From 01a36b6758e723f919420ef20cea5fca1fc06e2b Mon Sep 17 00:00:00 2001 From: Gang He Date: Tue, 2 Aug 2016 14:02:07 -0700 Subject: ocfs2: ensure that dlm lockspace is created by kernel module We encountered a bug from the customer, the user did a fsck.ocfs2 on the file system and exited unusually, the lockspace (with LVB size = 32) was left in the kernel space, next, the user mounted this file system, the kernel module did not create a new lockspace (LVB size = 64) via calling dlm_new_lockspace() function in mounting stage, just used the existing lockspace, created by the user space tool, this would lead the user was not able to mount this file system from the other nodes, with the error message like: dlm: 032F5......: config mismatch: 64,0 nodeid 177127961: 32,0 (mount.ocfs2,26981,46):ocfs2_dlm_init:2995 ERROR: status = -71 ocfs2_mount_volume:1881 ERROR: status = -71 ocfs2_fill_super:1236 ERROR: status = -71 The user found it very difficult to find the root cause, then, we brought out this patch to relieve such problem. First, we add one more flag in calling dlm_new_lockspace() function, to make sure the lockspace is created by kernel module itself, and this change will not affect the backward compatibility. Second, the obvious error message is reported in the kernel log, let the user be more easy to find the root cause. This patch will be used to insure the dlm lockspace is created by kernel module when mounting a ocfs2 file system. There are two ways to create a lockspace, from user space and kernel space, but the same name lockspaces probably have different lvblen lengths/flags. To avoid this mix using, we add one more flag DLM_LSFL_NEWEXCL, it will make sure the dlm lockspace is created by kernel module when mounting. Secondly, if a user space program (ocfs2-tools) is running on a file system, the user tries to mount this file system in the cluster, DLM module will return a -EEXIST or -EPROTO errno, we should give the user a obvious error message, then, the user can let that user space tool exit before mounting the file system again. Link: http://lkml.kernel.org/r/1463731940-13044-2-git-send-email-ghe@suse.com Signed-off-by: Gang He Reviewed-by: Goldwyn Rodrigues Reviewed-by: Mark Fasheh Cc: Joel Becker Cc: Junxiao Bi Cc: Joseph Qi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/ocfs2/stack_user.c b/fs/ocfs2/stack_user.c index ced70c8..c9e828e 100644 --- a/fs/ocfs2/stack_user.c +++ b/fs/ocfs2/stack_user.c @@ -1007,10 +1007,17 @@ static int user_cluster_connect(struct ocfs2_cluster_connection *conn) lc->oc_type = NO_CONTROLD; rc = dlm_new_lockspace(conn->cc_name, conn->cc_cluster_name, - DLM_LSFL_FS, DLM_LVB_LEN, + DLM_LSFL_FS | DLM_LSFL_NEWEXCL, DLM_LVB_LEN, &ocfs2_ls_ops, conn, &ops_rv, &fsdlm); - if (rc) + if (rc) { + if (rc == -EEXIST || rc == -EPROTO) + printk(KERN_ERR "ocfs2: Unable to create the " + "lockspace %s (%d), because a ocfs2-tools " + "program is running on this file system " + "with the same name lockspace\n", + conn->cc_name, rc); goto out; + } if (ops_rv == -EOPNOTSUPP) { lc->oc_type = WITH_CONTROLD; -- cgit v0.10.2 From 2070ad1aebfff2c26190188844c38e55d2df2ae2 Mon Sep 17 00:00:00 2001 From: Eric Ren Date: Tue, 2 Aug 2016 14:02:10 -0700 Subject: ocfs2: retry on ENOSPC if sufficient space in truncate log The testcase "mmaptruncate" in ocfs2 test suite always fails with ENOSPC error on small volume (say less than 10G). This testcase repeatedly performs "extend" and "truncate" on a file. Continuously, it truncates the file to 1/2 of the size, and then extends to 100% of the size. The main bitmap will quickly run out of space because the "truncate" code prevent truncate log from being flushed by ocfs2_schedule_truncate_log_flush(osb, 1), while truncate log may have cached lots of clusters. So retry to allocate after flushing truncate log when ENOSPC is returned. And we cannot reuse the deleted blocks before the transaction committed. Fortunately, we already have a function to do this - ocfs2_try_to_free_truncate_log(). Just need to remove the "static" modifier and put it into the right place. The "unlock"/"lock" code isn't elegant, but there seems to be no better option. [zren@suse.com: locking fix] Link: http://lkml.kernel.org/r/1468031546-4797-1-git-send-email-zren@suse.com Link: http://lkml.kernel.org/r/1466586469-5541-1-git-send-email-zren@suse.com Signed-off-by: Eric Ren Reviewed-by: Gang He Reviewed-by: Joseph Qi Reviewed-by: Mark Fasheh Cc: Joel Becker Cc: Junxiao Bi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index 460c0ce..7dabbc3 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -6106,6 +6106,43 @@ void ocfs2_schedule_truncate_log_flush(struct ocfs2_super *osb, } } +/* + * Try to flush truncate logs if we can free enough clusters from it. + * As for return value, "< 0" means error, "0" no space and "1" means + * we have freed enough spaces and let the caller try to allocate again. + */ +int ocfs2_try_to_free_truncate_log(struct ocfs2_super *osb, + unsigned int needed) +{ + tid_t target; + int ret = 0; + unsigned int truncated_clusters; + + inode_lock(osb->osb_tl_inode); + truncated_clusters = osb->truncated_clusters; + inode_unlock(osb->osb_tl_inode); + + /* + * Check whether we can succeed in allocating if we free + * the truncate log. + */ + if (truncated_clusters < needed) + goto out; + + ret = ocfs2_flush_truncate_log(osb); + if (ret) { + mlog_errno(ret); + goto out; + } + + if (jbd2_journal_start_commit(osb->journal->j_journal, &target)) { + jbd2_log_wait_commit(osb->journal->j_journal, target); + ret = 1; + } +out: + return ret; +} + static int ocfs2_get_truncate_log_info(struct ocfs2_super *osb, int slot_num, struct inode **tl_inode, diff --git a/fs/ocfs2/alloc.h b/fs/ocfs2/alloc.h index f3dc1b0..4a5152e 100644 --- a/fs/ocfs2/alloc.h +++ b/fs/ocfs2/alloc.h @@ -188,6 +188,8 @@ int ocfs2_truncate_log_append(struct ocfs2_super *osb, u64 start_blk, unsigned int num_clusters); int __ocfs2_flush_truncate_log(struct ocfs2_super *osb); +int ocfs2_try_to_free_truncate_log(struct ocfs2_super *osb, + unsigned int needed); /* * Process local structure which describes the block unlinks done diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index af2adfc..98d3654 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -1645,43 +1645,6 @@ static int ocfs2_zero_tail(struct inode *inode, struct buffer_head *di_bh, return ret; } -/* - * Try to flush truncate logs if we can free enough clusters from it. - * As for return value, "< 0" means error, "0" no space and "1" means - * we have freed enough spaces and let the caller try to allocate again. - */ -static int ocfs2_try_to_free_truncate_log(struct ocfs2_super *osb, - unsigned int needed) -{ - tid_t target; - int ret = 0; - unsigned int truncated_clusters; - - inode_lock(osb->osb_tl_inode); - truncated_clusters = osb->truncated_clusters; - inode_unlock(osb->osb_tl_inode); - - /* - * Check whether we can succeed in allocating if we free - * the truncate log. - */ - if (truncated_clusters < needed) - goto out; - - ret = ocfs2_flush_truncate_log(osb); - if (ret) { - mlog_errno(ret); - goto out; - } - - if (jbd2_journal_start_commit(osb->journal->j_journal, &target)) { - jbd2_log_wait_commit(osb->journal->j_journal, target); - ret = 1; - } -out: - return ret; -} - int ocfs2_write_begin_nolock(struct address_space *mapping, loff_t pos, unsigned len, ocfs2_write_type_t type, struct page **pagep, void **fsdata, diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c index 2f19aee..ea47120 100644 --- a/fs/ocfs2/suballoc.c +++ b/fs/ocfs2/suballoc.c @@ -1164,7 +1164,8 @@ static int ocfs2_reserve_clusters_with_limit(struct ocfs2_super *osb, int flags, struct ocfs2_alloc_context **ac) { - int status; + int status, ret = 0; + int retried = 0; *ac = kzalloc(sizeof(struct ocfs2_alloc_context), GFP_KERNEL); if (!(*ac)) { @@ -1189,7 +1190,24 @@ static int ocfs2_reserve_clusters_with_limit(struct ocfs2_super *osb, } if (status == -ENOSPC) { +retry: status = ocfs2_reserve_cluster_bitmap_bits(osb, *ac); + /* Retry if there is sufficient space cached in truncate log */ + if (status == -ENOSPC && !retried) { + retried = 1; + ocfs2_inode_unlock((*ac)->ac_inode, 1); + inode_unlock((*ac)->ac_inode); + + ret = ocfs2_try_to_free_truncate_log(osb, bits_wanted); + if (ret == 1) + goto retry; + + if (ret < 0) + mlog_errno(ret); + + inode_lock((*ac)->ac_inode); + ocfs2_inode_lock((*ac)->ac_inode, NULL, 1); + } if (status < 0) { if (status != -ENOSPC) mlog_errno(status); -- cgit v0.10.2 From 86b652b93adb57d8fed8edd532ed2eb8a791950d Mon Sep 17 00:00:00 2001 From: piaojun Date: Tue, 2 Aug 2016 14:02:13 -0700 Subject: ocfs2/dlm: disable BUG_ON when DLM_LOCK_RES_DROPPING_REF is cleared before dlm_deref_lockres_done_handler We found a BUG situation in which DLM_LOCK_RES_DROPPING_REF is cleared unexpected that described below. To solve the bug, we disable the BUG_ON and purge lockres in dlm_do_local_recovery_cleanup. Node 1 Node 2(master) dlm_purge_lockres dlm_deref_lockres_handler DLM_LOCK_RES_SETREF_INPROG is set response DLM_DEREF_RESPONSE_INPROG receive DLM_DEREF_RESPONSE_INPROG stop puring in dlm_purge_lockres and wait for DLM_DEREF_RESPONSE_DONE dispatch dlm_deref_lockres_worker response DLM_DEREF_RESPONSE_DONE receive DLM_DEREF_RESPONSE_DONE and prepare to purge lockres Node 2 goes down find Node2 down and do local clean up for Node2: dlm_do_local_recovery_cleanup -> clear DLM_LOCK_RES_DROPPING_REF when purging lockres, BUG_ON happens because DLM_LOCK_RES_DROPPING_REF is clear: dlm_deref_lockres_done_handler ->BUG_ON(!(res->state & DLM_LOCK_RES_DROPPING_REF)); [akpm@linux-foundation.org: fix duplicated write to `ret'] Fixes: 60d663cb5273 ("ocfs2/dlm: add DEREF_DONE message") Link: http://lkml.kernel.org/r/57845055.9080702@huawei.com Signed-off-by: Jun Piao Reviewed-by: Joseph Qi Reviewed-by: Jiufei Xue Reviewed-by: Mark Fasheh Cc: Joel Becker Cc: Junxiao Bi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c index 13719d3..525dc06 100644 --- a/fs/ocfs2/dlm/dlmmaster.c +++ b/fs/ocfs2/dlm/dlmmaster.c @@ -2416,7 +2416,17 @@ int dlm_deref_lockres_done_handler(struct o2net_msg *msg, u32 len, void *data, } spin_lock(&res->spinlock); - BUG_ON(!(res->state & DLM_LOCK_RES_DROPPING_REF)); + if (!(res->state & DLM_LOCK_RES_DROPPING_REF)) { + spin_unlock(&res->spinlock); + spin_unlock(&dlm->spinlock); + mlog(ML_NOTICE, "%s:%.*s: node %u sends deref done " + "but it is already derefed!\n", dlm->name, + res->lockname.len, res->lockname.name, node); + dlm_lockres_put(res); + ret = 0; + goto done; + } + if (!list_empty(&res->purge)) { mlog(0, "%s: Removing res %.*s from purgelist\n", dlm->name, res->lockname.len, res->lockname.name); @@ -2456,7 +2466,6 @@ int dlm_deref_lockres_done_handler(struct o2net_msg *msg, u32 len, void *data, spin_unlock(&dlm->spinlock); ret = 0; - done: dlm_put(dlm); return ret; -- cgit v0.10.2 From 309e91911daede6adde0364f489e69909c3f6894 Mon Sep 17 00:00:00 2001 From: piaojun Date: Tue, 2 Aug 2016 14:02:16 -0700 Subject: ocfs2/dlm: solve a BUG when deref failed in dlm_drop_lockres_ref We found a BUG situation that lockres is migrated during deref described below. To solve the BUG, we could purge lockres directly when other node says I did not have a ref. Additionally, we'd better purge lockres if master goes down, as no one will response deref done. Node 1 Node 2(old master) Node3(new master) dlm_purge_lockres send deref to N2 leave domain migrate lockres to N3 finish migration send do assert master to N1 receive do assert msg form N3, but can not find lockres because DROPPING_REF is set, so the owner is still N2. receive deref from N1 and response -EINVAL because lockres is migrated BUG when receive -EINVAL in dlm_drop_lockres_ref Fixes: 842b90b62461d ("ocfs2/dlm: return in progress if master can not clear the refmap bit right now") Link: http://lkml.kernel.org/r/57845103.3070406@huawei.com Signed-off-by: Jun Piao Reviewed-by: Joseph Qi Reviewed-by: Jiufei Xue Reviewed-by: Mark Fasheh Cc: Joel Becker Cc: Junxiao Bi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c index 525dc06..553d220 100644 --- a/fs/ocfs2/dlm/dlmmaster.c +++ b/fs/ocfs2/dlm/dlmmaster.c @@ -2276,9 +2276,12 @@ int dlm_drop_lockres_ref(struct dlm_ctxt *dlm, struct dlm_lock_resource *res) mlog(ML_ERROR, "%s: res %.*s, DEREF to node %u got %d\n", dlm->name, namelen, lockname, res->owner, r); dlm_print_one_lock_resource(res); - BUG(); - } - return ret ? ret : r; + if (r == -ENOMEM) + BUG(); + } else + ret = r; + + return ret; } int dlm_deref_lockres_handler(struct o2net_msg *msg, u32 len, void *data, diff --git a/fs/ocfs2/dlm/dlmthread.c b/fs/ocfs2/dlm/dlmthread.c index 68d239b..ce39722 100644 --- a/fs/ocfs2/dlm/dlmthread.c +++ b/fs/ocfs2/dlm/dlmthread.c @@ -175,6 +175,15 @@ static void dlm_purge_lockres(struct dlm_ctxt *dlm, res->lockname.len, res->lockname.name, master); if (!master) { + if (res->state & DLM_LOCK_RES_DROPPING_REF) { + mlog(ML_NOTICE, "%s: res %.*s already in " + "DLM_LOCK_RES_DROPPING_REF state\n", + dlm->name, res->lockname.len, + res->lockname.name); + spin_unlock(&res->spinlock); + return; + } + res->state |= DLM_LOCK_RES_DROPPING_REF; /* drop spinlock... retake below */ spin_unlock(&res->spinlock); @@ -203,8 +212,8 @@ static void dlm_purge_lockres(struct dlm_ctxt *dlm, dlm->purge_count--; } - if (!master && ret != 0) { - mlog(0, "%s: deref %.*s in progress or master goes down\n", + if (!master && ret == DLM_DEREF_RESPONSE_INPROG) { + mlog(0, "%s: deref %.*s in progress\n", dlm->name, res->lockname.len, res->lockname.name); spin_unlock(&res->spinlock); return; -- cgit v0.10.2 From ee8f7fcbe638b07e8d1c3dc98e8be35e56306d05 Mon Sep 17 00:00:00 2001 From: piaojun Date: Tue, 2 Aug 2016 14:02:19 -0700 Subject: ocfs2/dlm: continue to purge recovery lockres when recovery master goes down We found a dlm-blocked situation caused by continuous breakdown of recovery masters described below. To solve this problem, we should purge recovery lock once detecting recovery master goes down. N3 N2 N1(reco master) go down pick up recovery lock and begin recoverying for N2 go down pick up recovery lock failed, then purge it: dlm_purge_lockres ->DROPPING_REF is set send deref to N1 failed, recovery lock is not purged find N1 go down, begin recoverying for N1, but blocked in dlm_do_recovery as DROPPING_REF is set: dlm_do_recovery ->dlm_pick_recovery_master ->dlmlock ->dlm_get_lock_resource ->__dlm_wait_on_lockres_flags(tmpres, DLM_LOCK_RES_DROPPING_REF); Fixes: 8c0343968163 ("ocfs2/dlm: clear DROPPING_REF flag when the master goes down") Link: http://lkml.kernel.org/r/578453AF.8030404@huawei.com Signed-off-by: Jun Piao Reviewed-by: Joseph Qi Reviewed-by: Jiufei Xue Reviewed-by: Mark Fasheh Cc: Joel Becker Cc: Junxiao Bi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/ocfs2/dlm/dlmcommon.h b/fs/ocfs2/dlm/dlmcommon.h index 8107d0d..e9f3705 100644 --- a/fs/ocfs2/dlm/dlmcommon.h +++ b/fs/ocfs2/dlm/dlmcommon.h @@ -1004,6 +1004,8 @@ int dlm_finalize_reco_handler(struct o2net_msg *msg, u32 len, void *data, int dlm_do_master_requery(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, u8 nodenum, u8 *real_master); +void __dlm_do_purge_lockres(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res); int dlm_dispatch_assert_master(struct dlm_ctxt *dlm, struct dlm_lock_resource *res, diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c index 553d220..6ea06f8 100644 --- a/fs/ocfs2/dlm/dlmmaster.c +++ b/fs/ocfs2/dlm/dlmmaster.c @@ -2425,51 +2425,20 @@ int dlm_deref_lockres_done_handler(struct o2net_msg *msg, u32 len, void *data, mlog(ML_NOTICE, "%s:%.*s: node %u sends deref done " "but it is already derefed!\n", dlm->name, res->lockname.len, res->lockname.name, node); - dlm_lockres_put(res); ret = 0; goto done; } - if (!list_empty(&res->purge)) { - mlog(0, "%s: Removing res %.*s from purgelist\n", - dlm->name, res->lockname.len, res->lockname.name); - list_del_init(&res->purge); - dlm_lockres_put(res); - dlm->purge_count--; - } - - if (!__dlm_lockres_unused(res)) { - mlog(ML_ERROR, "%s: res %.*s in use after deref\n", - dlm->name, res->lockname.len, res->lockname.name); - __dlm_print_one_lock_resource(res); - BUG(); - } - - __dlm_unhash_lockres(dlm, res); - - spin_lock(&dlm->track_lock); - if (!list_empty(&res->tracking)) - list_del_init(&res->tracking); - else { - mlog(ML_ERROR, "%s: Resource %.*s not on the Tracking list\n", - dlm->name, res->lockname.len, res->lockname.name); - __dlm_print_one_lock_resource(res); - } - spin_unlock(&dlm->track_lock); - - /* lockres is not in the hash now. drop the flag and wake up - * any processes waiting in dlm_get_lock_resource. - */ - res->state &= ~DLM_LOCK_RES_DROPPING_REF; + __dlm_do_purge_lockres(dlm, res); spin_unlock(&res->spinlock); wake_up(&res->wq); - dlm_lockres_put(res); - spin_unlock(&dlm->spinlock); ret = 0; done: + if (res) + dlm_lockres_put(res); dlm_put(dlm); return ret; } diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c index f6b3138..dd5cb8b 100644 --- a/fs/ocfs2/dlm/dlmrecovery.c +++ b/fs/ocfs2/dlm/dlmrecovery.c @@ -2343,6 +2343,7 @@ static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node) struct dlm_lock_resource *res; int i; struct hlist_head *bucket; + struct hlist_node *tmp; struct dlm_lock *lock; @@ -2365,7 +2366,7 @@ static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node) */ for (i = 0; i < DLM_HASH_BUCKETS; i++) { bucket = dlm_lockres_hash(dlm, i); - hlist_for_each_entry(res, bucket, hash_node) { + hlist_for_each_entry_safe(res, tmp, bucket, hash_node) { /* always prune any $RECOVERY entries for dead nodes, * otherwise hangs can occur during later recovery */ if (dlm_is_recovery_lock(res->lockname.name, @@ -2386,8 +2387,17 @@ static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node) break; } } - dlm_lockres_clear_refmap_bit(dlm, res, - dead_node); + + if ((res->owner == dead_node) && + (res->state & DLM_LOCK_RES_DROPPING_REF)) { + dlm_lockres_get(res); + __dlm_do_purge_lockres(dlm, res); + spin_unlock(&res->spinlock); + wake_up(&res->wq); + dlm_lockres_put(res); + continue; + } else if (res->owner == dlm->node_num) + dlm_lockres_clear_refmap_bit(dlm, res, dead_node); spin_unlock(&res->spinlock); continue; } @@ -2398,14 +2408,17 @@ static void dlm_do_local_recovery_cleanup(struct dlm_ctxt *dlm, u8 dead_node) if (res->state & DLM_LOCK_RES_DROPPING_REF) { mlog(0, "%s:%.*s: owned by " "dead node %u, this node was " - "dropping its ref when it died. " - "continue, dropping the flag.\n", + "dropping its ref when master died. " + "continue, purging the lockres.\n", dlm->name, res->lockname.len, res->lockname.name, dead_node); + dlm_lockres_get(res); + __dlm_do_purge_lockres(dlm, res); + spin_unlock(&res->spinlock); + wake_up(&res->wq); + dlm_lockres_put(res); + continue; } - res->state &= ~DLM_LOCK_RES_DROPPING_REF; - dlm_move_lockres_to_recovery_list(dlm, - res); } else if (res->owner == dlm->node_num) { dlm_free_dead_locks(dlm, res, dead_node); __dlm_lockres_calc_usage(dlm, res); diff --git a/fs/ocfs2/dlm/dlmthread.c b/fs/ocfs2/dlm/dlmthread.c index ce39722..838a06d 100644 --- a/fs/ocfs2/dlm/dlmthread.c +++ b/fs/ocfs2/dlm/dlmthread.c @@ -160,6 +160,52 @@ void dlm_lockres_calc_usage(struct dlm_ctxt *dlm, spin_unlock(&dlm->spinlock); } +/* + * Do the real purge work: + * unhash the lockres, and + * clear flag DLM_LOCK_RES_DROPPING_REF. + * It requires dlm and lockres spinlock to be taken. + */ +void __dlm_do_purge_lockres(struct dlm_ctxt *dlm, + struct dlm_lock_resource *res) +{ + assert_spin_locked(&dlm->spinlock); + assert_spin_locked(&res->spinlock); + + if (!list_empty(&res->purge)) { + mlog(0, "%s: Removing res %.*s from purgelist\n", + dlm->name, res->lockname.len, res->lockname.name); + list_del_init(&res->purge); + dlm_lockres_put(res); + dlm->purge_count--; + } + + if (!__dlm_lockres_unused(res)) { + mlog(ML_ERROR, "%s: res %.*s in use after deref\n", + dlm->name, res->lockname.len, res->lockname.name); + __dlm_print_one_lock_resource(res); + BUG(); + } + + __dlm_unhash_lockres(dlm, res); + + spin_lock(&dlm->track_lock); + if (!list_empty(&res->tracking)) + list_del_init(&res->tracking); + else { + mlog(ML_ERROR, "%s: Resource %.*s not on the Tracking list\n", + dlm->name, res->lockname.len, res->lockname.name); + __dlm_print_one_lock_resource(res); + } + spin_unlock(&dlm->track_lock); + + /* + * lockres is not in the hash now. drop the flag and wake up + * any processes waiting in dlm_get_lock_resource. + */ + res->state &= ~DLM_LOCK_RES_DROPPING_REF; +} + static void dlm_purge_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res) { @@ -176,10 +222,8 @@ static void dlm_purge_lockres(struct dlm_ctxt *dlm, if (!master) { if (res->state & DLM_LOCK_RES_DROPPING_REF) { - mlog(ML_NOTICE, "%s: res %.*s already in " - "DLM_LOCK_RES_DROPPING_REF state\n", - dlm->name, res->lockname.len, - res->lockname.name); + mlog(ML_NOTICE, "%s: res %.*s already in DLM_LOCK_RES_DROPPING_REF state\n", + dlm->name, res->lockname.len, res->lockname.name); spin_unlock(&res->spinlock); return; } -- cgit v0.10.2 From c5f88bd29ab42d5d1e77085b5f69d5c6da20324e Mon Sep 17 00:00:00 2001 From: Vegard Nossum Date: Tue, 2 Aug 2016 14:02:22 -0700 Subject: mm: fail prefaulting if page table allocation fails I ran into this: BUG: sleeping function called from invalid context at mm/page_alloc.c:3784 in_atomic(): 0, irqs_disabled(): 0, pid: 1434, name: trinity-c1 2 locks held by trinity-c1/1434: #0: (&mm->mmap_sem){......}, at: [] __do_page_fault+0x1ce/0x8f0 #1: (rcu_read_lock){......}, at: [] filemap_map_pages+0xd6/0xdd0 CPU: 0 PID: 1434 Comm: trinity-c1 Not tainted 4.7.0+ #58 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Ubuntu-1.8.2-1ubuntu1 04/01/2014 Call Trace: dump_stack+0x65/0x84 panic+0x185/0x2dd ___might_sleep+0x51c/0x600 __might_sleep+0x90/0x1a0 __alloc_pages_nodemask+0x5b1/0x2160 alloc_pages_current+0xcc/0x370 pte_alloc_one+0x12/0x90 __pte_alloc+0x1d/0x200 alloc_set_pte+0xe3e/0x14a0 filemap_map_pages+0x42b/0xdd0 handle_mm_fault+0x17d5/0x28b0 __do_page_fault+0x310/0x8f0 trace_do_page_fault+0x18d/0x310 do_async_page_fault+0x27/0xa0 async_page_fault+0x28/0x30 The important bits from the above is that filemap_map_pages() is calling into the page allocator while holding rcu_read_lock (sleeping is not allowed inside RCU read-side critical sections). According to Kirill Shutemov, the prefaulting code in do_fault_around() is supposed to take care of this, but missing error handling means that the allocation failure can go unnoticed. We don't need to return VM_FAULT_OOM (or any other error) here, since we can just let the normal fault path try again. Fixes: 7267ec008b5c ("mm: postpone page table allocation until we have page to map") Link: http://lkml.kernel.org/r/1469708107-11868-1-git-send-email-vegard.nossum@oracle.com Signed-off-by: Vegard Nossum Acked-by: Kirill A. Shutemov Cc: "Hillf Danton" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/memory.c b/mm/memory.c index 4425b60..0400483 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -3133,6 +3133,8 @@ static int do_fault_around(struct fault_env *fe, pgoff_t start_pgoff) if (pmd_none(*fe->pmd)) { fe->prealloc_pte = pte_alloc_one(fe->vma->vm_mm, fe->address); + if (!fe->prealloc_pte) + goto out; smp_wmb(); /* See comment in __pte_alloc() */ } -- cgit v0.10.2 From 1a8018fb4c6976559c3f04bcf760822381be501d Mon Sep 17 00:00:00 2001 From: Minchan Kim Date: Tue, 2 Aug 2016 14:02:25 -0700 Subject: mm: move swap-in anonymous page into active list Every swap-in anonymous page starts from inactive lru list's head. It should be activated unconditionally when VM decide to reclaim because page table entry for the page always usually has marked accessed bit. Thus, their window size for getting a new referece is 2 * NR_inactive + NR_active while others is NR_inactive + NR_active. It's not fair that it has more chance to be referenced compared to other newly allocated page which starts from active lru list's head. Johannes: : The page can still have a valid copy on the swap device, so prefering to : reclaim that page over a fresh one could make sense. But as you point : out, having it start inactive instead of active actually ends up giving it : *more* LRU time, and that seems to be without justification. Rik: : The reason newly read in swap cache pages start on the inactive list is : that we do some amount of read-around, and do not know which pages will : get used. : : However, immediately activating the ones that DO get used, like your patch : does, is the right thing to do. Link: http://lkml.kernel.org/r/1469762740-17860-1-git-send-email-minchan@kernel.org Signed-off-by: Minchan Kim Acked-by: Johannes Weiner Acked-by: Rik van Riel Cc: Nadav Amit Cc: Mel Gorman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/memory.c b/mm/memory.c index 0400483..83be99d 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -2642,6 +2642,7 @@ int do_swap_page(struct fault_env *fe, pte_t orig_pte) if (page == swapcache) { do_page_add_anon_rmap(page, vma, fe->address, exclusive); mem_cgroup_commit_charge(page, memcg, true, false); + activate_page(page); } else { /* ksm created a completely new copy */ page_add_new_anon_rmap(page, vma, fe->address, false); mem_cgroup_commit_charge(page, memcg, false, false); -- cgit v0.10.2 From 117dec978cf64e8e96f13d0cf4891ff77c9acf55 Mon Sep 17 00:00:00 2001 From: Valdis Kletnieks Date: Tue, 2 Aug 2016 14:02:28 -0700 Subject: tools/testing/radix-tree/linux/gfp.h: fix bitrotted value Apparently, the tools/testing version dates to a few flags ago, and we've sprouted 4 new ones since. Keep in sync with the value in the main tree... Link: http://lkml.kernel.org/r/23400.1469702675@turing-police.cc.vt.edu Signed-off-by: Valdis Kletnieks Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/tools/testing/radix-tree/linux/gfp.h b/tools/testing/radix-tree/linux/gfp.h index 0e37f7a..5201b91 100644 --- a/tools/testing/radix-tree/linux/gfp.h +++ b/tools/testing/radix-tree/linux/gfp.h @@ -1,7 +1,7 @@ #ifndef _GFP_H #define _GFP_H -#define __GFP_BITS_SHIFT 22 +#define __GFP_BITS_SHIFT 26 #define __GFP_BITS_MASK ((gfp_t)((1 << __GFP_BITS_SHIFT) - 1)) #define __GFP_WAIT 1 #define __GFP_ACCOUNT 0 -- cgit v0.10.2 From 649920c6ab93429b94bc7c1aa7c0e8395351be32 Mon Sep 17 00:00:00 2001 From: Jia He Date: Tue, 2 Aug 2016 14:02:31 -0700 Subject: mm/hugetlb: avoid soft lockup in set_max_huge_pages() In powerpc servers with large memory(32TB), we watched several soft lockups for hugepage under stress tests. The call traces are as follows: 1. get_page_from_freelist+0x2d8/0xd50 __alloc_pages_nodemask+0x180/0xc20 alloc_fresh_huge_page+0xb0/0x190 set_max_huge_pages+0x164/0x3b0 2. prep_new_huge_page+0x5c/0x100 alloc_fresh_huge_page+0xc8/0x190 set_max_huge_pages+0x164/0x3b0 This patch fixes such soft lockups. It is safe to call cond_resched() there because it is out of spin_lock/unlock section. Link: http://lkml.kernel.org/r/1469674442-14848-1-git-send-email-hejianet@gmail.com Signed-off-by: Jia He Reviewed-by: Naoya Horiguchi Acked-by: Michal Hocko Acked-by: Dave Hansen Cc: Mike Kravetz Cc: "Kirill A. Shutemov" Cc: Paul Gortmaker Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/hugetlb.c b/mm/hugetlb.c index f904246..619e00d 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -2216,6 +2216,10 @@ static unsigned long set_max_huge_pages(struct hstate *h, unsigned long count, * and reducing the surplus. */ spin_unlock(&hugetlb_lock); + + /* yield cpu to avoid soft lockup */ + cond_resched(); + if (hstate_is_gigantic(h)) ret = alloc_fresh_gigantic_page(h, nodes_allowed); else -- cgit v0.10.2 From 4e666314d286765a9e61818b488c7372326654ec Mon Sep 17 00:00:00 2001 From: Michal Hocko Date: Tue, 2 Aug 2016 14:02:34 -0700 Subject: mm, hugetlb: fix huge_pte_alloc BUG_ON Zhong Jiang has reported a BUG_ON from huge_pte_alloc hitting when he runs his database load with memory online and offline running in parallel. The reason is that huge_pmd_share might detect a shared pmd which is currently migrated and so it has migration pte which is !pte_huge. There doesn't seem to be any easy way to prevent from the race and in fact seeing the migration swap entry is not harmful. Both callers of huge_pte_alloc are prepared to handle them. copy_hugetlb_page_range will copy the swap entry and make it COW if needed. hugetlb_fault will back off and so the page fault is retries if the page is still under migration and waits for its completion in hugetlb_fault. That means that the BUG_ON is wrong and we should update it. Let's simply check that all present ptes are pte_huge instead. Link: http://lkml.kernel.org/r/20160721074340.GA26398@dhcp22.suse.cz Signed-off-by: Michal Hocko Reported-by: zhongjiang Acked-by: Naoya Horiguchi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 619e00d..ef96830 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -4310,7 +4310,7 @@ pte_t *huge_pte_alloc(struct mm_struct *mm, pte = (pte_t *)pmd_alloc(mm, pud, addr); } } - BUG_ON(pte && !pte_none(*pte) && !pte_huge(*pte)); + BUG_ON(pte && pte_present(*pte) && !pte_huge(*pte)); return pte; } -- cgit v0.10.2 From d6507ff5331c002430cc20ab25922479453baae7 Mon Sep 17 00:00:00 2001 From: Michal Hocko Date: Tue, 2 Aug 2016 14:02:37 -0700 Subject: memcg: put soft limit reclaim out of way if the excess tree is empty We've had a report about soft lockups caused by lock bouncing in the soft reclaim path: BUG: soft lockup - CPU#0 stuck for 22s! [kav4proxy-kavic:3128] RIP: 0010:[] [] _raw_spin_lock+0x18/0x20 Call Trace: mem_cgroup_soft_limit_reclaim+0x25a/0x280 shrink_zones+0xed/0x200 do_try_to_free_pages+0x74/0x320 try_to_free_pages+0x112/0x180 __alloc_pages_slowpath+0x3ff/0x820 __alloc_pages_nodemask+0x1e9/0x200 alloc_pages_vma+0xe1/0x290 do_wp_page+0x19f/0x840 handle_pte_fault+0x1cd/0x230 do_page_fault+0x1fd/0x4c0 page_fault+0x25/0x30 There are no memcgs created so there cannot be any in the soft limit excess obviously: [...] memory 0 1 1 so all this just seems to be mem_cgroup_largest_soft_limit_node trying to get spin_lock_irq(&mctz->lock) just to find out that the soft limit excess tree is empty. This is just pointless wasting of cycles and cache line bouncing during heavy parallel reclaim on large machines. The particular machine wasn't very healthy and most probably suffering from a memory leak which just caused the memory reclaim to trash heavily. But bouncing on the lock certainly didn't help... Fix this by optimistic lockless check and bail out early if the tree is empty. This is theoretically racy but that shouldn't matter all that much. First of all soft limit is a best effort feature and it is slowly getting deprecated and its usage should be really scarce. Bouncing on a lock without a good reason is surely much bigger problem, especially on large CPU machines. Link: http://lkml.kernel.org/r/1470073277-1056-1-git-send-email-mhocko@kernel.org Signed-off-by: Michal Hocko Acked-by: Vladimir Davydov Acked-by: Johannes Weiner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/memcontrol.c b/mm/memcontrol.c index c265212..66beca1 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -2559,6 +2559,15 @@ unsigned long mem_cgroup_soft_limit_reclaim(pg_data_t *pgdat, int order, return 0; mctz = soft_limit_tree_node(pgdat->node_id); + + /* + * Do not even bother to check the largest node if the root + * is empty. Do it lockless to prevent lock bouncing. Races + * are acceptable as soft limit is best effort anyway. + */ + if (RB_EMPTY_ROOT(&mctz->rb_root)) + return 0; + /* * This loop can run a while, specially if mem_cgroup's continuously * keep exceeding their soft limit and putting the system under -- cgit v0.10.2 From 4a3d308d6674fabf213bce9c1a661ef43a85e515 Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Tue, 2 Aug 2016 14:02:40 -0700 Subject: mm/kasan: fix corruptions and false positive reports Once an object is put into quarantine, we no longer own it, i.e. object could leave the quarantine and be reallocated. So having set_track() call after the quarantine_put() may corrupt slab objects. BUG kmalloc-4096 (Not tainted): Poison overwritten ----------------------------------------------------------------------------- Disabling lock debugging due to kernel taint INFO: 0xffff8804540de850-0xffff8804540de857. First byte 0xb5 instead of 0x6b ... INFO: Freed in qlist_free_all+0x42/0x100 age=75 cpu=3 pid=24492 __slab_free+0x1d6/0x2e0 ___cache_free+0xb6/0xd0 qlist_free_all+0x83/0x100 quarantine_reduce+0x177/0x1b0 kasan_kmalloc+0xf3/0x100 kasan_slab_alloc+0x12/0x20 kmem_cache_alloc+0x109/0x3e0 mmap_region+0x53e/0xe40 do_mmap+0x70f/0xa50 vm_mmap_pgoff+0x147/0x1b0 SyS_mmap_pgoff+0x2c7/0x5b0 SyS_mmap+0x1b/0x30 do_syscall_64+0x1a0/0x4e0 return_from_SYSCALL_64+0x0/0x7a INFO: Slab 0xffffea0011503600 objects=7 used=7 fp=0x (null) flags=0x8000000000004080 INFO: Object 0xffff8804540de848 @offset=26696 fp=0xffff8804540dc588 Redzone ffff8804540de840: bb bb bb bb bb bb bb bb ........ Object ffff8804540de848: 6b 6b 6b 6b 6b 6b 6b 6b b5 52 00 00 f2 01 60 cc kkkkkkkk.R....`. Similarly, poisoning after the quarantine_put() leads to false positive use-after-free reports: BUG: KASAN: use-after-free in anon_vma_interval_tree_insert+0x304/0x430 at addr ffff880405c540a0 Read of size 8 by task trinity-c0/3036 CPU: 0 PID: 3036 Comm: trinity-c0 Not tainted 4.7.0-think+ #9 Call Trace: dump_stack+0x68/0x96 kasan_report_error+0x222/0x600 __asan_report_load8_noabort+0x61/0x70 anon_vma_interval_tree_insert+0x304/0x430 anon_vma_chain_link+0x91/0xd0 anon_vma_clone+0x136/0x3f0 anon_vma_fork+0x81/0x4c0 copy_process.part.47+0x2c43/0x5b20 _do_fork+0x16d/0xbd0 SyS_clone+0x19/0x20 do_syscall_64+0x1a0/0x4e0 entry_SYSCALL64_slow_path+0x25/0x25 Fix this by putting an object in the quarantine after all other operations. Fixes: 80a9201a5965 ("mm, kasan: switch SLUB to stackdepot, enable memory quarantine for SLUB") Link: http://lkml.kernel.org/r/1470062715-14077-1-git-send-email-aryabinin@virtuozzo.com Signed-off-by: Andrey Ryabinin Reported-by: Dave Jones Reported-by: Vegard Nossum Reported-by: Sasha Levin Acked-by: Alexander Potapenko Cc: Dmitry Vyukov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/kasan/kasan.c b/mm/kasan/kasan.c index b6f99e8..3019cec 100644 --- a/mm/kasan/kasan.c +++ b/mm/kasan/kasan.c @@ -543,9 +543,9 @@ bool kasan_slab_free(struct kmem_cache *cache, void *object) switch (alloc_info->state) { case KASAN_STATE_ALLOC: alloc_info->state = KASAN_STATE_QUARANTINE; - quarantine_put(free_info, cache); set_track(&free_info->track, GFP_NOWAIT); kasan_poison_slab_free(cache, object); + quarantine_put(free_info, cache); return true; case KASAN_STATE_QUARANTINE: case KASAN_STATE_FREE: -- cgit v0.10.2 From 4b3ec5a3f4b1d5c9d64b9ab704042400d050d432 Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Tue, 2 Aug 2016 14:02:43 -0700 Subject: mm/kasan: don't reduce quarantine in atomic contexts Currently we call quarantine_reduce() for ___GFP_KSWAPD_RECLAIM (implied by __GFP_RECLAIM) allocation. So, basically we call it on almost every allocation. quarantine_reduce() sometimes is heavy operation, and calling it with disabled interrupts may trigger hard LOCKUP: NMI watchdog: Watchdog detected hard LOCKUP on cpu 2irq event stamp: 1411258 Call Trace: dump_stack+0x68/0x96 watchdog_overflow_callback+0x15b/0x190 __perf_event_overflow+0x1b1/0x540 perf_event_overflow+0x14/0x20 intel_pmu_handle_irq+0x36a/0xad0 perf_event_nmi_handler+0x2c/0x50 nmi_handle+0x128/0x480 default_do_nmi+0xb2/0x210 do_nmi+0x1aa/0x220 end_repeat_nmi+0x1a/0x1e <> __kernel_text_address+0x86/0xb0 print_context_stack+0x7b/0x100 dump_trace+0x12b/0x350 save_stack_trace+0x2b/0x50 set_track+0x83/0x140 free_debug_processing+0x1aa/0x420 __slab_free+0x1d6/0x2e0 ___cache_free+0xb6/0xd0 qlist_free_all+0x83/0x100 quarantine_reduce+0x177/0x1b0 kasan_kmalloc+0xf3/0x100 Reduce the quarantine_reduce iff direct reclaim is allowed. Fixes: 55834c59098d("mm: kasan: initial memory quarantine implementation") Link: http://lkml.kernel.org/r/1470062715-14077-2-git-send-email-aryabinin@virtuozzo.com Signed-off-by: Andrey Ryabinin Reported-by: Dave Jones Acked-by: Alexander Potapenko Cc: Dmitry Vyukov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/kasan/kasan.c b/mm/kasan/kasan.c index 3019cec..c99ef40 100644 --- a/mm/kasan/kasan.c +++ b/mm/kasan/kasan.c @@ -565,7 +565,7 @@ void kasan_kmalloc(struct kmem_cache *cache, const void *object, size_t size, unsigned long redzone_start; unsigned long redzone_end; - if (flags & __GFP_RECLAIM) + if (gfpflags_allow_blocking(flags)) quarantine_reduce(); if (unlikely(object == NULL)) @@ -596,7 +596,7 @@ void kasan_kmalloc_large(const void *ptr, size_t size, gfp_t flags) unsigned long redzone_start; unsigned long redzone_end; - if (flags & __GFP_RECLAIM) + if (gfpflags_allow_blocking(flags)) quarantine_reduce(); if (unlikely(ptr == NULL)) -- cgit v0.10.2 From f7376aed6c032aab820fa36806a89e16e353a0d9 Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Tue, 2 Aug 2016 14:02:46 -0700 Subject: mm/kasan, slub: don't disable interrupts when object leaves quarantine SLUB doesn't require disabled interrupts to call ___cache_free(). Link: http://lkml.kernel.org/r/1470062715-14077-3-git-send-email-aryabinin@virtuozzo.com Signed-off-by: Andrey Ryabinin Acked-by: Alexander Potapenko Cc: Dmitry Vyukov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/kasan/quarantine.c b/mm/kasan/quarantine.c index 65793f1..4852625 100644 --- a/mm/kasan/quarantine.c +++ b/mm/kasan/quarantine.c @@ -147,10 +147,14 @@ static void qlink_free(struct qlist_node *qlink, struct kmem_cache *cache) struct kasan_alloc_meta *alloc_info = get_alloc_info(cache, object); unsigned long flags; - local_irq_save(flags); + if (IS_ENABLED(CONFIG_SLAB)) + local_irq_save(flags); + alloc_info->state = KASAN_STATE_FREE; ___cache_free(cache, object, _THIS_IP_); - local_irq_restore(flags); + + if (IS_ENABLED(CONFIG_SLAB)) + local_irq_restore(flags); } static void qlist_free_all(struct qlist_head *q, struct kmem_cache *cache) -- cgit v0.10.2 From 47b5c2a0f021e90a79845d1a1353780e5edd0bce Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Tue, 2 Aug 2016 14:02:49 -0700 Subject: mm/kasan: get rid of ->alloc_size in struct kasan_alloc_meta Size of slab object already stored in cache->object_size. Note, that kmalloc() internally rounds up size of allocation, so object_size may be not equal to alloc_size, but, usually we don't need to know the exact size of allocated object. In case if we need that information, we still can figure it out from the report. The dump of shadow memory allows to identify the end of allocated memory, and thereby the exact allocation size. Link: http://lkml.kernel.org/r/1470062715-14077-4-git-send-email-aryabinin@virtuozzo.com Signed-off-by: Andrey Ryabinin Cc: Alexander Potapenko Cc: Dmitry Vyukov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/kasan/kasan.c b/mm/kasan/kasan.c index c99ef40..388e812 100644 --- a/mm/kasan/kasan.c +++ b/mm/kasan/kasan.c @@ -584,7 +584,6 @@ void kasan_kmalloc(struct kmem_cache *cache, const void *object, size_t size, get_alloc_info(cache, object); alloc_info->state = KASAN_STATE_ALLOC; - alloc_info->alloc_size = size; set_track(&alloc_info->track, flags); } } diff --git a/mm/kasan/kasan.h b/mm/kasan/kasan.h index 31972cd..aa17546 100644 --- a/mm/kasan/kasan.h +++ b/mm/kasan/kasan.h @@ -75,8 +75,7 @@ struct kasan_track { struct kasan_alloc_meta { struct kasan_track track; - u32 state : 2; /* enum kasan_state */ - u32 alloc_size : 30; + u32 state; }; struct qlist_node { diff --git a/mm/kasan/report.c b/mm/kasan/report.c index 861b977..d67a7e0 100644 --- a/mm/kasan/report.c +++ b/mm/kasan/report.c @@ -136,7 +136,9 @@ static void kasan_object_err(struct kmem_cache *cache, struct page *page, struct kasan_free_meta *free_info; dump_stack(); - pr_err("Object at %p, in cache %s\n", object, cache->name); + pr_err("Object at %p, in cache %s size: %d\n", object, cache->name, + cache->object_size); + if (!