtin-kmem.c b/tools/perf/builtin-kmem.c index 924a951..32edb6a 100644 --- a/tools/perf/builtin-kmem.c +++ b/tools/perf/builtin-kmem.c @@ -369,7 +369,8 @@ static void __print_result(struct rb_root *root, struct perf_session *session, if (is_caller) { addr = data->call_site; if (!raw_ip) - sym = map_groups__find_function(&session->kmaps, addr, NULL); + sym = map_groups__find_function(&session->kmaps, + addr, NULL, NULL); } else addr = data->ptr; diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index e21f980..37913b2 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -268,12 +268,38 @@ void map_groups__flush(struct map_groups *self) struct symbol *map_groups__find_symbol(struct map_groups *self, enum map_type type, u64 addr, + struct map **mapp, symbol_filter_t filter) { struct map *map = map_groups__find(self, type, addr); - if (map != NULL) + if (map != NULL) { + if (mapp != NULL) + *mapp = map; return map__find_symbol(map, map->map_ip(map, addr), filter); + } + + return NULL; +} + +struct symbol *map_groups__find_symbol_by_name(struct map_groups *self, + enum map_type type, + const char *name, + struct map **mapp, + symbol_filter_t filter) +{ + struct rb_node *nd; + + for (nd = rb_first(&self->maps[type]); nd; nd = rb_next(nd)) { + struct map *pos = rb_entry(nd, struct map, rb_node); + struct symbol *sym = map__find_symbol_by_name(pos, name, filter); + + if (sym == NULL) + continue; + if (mapp != NULL) + *mapp = pos; + return sym; + } return NULL; } diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h index 4e7a11d..2031278 100644 --- a/tools/perf/util/map.h +++ b/tools/perf/util/map.h @@ -119,13 +119,28 @@ static inline struct map *map_groups__find(struct map_groups *self, struct symbol *map_groups__find_symbol(struct map_groups *self, enum map_type type, u64 addr, + struct map **mapp, symbol_filter_t filter); -static inline struct symbol *map_groups__find_function(struct map_groups *self, - u64 addr, - symbol_filter_t filter) +struct symbol *map_groups__find_symbol_by_name(struct map_groups *self, + enum map_type type, + const char *name, + struct map **mapp, + symbol_filter_t filter); + +static inline +struct symbol *map_groups__find_function(struct map_groups *self, u64 addr, + struct map **mapp, symbol_filter_t filter) +{ + return map_groups__find_symbol(self, MAP__FUNCTION, addr, mapp, filter); +} + +static inline +struct symbol *map_groups__find_function_by_name(struct map_groups *self, + const char *name, struct map **mapp, + symbol_filter_t filter) { - return map_groups__find_symbol(self, MAP__FUNCTION, addr, filter); + return map_groups__find_symbol_by_name(self, MAP__FUNCTION, name, mapp, filter); } int map_groups__fixup_overlappings(struct map_groups *self, struct map *map, -- cgit v0.10.2 From 5f4d3f8816461300ce54505c9117bf85b3044aa0 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 26 Mar 2010 21:16:22 -0300 Subject: perf report: Add progress bars MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit For when we are processing the events and inserting the entries in the browser. Experimentation here: naming "ui_something" we may be treading into creating a TUI/GUI set of routines that can then be implemented in terms of multiple backends. Also the time it takes for adding things to the "browser" takes, visually (I guess I should do some profiling here ;-) ), more time than for processing the events... That means we probably need to create a custom hist_entry browser, so that we reuse the structures we have in place instead of duplicating them in newt. But progress was made and at least we can see something while long files are being loaded, that must be one of UI 101 bullet points :-) Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 6ab1698..3819185 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -303,13 +303,14 @@ static int __cmd_report(void) next = rb_first(&session->stats_by_id); while (next) { struct event_stat_id *stats; + u64 nr_hists; stats = rb_entry(next, struct event_stat_id, rb_node); perf_session__collapse_resort(&stats->hists); - perf_session__output_resort(&stats->hists, stats->stats.total); - + nr_hists = perf_session__output_resort(&stats->hists, + stats->stats.total); if (use_browser) - perf_session__browse_hists(&stats->hists, + perf_session__browse_hists(&stats->hists, nr_hists, stats->stats.total, help); else { if (rb_first(&session->stats_by_id) == diff --git a/tools/perf/util/debug.h b/tools/perf/util/debug.h index 0172edf..5cb0a1b 100644 --- a/tools/perf/util/debug.h +++ b/tools/perf/util/debug.h @@ -10,13 +10,29 @@ extern int dump_trace; int dump_printf(const char *fmt, ...) __attribute__((format(printf, 1, 2))); void trace_event(event_t *event); +struct ui_progress; + #ifdef NO_NEWT_SUPPORT static inline int browser__show_help(const char *format __used, va_list ap __used) { return 0; } + +static inline struct ui_progress *ui_progress__new(const char *title __used, + u64 total __used) +{ + return (struct ui_progress *)1; +} + +static inline void ui_progress__update(struct ui_progress *self __used, + u64 curr __used) {} + +static inline void ui_progress__delete(struct ui_progress *self __used) {} #else int browser__show_help(const char *format, va_list ap); +struct ui_progress *ui_progress__new(const char *title, u64 total); +void ui_progress__update(struct ui_progress *self, u64 curr); +void ui_progress__delete(struct ui_progress *self); #endif #endif /* __PERF_DEBUG_H */ diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index de31901..a46d093 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -185,12 +185,13 @@ static void perf_session__insert_output_hist_entry(struct rb_root *root, rb_insert_color(&he->rb_node, root); } -void perf_session__output_resort(struct rb_root *hists, u64 total_samples) +u64 perf_session__output_resort(struct rb_root *hists, u64 total_samples) { struct rb_root tmp; struct rb_node *next; struct hist_entry *n; u64 min_callchain_hits; + u64 nr_hists = 0; min_callchain_hits = total_samples * (callchain_param.min_percent / 100); @@ -205,9 +206,11 @@ void perf_session__output_resort(struct rb_root *hists, u64 total_samples) rb_erase(&n->rb_node, hists); perf_session__insert_output_hist_entry(&tmp, n, min_callchain_hits); + ++nr_hists; } *hists = tmp; + return nr_hists; } static size_t callchain__fprintf_left_margin(FILE *fp, int left_margin) diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h index fe366ce..da6a8c1 100644 --- a/tools/perf/util/hist.h +++ b/tools/perf/util/hist.h @@ -25,7 +25,7 @@ size_t hist_entry__fprintf(struct hist_entry *self, u64 session_total); void hist_entry__free(struct hist_entry *); -void perf_session__output_resort(struct rb_root *hists, u64 total_samples); +u64 perf_session__output_resort(struct rb_root *hists, u64 total_samples); void perf_session__collapse_resort(struct rb_root *hists); size_t perf_session__fprintf_hists(struct rb_root *hists, struct perf_session *pair, diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index e99bcc8..b0210ae 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -12,6 +12,72 @@ #include "sort.h" #include "symbol.h" +struct ui_progress { + newtComponent form, scale; +}; + +struct ui_progress *ui_progress__new(const char *title, u64 total) +{ + struct ui_progress *self = malloc(sizeof(*self)); + + if (self != NULL) { + int cols; + newtGetScreenSize(&cols, NULL); + cols -= 4; + newtCenteredWindow(cols, 1, title); + self->form = newtForm(NULL, NULL, 0); + if (self->form == NULL) + goto out_free_self; + self->scale = newtScale(0, 0, cols, total); + if (self->scale == NULL) + goto out_free_form; + newtFormAddComponents(self->form, self->scale, NULL); + newtRefresh(); + } + + return self; + +out_free_form: + newtFormDestroy(self->form); +out_free_self: + free(self); + return NULL; +} + +void ui_progress__update(struct ui_progress *self, u64 curr) +{ + newtScaleSet(self->scale, curr); + newtRefresh(); +} + +void ui_progress__delete(struct ui_progress *self) +{ + newtFormDestroy(self->form); + newtPopWindow(); + free(self); +} + +static char browser__last_msg[1024]; + +int browser__show_help(const char *format, va_list ap) +{ + int ret; + static int backlog; + + ret = vsnprintf(browser__last_msg + backlog, + sizeof(browser__last_msg) - backlog, format, ap); + backlog += ret; + + if (browser__last_msg[backlog - 1] == '\n') { + newtPopHelpLine(); + newtPushHelpLine(browser__last_msg); + newtRefresh(); + backlog = 0; + } + + return ret; +} + static void newt_form__set_exit_keys(newtComponent self) { newtFormAddHotKey(self, NEWT_KEY_ESCAPE); @@ -364,8 +430,8 @@ static void perf_session__selection(newtComponent self, void *data) *symbol_ptr = newt__symbol_tree_get_current(self); } -void perf_session__browse_hists(struct rb_root *hists, u64 session_total, - const char *helpline) +int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, + u64 session_total, const char *helpline) { struct sort_entry *se; struct rb_node *nd; @@ -378,6 +444,12 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, newtComponent form, tree; struct newtExitStruct es; const struct map_symbol *selection; + u64 curr_hist = 0; + struct ui_progress *progress; + + progress = ui_progress__new("Adding entries to the browser...", nr_hists); + if (progress == NULL) + return -1; snprintf(str, sizeof(str), "Samples: %Ld", session_total); newtDrawRootText(0, 0, str); @@ -419,8 +491,13 @@ void perf_session__browse_hists(struct rb_root *hists, u64 session_total, max_len = len; if (symbol_conf.use_callchain) hist_entry__append_callchain_browser(h, tree, session_total, idx++); + ++curr_hist; + if (curr_hist % 5) + ui_progress__update(progress, curr_hist); } + ui_progress__delete(progress); + if (max_len > cols) max_len = cols - 3; @@ -480,27 +557,7 @@ do_annotate: newtFormDestroy(form); newtPopWindow(); -} - -static char browser__last_msg[1024]; - -int browser__show_help(const char *format, va_list ap) -{ - int ret; - static int backlog; - - ret = vsnprintf(browser__last_msg + backlog, - sizeof(browser__last_msg) - backlog, format, ap); - backlog += ret; - - if (browser__last_msg[backlog - 1] == '\n') { - newtPopHelpLine(); - newtPushHelpLine(browser__last_msg); - newtRefresh(); - backlog = 0; - } - - return ret; + return 0; } void setup_browser(void) diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index 76b4ac6..32765cd 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -397,6 +397,10 @@ int __perf_session__process_events(struct perf_session *self, event_t *event; uint32_t size; char *buf; + struct ui_progress *progress = ui_progress__new("Processing events...", + self->size); + if (progress == NULL) + return -1; perf_event_ops__fill_defaults(ops); @@ -425,6 +429,7 @@ remap: more: event = (event_t *)(buf + head); + ui_progress__update(progress, offset); if (self->header.needs_swap) perf_event_header__bswap(&event->header); @@ -475,6 +480,7 @@ more: done: err = 0; out_err: + ui_progress__delete(progress); return err; } diff --git a/tools/perf/util/session.h b/tools/perf/util/session.h index 631f815..6a15dae 100644 --- a/tools/perf/util/session.h +++ b/tools/perf/util/session.h @@ -88,11 +88,15 @@ static inline struct map * } #ifdef NO_NEWT_SUPPORT -static inline void perf_session__browse_hists(struct rb_root *hists __used, +static inline int perf_session__browse_hists(struct rb_root *hists __used, + u64 nr_hists __used, u64 session_total __used, - const char *helpline __used) {} + const char *helpline __used) +{ + return 0; +} #else -void perf_session__browse_hists(struct rb_root *hists, u64 session_total, - const char *helpline); +int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, + u64 session_total, const char *helpline); #endif #endif /* __PERF_SESSION_H */ -- cgit v0.10.2 From 8b2c551f9635bf1c5c2d38de300137998915478f Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 27 Mar 2010 11:43:36 -0300 Subject: perf tools: Use -o $(BITBUCKET) in one more case MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit As described in 1703f2c some gcc versions has issues using /dev/null, so use the mechanism used elsewhere. Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/Makefile b/tools/perf/Makefile index 6903645..120736c 100644 --- a/tools/perf/Makefile +++ b/tools/perf/Makefile @@ -542,7 +542,7 @@ PYTHON_EMBED_LDOPTS = `python-config --ldflags 2>/dev/null` PYTHON_EMBED_CCOPTS = `python-config --cflags 2>/dev/null` endif -ifneq ($(shell sh -c "(echo '\#include '; echo 'int main(void) { Py_Initialize(); return 0; }') | $(CC) -x c - $(PYTHON_EMBED_CCOPTS) -o /dev/null $(PYTHON_EMBED_LDOPTS) > /dev/null 2>&1 && echo y"), y) +ifneq ($(shell sh -c "(echo '\#include '; echo 'int main(void) { Py_Initialize(); return 0; }') | $(CC) -x c - $(PYTHON_EMBED_CCOPTS) -o $(BITBUCKET) $(PYTHON_EMBED_LDOPTS) > /dev/null 2>&1 && echo y"), y) BASIC_CFLAGS += -DNO_LIBPYTHON else ALL_LDFLAGS += $(PYTHON_EMBED_LDOPTS) -- cgit v0.10.2 From c29ede615fd35a640e771fbbb1778e915fac43a7 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 27 Mar 2010 14:30:45 -0300 Subject: perf tools: Allow specifying O= to build files in a separate directory MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Avoiding polluting the source tree with build files. Reported-by: Steven Rostedt Cc: Steven Rostedt Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/Makefile b/tools/perf/Makefile index 120736c..93029c0 100644 --- a/tools/perf/Makefile +++ b/tools/perf/Makefile @@ -1,3 +1,7 @@ +ifeq ("$(origin O)", "command line") + OUTPUT := $(O)/ +endif + # The default target of this Makefile is... all:: @@ -153,9 +157,13 @@ all:: # # Define NO_DWARF if you do not want debug-info analysis feature at all. -PERF-VERSION-FILE: .FORCE-PERF-VERSION-FILE - @$(SHELL_PATH) util/PERF-VERSION-GEN --include PERF-VERSION-FILE +$(shell sh -c 'mkdir -p $(OUTPUT)scripts/python/Perf-Trace-Util/' 2> /dev/null) +$(shell sh -c 'mkdir -p $(OUTPUT)util/scripting-engines/' 2> /dev/null) +$(shell sh -c 'mkdir $(OUTPUT)bench' 2> /dev/null) + +$(OUTPUT)PERF-VERSION-FILE: .FORCE-PERF-VERSION-FILE + @$(SHELL_PATH) util/PERF-VERSION-GEN $(OUTPUT) +-include $(OUTPUT)PERF-VERSION-FILE uname_S := $(shell sh -c 'uname -s 2>/dev/null || echo not') uname_M := $(shell sh -c 'uname -m 2>/dev/null || echo not') @@ -310,7 +318,7 @@ PROGRAMS += $(EXTRA_PROGRAMS) # # Single 'perf' binary right now: # -PROGRAMS += perf +PROGRAMS += $(OUTPUT)perf # List built-in command $C whose implementation cmd_$C() is not in # builtin-$C.o but is linked in as part of some other command. @@ -320,7 +328,7 @@ PROGRAMS += perf ALL_PROGRAMS = $(PROGRAMS) $(SCRIPTS) # what 'all' will build but not install in perfexecdir -OTHER_PROGRAMS = perf$X +OTHER_PROGRAMS = $(OUTPUT)perf$X # Set paths to tools early so that they can be used for version tests. ifndef SHELL_PATH @@ -332,7 +340,7 @@ endif export PERL_PATH -LIB_FILE=libperf.a +LIB_FILE=$(OUTPUT)libperf.a LIB_H += ../../include/linux/perf_event.h LIB_H += ../../include/linux/rbtree.h @@ -393,77 +401,77 @@ LIB_H += util/probe-finder.h LIB_H += util/probe-event.h LIB_H += util/cpumap.h -LIB_OBJS += util/abspath.o -LIB_OBJS += util/alias.o -LIB_OBJS += util/build-id.o -LIB_OBJS += util/config.o -LIB_OBJS += util/ctype.o -LIB_OBJS += util/debugfs.o -LIB_OBJS += util/environment.o -LIB_OBJS += util/event.o -LIB_OBJS += util/exec_cmd.o -LIB_OBJS += util/help.o -LIB_OBJS += util/levenshtein.o -LIB_OBJS += util/parse-options.o -LIB_OBJS += util/parse-events.o -LIB_OBJS += util/path.o -LIB_OBJS += util/rbtree.o -LIB_OBJS += util/bitmap.o -LIB_OBJS += util/hweight.o -LIB_OBJS += util/find_next_bit.o -LIB_OBJS += util/run-command.o -LIB_OBJS += util/quote.o -LIB_OBJS += util/strbuf.o -LIB_OBJS += util/string.o -LIB_OBJS += util/strlist.o -LIB_OBJS += util/usage.o -LIB_OBJS += util/wrapper.o -LIB_OBJS += util/sigchain.o -LIB_OBJS += util/symbol.o -LIB_OBJS += util/color.o -LIB_OBJS += util/pager.o -LIB_OBJS += util/header.o -LIB_OBJS += util/callchain.o -LIB_OBJS += util/values.o -LIB_OBJS += util/debug.o -LIB_OBJS += util/map.o -LIB_OBJS += util/session.o -LIB_OBJS += util/thread.o -LIB_OBJS += util/trace-event-parse.o -LIB_OBJS += util/trace-event-read.o -LIB_OBJS += util/trace-event-info.o -LIB_OBJS += util/trace-event-scripting.o -LIB_OBJS += util/svghelper.o -LIB_OBJS += util/sort.o -LIB_OBJS += util/hist.o -LIB_OBJS += util/probe-event.o -LIB_OBJS += util/util.o -LIB_OBJS += util/cpumap.o - -BUILTIN_OBJS += builtin-annotate.o - -BUILTIN_OBJS += builtin-bench.o +LIB_OBJS += $(OUTPUT)util/abspath.o +LIB_OBJS += $(OUTPUT)util/alias.o +LIB_OBJS += $(OUTPUT)util/build-id.o +LIB_OBJS += $(OUTPUT)util/config.o +LIB_OBJS += $(OUTPUT)util/ctype.o +LIB_OBJS += $(OUTPUT)util/debugfs.o +LIB_OBJS += $(OUTPUT)util/environment.o +LIB_OBJS += $(OUTPUT)util/event.o +LIB_OBJS += $(OUTPUT)util/exec_cmd.o +LIB_OBJS += $(OUTPUT)util/help.o +LIB_OBJS += $(OUTPUT)util/levenshtein.o +LIB_OBJS += $(OUTPUT)util/parse-options.o +LIB_OBJS += $(OUTPUT)util/parse-events.o +LIB_OBJS += $(OUTPUT)util/path.o +LIB_OBJS += $(OUTPUT)util/rbtree.o +LIB_OBJS += $(OUTPUT)util/bitmap.o +LIB_OBJS += $(OUTPUT)util/hweight.o +LIB_OBJS += $(OUTPUT)util/find_next_bit.o +LIB_OBJS += $(OUTPUT)util/run-command.o +LIB_OBJS += $(OUTPUT)util/quote.o +LIB_OBJS += $(OUTPUT)util/strbuf.o +LIB_OBJS += $(OUTPUT)util/string.o +LIB_OBJS += $(OUTPUT)util/strlist.o +LIB_OBJS += $(OUTPUT)util/usage.o +LIB_OBJS += $(OUTPUT)util/wrapper.o +LIB_OBJS += $(OUTPUT)util/sigchain.o +LIB_OBJS += $(OUTPUT)util/symbol.o +LIB_OBJS += $(OUTPUT)util/color.o +LIB_OBJS += $(OUTPUT)util/pager.o +LIB_OBJS += $(OUTPUT)util/header.o +LIB_OBJS += $(OUTPUT)util/callchain.o +LIB_OBJS += $(OUTPUT)util/values.o +LIB_OBJS += $(OUTPUT)util/debug.o +LIB_OBJS += $(OUTPUT)util/map.o +LIB_OBJS += $(OUTPUT)util/session.o +LIB_OBJS += $(OUTPUT)util/thread.o +LIB_OBJS += $(OUTPUT)util/trace-event-parse.o +LIB_OBJS += $(OUTPUT)util/trace-event-read.o +LIB_OBJS += $(OUTPUT)util/trace-event-info.o +LIB_OBJS += $(OUTPUT)util/trace-event-scripting.o +LIB_OBJS += $(OUTPUT)util/svghelper.o +LIB_OBJS += $(OUTPUT)util/sort.o +LIB_OBJS += $(OUTPUT)util/hist.o +LIB_OBJS += $(OUTPUT)util/probe-event.o +LIB_OBJS += $(OUTPUT)util/util.o +LIB_OBJS += $(OUTPUT)util/cpumap.o + +BUILTIN_OBJS += $(OUTPUT)builtin-annotate.o + +BUILTIN_OBJS += $(OUTPUT)builtin-bench.o # Benchmark modules -BUILTIN_OBJS += bench/sched-messaging.o -BUILTIN_OBJS += bench/sched-pipe.o -BUILTIN_OBJS += bench/mem-memcpy.o - -BUILTIN_OBJS += builtin-diff.o -BUILTIN_OBJS += builtin-help.o -BUILTIN_OBJS += builtin-sched.o -BUILTIN_OBJS += builtin-buildid-list.o -BUILTIN_OBJS += builtin-buildid-cache.o -BUILTIN_OBJS += builtin-list.o -BUILTIN_OBJS += builtin-record.o -BUILTIN_OBJS += builtin-report.o -BUILTIN_OBJS += builtin-stat.o -BUILTIN_OBJS += builtin-timechart.o -BUILTIN_OBJS += builtin-top.o -BUILTIN_OBJS += builtin-trace.o -BUILTIN_OBJS += builtin-probe.o -BUILTIN_OBJS += builtin-kmem.o -BUILTIN_OBJS += builtin-lock.o +BUILTIN_OBJS += $(OUTPUT)bench/sched-messaging.o +BUILTIN_OBJS += $(OUTPUT)bench/sched-pipe.o +BUILTIN_OBJS += $(OUTPUT)bench/mem-memcpy.o + +BUILTIN_OBJS += $(OUTPUT)builtin-diff.o +BUILTIN_OBJS += $(OUTPUT)builtin-help.o +BUILTIN_OBJS += $(OUTPUT)builtin-sched.o +BUILTIN_OBJS += $(OUTPUT)builtin-buildid-list.o +BUILTIN_OBJS += $(OUTPUT)builtin-buildid-cache.o +BUILTIN_OBJS += $(OUTPUT)builtin-list.o +BUILTIN_OBJS += $(OUTPUT)builtin-record.o +BUILTIN_OBJS += $(OUTPUT)builtin-report.o +BUILTIN_OBJS += $(OUTPUT)builtin-stat.o +BUILTIN_OBJS += $(OUTPUT)builtin-timechart.o +BUILTIN_OBJS += $(OUTPUT)builtin-top.o +BUILTIN_OBJS += $(OUTPUT)builtin-trace.o +BUILTIN_OBJS += $(OUTPUT)builtin-probe.o +BUILTIN_OBJS += $(OUTPUT)builtin-kmem.o +BUILTIN_OBJS += $(OUTPUT)builtin-lock.o PERFLIBS = $(LIB_FILE) @@ -494,6 +502,10 @@ ifeq ($(uname_S),Darwin) PTHREAD_LIBS = endif +ifneq ($(OUTPUT),) + BASIC_CFLAGS += -I$(OUTPUT) +endif + ifeq ($(shell sh -c "(echo '\#include '; echo 'int main(void) { Elf * elf = elf_begin(0, ELF_C_READ, 0); return (long)elf; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) ifneq ($(shell sh -c "(echo '\#include '; echo 'int main(void) { const char * version = gnu_get_libc_version(); return (long)version; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -o $(BITBUCKET) $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y) msg := $(error No gnu/libc-version.h found, please install glibc-dev[el]/glibc-static); @@ -512,7 +524,7 @@ else ifndef NO_DWARF BASIC_CFLAGS += -I/usr/include/elfutils -DDWARF_SUPPORT EXTLIBS += -lelf -ldw - LIB_OBJS += util/probe-finder.o + LIB_OBJS += $(OUTPUT)util/probe-finder.o endif endif @@ -521,7 +533,7 @@ ifneq ($(shell sh -c "(echo '\#include '; echo 'int main(void) { newtIni BASIC_CFLAGS += -DNO_NEWT_SUPPORT else EXTLIBS += -lnewt - LIB_OBJS += util/newt.o + LIB_OBJS += $(OUTPUT)util/newt.o endif ifndef NO_LIBPERL @@ -533,8 +545,8 @@ ifneq ($(shell sh -c "(echo '\#include '; echo '\#include '; e BASIC_CFLAGS += -DNO_LIBPERL else ALL_LDFLAGS += $(PERL_EMBED_LDOPTS) - LIB_OBJS += util/scripting-engines/trace-event-perl.o - LIB_OBJS += scripts/perl/Perf-Trace-Util/Context.o + LIB_OBJS += $(OUTPUT)util/scripting-engines/trace-event-perl.o + LIB_OBJS += $(OUTPUT)scripts/perl/Perf-Trace-Util/Context.o endif ifndef NO_LIBPYTHON @@ -546,8 +558,8 @@ ifneq ($(shell sh -c "(echo '\#include '; echo 'int main(void) { Py_In BASIC_CFLAGS += -DNO_LIBPYTHON else ALL_LDFLAGS += $(PYTHON_EMBED_LDOPTS) - LIB_OBJS += util/scripting-engines/trace-event-python.o - LIB_OBJS += scripts/python/Perf-Trace-Util/Context.o + LIB_OBJS += $(OUTPUT)util/scripting-engines/trace-event-python.o + LIB_OBJS += $(OUTPUT)scripts/python/Perf-Trace-Util/Context.o endif ifdef NO_DEMANGLE @@ -618,53 +630,53 @@ ifdef NO_C99_FORMAT endif ifdef SNPRINTF_RETURNS_BOGUS COMPAT_CFLAGS += -DSNPRINTF_RETURNS_BOGUS - COMPAT_OBJS += compat/snprintf.o + COMPAT_OBJS += $(OUTPUT)compat/snprintf.o endif ifdef FREAD_READS_DIRECTORIES COMPAT_CFLAGS += -DFREAD_READS_DIRECTORIES - COMPAT_OBJS += compat/fopen.o + COMPAT_OBJS += $(OUTPUT)compat/fopen.o endif ifdef NO_SYMLINK_HEAD BASIC_CFLAGS += -DNO_SYMLINK_HEAD endif ifdef NO_STRCASESTR COMPAT_CFLAGS += -DNO_STRCASESTR - COMPAT_OBJS += compat/strcasestr.o + COMPAT_OBJS += $(OUTPUT)compat/strcasestr.o endif ifdef NO_STRTOUMAX COMPAT_CFLAGS += -DNO_STRTOUMAX - COMPAT_OBJS += compat/strtoumax.o + COMPAT_OBJS += $(OUTPUT)compat/strtoumax.o endif ifdef NO_STRTOULL COMPAT_CFLAGS += -DNO_STRTOULL endif ifdef NO_SETENV COMPAT_CFLAGS += -DNO_SETENV - COMPAT_OBJS += compat/setenv.o + COMPAT_OBJS += $(OUTPUT)compat/setenv.o endif ifdef NO_MKDTEMP COMPAT_CFLAGS += -DNO_MKDTEMP - COMPAT_OBJS += compat/mkdtemp.o + COMPAT_OBJS += $(OUTPUT)compat/mkdtemp.o endif ifdef NO_UNSETENV COMPAT_CFLAGS += -DNO_UNSETENV - COMPAT_OBJS += compat/unsetenv.o + COMPAT_OBJS += $(OUTPUT)compat/unsetenv.o endif ifdef NO_SYS_SELECT_H BASIC_CFLAGS += -DNO_SYS_SELECT_H endif ifdef NO_MMAP COMPAT_CFLAGS += -DNO_MMAP - COMPAT_OBJS += compat/mmap.o + COMPAT_OBJS += $(OUTPUT)compat/mmap.o else ifdef USE_WIN32_MMAP COMPAT_CFLAGS += -DUSE_WIN32_MMAP - COMPAT_OBJS += compat/win32mmap.o + COMPAT_OBJS += $(OUTPUT)compat/win32mmap.o endif endif ifdef NO_PREAD COMPAT_CFLAGS += -DNO_PREAD - COMPAT_OBJS += compat/pread.o + COMPAT_OBJS += $(OUTPUT)compat/pread.o endif ifdef NO_FAST_WORKING_DIRECTORY BASIC_CFLAGS += -DNO_FAST_WORKING_DIRECTORY @@ -686,10 +698,10 @@ else endif endif ifdef NO_INET_NTOP - LIB_OBJS += compat/inet_ntop.o + LIB_OBJS += $(OUTPUT)compat/inet_ntop.o endif ifdef NO_INET_PTON - LIB_OBJS += compat/inet_pton.o + LIB_OBJS += $(OUTPUT)compat/inet_pton.o endif ifdef NO_ICONV @@ -706,15 +718,15 @@ endif ifdef PPC_SHA1 SHA1_HEADER = "ppc/sha1.h" - LIB_OBJS += ppc/sha1.o ppc/sha1ppc.o + LIB_OBJS += $(OUTPUT)ppc/sha1.o ppc/sha1ppc.o else ifdef ARM_SHA1 SHA1_HEADER = "arm/sha1.h" - LIB_OBJS += arm/sha1.o arm/sha1_arm.o + LIB_OBJS += $(OUTPUT)arm/sha1.o $(OUTPUT)arm/sha1_arm.o else ifdef MOZILLA_SHA1 SHA1_HEADER = "mozilla-sha1/sha1.h" - LIB_OBJS += mozilla-sha1/sha1.o + LIB_OBJS += $(OUTPUT)mozilla-sha1/sha1.o else SHA1_HEADER = EXTLIBS += $(LIB_4_CRYPTO) @@ -726,15 +738,15 @@ ifdef NO_PERL_MAKEMAKER endif ifdef NO_HSTRERROR COMPAT_CFLAGS += -DNO_HSTRERROR - COMPAT_OBJS += compat/hstrerror.o + COMPAT_OBJS += $(OUTPUT)compat/hstrerror.o endif ifdef NO_MEMMEM COMPAT_CFLAGS += -DNO_MEMMEM - COMPAT_OBJS += compat/memmem.o + COMPAT_OBJS += $(OUTPUT)compat/memmem.o endif ifdef INTERNAL_QSORT COMPAT_CFLAGS += -DINTERNAL_QSORT - COMPAT_OBJS += compat/qsort.o + COMPAT_OBJS += $(OUTPUT)compat/qsort.o endif ifdef RUNTIME_PREFIX COMPAT_CFLAGS += -DRUNTIME_PREFIX @@ -814,7 +826,7 @@ export TAR INSTALL DESTDIR SHELL_PATH SHELL = $(SHELL_PATH) -all:: .perf.dev.null shell_compatibility_test $(ALL_PROGRAMS) $(BUILT_INS) $(OTHER_PROGRAMS) PERF-BUILD-OPTIONS +all:: .perf.dev.null shell_compatibility_test $(ALL_PROGRAMS) $(BUILT_INS) $(OTHER_PROGRAMS) $(OUTPUT)PERF-BUILD-OPTIONS ifneq (,$X) $(foreach p,$(patsubst %$X,%,$(filter %$X,$(ALL_PROGRAMS) $(BUILT_INS) perf$X)), test '$p' -ef '$p$X' || $(RM) '$p';) endif @@ -826,39 +838,39 @@ please_set_SHELL_PATH_to_a_more_modern_shell: shell_compatibility_test: please_set_SHELL_PATH_to_a_more_modern_shell -strip: $(PROGRAMS) perf$X - $(STRIP) $(STRIP_OPTS) $(PROGRAMS) perf$X +strip: $(PROGRAMS) $(OUTPUT)perf$X + $(STRIP) $(STRIP_OPTS) $(PROGRAMS) $(OUTPUT)perf$X -perf.o: perf.c common-cmds.h PERF-CFLAGS +$(OUTPUT)perf.o: perf.c $(OUTPUT)common-cmds.h $(OUTPUT)PERF-CFLAGS $(QUIET_CC)$(CC) -DPERF_VERSION='"$(PERF_VERSION)"' \ '-DPERF_HTML_PATH="$(htmldir_SQ)"' \ - $(ALL_CFLAGS) -c $(filter %.c,$^) + $(ALL_CFLAGS) -c $(filter %.c,$^) -o $@ -perf$X: perf.o $(BUILTIN_OBJS) $(PERFLIBS) - $(QUIET_LINK)$(CC) $(ALL_CFLAGS) -o $@ perf.o \ +$(OUTPUT)perf$X: $(OUTPUT)perf.o $(BUILTIN_OBJS) $(PERFLIBS) + $(QUIET_LINK)$(CC) $(ALL_CFLAGS) -o $@ $(OUTPUT)perf.o \ $(BUILTIN_OBJS) $(ALL_LDFLAGS) $(LIBS) -builtin-help.o: builtin-help.c common-cmds.h PERF-CFLAGS - $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) \ +$(OUTPUT)builtin-help.o: builtin-help.c $(OUTPUT)common-cmds.h $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) \ '-DPERF_HTML_PATH="$(htmldir_SQ)"' \ '-DPERF_MAN_PATH="$(mandir_SQ)"' \ '-DPERF_INFO_PATH="$(infodir_SQ)"' $< -builtin-timechart.o: builtin-timechart.c common-cmds.h PERF-CFLAGS - $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) \ +$(OUTPUT)builtin-timechart.o: builtin-timechart.c $(OUTPUT)common-cmds.h $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) \ '-DPERF_HTML_PATH="$(htmldir_SQ)"' \ '-DPERF_MAN_PATH="$(mandir_SQ)"' \ '-DPERF_INFO_PATH="$(infodir_SQ)"' $< -$(BUILT_INS): perf$X +$(BUILT_INS): $(OUTPUT)perf$X $(QUIET_BUILT_IN)$(RM) $@ && \ ln perf$X $@ 2>/dev/null || \ ln -s perf$X $@ 2>/dev/null || \ cp perf$X $@ -common-cmds.h: util/generate-cmdlist.sh command-list.txt +$(OUTPUT)common-cmds.h: util/generate-cmdlist.sh command-list.txt -common-cmds.h: $(wildcard Documentation/perf-*.txt) +$(OUTPUT)common-cmds.h: $(wildcard Documentation/perf-*.txt) $(QUIET_GEN). util/generate-cmdlist.sh > $@+ && mv $@+ $@ $(patsubst %.sh,%,$(SCRIPT_SH)) : % : %.sh @@ -870,7 +882,7 @@ $(patsubst %.sh,%,$(SCRIPT_SH)) : % : %.sh -e 's/@@NO_CURL@@/$(NO_CURL)/g' \ $@.sh >$@+ && \ chmod +x $@+ && \ - mv $@+ $@ + mv $@+ $(OUTPUT)$@ configure: configure.ac $(QUIET_GEN)$(RM) $@ $<+ && \ @@ -880,60 +892,60 @@ configure: configure.ac $(RM) $<+ # These can record PERF_VERSION -perf.o perf.spec \ +$(OUTPUT)perf.o perf.spec \ $(patsubst %.sh,%,$(SCRIPT_SH)) \ $(patsubst %.perl,%,$(SCRIPT_PERL)) \ - : PERF-VERSION-FILE + : $(OUTPUT)PERF-VERSION-FILE -%.o: %.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) $< -%.s: %.c PERF-CFLAGS +$(OUTPUT)%.o: %.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) $< +$(OUTPUT)%.s: %.c $(OUTPUT)PERF-CFLAGS $(QUIET_CC)$(CC) -S $(ALL_CFLAGS) $< -%.o: %.S - $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) $< +$(OUTPUT)%.o: %.S + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) $< -util/exec_cmd.o: util/exec_cmd.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) \ +$(OUTPUT)util/exec_cmd.o: util/exec_cmd.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) \ '-DPERF_EXEC_PATH="$(perfexecdir_SQ)"' \ '-DBINDIR="$(bindir_relative_SQ)"' \ '-DPREFIX="$(prefix_SQ)"' \ $< -builtin-init-db.o: builtin-init-db.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) -DDEFAULT_PERF_TEMPLATE_DIR='"$(template_dir_SQ)"' $< +$(OUTPUT)builtin-init-db.o: builtin-init-db.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DDEFAULT_PERF_TEMPLATE_DIR='"$(template_dir_SQ)"' $< -util/config.o: util/config.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o $*.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< +$(OUTPUT)util/config.o: util/config.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< -util/rbtree.o: ../../lib/rbtree.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o util/rbtree.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< +$(OUTPUT)util/rbtree.o: ../../lib/rbtree.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< # some perf warning policies can't fit to lib/bitmap.c, eg: it warns about variable shadowing # from that comes from kernel headers wrapping. KBITMAP_FLAGS=`echo $(ALL_CFLAGS) | sed s/-Wshadow// | sed s/-Wswitch-default// | sed s/-Wextra//` -util/bitmap.o: ../../lib/bitmap.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o util/bitmap.o -c $(KBITMAP_FLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< +$(OUTPUT)util/bitmap.o: ../../lib/bitmap.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(KBITMAP_FLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< -util/hweight.o: ../../lib/hweight.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o util/hweight.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< +$(OUTPUT)util/hweight.o: ../../lib/hweight.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< -util/find_next_bit.o: ../../lib/find_next_bit.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o util/find_next_bit.o -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< +$(OUTPUT)util/find_next_bit.o: ../../lib/find_next_bit.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< -util/scripting-engines/trace-event-perl.o: util/scripting-engines/trace-event-perl.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o util/scripting-engines/trace-event-perl.o -c $(ALL_CFLAGS) $(PERL_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-shadow $< +$(OUTPUT)util/scripting-engines/trace-event-perl.o: util/scripting-engines/trace-event-perl.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) $(PERL_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-shadow $< -scripts/perl/Perf-Trace-Util/Context.o: scripts/perl/Perf-Trace-Util/Context.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o scripts/perl/Perf-Trace-Util/Context.o -c $(ALL_CFLAGS) $(PERL_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-nested-externs $< +$(OUTPUT)scripts/perl/Perf-Trace-Util/Context.o: scripts/perl/Perf-Trace-Util/Context.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) $(PERL_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-nested-externs $< -util/scripting-engines/trace-event-python.o: util/scripting-engines/trace-event-python.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o util/scripting-engines/trace-event-python.o -c $(ALL_CFLAGS) $(PYTHON_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-shadow $< +$(OUTPUT)util/scripting-engines/trace-event-python.o: util/scripting-engines/trace-event-python.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) $(PYTHON_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-shadow $< -scripts/python/Perf-Trace-Util/Context.o: scripts/python/Perf-Trace-Util/Context.c PERF-CFLAGS - $(QUIET_CC)$(CC) -o scripts/python/Perf-Trace-Util/Context.o -c $(ALL_CFLAGS) $(PYTHON_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-nested-externs $< +$(OUTPUT)scripts/python/Perf-Trace-Util/Context.o: scripts/python/Perf-Trace-Util/Context.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) $(PYTHON_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-nested-externs $< -perf-%$X: %.o $(PERFLIBS) +$(OUTPUT)perf-%$X: %.o $(PERFLIBS) $(QUIET_LINK)$(CC) $(ALL_CFLAGS) -o $@ $(ALL_LDFLAGS) $(filter %.o,$^) $(LIBS) $(LIB_OBJS) $(BUILTIN_OBJS): $(LIB_H) @@ -974,17 +986,17 @@ cscope: TRACK_CFLAGS = $(subst ','\'',$(ALL_CFLAGS)):\ $(bindir_SQ):$(perfexecdir_SQ):$(template_dir_SQ):$(prefix_SQ) -PERF-CFLAGS: .FORCE-PERF-CFLAGS +$(OUTPUT)PERF-CFLAGS: .FORCE-PERF-CFLAGS @FLAGS='$(TRACK_CFLAGS)'; \ - if test x"$$FLAGS" != x"`cat PERF-CFLAGS 2>/dev/null`" ; then \ + if test x"$$FLAGS" != x"`cat $(OUTPUT)PERF-CFLAGS 2>/dev/null`" ; then \ echo 1>&2 " * new build flags or prefix"; \ - echo "$$FLAGS" >PERF-CFLAGS; \ + echo "$$FLAGS" >$(OUTPUT)PERF-CFLAGS; \ fi # We need to apply sq twice, once to protect from the shell -# that runs PERF-BUILD-OPTIONS, and then again to protect it +# that runs $(OUTPUT)PERF-BUILD-OPTIONS, and then again to protect it # and the first level quoting from the shell that runs "echo". -PERF-BUILD-OPTIONS: .FORCE-PERF-BUILD-OPTIONS +$(OUTPUT)PERF-BUILD-OPTIONS: .FORCE-PERF-BUILD-OPTIONS @echo SHELL_PATH=\''$(subst ','\'',$(SHELL_PATH_SQ))'\' >$@ @echo TAR=\''$(subst ','\'',$(subst ','\'',$(TAR)))'\' >>$@ @echo NO_CURL=\''$(subst ','\'',$(subst ','\'',$(NO_CURL)))'\' >>$@ @@ -1005,7 +1017,7 @@ all:: $(TEST_PROGRAMS) export NO_SVN_TESTS -check: common-cmds.h +check: $(OUTPUT)common-cmds.h if sparse; \ then \ for i in *.c */*.c; \ @@ -1039,10 +1051,10 @@ export perfexec_instdir install: all $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(bindir_SQ)' - $(INSTALL) perf$X '$(DESTDIR_SQ)$(bindir_SQ)' + $(INSTALL) $(OUTPUT)perf$X '$(DESTDIR_SQ)$(bindir_SQ)' $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/Perf-Trace-Util/lib/Perf/Trace' $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/bin' - $(INSTALL) perf-archive -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)' + $(INSTALL) $(OUTPUT)perf-archive -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)' $(INSTALL) scripts/perl/Perf-Trace-Util/lib/Perf/Trace/* -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/Perf-Trace-Util/lib/Perf/Trace' $(INSTALL) scripts/perl/*.pl -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl' $(INSTALL) scripts/perl/bin/* -t '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/scripts/perl/bin' @@ -1056,7 +1068,7 @@ ifdef BUILT_INS $(INSTALL) -d -m 755 '$(DESTDIR_SQ)$(perfexec_instdir_SQ)' $(INSTALL) $(BUILT_INS) '$(DESTDIR_SQ)$(perfexec_instdir_SQ)' ifneq (,$X) - $(foreach p,$(patsubst %$X,%,$(filter %$X,$(ALL_PROGRAMS) $(BUILT_INS) perf$X)), $(RM) '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/$p';) + $(foreach p,$(patsubst %$X,%,$(filter %$X,$(ALL_PROGRAMS) $(BUILT_INS) $(OUTPUT)perf$X)), $(RM) '$(DESTDIR_SQ)$(perfexec_instdir_SQ)/$p';) endif endif @@ -1140,14 +1152,14 @@ clean: $(RM) *.o */*.o */*/*.o */*/*/*.o $(LIB_FILE) $(RM) $(ALL_PROGRAMS) $(BUILT_INS) perf$X $(RM) $(TEST_PROGRAMS) - $(RM) *.spec *.pyc *.pyo */*.pyc */*.pyo common-cmds.h TAGS tags cscope* + $(RM) *.spec *.pyc *.pyo */*.pyc */*.pyo $(OUTPUT)common-cmds.h TAGS tags cscope* $(RM) -r autom4te.cache $(RM) config.log config.mak.autogen config.mak.append config.status config.cache $(RM) -r $(PERF_TARNAME) .doc-tmp-dir $(RM) $(PERF_TARNAME).tar.gz perf-core_$(PERF_VERSION)-*.tar.gz $(RM) $(htmldocs).tar.gz $(manpages).tar.gz $(MAKE) -C Documentation/ clean - $(RM) PERF-VERSION-FILE PERF-CFLAGS PERF-BUILD-OPTIONS + $(RM) $(OUTPUT)PERF-VERSION-FILE $(OUTPUT)PERF-CFLAGS $(OUTPUT)PERF-BUILD-OPTIONS .PHONY: all install clean strip .PHONY: shell_compatibility_test please_set_SHELL_PATH_to_a_more_modern_shell diff --git a/tools/perf/util/PERF-VERSION-GEN b/tools/perf/util/PERF-VERSION-GEN index 54552a0..49ece79 100755 --- a/tools/perf/util/PERF-VERSION-GEN +++ b/tools/perf/util/PERF-VERSION-GEN @@ -1,6 +1,10 @@ #!/bin/sh -GVF=PERF-VERSION-FILE +if [ $# -eq 1 ] ; then + OUTPUT=$1 +fi + +GVF=${OUTPUT}PERF-VERSION-FILE DEF_VER=v0.0.2.PERF LF=' -- cgit v0.10.2 From 70162138c91b040da3162fe1f34fe8aaf6506f10 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 30 Mar 2010 18:27:39 -0300 Subject: perf record: Add a fallback to the reference relocation symbol MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Usually "_text" is enough, but I received reports that its not always available, so fallback to "_stext" for the symbol we use to check if we need to apply any relocation to all the symbols in the kernel symtab, for when, for instance, kexec is being used. Reported-by: Darren Hart Cc: Darren Hart Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 60ecdd3..80dc444 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -575,6 +575,9 @@ static int __cmd_record(int argc, const char **argv) err = event__synthesize_kernel_mmap(process_synthesized_event, session, "_text"); + if (err < 0) + err = event__synthesize_kernel_mmap(process_synthesized_event, + session, "_stext"); if (err < 0) { pr_err("Couldn't record kernel reference relocation symbol.\n"); return err; -- cgit v0.10.2 From a4e3b956a820162b7c1d616117b4f23b6017f504 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 31 Mar 2010 11:33:40 -0300 Subject: perf hist: Replace ->print() routines by ->snprintf() equivalents MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Then hist_entry__fprintf will just us the newly introduced hist_entry__snprintf, add the newline and fprintf it to the supplied FILE descriptor. This allows us to remove the use_browser checking in the color_printf routines, that now got color_snprintf variants too. The newt TUI browser (and other GUIs that may come in the future) don't have to worry about stdio specific stuff in the strings they get from the se->snprintf routines and instead use whatever means to do the equivalent. Also the newt TUI browser don't have to use the fmemopen() hack, instead it can use the se->snprintf routines directly. For now tho use the hist_entry__snprintf routine to reduce the patch size. Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/color.c b/tools/perf/util/color.c index 9da0191..e191eb9 100644 --- a/tools/perf/util/color.c +++ b/tools/perf/util/color.c @@ -166,6 +166,31 @@ int perf_color_default_config(const char *var, const char *value, void *cb) return perf_default_config(var, value, cb); } +static int __color_vsnprintf(char *bf, size_t size, const char *color, + const char *fmt, va_list args, const char *trail) +{ + int r = 0; + + /* + * Auto-detect: + */ + if (perf_use_color_default < 0) { + if (isatty(1) || pager_in_use()) + perf_use_color_default = 1; + else + perf_use_color_default = 0; + } + + if (perf_use_color_default && *color) + r += snprintf(bf, size, "%s", color); + r += vsnprintf(bf + r, size - r, fmt, args); + if (perf_use_color_default && *color) + r += snprintf(bf + r, size - r, "%s", PERF_COLOR_RESET); + if (trail) + r += snprintf(bf + r, size - r, "%s", trail); + return r; +} + static int __color_vfprintf(FILE *fp, const char *color, const char *fmt, va_list args, const char *trail) { @@ -191,11 +216,28 @@ static int __color_vfprintf(FILE *fp, const char *color, const char *fmt, return r; } +int color_vsnprintf(char *bf, size_t size, const char *color, + const char *fmt, va_list args) +{ + return __color_vsnprintf(bf, size, color, fmt, args, NULL); +} + int color_vfprintf(FILE *fp, const char *color, const char *fmt, va_list args) { return __color_vfprintf(fp, color, fmt, args, NULL); } +int color_snprintf(char *bf, size_t size, const char *color, + const char *fmt, ...) +{ + va_list args; + int r; + + va_start(args, fmt); + r = color_vsnprintf(bf, size, color, fmt, args); + va_end(args); + return r; +} int color_fprintf(FILE *fp, const char *color, const char *fmt, ...) { @@ -203,10 +245,7 @@ int color_fprintf(FILE *fp, const char *color, const char *fmt, ...) int r; va_start(args, fmt); - if (use_browser) - r = vfprintf(fp, fmt, args); - else - r = color_vfprintf(fp, color, fmt, args); + r = color_vfprintf(fp, color, fmt, args); va_end(args); return r; } @@ -277,3 +316,9 @@ int percent_color_fprintf(FILE *fp, const char *fmt, double percent) return r; } + +int percent_color_snprintf(char *bf, size_t size, const char *fmt, double percent) +{ + const char *color = get_percent_color(percent); + return color_snprintf(bf, size, color, fmt, percent); +} diff --git a/tools/perf/util/color.h b/tools/perf/util/color.h index 24e8809..dea082b 100644 --- a/tools/perf/util/color.h +++ b/tools/perf/util/color.h @@ -32,10 +32,14 @@ int perf_color_default_config(const char *var, const char *value, void *cb); int perf_config_colorbool(const char *var, const char *value, int stdout_is_tty); void color_parse(const char *value, const char *var, char *dst); void color_parse_mem(const char *value, int len, const char *var, char *dst); +int color_vsnprintf(char *bf, size_t size, const char *color, + const char *fmt, va_list args); int color_vfprintf(FILE *fp, const char *color, const char *fmt, va_list args); int color_fprintf(FILE *fp, const char *color, const char *fmt, ...); +int color_snprintf(char *bf, size_t size, const char *color, const char *fmt, ...); int color_fprintf_ln(FILE *fp, const char *color, const char *fmt, ...); int color_fwrite_lines(FILE *fp, const char *color, size_t count, const char *buf); +int percent_color_snprintf(char *bf, size_t size, const char *fmt, double percent); int percent_color_fprintf(FILE *fp, const char *fmt, double percent); const char *get_percent_color(double percent); diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index a46d093..f079491 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -455,16 +455,17 @@ static size_t hist_entry_callchain__fprintf(FILE *fp, struct hist_entry *self, return ret; } -size_t hist_entry__fprintf(struct hist_entry *self, +int hist_entry__snprintf(struct hist_entry *self, + char *s, size_t size, struct perf_session *pair_session, bool show_displacement, - long displacement, FILE *fp, + long displacement, bool color, u64 session_total) { struct sort_entry *se; u64 count, total; const char *sep = symbol_conf.field_sep; - size_t ret; + int ret; if (symbol_conf.exclude_other && !self->parent) return 0; @@ -477,17 +478,22 @@ size_t hist_entry__fprintf(struct hist_entry *self, total = session_total; } - if (total) - ret = percent_color_fprintf(fp, sep ? "%.2f" : " %6.2f%%", - (count * 100.0) / total); - else - ret = fprintf(fp, sep ? "%lld" : "%12lld ", count); + if (total) { + if (color) + ret = percent_color_snprintf(s, size, + sep ? "%.2f" : " %6.2f%%", + (count * 100.0) / total); + else + ret = snprintf(s, size, sep ? "%.2f" : " %6.2f%%", + (count * 100.0) / total); + } else + ret = snprintf(s, size, sep ? "%lld" : "%12lld ", count); if (symbol_conf.show_nr_samples) { if (sep) - ret += fprintf(fp, "%c%lld", *sep, count); + ret += snprintf(s + ret, size - ret, "%c%lld", *sep, count); else - ret += fprintf(fp, "%11lld", count); + ret += snprintf(s + ret, size - ret, "%11lld", count); } if (pair_session) { @@ -507,9 +513,9 @@ size_t hist_entry__fprintf(struct hist_entry *self, snprintf(bf, sizeof(bf), " "); if (sep) - ret += fprintf(fp, "%c%s", *sep, bf); + ret += snprintf(s + ret, size - ret, "%c%s", *sep, bf); else - ret += fprintf(fp, "%11.11s", bf); + ret += snprintf(s + ret, size - ret, "%11.11s", bf); if (show_displacement) { if (displacement) @@ -518,9 +524,9 @@ size_t hist_entry__fprintf(struct hist_entry *self, snprintf(bf, sizeof(bf), " "); if (sep) - ret += fprintf(fp, "%c%s", *sep, bf); + ret += snprintf(s + ret, size - ret, "%c%s", *sep, bf); else - ret += fprintf(fp, "%6.6s", bf); + ret += snprintf(s + ret, size - ret, "%6.6s", bf); } } @@ -528,11 +534,25 @@ size_t hist_entry__fprintf(struct hist_entry *self, if (se->elide) continue; - ret += fprintf(fp, "%s", sep ?: " "); - ret += se->print(fp, self, se->width ? *se->width : 0); + ret += snprintf(s + ret, size - ret, "%s", sep ?: " "); + ret += se->snprintf(self, s + ret, size - ret, + se->width ? *se->width : 0); } - return ret + fprintf(fp, "\n"); + return ret; +} + +int hist_entry__fprintf(struct hist_entry *self, + struct perf_session *pair_session, + bool show_displacement, + long displacement, FILE *fp, + u64 session_total) +{ + char bf[512]; + hist_entry__snprintf(self, bf, sizeof(bf), pair_session, + show_displacement, displacement, + true, session_total); + return fprintf(fp, "%s\n", bf); } static size_t hist_entry__fprintf_callchain(struct hist_entry *self, FILE *fp, diff --git a/tools/perf/util/hist.h b/tools/perf/util/hist.h index da6a8c1..ad17f0a 100644 --- a/tools/perf/util/hist.h +++ b/tools/perf/util/hist.h @@ -18,11 +18,16 @@ struct hist_entry *__perf_session__add_hist_entry(struct rb_root *hists, u64 count, bool *hit); extern int64_t hist_entry__cmp(struct hist_entry *, struct hist_entry *); extern int64_t hist_entry__collapse(struct hist_entry *, struct hist_entry *); -size_t hist_entry__fprintf(struct hist_entry *self, +int hist_entry__fprintf(struct hist_entry *self, struct perf_session *pair_session, bool show_displacement, long displacement, FILE *fp, u64 session_total); +int hist_entry__snprintf(struct hist_entry *self, + char *bf, size_t size, + struct perf_session *pair_session, + bool show_displacement, long displacement, + bool color, u64 session_total); void hist_entry__free(struct hist_entry *); u64 perf_session__output_resort(struct rb_root *hists, u64 total_samples); diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index b0210ae..edd628f 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -294,60 +294,17 @@ static void hist_entry__append_callchain_browser(struct hist_entry *self, } } -/* - * FIXME: get lib/string.c linked with perf somehow - */ -static char *skip_spaces(const char *str) -{ - while (isspace(*str)) - ++str; - return (char *)str; -} - -static char *strim(char *s) -{ - size_t size; - char *end; - - s = skip_spaces(s); - size = strlen(s); - if (!size) - return s; - - end = s + size - 1; - while (end >= s && isspace(*end)) - end--; - *(end + 1) = '\0'; - - return s; -} - static size_t hist_entry__append_browser(struct hist_entry *self, newtComponent tree, u64 total) { - char bf[1024], *s; - FILE *fp; + char s[256]; + size_t ret; if (symbol_conf.exclude_other && !self->parent) return 0; - fp = fmemopen(bf, sizeof(bf), "w"); - if (fp == NULL) - return 0; - - hist_entry__fprintf(self, NULL, false, 0, fp, total); - fclose(fp); - - /* - * FIXME: We shouldn't need to trim, as the printing routines shouldn't - * add spaces it in the first place, the stdio output routines should - * call a __snprintf method instead of the current __print (that - * actually is a __fprintf) one, but get the raw string and _then_ add - * the newline, as this is a detail of stdio printing, not needed in - * other UIs, e.g. newt. - */ - s = strim(bf); - + ret = hist_entry__snprintf(self, s, sizeof(s), NULL, + false, 0, false, total); if (symbol_conf.use_callchain) { int indexes[2]; @@ -357,7 +314,7 @@ static size_t hist_entry__append_browser(struct hist_entry *self, } else newtListboxAppendEntry(tree, s, &self->ms); - return strlen(s); + return ret; } static void map_symbol__annotate_browser(const struct map_symbol *self) diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c index 9b80c13..31329a1 100644 --- a/tools/perf/util/sort.c +++ b/tools/perf/util/sort.c @@ -18,10 +18,21 @@ char * field_sep; LIST_HEAD(hist_entry__sort_list); +static int hist_entry__thread_snprintf(struct hist_entry *self, char *bf, + size_t size, unsigned int width); +static int hist_entry__comm_snprintf(struct hist_entry *self, char *bf, + size_t size, unsigned int width); +static int hist_entry__dso_snprintf(struct hist_entry *self, char *bf, + size_t size, unsigned int width); +static int hist_entry__sym_snprintf(struct hist_entry *self, char *bf, + size_t size, unsigned int width); +static int hist_entry__parent_snprintf(struct hist_entry *self, char *bf, + size_t size, unsigned int width); + struct sort_entry sort_thread = { .header = "Command: Pid", .cmp = sort__thread_cmp, - .print = sort__thread_print, + .snprintf = hist_entry__thread_snprintf, .width = &threads__col_width, }; @@ -29,27 +40,27 @@ struct sort_entry sort_comm = { .header = "Command", .cmp = sort__comm_cmp, .collapse = sort__comm_collapse, - .print = sort__comm_print, + .snprintf = hist_entry__comm_snprintf, .width = &comms__col_width, }; struct sort_entry sort_dso = { .header = "Shared Object", .cmp = sort__dso_cmp, - .print = sort__dso_print, + .snprintf = hist_entry__dso_snprintf, .width = &dsos__col_width, }; struct sort_entry sort_sym = { .header = "Symbol", .cmp = sort__sym_cmp, - .print = sort__sym_print, + .snprintf = hist_entry__sym_snprintf, }; struct sort_entry sort_parent = { .header = "Parent symbol", .cmp = sort__parent_cmp, - .print = sort__parent_print, + .snprintf = hist_entry__parent_snprintf, .width = &parent_symbol__col_width, }; @@ -85,45 +96,38 @@ sort__thread_cmp(struct hist_entry *left, struct hist_entry *right) return right->thread->pid - left->thread->pid; } -int repsep_fprintf(FILE *fp, const char *fmt, ...) +static int repsep_snprintf(char *bf, size_t size, const char *fmt, ...) { int n; va_list ap; va_start(ap, fmt); - if (!field_sep) - n = vfprintf(fp, fmt, ap); - else { - char *bf = NULL; - n = vasprintf(&bf, fmt, ap); - if (n > 0) { - char *sep = bf; - - while (1) { - sep = strchr(sep, *field_sep); - if (sep == NULL) - break; - *sep = '.'; - } + n = vsnprintf(bf, size, fmt, ap); + if (field_sep && n > 0) { + char *sep = bf; + + while (1) { + sep = strchr(sep, *field_sep); + if (sep == NULL) + break; + *sep = '.'; } - fputs(bf, fp); - free(bf); } va_end(ap); return n; } -size_t -sort__thread_print(FILE *fp, struct hist_entry *self, unsigned int width) +static int hist_entry__thread_snprintf(struct hist_entry *self, char *bf, + size_t size, unsigned int width) { - return repsep_fprintf(fp, "%*s:%5d", width - 6, + return repsep_snprintf(bf, size, "%*s:%5d", width, self->thread->comm ?: "", self->thread->pid); } -size_t -sort__comm_print(FILE *fp, struct hist_entry *self, unsigned int width) +static int hist_entry__comm_snprintf(struct hist_entry *self, char *bf, + size_t size, unsigned int width) { - return repsep_fprintf(fp, "%*s", width, self->thread->comm); + return repsep_snprintf(bf, size, "%*s", width, self->thread->comm); } /* --sort dso */ @@ -149,16 +153,16 @@ sort__dso_cmp(struct hist_entry *left, struct hist_entry *right) return strcmp(dso_name_l, dso_name_r); } -size_t -sort__dso_print(FILE *fp, struct hist_entry *self, unsigned int width) +static int hist_entry__dso_snprintf(struct hist_entry *self, char *bf, + size_t size, unsigned int width) { if (self->ms.map && self->ms.map->dso) { const char *dso_name = !verbose ? self->ms.map->dso->short_name : self->ms.map->dso->long_name; - return repsep_fprintf(fp, "%-*s", width, dso_name); + return repsep_snprintf(bf, size, "%-*s", width, dso_name); } - return repsep_fprintf(fp, "%*llx", width, (u64)self->ip); + return repsep_snprintf(bf, size, "%*Lx", width, self->ip); } /* --sort symbol */ @@ -177,22 +181,22 @@ sort__sym_cmp(struct hist_entry *left, struct hist_entry *right) return (int64_t)(ip_r - ip_l); } - -size_t -sort__sym_print(FILE *fp, struct hist_entry *self, unsigned int width __used) +static int hist_entry__sym_snprintf(struct hist_entry *self, char *bf, + size_t size, unsigned int width __used) { size_t ret = 0; if (verbose) { char o = self->ms.map ? dso__symtab_origin(self->ms.map->dso) : '!'; - ret += repsep_fprintf(fp, "%#018llx %c ", (u64)self->ip, o); + ret += repsep_snprintf(bf, size, "%#018llx %c ", self->ip, o); } - ret += repsep_fprintf(fp, "[%c] ", self->level); + ret += repsep_snprintf(bf + ret, size - ret, "[%c] ", self->level); if (self->ms.sym) - ret += repsep_fprintf(fp, "%s", self->ms.sym->name); + ret += repsep_snprintf(bf + ret, size - ret, "%s", + self->ms.sym->name); else - ret += repsep_fprintf(fp, "%#016llx", (u64)self->ip); + ret += repsep_snprintf(bf + ret, size - ret, "%#016llx", self->ip); return ret; } @@ -231,10 +235,10 @@ sort__parent_cmp(struct hist_entry *left, struct hist_entry *right) return strcmp(sym_l->name, sym_r->name); } -size_t -sort__parent_print(FILE *fp, struct hist_entry *self, unsigned int width) +static int hist_entry__parent_snprintf(struct hist_entry *self, char *bf, + size_t size, unsigned int width) { - return repsep_fprintf(fp, "%-*s", width, + return repsep_snprintf(bf, size, "%-*s", width, self->parent ? self->parent->name : "[other]"); } diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h index 5985686..439ec5f 100644 --- a/tools/perf/util/sort.h +++ b/tools/perf/util/sort.h @@ -76,7 +76,8 @@ struct sort_entry { int64_t (*cmp)(struct hist_entry *, struct hist_entry *); int64_t (*collapse)(struct hist_entry *, struct hist_entry *); - size_t (*print)(FILE *fp, struct hist_entry *, unsigned int width); + int (*snprintf)(struct hist_entry *self, char *bf, size_t size, + unsigned int width); unsigned int *width; bool elide; }; @@ -86,7 +87,6 @@ extern struct list_head hist_entry__sort_list; void setup_sorting(const char * const usagestr[], const struct option *opts); -extern int repsep_fprintf(FILE *fp, const char *fmt, ...); extern size_t sort__thread_print(FILE *, struct hist_entry *, unsigned int); extern size_t sort__comm_print(FILE *, struct hist_entry *, unsigned int); extern size_t sort__dso_print(FILE *, struct hist_entry *, unsigned int); -- cgit v0.10.2 From e727ca73f85d4c5be3547eda674168219d1c22d8 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 1 Apr 2010 19:12:13 -0300 Subject: perf kmem: Resolve kernel symbols again MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Due to the assumption in perf_session__new that the kernel maps would be created using the fake PERF_RECORD_MMAP event in a perf.data file 'perf kmem --stat caller', that doesn't have such event, ends up not being able to resolve the kernel addresses. Fix it by calling perf_session__create_kernel_maps() in __cmd_kmem(). LKML-Reference: Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/builtin-kmem.c b/tools/perf/builtin-kmem.c index 32edb6a..7cbb5eb 100644 --- a/tools/perf/builtin-kmem.c +++ b/tools/perf/builtin-kmem.c @@ -489,6 +489,9 @@ static int __cmd_kmem(void) if (session == NULL) return -ENOMEM; + if (perf_session__create_kernel_maps(session) < 0) + goto out_delete; + if (!perf_session__has_traces(session, "kmem record")) goto out_delete; diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index 32765cd..9da5e72 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -52,11 +52,6 @@ out_close: return -1; } -static inline int perf_session__create_kernel_maps(struct perf_session *self) -{ - return map_groups__create_kernel_maps(&self->kmaps, self->vmlinux_maps); -} - struct perf_session *perf_session__new(const char *filename, int mode, bool force) { size_t len = filename ? strlen(filename) + 1 : 0; diff --git a/tools/perf/util/session.h b/tools/perf/util/session.h index 6a15dae..dffaff5 100644 --- a/tools/perf/util/session.h +++ b/tools/perf/util/session.h @@ -80,6 +80,11 @@ static inline int __perf_session__create_kernel_maps(struct perf_session *self, self->vmlinux_maps, kernel); } +static inline int perf_session__create_kernel_maps(struct perf_session *self) +{ + return map_groups__create_kernel_maps(&self->kmaps, self->vmlinux_maps); +} + static inline struct map * perf_session__new_module_map(struct perf_session *self, u64 start, const char *filename) -- cgit v0.10.2 From 71cf8b8ff7d6a79af086be9e4c72628da9d62d58 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 1 Apr 2010 21:24:38 -0300 Subject: perf kmem: Fixup the symbol address before using it MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit We get absolute addresses in the events, but relative ones from the symbol subsystem, so calculate the absolute address by asking for the map where the symbol was found, that has the place where the DSO was actually loaded. For the core kernel this poses no problems if the kernel is not relocated by things like kexec, or if we use /proc/kallsyms, but for modules we were getting really large, negative offsets. LKML-Reference: Cc: Frédéric Weisbecker Cc: Li Zefan Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/builtin-kmem.c b/tools/perf/builtin-kmem.c index 7cbb5eb..513aa8a 100644 --- a/tools/perf/builtin-kmem.c +++ b/tools/perf/builtin-kmem.c @@ -363,6 +363,7 @@ static void __print_result(struct rb_root *root, struct perf_session *session, struct alloc_stat *data = rb_entry(next, struct alloc_stat, node); struct symbol *sym = NULL; + struct map *map; char buf[BUFSIZ]; u64 addr; @@ -370,13 +371,13 @@ static void __print_result(struct rb_root *root, struct perf_session *session, addr = data->call_site; if (!raw_ip) sym = map_groups__find_function(&session->kmaps, - addr, NULL, NULL); + addr, &map, NULL); } else addr = data->ptr; if (sym != NULL) snprintf(buf, sizeof(buf), "%s+%Lx", sym->name, - addr - sym->start); + addr - map->unmap_ip(map, sym->start)); else snprintf(buf, sizeof(buf), "%#Lx", addr); printf(" %-34s |", buf); -- cgit v0.10.2 From b9fb93047756c5e4129dfda7591612de61b0e877 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 2 Apr 2010 09:50:42 -0300 Subject: perf hist: Only allocate callchain_node if processing callchains MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The struct callchain_node size is 120 bytes, that are never used when there are no callchains or '-g none' is specified, so conditionally allocate it, reducing sizeof(struct hist_entry) from 210 bytes to only 96, greatly speeding the non-callchain processing. LKML-Reference: Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 3819185..1fb13e5 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -110,8 +110,8 @@ static int perf_session__add_hist_entry(struct perf_session *self, if (symbol_conf.use_callchain) { if (!hit) - callchain_init(&he->callchain); - err = append_chain(&he->callchain, data->callchain, syms); + callchain_init(he->callchain); + err = append_chain(he->callchain, data->callchain, syms); free(syms); if (err) diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index f079491..18cf8b3 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -50,7 +50,8 @@ struct hist_entry *__perf_session__add_hist_entry(struct rb_root *hists, p = &(*p)->rb_right; } - he = malloc(sizeof(*he)); + he = malloc(sizeof(*he) + (symbol_conf.use_callchain ? + sizeof(struct callchain_node) : 0)); if (!he) return NULL; *he = entry; @@ -168,7 +169,7 @@ static void perf_session__insert_output_hist_entry(struct rb_root *root, struct hist_entry *iter; if (symbol_conf.use_callchain) - callchain_param.sort(&he->sorted_chain, &he->callchain, + callchain_param.sort(&he->sorted_chain, he->callchain, min_callchain_hits, &callchain_param); while (*p != NULL) { diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h index 439ec5f..5bf2b74 100644 --- a/tools/perf/util/sort.h +++ b/tools/perf/util/sort.h @@ -49,12 +49,12 @@ struct hist_entry { u64 ip; char level; struct symbol *parent; - struct callchain_node callchain; union { unsigned long position; struct hist_entry *pair; struct rb_root sorted_chain; }; + struct callchain_node callchain[0]; }; enum sort_type { -- cgit v0.10.2 From ad5b217b152d99ca3922153500c619d9758dd87a Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 2 Apr 2010 10:04:18 -0300 Subject: perf session: Remove one more exit() call from library code MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Return NULL instead and make the caller propagate the error. LKML-Reference: Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 1fb13e5..6767f10 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -89,9 +89,12 @@ static int perf_session__add_hist_entry(struct perf_session *self, struct event_stat_id *stats; struct perf_event_attr *attr; - if ((sort__has_parent || symbol_conf.use_callchain) && data->callchain) + if ((sort__has_parent || symbol_conf.use_callchain) && data->callchain) { syms = perf_session__resolve_callchain(self, al->thread, data->callchain, &parent); + if (syms == NULL) + return -ENOMEM; + } attr = perf_header__find_attr(data->id, &self->header); if (attr) diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index 9da5e72..ddf288f 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -118,16 +118,11 @@ struct map_symbol *perf_session__resolve_callchain(struct perf_session *self, struct symbol **parent) { u8 cpumode = PERF_RECORD_MISC_USER; - struct map_symbol *syms = NULL; unsigned int i; + struct map_symbol *syms = calloc(chain->nr, sizeof(*syms)); - if (symbol_conf.use_callchain) { - syms = calloc(chain->nr, sizeof(*syms)); - if (!syms) { - fprintf(stderr, "Can't allocate memory for symbols\n"); - exit(-1); - } - } + if (!syms) + return NULL; for (i = 0; i < chain->nr; i++) { u64 ip = chain->ips[i]; -- cgit v0.10.2 From 2aefa4f733f2c5ce51dd2316ffecb258463fde71 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Fri, 2 Apr 2010 12:30:57 -0300 Subject: perf tools: sort_dimension__add shouldn't die MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Propagate error instead. LKML-Reference: Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 6767f10..b13a7e2 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -473,7 +473,8 @@ int cmd_report(int argc, const char **argv, const char *prefix __used) setup_sorting(report_usage, options); if (parent_pattern != default_parent_pattern) { - sort_dimension__add("parent"); + if (sort_dimension__add("parent") < 0) + return -1; sort_parent.elide = 1; } else symbol_conf.exclude_other = false; diff --git a/tools/perf/util/sort.c b/tools/perf/util/sort.c index 31329a1..9d24d4b 100644 --- a/tools/perf/util/sort.c +++ b/tools/perf/util/sort.c @@ -264,9 +264,8 @@ int sort_dimension__add(const char *tok) char err[BUFSIZ]; regerror(ret, &parent_regex, err, sizeof(err)); - fprintf(stderr, "Invalid regex: %s\n%s", - parent_pattern, err); - exit(-1); + pr_err("Invalid regex: %s\n%s", parent_pattern, err); + return -EINVAL; } sort__has_parent = 1; } -- cgit v0.10.2 From 227f98d98d2ed7929f41426adc21f57b927354a6 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 18 Feb 2010 08:27:24 -0800 Subject: nfsd4: preallocate nfs4_rpc_args Instead of allocating this small structure, just include it in the delegation. The nfsd4_callback structure isn't really necessary yet, but we plan to add to it all the information necessary to perform a callback. Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index ed12ad4..b99c3f0 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -78,11 +78,6 @@ enum nfs_cb_opnum4 { cb_sequence_dec_sz + \ op_dec_sz) -struct nfs4_rpc_args { - void *args_op; - struct nfsd4_cb_sequence args_seq; -}; - /* * Generic encode routines from fs/nfs/nfs4xdr.c */ @@ -676,7 +671,7 @@ static void nfsd4_cb_recall_done(struct rpc_task *task, void *calldata) break; default: /* success, or error we can't handle */ - goto done; + return; } if (dp->dl_retries--) { rpc_delay(task, 2*HZ); @@ -687,8 +682,6 @@ static void nfsd4_cb_recall_done(struct rpc_task *task, void *calldata) atomic_set(&clp->cl_cb_conn.cb_set, 0); warn_no_callback_path(clp, task->tk_status); } -done: - kfree(task->tk_msg.rpc_argp); } static void nfsd4_cb_recall_release(void *calldata) @@ -714,24 +707,19 @@ nfsd4_cb_recall(struct nfs4_delegation *dp) { struct nfs4_client *clp = dp->dl_client; struct rpc_clnt *clnt = clp->cl_cb_conn.cb_client; - struct nfs4_rpc_args *args; + struct nfs4_rpc_args *args = &dp->dl_recall.cb_args; struct rpc_message msg = { .rpc_proc = &nfs4_cb_procedures[NFSPROC4_CLNT_CB_RECALL], .rpc_cred = callback_cred }; - int status = -ENOMEM; + int status; - args = kzalloc(sizeof(*args), GFP_KERNEL); - if (!args) - goto out; args->args_op = dp; msg.rpc_argp = args; dp->dl_retries = 1; status = rpc_call_async(clnt, &msg, RPC_TASK_SOFT, &nfsd4_cb_recall_ops, dp); -out: if (status) { - kfree(args); put_nfs4_client(clp); nfs4_put_delegation(dp); } diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h index fefeae2..b854379 100644 --- a/fs/nfsd/state.h +++ b/fs/nfsd/state.h @@ -70,6 +70,15 @@ struct nfsd4_cb_sequence { struct nfs4_client *cbs_clp; }; +struct nfs4_rpc_args { + void *args_op; + struct nfsd4_cb_sequence args_seq; +}; + +struct nfsd4_callback { + struct nfs4_rpc_args cb_args; +}; + struct nfs4_delegation { struct list_head dl_perfile; struct list_head dl_perclnt; @@ -86,6 +95,7 @@ struct nfs4_delegation { stateid_t dl_stateid; struct knfsd_fh dl_fh; int dl_retries; + struct nfsd4_callback dl_recall; }; /* client delegation callback info */ -- cgit v0.10.2 From 147efd0dd702ce2f1ab44449bd70369405ef68fd Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 21 Feb 2010 17:41:19 -0800 Subject: nfsd4: shutdown callbacks on expiry Once we've expired the client, there's no further purpose to the callbacks; go ahead and shut down the callback client rather than waiting for the last reference to go. Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index efef7f2..9ce5831 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -697,9 +697,6 @@ shutdown_callback_client(struct nfs4_client *clp) static inline void free_client(struct nfs4_client *clp) { - shutdown_callback_client(clp); - if (clp->cl_cb_xprt) - svc_xprt_put(clp->cl_cb_xprt); if (clp->cl_cred.cr_group_info) put_group_info(clp->cl_cred.cr_group_info); kfree(clp->cl_principal); @@ -752,6 +749,9 @@ expire_client(struct nfs4_client *clp) se_perclnt); release_session(ses); } + shutdown_callback_client(clp); + if (clp->cl_cb_xprt) + svc_xprt_put(clp->cl_cb_xprt); put_nfs4_client(clp); } -- cgit v0.10.2 From 3df796dbe97a98a6a25e6b7b88e9d326e261f371 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 21 Feb 2010 17:51:53 -0800 Subject: nfsd4: remove dprintk I haven't found this useful. Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index 9ce5831..5d86df1 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -718,9 +718,6 @@ expire_client(struct nfs4_client *clp) struct nfs4_delegation *dp; struct list_head reaplist; - dprintk("NFSD: expire_client cl_count %d\n", - atomic_read(&clp->cl_count)); - INIT_LIST_HEAD(&reaplist); spin_lock(&recall_lock); while (!list_empty(&clp->cl_delegations)) { -- cgit v0.10.2 From 9045b4b9f7f340f43de0cf687b5b52f6feaaa984 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Sun, 21 Feb 2010 17:53:04 -0800 Subject: nfsd4: remove probe task's reference on client Any null probe rpc will be synchronously destroyed by the rpc_shutdown_client() in expire_client(), so the rpc task cannot outlast the nfs4 client. Therefore there's no need for that task to hold a reference on the client. Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index b99c3f0..91eb2ea 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -509,7 +509,6 @@ static void nfsd4_cb_probe_done(struct rpc_task *task, void *calldata) warn_no_callback_path(clp, task->tk_status); else atomic_set(&clp->cl_cb_conn.cb_set, 1); - put_nfs4_client(clp); } static const struct rpc_call_ops nfsd4_cb_probe_ops = { @@ -542,10 +541,8 @@ void do_probe_callback(struct nfs4_client *clp) status = rpc_call_async(cb->cb_client, &msg, RPC_TASK_SOFT | RPC_TASK_SOFTCONN, &nfsd4_cb_probe_ops, (void *)clp); - if (status) { + if (status) warn_no_callback_path(clp, status); - put_nfs4_client(clp); - } } /* @@ -563,10 +560,6 @@ nfsd4_probe_callback(struct nfs4_client *clp) warn_no_callback_path(clp, status); return; } - - /* the task holds a reference to the nfs4_client struct */ - atomic_inc(&clp->cl_count); - do_probe_callback(clp); } -- cgit v0.10.2 From 0e0d5e0c4bb0476d53a43bfc87d03a25ec4b5579 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Fri, 2 Apr 2010 16:07:19 -0700 Subject: ceph: fix ack counter reset on connection reset If in_seq_acked isn't reset along with in_seq, we don't ack received messages until we reach the old count, consuming gobs memory on the other end of the connection and introducing a large delay when those messages are eventually deleted. Signed-off-by: Sage Weil diff --git a/fs/ceph/messenger.c b/fs/ceph/messenger.c index a32f0f8..f35b494 100644 --- a/fs/ceph/messenger.c +++ b/fs/ceph/messenger.c @@ -332,6 +332,7 @@ static void reset_connection(struct ceph_connection *con) con->out_msg = NULL; } con->in_seq = 0; + con->in_seq_acked = 0; } /* -- cgit v0.10.2 From 9875557ee8247c3f7390d378c027b45c7535a224 Mon Sep 17 00:00:00 2001 From: Stefan Bader Date: Mon, 29 Mar 2010 17:53:12 +0200 Subject: drm/i915: Add no_lvds entry for the Clientron U800 BugLink: http://bugs.launchpad.net/ubuntu/bugs/544671 This system claims to have a LVDS but has not. Signed-off-by: Stephane Graber Signed-off-by: Stefan Bader CC: stable@kernel.org Signed-off-by: Eric Anholt diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c index 9d99ddc..8238b40 100644 --- a/drivers/gpu/drm/i915/intel_lvds.c +++ b/drivers/gpu/drm/i915/intel_lvds.c @@ -857,6 +857,14 @@ static const struct dmi_system_id intel_no_lvds[] = { DMI_MATCH(DMI_PRODUCT_VERSION, "AO00001JW"), }, }, + { + .callback = intel_no_lvds_dmi_callback, + .ident = "Clientron U800", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "Clientron"), + DMI_MATCH(DMI_PRODUCT_NAME, "U800"), + }, + }, { } /* terminating entry */ }; -- cgit v0.10.2 From e206d556c5793ac5e28c0aaba2e07432e5f9a098 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 3 Apr 2010 10:19:26 -0300 Subject: perf tools: Move the prototypes in util/string.h to util.h MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit So that we avoid conflict with libc's string.h header. Reviewed-by: KOSAKI Motohiro Suggested-by: KOSAKI Motohiro Cc: KOSAKI Motohiro Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/Makefile b/tools/perf/Makefile index 93029c0..9f5a47e 100644 --- a/tools/perf/Makefile +++ b/tools/perf/Makefile @@ -385,7 +385,6 @@ LIB_H += util/header.h LIB_H += util/help.h LIB_H += util/session.h LIB_H += util/strbuf.h -LIB_H += util/string.h LIB_H += util/strlist.h LIB_H += util/svghelper.h LIB_H += util/run-command.h diff --git a/tools/perf/bench/mem-memcpy.c b/tools/perf/bench/mem-memcpy.c index 8977317..52e646e 100644 --- a/tools/perf/bench/mem-memcpy.c +++ b/tools/perf/bench/mem-memcpy.c @@ -10,7 +10,6 @@ #include "../perf.h" #include "../util/util.h" #include "../util/parse-options.h" -#include "../util/string.h" #include "../util/header.h" #include "bench.h" diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c index 887e8e0..ee0d917 100644 --- a/tools/perf/builtin-annotate.c +++ b/tools/perf/builtin-annotate.c @@ -14,7 +14,6 @@ #include "util/cache.h" #include #include "util/symbol.h" -#include "util/string.h" #include "perf.h" #include "util/debug.h" diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index 80dc444..dc61f1b 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -15,7 +15,6 @@ #include "util/util.h" #include "util/parse-options.h" #include "util/parse-events.h" -#include "util/string.h" #include "util/header.h" #include "util/event.h" diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index b13a7e2..6615e09 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -14,7 +14,6 @@ #include "util/cache.h" #include #include "util/symbol.h" -#include "util/string.h" #include "util/callchain.h" #include "util/strlist.h" #include "util/values.h" diff --git a/tools/perf/builtin-timechart.c b/tools/perf/builtin-timechart.c index 0d4d8ff..266e7aa 100644 --- a/tools/perf/builtin-timechart.c +++ b/tools/perf/builtin-timechart.c @@ -21,7 +21,6 @@ #include "util/cache.h" #include #include "util/symbol.h" -#include "util/string.h" #include "util/callchain.h" #include "util/strlist.h" diff --git a/tools/perf/perf.c b/tools/perf/perf.c index 2826e70..d4be55b 100644 --- a/tools/perf/perf.c +++ b/tools/perf/perf.c @@ -13,7 +13,6 @@ #include "util/quote.h" #include "util/run-command.h" #include "util/parse-events.h" -#include "util/string.h" #include "util/debugfs.h" bool use_browser; diff --git a/tools/perf/util/string.h b/tools/perf/util/string.h deleted file mode 100644 index 7005824..0000000 --- a/tools/perf/util/string.h +++ /dev/null @@ -1,16 +0,0 @@ -#ifndef __PERF_STRING_H_ -#define __PERF_STRING_H_ - -#include -#include "types.h" - -s64 perf_atoll(const char *str); -char **argv_split(const char *str, int *argcp); -void argv_free(char **argv); -bool strglobmatch(const char *str, const char *pat); -bool strlazymatch(const char *str, const char *pat); - -#define _STR(x) #x -#define STR(x) _STR(x) - -#endif /* __PERF_STRING_H */ diff --git a/tools/perf/util/util.h b/tools/perf/util/util.h index 5270108..fbf45d1 100644 --- a/tools/perf/util/util.h +++ b/tools/perf/util/util.h @@ -42,12 +42,14 @@ #define _ALL_SOURCE 1 #define _GNU_SOURCE 1 #define _BSD_SOURCE 1 +#define HAS_BOOL #include #include #include #include #include +#include #include #include #include @@ -78,6 +80,7 @@ #include #include #include "../../../include/linux/magic.h" +#include "types.h" #ifndef NO_ICONV @@ -415,4 +418,13 @@ void git_qsort(void *base, size_t nmemb, size_t size, int mkdir_p(char *path, mode_t mode); int copyfile(const char *from, const char *to); +s64 perf_atoll(const char *str); +char **argv_split(const char *str, int *argcp); +void argv_free(char **argv); +bool strglobmatch(const char *str, const char *pat); +bool strlazymatch(const char *str, const char *pat); + +#define _STR(x) #x +#define STR(x) _STR(x) + #endif -- cgit v0.10.2 From 4af8b35db6634dd1e0d616de689582b6c93550af Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Sat, 3 Apr 2010 22:53:31 +1100 Subject: perf symbols: Fill in pgoff in mmap synthesized events When we synthesize mmap events we need to fill in the pgoff field. I wasn't able to test this completely since I couldn't find an executable region with a non 0 offset. We will see it when we start doing data profiling. Signed-off-by: Anton Blanchard Cc: David Miller Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: <20100403115331.GK5594@kryten> Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c index 052eaecc..571fb25 100644 --- a/tools/perf/util/event.c +++ b/tools/perf/util/event.c @@ -130,6 +130,7 @@ static int event__synthesize_mmap_events(pid_t pid, pid_t tgid, continue; pbf += n + 3; if (*pbf == 'x') { /* vm_exec */ + u64 vm_pgoff; char *execname = strchr(bf, '/'); /* Catch VDSO */ @@ -139,6 +140,14 @@ static int event__synthesize_mmap_events(pid_t pid, pid_t tgid, if (execname == NULL) continue; + pbf += 3; + n = hex2u64(pbf, &vm_pgoff); + /* pgoff is in bytes, not pages */ + if (n >= 0) + ev.mmap.pgoff = vm_pgoff << getpagesize(); + else + ev.mmap.pgoff = 0; + size = strlen(execname); execname[size - 1] = '\0'; /* Remove \n */ memcpy(ev.mmap.filename, execname, size); -- cgit v0.10.2 From fb6b893180faec03e1d32149ef5cc412df9714df Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 3 Apr 2010 11:04:55 -0300 Subject: perf newt: Remove useless column width calculation MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Not used in the TUI interface. Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index edd628f..a2ba3ff 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -390,11 +390,8 @@ static void perf_session__selection(newtComponent self, void *data) int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, u64 session_total, const char *helpline) { - struct sort_entry *se; struct rb_node *nd; char seq[] = "."; - unsigned int width; - char *col_width = symbol_conf.col_width_list_str; int rows, cols, idx; int max_len = 0; char str[1024]; @@ -423,23 +420,6 @@ int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, newtComponentAddCallback(tree, perf_session__selection, &selection); - list_for_each_entry(se, &hist_entry__sort_list, list) { - if (se->elide) - continue; - width = strlen(se->header); - if (se->width) { - if (symbol_conf.col_width_list_str) { - if (col_width) { - *se->width = atoi(col_width); - col_width = strchr(col_width, ','); - if (col_width) - ++col_width; - } - } - *se->width = max(*se->width, width); - } - } - idx = 0; for (nd = rb_first(hists); nd; nd = rb_next(nd)) { struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node); -- cgit v0.10.2 From e65713ea1e61e92d28284a55df2aa039ebe10003 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 3 Apr 2010 11:25:56 -0300 Subject: perf newt: Move the hist browser population bits to separare function MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Next patches will use that when applying filtes to then repopulate the browser with the narrowed vision. Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index a2ba3ff..509d921 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -381,53 +381,69 @@ static const void *newt__symbol_tree_get_current(newtComponent self) return newtListboxGetCurrent(self); } -static void perf_session__selection(newtComponent self, void *data) +static void hist_browser__selection(newtComponent self, void *data) { const struct map_symbol **symbol_ptr = data; *symbol_ptr = newt__symbol_tree_get_current(self); } -int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, - u64 session_total, const char *helpline) +struct hist_browser { + newtComponent form, tree; + const struct map_symbol *selection; +}; + +static struct hist_browser *hist_browser__new(void) +{ + struct hist_browser *self = malloc(sizeof(*self)); + + if (self != NULL) { + char seq[] = "."; + int rows; + + newtGetScreenSize(NULL, &rows); + + if (symbol_conf.use_callchain) + self->tree = newtCheckboxTreeMulti(0, 0, rows - 5, seq, + NEWT_FLAG_SCROLL); + else + self->tree = newtListbox(0, 0, rows - 5, + (NEWT_FLAG_SCROLL | + NEWT_FLAG_RETURNEXIT)); + newtComponentAddCallback(self->tree, hist_browser__selection, + &self->selection); + } + + return self; +} + +static void hist_browser__delete(struct hist_browser *self) +{ + newtFormDestroy(self->form); + newtPopWindow(); + free(self); +} + +static int hist_browser__populate(struct hist_browser *self, struct rb_root *hists, + u64 nr_hists, u64 session_total) { + int max_len = 0, idx, cols, rows; + struct ui_progress *progress; struct rb_node *nd; - char seq[] = "."; - int rows, cols, idx; - int max_len = 0; - char str[1024]; - newtComponent form, tree; - struct newtExitStruct es; - const struct map_symbol *selection; u64 curr_hist = 0; - struct ui_progress *progress; progress = ui_progress__new("Adding entries to the browser...", nr_hists); if (progress == NULL) return -1; - snprintf(str, sizeof(str), "Samples: %Ld", session_total); - newtDrawRootText(0, 0, str); - newtPushHelpLine(helpline); - - newtGetScreenSize(&cols, &rows); - - if (symbol_conf.use_callchain) - tree = newtCheckboxTreeMulti(0, 0, rows - 5, seq, - NEWT_FLAG_SCROLL); - else - tree = newtListbox(0, 0, rows - 5, (NEWT_FLAG_SCROLL | - NEWT_FLAG_RETURNEXIT)); - - newtComponentAddCallback(tree, perf_session__selection, &selection); - idx = 0; for (nd = rb_first(hists); nd; nd = rb_next(nd)) { struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node); - int len = hist_entry__append_browser(h, tree, session_total); + int len = hist_entry__append_browser(h, self->tree, session_total); if (len > max_len) max_len = len; if (symbol_conf.use_callchain) - hist_entry__append_callchain_browser(h, tree, session_total, idx++); + hist_entry__append_callchain_browser(h, self->tree, + session_total, idx++); ++curr_hist; if (curr_hist % 5) ui_progress__update(progress, curr_hist); @@ -435,27 +451,50 @@ int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, ui_progress__delete(progress); + newtGetScreenSize(&cols, &rows); + if (max_len > cols) max_len = cols - 3; if (!symbol_conf.use_callchain) - newtListboxSetWidth(tree, max_len); + newtListboxSetWidth(self->tree, max_len); newtCenteredWindow(max_len + (symbol_conf.use_callchain ? 5 : 0), rows - 5, "Report"); - form = newt_form__new(); - newtFormAddHotKey(form, 'A'); - newtFormAddHotKey(form, 'a'); - newtFormAddHotKey(form, NEWT_KEY_RIGHT); - newtFormAddComponents(form, tree, NULL); - selection = newt__symbol_tree_get_current(tree); + self->form = newt_form__new(); + newtFormAddHotKey(self->form, 'A'); + newtFormAddHotKey(self->form, 'a'); + newtFormAddHotKey(self->form, NEWT_KEY_RIGHT); + newtFormAddComponents(self->form, self->tree, NULL); + self->selection = newt__symbol_tree_get_current(self->tree); + + return 0; +} + +int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, + u64 session_total, const char *helpline) +{ + struct newtExitStruct es; + char str[1024]; + int err = -1; + struct hist_browser *browser = hist_browser__new(); + + if (browser == NULL) + return -1; + + snprintf(str, sizeof(str), "Samples: %Ld", session_total); + newtDrawRootText(0, 0, str); + newtPushHelpLine(helpline); + + if (hist_browser__populate(browser, hists, nr_hists, session_total) < 0) + goto out; while (1) { char annotate[512]; const char *options[2]; int nr_options = 0, choice = 0; - newtFormRun(form, &es); + newtFormRun(browser->form, &es); if (es.reason == NEWT_EXIT_HOTKEY) { if (toupper(es.u.key) == 'A') goto do_annotate; @@ -469,9 +508,9 @@ int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, } } - if (selection->sym != NULL) { + if (browser->selection->sym != NULL) { snprintf(annotate, sizeof(annotate), - "Annotate %s", selection->sym->name); + "Annotate %s", browser->selection->sym->name); options[nr_options++] = annotate; } @@ -480,21 +519,21 @@ int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, if (choice == nr_options - 1) break; do_annotate: - if (selection->sym != NULL && choice >= 0) { - if (selection->map->dso->origin == DSO__ORIG_KERNEL) { + if (browser->selection->sym != NULL && choice >= 0) { + if (browser->selection->map->dso->origin == DSO__ORIG_KERNEL) { newtPopHelpLine(); newtPushHelpLine("No vmlinux file found, can't " "annotate with just a " "kallsyms file"); continue; } - map_symbol__annotate_browser(selection); + map_symbol__annotate_browser(browser->selection); } } - - newtFormDestroy(form); - newtPopWindow(); - return 0; + err = 0; +out: + hist_browser__delete(browser); + return err; } void setup_browser(void) -- cgit v0.10.2 From 533c46c31c0e82f19dbb087c77d85eaccd6fefdb Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 3 Apr 2010 11:54:35 -0300 Subject: perf newt: Pass the input_name to perf_session__browse_hists MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit So that it can use it in the 'perf annotate' command line, otherwise it'll use the default and not the specified -i filename passed to 'perf report'. Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 6615e09..e93c69a 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -313,7 +313,8 @@ static int __cmd_report(void) stats->stats.total); if (use_browser) perf_session__browse_hists(&stats->hists, nr_hists, - stats->stats.total, help); + stats->stats.total, help, + input_name); else { if (rb_first(&session->stats_by_id) == rb_last(&session->stats_by_id)) diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index 509d921..c93bc2a 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -317,7 +317,8 @@ static size_t hist_entry__append_browser(struct hist_entry *self, return ret; } -static void map_symbol__annotate_browser(const struct map_symbol *self) +static void map_symbol__annotate_browser(const struct map_symbol *self, + const char *input_name) { FILE *fp; int cols, rows; @@ -331,8 +332,8 @@ static void map_symbol__annotate_browser(const struct map_symbol *self) if (self->sym == NULL) return; - if (asprintf(&str, "perf annotate -d \"%s\" %s 2>&1 | expand", - self->map->dso->name, self->sym->name) < 0) + if (asprintf(&str, "perf annotate -i \"%s\" -d \"%s\" %s 2>&1 | expand", + input_name, self->map->dso->name, self->sym->name) < 0) return; fp = popen(str, "r"); @@ -472,7 +473,8 @@ static int hist_browser__populate(struct hist_browser *self, struct rb_root *his } int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, - u64 session_total, const char *helpline) + u64 session_total, const char *helpline, + const char *input_name) { struct newtExitStruct es; char str[1024]; @@ -527,7 +529,8 @@ do_annotate: "kallsyms file"); continue; } - map_symbol__annotate_browser(browser->selection); + map_symbol__annotate_browser(browser->selection, + input_name); } } err = 0; diff --git a/tools/perf/util/session.h b/tools/perf/util/session.h index dffaff5..27f4c2d 100644 --- a/tools/perf/util/session.h +++ b/tools/perf/util/session.h @@ -96,12 +96,14 @@ static inline struct map * static inline int perf_session__browse_hists(struct rb_root *hists __used, u64 nr_hists __used, u64 session_total __used, - const char *helpline __used) + const char *helpline __used, + const char *input_name __used) { return 0; } #else int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, - u64 session_total, const char *helpline); + u64 session_total, const char *helpline, + const char *input_name); #endif #endif /* __PERF_SESSION_H */ -- cgit v0.10.2 From a24e2d7d8f512340991ef0a59cb5d08d491b8e98 Mon Sep 17 00:00:00 2001 From: Steve French Date: Sat, 3 Apr 2010 17:20:21 +0000 Subject: [CIFS] initialize nbytes at the beginning of CIFSSMBWrite() By doing this we always overwrite nbytes value that is being passed on to CIFSSMBWrite() and need not rely on the callers to initialize. CIFSSMBWrite2 is doing this already. CC: Stable Reviewed-by: Shirish Pargaonkar Reviewed-by: Jeff Layton Signed-off-by: Suresh Jayaraman Signed-off-by: Steve French diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 7cc7f83..e1f90a3 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -1430,6 +1430,8 @@ CIFSSMBWrite(const int xid, struct cifsTconInfo *tcon, __u32 bytes_sent; __u16 byte_count; + *nbytes = 0; + /* cFYI(1, ("write at %lld %d bytes", offset, count));*/ if (tcon->ses == NULL) return -ECONNABORTED; @@ -1512,7 +1514,6 @@ CIFSSMBWrite(const int xid, struct cifsTconInfo *tcon, cifs_stats_inc(&tcon->num_writes); if (rc) { cFYI(1, ("Send error in write = %d", rc)); - *nbytes = 0; } else { *nbytes = le16_to_cpu(pSMBr->CountHigh); *nbytes = (*nbytes) << 16; -- cgit v0.10.2 From 6513a81e9325d712f1bfb9a1d7b750134e49ff18 Mon Sep 17 00:00:00 2001 From: Suresh Jayaraman Date: Wed, 31 Mar 2010 12:00:03 +0530 Subject: cifs: Fix a kernel BUG with remote OS/2 server (try #3) While chasing a bug report involving a OS/2 server, I noticed the server sets pSMBr->CountHigh to a incorrect value even in case of normal writes. This results in 'nbytes' being computed wrongly and triggers a kernel BUG at mm/filemap.c. void iov_iter_advance(struct iov_iter *i, size_t bytes) { BUG_ON(i->count < bytes); <--- BUG here Why the server is setting 'CountHigh' is not clear but only does so after writing 64k bytes. Though this looks like the server bug, the client side crash may not be acceptable. The workaround is to mask off high 16 bits if the number of bytes written as returned by the server is greater than the bytes requested by the client as suggested by Jeff Layton. CC: Stable Reviewed-by: Jeff Layton Signed-off-by: Suresh Jayaraman Signed-off-by: Steve French diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index e1f90a3..f213b8a 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -1518,6 +1518,14 @@ CIFSSMBWrite(const int xid, struct cifsTconInfo *tcon, *nbytes = le16_to_cpu(pSMBr->CountHigh); *nbytes = (*nbytes) << 16; *nbytes += le16_to_cpu(pSMBr->Count); + + /* + * Mask off high 16 bits when bytes written as returned by the + * server is greater than bytes requested by the client. Some + * OS/2 servers are known to set incorrect CountHigh values. + */ + if (*nbytes > count) + *nbytes &= 0xFFFF; } cifs_buf_release(pSMB); @@ -1606,6 +1614,14 @@ CIFSSMBWrite2(const int xid, struct cifsTconInfo *tcon, *nbytes = le16_to_cpu(pSMBr->CountHigh); *nbytes = (*nbytes) << 16; *nbytes += le16_to_cpu(pSMBr->Count); + + /* + * Mask off high 16 bits when bytes written as returned by the + * server is greater than bytes requested by the client. OS/2 + * servers are known to set incorrect CountHigh values. + */ + if (*nbytes > count) + *nbytes &= 0xFFFF; } /* cifs_small_buf_release(pSMB); */ /* Freed earlier now in SendReceive2 */ -- cgit v0.10.2 From 8b472d739b2ddd8ab7fb278874f696cd95b25a5e Mon Sep 17 00:00:00 2001 From: Curt Wohlgemuth Date: Sat, 3 Apr 2010 16:45:06 -0400 Subject: ext4: Fix possible lost inode write in no journal mode In the no-journal case, ext4_write_inode() will fetch the bh and call sync_dirty_buffer() on it. However, if the bh has already been written and the bh reclaimed for some other purpose, AND if the inode is the only one in the inode table block in use, then ext4_get_inode_loc() will not read the inode table block from disk, but as an optimization, fill the block with zero's assuming that its caller will copy in the on-disk version of the inode. This is not done by ext4_write_inode(), so the contents of the inode can simply get lost. The fix is to use __ext4_get_inode_loc() with in_mem set to 0, instead of ext4_get_inode_loc(). Long term the API needs to be fixed so it's obvious why latter is not safe. Addresses-Google-Bug: #2526446 Signed-off-by: Curt Wohlgemuth Signed-off-by: "Theodore Ts'o" diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 11119e0..87e3c70 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -5374,7 +5374,7 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc) } else { struct ext4_iloc iloc; - err = ext4_get_inode_loc(inode, &iloc); + err = __ext4_get_inode_loc(inode, &iloc, 0); if (err) return err; if (wbc->sync_mode == WB_SYNC_ALL) -- cgit v0.10.2 From fd2dd9fbaf9e498ec63eef298921e36556f7214c Mon Sep 17 00:00:00 2001 From: Curt Wohlgemuth Date: Sat, 3 Apr 2010 17:44:16 -0400 Subject: ext4: Fix buffer head leaks after calls to ext4_get_inode_loc() Calls to ext4_get_inode_loc() returns with a reference to a buffer head in iloc->bh. The callers of this function in ext4_write_inode() when in no journal mode and in ext4_xattr_fiemap() don't release the buffer head after using it. Addresses-Google-Bug: #2548165 Signed-off-by: Curt Wohlgemuth Signed-off-by: "Theodore Ts'o" diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 94c8ee8..236b834 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -3879,6 +3879,7 @@ static int ext4_xattr_fiemap(struct inode *inode, physical += offset; length = EXT4_SB(inode->i_sb)->s_inode_size - offset; flags |= FIEMAP_EXTENT_DATA_INLINE; + brelse(iloc.bh); } else { /* external block */ physical = EXT4_I(inode)->i_file_acl << blockbits; length = inode->i_sb->s_blocksize; diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 87e3c70..ba1eee8 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -5385,6 +5385,7 @@ int ext4_write_inode(struct inode *inode, struct writeback_control *wbc) (unsigned long long)iloc.bh->b_blocknr); err = -EIO; } + brelse(iloc.bh); } return err; } -- cgit v0.10.2 From 83753190c136901c916df267703937e60f24b8b8 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 3 Apr 2010 16:30:44 -0300 Subject: perf newt: Add a "Zoom into foo.so DSO" and reverse operations MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Clicking on -> will bring as one of the popup menu options a "Zoom into CURRENT DSO", i.e. CURRENT will be replaced by the name of the DSO in the current line. Choosing this option will filter out all samples that didn't took place in a symbol in this DSO. After that the option reverts to "Zoom out of CURRENT DSO", to allow going back to the more compreensive view, not filtered by DSO. Future similar operations will include zooming into a particular thread, COMM, CPU, "last minute", "last N usecs", etc. Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index c93bc2a..bbf725d 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -94,7 +94,7 @@ static newtComponent newt_form__new(void) return self; } -static int popup_menu(int argc, const char *argv[]) +static int popup_menu(int argc, char * const argv[]) { struct newtExitStruct es; int i, rc = -1, max_len = 5; @@ -397,22 +397,8 @@ static struct hist_browser *hist_browser__new(void) { struct hist_browser *self = malloc(sizeof(*self)); - if (self != NULL) { - char seq[] = "."; - int rows; - - newtGetScreenSize(NULL, &rows); - - if (symbol_conf.use_callchain) - self->tree = newtCheckboxTreeMulti(0, 0, rows - 5, seq, - NEWT_FLAG_SCROLL); - else - self->tree = newtListbox(0, 0, rows - 5, - (NEWT_FLAG_SCROLL | - NEWT_FLAG_RETURNEXIT)); - newtComponentAddCallback(self->tree, hist_browser__selection, - &self->selection); - } + if (self != NULL) + self->form = NULL; return self; } @@ -431,6 +417,30 @@ static int hist_browser__populate(struct hist_browser *self, struct rb_root *his struct ui_progress *progress; struct rb_node *nd; u64 curr_hist = 0; + char seq[] = "."; + char str[256]; + + if (self->form) { + newtFormDestroy(self->form); + newtPopWindow(); + } + + snprintf(str, sizeof(str), "Samples: %Ld ", + session_total); + newtDrawRootText(0, 0, str); + + newtGetScreenSize(NULL, &rows); + + if (symbol_conf.use_callchain) + self->tree = newtCheckboxTreeMulti(0, 0, rows - 5, seq, + NEWT_FLAG_SCROLL); + else + self->tree = newtListbox(0, 0, rows - 5, + (NEWT_FLAG_SCROLL | + NEWT_FLAG_RETURNEXIT)); + + newtComponentAddCallback(self->tree, hist_browser__selection, + &self->selection); progress = ui_progress__new("Adding entries to the browser...", nr_hists); if (progress == NULL) @@ -439,7 +449,12 @@ static int hist_browser__populate(struct hist_browser *self, struct rb_root *his idx = 0; for (nd = rb_first(hists); nd; nd = rb_next(nd)) { struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node); - int len = hist_entry__append_browser(h, self->tree, session_total); + int len; + + if (h->filtered) + continue; + + len = hist_entry__append_browser(h, self->tree, session_total); if (len > max_len) max_len = len; if (symbol_conf.use_callchain) @@ -463,6 +478,9 @@ static int hist_browser__populate(struct hist_browser *self, struct rb_root *his newtCenteredWindow(max_len + (symbol_conf.use_callchain ? 5 : 0), rows - 5, "Report"); self->form = newt_form__new(); + if (self->form == NULL) + return -1; + newtFormAddHotKey(self->form, 'A'); newtFormAddHotKey(self->form, 'a'); newtFormAddHotKey(self->form, NEWT_KEY_RIGHT); @@ -472,29 +490,50 @@ static int hist_browser__populate(struct hist_browser *self, struct rb_root *his return 0; } +static u64 hists__filter_by_dso(struct rb_root *hists, struct dso *dso, + u64 *session_total) +{ + struct rb_node *nd; + u64 nr_hists = 0; + + *session_total = 0; + + for (nd = rb_first(hists); nd; nd = rb_next(nd)) { + struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node); + + if (dso != NULL && (h->ms.map == NULL || h->ms.map->dso != dso)) { + h->filtered = true; + continue; + } + h->filtered = false; + ++nr_hists; + *session_total += h->count; + } + + return nr_hists; +} + int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, u64 session_total, const char *helpline, const char *input_name) { struct newtExitStruct es; - char str[1024]; + bool dso_filtered = false; int err = -1; struct hist_browser *browser = hist_browser__new(); if (browser == NULL) return -1; - snprintf(str, sizeof(str), "Samples: %Ld", session_total); - newtDrawRootText(0, 0, str); newtPushHelpLine(helpline); if (hist_browser__populate(browser, hists, nr_hists, session_total) < 0) goto out; while (1) { - char annotate[512]; - const char *options[2]; - int nr_options = 0, choice = 0; + char *options[16]; + int nr_options = 0, choice = 0, i, + annotate = -2, zoom_dso = -2; newtFormRun(browser->form, &es); if (es.reason == NEWT_EXIT_HOTKEY) { @@ -510,18 +549,29 @@ int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, } } - if (browser->selection->sym != NULL) { - snprintf(annotate, sizeof(annotate), - "Annotate %s", browser->selection->sym->name); - options[nr_options++] = annotate; - } + if (browser->selection->sym != NULL && + asprintf(&options[nr_options], "Annotate %s", + browser->selection->sym->name) > 0) + annotate = nr_options++; + + if (browser->selection->map != NULL && + asprintf(&options[nr_options], "Zoom %s %s DSO", + dso_filtered ? "out of" : "into", + (browser->selection->map->dso->kernel ? "the Kernel" : + browser->selection->map->dso->short_name)) > 0) + zoom_dso = nr_options++; + + options[nr_options++] = (char *)"Exit"; - options[nr_options++] = "Exit"; choice = popup_menu(nr_options, options); + + for (i = 0; i < nr_options - 1; ++i) + free(options[i]); + if (choice == nr_options - 1) break; do_annotate: - if (browser->selection->sym != NULL && choice >= 0) { + if (choice == annotate) { if (browser->selection->map->dso->origin == DSO__ORIG_KERNEL) { newtPopHelpLine(); newtPushHelpLine("No vmlinux file found, can't " @@ -531,6 +581,13 @@ do_annotate: } map_symbol__annotate_browser(browser->selection, input_name); + } if (choice == zoom_dso) { + hists__filter_by_dso(hists, + dso_filtered ? NULL : browser->selection->map->dso, + &session_total); + dso_filtered = !dso_filtered; + if (hist_browser__populate(browser, hists, nr_hists, session_total) < 0) + goto out; } } err = 0; diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h index 5bf2b74..dce79d3 100644 --- a/tools/perf/util/sort.h +++ b/tools/perf/util/sort.h @@ -48,7 +48,8 @@ struct hist_entry { struct map_symbol ms; u64 ip; char level; - struct symbol *parent; + bool filtered; + struct symbol *parent; union { unsigned long position; struct hist_entry *pair; -- cgit v0.10.2 From a5e29aca02fcecd086ac160ea29244cae6b4305e Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Sat, 3 Apr 2010 22:44:37 -0300 Subject: perf TUI: Add a "Zoom into COMM(PID) thread" and reverse operations MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Now one can press the right arrow key and in addition to being able to filter by DSO, filter out by thread too, or a combination of both filters. With this one can start collecting events for the whole system, then focus on a subset of the collected data quickly. Cc: Avi Kivity Cc: Frédéric Weisbecker Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras LKML-Reference: Signed-off-by: Arnaldo Carvalho de Melo diff --git a/tools/perf/util/newt.c b/tools/perf/util/newt.c index bbf725d..6d6e022 100644 --- a/tools/perf/util/newt.c +++ b/tools/perf/util/newt.c @@ -490,6 +490,11 @@ static int hist_browser__populate(struct hist_browser *self, struct rb_root *his return 0; } +enum hist_filter { + HIST_FILTER__DSO, + HIST_FILTER__THREAD, +}; + static u64 hists__filter_by_dso(struct rb_root *hists, struct dso *dso, u64 *session_total) { @@ -502,10 +507,10 @@ static u64 hists__filter_by_dso(struct rb_root *hists, struct dso *dso, struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node); if (dso != NULL && (h->ms.map == NULL || h->ms.map->dso != dso)) { - h->filtered = true; + h->filtered |= (1 << HIST_FILTER__DSO); continue; } - h->filtered = false; + h->filtered &= ~(1 << HIST_FILTER__DSO); ++nr_hists; *session_total += h->count; } @@ -513,12 +518,54 @@ static u64 hists__filter_by_dso(struct rb_root *hists, struct dso *dso, return nr_hists; } +static u64 hists__filter_by_thread(struct rb_root *hists, const struct thread *thread, + u64 *session_total) +{ + struct rb_node *nd; + u64 nr_hists = 0; + + *session_total = 0; + + for (nd = rb_first(hists); nd; nd = rb_next(nd)) { + struct hist_entry *h = rb_entry(nd, struct hist_entry, rb_node); + + if (thread != NULL && h->thread != thread) { + h->filtered |= (1 << HIST_FILTER__THREAD); + continue; + } + h->filtered &= ~(1 << HIST_FILTER__THREAD); + ++nr_hists; + *session_total += h->count; + } + + return nr_hists; +} + +static struct thread *hist_browser__selected_thread(struct hist_browser *self) +{ + int *indexes; + + if (!symbol_conf.use_callchain) + goto out; + + indexes = newtCheckboxTreeFindItem(self->tree, (void *)self->selection); + if (indexes) { + bool is_hist_entry = indexes[1] == NEWT_ARG_LAST; + free(indexes); + if (is_hist_entry) + goto out; + } + return NULL; +out: + return *(struct thread **)(self->selection + 1); +} + int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, u64 session_total, const char *helpline, const char *input_name) { struct newtExitStruct es; - bool dso_filtered = false; + bool dso_filtered = false, thread_filtered = false; int err = -1; struct hist_browser *browser = hist_browser__new(); @@ -531,9 +578,10 @@ int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, goto out; while (1) { + const struct thread *thread; char *options[16]; int nr_options = 0, choice = 0, i, - annotate = -2, zoom_dso = -2; + annotate = -2, zoom_dso = -2, zoom_thread = -2; newtFormRun(browser->form, &es); if (es.reason == NEWT_EXIT_HOTKEY) { @@ -561,6 +609,13 @@ int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, browser->selection->map->dso->short_name)) > 0) zoom_dso = nr_options++; + thread = hist_browser__selected_thread(browser); + if (thread != NULL && + asprintf(&options[nr_options], "Zoom %s %s(%d) thread", + (thread_filtered ? "out of" : "into"), + (thread->comm_set ? thread->comm : ""), thread->pid) > 0) + zoom_thread = nr_options++; + options[nr_options++] = (char *)"Exit"; choice = popup_menu(nr_options, options); @@ -570,6 +625,9 @@ int perf_session__browse_hists(struct rb_root *hists, u64 nr_hists, if (choice == nr_options - 1) break; + + if (choice == -1) + continue; do_annotate: if (choice == annotate) { if (browser->selection->map->dso->origin == DSO__ORIG_KERNEL) { @@ -581,13 +639,21 @@ do_annotate: } map_symbol__annotate_browser(browser->selection, input_name); - } if (choice == zoom_dso) { - hists__filter_by_dso(hists, - dso_filtered ? NULL : browser->selection->map->dso, - &session_total); + } else if (choice == zoom_dso) { + nr_hists = hists__filter_by_dso(hists, + (dso_filtered ? NULL : + browser->selection->map->dso), + &session_total); dso_filtered = !dso_filtered; if (hist_browser__populate(browser, hists, nr_hists, session_total) < 0) goto out; + } else if (choice == zoom_thread) { + nr_hists = hists__filter_by_thread(hists, + (thread_filtered ? NULL : thread), + &session_total); + thread_filtered = !thread_filtered; + if (hist_browser__populate(browser, hists, nr_hists, session_total) < 0) + goto out; } } err = 0; diff --git a/tools/perf/util/sort.h b/tools/perf/util/sort.h index dce79d3..6d7b4be 100644 --- a/tools/perf/util/sort.h +++ b/tools/perf/util/sort.h @@ -44,11 +44,16 @@ extern enum sort_type sort__first_dimension; struct hist_entry { struct rb_node rb_node; u64 count; - struct thread *thread; + /* + * XXX WARNING! + * thread _has_ to come after ms, see + * hist_browser__selected_thread in util/newt.c + */ struct map_symbol ms; + struct thread *thread; u64 ip; char level; - bool filtered; + u8 filtered; struct symbol *parent; union { unsigned long position; -- cgit v0.10.2 From b7b30de53aef6ce773d34837ba7d8422bd3baeec Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Wed, 24 Mar 2010 10:44:33 -0600 Subject: ACPI: use _HID when supplied by root-level devices Previously, we assumed the only Device object immediately below the root was the \_SB Scope (which the ACPI CA treats as a Device), so we forced the HID of all such objects to ACPI_BUS_HID ("LNXSYBUS"). However, there are DSDTs that supply root-level Device objects with _HIDs. This patch makes us pay attention to those _HIDs and only add the synthetic ACPI_BUS_HID for root-level objects that do not supply their own _HID. For example, this DSDT: https://bugzilla.kernel.org/show_bug.cgi?id=15605 contains: Scope (_SB) { ... } Device (AMW0) { Name (_HID, EisaId ("PNP0C14")) ... } and we should use "PNP0C14" for the AMW0 device, not "LNXSYBUS". Signed-off-by: Bjorn Helgaas Acked-by: Zhang Rui Tested-by: Yong Wang Signed-off-by: Len Brown diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c index 189cbc2..95c90ff 100644 --- a/drivers/acpi/scan.c +++ b/drivers/acpi/scan.c @@ -1080,12 +1080,6 @@ static void acpi_device_set_id(struct acpi_device *device) if (ACPI_IS_ROOT_DEVICE(device)) { acpi_add_id(device, ACPI_SYSTEM_HID); break; - } else if (ACPI_IS_ROOT_DEVICE(device->parent)) { - /* \_SB_, the only root-level namespace device */ - acpi_add_id(device, ACPI_BUS_HID); - strcpy(device->pnp.device_name, ACPI_BUS_DEVICE_NAME); - strcpy(device->pnp.device_class, ACPI_BUS_CLASS); - break; } status = acpi_get_object_info(device->handle, &info); @@ -1120,6 +1114,12 @@ static void acpi_device_set_id(struct acpi_device *device) acpi_add_id(device, ACPI_DOCK_HID); else if (!acpi_ibm_smbus_match(device)) acpi_add_id(device, ACPI_SMBUS_IBM_HID); + else if (!acpi_device_hid(device) && + ACPI_IS_ROOT_DEVICE(device->parent)) { + acpi_add_id(device, ACPI_BUS_HID); /* \_SB, LNXSYBUS */ + strcpy(device->pnp.device_name, ACPI_BUS_DEVICE_NAME); + strcpy(device->pnp.device_class, ACPI_BUS_CLASS); + } break; case ACPI_BUS_TYPE_POWER: -- cgit v0.10.2 From 0f9b75ef3722814134f307f51c19e0791da40e69 Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Thu, 25 Mar 2010 16:33:04 -0700 Subject: ACPI: NUMA: map pxms to low node ids pxms are mapped to low node ids to maintain generic kernel use of functions such as pxm_to_node() that are used to determine device affinity. Otherwise, there is no pxm-to-node and node-to-pxm matching rule for x86_64 users of NUMA emulation where a single pxm may be bound to multiple NUMA nodes. Signed-off-by: David Rientjes Signed-off-by: Len Brown diff --git a/drivers/acpi/numa.c b/drivers/acpi/numa.c index b872546..b0337d3 100644 --- a/drivers/acpi/numa.c +++ b/drivers/acpi/numa.c @@ -61,8 +61,10 @@ int node_to_pxm(int node) void __acpi_map_pxm_to_node(int pxm, int node) { - pxm_to_node_map[pxm] = node; - node_to_pxm_map[node] = pxm; + if (pxm_to_node_map[pxm] == NUMA_NO_NODE || node < pxm_to_node_map[pxm]) + pxm_to_node_map[pxm] = node; + if (node_to_pxm_map[node] == PXM_INVAL || pxm < node_to_pxm_map[node]) + node_to_pxm_map[node] = pxm; } int acpi_map_pxm_to_node(int pxm) -- cgit v0.10.2 From 17c452f99cf5e073b219a069a419b7b8ab3cfa97 Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Fri, 11 Dec 2009 17:40:46 -0500 Subject: ACPI: Don't send KEY_UNKNOWN for random video notifications I have a machine here that's sending 0xD1 notifications on the video device once every second or so. I have no idea why (it's a prototype, it may be broken), but sending KEY_UNKNOWN is unhelpful and results in the console becoming unusable. Let's not report keys unless we have something useful to say about them. Signed-off-by: Matthew Garrett Acked-by: Zhang Rui Signed-off-by: Len Brown diff --git a/drivers/acpi/video.c b/drivers/acpi/video.c index cbe6f39..2c7ca7a 100644 --- a/drivers/acpi/video.c +++ b/drivers/acpi/video.c @@ -2122,7 +2122,7 @@ static void acpi_video_bus_notify(struct acpi_device *device, u32 event) { struct acpi_video_bus *video = acpi_driver_data(device); struct input_dev *input; - int keycode; + int keycode = 0; if (!video) return; @@ -2158,17 +2158,19 @@ static void acpi_video_bus_notify(struct acpi_device *device, u32 event) break; default: - keycode = KEY_UNKNOWN; ACPI_DEBUG_PRINT((ACPI_DB_INFO, "Unsupported event [0x%x]\n", event)); break; } acpi_notifier_call_chain(device, event, 0); - input_report_key(input, keycode, 1); - input_sync(input); - input_report_key(input, keycode, 0); - input_sync(input); + + if (keycode) { + input_report_key(input, keycode, 1); + input_sync(input); + input_report_key(input, keycode, 0); + input_sync(input); + } return; } @@ -2179,7 +2181,7 @@ static void acpi_video_device_notify(acpi_handle handle, u32 event, void *data) struct acpi_device *device = NULL; struct acpi_video_bus *bus; struct input_dev *input; - int keycode; + int keycode = 0; if (!video_device) return; @@ -2220,17 +2222,19 @@ static void acpi_video_device_notify(acpi_handle handle, u32 event, void *data) keycode = KEY_DISPLAY_OFF; break; default: - keycode = KEY_UNKNOWN; ACPI_DEBUG_PRINT((ACPI_DB_INFO, "Unsupported event [0x%x]\n", event)); break; } acpi_notifier_call_chain(device, event, 0); - input_report_key(input, keycode, 1); - input_sync(input); - input_report_key(input, keycode, 0); - input_sync(input); + + if (keycode) { + input_report_key(input, keycode, 1); + input_sync(input); + input_report_key(input, keycode, 0); + input_sync(input); + } return; } @@ -2357,7 +2361,6 @@ static int acpi_video_bus_add(struct acpi_device *device) set_bit(KEY_BRIGHTNESSDOWN, input->keybit); set_bit(KEY_BRIGHTNESS_ZERO, input->keybit); set_bit(KEY_DISPLAY_OFF, input->keybit); - set_bit(KEY_UNKNOWN, input->keybit); error = input_register_device(input); if (error) -- cgit v0.10.2 From 3162b6f0c5e1fcad372d64194fb3cb968941b428 Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Thu, 25 Mar 2010 10:32:49 -0600 Subject: PNPACPI: truncate _CRS windows with _LEN > _MAX - _MIN + 1 The ACPI spec (sec 6.4.3.5 in v4.0) requires that for Address Space Resource Descriptors, _LEN <= _MAX - _MIN + 1 in all cases, but there are BIOSes that violate this. We experimentally determined that Windows truncates the resource so it doesn't extend past _MAX, so let's do the same thing in Linux. http://bugzilla.kernel.org/show_bug.cgi?id=15480 Signed-off-by: Bjorn Helgaas Acked-by: Rafael J. Wysocki Signed-off-by: Len Brown diff --git a/drivers/pnp/pnpacpi/rsparser.c b/drivers/pnp/pnpacpi/rsparser.c index 54514aa..fe5bfa9 100644 --- a/drivers/pnp/pnpacpi/rsparser.c +++ b/drivers/pnp/pnpacpi/rsparser.c @@ -273,12 +273,33 @@ static void pnpacpi_parse_allocated_busresource(struct pnp_dev *dev, pnp_add_bus_resource(dev, start, end); } +static u64 addr_space_length(struct pnp_dev *dev, u64 min, u64 max, u64 len) +{ + u64 max_len; + + max_len = max - min + 1; + if (len <= max_len) + return len; + + /* + * Per 6.4.3.5, _LEN cannot exceed _MAX - _MIN + 1, but some BIOSes + * don't do this correctly, e.g., + * https://bugzilla.kernel.org/show_bug.cgi?id=15480 + */ + dev_info(&dev->dev, + "resource length %#llx doesn't fit in %#llx-%#llx, trimming\n", + (unsigned long long) len, (unsigned long long) min, + (unsigned long long) max); + return max_len; +} + static void pnpacpi_parse_allocated_address_space(struct pnp_dev *dev, struct acpi_resource *res) { struct acpi_resource_address64 addr, *p = &addr; acpi_status status; int window; + u64 len; status = acpi_resource_to_address64(res, p); if (!ACPI_SUCCESS(status)) { @@ -287,20 +308,18 @@ static void pnpacpi_parse_allocated_address_space(struct pnp_dev *dev, return; } + len = addr_space_length(dev, p->minimum, p->maximum, p->address_length); window = (p->producer_consumer == ACPI_PRODUCER) ? 1 : 0; if (p->resource_type == ACPI_MEMORY_RANGE) - pnpacpi_parse_allocated_memresource(dev, - p->minimum, p->address_length, + pnpacpi_parse_allocated_memresource(dev, p->minimum, len, p->info.mem.write_protect, window); else if (p->resource_type == ACPI_IO_RANGE) - pnpacpi_parse_allocated_ioresource(dev, - p->minimum, p->address_length, + pnpacpi_parse_allocated_ioresource(dev, p->minimum, len, p->granularity == 0xfff ? ACPI_DECODE_10 : ACPI_DECODE_16, window); else if (p->resource_type == ACPI_BUS_NUMBER_RANGE) - pnpacpi_parse_allocated_busresource(dev, p->minimum, - p->address_length); + pnpacpi_parse_allocated_busresource(dev, p->minimum, len); } static void pnpacpi_parse_allocated_ext_address_space(struct pnp_dev *dev, @@ -308,21 +327,20 @@ static void pnpacpi_parse_allocated_ext_address_space(struct pnp_dev *dev, { struct acpi_resource_extended_address64 *p = &res->data.ext_address64; int window; + u64 len; + len = addr_space_length(dev, p->minimum, p->maximum, p->address_length); window = (p->producer_consumer == ACPI_PRODUCER) ? 1 : 0; if (p->resource_type == ACPI_MEMORY_RANGE) - pnpacpi_parse_allocated_memresource(dev, - p->minimum, p->address_length, + pnpacpi_parse_allocated_memresource(dev, p->minimum, len, p->info.mem.write_protect, window); else if (p->resource_type == ACPI_IO_RANGE) - pnpacpi_parse_allocated_ioresource(dev, - p->minimum, p->address_length, + pnpacpi_parse_allocated_ioresource(dev, p->minimum, len, p->granularity == 0xfff ? ACPI_DECODE_10 : ACPI_DECODE_16, window); else if (p->resource_type == ACPI_BUS_NUMBER_RANGE) - pnpacpi_parse_allocated_busresource(dev, p->minimum, - p->address_length); + pnpacpi_parse_allocated_busresource(dev, p->minimum, len); } static acpi_status pnpacpi_allocated_resource(struct acpi_resource *res, -- cgit v0.10.2 From 3815595e78d2baae6feb866e737f92d8ef48b337 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Sun, 4 Apr 2010 12:14:03 +0200 Subject: ALSA: hda - Add MSI blacklist for Aopen MZ915-M The device needs MSI disablement. Added to the quirk list. Reported-by: Harald Dunkel Cc: Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c index 4bb9067..f8fd586 100644 --- a/sound/pci/hda/hda_intel.c +++ b/sound/pci/hda/hda_intel.c @@ -2362,6 +2362,7 @@ static struct snd_pci_quirk msi_black_list[] __devinitdata = { SND_PCI_QUIRK(0x1043, 0x81f6, "ASUS", 0), /* nvidia */ SND_PCI_QUIRK(0x1043, 0x822d, "ASUS", 0), /* Athlon64 X2 + nvidia MCP55 */ SND_PCI_QUIRK(0x1849, 0x0888, "ASRock", 0), /* Athlon64 X2 + nvidia */ + SND_PCI_QUIRK(0xa0a0, 0x0575, "Aopen MZ915-M", 0), /* ICH6 */ {} }; -- cgit v0.10.2 From f11947c7c5b8abffd328739996dfdffef2b3e03f Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Fri, 2 Apr 2010 14:29:23 +0300 Subject: ALSA: i2c: cleanup: change parameter to pointer We actually pass an array of 7 chars not 5. This silences a smatch warning. Signed-off-by: Dan Carpenter Signed-off-by: Takashi Iwai diff --git a/include/sound/ak4113.h b/include/sound/ak4113.h index 8988eda..2609048 100644 --- a/include/sound/ak4113.h +++ b/include/sound/ak4113.h @@ -307,7 +307,7 @@ struct ak4113 { int snd_ak4113_create(struct snd_card *card, ak4113_read_t *read, ak4113_write_t *write, - const unsigned char pgm[AK4113_WRITABLE_REGS], + const unsigned char *pgm, void *private_data, struct ak4113 **r_ak4113); void snd_ak4113_reg_write(struct ak4113 *ak4113, unsigned char reg, unsigned char mask, unsigned char val); diff --git a/sound/i2c/other/ak4113.c b/sound/i2c/other/ak4113.c index fff62cc..971a84a 100644 --- a/sound/i2c/other/ak4113.c +++ b/sound/i2c/other/ak4113.c @@ -70,7 +70,7 @@ static int snd_ak4113_dev_free(struct snd_device *device) } int snd_ak4113_create(struct snd_card *card, ak4113_read_t *read, - ak4113_write_t *write, const unsigned char pgm[5], + ak4113_write_t *write, const unsigned char *pgm, void *private_data, struct ak4113 **r_ak4113) { struct ak4113 *chip; -- cgit v0.10.2 From a0fd4345f928d72a56e27b23e4cd28c94bf36be5 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Fri, 2 Apr 2010 14:47:59 +0200 Subject: ALSA: echoaudio - Eliminate use after free Use the call to snd_card_free in the error handling code at the end of the function, as in the other error cases. A simplified version of the semantic patch that finds this problem is as follows: (http://coccinelle.lip6.fr/) // @@ expression E,E2; @@ snd_card_free(E) ... ( E = E2 | * E ) // Signed-off-by: Julia Lawall Signed-off-by: Takashi Iwai diff --git a/sound/pci/echoaudio/echoaudio.c b/sound/pci/echoaudio/echoaudio.c index 8dab82d..668a5ec 100644 --- a/sound/pci/echoaudio/echoaudio.c +++ b/sound/pci/echoaudio/echoaudio.c @@ -2184,10 +2184,9 @@ static int __devinit snd_echo_probe(struct pci_dev *pci, goto ctl_error; #endif - if ((err = snd_card_register(card)) < 0) { - snd_card_free(card); + err = snd_card_register(card); + if (err < 0) goto ctl_error; - } snd_printk(KERN_INFO "Card registered: %s\n", card->longname); pci_set_drvdata(pci, chip); -- cgit v0.10.2 From 6f4dee06fbf0133917f3d76fa3fb50e18b10c1f5 Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Thu, 18 Mar 2010 23:47:01 +0100 Subject: perf: Drop the frame reliablity check It is useless now that we have a pure stack frame walker, as given addr are always reliable. Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 65e9c5e..353a174 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -1602,8 +1602,7 @@ static void backtrace_address(void *data, unsigned long addr, int reliable) { struct perf_callchain_entry *entry = data; - if (reliable) - callchain_store(entry, addr); + callchain_store(entry, addr); } static const struct stacktrace_ops backtrace_ops = { -- cgit v0.10.2 From 6cc8a7c1d8560c042f486b23318a6291569ab96b Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Fri, 19 Mar 2010 01:23:53 +0100 Subject: perf: Fetch hot regs from the template caller Trace events can be defined from a template using DECLARE_EVENT_CLASS/DEFINE_EVENT or directly with TRACE_EVENT. In both cases we have a template tracepoint handler, used to record the trace, to which we pass our ftrace event instance. In the function level, if the class is named "foo" and the event is named "blah", we have the following chain of calls: perf_trace_blah() -> perf_trace_templ_foo() In the case we have several events sharing the class "blah", we'll have multiple users of perf_trace_templ_foo(), and it won't be inlined by the compiler. This is usually what happens with the DECLARE_EVENT_CLASS/DEFINE_EVENT based definition. But if perf_trace_blah() is the only caller of perf_trace_templ_foo() there are fair chances that it will be inlined. The problem is that we fetch the regs from perf_trace_templ_foo() after we rewinded the frame pointer to the second caller, we want to reach the caller of perf_trace_blah() to get the right source of the event. And we do this by always assuming that perf_trace_templ_foo() is not inlined. But as shown above this is not always true. And if it is inlined we miss the first caller, losing the most important level of precision. We get: 61.31% ls [kernel.kallsyms] [k] do_softirq | --- do_softirq irq_exit do_IRQ common_interrupt | |--25.00%-- tty_buffer_request_room Instead of: 61.31% ls [kernel.kallsyms] [k] __do_softirq | --- __do_softirq do_softirq irq_exit do_IRQ common_interrupt | |--25.00%-- tty_buffer_request_room To fix this, we fetch the regs from perf_trace_blah() rather than perf_trace_templ_foo() so that we don't have to deal with inlining surprises. That also bring us the advantage of having the true source of the event even if we don't have frame pointers. Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Ingo Molnar diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index ea6f9d4..882c648 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -758,13 +758,12 @@ __attribute__((section("_ftrace_events"))) event_##call = { \ #define DECLARE_EVENT_CLASS(call, proto, args, tstruct, assign, print) \ static notrace void \ perf_trace_templ_##call(struct ftrace_event_call *event_call, \ - proto) \ + struct pt_regs *__regs, proto) \ { \ struct ftrace_data_offsets_##call __maybe_unused __data_offsets;\ struct ftrace_raw_##call *entry; \ u64 __addr = 0, __count = 1; \ unsigned long irq_flags; \ - struct pt_regs *__regs; \ int __entry_size; \ int __data_size; \ int rctx; \ @@ -785,20 +784,22 @@ perf_trace_templ_##call(struct ftrace_event_call *event_call, \ \ { assign; } \ \ - __regs = &__get_cpu_var(perf_trace_regs); \ - perf_fetch_caller_regs(__regs, 2); \ - \ perf_trace_buf_submit(entry, __entry_size, rctx, __addr, \ __count, irq_flags, __regs); \ } #undef DEFINE_EVENT -#define DEFINE_EVENT(template, call, proto, args) \ -static notrace void perf_trace_##call(proto) \ -{ \ - struct ftrace_event_call *event_call = &event_##call; \ - \ - perf_trace_templ_##template(event_call, args); \ +#define DEFINE_EVENT(template, call, proto, args) \ +static notrace void perf_trace_##call(proto) \ +{ \ + struct ftrace_event_call *event_call = &event_##call; \ + struct pt_regs *__regs = &get_cpu_var(perf_trace_regs); \ + \ + perf_fetch_caller_regs(__regs, 1); \ + \ + perf_trace_templ_##template(event_call, __regs, args); \ + \ + put_cpu_var(perf_trace_regs); \ } #undef DEFINE_EVENT_PRINT -- cgit v0.10.2 From 3326c1ceee234e63160852720d48be8a8f7a6d08 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 23 Mar 2010 19:09:33 +0100 Subject: perf_event: Make perf fd non seekable Perf_event does not need seeking, so prevent it in order to get rid of default_llseek, which uses the BKL. Signed-off-by: Arnd Bergmann Cc: Peter Zijlstra Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Ingo Molnar [drop the nonseekable_open, not needed for anon inodes] Signed-off-by: Frederic Weisbecker diff --git a/kernel/perf_event.c b/kernel/perf_event.c index 63fbce1..4aa50ff 100644 --- a/kernel/perf_event.c +++ b/kernel/perf_event.c @@ -2645,6 +2645,7 @@ static int perf_fasync(int fd, struct file *filp, int on) } static const struct file_operations perf_fops = { + .llseek = no_llseek, .release = perf_release, .read = perf_read, .poll = perf_poll, -- cgit v0.10.2 From 8141d0050d76e5695011b5ab577ec66fb51a998c Mon Sep 17 00:00:00 2001 From: Hitoshi Mitake Date: Sun, 4 Apr 2010 17:13:18 +0900 Subject: perf: Swap inclusion order of util.h and string.h in util/string.c MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Currently util/string.c includes headers in this order: string.h, util.h But this causes a build error because __USE_GNU definition is needed for strndup() definition: % make -j touch .perf.dev.null CC util/string.o cc1: warnings being treated as errors util/string.c: In function ‘argv_split’: util/string.c:171: error: implicit declaration of function ‘strndup’ util/string.c:171: error: incompatible implicit declaration of built-in function ‘strndup’ So this patch swaps the headers inclusion order. util.h defines _GNU_SOURCE, and /usr/include/features.h defines __USE_GNU as 1 if _GNU_SOURCE is defined. Signed-off-by: Hitoshi Mitake Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo LKML-Reference: <1270368798-27232-1-git-send-email-mitake@dcl.info.waseda.ac.jp> Signed-off-by: Frederic Weisbecker diff --git a/tools/perf/util/string.c b/tools/perf/util/string.c index d438924..0409fc7 100644 --- a/tools/perf/util/string.c +++ b/tools/perf/util/string.c @@ -1,5 +1,5 @@ -#include "string.h" #include "util.h" +#include "string.h" #define K 1024LL /* -- cgit v0.10.2 From e363a755e8033e18f733fc0d1687039df8efade0 Mon Sep 17 00:00:00 2001 From: Alan Jenkins Date: Tue, 30 Jun 2009 14:35:05 +0000 Subject: ACPI: battery: Fix CONFIG_ACPI_SYSFS_POWER=n Disabling CONFIG_ACPI_SYSFS_POWER changes the behaviour of acpi_battery_update(). It will call acpi_battery_get_info() even if the battery is not present. I haven't noticed this causing any problem, but it does look like a bad idea. Signed-off-by: Alan Jenkins Acked-by: Alexey Starikovskiy Signed-off-by: Len Brown diff --git a/drivers/acpi/battery.c b/drivers/acpi/battery.c index 75f39f2..52df994 100644 --- a/drivers/acpi/battery.c +++ b/drivers/acpi/battery.c @@ -567,13 +567,13 @@ static int acpi_battery_update(struct acpi_battery *battery) result = acpi_battery_get_status(battery); if (result) return result; -#ifdef CONFIG_ACPI_SYSFS_POWER if (!acpi_battery_present(battery)) { +#ifdef CONFIG_ACPI_SYSFS_POWER sysfs_remove_battery(battery); +#endif battery->update_time = 0; return 0; } -#endif if (!battery->update_time || old_present != acpi_battery_present(battery)) { result = acpi_battery_get_info(battery); -- cgit v0.10.2 From f79e1cec8c8aa64895fd7b595dc7b48157df0754 Mon Sep 17 00:00:00 2001 From: Alan Jenkins Date: Tue, 30 Jun 2009 14:36:16 +0000 Subject: ACPI: battery drivers should call power_supply_changed() Calling kobject_uevent() directly is a layering violation. In particular, it means we'll miss updating the generic LED trigger. Signed-off-by: Alan Jenkins Acked-by: Alexey Starikovskiy Signed-off-by: Len Brown diff --git a/drivers/acpi/battery.c b/drivers/acpi/battery.c index 52df994..db78b6e 100644 --- a/drivers/acpi/battery.c +++ b/drivers/acpi/battery.c @@ -879,7 +879,7 @@ static void acpi_battery_notify(struct acpi_device *device, u32 event) #ifdef CONFIG_ACPI_SYSFS_POWER /* acpi_battery_update could remove power_supply object */ if (battery->bat.dev) - kobject_uevent(&battery->bat.dev->kobj, KOBJ_CHANGE); + power_supply_changed(&battery->bat); #endif } -- cgit v0.10.2 From 1638bca898e55f1c89f18b2e5accfac8591fed61 Mon Sep 17 00:00:00 2001 From: Chase Douglas Date: Mon, 22 Mar 2010 15:08:09 -0400 Subject: ACPI: Reduce ACPI resource conflict message to KERN_WARNING, printk cleanup By default, ACPI resource conflict messages are logged at level KERN_ERR. This is a rather high level for a message that is more a warning than an indication of a real kernel error. Also, KERN_ERR level messages can appear over some boot splash screens, and this message is not serious enough to warrant such treatment. Thus, the log level has been reduced to KERN_WARNING. [lenb changed to KERN_WARNING rather than all the way to KERN_INFO] Also, cleanup message to use %pR resource printing format. Signed-off-by: Chase Douglas Signed-off-by: Len Brown diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c index 8e6d866..6e49f62 100644 --- a/drivers/acpi/osl.c +++ b/drivers/acpi/osl.c @@ -1151,16 +1151,10 @@ int acpi_check_resource_conflict(const struct resource *res) if (clash) { if (acpi_enforce_resources != ENFORCE_RESOURCES_NO) { - printk("%sACPI: %s resource %s [0x%llx-0x%llx]" - " conflicts with ACPI region %s" - " [0x%llx-0x%llx]\n", - acpi_enforce_resources == ENFORCE_RESOURCES_LAX - ? KERN_WARNING : KERN_ERR, - ioport ? "I/O" : "Memory", res->name, - (long long) res->start, (long long) res->end, - res_list_elem->name, - (long long) res_list_elem->start, - (long long) res_list_elem->end); + printk(KERN_WARNING "ACPI: resource %s %pR" + " conflicts with ACPI region %s %pR\n", + res->name, res, res_list_elem->name, + res_list_elem); if (acpi_enforce_resources == ENFORCE_RESOURCES_LAX) printk(KERN_NOTICE "ACPI: This conflict may" " cause random problems and system" -- cgit v0.10.2 From aa27497c2fb4c7f57706099bd489e683e5cc3e3b Mon Sep 17 00:00:00 2001 From: Lai Jiangshan Date: Mon, 5 Apr 2010 17:11:05 +0800 Subject: tracing: Fix uninitialized variable of tracing/trace output Because a local variable is not initialized, I got these when I did 'cat tracing/trace'. (not trace_pipe): CPU:0 [LOST 18446744071579453134 EVENTS] ps-3099 [000] 560.770221: lock_acquire: ffff880030865010 &(&dentry->d_lock)->rlock CPU:0 [LOST 18446744071579453134 EVENTS] ps-3099 [000] 560.770221: lock_release: ffff880030865010 &(&dentry->d_lock)->rlock CPU:0 [LOST 18446612133255294080 EVENTS] ps-3099 [000] 560.770221: lock_acquire: ffff880030865010 &(&dentry->d_lock)->rlock CPU:0 [LOST 18446744071579453134 EVENTS] ps-3099 [000] 560.770222: lock_release: ffff880030865010 &(&dentry->d_lock)->rlock CPU:0 [LOST 18446744071579453134 EVENTS] ps-3099 [000] 560.770222: lock_release: ffffffff816cfb98 dcache_lock See peek_next_entry(), it does not set *lost_events when we 'cat tracing/trace' Signed-off-by: Lai Jiangshan LKML-Reference: <4BB9A929.2000303@cn.fujitsu.com> Signed-off-by: Steven Rostedt diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index 0498beb..b9be232 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -1571,7 +1571,7 @@ __find_next_entry(struct trace_iterator *iter, int *ent_cpu, { struct ring_buffer *buffer = iter->tr->buffer; struct trace_entry *ent, *next = NULL; - unsigned long lost_events, next_lost = 0; + unsigned long lost_events = 0, next_lost = 0; int cpu_file = iter->cpu_file; u64 next_ts = 0, ts; int next_cpu = -1; -- cgit v0.10.2 From 3fa49e3ad9ac20b15edfb0c51bbad36e45a84b17 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Mon, 29 Mar 2010 15:24:40 +0100 