(cache->flags & SLAB_KASAN)) return; switch (alloc_info->state) { @@ -144,15 +146,11 @@ static void kasan_object_err(struct kmem_cache *cache, struct page *page, pr_err("Object not allocated yet\n"); break; case KASAN_STATE_ALLOC: - pr_err("Object allocated with size %u bytes.\n", - alloc_info->alloc_size); pr_err("Allocation:\n"); print_track(&alloc_info->track); break; case KASAN_STATE_FREE: case KASAN_STATE_QUARANTINE: - pr_err("Object freed, allocated with size %u bytes\n", - alloc_info->alloc_size); free_info = get_free_info(cache, object); pr_err("Allocation:\n"); print_track(&alloc_info->track); -- cgit v0.10.2 From b3cbd9bf77cd1888114dbee1653e79aa23fd4068 Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Tue, 2 Aug 2016 14:02:52 -0700 Subject: mm/kasan: get rid of ->state in struct kasan_alloc_meta The state of object currently tracked in two places - shadow memory, and the ->state field in struct kasan_alloc_meta. We can get rid of the latter. The will save us a little bit of memory. Also, this allow us to move free stack into struct kasan_alloc_meta, without increasing memory consumption. So now we should always know when the last time the object was freed. This may be useful for long delayed use-after-free bugs. As a side effect this fixes following UBSAN warning: UBSAN: Undefined behaviour in mm/kasan/quarantine.c:102:13 member access within misaligned address ffff88000d1efebc for type 'struct qlist_node' which requires 8 byte alignment Link: http://lkml.kernel.org/r/1470062715-14077-5-git-send-email-aryabinin@virtuozzo.com Reported-by: kernel test robot Signed-off-by: Andrey Ryabinin Cc: Alexander Potapenko Cc: Dmitry Vyukov Cc: Christoph Lameter Cc: Pekka Enberg Cc: David Rientjes Cc: Joonsoo Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/kasan.h b/include/linux/kasan.h index c9cf374..d600303 100644 --- a/include/linux/kasan.h +++ b/include/linux/kasan.h @@ -56,6 +56,7 @@ void kasan_cache_destroy(struct kmem_cache *cache); void kasan_poison_slab(struct page *page); void kasan_unpoison_object_data(struct kmem_cache *cache, void *object); void kasan_poison_object_data(struct kmem_cache *cache, void *object); +void kasan_init_slab_obj(struct kmem_cache *cache, const void *object); void kasan_kmalloc_large(const void *ptr, size_t size, gfp_t flags); void kasan_kfree_large(const void *ptr); @@ -102,6 +103,8 @@ static inline void kasan_unpoison_object_data(struct kmem_cache *cache, void *object) {} static inline void kasan_poison_object_data(struct kmem_cache *cache, void *object) {} +static inline void kasan_init_slab_obj(struct kmem_cache *cache, + const void *object) {} static inline void kasan_kmalloc_large(void *ptr, size_t size, gfp_t flags) {} static inline void kasan_kfree_large(const void *ptr) {} diff --git a/mm/kasan/kasan.c b/mm/kasan/kasan.c index 388e812..92750e3 100644 --- a/mm/kasan/kasan.c +++ b/mm/kasan/kasan.c @@ -442,11 +442,6 @@ void kasan_poison_object_data(struct kmem_cache *cache, void *object) kasan_poison_shadow(object, round_up(cache->object_size, KASAN_SHADOW_SCALE_SIZE), KASAN_KMALLOC_REDZONE); - if (cache->flags & SLAB_KASAN) { - struct kasan_alloc_meta *alloc_info = - get_alloc_info(cache, object); - alloc_info->state = KASAN_STATE_INIT; - } } static inline int in_irqentry_text(unsigned long ptr) @@ -510,6 +505,17 @@ struct kasan_free_meta *get_free_info(struct kmem_cache *cache, return (void *)object + cache->kasan_info.free_meta_offset; } +void kasan_init_slab_obj(struct kmem_cache *cache, const void *object) +{ + struct kasan_alloc_meta *alloc_info; + + if (!(cache->flags & SLAB_KASAN)) + return; + + alloc_info = get_alloc_info(cache, object); + __memset(alloc_info, 0, sizeof(*alloc_info)); +} + void kasan_slab_alloc(struct kmem_cache *cache, void *object, gfp_t flags) { kasan_kmalloc(cache, object, cache->object_size, flags); @@ -529,34 +535,27 @@ static void kasan_poison_slab_free(struct kmem_cache *cache, void *object) bool kasan_slab_free(struct kmem_cache *cache, void *object) { + s8 shadow_byte; + /* RCU slabs could be legally used after free within the RCU period */ if (unlikely(cache->flags & SLAB_DESTROY_BY_RCU)) return false; - if (likely(cache->flags & SLAB_KASAN)) { - struct kasan_alloc_meta *alloc_info; - struct kasan_free_meta *free_info; + shadow_byte = READ_ONCE(*(s8 *)kasan_mem_to_shadow(object)); + if (shadow_byte < 0 || shadow_byte >= KASAN_SHADOW_SCALE_SIZE) { + pr_err("Double free"); + dump_stack(); + return true; + } - alloc_info = get_alloc_info(cache, object); - free_info = get_free_info(cache, object); + kasan_poison_slab_free(cache, object); - switch (alloc_info->state) { - case KASAN_STATE_ALLOC: - alloc_info->state = KASAN_STATE_QUARANTINE; - set_track(&free_info->track, GFP_NOWAIT); - kasan_poison_slab_free(cache, object); - quarantine_put(free_info, cache); - return true; - case KASAN_STATE_QUARANTINE: - case KASAN_STATE_FREE: - pr_err("Double free"); - dump_stack(); - break; - default: - break; - } - } - return false; + if (unlikely(!(cache->flags & SLAB_KASAN))) + return false; + + set_track(&get_alloc_info(cache, object)->free_track, GFP_NOWAIT); + quarantine_put(get_free_info(cache, object), cache); + return true; } void kasan_kmalloc(struct kmem_cache *cache, const void *object, size_t size, @@ -579,13 +578,9 @@ void kasan_kmalloc(struct kmem_cache *cache, const void *object, size_t size, kasan_unpoison_shadow(object, size); kasan_poison_shadow((void *)redzone_start, redzone_end - redzone_start, KASAN_KMALLOC_REDZONE); - if (cache->flags & SLAB_KASAN) { - struct kasan_alloc_meta *alloc_info = - get_alloc_info(cache, object); - alloc_info->state = KASAN_STATE_ALLOC; - set_track(&alloc_info->track, flags); - } + if (cache->flags & SLAB_KASAN) + set_track(&get_alloc_info(cache, object)->alloc_track, flags); } EXPORT_SYMBOL(kasan_kmalloc); diff --git a/mm/kasan/kasan.h b/mm/kasan/kasan.h index aa17546..9b7b31e 100644 --- a/mm/kasan/kasan.h +++ b/mm/kasan/kasan.h @@ -59,13 +59,6 @@ struct kasan_global { * Structures to keep alloc and free tracks * */ -enum kasan_state { - KASAN_STATE_INIT, - KASAN_STATE_ALLOC, - KASAN_STATE_QUARANTINE, - KASAN_STATE_FREE -}; - #define KASAN_STACK_DEPTH 64 struct kasan_track { @@ -74,8 +67,8 @@ struct kasan_track { }; struct kasan_alloc_meta { - struct kasan_track track; - u32 state; + struct kasan_track alloc_track; + struct kasan_track free_track; }; struct qlist_node { @@ -86,7 +79,6 @@ struct kasan_free_meta { * Otherwise it might be used for the allocator freelist. */ struct qlist_node quarantine_link; - struct kasan_track track; }; struct kasan_alloc_meta *get_alloc_info(struct kmem_cache *cache, diff --git a/mm/kasan/quarantine.c b/mm/kasan/quarantine.c index 4852625..7fd121d 100644 --- a/mm/kasan/quarantine.c +++ b/mm/kasan/quarantine.c @@ -144,13 +144,11 @@ static void *qlink_to_object(struct qlist_node *qlink, struct kmem_cache *cache) static void qlink_free(struct qlist_node *qlink, struct kmem_cache *cache) { void *object = qlink_to_object(qlink, cache); - struct kasan_alloc_meta *alloc_info = get_alloc_info(cache, object); unsigned long flags; if (IS_ENABLED(CONFIG_SLAB)) local_irq_save(flags); - alloc_info->state = KASAN_STATE_FREE; ___cache_free(cache, object, _THIS_IP_); if (IS_ENABLED(CONFIG_SLAB)) diff --git a/mm/kasan/report.c b/mm/kasan/report.c index d67a7e0..f437398 100644 --- a/mm/kasan/report.c +++ b/mm/kasan/report.c @@ -133,7 +133,6 @@ static void kasan_object_err(struct kmem_cache *cache, struct page *page, void *object, char *unused_reason) { struct kasan_alloc_meta *alloc_info = get_alloc_info(cache, object); - struct kasan_free_meta *free_info; dump_stack(); pr_err("Object at %p, in cache %s size: %d\n", object, cache->name, @@ -141,23 +140,11 @@ static void kasan_object_err(struct kmem_cache *cache, struct page *page, if (!(cache->flags & SLAB_KASAN)) return; - switch (alloc_info->state) { - case KASAN_STATE_INIT: - pr_err("Object not allocated yet\n"); - break; - case KASAN_STATE_ALLOC: - pr_err("Allocation:\n"); - print_track(&alloc_info->track); - break; - case KASAN_STATE_FREE: - case KASAN_STATE_QUARANTINE: - free_info = get_free_info(cache, object); - pr_err("Allocation:\n"); - print_track(&alloc_info->track); - pr_err("Deallocation:\n"); - print_track(&free_info->track); - break; - } + + pr_err("Allocated:\n"); + print_track(&alloc_info->alloc_track); + pr_err("Freed:\n"); + print_track(&alloc_info->free_track); } static void print_address_description(struct kasan_access_info *info) diff --git a/mm/slab.c b/mm/slab.c index 09771ed..ca135bd 100644 --- a/mm/slab.c +++ b/mm/slab.c @@ -2604,9 +2604,11 @@ static void cache_init_objs(struct kmem_cache *cachep, } for (i = 0; i < cachep->num; i++) { + objp = index_to_obj(cachep, page, i); + kasan_init_slab_obj(cachep, objp); + /* constructor could break poison info */ if (DEBUG == 0 && cachep->ctor) { - objp = index_to_obj(cachep, page, i); kasan_unpoison_object_data(cachep, objp); cachep->ctor(objp); kasan_poison_object_data(cachep, objp); diff --git a/mm/slub.c b/mm/slub.c index 74e7c8c..26eb6a99 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -1384,6 +1384,7 @@ static void setup_object(struct kmem_cache *s, struct page *page, void *object) { setup_object_debug(s, page, object); + kasan_init_slab_obj(s, object); if (unlikely(s->ctor)) { kasan_unpoison_object_data(s, object); s->ctor(object); -- cgit v0.10.2 From 7e088978933ee186533355ae03a9dc1de99cf6c7 Mon Sep 17 00:00:00 2001 From: Andrey Ryabinin Date: Tue, 2 Aug 2016 14:02:55 -0700 Subject: kasan: improve double-free reports Currently we just dump stack in case of double free bug. Let's dump all info about the object that we have. [aryabinin@virtuozzo.com: change double free message per Alexander] Link: http://lkml.kernel.org/r/1470153654-30160-1-git-send-email-aryabinin@virtuozzo.com Link: http://lkml.kernel.org/r/1470062715-14077-6-git-send-email-aryabinin@virtuozzo.com Signed-off-by: Andrey Ryabinin Cc: Alexander Potapenko Cc: Dmitry Vyukov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/kasan/kasan.c b/mm/kasan/kasan.c index 92750e3..88af13c 100644 --- a/mm/kasan/kasan.c +++ b/mm/kasan/kasan.c @@ -543,8 +543,7 @@ bool kasan_slab_free(struct kmem_cache *cache, void *object) shadow_byte = READ_ONCE(*(s8 *)kasan_mem_to_shadow(object)); if (shadow_byte < 0 || shadow_byte >= KASAN_SHADOW_SCALE_SIZE) { - pr_err("Double free"); - dump_stack(); + kasan_report_double_free(cache, object, shadow_byte); return true; } diff --git a/mm/kasan/kasan.h b/mm/kasan/kasan.h index 9b7b31e..e5c2181 100644 --- a/mm/kasan/kasan.h +++ b/mm/kasan/kasan.h @@ -99,6 +99,8 @@ static inline bool kasan_report_enabled(void) void kasan_report(unsigned long addr, size_t size, bool is_write, unsigned long ip); +void kasan_report_double_free(struct kmem_cache *cache, void *object, + s8 shadow); #if defined(CONFIG_SLAB) || defined(CONFIG_SLUB) void quarantine_put(struct kasan_free_meta *info, struct kmem_cache *cache); diff --git a/mm/kasan/report.c b/mm/kasan/report.c index f437398..24c1211 100644 --- a/mm/kasan/report.c +++ b/mm/kasan/report.c @@ -116,6 +116,26 @@ static inline bool init_task_stack_addr(const void *addr) sizeof(init_thread_union.stack)); } +static DEFINE_SPINLOCK(report_lock); + +static void kasan_start_report(unsigned long *flags) +{ + /* + * Make sure we don't end up in loop. + */ + kasan_disable_current(); + spin_lock_irqsave(&report_lock, *flags); + pr_err("==================================================================\n"); +} + +static void kasan_end_report(unsigned long *flags) +{ + pr_err("==================================================================\n"); + add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); + spin_unlock_irqrestore(&report_lock, *flags); + kasan_enable_current(); +} + static void print_track(struct kasan_track *track) { pr_err("PID = %u\n", track->pid); @@ -129,8 +149,7 @@ static void print_track(struct kasan_track *track) } } -static void kasan_object_err(struct kmem_cache *cache, struct page *page, - void *object, char *unused_reason) +static void kasan_object_err(struct kmem_cache *cache, void *object) { struct kasan_alloc_meta *alloc_info = get_alloc_info(cache, object); @@ -147,6 +166,18 @@ static void kasan_object_err(struct kmem_cache *cache, struct page *page, print_track(&alloc_info->free_track); } +void kasan_report_double_free(struct kmem_cache *cache, void *object, + s8 shadow) +{ + unsigned long flags; + + kasan_start_report(&flags); + pr_err("BUG: Double free or freeing an invalid pointer\n"); + pr_err("Unexpected shadow byte: 0x%hhX\n", shadow); + kasan_object_err(cache, object); + kasan_end_report(&flags); +} + static void print_address_description(struct kasan_access_info *info) { const void *addr = info->access_addr; @@ -160,8 +191,7 @@ static void print_address_description(struct kasan_access_info *info) struct kmem_cache *cache = page->slab_cache; object = nearest_obj(cache, page, (void *)info->access_addr); - kasan_object_err(cache, page, object, - "kasan: bad access detected"); + kasan_object_err(cache, object); return; } dump_page(page, "kasan: bad access detected"); @@ -226,19 +256,13 @@ static void print_shadow_for_address(const void *addr) } } -static DEFINE_SPINLOCK(report_lock); - static void kasan_report_error(struct kasan_access_info *info) { unsigned long flags; const char *bug_type; - /* - * Make sure we don't end up in loop. - */ - kasan_disable_current(); - spin_lock_irqsave(&report_lock, flags); - pr_err("==================================================================\n"); + kasan_start_report(&flags); + if (info->access_addr < kasan_shadow_to_mem((void *)KASAN_SHADOW_START)) { if ((unsigned long)info->access_addr < PAGE_SIZE) @@ -259,10 +283,8 @@ static void kasan_report_error(struct kasan_access_info *info) print_address_description(info); print_shadow_for_address(info->first_bad_addr); } - pr_err("==================================================================\n"); - add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE); - spin_unlock_irqrestore(&report_lock, flags); - kasan_enable_current(); + + kasan_end_report(&flags); } void kasan_report(unsigned long addr, size_t size, -- cgit v0.10.2 From c3cee372282cb6bcdf19ac1457581d5dd5ecb554 Mon Sep 17 00:00:00 2001 From: Alexander Potapenko Date: Tue, 2 Aug 2016 14:02:58 -0700 Subject: kasan: avoid overflowing quarantine size on low memory systems If the total amount of memory assigned to quarantine is less than the amount of memory assigned to per-cpu quarantines, |new_quarantine_size| may overflow. Instead, set it to zero. [akpm@linux-foundation.org: cleanup: use WARN_ONCE return value] Link: http://lkml.kernel.org/r/1470063563-96266-1-git-send-email-glider@google.com Fixes: 55834c59098d ("mm: kasan: initial memory quarantine implementation") Signed-off-by: Alexander Potapenko Reported-by: Dmitry Vyukov Cc: Andrey Ryabinin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/kasan/quarantine.c b/mm/kasan/quarantine.c index 7fd121d..b6728a3 100644 --- a/mm/kasan/quarantine.c +++ b/mm/kasan/quarantine.c @@ -198,7 +198,7 @@ void quarantine_put(struct kasan_free_meta *info, struct kmem_cache *cache) void quarantine_reduce(void) { - size_t new_quarantine_size; + size_t new_quarantine_size, percpu_quarantines; unsigned long flags; struct qlist_head to_free = QLIST_INIT; size_t size_to_free = 0; @@ -216,7 +216,12 @@ void quarantine_reduce(void) */ new_quarantine_size = (READ_ONCE(totalram_pages) << PAGE_SHIFT) / QUARANTINE_FRACTION; - new_quarantine_size -= QUARANTINE_PERCPU_SIZE * num_online_cpus(); + percpu_quarantines = QUARANTINE_PERCPU_SIZE * num_online_cpus(); + if (WARN_ONCE(new_quarantine_size < percpu_quarantines, + "Too little memory, disabling global KASAN quarantine.\n")) + new_quarantine_size = 0; + else + new_quarantine_size -= percpu_quarantines; WRITE_ONCE(quarantine_size, new_quarantine_size); last = global_quarantine.head; -- cgit v0.10.2 From 05eb6e7263185a6bb0de9501ccf2addc52429414 Mon Sep 17 00:00:00 2001 From: Vladimir Davydov Date: Tue, 2 Aug 2016 14:03:01 -0700 Subject: radix-tree: account nodes to memcg only if explicitly requested Radix trees may be used not only for storing page cache pages, so unconditionally accounting radix tree nodes to the current memory cgroup is bad: if a radix tree node is used for storing data shared among different cgroups we risk pinning dead memory cgroups forever. So let's only account radix tree nodes if it was explicitly requested by passing __GFP_ACCOUNT to INIT_RADIX_TREE. Currently, we only want to account page cache entries, so mark mapping->page_tree so. Fixes: 58e698af4c63 ("radix-tree: account radix_tree_node to memory cgroup") Link: http://lkml.kernel.org/r/1470057188-7864-1-git-send-email-vdavydov@virtuozzo.com Signed-off-by: Vladimir Davydov Acked-by: Johannes Weiner Acked-by: Michal Hocko Cc: [4.6+] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/inode.c b/fs/inode.c index 9cef4e1..ad44554 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -345,7 +345,7 @@ EXPORT_SYMBOL(inc_nlink); void address_space_init_once(struct address_space *mapping) { memset(mapping, 0, sizeof(*mapping)); - INIT_RADIX_TREE(&mapping->page_tree, GFP_ATOMIC); + INIT_RADIX_TREE(&mapping->page_tree, GFP_ATOMIC | __GFP_ACCOUNT); spin_lock_init(&mapping->tree_lock); init_rwsem(&mapping->i_mmap_rwsem); INIT_LIST_HEAD(&mapping->private_list); diff --git a/lib/radix-tree.c b/lib/radix-tree.c index 61b8fb5..1b7bf73 100644 --- a/lib/radix-tree.c +++ b/lib/radix-tree.c @@ -277,10 +277,11 @@ radix_tree_node_alloc(struct radix_tree_root *root) /* * Even if the caller has preloaded, try to allocate from the - * cache first for the new node to get accounted. + * cache first for the new node to get accounted to the memory + * cgroup. */ ret = kmem_cache_alloc(radix_tree_node_cachep, - gfp_mask | __GFP_ACCOUNT | __GFP_NOWARN); + gfp_mask | __GFP_NOWARN); if (ret) goto out; @@ -303,8 +304,7 @@ radix_tree_node_alloc(struct radix_tree_root *root) kmemleak_update_trace(ret); goto out; } - ret = kmem_cache_alloc(radix_tree_node_cachep, - gfp_mask | __GFP_ACCOUNT); + ret = kmem_cache_alloc(radix_tree_node_cachep, gfp_mask); out: BUG_ON(radix_tree_is_internal_node(ret)); return ret; @@ -351,6 +351,12 @@ static int __radix_tree_preload(gfp_t gfp_mask, int nr) struct radix_tree_node *node; int ret = -ENOMEM; + /* + * Nodes preloaded by one cgroup can be be used by another cgroup, so + * they should never be accounted to any particular memory cgroup. + */ + gfp_mask &= ~__GFP_ACCOUNT; + preempt_disable(); rtp = this_cpu_ptr(&radix_tree_preloads); while (rtp->nr < nr) { -- cgit v0.10.2 From b5afba2974f9ebbaaa11b5a633e55db9be3cc363 Mon Sep 17 00:00:00 2001 From: Vladimir Davydov Date: Tue, 2 Aug 2016 14:03:04 -0700 Subject: mm: vmscan: fix memcg-aware shrinkers not called on global reclaim We must call shrink_slab() for each memory cgroup on both global and memcg reclaim in shrink_node_memcg(). Commit d71df22b55099 accidentally changed that so that now shrink_slab() is only called with memcg != NULL on memcg reclaim. As a result, memcg-aware shrinkers (including dentry/inode) are never invoked on global reclaim. Fix that. Fixes: b2e18757f2c9 ("mm, vmscan: begin reclaiming pages on a per-node basis") Link: http://lkml.kernel.org/r/1470056590-7177-1-git-send-email-vdavydov@virtuozzo.com Signed-off-by: Vladimir Davydov Acked-by: Johannes Weiner Acked-by: Michal Hocko Cc: Mel Gorman Cc: Hillf Danton Cc: Vlastimil Babka Cc: Joonsoo Kim Cc: Minchan Kim Cc: Rik van Riel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/vmscan.c b/mm/vmscan.c index 650d268..374d95d 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -2561,7 +2561,7 @@ static bool shrink_node(pg_data_t *pgdat, struct scan_control *sc) shrink_node_memcg(pgdat, memcg, sc, &lru_pages); node_lru_pages += lru_pages; - if (!global_reclaim(sc)) + if (memcg) shrink_slab(sc->gfp_mask, pgdat->node_id, memcg, sc->nr_scanned - scanned, lru_pages); -- cgit v0.10.2 From 9b24fef9f0410fb5364245d6cc2bd044cc064007 Mon Sep 17 00:00:00 2001 From: Fabian Frederick Date: Tue, 2 Aug 2016 14:03:07 -0700 Subject: sysv, ipc: fix security-layer leaking Commit 53dad6d3a8e5 ("ipc: fix race with LSMs") updated ipc_rcu_putref() to receive rcu freeing function but used generic ipc_rcu_free() instead of msg_rcu_free() which does security cleaning. Running LTP msgsnd06 with kmemleak gives the following: cat /sys/kernel/debug/kmemleak unreferenced object 0xffff88003c0a11f8 (size 8): comm "msgsnd06", pid 1645, jiffies 4294672526 (age 6.549s) hex dump (first 8 bytes): 1b 00 00 00 01 00 00 00 ........ backtrace: kmemleak_alloc+0x23/0x40 kmem_cache_alloc_trace+0xe1/0x180 selinux_msg_queue_alloc_security+0x3f/0xd0 security_msg_queue_alloc+0x2e/0x40 newque+0x4e/0x150 ipcget+0x159/0x1b0 SyS_msgget+0x39/0x40 entry_SYSCALL_64_fastpath+0x13/0x8f Manfred Spraul suggested to fix sem.c as well and Davidlohr Bueso to only use ipc_rcu_free in case of security allocation failure in newary() Fixes: 53dad6d3a8e ("ipc: fix race with LSMs") Link: http://lkml.kernel.org/r/1470083552-22966-1-git-send-email-fabf@skynet.be Signed-off-by: Fabian Frederick Cc: Davidlohr Bueso Cc: Manfred Spraul Cc: [3.12+] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/ipc/msg.c b/ipc/msg.c index 1471db9..c6521c2 100644 --- a/ipc/msg.c +++ b/ipc/msg.c @@ -680,7 +680,7 @@ long do_msgsnd(int msqid, long mtype, void __user *mtext, rcu_read_lock(); ipc_lock_object(&msq->q_perm); - ipc_rcu_putref(msq, ipc_rcu_free); + ipc_rcu_putref(msq, msg_rcu_free); /* raced with RMID? */ if (!ipc_valid_object(&msq->q_perm)) { err = -EIDRM; diff --git a/ipc/sem.c b/ipc/sem.c index ae72b3c..7c9d4f7 100644 --- a/ipc/sem.c +++ b/ipc/sem.c @@ -438,7 +438,7 @@ static inline struct sem_array *sem_obtain_object_check(struct ipc_namespace *ns static inline void sem_lock_and_putref(struct sem_array *sma) { sem_lock(sma, NULL, -1); - ipc_rcu_putref(sma, ipc_rcu_free); + ipc_rcu_putref(sma, sem_rcu_free); } static inline void sem_rmid(struct ipc_namespace *ns, struct sem_array *s) @@ -1381,7 +1381,7 @@ static int semctl_main(struct ipc_namespace *ns, int semid, int semnum, rcu_read_unlock(); sem_io = ipc_alloc(sizeof(ushort)*nsems); if (sem_io == NULL) { - ipc_rcu_putref(sma, ipc_rcu_free); + ipc_rcu_putref(sma, sem_rcu_free); return -ENOMEM; } @@ -1415,20 +1415,20 @@ static int semctl_main(struct ipc_namespace *ns, int semid, int semnum, if (nsems > SEMMSL_FAST) { sem_io = ipc_alloc(sizeof(ushort)*nsems); if (sem_io == NULL) { - ipc_rcu_putref(sma, ipc_rcu_free); + ipc_rcu_putref(sma, sem_rcu_free); return -ENOMEM; } } if (copy_from_user(sem_io, p, nsems*sizeof(ushort))) { - ipc_rcu_putref(sma, ipc_rcu_free); + ipc_rcu_putref(sma, sem_rcu_free); err = -EFAULT; goto out_free; } for (i = 0; i < nsems; i++) { if (sem_io[i] > SEMVMX) { - ipc_rcu_putref(sma, ipc_rcu_free); + ipc_rcu_putref(sma, sem_rcu_free); err = -ERANGE; goto out_free; } @@ -1720,7 +1720,7 @@ static struct sem_undo *find_alloc_undo(struct ipc_namespace *ns, int semid) /* step 2: allocate new undo structure */ new = kzalloc(sizeof(struct sem_undo) + sizeof(short)*nsems, GFP_KERNEL); if (!new) { - ipc_rcu_putref(sma, ipc_rcu_free); + ipc_rcu_putref(sma, sem_rcu_free); return ERR_PTR(-ENOMEM); } -- cgit v0.10.2 From 901d805c33fc4c029fc6b2d94ee5fb7d30278045 Mon Sep 17 00:00:00 2001 From: Nicolas Iooss Date: Tue, 2 Aug 2016 14:03:10 -0700 Subject: UBSAN: fix typo in format string handle_object_size_mismatch() used %pk to format a kernel pointer with pr_err(). This seemed to be a misspelling for %pK, but using this to format a kernel pointer does not make much sence here. Therefore use %p instead, like in handle_missaligned_access(). Link: http://lkml.kernel.org/r/20160730083010.11569-1-nicolas.iooss_linux@m4x.org Signed-off-by: Nicolas Iooss Acked-by: Andrey Ryabinin Cc: Joe Perches Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/lib/ubsan.c b/lib/ubsan.c index 8799ae5..fb0409d 100644 --- a/lib/ubsan.c +++ b/lib/ubsan.c @@ -308,7 +308,7 @@ static void handle_object_size_mismatch(struct type_mismatch_data *data, return; ubsan_prologue(&data->location, &flags); - pr_err("%s address %pk with insufficient space\n", + pr_err("%s address %p with insufficient space\n", type_check_kinds[data->type_check_kind], (void *) ptr); pr_err("for an object of type %s\n", data->type->type_name); -- cgit v0.10.2 From 9991a9c8dbd2a45b7e09176ff54ffc8c40ae7791 Mon Sep 17 00:00:00 2001 From: "seokhoon.yoon" Date: Tue, 2 Aug 2016 14:03:13 -0700 Subject: cgroup: update cgroup's document path cgroup's document path is changed to "cgroup-v1". update it. Link: http://lkml.kernel.org/r/1470148443-6509-1-git-send-email-iamyooon@gmail.com Signed-off-by: seokhoon.yoon Cc: Joonsoo Kim Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/init/Kconfig b/init/Kconfig index 46f817a..79c6aad 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -952,7 +952,7 @@ menuconfig CGROUPS controls or device isolation. See - Documentation/scheduler/sched-design-CFS.txt (CFS) - - Documentation/cgroups/ (features for grouping, isolation + - Documentation/cgroup-v1/ (features for grouping, isolation and resource control) Say N if unsure. @@ -1009,7 +1009,7 @@ config BLK_CGROUP CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set CONFIG_BLK_DEV_THROTTLING=y. - See Documentation/cgroups/blkio-controller.txt for more information. + See Documentation/cgroup-v1/blkio-controller.txt for more information. config DEBUG_BLK_CGROUP bool "IO controller debugging" -- cgit v0.10.2 From db4ad0360c54a4e3bf07de9dfe429259ca8dc223 Mon Sep 17 00:00:00 2001 From: Luis de Bethencourt Date: Tue, 2 Aug 2016 14:03:16 -0700 Subject: MAINTAINERS: befs: add new maintainers Salah Triki and Luis de Bethencourt are taking over maintainership of befs. Link: http://lkml.kernel.org/r/1469651079-32455-1-git-send-email-luisbg@osg.samsung.com Signed-off-by: Luis de Bethencourt Acked-by: Greg Kroah-Hartman Acked-by: Theodore Ts'o Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/MAINTAINERS b/MAINTAINERS index 25f4320..bb51bbb 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2346,7 +2346,10 @@ S: Supported F: drivers/media/platform/sti/bdisp BEFS FILE SYSTEM -S: Orphan +M: Luis de Bethencourt +M: Salah Triki +S: Maintained +T: git git://github.com/luisbg/linux-befs.git F: Documentation/filesystems/befs.txt F: fs/befs/ -- cgit v0.10.2 From ef419398b68925f21fd3d8463c7bf6934d2ec926 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Tue, 2 Aug 2016 14:03:19 -0700 Subject: proc_oom_score: remove tasklist_lock and pid_alive() This was needed before to ensure that ->signal != 0 and do_each_thread() is safe, see commit b95c35e76b29b ("oom: fix the unsafe usage of badness() in proc_oom_score()") for details. Today tsk->signal can't go away and for_each_thread(tsk) is always safe. Link: http://lkml.kernel.org/r/20160608211921.GA15508@redhat.com Signed-off-by: Oleg Nesterov Acked-by: David Rientjes Acked-by: Michal Hocko Cc: Tetsuo Handa Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/proc/base.c b/fs/proc/base.c index 31370da..54e2702 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -579,11 +579,8 @@ static int proc_oom_score(struct seq_file *m, struct pid_namespace *ns, unsigned long totalpages = totalram_pages + total_swap_pages; unsigned long points = 0; - read_lock(&tasklist_lock); - if (pid_alive(task)) - points = oom_badness(task, NULL, NULL, totalpages) * - 1000 / totalpages; - read_unlock(&tasklist_lock); + points = oom_badness(task, NULL, NULL, totalpages) * + 1000 / totalpages; seq_printf(m, "%lu\n", points); return 0; -- cgit v0.10.2 From 519ded5a89ec0e46e5b0867ba9f5752239b73898 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 2 Aug 2016 14:03:22 -0700 Subject: procfs: avoid 32-bit time_t in /proc/*/stat /proc/stat shows (among lots of other things) the current boottime (i.e. number of seconds since boot). While a 32-bit number is sufficient for this particular case, we want to get rid of the 'struct timespec' suffers from a 32-bit overflow in 2038. This changes the code to use a struct timespec64, which is known to be safe in all cases. Link: http://lkml.kernel.org/r/20160617201247.2292101-1-arnd@arndb.de Signed-off-by: Arnd Bergmann Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/proc/stat.c b/fs/proc/stat.c index 510413eb..7907e45 100644 --- a/fs/proc/stat.c +++ b/fs/proc/stat.c @@ -80,19 +80,17 @@ static u64 get_iowait_time(int cpu) static int show_stat(struct seq_file *p, void *v) { int i, j; - unsigned long jif; u64 user, nice, system, idle, iowait, irq, softirq, steal; u64 guest, guest_nice; u64 sum = 0; u64 sum_softirq = 0; unsigned int per_softirq_sums[NR_SOFTIRQS] = {0}; - struct timespec boottime; + struct timespec64 boottime; user = nice = system = idle = iowait = irq = softirq = steal = 0; guest = guest_nice = 0; - getboottime(&boottime); - jif = boottime.tv_sec; + getboottime64(&boottime); for_each_possible_cpu(i) { user += kcpustat_cpu(i).cpustat[CPUTIME_USER]; @@ -163,12 +161,12 @@ static int show_stat(struct seq_file *p, void *v) seq_printf(p, "\nctxt %llu\n" - "btime %lu\n" + "btime %llu\n" "processes %lu\n" "procs_running %lu\n" "procs_blocked %lu\n", nr_context_switches(), - (unsigned long)jif, + (unsigned long long)boottime.tv_sec, total_forks, nr_running(), nr_iowait()); -- cgit v0.10.2 From ca52953f5f24aff0aa8fa8de750b76ba0302142d Mon Sep 17 00:00:00 2001 From: Valdis Kletnieks Date: Tue, 2 Aug 2016 14:03:25 -0700 Subject: fs/proc/task_mmu.c: suppress compilation warnings with W=1 Suppress a bunch of warnings of the form: fs/proc/task_mmu.c: In function 'show_smap_vma_flags': fs/proc/task_mmu.c:635:22: warning: initialized field overwritten [-Wt override-init] [ilog2(VM_READ)] = "rd", ^~~~ fs/proc/task_mmu.c:635:22: note: (near initialization for 'mnemonics[0]') They happen because of the way we intentionally build the table, so silence the warning when building with 'make W=1'. Link: http://lkml.kernel.org/r/8727.1470022083@turing-police.cc.vt.edu Signed-off-by: Valdis Kletnieks Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/proc/Makefile b/fs/proc/Makefile index 7151ea4..a8c1360 100644 --- a/fs/proc/Makefile +++ b/fs/proc/Makefile @@ -4,6 +4,7 @@ obj-y += proc.o +CFLAGS_task_mmu.o += -Wno-override-init proc-y := nommu.o task_nommu.o proc-$(CONFIG_MMU) := task_mmu.o -- cgit v0.10.2 From bc083a64b6c035135c0f80718f9e9192cc0867c6 Mon Sep 17 00:00:00 2001 From: Richard Weinberger Date: Tue, 2 Aug 2016 14:03:27 -0700 Subject: init/Kconfig: make COMPILE_TEST depend on !UML UML is a bit special since it does not have iomem nor dma. That means a lot of drivers will not build if they miss a dependency on HAS_IOMEM. s390 used to have the same issues but since it gained PCI support UML is the only stranger. We are tired of patching dozens of new drivers after every merge window just to un-break allmod/yesconfig UML builds. One could argue that a decent driver has to know on what it depends and therefore a missing HAS_IOMEM dependency is a clear driver bug. But the dependency not obvious and not everyone does UML builds with COMPILE_TEST enabled when developing a device driver. A possible solution to make these builds succeed on UML would be providing stub functions for ioremap() and friends which fail upon runtime. Another one is simply disabling COMPILE_TEST for UML. Since it is the least hassle and does not force use to fake iomem support let's do the latter. Link: http://lkml.kernel.org/r/1466152995-28367-1-git-send-email-richard@nod.at Signed-off-by: Richard Weinberger Acked-by: Arnd Bergmann Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/init/Kconfig b/init/Kconfig index 79c6aad..8f08f49 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -55,6 +55,7 @@ config CROSS_COMPILE config COMPILE_TEST bool "Compile also drivers which will not load" + depends on !UML default n help Some drivers can be compiled on a different platform than they are -- cgit v0.10.2 From ca945e71529c69f71b773b31f03a681876872117 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Tue, 2 Aug 2016 14:03:30 -0700 Subject: memstick: don't allocate unused major for ms_block When alloc_disk(0) is used the ->major number is completely ignored. All devices are allocated with a major of BLOCK_EXT_MAJOR. So remove registration and deregistration of 'major'. Link: http://lkml.kernel.org/r/20160602064318.4403.49955.stgit@noble Signed-off-by: NeilBrown Cc: Keith Busch Cc: Jens Axboe Cc: Maxim Levitsky Cc: Greg Kroah-Hartman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/memstick/core/ms_block.c b/drivers/memstick/core/ms_block.c index 40bb8ae..aacf584 100644 --- a/drivers/memstick/core/ms_block.c +++ b/drivers/memstick/core/ms_block.c @@ -2338,23 +2338,11 @@ static struct memstick_driver msb_driver = { .resume = msb_resume }; -static int major; - static int __init msb_init(void) { - int rc = register_blkdev(0, DRIVER_NAME); - - if (rc < 0) { - pr_err("failed to register major (error %d)\n", rc); - return rc; - } - - major = rc; - rc = memstick_register_driver(&msb_driver); - if (rc) { - unregister_blkdev(major, DRIVER_NAME); + int rc = memstick_register_driver(&msb_driver); + if (rc) pr_err("failed to register memstick driver (error %d)\n", rc); - } return rc; } @@ -2362,7 +2350,6 @@ static int __init msb_init(void) static void __exit msb_exit(void) { memstick_unregister_driver(&msb_driver); - unregister_blkdev(major, DRIVER_NAME); idr_destroy(&msb_disk_idr); } -- cgit v0.10.2 From bd721ea73e1f965569b40620538c942001f76294 Mon Sep 17 00:00:00 2001 From: Fabian Frederick Date: Tue, 2 Aug 2016 14:03:33 -0700 Subject: treewide: replace obsolete _refok by __ref There was only one use of __initdata_refok and __exit_refok __init_refok was used 46 times against 82 for __ref. Those definitions are obsolete since commit 312b1485fb50 ("Introduce new section reference annotations tags: __ref, __refdata, __refconst") This patch removes the following compatibility definitions and replaces them treewide. /* compatibility defines */ #define __init_refok __ref #define __initdata_refok __refdata #define __exit_refok __ref I can also provide separate patches if necessary. (One patch per tree and check in 1 month or 2 to remove old definitions) [akpm@linux-foundation.org: coding-style fixes] Link: http://lkml.kernel.org/r/1466796271-3043-1-git-send-email-fabf@skynet.be Signed-off-by: Fabian Frederick Cc: Ingo Molnar Cc: Sam Ravnborg Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/alpha/kernel/machvec_impl.h b/arch/alpha/kernel/machvec_impl.h index f54bdf6..d3398f6 100644 --- a/arch/alpha/kernel/machvec_impl.h +++ b/arch/alpha/kernel/machvec_impl.h @@ -137,7 +137,7 @@ #define __initmv __initdata #define ALIAS_MV(x) #else -#define __initmv __initdata_refok +#define __initmv __refdata /* GCC actually has a syntax for defining aliases, but is under some delusion that you shouldn't be able to declare it extern somewhere diff --git a/arch/arc/mm/init.c b/arch/arc/mm/init.c index 8be9303..399e2f2 100644 --- a/arch/arc/mm/init.c +++ b/arch/arc/mm/init.c @@ -220,7 +220,7 @@ void __init mem_init(void) /* * free_initmem: Free all the __init memory. */ -void __init_refok free_initmem(void) +void __ref free_initmem(void) { free_initmem_default(-1); } diff --git a/arch/arm/mach-integrator/impd1.c b/arch/arm/mach-integrator/impd1.c index 38b0da3..ed9a014 100644 --- a/arch/arm/mach-integrator/impd1.c +++ b/arch/arm/mach-integrator/impd1.c @@ -320,11 +320,11 @@ static struct impd1_device impd1_devs[] = { #define IMPD1_VALID_IRQS 0x00000bffU /* - * As this module is bool, it is OK to have this as __init_refok() - no + * As this module is bool, it is OK to have this as __ref() - no * probe calls will be done after the initial system bootup, as devices * are discovered as part of the machine startup. */ -static int __init_refok impd1_probe(struct lm_device *dev) +static int __ref impd1_probe(struct lm_device *dev) { struct impd1_module *impd1; int irq_base; diff --git a/arch/arm/mach-mv78xx0/common.c b/arch/arm/mach-mv78xx0/common.c index 45a0520..6af5430 100644 --- a/arch/arm/mach-mv78xx0/common.c +++ b/arch/arm/mach-mv78xx0/common.c @@ -343,7 +343,7 @@ void __init mv78xx0_init_early(void) DDR_WINDOW_CPU1_BASE, DDR_WINDOW_CPU_SZ); } -void __init_refok mv78xx0_timer_init(void) +void __ref mv78xx0_timer_init(void) { orion_time_init(BRIDGE_VIRT_BASE, BRIDGE_INT_TIMER1_CLR, IRQ_MV78XX0_TIMER_1, get_tclk()); diff --git a/arch/blackfin/mm/init.c b/arch/blackfin/mm/init.c index 166842d..b59cd7c 100644 --- a/arch/blackfin/mm/init.c +++ b/arch/blackfin/mm/init.c @@ -112,7 +112,7 @@ void __init free_initrd_mem(unsigned long start, unsigned long end) } #endif -void __init_refok free_initmem(void) +void __ref free_initmem(void) { #if defined CONFIG_RAMKERNEL && !defined CONFIG_MPU free_initmem_default(-1); diff --git a/arch/hexagon/mm/init.c b/arch/hexagon/mm/init.c index 88977e4..192584d 100644 --- a/arch/hexagon/mm/init.c +++ b/arch/hexagon/mm/init.c @@ -93,7 +93,7 @@ void __init mem_init(void) * Todo: free pages between __init_begin and __init_end; possibly * some devtree related stuff as well. */ -void __init_refok free_initmem(void) +void __ref free_initmem(void) { } diff --git a/arch/ia64/kernel/mca.c b/arch/ia64/kernel/mca.c index 07a4e32..eb9220c 100644 --- a/arch/ia64/kernel/mca.c +++ b/arch/ia64/kernel/mca.c @@ -1831,7 +1831,7 @@ format_mca_init_stack(void *mca_data, unsigned long offset, } /* Caller prevents this from being called after init */ -static void * __init_refok mca_bootmem(void) +static void * __ref mca_bootmem(void) { return __alloc_bootmem(sizeof(struct ia64_mca_cpu), KERNEL_STACK_SIZE, 0); diff --git a/arch/microblaze/mm/init.c b/arch/microblaze/mm/init.c index 77bc7c7..434639f 100644 --- a/arch/microblaze/mm/init.c +++ b/arch/microblaze/mm/init.c @@ -414,7 +414,7 @@ void __init *early_get_page(void) #endif /* CONFIG_MMU */ -void * __init_refok alloc_maybe_bootmem(size_t size, gfp_t mask) +void * __ref alloc_maybe_bootmem(size_t size, gfp_t mask) { if (mem_init_done) return kmalloc(size, mask); @@ -422,7 +422,7 @@ void * __init_refok alloc_maybe_bootmem(size_t size, gfp_t mask) return alloc_bootmem(size); } -void * __init_refok zalloc_maybe_bootmem(size_t size, gfp_t mask) +void * __ref zalloc_maybe_bootmem(size_t size, gfp_t mask) { void *p; diff --git a/arch/microblaze/mm/pgtable.c b/arch/microblaze/mm/pgtable.c index eb99fcc..cc732fe 100644 --- a/arch/microblaze/mm/pgtable.c +++ b/arch/microblaze/mm/pgtable.c @@ -234,7 +234,7 @@ unsigned long iopa(unsigned long addr) return pa; } -__init_refok pte_t *pte_alloc_one_kernel(struct mm_struct *mm, +__ref pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address) { pte_t *pte; diff --git a/arch/mips/mm/init.c b/arch/mips/mm/init.c index 9b58eb5..a5509e7 100644 --- a/arch/mips/mm/init.c +++ b/arch/mips/mm/init.c @@ -504,7 +504,7 @@ void free_initrd_mem(unsigned long start, unsigned long end) void (*free_init_pages_eva)(void *begin, void *end) = NULL; -void __init_refok free_initmem(void) +void __ref free_initmem(void) { prom_free_prom_memory(); /* diff --git a/arch/mips/txx9/generic/pci.c b/arch/mips/txx9/generic/pci.c index a77698f..1f6bc9a 100644 --- a/arch/mips/txx9/generic/pci.c +++ b/arch/mips/txx9/generic/pci.c @@ -268,7 +268,7 @@ static int txx9_i8259_irq_setup(int irq) return err; } -static void __init_refok quirk_slc90e66_bridge(struct pci_dev *dev) +static void __ref quirk_slc90e66_bridge(struct pci_dev *dev) { int irq; /* PCI/ISA Bridge interrupt */ u8 reg_64; diff --git a/arch/nios2/mm/init.c b/arch/nios2/mm/init.c index e75c75d..c92fe42 100644 --- a/arch/nios2/mm/init.c +++ b/arch/nios2/mm/init.c @@ -89,7 +89,7 @@ void __init free_initrd_mem(unsigned long start, unsigned long end) } #endif -void __init_refok free_initmem(void) +void __ref free_initmem(void) { free_initmem_default(-1); } diff --git a/arch/openrisc/mm/ioremap.c b/arch/openrisc/mm/ioremap.c index 5b2a9511..fa60b81 100644 --- a/arch/openrisc/mm/ioremap.c +++ b/arch/openrisc/mm/ioremap.c @@ -38,7 +38,7 @@ static unsigned int fixmaps_used __initdata; * have to convert them into an offset in a page-aligned mapping, but the * caller shouldn't need to know that small detail. */ -void __iomem *__init_refok +void __iomem *__ref __ioremap(phys_addr_t addr, unsigned long size, pgprot_t prot) { phys_addr_t p; @@ -116,7 +116,7 @@ void iounmap(void *addr) * the memblock infrastructure. */ -pte_t __init_refok *pte_alloc_one_kernel(struct mm_struct *mm, +pte_t __ref *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address) { pte_t *pte; diff --git a/arch/powerpc/lib/alloc.c b/arch/powerpc/lib/alloc.c index 60b0b3f..a58abe4 100644 --- a/arch/powerpc/lib/alloc.c +++ b/arch/powerpc/lib/alloc.c @@ -6,7 +6,7 @@ #include -void * __init_refok zalloc_maybe_bootmem(size_t size, gfp_t mask) +void * __ref zalloc_maybe_bootmem(size_t size, gfp_t mask) { void *p; diff --git a/arch/powerpc/mm/pgtable_32.c b/arch/powerpc/mm/pgtable_32.c index 7f922f5..0ae0572 100644 --- a/arch/powerpc/mm/pgtable_32.c +++ b/arch/powerpc/mm/pgtable_32.c @@ -79,7 +79,7 @@ void pgd_free(struct mm_struct *mm, pgd_t *pgd) #endif } -__init_refok pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address) +__ref pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address) { pte_t *pte; diff --git a/arch/powerpc/platforms/powermac/setup.c b/arch/powerpc/platforms/powermac/setup.c index 3de4a7c..6b4e9d18 100644 --- a/arch/powerpc/platforms/powermac/setup.c +++ b/arch/powerpc/platforms/powermac/setup.c @@ -353,12 +353,12 @@ static int pmac_late_init(void) machine_late_initcall(powermac, pmac_late_init); /* - * This is __init_refok because we check for "initializing" before + * This is __ref because we check for "initializing" before * touching any of the __init sensitive things and "initializing" * will be false after __init time. This can't be __init because it * can be called whenever a disk is first accessed. */ -void __init_refok note_bootable_part(dev_t dev, int part, int goodness) +void __ref note_bootable_part(dev_t dev, int part, int goodness) { char *p; diff --git a/arch/powerpc/platforms/ps3/device-init.c b/arch/powerpc/platforms/ps3/device-init.c index 3f175e8..57caaf1 100644 --- a/arch/powerpc/platforms/ps3/device-init.c +++ b/arch/powerpc/platforms/ps3/device-init.c @@ -189,7 +189,7 @@ fail_malloc: return result; } -static int __init_refok ps3_setup_uhc_device( +static int __ref ps3_setup_uhc_device( const struct ps3_repository_device *repo, enum ps3_match_id match_id, enum ps3_interrupt_type interrupt_type, enum ps3_reg_type reg_type) { diff --git a/arch/powerpc/sysdev/msi_bitmap.c b/arch/powerpc/sysdev/msi_bitmap.c index ed5234e..5ebd3f0 100644 --- a/arch/powerpc/sysdev/msi_bitmap.c +++ b/arch/powerpc/sysdev/msi_bitmap.c @@ -112,7 +112,7 @@ int msi_bitmap_reserve_dt_hwirqs(struct msi_bitmap *bmp) return 0; } -int __init_refok msi_bitmap_alloc(struct msi_bitmap *bmp, unsigned int irq_count, +int __ref msi_bitmap_alloc(struct msi_bitmap *bmp, unsigned int irq_count, struct device_node *of_node) { int size; diff --git a/arch/score/mm/init.c b/arch/score/mm/init.c index 9fbce49..444c26c 100644 --- a/arch/score/mm/init.c +++ b/arch/score/mm/init.c @@ -91,7 +91,7 @@ void free_initrd_mem(unsigned long start, unsigned long end) } #endif -void __init_refok free_initmem(void) +void __ref free_initmem(void) { free_initmem_default(POISON_FREE_INITMEM); } diff --git a/arch/sh/drivers/pci/pci.c b/arch/sh/drivers/pci/pci.c index d5462b7..84563e3 100644 --- a/arch/sh/drivers/pci/pci.c +++ b/arch/sh/drivers/pci/pci.c @@ -221,7 +221,7 @@ pcibios_bus_report_status_early(struct pci_channel *hose, * We can't use pci_find_device() here since we are * called from interrupt context. */ -static void __init_refok +static void __ref pcibios_bus_report_status(struct pci_bus *bus, unsigned int status_mask, int warn) { @@ -256,7 +256,7 @@ pcibios_bus_report_status(struct pci_bus *bus, unsigned int status_mask, pcibios_bus_report_status(dev->subordinate, status_mask, warn); } -void __init_refok pcibios_report_status(unsigned int status_mask, int warn) +void __ref pcibios_report_status(unsigned int status_mask, int warn) { struct pci_channel *hose; diff --git a/arch/sh/mm/ioremap.c b/arch/sh/mm/ioremap.c index 0c99ec2..d09ddfe 100644 --- a/arch/sh/mm/ioremap.c +++ b/arch/sh/mm/ioremap.c @@ -34,7 +34,7 @@ * have to convert them into an offset in a page-aligned mapping, but the * caller shouldn't need to know that small detail. */ -void __iomem * __init_refok +void __iomem * __ref __ioremap_caller(phys_addr_t phys_addr, unsigned long size, pgprot_t pgprot, void *caller) { diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c index fb4c1b4..6209289 100644 --- a/arch/x86/mm/init.c +++ b/arch/x86/mm/init.c @@ -208,7 +208,7 @@ static int __meminit save_mr(struct map_range *mr, int nr_range, * adjust the page_size_mask for small range to go with * big page size instead small one if nearby are ram too. */ -static void __init_refok adjust_range_page_size_mask(struct map_range *mr, +static void __ref adjust_range_page_size_mask(struct map_range *mr, int nr_range) { int i; @@ -396,7 +396,7 @@ bool pfn_range_is_mapped(unsigned long start_pfn, unsigned long end_pfn) * This runs before bootmem is initialized and gets pages directly from * the physical memory. To access them they are temporarily mapped. */ -unsigned long __init_refok init_memory_mapping(unsigned long start, +unsigned long __ref init_memory_mapping(unsigned long start, unsigned long end) { struct map_range mr[NR_RANGE_MR]; diff --git a/arch/x86/platform/efi/early_printk.c b/arch/x86/platform/efi/early_printk.c index 5241421..5fdacb3 100644 --- a/arch/x86/platform/efi/early_printk.c +++ b/arch/x86/platform/efi/early_printk.c @@ -44,7 +44,7 @@ early_initcall(early_efi_map_fb); * In case earlyprintk=efi,keep we have the whole framebuffer mapped already * so just return the offset efi_fb + start. */ -static __init_refok void *early_efi_map(unsigned long start, unsigned long len) +static __ref void *early_efi_map(unsigned long start, unsigned long len) { unsigned long base; @@ -56,7 +56,7 @@ static __init_refok void *early_efi_map(unsigned long start, unsigned long len) return early_ioremap(base + start, len); } -static __init_refok void early_efi_unmap(void *addr, unsigned long len) +static __ref void early_efi_unmap(void *addr, unsigned long len) { if (!efi_fb) early_iounmap(addr, len); diff --git a/drivers/acpi/osl.c b/drivers/acpi/osl.c index b108f13..4305ee9 100644 --- a/drivers/acpi/osl.c +++ b/drivers/acpi/osl.c @@ -309,7 +309,7 @@ static void acpi_unmap(acpi_physical_address pg_off, void __iomem *vaddr) * During early init (when acpi_gbl_permanent_mmap has not been set yet) this * routine simply calls __acpi_map_table() to get the job done. */ -void __iomem *__init_refok +void __iomem *__ref acpi_os_map_iomem(acpi_physical_address phys, acpi_size size) { struct acpi_ioremap *map; @@ -362,8 +362,7 @@ out: } EXPORT_SYMBOL_GPL(acpi_os_map_iomem); -void *__init_refok -acpi_os_map_memory(acpi_physical_address phys, acpi_size size) +void *__ref acpi_os_map_memory(acpi_physical_address phys, acpi_size size) { return (void *)acpi_os_map_iomem(phys, size); } diff --git a/drivers/base/node.c b/drivers/base/node.c index 29cd966..5548f96 100644 --- a/drivers/base/node.c +++ b/drivers/base/node.c @@ -370,7 +370,7 @@ int unregister_cpu_under_node(unsigned int cpu, unsigned int nid) #ifdef CONFIG_MEMORY_HOTPLUG_SPARSE #define page_initialized(page) (page->lru.next) -static int __init_refok get_nid_for_pfn(unsigned long pfn) +static int __ref get_nid_for_pfn(unsigned long pfn) { struct page *page; diff --git a/drivers/clk/clkdev.c b/drivers/clk/clkdev.c index 89cc700..97ae60f 100644 --- a/drivers/clk/clkdev.c +++ b/drivers/clk/clkdev.c @@ -250,7 +250,7 @@ struct clk_lookup_alloc { char con_id[MAX_CON_ID]; }; -static struct clk_lookup * __init_refok +static struct clk_lookup * __ref vclkdev_alloc(struct clk_hw *hw, const char *con_id, const char *dev_fmt, va_list ap) { @@ -287,7 +287,7 @@ vclkdev_create(struct clk_hw *hw, const char *con_id, const char *dev_fmt, return cl; } -struct clk_lookup * __init_refok +struct clk_lookup * __ref clkdev_alloc(struct clk *clk, const char *con_id, const char *dev_fmt, ...) { struct clk_lookup *cl; diff --git a/drivers/pci/xen-pcifront.c b/drivers/pci/xen-pcifront.c index 5f70fee..d6ff5e8 100644 --- a/drivers/pci/xen-pcifront.c +++ b/drivers/pci/xen-pcifront.c @@ -1086,7 +1086,7 @@ out: return err; } -static void __init_refok pcifront_backend_changed(struct xenbus_device *xdev, +static void __ref pcifront_backend_changed(struct xenbus_device *xdev, enum xenbus_state be_state) { struct pcifront_device *pdev = dev_get_drvdata(&xdev->dev); diff --git a/drivers/video/logo/logo.c b/drivers/video/logo/logo.c index 10fbfd8..b6bc4a0 100644 --- a/drivers/video/logo/logo.c +++ b/drivers/video/logo/logo.c @@ -36,11 +36,11 @@ static int __init fb_logo_late_init(void) late_initcall(fb_logo_late_init); -/* logo's are marked __initdata. Use __init_refok to tell +/* logo's are marked __initdata. Use __ref to tell * modpost that it is intended that this function uses data * marked __initdata. */ -const struct linux_logo * __init_refok fb_find_logo(int depth) +const struct linux_logo * __ref fb_find_logo(int depth) { const struct linux_logo *logo = NULL; diff --git a/include/acpi/acpi_io.h b/include/acpi/acpi_io.h index dd86c5f..d7d0f49 100644 --- a/include/acpi/acpi_io.h +++ b/include/acpi/acpi_io.h @@ -13,7 +13,7 @@ static inline void __iomem *acpi_os_ioremap(acpi_physical_address phys, } #endif -void __iomem *__init_refok +void __iomem *__ref acpi_os_map_iomem(acpi_physical_address phys, acpi_size size); void __ref acpi_os_unmap_iomem(void __iomem *virt, acpi_size size); void __iomem *acpi_os_get_iomem(acpi_physical_address phys, unsigned int size); diff --git a/include/linux/init.h b/include/linux/init.h index aedb254..6935d02 100644 --- a/include/linux/init.h +++ b/include/linux/init.h @@ -77,12 +77,6 @@ #define __refdata __section(.ref.data) #define __refconst __constsection(.ref.rodata) -/* compatibility defines */ -#define __init_refok __ref -#define __initdata_refok __refdata -#define __exit_refok __ref - - #ifdef MODULE #define __exitused #else diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 4089abc..0933c74 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -275,7 +275,7 @@ static inline struct net *read_pnet(const possible_net_t *pnet) #define __net_initconst #else #define __net_init __init -#define __net_exit __exit_refok +#define __net_exit __ref #define __net_initdata __initdata #define __net_initconst __initconst #endif diff --git a/init/main.c b/init/main.c index eae02aa..e7345dc 100644 --- a/init/main.c +++ b/init/main.c @@ -380,7 +380,7 @@ static void __init setup_command_line(char *command_line) static __initdata DECLARE_COMPLETION(kthreadd_done); -static noinline void __init_refok rest_init(void) +static noinline void __ref rest_init(void) { int pid; diff --git a/mm/page_alloc.c b/mm/page_alloc.c index ea759b9..39a372a 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -5276,7 +5276,7 @@ void __init setup_per_cpu_pageset(void) setup_zone_pageset(zone); } -static noinline __init_refok +static noinline __ref int zone_wait_table_init(struct zone *zone, unsigned long zone_size_pages) { int i; @@ -5903,7 +5903,7 @@ static void __paginginit free_area_init_core(struct pglist_data *pgdat) } } -static void __init_refok alloc_node_mem_map(struct pglist_data *pgdat) +static void __ref alloc_node_mem_map(struct pglist_data *pgdat) { unsigned long __maybe_unused start = 0; unsigned long __maybe_unused offset = 0; diff --git a/mm/slab.c b/mm/slab.c index ca135bd..261147b 100644 --- a/mm/slab.c +++ b/mm/slab.c @@ -1877,7 +1877,7 @@ static struct array_cache __percpu *alloc_kmem_cache_cpus( return cpu_cache; } -static int __init_refok setup_cpu_cache(struct kmem_cache *cachep, gfp_t gfp) +static int __ref setup_cpu_cache(struct kmem_cache *cachep, gfp_t gfp) { if (slab_state >= FULL) return enable_cpucache(cachep, gfp); diff --git a/mm/sparse-vmemmap.c b/mm/sparse-vmemmap.c index 68885dc..574c67b 100644 --- a/mm/sparse-vmemmap.c +++ b/mm/sparse-vmemmap.c @@ -36,7 +36,7 @@ * Uses the main allocators if they are available, else bootmem. */ -static void * __init_refok __earlyonly_bootmem_alloc(int node, +static void * __ref __earlyonly_bootmem_alloc(int node, unsigned long size, unsigned long align, unsigned long goal) diff --git a/mm/sparse.c b/mm/sparse.c index 36d7bbb..1e168bf 100644 --- a/mm/sparse.c +++ b/mm/sparse.c @@ -59,7 +59,7 @@ static inline void set_section_nid(unsigned long section_nr, int nid) #endif #ifdef CONFIG_SPARSEMEM_EXTREME -static struct mem_section noinline __init_refok *sparse_index_alloc(int nid) +static noinline struct mem_section __ref *sparse_index_alloc(int nid) { struct mem_section *section = NULL; unsigned long array_size = SECTIONS_PER_ROOT * -- cgit v0.10.2 From db3f60012482756f46cc4d7d9ad7d793ae30360c Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Tue, 2 Aug 2016 14:03:36 -0700 Subject: uapi: move forward declarations of internal structures Don't user forward declarations of internal kernel structures in headers exported to userspace. Move "struct completion;". Move "struct task_struct;". Link: http://lkml.kernel.org/r/20160713215808.GA22486@p183.telecom.by Signed-off-by: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/capability.h b/include/linux/capability.h index 5f3c63d..dbc21c7 100644 --- a/include/linux/capability.h +++ b/include/linux/capability.h @@ -38,6 +38,7 @@ struct cpu_vfs_cap_data { struct file; struct inode; struct dentry; +struct task_struct; struct user_namespace; extern const kernel_cap_t __cap_empty_set; diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h index fa7bc29..697e160 100644 --- a/include/linux/sysctl.h +++ b/include/linux/sysctl.h @@ -28,6 +28,7 @@ #include /* For the /proc/sys support */ +struct completion; struct ctl_table; struct nsproxy; struct ctl_table_root; diff --git a/include/uapi/linux/capability.h b/include/uapi/linux/capability.h index 12c37a1..49bc062 100644 --- a/include/uapi/linux/capability.h +++ b/include/uapi/linux/capability.h @@ -15,8 +15,6 @@ #include -struct task_struct; - /* User-level do most of the mapping between kernel and user capabilities based on the version tag given by the kernel. The kernel might be somewhat backwards compatible, but don't bet on diff --git a/include/uapi/linux/sysctl.h b/include/uapi/linux/sysctl.h index 0956373..d2b1215 100644 --- a/include/uapi/linux/sysctl.h +++ b/include/uapi/linux/sysctl.h @@ -26,8 +26,6 @@ #include #include -struct completion; - #define CTL_MAXNAME 10 /* how many path components do we allow in a call to sysctl? In other words, what is the largest acceptable value for the nlen -- cgit v0.10.2 From bd804ba12525c13a096f64f305a169c654a706e7 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Linus=20L=C3=BCssing?= Date: Tue, 2 Aug 2016 14:03:39 -0700 Subject: =?UTF-8?q?mailmap:=20add=20Linus=20L=C3=BCssing?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit For one thing, summarizes all non-umlaut versions into the umlaut one (Linus Luessing -> Linus Lüssing). For another, maps obsolete email addresses to the current @c0d3.blue one. Link: http://lkml.kernel.org/r/1467805371-2773-1-git-send-email-linus.luessing@c0d3.blue Signed-off-by: Linus Lüssing Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/.mailmap b/.mailmap index d2acafb..c0d5704 100644 --- a/.mailmap +++ b/.mailmap @@ -92,6 +92,8 @@ Krzysztof Kozlowski Kuninori Morimoto Leonid I Ananiev Linas Vepstas +Linus Lüssing +Linus Lüssing Mark Brown Matthieu CASTET Mauro Carvalho Chehab -- cgit v0.10.2 From 949bed2f5764435715e3d6dd3ab6dd4dbd890a71 Mon Sep 17 00:00:00 2001 From: Chen Gang Date: Tue, 2 Aug 2016 14:03:42 -0700 Subject: include: mman: use bool instead of int for the return value of arch_validate_prot For pure bool function's return value, bool is a little better more or less than int. Link: http://lkml.kernel.org/r/1469331815-2026-1-git-send-email-chengang@emindsoft.com.cn Signed-off-by: Chen Gang Acked-by: Michael Ellerman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/powerpc/include/asm/mman.h b/arch/powerpc/include/asm/mman.h index 2563c43..fc420ce 100644 --- a/arch/powerpc/include/asm/mman.h +++ b/arch/powerpc/include/asm/mman.h @@ -31,13 +31,13 @@ static inline pgprot_t arch_vm_get_page_prot(unsigned long vm_flags) } #define arch_vm_get_page_prot(vm_flags) arch_vm_get_page_prot(vm_flags) -static inline int arch_validate_prot(unsigned long prot) +static inline bool arch_validate_prot(unsigned long prot) { if (prot & ~(PROT_READ | PROT_WRITE | PROT_EXEC | PROT_SEM | PROT_SAO)) - return 0; + return false; if ((prot & PROT_SAO) && !cpu_has_feature(CPU_FTR_SAO)) - return 0; - return 1; + return false; + return true; } #define arch_validate_prot(prot) arch_validate_prot(prot) diff --git a/include/linux/mman.h b/include/linux/mman.h index 33e17f6..634c4c5 100644 --- a/include/linux/mman.h +++ b/include/linux/mman.h @@ -49,7 +49,7 @@ static inline void vm_unacct_memory(long pages) * * Returns true if the prot flags are valid */ -static inline int arch_validate_prot(unsigned long prot) +static inline bool arch_validate_prot(unsigned long prot) { return (prot & ~(PROT_READ | PROT_WRITE | PROT_EXEC | PROT_SEM)) == 0; } -- cgit v0.10.2 From 61e96496d3c949701a48b908f99f4ed891cd1101 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Tue, 2 Aug 2016 14:03:44 -0700 Subject: task_work: use READ_ONCE/lockless_dereference, avoid pi_lock if !task_works Change task_work_cancel() to use lockless_dereference(), this is what the code really wants but we didn't have this helper when it was written. Also add the fast-path task->task_works == NULL check, in the likely case this task has no pending works and we can avoid spin_lock(task->pi_lock). While at it, change other users of ACCESS_ONCE() to use READ_ONCE(). Link: http://lkml.kernel.org/r/20160610150042.GA13868@redhat.com Signed-off-by: Oleg Nesterov Cc: Andrea Parri Cc: Peter Zijlstra Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/task_work.c b/kernel/task_work.c index 6ab4842..d513051 100644 --- a/kernel/task_work.c +++ b/kernel/task_work.c @@ -29,7 +29,7 @@ task_work_add(struct task_struct *task, struct callback_head *work, bool notify) struct callback_head *head; do { - head = ACCESS_ONCE(task->task_works); + head = READ_ONCE(task->task_works); if (unlikely(head == &work_exited)) return -ESRCH; work->next = head; @@ -57,6 +57,9 @@ task_work_cancel(struct task_struct *task, task_work_func_t func) struct callback_head **pprev = &task->task_works; struct callback_head *work; unsigned long flags; + + if (likely(!task->task_works)) + return NULL; /* * If cmpxchg() fails we continue without updating pprev. * Either we raced with task_work_add() which added the @@ -64,8 +67,7 @@ task_work_cancel(struct task_struct *task, task_work_func_t func) * we raced with task_work_run(), *pprev == NULL/exited. */ raw_spin_lock_irqsave(&task->pi_lock, flags); - while ((work = ACCESS_ONCE(*pprev))) { - smp_read_barrier_depends(); + while ((work = lockless_dereference(*pprev))) { if (work->func != func) pprev = &work->next; else if (cmpxchg(pprev, work, work->next) == work) @@ -95,7 +97,7 @@ void task_work_run(void) * work_exited unless the list is empty. */ do { - work = ACCESS_ONCE(task->task_works); + work = READ_ONCE(task->task_works); head = !work && (task->flags & PF_EXITING) ? &work_exited : NULL; } while (cmpxchg(&task->task_works, work, head) != work); -- cgit v0.10.2 From 9d5059c959ac739dbf837cec14586e58e7a67292 Mon Sep 17 00:00:00 2001 From: Luis de Bethencourt Date: Tue, 2 Aug 2016 14:03:47 -0700 Subject: dynamic_debug: only add header when used kernel.h header doesn't directly use dynamic debug, instead we can include it in module.c (which used it via kernel.h). printk.h only uses it if CONFIG_DYNAMIC_DEBUG is on, changing the inclusion to only happen in that case. Link: http://lkml.kernel.org/r/1468429793-16917-1-git-send-email-luisbg@osg.samsung.com [luisbg@osg.samsung.com: include dynamic_debug.h in drb_int.h] Link: http://lkml.kernel.org/r/1468447828-18558-2-git-send-email-luisbg@osg.samsung.com Signed-off-by: Luis de Bethencourt Cc: Rusty Russell Cc: Hidehiro Kawai Cc: Borislav Petkov Cc: Michal Nazarewicz Cc: Rasmus Villemoes Cc: Joe Perches Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/block/drbd/drbd_actlog.c b/drivers/block/drbd/drbd_actlog.c index 0a1aaf8..2d3d50a 100644 --- a/drivers/block/drbd/drbd_actlog.c +++ b/drivers/block/drbd/drbd_actlog.c @@ -27,7 +27,6 @@ #include #include #include -#include #include "drbd_int.h" diff --git a/drivers/block/drbd/drbd_int.h b/drivers/block/drbd/drbd_int.h index 7b54354..4cb8f21 100644 --- a/drivers/block/drbd/drbd_int.h +++ b/drivers/block/drbd/drbd_int.h @@ -41,6 +41,7 @@ #include #include #include +#include #include #include #include diff --git a/include/linux/kernel.h b/include/linux/kernel.h index c420821..d96a611 100644 --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -11,7 +11,6 @@ #include #include #include -#include #include #include diff --git a/include/linux/printk.h b/include/linux/printk.h index f136b22..987c65e 100644 --- a/include/linux/printk.h +++ b/include/linux/printk.h @@ -289,10 +289,11 @@ extern asmlinkage void dump_stack(void) __cold; no_printk(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__) #endif -#include /* If you are writing a driver, please use dev_dbg instead */ #if defined(CONFIG_DYNAMIC_DEBUG) +#include + /* dynamic_pr_debug() uses pr_fmt() internally so we don't need it here */ #define pr_debug(fmt, ...) \ dynamic_pr_debug(fmt, ##__VA_ARGS__) diff --git a/kernel/module.c b/kernel/module.c index 5f71aa6..a0f48b8 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -60,6 +60,7 @@ #include #include #include +#include #include #include "module-internal.h" -- cgit v0.10.2 From bebca05281d039e4144e1c51f402fd1d5f54b5e2 Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Tue, 2 Aug 2016 14:03:50 -0700 Subject: printk: do not include interrupt.h A trivial cosmetic change: interrupt.h header is redundant since commit 6b898c07cb1d ("console: use might_sleep in console_lock"). Link: http://lkml.kernel.org/r/20160620132847.21930-1-sergey.senozhatsky@gmail.com Signed-off-by: Sergey Senozhatsky Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index d4de339..09af62e 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -26,7 +26,6 @@ #include #include #include -#include /* For in_interrupt() */ #include #include #include -- cgit v0.10.2 From 874f9c7da9a4acbc1b9e12ca722579fb50e4d142 Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Tue, 2 Aug 2016 14:03:53 -0700 Subject: printk: create pr_ functions Using functions instead of macros can reduce overall code size by eliminating unnecessary "KERN_SOH" prefixes from format strings. defconfig x86-64: $ size vmlinux* text data bss dec hex filename 10193570 4331464 1105920 15630954 ee826a vmlinux.new 10192623 4335560 1105920 15634103 ee8eb7 vmlinux.old As the return value are unimportant and unused in the kernel tree, these new functions return void. Miscellanea: - change pr_ macros to call new __pr_ functions - change vprintk_nmi and vprintk_default to add LOGLEVEL_ argument [akpm@linux-foundation.org: fix LOGLEVEL_INFO, per Joe] Link: http://lkml.kernel.org/r/e16cc34479dfefcae37c98b481e6646f0f69efc3.1466718827.git.joe@perches.com Signed-off-by: Joe Perches Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/printk.h b/include/linux/printk.h index 987c65e..c2158f0 100644 --- a/include/linux/printk.h +++ b/include/linux/printk.h @@ -257,21 +257,39 @@ extern asmlinkage void dump_stack(void) __cold; * and other debug macros are compiled out unless either DEBUG is defined * or CONFIG_DYNAMIC_DEBUG is set. */ -#define pr_emerg(fmt, ...) \ - printk(KERN_EMERG pr_fmt(fmt), ##__VA_ARGS__) -#define pr_alert(fmt, ...) \ - printk(KERN_ALERT pr_fmt(fmt), ##__VA_ARGS__) -#define pr_crit(fmt, ...) \ - printk(KERN_CRIT pr_fmt(fmt), ##__VA_ARGS__) -#define pr_err(fmt, ...) \ - printk(KERN_ERR pr_fmt(fmt), ##__VA_ARGS__) -#define pr_warning(fmt, ...) \ - printk(KERN_WARNING pr_fmt(fmt), ##__VA_ARGS__) -#define pr_warn pr_warning -#define pr_notice(fmt, ...) \ - printk(KERN_NOTICE pr_fmt(fmt), ##__VA_ARGS__) -#define pr_info(fmt, ...) \ - printk(KERN_INFO pr_fmt(fmt), ##__VA_ARGS__) + +#ifdef CONFIG_PRINTK + +asmlinkage __printf(1, 2) __cold void __pr_emerg(const char *fmt, ...); +asmlinkage __printf(1, 2) __cold void __pr_alert(const char *fmt, ...); +asmlinkage __printf(1, 2) __cold void __pr_crit(const char *fmt, ...); +asmlinkage __printf(1, 2) __cold void __pr_err(const char *fmt, ...); +asmlinkage __printf(1, 2) __cold void __pr_warn(const char *fmt, ...); +asmlinkage __printf(1, 2) __cold void __pr_notice(const char *fmt, ...); +asmlinkage __printf(1, 2) __cold void __pr_info(const char *fmt, ...); + +#define pr_emerg(fmt, ...) __pr_emerg(pr_fmt(fmt), ##__VA_ARGS__) +#define pr_alert(fmt, ...) __pr_alert(pr_fmt(fmt), ##__VA_ARGS__) +#define pr_crit(fmt, ...) __pr_crit(pr_fmt(fmt), ##__VA_ARGS__) +#define pr_err(fmt, ...) __pr_err(pr_fmt(fmt), ##__VA_ARGS__) +#define pr_warn(fmt, ...) __pr_warn(pr_fmt(fmt), ##__VA_ARGS__) +#define pr_notice(fmt, ...) __pr_notice(pr_fmt(fmt), ##__VA_ARGS__) +#define pr_info(fmt, ...) __pr_info(pr_fmt(fmt), ##__VA_ARGS__) + +#else + +#define pr_emerg(fmt, ...) printk(KERN_EMERG pr_fmt(fmt), ##__VA_ARGS__) +#define pr_alert(fmt, ...) printk(KERN_ALERT pr_fmt(fmt), ##__VA_ARGS__) +#define pr_crit(fmt, ...) printk(KERN_CRIT pr_fmt(fmt), ##__VA_ARGS__) +#define pr_err(fmt, ...) printk(KERN_ERR pr_fmt(fmt), ##__VA_ARGS__) +#define pr_warn(fmt, ...) printk(KERN_WARNING pr_fmt(fmt), ##__VA_ARGS__) +#define pr_notice(fmt, ...) printk(KERN_NOTICE pr_fmt(fmt), ##__VA_ARGS__) +#define pr_info(fmt, ...) printk(KERN_INFO pr_fmt(fmt), ##__VA_ARGS__) + +#endif + +#define pr_warning pr_warn + /* * Like KERN_CONT, pr_cont() should only be used when continuing * a line with no newline ('\n') enclosed. Otherwise it defaults diff --git a/kernel/printk/internal.h b/kernel/printk/internal.h index 7fd2838..5d4505f 100644 --- a/kernel/printk/internal.h +++ b/kernel/printk/internal.h @@ -16,9 +16,11 @@ */ #include -typedef __printf(1, 0) int (*printk_func_t)(const char *fmt, va_list args); +typedef __printf(2, 0) int (*printk_func_t)(int level, const char *fmt, + va_list args); -int __printf(1, 0) vprintk_default(const char *fmt, va_list args); +__printf(2, 0) +int vprintk_default(int level, const char *fmt, va_list args); #ifdef CONFIG_PRINTK_NMI @@ -31,9 +33,10 @@ extern raw_spinlock_t logbuf_lock; * via per-CPU variable. */ DECLARE_PER_CPU(printk_func_t, printk_func); -static inline __printf(1, 0) int vprintk_func(const char *fmt, va_list args) +__printf(2, 0) +static inline int vprintk_func(int level, const char *fmt, va_list args) { - return this_cpu_read(printk_func)(fmt, args); + return this_cpu_read(printk_func)(level, fmt, args); } extern atomic_t nmi_message_lost; @@ -44,9 +47,10 @@ static inline int get_nmi_message_lost(void) #else /* CONFIG_PRINTK_NMI */ -static inline __printf(1, 0) int vprintk_func(const char *fmt, va_list args) +__printf(2, 0) +static inline int vprintk_func(int level, const char *fmt, va_list args) { - return vprintk_default(fmt, args); + return vprintk_default(level, fmt, args); } static inline int get_nmi_message_lost(void) diff --git a/kernel/printk/nmi.c b/kernel/printk/nmi.c index b69eb8a..bc3eeb1 100644 --- a/kernel/printk/nmi.c +++ b/kernel/printk/nmi.c @@ -58,7 +58,7 @@ static DEFINE_PER_CPU(struct nmi_seq_buf, nmi_print_seq); * one writer running. But the buffer might get flushed from another * CPU, so we need to be careful. */ -static int vprintk_nmi(const char *fmt, va_list args) +static int vprintk_nmi(int level, const char *fmt, va_list args) { struct nmi_seq_buf *s = this_cpu_ptr(&nmi_print_seq); int add = 0; @@ -79,7 +79,16 @@ again: if (!len) smp_rmb(); - add = vsnprintf(s->buffer + len, sizeof(s->buffer) - len, fmt, args); + if (level != LOGLEVEL_DEFAULT) { + add = snprintf(s->buffer + len, sizeof(s->buffer) - len, + KERN_SOH "%c", '0' + level); + add += vsnprintf(s->buffer + len + add, + sizeof(s->buffer) - len - add, + fmt, args); + } else { + add = vsnprintf(s->buffer + len, sizeof(s->buffer) - len, + fmt, args); + } /* * Do it once again if the buffer has been flushed in the meantime. diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index 09af62e..d2accf2 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -1801,7 +1801,28 @@ asmlinkage int printk_emit(int facility, int level, } EXPORT_SYMBOL(printk_emit); -int vprintk_default(const char *fmt, va_list args) +#ifdef CONFIG_PRINTK +#define define_pr_level(func, loglevel) \ +asmlinkage __visible void func(const char *fmt, ...) \ +{ \ + va_list args; \ + \ + va_start(args, fmt); \ + vprintk_default(loglevel, fmt, args); \ + va_end(args); \ +} \ +EXPORT_SYMBOL(func) + +define_pr_level(__pr_emerg, LOGLEVEL_EMERG); +define_pr_level(__pr_alert, LOGLEVEL_ALERT); +define_pr_level(__pr_crit, LOGLEVEL_CRIT); +define_pr_level(__pr_err, LOGLEVEL_ERR); +define_pr_level(__pr_warn, LOGLEVEL_WARNING); +define_pr_level(__pr_notice, LOGLEVEL_NOTICE); +define_pr_level(__pr_info, LOGLEVEL_INFO); +#endif + +int vprintk_default(int level, const char *fmt, va_list args) { int r; @@ -1811,7 +1832,7 @@ int vprintk_default(const char *fmt, va_list args) return r; } #endif - r = vprintk_emit(0, LOGLEVEL_DEFAULT, NULL, 0, fmt, args); + r = vprintk_emit(0, level, NULL, 0, fmt, args); return r; } @@ -1844,7 +1865,7 @@ asmlinkage __visible int printk(const char *fmt, ...) int r; va_start(args, fmt); - r = vprintk_func(fmt, args); + r = vprintk_func(LOGLEVEL_DEFAULT, fmt, args); va_end(args); return r; -- cgit v0.10.2 From cf7754441c563230ed74096fcd4b8cca49910550 Mon Sep 17 00:00:00 2001 From: Sergey Senozhatsky Date: Tue, 2 Aug 2016 14:03:56 -0700 Subject: printk: introduce suppress_message_printing() Messages' levels and console log level are inspected when the actual printing occurs, which may provoke console_unlock() and console_cont_flush() to waste CPU cycles on every message that has loglevel above the current console_loglevel. Schematically, console_unlock() does the following: console_unlock() { ... for (;;) { ... raw_spin_lock_irqsave(&logbuf_lock, flags); skip: msg = log_from_idx(console_idx); if (msg->flags & LOG_NOCONS) { ... goto skip; } level = msg->level; len += msg_print_text(); >> sprintfs memcpy, etc. if (nr_ext_console_drivers) { ext_len = msg_print_ext_header(); >> scnprintf ext_len += msg_print_ext_body(); >> scnprintfs etc. } ... raw_spin_unlock(&logbuf_lock); call_console_drivers(level, ext_text, ext_len, text, len) { if (level >= console_loglevel && >> drop the message !ignore_loglevel) return; console->write(...); } local_irq_restore(flags); } ... } The thing here is this deferred `level >= console_loglevel' check. We are wasting CPU cycles on sprintfs/memcpy/etc. preparing the messages that we will eventually drop. This can be huge when we register a new CON_PRINTBUFFER console, for instance. For every such a console register_console() resets the console_seq, console_idx, console_prev and sets a `exclusive console' pointer to replay the log buffer to that just-registered console. And there can be a lot of messages to replay, in the worst case most of which can be dropped after console_loglevel test. We know messages' levels long before we call msg_print_text() and friends, so we can just move console_loglevel check out of call_console_drivers() and format a new message only if we are sure that it won't be dropped. The patch factors out loglevel check into suppress_message_printing() function and tests message->level and console_loglevel before formatting functions in console_unlock() and console_cont_flush() are getting executed. This improves things not only for exclusive CON_PRINTBUFFER consoles, but for every console_unlock() that attempts to print a message of level above the console_loglevel. Link: http://lkml.kernel.org/r/20160627135012.8229-1-sergey.senozhatsky@gmail.com Signed-off-by: Sergey Senozhatsky Reviewed-by: Petr Mladek Cc: Tejun Heo Cc: Jan Kara Cc: Calvin Owens Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index d2accf2..8bdce14 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -985,6 +985,11 @@ module_param(ignore_loglevel, bool, S_IRUGO | S_IWUSR); MODULE_PARM_DESC(ignore_loglevel, "ignore loglevel setting (prints all kernel messages to the console)"); +static bool suppress_message_printing(int level) +{ + return (level >= console_loglevel && !ignore_loglevel); +} + #ifdef CONFIG_BOOT_PRINTK_DELAY static int boot_delay; /* msecs delay after each printk during bootup */ @@ -1014,7 +1019,7 @@ static void boot_delay_msec(int level) unsigned long timeout; if ((boot_delay == 0 || system_state != SYSTEM_BOOTING) - || (level >= console_loglevel && !ignore_loglevel)) { + || suppress_message_printing(level)) { return; } @@ -1438,8 +1443,6 @@ static void call_console_drivers(int level, trace_console(text, len); - if (level >= console_loglevel && !ignore_loglevel) - return; if (!console_drivers) return; @@ -1908,6 +1911,7 @@ static void call_console_drivers(int level, static size_t msg_print_text(const struct printk_log *msg, enum log_flags prev, bool syslog, char *buf, size_t size) { return 0; } static size_t cont_print_text(char *text, size_t size) { return 0; } +static bool suppress_message_printing(int level) { return false; } /* Still needs to be defined for users */ DEFINE_PER_CPU(printk_func_t, printk_func); @@ -2187,6 +2191,13 @@ static void console_cont_flush(char *text, size_t size) if (!cont.len) goto out; + if (suppress_message_printing(cont.level)) { + cont.cons = cont.len; + if (cont.flushed) + cont.len = 0; + goto out; + } + /* * We still queue earlier records, likely because the console was * busy. The earlier ones need to be printed before this one, we @@ -2290,10 +2301,13 @@ skip: break; msg = log_from_idx(console_idx); - if (msg->flags & LOG_NOCONS) { + level = msg->level; + if ((msg->flags & LOG_NOCONS) || + suppress_message_printing(level)) { /* * Skip record we have buffered and already printed - * directly to the console when we received it. + * directly to the console when we received it, and + * record that has level above the console loglevel. */ console_idx = log_next(console_idx); console_seq++; @@ -2307,7 +2321,6 @@ skip: goto skip; } - level = msg->level; len += msg_print_text(msg, console_prev, false, text + len, sizeof(text) - len); if (nr_ext_console_drivers) { -- cgit v0.10.2 From 40a7d9f5f90681c6d7890b6a07f230bb4afe7e39 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Tue, 2 Aug 2016 14:03:59 -0700 Subject: printk: include instead of asm-generic headers are generic implementations for architecture specific code and should not be included by common code. Thus use the asm/ version of sections.h to get at the linker sections. Link: http://lkml.kernel.org/r/1468285008-7331-1-git-send-email-hch@lst.de Signed-off-by: Christoph Hellwig Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index 8bdce14..70c66c5 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -47,7 +47,7 @@ #include #include -#include +#include #define CREATE_TRACE_POINTS #include -- cgit v0.10.2 From b5644a153d2701ffc335cfb9ef49967bd5b6a3c2 Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Tue, 2 Aug 2016 14:04:01 -0700 Subject: fbdev/bfin_adv7393fb: move DRIVER_NAME before its first use Move the DRIVER_NAME macro definition before the first usage site and fix build error. Link: http://lkml.kernel.org/r/20160801163937.GA28119@nazgul.tnic Signed-off-by: Borislav Petkov Reported-by: kbuild test robot Cc: Jean-Christophe Plagniol-Villard Cc: Tomi Valkeinen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/video/fbdev/bfin_adv7393fb.c b/drivers/video/fbdev/bfin_adv7393fb.c index 8fe41ca..e2d7d03 100644 --- a/drivers/video/fbdev/bfin_adv7393fb.c +++ b/drivers/video/fbdev/bfin_adv7393fb.c @@ -10,6 +10,8 @@ * TODO: Code Cleanup */ +#define DRIVER_NAME "bfin-adv7393" + #define pr_fmt(fmt) DRIVER_NAME ": " fmt #include diff --git a/drivers/video/fbdev/bfin_adv7393fb.h b/drivers/video/fbdev/bfin_adv7393fb.h index cd591b5..afd0380 100644 --- a/drivers/video/fbdev/bfin_adv7393fb.h +++ b/drivers/video/fbdev/bfin_adv7393fb.h @@ -59,8 +59,6 @@ enum { BLANK_OFF, }; -#define DRIVER_NAME "bfin-adv7393" - struct adv7393fb_modes { const s8 name[25]; /* Full name */ u16 xres; /* Active Horizonzal Pixels */ -- cgit v0.10.2 From 6b1d174b0c27b5de421eda55c2731f32b6bd9852 Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Tue, 2 Aug 2016 14:04:04 -0700 Subject: ratelimit: extend to print suppressed messages on release MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Extend the ratelimiting facility to print the amount of suppressed lines when it is being released. This use case is aimed at short-termed, burst-like users for which we want to output the suppressed lines stats only once, after it has been disposed of. For an example, see /dev/kmsg usage in a follow-on patch. Also, change the printk() line we issue on release to not use "callbacks" as it is misleading: we're not suppressing callbacks but printk() calls. This has been separated from a previous patch by Linus. Link: http://lkml.kernel.org/r/20160716061745.15795-2-bp@alien8.de Signed-off-by: Borislav Petkov Cc: Dave Young Cc: Franck Bui Cc: Greg Kroah-Hartman Cc: Ingo Molnar Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Steven Rostedt Cc: Uwe Kleine-König Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/ratelimit.h b/include/linux/ratelimit.h index 1810252..57c9e06 100644 --- a/include/linux/ratelimit.h +++ b/include/linux/ratelimit.h @@ -2,11 +2,15 @@ #define _LINUX_RATELIMIT_H #include +#include #include #define DEFAULT_RATELIMIT_INTERVAL (5 * HZ) #define DEFAULT_RATELIMIT_BURST 10 +/* issue num suppressed message on exit */ +#define RATELIMIT_MSG_ON_RELEASE BIT(0) + struct ratelimit_state { raw_spinlock_t lock; /* protect the state */ @@ -15,6 +19,7 @@ struct ratelimit_state { int printed; int missed; unsigned long begin; + unsigned long flags; }; #define RATELIMIT_STATE_INIT(name, interval_init, burst_init) { \ @@ -34,12 +39,35 @@ struct ratelimit_state { static inline void ratelimit_state_init(struct ratelimit_state *rs, int interval, int burst) { + memset(rs, 0, sizeof(*rs)); + raw_spin_lock_init(&rs->lock); - rs->interval = interval; - rs->burst = burst; - rs->printed = 0; - rs->missed = 0; - rs->begin = 0; + rs->interval = interval; + rs->burst = burst; +} + +static inline void ratelimit_default_init(struct ratelimit_state *rs) +{ + return ratelimit_state_init(rs, DEFAULT_RATELIMIT_INTERVAL, + DEFAULT_RATELIMIT_BURST); +} + +static inline void ratelimit_state_exit(struct ratelimit_state *rs) +{ + if (!