Subject: ASoC: Avoid wraparound in wm_hubs DC servo correction If the correction wraps around then a substantial offset would be introduced. Signed-off-by: Mark Brown Acked-by: Liam Girdwood diff --git a/sound/soc/codecs/wm_hubs.c b/sound/soc/codecs/wm_hubs.c index 486bdd2..3729a12 100644 --- a/sound/soc/codecs/wm_hubs.c +++ b/sound/soc/codecs/wm_hubs.c @@ -113,13 +113,15 @@ static void calibrate_dc_servo(struct snd_soc_codec *codec) /* HPOUT1L */ reg = snd_soc_read(codec, WM8993_DC_SERVO_READBACK_1) & WM8993_DCS_INTEG_CHAN_0_MASK;; - reg += hubs->dcs_codes; + if (reg + hubs->dcs_codes > 0 && reg + hubs->dcs_codes < 0xff) + reg += hubs->dcs_codes; dcs_cfg = reg << WM8993_DCS_DAC_WR_VAL_1_SHIFT; /* HPOUT1R */ reg = snd_soc_read(codec, WM8993_DC_SERVO_READBACK_2) & WM8993_DCS_INTEG_CHAN_1_MASK; - reg += hubs->dcs_codes; + if (reg + hubs->dcs_codes > 0 && reg + hubs->dcs_codes < 0xff) + reg += hubs->dcs_codes; dcs_cfg |= reg; /* Do it */ -- cgit v0.10.2 From 8437f7006b9cfa249791e2fd57596683d4561843 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Mon, 29 Mar 2010 17:09:45 +0100 Subject: ASoC: Support second DC servo readback method for wm_hubs More recent Wolfson hubs devices add the ability to read back the DC servo calibration information from the register used to write offsets, and later still ones remove the old readback registers. Add support for the new scheme, and use it for WM8994 device revisions that support it. Signed-off-by: Mark Brown Acked-by: Liam Girdwood diff --git a/sound/soc/codecs/wm8994.c b/sound/soc/codecs/wm8994.c index d10d651..c80218f 100644 --- a/sound/soc/codecs/wm8994.c +++ b/sound/soc/codecs/wm8994.c @@ -3730,11 +3730,12 @@ static int wm8994_codec_probe(struct platform_device *pdev) case 3: wm8994->hubs.dcs_codes = -5; wm8994->hubs.hp_startup_mode = 1; + wm8994->hubs.dcs_readback_mode = 1; break; default: + wm8994->hubs.dcs_readback_mode = 1; break; } - /* Remember if AIFnLRCLK is configured as a GPIO. This should be * configured on init - if a system wants to do this dynamically diff --git a/sound/soc/codecs/wm_hubs.c b/sound/soc/codecs/wm_hubs.c index 3729a12..2b5c092 100644 --- a/sound/soc/codecs/wm_hubs.c +++ b/sound/soc/codecs/wm_hubs.c @@ -86,7 +86,7 @@ static void wait_for_dc_servo(struct snd_soc_codec *codec) static void calibrate_dc_servo(struct snd_soc_codec *codec) { struct wm_hubs_data *hubs = codec->private_data; - u16 reg, dcs_cfg; + u16 reg, reg_l, reg_r, dcs_cfg; /* Set for 32 series updates */ snd_soc_update_bits(codec, WM8993_DC_SERVO_1, @@ -110,19 +110,38 @@ static void calibrate_dc_servo(struct snd_soc_codec *codec) dev_dbg(codec->dev, "Applying %d code DC servo correction\n", hubs->dcs_codes); + /* Different chips in the family support different + * readback methods. + */ + switch (hubs->dcs_readback_mode) { + case 0: + reg_l = snd_soc_read(codec, WM8993_DC_SERVO_READBACK_1) + & WM8993_DCS_INTEG_CHAN_0_MASK;; + reg_r = snd_soc_read(codec, WM8993_DC_SERVO_READBACK_2) + & WM8993_DCS_INTEG_CHAN_1_MASK; + break; + case 1: + reg = snd_soc_read(codec, WM8993_DC_SERVO_3); + reg_l = (reg & WM8993_DCS_DAC_WR_VAL_1_MASK) + >> WM8993_DCS_DAC_WR_VAL_1_SHIFT; + reg_r = reg & WM8993_DCS_DAC_WR_VAL_0_MASK; + break; + default: + WARN(1, "Unknown DCS readback method"); + break; + } + /* HPOUT1L */ - reg = snd_soc_read(codec, WM8993_DC_SERVO_READBACK_1) & - WM8993_DCS_INTEG_CHAN_0_MASK;; - if (reg + hubs->dcs_codes > 0 && reg + hubs->dcs_codes < 0xff) - reg += hubs->dcs_codes; - dcs_cfg = reg << WM8993_DCS_DAC_WR_VAL_1_SHIFT; + if (reg_l + hubs->dcs_codes > 0 && + reg_l + hubs->dcs_codes < 0xff) + reg_l += hubs->dcs_codes; + dcs_cfg = reg_l << WM8993_DCS_DAC_WR_VAL_1_SHIFT; /* HPOUT1R */ - reg = snd_soc_read(codec, WM8993_DC_SERVO_READBACK_2) & - WM8993_DCS_INTEG_CHAN_1_MASK; - if (reg + hubs->dcs_codes > 0 && reg + hubs->dcs_codes < 0xff) - reg += hubs->dcs_codes; - dcs_cfg |= reg; + if (reg_r + hubs->dcs_codes > 0 && + reg_r + hubs->dcs_codes < 0xff) + reg_r += hubs->dcs_codes; + dcs_cfg |= reg_r; /* Do it */ snd_soc_write(codec, WM8993_DC_SERVO_3, dcs_cfg); diff --git a/sound/soc/codecs/wm_hubs.h b/sound/soc/codecs/wm_hubs.h index 420104f..e51c166 100644 --- a/sound/soc/codecs/wm_hubs.h +++ b/sound/soc/codecs/wm_hubs.h @@ -21,6 +21,7 @@ extern const unsigned int wm_hubs_spkmix_tlv[]; /* This *must* be the first element of the codec->private_data struct */ struct wm_hubs_data { int dcs_codes; + int dcs_readback_mode; int hp_startup_mode; }; -- cgit v0.10.2 From ae9d8607fe24253efc9f14b696f51cfd683801be Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Mon, 29 Mar 2010 16:34:42 +0100 Subject: ASoC: Don't do runtime wm_hubs DC servo updates if using offset correction If we need to offset correct the DC servo then don't use runtime recalibration since that is likely to introduce further offsets which will be evident on powerdown. Signed-off-by: Mark Brown Acked-by: Liam Girdwood diff --git a/sound/soc/codecs/wm_hubs.c b/sound/soc/codecs/wm_hubs.c index 2b5c092..e81ba6d 100644 --- a/sound/soc/codecs/wm_hubs.c +++ b/sound/soc/codecs/wm_hubs.c @@ -162,10 +162,16 @@ static int wm8993_put_dc_servo(struct snd_kcontrol *kcontrol, struct snd_ctl_elem_value *ucontrol) { struct snd_soc_codec *codec = snd_kcontrol_chip(kcontrol); + struct wm_hubs_data *hubs = codec->private_data; int ret; ret = snd_soc_put_volsw_2r(kcontrol, ucontrol); + /* If we're applying an offset correction then updating the + * callibration would be likely to introduce further offsets. */ + if (hubs->dcs_codes) + return ret; + /* Only need to do this if the outputs are active */ if (snd_soc_read(codec, WM8993_POWER_MANAGEMENT_1) & (WM8993_HPOUT1L_ENA | WM8993_HPOUT1R_ENA)) -- cgit v0.10.2 From 4dcc93d0ede49fae32dd0ee41c685db1be14c529 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Mon, 29 Mar 2010 17:18:41 +0100 Subject: ASoC: Don't use DCS_DATAPATH_BUSY for WM hubs devices The DCS_DATAPATH_BUSY bit used to monitor the completion of DC servo operations has been deprecated and with some more recente revisions may perform incorrectly, especially when only analogue bypass paths are in use. Switch to using readback from the DC servo command register instead, which is supported for all devices. Without this unacceptably long timeouts may be observed in some circumstances. Signed-off-by: Mark Brown Acked-by: Liam Girdwood diff --git a/sound/soc/codecs/wm_hubs.c b/sound/soc/codecs/wm_hubs.c index e81ba6d..e1f225a 100644 --- a/sound/soc/codecs/wm_hubs.c +++ b/sound/soc/codecs/wm_hubs.c @@ -62,21 +62,27 @@ static const char *speaker_mode_text[] = { static const struct soc_enum speaker_mode = SOC_ENUM_SINGLE(WM8993_SPKMIXR_ATTENUATION, 8, 2, speaker_mode_text); -static void wait_for_dc_servo(struct snd_soc_codec *codec) +static void wait_for_dc_servo(struct snd_soc_codec *codec, unsigned int op) { unsigned int reg; int count = 0; + unsigned int val; + + val = op | WM8993_DCS_ENA_CHAN_0 | WM8993_DCS_ENA_CHAN_1; + + /* Trigger the command */ + snd_soc_write(codec, WM8993_DC_SERVO_0, val); dev_dbg(codec->dev, "Waiting for DC servo...\n"); do { count++; msleep(1); - reg = snd_soc_read(codec, WM8993_DC_SERVO_READBACK_0); + reg = snd_soc_read(codec, WM8993_DC_SERVO_0); dev_dbg(codec->dev, "DC servo: %x\n", reg); - } while (reg & WM8993_DCS_DATAPATH_BUSY && count < 400); + } while (reg & op && count < 400); - if (reg & WM8993_DCS_DATAPATH_BUSY) + if (reg & op) dev_err(codec->dev, "Timed out waiting for DC Servo\n"); } @@ -92,18 +98,8 @@ static void calibrate_dc_servo(struct snd_soc_codec *codec) snd_soc_update_bits(codec, WM8993_DC_SERVO_1, WM8993_DCS_SERIES_NO_01_MASK, 32 << WM8993_DCS_SERIES_NO_01_SHIFT); - - /* Enable the DC servo. Write all bits to avoid triggering startup - * or write calibration. - */ - snd_soc_update_bits(codec, WM8993_DC_SERVO_0, - 0xFFFF, - WM8993_DCS_ENA_CHAN_0 | - WM8993_DCS_ENA_CHAN_1 | - WM8993_DCS_TRIG_SERIES_1 | - WM8993_DCS_TRIG_SERIES_0); - - wait_for_dc_servo(codec); + wait_for_dc_servo(codec, + WM8993_DCS_TRIG_SERIES_0 | WM8993_DCS_TRIG_SERIES_1); /* Apply correction to DC servo result */ if (hubs->dcs_codes) { @@ -145,13 +141,9 @@ static void calibrate_dc_servo(struct snd_soc_codec *codec) /* Do it */ snd_soc_write(codec, WM8993_DC_SERVO_3, dcs_cfg); - snd_soc_update_bits(codec, WM8993_DC_SERVO_0, - WM8993_DCS_TRIG_DAC_WR_0 | - WM8993_DCS_TRIG_DAC_WR_1, - WM8993_DCS_TRIG_DAC_WR_0 | - WM8993_DCS_TRIG_DAC_WR_1); - - wait_for_dc_servo(codec); + wait_for_dc_servo(codec, + WM8993_DCS_TRIG_DAC_WR_0 | + WM8993_DCS_TRIG_DAC_WR_1); } } -- cgit v0.10.2 From d522ffbfb9fccf6eca283cd2e8b03cf3d21fb616 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Tue, 30 Mar 2010 14:29:14 +0100 Subject: ASoC: Only do WM8994 bias off transition from standby Otherwise we may try to power down multiple times when the using idle bias off and the driver is removed. Signed-off-by: Mark Brown Acked-by: Liam Girdwood diff --git a/sound/soc/codecs/wm8994.c b/sound/soc/codecs/wm8994.c index c80218f..f8355ac 100644 --- a/sound/soc/codecs/wm8994.c +++ b/sound/soc/codecs/wm8994.c @@ -3007,34 +3007,39 @@ static int wm8994_set_bias_level(struct snd_soc_codec *codec, break; case SND_SOC_BIAS_OFF: - /* Switch over to startup biases */ - snd_soc_update_bits(codec, WM8994_ANTIPOP_2, - WM8994_BIAS_SRC | WM8994_STARTUP_BIAS_ENA | - WM8994_VMID_BUF_ENA | - WM8994_VMID_RAMP_MASK, - WM8994_BIAS_SRC | WM8994_STARTUP_BIAS_ENA | - WM8994_VMID_BUF_ENA | - (1 << WM8994_VMID_RAMP_SHIFT)); - - /* Disable main biases */ - snd_soc_update_bits(codec, WM8994_POWER_MANAGEMENT_1, - WM8994_BIAS_ENA | WM8994_VMID_SEL_MASK, 0); + if (codec->bias_level == SND_SOC_BIAS_STANDBY) { + /* Switch over to startup biases */ + snd_soc_update_bits(codec, WM8994_ANTIPOP_2, + WM8994_BIAS_SRC | + WM8994_STARTUP_BIAS_ENA | + WM8994_VMID_BUF_ENA | + WM8994_VMID_RAMP_MASK, + WM8994_BIAS_SRC | + WM8994_STARTUP_BIAS_ENA | + WM8994_VMID_BUF_ENA | + (1 << WM8994_VMID_RAMP_SHIFT)); - /* Discharge line */ - snd_soc_update_bits(codec, WM8994_ANTIPOP_1, - WM8994_LINEOUT1_DISCH | - WM8994_LINEOUT2_DISCH, - WM8994_LINEOUT1_DISCH | - WM8994_LINEOUT2_DISCH); + /* Disable main biases */ + snd_soc_update_bits(codec, WM8994_POWER_MANAGEMENT_1, + WM8994_BIAS_ENA | + WM8994_VMID_SEL_MASK, 0); - msleep(5); + /* Discharge line */ + snd_soc_update_bits(codec, WM8994_ANTIPOP_1, + WM8994_LINEOUT1_DISCH | + WM8994_LINEOUT2_DISCH, + WM8994_LINEOUT1_DISCH | + WM8994_LINEOUT2_DISCH); - /* Switch off startup biases */ - snd_soc_update_bits(codec, WM8994_ANTIPOP_2, - WM8994_BIAS_SRC | WM8994_STARTUP_BIAS_ENA | - WM8994_VMID_BUF_ENA | - WM8994_VMID_RAMP_MASK, 0); + msleep(5); + /* Switch off startup biases */ + snd_soc_update_bits(codec, WM8994_ANTIPOP_2, + WM8994_BIAS_SRC | + WM8994_STARTUP_BIAS_ENA | + WM8994_VMID_BUF_ENA | + WM8994_VMID_RAMP_MASK, 0); + } break; } codec->bias_level = level; -- cgit v0.10.2 From bade732a2848e640482c1e99629a90e085d574dd Mon Sep 17 00:00:00 2001 From: Tom Tucker Date: Sat, 3 Apr 2010 08:27:29 -0500 Subject: svcrdma: RDMA support not yet compatible with RPC6 RPC6 requires that it be possible to create endpoints that listen exclusively for IPv4 or IPv6 connection requests. This is not currently supported by the RDMA API. This fixes a server RDMA regression introduced by 37498292a "NFSD: Create PF_INET6 listener in write_ports". Signed-off-by: Tom Tucker Tested-by: Steve Wise Reviewed-by: Chuck Lever Signed-off-by: J. Bruce Fields diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c index 3fa5751..4e6bbf9 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_transport.c +++ b/net/sunrpc/xprtrdma/svc_rdma_transport.c @@ -678,7 +678,10 @@ static struct svc_xprt *svc_rdma_create(struct svc_serv *serv, int ret; dprintk("svcrdma: Creating RDMA socket\n"); - + if (sa->sa_family != AF_INET) { + dprintk("svcrdma: Address family %d is not supported.\n", sa->sa_family); + return ERR_PTR(-EAFNOSUPPORT); + } cma_xprt = rdma_create_xprt(serv, 1); if (!cma_xprt) return ERR_PTR(-ENOMEM); -- cgit v0.10.2 From a74b2adae06265b8cfa335d7d40d4a5abd11e977 Mon Sep 17 00:00:00 2001 From: Ricky Benitez Date: Mon, 5 Apr 2010 18:22:17 +0200 Subject: block: expose the statistics in blkio.time and blkio.sectors for the root cgroup Currently, the io statistics for the root cgroup are maintained, but they are not shown because the device information is not available at the point that the root blkio cgroup is created. This patch updates the device information when the statistics are updated so that the statistics become visible. Signed-off-by: Ricky Benitez Acked-by: Vivek Goyal Signed-off-by: Jens Axboe diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index 2c7a0f4..7104ac8 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -948,6 +948,11 @@ cfq_find_alloc_cfqg(struct cfq_data *cfqd, struct cgroup *cgroup, int create) unsigned int major, minor; cfqg = cfqg_of_blkg(blkiocg_lookup_group(blkcg, key)); + if (cfqg && !cfqg->blkg.dev && bdi->dev && dev_name(bdi->dev)) { + sscanf(dev_name(bdi->dev), "%u:%u", &major, &minor); + cfqg->blkg.dev = MKDEV(major, minor); + goto done; + } if (cfqg || !create) goto done; -- cgit v0.10.2 From d12841827a6de120199609dadb6ff4ec99bd90ea Mon Sep 17 00:00:00 2001 From: Tony Vroon Date: Mon, 5 Apr 2010 16:30:43 +0100 Subject: ALSA: hda - Enable amplifiers on Acer Inspire 6530G After more tests it appears that EAPD needs to be enabled on both the 0x14 and 0x15 NIDs to enable the main speaker and headphone amplifiers. The maximum volume setting is now equal to what the machine achieves under other operating systems. Disabling Front or LFE playback triggers EAPD and disables the amplifier. As such, these two playback switches have been removed from the mixer. Signed-off-by: Tony Vroon Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index ca93c4c..5472062 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -1621,6 +1621,11 @@ static struct hda_verb alc888_acer_aspire_4930g_verbs[] = { */ static struct hda_verb alc888_acer_aspire_6530g_verbs[] = { +/* Route to built-in subwoofer as well as speakers */ + {0x0c, AC_VERB_SET_AMP_GAIN_MUTE, AMP_IN_UNMUTE(0)}, + {0x0c, AC_VERB_SET_AMP_GAIN_MUTE, AMP_IN_UNMUTE(1)}, + {0x0f, AC_VERB_SET_AMP_GAIN_MUTE, AMP_IN_UNMUTE(0)}, + {0x0f, AC_VERB_SET_AMP_GAIN_MUTE, AMP_IN_UNMUTE(1)}, /* Bias voltage on for external mic port */ {0x18, AC_VERB_SET_PIN_WIDGET_CONTROL, PIN_IN | PIN_VREF80}, /* Front Mic: set to PIN_IN (empty by default) */ @@ -1632,10 +1637,12 @@ static struct hda_verb alc888_acer_aspire_6530g_verbs[] = { /* Enable speaker output */ {0x14, AC_VERB_SET_PIN_WIDGET_CONTROL, PIN_OUT}, {0x14, AC_VERB_SET_AMP_GAIN_MUTE, AMP_OUT_UNMUTE}, + {0x14, AC_VERB_SET_EAPD_BTLENABLE, 2}, /* Enable headphone output */ {0x15, AC_VERB_SET_PIN_WIDGET_CONTROL, PIN_OUT | PIN_HP}, {0x15, AC_VERB_SET_AMP_GAIN_MUTE, AMP_OUT_UNMUTE}, {0x15, AC_VERB_SET_CONNECT_SEL, 0x00}, + {0x15, AC_VERB_SET_EAPD_BTLENABLE, 2}, { } }; @@ -8462,9 +8469,7 @@ static struct snd_kcontrol_new alc883_acer_aspire_mixer[] = { static struct snd_kcontrol_new alc888_acer_aspire_6530_mixer[] = { HDA_CODEC_VOLUME("Front Playback Volume", 0x0c, 0x0, HDA_OUTPUT), - HDA_BIND_MUTE("Front Playback Switch", 0x0c, 2, HDA_INPUT), HDA_CODEC_VOLUME("LFE Playback Volume", 0x0f, 0x0, HDA_OUTPUT), - HDA_BIND_MUTE("LFE Playback Switch", 0x0f, 2, HDA_INPUT), HDA_CODEC_VOLUME("Line Playback Volume", 0x0b, 0x02, HDA_INPUT), HDA_CODEC_MUTE("Line Playback Switch", 0x0b, 0x02, HDA_INPUT), HDA_CODEC_VOLUME("CD Playback Volume", 0x0b, 0x04, HDA_INPUT), -- cgit v0.10.2 From 5f712b2b73a9fc87fcc52124cfe8adefaa0c92f5 Mon Sep 17 00:00:00 2001 From: Daniel Mack Date: Mon, 22 Mar 2010 10:11:15 +0100 Subject: ALSA: ASoC: move dma_data from snd_soc_dai to snd_soc_pcm_stream This fixes a memory corruption when ASoC devices are used in full-duplex mode. Specifically for pxa-ssp code, where this pointer is dynamically allocated for each direction and destroyed upon each stream start. All other platforms are fixed blindly, I couldn't even compile-test them. Sorry for any breakage I may have caused. [Note that this is a backported version for 2.6.34. Upstream commit is fd23b7dee] Signed-off-by: Daniel Mack Reported-by: Sven Neumann Reported-by: Michael Hirsch Acked-by: Liam Girdwood Signed-off-by: Mark Brown diff --git a/include/sound/soc-dai.h b/include/sound/soc-dai.h index 061f16d..0a0b019 100644 --- a/include/sound/soc-dai.h +++ b/include/sound/soc-dai.h @@ -219,7 +219,6 @@ struct snd_soc_dai { struct snd_soc_codec *codec; unsigned int active; unsigned char pop_wait:1; - void *dma_data; /* DAI private data */ void *private_data; @@ -230,4 +229,21 @@ struct snd_soc_dai { struct list_head list; }; +static inline void *snd_soc_dai_get_dma_data(const struct snd_soc_dai *dai, + const struct snd_pcm_substream *ss) +{ + return (ss->stream == SNDRV_PCM_STREAM_PLAYBACK) ? + dai->playback.dma_data : dai->capture.dma_data; +} + +static inline void snd_soc_dai_set_dma_data(struct snd_soc_dai *dai, + const struct snd_pcm_substream *ss, + void *data) +{ + if (ss->stream == SNDRV_PCM_STREAM_PLAYBACK) + dai->playback.dma_data = data; + else + dai->capture.dma_data = data; +} + #endif diff --git a/include/sound/soc.h b/include/sound/soc.h index 5d234a8..a57fbfc 100644 --- a/include/sound/soc.h +++ b/include/sound/soc.h @@ -375,6 +375,7 @@ struct snd_soc_pcm_stream { unsigned int channels_min; /* min channels */ unsigned int channels_max; /* max channels */ unsigned int active:1; /* stream is in use */ + void *dma_data; /* used by platform code */ }; /* SoC audio ops */ diff --git a/sound/soc/atmel/atmel-pcm.c b/sound/soc/atmel/atmel-pcm.c index 9ef6b96..3e6628c8 100644 --- a/sound/soc/atmel/atmel-pcm.c +++ b/sound/soc/atmel/atmel-pcm.c @@ -180,7 +180,7 @@ static int atmel_pcm_hw_params(struct snd_pcm_substream *substream, snd_pcm_set_runtime_buffer(substream, &substream->dma_buffer); runtime->dma_bytes = params_buffer_bytes(params); - prtd->params = rtd->dai->cpu_dai->dma_data; + prtd->params = snd_soc_dai_get_dma_data(rtd->dai->cpu_dai, substream); prtd->params->dma_intr_handler = atmel_pcm_dma_irq; prtd->dma_buffer = runtime->dma_addr; diff --git a/sound/soc/atmel/atmel_ssc_dai.c b/sound/soc/atmel/atmel_ssc_dai.c index e588e63..0b59806 100644 --- a/sound/soc/atmel/atmel_ssc_dai.c +++ b/sound/soc/atmel/atmel_ssc_dai.c @@ -363,12 +363,12 @@ static int atmel_ssc_hw_params(struct snd_pcm_substream *substream, ssc_p->dma_params[dir] = dma_params; /* - * The cpu_dai->dma_data field is only used to communicate the - * appropriate DMA parameters to the pcm driver hw_params() + * The snd_soc_pcm_stream->dma_data field is only used to communicate + * the appropriate DMA parameters to the pcm driver hw_params() * function. It should not be used for other purposes * as it is common to all substreams. */ - rtd->dai->cpu_dai->dma_data = dma_params; + snd_soc_dai_set_dma_data(rtd->dai->cpu_dai, substream, dma_params); channels = params_channels(params); diff --git a/sound/soc/davinci/davinci-i2s.c b/sound/soc/davinci/davinci-i2s.c index 6362ca0..4aad7ecc 100644 --- a/sound/soc/davinci/davinci-i2s.c +++ b/sound/soc/davinci/davinci-i2s.c @@ -585,7 +585,8 @@ static int davinci_i2s_probe(struct platform_device *pdev) dev->dma_params[SNDRV_PCM_STREAM_CAPTURE].channel = res->start; davinci_i2s_dai.private_data = dev; - davinci_i2s_dai.dma_data = dev->dma_params; + davinci_i2s_dai.capture.dma_data = dev->dma_params; + davinci_i2s_dai.playback.dma_data = dev->dma_params; ret = snd_soc_register_dai(&davinci_i2s_dai); if (ret != 0) goto err_free_mem; diff --git a/sound/soc/davinci/davinci-mcasp.c b/sound/soc/davinci/davinci-mcasp.c index ab6518d..c056bfb 100644 --- a/sound/soc/davinci/davinci-mcasp.c +++ b/sound/soc/davinci/davinci-mcasp.c @@ -917,7 +917,8 @@ static int davinci_mcasp_probe(struct platform_device *pdev) dma_data->channel = res->start; davinci_mcasp_dai[pdata->op_mode].private_data = dev; - davinci_mcasp_dai[pdata->op_mode].dma_data = dev->dma_params; + davinci_mcasp_dai[pdata->op_mode].capture.dma_data = dev->dma_params; + davinci_mcasp_dai[pdata->op_mode].playback.dma_data = dev->dma_params; davinci_mcasp_dai[pdata->op_mode].dev = &pdev->dev; ret = snd_soc_register_dai(&davinci_mcasp_dai[pdata->op_mode]); diff --git a/sound/soc/davinci/davinci-pcm.c b/sound/soc/davinci/davinci-pcm.c index 80c7fdf..2dc406f 100644 --- a/sound/soc/davinci/davinci-pcm.c +++ b/sound/soc/davinci/davinci-pcm.c @@ -649,8 +649,10 @@ static int davinci_pcm_open(struct snd_pcm_substream *substream) struct snd_pcm_hardware *ppcm; int ret = 0; struct snd_soc_pcm_runtime *rtd = substream->private_data; - struct davinci_pcm_dma_params *pa = rtd->dai->cpu_dai->dma_data; + struct davinci_pcm_dma_params *pa; struct davinci_pcm_dma_params *params; + + pa = snd_soc_dai_get_dma_data(rtd->dai->cpu_dai, substream); if (!pa) return -ENODEV; params = &pa[substream->stream]; diff --git a/sound/soc/imx/imx-pcm-dma-mx2.c b/sound/soc/imx/imx-pcm-dma-mx2.c index 19452e4..c78c000 100644 --- a/sound/soc/imx/imx-pcm-dma-mx2.c +++ b/sound/soc/imx/imx-pcm-dma-mx2.c @@ -83,11 +83,13 @@ static void snd_imx_dma_err_callback(int channel, void *data, int err) static int imx_ssi_dma_alloc(struct snd_pcm_substream *substream) { struct snd_soc_pcm_runtime *rtd = substream->private_data; - struct imx_pcm_dma_params *dma_params = rtd->dai->cpu_dai->dma_data; + struct imx_pcm_dma_params *dma_params; struct snd_pcm_runtime *runtime = substream->runtime; struct imx_pcm_runtime_data *iprtd = runtime->private_data; int ret; + dma_params = snd_soc_get_dma_data(rtd->dai->cpu_dai, substream); + iprtd->dma = imx_dma_request_by_prio(DRV_NAME, DMA_PRIO_HIGH); if (iprtd->dma < 0) { pr_err("Failed to claim the audio DMA\n"); @@ -192,10 +194,12 @@ static int snd_imx_pcm_prepare(struct snd_pcm_substream *substream) { struct snd_pcm_runtime *runtime = substream->runtime; struct snd_soc_pcm_runtime *rtd = substream->private_data; - struct imx_pcm_dma_params *dma_params = rtd->dai->cpu_dai->dma_data; + struct imx_pcm_dma_params *dma_params; struct imx_pcm_runtime_data *iprtd = runtime->private_data; int err; + dma_params = snd_soc_get_dma_data(rtd->dai->cpu_dai, substream); + iprtd->substream = substream; iprtd->buf = (unsigned int *)substream->dma_buffer.area; iprtd->period_cnt = 0; diff --git a/sound/soc/imx/imx-ssi.c b/sound/soc/imx/imx-ssi.c index 56f46a7..28e55c7 100644 --- a/sound/soc/imx/imx-ssi.c +++ b/sound/soc/imx/imx-ssi.c @@ -234,17 +234,20 @@ static int imx_ssi_hw_params(struct snd_pcm_substream *substream, struct snd_soc_dai *cpu_dai) { struct imx_ssi *ssi = cpu_dai->private_data; + struct imx_pcm_dma_params *dma_data; u32 reg, sccr; /* Tx/Rx config */ if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) { reg = SSI_STCCR; - cpu_dai->dma_data = &ssi->dma_params_tx; + dma_data = &ssi->dma_params_tx; } else { reg = SSI_SRCCR; - cpu_dai->dma_data = &ssi->dma_params_rx; + dma_data = &ssi->dma_params_rx; } + snd_soc_dai_set_dma_data(cpu_dai, substream, dma_data); + sccr = readl(ssi->base + reg) & ~SSI_STCCR_WL_MASK; /* DAI data (word) size */ diff --git a/sound/soc/omap/omap-mcbsp.c b/sound/soc/omap/omap-mcbsp.c index e814a95..8ad9dc9 100644 --- a/sound/soc/omap/omap-mcbsp.c +++ b/sound/soc/omap/omap-mcbsp.c @@ -297,7 +297,9 @@ static int omap_mcbsp_dai_hw_params(struct snd_pcm_substream *substream, omap_mcbsp_dai_dma_params[id][substream->stream].sync_mode = sync_mode; omap_mcbsp_dai_dma_params[id][substream->stream].data_type = OMAP_DMA_DATA_TYPE_S16; - cpu_dai->dma_data = &omap_mcbsp_dai_dma_params[id][substream->stream]; + + snd_soc_dai_set_dma_data(cpu_dai, substream, + &omap_mcbsp_dai_dma_params[id][substream->stream]); if (mcbsp_data->configured) { /* McBSP already configured by another stream */ diff --git a/sound/soc/omap/omap-mcpdm.c b/sound/soc/omap/omap-mcpdm.c index 25f19e4..b7f4f7e 100644 --- a/sound/soc/omap/omap-mcpdm.c +++ b/sound/soc/omap/omap-mcpdm.c @@ -150,7 +150,8 @@ static int omap_mcpdm_dai_hw_params(struct snd_pcm_substream *substream, int stream = substream->stream; int channels, err, link_mask = 0; - cpu_dai->dma_data = &omap_mcpdm_dai_dma_params[stream]; + snd_soc_dai_set_dma_data(cpu_dai, substream, + &omap_mcpdm_dai_dma_params[stream]); channels = params_channels(params); switch (channels) { diff --git a/sound/soc/omap/omap-pcm.c b/sound/soc/omap/omap-pcm.c index bdd1097..3945644 100644 --- a/sound/soc/omap/omap-pcm.c +++ b/sound/soc/omap/omap-pcm.c @@ -99,9 +99,11 @@ static int omap_pcm_hw_params(struct snd_pcm_substream *substream, struct snd_pcm_runtime *runtime = substream->runtime; struct snd_soc_pcm_runtime *rtd = substream->private_data; struct omap_runtime_data *prtd = runtime->private_data; - struct omap_pcm_dma_data *dma_data = rtd->dai->cpu_dai->dma_data; + struct omap_pcm_dma_data *dma_data; int err = 0; + dma_data = snd_soc_dai_get_dma_data(rtd->dai->cpu_dai, substream); + /* return if this is a bufferless transfer e.g. * codec <--> BT codec or GSM modem -- lg FIXME */ if (!dma_data) diff --git a/sound/soc/pxa/pxa-ssp.c b/sound/soc/pxa/pxa-ssp.c index 9e95e51..6959c51 100644 --- a/sound/soc/pxa/pxa-ssp.c +++ b/sound/soc/pxa/pxa-ssp.c @@ -121,10 +121,9 @@ static int pxa_ssp_startup(struct snd_pcm_substream *substream, ssp_disable(ssp); } - if (cpu_dai->dma_data) { - kfree(cpu_dai->dma_data); - cpu_dai->dma_data = NULL; - } + kfree(snd_soc_dai_get_dma_data(cpu_dai, substream)); + snd_soc_dai_set_dma_data(cpu_dai, substream, NULL); + return ret; } @@ -141,10 +140,8 @@ static void pxa_ssp_shutdown(struct snd_pcm_substream *substream, clk_disable(ssp->clk); } - if (cpu_dai->dma_data) { - kfree(cpu_dai->dma_data); - cpu_dai->dma_data = NULL; - } + kfree(snd_soc_dai_get_dma_data(cpu_dai, substream)); + snd_soc_dai_set_dma_data(cpu_dai, substream, NULL); } #ifdef CONFIG_PM @@ -569,19 +566,23 @@ static int pxa_ssp_hw_params(struct snd_pcm_substream *substream, u32 sspsp; int width = snd_pcm_format_physical_width(params_format(params)); int ttsa = ssp_read_reg(ssp, SSTSA) & 0xf; + struct pxa2xx_pcm_dma_params *dma_data; + + dma_data = snd_soc_dai_get_dma_data(dai, substream); /* generate correct DMA params */ - if (cpu_dai->dma_data) - kfree(cpu_dai->dma_data); + kfree(dma_data); /* Network mode with one active slot (ttsa == 1) can be used * to force 16-bit frame width on the wire (for S16_LE), even * with two channels. Use 16-bit DMA transfers for this case. */ - cpu_dai->dma_data = ssp_get_dma_params(ssp, + dma_data = ssp_get_dma_params(ssp, ((chn == 2) && (ttsa != 1)) || (width == 32), substream->stream == SNDRV_PCM_STREAM_PLAYBACK); + snd_soc_dai_set_dma_data(dai, substream, dma_data); + /* we can only change the settings if the port is not in use */ if (ssp_read_reg(ssp, SSCR0) & SSCR0_SSE) return 0; diff --git a/sound/soc/pxa/pxa2xx-ac97.c b/sound/soc/pxa/pxa2xx-ac97.c index e9ae7b3..d314115 100644 --- a/sound/soc/pxa/pxa2xx-ac97.c +++ b/sound/soc/pxa/pxa2xx-ac97.c @@ -122,11 +122,14 @@ static int pxa2xx_ac97_hw_params(struct snd_pcm_substream *substream, { struct snd_soc_pcm_runtime *rtd = substream->private_data; struct snd_soc_dai *cpu_dai = rtd->dai->cpu_dai; + struct pxa2xx_pcm_dma_params *dma_data; if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) - cpu_dai->dma_data = &pxa2xx_ac97_pcm_stereo_out; + dma_data = &pxa2xx_ac97_pcm_stereo_out; else - cpu_dai->dma_data = &pxa2xx_ac97_pcm_stereo_in; + dma_data = &pxa2xx_ac97_pcm_stereo_in; + + snd_soc_dai_set_dma_data(cpu_dai, substream, dma_data); return 0; } @@ -137,11 +140,14 @@ static int pxa2xx_ac97_hw_aux_params(struct snd_pcm_substream *substream, { struct snd_soc_pcm_runtime *rtd = substream->private_data; struct snd_soc_dai *cpu_dai = rtd->dai->cpu_dai; + struct pxa2xx_pcm_dma_params *dma_data; if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) - cpu_dai->dma_data = &pxa2xx_ac97_pcm_aux_mono_out; + dma_data = &pxa2xx_ac97_pcm_aux_mono_out; else - cpu_dai->dma_data = &pxa2xx_ac97_pcm_aux_mono_in; + dma_data = &pxa2xx_ac97_pcm_aux_mono_in; + + snd_soc_dai_set_dma_data(cpu_dai, substream, dma_data); return 0; } @@ -156,7 +162,8 @@ static int pxa2xx_ac97_hw_mic_params(struct snd_pcm_substream *substream, if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) return -ENODEV; else - cpu_dai->dma_data = &pxa2xx_ac97_pcm_mic_mono_in; + snd_soc_dai_set_dma_data(cpu_dai, substream, + &pxa2xx_ac97_pcm_mic_mono_in); return 0; } diff --git a/sound/soc/pxa/pxa2xx-i2s.c b/sound/soc/pxa/pxa2xx-i2s.c index 6b8f655..c1a52757 100644 --- a/sound/soc/pxa/pxa2xx-i2s.c +++ b/sound/soc/pxa/pxa2xx-i2s.c @@ -164,6 +164,7 @@ static int pxa2xx_i2s_hw_params(struct snd_pcm_substream *substream, { struct snd_soc_pcm_runtime *rtd = substream->private_data; struct snd_soc_dai *cpu_dai = rtd->dai->cpu_dai; + struct pxa2xx_pcm_dma_params *dma_data; BUG_ON(IS_ERR(clk_i2s)); clk_enable(clk_i2s); @@ -171,9 +172,11 @@ static int pxa2xx_i2s_hw_params(struct snd_pcm_substream *substream, pxa_i2s_wait(); if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) - cpu_dai->dma_data = &pxa2xx_i2s_pcm_stereo_out; + dma_data = &pxa2xx_i2s_pcm_stereo_out; else - cpu_dai->dma_data = &pxa2xx_i2s_pcm_stereo_in; + dma_data = &pxa2xx_i2s_pcm_stereo_in; + + snd_soc_dai_set_dma_data(cpu_dai, substream, dma_data); /* is port used by another stream */ if (!(SACR0 & SACR0_ENB)) { diff --git a/sound/soc/pxa/pxa2xx-pcm.c b/sound/soc/pxa/pxa2xx-pcm.c index d38e395..adc7e6f 100644 --- a/sound/soc/pxa/pxa2xx-pcm.c +++ b/sound/soc/pxa/pxa2xx-pcm.c @@ -25,9 +25,11 @@ static int pxa2xx_pcm_hw_params(struct snd_pcm_substream *substream, struct snd_pcm_runtime *runtime = substream->runtime; struct pxa2xx_runtime_data *prtd = runtime->private_data; struct snd_soc_pcm_runtime *rtd = substream->private_data; - struct pxa2xx_pcm_dma_params *dma = rtd->dai->cpu_dai->dma_data; + struct pxa2xx_pcm_dma_params *dma; int ret; + dma = snd_soc_dai_get_dma_data(rtd->dai->cpu_dai, substream); + /* return if this is a bufferless transfer e.g. * codec <--> BT codec or GSM modem -- lg FIXME */ if (!dma) diff --git a/sound/soc/s3c24xx/s3c-ac97.c b/sound/soc/s3c24xx/s3c-ac97.c index ee8ed9d..ecf4fd0 100644 --- a/sound/soc/s3c24xx/s3c-ac97.c +++ b/sound/soc/s3c24xx/s3c-ac97.c @@ -224,11 +224,14 @@ static int s3c_ac97_hw_params(struct snd_pcm_substream *substream, { struct snd_soc_pcm_runtime *rtd = substream->private_data; struct snd_soc_dai *cpu_dai = rtd->dai->cpu_dai; + struct s3c_dma_params *dma_data; if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) - cpu_dai->dma_data = &s3c_ac97_pcm_out; + dma_data = &s3c_ac97_pcm_out; else - cpu_dai->dma_data = &s3c_ac97_pcm_in; + dma_data = &s3c_ac97_pcm_in; + + snd_soc_dai_set_dma_data(cpu_dai, substream, dma_data); return 0; } @@ -238,8 +241,8 @@ static int s3c_ac97_trigger(struct snd_pcm_substream *substream, int cmd, { u32 ac_glbctrl; struct snd_soc_pcm_runtime *rtd = substream->private_data; - int channel = ((struct s3c_dma_params *) - rtd->dai->cpu_dai->dma_data)->channel; + struct s3c_dma_params *dma_data = + snd_soc_dai_get_dma_data(rtd->dai->cpu_dai, substream); ac_glbctrl = readl(s3c_ac97.regs + S3C_AC97_GLBCTRL); if (substream->stream == SNDRV_PCM_STREAM_CAPTURE) @@ -265,7 +268,7 @@ static int s3c_ac97_trigger(struct snd_pcm_substream *substream, int cmd, writel(ac_glbctrl, s3c_ac97.regs + S3C_AC97_GLBCTRL); - s3c2410_dma_ctrl(channel, S3C2410_DMAOP_STARTED); + s3c2410_dma_ctrl(dma_data->channel, S3C2410_DMAOP_STARTED); return 0; } @@ -280,7 +283,7 @@ static int s3c_ac97_hw_mic_params(struct snd_pcm_substream *substream, if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) return -ENODEV; else - cpu_dai->dma_data = &s3c_ac97_mic_in; + snd_soc_dai_set_dma_data(cpu_dai, substream, &s3c_ac97_mic_in); return 0; } @@ -290,8 +293,8 @@ static int s3c_ac97_mic_trigger(struct snd_pcm_substream *substream, { u32 ac_glbctrl; struct snd_soc_pcm_runtime *rtd = substream->private_data; - int channel = ((struct s3c_dma_params *) - rtd->dai->cpu_dai->dma_data)->channel; + struct s3c_dma_params *dma_data = + snd_soc_dai_get_dma_data(rtd->dai->cpu_dai, substream); ac_glbctrl = readl(s3c_ac97.regs + S3C_AC97_GLBCTRL); ac_glbctrl &= ~S3C_AC97_GLBCTRL_MICINTM_MASK; @@ -311,7 +314,7 @@ static int s3c_ac97_mic_trigger(struct snd_pcm_substream *substream, writel(ac_glbctrl, s3c_ac97.regs + S3C_AC97_GLBCTRL); - s3c2410_dma_ctrl(channel, S3C2410_DMAOP_STARTED); + s3c2410_dma_ctrl(dma_data->channel, S3C2410_DMAOP_STARTED); return 0; } diff --git a/sound/soc/s3c24xx/s3c-dma.c b/sound/soc/s3c24xx/s3c-dma.c index 7725e26..1b61c23 100644 --- a/sound/soc/s3c24xx/s3c-dma.c +++ b/sound/soc/s3c24xx/s3c-dma.c @@ -145,10 +145,12 @@ static int s3c_dma_hw_params(struct snd_pcm_substream *substream, struct snd_pcm_runtime *runtime = substream->runtime; struct s3c24xx_runtime_data *prtd = runtime->private_data; struct snd_soc_pcm_runtime *rtd = substream->private_data; - struct s3c_dma_params *dma = rtd->dai->cpu_dai->dma_data; unsigned long totbytes = params_buffer_bytes(params); + struct s3c_dma_params *dma = + snd_soc_dai_get_dma_data(rtd->dai->cpu_dai, substream); int ret = 0; + pr_debug("Entered %s\n", __func__); /* return if this is a bufferless transfer e.g. diff --git a/sound/soc/s3c24xx/s3c-i2s-v2.c b/sound/soc/s3c24xx/s3c-i2s-v2.c index e994d83..8851594 100644 --- a/sound/soc/s3c24xx/s3c-i2s-v2.c +++ b/sound/soc/s3c24xx/s3c-i2s-v2.c @@ -339,14 +339,17 @@ static int s3c2412_i2s_hw_params(struct snd_pcm_substream *substream, struct snd_soc_pcm_runtime *rtd = substream->private_data; struct snd_soc_dai_link *dai = rtd->dai; struct s3c_i2sv2_info *i2s = to_info(dai->cpu_dai); + struct s3c_dma_params *dma_data; u32 iismod; pr_debug("Entered %s\n", __func__); if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) - dai->cpu_dai->dma_data = i2s->dma_playback; + dma_data = i2s->dma_playback; else - dai->cpu_dai->dma_data = i2s->dma_capture; + dma_data = i2s->dma_capture; + + snd_soc_dai_set_dma_data(dai->cpu_dai, substream, dma_data); /* Working copies of register */ iismod = readl(i2s->regs + S3C2412_IISMOD); @@ -394,8 +397,8 @@ static int s3c2412_i2s_trigger(struct snd_pcm_substream *substream, int cmd, int capture = (substream->stream == SNDRV_PCM_STREAM_CAPTURE); unsigned long irqs; int ret = 0; - int channel = ((struct s3c_dma_params *) - rtd->dai->cpu_dai->dma_data)->channel; + struct s3c_dma_params *dma_data = + snd_soc_dai_get_dma_data(rtd->dai->cpu_dai, substream); pr_debug("Entered %s\n", __func__); @@ -431,7 +434,7 @@ static int s3c2412_i2s_trigger(struct snd_pcm_substream *substream, int cmd, * of the auto reload mechanism of S3C24XX. * This call won't bother S3C64XX. */ - s3c2410_dma_ctrl(channel, S3C2410_DMAOP_STARTED); + s3c2410_dma_ctrl(dma_data->channel, S3C2410_DMAOP_STARTED); break; diff --git a/sound/soc/s3c24xx/s3c-pcm.c b/sound/soc/s3c24xx/s3c-pcm.c index a98f40c..326f0a9 100644 --- a/sound/soc/s3c24xx/s3c-pcm.c +++ b/sound/soc/s3c24xx/s3c-pcm.c @@ -178,6 +178,7 @@ static int s3c_pcm_hw_params(struct snd_pcm_substream *substream, struct snd_soc_pcm_runtime *rtd = substream->private_data; struct snd_soc_dai_link *dai = rtd->dai; struct s3c_pcm_info *pcm = to_info(dai->cpu_dai); + struct s3c_dma_params *dma_data; void __iomem *regs = pcm->regs; struct clk *clk; int sclk_div, sync_div; @@ -187,9 +188,11 @@ static int s3c_pcm_hw_params(struct snd_pcm_substream *substream, dev_dbg(pcm->dev, "Entered %s\n", __func__); if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) - dai->cpu_dai->dma_data = pcm->dma_playback; + dma_data = pcm->dma_playback; else - dai->cpu_dai->dma_data = pcm->dma_capture; + dma_data = pcm->dma_capture; + + snd_soc_dai_set_dma_data(dai->cpu_dai, substream, dma_data); /* Strictly check for sample size */ switch (params_format(params)) { diff --git a/sound/soc/s3c24xx/s3c24xx-i2s.c b/sound/soc/s3c24xx/s3c24xx-i2s.c index 0bc5950..c3ac890 100644 --- a/sound/soc/s3c24xx/s3c24xx-i2s.c +++ b/sound/soc/s3c24xx/s3c24xx-i2s.c @@ -242,14 +242,17 @@ static int s3c24xx_i2s_hw_params(struct snd_pcm_substream *substream, struct snd_soc_dai *dai) { struct snd_soc_pcm_runtime *rtd = substream->private_data; + struct s3c_dma_params *dma_data; u32 iismod; pr_debug("Entered %s\n", __func__); if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) - rtd->dai->cpu_dai->dma_data = &s3c24xx_i2s_pcm_stereo_out; + dma_data = &s3c24xx_i2s_pcm_stereo_out; else - rtd->dai->cpu_dai->dma_data = &s3c24xx_i2s_pcm_stereo_in; + dma_data = &s3c24xx_i2s_pcm_stereo_in; + + snd_soc_dai_set_dma_data(rtd->dai->cpu_dai, substream, dma_data); /* Working copies of register */ iismod = readl(s3c24xx_i2s.regs + S3C2410_IISMOD); @@ -258,13 +261,11 @@ static int s3c24xx_i2s_hw_params(struct snd_pcm_substream *substream, switch (params_format(params)) { case SNDRV_PCM_FORMAT_S8: iismod &= ~S3C2410_IISMOD_16BIT; - ((struct s3c_dma_params *) - rtd->dai->cpu_dai->dma_data)->dma_size = 1; + dma_data->dma_size = 1; break; case SNDRV_PCM_FORMAT_S16_LE: iismod |= S3C2410_IISMOD_16BIT; - ((struct s3c_dma_params *) - rtd->dai->cpu_dai->dma_data)->dma_size = 2; + dma_data->dma_size = 2; break; default: return -EINVAL; @@ -280,8 +281,8 @@ static int s3c24xx_i2s_trigger(struct snd_pcm_substream *substream, int cmd, { int ret = 0; struct snd_soc_pcm_runtime *rtd = substream->private_data; - int channel = ((struct s3c_dma_params *) - rtd->dai->cpu_dai->dma_data)->channel; + struct s3c_dma_params *dma_data = + snd_soc_dai_get_dma_data(rtd->dai->cpu_dai, substream); pr_debug("Entered %s\n", __func__); @@ -300,7 +301,7 @@ static int s3c24xx_i2s_trigger(struct snd_pcm_substream *substream, int cmd, else s3c24xx_snd_txctrl(1); - s3c2410_dma_ctrl(channel, S3C2410_DMAOP_STARTED); + s3c2410_dma_ctrl(dma_data->channel, S3C2410_DMAOP_STARTED); break; case SNDRV_PCM_TRIGGER_STOP: case SNDRV_PCM_TRIGGER_SUSPEND: diff --git a/sound/soc/s6000/s6000-i2s.c b/sound/soc/s6000/s6000-i2s.c index c5cda18..fa23854 100644 --- a/sound/soc/s6000/s6000-i2s.c +++ b/sound/soc/s6000/s6000-i2s.c @@ -518,7 +518,8 @@ static int __devinit s6000_i2s_probe(struct platform_device *pdev) s6000_i2s_dai.dev = &pdev->dev; s6000_i2s_dai.private_data = dev; - s6000_i2s_dai.dma_data = &dev->dma_params; + s6000_i2s_dai.capture.dma_data = &dev->dma_params; + s6000_i2s_dai.playback.dma_data = &dev->dma_params; dev->sifbase = sifmem->start; dev->scbbase = mmio; diff --git a/sound/soc/s6000/s6000-pcm.c b/sound/soc/s6000/s6000-pcm.c index 1d61109..9c7f7f0 100644 --- a/sound/soc/s6000/s6000-pcm.c +++ b/sound/soc/s6000/s6000-pcm.c @@ -58,13 +58,15 @@ static void s6000_pcm_enqueue_dma(struct snd_pcm_substream *substream) struct snd_pcm_runtime *runtime = substream->runtime; struct s6000_runtime_data *prtd = runtime->private_data; struct snd_soc_pcm_runtime *soc_runtime = substream->private_data; - struct s6000_pcm_dma_params *par = soc_runtime->dai->cpu_dai->dma_data; + struct s6000_pcm_dma_params *par; int channel; unsigned int period_size; unsigned int dma_offset; dma_addr_t dma_pos; dma_addr_t src, dst; + par = snd_soc_dai_get_dma_data(soc_runtime->dai->cpu_dai, substream); + period_size = snd_pcm_lib_period_bytes(substream); dma_offset = prtd->period * period_size; dma_pos = runtime->dma_addr + dma_offset; @@ -101,7 +103,8 @@ static irqreturn_t s6000_pcm_irq(int irq, void *data) { struct snd_pcm *pcm = data; struct snd_soc_pcm_runtime *runtime = pcm->private_data; - struct s6000_pcm_dma_params *params = runtime->dai->cpu_dai->dma_data; + struct s6000_pcm_dma_params *params = + snd_soc_dai_get_dma_data(soc_runtime->dai->cpu_dai, substream); struct s6000_runtime_data *prtd; unsigned int has_xrun; int i, ret = IRQ_NONE; @@ -172,11 +175,13 @@ static int s6000_pcm_start(struct snd_pcm_substream *substream) { struct s6000_runtime_data *prtd = substream->runtime->private_data; struct snd_soc_pcm_runtime *soc_runtime = substream->private_data; - struct s6000_pcm_dma_params *par = soc_runtime->dai->cpu_dai->dma_data; + struct s6000_pcm_dma_params *par; unsigned long flags; int srcinc; u32 dma; + par = snd_soc_dai_get_dma_data(soc_runtime->dai->cpu_dai, substream); + spin_lock_irqsave(&prtd->lock, flags); if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) { @@ -212,10 +217,12 @@ static int s6000_pcm_stop(struct snd_pcm_substream *substream) { struct s6000_runtime_data *prtd = substream->runtime->private_data; struct snd_soc_pcm_runtime *soc_runtime = substream->private_data; - struct s6000_pcm_dma_params *par = soc_runtime->dai->cpu_dai->dma_data; + struct s6000_pcm_dma_params *par; unsigned long flags; u32 channel; + par = snd_soc_dai_get_dma_data(soc_runtime->dai->cpu_dai, substream); + if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) channel = par->dma_out; else @@ -236,9 +243,11 @@ static int s6000_pcm_stop(struct snd_pcm_substream *substream) static int s6000_pcm_trigger(struct snd_pcm_substream *substream, int cmd) { struct snd_soc_pcm_runtime *soc_runtime = substream->private_data; - struct s6000_pcm_dma_params *par = soc_runtime->dai->cpu_dai->dma_data; + struct s6000_pcm_dma_params *par; int ret; + par = snd_soc_dai_get_dma_data(soc_runtime->dai->cpu_dai, substream); + ret = par->trigger(substream, cmd, 0); if (ret < 0) return ret; @@ -275,13 +284,15 @@ static int s6000_pcm_prepare(struct snd_pcm_substream *substream) static snd_pcm_uframes_t s6000_pcm_pointer(struct snd_pcm_substream *substream) { struct snd_soc_pcm_runtime *soc_runtime = substream->private_data; - struct s6000_pcm_dma_params *par = soc_runtime->dai->cpu_dai->dma_data; + struct s6000_pcm_dma_params *par; struct snd_pcm_runtime *runtime = substream->runtime; struct s6000_runtime_data *prtd = runtime->private_data; unsigned long flags; unsigned int offset; dma_addr_t count; + par = snd_soc_dai_get_dma_data(soc_runtime->dai->cpu_dai, substream); + spin_lock_irqsave(&prtd->lock, flags); if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) @@ -305,11 +316,12 @@ static snd_pcm_uframes_t s6000_pcm_pointer(struct snd_pcm_substream *substream) static int s6000_pcm_open(struct snd_pcm_substream *substream) { struct snd_soc_pcm_runtime *soc_runtime = substream->private_data; - struct s6000_pcm_dma_params *par = soc_runtime->dai->cpu_dai->dma_data; + struct s6000_pcm_dma_params *par; struct snd_pcm_runtime *runtime = substream->runtime; struct s6000_runtime_data *prtd; int ret; + par = snd_soc_dai_get_dma_data(soc_runtime->dai->cpu_dai, substream); snd_soc_set_runtime_hwparams(substream, &s6000_pcm_hardware); ret = snd_pcm_hw_constraint_step(runtime, 0, @@ -364,7 +376,7 @@ static int s6000_pcm_hw_params(struct