(rs->flags & RATELIMIT_MSG_ON_RELEASE)) + return; + + if (rs->missed) { + pr_warn("%s: %d output lines suppressed due to ratelimiting\n", + current->comm, rs->missed); + rs->missed = 0; + } +} + +static inline void +ratelimit_set_flags(struct ratelimit_state *rs, unsigned long flags) +{ + rs->flags = flags; } extern struct ratelimit_state printk_ratelimit_state; diff --git a/lib/ratelimit.c b/lib/ratelimit.c index 2c5de86..08f8043 100644 --- a/lib/ratelimit.c +++ b/lib/ratelimit.c @@ -46,12 +46,14 @@ int ___ratelimit(struct ratelimit_state *rs, const char *func) rs->begin = jiffies; if (time_is_before_jiffies(rs->begin + rs->interval)) { - if (rs->missed) - printk(KERN_WARNING "%s: %d callbacks suppressed\n", - func, rs->missed); + if (rs->missed) { + if (!(rs->flags & RATELIMIT_MSG_ON_RELEASE)) { + pr_warn("%s: %d callbacks suppressed\n", func, rs->missed); + rs->missed = 0; + } + } rs->begin = jiffies; rs->printed = 0; - rs->missed = 0; } if (rs->burst && rs->burst > rs->printed) { rs->printed++; -- cgit v0.10.2 From 750afe7babd117daabebf4855da18e4418ea845e Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Tue, 2 Aug 2016 14:04:07 -0700 Subject: printk: add kernel parameter to control writes to /dev/kmsg MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Add a "printk.devkmsg" kernel command line parameter which controls how userspace writes into /dev/kmsg. It has three options: * ratelimit - ratelimit logging from userspace. * on - unlimited logging from userspace * off - logging from userspace gets ignored The default setting is to ratelimit the messages written to it. This changes the kernel default setting of "on" to "ratelimit" and we do that because we want to keep userspace spamming /dev/kmsg to sane levels. This is especially moot when a small kernel log buffer wraps around and messages get lost. So the ratelimiting setting should be a sane setting where kernel messages should have a bit higher chance of survival from all the spamming. It additionally does not limit logging to /dev/kmsg while the system is booting if we haven't disabled it on the command line. Furthermore, we can control the logging from a lower priority sysctl interface - kernel.printk_devkmsg. That interface will succeed only if printk.devkmsg *hasn't* been supplied on the command line. If it has, then printk.devkmsg is a one-time setting which remains for the duration of the system lifetime. This "locking" of the setting is to prevent userspace from changing the logging on us through sysctl(2). This patch is based on previous patches from Linus and Steven. [bp@suse.de: fixes] Link: http://lkml.kernel.org/r/20160719072344.GC25563@nazgul.tnic Link: http://lkml.kernel.org/r/20160716061745.15795-3-bp@alien8.de Signed-off-by: Borislav Petkov Cc: Dave Young Cc: Franck Bui Cc: Greg Kroah-Hartman Cc: Ingo Molnar Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Steven Rostedt Cc: Uwe Kleine-König Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt index e24aa11..b240540 100644 --- a/Documentation/kernel-parameters.txt +++ b/Documentation/kernel-parameters.txt @@ -3173,6 +3173,13 @@ bytes respectively. Such letter suffixes can also be entirely omitted. Format: (1/Y/y=enable, 0/N/n=disable) default: disabled + printk.devkmsg={on,off,ratelimit} + Control writing to /dev/kmsg. + on - unlimited logging to /dev/kmsg from userspace + off - logging to /dev/kmsg disabled + ratelimit - ratelimit the logging + Default: ratelimit + printk.time= Show timing data prefixed to each printk message line Format: (1/Y/y=enable, 0/N/n=disable) diff --git a/Documentation/sysctl/kernel.txt b/Documentation/sysctl/kernel.txt index 3320460..ffab8b5 100644 --- a/Documentation/sysctl/kernel.txt +++ b/Documentation/sysctl/kernel.txt @@ -764,6 +764,20 @@ send before ratelimiting kicks in. ============================================================== +printk_devkmsg: + +Control the logging to /dev/kmsg from userspace: + +ratelimit: default, ratelimited +on: unlimited logging to /dev/kmsg from userspace +off: logging to /dev/kmsg disabled + +The kernel command line parameter printk.devkmsg= overrides this and is +a one-time setting until next reboot: once set, it cannot be changed by +this sysctl interface anymore. + +============================================================== + randomize_va_space: This option can be used to select the type of process address diff --git a/include/linux/printk.h b/include/linux/printk.h index c2158f0..8dc155d 100644 --- a/include/linux/printk.h +++ b/include/linux/printk.h @@ -61,6 +61,11 @@ static inline void console_verbose(void) console_loglevel = CONSOLE_LOGLEVEL_MOTORMOUTH; } +/* strlen("ratelimit") + 1 */ +#define DEVKMSG_STR_MAX_SIZE 10 +extern char devkmsg_log_str[]; +struct ctl_table; + struct va_format { const char *fmt; va_list *va; @@ -175,6 +180,10 @@ extern int printk_delay_msec; extern int dmesg_restrict; extern int kptr_restrict; +extern int +devkmsg_sysctl_set_loglvl(struct ctl_table *table, int write, void __user *buf, + size_t *lenp, loff_t *ppos); + extern void wake_up_klogd(void); char *log_buf_addr_get(void); diff --git a/kernel/printk/printk.c b/kernel/printk/printk.c index 70c66c5..a5ef95c 100644 --- a/kernel/printk/printk.c +++ b/kernel/printk/printk.c @@ -85,6 +85,111 @@ static struct lockdep_map console_lock_dep_map = { }; #endif +enum devkmsg_log_bits { + __DEVKMSG_LOG_BIT_ON = 0, + __DEVKMSG_LOG_BIT_OFF, + __DEVKMSG_LOG_BIT_LOCK, +}; + +enum devkmsg_log_masks { + DEVKMSG_LOG_MASK_ON = BIT(__DEVKMSG_LOG_BIT_ON), + DEVKMSG_LOG_MASK_OFF = BIT(__DEVKMSG_LOG_BIT_OFF), + DEVKMSG_LOG_MASK_LOCK = BIT(__DEVKMSG_LOG_BIT_LOCK), +}; + +/* Keep both the 'on' and 'off' bits clear, i.e. ratelimit by default: */ +#define DEVKMSG_LOG_MASK_DEFAULT 0 + +static unsigned int __read_mostly devkmsg_log = DEVKMSG_LOG_MASK_DEFAULT; + +static int __control_devkmsg(char *str) +{ + if (!str) + return -EINVAL; + + if (!strncmp(str, "on", 2)) { + devkmsg_log = DEVKMSG_LOG_MASK_ON; + return 2; + } else if (!strncmp(str, "off", 3)) { + devkmsg_log = DEVKMSG_LOG_MASK_OFF; + return 3; + } else if (!strncmp(str, "ratelimit", 9)) { + devkmsg_log = DEVKMSG_LOG_MASK_DEFAULT; + return 9; + } + return -EINVAL; +} + +static int __init control_devkmsg(char *str) +{ + if (__control_devkmsg(str) < 0) + return 1; + + /* + * Set sysctl string accordingly: + */ + if (devkmsg_log == DEVKMSG_LOG_MASK_ON) { + memset(devkmsg_log_str, 0, DEVKMSG_STR_MAX_SIZE); + strncpy(devkmsg_log_str, "on", 2); + } else if (devkmsg_log == DEVKMSG_LOG_MASK_OFF) { + memset(devkmsg_log_str, 0, DEVKMSG_STR_MAX_SIZE); + strncpy(devkmsg_log_str, "off", 3); + } + /* else "ratelimit" which is set by default. */ + + /* + * Sysctl cannot change it anymore. The kernel command line setting of + * this parameter is to force the setting to be permanent throughout the + * runtime of the system. This is a precation measure against userspace + * trying to be a smarta** and attempting to change it up on us. + */ + devkmsg_log |= DEVKMSG_LOG_MASK_LOCK; + + return 0; +} +__setup("printk.devkmsg=", control_devkmsg); + +char devkmsg_log_str[DEVKMSG_STR_MAX_SIZE] = "ratelimit"; + +int devkmsg_sysctl_set_loglvl(struct ctl_table *table, int write, + void __user *buffer, size_t *lenp, loff_t *ppos) +{ + char old_str[DEVKMSG_STR_MAX_SIZE]; + unsigned int old; + int err; + + if (write) { + if (devkmsg_log & DEVKMSG_LOG_MASK_LOCK) + return -EINVAL; + + old = devkmsg_log; + strncpy(old_str, devkmsg_log_str, DEVKMSG_STR_MAX_SIZE); + } + + err = proc_dostring(table, write, buffer, lenp, ppos); + if (err) + return err; + + if (write) { + err = __control_devkmsg(devkmsg_log_str); + + /* + * Do not accept an unknown string OR a known string with + * trailing crap... + */ + if (err < 0 || (err + 1 != *lenp)) { + + /* ... and restore old setting. */ + devkmsg_log = old; + strncpy(devkmsg_log_str, old_str, DEVKMSG_STR_MAX_SIZE); + + return -EINVAL; + } + } + + return 0; +} + /* * Number of registered extended console drivers. * @@ -613,6 +718,7 @@ struct devkmsg_user { u64 seq; u32 idx; enum log_flags prev; + struct ratelimit_state rs; struct mutex lock; char buf[CONSOLE_EXT_LOG_MAX]; }; @@ -622,11 +728,24 @@ static ssize_t devkmsg_write(struct kiocb *iocb, struct iov_iter *from) char *buf, *line; int level = default_message_loglevel; int facility = 1; /* LOG_USER */ + struct file *file = iocb->ki_filp; + struct devkmsg_user *user = file->private_data; size_t len = iov_iter_count(from); ssize_t ret = len; - if (len > LOG_LINE_MAX) + if (!user || len > LOG_LINE_MAX) return -EINVAL; + + /* Ignore when user logging is disabled. */ + if (devkmsg_log & DEVKMSG_LOG_MASK_OFF) + return len; + + /* Ratelimit when not explicitly enabled. */ + if (!(devkmsg_log & DEVKMSG_LOG_MASK_ON)) { + if (!___ratelimit(&user->rs, current->comm)) + return ret; + } + buf = kmalloc(len+1, GFP_KERNEL); if (buf == NULL) return -ENOMEM; @@ -799,19 +918,24 @@ static int devkmsg_open(struct inode *inode, struct file *file) struct devkmsg_user *user; int err; - /* write-only does not need any file context */ - if ((file->f_flags & O_ACCMODE) == O_WRONLY) - return 0; + if (devkmsg_log & DEVKMSG_LOG_MASK_OFF) + return -EPERM; - err = check_syslog_permissions(SYSLOG_ACTION_READ_ALL, - SYSLOG_FROM_READER); - if (err) - return err; + /* write-only does not need any file context */ + if ((file->f_flags & O_ACCMODE) != O_WRONLY) { + err = check_syslog_permissions(SYSLOG_ACTION_READ_ALL, + SYSLOG_FROM_READER); + if (err) + return err; + } user = kmalloc(sizeof(struct devkmsg_user), GFP_KERNEL); if (!user) return -ENOMEM; + ratelimit_default_init(&user->rs); + ratelimit_set_flags(&user->rs, RATELIMIT_MSG_ON_RELEASE); + mutex_init(&user->lock); raw_spin_lock_irq(&logbuf_lock); @@ -830,6 +954,8 @@ static int devkmsg_release(struct inode *inode, struct file *file) if (!user) return 0; + ratelimit_state_exit(&user->rs); + mutex_destroy(&user->lock); kfree(user); return 0; diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 5395463..b43d0b2 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -814,6 +814,13 @@ static struct ctl_table kern_table[] = { .extra2 = &ten_thousand, }, { + .procname = "printk_devkmsg", + .data = devkmsg_log_str, + .maxlen = DEVKMSG_STR_MAX_SIZE, + .mode = 0644, + .proc_handler = devkmsg_sysctl_set_loglvl, + }, + { .procname = "dmesg_restrict", .data = &dmesg_restrict, .maxlen = sizeof(int), -- cgit v0.10.2 From 4cad35a7ca690eabf0d241062ce9e59693ec03e7 Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Tue, 2 Aug 2016 14:04:10 -0700 Subject: get_maintainer.pl: reduce need for command-line option -f If a vcs is used, look to see if the vcs tracks the file specified and so the -f option becomes optional. Link: http://lkml.kernel.org/r/7c86a8df0d48770c45778a43b6b3e4627b2a90ee.1469746395.git.joe@perches.com Signed-off-by: Joe Perches Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/scripts/get_maintainer.pl b/scripts/get_maintainer.pl index 1873421..122fcda 100755 --- a/scripts/get_maintainer.pl +++ b/scripts/get_maintainer.pl @@ -133,6 +133,7 @@ my %VCS_cmds_git = ( "author_pattern" => "^GitAuthor: (.*)", "subject_pattern" => "^GitSubject: (.*)", "stat_pattern" => "^(\\d+)\\t(\\d+)\\t\$file\$", + "file_exists_cmd" => "git ls-files \$file", ); my %VCS_cmds_hg = ( @@ -161,6 +162,7 @@ my %VCS_cmds_hg = ( "author_pattern" => "^HgAuthor: (.*)", "subject_pattern" => "^HgSubject: (.*)", "stat_pattern" => "^(\\d+)\t(\\d+)\t\$file\$", + "file_exists_cmd" => "hg files \$file", ); my $conf = which_conf(".get_maintainer.conf"); @@ -430,7 +432,7 @@ foreach my $file (@ARGV) { die "$P: file '${file}' not found\n"; } } - if ($from_filename) { + if ($from_filename || vcs_file_exists($file)) { $file =~ s/^\Q${cur_path}\E//; #strip any absolute path $file =~ s/^\Q${lk_path}\E//; #or the path to the lk tree push(@files, $file); @@ -2124,6 +2126,22 @@ sub vcs_file_blame { } } +sub vcs_file_exists { + my ($file) = @_; + + my $exists; + + my $vcs_used = vcs_exists(); + return 0 if (!$vcs_used); + + my $cmd = $VCS_cmds{"file_exists_cmd"}; + $cmd =~ s/(\$\w+)/$1/eeg; # interpolate $cmd + + $exists = &{$VCS_cmds{"execute_cmd"}}($cmd); + + return $exists; +} + sub uniq { my (@parms) = @_; -- cgit v0.10.2 From f003a1f182bb821f13775338a4bf8711830f927a Mon Sep 17 00:00:00 2001 From: Sebastian Ott Date: Tue, 2 Aug 2016 14:04:13 -0700 Subject: lib/iommu-helper: skip to next segment When a large enough area in the iommu bitmap is found but would span a boundary we continue the search starting from the next bit position. For large allocations this can lead to several useless invocations of bitmap_find_next_zero_area() and iommu_is_span_boundary(). Continue the search from the start of the next segment (which is the next bit position such that we'll not cross the same segment boundary again). Link: http://lkml.kernel.org/r/alpine.LFD.2.20.1606081910070.3211@schleppi Signed-off-by: Sebastian Ott Reviewed-by: Gerald Schaefer Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/lib/iommu-helper.c b/lib/iommu-helper.c index c27e269..a816f3a 100644 --- a/lib/iommu-helper.c +++ b/lib/iommu-helper.c @@ -29,8 +29,7 @@ again: index = bitmap_find_next_zero_area(map, size, start, nr, align_mask); if (index < size) { if (iommu_is_span_boundary(index, nr, shift, boundary_size)) { - /* we could do more effectively */ - start = index + 1; + start = ALIGN(shift + index, boundary_size) - shift; goto again; } bitmap_set(map, index, nr); -- cgit v0.10.2 From a9bfd3321713ecec86282dd2bec04212189f91f1 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Tue, 2 Aug 2016 14:04:16 -0700 Subject: crc32: use ktime_get_ns() for measurement The crc32 test function measures the elapsed time in nanoseconds, but uses 'struct timespec' for that. We want to remove timespec from the kernel for y2038 compatibility, and ktime_get_ns() also helps make the code simpler here. It is also slightly better to use monontonic time, as we are only interested in the time difference. Link: http://lkml.kernel.org/r/20160617143932.3289626-1-arnd@arndb.de Signed-off-by: Arnd Bergmann Cc: "David S . Miller" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/lib/crc32.c b/lib/crc32.c index 9a907d4..7fbd1a1 100644 --- a/lib/crc32.c +++ b/lib/crc32.c @@ -979,7 +979,6 @@ static int __init crc32c_test(void) int i; int errors = 0; int bytes = 0; - struct timespec start, stop; u64 nsec; unsigned long flags; @@ -999,20 +998,17 @@ static int __init crc32c_test(void) local_irq_save(flags); local_irq_disable(); - getnstimeofday(&start); + nsec = ktime_get_ns(); for (i = 0; i < 100; i++) { if (test[i].crc32c_le != __crc32c_le(test[i].crc, test_buf + test[i].start, test[i].length)) errors++; } - getnstimeofday(&stop); + nsec = ktime_get_ns() - nsec; local_irq_restore(flags); local_irq_enable(); - nsec = stop.tv_nsec - start.tv_nsec + - 1000000000 * (stop.tv_sec - start.tv_sec); - pr_info("crc32c: CRC_LE_BITS = %d\n", CRC_LE_BITS); if (errors) @@ -1065,7 +1061,6 @@ static int __init crc32_test(void) int i; int errors = 0; int bytes = 0; - struct timespec start, stop; u64 nsec; unsigned long flags; @@ -1088,7 +1083,7 @@ static int __init crc32_test(void) local_irq_save(flags); local_irq_disable(); - getnstimeofday(&start); + nsec = ktime_get_ns(); for (i = 0; i < 100; i++) { if (test[i].crc_le != crc32_le(test[i].crc, test_buf + test[i].start, test[i].length)) @@ -1098,14 +1093,11 @@ static int __init crc32_test(void) test[i].start, test[i].length)) errors++; } - getnstimeofday(&stop); + nsec = ktime_get_ns() - nsec; local_irq_restore(flags); local_irq_enable(); - nsec = stop.tv_nsec - start.tv_nsec + - 1000000000 * (stop.tv_sec - start.tv_sec); - pr_info("crc32: CRC_LE_BITS = %d, CRC_BE BITS = %d\n", CRC_LE_BITS, CRC_BE_BITS); -- cgit v0.10.2 From a23216a2f1f8a30a3b6588c743681651e4a6aa94 Mon Sep 17 00:00:00 2001 From: Ross Zwisler Date: Tue, 2 Aug 2016 14:04:19 -0700 Subject: radix-tree: fix comment about "exceptional" bits The bottom two bits of radix tree entries are reserved for special use by the radix tree code itself. A comment detailing their usage was added by commit 3bcadd6fa6c4 ("radix-tree: free up the bottom bit of exceptional entries for reuse") This comment states that if the bottom two bits are '11', this means that this is a locked exceptional entry. It turns out that this bit combination was never actually used. Radix tree locking for DAX was indeed implemented, but it actually used the third LSB: /* We use lowest available exceptional entry bit for locking */ #define RADIX_DAX_ENTRY_LOCK (1 << RADIX_TREE_EXCEPTIONAL_SHIFT) This locking code was also made specific to the DAX code instead of being generally implemented in radix-tree.h. So, fix the comment. Link: http://lkml.kernel.org/r/1468997731-2155-1-git-send-email-ross.zwisler@linux.intel.com Signed-off-by: Ross Zwisler Reviewed-by: Johannes Thumshirn Cc: Dave Chinner Cc: Jan Kara Cc: Konstantin Khlebnikov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/radix-tree.h b/include/linux/radix-tree.h index cbfee50..4c45105 100644 --- a/include/linux/radix-tree.h +++ b/include/linux/radix-tree.h @@ -35,7 +35,7 @@ * 00 - data pointer * 01 - internal entry * 10 - exceptional entry - * 11 - locked exceptional entry + * 11 - this bit combination is currently unused/reserved * * The internal entry may be a pointer to the next level in the tree, a * sibling entry, or an indicator that the entry in this slot has been moved -- cgit v0.10.2 From 9ccf98119821defe66ee2ee21f8a11071f63fa65 Mon Sep 17 00:00:00 2001 From: Stephen Boyd Date: Tue, 2 Aug 2016 14:04:22 -0700 Subject: firmware: consolidate kmap/read/write logic Some systems are memory constrained but they need to load very large firmwares. The firmware subsystem allows drivers to request this firmware be loaded from the filesystem, but this requires that the entire firmware be loaded into kernel memory first before it's provided to the driver. This can lead to a situation where we map the firmware twice, once to load the firmware into kernel memory and once to copy the firmware into the final resting place. This design creates needless memory pressure and delays loading because we have to copy from kernel memory to somewhere else. This patch sets adds support to the request firmware API to load the firmware directly into a pre-allocated buffer, skipping the intermediate copying step and alleviating memory pressure during firmware loading. The drawback is that we can't use the firmware caching feature because the memory for the firmware cache is not managed by the firmware layer. This patch (of 3): We use similar structured code to read and write the kmapped firmware pages. The only difference is read copies from the kmap region and write copies to it. Consolidate this into one function to reduce duplication. Link: http://lkml.kernel.org/r/20160607164741.31849-2-stephen.boyd@linaro.org Signed-off-by: Stephen Boyd Cc: Vikram Mulukutla Cc: Mimi Zohar Cc: Mark Brown Cc: Ming Lei Cc: Greg Kroah-Hartman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/base/firmware_class.c b/drivers/base/firmware_class.c index 773fc30..01d5572 100644 --- a/drivers/base/firmware_class.c +++ b/drivers/base/firmware_class.c @@ -691,6 +691,29 @@ out: static DEVICE_ATTR(loading, 0644, firmware_loading_show, firmware_loading_store); +static void firmware_rw(struct firmware_buf *buf, char *buffer, + loff_t offset, size_t count, bool read) +{ + while (count) { + void *page_data; + int page_nr = offset >> PAGE_SHIFT; + int page_ofs = offset & (PAGE_SIZE-1); + int page_cnt = min_t(size_t, PAGE_SIZE - page_ofs, count); + + page_data = kmap(buf->pages[page_nr]); + + if (read) + memcpy(buffer, page_data + page_ofs, page_cnt); + else + memcpy(page_data + page_ofs, buffer, page_cnt); + + kunmap(buf->pages[page_nr]); + buffer += page_cnt; + offset += page_cnt; + count -= page_cnt; + } +} + static ssize_t firmware_data_read(struct file *filp, struct kobject *kobj, struct bin_attribute *bin_attr, char *buffer, loff_t offset, size_t count) @@ -715,21 +738,8 @@ static ssize_t firmware_data_read(struct file *filp, struct kobject *kobj, ret_count = count; - while (count) { - void *page_data; - int page_nr = offset >> PAGE_SHIFT; - int page_ofs = offset & (PAGE_SIZE-1); - int page_cnt = min_t(size_t, PAGE_SIZE - page_ofs, count); - - page_data = kmap(buf->pages[page_nr]); - - memcpy(buffer, page_data + page_ofs, page_cnt); + firmware_rw(buf, buffer, offset, count, true); - kunmap(buf->pages[page_nr]); - buffer += page_cnt; - offset += page_cnt; - count -= page_cnt; - } out: mutex_unlock(&fw_lock); return ret_count; @@ -809,24 +819,9 @@ static ssize_t firmware_data_write(struct file *filp, struct kobject *kobj, goto out; retval = count; + firmware_rw(buf, buffer, offset, count, false); - while (count) { - void *page_data; - int page_nr = offset >> PAGE_SHIFT; - int page_ofs = offset & (PAGE_SIZE - 1); - int page_cnt = min_t(size_t, PAGE_SIZE - page_ofs, count); - - page_data = kmap(buf->pages[page_nr]); - - memcpy(page_data + page_ofs, buffer, page_cnt); - - kunmap(buf->pages[page_nr]); - buffer += page_cnt; - offset += page_cnt; - count -= page_cnt; - } - - buf->size = max_t(size_t, offset, buf->size); + buf->size = max_t(size_t, offset + count, buf->size); out: mutex_unlock(&fw_lock); return retval; -- cgit v0.10.2 From 0e742e927571946e08e877d3629e6efd4891ed95 Mon Sep 17 00:00:00 2001 From: Vikram Mulukutla Date: Tue, 2 Aug 2016 14:04:25 -0700 Subject: firmware: provide infrastructure to make fw caching optional Some low memory systems with complex peripherals cannot afford to have the relatively large firmware images taking up valuable memory during suspend and resume. Change the internal implementation of firmware_class to disallow caching based on a configurable option. In the near future, variants of request_firmware will take advantage of this feature. Link: http://lkml.kernel.org/r/20160607164741.31849-3-stephen.boyd@linaro.org [stephen.boyd@linaro.org: Drop firmware_desc design and use flags] Signed-off-by: Vikram Mulukutla Signed-off-by: Stephen Boyd Cc: Mimi Zohar Cc: Mark Brown Cc: Ming Lei Cc: Greg Kroah-Hartman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/base/firmware_class.c b/drivers/base/firmware_class.c index 01d5572..45ed20c 100644 --- a/drivers/base/firmware_class.c +++ b/drivers/base/firmware_class.c @@ -112,6 +112,7 @@ static inline long firmware_loading_timeout(void) #define FW_OPT_FALLBACK 0 #endif #define FW_OPT_NO_WARN (1U << 3) +#define FW_OPT_NOCACHE (1U << 4) struct firmware_cache { /* firmware_buf instance will be added into the below list */ @@ -1065,14 +1066,16 @@ static int assign_firmware_buf(struct firmware *fw, struct device *device, * should be fixed in devres or driver core. */ /* don't cache firmware handled without uevent */ - if (device && (opt_flags & FW_OPT_UEVENT)) + if (device && (opt_flags & FW_OPT_UEVENT) && + !(opt_flags & FW_OPT_NOCACHE)) fw_add_devm_name(device, buf->fw_id); /* * After caching firmware image is started, let it piggyback * on request firmware. */ - if (buf->fwc->state == FW_LOADER_START_CACHE) { + if (!(opt_flags & FW_OPT_NOCACHE) && + buf->fwc->state == FW_LOADER_START_CACHE) { if (fw_cache_piggyback_on_request(buf->fw_id)) kref_get(&buf->ref); } -- cgit v0.10.2 From a098ecd2fa7db8fa4fcc178a43627b29b226edb9 Mon Sep 17 00:00:00 2001 From: Stephen Boyd Date: Tue, 2 Aug 2016 14:04:28 -0700 Subject: firmware: support loading into a pre-allocated buffer Some systems are memory constrained but they need to load very large firmwares. The firmware subsystem allows drivers to request this firmware be loaded from the filesystem, but this requires that the entire firmware be loaded into kernel memory first before it's provided to the driver. This can lead to a situation where we map the firmware twice, once to load the firmware into kernel memory and once to copy the firmware into the final resting place. This creates needless memory pressure and delays loading because we have to copy from kernel memory to somewhere else. Let's add a request_firmware_into_buf() API that allows drivers to request firmware be loaded directly into a pre-allocated buffer. This skips the intermediate step of allocating a buffer in kernel memory to hold the firmware image while it's read from the filesystem. It also requires that drivers know how much memory they'll require before requesting the firmware and negates any benefits of firmware caching because the firmware layer doesn't manage the buffer lifetime. For a 16MB buffer, about half the time is spent performing a memcpy from the buffer to the final resting place. I see loading times go from 0.081171 seconds to 0.047696 seconds after applying this patch. Plus the vmalloc pressure is reduced. This is based on a patch from Vikram Mulukutla on codeaurora.org: https://www.codeaurora.org/cgit/quic/la/kernel/msm-3.18/commit/drivers/base/firmware_class.c?h=rel/msm-3.18&id=0a328c5f6cd999f5c591f172216835636f39bcb5 Link: http://lkml.kernel.org/r/20160607164741.31849-4-stephen.boyd@linaro.org Signed-off-by: Stephen Boyd Cc: Mimi Zohar Cc: Vikram Mulukutla Cc: Mark Brown Cc: Ming Lei Cc: Greg Kroah-Hartman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/base/firmware_class.c b/drivers/base/firmware_class.c index 45ed20c..22d1760 100644 --- a/drivers/base/firmware_class.c +++ b/drivers/base/firmware_class.c @@ -46,7 +46,8 @@ MODULE_LICENSE("GPL"); extern struct builtin_fw __start_builtin_fw[]; extern struct builtin_fw __end_builtin_fw[]; -static bool fw_get_builtin_firmware(struct firmware *fw, const char *name) +static bool fw_get_builtin_firmware(struct firmware *fw, const char *name, + void *buf, size_t size) { struct builtin_fw *b_fw; @@ -54,6 +55,9 @@ static bool fw_get_builtin_firmware(struct firmware *fw, const char *name) if (strcmp(name, b_fw->name) == 0) { fw->size = b_fw->size; fw->data = b_fw->data; + + if (buf && fw->size <= size) + memcpy(buf, fw->data, fw->size); return true; } } @@ -74,7 +78,9 @@ static bool fw_is_builtin_firmware(const struct firmware *fw) #else /* Module case - no builtin firmware support */ -static inline bool fw_get_builtin_firmware(struct firmware *fw, const char *name) +static inline bool fw_get_builtin_firmware(struct firmware *fw, + const char *name, void *buf, + size_t size) { return false; } @@ -144,6 +150,7 @@ struct firmware_buf { unsigned long status; void *data; size_t size; + size_t allocated_size; #ifdef CONFIG_FW_LOADER_USER_HELPER bool is_paged_buf; bool need_uevent; @@ -179,7 +186,8 @@ static DEFINE_MUTEX(fw_lock); static struct firmware_cache fw_cache; static struct firmware_buf *__allocate_fw_buf(const char *fw_name, - struct firmware_cache *fwc) + struct firmware_cache *fwc, + void *dbuf, size_t size) { struct firmware_buf *buf; @@ -195,6 +203,8 @@ static struct firmware_buf *__allocate_fw_buf(const char *fw_name, kref_init(&buf->ref); buf->fwc = fwc; + buf->data = dbuf; + buf->allocated_size = size; init_completion(&buf->completion); #ifdef CONFIG_FW_LOADER_USER_HELPER INIT_LIST_HEAD(&buf->pending_list); @@ -218,7 +228,8 @@ static struct firmware_buf *__fw_lookup_buf(const char *fw_name) static int fw_lookup_and_allocate_buf(const char *fw_name, struct firmware_cache *fwc, - struct firmware_buf **buf) + struct firmware_buf **buf, void *dbuf, + size_t size) { struct firmware_buf *tmp; @@ -230,7 +241,7 @@ static int fw_lookup_and_allocate_buf(const char *fw_name, *buf = tmp; return 1; } - tmp = __allocate_fw_buf(fw_name, fwc); + tmp = __allocate_fw_buf(fw_name, fwc, dbuf, size); if (tmp) list_add(&tmp->list, &fwc->head); spin_unlock(&fwc->lock); @@ -262,6 +273,7 @@ static void __fw_free_buf(struct kref *ref) vfree(buf->pages); } else #endif + if (!buf->allocated_size) vfree(buf->data); kfree_const(buf->fw_id); kfree(buf); @@ -302,13 +314,21 @@ static void fw_finish_direct_load(struct device *device, mutex_unlock(&fw_lock); } -static int fw_get_filesystem_firmware(struct device *device, - struct firmware_buf *buf) +static int +fw_get_filesystem_firmware(struct device *device, struct firmware_buf *buf) { loff_t size; int i, len; int rc = -ENOENT; char *path; + enum kernel_read_file_id id = READING_FIRMWARE; + size_t msize = INT_MAX; + + /* Already populated data member means we're loading into a buffer */ + if (buf->data) { + id = READING_FIRMWARE_PREALLOC_BUFFER; + msize = buf->allocated_size; + } path = __getname(); if (!path) @@ -327,8 +347,8 @@ static int fw_get_filesystem_firmware(struct device *device, } buf->size = 0; - rc = kernel_read_file_from_path(path, &buf->data, &size, - INT_MAX, READING_FIRMWARE); + rc = kernel_read_file_from_path(path, &buf->data, &size, msize, + id); if (rc) { if (rc == -ENOENT) dev_dbg(device, "loading %s failed with error %d\n", @@ -692,6 +712,15 @@ out: static DEVICE_ATTR(loading, 0644, firmware_loading_show, firmware_loading_store); +static void firmware_rw_buf(struct firmware_buf *buf, char *buffer, + loff_t offset, size_t count, bool read) +{ + if (read) + memcpy(buffer, buf->data + offset, count); + else + memcpy(buf->data + offset, buffer, count); +} + static void firmware_rw(struct firmware_buf *buf, char *buffer, loff_t offset, size_t count, bool read) { @@ -739,7 +768,10 @@ static ssize_t firmware_data_read(struct file *filp, struct kobject *kobj, ret_count = count; - firmware_rw(buf, buffer, offset, count, true); + if (buf->data) + firmware_rw_buf(buf, buffer, offset, count, true); + else + firmware_rw(buf, buffer, offset, count, true); out: mutex_unlock(&fw_lock); @@ -815,12 +847,21 @@ static ssize_t firmware_data_write(struct file *filp, struct kobject *kobj, goto out; } - retval = fw_realloc_buffer(fw_priv, offset + count); - if (retval) - goto out; + if (buf->data) { + if (offset + count > buf->allocated_size) { + retval = -ENOMEM; + goto out; + } + firmware_rw_buf(buf, buffer, offset, count, false); + retval = count; + } else { + retval = fw_realloc_buffer(fw_priv, offset + count); + if (retval) + goto out; - retval = count; - firmware_rw(buf, buffer, offset, count, false); + retval = count; + firmware_rw(buf, buffer, offset, count, false); + } buf->size = max_t(size_t, offset + count, buf->size); out: @@ -890,7 +931,8 @@ static int _request_firmware_load(struct firmware_priv *fw_priv, struct firmware_buf *buf = fw_priv->buf; /* fall back on userspace loading */ - buf->is_paged_buf = true; + if (!buf->data) + buf->is_paged_buf = true; dev_set_uevent_suppress(f_dev, true); @@ -925,7 +967,7 @@ static int _request_firmware_load(struct firmware_priv *fw_priv, if (is_fw_load_aborted(buf)) retval = -EAGAIN; - else if (!buf->data) + else if (buf->is_paged_buf && !buf->data) retval = -ENOMEM; device_del(f_dev); @@ -1008,7 +1050,7 @@ static int sync_cached_firmware_buf(struct firmware_buf *buf) */ static int _request_firmware_prepare(struct firmware **firmware_p, const char *name, - struct device *device) + struct device *device, void *dbuf, size_t size) { struct firmware *firmware; struct firmware_buf *buf; @@ -1021,12 +1063,12 @@ _request_firmware_prepare(struct firmware **firmware_p, const char *name, return -ENOMEM; } - if (fw_get_builtin_firmware(firmware, name)) { + if (fw_get_builtin_firmware(firmware, name, dbuf, size)) { dev_dbg(device, "using built-in %s\n", name); return 0; /* assigned */ } - ret = fw_lookup_and_allocate_buf(name, &fw_cache, &buf); + ret = fw_lookup_and_allocate_buf(name, &fw_cache, &buf, dbuf, size); /* * bind with 'buf' now to avoid warning in failure path @@ -1089,7 +1131,8 @@ static int assign_firmware_buf(struct firmware *fw, struct device *device, /* called from request_firmware() and request_firmware_work_func() */ static int _request_firmware(const struct firmware **firmware_p, const char *name, - struct device *device, unsigned int opt_flags) + struct device *device, void *buf, size_t size, + unsigned int opt_flags) { struct firmware *fw = NULL; long timeout; @@ -1103,7 +1146,7 @@ _request_firmware(const struct firmware **firmware_p, const char *name, goto out; } - ret = _request_firmware_prepare(&fw, name, device); + ret = _request_firmware_prepare(&fw, name, device, buf, size); if (ret <= 0) /* error or already assigned */ goto out; @@ -1182,7 +1225,7 @@ request_firmware(const struct firmware **firmware_p, const char *name, /* Need to pin this module until return */ __module_get(THIS_MODULE); - ret = _request_firmware(firmware_p, name, device, + ret = _request_firmware(firmware_p, name, device, NULL, 0, FW_OPT_UEVENT | FW_OPT_FALLBACK); module_put(THIS_MODULE); return ret; @@ -1206,7 +1249,7 @@ int request_firmware_direct(const struct firmware **firmware_p, int ret; __module_get(THIS_MODULE); - ret = _request_firmware(firmware_p, name, device, + ret = _request_firmware(firmware_p, name, device, NULL, 0, FW_OPT_UEVENT | FW_OPT_NO_WARN); module_put(THIS_MODULE); return ret; @@ -1214,6 +1257,36 @@ int request_firmware_direct(const struct firmware **firmware_p, EXPORT_SYMBOL_GPL(request_firmware_direct); /** + * request_firmware_into_buf - load firmware into a previously allocated buffer + * @firmware_p: pointer to firmware image + * @name: name of firmware file + * @device: device for which firmware is being loaded and DMA region allocated + * @buf: address of buffer to load firmware into + * @size: size of buffer + * + * This function works pretty much like request_firmware(), but it doesn't + * allocate a buffer to hold the firmware data. Instead, the firmware + * is loaded directly into the buffer pointed to by @buf and the @firmware_p + * data member is pointed at @buf. + * + * This function doesn't cache firmware either. + */ +int +request_firmware_into_buf(const struct firmware **firmware_p, const char *name, + struct device *device, void *buf, size_t size) +{ + int ret; + + __module_get(THIS_MODULE); + ret = _request_firmware(firmware_p, name, device, buf, size, + FW_OPT_UEVENT | FW_OPT_FALLBACK | + FW_OPT_NOCACHE); + module_put(THIS_MODULE); + return ret; +} +EXPORT_SYMBOL(request_firmware_into_buf); + +/** * release_firmware: - release the resource associated with a firmware image * @fw: firmware resource to release **/ @@ -1245,7 +1318,7 @@ static void request_firmware_work_func(struct work_struct *work) fw_work = container_of(work, struct firmware_work, work); - _request_firmware(&fw, fw_work->name, fw_work->device, + _request_firmware(&fw, fw_work->name, fw_work->device, NULL, 0, fw_work->opt_flags); fw_work->cont(fw, fw_work->context); put_device(fw_work->device); /* taken in request_firmware_nowait() */ @@ -1378,7 +1451,7 @@ static int uncache_firmware(const char *fw_name) pr_debug("%s: %s\n", __func__, fw_name); - if (fw_get_builtin_firmware(&fw, fw_name)) + if (fw_get_builtin_firmware(&fw, fw_name, NULL, 0)) return 0; buf = fw_lookup_buf(fw_name); diff --git a/fs/exec.c b/fs/exec.c index ca239fc..a1789cd 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -866,7 +866,8 @@ int kernel_read_file(struct file *file, void **buf, loff_t *size, goto out; } - *buf = vmalloc(i_size); + if (id != READING_FIRMWARE_PREALLOC_BUFFER) + *buf = vmalloc(i_size); if (!