snd_pcm_substream *substream, struct snd_pcm_hw_params *hw_params) { struct snd_soc_pcm_runtime *soc_runtime = substream->private_data; - struct s6000_pcm_dma_params *par = soc_runtime->dai->cpu_dai->dma_data; + struct s6000_pcm_dma_params *par; int ret; ret = snd_pcm_lib_malloc_pages(substream, params_buffer_bytes(hw_params)); @@ -373,6 +385,8 @@ static int s6000_pcm_hw_params(struct snd_pcm_substream *substream, return ret; } + par = snd_soc_dai_get_dma_data(soc_runtime->dai->cpu_dai, substream); + if (par->same_rate) { spin_lock(&par->lock); if (par->rate == -1 || @@ -392,7 +406,8 @@ static int s6000_pcm_hw_params(struct snd_pcm_substream *substream, static int s6000_pcm_hw_free(struct snd_pcm_substream *substream) { struct snd_soc_pcm_runtime *soc_runtime = substream->private_data; - struct s6000_pcm_dma_params *par = soc_runtime->dai->cpu_dai->dma_data; + struct s6000_pcm_dma_params *par = + snd_soc_dai_get_dma_data(soc_runtime->dai->cpu_dai, substream); spin_lock(&par->lock); par->in_use &= ~(1 << substream->stream); @@ -417,7 +432,8 @@ static struct snd_pcm_ops s6000_pcm_ops = { static void s6000_pcm_free(struct snd_pcm *pcm) { struct snd_soc_pcm_runtime *runtime = pcm->private_data; - struct s6000_pcm_dma_params *params = runtime->dai->cpu_dai->dma_data; + struct s6000_pcm_dma_params *params = + snd_soc_dai_get_dma_data(soc_runtime->dai->cpu_dai, substream); free_irq(params->irq, pcm); snd_pcm_lib_preallocate_free_for_all(pcm); @@ -429,9 +445,11 @@ static int s6000_pcm_new(struct snd_card *card, struct snd_soc_dai *dai, struct snd_pcm *pcm) { struct snd_soc_pcm_runtime *runtime = pcm->private_data; - struct s6000_pcm_dma_params *params = runtime->dai->cpu_dai->dma_data; + struct s6000_pcm_dma_params *params; int res; + params = snd_soc_dai_get_dma_data(soc_runtime->dai->cpu_dai, substream); + if (!card->dev->dma_mask) card->dev->dma_mask = &s6000_pcm_dmamask; if (!card->dev->coherent_dma_mask) -- cgit v0.10.2 From ab6e24103cbd215e922938a4f58c75194761a60e Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 19 Mar 2010 14:38:13 +0000 Subject: Btrfs: fix data enospc check overflow Because we account for reserved space we get from the allocator before we actually account for allocating delalloc space, we can have a small window where the amount of "used" space in a space_info is more than the total amount of space in the space_info. This will cause a overflow in our check, so it will seem like we have _tons_ of free space, and we'll allow reservations to occur that will end up larger than the amount of space we have. I've seen users report ENOSPC panic's in cow_file_range a few times recently, so I tried to reproduce this problem and found I could reproduce it if I ran one of my tests in a loop for like 20 minutes. With this patch my test ran all night without issues. Thanks, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 101041d..786899d 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3234,7 +3234,8 @@ int btrfs_check_data_free_space(struct btrfs_root *root, struct inode *inode, u64 bytes) { struct btrfs_space_info *data_sinfo; - int ret = 0, committed = 0; + u64 used; + int ret = 0, committed = 0, flushed = 0; /* make sure bytes are sectorsize aligned */ bytes = (bytes + root->sectorsize - 1) & ~((u64)root->sectorsize - 1); @@ -3246,12 +3247,21 @@ int btrfs_check_data_free_space(struct btrfs_root *root, struct inode *inode, again: /* make sure we have enough space to handle the data first */ spin_lock(&data_sinfo->lock); - if (data_sinfo->total_bytes - data_sinfo->bytes_used - - data_sinfo->bytes_delalloc - data_sinfo->bytes_reserved - - data_sinfo->bytes_pinned - data_sinfo->bytes_readonly - - data_sinfo->bytes_may_use - data_sinfo->bytes_super < bytes) { + used = data_sinfo->bytes_used + data_sinfo->bytes_delalloc + + data_sinfo->bytes_reserved + data_sinfo->bytes_pinned + + data_sinfo->bytes_readonly + data_sinfo->bytes_may_use + + data_sinfo->bytes_super; + + if (used + bytes > data_sinfo->total_bytes) { struct btrfs_trans_handle *trans; + if (!flushed) { + spin_unlock(&data_sinfo->lock); + flush_delalloc(root, data_sinfo); + flushed = 1; + goto again; + } + /* * if we don't have enough free bytes in this space then we need * to alloc a new chunk. -- cgit v0.10.2 From f9700d5a4575e7fb343df10a1d29d425e4b81082 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 5 Apr 2010 23:25:13 +0200 Subject: ALSA: hda - Fix a wrong array range check in patch_realtek.c The commit 6a4f2ccb467e00281470cde2dee08fe5ecde62d1 introduced a wrong comparision for the array range check, which effectively skips the whole initialization of DAC connections. Fixed now. Reference: bko#15689 https://bugzilla.kernel.org/show_bug.cgi?id=15689 Reported-by: Adrian Ulrich Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 5472062..c7730db 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -10110,13 +10110,12 @@ static void alc882_auto_set_output_and_unmute(struct hda_codec *codec, int idx; alc_set_pin_output(codec, nid, pin_type); + if (dac_idx >= spec->multiout.num_dacs) + return; if (spec->multiout.dac_nids[dac_idx] == 0x25) idx = 4; - else { - if (spec->multiout.num_dacs >= dac_idx) - return; + else idx = spec->multiout.dac_nids[dac_idx] - 2; - } snd_hda_codec_write(codec, nid, 0, AC_VERB_SET_CONNECT_SEL, idx); } -- cgit v0.10.2 From bd6d29c25bb1a24a4c160ec5de43e0004e01f72b Mon Sep 17 00:00:00 2001 From: Frederic Weisbecker Date: Tue, 6 Apr 2010 00:10:17 +0200 Subject: lockstat: Make lockstat counting per cpu Locking statistics are implemented using global atomic variables. This is usually fine unless some path write them very often. This is the case for the function and function graph tracers that disable irqs for each entry saved (except if the function tracer is in preempt disabled only mode). And calls to local_irq_save/restore() increment hardirqs_on_events and hardirqs_off_events stats (or similar stats for redundant versions). Incrementing these global vars for each function ends up in too much cache bouncing if lockstats are enabled. To solve this, implement the debug_atomic_*() operations using per cpu vars. -v2: Use per_cpu() instead of get_cpu_var() to fetch the desired cpu vars on debug_atomic_read() -v3: Store the stats in a structure. No need for local_t as we are NMI/irq safe. -v4: Fix tons of build errors. I thought I had tested it but I probably forgot to select the relevant config. Suggested-by: Steven Rostedt Signed-off-by: Frederic Weisbecker Cc: Peter Zijlstra Cc: Steven Rostedt LKML-Reference: <1270505417-8144-1-git-send-regression-fweisbec@gmail.com> Signed-off-by: Ingo Molnar Cc: Peter Zijlstra Cc: Steven Rostedt diff --git a/kernel/lockdep.c b/kernel/lockdep.c index 0c30d04..069af02 100644 --- a/kernel/lockdep.c +++ b/kernel/lockdep.c @@ -430,20 +430,7 @@ static struct stack_trace lockdep_init_trace = { /* * Various lockdep statistics: */ -atomic_t chain_lookup_hits; -atomic_t chain_lookup_misses; -atomic_t hardirqs_on_events; -atomic_t hardirqs_off_events; -atomic_t redundant_hardirqs_on; -atomic_t redundant_hardirqs_off; -atomic_t softirqs_on_events; -atomic_t softirqs_off_events; -atomic_t redundant_softirqs_on; -atomic_t redundant_softirqs_off; -atomic_t nr_unused_locks; -atomic_t nr_cyclic_checks; -atomic_t nr_find_usage_forwards_checks; -atomic_t nr_find_usage_backwards_checks; +DEFINE_PER_CPU(struct lockdep_stats, lockdep_stats); #endif /* @@ -758,7 +745,7 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) return NULL; } class = lock_classes + nr_lock_classes++; - debug_atomic_inc(&nr_unused_locks); + debug_atomic_inc(nr_unused_locks); class->key = key; class->name = lock->name; class->subclass = subclass; @@ -1215,7 +1202,7 @@ check_noncircular(struct lock_list *root, struct lock_class *target, { int result; - debug_atomic_inc(&nr_cyclic_checks); + debug_atomic_inc(nr_cyclic_checks); result = __bfs_forwards(root, target, class_equal, target_entry); @@ -1252,7 +1239,7 @@ find_usage_forwards(struct lock_list *root, enum lock_usage_bit bit, { int result; - debug_atomic_inc(&nr_find_usage_forwards_checks); + debug_atomic_inc(nr_find_usage_forwards_checks); result = __bfs_forwards(root, (void *)bit, usage_match, target_entry); @@ -1275,7 +1262,7 @@ find_usage_backwards(struct lock_list *root, enum lock_usage_bit bit, { int result; - debug_atomic_inc(&nr_find_usage_backwards_checks); + debug_atomic_inc(nr_find_usage_backwards_checks); result = __bfs_backwards(root, (void *)bit, usage_match, target_entry); @@ -1835,7 +1822,7 @@ static inline int lookup_chain_cache(struct task_struct *curr, list_for_each_entry(chain, hash_head, entry) { if (chain->chain_key == chain_key) { cache_hit: - debug_atomic_inc(&chain_lookup_hits); + debug_atomic_inc(chain_lookup_hits); if (very_verbose(class)) printk("\nhash chain already cached, key: " "%016Lx tail class: [%p] %s\n", @@ -1900,7 +1887,7 @@ cache_hit: chain_hlocks[chain->base + j] = class - lock_classes; } list_add_tail_rcu(&chain->entry, hash_head); - debug_atomic_inc(&chain_lookup_misses); + debug_atomic_inc(chain_lookup_misses); inc_chains(); return 1; @@ -2321,7 +2308,7 @@ void trace_hardirqs_on_caller(unsigned long ip) return; if (unlikely(curr->hardirqs_enabled)) { - debug_atomic_inc(&redundant_hardirqs_on); + debug_atomic_inc(redundant_hardirqs_on); return; } /* we'll do an OFF -> ON transition: */ @@ -2348,7 +2335,7 @@ void trace_hardirqs_on_caller(unsigned long ip) curr->hardirq_enable_ip = ip; curr->hardirq_enable_event = ++curr->irq_events; - debug_atomic_inc(&hardirqs_on_events); + debug_atomic_inc(hardirqs_on_events); } EXPORT_SYMBOL(trace_hardirqs_on_caller); @@ -2380,9 +2367,9 @@ void trace_hardirqs_off_caller(unsigned long ip) curr->hardirqs_enabled = 0; curr->hardirq_disable_ip = ip; curr->hardirq_disable_event = ++curr->irq_events; - debug_atomic_inc(&hardirqs_off_events); + debug_atomic_inc(hardirqs_off_events); } else - debug_atomic_inc(&redundant_hardirqs_off); + debug_atomic_inc(redundant_hardirqs_off); } EXPORT_SYMBOL(trace_hardirqs_off_caller); @@ -2406,7 +2393,7 @@ void trace_softirqs_on(unsigned long ip) return; if (curr->softirqs_enabled) { - debug_atomic_inc(&redundant_softirqs_on); + debug_atomic_inc(redundant_softirqs_on); return; } @@ -2416,7 +2403,7 @@ void trace_softirqs_on(unsigned long ip) curr->softirqs_enabled = 1; curr->softirq_enable_ip = ip; curr->softirq_enable_event = ++curr->irq_events; - debug_atomic_inc(&softirqs_on_events); + debug_atomic_inc(softirqs_on_events); /* * We are going to turn softirqs on, so set the * usage bit for all held locks, if hardirqs are @@ -2446,10 +2433,10 @@ void trace_softirqs_off(unsigned long ip) curr->softirqs_enabled = 0; curr->softirq_disable_ip = ip; curr->softirq_disable_event = ++curr->irq_events; - debug_atomic_inc(&softirqs_off_events); + debug_atomic_inc(softirqs_off_events); DEBUG_LOCKS_WARN_ON(!softirq_count()); } else - debug_atomic_inc(&redundant_softirqs_off); + debug_atomic_inc(redundant_softirqs_off); } static void __lockdep_trace_alloc(gfp_t gfp_mask, unsigned long flags) @@ -2654,7 +2641,7 @@ static int mark_lock(struct task_struct *curr, struct held_lock *this, return 0; break; case LOCK_USED: - debug_atomic_dec(&nr_unused_locks); + debug_atomic_dec(nr_unused_locks); break; default: if (!debug_locks_off_graph_unlock()) @@ -2760,7 +2747,7 @@ static int __lock_acquire(struct lockdep_map *lock, unsigned int subclass, if (!class) return 0; } - debug_atomic_inc((atomic_t *)&class->ops); + atomic_inc((atomic_t *)&class->ops); if (very_verbose(class)) { printk("\nacquire class [%p] %s", class->key, class->name); if (class->name_version > 1) diff --git a/kernel/lockdep_internals.h b/kernel/lockdep_internals.h index a2ee95a..8d7d4b6 100644 --- a/kernel/lockdep_internals.h +++ b/kernel/lockdep_internals.h @@ -110,29 +110,61 @@ lockdep_count_backward_deps(struct lock_class *class) #endif #ifdef CONFIG_DEBUG_LOCKDEP + +#include /* - * Various lockdep statistics: + * Various lockdep statistics. + * We want them per cpu as they are often accessed in fast path + * and we want to avoid too much cache bouncing. */ -extern atomic_t chain_lookup_hits; -extern atomic_t chain_lookup_misses; -extern atomic_t hardirqs_on_events; -extern atomic_t hardirqs_off_events; -extern atomic_t redundant_hardirqs_on; -extern atomic_t redundant_hardirqs_off; -extern atomic_t softirqs_on_events; -extern atomic_t softirqs_off_events; -extern atomic_t redundant_softirqs_on; -extern atomic_t redundant_softirqs_off; -extern atomic_t nr_unused_locks; -extern atomic_t nr_cyclic_checks; -extern atomic_t nr_cyclic_check_recursions; -extern atomic_t nr_find_usage_forwards_checks; -extern atomic_t nr_find_usage_forwards_recursions; -extern atomic_t nr_find_usage_backwards_checks; -extern atomic_t nr_find_usage_backwards_recursions; -# define debug_atomic_inc(ptr) atomic_inc(ptr) -# define debug_atomic_dec(ptr) atomic_dec(ptr) -# define debug_atomic_read(ptr) atomic_read(ptr) +struct lockdep_stats { + int chain_lookup_hits; + int chain_lookup_misses; + int hardirqs_on_events; + int hardirqs_off_events; + int redundant_hardirqs_on; + int redundant_hardirqs_off; + int softirqs_on_events; + int softirqs_off_events; + int redundant_softirqs_on; + int redundant_softirqs_off; + int nr_unused_locks; + int nr_cyclic_checks; + int nr_cyclic_check_recursions; + int nr_find_usage_forwards_checks; + int nr_find_usage_forwards_recursions; + int nr_find_usage_backwards_checks; + int nr_find_usage_backwards_recursions; +}; + +DECLARE_PER_CPU(struct lockdep_stats, lockdep_stats); + +#define debug_atomic_inc(ptr) { \ + struct lockdep_stats *__cpu_lockdep_stats; \ + \ + WARN_ON_ONCE(!irqs_disabled()); \ + __cpu_lockdep_stats = &__get_cpu_var(lockdep_stats); \ + __cpu_lockdep_stats->ptr++; \ +} + +#define debug_atomic_dec(ptr) { \ + struct lockdep_stats *__cpu_lockdep_stats; \ + \ + WARN_ON_ONCE(!irqs_disabled()); \ + __cpu_lockdep_stats = &__get_cpu_var(lockdep_stats); \ + __cpu_lockdep_stats->ptr--; \ +} + +#define debug_atomic_read(ptr) ({ \ + struct lockdep_stats *__cpu_lockdep_stats; \ + unsigned long long __total = 0; \ + int __cpu; \ + for_each_possible_cpu(__cpu) { \ + __cpu_lockdep_stats = &per_cpu(lockdep_stats, __cpu); \ + __total += __cpu_lockdep_stats->ptr; \ + } \ + __total; \ +}) #else # define debug_atomic_inc(ptr) do { } while (0) # define debug_atomic_dec(ptr) do { } while (0) diff --git a/kernel/lockdep_proc.c b/kernel/lockdep_proc.c index d4aba4f..59b76c8 100644 --- a/kernel/lockdep_proc.c +++ b/kernel/lockdep_proc.c @@ -184,34 +184,34 @@ static const struct file_operations proc_lockdep_chains_operations = { static void lockdep_stats_debug_show(struct seq_file *m) { #ifdef CONFIG_DEBUG_LOCKDEP - unsigned int hi1 = debug_atomic_read(&hardirqs_on_events), - hi2 = debug_atomic_read(&hardirqs_off_events), - hr1 = debug_atomic_read(&redundant_hardirqs_on), - hr2 = debug_atomic_read(&redundant_hardirqs_off), - si1 = debug_atomic_read(&softirqs_on_events), - si2 = debug_atomic_read(&softirqs_off_events), - sr1 = debug_atomic_read(&redundant_softirqs_on), - sr2 = debug_atomic_read(&redundant_softirqs_off); - - seq_printf(m, " chain lookup misses: %11u\n", - debug_atomic_read(&chain_lookup_misses)); - seq_printf(m, " chain lookup hits: %11u\n", - debug_atomic_read(&chain_lookup_hits)); - seq_printf(m, " cyclic checks: %11u\n", - debug_atomic_read(&nr_cyclic_checks)); - seq_printf(m, " find-mask forwards checks: %11u\n", - debug_atomic_read(&nr_find_usage_forwards_checks)); - seq_printf(m, " find-mask backwards checks: %11u\n", - debug_atomic_read(&nr_find_usage_backwards_checks)); - - seq_printf(m, " hardirq on events: %11u\n", hi1); - seq_printf(m, " hardirq off events: %11u\n", hi2); - seq_printf(m, " redundant hardirq ons: %11u\n", hr1); - seq_printf(m, " redundant hardirq offs: %11u\n", hr2); - seq_printf(m, " softirq on events: %11u\n", si1); - seq_printf(m, " softirq off events: %11u\n", si2); - seq_printf(m, " redundant softirq ons: %11u\n", sr1); - seq_printf(m, " redundant softirq offs: %11u\n", sr2); + unsigned long long hi1 = debug_atomic_read(hardirqs_on_events), + hi2 = debug_atomic_read(hardirqs_off_events), + hr1 = debug_atomic_read(redundant_hardirqs_on), + hr2 = debug_atomic_read(redundant_hardirqs_off), + si1 = debug_atomic_read(softirqs_on_events), + si2 = debug_atomic_read(softirqs_off_events), + sr1 = debug_atomic_read(redundant_softirqs_on), + sr2 = debug_atomic_read(redundant_softirqs_off); + + seq_printf(m, " chain lookup misses: %11llu\n", + debug_atomic_read(chain_lookup_misses)); + seq_printf(m, " chain lookup hits: %11llu\n", + debug_atomic_read(chain_lookup_hits)); + seq_printf(m, " cyclic checks: %11llu\n", + debug_atomic_read(nr_cyclic_checks)); + seq_printf(m, " find-mask forwards checks: %11llu\n", + debug_atomic_read(nr_find_usage_forwards_checks)); + seq_printf(m, " find-mask backwards checks: %11llu\n", + debug_atomic_read(nr_find_usage_backwards_checks)); + + seq_printf(m, " hardirq on events: %11llu\n", hi1); + seq_printf(m, " hardirq off events: %11llu\n", hi2); + seq_printf(m, " redundant hardirq ons: %11llu\n", hr1); + seq_printf(m, " redundant hardirq offs: %11llu\n", hr2); + seq_printf(m, " softirq on events: %11llu\n", si1); + seq_printf(m, " softirq off events: %11llu\n", si2); + seq_printf(m, " redundant softirq ons: %11llu\n", sr1); + seq_printf(m, " redundant softirq offs: %11llu\n", sr2); #endif } @@ -263,7 +263,7 @@ static int lockdep_stats_show(struct seq_file *m, void *v) #endif } #ifdef CONFIG_DEBUG_LOCKDEP - DEBUG_LOCKS_WARN_ON(debug_atomic_read(&nr_unused_locks) != nr_unused); + DEBUG_LOCKS_WARN_ON(debug_atomic_read(nr_unused_locks) != nr_unused); #endif seq_printf(m, " lock-classes: %11lu [max: %lu]\n", nr_lock_classes, MAX_LOCKDEP_KEYS); -- cgit v0.10.2 From d99c3871148cb643cffe138e9427b7a743aca367 Mon Sep 17 00:00:00 2001 From: Kevin Hilman Date: Thu, 11 Mar 2010 14:57:35 -0800 Subject: davinci: timers: don't enable timer until clocksource is initialized On da830, when the same timer is used for clocksource and clockevent, the timer can be started before the clockevent is registered/initialzed. This creates a window where a timer interrupt might fire before the clockevent handler has been setup and causes a crash. This patch moves the actual enable/start of the timer after the clockevent has ben registered. Signed-off-by: Kevin Hilman diff --git a/arch/arm/mach-davinci/time.c b/arch/arm/mach-davinci/time.c index 42d985b..9e0b106 100644 --- a/arch/arm/mach-davinci/time.c +++ b/arch/arm/mach-davinci/time.c @@ -253,8 +253,6 @@ static void __init timer_init(void) irq = USING_COMPARE(t) ? dtip[i].cmp_irq : irq; setup_irq(irq, &t->irqaction); } - - timer32_config(&timers[i]); } } @@ -331,6 +329,7 @@ static void __init davinci_timer_init(void) unsigned int clocksource_id; static char err[] __initdata = KERN_ERR "%s: can't register clocksource!\n"; + int i; clockevent_id = soc_info->timer_info->clockevent_id; clocksource_id = soc_info->timer_info->clocksource_id; @@ -389,6 +388,9 @@ static void __init davinci_timer_init(void) clockevent_davinci.cpumask = cpumask_of(0); clockevents_register_device(&clockevent_davinci); + + for (i=0; i< ARRAY_SIZE(timers); i++) + timer32_config(&timers[i]); } struct sys_timer davinci_timer = { -- cgit v0.10.2 From 5eb2e891f038c8e04757f4536b9693faf35978a3 Mon Sep 17 00:00:00 2001 From: Sekhar Nori Date: Wed, 10 Mar 2010 15:11:24 +0530 Subject: davinci: da8xx/omap-l1: fix build error when CONFIG_DAVINCI_MUX is undefined The da8xx/omap-l1 boards refuse to build when CONFIG_DAVINCI_MUX is undefined because arch/arm/mach-davinci/mux.c:da8xx_pinmux_setup() is not defined. This patch fixes this issue. This is build tested with davinci_all_defconfig and da8xx_omapl_defconfig and boot tested on DA830 EVM. Reported-by: Shanmuga Sundaram Mahendran Signed-off-by: Sekhar Nori Signed-off-by: Kevin Hilman diff --git a/arch/arm/mach-davinci/include/mach/da8xx.h b/arch/arm/mach-davinci/include/mach/da8xx.h index cc9be7f..b87a6ba 100644 --- a/arch/arm/mach-davinci/include/mach/da8xx.h +++ b/arch/arm/mach-davinci/include/mach/da8xx.h @@ -144,6 +144,10 @@ extern const short da850_mmcsd0_pins[]; extern const short da850_nand_pins[]; extern const short da850_nor_pins[]; +#ifdef CONFIG_DAVINCI_MUX int da8xx_pinmux_setup(const short pins[]); +#else +static inline int da8xx_pinmux_setup(const short pins[]) { return 0; } +#endif #endif /* __ASM_ARCH_DAVINCI_DA8XX_H */ -- cgit v0.10.2 From bb17ef102b72e38557bfa829511e06c4bd3bb972 Mon Sep 17 00:00:00 2001 From: Brian Niebuhr Date: Tue, 9 Mar 2010 16:48:03 -0600 Subject: davinci: edma: clear events in edma_start() This patch fixes an issue where a DMA channel can erroneously process an event generated by a previous transfer. A failure case is where DMA is being used for SPI transmit and receive channels on OMAP L138. In this case there is a single bit that controls all event generation from the SPI peripheral. Therefore it is possible that between when edma_stop() has been called for the transmit channel on a previous transfer and edma_start() is called for the transmit channel on a subsequent transfer, that a transmit event has been generated. The fix is to clear events in edma_start(). This prevents false events from being processed when events are enabled for that channel. Signed-off-by: Brian Niebuhr Signed-off-by: Kevin Hilman diff --git a/arch/arm/mach-davinci/dma.c b/arch/arm/mach-davinci/dma.c index 15dd886..260485c 100644 --- a/arch/arm/mach-davinci/dma.c +++ b/arch/arm/mach-davinci/dma.c @@ -1266,7 +1266,8 @@ int edma_start(unsigned channel) /* EDMA channel with event association */ pr_debug("EDMA: ER%d %08x\n", j, edma_shadow0_read_array(ctlr, SH_ER, j)); - /* Clear any pending error */ + /* Clear any pending event or error */ + edma_write_array(ctlr, EDMA_ECR, j, mask); edma_write_array(ctlr, EDMA_EMCR, j, mask); /* Clear any SER */ edma_shadow0_write_array(ctlr, SH_SECR, j, mask); -- cgit v0.10.2 From c8e42de6efcf4e9bed79a8b030261d799ea1960a Mon Sep 17 00:00:00 2001 From: Kevin Hilman Date: Thu, 25 Feb 2010 16:24:41 -0800 Subject: davinci: DM365: fix duplicate default IRQ priorities IRQ 29 has two possible interrupts DDRINT and RTC, but having both in the default priority table is confusing (and triggers a warning from sparse.) This patch removes the lower priority DDRINT from the default priority table leaving the RTC setting as the default. Signed-off-by: Kevin Hilman diff --git a/arch/arm/mach-davinci/dm365.c b/arch/arm/mach-davinci/dm365.c index 27772e1..0d6ee58 100644 --- a/arch/arm/mach-davinci/dm365.c +++ b/arch/arm/mach-davinci/dm365.c @@ -758,7 +758,6 @@ static u8 dm365_default_priorities[DAVINCI_N_AINTC_IRQ] = { [IRQ_MMCINT] = 7, [IRQ_DM365_MMCINT1] = 7, [IRQ_DM365_PWMINT3] = 7, - [IRQ_DDRINT] = 4, [IRQ_AEMIFINT] = 2, [IRQ_DM365_SDIOINT1] = 2, [IRQ_TINT0_TINT12] = 7, -- cgit v0.10.2 From 57b54ea6b7863ccfeb41851b5f58f9fd1b83c79e Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Michel=20D=C3=A4nzer?= Date: Fri, 2 Apr 2010 16:59:06 +0000 Subject: drm/radeon: R300 AD only has one quad pipe. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Gleaned from the Mesa code. Fixes https://bugs.freedesktop.org/show_bug.cgi?id=27355 . Signed-off-by: Michel Dänzer Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/r300.c b/drivers/gpu/drm/radeon/r300.c index 1023eeb..0e9eb76 100644 --- a/drivers/gpu/drm/radeon/r300.c +++ b/drivers/gpu/drm/radeon/r300.c @@ -324,11 +324,12 @@ void r300_gpu_init(struct radeon_device *rdev) r100_hdp_reset(rdev); /* FIXME: rv380 one pipes ? */ - if ((rdev->family == CHIP_R300) || (rdev->family == CHIP_R350)) { + if ((rdev->family == CHIP_R300 && rdev->pdev->device != 0x4144) || + (rdev->family == CHIP_R350)) { /* r300,r350 */ rdev->num_gb_pipes = 2; } else { - /* rv350,rv370,rv380 */ + /* rv350,rv370,rv380,r300 AD */ rdev->num_gb_pipes = 1; } rdev->num_z_pipes = 1; diff --git a/drivers/gpu/drm/radeon/radeon_cp.c b/drivers/gpu/drm/radeon/radeon_cp.c index dc6eba6..419630d 100644 --- a/drivers/gpu/drm/radeon/radeon_cp.c +++ b/drivers/gpu/drm/radeon/radeon_cp.c @@ -417,8 +417,9 @@ static int radeon_do_wait_for_idle(drm_radeon_private_t * dev_priv) return -EBUSY; } -static void radeon_init_pipes(drm_radeon_private_t *dev_priv) +static void radeon_init_pipes(struct drm_device *dev) { + drm_radeon_private_t *dev_priv = dev->dev_private; uint32_t gb_tile_config, gb_pipe_sel = 0; if ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV530) { @@ -436,11 +437,12 @@ static void radeon_init_pipes(drm_radeon_private_t *dev_priv) dev_priv->num_gb_pipes = ((gb_pipe_sel >> 12) & 0x3) + 1; } else { /* R3xx */ - if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_R300) || + if (((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_R300 && + dev->pdev->device != 0x4144) || ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_R350)) { dev_priv->num_gb_pipes = 2; } else { - /* R3Vxx */ + /* RV3xx/R300 AD */ dev_priv->num_gb_pipes = 1; } } @@ -736,7 +738,7 @@ static int radeon_do_engine_reset(struct drm_device * dev) /* setup the raster pipes */ if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_R300) - radeon_init_pipes(dev_priv); + radeon_init_pipes(dev); /* Reset the CP ring */ radeon_do_cp_reset(dev_priv); -- cgit v0.10.2 From 95beb690170e6ce918fe53c73a0fcc7cf64d704a Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Thu, 1 Apr 2010 19:08:47 +0000 Subject: drm/radeon/kms/atom: fix gpio i2c table overrun (v2) The GPIO_I2C_INFO table does not always have ATOM_MAX_SUPPORTED_DEVICE entries. Limit the number of indices to the size of the table. Should fix Novell bug 589022. v2: fix typo Signed-off-by: Alex Deucher Cc: Stable Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c index 1fff955..5673665 100644 --- a/drivers/gpu/drm/radeon/radeon_atombios.c +++ b/drivers/gpu/drm/radeon/radeon_atombios.c @@ -69,16 +69,19 @@ static inline struct radeon_i2c_bus_rec radeon_lookup_i2c_gpio(struct radeon_dev struct radeon_i2c_bus_rec i2c; int index = GetIndexIntoMasterTable(DATA, GPIO_I2C_Info); struct _ATOM_GPIO_I2C_INFO *i2c_info; - uint16_t data_offset; - int i; + uint16_t data_offset, size; + int i, num_indices; memset(&i2c, 0, sizeof(struct radeon_i2c_bus_rec)); i2c.valid = false; - if (atom_parse_data_header(ctx, index, NULL, NULL, NULL, &data_offset)) { + if (atom_parse_data_header(ctx, index, &size, NULL, NULL, &data_offset)) { i2c_info = (struct _ATOM_GPIO_I2C_INFO *)(ctx->bios + data_offset); - for (i = 0; i < ATOM_MAX_SUPPORTED_DEVICE; i++) { + num_indices = (size - sizeof(ATOM_COMMON_TABLE_HEADER)) / + sizeof(ATOM_GPIO_I2C_ASSIGMENT); + + for (i = 0; i < num_indices; i++) { gpio = &i2c_info->asGPIO_Info[i]; if (gpio->sucI2cId.ucAccess == id) { -- cgit v0.10.2 From 643acacf02679befd0f98ac3c5fecb805f1c9548 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Mon, 5 Apr 2010 23:57:52 -0400 Subject: drm/radeon/kms: fix washed out image on legacy tv dac bad cast was overwriting the tvdac adj values Fixes fdo bug 27478 Signed-off-by: Alex Deucher Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c index 60d5981..3fba505 100644 --- a/drivers/gpu/drm/radeon/radeon_connectors.c +++ b/drivers/gpu/drm/radeon/radeon_connectors.c @@ -315,7 +315,7 @@ int radeon_connector_set_property(struct drm_connector *connector, struct drm_pr radeon_encoder = to_radeon_encoder(encoder); if (!radeon_encoder->enc_priv) return 0; - if (rdev->is_atom_bios) { + if (ASIC_IS_AVIVO(rdev) || radeon_r4xx_atom) { struct radeon_encoder_atom_dac *dac_int; dac_int = radeon_encoder->enc_priv; dac_int->tv_std = val; -- cgit v0.10.2 From 2c860a1101471a69f7a6778b7b1fb43344c38619 Mon Sep 17 00:00:00 2001 From: Dominik Brodowski Date: Mon, 5 Apr 2010 22:29:09 -0700 Subject: Input: i8042 - spelling fix Signed-off-by: Dominik Brodowski Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/serio/i8042.c b/drivers/input/serio/i8042.c index fce8ab7..90e5668 100644 --- a/drivers/input/serio/i8042.c +++ b/drivers/input/serio/i8042.c @@ -38,7 +38,7 @@ MODULE_PARM_DESC(noaux, "Do not probe or use AUX (mouse) port."); static bool i8042_nomux; module_param_named(nomux, i8042_nomux, bool, 0); -MODULE_PARM_DESC(nomux, "Do not check whether an active multiplexing conrtoller is present."); +MODULE_PARM_DESC(nomux, "Do not check whether an active multiplexing controller is present."); static bool i8042_unlock; module_param_named(unlock, i8042_unlock, bool, 0); -- cgit v0.10.2 From 5e28d8eb68c12eab9c4a47b42ba993a6420d71d3 Mon Sep 17 00:00:00 2001 From: Chase Douglas Date: Mon, 5 Apr 2010 22:29:08 -0700 Subject: Input: ALPS - add signature for HP Pavilion dm3 laptops Tested by a user running Ubuntu 9.10 in the following bug report. BugLink: http://bugs.launchpad.net/bugs/545307 Signed-off-by: Chase Douglas Cc: stable@kernel.org Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/mouse/alps.c b/drivers/input/mouse/alps.c index 7490f1d..2a791e0 100644 --- a/drivers/input/mouse/alps.c +++ b/drivers/input/mouse/alps.c @@ -63,6 +63,7 @@ static const struct alps_model_info alps_model_data[] = { { { 0x62, 0x02, 0x14 }, 0xcf, 0xcf, ALPS_PASS | ALPS_DUALPOINT | ALPS_PS2_INTERLEAVED }, { { 0x73, 0x02, 0x50 }, 0xcf, 0xcf, ALPS_FOUR_BUTTONS }, /* Dell Vostro 1400 */ + { { 0x73, 0x02, 0x64 }, 0xf8, 0xf8, 0 }, /* HP Pavilion dm3 */ { { 0x52, 0x01, 0x14 }, 0xff, 0xff, ALPS_PASS | ALPS_DUALPOINT | ALPS_PS2_INTERLEAVED }, /* Toshiba Tecra A11-11L */ }; -- cgit v0.10.2 From 9d32c30542f9ecdb4b96a1a960924c9f403e3562 Mon Sep 17 00:00:00 2001 From: H Hartley Sweeten Date: Mon, 5 Apr 2010 22:29:09 -0700 Subject: Input: matrix_keypad - allow platform to disable key autorepeat In an embedded system the matrix_keypad driver might be used to interface with an external control panel and not an actual keyboard. On the control panel some of the keys could be used to turn on/off various functions. If key autorepeat is enabled this causes the function to quickly toggle between the on and off states and makes operation difficult. Add an option in the platform-specific data to disable the key autorepeat. Signed-off-by: H Hartley Sweeten Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/keyboard/matrix_keypad.c b/drivers/input/keyboard/matrix_keypad.c index d3c8b61..f9d86cf 100644 --- a/drivers/input/keyboard/matrix_keypad.c +++ b/drivers/input/keyboard/matrix_keypad.c @@ -373,7 +373,9 @@ static int __devinit matrix_keypad_probe(struct platform_device *pdev) input_dev->name = pdev->name; input_dev->id.bustype = BUS_HOST; input_dev->dev.parent = &pdev->dev; - input_dev->evbit[0] = BIT_MASK(EV_KEY) | BIT_MASK(EV_REP); + input_dev->evbit[0] = BIT_MASK(EV_KEY); + if (!pdata->no_autorepeat) + input_dev->evbit[0] |= BIT_MASK(EV_REP); input_dev->open = matrix_keypad_start; input_dev->close = matrix_keypad_stop; diff --git a/include/linux/input/matrix_keypad.h b/include/linux/input/matrix_keypad.h index 3bd018b..c964cd7 100644 --- a/include/linux/input/matrix_keypad.h +++ b/include/linux/input/matrix_keypad.h @@ -44,6 +44,7 @@ struct matrix_keymap_data { * @active_low: gpio polarity * @wakeup: controls whether the device should be set up as wakeup * source + * @no_autorepeat: disable key autorepeat * * This structure represents platform-specific data that use used by * matrix_keypad driver to perform proper initialization. @@ -64,6 +65,7 @@ struct matrix_keypad_platform_data { bool active_low; bool wakeup; + bool no_autorepeat; }; /** -- cgit v0.10.2 From 84fba5ec91f11c0efb27d0ed6098f7447491f0df Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Tue, 6 Apr 2010 17:02:19 +1000 Subject: sched: Fix sched_getaffinity() taskset on 2.6.34-rc3 fails on one of my ppc64 test boxes with the following error: sched_getaffinity(0, 16, 0x10029650030) = -1 EINVAL (Invalid argument) This box has 128 threads and 16 bytes is enough to cover it. Commit cd3d8031eb4311e516329aee03c79a08333141f1 (sched: sched_getaffinity(): Allow less than NR_CPUS length) is comparing this 16 bytes agains nr_cpu_ids. Fix it by comparing nr_cpu_ids to the number of bits in the cpumask we pass in. Signed-off-by: Anton Blanchard Reviewed-by: KOSAKI Motohiro Cc: Sharyathi Nagesh Cc: Ulrich Drepper Cc: Peter Zijlstra Cc: Linus Torvalds Cc: Jack Steiner Cc: Russ Anderson Cc: Mike Travis LKML-Reference: <20100406070218.GM5594@kryten> Signed-off-by: Ingo Molnar diff --git a/kernel/sched.c b/kernel/sched.c index 528a105..eaf5c73 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -4902,7 +4902,7 @@ SYSCALL_DEFINE3(sched_getaffinity, pid_t, pid, unsigned int, len, int ret; cpumask_var_t mask; - if (len < nr_cpu_ids) + if ((len * BITS_PER_BYTE) < nr_cpu_ids) return -EINVAL; if (len & (sizeof(unsigned long)-1)) return -EINVAL; -- cgit v0.10.2 From 9f680ce04ea19dabbbafe01b57b61930a9b70741 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Tue, 6 Apr 2010 09:37:47 -0400 Subject: Btrfs: make sure the chunk allocator doesn't create zero length chunks A recent commit allowed for smaller chunks to be created, but didn't make sure they were always bigger than a stripe. After some divides, this led to zero length stripes. Signed-off-by: Chris Mason diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 9bf1f58..b584e9a 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -2249,6 +2249,12 @@ again: if (!looped) calc_size = max_t(u64, min_stripe_size, calc_size); + /* + * we're about to do_div by the stripe_len so lets make sure + * we end up with something bigger than a stripe + */ + calc_size = max_t(u64, calc_size, stripe_len * 4); + do_div(calc_size, stripe_len); calc_size *= stripe_len; -- cgit v0.10.2 From 8c40041f75a202ed6a3b38143b823cb80f6d6b7c Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 6 Apr 2010 10:37:33 -0300 Subject: perf kmem: Fix breakage introduced by 5a0e3ad slab.h script MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit 5a0e3ad ("include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h") added a '#include ' to tools/perf/builtin-kmem.h because: that tool has lines like this: if (!strcmp(event->name, "kmalloc") || !strcmp(event->name, "kmem_cache_alloc")) { process_alloc_event(data, event, cpu, timestamp, thread, 0); return; } So, using the script regex: >>> import re >>> s = re.compile(r'^(|.*[^a-zA-Z0-9_])_*(slab_is_available|kmem_cache_|k[mzc]alloc|krealloc|kz?free|ksize|__getname|putname)') >>> l = ' !strcmp(event->name, "kmem_cache_alloc")) {' >>> s.search(l) <_sre.SRE_Match object at 0xb77b1ad0> >>> Remove that file that is not available in the tools/perf include path and thus builtin-kmem.c couldn't be compiled. Reported-by: Peter Zijlstra Cc: Christoph Lameter Cc: Frédéric Weisbecker Cc: Lee Schermerhorn Cc: Mike Galbraith Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Tejun Heo Cc: Linus Torvalds LKML-Reference: <1270561053-14308-1-git-send-email-acme@infradead.org> Signed-off-by: Ingo Molnar diff --git a/tools/perf/builtin-kmem.c b/tools/perf/builtin-kmem.c index 7d9e3a7..924a951 100644 --- a/tools/perf/builtin-kmem.c +++ b/tools/perf/builtin-kmem.c @@ -14,7 +14,6 @@ #include "util/debug.h" #include -#include struct alloc_stat; typedef int (*sort_fn_t)(struct alloc_stat *, struct alloc_stat *); -- cgit v0.10.2 From 134fbadf028a5977a1b06b0253d3ee33e6f0c642 Mon Sep 17 00:00:00 2001 From: Vince Weaver Date: Tue, 6 Apr 2010 10:01:19 -0400 Subject: perf, x86: Enable Nehalem-EX support According to Intel Software Devel Manual Volume 3B, the Nehalem-EX PMU is just like regular Nehalem (except for the uncore support, which is completely different). Signed-off-by: Vince Weaver Cc: Peter Zijlstra Cc: Paul Mackerras Cc: Arnaldo Carvalho de Melo Cc: "H. Peter Anvin" Cc: Arjan van de Ven Cc: Lin Ming LKML-Reference: Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c index 84bfde6..9c794ac 100644 --- a/arch/x86/kernel/cpu/perf_event_intel.c +++ b/arch/x86/kernel/cpu/perf_event_intel.c @@ -936,6 +936,7 @@ static __init int intel_pmu_init(void) case 26: /* 45 nm nehalem, "Bloomfield" */ case 30: /* 45 nm nehalem, "Lynnfield" */ + case 46: /* 45 nm nehalem-ex, "Beckton" */ memcpy(hw_cache_event_ids, nehalem_hw_cache_event_ids, sizeof(hw_cache_event_ids)); -- cgit v0.10.2 From b0cc58a25d04160d39a80e436847eaa2fbc5aa09 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 6 Apr 2010 19:31:26 +0300 Subject: ALSA: mixart: range checking proc file The original code doesn't take into consideration that the value of MIXART_BA0_SIZE - pos can be less than zero which would lead to a large unsigned value for "count". Also I moved the check that read size is a multiple of 4 bytes below the code that adjusts "count". Signed-off-by: Dan Carpenter Cc: Acked-by: Linus Torvalds Signed-off-by: Takashi Iwai diff --git a/sound/pci/mixart/mixart.c b/sound/pci/mixart/mixart.c index 7e8e7da..ea4256b 100644 --- a/sound/pci/mixart/mixart.c +++ b/sound/pci/mixart/mixart.c @@ -1161,13 +1161,15 @@ static long snd_mixart_BA0_read(struct snd_info_entry *entry, void *file_private unsigned long count, unsigned long pos) { struct mixart_mgr *mgr = entry->private_data; + unsigned long maxsize; - count = count & ~3; /* make sure the read size is a multiple of 4 bytes */ - if(count <= 0) + if (pos >= MIXART_BA0_SIZE) return 0; - if(pos + count > MIXART_BA0_SIZE) - count = (long)(MIXART_BA0_SIZE - pos); - if(copy_to_user_fromio(buf, MIXART_MEM( mgr, pos ), count)) + maxsize = MIXART_BA0_SIZE - pos; + if (count > maxsize) + count = maxsize; + count = count & ~3; /* make sure the read size is a multiple of 4 bytes */ + if (copy_to_user_fromio(buf, MIXART_MEM(mgr, pos), count)) return -EFAULT; return count; } @@ -1180,13 +1182,15 @@ static long snd_mixart_BA1_read(struct snd_info_entry *entry, void *file_private unsigned long count, unsigned long pos) { struct mixart_mgr *mgr = entry->private_data; + unsigned long maxsize; - count = count & ~3; /* make sure the read size is a multiple of 4 bytes */ - if(count <= 0) + if (pos > MIXART_BA1_SIZE) return 0; - if(pos + count > MIXART_BA1_SIZE) - count = (long)(MIXART_BA1_SIZE - pos); - if(copy_to_user_fromio(buf, MIXART_REG( mgr, pos ), count)) + maxsize = MIXART_BA1_SIZE - pos; + if (count > maxsize) + count = maxsize; + count = count & ~3; /* make sure the read size is a multiple of 4 bytes */ + if (copy_to_user_fromio(buf, MIXART_REG(mgr, pos), count)) return -EFAULT; return count; } -- cgit v0.10.2 From f05337c6ac48d19d354e0640a8eb8fc884f82bcc Mon Sep 17 00:00:00 2001 From: Pavel Shilovsky Date: Mon, 5 Apr 2010 09:59:14 +0400 Subject: not overwriting file_lock structure after GET_LK If we have preventing lock, cifs should overwrite file_lock structure with info about preventing lock. If we haven't preventing lock, cifs should leave it unchanged except for the lock type (change it to F_UNLCK). Signed-off-by: Pavel Shilovsky Reviewed-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index f213b8a..184a399 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -1810,8 +1810,21 @@ CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, } parm_data = (struct cifs_posix_lock *) ((char *)&pSMBr->hdr.Protocol + data_offset); - if (parm_data->lock_type == cpu_to_le16(CIFS_UNLCK)) + if (parm_data->lock_type == __constant_cpu_to_le16(CIFS_UNLCK)) pLockData->fl_type = F_UNLCK; + else { + if (parm_data->lock_type == + __constant_cpu_to_le16(CIFS_RDLCK)) + pLockData->fl_type = F_RDLCK; + else if (parm_data->lock_type == + __constant_cpu_to_le16(CIFS_WRLCK)) + pLockData->fl_type = F_WRLCK; + + pLockData->fl_start = parm_data->start; + pLockData->fl_end = parm_data->start + + parm_data->length - 1; + pLockData->fl_pid = parm_data->pid; + } } plk_err_exit: diff --git a/fs/cifs/file.c b/fs/cifs/file.c index ca2ba7a..d9e8650 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -838,8 +838,32 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) } else { /* if rc == ERR_SHARING_VIOLATION ? */ - rc = 0; /* do not change lock type to unlock - since range in use */ + rc = 0; + + if (lockType & LOCKING_ANDX_SHARED_LOCK) { + pfLock->fl_type = F_WRLCK; + } else { + rc = CIFSSMBLock(xid, tcon, netfid, length, + pfLock->fl_start, 0, 1, + lockType | LOCKING_ANDX_SHARED_LOCK, + 0 /* wait flag */); + if (rc == 0) { + rc = CIFSSMBLock(xid, tcon, netfid, + length, pfLock->fl_start, 1, 0, + lockType | + LOCKING_ANDX_SHARED_LOCK, + 0 /* wait flag */); + pfLock->fl_type = F_RDLCK; + if (rc != 0) + cERROR(1, ("Error unlocking " + "previously locked range %d " + "during test of lock", rc)); + rc = 0; + } else { + pfLock->fl_type = F_WRLCK; + rc = 0; + } + } } FreeXid(xid); -- cgit v0.10.2 From 3bbb9ec946428b96657126768f65487a48dd090c Mon Sep 17 00:00:00 2001 From: Arjan van de Ven Date: Thu, 11 Mar 2010 14:04:36 -0800 Subject: timers: Introduce the concept of timer slack for legacy timers While HR timers have had the concept of timer slack for quite some time now, the legacy timers lacked this concept, and had to make do with round_jiffies() and friends. Timer slack is important for power management; grouping timers reduces the number of wakeups which in turn reduces power consumption. This patch introduces timer slack to the legacy timers using the following pieces: * A slack field in the timer struct * An api (set_timer_slack) that callers can use to set explicit timer slack * A default slack of 0.4% of the requested delay for callers that do not set any explicit slack * Rounding code that is part of mod_timer() that tries to group timers around jiffies values every 'power of two' (so quick timers will group around every 2, but longer timers will group around every 4, 8, 16, 32 etc) Signed-off-by: Arjan van de Ven Cc: johnstul@us.ibm.com Cc: Ingo Molnar Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner diff --git a/include/linux/timer.h b/include/linux/timer.h index a2d1eb6..ea965b8 100644 --- a/include/linux/timer.h +++ b/include/linux/timer.h @@ -10,13 +10,19 @@ struct tvec_base; struct timer_list { + /* + * All fields that change during normal runtime grouped to the + * same cacheline + */ struct list_head entry; unsigned long expires; + struct tvec_base *base; void (*function)(unsigned long); unsigned long data; - struct tvec_base *base; + int slack; + #ifdef CONFIG_TIMER_STATS void *start_site; char start_comm[16]; @@ -165,6 +171,8 @@ extern int mod_timer(struct timer_list *timer, unsigned long expires); extern int mod_timer_pending(struct timer_list *timer, unsigned long expires); extern int mod_timer_pinned(struct timer_list *timer, unsigned long expires); +extern void set_timer_slack(struct timer_list *time, int slack_hz); + #define TIMER_NOT_PINNED 0 #define TIMER_PINNED 1 /* diff --git a/kernel/timer.c b/kernel/timer.c index 7e12e7b..49773f3 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -318,6 +318,24 @@ unsigned long round_jiffies_up_relative(unsigned long j) } EXPORT_SYMBOL_GPL(round_jiffies_up_relative); +/** + * set_timer_slack - set the allowed slack for a timer + * @slack_hz: the amount of time (in jiffies) allowed for rounding + * + * Set the amount of time, in jiffies, that a certain timer has + * in terms of slack. By setting this value, the timer subsystem + * will schedule the actual timer somewhere between + * the time mod_timer() asks for, and that time plus the slack. + * + * By setting the slack to -1, a percentage of the delay is used + * instead. + */ +void set_timer_slack(struct timer_list *timer, int slack_hz) +{ + timer->slack = slack_hz; +} +EXPORT_SYMBOL_GPL(set_timer_slack); + static inline void set_running_timer(struct tvec_base *base, struct timer_list *timer) @@ -549,6 +567,7 @@ static void __init_timer(struct timer_list *timer, { timer->entry.next = NULL; timer->base = __raw_get_cpu_var(tvec_bases); + timer->slack = -1; #ifdef CONFIG_TIMER_STATS timer->start_site = NULL; timer->start_pid = -1; @@ -714,6 +733,41 @@ int mod_timer_pending(struct timer_list *timer, unsigned long expires) } EXPORT_SYMBOL(mod_timer_pending); +/* + * Decide where to put the timer while taking the slack into account + * + * Algorithm: + * 1) calculate the maximum (absolute) time + * 2) calculate the highest bit where the expires and new max are different + * 3) use this bit to make a mask + * 4) use the bitmask to round down the maximum time, so that all last + * bits are zeros + */ +static inline +unsigned long apply_slack(struct timer_list *timer, unsigned long expires) +{ + unsigned long expires_limit, mask; + int bit; + + expires_limit = expires + timer->slack; + + if (timer->slack < 0) /* auto slack: use 0.4% */ + expires_limit = expires + (expires - jiffies)/256; + + mask = expires ^ expires_limit; + + if (mask == 0) + return expires; + + bit = find_last_bit(&mask, BITS_PER_LONG); + + mask = (1 << bit) - 1; + + expires_limit = expires_limit & ~(mask); + + return expires_limit; +} + /** * mod_timer - modify a timer's timeout * @timer: the timer to be modified @@ -744,6 +798,8 @@ int mod_timer(struct timer_list *timer, unsigned long expires) if (timer_pending(timer) && timer->expires == expires) return 1; + expires = apply_slack(timer, expires); + return __mod_timer(timer, expires, false, TIMER_NOT_PINNED); } EXPORT_SYMBOL(mod_timer); -- cgit v0.10.2 From 351b3f7a21e413a9b14d0393171497d2373bd702 Mon Sep 17 00:00:00 2001 From: Carsten Emde Date: Fri, 2 Apr 2010 22:40:19 +0200 Subject: hrtimers: Provide schedule_hrtimeout for CLOCK_REALTIME The current version of schedule_hrtimeout() always uses the monotonic clock. Some system calls such as mq_timedsend() and mq_timedreceive(), however, require the use of the wall clock due to the definition of the system call. This patch provides the infrastructure to use schedule_hrtimeout() with a CLOCK_REALTIME timer. Signed-off-by: Carsten Emde Tested-by: Pradyumna Sampath Cc: Andrew Morton Cc: Arjan van de Veen LKML-Reference: <20100402204331.167439615@osadl.org> Signed-off-by: Thomas Gleixner diff --git a/include/linux/hrtimer.h b/include/linux/hrtimer.h index 5d86fb2..fd0c1b8 100644 --- a/include/linux/hrtimer.h +++ b/include/linux/hrtimer.h @@ -422,6 +422,8 @@ extern void hrtimer_init_sleeper(struct hrtimer_sleeper *sl, extern int schedule_hrtimeout_range(ktime_t *expires, unsigned long delta, const enum hrtimer_mode mode); +extern int schedule_hrtimeout_range_clock(ktime_t *expires, + unsigned long delta, const enum hrtimer_mode mode, int clock); extern int schedule_hrtimeout(ktime_t *expires, const enum hrtimer_mode mode); /* Soft interrupt function to run the hrtimer queues: */ diff --git a/kernel/hrtimer.c b/kernel/hrtimer.c index 0086628..b9b134b 100644 --- a/kernel/hrtimer.c +++ b/kernel/hrtimer.c @@ -1749,35 +1749,15 @@ void __init hrtimers_init(void) } /** - * schedule_hrtimeout_range - sleep until timeout + * schedule_hrtimeout_range_clock - sleep until timeout * @expires: timeout value (ktime_t) * @delta: slack in expires timeout (ktime_t) * @mode: timer mode, HRTIMER_MODE_ABS or HRTIMER_MODE_REL - * - * Make the current task sleep until the given expiry time has - * elapsed. The routine will return immediately unless - * the current task state has been set (see set_current_state()). - * - * The @delta argument gives the kernel the freedom to schedule the - * actual wakeup to a time that is both power and performance friendly. - * The kernel give the normal best effort behavior for "@expires+@delta", - * but may decide to fire the timer earlier, but no earlier than @expires. - * - * You can set the task state as follows - - * - * %TASK_UNINTERRUPTIBLE - at least @timeout time is guaranteed to - * pass before the routine returns. - * - * %TASK_INTERRUPTIBLE - the routine may return early if a signal is - * delivered to the current task. - * - * The current task state is guaranteed to be TASK_RUNNING when this - * routine returns. - * - * Returns 0 when the timer has expired otherwise -EINTR + * @clock: timer clock, CLOCK_MONOTONIC or CLOCK_REALTIME */ -int __sched schedule_hrtimeout_range(ktime_t *expires, unsigned long delta, - const enum hrtimer_mode mode) +int __sched +schedule_hrtimeout_range_clock(ktime_t *expires, unsigned long delta, + const enum hrtimer_mode mode, int clock) { struct hrtimer_sleeper t; @@ -1799,7 +1779,7 @@ int __sched schedule_hrtimeout_range(ktime_t *expires, unsigned long delta, return -EINTR; } - hrtimer_init_on_stack(&t.timer, CLOCK_MONOTONIC, mode); + hrtimer_init_on_stack(&t.timer, clock, mode); hrtimer_set_expires_range_ns(&t.timer, *expires, delta); hrtimer_init_sleeper(&t, current); @@ -1818,6 +1798,41 @@ int __sched schedule_hrtimeout_range(ktime_t *expires, unsigned long delta, return !t.task ? 0 : -EINTR; } + +/** + * schedule_hrtimeout_range - sleep until timeout + * @expires: timeout value (ktime_t) + * @delta: slack in expires timeout (ktime_t) + * @mode: timer mode, HRTIMER_MODE_ABS or HRTIMER_MODE_REL + * + * Make the current task sleep until the given expiry time has + * elapsed. The routine will return immediately unless + * the current task state has been set (see set_current_state()). + * + * The @delta argument gives the kernel the freedom to schedule the + * actual wakeup to a time that is both power and performance friendly. + * The kernel give the normal best effort behavior for "@expires+@delta", + * but may decide to fire the timer earlier, but no earlier than @expires. + * + * You can set the task state as follows - + * + * %TASK_UNINTERRUPTIBLE - at least @timeout time is guaranteed to + * pass before the routine returns. + * + * %TASK_INTERRUPTIBLE - the routine may return early if a signal is + * delivered to the current task. + * + * The current task state is guaranteed to be TASK_RUNNING when this + * routine returns. + * + * Returns 0 when the timer has expired otherwise -EINTR + */ +int __sched schedule_hrtimeout_range(ktime_t *expires, unsigned long delta, + const enum hrtimer_mode mode) +{ + return schedule_hrtimeout_range_clock(expires, delta, mode, + CLOCK_MONOTONIC); +} EXPORT_SYMBOL_GPL(schedule_hrtimeout_range); /** -- cgit v0.10.2 From 9ca7d8e6834c40a99622bbe4a88aaf64313ae43c Mon Sep 17 00:00:00 2001 From: Carsten Emde Date: Fri, 2 Apr 2010 22:40:20 +0200 Subject: mqueue: Convert message queue timeout to use hrtimers The message queue functions mq_timedsend() and mq_timedreceive() have not yet been converted to use the hrtimer interface. This patch replaces the call to schedule_timeout() by a call to schedule_hrtimeout() and transforms the expiration time from timespec to ktime as required. [ tglx: Fixed whitespace wreckage ] Signed-off-by: Carsten Emde Tested-by: Pradyumna Sampath Cc: Arjan van de Veen Cc: Andrew Morton LKML-Reference: <20100402204331.715783034@osadl.org> Signed-off-by: Thomas Gleixner diff --git a/ipc/mqueue.c b/ipc/mqueue.c index e4e3f04..a9d8b0c 100644 --- a/ipc/mqueue.c +++ b/ipc/mqueue.c @@ -428,7 +428,7 @@ static void wq_add(struct mqueue_inode_info *info, int sr, * sr: SEND or RECV */ static int wq_sleep(struct mqueue_inode_info *info, int sr, - long timeout, struct ext_wait_queue *ewp) + ktime_t *timeout, struct ext_wait_queue *ewp) { int retval; signed long time; @@ -439,7 +439,8 @@ static int wq_sleep(struct mqueue_inode_info *info, int sr, set_current_state(TASK_INTERRUPTIBLE); spin_unlock(&info->lock); - time = schedule_timeout(timeout); + time = schedule_hrtimeout_range_clock(timeout, + HRTIMER_MODE_ABS, 0, CLOCK_REALTIME); while (ewp->state == STATE_PENDING) cpu_relax(); @@ -551,31 +552,16 @@ static void __do_notify(struct mqueue_inode_info *info) wake_up(&info->wait_q); } -static long prepare_timeout(struct timespec *p) +static int prepare_timeout(const struct timespec __user *u_abs_timeout, + ktime_t *expires, struct timespec *ts) { - struct timespec nowts; - long timeout; - - if (p) { - if (unlikely(p->tv_nsec < 0 || p->tv_sec < 0 - || p->tv_nsec >= NSEC_PER_SEC)) - return -EINVAL; - nowts = CURRENT_TIME; - /* first subtract as jiffies can't be too big */ - p->tv_sec -= nowts.tv_sec; - if (p->tv_nsec < nowts.tv_nsec) { - p->tv_nsec += NSEC_PER_SEC; - p->tv_sec--; - } - p->tv_nsec -= nowts.tv_nsec; - if (p->tv_sec < 0) - return 0; - - timeout = timespec_to_jiffies(p) + 1; - } else - return MAX_SCHEDULE_TIMEOUT; + if (copy_from_user(ts, u_abs_timeout, sizeof(struct timespec))) + return -EFAULT; + if (!timespec_valid(ts)) + return -EINVAL; - return timeout; + *expires = timespec_to_ktime(*ts); + return 0; } static void remove_notification(struct mqueue_inode_info *info) @@ -861,22 +847,21 @@ SYSCALL_DEFINE5(mq_timedsend, mqd_t, mqdes, const char __user *, u_msg_ptr, struct ext_wait_queue *receiver; struct msg_msg *msg_ptr; struct mqueue_inode_info *info; - struct timespec ts, *p = NULL; - long timeout; + ktime_t expires, *timeout = NULL; + struct timespec ts; int ret; if (u_abs_timeout) { - if (copy_from_user(&ts, u_abs_timeout, - sizeof(struct timespec))) - return -EFAULT; - p = &ts; + int res = prepare_timeout(u_abs_timeout, &expires, &ts); + if (res) + return res; + timeout = &expires; } if (unlikely(msg_prio >= (unsigned long) MQ_PRIO_MAX)) return -EINVAL; - audit_mq_sendrecv(mqdes, msg_len, msg_prio, p); - timeout = prepare_timeout(p); + audit_mq_sendrecv(mqdes, msg_len, msg_prio, timeout ? &ts : NULL); filp = fget(mqdes); if (unlikely(!filp)) { @@ -918,9 +903,6 @@ SYSCALL_DEFINE5(mq_timedsend, mqd_t, mqdes, const char __user *, u_msg_ptr, if (filp->f_flags & O_NONBLOCK) { spin_unlock(&info->lock); ret = -EAGAIN; - } else if (unlikely(timeout < 0)) { - spin_unlock(&info->lock); - ret = timeout; } else { wait.task = current; wait.msg = (void *) msg_ptr; @@ -953,24 +935,23 @@ SYSCALL_DEFINE5(mq_timedreceive, mqd_t, mqdes, char __user *, u_msg_ptr, size_t, msg_len, unsigned int __user *, u_msg_prio, const struct timespec __user *, u_abs_timeout) { - long timeout; ssize_t ret; struct msg_msg *msg_ptr; struct file *filp; struct inode *inode; struct mqueue_inode_info *info; struct ext_wait_queue wait; - struct timespec ts, *p = NULL; + ktime_t expires, *timeout = NULL; + struct timespec ts; if (u_abs_timeout) { - if (copy_from_user(&ts, u_abs_timeout, - sizeof(struct timespec))) - return -EFAULT; - p = &ts; + int res = prepare_timeout(u_abs_timeout, &expires, &ts); + if (res) + return res; + timeout = &expires; } - audit_mq_sendrecv(mqdes, msg_len, 0, p); - timeout = prepare_timeout(p); + audit_mq_sendrecv(mqdes, msg_len, 0, timeout ? &ts : NULL); filp = fget(mqdes); if (unlikely(!filp)) { @@ -1002,11 +983,6 @@ SYSCALL_DEFINE5(mq_timedreceive, mqd_t, mqdes, char __user *, u_msg_ptr, if (filp->f_flags & O_NONBLOCK) { spin_unlock(&info->lock); ret = -EAGAIN; - msg_ptr = NULL; - } else if (unlikely(timeout < 0)) { - spin_unlock(&info->lock); - ret = timeout; - msg_ptr = NULL; } else { wait.task = current; wait.state = STATE_NONE; -- cgit v0.10.2 From 1cb561f83793191cf86a2db3948d28f5f42df9ff Mon Sep 17 00:00:00 2001 From: Javier Cardona Date: Mon, 29 Mar 2010 11:00:20 -0700 Subject: mac80211: Handle mesh action frames in ieee80211_rx_h_action This fixes the problem introduced in commit 8404080568613d93ad7cf0a16dfb68 which broke mesh peer link establishment. changes: v2 Added missing break (Johannes) v3 Broke original patch into two (Johannes) Signed-off-by: Javier Cardona Cc: stable@kernel.org Reviewed-by: Johannes Berg Signed-off-by: John W. Linville diff --git a/net/mac80211/mesh.c b/net/mac80211/mesh.c index 61080c5..7a6bebc 100644 --- a/net/mac80211/mesh.c +++ b/net/mac80211/mesh.c @@ -749,9 +749,6 @@ ieee80211_mesh_rx_mgmt(struct ieee80211_sub_if_data *sdata, struct sk_buff *skb) switch (fc & IEEE80211_FCTL_STYPE) { case IEEE80211_STYPE_ACTION: - if (skb->len < IEEE80211_MIN_ACTION_SIZE) - return RX_DROP_MONITOR; - /* fall through */ case IEEE80211_STYPE_PROBE_RESP: case IEEE80211_STYPE_BEACON: skb_queue_tail(&ifmsh->skb_queue, skb); diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index b5c48de..13fcd2d 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -1973,6 +1973,11 @@ ieee80211_rx_h_action(struct ieee80211_rx_data *rx) goto handled; } break; + case MESH_PLINK_CATEGORY: + case MESH_PATH_SEL_CATEGORY: + if (ieee80211_vif_is_mesh(&sdata->vif)) + return ieee80211_mesh_rx_mgmt(sdata, rx->skb); + break; } /* -- cgit v0.10.2 From 0379185b6c0d1e8252023698cf1091da92a3dc03 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 6 Apr 2010 11:18:42 +0200 Subject: mac80211: annotate station rcu dereferences The new RCU lockdep support warns about these in some contexts -- make it aware of the locks used to protect all this. Different locks are used in different contexts which unfortunately means we can't get perfect checking. Also remove rcu_dereference() from two places that don't actually dereference the pointers. Signed-off-by: Johannes Berg Signed-off-by: John W. Linville diff --git a/net/mac80211/main.c b/net/mac80211/main.c index 06c33b6..b887e48 100644 --- a/net/mac80211/main.c +++ b/net/mac80211/main.c @@ -225,11 +225,11 @@ void ieee80211_bss_info_change_notify(struct ieee80211_sub_if_data *sdata, switch (sdata->vif.type) { case NL80211_IFTYPE_AP: sdata->vif.bss_conf.enable_beacon = - !!rcu_dereference(sdata->u.ap.beacon); + !!sdata->u.ap.beacon; break; case NL80211_IFTYPE_ADHOC: sdata->vif.bss_conf.enable_beacon = - !!rcu_dereference(sdata->u.ibss.presp); + !!sdata->u.ibss.presp; break; case NL80211_IFTYPE_MESH_POINT: sdata->vif.bss_conf.enable_beacon = true; diff --git a/net/mac80211/sta_info.c b/net/mac80211/sta_info.c index 56422d8..fb12cec 100644 --- a/net/mac80211/sta_info.c +++ b/net/mac80211/sta_info.c @@ -93,12 +93,18 @@ struct sta_info *sta_info_get(struct ieee80211_sub_if_data *sdata, struct ieee80211_local *local = sdata->local; struct sta_info *sta; - sta = rcu_dereference(local->sta_hash[STA_HASH(addr)]); + sta = rcu_dereference_check(local->sta_hash[STA_HASH(addr)], + rcu_read_lock_held() || + lockdep_is_held(&local->sta_lock) || + lockdep_is_held(&local->sta_mtx)); while (sta) { if (sta->sdata == sdata && memcmp(sta->sta.addr, addr, ETH_ALEN) == 0) break; - sta = rcu_dereference(sta->hnext); + sta = rcu_dereference_check(sta->hnext, + rcu_read_lock_held() || + lockdep_is_held(&local->sta_lock) || + lockdep_is_held(&local->sta_mtx)); } return sta; } @@ -113,13 +119,19 @@ struct sta_info *sta_info_get_bss(struct ieee80211_sub_if_data *sdata, struct ieee80211_local *local = sdata->local; struct sta_info *sta; - sta = rcu_dereference(local->sta_hash[STA_HASH(addr)]); + sta = rcu_dereference_check(local->sta_hash[STA_HASH(addr)], + rcu_read_lock_held() || + lockdep_is_held(&local->sta_lock) || + lockdep_is_held(&local->sta_mtx)); while (sta) { if ((sta->sdata == sdata || sta->sdata->bss == sdata->bss) && memcmp(sta->sta.addr, addr, ETH_ALEN) == 0) break; - sta = rcu_dereference(sta->hnext); + sta = rcu_dereference_check(sta->hnext, + rcu_read_lock_held() || + lockdep_is_held(&local->sta_lock) || + lockdep_is_held(&local->sta_mtx)); } return sta; } -- cgit v0.10.2 From 1144601118507f8b3b676a9a392584d216d3f2cc Mon Sep 17 00:00:00 2001 From: Felix Fietkau Date: Tue, 6 Apr 2010 12:05:01 -0700 Subject: ath9k: fix double calls to ath_radio_enable With the enable_radio being uninitialized, ath_radio_enable() might be called twice, which can leave some hardware in an undefined state. Signed-off-by: Felix Fietkau Cc: stable@kernel.org Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c index 67ca4e5..115e1ae 100644 --- a/drivers/net/wireless/ath/ath9k/main.c +++ b/drivers/net/wireless/ath/ath9k/main.c @@ -1532,8 +1532,7 @@ static int ath9k_config(struct ieee80211_hw *hw, u32 changed) all_wiphys_idle = ath9k_all_wiphys_idle(sc); ath9k_set_wiphy_idle(aphy, idle); - if (!idle && all_wiphys_idle) - enable_radio = true; + enable_radio = (!idle && all_wiphys_idle); /* * After we unlock here its possible another wiphy -- cgit v0.10.2 From ac7729da880e742613129ee6dea0045328670d2d Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Mon, 5 Apr 2010 01:43:51 +0200 Subject: ACPI / PM: Move ACPI video resume to a PM notifier MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit There is a problem with the ACPI video resume routine that it's executed before the GPU that may be accessed by it. To fix this issue, move the ACPI video resume to a power management notifier, so that's executed after resuming all devices, including the GPU. Fixes https://bugzilla.kernel.org/show_bug.cgi?id=15096, which is a listed regression from 2.6.31. Signed-off-by: Rafael J. Wysocki Tested-by: Rafał Miłecki Acked-by: Matthew Garrett Signed-off-by: Len Brown diff --git a/drivers/acpi/video.c b/drivers/acpi/video.c index 6a01437..416eb03 100644 --- a/drivers/acpi/video.c +++ b/drivers/acpi/video.c @@ -44,6 +44,7 @@ #include #include #include +#include #define PREFIX "ACPI: " @@ -89,7 +90,6 @@ module_param(allow_duplicates, bool, 0644); static int register_count = 0; static int acpi_video_bus_add(struct acpi_device *device); static int acpi_video_bus_remove(struct acpi_device *device, int type); -static int acpi_video_resume(struct acpi_device *device); static void acpi_video_bus_notify(struct acpi_device *device, u32 event); static const struct acpi_device_id video_device_ids[] = { @@ -105,7 +105,6 @@ static struct acpi_driver acpi_video_bus = { .ops = { .add = acpi_video_bus_add, .remove = acpi_video_bus_remove, - .resume = acpi_video_resume, .notify = acpi_video_bus_notify, }, }; @@ -160,6 +159,7 @@ struct acpi_video_bus { struct proc_dir_entry *dir; struct input_dev *input; char phys[32]; /* for input device */ + struct notifier_block pm_nb; }; struct acpi_video_device_flags { @@ -1021,6 +1021,13 @@ static void acpi_video_device_find_cap(struct acpi_video_device *device) if (IS_ERR(device->backlight)) return; + /* + * Save current brightness level in case we have to restore it + * before acpi_video_device_lcd_set_level() is called next time. + */ + device->backlight->props.brightness = + acpi_video_get_brightness(device->backlight); + result = sysfs_create_link(&device->backlight->dev.kobj, &device->dev->dev.kobj, "device"); if (result) @@ -2236,24 +2243,31 @@ static void acpi_video_device_notify(acpi_handle handle, u32 event, void *data) return; } -static int instance; -static int acpi_video_resume(struct acpi_device *device) +static int acpi_video_resume(struct notifier_block *nb, + unsigned long val, void *ign) { struct acpi_video_bus *video; struct acpi_video_device *video_device; int i; - if (!device || !acpi_driver_data(device)) - return -EINVAL; + switch (val) { + case PM_HIBERNATION_PREPARE: + case PM_SUSPEND_PREPARE: + case PM_RESTORE_PREPARE: + return NOTIFY_DONE; + } - video = acpi_driver_data(device); + video = container_of(nb, struct acpi_video_bus, pm_nb); + + dev_info(&video->device->dev, "Restoring backlight state\n"); for (i = 0; i < video->attached_count; i++) { video_device = video->attached_array[i].bind_info; if (video_device && video_device->backlight) acpi_video_set_brightness(video_device->backlight); } - return AE_OK; + + return NOTIFY_OK; } static acpi_status @@ -2277,6 +2291,8 @@ acpi_video_bus_match(acpi_handle handle, u32 level, void *context, return AE_OK; } +static int instance; + static int acpi_video_bus_add(struct acpi_device *device) { struct acpi_video_bus *video; @@ -2370,6 +2386,10 @@ static int acpi_video_bus_add(struct acpi_device *device) video->flags.rom ? "yes" : "no", video->flags.post ? "yes" : "no"); + video->pm_nb.notifier_call = acpi_video_resume; + video->pm_nb.priority = 0; + register_pm_notifier(&video->pm_nb); + return 0; err_free_input_dev: @@ -2396,6 +2416,8 @@ static int acpi_video_bus_remove(struct acpi_device *device, int type) video = acpi_driver_data(device); + unregister_pm_notifier(&video->pm_nb); + acpi_video_bus_stop_devices(video); acpi_video_bus_put_devices(video); acpi_video_bus_remove_fs(device); -- cgit v0.10.2 From 6549981bc54777c37eccf987e227aff47022ab7c Mon Sep 17 00:00:00 2001 From: Stephane Chatty Date: Tue, 6 Apr 2010 22:22:58 +0200 Subject: HID: fix N-trig touch panel with recent firmware Added an init message that avoids device freeze with recent firmware. Signed-off-by: Stephane Chatty Tested-by: Rafi Rubin Signed-off-by: Jiri Kosina diff --git a/drivers/hid/hid-ntrig.c b/drivers/hid/hid-ntrig.c index 9b24fc5..513db0a 100644 --- a/drivers/hid/hid-ntrig.c +++ b/drivers/hid/hid-ntrig.c @@ -1,8 +1,8 @@ /* * HID driver for N-Trig touchscreens * - * Copyright (c) 2008 Rafi Rubin - * Copyright (c) 2009 Stephane Chatty + * Copyright (c) 2008-2010 Rafi Rubin + * Copyright (c) 2009-2010 Stephane Chatty * */ @@ -15,6 +15,8 @@ #include #include +#include +#include "usbhid/usbhid.h" #include #include @@ -286,6 +288,7 @@ static int ntrig_probe(struct hid_device *hdev, const struct hid_device_id *id) struct ntrig_data *nd; struct hid_input *hidinput; struct input_dev *input; + struct hid_report *report; if (id->driver_data) hdev->quirks |= HID_QUIRK_MULTI_INPUT; @@ -349,6 +352,11 @@ static int ntrig_probe(struct hid_device *hdev, const struct hid_device_id *id) } } + report = hdev->report_enum[HID_FEATURE_REPORT].report_id_hash[0x0a]; + if (report) + usbhid_submit_report(hdev, report, USB_DIR_OUT); + + return 0; err_free: kfree(nd); -- cgit v0.10.2 From 18e5b539b451158be7aae6c390a20f0d3e5b9213 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Tue, 6 Apr 2010 19:23:33 +0900 Subject: nodemask: include slab.h from drivers/base/node.c NODEMASK_ALLOC/FREE are mapped to kmalloc/free if NODES_SHIFT > 8. Among its several users, drivers/base/node.c wasn't including slab.h leading to build failure if NODES_SHIFT > 8. Include slab.h from drivers/base/node.c. This isn't an ideal solution but including slab.h directly from nodemask.h is not an option because nodemask.h gets included everywhere. For now, make it work by including slab.h from its users. Signed-off-by: Tejun Heo Reported-by: Ingo Molnar diff --git a/drivers/base/node.c b/drivers/base/node.c index 985abd7..057979a 100644 --- a/drivers/base/node.c +++ b/drivers/base/node.c @@ -15,7 +15,7 @@ #include #include #include -#include +#include static struct sysdev_class_attribute *node_state_attrs[]; -- cgit v0.10.2 From 1527bc8b928dd1399c3d3467dd47d9ede210978a Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 1 Feb 2010 15:03:07 +0100 Subject: bitops: Optimize hweight() by making use of compile-time evaluation Rename the extisting runtime hweight() implementations to __arch_hweight(), rename the compile-time versions to __const_hweight() and then have hweight() pick between them. Suggested-by: H. Peter Anvin Signed-off-by: Peter Zijlstra LKML-Reference: <20100318111929.GB11152@aftab> Acked-by: H. Peter Anvin LKML-Reference: <1265028224.24455.154.camel@laptop> Signed-off-by: H. Peter Anvin diff --git a/arch/alpha/include/asm/bitops.h b/arch/alpha/include/asm/bitops.h index 15f3ae2..296da1d 100644 --- a/arch/alpha/include/asm/bitops.h +++ b/arch/alpha/include/asm/bitops.h @@ -405,29 +405,31 @@ static inline int fls(int x) #if defined(CONFIG_ALPHA_EV6) && defined(CONFIG_ALPHA_EV67) /* Whee. EV67 can calculate it directly. */ -static inline unsigned long hweight64(unsigned long w) +static inline unsigned long __arch_hweight64(unsigned long w) { return __kernel_ctpop(w); } -static inline unsigned int hweight32(unsigned int w) +static inline unsigned int __arch_weight32(unsigned int w) { - return hweight64(w); + return __arch_hweight64(w); } -static inline unsigned int hweight16(unsigned int w) +static inline unsigned int __arch_hweight16(unsigned int w) { - return hweight64(w & 0xffff); + return __arch_hweight64(w & 0xffff); } -static inline unsigned int hweight8(unsigned int w) +static inline unsigned int __arch_hweight8(unsigned int w) { - return hweight64(w & 0xff); + return __arch_hweight64(w & 0xff); } #else -#include +#include #endif +#include + #endif /* __KERNEL__ */ #include diff --git a/arch/ia64/include/asm/bitops.h b/arch/ia64/include/asm/bitops.h index 6ebc229..9da3df6 100644 --- a/arch/ia64/include/asm/bitops.h +++ b/arch/ia64/include/asm/bitops.h @@ -437,17 +437,18 @@ __fls (unsigned long x) * hweightN: returns the hamming weight (i.e. the number * of bits set) of a N-bit word */ -static __inline__ unsigned long -hweight64 (unsigned long x) +static __inline__ unsigned long __arch_hweight64(unsigned long x) { unsigned long result; result = ia64_popcnt(x); return result; } -#define hweight32(x) (unsigned int) hweight64((x) & 0xfffffffful) -#define hweight16(x) (unsigned int) hweight64((x) & 0xfffful) -#define hweight8(x) (unsigned int) hweight64((x) & 0xfful) +#define __arch_hweight32(x) ((unsigned int) __arch_hweight64((x) & 0xfffffffful)) +#define __arch_hweight16(x) ((unsigned int) __arch_hweight64((x) & 0xfffful)) +#define __arch_hweight8(x) ((unsigned int) __arch_hweight64((x) & 0xfful)) + +#include #endif /* __KERNEL__ */ diff --git a/arch/sparc/include/asm/bitops_64.h b/arch/sparc/include/asm/bitops_64.h index e72ac9c..766121a 100644 --- a/arch/sparc/include/asm/bitops_64.h +++ b/arch/sparc/include/asm/bitops_64.h @@ -44,7 +44,7 @@ extern void change_bit(unsigned long nr, volatile unsigned long *addr); #ifdef ULTRA_HAS_POPULATION_COUNT -static inline unsigned int hweight64(unsigned long w) +static inline unsigned int __arch_hweight64(unsigned long w) { unsigned int res; @@ -52,7 +52,7 @@ static inline unsigned int hweight64(unsigned long w) return res; } -static inline unsigned int hweight32(unsigned int w) +static inline unsigned int __arch_hweight32(unsigned int w) { unsigned int res; @@ -60,7 +60,7 @@ static inline unsigned int hweight32(unsigned int w) return res; } -static inline unsigned int hweight16(unsigned int w) +static inline unsigned int __arch_hweight16(unsigned int w) { unsigned int res; @@ -68,7 +68,7 @@ static inline unsigned int hweight16(unsigned int w) return res; } -static inline unsigned int hweight8(unsigned int w) +static inline unsigned int __arch_hweight8(unsigned int w) { unsigned int res; @@ -78,9 +78,10 @@ static inline unsigned int hweight8(unsigned int w) #else -#include +#include #endif +#include #include #endif /* __KERNEL__ */ diff --git a/include/asm-generic/bitops/arch_hweight.h b/include/asm-generic/bitops/arch_hweight.h new file mode 100644 index 0000000..3a7be84 --- /dev/null +++ b/include/asm-generic/bitops/arch_hweight.h @@ -0,0 +1,11 @@ +#ifndef _ASM_GENERIC_BITOPS_ARCH_HWEIGHT_H_ +#define _ASM_GENERIC_BITOPS_ARCH_HWEIGHT_H_ + +#include + +extern unsigned int __arch_hweight32(unsigned int w); +extern unsigned int __arch_hweight16(unsigned int w); +extern unsigned int __arch_hweight8(unsigned int w); +extern unsigned long __arch_hweight64(__u64 w); + +#endif /* _ASM_GENERIC_BITOPS_HWEIGHT_H_ */ diff --git a/include/asm-generic/bitops/const_hweight.h b/include/asm-generic/bitops/const_hweight.h new file mode 100644 index 0000000..fa2a50b --- /dev/null +++ b/include/asm-generic/bitops/const_hweight.h @@ -0,0 +1,42 @@ +#ifndef _ASM_GENERIC_BITOPS_CONST_HWEIGHT_H_ +#define _ASM_GENERIC_BITOPS_CONST_HWEIGHT_H_ + +/* + * Compile time versions of __arch_hweightN() + */ +#define __const_hweight8(w) \ + ( (!!((w) & (1ULL << 0))) + \ + (!!((w) & (1ULL << 1))) + \ + (!!((w) & (1ULL << 2))) + \ + (!!((w) & (1ULL << 3))) + \ + (!!((w) & (1ULL << 4))) + \ + (!!((w) & (1ULL << 5))) + \ + (!!((w) & (1ULL << 6))) + \ + (!!((w) & (1ULL << 7))) ) + +#define __const_hweight16(w) (__const_hweight8(w) + __const_hweight8((w) >> 8 )) +#define __const_hweight32(w) (__const_hweight16(w) + __const_hweight16((w) >> 16)) +#define __const_hweight64(w) (__const_hweight32(w) + __const_hweight32((w) >> 32)) + +/* + * Generic interface. + */ +#define hweight8(w) (__builtin_constant_p(w) ? __const_hweight8(w) : __arch_hweight8(w)) +#define hweight16(w) (__builtin_constant_p(w) ? __const_hweight16(w) : __arch_hweight16(w)) +#define hweight32(w) (__builtin_constant_p(w) ? __const_hweight32(w) : __arch_hweight32(w)) +#define hweight64(w) (__builtin_constant_p(w) ? __const_hweight64(w) : __arch_hweight64(w)) + +/* + * Interface for known constant arguments + */ +#define HWEIGHT8(w) (BUILD_BUG_ON_ZERO(!__builtin_constant_p(w)) + __const_hweight8(w)) +#define HWEIGHT16(w) (BUILD_BUG_ON_ZERO(!__builtin_constant_p(w)) + __const_hweight16(w)) +#define HWEIGHT32(w) (BUILD_BUG_ON_ZERO(!__builtin_constant_p(w)) + __const_hweight32(w)) +#define HWEIGHT64(w) (BUILD_BUG_ON_ZERO(!__builtin_constant_p(w)) + __const_hweight64(w)) + +/* + * Type invariant interface to the compile time constant hweight functions. + */ +#define HWEIGHT(w) HWEIGHT64((u64)w) + +#endif /* _ASM_GENERIC_BITOPS_CONST_HWEIGHT_H_ */ diff --git a/include/asm-generic/bitops/hweight.h b/include/asm-generic/bitops/hweight.h index fbbc383..a94d651 100644 --- a/include/asm-generic/bitops/hweight.h +++ b/include/asm-generic/bitops/hweight.h @@ -1,11 +1,7 @@ #ifndef _ASM_GENERIC_BITOPS_HWEIGHT_H_ #define _ASM_GENERIC_BITOPS_HWEIGHT_H_ -#include - -extern unsigned int hweight32(unsigned int w); -extern unsigned int hweight16(unsigned int w); -extern unsigned int hweight8(unsigned int w); -extern unsigned long hweight64(__u64 w); +#include +#include #endif /* _ASM_GENERIC_BITOPS_HWEIGHT_H_ */ diff --git a/include/linux/bitops.h b/include/linux/bitops.h index b793898..c55d5bc 100644 --- a/include/linux/bitops.h +++ b/include/linux/bitops.h @@ -47,31 +47,6 @@ static inline unsigned long hweight_long(unsigned long w) return sizeof(w) == 4 ? hweight32(w) : hweight64(w); } -/* - * Clearly slow versions of the hweightN() functions, their benefit is - * of course compile time evaluation of constant arguments. - */ -#define HWEIGHT8(w) \ - ( BUILD_BUG_ON_ZERO(!__builtin_constant_p(w)) + \ - (!!((w) & (1ULL << 0))) + \ - (!!((w) & (1ULL << 1))) + \ - (!!((w) & (1ULL << 2))) + \ - (!!((w) & (1ULL << 3))) + \ - (!!((w) & (1ULL << 4))) + \ - (!!((w) & (1ULL << 5))) + \ - (!!((w) & (1ULL << 6))) + \ - (!!((w) & (1ULL << 7))) ) - -#define HWEIGHT16(w) (HWEIGHT8(w) + HWEIGHT8((w) >> 8)) -#define HWEIGHT32(w) (HWEIGHT16(w) + HWEIGHT16((w) >> 16)) -#define HWEIGHT64(w) (HWEIGHT32(w) + HWEIGHT32((w) >> 32)) - -/* - * Type invariant version that simply casts things to the - * largest type. - */ -#define HWEIGHT(w) HWEIGHT64((u64)(w)) - /** * rol32 - rotate a 32-bit value left * @word: value to rotate diff --git a/lib/hweight.c b/lib/hweight.c index 63ee4eb..a6927e7 100644 --- a/lib/hweight.c +++ b/lib/hweight.c @@ -9,7 +9,7 @@ * The Hamming Weight of a number is the total number of bits set in it. */ -unsigned int hweight32(unsigned int w) +unsigned int __arch_hweight32(unsigned int w) { #ifdef ARCH_HAS_FAST_MULTIPLIER w -= (w >> 1) & 0x55555555; @@ -24,29 +24,30 @@ unsigned int hweight32(unsigned int w) return (res + (res >> 16)) & 0x000000FF; #endif } -EXPORT_SYMBOL(hweight32); +EXPORT_SYMBOL(__arch_hweight32); -unsigned int hweight16(unsigned int w) +unsigned int __arch_hweight16(unsigned int w) { unsigned int res = w - ((w >> 1) & 0x5555); res = (res & 0x3333) + ((res >> 2) & 0x3333); res = (res + (res >> 4)) & 0x0F0F; return (res + (res >> 8)) & 0x00FF; } -EXPORT_SYMBOL(hweight16); +EXPORT_SYMBOL(__arch_hweight16); -unsigned int hweight8(unsigned int w) +unsigned int __arch_hweight8(unsigned int w) { unsigned int res = w - ((w >> 1) & 0x55); res = (res & 0x33) + ((res >> 2) & 0x33); return (res + (res >> 4)) & 0x0F; } -EXPORT_SYMBOL(hweight8); +EXPORT_SYMBOL(__arch_hweight8); -unsigned long hweight64(__u64 w) +unsigned long __arch_hweight64(__u64 w) { #if BITS_PER_LONG == 32 - return hweight32((unsigned int)(w >> 32)) + hweight32((unsigned int)w); + return __arch_hweight32((unsigned int)(w >> 32)) + + __arch_hweight32((unsigned int)w); #elif BITS_PER_LONG == 64 #ifdef ARCH_HAS_FAST_MULTIPLIER w -= (w >> 1) & 0x5555555555555555ul; @@ -63,4 +64,4 @@ unsigned long hweight64(__u64 w) #endif #endif } -EXPORT_SYMBOL(hweight64); +EXPORT_SYMBOL(__arch_hweight64); -- cgit v0.10.2 From d61931d89be506372d01a90d1755f6d0a9fafe2d Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Fri, 5 Mar 2010 17:34:46 +0100 Subject: x86: Add optimized popcnt variants Add support for the hardware version of the Hamming weight function, popcnt, present in CPUs which advertize it under CPUID, Function 0x0000_0001_ECX[23]. On CPUs which don't support it, we fallback to the default lib/hweight.c sw versions. A synthetic benchmark comparing popcnt with __sw_hweight64 showed almost a 3x speedup on a F10h machine. Signed-off-by: Borislav Petkov LKML-Reference: <20100318112015.GC11152@aftab> Signed-off-by: H. Peter Anvin diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 0eacb1f..89d8c54 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -238,6 +238,11 @@ config X86_32_LAZY_GS def_bool y depends on X86_32 && !CC_STACKPROTECTOR +config ARCH_HWEIGHT_CFLAGS + string + default "-fcall-saved-ecx -fcall-saved-edx" if X86_32 + default "-fcall-saved-rdi -fcall-saved-rsi -fcall-saved-rdx -fcall-saved-rcx -fcall-saved-r8 -fcall-saved-r9 -fcall-saved-r10 -fcall-saved-r11" if X86_64 + config KTIME_SCALAR def_bool X86_32 source "init/Kconfig" diff --git a/arch/x86/include/asm/alternative.h b/arch/x86/include/asm/alternative.h index b09ec55..67dae51 100644 --- a/arch/x86/include/asm/alternative.h +++ b/arch/x86/include/asm/alternative.h @@ -39,9 +39,6 @@ #define LOCK_PREFIX "" #endif -/* This must be included *after* the definition of LOCK_PREFIX */ -#include - struct alt_instr { u8 *instr; /* original instruction */ u8 *replacement; @@ -96,6 +93,12 @@ static inline int alternatives_text_reserved(void *start, void *end) ".previous" /* + * This must be included *after* the definition of ALTERNATIVE due to + * + */ +#include + +/* * Alternative instructions for different CPU types or capabilities. * * This allows to use optimized instructions even on generic binary diff --git a/arch/x86/include/asm/arch_hweight.h b/arch/x86/include/asm/arch_hweight.h new file mode 100644 index 0000000..d1fc3c2 --- /dev/null +++ b/arch/x86/include/asm/arch_hweight.h @@ -0,0 +1,59 @@ +#ifndef _ASM_X86_HWEIGHT_H +#define _ASM_X86_HWEIGHT_H + +#ifdef CONFIG_64BIT +/* popcnt %rdi, %rax */ +#define POPCNT ".byte 0xf3,0x48,0x0f,0xb8,0xc7" +#define REG_IN "D" +#define REG_OUT "a" +#else +/* popcnt %eax, %eax */ +#define POPCNT ".byte 0xf3,0x0f,0xb8,0xc0" +#define REG_IN "a" +#define REG_OUT "a" +#endif + +/* + * __sw_hweightXX are called from within the alternatives below + * and callee-clobbered registers need to be taken care of. See + * ARCH_HWEIGHT_CFLAGS in for the respective + * compiler switches. + */ +static inline unsigned int __arch_hweight32(unsigned int w) +{ + unsigned int res = 0; + + asm (ALTERNATIVE("call __sw_hweight32", POPCNT, X86_FEATURE_POPCNT) + : "="REG_OUT (res) + : REG_IN (w)); + + return res; +} + +static inline unsigned int __arch_hweight16(unsigned int w) +{ + return __arch_hweight32(w & 0xffff); +} + +static inline unsigned int __arch_hweight8(unsigned int w) +{ + return __arch_hweight32(w & 0xff); +} + +static inline unsigned long __arch_hweight64(__u64 w) +{ + unsigned long res = 0; + +#ifdef CONFIG_X86_32 + return __arch_hweight32((u32)w) + + __arch_hweight32((u32)(w >> 32)); +#else + asm (ALTERNATIVE("call __sw_hweight64", POPCNT, X86_FEATURE_POPCNT) + : "="REG_OUT (res) + : REG_IN (w)); +#endif /* CONFIG_X86_32 */ + + return res; +} + +#endif diff --git a/arch/x86/include/asm/bitops.h b/arch/x86/include/asm/bitops.h index 02b47a6..545776e 100644 --- a/arch/x86/include/asm/bitops.h +++ b/arch/x86/include/asm/bitops.h @@ -444,7 +444,9 @@ static inline int fls(int x) #define ARCH_HAS_FAST_MULTIPLIER 1 -#include +#include + +#include #endif /* __KERNEL__ */ diff --git a/include/asm-generic/bitops/arch_hweight.h b/include/asm-generic/bitops/arch_hweight.h index 3a7be84..9a81c1e 100644 --- a/include/asm-generic/bitops/arch_hweight.h +++ b/include/asm-generic/bitops/arch_hweight.h @@ -3,9 +3,23 @@ #include -extern unsigned int __arch_hweight32(unsigned int w); -extern unsigned int __arch_hweight16(unsigned int w); -extern unsigned int __arch_hweight8(unsigned int w); -extern unsigned long __arch_hweight64(__u64 w); +inline unsigned int __arch_hweight32(unsigned int w) +{ + return __sw_hweight32(w); +} +inline unsigned int __arch_hweight16(unsigned int w) +{ + return __sw_hweight16(w); +} + +inline unsigned int __arch_hweight8(unsigned int w) +{ + return __sw_hweight8(w); +} + +inline unsigned long __arch_hweight64(__u64 w) +{ + return __sw_hweight64(w); +} #endif /* _ASM_GENERIC_BITOPS_HWEIGHT_H_ */ diff --git a/lib/Makefile b/lib/Makefile index 2e152ae..abe63a8 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -39,7 +39,10 @@ lib-$(CONFIG_RWSEM_XCHGADD_ALGORITHM) += rwsem.o lib-$(CONFIG_GENERIC_FIND_FIRST_BIT) += find_next_bit.o lib-$(CONFIG_GENERIC_FIND_NEXT_BIT) += find_next_bit.o obj-$(CONFIG_GENERIC_FIND_LAST_BIT) += find_last_bit.o + +CFLAGS_hweight.o = $(subst $(quote),,$(CONFIG_ARCH_HWEIGHT_CFLAGS)) obj-$(CONFIG_GENERIC_HWEIGHT) += hweight.o + obj-$(CONFIG_LOCK_KERNEL) += kernel_lock.o obj-$(CONFIG_BTREE) += btree.o obj-$(CONFIG_DEBUG_PREEMPT) += smp_processor_id.o diff --git a/lib/hweight.c b/lib/hweight.c index a6927e7..3c79d50 100644 --- a/lib/hweight.c +++ b/lib/hweight.c @@ -9,7 +9,7 @@ * The Hamming Weight of a number is the total number of bits set in it. */ -unsigned int __arch_hweight32(unsigned int w) +unsigned int __sw_hweight32(unsigned int w) { #ifdef ARCH_HAS_FAST_MULTIPLIER w -= (w >> 1) & 0x55555555; @@ -24,30 +24,30 @@ unsigned int __arch_hweight32(unsigned int w) return (res + (res >> 16)) & 0x000000FF; #endif } -EXPORT_SYMBOL(__arch_hweight32); +EXPORT_SYMBOL(__sw_hweight32); -unsigned int __arch_hweight16(unsigned int w) +unsigned int __sw_hweight16(unsigned int w) { unsigned int res = w - ((w >> 1) & 0x5555); res = (res & 0x3333) + ((res >> 2) & 0x3333); res = (res + (res >> 4)) & 0x0F0F; return (res + (res >> 8)) & 0x00FF; } -EXPORT_SYMBOL(__arch_hweight16); +EXPORT_SYMBOL(__sw_hweight16); -unsigned int __arch_hweight8(unsigned int w) +unsigned int __sw_hweight8(unsigned int w) { unsigned int res = w - ((w >> 1) & 0x55); res = (res & 0x33) + ((res >> 2) & 0x33); return (res + (res >> 4)) & 0x0F; } -EXPORT_SYMBOL(__arch_hweight8); +EXPORT_SYMBOL(__sw_hweight8); -unsigned long __arch_hweight64(__u64 w) +unsigned long __sw_hweight64(__u64 w) { #if BITS_PER_LONG == 32 - return __arch_hweight32((unsigned int)(w >> 32)) + - __arch_hweight32((unsigned int)w); + return __sw_hweight32((unsigned int)(w >> 32)) + + __sw_hweight32((unsigned int)w); #elif BITS_PER_LONG == 64 #ifdef ARCH_HAS_FAST_MULTIPLIER w -= (w >> 1) & 0x5555555555555555ul; @@ -64,4 +64,4 @@ unsigned long __arch_hweight64(__u64 w) #endif #endif } -EXPORT_SYMBOL(__arch_hweight64); +EXPORT_SYMBOL(__sw_hweight64); diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib index f9bdf26..cbcd654 100644 --- a/scripts/Makefile.lib +++ b/scripts/Makefile.lib @@ -245,3 +245,7 @@ quiet_cmd_lzo = LZO $@ cmd_lzo = (cat $(filter-out FORCE,$^) | \ lzop -9 && $(call size_append, $(filter-out FORCE,$^))) > $@ || \ (rm -f $@ ; false) + +# misc stuff +# --------------------------------------------------------------------------- +quote:=" -- cgit v0.10.2 From 720fc22a7af79d91ec460c80efa92c65c12d105e Mon Sep 17 00:00:00 2001 From: Mikulas Patocka Date: Tue, 6 Apr 2010 16:43:33 -0700 Subject: ide: Fix IDE taskfile with cfq scheduler When ide taskfile access is being used (for example with hdparm --security commands) and cfq scheduler is selected, the scheduler crashes on BUG in cfq_put_request. The reason is that the cfq scheduler is tracking counts of read and write requests separately; the ide-taskfile subsystem allocates a read request and then flips the flag to make it a write request. The counters in cfq will mismatch. This patch changes ide-taskfile to allocate the READ or WRITE request as required and don't change the flag later. Signed-off-by: Mikulas Patocka Signed-off-by: David S. Miller diff --git a/drivers/ide/ide-taskfile.c b/drivers/ide/ide-taskfile.c index cc8633c..67fb735 100644 --- a/drivers/ide/ide-taskfile.c +++ b/drivers/ide/ide-taskfile.c @@ -428,13 +428,11 @@ int ide_raw_taskfile(ide_drive_t *drive, struct ide_cmd *cmd, u8 *buf, { struct request *rq; int error; + int rw = !