*buf) { ret = -ENOMEM; goto out; @@ -897,8 +898,10 @@ int kernel_read_file(struct file *file, void **buf, loff_t *size, out_free: if (ret < 0) { - vfree(*buf); - *buf = NULL; + if (id != READING_FIRMWARE_PREALLOC_BUFFER) { + vfree(*buf); + *buf = NULL; + } } out: diff --git a/include/linux/firmware.h b/include/linux/firmware.h index 5c41c5e..b1f9f0c 100644 --- a/include/linux/firmware.h +++ b/include/linux/firmware.h @@ -47,6 +47,8 @@ int request_firmware_nowait( void (*cont)(const struct firmware *fw, void *context)); int request_firmware_direct(const struct firmware **fw, const char *name, struct device *device); +int request_firmware_into_buf(const struct firmware **firmware_p, + const char *name, struct device *device, void *buf, size_t size); void release_firmware(const struct firmware *fw); #else @@ -75,5 +77,11 @@ static inline int request_firmware_direct(const struct firmware **fw, return -EINVAL; } +static inline int request_firmware_into_buf(const struct firmware **firmware_p, + const char *name, struct device *device, void *buf, size_t size) +{ + return -EINVAL; +} + #endif #endif diff --git a/include/linux/fs.h b/include/linux/fs.h index 577365a..f3f0b4c8 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2652,6 +2652,7 @@ extern int do_pipe_flags(int *, int); #define __kernel_read_file_id(id) \ id(UNKNOWN, unknown) \ id(FIRMWARE, firmware) \ + id(FIRMWARE_PREALLOC_BUFFER, firmware) \ id(MODULE, kernel-module) \ id(KEXEC_IMAGE, kexec-image) \ id(KEXEC_INITRAMFS, kexec-initramfs) \ -- cgit v0.10.2 From d560a5f8a46e98c2f83c5fe699e1d6f6393a14cf Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Tue, 2 Aug 2016 14:04:31 -0700 Subject: checkpatch: skip long lines that use an EFI_GUID macro These are also possible single line uses that exceed the generic maximum line length (typically 80 columns) Link: http://lkml.kernel.org/r/32a6a85fbd6161f1bb55ce176a464e44591afc5b.1468368420.git.joe@perches.com Signed-off-by: Joe Perches Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl index 24a0836..a4476b6 100755 --- a/scripts/checkpatch.pl +++ b/scripts/checkpatch.pl @@ -2763,6 +2763,10 @@ sub process { $line =~ /^\+\s*#\s*define\s+\w+\s+$String$/) { $msg_type = ""; + # EFI_GUID is another special case + } elsif ($line =~ /^\+.*\bEFI_GUID\s*\(/) { + $msg_type = ""; + # Otherwise set the alternate message types # a comment starts before $max_line_length -- cgit v0.10.2 From dadf680de3c2eb4cba9840619991eda0cfe98778 Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Tue, 2 Aug 2016 14:04:33 -0700 Subject: checkpatch: allow c99 style // comments Sanitise the lines that contain c99 comments so that the error doesn't get emitted. Link: http://lkml.kernel.org/r/d4d22c34ad7bcc1bceb52f0742f76b7a6d585235.1468368420.git.joe@perches.com Signed-off-by: Joe Perches Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl index a4476b6..7927300 100755 --- a/scripts/checkpatch.pl +++ b/scripts/checkpatch.pl @@ -55,6 +55,7 @@ my $spelling_file = "$D/spelling.txt"; my $codespell = 0; my $codespellfile = "/usr/share/codespell/dictionary.txt"; my $color = 1; +my $allow_c99_comments = 1; sub help { my ($exitcode) = @_; @@ -1144,6 +1145,11 @@ sub sanitise_line { $res =~ s@(\#\s*(?:error|warning)\s+).*@$1$clean@; } + if ($allow_c99_comments && $res =~ m@(//.*$)@) { + my $match = $1; + $res =~ s/\Q$match\E/"$;" x length($match)/e; + } + return $res; } -- cgit v0.10.2 From aab38f516aa99e8132f906a526bf44fa59e9daa3 Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Tue, 2 Aug 2016 14:04:36 -0700 Subject: checkpatch: yet another commit id improvement Using \b isn't good enough to isolate what appears to be a commit id in a commit message. Make sure there is a space or a quote like character after a continuous run of hexadecimal characters that could be a commit id. Link: http://lkml.kernel.org/r/fdd22b47463a21c21132edbb8aa35e372950a1e6.1468869915.git.joe@perches.com Signed-off-by: Joe Perches Cc: "Zhuo, Qiuxu" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl index 7927300..7a28775 100755 --- a/scripts/checkpatch.pl +++ b/scripts/checkpatch.pl @@ -2459,9 +2459,9 @@ sub process { # Check for git id commit length and improperly formed commit descriptions if ($in_commit_log && !$commit_log_possible_stack_dump && - $line !~ /^\s*(?:Link|Patchwork|http|BugLink):/i && + $line !~ /^\s*(?:Link|Patchwork|http|https|BugLink):/i && ($line =~ /\bcommit\s+[0-9a-f]{5,}\b/i || - ($line =~ /\b[0-9a-f]{12,40}\b/i && + ($line =~ /(?:\s|^)[0-9a-f]{12,40}(?:[\s"'\(\[]|$)/i && $line !~ /[\<\[][0-9a-f]{12,40}[\>\]]/i && $line !~ /\bfixes:\s*[0-9a-f]{12,40}/i))) { my $init_char = "c"; -- cgit v0.10.2 From cec3aaa56638c7aad763630b9cbe591f2e791a3b Mon Sep 17 00:00:00 2001 From: Tomas Winkler Date: Tue, 2 Aug 2016 14:04:39 -0700 Subject: checkpatch: don't complain about BIT macro in uapi BIT macro cannot be exported to UAPI, don't complain about it. Link: http://lkml.kernel.org/r/1468707033-16173-1-git-send-email-tomas.winkler@intel.com Signed-off-by: Tomas Winkler Acked-by: Joe Perches Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl index 7a28775..77915e0 100755 --- a/scripts/checkpatch.pl +++ b/scripts/checkpatch.pl @@ -5732,8 +5732,9 @@ sub process { } } -# check for #defines like: 1 << that could be BIT(digit) - if ($line =~ /#\s*define\s+\w+\s+\(?\s*1\s*([ulUL]*)\s*\<\<\s*(?:\d+|$Ident)\s*\)?/) { +# check for #defines like: 1 << that could be BIT(digit), it is not exported to uapi + if ($realfile !~ m@^include/uapi/@ && + $line =~ /#\s*define\s+\w+\s+\(?\s*1\s*([ulUL]*)\s*\<\<\s*(?:\d+|$Ident)\s*\)?/) { my $ull = ""; $ull = "_ULL" if (defined($1) && $1 =~ /ll/i); if (CHK("BIT_MACRO", -- cgit v0.10.2 From c844711575086231890084390a275d06f11a623a Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Tue, 2 Aug 2016 14:04:42 -0700 Subject: checkpatch: improve 'bare use of' signed/unsigned types warning Fix false positive warning of identifiers ending in signed with an = assignment of WARNING: Prefer 'signed int' to bare use of 'signed'. Link: http://lkml.kernel.org/r/6a0e24c3e9102337528ecfcbbe91a0eb5b4820ed.1469529497.git.joe@perches.com Signed-off-by: Joe Perches Reported-by: Alan Douglas Acked-by: Andy Whitcroft Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl index 77915e0..1d5b09d 100755 --- a/scripts/checkpatch.pl +++ b/scripts/checkpatch.pl @@ -3347,7 +3347,7 @@ sub process { next if ($line =~ /^[^\+]/); # check for declarations of signed or unsigned without int - while ($line =~ m{($Declare)\s*(?!char\b|short\b|int\b|long\b)\s*($Ident)?\s*[=,;\[\)\(]}g) { + while ($line =~ m{\b($Declare)\s*(?!char\b|short\b|int\b|long\b)\s*($Ident)?\s*[=,;\[\)\(]}g) { my $type = $1; my $var = $2; $var = "" if (!defined $var); -- cgit v0.10.2 From ed43c4e58a6d3061e3329c41d7b880f11541245a Mon Sep 17 00:00:00 2001 From: Allen Hubbe Date: Tue, 2 Aug 2016 14:04:45 -0700 Subject: checkpatch: check signoff when reading stdin Signoff was not checked if the filename is '-', indicating reading the patch from stdin. Commands such as the below would not warn about a missing signoff, because the patch filename is '-'. This change allows checkpatch to warn about a missing signoff, even if the input filename is '-', but only if the patch has a commit message. git show --pretty=email | scripts/checkpatch.pl - A more common use of checkpatch with stdin is for piping git diff through checkpatch. The diff output would not contain a commit message, and therefore it would not contain a signoff line. For this common use case, a warning should not be printed about the missing signoff. With this change we will only warn about a missing signoff if the input contains a commit message. git diff | scripts/checkpatch.pl - Before this patch, a workaround for the first command was to refer to stdin by a name other than '-'. The workaround is not an elegant solution, because elsewhere checkpatch uses the fact that filename equals '-', such as in setting '$vname' to 'Your patch' for stdin. The command below would report "/dev/stdin has style problems" instead of "Your patch has style problems." git show --pretty=email | scripts/checkpatch.pl /dev/stdin Link: http://lkml.kernel.org/r/48be31e414bddc65bccfa6b1322359be9ba032eb.1469670589.git.allenbh@gmail.com Signed-off-by: Allen Hubbe Acked-by: Joe Perches Cc: Andy Whitcroft Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl index 1d5b09d..6f2ce0c 100755 --- a/scripts/checkpatch.pl +++ b/scripts/checkpatch.pl @@ -2069,6 +2069,7 @@ sub process { my $is_patch = 0; my $in_header_lines = $file ? 0 : 1; my $in_commit_log = 0; #Scanning lines before patch + my $has_commit_log = 0; #Encountered lines before patch my $commit_log_possible_stack_dump = 0; my $commit_log_long_line = 0; my $commit_log_has_diff = 0; @@ -2566,6 +2567,7 @@ sub process { $rawline =~ /^(commit\b|from\b|[\w-]+:).*$/i)) { $in_header_lines = 0; $in_commit_log = 1; + $has_commit_log = 1; } # Check if there is UTF-8 in a commit log when a mail header has explicitly @@ -6055,7 +6057,7 @@ sub process { ERROR("NOT_UNIFIED_DIFF", "Does not appear to be a unified-diff format patch\n"); } - if ($is_patch && $filename ne '-' && $chk_signoff && $signoff == 0) { + if ($is_patch && $has_commit_log && $chk_signoff && $signoff == 0) { ERROR("MISSING_SIGN_OFF", "Missing Signed-off-by: line(s)\n"); } -- cgit v0.10.2 From 45107ff6d5265b9786c62b694140d839bc3d2433 Mon Sep 17 00:00:00 2001 From: Allen Hubbe Date: Tue, 2 Aug 2016 14:04:47 -0700 Subject: checkpatch: if no filenames then read stdin If no filenames are given, then read the patch from stdin. Link: http://lkml.kernel.org/r/a8784f291ccb5067361992bf5d41ff6cfb0ce5cb.1469830917.git.allenbh@gmail.com Signed-off-by: Allen Hubbe Acked-by: Joe Perches Cc: Andy Whitcroft Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/scripts/checkpatch.pl b/scripts/checkpatch.pl index 6f2ce0c..4de3cc4 100755 --- a/scripts/checkpatch.pl +++ b/scripts/checkpatch.pl @@ -228,9 +228,9 @@ if ($^V && $^V lt $minimum_perl_version) { } } +#if no filenames are given, push '-' to read patch from stdin if ($#ARGV < 0) { - print "$P: no input files\n"; - exit(1); + push(@ARGV, '-'); } sub hash_save_array_words { -- cgit v0.10.2 From 0036d1f7eb95bcc52977f15507f00dd07018e7e2 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Tue, 2 Aug 2016 14:04:51 -0700 Subject: binfmt_elf: fix calculations for bss padding A double-bug exists in the bss calculation code, where an overflow can happen in the "last_bss - elf_bss" calculation, but vm_brk internally aligns the argument, underflowing it, wrapping back around safe. We shouldn't depend on these bugs staying in sync, so this cleans up the bss padding handling to avoid the overflow. This moves the bss padzero() before the last_bss > elf_bss case, since the zero-filling of the ELF_PAGE should have nothing to do with the relationship of last_bss and elf_bss: any trailing portion should be zeroed, and a zero size is already handled by padzero(). Then it handles the math on elf_bss vs last_bss correctly. These need to both be ELF_PAGE aligned to get the comparison correct, since that's the expected granularity of the mappings. Since elf_bss already had alignment-based padding happen in padzero(), the "start" of the new vm_brk() should be moved forward as done in the original code. However, since the "end" of the vm_brk() area will already become PAGE_ALIGNed in vm_brk() then last_bss should get aligned here to avoid hiding it as a side-effect. Additionally makes a cosmetic change to the initial last_bss calculation so it's easier to read in comparison to the load_addr calculation above it (i.e. the only difference is p_filesz vs p_memsz). Link: http://lkml.kernel.org/r/1468014494-25291-2-git-send-email-keescook@chromium.org Signed-off-by: Kees Cook Reported-by: Hector Marco-Gisbert Cc: Ismael Ripoll Ripoll Cc: Alexander Viro Cc: "Kirill A. Shutemov" Cc: Oleg Nesterov Cc: Chen Gang Cc: Michal Hocko Cc: Konstantin Khlebnikov Cc: Andrea Arcangeli Cc: Andrey Ryabinin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index a7a28110..7f6aff3f 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -605,28 +605,30 @@ static unsigned long load_elf_interp(struct elfhdr *interp_elf_ex, * Do the same thing for the memory mapping - between * elf_bss and last_bss is the bss section. */ - k = load_addr + eppnt->p_memsz + eppnt->p_vaddr; + k = load_addr + eppnt->p_vaddr + eppnt->p_memsz; if (k > last_bss) last_bss = k; } } + /* + * Now fill out the bss section: first pad the last page from + * the file up to the page boundary, and zero it from elf_bss + * up to the end of the page. + */ + if (padzero(elf_bss)) { + error = -EFAULT; + goto out; + } + /* + * Next, align both the file and mem bss up to the page size, + * since this is where elf_bss was just zeroed up to, and where + * last_bss will end after the vm_brk() below. + */ + elf_bss = ELF_PAGEALIGN(elf_bss); + last_bss = ELF_PAGEALIGN(last_bss); + /* Finally, if there is still more bss to allocate, do it. */ if (last_bss > elf_bss) { - /* - * Now fill out the bss section. First pad the last page up - * to the page boundary, and then perform a mmap to make sure - * that there are zero-mapped pages up to and including the - * last bss page. - */ - if (padzero(elf_bss)) { - error = -EFAULT; - goto out; - } - - /* What we have mapped so far */ - elf_bss = ELF_PAGESTART(elf_bss + ELF_MIN_ALIGN - 1); - - /* Map the last of the bss segment */ error = vm_brk(elf_bss, last_bss - elf_bss); if (error) goto out; -- cgit v0.10.2 From ba093a6d9397da8eafcfbaa7d95bd34255da39a0 Mon Sep 17 00:00:00 2001 From: Kees Cook Date: Tue, 2 Aug 2016 14:04:54 -0700 Subject: mm: refuse wrapped vm_brk requests The vm_brk() alignment calculations should refuse to overflow. The ELF loader depending on this, but it has been fixed now. No other unsafe callers have been found. Link: http://lkml.kernel.org/r/1468014494-25291-3-git-send-email-keescook@chromium.org Signed-off-by: Kees Cook Reported-by: Hector Marco-Gisbert Cc: Ismael Ripoll Ripoll Cc: Alexander Viro Cc: "Kirill A. Shutemov" Cc: Oleg Nesterov Cc: Chen Gang Cc: Michal Hocko Cc: Konstantin Khlebnikov Cc: Andrea Arcangeli Cc: Andrey Ryabinin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/mmap.c b/mm/mmap.c index d44bee9..ca9d91b 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -2653,16 +2653,18 @@ static inline void verify_mm_writelocked(struct mm_struct *mm) * anonymous maps. eventually we may be able to do some * brk-specific accounting here. */ -static int do_brk(unsigned long addr, unsigned long len) +static int do_brk(unsigned long addr, unsigned long request) { struct mm_struct *mm = current->mm; struct vm_area_struct *vma, *prev; - unsigned long flags; + unsigned long flags, len; struct rb_node **rb_link, *rb_parent; pgoff_t pgoff = addr >> PAGE_SHIFT; int error; - len = PAGE_ALIGN(len); + len = PAGE_ALIGN(request); + if (len < request) + return -ENOMEM; if (!len) return 0; -- cgit v0.10.2 From a310dcb7a43f971ee7d810fbbe36bd766a299717 Mon Sep 17 00:00:00 2001 From: Daniel Wagner Date: Tue, 2 Aug 2016 14:04:57 -0700 Subject: fs/binfmt_em86.c: fix incompatible pointer type Since the -Wincompatible-pointer-types is reported as error, alpha doesn't build anymore. Let's fix it in a minimal way. fs/binfmt_em86.c:73:35: error: passing argument 2 of `copy_strings_kernel' from incompatible pointer type [-Werror=incompatible-pointer-types] retval = copy_strings_kernel(1, &i_arg, bprm); ^ ^ fs/binfmt_em86.c:77:34: error: passing argument 2 of `copy_strings_kernel' from incompatible pointer type [-Werror=incompatible-pointer-types] retval = copy_strings_kernel(1, &i_name, bprm); ^ Link: http://lkml.kernel.org/r/1469525978-23359-1-git-send-email-wagi@monom.org Signed-off-by: Daniel Wagner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/binfmt_em86.c b/fs/binfmt_em86.c index 4905385..dd2d3f0 100644 --- a/fs/binfmt_em86.c +++ b/fs/binfmt_em86.c @@ -24,7 +24,8 @@ static int load_em86(struct linux_binprm *bprm) { - char *interp, *i_name, *i_arg; + const char *i_name, *i_arg; + char *interp; struct file * file; int retval; struct elfhdr elf_ex; -- cgit v0.10.2 From cae3d4ca6fd6872d8e9c21eff0e56398c938100a Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Tue, 2 Aug 2016 14:05:00 -0700 Subject: nilfs2: hide function name argument from nilfs_error() Simplify nilfs_error(), an output function used to report critical issues in file system. This renames the original nilfs_error() function to __nilfs_error() and redefines it as a macro to hide its function name argument within the macro. Every call site of nilfs_error() is changed to strip __func__ argument except nilfs_bmap_convert_error(); nilfs_bmap_convert_error() directly calls __nilfs_error() because it inherits caller's function name. Link: http://lkml.kernel.org/r/1464875891-5443-2-git-send-email-konishi.ryusuke@lab.ntt.co.jp Signed-off-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/nilfs2/bmap.c b/fs/nilfs2/bmap.c index f2a7877..01fb183 100644 --- a/fs/nilfs2/bmap.c +++ b/fs/nilfs2/bmap.c @@ -41,8 +41,8 @@ static int nilfs_bmap_convert_error(struct nilfs_bmap *bmap, struct inode *inode = bmap->b_inode; if (err == -EINVAL) { - nilfs_error(inode->i_sb, fname, - "broken bmap (inode number=%lu)", inode->i_ino); + __nilfs_error(inode->i_sb, fname, + "broken bmap (inode number=%lu)", inode->i_ino); err = -EIO; } return err; diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c index e506f4f..746956d 100644 --- a/fs/nilfs2/dir.c +++ b/fs/nilfs2/dir.c @@ -140,10 +140,9 @@ out: /* Too bad, we had an error */ Ebadsize: - nilfs_error(sb, "nilfs_check_page", + nilfs_error(sb, "size of directory #%lu is not a multiple of chunk size", - dir->i_ino - ); + dir->i_ino); goto fail; Eshort: error = "rec_len is smaller than minimal"; @@ -157,19 +156,18 @@ Enamelen: Espan: error = "directory entry across blocks"; bad_entry: - nilfs_error(sb, "nilfs_check_page", "bad entry in directory #%lu: %s - " - "offset=%lu, inode=%lu, rec_len=%d, name_len=%d", - dir->i_ino, error, (page->index<inode), + nilfs_error(sb, + "bad entry in directory #%lu: %s - offset=%lu, inode=%lu, rec_len=%d, name_len=%d", + dir->i_ino, error, (page->index << PAGE_SHIFT) + offs, + (unsigned long)le64_to_cpu(p->inode), rec_len, p->name_len); goto fail; Eend: p = (struct nilfs_dir_entry *)(kaddr + offs); - nilfs_error(sb, "nilfs_check_page", - "entry in directory #%lu spans the page boundary" - "offset=%lu, inode=%lu", - dir->i_ino, (page->index<inode)); + nilfs_error(sb, + "entry in directory #%lu spans the page boundary offset=%lu, inode=%lu", + dir->i_ino, (page->index << PAGE_SHIFT) + offs, + (unsigned long)le64_to_cpu(p->inode)); fail: SetPageError(page); return false; @@ -267,8 +265,7 @@ static int nilfs_readdir(struct file *file, struct dir_context *ctx) struct page *page = nilfs_get_page(inode, n); if (IS_ERR(page)) { - nilfs_error(sb, __func__, "bad page in #%lu", - inode->i_ino); + nilfs_error(sb, "bad page in #%lu", inode->i_ino); ctx->pos += PAGE_SIZE - offset; return -EIO; } @@ -278,8 +275,7 @@ static int nilfs_readdir(struct file *file, struct dir_context *ctx) NILFS_DIR_REC_LEN(1); for ( ; (char *)de <= limit; de = nilfs_next_entry(de)) { if (de->rec_len == 0) { - nilfs_error(sb, __func__, - "zero-length directory entry"); + nilfs_error(sb, "zero-length directory entry"); nilfs_put_page(page); return -EIO; } @@ -345,7 +341,7 @@ nilfs_find_entry(struct inode *dir, const struct qstr *qstr, kaddr += nilfs_last_byte(dir, n) - reclen; while ((char *) de <= kaddr) { if (de->rec_len == 0) { - nilfs_error(dir->i_sb, __func__, + nilfs_error(dir->i_sb, "zero-length directory entry"); nilfs_put_page(page); goto out; @@ -360,7 +356,7 @@ nilfs_find_entry(struct inode *dir, const struct qstr *qstr, n = 0; /* next page is past the blocks we've got */ if (unlikely(n > (dir->i_blocks >> (PAGE_SHIFT - 9)))) { - nilfs_error(dir->i_sb, __func__, + nilfs_error(dir->i_sb, "dir %lu size %lld exceeds block count %llu", dir->i_ino, dir->i_size, (unsigned long long)dir->i_blocks); @@ -469,7 +465,7 @@ int nilfs_add_link(struct dentry *dentry, struct inode *inode) goto got_it; } if (de->rec_len == 0) { - nilfs_error(dir->i_sb, __func__, + nilfs_error(dir->i_sb, "zero-length directory entry"); err = -EIO; goto out_unlock; @@ -541,7 +537,7 @@ int nilfs_delete_entry(struct nilfs_dir_entry *dir, struct page *page) while ((char *)de < (char *)dir) { if (de->rec_len == 0) { - nilfs_error(inode->i_sb, __func__, + nilfs_error(inode->i_sb, "zero-length directory entry"); err = -EIO; goto out; @@ -628,7 +624,7 @@ int nilfs_empty_dir(struct inode *inode) while ((char *)de <= kaddr) { if (de->rec_len == 0) { - nilfs_error(inode->i_sb, __func__, + nilfs_error(inode->i_sb, "zero-length directory entry (kaddr=%p, de=%p)", kaddr, de); goto not_empty; diff --git a/fs/nilfs2/ifile.c b/fs/nilfs2/ifile.c index 1d2b180..b1c9628 100644 --- a/fs/nilfs2/ifile.c +++ b/fs/nilfs2/ifile.c @@ -145,8 +145,7 @@ int nilfs_ifile_get_inode_block(struct inode *ifile, ino_t ino, int err; if (unlikely(!NILFS_VALID_INODE(sb, ino))) { - nilfs_error(sb, __func__, "bad inode number: %lu", - (unsigned long) ino); + nilfs_error(sb, "bad inode number: %lu", (unsigned long)ino); return -EINVAL; } diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h index b1d48bc..e482c78 100644 --- a/fs/nilfs2/nilfs.h +++ b/fs/nilfs2/nilfs.h @@ -299,10 +299,28 @@ static inline int nilfs_mark_inode_dirty_sync(struct inode *inode) /* super.c */ extern struct inode *nilfs_alloc_inode(struct super_block *); extern void nilfs_destroy_inode(struct inode *); + extern __printf(3, 4) -void nilfs_error(struct super_block *, const char *, const char *, ...); +void __nilfs_error(struct super_block *sb, const char *function, + const char *fmt, ...); extern __printf(3, 4) void nilfs_warning(struct super_block *, const char *, const char *, ...); + +#ifdef CONFIG_PRINTK + +#define nilfs_error(sb, fmt, ...) \ + __nilfs_error(sb, __func__, fmt, ##__VA_ARGS__) + +#else + +#define nilfs_error(sb, fmt, ...) \ + do { \ + no_printk(fmt, ##__VA_ARGS__); \ + __nilfs_error(sb, "", " "); \ + } while (0) + +#endif /* CONFIG_PRINTK */ + extern struct nilfs_super_block * nilfs_read_super_block(struct super_block *, u64, int, struct buffer_head **); extern int nilfs_store_magic_and_option(struct super_block *, diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 666107a..7fe497e 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -91,19 +91,21 @@ static void nilfs_set_error(struct super_block *sb) } /** - * nilfs_error() - report failure condition on a filesystem + * __nilfs_error() - report failure condition on a filesystem * - * nilfs_error() sets an ERROR_FS flag on the superblock as well as - * reporting an error message. It should be called when NILFS detects - * incoherences or defects of meta data on disk. As for sustainable - * errors such as a single-shot I/O error, nilfs_warning() or the printk() - * function should be used instead. + * __nilfs_error() sets an ERROR_FS flag on the superblock as well as + * reporting an error message. This function should be called when + * NILFS detects incoherences or defects of meta data on disk. * - * The segment constructor must not call this function because it can - * kill itself. + * This implements the body of nilfs_error() macro. Normally, + * nilfs_error() should be used. As for sustainable errors such as a + * single-shot I/O error, nilfs_warning() or printk() should be used + * instead. + * + * Callers should not add a trailing newline since this will do it. */ -void nilfs_error(struct super_block *sb, const char *function, - const char *fmt, ...) +void __nilfs_error(struct super_block *sb, const char *function, + const char *fmt, ...) { struct the_nilfs *nilfs = sb->s_fs_info; struct va_format vaf; -- cgit v0.10.2 From a66dfb0a91c211c77b5d4e503d3e760e2e566189 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Tue, 2 Aug 2016 14:05:02 -0700 Subject: nilfs2: add nilfs_msg() message interface Define an own output routine to replace bare use of printk() function. The output routine is implemented with a macro and a helper function, which are named nilfs_msg() and __nilfs_msg(), respectively. __nilfs_msg() formats a message like "NILFS (): ", prefixing it with a given log level, and terminates the statement with a newline. The "device-name" is optional to make it available in early stages; it will be omitted if a NULL pointer is passed to super block instance argument. nilfs_msg() wraps __nilfs_msg() and is removed if CONFIG_PRINTK is not set. Link: http://lkml.kernel.org/r/1464875891-5443-3-git-send-email-konishi.ryusuke@lab.ntt.co.jp Signed-off-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h index e482c78..b57ce41e 100644 --- a/fs/nilfs2/nilfs.h +++ b/fs/nilfs2/nilfs.h @@ -301,6 +301,9 @@ extern struct inode *nilfs_alloc_inode(struct super_block *); extern void nilfs_destroy_inode(struct inode *); extern __printf(3, 4) +void __nilfs_msg(struct super_block *sb, const char *level, + const char *fmt, ...); +extern __printf(3, 4) void __nilfs_error(struct super_block *sb, const char *function, const char *fmt, ...); extern __printf(3, 4) @@ -308,11 +311,15 @@ void nilfs_warning(struct super_block *, const char *, const char *, ...); #ifdef CONFIG_PRINTK +#define nilfs_msg(sb, level, fmt, ...) \ + __nilfs_msg(sb, level, fmt, ##__VA_ARGS__) #define nilfs_error(sb, fmt, ...) \ __nilfs_error(sb, __func__, fmt, ##__VA_ARGS__) #else +#define nilfs_msg(sb, level, fmt, ...) \ + no_printk(fmt, ##__VA_ARGS__) #define nilfs_error(sb, fmt, ...) \ do { \ no_printk(fmt, ##__VA_ARGS__); \ diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 7fe497e..86e3c00 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -71,6 +71,22 @@ struct kmem_cache *nilfs_btree_path_cache; static int nilfs_setup_super(struct super_block *sb, int is_mount); static int nilfs_remount(struct super_block *sb, int *flags, char *data); +void __nilfs_msg(struct super_block *sb, const char *level, const char *fmt, + ...) +{ + struct va_format vaf; + va_list args; + + va_start(args, fmt); + vaf.fmt = fmt; + vaf.va = &args; + if (sb) + printk("%sNILFS (%s): %pV\n", level, sb->s_id, &vaf); + else + printk("%sNILFS: %pV\n", level, &vaf); + va_end(args); +} + static void nilfs_set_error(struct super_block *sb) { struct the_nilfs *nilfs = sb->s_fs_info; -- cgit v0.10.2 From 6625689e159fa1d43572ee113713ab23bec03131 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Tue, 2 Aug 2016 14:05:06 -0700 Subject: nilfs2: embed a back pointer to super block instance in nilfs object Insert a back pointer to super block instance in nilfs object so that functions of nilfs2 easily refer to the super block instance. This simplifies replacement of printk() in the successive change. Link: http://lkml.kernel.org/r/1464875891-5443-4-git-send-email-konishi.ryusuke@lab.ntt.co.jp Signed-off-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 86e3c00..2d4d0be 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -1076,7 +1076,7 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent) __u64 cno; int err; - nilfs = alloc_nilfs(sb->s_bdev); + nilfs = alloc_nilfs(sb); if (!nilfs) return -ENOMEM; diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c index e9fd241..7021151 100644 --- a/fs/nilfs2/the_nilfs.c +++ b/fs/nilfs2/the_nilfs.c @@ -56,12 +56,12 @@ void nilfs_set_last_segment(struct the_nilfs *nilfs, /** * alloc_nilfs - allocate a nilfs object - * @bdev: block device to which the_nilfs is related + * @sb: super block instance * * Return Value: On success, pointer to the_nilfs is returned. * On error, NULL is returned. */ -struct the_nilfs *alloc_nilfs(struct block_device *bdev) +struct the_nilfs *alloc_nilfs(struct super_block *sb) { struct the_nilfs *nilfs; @@ -69,7 +69,8 @@ struct the_nilfs *alloc_nilfs(struct block_device *bdev) if (!nilfs) return NULL; - nilfs->ns_bdev = bdev; + nilfs->ns_sb = sb; + nilfs->ns_bdev = sb->s_bdev; atomic_set(&nilfs->ns_ndirtyblks, 0); init_rwsem(&nilfs->ns_sem); mutex_init(&nilfs->ns_snapshot_mount_mutex); diff --git a/fs/nilfs2/the_nilfs.h b/fs/nilfs2/the_nilfs.h index 79369fd..79d1421 100644 --- a/fs/nilfs2/the_nilfs.h +++ b/fs/nilfs2/the_nilfs.h @@ -43,6 +43,7 @@ enum { * struct the_nilfs - struct to supervise multiple nilfs mount points * @ns_flags: flags * @ns_flushed_device: flag indicating if all volatile data was flushed + * @ns_sb: back pointer to super block instance * @ns_bdev: block device * @ns_sem: semaphore for shared states * @ns_snapshot_mount_mutex: mutex to protect snapshot mounts @@ -102,6 +103,7 @@ struct the_nilfs { unsigned long ns_flags; int ns_flushed_device; + struct super_block *ns_sb; struct block_device *ns_bdev; struct rw_semaphore ns_sem; struct mutex ns_snapshot_mount_mutex; @@ -281,7 +283,7 @@ static inline int nilfs_sb_will_flip(struct the_nilfs *nilfs) } void nilfs_set_last_segment(struct the_nilfs *, sector_t, u64, __u64); -struct the_nilfs *alloc_nilfs(struct block_device *bdev); +struct the_nilfs *alloc_nilfs(struct super_block *sb); void destroy_nilfs(struct the_nilfs *nilfs); int init_nilfs(struct the_nilfs *nilfs, struct super_block *sb, char *data); int load_nilfs(struct the_nilfs *nilfs, struct super_block *sb); -- cgit v0.10.2 From feee880fa58254fcc1c78bc8b6446a435cc1baf0 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Tue, 2 Aug 2016 14:05:10 -0700 Subject: nilfs2: reduce bare use of printk() with nilfs_msg() Replace most use of printk() in nilfs2 implementation with nilfs_msg(), and reduce the following checkpatch.pl warning: "WARNING: Prefer [subsystem eg: netdev]_crit([subsystem]dev, ... then dev_crit(dev, ... then pr_crit(... to printk(KERN_CRIT ..." This patch also fixes a minor checkpatch warning "WARNING: quoted string split across lines" that often accompanies the prior warning, and amends message format as needed. Link: http://lkml.kernel.org/r/1464875891-5443-5-git-send-email-konishi.ryusuke@lab.ntt.co.jp Signed-off-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/nilfs2/btree.c b/fs/nilfs2/btree.c index 982d1e3..2c52693 100644 --- a/fs/nilfs2/btree.c +++ b/fs/nilfs2/btree.c @@ -339,12 +339,14 @@ static int nilfs_btree_node_lookup(const struct nilfs_btree_node *node, * nilfs_btree_node_broken - verify consistency of btree node * @node: btree node block to be examined * @size: node size (in bytes) + * @inode: host inode of btree * @blocknr: block number * * Return Value: If node is broken, 1 is returned. Otherwise, 0 is returned. */ static int nilfs_btree_node_broken(const struct nilfs_btree_node *node, - size_t size, sector_t blocknr) + size_t size, struct inode *inode, + sector_t blocknr) { int level, flags, nchildren; int ret = 0; @@ -358,9 +360,10 @@ static int nilfs_btree_node_broken(const struct nilfs_btree_node *node, (flags & NILFS_BTREE_NODE_ROOT) || nchildren < 0 || nchildren > NILFS_BTREE_NODE_NCHILDREN_MAX(size))) { - printk(KERN_CRIT "NILFS: bad btree node (blocknr=%llu): " - "level = %d, flags = 0x%x, nchildren = %d\n", - (unsigned long long)blocknr, level, flags, nchildren); + nilfs_msg(inode->i_sb, KERN_CRIT, + "bad btree node (ino=%lu, blocknr=%llu): level = %d, flags = 0x%x, nchildren = %d", + inode->i_ino, (unsigned long long)blocknr, level, + flags, nchildren); ret = 1; } return ret; @@ -369,12 +372,12 @@ static int nilfs_btree_node_broken(const struct nilfs_btree_node *node, /** * nilfs_btree_root_broken - verify consistency of btree root node * @node: btree root node to be examined - * @ino: inode number + * @inode: host inode of btree * * Return Value: If node is broken, 1 is returned. Otherwise, 0 is returned. */ static int nilfs_btree_root_broken(const struct nilfs_btree_node *node, - unsigned long ino) + struct inode *inode) { int level, flags, nchildren; int ret = 0; @@ -387,8 +390,9 @@ static int nilfs_btree_root_broken(const struct nilfs_btree_node *node, level >= NILFS_BTREE_LEVEL_MAX || nchildren < 0 || nchildren > NILFS_BTREE_ROOT_NCHILDREN_MAX)) { - pr_crit("NILFS: bad btree root (inode number=%lu): level = %d, flags = 0x%x, nchildren = %d\n", - ino, level, flags, nchildren); + nilfs_msg(inode->i_sb, KERN_CRIT, + "bad btree root (ino=%lu): level = %d, flags = 0x%x, nchildren = %d", + inode->i_ino, level, flags, nchildren); ret = 1; } return ret; @@ -396,13 +400,15 @@ static int nilfs_btree_root_broken(const struct nilfs_btree_node *node, int nilfs_btree_broken_node_block(struct buffer_head *bh) { + struct inode *inode; int ret; if (buffer_nilfs_checked(bh)) return 0; + inode = bh->b_page->mapping->host; ret = nilfs_btree_node_broken((struct nilfs_btree_node *)bh->b_data, - bh->b_size, bh->b_blocknr); + bh->b_size, inode, bh->b_blocknr); if (likely(!ret)) set_buffer_nilfs_checked(bh); return ret; @@ -448,13 +454,15 @@ nilfs_btree_get_node(const struct nilfs_bmap *btree, return node; } -static int -nilfs_btree_bad_node(struct nilfs_btree_node *node, int level) +static int nilfs_btree_bad_node(const struct nilfs_bmap *btree, + struct nilfs_btree_node *node, int level) { if (unlikely(nilfs_btree_node_get_level(node) != level)) { dump_stack(); - printk(KERN_CRIT "NILFS: btree level mismatch: %d != %d\n", - nilfs_btree_node_get_level(node), level); + nilfs_msg(btree->b_inode->i_sb, KERN_CRIT, + "btree level mismatch (ino=%lu): %d != %d", + btree->b_inode->i_ino, + nilfs_btree_node_get_level(node), level); return 1; } return 0; @@ -568,7 +576,7 @@ static int nilfs_btree_do_lookup(const struct nilfs_bmap *btree, return ret; node = nilfs_btree_get_nonroot_node(path, level); - if (nilfs_btree_bad_node(node, level)) + if (nilfs_btree_bad_node(btree, node, level)) return -EINVAL; if (!found) found = nilfs_btree_node_lookup(node, key, &index); @@ -616,7 +624,7 @@ static int nilfs_btree_do_lookup_last(const struct nilfs_bmap *btree, if (ret < 0) return ret; node = nilfs_btree_get_nonroot_node(path, level); - if (nilfs_btree_bad_node(node, level)) + if (nilfs_btree_bad_node(btree, node, level)) return -EINVAL; index = nilfs_btree_node_get_nchildren(node) - 1; ptr = nilfs_btree_node_get_ptr(node, index, ncmax); @@ -2072,8 +2080,10 @@ static int nilfs_btree_propagate(struct nilfs_bmap *btree, ret = nilfs_btree_do_lookup(btree, path, key, NULL, level + 1, 0); if (ret < 0) { if (unlikely(ret == -ENOENT)) - printk(KERN_CRIT "%s: key = %llu, level == %d\n", - __func__, (unsigned long long)key, level); + nilfs_msg(btree->b_inode->i_sb, KERN_CRIT, + "writing node/leaf block does not appear in b-tree (ino=%lu) at key=%llu, level=%d", + btree->b_inode->i_ino, + (unsigned long long)key, level); goto out; } @@ -2110,12 +2120,11 @@ static void nilfs_btree_add_dirty_buffer(struct nilfs_bmap *btree, if (level < NILFS_BTREE_LEVEL_NODE_MIN || level >= NILFS_BTREE_LEVEL_MAX) { dump_stack(); - printk(KERN_WARNING - "%s: invalid btree level: %d (key=%llu, ino=%lu, " - "blocknr=%llu)\n", - __func__, level, (unsigned long long)key, - NILFS_BMAP_I(btree)->vfs_inode.i_ino, - (unsigned long long)bh->b_blocknr); + nilfs_msg(btree->b_inode->i_sb, KERN_WARNING, + "invalid btree level: %d (key=%llu, ino=%lu, blocknr=%llu)", + level, (unsigned long long)key, + btree->b_inode->i_ino, + (unsigned long long)bh->b_blocknr); return; } @@ -2394,8 +2403,7 @@ int nilfs_btree_init(struct nilfs_bmap *bmap) __nilfs_btree_init(bmap); - if (nilfs_btree_root_broken(nilfs_btree_get_root(bmap), - bmap->b_inode->i_ino)) + if (nilfs_btree_root_broken(nilfs_btree_get_root(bmap), bmap->b_inode)) ret = -EIO; return ret; } diff --git a/fs/nilfs2/cpfile.c b/fs/nilfs2/cpfile.c index 8a3d3b6..19d9f4a 100644 --- a/fs/nilfs2/cpfile.c +++ b/fs/nilfs2/cpfile.c @@ -332,9 +332,9 @@ int nilfs_cpfile_delete_checkpoints(struct inode *cpfile, int ret, ncps, nicps, nss, count, i; if (unlikely(start == 0 || start > end)) { - printk(KERN_ERR "%s: invalid range of checkpoint numbers: " - "[%llu, %llu)\n", __func__, - (unsigned long long)start, (unsigned long long)end); + nilfs_msg(cpfile->i_sb, KERN_ERR, + "cannot delete checkpoints: invalid range [%llu, %llu)", + (unsigned long long)start, (unsigned long long)end); return -EINVAL; } @@ -386,9 +386,9 @@ int nilfs_cpfile_delete_checkpoints(struct inode *cpfile, cpfile, cno); if (ret == 0) continue; - printk(KERN_ERR - "%s: cannot delete block\n", - __func__); + nilfs_msg(cpfile->i_sb, KERN_ERR, + "error %d deleting checkpoint block", + ret); break; } } @@ -991,14 +991,12 @@ int nilfs_cpfile_read(struct super_block *sb, size_t cpsize, int err; if (cpsize > sb->s_blocksize) { - printk(KERN_ERR - "NILFS: too large checkpoint size: %zu bytes.