(cmd->tf_flags & IDE_TFLAG_WRITE) ? READ : WRITE; - rq = blk_get_request(drive->queue, READ, __GFP_WAIT); + rq = blk_get_request(drive->queue, rw, __GFP_WAIT); rq->cmd_type = REQ_TYPE_ATA_TASKFILE; - if (cmd->tf_flags & IDE_TFLAG_WRITE) - rq->cmd_flags |= REQ_RW; - /* * (ks) We transfer currently only whole sectors. * This is suffient for now. But, it would be great, -- cgit v0.10.2 From ba1163de2f74d624e7b0e530c4104c98ede0045a Mon Sep 17 00:00:00 2001 From: Adam Jackson Date: Tue, 6 Apr 2010 16:11:00 +0000 Subject: drm/edid/quirks: Envision EN2028 Claims 1280x1024 preferred, physically 1600x1200 cf. http://bugzilla.redhat.com/530399 Signed-off-by: Adam Jackson Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c index 7e608f4..e661e4c 100644 --- a/drivers/gpu/drm/drm_edid.c +++ b/drivers/gpu/drm/drm_edid.c @@ -84,6 +84,8 @@ static struct edid_quirk { /* Envision Peripherals, Inc. EN-7100e */ { "EPI", 59264, EDID_QUIRK_135_CLOCK_TOO_HIGH }, + /* Envision EN2028 */ + { "EPI", 8232, EDID_QUIRK_PREFER_LARGE_60 }, /* Funai Electronics PM36B */ { "FCM", 13600, EDID_QUIRK_PREFER_LARGE_75 | -- cgit v0.10.2 From 01a356fd2a5f5e72e783312037ace05df4ab4e32 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 6 Apr 2010 10:55:33 +0000 Subject: drm/radeon/kms: small memory leak in atom exit code This is an unlikely memory leak, but we may as well fix it. It's easy to fix and every static checker will complain if we don't. Signed-off-by: Dan Carpenter Reviewed-by: Matt Turner Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/atom.c b/drivers/gpu/drm/radeon/atom.c index 247f8ee..58845e0 100644 --- a/drivers/gpu/drm/radeon/atom.c +++ b/drivers/gpu/drm/radeon/atom.c @@ -1136,6 +1136,7 @@ static int atom_execute_table_locked(struct atom_context *ctx, int index, uint32 int len, ws, ps, ptr; unsigned char op; atom_exec_context ectx; + int ret = 0; if (!base) return -EINVAL; @@ -1168,7 +1169,8 @@ static int atom_execute_table_locked(struct atom_context *ctx, int index, uint32 if (ectx.abort) { DRM_ERROR("atombios stuck executing %04X (len %d, WS %d, PS %d) @ 0x%04X\n", base, len, ws, ps, ptr - 1); - return -EINVAL; + ret = -EINVAL; + goto free; } if (op < ATOM_OP_CNT && op > 0) @@ -1183,9 +1185,10 @@ static int atom_execute_table_locked(struct atom_context *ctx, int index, uint32 debug_depth--; SDEBUG("<<\n"); +free: if (ws) kfree(ectx.ws); - return 0; + return ret; } int atom_execute_table(struct atom_context *ctx, int index, uint32_t * params) -- cgit v0.10.2 From 7741618766417e77f49013400d3672d65578928a Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 6 Apr 2010 00:05:46 -0400 Subject: drm/radeon/kms: legacy tv dac cleanup - fix formatting - clean up tv_dac_cntl handling for tv Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_legacy_encoders.c b/drivers/gpu/drm/radeon/radeon_legacy_encoders.c index cf389ce..2441cca 100644 --- a/drivers/gpu/drm/radeon/radeon_legacy_encoders.c +++ b/drivers/gpu/drm/radeon/radeon_legacy_encoders.c @@ -830,8 +830,8 @@ static void radeon_legacy_tv_dac_dpms(struct drm_encoder *encoder, int mode) crtc2_gen_cntl &= ~RADEON_CRTC2_CRT2_ON; if (rdev->family == CHIP_R420 || - rdev->family == CHIP_R423 || - rdev->family == CHIP_RV410) + rdev->family == CHIP_R423 || + rdev->family == CHIP_RV410) tv_dac_cntl |= (R420_TV_DAC_RDACPD | R420_TV_DAC_GDACPD | R420_TV_DAC_BDACPD | @@ -907,35 +907,43 @@ static void radeon_legacy_tv_dac_mode_set(struct drm_encoder *encoder, if (rdev->family != CHIP_R200) { tv_dac_cntl = RREG32(RADEON_TV_DAC_CNTL); if (rdev->family == CHIP_R420 || - rdev->family == CHIP_R423 || - rdev->family == CHIP_RV410) { + rdev->family == CHIP_R423 || + rdev->family == CHIP_RV410) { tv_dac_cntl &= ~(RADEON_TV_DAC_STD_MASK | - RADEON_TV_DAC_BGADJ_MASK | - R420_TV_DAC_DACADJ_MASK | - R420_TV_DAC_RDACPD | - R420_TV_DAC_GDACPD | - R420_TV_DAC_BDACPD | - R420_TV_DAC_TVENABLE); + RADEON_TV_DAC_BGADJ_MASK | + R420_TV_DAC_DACADJ_MASK | + R420_TV_DAC_RDACPD | + R420_TV_DAC_GDACPD | + R420_TV_DAC_BDACPD | + R420_TV_DAC_TVENABLE); } else { tv_dac_cntl &= ~(RADEON_TV_DAC_STD_MASK | - RADEON_TV_DAC_BGADJ_MASK | - RADEON_TV_DAC_DACADJ_MASK | - RADEON_TV_DAC_RDACPD | - RADEON_TV_DAC_GDACPD | - RADEON_TV_DAC_BDACPD); + RADEON_TV_DAC_BGADJ_MASK | + RADEON_TV_DAC_DACADJ_MASK | + RADEON_TV_DAC_RDACPD | + RADEON_TV_DAC_GDACPD | + RADEON_TV_DAC_BDACPD); } - /* FIXME TV */ - if (tv_dac) { - struct radeon_encoder_tv_dac *tv_dac = radeon_encoder->enc_priv; - tv_dac_cntl |= (RADEON_TV_DAC_NBLANK | - RADEON_TV_DAC_NHOLD | - RADEON_TV_DAC_STD_PS2 | - tv_dac->ps2_tvdac_adj); + tv_dac_cntl |= RADEON_TV_DAC_NBLANK | RADEON_TV_DAC_NHOLD; + + if (is_tv) { + if (tv_dac->tv_std == TV_STD_NTSC || + tv_dac->tv_std == TV_STD_NTSC_J || + tv_dac->tv_std == TV_STD_PAL_M || + tv_dac->tv_std == TV_STD_PAL_60) + tv_dac_cntl |= tv_dac->ntsc_tvdac_adj; + else + tv_dac_cntl |= tv_dac->pal_tvdac_adj; + + if (tv_dac->tv_std == TV_STD_NTSC || + tv_dac->tv_std == TV_STD_NTSC_J) + tv_dac_cntl |= RADEON_TV_DAC_STD_NTSC; + else + tv_dac_cntl |= RADEON_TV_DAC_STD_PAL; } else - tv_dac_cntl |= (RADEON_TV_DAC_NBLANK | - RADEON_TV_DAC_NHOLD | - RADEON_TV_DAC_STD_PS2); + tv_dac_cntl |= (RADEON_TV_DAC_STD_PS2 | + tv_dac->ps2_tvdac_adj); WREG32(RADEON_TV_DAC_CNTL, tv_dac_cntl); } -- cgit v0.10.2 From affd858907aae7bf7d7d7fa02ff19f35de2ff1d8 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 6 Apr 2010 01:22:41 -0400 Subject: drm/radeon/kms: clean up atom dac handling - make sure legacy dac1 has an enc priv - remove unused num var - no need for extra tv_dac var in atom dac functions Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c index 52d6f96..c52fc30 100644 --- a/drivers/gpu/drm/radeon/radeon_encoders.c +++ b/drivers/gpu/drm/radeon/radeon_encoders.c @@ -317,12 +317,8 @@ atombios_dac_setup(struct drm_encoder *encoder, int action) struct radeon_device *rdev = dev->dev_private; struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder); DAC_ENCODER_CONTROL_PS_ALLOCATION args; - int index = 0, num = 0; + int index = 0; struct radeon_encoder_atom_dac *dac_info = radeon_encoder->enc_priv; - enum radeon_tv_std tv_std = TV_STD_NTSC; - - if (dac_info->tv_std) - tv_std = dac_info->tv_std; memset(&args, 0, sizeof(args)); @@ -330,12 +326,10 @@ atombios_dac_setup(struct drm_encoder *encoder, int action) case ENCODER_OBJECT_ID_INTERNAL_DAC1: case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC1: index = GetIndexIntoMasterTable(COMMAND, DAC1EncoderControl); - num = 1; break; case ENCODER_OBJECT_ID_INTERNAL_DAC2: case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC2: index = GetIndexIntoMasterTable(COMMAND, DAC2EncoderControl); - num = 2; break; } @@ -346,7 +340,7 @@ atombios_dac_setup(struct drm_encoder *encoder, int action) else if (radeon_encoder->active_device & (ATOM_DEVICE_CV_SUPPORT)) args.ucDacStandard = ATOM_DAC1_CV; else { - switch (tv_std) { + switch (dac_info->tv_std) { case TV_STD_PAL: case TV_STD_PAL_M: case TV_STD_SCART_PAL: @@ -377,10 +371,6 @@ atombios_tv_setup(struct drm_encoder *encoder, int action) TV_ENCODER_CONTROL_PS_ALLOCATION args; int index = 0; struct radeon_encoder_atom_dac *dac_info = radeon_encoder->enc_priv; - enum radeon_tv_std tv_std = TV_STD_NTSC; - - if (dac_info->tv_std) - tv_std = dac_info->tv_std; memset(&args, 0, sizeof(args)); @@ -391,7 +381,7 @@ atombios_tv_setup(struct drm_encoder *encoder, int action) if (radeon_encoder->active_device & (ATOM_DEVICE_CV_SUPPORT)) args.sTVEncoder.ucTvStandard = ATOM_TV_CV; else { - switch (tv_std) { + switch (dac_info->tv_std) { case TV_STD_NTSC: args.sTVEncoder.ucTvStandard = ATOM_TV_NTSC; break; @@ -1558,12 +1548,14 @@ static const struct drm_encoder_funcs radeon_atom_enc_funcs = { struct radeon_encoder_atom_dac * radeon_atombios_set_dac_info(struct radeon_encoder *radeon_encoder) { + struct drm_device *dev = radeon_encoder->base.dev; + struct radeon_device *rdev = dev->dev_private; struct radeon_encoder_atom_dac *dac = kzalloc(sizeof(struct radeon_encoder_atom_dac), GFP_KERNEL); if (!dac) return NULL; - dac->tv_std = TV_STD_NTSC; + dac->tv_std = radeon_atombios_get_tv_info(rdev); return dac; } @@ -1641,6 +1633,7 @@ radeon_add_atom_encoder(struct drm_device *dev, uint32_t encoder_id, uint32_t su break; case ENCODER_OBJECT_ID_INTERNAL_DAC1: drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_DAC); + radeon_encoder->enc_priv = radeon_atombios_set_dac_info(radeon_encoder); drm_encoder_helper_add(encoder, &radeon_atom_dac_helper_funcs); break; case ENCODER_OBJECT_ID_INTERNAL_DAC2: -- cgit v0.10.2 From 3a89b4a9ca7ce11e3b7d5119aea917b9fc29a302 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 6 Apr 2010 12:35:26 -0400 Subject: drm/radeon/kms/combios: verify dac_adj values are valid Some vbios dac_adj tables are all zeros. Check for that case and use the default table if so. Should fix fdo bug 27478. Signed-off-by: Alex Deucher Cc: stable Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_combios.c b/drivers/gpu/drm/radeon/radeon_combios.c index 2becded..37db8ad 100644 --- a/drivers/gpu/drm/radeon/radeon_combios.c +++ b/drivers/gpu/drm/radeon/radeon_combios.c @@ -760,7 +760,9 @@ struct radeon_encoder_primary_dac *radeon_combios_get_primary_dac_info(struct dac = RBIOS8(dac_info + 0x3) & 0xf; p_dac->ps2_pdac_adj = (bg << 8) | (dac); } - found = 1; + /* if the values are all zeros, use the table */ + if (p_dac->ps2_pdac_adj) + found = 1; } if (!found) /* fallback to defaults */ @@ -895,7 +897,9 @@ struct radeon_encoder_tv_dac *radeon_combios_get_tv_dac_info(struct bg = RBIOS8(dac_info + 0x10) & 0xf; dac = RBIOS8(dac_info + 0x11) & 0xf; tv_dac->ntsc_tvdac_adj = (bg << 16) | (dac << 20); - found = 1; + /* if the values are all zeros, use the table */ + if (tv_dac->ps2_tvdac_adj) + found = 1; } else if (rev > 1) { bg = RBIOS8(dac_info + 0xc) & 0xf; dac = (RBIOS8(dac_info + 0xc) >> 4) & 0xf; @@ -908,7 +912,9 @@ struct radeon_encoder_tv_dac *radeon_combios_get_tv_dac_info(struct bg = RBIOS8(dac_info + 0xe) & 0xf; dac = (RBIOS8(dac_info + 0xe) >> 4) & 0xf; tv_dac->ntsc_tvdac_adj = (bg << 16) | (dac << 20); - found = 1; + /* if the values are all zeros, use the table */ + if (tv_dac->ps2_tvdac_adj) + found = 1; } tv_dac->tv_std = radeon_combios_get_tv_info(rdev); } @@ -925,7 +931,9 @@ struct radeon_encoder_tv_dac *radeon_combios_get_tv_dac_info(struct (bg << 16) | (dac << 20); tv_dac->pal_tvdac_adj = tv_dac->ps2_tvdac_adj; tv_dac->ntsc_tvdac_adj = tv_dac->ps2_tvdac_adj; - found = 1; + /* if the values are all zeros, use the table */ + if (tv_dac->ps2_tvdac_adj) + found = 1; } else { bg = RBIOS8(dac_info + 0x4) & 0xf; dac = RBIOS8(dac_info + 0x5) & 0xf; @@ -933,7 +941,9 @@ struct radeon_encoder_tv_dac *radeon_combios_get_tv_dac_info(struct (bg << 16) | (dac << 20); tv_dac->pal_tvdac_adj = tv_dac->ps2_tvdac_adj; tv_dac->ntsc_tvdac_adj = tv_dac->ps2_tvdac_adj; - found = 1; + /* if the values are all zeros, use the table */ + if (tv_dac->ps2_tvdac_adj) + found = 1; } } else { DRM_INFO("No TV DAC info found in BIOS\n"); -- cgit v0.10.2 From d2bf98e6fc2612cbd2fc40f75a1e2c5d60262331 Mon Sep 17 00:00:00 2001 From: "Steven J. Magnani" Date: Fri, 2 Apr 2010 13:29:39 -0500 Subject: microblaze: Fix ftrace_update_ftrace_func panic The Microblaze dynamic ftrace code assumes a call ordering that is not met in all scenarios. Specifically, executing a command similar to: echo 105 > /sys/kernel/debug/tracing/set_ftrace_pid before any other tracing-related commands results in a kernel panic: BUG: failure at arch/microblaze/kernel/ftrace.c:198/ftrace_update_ftrace_func()! Recoding ftrace_update_ftrace_func() to use &ftrace_caller directly eliminates the need to capture its address elsewhere (and thus rely on a particular call sequence). Signed-off-by: Steven J. Magnani Signed-off-by: Michal Simek diff --git a/arch/microblaze/kernel/ftrace.c b/arch/microblaze/kernel/ftrace.c index 388b31c..515feb4 100644 --- a/arch/microblaze/kernel/ftrace.c +++ b/arch/microblaze/kernel/ftrace.c @@ -151,13 +151,10 @@ int ftrace_make_nop(struct module *mod, return ret; } -static int ret_addr; /* initialized as 0 by default */ - /* I believe that first is called ftrace_make_nop before this function */ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr) { int ret; - ret_addr = addr; /* saving where the barrier jump is */ pr_debug("%s: addr:0x%x, rec->ip: 0x%x, imm:0x%x\n", __func__, (unsigned int)addr, (unsigned int)rec->ip, imm); ret = ftrace_modify_code(rec->ip, imm); @@ -194,12 +191,9 @@ int ftrace_update_ftrace_func(ftrace_func_t func) ret = ftrace_modify_code(ip, upper); ret += ftrace_modify_code(ip + 4, lower); - /* We just need to remove the rtsd r15, 8 by NOP */ - BUG_ON(!ret_addr); - if (ret_addr) - ret += ftrace_modify_code(ret_addr, MICROBLAZE_NOP); - else - ret = 1; /* fault */ + /* We just need to replace the rtsd r15, 8 with NOP */ + ret += ftrace_modify_code((unsigned long)&ftrace_caller, + MICROBLAZE_NOP); /* All changes are done - lets do caches consistent */ flush_icache(); -- cgit v0.10.2 From 4bb30baa6d5e7660c06e3b50d8a8a76d402c7170 Mon Sep 17 00:00:00 2001 From: Michal Simek Date: Tue, 6 Apr 2010 18:51:37 +0200 Subject: microblaze: Fix futex code Use logical and NOT for ANDN. It was likely type fault. Signed-off-by: Michal Simek diff --git a/arch/microblaze/include/asm/futex.h b/arch/microblaze/include/asm/futex.h index 8dbb6e7..ad3fd61 100644 --- a/arch/microblaze/include/asm/futex.h +++ b/arch/microblaze/include/asm/futex.h @@ -55,7 +55,7 @@ futex_atomic_op_inuser(int encoded_op, int __user *uaddr) __futex_atomic_op("or %1,%0,%4;", ret, oldval, uaddr, oparg); break; case FUTEX_OP_ANDN: - __futex_atomic_op("and %1,%0,%4;", ret, oldval, uaddr, oparg); + __futex_atomic_op("andn %1,%0,%4;", ret, oldval, uaddr, oparg); break; case FUTEX_OP_XOR: __futex_atomic_op("xor %1,%0,%4;", ret, oldval, uaddr, oparg); -- cgit v0.10.2 From 47c4c864af60d14926f4017d23968a8341b8ab9f Mon Sep 17 00:00:00 2001 From: Michal Simek Date: Tue, 6 Apr 2010 10:19:30 +0200 Subject: microblaze: Enable memory leak detector Enable DEBUG_KMEMLEAK for microblaze Signed-off-by: Michal Simek diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index 1fafb4b..ff01710 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -356,7 +356,7 @@ config SLUB_STATS config DEBUG_KMEMLEAK bool "Kernel memory leak detector" depends on DEBUG_KERNEL && EXPERIMENTAL && !MEMORY_HOTPLUG && \ - (X86 || ARM || PPC || S390 || SUPERH) + (X86 || ARM || PPC || S390 || SUPERH || MICROBLAZE) select DEBUG_FS if SYSFS select STACKTRACE if STACKTRACE_SUPPORT -- cgit v0.10.2 From 99dc5c0ca953fb154a3c75ebc6801e0a3c2fe3fe Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Wed, 7 Apr 2010 17:08:32 +0900 Subject: sh: Disable MMUCR_AT for SH-4 nommu. Presently address translation is default-enabled regardless of whether CONFIG_MMU is set or not in the SH-4 case, this fixes it up, and also makes the control init word a bit more readable in the process. Signed-off-by: Paul Mundt diff --git a/arch/sh/include/cpu-sh4/cpu/mmu_context.h b/arch/sh/include/cpu-sh4/cpu/mmu_context.h index 2941be6..e46ec70 100644 --- a/arch/sh/include/cpu-sh4/cpu/mmu_context.h +++ b/arch/sh/include/cpu-sh4/cpu/mmu_context.h @@ -32,7 +32,13 @@ #define MMU_UTLB_DATA_ARRAY2 0xF7800000 #define MMU_PAGE_ASSOC_BIT 0x80 -#define MMUCR_TI (1<<2) +#ifdef CONFIG_MMU +#define MMUCR_AT (1 << 0) +#else +#define MMUCR_AT (0) +#endif + +#define MMUCR_TI (1 << 2) #define MMUCR_URB 0x00FC0000 #define MMUCR_URB_SHIFT 18 @@ -65,7 +71,8 @@ #endif #define MMU_NTLB_ENTRIES 64 -#define MMU_CONTROL_INIT (0x05|MMUCR_SQMD|MMUCR_ME|MMUCR_SE|MMUCR_AEX) +#define MMU_CONTROL_INIT (MMUCR_AT | MMUCR_TI | MMUCR_SQMD | \ + MMUCR_ME | MMUCR_SE | MMUCR_AEX) #define TRA 0xff000020 #define EXPEVT 0xff000024 -- cgit v0.10.2 From 75f66533bc883f761a7adcab3281fe3323efbc90 Mon Sep 17 00:00:00 2001 From: Chris Wright Date: Fri, 2 Apr 2010 18:27:52 -0700 Subject: x86/amd-iommu: enable iommu before attaching devices Hit another kdump problem as reported by Neil Horman. When initializaing the IOMMU, we attach devices to their domains before the IOMMU is fully (re)initialized. Attaching a device will issue some important invalidations. In the context of the newly kexec'd kdump kernel, the IOMMU may have stale cached data from the original kernel. Because we do the attach too early, the invalidation commands are placed in the new command buffer before the IOMMU is updated w/ that buffer. This leaves the stale entries in the kdump context and can renders device unusable. Simply enable the IOMMU before we do the attach. Cc: stable@kernel.org Cc: Neil Horman Cc: Vivek Goyal Signed-off-by: Chris Wright Signed-off-by: Joerg Roedel diff --git a/arch/x86/kernel/amd_iommu_init.c b/arch/x86/kernel/amd_iommu_init.c index feaf471..8975965 100644 --- a/arch/x86/kernel/amd_iommu_init.c +++ b/arch/x86/kernel/amd_iommu_init.c @@ -1304,6 +1304,8 @@ static int __init amd_iommu_init(void) if (ret) goto free; + enable_iommus(); + if (iommu_pass_through) ret = amd_iommu_init_passthrough(); else @@ -1316,8 +1318,6 @@ static int __init amd_iommu_init(void) amd_iommu_init_notifier(); - enable_iommus(); - if (iommu_pass_through) goto out; @@ -1331,6 +1331,7 @@ out: return ret; free: + disable_iommus(); amd_iommu_uninit_devices(); -- cgit v0.10.2 From 549c90dc9a6d659e792b2a42a0930c7da015ea4a Mon Sep 17 00:00:00 2001 From: Chris Wright Date: Fri, 2 Apr 2010 18:27:53 -0700 Subject: x86/amd-iommu: warn when issuing command to uninitialized cmd buffer To catch future potential issues we can add a warning whenever we issue a command before the command buffer is fully initialized. Signed-off-by: Chris Wright Signed-off-by: Joerg Roedel diff --git a/arch/x86/include/asm/amd_iommu_types.h b/arch/x86/include/asm/amd_iommu_types.h index 5e46e78..86a0ff0 100644 --- a/arch/x86/include/asm/amd_iommu_types.h +++ b/arch/x86/include/asm/amd_iommu_types.h @@ -141,6 +141,7 @@ /* constants to configure the command buffer */ #define CMD_BUFFER_SIZE 8192 +#define CMD_BUFFER_UNINITIALIZED 1 #define CMD_BUFFER_ENTRIES 512 #define MMIO_CMD_SIZE_SHIFT 56 #define MMIO_CMD_SIZE_512 (0x9ULL << MMIO_CMD_SIZE_SHIFT) diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index b06f29e..71dfc0a 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -392,6 +392,7 @@ static int __iommu_queue_command(struct amd_iommu *iommu, struct iommu_cmd *cmd) u32 tail, head; u8 *target; + WARN_ON(iommu->cmd_buf_size & CMD_BUFFER_UNINITIALIZED); tail = readl(iommu->mmio_base + MMIO_CMD_TAIL_OFFSET); target = iommu->cmd_buf + tail; memcpy_toio(target, cmd, sizeof(*cmd)); diff --git a/arch/x86/kernel/amd_iommu_init.c b/arch/x86/kernel/amd_iommu_init.c index 8975965..5edf41c 100644 --- a/arch/x86/kernel/amd_iommu_init.c +++ b/arch/x86/kernel/amd_iommu_init.c @@ -438,7 +438,7 @@ static u8 * __init alloc_command_buffer(struct amd_iommu *iommu) if (cmd_buf == NULL) return NULL; - iommu->cmd_buf_size = CMD_BUFFER_SIZE; + iommu->cmd_buf_size = CMD_BUFFER_SIZE | CMD_BUFFER_UNINITIALIZED; return cmd_buf; } @@ -474,12 +474,13 @@ static void iommu_enable_command_buffer(struct amd_iommu *iommu) &entry, sizeof(entry)); amd_iommu_reset_cmd_buffer(iommu); + iommu->cmd_buf_size &= ~(CMD_BUFFER_UNINITIALIZED); } static void __init free_command_buffer(struct amd_iommu *iommu) { free_pages((unsigned long)iommu->cmd_buf, - get_order(iommu->cmd_buf_size)); + get_order(iommu->cmd_buf_size & ~(CMD_BUFFER_UNINITIALIZED))); } /* allocates the memory where the IOMMU will log its events to */ -- cgit v0.10.2 From 8f9f55e83e939724490d7cde3833c4883c6d1310 Mon Sep 17 00:00:00 2001 From: Chris Wright Date: Fri, 2 Apr 2010 18:27:54 -0700 Subject: Revert "x86: disable IOMMUs on kernel crash" This effectively reverts commit 61d047be99757fd9b0af900d7abce9a13a337488. Disabling the IOMMU can potetially allow DMA transactions to complete without being translated. Leave it enabled, and allow crash kernel to do the IOMMU reinitialization properly. Cc: stable@kernel.org Cc: Joerg Roedel Cc: Eric Biederman Cc: Neil Horman Cc: Vivek Goyal Signed-off-by: Chris Wright Signed-off-by: Joerg Roedel diff --git a/arch/x86/kernel/crash.c b/arch/x86/kernel/crash.c index a4849c1..ebd4c51 100644 --- a/arch/x86/kernel/crash.c +++ b/arch/x86/kernel/crash.c @@ -27,7 +27,6 @@ #include #include #include -#include #if defined(CONFIG_SMP) && defined(CONFIG_X86_LOCAL_APIC) @@ -103,10 +102,5 @@ void native_machine_crash_shutdown(struct pt_regs *regs) #ifdef CONFIG_HPET_TIMER hpet_disable(); #endif - -#ifdef CONFIG_X86_64 - x86_platform.iommu_shutdown(); -#endif - crash_save_cpu(regs, safe_smp_processor_id()); } -- cgit v0.10.2 From d18c69d3898985c66cd6e878b8f576fd9a21ab39 Mon Sep 17 00:00:00 2001 From: Chris Wright Date: Fri, 2 Apr 2010 18:27:55 -0700 Subject: x86/amd-iommu: use for_each_pci_dev Replace open coded version with for_each_pci_dev Signed-off-by: Chris Wright Signed-off-by: Joerg Roedel diff --git a/arch/x86/kernel/amd_iommu.c b/arch/x86/kernel/amd_iommu.c index 71dfc0a..4949568 100644 --- a/arch/x86/kernel/amd_iommu.c +++ b/arch/x86/kernel/amd_iommu.c @@ -2187,7 +2187,7 @@ static void prealloc_protection_domains(void) struct dma_ops_domain *dma_dom; u16 devid; - while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) { + for_each_pci_dev(dev) { /* Do we handle this device? */ if (!check_device(&dev->dev)) -- cgit v0.10.2 From c0858552c088616c18879c347d9e0daa98cf2b15 Mon Sep 17 00:00:00 2001 From: Jiri Kosina Date: Wed, 7 Apr 2010 12:10:29 +0200 Subject: HID: ntrig: explain firmware quirk Commit 6549981bc54777c ("HID: fix N-trig touch panel with recent firmware") adds a quirk that is needed for devices with more recent firmware so that they become operational. As it's not directly obvious from the code why is it needed, a comment is worthwile. Signed-off-by: Jiri Kosina diff --git a/drivers/hid/hid-ntrig.c b/drivers/hid/hid-ntrig.c index 513db0a..a418f9e 100644 --- a/drivers/hid/hid-ntrig.c +++ b/drivers/hid/hid-ntrig.c @@ -352,7 +352,8 @@ static int ntrig_probe(struct hid_device *hdev, const struct hid_device_id *id) } } - report = hdev->report_enum[HID_FEATURE_REPORT].report_id_hash[0x0a]; + /* This is needed for devices with more recent firmware versions */ + report = hdev->report_enum[HID_FEATURE_REPORT].report_id_hash[0x0a]; if (report) usbhid_submit_report(hdev, report, USB_DIR_OUT); -- cgit v0.10.2 From 20a1cfba340f23a7ca62391e199c40c86b762ea3 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Wed, 7 Apr 2010 14:28:26 +0200 Subject: x86/amd-iommu: Remove obsolete parameter documentation Support for the share and fullflush parameters was removed. Remove the documentation about them too. Signed-off-by: Joerg Roedel diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index e7848a0..ccea846 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -323,11 +323,6 @@ and is between 256 and 4096 characters. It is defined in the file amd_iommu= [HW,X86-84] Pass parameters to the AMD IOMMU driver in the system. Possible values are: - isolate - enable device isolation (each device, as far - as possible, will get its own protection - domain) [default] - share - put every device behind one IOMMU into the - same protection domain fullflush - enable flushing of IO/TLB entries when they are unmapped. Otherwise they are flushed before they will be reused, which -- cgit v0.10.2 From 39a37ce1cc5eef420604fa68b776ff5dab400340 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 6 Apr 2010 19:45:12 +0300 Subject: dma-debug: Cleanup for copy-loop in filter_write() Earlier in this function we set the last byte of "buf" to NULL so we always hit the break statement and "i" is never equal to NAME_MAX_LEN. This patch doesn't change how the driver works but it silences a Smatch warning and it makes it clearer that we don't write past the end of the array. Signed-off-by: Dan Carpenter Signed-off-by: Joerg Roedel diff --git a/lib/dma-debug.c b/lib/dma-debug.c index ba8b670..01e6427 100644 --- a/lib/dma-debug.c +++ b/lib/dma-debug.c @@ -570,7 +570,7 @@ static ssize_t filter_write(struct file *file, const char __user *userbuf, * Now parse out the first token and use it as the name for the * driver to filter for. */ - for (i = 0; i < NAME_MAX_LEN; ++i) { + for (i = 0; i < NAME_MAX_LEN - 1; ++i) { current_driver_name[i] = buf[i]; if (isspace(buf[i]) || buf[i] == ' ' || buf[i] == 0) break; -- cgit v0.10.2 From 4b83873d3da0704987cb116833818ed96214ee29 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Wed, 7 Apr 2010 12:57:35 +0200 Subject: x86/gart: Disable GART explicitly before initialization If we boot into a crash-kernel the gart might still be enabled and its caches might be dirty. This can result in undefined behavior later. Fix it by explicitly disabling the gart hardware before initialization and flushing the caches after enablement. Signed-off-by: Joerg Roedel diff --git a/arch/x86/kernel/aperture_64.c b/arch/x86/kernel/aperture_64.c index 3704997..b5d8b0b 100644 --- a/arch/x86/kernel/aperture_64.c +++ b/arch/x86/kernel/aperture_64.c @@ -393,6 +393,7 @@ void __init gart_iommu_hole_init(void) for (i = 0; i < ARRAY_SIZE(bus_dev_ranges); i++) { int bus; int dev_base, dev_limit; + u32 ctl; bus = bus_dev_ranges[i].bus; dev_base = bus_dev_ranges[i].dev_base; @@ -406,7 +407,19 @@ void __init gart_iommu_hole_init(void) gart_iommu_aperture = 1; x86_init.iommu.iommu_init = gart_iommu_init; - aper_order = (read_pci_config(bus, slot, 3, AMD64_GARTAPERTURECTL) >> 1) & 7; + ctl = read_pci_config(bus, slot, 3, + AMD64_GARTAPERTURECTL); + + /* + * Before we do anything else disable the GART. It may + * still be enabled if we boot into a crash-kernel here. + * Reconfiguring the GART while it is enabled could have + * unknown side-effects. + */ + ctl &= ~GARTEN; + write_pci_config(bus, slot, 3, AMD64_GARTAPERTURECTL, ctl); + + aper_order = (ctl >> 1) & 7; aper_size = (32 * 1024 * 1024) << aper_order; aper_base = read_pci_config(bus, slot, 3, AMD64_GARTAPERTUREBASE) & 0x7fff; aper_base <<= 25; diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c index f3af115..0ae24d9b 100644 --- a/arch/x86/kernel/pci-gart_64.c +++ b/arch/x86/kernel/pci-gart_64.c @@ -564,6 +564,9 @@ static void enable_gart_translations(void) enable_gart_translation(dev, __pa(agp_gatt_table)); } + + /* Flush the GART-TLB to remove stale entries */ + k8_flush_garts(); } /* -- cgit v0.10.2 From 7ad7b218f4aae4f395b3b4cef261572556bbd20a Mon Sep 17 00:00:00 2001 From: Maurus Cuelenaere Date: Tue, 6 Apr 2010 18:12:52 +0200 Subject: ALSA: hda: Add support for Medion WIM2160 This adds support for the Medion WIM2160 soundcard. There's no PCI quirk added because it has the same PCI id as the Medion MD2. Signed-off-by: Maurus Cuelenaere Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index c7730db..2971e48 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -230,6 +230,7 @@ enum { ALC888_ACER_ASPIRE_7730G, ALC883_MEDION, ALC883_MEDION_MD2, + ALC883_MEDION_WIM2160, ALC883_LAPTOP_EAPD, ALC883_LENOVO_101E_2ch, ALC883_LENOVO_NB0763, @@ -8455,6 +8456,42 @@ static struct snd_kcontrol_new alc883_medion_md2_mixer[] = { { } /* end */ }; +static struct snd_kcontrol_new alc883_medion_wim2160_mixer[] = { + HDA_CODEC_VOLUME("Front Playback Volume", 0x0c, 0x0, HDA_OUTPUT), + HDA_BIND_MUTE("Front Playback Switch", 0x0c, 2, HDA_INPUT), + HDA_CODEC_MUTE("Speaker Playback Switch", 0x15, 0x0, HDA_OUTPUT), + HDA_CODEC_MUTE("Headphone Playback Switch", 0x1a, 0x0, HDA_OUTPUT), + HDA_CODEC_VOLUME("Line Playback Volume", 0x08, 0x0, HDA_INPUT), + HDA_CODEC_MUTE("Line Playback Switch", 0x08, 0x0, HDA_INPUT), + { } /* end */ +}; + +static struct hda_verb alc883_medion_wim2160_verbs[] = { + /* Unmute front mixer */ + {0x0c, AC_VERB_SET_AMP_GAIN_MUTE, AMP_IN_UNMUTE(0)}, + {0x0c, AC_VERB_SET_AMP_GAIN_MUTE, AMP_IN_UNMUTE(1)}, + + /* Set speaker pin to front mixer */ + {0x15, AC_VERB_SET_CONNECT_SEL, 0x00}, + + /* Init headphone pin */ + {0x1a, AC_VERB_SET_PIN_WIDGET_CONTROL, PIN_HP}, + {0x1a, AC_VERB_SET_AMP_GAIN_MUTE, AMP_OUT_UNMUTE}, + {0x1a, AC_VERB_SET_CONNECT_SEL, 0x00}, + {0x1a, AC_VERB_SET_UNSOLICITED_ENABLE, ALC880_HP_EVENT | AC_USRSP_EN}, + + { } /* end */ +}; + +/* toggle speaker-output according to the hp-jack state */ +static void alc883_medion_wim2160_setup(struct hda_codec *codec) +{ + struct alc_spec *spec = codec->spec; + + spec->autocfg.hp_pins[0] = 0x1a; + spec->autocfg.speaker_pins[0] = 0x15; +} + static struct snd_kcontrol_new alc883_acer_aspire_mixer[] = { HDA_CODEC_VOLUME("Front Playback Volume", 0x0c, 0x0, HDA_OUTPUT), HDA_BIND_MUTE("Front Playback Switch", 0x0c, 2, HDA_INPUT), @@ -9164,6 +9201,7 @@ static const char *alc882_models[ALC882_MODEL_LAST] = { [ALC888_ACER_ASPIRE_7730G] = "acer-aspire-7730g", [ALC883_MEDION] = "medion", [ALC883_MEDION_MD2] = "medion-md2", + [ALC883_MEDION_WIM2160] = "medion-wim2160", [ALC883_LAPTOP_EAPD] = "laptop-eapd", [ALC883_LENOVO_101E_2ch] = "lenovo-101e", [ALC883_LENOVO_NB0763] = "lenovo-nb0763", @@ -9818,6 +9856,21 @@ static struct alc_config_preset alc882_presets[] = { .setup = alc883_medion_md2_setup, .init_hook = alc_automute_amp, }, + [ALC883_MEDION_WIM2160] = { + .mixers = { alc883_medion_wim2160_mixer }, + .init_verbs = { alc883_init_verbs, alc883_medion_wim2160_verbs }, + .num_dacs = ARRAY_SIZE(alc883_dac_nids), + .dac_nids = alc883_dac_nids, + .dig_out_nid = ALC883_DIGOUT_NID, + .num_adc_nids = ARRAY_SIZE(alc883_adc_nids), + .adc_nids = alc883_adc_nids, + .num_channel_mode = ARRAY_SIZE(alc883_3ST_2ch_modes), + .channel_mode = alc883_3ST_2ch_modes, + .input_mux = &alc883_capture_source, + .unsol_event = alc_automute_amp_unsol_event, + .setup = alc883_medion_wim2160_setup, + .init_hook = alc_automute_amp, + }, [ALC883_LAPTOP_EAPD] = { .mixers = { alc883_base_mixer }, .init_verbs = { alc883_init_verbs, alc882_eapd_verbs }, -- cgit v0.10.2 From 7a325aa26cf8eb367b5d8b675a9425bef18d351a Mon Sep 17 00:00:00 2001 From: John Linn Date: Wed, 7 Apr 2010 11:03:15 +0200 Subject: microblaze: io.h: Add io big-endian function Add new io big-endian function. They will be used for uartlite and spi driver. Signed-off-by: John Linn Signed-off-by: Michal Simek diff --git a/arch/microblaze/include/asm/io.h b/arch/microblaze/include/asm/io.h index 32d621a..e45a6ee 100644 --- a/arch/microblaze/include/asm/io.h +++ b/arch/microblaze/include/asm/io.h @@ -108,6 +108,11 @@ static inline void writel(unsigned int v, volatile void __iomem *addr) #define iowrite16(v, addr) __raw_writew((u16)(v), (u16 *)(addr)) #define iowrite32(v, addr) __raw_writel((u32)(v), (u32 *)(addr)) +#define ioread16be(addr) __raw_readw((u16 *)(addr)) +#define ioread32be(addr) __raw_readl((u32 *)(addr)) +#define iowrite16be(v, addr) __raw_writew((u16)(v), (u16 *)(addr)) +#define iowrite32be(v, addr) __raw_writel((u32)(v), (u32 *)(addr)) + /* These are the definitions for the x86 IO instructions * inb/inw/inl/outb/outw/outl, the "string" versions * insb/insw/insl/outsb/outsw/outsl, and the "pausing" versions -- cgit v0.10.2 From af2b021fd6dc3e145d20f5eb927af9590c80bc44 Mon Sep 17 00:00:00 2001 From: Michal Simek Date: Wed, 7 Apr 2010 11:44:00 +0200 Subject: microblaze: Remove unused variable from ptrace Patch b3c1e01a09d6af2dd7811a066ffcfc5171be2bed should do it. Signed-off-by: Michal Simek diff --git a/arch/microblaze/kernel/ptrace.c b/arch/microblaze/kernel/ptrace.c index 6d6349a..a4a7770 100644 --- a/arch/microblaze/kernel/ptrace.c +++ b/arch/microblaze/kernel/ptrace.c @@ -75,7 +75,6 @@ long arch_ptrace(struct task_struct *child, long request, long addr, long data) { int rval; unsigned long val = 0; - unsigned long copied; switch (request) { /* Read/write the word at location ADDR in the registers. */ -- cgit v0.10.2 From 179b284e2fc0c638035843968f7d7ab8ab701525 Mon Sep 17 00:00:00 2001 From: Jeff Dike Date: Wed, 7 Apr 2010 09:59:10 -0400 Subject: vhost-net: fix vq_memory_access_ok error checking vq_memory_access_ok needs to check whether mem == NULL Signed-off-by: Jeff Dike Signed-off-by: Michael S. Tsirkin diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c index 7bd7a1e..b8e1127 100644 --- a/drivers/vhost/vhost.c +++ b/drivers/vhost/vhost.c @@ -235,6 +235,10 @@ static int vq_memory_access_ok(void __user *log_base, struct vhost_memory *mem, int log_all) { int i; + + if (!mem) + return 0; + for (i = 0; i < mem->nregions; ++i) { struct vhost_memory_region *m = mem->regions + i; unsigned long a = m->userspace_addr; -- cgit v0.10.2 From b01d0942c2b7a3026d2b7d38b5773d3d00420e06 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Tue, 6 Apr 2010 14:34:41 -0700 Subject: bitops: remove temporary for_each_bit() Migration has been completed so remove this now. There's one straggler in linux-next's drivers/mtd/sm_ftl.c. A patch has been sent. Cc: Akinobu Mita Cc: Stephen Rothwell Cc: David Woodhouse Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/bitops.h b/include/linux/bitops.h index b793898..b796eab 100644 --- a/include/linux/bitops.h +++ b/include/linux/bitops.h @@ -21,9 +21,6 @@ (bit) < (size); \ (bit) = find_next_bit((addr), (size), (bit) + 1)) -/* Temporary */ -#define for_each_bit(bit, addr, size) for_each_set_bit(bit, addr, size) - static __inline__ int get_bitmask_order(unsigned int count) { int order; -- cgit v0.10.2 From a3a2e76c77fa22b114e421ac11dec0c56c3503fb Mon Sep 17 00:00:00 2001 From: KAMEZAWA Hiroyuki Date: Tue, 6 Apr 2010 14:34:42 -0700 Subject: mm: avoid null-pointer deref in sync_mm_rss() - We weren't zeroing p->rss_stat[] at fork() - Consequently sync_mm_rss() was dereferencing tsk->mm for kernel threads and was oopsing. - Make __sync_task_rss_stat() static, too. Addresses https://bugzilla.kernel.org/show_bug.cgi?id=15648 [akpm@linux-foundation.org: remove the BUG_ON(!mm->rss)] Reported-by: Troels Liebe Bentsen Signed-off-by: KAMEZAWA Hiroyuki "Michael S. Tsirkin" Cc: Andrea Arcangeli Cc: Rik van Riel Cc: Minchan Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/exit.c b/kernel/exit.c index cce59cb..7f2683a 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -953,7 +953,8 @@ NORET_TYPE void do_exit(long code) acct_update_integrals(tsk); /* sync mm's RSS info before statistics gathering */ - sync_mm_rss(tsk, tsk->mm); + if (tsk->mm) + sync_mm_rss(tsk, tsk->mm); group_dead = atomic_dec_and_test(&tsk->signal->live); if (group_dead) { hrtimer_cancel(&tsk->signal->real_timer); diff --git a/kernel/fork.c b/kernel/fork.c index 4799c5f..44b0791 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -1052,6 +1052,9 @@ static struct task_struct *copy_process(unsigned long clone_flags, p->prev_utime = cputime_zero; p->prev_stime = cputime_zero; #endif +#if defined(SPLIT_RSS_COUNTING) + memset(&p->rss_stat, 0, sizeof(p->rss_stat)); +#endif p->default_timer_slack_ns = current->timer_slack_ns; diff --git a/mm/memory.c b/mm/memory.c index 1d2ea39..833952d 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -125,13 +125,12 @@ core_initcall(init_zero_pfn); #if defined(SPLIT_RSS_COUNTING) -void __sync_task_rss_stat(struct task_struct *task, struct mm_struct *mm) +static void __sync_task_rss_stat(struct task_struct *task, struct mm_struct *mm) { int i; for (i = 0; i < NR_MM_COUNTERS; i++) { if (task->rss_stat.count[i]) { - BUG_ON(!mm); add_mm_counter(mm, i, task->rss_stat.count[i]); task->rss_stat.count[i] = 0; } -- cgit v0.10.2 From 530cd330dc3865e3107304a6e84fdc332aa72f7d Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?David=20H=C3=A4rdeman?= Date: Tue, 6 Apr 2010 14:34:43 -0700 Subject: include/linux/kfifo.h: fix INIT_KFIFO() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit DECLARE_KFIFO creates a union with a struct kfifo and a buffer array with size [size + sizeof(struct kfifo)]. INIT_KFIFO then sets the buffer pointer in struct kfifo to point to the beginning of the buffer array which means that the first call to kfifo_in will overwrite members of the struct kfifo. Signed-off-by: David Härdeman Acked-by: Stefani Seibold Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/kfifo.h b/include/linux/kfifo.h index ece0b1c..e117b1a 100644 --- a/include/linux/kfifo.h +++ b/include/linux/kfifo.h @@ -86,7 +86,8 @@ union { \ */ #define INIT_KFIFO(name) \ name = __kfifo_initializer(sizeof(name##kfifo_buffer) - \ - sizeof(struct kfifo), name##kfifo_buffer) + sizeof(struct kfifo), \ + name##kfifo_buffer + sizeof(struct kfifo)) /** * DEFINE_KFIFO - macro to define and initialize a kfifo -- cgit v0.10.2 From fc95c6d9b4433525ed9484089b6e092f3516a92b Mon Sep 17 00:00:00 2001 From: Jan Beulich Date: Tue, 6 Apr 2010 14:34:44 -0700 Subject: vesafb: use platform_driver_probe() instead of platform_driver_register() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit c2e13037e6794bd0d9de3f9ecabf5615f15c160b ("platform-drivers: move probe to .devinit.text in drivers/video") introduced a huge amount of section mismatch warnings in vesafb code. Rather than converting all of the annotations, do the obvious and revert the __init -> __devinit change, and use the recommended (in that patch) alternative to calling platform_driver_register(): vesafb depends on information obtained from by kernel at boot time, cannot be a module, and no post-boot devices can ever show up. Signed-off-by: Jan Beulich Cc: Greg KH Acked-by: Uwe Kleine-König Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/video/vesafb.c b/drivers/video/vesafb.c index 54ac91d..0cadf7a 100644 --- a/drivers/video/vesafb.c +++ b/drivers/video/vesafb.c @@ -225,7 +225,7 @@ static int __init vesafb_setup(char *options) return 0; } -static int __devinit vesafb_probe(struct platform_device *dev) +static int __init vesafb_probe(struct platform_device *dev) { struct fb_info *info; int i, err; @@ -476,7 +476,6 @@ err: } static struct platform_driver vesafb_driver = { - .probe = vesafb_probe, .driver = { .name = "vesafb", }, @@ -492,20 +491,21 @@ static int __init vesafb_init(void) /* ignore error return of fb_get_options */ fb_get_options("vesafb", &option); vesafb_setup(option); - ret = platform_driver_register(&vesafb_driver); - if (!ret) { - vesafb_device = platform_device_alloc("vesafb", 0); + vesafb_device = platform_device_alloc("vesafb", 0); + if (!vesafb_device) + return -ENOMEM; - if (vesafb_device) - ret = platform_device_add(vesafb_device); - else - ret = -ENOMEM; + ret = platform_device_add(vesafb_device); + if (!ret) { + ret = platform_driver_probe(&vesafb_driver, vesafb_probe); + if (ret) + platform_device_del(vesafb_device); + } - if (ret) { - platform_device_put(vesafb_device); - platform_driver_unregister(&vesafb_driver); - } + if (ret) { + platform_device_put(vesafb_device); + vesafb_device = NULL; } return ret; -- cgit v0.10.2 From a8557dc71949e80c298ec298b902ac6ebbc5d9dd Mon Sep 17 00:00:00 2001 From: "Justin P. Mattock" Date: Tue, 6 Apr 2010 14:34:45 -0700 Subject: fbdev: rename imacfb.txt to efifb.txt and change imacfb to efifb. Rename imacfb.txt to efifb.txt since imacfb was moved to efifb,and change imacfb to efifb. Signed-off-by: Justin P. Mattock Cc: Geert Uytterhoeven Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/Documentation/fb/efifb.txt b/Documentation/fb/efifb.txt new file mode 100644 index 0000000..a59916c --- /dev/null +++ b/Documentation/fb/efifb.txt @@ -0,0 +1,31 @@ + +What is efifb? +=============== + +This is a generic EFI platform driver for Intel based Apple computers. +efifb is only for EFI booted Intel Macs. + +Supported Hardware +================== + +iMac 17"/20" +Macbook +Macbook Pro 15"/17" +MacMini + +How to use it? +============== + +efifb does not have any kind of autodetection of your machine. +You have to add the following kernel parameters in your elilo.conf: + Macbook : + video=efifb:macbook + MacMini : + video=efifb:mini + Macbook Pro 15", iMac 17" : + video=efifb:i17 + Macbook Pro 17", iMac 20" : + video=efifb:i20 + +-- +Edgar Hucek diff --git a/Documentation/fb/imacfb.txt b/Documentation/fb/imacfb.txt deleted file mode 100644 index 316ec9b..0000000 --- a/Documentation/fb/imacfb.txt +++ /dev/null @@ -1,31 +0,0 @@ - -What is imacfb? -=============== - -This is a generic EFI platform driver for Intel based Apple computers. -Imacfb is only for EFI booted Intel Macs. - -Supported Hardware -================== - -iMac 17"/20" -Macbook -Macbook Pro 15"/17" -MacMini - -How to use it? -============== - -Imacfb does not have any kind of autodetection of your machine. -You have to add the following kernel parameters in your elilo.conf: - Macbook : - video=imacfb:macbook - MacMini : - video=imacfb:mini - Macbook Pro 15", iMac 17" : - video=imacfb:i17 - Macbook Pro 17", iMac 20" : - video=imacfb:i20 - --- -Edgar Hucek -- cgit v0.10.2 From c70c036f04b7b74282477d6a51bbcc8bb8897e03 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Tue, 6 Apr 2010 14:34:46 -0700 Subject: drivers/char/amiserial.c: add missing local_irq_restore rs_init() is failing to restore interrupts on two error paths, and is incorrectly calling tty_unregister_driver() with local interrupts disabled. Fix these things by disabling interrupts later, after the reauest_irq() calls. A simplified version of the semantic patch that finds this problem is as follows: (http://coccinelle.lip6.fr/) // @r exists@ expression E1; identifier f; @@ f (...) { <+... * local_irq_save (E1,...); ... when != E1 * return ...; ...+> } // [akpm@linux-foundation.org: reimplement the fix] Signed-off-by: Julia Lawall Cc: Thadeu Lima de Souza Cascardo Cc: Greg KH Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/char/amiserial.c b/drivers/char/amiserial.c index 6c32fbf..56b2767 100644 --- a/drivers/char/amiserial.c +++ b/drivers/char/amiserial.c @@ -2021,8 +2021,6 @@ static int __init rs_init(void) state->baud_base = amiga_colorclock; state->xmit_fifo_size = 1; - local_irq_save(flags); - /* set ISRs, and then disable the rx interrupts */ error = request_irq(IRQ_AMIGA_TBE, ser_tx_int, 0, "serial TX", state); if (error) @@ -2033,6 +2031,8 @@ static int __init rs_init(void) if (error) goto fail_free_irq; + local_irq_save(flags); + /* turn off Rx and Tx interrupts */ custom.intena = IF_RBF | IF_TBE; mb(); -- cgit v0.10.2 From 2a481800caf78f4750cc673c8baed12b5d703ff6 Mon Sep 17 00:00:00 2001 From: Julia Lawall Date: Tue, 6 Apr 2010 14:34:48 -0700 Subject: drivers/gpio/timbgpio.c: add missing unlock In an error handling case the lock is not unlocked. The return is converted to a goto, to share the unlock at the end of the function. A simplified version of the semantic patch that finds this problem is as follows: (http://coccinelle.lip6.fr/) // @r exists@ expression E1; identifier f; @@ f (...) { <+... * spin_lock_irqsave (E1,...); ... when != E1 * return ...; ...