\n", - cpsize); + nilfs_msg(sb, KERN_ERR, + "too large checkpoint size: %zu bytes", cpsize); return -EINVAL; } else if (cpsize < NILFS_MIN_CHECKPOINT_SIZE) { - printk(KERN_ERR - "NILFS: too small checkpoint size: %zu bytes.\n", - cpsize); + nilfs_msg(sb, KERN_ERR, + "too small checkpoint size: %zu bytes", cpsize); return -EINVAL; } diff --git a/fs/nilfs2/dat.c b/fs/nilfs2/dat.c index 7367610..dffedb2 100644 --- a/fs/nilfs2/dat.c +++ b/fs/nilfs2/dat.c @@ -349,10 +349,11 @@ int nilfs_dat_move(struct inode *dat, __u64 vblocknr, sector_t blocknr) kaddr = kmap_atomic(entry_bh->b_page); entry = nilfs_palloc_block_get_entry(dat, vblocknr, entry_bh, kaddr); if (unlikely(entry->de_blocknr == cpu_to_le64(0))) { - printk(KERN_CRIT "%s: vbn = %llu, [%llu, %llu)\n", __func__, - (unsigned long long)vblocknr, - (unsigned long long)le64_to_cpu(entry->de_start), - (unsigned long long)le64_to_cpu(entry->de_end)); + nilfs_msg(dat->i_sb, KERN_CRIT, + "%s: invalid vblocknr = %llu, [%llu, %llu)", + __func__, (unsigned long long)vblocknr, + (unsigned long long)le64_to_cpu(entry->de_start), + (unsigned long long)le64_to_cpu(entry->de_end)); kunmap_atomic(kaddr); brelse(entry_bh); return -EINVAL; @@ -479,14 +480,12 @@ int nilfs_dat_read(struct super_block *sb, size_t entry_size, int err; if (entry_size > sb->s_blocksize) { - printk(KERN_ERR - "NILFS: too large DAT entry size: %zu bytes.\n", - entry_size); + nilfs_msg(sb, KERN_ERR, "too large DAT entry size: %zu bytes", + entry_size); return -EINVAL; } else if (entry_size < NILFS_MIN_DAT_ENTRY_SIZE) { - printk(KERN_ERR - "NILFS: too small DAT entry size: %zu bytes.\n", - entry_size); + nilfs_msg(sb, KERN_ERR, "too small DAT entry size: %zu bytes", + entry_size); return -EINVAL; } diff --git a/fs/nilfs2/direct.c b/fs/nilfs2/direct.c index 251a4492..96e3ed0 100644 --- a/fs/nilfs2/direct.c +++ b/fs/nilfs2/direct.c @@ -337,14 +337,16 @@ static int nilfs_direct_assign(struct nilfs_bmap *bmap, key = nilfs_bmap_data_get_key(bmap, *bh); if (unlikely(key > NILFS_DIRECT_KEY_MAX)) { - printk(KERN_CRIT "%s: invalid key: %llu\n", __func__, - (unsigned long long)key); + nilfs_msg(bmap->b_inode->i_sb, KERN_CRIT, + "%s (ino=%lu): invalid key: %llu", __func__, + bmap->b_inode->i_ino, (unsigned long long)key); return -EINVAL; } ptr = nilfs_direct_get_ptr(bmap, key); if (unlikely(ptr == NILFS_BMAP_INVALID_PTR)) { - printk(KERN_CRIT "%s: invalid pointer: %llu\n", __func__, - (unsigned long long)ptr); + nilfs_msg(bmap->b_inode->i_sb, KERN_CRIT, + "%s (ino=%lu): invalid pointer: %llu", __func__, + bmap->b_inode->i_ino, (unsigned long long)ptr); return -EINVAL; } diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c index a0ebdb1..a965fcf 100644 --- a/fs/nilfs2/inode.c +++ b/fs/nilfs2/inode.c @@ -112,13 +112,10 @@ int nilfs_get_block(struct inode *inode, sector_t blkoff, * However, the page having this block must * be locked in this case. */ - printk(KERN_WARNING - "nilfs_get_block: a race condition " - "while inserting a data block. " - "(inode number=%lu, file block " - "offset=%llu)\n", - inode->i_ino, - (unsigned long long)blkoff); + nilfs_msg(inode->i_sb, KERN_WARNING, + "%s (ino=%lu): a race condition while inserting a data block at offset=%llu", + __func__, inode->i_ino, + (unsigned long long)blkoff); err = 0; } nilfs_transaction_abort(inode->i_sb); diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c index 358b57e..827283f 100644 --- a/fs/nilfs2/ioctl.c +++ b/fs/nilfs2/ioctl.c @@ -584,27 +584,25 @@ static int nilfs_ioctl_move_inode_block(struct inode *inode, if (unlikely(ret < 0)) { if (ret == -ENOENT) - printk(KERN_CRIT - "%s: invalid virtual block address (%s): " - "ino=%llu, cno=%llu, offset=%llu, " - "blocknr=%llu, vblocknr=%llu\n", - __func__, vdesc->vd_flags ? "node" : "data", - (unsigned long long)vdesc->vd_ino, - (unsigned long long)vdesc->vd_cno, - (unsigned long long)vdesc->vd_offset, - (unsigned long long)vdesc->vd_blocknr, - (unsigned long long)vdesc->vd_vblocknr); + nilfs_msg(inode->i_sb, KERN_CRIT, + "%s: invalid virtual block address (%s): ino=%llu, cno=%llu, offset=%llu, blocknr=%llu, vblocknr=%llu", + __func__, vdesc->vd_flags ? "node" : "data", + (unsigned long long)vdesc->vd_ino, + (unsigned long long)vdesc->vd_cno, + (unsigned long long)vdesc->vd_offset, + (unsigned long long)vdesc->vd_blocknr, + (unsigned long long)vdesc->vd_vblocknr); return ret; } if (unlikely(!list_empty(&bh->b_assoc_buffers))) { - printk(KERN_CRIT "%s: conflicting %s buffer: ino=%llu, " - "cno=%llu, offset=%llu, blocknr=%llu, vblocknr=%llu\n", - __func__, vdesc->vd_flags ? "node" : "data", - (unsigned long long)vdesc->vd_ino, - (unsigned long long)vdesc->vd_cno, - (unsigned long long)vdesc->vd_offset, - (unsigned long long)vdesc->vd_blocknr, - (unsigned long long)vdesc->vd_vblocknr); + nilfs_msg(inode->i_sb, KERN_CRIT, + "%s: conflicting %s buffer: ino=%llu, cno=%llu, offset=%llu, blocknr=%llu, vblocknr=%llu", + __func__, vdesc->vd_flags ? "node" : "data", + (unsigned long long)vdesc->vd_ino, + (unsigned long long)vdesc->vd_cno, + (unsigned long long)vdesc->vd_offset, + (unsigned long long)vdesc->vd_blocknr, + (unsigned long long)vdesc->vd_vblocknr); brelse(bh); return -EEXIST; } @@ -854,8 +852,8 @@ int nilfs_ioctl_prepare_clean_segments(struct the_nilfs *nilfs, return 0; failed: - printk(KERN_ERR "NILFS: GC failed during preparation: %s: err=%d\n", - msg, ret); + nilfs_msg(nilfs->ns_sb, KERN_ERR, "error %d preparing GC: %s", ret, + msg); return ret; } @@ -963,10 +961,11 @@ static int nilfs_ioctl_clean_segments(struct inode *inode, struct file *filp, } ret = nilfs_ioctl_move_blocks(inode->i_sb, &argv[0], kbufs[0]); - if (ret < 0) - printk(KERN_ERR "NILFS: GC failed during preparation: " - "cannot read source blocks: err=%d\n", ret); - else { + if (ret < 0) { + nilfs_msg(inode->i_sb, KERN_ERR, + "error %d preparing GC: cannot read source blocks", + ret); + } else { if (nilfs_sb_need_update(nilfs)) set_nilfs_discontinued(nilfs); ret = nilfs_clean_segments(inode->i_sb, argv, kbufs); diff --git a/fs/nilfs2/recovery.c b/fs/nilfs2/recovery.c index d893dc9..5139efe 100644 --- a/fs/nilfs2/recovery.c +++ b/fs/nilfs2/recovery.c @@ -54,38 +54,37 @@ struct nilfs_recovery_block { }; -static int nilfs_warn_segment_error(int err) +static int nilfs_warn_segment_error(struct super_block *sb, int err) { + const char *msg = NULL; + switch (err) { case NILFS_SEG_FAIL_IO: - printk(KERN_WARNING - "NILFS warning: I/O error on loading last segment\n"); + nilfs_msg(sb, KERN_ERR, "I/O error reading segment"); return -EIO; case NILFS_SEG_FAIL_MAGIC: - printk(KERN_WARNING - "NILFS warning: Segment magic number invalid\n"); + msg = "Magic number mismatch"; break; case NILFS_SEG_FAIL_SEQ: - printk(KERN_WARNING - "NILFS warning: Sequence number mismatch\n"); + msg = "Sequence number mismatch"; break; case NILFS_SEG_FAIL_CHECKSUM_SUPER_ROOT: - printk(KERN_WARNING - "NILFS warning: Checksum error in super root\n"); + msg = "Checksum error in super root"; break; case NILFS_SEG_FAIL_CHECKSUM_FULL: - printk(KERN_WARNING - "NILFS warning: Checksum error in segment payload\n"); + msg = "Checksum error in segment payload"; break; case NILFS_SEG_FAIL_CONSISTENCY: - printk(KERN_WARNING - "NILFS warning: Inconsistent segment\n"); + msg = "Inconsistency found"; break; case NILFS_SEG_NO_SUPER_ROOT: - printk(KERN_WARNING - "NILFS warning: No super root in the last segment\n"); + msg = "No super root in the last segment"; break; + default: + nilfs_msg(sb, KERN_ERR, "unrecognized segment error %d", err); + return -EINVAL; } + nilfs_msg(sb, KERN_WARNING, "invalid segment: %s", msg); return -EINVAL; } @@ -178,7 +177,7 @@ int nilfs_read_super_root_block(struct the_nilfs *nilfs, sector_t sr_block, brelse(bh_sr); failed: - return nilfs_warn_segment_error(ret); + return nilfs_warn_segment_error(nilfs->ns_sb, ret); } /** @@ -553,11 +552,10 @@ static int nilfs_recover_dsync_blocks(struct the_nilfs *nilfs, put_page(page); failed_inode: - printk(KERN_WARNING - "NILFS warning: error recovering data block " - "(err=%d, ino=%lu, block-offset=%llu)\n", - err, (unsigned long)rb->ino, - (unsigned long long)rb->blkoff); + nilfs_msg(sb, KERN_WARNING, + "error %d recovering data block (ino=%lu, block-offset=%llu)", + err, (unsigned long)rb->ino, + (unsigned long long)rb->blkoff); if (!err2) err2 = err; next: @@ -680,8 +678,8 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs, } if (nsalvaged_blocks) { - printk(KERN_INFO "NILFS (device %s): salvaged %lu blocks\n", - sb->s_id, nsalvaged_blocks); + nilfs_msg(sb, KERN_INFO, "salvaged %lu blocks", + nsalvaged_blocks); ri->ri_need_recovery = NILFS_RECOVERY_ROLLFORWARD_DONE; } out: @@ -692,10 +690,9 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs, confused: err = -EINVAL; failed: - printk(KERN_ERR - "NILFS (device %s): Error roll-forwarding " - "(err=%d, pseg block=%llu). ", - sb->s_id, err, (unsigned long long)pseg_start); + nilfs_msg(sb, KERN_ERR, + "error %d roll-forwarding partial segment at blocknr = %llu", + err, (unsigned long long)pseg_start); goto out; } @@ -715,9 +712,8 @@ static void nilfs_finish_roll_forward(struct the_nilfs *nilfs, set_buffer_dirty(bh); err = sync_dirty_buffer(bh); if (unlikely(err)) - printk(KERN_WARNING - "NILFS warning: buffer sync write failed during " - "post-cleaning of recovery.\n"); + nilfs_msg(nilfs->ns_sb, KERN_WARNING, + "buffer sync write failed during post-cleaning of recovery."); brelse(bh); } @@ -752,8 +748,8 @@ int nilfs_salvage_orphan_logs(struct the_nilfs *nilfs, err = nilfs_attach_checkpoint(sb, ri->ri_cno, true, &root); if (unlikely(err)) { - printk(KERN_ERR - "NILFS: error loading the latest checkpoint.\n"); + nilfs_msg(sb, KERN_ERR, + "error %d loading the latest checkpoint", err); return err; } @@ -764,8 +760,9 @@ int nilfs_salvage_orphan_logs(struct the_nilfs *nilfs, if (ri->ri_need_recovery == NILFS_RECOVERY_ROLLFORWARD_DONE) { err = nilfs_prepare_segment_for_recovery(nilfs, sb, ri); if (unlikely(err)) { - printk(KERN_ERR "NILFS: Error preparing segments for " - "recovery.\n"); + nilfs_msg(sb, KERN_ERR, + "error %d preparing segment for recovery", + err); goto failed; } @@ -778,8 +775,9 @@ int nilfs_salvage_orphan_logs(struct the_nilfs *nilfs, nilfs_detach_log_writer(sb); if (unlikely(err)) { - printk(KERN_ERR "NILFS: Oops! recovery failed. " - "(err=%d)\n", err); + nilfs_msg(sb, KERN_ERR, + "error %d writing segment for recovery", + err); goto failed; } @@ -961,5 +959,5 @@ int nilfs_search_super_root(struct the_nilfs *nilfs, failed: brelse(bh_sum); nilfs_dispose_segment_list(&segments); - return (ret < 0) ? ret : nilfs_warn_segment_error(ret); + return ret < 0 ? ret : nilfs_warn_segment_error(nilfs->ns_sb, ret); } diff --git a/fs/nilfs2/segbuf.c b/fs/nilfs2/segbuf.c index a962d7d..6f87b2a 100644 --- a/fs/nilfs2/segbuf.c +++ b/fs/nilfs2/segbuf.c @@ -514,7 +514,11 @@ static int nilfs_segbuf_wait(struct nilfs_segment_buffer *segbuf) } while (--segbuf->sb_nbio > 0); if (unlikely(atomic_read(&segbuf->sb_err) > 0)) { - printk(KERN_ERR "NILFS: IO error writing segment\n"); + nilfs_msg(segbuf->sb_super, KERN_ERR, + "I/O error writing log (start-blocknr=%llu, block-count=%lu) in segment %llu", + (unsigned long long)segbuf->sb_pseg_start, + segbuf->sb_sum.nblocks, + (unsigned long long)segbuf->sb_segnum); err = -EIO; } return err; diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index e78b68a8..1cc9685 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -150,7 +150,8 @@ static void nilfs_dispose_list(struct the_nilfs *, struct list_head *, int); #define nilfs_cnt32_lt(a, b) nilfs_cnt32_gt(b, a) #define nilfs_cnt32_le(a, b) nilfs_cnt32_ge(b, a) -static int nilfs_prepare_segment_lock(struct nilfs_transaction_info *ti) +static int nilfs_prepare_segment_lock(struct super_block *sb, + struct nilfs_transaction_info *ti) { struct nilfs_transaction_info *cur_ti = current->journal_info; void *save = NULL; @@ -164,8 +165,7 @@ static int nilfs_prepare_segment_lock(struct nilfs_transaction_info *ti) * it is saved and will be restored on * nilfs_transaction_commit(). */ - printk(KERN_WARNING - "NILFS warning: journal info from a different FS\n"); + nilfs_msg(sb, KERN_WARNING, "journal info from a different FS"); save = current->journal_info; } if (!ti) { @@ -215,7 +215,7 @@ int nilfs_transaction_begin(struct super_block *sb, int vacancy_check) { struct the_nilfs *nilfs; - int ret = nilfs_prepare_segment_lock(ti); + int ret = nilfs_prepare_segment_lock(sb, ti); struct nilfs_transaction_info *trace_ti; if (unlikely(ret < 0)) @@ -2467,9 +2467,9 @@ int nilfs_clean_segments(struct super_block *sb, struct nilfs_argv *argv, int ret = nilfs_discard_segments(nilfs, sci->sc_freesegs, sci->sc_nfreesegs); if (ret) { - printk(KERN_WARNING - "NILFS warning: error %d on discard request, " - "turning discards off for the device\n", ret); + nilfs_msg(sb, KERN_WARNING, + "error %d on discard request, turning discards off for the device", + ret); nilfs_clear_opt(nilfs, DISCARD); } } @@ -2551,10 +2551,9 @@ static int nilfs_segctor_thread(void *arg) /* start sync. */ sci->sc_task = current; wake_up(&sci->sc_wait_task); /* for nilfs_segctor_start_thread() */ - printk(KERN_INFO - "segctord starting. Construction interval = %lu seconds, " - "CP frequency < %lu seconds\n", - sci->sc_interval / HZ, sci->sc_mjcp_freq / HZ); + nilfs_msg(sci->sc_super, KERN_INFO, + "segctord starting. Construction interval = %lu seconds, CP frequency < %lu seconds", + sci->sc_interval / HZ, sci->sc_mjcp_freq / HZ); spin_lock(&sci->sc_state_lock); loop: @@ -2628,8 +2627,8 @@ static int nilfs_segctor_start_thread(struct nilfs_sc_info *sci) if (IS_ERR(t)) { int err = PTR_ERR(t); - printk(KERN_ERR "NILFS: error %d creating segctord thread\n", - err); + nilfs_msg(sci->sc_super, KERN_ERR, + "error %d creating segctord thread", err); return err; } wait_event(sci->sc_wait_task, sci->sc_task != NULL); diff --git a/fs/nilfs2/sufile.c b/fs/nilfs2/sufile.c index 1963595..5b495c4 100644 --- a/fs/nilfs2/sufile.c +++ b/fs/nilfs2/sufile.c @@ -181,9 +181,9 @@ int nilfs_sufile_updatev(struct inode *sufile, __u64 *segnumv, size_t nsegs, down_write(&NILFS_MDT(sufile)->mi_sem); for (seg = segnumv; seg < segnumv + nsegs; seg++) { if (unlikely(*seg >= nilfs_sufile_get_nsegments(sufile))) { - printk(KERN_WARNING - "%s: invalid segment number: %llu\n", __func__, - (unsigned long long)*seg); + nilfs_msg(sufile->i_sb, KERN_WARNING, + "%s: invalid segment number: %llu", + __func__, (unsigned long long)*seg); nerr++; } } @@ -240,8 +240,9 @@ int nilfs_sufile_update(struct inode *sufile, __u64 segnum, int create, int ret; if (unlikely(segnum >= nilfs_sufile_get_nsegments(sufile))) { - printk(KERN_WARNING "%s: invalid segment number: %llu\n", - __func__, (unsigned long long)segnum); + nilfs_msg(sufile->i_sb, KERN_WARNING, + "%s: invalid segment number: %llu", + __func__, (unsigned long long)segnum); return -EINVAL; } down_write(&NILFS_MDT(sufile)->mi_sem); @@ -419,8 +420,9 @@ void nilfs_sufile_do_cancel_free(struct inode *sufile, __u64 segnum, kaddr = kmap_atomic(su_bh->b_page); su = nilfs_sufile_block_get_segment_usage(sufile, segnum, su_bh, kaddr); if (unlikely(!nilfs_segment_usage_clean(su))) { - printk(KERN_WARNING "%s: segment %llu must be clean\n", - __func__, (unsigned long long)segnum); + nilfs_msg(sufile->i_sb, KERN_WARNING, + "%s: segment %llu must be clean", __func__, + (unsigned long long)segnum); kunmap_atomic(kaddr); return; } @@ -476,8 +478,9 @@ void nilfs_sufile_do_free(struct inode *sufile, __u64 segnum, kaddr = kmap_atomic(su_bh->b_page); su = nilfs_sufile_block_get_segment_usage(sufile, segnum, su_bh, kaddr); if (nilfs_segment_usage_clean(su)) { - printk(KERN_WARNING "%s: segment %llu is already clean\n", - __func__, (unsigned long long)segnum); + nilfs_msg(sufile->i_sb, KERN_WARNING, + "%s: segment %llu is already clean", + __func__, (unsigned long long)segnum); kunmap_atomic(kaddr); return; } @@ -1175,14 +1178,12 @@ int nilfs_sufile_read(struct super_block *sb, size_t susize, int err; if (susize > sb->s_blocksize) { - printk(KERN_ERR - "NILFS: too large segment usage size: %zu bytes.\n", - susize); + nilfs_msg(sb, KERN_ERR, + "too large segment usage size: %zu bytes", susize); return -EINVAL; } else if (susize < NILFS_MIN_SEGMENT_USAGE_SIZE) { - printk(KERN_ERR - "NILFS: too small segment usage size: %zu bytes.\n", - susize); + nilfs_msg(sb, KERN_ERR, + "too small segment usage size: %zu bytes", susize); return -EINVAL; } diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 2d4d0be..90c62b4 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -214,8 +214,8 @@ static int nilfs_sync_super(struct super_block *sb, int flag) } if (unlikely(err)) { - printk(KERN_ERR - "NILFS: unable to write superblock (err=%d)\n", err); + nilfs_msg(sb, KERN_ERR, "unable to write superblock: err=%d", + err); if (err == -EIO && nilfs->ns_sbh[1]) { /* * sbp[0] points to newer log than sbp[1], @@ -285,8 +285,7 @@ struct nilfs_super_block **nilfs_prepare_super(struct super_block *sb, sbp[1]->s_magic == cpu_to_le16(NILFS_SUPER_MAGIC)) { memcpy(sbp[0], sbp[1], nilfs->ns_sbsize); } else { - printk(KERN_CRIT "NILFS: superblock broke on dev %s\n", - sb->s_id); + nilfs_msg(sb, KERN_CRIT, "superblock broke"); return NULL; } } else if (sbp[1] && @@ -396,9 +395,9 @@ static int nilfs_move_2nd_super(struct super_block *sb, loff_t sb2off) offset = sb2off & (nilfs->ns_blocksize - 1); nsbh = sb_getblk(sb, newblocknr); if (!nsbh) { - printk(KERN_WARNING - "NILFS warning: unable to move secondary superblock " - "to block %llu\n", (unsigned long long)newblocknr); + nilfs_msg(sb, KERN_WARNING, + "unable to move secondary superblock to block %llu", + (unsigned long long)newblocknr); ret = -EIO; goto out; } @@ -561,10 +560,9 @@ int nilfs_attach_checkpoint(struct super_block *sb, __u64 cno, int curr_mnt, up_read(&nilfs->ns_segctor_sem); if (unlikely(err)) { if (err == -ENOENT || err == -EINVAL) { - printk(KERN_ERR - "NILFS: Invalid checkpoint " - "(checkpoint number=%llu)\n", - (unsigned long long)cno); + nilfs_msg(sb, KERN_ERR, + "Invalid checkpoint (checkpoint number=%llu)", + (unsigned long long)cno); err = -EINVAL; } goto failed; @@ -660,9 +658,8 @@ static int nilfs_statfs(struct dentry *dentry, struct kstatfs *buf) err = nilfs_ifile_count_free_inodes(root->ifile, &nmaxinodes, &nfreeinodes); if (unlikely(err)) { - printk(KERN_WARNING - "NILFS warning: fail to count free inodes: err %d.\n", - err); + nilfs_msg(sb, KERN_WARNING, + "failed to count free inodes: err=%d", err); if (err == -ERANGE) { /* * If nilfs_palloc_count_max_entries() returns @@ -794,9 +791,9 @@ static int parse_options(char *options, struct super_block *sb, int is_remount) break; case Opt_snapshot: if (is_remount) { - printk(KERN_ERR - "NILFS: \"%s\" option is invalid " - "for remount.\n", p); + nilfs_msg(sb, KERN_ERR, + "\"%s\" option is invalid for remount", + p); return 0; } break; @@ -810,8 +807,8 @@ static int parse_options(char *options, struct super_block *sb, int is_remount) nilfs_clear_opt(nilfs, DISCARD); break; default: - printk(KERN_ERR - "NILFS: Unrecognized mount option \"%s\"\n", p); + nilfs_msg(sb, KERN_ERR, + "unrecognized mount option \"%s\"", p); return 0; } } @@ -847,12 +844,10 @@ static int nilfs_setup_super(struct super_block *sb, int is_mount) mnt_count = le16_to_cpu(sbp[0]->s_mnt_count); if (nilfs->ns_mount_state & NILFS_ERROR_FS) { - printk(KERN_WARNING - "NILFS warning: mounting fs with errors\n"); + nilfs_msg(sb, KERN_WARNING, "mounting fs with errors"); #if 0 } else if (max_mnt_count >= 0 && mnt_count >= max_mnt_count) { - printk(KERN_WARNING - "NILFS warning: maximal mount count reached\n"); + nilfs_msg(sb, KERN_WARNING, "maximal mount count reached"); #endif } if (!max_mnt_count) @@ -915,17 +910,17 @@ int nilfs_check_feature_compatibility(struct super_block *sb, features = le64_to_cpu(sbp->s_feature_incompat) & ~NILFS_FEATURE_INCOMPAT_SUPP; if (features) { - printk(KERN_ERR "NILFS: couldn't mount because of unsupported " - "optional features (%llx)\n", - (unsigned long long)features); + nilfs_msg(sb, KERN_ERR, + "couldn't mount because of unsupported optional features (%llx)", + (unsigned long long)features); return -EINVAL; } features = le64_to_cpu(sbp->s_feature_compat_ro) & ~NILFS_FEATURE_COMPAT_RO_SUPP; if (!(sb->s_flags & MS_RDONLY) && features) { - printk(KERN_ERR "NILFS: couldn't mount RDWR because of " - "unsupported optional features (%llx)\n", - (unsigned long long)features); + nilfs_msg(sb, KERN_ERR, + "couldn't mount RDWR because of unsupported optional features (%llx)", + (unsigned long long)features); return -EINVAL; } return 0; @@ -941,13 +936,13 @@ static int nilfs_get_root_dentry(struct super_block *sb, inode = nilfs_iget(sb, root, NILFS_ROOT_INO); if (IS_ERR(inode)) { - printk(KERN_ERR "NILFS: get root inode failed\n"); ret = PTR_ERR(inode); + nilfs_msg(sb, KERN_ERR, "error %d getting root inode", ret); goto out; } if (!S_ISDIR(inode->i_mode) || !inode->i_blocks || !inode->i_size) { iput(inode); - printk(KERN_ERR "NILFS: corrupt root inode.\n"); + nilfs_msg(sb, KERN_ERR, "corrupt root inode"); ret = -EINVAL; goto out; } @@ -975,7 +970,7 @@ static int nilfs_get_root_dentry(struct super_block *sb, return ret; failed_dentry: - printk(KERN_ERR "NILFS: get root dentry failed\n"); + nilfs_msg(sb, KERN_ERR, "error %d getting root dentry", ret); goto out; } @@ -995,18 +990,18 @@ static int nilfs_attach_snapshot(struct super_block *s, __u64 cno, ret = (ret == -ENOENT) ? -EINVAL : ret; goto out; } else if (!ret) { - printk(KERN_ERR "NILFS: The specified checkpoint is " - "not a snapshot (checkpoint number=%llu).\n", - (unsigned long long)cno); + nilfs_msg(s, KERN_ERR, + "The specified checkpoint is not a snapshot (checkpoint number=%llu)", + (unsigned long long)cno); ret = -EINVAL; goto out; } ret = nilfs_attach_checkpoint(s, cno, false, &root); if (ret) { - printk(KERN_ERR "NILFS: error loading snapshot " - "(checkpoint number=%llu).\n", - (unsigned long long)cno); + nilfs_msg(s, KERN_ERR, + "error %d while loading snapshot (checkpoint number=%llu)", + ret, (unsigned long long)cno); goto out; } ret = nilfs_get_root_dentry(s, root, root_dentry); @@ -1101,8 +1096,9 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent) cno = nilfs_last_cno(nilfs); err = nilfs_attach_checkpoint(sb, cno, true, &fsroot); if (err) { - printk(KERN_ERR "NILFS: error loading last checkpoint " - "(checkpoint number=%llu).\n", (unsigned long long)cno); + nilfs_msg(sb, KERN_ERR, + "error %d while loading last checkpoint (checkpoint number=%llu)", + err, (unsigned long long)cno); goto failed_unload; } @@ -1162,9 +1158,8 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data) err = -EINVAL; if (!nilfs_valid_fs(nilfs)) { - printk(KERN_WARNING "NILFS (device %s): couldn't " - "remount because the filesystem is in an " - "incomplete recovery state.\n", sb->s_id); + nilfs_msg(sb, KERN_WARNING, + "couldn't remount because the filesystem is in an incomplete recovery state"); goto restore_opts; } @@ -1196,10 +1191,9 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data) ~NILFS_FEATURE_COMPAT_RO_SUPP; up_read(&nilfs->ns_sem); if (features) { - printk(KERN_WARNING "NILFS (device %s): couldn't " - "remount RDWR because of unsupported optional " - "features (%llx)\n", - sb->s_id, (unsigned long long)features); + nilfs_msg(sb, KERN_WARNING, + "couldn't remount RDWR because of unsupported optional features (%llx)", + (unsigned long long)features); err = -EROFS; goto restore_opts; } @@ -1262,8 +1256,8 @@ static int nilfs_identify(char *data, struct nilfs_super_data *sd) } } if (ret) - printk(KERN_ERR - "NILFS: invalid mount option: %s\n", p); + nilfs_msg(NULL, KERN_ERR, + "invalid mount option: %s", p); } if (!options) break; @@ -1344,10 +1338,10 @@ nilfs_mount(struct file_system_type *fs_type, int flags, } else if (!sd.cno) { if (nilfs_tree_is_busy(s->s_root)) { if ((flags ^ s->s_flags) & MS_RDONLY) { - printk(KERN_ERR "NILFS: the device already " - "has a %s mount.\n", - (s->s_flags & MS_RDONLY) ? - "read-only" : "read/write"); + nilfs_msg(s, KERN_ERR, + "the device already has a %s mount.", + (s->s_flags & MS_RDONLY) ? + "read-only" : "read/write"); err = -EBUSY; goto failed_super; } diff --git a/fs/nilfs2/sysfs.c b/fs/nilfs2/sysfs.c index 8ffa42b..8e57bb9 100644 --- a/fs/nilfs2/sysfs.c +++ b/fs/nilfs2/sysfs.c @@ -272,8 +272,8 @@ nilfs_checkpoints_checkpoints_number_show(struct nilfs_checkpoints_attr *attr, err = nilfs_cpfile_get_stat(nilfs->ns_cpfile, &cpstat); up_read(&nilfs->ns_segctor_sem); if (err < 0) { - printk(KERN_ERR "NILFS: unable to get checkpoint stat: err=%d\n", - err); + nilfs_msg(nilfs->ns_sb, KERN_ERR, + "unable to get checkpoint stat: err=%d", err); return err; } @@ -295,8 +295,8 @@ nilfs_checkpoints_snapshots_number_show(struct nilfs_checkpoints_attr *attr, err = nilfs_cpfile_get_stat(nilfs->ns_cpfile, &cpstat); up_read(&nilfs->ns_segctor_sem); if (err < 0) { - printk(KERN_ERR "NILFS: unable to get checkpoint stat: err=%d\n", - err); + nilfs_msg(nilfs->ns_sb, KERN_ERR, + "unable to get checkpoint stat: err=%d", err); return err; } @@ -414,8 +414,8 @@ nilfs_segments_dirty_segments_show(struct nilfs_segments_attr *attr, err = nilfs_sufile_get_stat(nilfs->ns_sufile, &sustat); up_read(&nilfs->ns_segctor_sem); if (err < 0) { - printk(KERN_ERR "NILFS: unable to get segment stat: err=%d\n", - err); + nilfs_msg(nilfs->ns_sb, KERN_ERR, + "unable to get segment stat: err=%d", err); return err; } @@ -789,14 +789,15 @@ nilfs_superblock_sb_update_frequency_store(struct nilfs_superblock_attr *attr, err = kstrtouint(skip_spaces(buf), 0, &val); if (err) { - printk(KERN_ERR "NILFS: unable to convert string: err=%d\n", - err); + nilfs_msg(nilfs->ns_sb, KERN_ERR, + "unable to convert string: err=%d", err); return err; } if (val < NILFS_SB_FREQ) { val = NILFS_SB_FREQ; - printk(KERN_WARNING "NILFS: superblock update frequency cannot be lesser than 10 seconds\n"); + nilfs_msg(nilfs->ns_sb, KERN_WARNING, + "superblock update frequency cannot be lesser than 10 seconds"); } down_write(&nilfs->ns_sem); @@ -999,7 +1000,8 @@ int nilfs_sysfs_create_device_group(struct super_block *sb) nilfs->ns_dev_subgroups = kzalloc(devgrp_size, GFP_KERNEL); if (unlikely(!nilfs->ns_dev_subgroups)) { err = -ENOMEM; - printk(KERN_ERR "NILFS: unable to allocate memory for device group\n"); + nilfs_msg(sb, KERN_ERR, + "unable to allocate memory for device group"); goto failed_create_device_group; } @@ -1109,15 +1111,15 @@ int __init nilfs_sysfs_init(void) nilfs_kset = kset_create_and_add(NILFS_ROOT_GROUP_NAME, NULL, fs_kobj); if (!nilfs_kset) { err = -ENOMEM; - printk(KERN_ERR "NILFS: unable to create sysfs entry: err %d\n", - err); + nilfs_msg(NULL, KERN_ERR, + "unable to create sysfs entry: err=%d", err); goto failed_sysfs_init; } err = sysfs_create_group(&nilfs_kset->kobj, &nilfs_feature_attr_group); if (unlikely(err)) { - printk(KERN_ERR "NILFS: unable to create feature group: err %d\n", - err); + nilfs_msg(NULL, KERN_ERR, + "unable to create feature group: err=%d", err); goto cleanup_sysfs_init; } diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c index 7021151..2dd75bf 100644 --- a/fs/nilfs2/the_nilfs.c +++ b/fs/nilfs2/the_nilfs.c @@ -192,7 +192,10 @@ static int nilfs_store_log_cursor(struct the_nilfs *nilfs, nilfs_get_segnum_of_block(nilfs, nilfs->ns_last_pseg); nilfs->ns_cno = nilfs->ns_last_cno + 1; if (nilfs->ns_segnum >= nilfs->ns_nsegments) { - printk(KERN_ERR "NILFS invalid last segment number.\n"); + nilfs_msg(nilfs->ns_sb, KERN_ERR, + "pointed segment number is out of range: segnum=%llu, nsegments=%lu", + (unsigned long long)nilfs->ns_segnum, + nilfs->ns_nsegments); ret = -EINVAL; } return ret; @@ -216,12 +219,12 @@ int load_nilfs(struct the_nilfs *nilfs, struct super_block *sb) int err; if (!valid_fs) { - printk(KERN_WARNING "NILFS warning: mounting unchecked fs\n"); + nilfs_msg(sb, KERN_WARNING, "mounting unchecked fs"); if (s_flags & MS_RDONLY) { - printk(KERN_INFO "NILFS: INFO: recovery " - "required for readonly filesystem.\n"); - printk(KERN_INFO "NILFS: write access will " - "be enabled during recovery.\n"); + nilfs_msg(sb, KERN_INFO, + "recovery required for readonly filesystem"); + nilfs_msg(sb, KERN_INFO, + "write access will be enabled during recovery"); } } @@ -236,13 +239,12 @@ int load_nilfs(struct the_nilfs *nilfs, struct super_block *sb) goto scan_error; if (!nilfs_valid_sb(sbp[1])) { - printk(KERN_WARNING - "NILFS warning: unable to fall back to spare" - "super block\n"); + nilfs_msg(sb, KERN_WARNING, + "unable to fall back to spare super block"); goto scan_error; } - printk(KERN_INFO - "NILFS: try rollback from an earlier position\n"); + nilfs_msg(sb, KERN_INFO, + "trying rollback from an earlier position"); /* * restore super block with its spare and reconfigure @@ -255,10 +257,9 @@ int load_nilfs(struct the_nilfs *nilfs, struct super_block *sb) /* verify consistency between two super blocks */ blocksize = BLOCK_SIZE << le32_to_cpu(sbp[0]->s_log_block_size); if (blocksize != nilfs->ns_blocksize) { - printk(KERN_WARNING - "NILFS warning: blocksize differs between " - "two super blocks (%d != %d)\n", - blocksize, nilfs->ns_blocksize); + nilfs_msg(sb, KERN_WARNING, + "blocksize differs between two super blocks (%d != %d)", + blocksize, nilfs->ns_blocksize); goto scan_error; } @@ -277,7 +278,8 @@ int load_nilfs(struct the_nilfs *nilfs, struct super_block *sb) err = nilfs_load_super_root(nilfs, sb, ri.ri_super_root); if (unlikely(err)) { - printk(KERN_ERR "NILFS: error loading super root.\n"); + nilfs_msg(sb, KERN_ERR, "error %d while loading super root", + err); goto failed; } @@ -288,30 +290,29 @@ int load_nilfs(struct the_nilfs *nilfs, struct super_block *sb) __u64 features; if (nilfs_test_opt(nilfs, NORECOVERY)) { - printk(KERN_INFO "NILFS: norecovery option specified. " - "skipping roll-forward recovery\n"); + nilfs_msg(sb, KERN_INFO, + "norecovery option specified, skipping roll-forward recovery"); goto skip_recovery; } features = le64_to_cpu(nilfs->ns_sbp[0]->s_feature_compat_ro) & ~NILFS_FEATURE_COMPAT_RO_SUPP; if (features) { - printk(KERN_ERR "NILFS: couldn't proceed with " - "recovery because of unsupported optional " - "features (%llx)\n", - (unsigned long long)features); + nilfs_msg(sb, KERN_ERR, + "couldn't proceed with recovery because of unsupported optional features (%llx)", + (unsigned long long)features); err = -EROFS; goto failed_unload; } if (really_read_only) { - printk(KERN_ERR "NILFS: write access " - "unavailable, cannot proceed.\n"); + nilfs_msg(sb, KERN_ERR, + "write access unavailable, cannot proceed"); err = -EROFS; goto failed_unload; } sb->s_flags &= ~MS_RDONLY; } else if (nilfs_test_opt(nilfs, NORECOVERY)) { - printk(KERN_ERR "NILFS: recovery cancelled because norecovery " - "option was specified for a read/write mount\n"); + nilfs_msg(sb, KERN_ERR, + "recovery cancelled because norecovery option was specified for a read/write mount"); err = -EINVAL; goto failed_unload; } @@ -326,11 +327,12 @@ int load_nilfs(struct the_nilfs *nilfs, struct super_block *sb) up_write(&nilfs->ns_sem); if (err) { - printk(KERN_ERR "NILFS: failed to update super block. " - "recovery unfinished.\n"); + nilfs_msg(sb, KERN_ERR, + "error %d updating super block. recovery unfinished.", + err); goto failed_unload; } - printk(KERN_INFO "NILFS: recovery complete.\n"); + nilfs_msg(sb, KERN_INFO, "recovery complete"); skip_recovery: nilfs_clear_recovery_info(&ri); @@ -338,7 +340,7 @@ int load_nilfs(struct the_nilfs *nilfs, struct super_block *sb) return 0; scan_error: - printk(KERN_ERR "NILFS: error searching super root.\n"); + nilfs_msg(sb, KERN_ERR, "error %d while searching super root", err); goto failed; failed_unload: @@ -385,12 +387,11 @@ static int nilfs_store_disk_layout(struct the_nilfs *nilfs, struct nilfs_super_block *sbp) { if (le32_to_cpu(sbp->s_rev_level) < NILFS_MIN_SUPP_REV) { - printk(KERN_ERR "NILFS: unsupported revision " - "(superblock rev.=%d.%d, current rev.=%d.%d). " - "Please check the version of mkfs.nilfs.\n", - le32_to_cpu(sbp->s_rev_level), - le16_to_cpu(sbp->s_minor_rev_level), - NILFS_CURRENT_REV, NILFS_MINOR_REV); + nilfs_msg(nilfs->ns_sb, KERN_ERR, + "unsupported revision (superblock rev.=%d.%d, current rev.=%d.%d). Please check the version of mkfs.nilfs(2).", + le32_to_cpu(sbp->s_rev_level), + le16_to_cpu(sbp->s_minor_rev_level), + NILFS_CURRENT_REV, NILFS_MINOR_REV); return -EINVAL; } nilfs->ns_sbsize = le16_to_cpu(sbp->s_bytes); @@ -399,12 +400,14 @@ static int nilfs_store_disk_layout(struct the_nilfs *nilfs, nilfs->ns_inode_size = le16_to_cpu(sbp->s_inode_size); if (nilfs->ns_inode_size > nilfs->ns_blocksize) { - printk(KERN_ERR "NILFS: too large inode size: %d bytes.\n", - nilfs->ns_inode_size); + nilfs_msg(nilfs->ns_sb, KERN_ERR, + "too large inode size: %d bytes", + nilfs->ns_inode_size); return -EINVAL; } else if (nilfs->ns_inode_size < NILFS_MIN_INODE_SIZE) { - printk(KERN_ERR "NILFS: too small inode size: %d bytes.\n", - nilfs->ns_inode_size); + nilfs_msg(nilfs->ns_sb, KERN_ERR, + "too small inode size: %d bytes", + nilfs->ns_inode_size); return -EINVAL; } @@ -412,7 +415,9 @@ static int nilfs_store_disk_layout(struct the_nilfs *nilfs, nilfs->ns_blocks_per_segment = le32_to_cpu(sbp->s_blocks_per_segment); if (nilfs->ns_blocks_per_segment < NILFS_SEG_MIN_BLOCKS) { - printk(KERN_ERR "NILFS: too short segment.\n"); + nilfs_msg(nilfs->ns_sb, KERN_ERR, + "too short segment: %lu blocks", + nilfs->ns_blocks_per_segment); return -EINVAL; } @@ -421,7 +426,9 @@ static int nilfs_store_disk_layout(struct the_nilfs *nilfs, le32_to_cpu(sbp->s_r_segments_percentage); if (nilfs->ns_r_segments_percentage < 1 || nilfs->ns_r_segments_percentage > 99) { - printk(KERN_ERR "NILFS: invalid reserved segments percentage.\n"); + nilfs_msg(nilfs->ns_sb, KERN_ERR, + "invalid reserved segments percentage: %lu", + nilfs->ns_r_segments_percentage); return -EINVAL; } @@ -505,16 +512,16 @@ static int nilfs_load_super_block(struct the_nilfs *nilfs, if (!sbp[0]) { if (!sbp[1]) { - printk(KERN_ERR "NILFS: unable to read superblock\n"); + nilfs_msg(sb, KERN_ERR, "unable to read superblock"); return -EIO; } - printk(KERN_WARNING - "NILFS warning: unable to read primary superblock " - "(blocksize = %d)\n", blocksize); + nilfs_msg(sb, KERN_WARNING, + "unable to read primary superblock (blocksize = %d)", + blocksize); } else if (!sbp[1]) { - printk(KERN_WARNING - "NILFS warning: unable to read secondary superblock " - "(blocksize = %d)\n", blocksize); + nilfs_msg(sb, KERN_WARNING, + "unable to read secondary superblock (blocksize = %d)", + blocksize); } /* @@ -536,14 +543,14 @@ static int nilfs_load_super_block(struct the_nilfs *nilfs, } if (!valid[swp]) { nilfs_release_super_block(nilfs); - printk(KERN_ERR "NILFS: Can't find nilfs on dev %s.\n", - sb->s_id); + nilfs_msg(sb, KERN_ERR, "couldn't find nilfs on the device"); return -EINVAL; } if (!valid[!swp]) - printk(KERN_WARNING "NILFS warning: broken superblock. " - "using spare superblock (blocksize = %d).\n", blocksize); + nilfs_msg(sb, KERN_WARNING, + "broken superblock, retrying with spare superblock (blocksize = %d)", + blocksize); if (swp) nilfs_swap_super_block(nilfs); @@ -577,7 +584,7 @@ int init_nilfs(struct the_nilfs *nilfs, struct super_block *sb, char *data) blocksize = sb_min_blocksize(sb, NILFS_MIN_BLOCK_SIZE); if (!blocksize) { - printk(KERN_ERR "NILFS: unable to set blocksize\n"); + nilfs_msg(sb, KERN_ERR, "unable to set blocksize"); err = -EINVAL; goto out; } @@ -596,8 +603,9 @@ int init_nilfs(struct the_nilfs *nilfs, struct super_block *sb, char *data) blocksize = BLOCK_SIZE << le32_to_cpu(sbp->s_log_block_size); if (blocksize < NILFS_MIN_BLOCK_SIZE || blocksize > NILFS_MAX_BLOCK_SIZE) { - printk(KERN_ERR "NILFS: couldn't mount because of unsupported " - "filesystem blocksize %d\n", blocksize); + nilfs_msg(sb, KERN_ERR, + "couldn't mount because of unsupported filesystem blocksize %d", + blocksize); err = -EINVAL; goto failed_sbh; } @@ -605,10 +613,9 @@ int init_nilfs(struct the_nilfs *nilfs, struct super_block *sb, char *data) int hw_blocksize = bdev_logical_block_size(sb->s_bdev); if (blocksize < hw_blocksize) { - printk(KERN_ERR - "NILFS: blocksize %d too small for device " - "(sector-size = %d).\n", - blocksize, hw_blocksize); + nilfs_msg(sb, KERN_ERR, + "blocksize %d too small for device (sector-size = %d)", + blocksize, hw_blocksize); err = -EINVAL; goto failed_sbh; } -- cgit v0.10.2 From d6517deb014954d3229910e46f3b85b7ad80db3e Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Tue, 2 Aug 2016 14:05:14 -0700 Subject: nilfs2: replace nilfs_warning() with nilfs_msg() Use nilfs_msg() to output warning messages and get rid of nilfs_warning() function. This also removes function names from the messages unless we embed them explicitly in format strings. Instead, some messages are revised to clarify the context. [arnd@arndb.de: avoid warning about unused variables] Link: http://lkml.kernel.org/r/20160615201945.3348205-1-arnd@arndb.de Link: http://lkml.kernel.org/r/1464875891-5443-6-git-send-email-konishi.ryusuke@lab.ntt.co.jp Signed-off-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/nilfs2/alloc.c b/fs/nilfs2/alloc.c index 1a85d94..2c90e28 100644 --- a/fs/nilfs2/alloc.c +++ b/fs/nilfs2/alloc.c @@ -622,10 +622,10 @@ void nilfs_palloc_commit_free_entry(struct inode *inode, lock = nilfs_mdt_bgl_lock(inode, group); if (!nilfs_clear_bit_atomic(lock, group_offset, bitmap)) - nilfs_warning(inode->i_sb, __func__, - "entry number %llu already freed: ino=%lu", - (unsigned long long)req->pr_entry_nr, - (unsigned long)inode->i_ino); + nilfs_msg(inode->i_sb, KERN_WARNING, + "%s (ino=%lu): entry number %llu already freed", + __func__, inode->i_ino, + (unsigned long long)req->pr_entry_nr); else nilfs_palloc_group_desc_add_entries(desc, lock, 1); @@ -663,10 +663,10 @@ void nilfs_palloc_abort_alloc_entry(struct inode *inode, lock = nilfs_mdt_bgl_lock(inode, group); if (!nilfs_clear_bit_atomic(lock, group_offset, bitmap)) - nilfs_warning(inode->i_sb, __func__, - "entry number %llu already freed: ino=%lu", - (unsigned long long)req->pr_entry_nr, - (unsigned long)inode->i_ino); + nilfs_msg(inode->i_sb, KERN_WARNING, + "%s (ino=%lu): entry number %llu already freed", + __func__, inode->i_ino, + (unsigned long long)req->pr_entry_nr); else nilfs_palloc_group_desc_add_entries(desc, lock, 1); @@ -772,10 +772,10 @@ int nilfs_palloc_freev(struct inode *inode, __u64 *entry_nrs, size_t nitems) do { if (!nilfs_clear_bit_atomic(lock, group_offset, bitmap)) { - nilfs_warning(inode->i_sb, __func__, - "entry number %llu already freed: ino=%lu", - (unsigned long long)entry_nrs[j], - (unsigned long)inode->i_ino); + nilfs_msg(inode->i_sb, KERN_WARNING, + "%s (ino=%lu): entry number %llu already freed", + __func__, inode->i_ino, + (unsigned long long)entry_nrs[j]); } else { n++; } @@ -816,12 +816,11 @@ int nilfs_palloc_freev(struct inode *inode, __u64 *entry_nrs, size_t nitems) for (k = 0; k < nempties; k++) { ret = nilfs_palloc_delete_entry_block(inode, last_nrs[k]); - if (ret && ret != -ENOENT) { - nilfs_warning(inode->i_sb, __func__, - "failed to delete block of entry %llu: ino=%lu, err=%d", - (unsigned long long)last_nrs[k], - (unsigned long)inode->i_ino, ret); - } + if (ret && ret != -ENOENT) + nilfs_msg(inode->i_sb, KERN_WARNING, + "error %d deleting block that object (entry=%llu, ino=%lu) belongs to", + ret, (unsigned long long)last_nrs[k], + inode->i_ino); } desc_kaddr = kmap_atomic(desc_bh->b_page); @@ -835,12 +834,10 @@ int nilfs_palloc_freev(struct inode *inode, __u64 *entry_nrs, size_t nitems) if (nfree == nilfs_palloc_entries_per_group(inode)) { ret = nilfs_palloc_delete_bitmap_block(inode, group); - if (ret && ret != -ENOENT) { - nilfs_warning(inode->i_sb, __func__, - "failed to delete bitmap block of group %lu: ino=%lu, err=%d", - group, - (unsigned long)inode->i_ino, ret); - } + if (ret && ret != -ENOENT) + nilfs_msg(inode->i_sb, KERN_WARNING, + "error %d deleting bitmap block of group=%lu, ino=%lu", + ret, group, inode->i_ino); } } return 0; diff --git a/fs/nilfs2/ifile.c b/fs/nilfs2/ifile.c index b1c9628..b8fa45c 100644 --- a/fs/nilfs2/ifile.c +++ b/fs/nilfs2/ifile.c @@ -151,8 +151,8 @@ int nilfs_ifile_get_inode_block(struct inode *ifile, ino_t ino, err = nilfs_palloc_get_entry_block(ifile, ino, 0, out_bh); if (unlikely(err)) - nilfs_warning(sb, __func__, "unable to read inode: %lu", - (unsigned long) ino); + nilfs_msg(sb, KERN_WARNING, "error %d reading inode: ino=%lu", + err, (unsigned long)ino); return err; } diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c index a965fcf..b286b35 100644 --- a/fs/nilfs2/inode.c +++ b/fs/nilfs2/inode.c @@ -723,9 +723,9 @@ repeat: goto repeat; failed: - nilfs_warning(ii->vfs_inode.i_sb, __func__, - "failed to truncate bmap (ino=%lu, err=%d)", - ii->vfs_inode.i_ino, ret); + nilfs_msg(ii->vfs_inode.i_sb, KERN_WARNING, + "error %d truncating bmap (ino=%lu)", ret, + ii->vfs_inode.i_ino); } void nilfs_truncate(struct inode *inode) @@ -936,9 +936,9 @@ int nilfs_set_file_dirty(struct inode *inode, unsigned int nr_dirty) * This will happen when somebody is freeing * this inode. */ - nilfs_warning(inode->i_sb, __func__, - "cannot get inode (ino=%lu)", - inode->i_ino); + nilfs_msg(inode->i_sb, KERN_WARNING, + "cannot set file dirty (ino=%lu): the file is being freed", + inode->i_ino); spin_unlock(&nilfs->ns_inode_lock); return -EINVAL; /* * NILFS_I_DIRTY may remain for @@ -959,8 +959,9 @@ int __nilfs_mark_inode_dirty(struct inode *inode, int flags) err = nilfs_load_inode_block(inode, &ibh); if (unlikely(err)) { - nilfs_warning(inode->i_sb, __func__, - "failed to reget inode block."); + nilfs_msg(inode->i_sb, KERN_WARNING, + "cannot mark inode dirty (ino=%lu): error %d loading inode block", + inode->i_ino, err); return err; } nilfs_update_inode(inode, ibh, flags); @@ -986,8 +987,8 @@ void nilfs_dirty_inode(struct inode *inode, int flags) struct nilfs_mdt_info *mdi = NILFS_MDT(inode); if (is_bad_inode(inode)) { - nilfs_warning(inode->i_sb, __func__, - "tried to mark bad_inode dirty. ignored."); + nilfs_msg(inode->i_sb, KERN_WARNING, + "tried to mark bad_inode dirty. ignored."); dump_stack(); return; } diff --git a/fs/nilfs2/namei.c b/fs/nilfs2/namei.c index 1ec8ae5..dbcf1dc 100644 --- a/fs/nilfs2/namei.c +++ b/fs/nilfs2/namei.c @@ -283,9 +283,9 @@ static int nilfs_do_unlink(struct inode *dir, struct dentry *dentry) goto out; if (!inode->i_nlink) { - nilfs_warning(inode->i_sb, __func__, - "deleting nonexistent file (%lu), %d", - inode->i_ino, inode->i_nlink); + nilfs_msg(inode->i_sb, KERN_WARNING, + "deleting nonexistent file (ino=%lu), %d", + inode->i_ino, inode->i_nlink); set_nlink(inode, 1); } err = nilfs_delete_entry(de, page); diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h index b57ce41e..46fbd4e 100644 --- a/fs/nilfs2/nilfs.h +++ b/fs/nilfs2/nilfs.h @@ -306,8 +306,6 @@ void __nilfs_msg(struct super_block *sb, const char *level, extern __printf(3, 4) void __nilfs_error(struct super_block *sb, const char *function, const char *fmt, ...); -extern __printf(3, 4) -void nilfs_warning(struct super_block *, const char *, const char *, ...); #ifdef CONFIG_PRINTK @@ -319,7 +317,10 @@ void nilfs_warning(struct super_block *, const char *, const char *, ...); #else #define nilfs_msg(sb, level, fmt, ...) \ - no_printk(fmt, ##__VA_ARGS__) + do { \ + no_printk(fmt, ##__VA_ARGS__); \ + (void)(sb); \ + } while (0) #define nilfs_error(sb, fmt, ...) \ do { \ no_printk(fmt, ##__VA_ARGS__); \ diff --git a/fs/nilfs2/page.c b/fs/nilfs2/page.c index d97ba5f..eaccf12 100644 --- a/fs/nilfs2/page.c +++ b/fs/nilfs2/page.c @@ -403,11 +403,10 @@ void nilfs_clear_dirty_page(struct page *page, bool silent) BUG_ON(!PageLocked(page)); - if (!silent) { - nilfs_warning(sb, __func__, - "discard page: offset %lld, ino %lu", - page_offset(page), inode->i_ino); - } + if (!silent) + nilfs_msg(sb, KERN_WARNING, + "discard dirty page: offset=%lld, ino=%lu", + page_offset(page), inode->i_ino); ClearPageUptodate(page); ClearPageMappedToDisk(page); @@ -422,11 +421,11 @@ void nilfs_clear_dirty_page(struct page *page, bool silent) bh = head = page_buffers(page); do { lock_buffer(bh); - if (!silent) { - nilfs_warning(sb, __func__, - "discard block %llu, size %zu", - (u64)bh->b_blocknr, bh->b_size); - } + if (!silent) + nilfs_msg(sb, KERN_WARNING, + "discard dirty block: blocknr=%llu, size=%zu", + (u64)bh->b_blocknr, bh->b_size); + set_mask_bits(&bh->b_state, clear_bits, 0); unlock_buffer(bh); } while (bh = bh->b_this_page, bh != head); diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index 1cc9685..7e1864c 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -1951,8 +1951,9 @@ static int nilfs_segctor_collect_dirty_files(struct nilfs_sc_info *sci, err = nilfs_ifile_get_inode_block( ifile, ii->vfs_inode.i_ino, &ibh); if (unlikely(err)) { - nilfs_warning(sci->sc_super, __func__, - "failed to get inode block."); + nilfs_msg(sci->sc_super, KERN_WARNING, + "log writer: error %d getting inode block (ino=%lu)", + err, ii->vfs_inode.i_ino); return err; } mark_buffer_dirty(ibh); @@ -2458,8 +2459,7 @@ int nilfs_clean_segments(struct super_block *sb, struct nilfs_argv *argv, if (likely(!err)) break; - nilfs_warning(sb, __func__, - "segment construction failed. (err=%d)", err); + nilfs_msg(sb, KERN_WARNING, "error %d cleaning segments", err); set_current_state(TASK_INTERRUPTIBLE); schedule_timeout(sci->sc_interval); } @@ -2738,14 +2738,14 @@ static void nilfs_segctor_destroy(struct nilfs_sc_info *sci) nilfs_segctor_write_out(sci); if (!list_empty(&sci->sc_dirty_files)) { - nilfs_warning(sci->sc_super, __func__, - "dirty file(s) after the final construction"); + nilfs_msg(sci->sc_super, KERN_WARNING, + "disposed unprocessed dirty file(s) when stopping log writer"); nilfs_dispose_list(nilfs, &sci->sc_dirty_files, 1); } if (!list_empty(&sci->sc_iput_queue)) { - nilfs_warning(sci->sc_super, __func__, - "iput queue is not empty"); + nilfs_msg(sci->sc_super, KERN_WARNING, + "disposed unprocessed inode(s) in iput queue when stopping log writer"); nilfs_dispose_list(nilfs, &sci->sc_iput_queue, 1); } @@ -2821,8 +2821,8 @@ void nilfs_detach_log_writer(struct super_block *sb) spin_lock(&nilfs->ns_inode_lock); if (!list_empty(&nilfs->ns_dirty_files)) { list_splice_init(&nilfs->ns_dirty_files, &garbage_list); - nilfs_warning(sb, __func__, - "Hit dirty file after stopped log writer"); + nilfs_msg(sb, KERN_WARNING, + "disposed unprocessed dirty file(s) when detaching log writer"); } spin_unlock(&nilfs->ns_inode_lock); up_write(&nilfs->ns_segctor_sem); diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 90c62b4..33ba6f7 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -115,8 +115,7 @@ static void nilfs_set_error(struct super_block *sb) * * This implements the body of nilfs_error() macro. Normally, * nilfs_error() should be used. As for sustainable errors such as a - * single-shot I/O error, nilfs_warning() or printk() should be used - * instead. + * single-shot I/O error, nilfs_msg() should be used instead. * * Callers should not add a trailing newline since this will do it. */ @@ -151,24 +150,6 @@ void __nilfs_error(struct super_block *sb, const char *function, sb->s_id); } -void nilfs_warning(struct super_block *sb, const char *function, - const char *fmt, ...) -{ - struct va_format vaf; - va_list args; - - va_start(args, fmt); - - vaf.fmt = fmt; - vaf.va = &args; - - printk(KERN_WARNING "NILFS warning (device %s): %s: %pV\n", - sb->s_id, function, &vaf); - - va_end(args); -} - - struct inode *nilfs_alloc_inode(struct super_block *sb) { struct nilfs_inode_info *ii; -- cgit v0.10.2 From 39a9dcca61a3d1375b9440676cbfc541804cd217 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Tue, 2 Aug 2016 14:05:17 -0700 Subject: nilfs2: emit error message when I/O error is detected When nilfs returned -EIO as an error code, it's not always clear if it came from the underlying block device or not. This will mend the issue by having low level I/O routines of nilfs output an error message when they detected an I/O error. Link: http://lkml.kernel.org/r/1464875891-5443-7-git-send-email-konishi.ryusuke@lab.ntt.co.jp Signed-off-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/nilfs2/btree.c b/fs/nilfs2/btree.c index 2c52693..2e315f9 100644 --- a/fs/nilfs2/btree.c +++ b/fs/nilfs2/btree.c @@ -517,6 +517,9 @@ static int __nilfs_btree_get_block(const struct nilfs_bmap *btree, __u64 ptr, out_no_wait: if (!buffer_uptodate(bh)) { + nilfs_msg(btree->b_inode->i_sb, KERN_ERR, + "I/O error reading b-tree node block (ino=%lu, blocknr=%llu)", + btree->b_inode->i_ino, (unsigned long long)ptr); brelse(bh); return -EIO; } diff --git a/fs/nilfs2/gcinode.c b/fs/nilfs2/gcinode.c index e9148f9..853a831 100644 --- a/fs/nilfs2/gcinode.c +++ b/fs/nilfs2/gcinode.c @@ -148,8 +148,15 @@ int nilfs_gccache_submit_read_node(struct inode *inode, sector_t pbn, int nilfs_gccache_wait_and_mark_dirty(struct buffer_head *bh) { wait_on_buffer(bh); - if (!buffer_uptodate(bh)) + if (!buffer_uptodate(bh)) { + struct inode *inode = bh->b_page->mapping->host; + + nilfs_msg(inode->i_sb, KERN_ERR, + "I/O error reading %s block for GC (ino=%lu, vblocknr=%llu)", + buffer_nilfs_node(bh) ? "node" : "data", + inode->i_ino, (unsigned long long)bh->b_blocknr); return -EIO; + } if (buffer_dirty(bh)) return -EEXIST; diff --git a/fs/nilfs2/mdt.c b/fs/nilfs2/mdt.c index 0d7b71f..d56d3a5 100644 --- a/fs/nilfs2/mdt.c +++ b/fs/nilfs2/mdt.c @@ -207,8 +207,12 @@ static int nilfs_mdt_read_block(struct inode *inode, unsigned long block, out_no_wait: err = -EIO; - if (!buffer_uptodate(first_bh)) + if (!buffer_uptodate(first_bh)) { + nilfs_msg(inode->i_sb, KERN_ERR, + "I/O error reading meta-data file (ino=%lu, block-offset=%lu)", + inode->i_ino, block); goto failed_bh; + } out: *out_bh = first_bh; return 0; -- cgit v0.10.2 From aceb4170bb2ba88c5327cc69b9c91a708c7f7046 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Tue, 2 Aug 2016 14:05:19 -0700 Subject: nilfs2: do not use yield() Use cond_resched() instead of yield() in the loop of nilfs_transaction_lock() since the usage corresponds to the "be nice for others" case that the comment of yield() says. This removes the following checkpatch.pl warning: "WARNING: Using yield() is generally wrong. See yield() kernel-doc (sched/core.c)" Link: http://lkml.kernel.org/r/1464875891-5443-8-git-send-email-konishi.ryusuke@lab.ntt.co.jp Signed-off-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index 7e1864c..5a97282 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -373,7 +373,7 @@ static void nilfs_transaction_lock(struct super_block *sb, nilfs_segctor_do_immediate_flush(sci); up_write(&nilfs->ns_segctor_sem); - yield(); + cond_resched(); } if (gcflag) ti->ti_flags |= NILFS_TI_GC; -- cgit v0.10.2 From a7d3f104da57eecb2b9881127d6bdf9abe7fde99 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Tue, 2 Aug 2016 14:05:22 -0700 Subject: nilfs2: refactor parser of snapshot mount option Move parser of snapshot mount option to a separate function nilfs_parse_snapshot_option(), replace simple_strtoull() with kstrtoull() to avoid checkpatch.pl warning "WARNING: simple_strtoull is obsolete, use kstrtoull instead", and refine the error message of the parser. Link: http://lkml.kernel.org/r/1464875891-5443-9-git-send-email-konishi.ryusuke@lab.ntt.co.jp Signed-off-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 33ba6f7..c95d369 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -1205,6 +1205,38 @@ struct nilfs_super_data { int flags; }; +static int nilfs_parse_snapshot_option(const char *option, + const substring_t *arg, + struct nilfs_super_data *sd) +{ + unsigned long long val; + const char *msg = NULL; + int err; + + if (!(sd->flags & MS_RDONLY)) { + msg = "read-only option is not specified"; + goto parse_error; + } + + err = kstrtoull(arg->from, 0, &val); + if (err) { + if (err == -ERANGE) + msg = "too large checkpoint number"; + else + msg = "malformed argument"; + goto parse_error; + } else if (val == 0) { + msg = "invalid checkpoint number 0"; + goto parse_error; + } + sd->cno = val; + return 0; + +parse_error: + nilfs_msg(NULL, KERN_ERR, "invalid option \"%s\": %s", option, msg); + return 1; +} + /** * nilfs_identify - pre-read mount options needed to identify mount instance * @data: mount options @@ -1221,24 +1253,9 @@ static int nilfs_identify(char *data, struct nilfs_super_data *sd) p = strsep(&options, ","); if (p != NULL && *p) { token = match_token(p, tokens, args); - if (token == Opt_snapshot) { - if (!(sd->flags & MS_RDONLY)) { - ret++; - } else { - sd->cno = simple_strtoull(args[0].from, - NULL, 0); - /* - * No need to see the end pointer; - * match_token() has done syntax - * checking. - */ - if (sd->cno == 0) - ret++; - } - } - if (ret) - nilfs_msg(NULL, KERN_ERR, - "invalid mount option: %s", p); + if (token == Opt_snapshot) + ret = nilfs_parse_snapshot_option(p, &args[0], + sd); } if (!options) break; -- cgit v0.10.2 From ad980c9ab77cc4e0edc5dd3361fd69daabeb99f9 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Tue, 2 Aug 2016 14:05:25 -0700 Subject: nilfs2: fix misuse of a semaphore in sysfs code Variables ns_seg_seq, ns_segnum, ns_nextnum, ns_pseg_offset, ns_cno, ns_ctime, ns_nongc_ctime, and ns_ndirtyblks, are protected by ns_segctor_sem, but ns_sem is wrongly used by the nilfs sysfs code when reading these variables. This fixes the misuse and clarifies which semaphore protects them in the comment of the_nilfs struct. Link: http://lkml.kernel.org/r/1465825507-3407-2-git-send-email-konishi.ryusuke@lab.ntt.co.jp Signed-off-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/nilfs2/sysfs.c b/fs/nilfs2/sysfs.c index 8e57bb9..490303e 100644 --- a/fs/nilfs2/sysfs.c +++ b/fs/nilfs2/sysfs.c @@ -326,9 +326,9 @@ nilfs_checkpoints_next_checkpoint_show(struct nilfs_checkpoints_attr *attr, { __u64 cno; - down_read(&nilfs->ns_sem); + down_read(&nilfs->ns_segctor_sem); cno = nilfs->ns_cno; - up_read(&nilfs->ns_sem); + up_read(&nilfs->ns_segctor_sem); return snprintf(buf, PAGE_SIZE, "%llu\n", cno); } @@ -511,9 +511,9 @@ nilfs_segctor_current_seg_sequence_show(struct nilfs_segctor_attr *attr, { u64 seg_seq; - down_read(&nilfs->ns_sem); + down_read(&nilfs->ns_segctor_sem); seg_seq = nilfs->ns_seg_seq; - up_read(&nilfs->ns_sem); + up_read(&nilfs->ns_segctor_sem); return snprintf(buf, PAGE_SIZE, "%llu\n", seg_seq); } @@ -525,9 +525,9 @@ nilfs_segctor_current_last_full_seg_show(struct nilfs_segctor_attr *attr, { __u64 segnum; - down_read(&nilfs->ns_sem); + down_read(&nilfs->ns_segctor_sem); segnum = nilfs->ns_segnum; - up_read(&nilfs->ns_sem); + up_read(&nilfs->ns_segctor_sem); return snprintf(buf, PAGE_SIZE, "%llu\n", segnum); } @@ -539,9 +539,9 @@ nilfs_segctor_next_full_seg_show(struct nilfs_segctor_attr *attr, { __u64 nextnum; - down_read(&nilfs->ns_sem); + down_read(&nilfs->ns_segctor_sem); nextnum = nilfs->ns_nextnum; - up_read(&nilfs->ns_sem); + up_read(&nilfs->ns_segctor_sem); return snprintf(buf, PAGE_SIZE, "%llu\n", nextnum); } @@ -553,9 +553,9 @@ nilfs_segctor_next_pseg_offset_show(struct nilfs_segctor_attr *attr, { unsigned long pseg_offset; - down_read(&nilfs->ns_sem); + down_read(&nilfs->ns_segctor_sem); pseg_offset = nilfs->ns_pseg_offset; - up_read(&nilfs->ns_sem); + up_read(&nilfs->ns_segctor_sem); return snprintf(buf, PAGE_SIZE, "%lu\n", pseg_offset); } @@ -567,9 +567,9 @@ nilfs_segctor_next_checkpoint_show(struct nilfs_segctor_attr *attr, { __u64 cno; - down_read(&nilfs->ns_sem); + down_read(&nilfs->ns_segctor_sem); cno = nilfs->ns_cno; - up_read(&nilfs->ns_sem); + up_read(&nilfs->ns_segctor_sem); return snprintf(buf, PAGE_SIZE, "%llu\n", cno); } @@ -581,9 +581,9 @@ nilfs_segctor_last_seg_write_time_show(struct nilfs_segctor_attr *attr, { time_t ctime; - down_read(&nilfs->ns_sem); + down_read(&nilfs->ns_segctor_sem); ctime = nilfs->ns_ctime; - up_read(&nilfs->ns_sem); + up_read(&nilfs->ns_segctor_sem); return NILFS_SHOW_TIME(ctime, buf); } @@ -595,9 +595,9 @@ nilfs_segctor_last_seg_write_time_secs_show(struct nilfs_segctor_attr *attr, { time_t ctime; - down_read(&nilfs->ns_sem); + down_read(&nilfs->ns_segctor_sem); ctime = nilfs->ns_ctime; - up_read(&nilfs->ns_sem); + up_read(&nilfs->ns_segctor_sem); return snprintf(buf, PAGE_SIZE, "%llu\n", (unsigned long long)ctime); } @@ -609,9 +609,9 @@ nilfs_segctor_last_nongc_write_time_show(struct nilfs_segctor_attr *attr, { time_t nongc_ctime; - down_read(&nilfs->ns_sem); + down_read(&nilfs->ns_segctor_sem); nongc_ctime = nilfs->ns_nongc_ctime; - up_read(&nilfs->ns_sem); + up_read(&nilfs->ns_segctor_sem); return NILFS_SHOW_TIME(nongc_ctime, buf); } @@ -623,9 +623,9 @@ nilfs_segctor_last_nongc_write_time_secs_show(struct nilfs_segctor_attr *attr, { time_t nongc_ctime; - down_read(&nilfs->ns_sem); + down_read(&nilfs->ns_segctor_sem); nongc_ctime = nilfs->ns_nongc_ctime; - up_read(&nilfs->ns_sem); + up_read(&nilfs->ns_segctor_sem); return snprintf(buf, PAGE_SIZE, "%llu\n", (unsigned long long)nongc_ctime); @@ -638,9 +638,9 @@ nilfs_segctor_dirty_data_blocks_count_show(struct nilfs_segctor_attr *attr, { u32 ndirtyblks; - down_read(&nilfs->ns_sem); + down_read(&nilfs->ns_segctor_sem); ndirtyblks = atomic_read(&nilfs->ns_ndirtyblks); - up_read(&nilfs->ns_sem); + up_read(&nilfs->ns_segctor_sem); return snprintf(buf, PAGE_SIZE, "%u\n", ndirtyblks); } diff --git a/fs/nilfs2/the_nilfs.h b/fs/nilfs2/the_nilfs.h index 79d1421..b305c6f 100644 --- a/fs/nilfs2/the_nilfs.h +++ b/fs/nilfs2/the_nilfs.h @@ -122,11 +122,8 @@ struct the_nilfs { unsigned int ns_sb_update_freq; /* - * Following fields are dedicated to a writable FS-instance. - * Except for the period seeking checkpoint, code outside the segment - * constructor must lock a segment semaphore while accessing these - * fields. - * The writable FS-instance is sole during a lifetime of the_nilfs. + * The following fields are updated by a writable FS-instance. + * These fields are protected by ns_segctor_sem outside load_nilfs(). */ u64 ns_seg_seq; __u64 ns_segnum; -- cgit v0.10.2 From 4ce5c3426cbe9193f82345fb103e17dc3335eb4f Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Tue, 2 Aug 2016 14:05:28 -0700 Subject: nilfs2: use BIT() macro Replace bit shifts by BIT macro for clarity. Link: http://lkml.kernel.org/r/1465825507-3407-3-git-send-email-konishi.ryusuke@lab.ntt.co.jp Signed-off-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/nilfs2/btnode.c b/fs/nilfs2/btnode.c index 4cca998..d5c23da 100644 --- a/fs/nilfs2/btnode.c +++ b/fs/nilfs2/btnode.c @@ -41,7 +41,7 @@ nilfs_btnode_create_block(struct address_space *btnc, __u64 blocknr) struct inode *inode = NILFS_BTNC_I(btnc); struct buffer_head *bh; - bh = nilfs_grab_buffer(inode, btnc, blocknr, 1 << BH_NILFS_Node); + bh = nilfs_grab_buffer(inode, btnc, blocknr, BIT(BH_NILFS_Node)); if (unlikely(!bh)) return NULL; @@ -70,7 +70,7 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr, struct page *page; int err; - bh = nilfs_grab_buffer(inode, btnc, blocknr, 1 << BH_NILFS_Node); + bh = nilfs_grab_buffer(inode, btnc, blocknr, BIT(BH_NILFS_Node)); if (unlikely(!bh)) return -ENOMEM; diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c index b286b35..af04f55 100644 --- a/fs/nilfs2/inode.c +++ b/fs/nilfs2/inode.c @@ -356,7 +356,7 @@ struct inode *nilfs_new_inode(struct inode *dir, umode_t mode) root = NILFS_I(dir)->i_root; ii = NILFS_I(inode); - ii->i_state = 1 << NILFS_I_NEW; + ii->i_state = BIT(NILFS_I_NEW); ii->i_root = root; err = nilfs_ifile_create_inode(root->ifile, &ino, &ii->i_bh); @@ -555,7 +555,7 @@ static int nilfs_iget_set(struct inode *inode, void *opaque) inode->i_ino = args->ino; if (args->for_gc) { - NILFS_I(inode)->i_state = 1 << NILFS_I_GCINODE; + NILFS_I(inode)->i_state = BIT(NILFS_I_GCINODE); NILFS_I(inode)->i_cno = args->cno; NILFS_I(inode)->i_root = NULL; } else { diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h index 46fbd4e..2ba8a14 100644 --- a/fs/nilfs2/nilfs.h +++ b/fs/nilfs2/nilfs.h @@ -119,20 +119,19 @@ enum { /* * Macros to check inode numbers */ -#define NILFS_MDT_INO_BITS \ - ((unsigned int)(1 << NILFS_DAT_INO | 1 << NILFS_CPFILE_INO | \ - 1 << NILFS_SUFILE_INO | 1 << NILFS_IFILE_INO | \ - 1 << NILFS_ATIME_INO | 1 << NILFS_SKETCH_INO)) +#define NILFS_MDT_INO_BITS \ + (BIT(NILFS_DAT_INO) | BIT(NILFS_CPFILE_INO) | \ + BIT(NILFS_SUFILE_INO) | BIT(NILFS_IFILE_INO) | \ + BIT(NILFS_ATIME_INO) | BIT(NILFS_SKETCH_INO)) -#define NILFS_SYS_INO_BITS \ - ((unsigned int)(1 << NILFS_ROOT_INO) | NILFS_MDT_INO_BITS) +#define NILFS_SYS_INO_BITS (BIT(NILFS_ROOT_INO) | NILFS_MDT_INO_BITS) #define NILFS_FIRST_INO(sb) (((struct the_nilfs *)sb->s_fs_info)->ns_first_ino) #define NILFS_MDT_INODE(sb, ino) \ - ((ino) < NILFS_FIRST_INO(sb) && (NILFS_MDT_INO_BITS & (1 << (ino)))) + ((ino) < NILFS_FIRST_INO(sb) && (NILFS_MDT_INO_BITS & BIT(ino))) #define NILFS_VALID_INODE(sb, ino) \ - ((ino) >= NILFS_FIRST_INO(sb) || (NILFS_SYS_INO_BITS & (1 << (ino)))) + ((ino) >= NILFS_FIRST_INO(sb) || (NILFS_SYS_INO_BITS & BIT(ino))) /** * struct nilfs_transaction_info: context information for synchronization diff --git a/fs/nilfs2/page.c b/fs/nilfs2/page.c index eaccf12..f11a3ad 100644 --- a/fs/nilfs2/page.c +++ b/fs/nilfs2/page.c @@ -30,9 +30,9 @@ #include "mdt.h" -#define NILFS_BUFFER_INHERENT_BITS \ - ((1UL << BH_Uptodate) | (1UL << BH_Mapped) | (1UL << BH_NILFS_Node) | \ - (1UL << BH_NILFS_Volatile) | (1UL << BH_NILFS_Checked)) +#define NILFS_BUFFER_INHERENT_BITS \ + (BIT(BH_Uptodate) | BIT(BH_Mapped) | BIT(BH_NILFS_Node) | \ + BIT(BH_NILFS_Volatile) | BIT(BH_NILFS_Checked)) static struct buffer_head * __nilfs_get_page_block(struct page *page, unsigned long block, pgoff_t index, @@ -85,9 +85,9 @@ void nilfs_forget_buffer(struct buffer_head *bh) { struct page *page = bh->b_page; const unsigned long clear_bits = - (1 << BH_Uptodate | 1 << BH_Dirty | 1 << BH_Mapped | - 1 << BH_Async_Write | 1 << BH_NILFS_Volatile | - 1 << BH_NILFS_Checked | 1 << BH_NILFS_Redirected); + (BIT(BH_Uptodate) | BIT(BH_Dirty) | BIT(BH_Mapped) | + BIT(BH_Async_Write) | BIT(BH_NILFS_Volatile) | + BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected)); lock_buffer(bh); set_mask_bits(&bh->b_state, clear_bits, 0); @@ -124,17 +124,17 @@ void nilfs_copy_buffer(struct buffer_head *dbh, struct buffer_head *sbh) dbh->b_bdev = sbh->b_bdev; bh = dbh; - bits = sbh->b_state & ((1UL << BH_Uptodate) | (1UL << BH_Mapped)); + bits = sbh->b_state & (BIT(BH_Uptodate) | BIT(BH_Mapped)); while ((bh = bh->b_this_page) != dbh) { lock_buffer(bh); bits &= bh->b_state; unlock_buffer(bh); } - if (bits & (1UL << BH_Uptodate)) + if (bits & BIT(BH_Uptodate)) SetPageUptodate(dpage); else ClearPageUptodate(dpage); - if (bits & (1UL << BH_Mapped)) + if (bits & BIT(BH_Mapped)) SetPageMappedToDisk(dpage); else ClearPageMappedToDisk(dpage); @@ -215,7 +215,7 @@ static void nilfs_copy_page(struct page *dst, struct page *src, int copy_dirty) create_empty_buffers(dst, sbh->b_size, 0); if (copy_dirty) - mask |= (1UL << BH_Dirty); + mask |= BIT(BH_Dirty); dbh = dbufs = page_buffers(dst); do { @@ -414,9 +414,9 @@ void nilfs_clear_dirty_page(struct page *page, bool silent) if (page_has_buffers(page)) { struct buffer_head *bh, *head; const unsigned long clear_bits = - (1 << BH_Uptodate | 1 << BH_Dirty | 1 << BH_Mapped | - 1 << BH_Async_Write | 1 << BH_NILFS_Volatile | - 1 << BH_NILFS_Checked | 1 << BH_NILFS_Redirected); + (BIT(BH_Uptodate) | BIT(BH_Dirty) | BIT(BH_Mapped) | + BIT(BH_Async_Write) | BIT(BH_NILFS_Volatile) | + BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected)); bh = head = page_buffers(page); do { diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index 5a97282..bedcae2 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -1858,11 +1858,11 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci) */ list_for_each_entry(bh, &segbuf->sb_payload_buffers, b_assoc_buffers) { - const unsigned long set_bits = (1 << BH_Uptodate); + const unsigned long set_bits = BIT(BH_Uptodate); const unsigned long clear_bits = - (1 << BH_Dirty | 1 << BH_Async_Write | - 1 << BH_Delay | 1 << BH_NILFS_Volatile | - 1 << BH_NILFS_Redirected); + (BIT(BH_Dirty) | BIT(BH_Async_Write) | + BIT(BH_Delay) | BIT(BH_NILFS_Volatile) | + BIT(BH_NILFS_Redirected)); set_mask_bits(&bh->b_state, clear_bits, set_bits); if (bh == segbuf->sb_super_root) { @@ -2132,10 +2132,10 @@ static void nilfs_segctor_start_timer(struct nilfs_sc_info *sci) static void nilfs_segctor_do_flush(struct nilfs_sc_info *sci, int bn) { spin_lock(&sci->sc_state_lock); - if (!(sci->sc_flush_request & (1 << bn))) { + if (!(sci->sc_flush_request & BIT(bn))) { unsigned long prev_req = sci->sc_flush_request; - sci->sc_flush_request |= (1 << bn); + sci->sc_flush_request |= BIT(bn); if (!prev_req) wake_up(&sci->sc_wait_daemon); } @@ -2319,7 +2319,7 @@ int nilfs_construct_dsync_segment(struct super_block *sb, struct inode *inode, } #define FLUSH_FILE_BIT (0x1) /* data file only */ -#define FLUSH_DAT_BIT (1 << NILFS_DAT_INO) /* DAT only */ +#define FLUSH_DAT_BIT BIT(NILFS_DAT_INO) /* DAT only */ /** * nilfs_segctor_accept - record accepted sequence count of log-write requests diff --git a/fs/nilfs2/sufile.c b/fs/nilfs2/sufile.c index 5b495c4..12d11de 100644 --- a/fs/nilfs2/sufile.c +++ b/fs/nilfs2/sufile.c @@ -446,7 +446,7 @@ void nilfs_sufile_do_scrap(struct inode *sufile, __u64 segnum, kaddr = kmap_atomic(su_bh->b_page); su = nilfs_sufile_block_get_segment_usage(sufile, segnum, su_bh, kaddr); - if (su->su_flags == cpu_to_le32(1UL << NILFS_SEGMENT_USAGE_DIRTY) && + if (su->su_flags == cpu_to_le32(BIT(NILFS_SEGMENT_USAGE_DIRTY)) && su->su_nblocks == cpu_to_le32(0)) { kunmap_atomic(kaddr); return; @@ -457,7 +457,7 @@ void nilfs_sufile_do_scrap(struct inode *sufile, __u64 segnum, /* make the segment garbage */ su->su_lastmod = cpu_to_le64(0); su->su_nblocks = cpu_to_le32(0); - su->su_flags = cpu_to_le32(1UL << NILFS_SEGMENT_USAGE_DIRTY); + su->su_flags = cpu_to_le32(BIT(NILFS_SEGMENT_USAGE_DIRTY)); kunmap_atomic(kaddr); nilfs_sufile_mod_counter(header_bh, clean ? (u64)-1 : 0, dirty ? 0 : 1); @@ -695,7 +695,7 @@ static int nilfs_sufile_truncate_range(struct inode *sufile, su2 = su; for (j = 0; j < n; j++, su = (void *)su + susz) { if ((le32_to_cpu(su->su_flags) & - ~(1UL << NILFS_SEGMENT_USAGE_ERROR)) || + ~BIT(NILFS_SEGMENT_USAGE_ERROR)) || nilfs_segment_is_active(nilfs, segnum + j)) { ret = -EBUSY; kunmap_atomic(kaddr); @@ -862,10 +862,10 @@ ssize_t nilfs_sufile_get_suinfo(struct inode *sufile, __u64 segnum, void *buf, si->sui_lastmod = le64_to_cpu(su->su_lastmod); si->sui_nblocks = le32_to_cpu(su->su_nblocks); si->sui_flags = le32_to_cpu(su->su_flags) & - ~(1UL << NILFS_SEGMENT_USAGE_ACTIVE); + ~BIT(NILFS_SEGMENT_USAGE_ACTIVE); if (nilfs_segment_is_active(nilfs, segnum + j)) si->sui_flags |= - (1UL << NILFS_SEGMENT_USAGE_ACTIVE); + BIT(NILFS_SEGMENT_USAGE_ACTIVE); } kunmap_atomic(kaddr); brelse(su_bh); @@ -953,7 +953,7 @@ ssize_t nilfs_sufile_set_suinfo(struct inode *sufile, void *buf, * disk. */ sup->sup_sui.sui_flags &= - ~(1UL << NILFS_SEGMENT_USAGE_ACTIVE); + ~BIT(NILFS_SEGMENT_USAGE_ACTIVE); cleansi = nilfs_suinfo_clean(&sup->sup_sui); cleansu = nilfs_segment_usage_clean(su); -- cgit v0.10.2 From e63e88bc53bac7e4c3f592f8126c51a7569be673 Mon Sep 17 00:00:00 2001 From: Ryusuke Konishi Date: Tue, 2 Aug 2016 14:05:30 -0700 Subject: nilfs2: move ioctl interface and disk layout to uapi separately The header file "include/linux/nilfs2_fs.h" is composed of parts for ioctl and disk format, and both are intended to be shared with user space programs. This moves them to the uapi directory "include/uapi/linux" splitting the file to "nilfs2_api.h" and "nilfs2_ondisk.h". The following minor changes are accompanied by this migration: - nilfs_direct_node struct in nilfs2/direct.h is converged to nilfs2_ondisk.h because it's an on-disk structure. - inline functions nilfs_rec_len_from_disk() and nilfs_rec_len_to_disk() are moved to nilfs2/dir.c. Link: http://lkml.kernel.org/r/1465825507-3407-4-git-send-email-konishi.ryusuke@lab.ntt.co.jp Signed-off-by: Ryusuke Konishi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/Documentation/filesystems/nilfs2.txt b/Documentation/filesystems/nilfs2.txt index 5b21ef7..c0727dc 100644 --- a/Documentation/filesystems/nilfs2.txt +++ b/Documentation/filesystems/nilfs2.txt @@ -267,7 +267,8 @@ among NILFS2 files can be depicted as follows: `-- file (ino=yy) ( regular file, directory, or symlink ) -For detail on the format of each file, please see include/linux/nilfs2_fs.h. +For detail on the format of each file, please see nilfs2_ondisk.h +located at include/uapi/linux directory. There are no patents or other intellectual property that we protect with regard to the design of NILFS2. It is allowed to replicate the diff --git a/Documentation/ioctl/ioctl-number.txt b/Documentation/ioctl/ioctl-number.txt index 56af5e4..81c7f2b 100644 --- a/Documentation/ioctl/ioctl-number.txt +++ b/Documentation/ioctl/ioctl-number.txt @@ -248,7 +248,7 @@ Code Seq#(hex) Include File Comments 'm' 00 drivers/scsi/megaraid/megaraid_ioctl.h conflict! 'm' 00-1F net/irda/irmod.h conflict! 