+> } // Signed-off-by: Julia Lawall Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/gpio/timbgpio.c b/drivers/gpio/timbgpio.c index ac4d0f0..ddd0531 100644 --- a/drivers/gpio/timbgpio.c +++ b/drivers/gpio/timbgpio.c @@ -131,6 +131,7 @@ static int timbgpio_irq_type(unsigned irq, unsigned trigger) unsigned long flags; u32 lvr, flr, bflr = 0; u32 ver; + int ret = 0; if (offset < 0 || offset > tgpio->gpio.ngpio) return -EINVAL; @@ -154,8 +155,10 @@ static int timbgpio_irq_type(unsigned irq, unsigned trigger) } if ((trigger & IRQ_TYPE_EDGE_BOTH) == IRQ_TYPE_EDGE_BOTH) { - if (ver < 3) - return -EINVAL; + if (ver < 3) { + ret = -EINVAL; + goto out; + } else { flr |= 1 << offset; bflr |= 1 << offset; @@ -175,9 +178,10 @@ static int timbgpio_irq_type(unsigned irq, unsigned trigger) iowrite32(bflr, tgpio->membase + TGPIO_BFLR); iowrite32(1 << offset, tgpio->membase + TGPIO_ICR); - spin_unlock_irqrestore(&tgpio->lock, flags); - return 0; +out: + spin_unlock_irqrestore(&tgpio->lock, flags); + return ret; } static void timbgpio_irq(unsigned int irq, struct irq_desc *desc) -- cgit v0.10.2 From cdeebaddb7f97298c73df4d43d31e164fc8aa0e4 Mon Sep 17 00:00:00 2001 From: Madhusudhan Chikkature Date: Tue, 6 Apr 2010 14:34:49 -0700 Subject: omap hsmmc: fix a bug in card remove scenario The reset of data lines when the card is removed from the cage results in a failure.The failure is seen if the card is removed from the cage when TC is pending after a CMD with data received CC.The reset logic leaves the controller in a state where niether a TC is received nor DTO. The rest code can be safely removed here since it is taken care in the IRQ handler. Signed-off-by: Madhusudhan Chikkature Cc: Adrian Hunter Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/mmc/host/omap_hsmmc.c b/drivers/mmc/host/omap_hsmmc.c index 83f0aff..e9caf69 100644 --- a/drivers/mmc/host/omap_hsmmc.c +++ b/drivers/mmc/host/omap_hsmmc.c @@ -1179,15 +1179,10 @@ static void omap_hsmmc_detect(struct work_struct *work) carddetect = -ENOSYS; } - if (carddetect) { + if (carddetect) mmc_detect_change(host->mmc, (HZ * 200) / 1000); - } else { - mmc_host_enable(host->mmc); - omap_hsmmc_reset_controller_fsm(host, SRD); - mmc_host_lazy_disable(host->mmc); - + else mmc_detect_change(host->mmc, (HZ * 50) / 1000); - } } /* -- cgit v0.10.2 From c6eb69acfdfe258b4a3f69fa5ced8928bbebdba9 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 6 Apr 2010 14:34:50 -0700 Subject: mxser: spin_lock() => spin_lock_irq() This should be spin_lock_irq() to match the spin_unlock_irq(). Originally it was a lock_kernel() but we switched everything to spin_lock_irq() last November. [akpm@linux-foundation.org: fix the MOXA_ASPP_MON case too (per Jiri)] Signed-off-by: Dan Carpenter Cc: Jiri Slaby Cc: Alan Cox Cc: Greg KH Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/char/mxser.c b/drivers/char/mxser.c index 95c9f54..4702305 100644 --- a/drivers/char/mxser.c +++ b/drivers/char/mxser.c @@ -1768,7 +1768,7 @@ static int mxser_ioctl(struct tty_struct *tty, struct file *file, int len, lsr; len = mxser_chars_in_buffer(tty); - spin_lock(&info->slock); + spin_lock_irq(&info->slock); lsr = inb(info->ioaddr + UART_LSR) & UART_LSR_THRE; spin_unlock_irq(&info->slock); len += (lsr ? 0 : 1); @@ -1778,12 +1778,12 @@ static int mxser_ioctl(struct tty_struct *tty, struct file *file, case MOXA_ASPP_MON: { int mcr, status; - spin_lock(&info->slock); + spin_lock_irq(&info->slock); status = mxser_get_msr(info->ioaddr, 1, tty->index); mxser_check_modem_status(tty, info, status); mcr = inb(info->ioaddr + UART_MCR); - spin_unlock(&info->slock); + spin_unlock_irq(&info->slock); if (mcr & MOXA_MUST_MCR_XON_FLAG) info->mon_data.hold_reason &= ~NPPI_NOTIFY_XOFFHOLD; -- cgit v0.10.2 From 829f46af39d50a43e260adaa9e7bbdff74a9f696 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 6 Apr 2010 14:34:50 -0700 Subject: cciss: unlock on error path We take the spin_lock again in fail_all_cmds() so we need to unlock here. Signed-off-by: Dan Carpenter Acked-by: Steve Cameron Cc: Jens Axboe Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index 9e3af30..eb5ff05 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -3341,6 +3341,7 @@ static irqreturn_t do_cciss_intr(int irq, void *dev_id) printk(KERN_WARNING "cciss: controller cciss%d failed, stopping.\n", h->ctlr); + spin_unlock_irqrestore(CCISS_LOCK(h->ctlr), flags); fail_all_cmds(h->ctlr); return IRQ_HANDLED; } -- cgit v0.10.2 From 975f8c5653acba461229e671202113da69b87be1 Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Tue, 6 Apr 2010 14:34:51 -0700 Subject: drivers/thermal/thermal_sys.c: fix 'key f70f4b50 not in .data' in thermal_sys Initialize sysfs attributes before device_create_file call. Addresses https://bugzilla.kernel.org/show_bug.cgi?id=15548 Signed-off-by: Wolfram Sang Signed-off-by: Sergey Senozhatsky Cc: "Eric W. Biederman" Cc: Greg KH Cc: Zhang Rui Cc: Len Brown Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/thermal/thermal_sys.c b/drivers/thermal/thermal_sys.c index 9b6297f..13c72c6 100644 --- a/drivers/thermal/thermal_sys.c +++ b/drivers/thermal/thermal_sys.c @@ -506,6 +506,7 @@ thermal_add_hwmon_sysfs(struct thermal_zone_device *tz) tz->temp_input.attr.attr.name = tz->temp_input.name; tz->temp_input.attr.attr.mode = 0444; tz->temp_input.attr.show = temp_input_show; + sysfs_attr_init(&tz->temp_input.attr.attr); result = device_create_file(hwmon->device, &tz->temp_input.attr); if (result) goto unregister_hwmon_device; @@ -518,6 +519,7 @@ thermal_add_hwmon_sysfs(struct thermal_zone_device *tz) tz->temp_crit.attr.attr.name = tz->temp_crit.name; tz->temp_crit.attr.attr.mode = 0444; tz->temp_crit.attr.show = temp_crit_show; + sysfs_attr_init(&tz->temp_crit.attr.attr); result = device_create_file(hwmon->device, &tz->temp_crit.attr); if (result) @@ -726,6 +728,7 @@ int thermal_zone_bind_cooling_device(struct thermal_zone_device *tz, goto release_idr; sprintf(dev->attr_name, "cdev%d_trip_point", dev->id); + sysfs_attr_init(&dev->attr.attr); dev->attr.attr.name = dev->attr_name; dev->attr.attr.mode = 0444; dev->attr.show = thermal_cooling_device_trip_point_show; -- cgit v0.10.2 From 12765517d9dbb477a2432375938f1eb5bdbcb532 Mon Sep 17 00:00:00 2001 From: Wolfram Sang Date: Tue, 6 Apr 2010 14:34:52 -0700 Subject: device_attributes: add sysfs_attr_init() for dynamic attributes Made necessary by 6992f5334995af474c2b58d010d08bc597f0f2fe ("sysfs: Use one lockdep class per sysfs attribute"). Prevents further "key xxx not in .data" bug-reports. Although some attributes could probably be converted to static ones, this is left for people having hardware to test. Found by this semantic patch: @ init @ type T; identifier A; @@ T { ... struct device_attribute A; ... }; @ main extends init @ expression E; statement S; identifier err; T *name; @@ ... when != sysfs_attr_init(&name->A.attr); ( + sysfs_attr_init(&name->A.attr); if (device_create_file(E, &name->A)) S | + sysfs_attr_init(&name->A.attr); err = device_create_file(E, &name->A); ) While reviewing, I put the initialization to apropriate places. Signed-off-by: Wolfram Sang Cc: Eric W. Biederman Cc: Greg KH Cc: Benjamin Herrenschmidt Cc: Grant Likely Cc: Mike Isely Cc: Mauro Carvalho Chehab Cc: Sujith Thomas Cc: Matthew Garrett Cc: Len Brown Cc: Krzysztof Helt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/macintosh/windfarm_core.c b/drivers/macintosh/windfarm_core.c index c092354..ce88979 100644 --- a/drivers/macintosh/windfarm_core.c +++ b/drivers/macintosh/windfarm_core.c @@ -210,6 +210,7 @@ int wf_register_control(struct wf_control *new_ct) kref_init(&new_ct->ref); list_add(&new_ct->link, &wf_controls); + sysfs_attr_init(&new_ct->attr.attr); new_ct->attr.attr.name = new_ct->name; new_ct->attr.attr.mode = 0644; new_ct->attr.show = wf_show_control; diff --git a/drivers/media/video/pvrusb2/pvrusb2-sysfs.c b/drivers/media/video/pvrusb2/pvrusb2-sysfs.c index 6c23456..71f5056 100644 --- a/drivers/media/video/pvrusb2/pvrusb2-sysfs.c +++ b/drivers/media/video/pvrusb2/pvrusb2-sysfs.c @@ -423,10 +423,12 @@ static void pvr2_sysfs_add_debugifc(struct pvr2_sysfs *sfp) dip = kzalloc(sizeof(*dip),GFP_KERNEL); if (!dip) return; + sysfs_attr_init(&dip->attr_debugcmd.attr); dip->attr_debugcmd.attr.name = "debugcmd"; dip->attr_debugcmd.attr.mode = S_IRUGO|S_IWUSR|S_IWGRP; dip->attr_debugcmd.show = debugcmd_show; dip->attr_debugcmd.store = debugcmd_store; + sysfs_attr_init(&dip->attr_debuginfo.attr); dip->attr_debuginfo.attr.name = "debuginfo"; dip->attr_debuginfo.attr.mode = S_IRUGO; dip->attr_debuginfo.show = debuginfo_show; @@ -644,6 +646,7 @@ static void class_dev_create(struct pvr2_sysfs *sfp, return; } + sysfs_attr_init(&sfp->attr_v4l_minor_number.attr); sfp->attr_v4l_minor_number.attr.name = "v4l_minor_number"; sfp->attr_v4l_minor_number.attr.mode = S_IRUGO; sfp->attr_v4l_minor_number.show = v4l_minor_number_show; @@ -658,6 +661,7 @@ static void class_dev_create(struct pvr2_sysfs *sfp, sfp->v4l_minor_number_created_ok = !0; } + sysfs_attr_init(&sfp->attr_v4l_radio_minor_number.attr); sfp->attr_v4l_radio_minor_number.attr.name = "v4l_radio_minor_number"; sfp->attr_v4l_radio_minor_number.attr.mode = S_IRUGO; sfp->attr_v4l_radio_minor_number.show = v4l_radio_minor_number_show; @@ -672,6 +676,7 @@ static void class_dev_create(struct pvr2_sysfs *sfp, sfp->v4l_radio_minor_number_created_ok = !0; } + sysfs_attr_init(&sfp->attr_unit_number.attr); sfp->attr_unit_number.attr.name = "unit_number"; sfp->attr_unit_number.attr.mode = S_IRUGO; sfp->attr_unit_number.show = unit_number_show; @@ -685,6 +690,7 @@ static void class_dev_create(struct pvr2_sysfs *sfp, sfp->unit_number_created_ok = !0; } + sysfs_attr_init(&sfp->attr_bus_info.attr); sfp->attr_bus_info.attr.name = "bus_info_str"; sfp->attr_bus_info.attr.mode = S_IRUGO; sfp->attr_bus_info.show = bus_info_show; @@ -699,6 +705,7 @@ static void class_dev_create(struct pvr2_sysfs *sfp, sfp->bus_info_created_ok = !0; } + sysfs_attr_init(&sfp->attr_hdw_name.attr); sfp->attr_hdw_name.attr.name = "device_hardware_type"; sfp->attr_hdw_name.attr.mode = S_IRUGO; sfp->attr_hdw_name.show = hdw_name_show; @@ -713,6 +720,7 @@ static void class_dev_create(struct pvr2_sysfs *sfp, sfp->hdw_name_created_ok = !0; } + sysfs_attr_init(&sfp->attr_hdw_desc.attr); sfp->attr_hdw_desc.attr.name = "device_hardware_description"; sfp->attr_hdw_desc.attr.mode = S_IRUGO; sfp->attr_hdw_desc.show = hdw_desc_show; diff --git a/drivers/platform/x86/intel_menlow.c b/drivers/platform/x86/intel_menlow.c index 1190bad..2f795ce 100644 --- a/drivers/platform/x86/intel_menlow.c +++ b/drivers/platform/x86/intel_menlow.c @@ -397,6 +397,7 @@ static int intel_menlow_add_one_attribute(char *name, int mode, void *show, if (!attr) return -ENOMEM; + sysfs_attr_init(&attr->attr.attr); /* That is consistent naming :D */ attr->attr.attr.name = name; attr->attr.attr.mode = mode; attr->attr.show = show; diff --git a/drivers/video/fsl-diu-fb.c b/drivers/video/fsl-diu-fb.c index 4637bcb..994358a 100644 --- a/drivers/video/fsl-diu-fb.c +++ b/drivers/video/fsl-diu-fb.c @@ -1536,6 +1536,7 @@ static int __devinit fsl_diu_probe(struct of_device *ofdev, goto error; } + sysfs_attr_init(&machine_data->dev_attr.attr); machine_data->dev_attr.attr.name = "monitor"; machine_data->dev_attr.attr.mode = S_IRUGO|S_IWUSR; machine_data->dev_attr.show = show_monitor; -- cgit v0.10.2 From 70655c06bd3f25111312d63985888112aed15ac5 Mon Sep 17 00:00:00 2001 From: Wu Fengguang Date: Tue, 6 Apr 2010 14:34:53 -0700 Subject: readahead: fix NULL filp dereference btrfs relocate_file_extent_cluster() calls us with NULL filp: [ 4005.426805] BUG: unable to handle kernel NULL pointer dereference at 00000021 [ 4005.426818] IP: [] page_cache_sync_readahead+0x18/0x3e Signed-off-by: Wu Fengguang Cc: Yan Zheng Reported-by: Kirill A. Shutemov Tested-by: Kirill A. Shutemov Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/readahead.c b/mm/readahead.c index 999b54b..dfa9a1a 100644 --- a/mm/readahead.c +++ b/mm/readahead.c @@ -503,7 +503,7 @@ void page_cache_sync_readahead(struct address_space *mapping, return; /* be dumb */ - if (filp->f_mode & FMODE_RANDOM) { + if (filp && (filp->f_mode & FMODE_RANDOM)) { force_page_cache_readahead(mapping, filp, offset, req_size); return; } -- cgit v0.10.2 From 6e191f7bb083544dc4fa3879ff81caf97c65d197 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Tue, 6 Apr 2010 14:34:55 -0700 Subject: devmem: handle class_create() failure I hit this when we had a bug in IDR for a few days. Basically sysfs would fail to create new inodes since it uses an IDR and therefore class_create would fail. While we are unlikely to see this fail we may as well handle it instead of oopsing. Signed-off-by: Anton Blanchard Reviewed-by: Wu Fengguang Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/char/mem.c b/drivers/char/mem.c index 1f3215a..47c8452 100644 --- a/drivers/char/mem.c +++ b/drivers/char/mem.c @@ -908,6 +908,9 @@ static int __init chr_dev_init(void) printk("unable to get major %d for memory devs\n", MEM_MAJOR); mem_class = class_create(THIS_MODULE, "mem"); + if (IS_ERR(mem_class)) + return PTR_ERR(mem_class); + mem_class->devnode = mem_devnode; for (minor = 1; minor < ARRAY_SIZE(devlist); minor++) { if (!devlist[minor].name) -- cgit v0.10.2 From d6da1a5abc2bf3a06a5bda08e0f6833409234666 Mon Sep 17 00:00:00 2001 From: KOSAKI Motohiro Date: Tue, 6 Apr 2010 14:34:56 -0700 Subject: mm: revert "vmscan: get_scan_ratio() cleanup" Shaohua Li reported his tmpfs streaming I/O test can lead to make oom. The test uses a 6G tmpfs in a system with 3G memory. In the tmpfs, there are 6 copies of kernel source and the test does kbuild for each copy. His investigation shows the test has a lot of rotated anon pages and quite few file pages, so get_scan_ratio calculates percent[0] (i.e. scanning percent for anon) to be zero. Actually the percent[0] shoule be a big value, but our calculation round it to zero. Although before commit 84b18490 ("vmscan: get_scan_ratio() cleanup") , we have the same problem too. But the old logic can rescue percent[0]==0 case only when priority==0. It had hided the real issue. I didn't think merely streaming io can makes percent[0]==0 && priority==0 situation. but I was wrong. So, definitely we have to fix such tmpfs streaming io issue. but anyway I revert the regression commit at first. This reverts commit 84b18490d1f1bc7ed5095c929f78bc002eb70f26. Signed-off-by: KOSAKI Motohiro Reported-by: Shaohua Li Cc: Rik van Riel Cc: KAMEZAWA Hiroyuki Cc: Minchan Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/vmscan.c b/mm/vmscan.c index e0e5f15..3ff3311 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -1535,13 +1535,6 @@ static void get_scan_ratio(struct zone *zone, struct scan_control *sc, unsigned long ap, fp; struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc); - /* If we have no swap space, do not bother scanning anon pages. */ - if (!sc->may_swap || (nr_swap_pages <= 0)) { - percent[0] = 0; - percent[1] = 100; - return; - } - anon = zone_nr_lru_pages(zone, sc, LRU_ACTIVE_ANON) + zone_nr_lru_pages(zone, sc, LRU_INACTIVE_ANON); file = zone_nr_lru_pages(zone, sc, LRU_ACTIVE_FILE) + @@ -1639,20 +1632,22 @@ static void shrink_zone(int priority, struct zone *zone, unsigned long nr_reclaimed = sc->nr_reclaimed; unsigned long nr_to_reclaim = sc->nr_to_reclaim; struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc); + int noswap = 0; - get_scan_ratio(zone, sc, percent); + /* If we have no swap space, do not bother scanning anon pages. */ + if (!sc->may_swap || (nr_swap_pages <= 0)) { + noswap = 1; + percent[0] = 0; + percent[1] = 100; + } else + get_scan_ratio(zone, sc, percent); for_each_evictable_lru(l) { int file = is_file_lru(l); unsigned long scan; - if (percent[file] == 0) { - nr[l] = 0; - continue; - } - scan = zone_nr_lru_pages(zone, sc, l); - if (priority) { + if (priority || noswap) { scan >>= priority; scan = (scan * percent[file]) / 100; } -- cgit v0.10.2 From a71dc148c2674bbb5988ea924702137993b3305a Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Tue, 6 Apr 2010 14:34:57 -0700 Subject: mb862xxfb: fix acceleration module license mb862xxfb_accel built as a separate module, but it does not have a MODULE_LICENSE, so it taints the kernel. Add a MODULE_LICENSE to it (same as mb862xxfb license). mb862xxfb_accel: module license 'unspecified' taints kernel. Or should mb862xxfb_accel be built into the mb862xxfb binary file instead? Signed-off-by: Randy Dunlap Acked-by: Alexander Shishkin Cc: Valentin Sitdikov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/video/mb862xx/mb862xxfb_accel.c b/drivers/video/mb862xx/mb862xxfb_accel.c index 8414249..b0bf4d8 100644 --- a/drivers/video/mb862xx/mb862xxfb_accel.c +++ b/drivers/video/mb862xx/mb862xxfb_accel.c @@ -16,6 +16,7 @@ #include #include #include +#include #include #include #if defined(CONFIG_OF) @@ -330,3 +331,5 @@ void mb862xxfb_init_accel(struct fb_info *info, int xres) info->fix.accel = 0xff; /*FIXME: add right define */ } EXPORT_SYMBOL(mb862xxfb_init_accel); + +MODULE_LICENSE("GPL v2"); -- cgit v0.10.2 From da258016293f5e82b36db67ac3db3931a4fbbc4d Mon Sep 17 00:00:00 2001 From: Alexander Shishkin Date: Tue, 6 Apr 2010 14:34:57 -0700 Subject: mb862xxfb: update Valentin's email address Since Valentin's email address @siemens.com is no longer valid, it's time to change it to the one that actually works so that I don't have to manually forward patches against mb862xx to him every time. Signed-off-by: Alexander Shishkin Acked-by: Valentin Sitdikov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/video/mb862xx/mb862xxfb_accel.c b/drivers/video/mb862xx/mb862xxfb_accel.c index b0bf4d8..fe92eed 100644 --- a/drivers/video/mb862xx/mb862xxfb_accel.c +++ b/drivers/video/mb862xx/mb862xxfb_accel.c @@ -4,7 +4,7 @@ * Fujitsu Carmine/Coral-P(A)/Lime framebuffer driver acceleration support * * (C) 2007 Alexander Shishkin - * (C) 2009 Valentin Sitdikov + * (C) 2009 Valentin Sitdikov * (C) 2009 Siemens AG * * This program is free software; you can redistribute it and/or modify -- cgit v0.10.2 From 55ab3a1ff843e3f0e24d2da44e71bffa5d853010 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Tue, 6 Apr 2010 14:34:58 -0700 Subject: raw: fsync method is now required Commit 148f948ba877f4d3cdef036b1ff6d9f68986706a (vfs: Introduce new helpers for syncing after writing to O_SYNC file or IS_SYNC inode) broke the raw driver. We now call through generic_file_aio_write -> generic_write_sync -> vfs_fsync_range. vfs_fsync_range has: if (!fop || !fop->fsync) { ret = -EINVAL; goto out; } But drivers/char/raw.c doesn't set an fsync method. We have two options: fix it or remove the raw driver completely. I'm happy to do either, the fact this has been broken for so long suggests it is rarely used. The patch below adds an fsync method to the raw driver. My knowledge of the block layer is pretty sketchy so this could do with a once over. If we instead decide to remove the raw driver, this patch might still be useful as a backport to 2.6.33 and 2.6.32. Signed-off-by: Anton Blanchard Reviewed-by: Jan Kara Cc: Christoph Hellwig Cc: Alexander Viro Cc: Jens Axboe Reviewed-by: Jeff Moyer Tested-by: Jeff Moyer Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/char/raw.c b/drivers/char/raw.c index d331c59..f5a3231 100644 --- a/drivers/char/raw.c +++ b/drivers/char/raw.c @@ -248,6 +248,7 @@ static const struct file_operations raw_fops = { .aio_read = generic_file_aio_read, .write = do_sync_write, .aio_write = blkdev_aio_write, + .fsync = block_fsync, .open = raw_open, .release= raw_release, .ioctl = raw_ioctl, diff --git a/fs/block_dev.c b/fs/block_dev.c index d11d028..8db62b2 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -404,7 +404,7 @@ static loff_t block_llseek(struct file *file, loff_t offset, int origin) * NULL first argument is nfsd_sync_dir() and that's not a directory. */ -static int block_fsync(struct file *filp, struct dentry *dentry, int datasync) +int block_fsync(struct file *filp, struct dentry *dentry, int datasync) { struct block_device *bdev = I_BDEV(filp->f_mapping->host); int error; @@ -418,6 +418,7 @@ static int block_fsync(struct file *filp, struct dentry *dentry, int datasync) error = 0; return error; } +EXPORT_SYMBOL(block_fsync); /* * pseudo-fs diff --git a/include/linux/fs.h b/include/linux/fs.h index 10b8ded..5d9c7e2 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2212,6 +2212,7 @@ extern int generic_segment_checks(const struct iovec *iov, /* fs/block_dev.c */ extern ssize_t blkdev_aio_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos); +extern int block_fsync(struct file *filp, struct dentry *dentry, int datasync); /* fs/splice.c */ extern ssize_t generic_file_splice_read(struct file *, loff_t *, -- cgit v0.10.2 From b1dd3b2843b3b73b7fc2ee47d96310cd1c051371 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Tue, 6 Apr 2010 14:35:00 -0700 Subject: vfs: rename block_fsync() to blkdev_fsync() Requested by hch, for consistency now it is exported. Cc: Alexander Viro Cc: Anton Blanchard Cc: Christoph Hellwig Cc: Jan Kara Cc: Jeff Moyer Cc: Jens Axboe Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/char/raw.c b/drivers/char/raw.c index f5a3231..8756ab0 100644 --- a/drivers/char/raw.c +++ b/drivers/char/raw.c @@ -248,7 +248,7 @@ static const struct file_operations raw_fops = { .aio_read = generic_file_aio_read, .write = do_sync_write, .aio_write = blkdev_aio_write, - .fsync = block_fsync, + .fsync = blkdev_fsync, .open = raw_open, .release= raw_release, .ioctl = raw_ioctl, diff --git a/fs/block_dev.c b/fs/block_dev.c index 8db62b2..2a6d019 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -404,7 +404,7 @@ static loff_t block_llseek(struct file *file, loff_t offset, int origin) * NULL first argument is nfsd_sync_dir() and that's not a directory. */ -int block_fsync(struct file *filp, struct dentry *dentry, int datasync) +int blkdev_fsync(struct file *filp, struct dentry *dentry, int datasync) { struct block_device *bdev = I_BDEV(filp->f_mapping->host); int error; @@ -418,7 +418,7 @@ int block_fsync(struct file *filp, struct dentry *dentry, int datasync) error = 0; return error; } -EXPORT_SYMBOL(block_fsync); +EXPORT_SYMBOL(blkdev_fsync); /* * pseudo-fs @@ -1482,7 +1482,7 @@ const struct file_operations def_blk_fops = { .aio_read = generic_file_aio_read, .aio_write = blkdev_aio_write, .mmap = generic_file_mmap, - .fsync = block_fsync, + .fsync = blkdev_fsync, .unlocked_ioctl = block_ioctl, #ifdef CONFIG_COMPAT .compat_ioctl = compat_blkdev_ioctl, diff --git a/include/linux/fs.h b/include/linux/fs.h index 5d9c7e2..39d57bc 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2212,7 +2212,7 @@ extern int generic_segment_checks(const struct iovec *iov, /* fs/block_dev.c */ extern ssize_t blkdev_aio_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos); -extern int block_fsync(struct file *filp, struct dentry *dentry, int datasync); +extern int blkdev_fsync(struct file *filp, struct dentry *dentry, int datasync); /* fs/splice.c */ extern ssize_t generic_file_splice_read(struct file *, loff_t *, -- cgit v0.10.2 From ee5d2acd5ca1534f40e06d4f0d41a940b17beb54 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 6 Apr 2010 14:35:01 -0700 Subject: /dev/mem: allow rewinding commit dcefafb6 ("/dev/mem: dont allow seek to last page") inadvertently disabled rewinding on /dev/mem. This broke x86info for example. Signed-off-by: Eric Dumazet Acked-by: Wu Fengguang Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/char/mem.c b/drivers/char/mem.c index 47c8452..9cdbadc 100644 --- a/drivers/char/mem.c +++ b/drivers/char/mem.c @@ -710,11 +710,6 @@ static loff_t memory_lseek(struct file *file, loff_t offset, int orig) switch (orig) { case SEEK_CUR: offset += file->f_pos; - if ((unsigned long long)offset < - (unsigned long long)file->f_pos) { - ret = -EOVERFLOW; - break; - } case SEEK_SET: /* to avoid userland mistaking f_pos=-9 as -EBADF=-9 */ if ((unsigned long long)offset >= ~0xFFFULL) { -- cgit v0.10.2 From 2a7268abc48c8009b284bd88605d14fcb65772ec Mon Sep 17 00:00:00 2001 From: Yong Zhang Date: Tue, 6 Apr 2010 14:35:01 -0700 Subject: ratelimit: annotate ___ratelimit() To prevent from wrongly using the return value. [akpm@linux-foundation.org: fix spello] Signed-off-by: Yong Zhang Cc: Ingo Molnar Cc: Dave Young Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/lib/ratelimit.c b/lib/ratelimit.c index 09f5ce1..b4cd24f 100644 --- a/lib/ratelimit.c +++ b/lib/ratelimit.c @@ -16,9 +16,14 @@ /* * __ratelimit - rate limiting * @rs: ratelimit_state data + * @func: name of calling function * - * This enforces a rate limit: not more than @rs->ratelimit_burst callbacks - * in every @rs->ratelimit_jiffies + * This enforces a rate limit: not more than @rs->burst callbacks + * in every @rs->interval + * + * RETURNS: + * 0 means callbacks will be suppressed. + * 1 means go ahead and do it. */ int ___ratelimit(struct ratelimit_state *rs, const char *func) { -- cgit v0.10.2 From bb1dc0bacb8ddd7ba6a5906c678a5a5a110cf695 Mon Sep 17 00:00:00 2001 From: Yong Zhang Date: Tue, 6 Apr 2010 14:35:02 -0700 Subject: kernel.h: fix wrong usage of __ratelimit() When __ratelimit() returns 1 this means that we can go ahead. Signed-off-by: Yong Zhang Cc: Ingo Molnar Cc: Joe Perches Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/kernel.h b/include/linux/kernel.h index 7f07074..9365227 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -426,7 +426,7 @@ static inline char *pack_hex_byte(char *buf, u8 byte) .burst = DEFAULT_RATELIMIT_BURST, \ }; \ \ - if (!__ratelimit(&_rs)) \ + if (__ratelimit(&_rs)) \ printk(fmt, ##__VA_ARGS__); \ }) #else -- cgit v0.10.2 From 57119c34e53bbb8d244c3ff1335ef5145768538f Mon Sep 17 00:00:00 2001 From: Yong Zhang Date: Tue, 6 Apr 2010 14:35:03 -0700 Subject: ratelimit: fix the return value when __ratelimit() fails to acquire the lock The log of commit edaac8e3167501cda336231d00611bf59c164346 ("ratelimit: Fix/allow use in atomic contexts"), indicates that we want to suppress the callback when the trylock fails. Signed-off-by: Yong Zhang Cc: Ingo Molnar Cc: Christian Borntraeger Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/lib/ratelimit.c b/lib/ratelimit.c index b4cd24f..027a03f 100644 --- a/lib/ratelimit.c +++ b/lib/ratelimit.c @@ -40,7 +40,7 @@ int ___ratelimit(struct ratelimit_state *rs, const char *func) * the entity that is holding the lock already: */ if (!spin_trylock_irqsave(&rs->lock, flags)) - return 1; + return 0; if (!rs->begin) rs->begin = jiffies; -- cgit v0.10.2 From 116354d177ba2da37e91cf884e3d11e67f825efd Mon Sep 17 00:00:00 2001 From: Naoya Horiguchi Date: Tue, 6 Apr 2010 14:35:04 -0700 Subject: pagemap: fix pfn calculation for hugepage When we look into pagemap using page-types with option -p, the value of pfn for hugepages looks wrong (see below.) This is because pte was evaluated only once for one vma although it should be updated for each hugepage. This patch fixes it. $ page-types -p 3277 -Nl -b huge voffset offset len flags 7f21e8a00 11e400 1 ___U___________H_G________________ 7f21e8a01 11e401 1ff ________________TG________________ ^^^ 7f21e8c00 11e400 1 ___U___________H_G________________ 7f21e8c01 11e401 1ff ________________TG________________ ^^^ One hugepage contains 1 head page and 511 tail pages in x86_64 and each two lines represent each hugepage. Voffset and offset mean virtual address and physical address in the page unit, respectively. The different hugepages should not have the same offset value. With this patch applied: $ page-types -p 3386 -Nl -b huge voffset offset len flags 7fec7a600 112c00 1 ___UD__________H_G________________ 7fec7a601 112c01 1ff ________________TG________________ ^^^ 7fec7a800 113200 1 ___UD__________H_G________________ 7fec7a801 113201 1ff ________________TG________________ ^^^ OK More info: - This patch modifies walk_page_range()'s hugepage walker. But the change only affects pagemap_read(), which is the only caller of hugepage callback. - Without this patch, hugetlb_entry() callback is called per vma, that doesn't match the natural expectation from its name. - With this patch, hugetlb_entry() is called per hugepte entry and the callback can become much simpler. Signed-off-by: Naoya Horiguchi Signed-off-by: KAMEZAWA Hiroyuki Acked-by: Matt Mackall Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index a05a669..0705534 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -662,31 +662,18 @@ static u64 huge_pte_to_pagemap_entry(pte_t pte, int offset) return pme; } -static int pagemap_hugetlb_range(pte_t *pte, unsigned long addr, - unsigned long end, struct mm_walk *walk) +/* This function walks within one hugetlb entry in the single call */ +static int pagemap_hugetlb_range(pte_t *pte, unsigned long hmask, + unsigned long addr, unsigned long end, + struct mm_walk *walk) { - struct vm_area_struct *vma; struct pagemapread *pm = walk->private; - struct hstate *hs = NULL; int err = 0; + u64 pfn; - vma = find_vma(walk->mm, addr); - if (vma) - hs = hstate_vma(vma); for (; addr != end; addr += PAGE_SIZE) { - u64 pfn = PM_NOT_PRESENT; - - if (vma && (addr >= vma->vm_end)) { - vma = find_vma(walk->mm, addr); - if (vma) - hs = hstate_vma(vma); - } - - if (vma && (vma->vm_start <= addr) && is_vm_hugetlb_page(vma)) { - /* calculate pfn of the "raw" page in the hugepage. */ - int offset = (addr & ~huge_page_mask(hs)) >> PAGE_SHIFT; - pfn = huge_pte_to_pagemap_entry(*pte, offset); - } + int offset = (addr & ~hmask) >> PAGE_SHIFT; + pfn = huge_pte_to_pagemap_entry(*pte, offset); err = add_to_pagemap(addr, pfn, pm); if (err) return err; diff --git a/include/linux/mm.h b/include/linux/mm.h index e70f21b..462acaf 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -783,8 +783,8 @@ struct mm_walk { int (*pmd_entry)(pmd_t *, unsigned long, unsigned long, struct mm_walk *); int (*pte_entry)(pte_t *, unsigned long, unsigned long, struct mm_walk *); int (*pte_hole)(unsigned long, unsigned long, struct mm_walk *); - int (*hugetlb_entry)(pte_t *, unsigned long, unsigned long, - struct mm_walk *); + int (*hugetlb_entry)(pte_t *, unsigned long, + unsigned long, unsigned long, struct mm_walk *); struct mm_struct *mm; void *private; }; diff --git a/mm/pagewalk.c b/mm/pagewalk.c index 7b47a57..8b1a2ce 100644 --- a/mm/pagewalk.c +++ b/mm/pagewalk.c @@ -80,6 +80,37 @@ static int walk_pud_range(pgd_t *pgd, unsigned long addr, unsigned long end, return err; } +#ifdef CONFIG_HUGETLB_PAGE +static unsigned long hugetlb_entry_end(struct hstate *h, unsigned long addr, + unsigned long end) +{ + unsigned long boundary = (addr & huge_page_mask(h)) + huge_page_size(h); + return boundary < end ? boundary : end; +} + +static int walk_hugetlb_range(struct vm_area_struct *vma, + unsigned long addr, unsigned long end, + struct mm_walk *walk) +{ + struct hstate *h = hstate_vma(vma); + unsigned long next; + unsigned long hmask = huge_page_mask(h); + pte_t *pte; + int err = 0; + + do { + next = hugetlb_entry_end(h, addr, end); + pte = huge_pte_offset(walk->mm, addr & hmask); + if (pte && walk->hugetlb_entry) + err = walk->hugetlb_entry(pte, hmask, addr, next, walk); + if (err) + return err; + } while (addr = next, addr != end); + + return 0; +} +#endif + /** * walk_page_range - walk a memory map's page tables with a callback * @mm: memory map to walk @@ -128,20 +159,16 @@ int walk_page_range(unsigned long addr, unsigned long end, vma = find_vma(walk->mm, addr); #ifdef CONFIG_HUGETLB_PAGE if (vma && is_vm_hugetlb_page(vma)) { - pte_t *pte; - struct hstate *hs; - if (vma->vm_end < next) next = vma->vm_end; - hs = hstate_vma(vma); - pte = huge_pte_offset(walk->mm, - addr & huge_page_mask(hs)); - if (pte && !huge_pte_none(huge_ptep_get(pte)) - && walk->hugetlb_entry) - err = walk->hugetlb_entry(pte, addr, - next, walk); + /* + * Hugepage is very tightly coupled with vma, so + * walk through hugetlb entries within a given vma. + */ + err = walk_hugetlb_range(vma, addr, next, walk); if (err) break; + pgd = pgd_offset(walk->mm, next); continue; } #endif -- cgit v0.10.2 From 8725d5416213a145ccc9c236dbd26830ba409e00 Mon Sep 17 00:00:00 2001 From: KAMEZAWA Hiroyuki Date: Tue, 6 Apr 2010 14:35:05 -0700 Subject: memcg: fix race in file_mapped accounting Presently, memcg's FILE_MAPPED accounting has following race with move_account (happens at rmdir()). increment page->mapcount (rmap.c) mem_cgroup_update_file_mapped() move_account() lock_page_cgroup() check page_mapped() if page_mapped(page)>1 { FILE_MAPPED -1 from old memcg FILE_MAPPED +1 to old memcg } ..... overwrite pc->mem_cgroup unlock_page_cgroup() lock_page_cgroup() FILE_MAPPED + 1 to pc->mem_cgroup unlock_page_cgroup() Then, old memcg (-1 file mapped) new memcg (+2 file mapped) This happens because move_account see page_mapped() which is not guarded by lock_page_cgroup(). This patch adds FILE_MAPPED flag to page_cgroup and move account information based on it. Now, all checks are synchronous with lock_page_cgroup(). Signed-off-by: KAMEZAWA Hiroyuki Reviewed-by: Balbir Singh Reviewed-by: Daisuke Nishimura Cc: Andrea Righi Cc: Andrea Arcangeli Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/page_cgroup.h b/include/linux/page_cgroup.h index 30b0813..aef22ae 100644 --- a/include/linux/page_cgroup.h +++ b/include/linux/page_cgroup.h @@ -39,6 +39,7 @@ enum { PCG_CACHE, /* charged as cache */ PCG_USED, /* this object is in use. */ PCG_ACCT_LRU, /* page has been accounted for */ + PCG_FILE_MAPPED, /* page is accounted as "mapped" */ }; #define TESTPCGFLAG(uname, lname) \ @@ -73,6 +74,11 @@ CLEARPCGFLAG(AcctLRU, ACCT_LRU) TESTPCGFLAG(AcctLRU, ACCT_LRU) TESTCLEARPCGFLAG(AcctLRU, ACCT_LRU) + +SETPCGFLAG(FileMapped, FILE_MAPPED) +CLEARPCGFLAG(FileMapped, FILE_MAPPED) +TESTPCGFLAG(FileMapped, FILE_MAPPED) + static inline int page_cgroup_nid(struct page_cgroup *pc) { return page_to_nid(pc->page); diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 9ed760d..f4ede99 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -1359,16 +1359,19 @@ void mem_cgroup_update_file_mapped(struct page *page, int val) lock_page_cgroup(pc); mem = pc->mem_cgroup; - if (!mem) - goto done; - - if (!PageCgroupUsed(pc)) + if (!mem || !PageCgroupUsed(pc)) goto done; /* * Preemption is already disabled. We can use __this_cpu_xxx */ - __this_cpu_add(mem->stat->count[MEM_CGROUP_STAT_FILE_MAPPED], val); + if (val > 0) { + __this_cpu_inc(mem->stat->count[MEM_CGROUP_STAT_FILE_MAPPED]); + SetPageCgroupFileMapped(pc); + } else { + __this_cpu_dec(mem->stat->count[MEM_CGROUP_STAT_FILE_MAPPED]); + ClearPageCgroupFileMapped(pc); + } done: unlock_page_cgroup(pc); @@ -1801,16 +1804,13 @@ static void __mem_cgroup_commit_charge(struct mem_cgroup *mem, static void __mem_cgroup_move_account(struct page_cgroup *pc, struct mem_cgroup *from, struct mem_cgroup *to, bool uncharge) { - struct page *page; - VM_BUG_ON(from == to); VM_BUG_ON(PageLRU(pc->page)); VM_BUG_ON(!PageCgroupLocked(pc)); VM_BUG_ON(!PageCgroupUsed(pc)); VM_BUG_ON(pc->mem_cgroup != from); - page = pc->page; - if (page_mapped(page) && !PageAnon(page)) { + if (PageCgroupFileMapped(pc)) { /* Update mapped_file data for mem_cgroup */ preempt_disable(); __this_cpu_dec(from->stat->count[MEM_CGROUP_STAT_FILE_MAPPED]); -- cgit v0.10.2 From c783a29efcbc90a90ccfec8956c7ff0854301c34 Mon Sep 17 00:00:00 2001 From: Vladimir Zapolskiy Date: Tue, 6 Apr 2010 14:35:07 -0700 Subject: rtc-mxc: multiple fixes in rtc-mxc probe method On exit paths in mxc_rtc_probe() method some resources are not freed correctly. This patch fixes: * unrequested memory region containing imx RTC registers * iounmap() isn't called on exit_free_pdata branch * clock get rate is called for freed clock source * clock isn't disabled on exit_put_clk branch To simplify the fix managed device resources are used. [akpm@linux-foundation.org: coding-style fixes] Signed-off-by: Vladimir Zapolskiy Cc: Alessandro Zummo Cc: Daniel Mack Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rtc/rtc-mxc.c b/drivers/rtc/rtc-mxc.c index c77f6f7..d71fe61 100644 --- a/drivers/rtc/rtc-mxc.c +++ b/drivers/rtc/rtc-mxc.c @@ -384,21 +384,26 @@ static int __init mxc_rtc_probe(struct platform_device *pdev) struct rtc_device *rtc; struct rtc_plat_data *pdata = NULL; u32 reg; - int ret, rate; + unsigned long rate; + int ret; res = platform_get_resource(pdev, IORESOURCE_MEM, 0); if (!res) return -ENODEV; - pdata = kzalloc(sizeof(*pdata), GFP_KERNEL); + pdata = devm_kzalloc(&pdev->dev, sizeof(*pdata), GFP_KERNEL); if (!pdata) return -ENOMEM; - pdata->ioaddr = ioremap(res->start, resource_size(res)); + if (!devm_request_mem_region(&pdev->dev, res->start, + resource_size(res), pdev->name)) + return -EBUSY; + + pdata->ioaddr = devm_ioremap(&pdev->dev, res->start, + resource_size(res)); clk = clk_get(&pdev->dev, "ckil"); if (IS_ERR(clk)) { - iounmap(pdata->ioaddr); ret = PTR_ERR(clk); goto exit_free_pdata; } @@ -413,8 +418,7 @@ static int __init mxc_rtc_probe(struct platform_device *pdev) else if (rate == 38400) reg = RTC_INPUT_CLK_38400HZ; else { - dev_err(&pdev->dev, "rtc clock is not valid (%lu)\n", - clk_get_rate(clk)); + dev_err(&pdev->dev, "rtc clock is not valid (%lu)\n", rate); ret = -EINVAL; goto exit_free_pdata; } @@ -450,8 +454,8 @@ static int __init mxc_rtc_probe(struct platform_device *pdev) pdata->irq = platform_get_irq(pdev, 0); if (pdata->irq >= 0 && - request_irq(pdata->irq, mxc_rtc_interrupt, IRQF_SHARED, - pdev->name, pdev) < 0) { + devm_request_irq(&pdev->dev, pdata->irq, mxc_rtc_interrupt, + IRQF_SHARED, pdev->name, pdev) < 0) { dev_warn(&pdev->dev, "interrupt not available.\n"); pdata->irq = -1; } @@ -459,10 +463,10 @@ static int __init mxc_rtc_probe(struct platform_device *pdev) return 0; exit_put_clk: + clk_disable(pdata->clk); clk_put(pdata->clk); exit_free_pdata: - kfree(pdata); return ret; } @@ -473,12 +477,8 @@ static int __exit mxc_rtc_remove(struct platform_device *pdev) rtc_device_unregister(pdata->rtc); - if (pdata->irq >= 0) - free_irq(pdata->irq, pdev); - clk_disable(pdata->clk); clk_put(pdata->clk); - kfree(pdata); platform_set_drvdata(pdev, NULL); return 0; -- cgit v0.10.2 From ea56f411ec2d4d8689eb7f3bcd24e860acf87c47 Mon Sep 17 00:00:00 2001 From: David Howells Date: Tue, 6 Apr 2010 14:35:08 -0700 Subject: frv: hide uncached_access() when pgprot_noncached is not #defined Hide uncached_access() when pgprot_noncached is not #defined. This prevents the following warning: CC drivers/char/mem.o drivers/char/mem.c:229: warning: 'uncached_access' defined but not used Repairs d7d4d849b4e3acc405ec222884936800ffb26d48 ("drivers/char/mem.c: cleanups"). Signed-off-by: David Howells Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/char/mem.c b/drivers/char/mem.c index 9cdbadc..f54dab8 100644 --- a/drivers/char/mem.c +++ b/drivers/char/mem.c @@ -225,6 +225,7 @@ int __weak phys_mem_access_prot_allowed(struct file *file, * outside of main memory. * */ +#ifdef pgprot_noncached static int uncached_access(struct file *file, unsigned long addr) { #if defined(CONFIG_IA64) @@ -251,6 +252,7 @@ static int uncached_access(struct file *file, unsigned long addr) return addr >= __pa(high_memory); #endif } +#endif static pgprot_t phys_mem_access_prot(struct file *file, unsigned long pfn, unsigned long size, pgprot_t vma_prot) -- cgit v0.10.2 From 08dc179b9b00cfbed006ae6f5d0e471ef79dc69c Mon Sep 17 00:00:00 2001 From: David Howells Date: Tue, 6 Apr 2010 14:35:09 -0700 Subject: frv: fix kernel/user segment handling in NOMMU mode In NOMMU mode, the FRV segment handling is broken because KERNEL_DS == USER_DS. This causes tests of the following sort: /* don't pin down non-user-based iovecs */ if (segment_eq(get_fs(), KERNEL_DS)) return NULL; to malfunction. To fix this, make USER_DS the top of RAM instead of the top of the non-IO address space, and make KERNEL_DS one more than the top of the non-IO address space. Also get rid of FRV's __addr_ok() as nothing uses it. Signed-off-by: David Howells Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/frv/include/asm/segment.h b/arch/frv/include/asm/segment.h index e3616a6..a2320a4 100644 --- a/arch/frv/include/asm/segment.h +++ b/arch/frv/include/asm/segment.h @@ -21,12 +21,12 @@ typedef struct { #define MAKE_MM_SEG(s) ((mm_segment_t) { (s) }) -#define KERNEL_DS MAKE_MM_SEG(0xdfffffffUL) - #ifdef CONFIG_MMU #define USER_DS MAKE_MM_SEG(TASK_SIZE - 1) +#define KERNEL_DS MAKE_MM_SEG(0xdfffffffUL) #else -#define USER_DS KERNEL_DS +#define USER_DS MAKE_MM_SEG(memory_end) +#define KERNEL_DS MAKE_MM_SEG(0xe0000000UL) #endif #define get_ds() (KERNEL_DS) diff --git a/arch/frv/include/asm/uaccess.h b/arch/frv/include/asm/uaccess.h index 53650c9..0b67ec5 100644 --- a/arch/frv/include/asm/uaccess.h +++ b/arch/frv/include/asm/uaccess.h @@ -27,8 +27,6 @@ #define VERIFY_READ 0 #define VERIFY_WRITE 1 -#define __addr_ok(addr) ((unsigned long)(addr) < get_addr_limit()) - /* * check that a range of addresses falls within the current address limit */ -- cgit v0.10.2 From cc4fc29e59c386919a7674e203be7822dc968dc0 Mon Sep 17 00:00:00 2001 From: David Howells Date: Tue, 6 Apr 2010 14:35:09 -0700 Subject: fs-cache: order the debugfs stats correctly Order the debugfs statistics correctly. The values displayed through a seq_printf() statement should be in the same order as the names in the format string. In the 'Lookups' line, objects created ('crt=') and lookups timed out ('tmo=') have their values transposed. Signed-off-by: David Howells Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/fscache/stats.c b/fs/fscache/stats.c index 46435f3a..4765190 100644 --- a/fs/fscache/stats.c +++ b/fs/fscache/stats.c @@ -165,8 +165,8 @@ static int fscache_stats_show(struct seq_file *m, void *v) atomic_read(&fscache_n_object_lookups), atomic_read(&fscache_n_object_lookups_negative), atomic_read(&fscache_n_object_lookups_positive), - atomic_read(&fscache_n_object_lookups_timed_out), - atomic_read(&fscache_n_object_created)); + atomic_read(&fscache_n_object_created), + atomic_read(&fscache_n_object_lookups_timed_out)); seq_printf(m, "Updates: n=%u nul=%u run=%u\n", atomic_read(&fscache_n_updates), -- cgit v0.10.2 From 75392dd3041b8fb7ee4cef6bd9b71154ae34f4e1 Mon Sep 17 00:00:00 2001 From: Sergei Shtylyov Date: Fri, 26 Mar 2010 17:56:58 +0300 Subject: davinci: fix compile warning: : #include MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This hushes the following warning: arch/arm/mach-davinci/include/mach/da8xx.h:104: warning: ‘struct platform_device’ declared inside parameter list arch/arm/mach-davinci/include/mach/da8xx.h:104: warning: its scope is only this definition or declaration, which is probably not what you want Signed-off-by: Sergei Shtylyov Signed-off-by: Kevin Hilman diff --git a/arch/arm/mach-davinci/include/mach/da8xx.h b/arch/arm/mach-davinci/include/mach/da8xx.h index b87a6ba..03acfd3 100644 --- a/arch/arm/mach-davinci/include/mach/da8xx.h +++ b/arch/arm/mach-davinci/include/mach/da8xx.h @@ -3,7 +3,7 @@ * * Author: Mark A. Greer * - * 2007, 2009 (c) MontaVista Software, Inc. This file is licensed under + * 2007, 2009-2010 (c) MontaVista Software, Inc. This file is licensed under * the terms of the GNU General Public License version 2. This program * is licensed "as is" without any warranty of any kind, whether express * or implied. @@ -13,7 +13,9 @@ #include