'n' 00-7F linux/ncp_fs.h and fs/ncpfs/ioctl.c -'n' 80-8F linux/nilfs2_fs.h NILFS2 +'n' 80-8F uapi/linux/nilfs2_api.h NILFS2 'n' E0-FF linux/matroxfb.h matroxfb 'o' 00-1F fs/ocfs2/ocfs2_fs.h OCFS2 'o' 00-03 mtd/ubi-user.h conflict! (OCFS2 and UBI overlaps) diff --git a/MAINTAINERS b/MAINTAINERS index bb51bbb..e9eacac 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -8258,8 +8258,9 @@ T: git git://github.com/konis/nilfs2.git S: Supported F: Documentation/filesystems/nilfs2.txt F: fs/nilfs2/ -F: include/linux/nilfs2_fs.h F: include/trace/events/nilfs2.h +F: include/uapi/linux/nilfs2_api.h +F: include/uapi/linux/nilfs2_ondisk.h NINJA SCSI-3 / NINJA SCSI-32Bi (16bit/CardBus) PCMCIA SCSI HOST ADAPTER DRIVER M: YOKOTA Hiroshi diff --git a/fs/nilfs2/bmap.h b/fs/nilfs2/bmap.h index b6a4c8f..2b6ffbe 100644 --- a/fs/nilfs2/bmap.h +++ b/fs/nilfs2/bmap.h @@ -22,7 +22,7 @@ #include #include #include -#include +#include /* nilfs_binfo, nilfs_inode, etc */ #include "alloc.h" #include "dat.h" diff --git a/fs/nilfs2/btree.h b/fs/nilfs2/btree.h index df1a25f..2184e47 100644 --- a/fs/nilfs2/btree.h +++ b/fs/nilfs2/btree.h @@ -22,7 +22,7 @@ #include #include #include -#include +#include /* nilfs_btree_node */ #include "btnode.h" #include "bmap.h" diff --git a/fs/nilfs2/cpfile.c b/fs/nilfs2/cpfile.c index 19d9f4a..a15a160 100644 --- a/fs/nilfs2/cpfile.c +++ b/fs/nilfs2/cpfile.c @@ -21,7 +21,6 @@ #include #include #include -#include #include "mdt.h" #include "cpfile.h" diff --git a/fs/nilfs2/cpfile.h b/fs/nilfs2/cpfile.h index 0249744..6eca972 100644 --- a/fs/nilfs2/cpfile.h +++ b/fs/nilfs2/cpfile.h @@ -21,7 +21,8 @@ #include #include -#include +#include /* nilfs_cpstat */ +#include /* nilfs_inode, nilfs_checkpoint */ int nilfs_cpfile_get_checkpoint(struct inode *, __u64, int, diff --git a/fs/nilfs2/dat.h b/fs/nilfs2/dat.h index abbfdab..57dc6cf 100644 --- a/fs/nilfs2/dat.h +++ b/fs/nilfs2/dat.h @@ -22,6 +22,7 @@ #include #include #include +#include /* nilfs_inode, nilfs_checkpoint */ struct nilfs_palloc_req; diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c index 746956d..908ebbf 100644 --- a/fs/nilfs2/dir.c +++ b/fs/nilfs2/dir.c @@ -42,6 +42,28 @@ #include "nilfs.h" #include "page.h" +static inline unsigned int nilfs_rec_len_from_disk(__le16 dlen) +{ + unsigned int len = le16_to_cpu(dlen); + +#if (PAGE_SIZE >= 65536) + if (len == NILFS_MAX_REC_LEN) + return 1 << 16; +#endif + return len; +} + +static inline __le16 nilfs_rec_len_to_disk(unsigned int len) +{ +#if (PAGE_SIZE >= 65536) + if (len == (1 << 16)) + return cpu_to_le16(NILFS_MAX_REC_LEN); + + BUG_ON(len > (1 << 16)); +#endif + return cpu_to_le16(len); +} + /* * nilfs uses block-sized chunks. Arguably, sector-sized ones would be * more robust, but we have what we have diff --git a/fs/nilfs2/direct.h b/fs/nilfs2/direct.h index 3015a6e..cfe85e8 100644 --- a/fs/nilfs2/direct.h +++ b/fs/nilfs2/direct.h @@ -24,16 +24,6 @@ #include "bmap.h" -/** - * struct nilfs_direct_node - direct node - * @dn_flags: flags - * @dn_pad: padding - */ -struct nilfs_direct_node { - __u8 dn_flags; - __u8 pad[7]; -}; - #define NILFS_DIRECT_NBLOCKS (NILFS_BMAP_SIZE / sizeof(__le64) - 1) #define NILFS_DIRECT_KEY_MIN 0 #define NILFS_DIRECT_KEY_MAX (NILFS_DIRECT_NBLOCKS - 1) diff --git a/fs/nilfs2/ifile.h b/fs/nilfs2/ifile.h index 23ad2f0..188b94f 100644 --- a/fs/nilfs2/ifile.h +++ b/fs/nilfs2/ifile.h @@ -23,7 +23,6 @@ #include #include -#include #include "mdt.h" #include "alloc.h" diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c index 827283f..f1d7989 100644 --- a/fs/nilfs2/ioctl.c +++ b/fs/nilfs2/ioctl.c @@ -25,7 +25,6 @@ #include /* compat_ptr() */ #include /* mnt_want_write_file(), mnt_drop_write_file() */ #include -#include #include "nilfs.h" #include "segment.h" #include "bmap.h" diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h index 2ba8a14..33f8c8f 100644 --- a/fs/nilfs2/nilfs.h +++ b/fs/nilfs2/nilfs.h @@ -23,7 +23,8 @@ #include #include #include -#include +#include +#include #include "the_nilfs.h" #include "bmap.h" diff --git a/fs/nilfs2/segment.h b/fs/nilfs2/segment.h index 6565c10..1060949 100644 --- a/fs/nilfs2/segment.h +++ b/fs/nilfs2/segment.h @@ -23,7 +23,6 @@ #include #include #include -#include #include "nilfs.h" struct nilfs_root; diff --git a/fs/nilfs2/sufile.c b/fs/nilfs2/sufile.c index 12d11de..1541a1e 100644 --- a/fs/nilfs2/sufile.c +++ b/fs/nilfs2/sufile.c @@ -22,7 +22,6 @@ #include #include #include -#include #include "mdt.h" #include "sufile.h" diff --git a/fs/nilfs2/sufile.h b/fs/nilfs2/sufile.h index 46e8987..158a919 100644 --- a/fs/nilfs2/sufile.h +++ b/fs/nilfs2/sufile.h @@ -21,7 +21,6 @@ #include #include -#include #include "mdt.h" diff --git a/include/linux/nilfs2_fs.h b/include/linux/nilfs2_fs.h deleted file mode 100644 index 5988dd5..0000000 --- a/include/linux/nilfs2_fs.h +++ /dev/null @@ -1,934 +0,0 @@ -/* - * nilfs2_fs.h - NILFS2 on-disk structures and common declarations. - * - * Copyright (C) 2005-2008 Nippon Telegraph and Telephone Corporation. - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU Lesser General Public License as published - * by the Free Software Foundation; either version 2.1 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU Lesser General Public License for more details. - * - * Written by Koji Sato and Ryusuke Konishi. - */ -/* - * linux/include/linux/ext2_fs.h - * - * Copyright (C) 1992, 1993, 1994, 1995 - * Remy Card (card@masi.ibp.fr) - * Laboratoire MASI - Institut Blaise Pascal - * Universite Pierre et Marie Curie (Paris VI) - * - * from - * - * linux/include/linux/minix_fs.h - * - * Copyright (C) 1991, 1992 Linus Torvalds - */ - -#ifndef _LINUX_NILFS_FS_H -#define _LINUX_NILFS_FS_H - -#include -#include -#include -#include - - -#define NILFS_INODE_BMAP_SIZE 7 -/** - * struct nilfs_inode - structure of an inode on disk - * @i_blocks: blocks count - * @i_size: size in bytes - * @i_ctime: creation time (seconds) - * @i_mtime: modification time (seconds) - * @i_ctime_nsec: creation time (nano seconds) - * @i_mtime_nsec: modification time (nano seconds) - * @i_uid: user id - * @i_gid: group id - * @i_mode: file mode - * @i_links_count: links count - * @i_flags: file flags - * @i_bmap: block mapping - * @i_xattr: extended attributes - * @i_generation: file generation (for NFS) - * @i_pad: padding - */ -struct nilfs_inode { - __le64 i_blocks; - __le64 i_size; - __le64 i_ctime; - __le64 i_mtime; - __le32 i_ctime_nsec; - __le32 i_mtime_nsec; - __le32 i_uid; - __le32 i_gid; - __le16 i_mode; - __le16 i_links_count; - __le32 i_flags; - __le64 i_bmap[NILFS_INODE_BMAP_SIZE]; -#define i_device_code i_bmap[0] - __le64 i_xattr; - __le32 i_generation; - __le32 i_pad; -}; - -#define NILFS_MIN_INODE_SIZE 128 - -/** - * struct nilfs_super_root - structure of super root - * @sr_sum: check sum - * @sr_bytes: byte count of the structure - * @sr_flags: flags (reserved) - * @sr_nongc_ctime: write time of the last segment not for cleaner operation - * @sr_dat: DAT file inode - * @sr_cpfile: checkpoint file inode - * @sr_sufile: segment usage file inode - */ -struct nilfs_super_root { - __le32 sr_sum; - __le16 sr_bytes; - __le16 sr_flags; - __le64 sr_nongc_ctime; - struct nilfs_inode sr_dat; - struct nilfs_inode sr_cpfile; - struct nilfs_inode sr_sufile; -}; - -#define NILFS_SR_MDT_OFFSET(inode_size, i) \ - ((unsigned long)&((struct nilfs_super_root *)0)->sr_dat + \ - (inode_size) * (i)) -#define NILFS_SR_DAT_OFFSET(inode_size) NILFS_SR_MDT_OFFSET(inode_size, 0) -#define NILFS_SR_CPFILE_OFFSET(inode_size) NILFS_SR_MDT_OFFSET(inode_size, 1) -#define NILFS_SR_SUFILE_OFFSET(inode_size) NILFS_SR_MDT_OFFSET(inode_size, 2) -#define NILFS_SR_BYTES(inode_size) NILFS_SR_MDT_OFFSET(inode_size, 3) - -/* - * Maximal mount counts - */ -#define NILFS_DFL_MAX_MNT_COUNT 50 /* 50 mounts */ - -/* - * File system states (sbp->s_state, nilfs->ns_mount_state) - */ -#define NILFS_VALID_FS 0x0001 /* Unmounted cleanly */ -#define NILFS_ERROR_FS 0x0002 /* Errors detected */ -#define NILFS_RESIZE_FS 0x0004 /* Resize required */ - -/* - * Mount flags (sbi->s_mount_opt) - */ -#define NILFS_MOUNT_ERROR_MODE 0x0070 /* Error mode mask */ -#define NILFS_MOUNT_ERRORS_CONT 0x0010 /* Continue on errors */ -#define NILFS_MOUNT_ERRORS_RO 0x0020 /* Remount fs ro on errors */ -#define NILFS_MOUNT_ERRORS_PANIC 0x0040 /* Panic on errors */ -#define NILFS_MOUNT_BARRIER 0x1000 /* Use block barriers */ -#define NILFS_MOUNT_STRICT_ORDER 0x2000 /* - * Apply strict in-order - * semantics also for data - */ -#define NILFS_MOUNT_NORECOVERY 0x4000 /* - * Disable write access during - * mount-time recovery - */ -#define NILFS_MOUNT_DISCARD 0x8000 /* Issue DISCARD requests */ - - -/** - * struct nilfs_super_block - structure of super block on disk - */ -struct nilfs_super_block { -/*00*/ __le32 s_rev_level; /* Revision level */ - __le16 s_minor_rev_level; /* minor revision level */ - __le16 s_magic; /* Magic signature */ - - __le16 s_bytes; /* - * Bytes count of CRC calculation - * for this structure. s_reserved - * is excluded. - */ - __le16 s_flags; /* flags */ - __le32 s_crc_seed; /* Seed value of CRC calculation */ -/*10*/ __le32 s_sum; /* Check sum of super block */ - - __le32 s_log_block_size; /* - * Block size represented as follows - * blocksize = - * 1 << (s_log_block_size + 10) - */ - __le64 s_nsegments; /* Number of segments in filesystem */ -/*20*/ __le64 s_dev_size; /* block device size in bytes */ - __le64 s_first_data_block; /* 1st seg disk block number */ -/*30*/ __le32 s_blocks_per_segment; /* number of blocks per full segment */ - __le32 s_r_segments_percentage; /* Reserved segments percentage */ - - __le64 s_last_cno; /* Last checkpoint number */ -/*40*/ __le64 s_last_pseg; /* disk block addr pseg written last */ - __le64 s_last_seq; /* seq. number of seg written last */ -/*50*/ __le64 s_free_blocks_count; /* Free blocks count */ - - __le64 s_ctime; /* - * Creation time (execution time of - * newfs) - */ -/*60*/ __le64 s_mtime; /* Mount time */ - __le64 s_wtime; /* Write time */ -/*70*/ __le16 s_mnt_count; /* Mount count */ - __le16 s_max_mnt_count; /* Maximal mount count */ - __le16 s_state; /* File system state */ - __le16 s_errors; /* Behaviour when detecting errors */ - __le64 s_lastcheck; /* time of last check */ - -/*80*/ __le32 s_checkinterval; /* max. time between checks */ - __le32 s_creator_os; /* OS */ - __le16 s_def_resuid; /* Default uid for reserved blocks */ - __le16 s_def_resgid; /* Default gid for reserved blocks */ - __le32 s_first_ino; /* First non-reserved inode */ - -/*90*/ __le16 s_inode_size; /* Size of an inode */ - __le16 s_dat_entry_size; /* Size of a dat entry */ - __le16 s_checkpoint_size; /* Size of a checkpoint */ - __le16 s_segment_usage_size; /* Size of a segment usage */ - -/*98*/ __u8 s_uuid[16]; /* 128-bit uuid for volume */ -/*A8*/ char s_volume_name[80]; /* volume name */ - -/*F8*/ __le32 s_c_interval; /* Commit interval of segment */ - __le32 s_c_block_max; /* - * Threshold of data amount for - * the segment construction - */ -/*100*/ __le64 s_feature_compat; /* Compatible feature set */ - __le64 s_feature_compat_ro; /* Read-only compatible feature set */ - __le64 s_feature_incompat; /* Incompatible feature set */ - __u32 s_reserved[186]; /* padding to the end of the block */ -}; - -/* - * Codes for operating systems - */ -#define NILFS_OS_LINUX 0 -/* Codes from 1 to 4 are reserved to keep compatibility with ext2 creator-OS */ - -/* - * Revision levels - */ -#define NILFS_CURRENT_REV 2 /* current major revision */ -#define NILFS_MINOR_REV 0 /* minor revision */ -#define NILFS_MIN_SUPP_REV 2 /* minimum supported revision */ - -/* - * Feature set definitions - * - * If there is a bit set in the incompatible feature set that the kernel - * doesn't know about, it should refuse to mount the filesystem. - */ -#define NILFS_FEATURE_COMPAT_RO_BLOCK_COUNT 0x00000001ULL - -#define NILFS_FEATURE_COMPAT_SUPP 0ULL -#define NILFS_FEATURE_COMPAT_RO_SUPP NILFS_FEATURE_COMPAT_RO_BLOCK_COUNT -#define NILFS_FEATURE_INCOMPAT_SUPP 0ULL - -/* - * Bytes count of super_block for CRC-calculation - */ -#define NILFS_SB_BYTES \ - ((long)&((struct nilfs_super_block *)0)->s_reserved) - -/* - * Special inode number - */ -#define NILFS_ROOT_INO 2 /* Root file inode */ -#define NILFS_DAT_INO 3 /* DAT file */ -#define NILFS_CPFILE_INO 4 /* checkpoint file */ -#define NILFS_SUFILE_INO 5 /* segment usage file */ -#define NILFS_IFILE_INO 6 /* ifile */ -#define NILFS_ATIME_INO 7 /* Atime file (reserved) */ -#define NILFS_XATTR_INO 8 /* Xattribute file (reserved) */ -#define NILFS_SKETCH_INO 10 /* Sketch file */ -#define NILFS_USER_INO 11 /* Fisrt user's file inode number */ - -#define NILFS_SB_OFFSET_BYTES 1024 /* byte offset of nilfs superblock */ - -#define NILFS_SEG_MIN_BLOCKS 16 /* - * Minimum number of blocks in - * a full segment - */ -#define NILFS_PSEG_MIN_BLOCKS 2 /* - * Minimum number of blocks in - * a partial segment - */ -#define NILFS_MIN_NRSVSEGS 8 /* - * Minimum number of reserved - * segments - */ - -/* - * We call DAT, cpfile, and sufile root metadata files. Inodes of - * these files are written in super root block instead of ifile, and - * garbage collector doesn't keep any past versions of these files. - */ -#define NILFS_ROOT_METADATA_FILE(ino) \ - ((ino) >= NILFS_DAT_INO && (ino) <= NILFS_SUFILE_INO) - -/* - * bytes offset of secondary super block - */ -#define NILFS_SB2_OFFSET_BYTES(devsize) ((((devsize) >> 12) - 1) << 12) - -/* - * Maximal count of links to a file - */ -#define NILFS_LINK_MAX 32000 - -/* - * Structure of a directory entry - * (Same as ext2) - */ - -#define NILFS_NAME_LEN 255 - -/* - * Block size limitations - */ -#define NILFS_MIN_BLOCK_SIZE 1024 -#define NILFS_MAX_BLOCK_SIZE 65536 - -/* - * The new version of the directory entry. Since V0 structures are - * stored in intel byte order, and the name_len field could never be - * bigger than 255 chars, it's safe to reclaim the extra byte for the - * file_type field. - */ -struct nilfs_dir_entry { - __le64 inode; /* Inode number */ - __le16 rec_len; /* Directory entry length */ - __u8 name_len; /* Name length */ - __u8 file_type; /* Dir entry type (file, dir, etc) */ - char name[NILFS_NAME_LEN]; /* File name */ - char pad; -}; - -/* - * NILFS directory file types. Only the low 3 bits are used. The - * other bits are reserved for now. - */ -enum { - NILFS_FT_UNKNOWN, - NILFS_FT_REG_FILE, - NILFS_FT_DIR, - NILFS_FT_CHRDEV, - NILFS_FT_BLKDEV, - NILFS_FT_FIFO, - NILFS_FT_SOCK, - NILFS_FT_SYMLINK, - NILFS_FT_MAX -}; - -/* - * NILFS_DIR_PAD defines the directory entries boundaries - * - * NOTE: It must be a multiple of 8 - */ -#define NILFS_DIR_PAD 8 -#define NILFS_DIR_ROUND (NILFS_DIR_PAD - 1) -#define NILFS_DIR_REC_LEN(name_len) (((name_len) + 12 + NILFS_DIR_ROUND) & \ - ~NILFS_DIR_ROUND) -#define NILFS_MAX_REC_LEN ((1<<16)-1) - -static inline unsigned int nilfs_rec_len_from_disk(__le16 dlen) -{ - unsigned int len = le16_to_cpu(dlen); - -#if !defined(__KERNEL__) || (PAGE_SIZE >= 65536) - if (len == NILFS_MAX_REC_LEN) - return 1 << 16; -#endif - return len; -} - -static inline __le16 nilfs_rec_len_to_disk(unsigned int len) -{ -#if !defined(__KERNEL__) || (PAGE_SIZE >= 65536) - if (len == (1 << 16)) - return cpu_to_le16(NILFS_MAX_REC_LEN); - else if (len > (1 << 16)) - BUG(); -#endif - return cpu_to_le16(len); -} - -/** - * struct nilfs_finfo - file information - * @fi_ino: inode number - * @fi_cno: checkpoint number - * @fi_nblocks: number of blocks (including intermediate blocks) - * @fi_ndatablk: number of file data blocks - */ -struct nilfs_finfo { - __le64 fi_ino; - __le64 fi_cno; - __le32 fi_nblocks; - __le32 fi_ndatablk; - /* array of virtual block numbers */ -}; - -/** - * struct nilfs_binfo_v - information for the block to which a virtual block number is assigned - * @bi_vblocknr: virtual block number - * @bi_blkoff: block offset - */ -struct nilfs_binfo_v { - __le64 bi_vblocknr; - __le64 bi_blkoff; -}; - -/** - * struct nilfs_binfo_dat - information for the block which belongs to the DAT file - * @bi_blkoff: block offset - * @bi_level: level - * @bi_pad: padding - */ -struct nilfs_binfo_dat { - __le64 bi_blkoff; - __u8 bi_level; - __u8 bi_pad[7]; -}; - -/** - * union nilfs_binfo: block information - * @bi_v: nilfs_binfo_v structure - * @bi_dat: nilfs_binfo_dat structure - */ -union nilfs_binfo { - struct nilfs_binfo_v bi_v; - struct nilfs_binfo_dat bi_dat; -}; - -/** - * struct nilfs_segment_summary - segment summary header - * @ss_datasum: checksum of data - * @ss_sumsum: checksum of segment summary - * @ss_magic: magic number - * @ss_bytes: size of this structure in bytes - * @ss_flags: flags - * @ss_seq: sequence number - * @ss_create: creation timestamp - * @ss_next: next segment - * @ss_nblocks: number of blocks - * @ss_nfinfo: number of finfo structures - * @ss_sumbytes: total size of segment summary in bytes - * @ss_pad: padding - * @ss_cno: checkpoint number - */ -struct nilfs_segment_summary { - __le32 ss_datasum; - __le32 ss_sumsum; - __le32 ss_magic; - __le16 ss_bytes; - __le16 ss_flags; - __le64 ss_seq; - __le64 ss_create; - __le64 ss_next; - __le32 ss_nblocks; - __le32 ss_nfinfo; - __le32 ss_sumbytes; - __le32 ss_pad; - __le64 ss_cno; - /* array of finfo structures */ -}; - -#define NILFS_SEGSUM_MAGIC 0x1eaffa11 /* segment summary magic number */ - -/* - * Segment summary flags - */ -#define NILFS_SS_LOGBGN 0x0001 /* begins a logical segment */ -#define NILFS_SS_LOGEND 0x0002 /* ends a logical segment */ -#define NILFS_SS_SR 0x0004 /* has super root */ -#define NILFS_SS_SYNDT 0x0008 /* includes data only updates */ -#define NILFS_SS_GC 0x0010 /* segment written for cleaner operation */ - -/** - * struct nilfs_btree_node - B-tree node - * @bn_flags: flags - * @bn_level: level - * @bn_nchildren: number of children - * @bn_pad: padding - */ -struct nilfs_btree_node { - __u8 bn_flags; - __u8 bn_level; - __le16 bn_nchildren; - __le32 bn_pad; -}; - -/* flags */ -#define NILFS_BTREE_NODE_ROOT 0x01 - -/* level */ -#define NILFS_BTREE_LEVEL_DATA 0 -#define NILFS_BTREE_LEVEL_NODE_MIN (NILFS_BTREE_LEVEL_DATA + 1) -#define NILFS_BTREE_LEVEL_MAX 14 /* Max level (exclusive) */ - -/** - * struct nilfs_palloc_group_desc - block group descriptor - * @pg_nfrees: number of free entries in block group - */ -struct nilfs_palloc_group_desc { - __le32 pg_nfrees; -}; - -/** - * struct nilfs_dat_entry - disk address translation entry - * @de_blocknr: block number - * @de_start: start checkpoint number - * @de_end: end checkpoint number - * @de_rsv: reserved for future use - */ -struct nilfs_dat_entry { - __le64 de_blocknr; - __le64 de_start; - __le64 de_end; - __le64 de_rsv; -}; - -#define NILFS_MIN_DAT_ENTRY_SIZE 32 - -/** - * struct nilfs_snapshot_list - snapshot list - * @ssl_next: next checkpoint number on snapshot list - * @ssl_prev: previous checkpoint number on snapshot list - */ -struct nilfs_snapshot_list { - __le64 ssl_next; - __le64 ssl_prev; -}; - -/** - * struct nilfs_checkpoint - checkpoint structure - * @cp_flags: flags - * @cp_checkpoints_count: checkpoints count in a block - * @cp_snapshot_list: snapshot list - * @cp_cno: checkpoint number - * @cp_create: creation timestamp - * @cp_nblk_inc: number of blocks incremented by this checkpoint - * @cp_inodes_count: inodes count - * @cp_blocks_count: blocks count - * @cp_ifile_inode: inode of ifile - */ -struct nilfs_checkpoint { - __le32 cp_flags; - __le32 cp_checkpoints_count; - struct nilfs_snapshot_list cp_snapshot_list; - __le64 cp_cno; - __le64 cp_create; - __le64 cp_nblk_inc; - __le64 cp_inodes_count; - __le64 cp_blocks_count; - - /* - * Do not change the byte offset of ifile inode. - * To keep the compatibility of the disk format, - * additional fields should be added behind cp_ifile_inode. - */ - struct nilfs_inode cp_ifile_inode; -}; - -#define NILFS_MIN_CHECKPOINT_SIZE (64 + NILFS_MIN_INODE_SIZE) - -/* checkpoint flags */ -enum { - NILFS_CHECKPOINT_SNAPSHOT, - NILFS_CHECKPOINT_INVALID, - NILFS_CHECKPOINT_SKETCH, - NILFS_CHECKPOINT_MINOR, -}; - -#define NILFS_CHECKPOINT_FNS(flag, name) \ -static inline void \ -nilfs_checkpoint_set_##name(struct nilfs_checkpoint *cp) \ -{ \ - cp->cp_flags = cpu_to_le32(le32_to_cpu(cp->cp_flags) | \ - (1UL << NILFS_CHECKPOINT_##flag)); \ -} \ -static inline void \ -nilfs_checkpoint_clear_##name(struct nilfs_checkpoint *cp) \ -{ \ - cp->cp_flags = cpu_to_le32(le32_to_cpu(cp->cp_flags) & \ - ~(1UL << NILFS_CHECKPOINT_##flag)); \ -} \ -static inline int \ -nilfs_checkpoint_##name(const struct nilfs_checkpoint *cp) \ -{ \ - return !!(le32_to_cpu(cp->cp_flags) & \ - (1UL << NILFS_CHECKPOINT_##flag)); \ -} - -NILFS_CHECKPOINT_FNS(SNAPSHOT, snapshot) -NILFS_CHECKPOINT_FNS(INVALID, invalid) -NILFS_CHECKPOINT_FNS(MINOR, minor) - -/** - * struct nilfs_cpinfo - checkpoint information - * @ci_flags: flags - * @ci_pad: padding - * @ci_cno: checkpoint number - * @ci_create: creation timestamp - * @ci_nblk_inc: number of blocks incremented by this checkpoint - * @ci_inodes_count: inodes count - * @ci_blocks_count: blocks count - * @ci_next: next checkpoint number in snapshot list - */ -struct nilfs_cpinfo { - __u32 ci_flags; - __u32 ci_pad; - __u64 ci_cno; - __u64 ci_create; - __u64 ci_nblk_inc; - __u64 ci_inodes_count; - __u64 ci_blocks_count; - __u64 ci_next; -}; - -#define NILFS_CPINFO_FNS(flag, name) \ -static inline int \ -nilfs_cpinfo_##name(const struct nilfs_cpinfo *cpinfo) \ -{ \ - return !!(cpinfo->ci_flags & (1UL << NILFS_CHECKPOINT_##flag)); \ -} - -NILFS_CPINFO_FNS(SNAPSHOT, snapshot) -NILFS_CPINFO_FNS(INVALID, invalid) -NILFS_CPINFO_FNS(MINOR, minor) - - -/** - * struct nilfs_cpfile_header - checkpoint file header - * @ch_ncheckpoints: number of checkpoints - * @ch_nsnapshots: number of snapshots - * @ch_snapshot_list: snapshot list - */ -struct nilfs_cpfile_header { - __le64 ch_ncheckpoints; - __le64 ch_nsnapshots; - struct nilfs_snapshot_list ch_snapshot_list; -}; - -#define NILFS_CPFILE_FIRST_CHECKPOINT_OFFSET \ - ((sizeof(struct nilfs_cpfile_header) + \ - sizeof(struct nilfs_checkpoint) - 1) / \ - sizeof(struct nilfs_checkpoint)) - -/** - * struct nilfs_segment_usage - segment usage - * @su_lastmod: last modified timestamp - * @su_nblocks: number of blocks in segment - * @su_flags: flags - */ -struct nilfs_segment_usage { - __le64 su_lastmod; - __le32 su_nblocks; - __le32 su_flags; -}; - -#define NILFS_MIN_SEGMENT_USAGE_SIZE 16 - -/* segment usage flag */ -enum { - NILFS_SEGMENT_USAGE_ACTIVE, - NILFS_SEGMENT_USAGE_DIRTY, - NILFS_SEGMENT_USAGE_ERROR, - - /* ... */ -}; - -#define NILFS_SEGMENT_USAGE_FNS(flag, name) \ -static inline void \ -nilfs_segment_usage_set_##name(struct nilfs_segment_usage *su) \ -{ \ - su->su_flags = cpu_to_le32(le32_to_cpu(su->su_flags) | \ - (1UL << NILFS_SEGMENT_USAGE_##flag));\ -} \ -static inline void \ -nilfs_segment_usage_clear_##name(struct nilfs_segment_usage *su) \ -{ \ - su->su_flags = \ - cpu_to_le32(le32_to_cpu(su->su_flags) & \ - ~(1UL << NILFS_SEGMENT_USAGE_##flag)); \ -} \ -static inline int \ -nilfs_segment_usage_##name(const struct nilfs_segment_usage *su) \ -{ \ - return !!(le32_to_cpu(su->su_flags) & \ - (1UL << NILFS_SEGMENT_USAGE_##flag)); \ -} - -NILFS_SEGMENT_USAGE_FNS(ACTIVE, active) -NILFS_SEGMENT_USAGE_FNS(DIRTY, dirty) -NILFS_SEGMENT_USAGE_FNS(ERROR, error) - -static inline void -nilfs_segment_usage_set_clean(struct nilfs_segment_usage *su) -{ - su->su_lastmod = cpu_to_le64(0); - su->su_nblocks = cpu_to_le32(0); - su->su_flags = cpu_to_le32(0); -} - -static inline int -nilfs_segment_usage_clean(const struct nilfs_segment_usage *su) -{ - return !le32_to_cpu(su->su_flags); -} - -/** - * struct nilfs_sufile_header - segment usage file header - * @sh_ncleansegs: number of clean segments - * @sh_ndirtysegs: number of dirty segments - * @sh_last_alloc: last allocated segment number - */ -struct nilfs_sufile_header { - __le64 sh_ncleansegs; - __le64 sh_ndirtysegs; - __le64 sh_last_alloc; - /* ... */ -}; - -#define NILFS_SUFILE_FIRST_SEGMENT_USAGE_OFFSET \ - ((sizeof(struct nilfs_sufile_header) + \ - sizeof(struct nilfs_segment_usage) - 1) / \ - sizeof(struct nilfs_segment_usage)) - -/** - * nilfs_suinfo - segment usage information - * @sui_lastmod: timestamp of last modification - * @sui_nblocks: number of written blocks in segment - * @sui_flags: segment usage flags - */ -struct nilfs_suinfo { - __u64 sui_lastmod; - __u32 sui_nblocks; - __u32 sui_flags; -}; - -#define NILFS_SUINFO_FNS(flag, name) \ -static inline int \ -nilfs_suinfo_##name(const struct nilfs_suinfo *si) \ -{ \ - return si->sui_flags & (1UL << NILFS_SEGMENT_USAGE_##flag); \ -} - -NILFS_SUINFO_FNS(ACTIVE, active) -NILFS_SUINFO_FNS(DIRTY, dirty) -NILFS_SUINFO_FNS(ERROR, error) - -static inline int nilfs_suinfo_clean(const struct nilfs_suinfo *si) -{ - return !si->sui_flags; -} - -/* ioctl */ -/** - * nilfs_suinfo_update - segment usage information update - * @sup_segnum: segment number - * @sup_flags: flags for which fields are active in sup_sui - * @sup_reserved: reserved necessary for alignment - * @sup_sui: segment usage information - */ -struct nilfs_suinfo_update { - __u64 sup_segnum; - __u32 sup_flags; - __u32 sup_reserved; - struct nilfs_suinfo sup_sui; -}; - -enum { - NILFS_SUINFO_UPDATE_LASTMOD, - NILFS_SUINFO_UPDATE_NBLOCKS, - NILFS_SUINFO_UPDATE_FLAGS, - __NR_NILFS_SUINFO_UPDATE_FIELDS, -}; - -#define NILFS_SUINFO_UPDATE_FNS(flag, name) \ -static inline void \ -nilfs_suinfo_update_set_##name(struct nilfs_suinfo_update *sup) \ -{ \ - sup->sup_flags |= 1UL << NILFS_SUINFO_UPDATE_##flag; \ -} \ -static inline void \ -nilfs_suinfo_update_clear_##name(struct nilfs_suinfo_update *sup) \ -{ \ - sup->sup_flags &= ~(1UL << NILFS_SUINFO_UPDATE_##flag); \ -} \ -static inline int \ -nilfs_suinfo_update_##name(const struct nilfs_suinfo_update *sup) \ -{ \ - return !!(sup->sup_flags & (1UL << NILFS_SUINFO_UPDATE_##flag));\ -} - -NILFS_SUINFO_UPDATE_FNS(LASTMOD, lastmod) -NILFS_SUINFO_UPDATE_FNS(NBLOCKS, nblocks) -NILFS_SUINFO_UPDATE_FNS(FLAGS, flags) - -enum { - NILFS_CHECKPOINT, - NILFS_SNAPSHOT, -}; - -/** - * struct nilfs_cpmode - change checkpoint mode structure - * @cm_cno: checkpoint number - * @cm_mode: mode of checkpoint - * @cm_pad: padding - */ -struct nilfs_cpmode { - __u64 cm_cno; - __u32 cm_mode; - __u32 cm_pad; -}; - -/** - * struct nilfs_argv - argument vector - * @v_base: pointer on data array from userspace - * @v_nmembs: number of members in data array - * @v_size: size of data array in bytes - * @v_flags: flags - * @v_index: start number of target data items - */ -struct nilfs_argv { - __u64 v_base; - __u32 v_nmembs; /* number of members */ - __u16 v_size; /* size of members */ - __u16 v_flags; - __u64 v_index; -}; - -/** - * struct nilfs_period - period of checkpoint numbers - * @p_start: start checkpoint number (inclusive) - * @p_end: end checkpoint number (exclusive) - */ -struct nilfs_period { - __u64 p_start; - __u64 p_end; -}; - -/** - * struct nilfs_cpstat - checkpoint statistics - * @cs_cno: checkpoint number - * @cs_ncps: number of checkpoints - * @cs_nsss: number of snapshots - */ -struct nilfs_cpstat { - __u64 cs_cno; - __u64 cs_ncps; - __u64 cs_nsss; -}; - -/** - * struct nilfs_sustat - segment usage statistics - * @ss_nsegs: number of segments - * @ss_ncleansegs: number of clean segments - * @ss_ndirtysegs: number of dirty segments - * @ss_ctime: creation time of the last segment - * @ss_nongc_ctime: creation time of the last segment not for GC - * @ss_prot_seq: least sequence number of segments which must not be reclaimed - */ -struct nilfs_sustat { - __u64 ss_nsegs; - __u64 ss_ncleansegs; - __u64 ss_ndirtysegs; - __u64 ss_ctime; - __u64 ss_nongc_ctime; - __u64 ss_prot_seq; -}; - -/** - * struct nilfs_vinfo - virtual block number information - * @vi_vblocknr: virtual block number - * @vi_start: start checkpoint number (inclusive) - * @vi_end: end checkpoint number (exclusive) - * @vi_blocknr: disk block number - */ -struct nilfs_vinfo { - __u64 vi_vblocknr; - __u64 vi_start; - __u64 vi_end; - __u64 vi_blocknr; -}; - -/** - * struct nilfs_vdesc - descriptor of virtual block number - * @vd_ino: inode number - * @vd_cno: checkpoint number - * @vd_vblocknr: virtual block number - * @vd_period: period of checkpoint numbers - * @vd_blocknr: disk block number - * @vd_offset: logical block offset inside a file - * @vd_flags: flags (data or node block) - * @vd_pad: padding - */ -struct nilfs_vdesc { - __u64 vd_ino; - __u64 vd_cno; - __u64 vd_vblocknr; - struct nilfs_period vd_period; - __u64 vd_blocknr; - __u64 vd_offset; - __u32 vd_flags; - __u32 vd_pad; -}; - -/** - * struct nilfs_bdesc - descriptor of disk block number - * @bd_ino: inode number - * @bd_oblocknr: disk block address (for skipping dead blocks) - * @bd_blocknr: disk block address - * @bd_offset: logical block offset inside a file - * @bd_level: level in the b-tree organization - * @bd_pad: padding - */ -struct nilfs_bdesc { - __u64 bd_ino; - __u64 bd_oblocknr; - __u64 bd_blocknr; - __u64 bd_offset; - __u32 bd_level; - __u32 bd_pad; -}; - -#define NILFS_IOCTL_IDENT 'n' - -#define NILFS_IOCTL_CHANGE_CPMODE \ - _IOW(NILFS_IOCTL_IDENT, 0x80, struct nilfs_cpmode) -#define NILFS_IOCTL_DELETE_CHECKPOINT \ - _IOW(NILFS_IOCTL_IDENT, 0x81, __u64) -#define NILFS_IOCTL_GET_CPINFO \ - _IOR(NILFS_IOCTL_IDENT, 0x82, struct nilfs_argv) -#define NILFS_IOCTL_GET_CPSTAT \ - _IOR(NILFS_IOCTL_IDENT, 0x83, struct nilfs_cpstat) -#define NILFS_IOCTL_GET_SUINFO \ - _IOR(NILFS_IOCTL_IDENT, 0x84, struct nilfs_argv) -#define NILFS_IOCTL_GET_SUSTAT \ - _IOR(NILFS_IOCTL_IDENT, 0x85, struct nilfs_sustat) -#define NILFS_IOCTL_GET_VINFO \ - _IOWR(NILFS_IOCTL_IDENT, 0x86, struct nilfs_argv) -#define NILFS_IOCTL_GET_BDESCS \ - _IOWR(NILFS_IOCTL_IDENT, 0x87, struct nilfs_argv) -#define NILFS_IOCTL_CLEAN_SEGMENTS \ - _IOW(NILFS_IOCTL_IDENT, 0x88, struct nilfs_argv[5]) -#define NILFS_IOCTL_SYNC \ - _IOR(NILFS_IOCTL_IDENT, 0x8A, __u64) -#define NILFS_IOCTL_RESIZE \ - _IOW(NILFS_IOCTL_IDENT, 0x8B, __u64) -#define NILFS_IOCTL_SET_ALLOC_RANGE \ - _IOW(NILFS_IOCTL_IDENT, 0x8C, __u64[2]) -#define NILFS_IOCTL_SET_SUINFO \ - _IOW(NILFS_IOCTL_IDENT, 0x8D, struct nilfs_argv) - -#endif /* _LINUX_NILFS_FS_H */ diff --git a/include/uapi/linux/nilfs2_api.h b/include/uapi/linux/nilfs2_api.h new file mode 100644 index 0000000..ef4c1de --- /dev/null +++ b/include/uapi/linux/nilfs2_api.h @@ -0,0 +1,292 @@ +/* + * nilfs2_api.h - NILFS2 user space API + * + * Copyright (C) 2005-2008 Nippon Telegraph and Telephone Corporation. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU Lesser General Public License as published + * by the Free Software Foundation; either version 2.1 of the License, or + * (at your option) any later version. + */ + +#ifndef _LINUX_NILFS2_API_H +#define _LINUX_NILFS2_API_H + +#include +#include + +/** + * struct nilfs_cpinfo - checkpoint information + * @ci_flags: flags + * @ci_pad: padding + * @ci_cno: checkpoint number + * @ci_create: creation timestamp + * @ci_nblk_inc: number of blocks incremented by this checkpoint + * @ci_inodes_count: inodes count + * @ci_blocks_count: blocks count + * @ci_next: next checkpoint number in snapshot list + */ +struct nilfs_cpinfo { + __u32 ci_flags; + __u32 ci_pad; + __u64 ci_cno; + __u64 ci_create; + __u64 ci_nblk_inc; + __u64 ci_inodes_count; + __u64 ci_blocks_count; + __u64 ci_next; +}; + +/* checkpoint flags */ +enum { + NILFS_CPINFO_SNAPSHOT, + NILFS_CPINFO_INVALID, + NILFS_CPINFO_SKETCH, + NILFS_CPINFO_MINOR, +}; + +#define NILFS_CPINFO_FNS(flag, name) \ +static inline int \ +nilfs_cpinfo_##name(const struct nilfs_cpinfo *cpinfo) \ +{ \ + return !!(cpinfo->ci_flags & (1UL << NILFS_CPINFO_##flag)); \ +} + +NILFS_CPINFO_FNS(SNAPSHOT, snapshot) +NILFS_CPINFO_FNS(INVALID, invalid) +NILFS_CPINFO_FNS(MINOR, minor) + +/** + * nilfs_suinfo - segment usage information + * @sui_lastmod: timestamp of last modification + * @sui_nblocks: number of written blocks in segment + * @sui_flags: segment usage flags + */ +struct nilfs_suinfo { + __u64 sui_lastmod; + __u32 sui_nblocks; + __u32 sui_flags; +}; + +/* segment usage flags */ +enum { + NILFS_SUINFO_ACTIVE, + NILFS_SUINFO_DIRTY, + NILFS_SUINFO_ERROR, +}; + +#define NILFS_SUINFO_FNS(flag, name) \ +static inline int \ +nilfs_suinfo_##name(const struct nilfs_suinfo *si) \ +{ \ + return si->sui_flags & (1UL << NILFS_SUINFO_##flag); \ +} + +NILFS_SUINFO_FNS(ACTIVE, active) +NILFS_SUINFO_FNS(DIRTY, dirty) +NILFS_SUINFO_FNS(ERROR, error) + +static inline int nilfs_suinfo_clean(const struct nilfs_suinfo *si) +{ + return !si->sui_flags; +} + +/** + * nilfs_suinfo_update - segment usage information update + * @sup_segnum: segment number + * @sup_flags: flags for which fields are active in sup_sui + * @sup_reserved: reserved necessary for alignment + * @sup_sui: segment usage information + */ +struct nilfs_suinfo_update { + __u64 sup_segnum; + __u32 sup_flags; + __u32 sup_reserved; + struct nilfs_suinfo sup_sui; +}; + +enum { + NILFS_SUINFO_UPDATE_LASTMOD, + NILFS_SUINFO_UPDATE_NBLOCKS, + NILFS_SUINFO_UPDATE_FLAGS, + __NR_NILFS_SUINFO_UPDATE_FIELDS, +}; + +#define NILFS_SUINFO_UPDATE_FNS(flag, name) \ +static inline void \ +nilfs_suinfo_update_set_##name(struct nilfs_suinfo_update *sup) \ +{ \ + sup->sup_flags |= 1UL << NILFS_SUINFO_UPDATE_##flag; \ +} \ +static inline void \ +nilfs_suinfo_update_clear_##name(struct nilfs_suinfo_update *sup) \ +{ \ + sup->sup_flags &= ~(1UL << NILFS_SUINFO_UPDATE_##flag); \ +} \ +static inline int \ +nilfs_suinfo_update_##name(const struct nilfs_suinfo_update *sup) \ +{ \ + return !!(sup->sup_flags & (1UL << NILFS_SUINFO_UPDATE_##flag));\ +} + +NILFS_SUINFO_UPDATE_FNS(LASTMOD, lastmod) +NILFS_SUINFO_UPDATE_FNS(NBLOCKS, nblocks) +NILFS_SUINFO_UPDATE_FNS(FLAGS, flags) + +enum { + NILFS_CHECKPOINT, + NILFS_SNAPSHOT, +}; + +/** + * struct nilfs_cpmode - change checkpoint mode structure + * @cm_cno: checkpoint number + * @cm_mode: mode of checkpoint + * @cm_pad: padding + */ +struct nilfs_cpmode { + __u64 cm_cno; + __u32 cm_mode; + __u32 cm_pad; +}; + +/** + * struct nilfs_argv - argument vector + * @v_base: pointer on data array from userspace + * @v_nmembs: number of members in data array + * @v_size: size of data array in bytes + * @v_flags: flags + * @v_index: start number of target data items + */ +struct nilfs_argv { + __u64 v_base; + __u32 v_nmembs; /* number of members */ + __u16 v_size; /* size of members */ + __u16 v_flags; + __u64 v_index; +}; + +/** + * struct nilfs_period - period of checkpoint numbers + * @p_start: start checkpoint number (inclusive) + * @p_end: end checkpoint number (exclusive) + */ +struct nilfs_period { + __u64 p_start; + __u64 p_end; +}; + +/** + * struct nilfs_cpstat - checkpoint statistics + * @cs_cno: checkpoint number + * @cs_ncps: number of checkpoints + * @cs_nsss: number of snapshots + */ +struct nilfs_cpstat { + __u64 cs_cno; + __u64 cs_ncps; + __u64 cs_nsss; +}; + +/** + * struct nilfs_sustat - segment usage statistics + * @ss_nsegs: number of segments + * @ss_ncleansegs: number of clean segments + * @ss_ndirtysegs: number of dirty segments + * @ss_ctime: creation time of the last segment + * @ss_nongc_ctime: creation time of the last segment not for GC + * @ss_prot_seq: least sequence number of segments which must not be reclaimed + */ +struct nilfs_sustat { + __u64 ss_nsegs; + __u64 ss_ncleansegs; + __u64 ss_ndirtysegs; + __u64 ss_ctime; + __u64 ss_nongc_ctime; + __u64 ss_prot_seq; +}; + +/** + * struct nilfs_vinfo - virtual block number information + * @vi_vblocknr: virtual block number + * @vi_start: start checkpoint number (inclusive) + * @vi_end: end checkpoint number (exclusive) + * @vi_blocknr: disk block number + */ +struct nilfs_vinfo { + __u64 vi_vblocknr; + __u64 vi_start; + __u64 vi_end; + __u64 vi_blocknr; +}; + +/** + * struct nilfs_vdesc - descriptor of virtual block number + * @vd_ino: inode number + * @vd_cno: checkpoint number + * @vd_vblocknr: virtual block number + * @vd_period: period of checkpoint numbers + * @vd_blocknr: disk block number + * @vd_offset: logical block offset inside a file + * @vd_flags: flags (data or node block) + * @vd_pad: padding + */ +struct nilfs_vdesc { + __u64 vd_ino; + __u64 vd_cno; + __u64 vd_vblocknr; + struct nilfs_period vd_period; + __u64 vd_blocknr; + __u64 vd_offset; + __u32 vd_flags; + __u32 vd_pad; +}; + +/** + * struct nilfs_bdesc - descriptor of disk block number + * @bd_ino: inode number + * @bd_oblocknr: disk block address (for skipping dead blocks) + * @bd_blocknr: disk block address + * @bd_offset: logical block offset inside a file + * @bd_level: level in the b-tree organization + * @bd_pad: padding + */ +struct nilfs_bdesc { + __u64 bd_ino; + __u64 bd_oblocknr; + __u64 bd_blocknr; + __u64 bd_offset; + __u32 bd_level; + __u32 bd_pad; +}; + +#define NILFS_IOCTL_IDENT 'n' + +#define NILFS_IOCTL_CHANGE_CPMODE \ + _IOW(NILFS_IOCTL_IDENT, 0x80, struct nilfs_cpmode) +#define NILFS_IOCTL_DELETE_CHECKPOINT \ + _IOW(NILFS_IOCTL_IDENT, 0x81, __u64) +#define NILFS_IOCTL_GET_CPINFO \ + _IOR(NILFS_IOCTL_IDENT, 0x82, struct nilfs_argv) +#define NILFS_IOCTL_GET_CPSTAT \ + _IOR(NILFS_IOCTL_IDENT, 0x83, struct nilfs_cpstat) +#define NILFS_IOCTL_GET_SUINFO \ + _IOR(NILFS_IOCTL_IDENT, 0x84, struct nilfs_argv) +#define NILFS_IOCTL_GET_SUSTAT \ + _IOR(NILFS_IOCTL_IDENT, 0x85, struct nilfs_sustat) +#define NILFS_IOCTL_GET_VINFO \ + _IOWR(NILFS_IOCTL_IDENT, 0x86, struct nilfs_argv) +#define NILFS_IOCTL_GET_BDESCS \ + _IOWR(NILFS_IOCTL_IDENT, 0x87, struct nilfs_argv) +#define NILFS_IOCTL_CLEAN_SEGMENTS \ + _IOW(NILFS_IOCTL_IDENT, 0x88, struct nilfs_argv[5]) +#define NILFS_IOCTL_SYNC \ + _IOR(NILFS_IOCTL_IDENT, 0x8A, __u64) +#define NILFS_IOCTL_RESIZE \ + _IOW(NILFS_IOCTL_IDENT, 0x8B, __u64) +#define NILFS_IOCTL_SET_ALLOC_RANGE \ + _IOW(NILFS_IOCTL_IDENT, 0x8C, __u64[2]) +#define NILFS_IOCTL_SET_SUINFO \ + _IOW(NILFS_IOCTL_IDENT, 0x8D, struct nilfs_argv) + +#endif /* _LINUX_NILFS2_API_H */ diff --git a/include/uapi/linux/nilfs2_ondisk.h b/include/uapi/linux/nilfs2_ondisk.h new file mode 100644 index 0000000..2a8a3ad --- /dev/null +++ b/include/uapi/linux/nilfs2_ondisk.h @@ -0,0 +1,650 @@ +/* + * nilfs2_ondisk.h - NILFS2 on-disk structures + * + * Copyright (C) 2005-2008 Nippon Telegraph and Telephone Corporation. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU Lesser General Public License as published + * by the Free Software Foundation; either version 2.1 of the License, or + * (at your option) any later version. + */ +/* + * linux/include/linux/ext2_fs.h + * + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * + * from + * + * linux/include/linux/minix_fs.h + * + * Copyright (C) 1991, 1992 Linus Torvalds + */ + +#ifndef _LINUX_NILFS2_ONDISK_H +#define _LINUX_NILFS2_ONDISK_H + +#include +#include + + +#define NILFS_INODE_BMAP_SIZE 7 + +/** + * struct nilfs_inode - structure of an inode on disk + * @i_blocks: blocks count + * @i_size: size in bytes + * @i_ctime: creation time (seconds) + * @i_mtime: modification time (seconds) + * @i_ctime_nsec: creation time (nano seconds) + * @i_mtime_nsec: modification time (nano seconds) + * @i_uid: user id + * @i_gid: group id + * @i_mode: file mode + * @i_links_count: links count + * @i_flags: file flags + * @i_bmap: block mapping + * @i_xattr: extended attributes + * @i_generation: file generation (for NFS) + * @i_pad: padding + */ +struct nilfs_inode { + __le64 i_blocks; + __le64 i_size; + __le64 i_ctime; + __le64 i_mtime; + __le32 i_ctime_nsec; + __le32 i_mtime_nsec; + __le32 i_uid; + __le32 i_gid; + __le16 i_mode; + __le16 i_links_count; + __le32 i_flags; + __le64 i_bmap[NILFS_INODE_BMAP_SIZE]; +#define i_device_code i_bmap[0] + __le64 i_xattr; + __le32 i_generation; + __le32 i_pad; +}; + +#define NILFS_MIN_INODE_SIZE 128 + +/** + * struct nilfs_super_root - structure of super root + * @sr_sum: check sum + * @sr_bytes: byte count of the structure + * @sr_flags: flags (reserved) + * @sr_nongc_ctime: write time of the last segment not for cleaner operation + * @sr_dat: DAT file inode + * @sr_cpfile: checkpoint file inode + * @sr_sufile: segment usage file inode + */ +struct nilfs_super_root { + __le32 sr_sum; + __le16 sr_bytes; + __le16 sr_flags; + __le64 sr_nongc_ctime; + struct nilfs_inode sr_dat; + struct nilfs_inode sr_cpfile; + struct nilfs_inode sr_sufile; +}; + +#define NILFS_SR_MDT_OFFSET(inode_size, i) \ + ((unsigned long)&((struct nilfs_super_root *)0)->sr_dat + \ + (inode_size) * (i)) +#define NILFS_SR_DAT_OFFSET(inode_size) NILFS_SR_MDT_OFFSET(inode_size, 0) +#define NILFS_SR_CPFILE_OFFSET(inode_size) NILFS_SR_MDT_OFFSET(inode_size, 1) +#define NILFS_SR_SUFILE_OFFSET(inode_size) NILFS_SR_MDT_OFFSET(inode_size, 2) +#define NILFS_SR_BYTES(inode_size) NILFS_SR_MDT_OFFSET(inode_size, 3) + +/* + * Maximal mount counts + */ +#define NILFS_DFL_MAX_MNT_COUNT 50 /* 50 mounts */ + +/* + * File system states (sbp->s_state, nilfs->ns_mount_state) + */ +#define NILFS_VALID_FS 0x0001 /* Unmounted cleanly */ +#define NILFS_ERROR_FS 0x0002 /* Errors detected */ +#define NILFS_RESIZE_FS 0x0004 /* Resize required */ + +/* + * Mount flags (sbi->s_mount_opt) + */ +#define NILFS_MOUNT_ERROR_MODE 0x0070 /* Error mode mask */ +#define NILFS_MOUNT_ERRORS_CONT 0x0010 /* Continue on errors */ +#define NILFS_MOUNT_ERRORS_RO 0x0020 /* Remount fs ro on errors */ +#define NILFS_MOUNT_ERRORS_PANIC 0x0040 /* Panic on errors */ +#define NILFS_MOUNT_BARRIER 0x1000 /* Use block barriers */ +#define NILFS_MOUNT_STRICT_ORDER 0x2000 /* + * Apply strict in-order + * semantics also for data + */ +#define NILFS_MOUNT_NORECOVERY 0x4000 /* + * Disable write access during + * mount-time recovery + */ +#define NILFS_MOUNT_DISCARD 0x8000 /* Issue DISCARD requests */ + + +/** + * struct nilfs_super_block - structure of super block on disk + */ +struct nilfs_super_block { +/*00*/ __le32 s_rev_level; /* Revision level */ + __le16 s_minor_rev_level; /* minor revision level */ + __le16 s_magic; /* Magic signature */ + + __le16 s_bytes; /* + * Bytes count of CRC calculation + * for this structure. s_reserved + * is excluded. + */ + __le16 s_flags; /* flags */ + __le32 s_crc_seed; /* Seed value of CRC calculation */ +/*10*/ __le32 s_sum; /* Check sum of super block */ + + __le32 s_log_block_size; /* + * Block size represented as follows + * blocksize = + * 1 << (s_log_block_size + 10) + */ + __le64 s_nsegments; /* Number of segments in filesystem */ +/*20*/ __le64 s_dev_size; /* block device size in bytes */ + __le64 s_first_data_block; /* 1st seg disk block number */ +/*30*/ __le32 s_blocks_per_segment; /* number of blocks per full segment */ + __le32 s_r_segments_percentage; /* Reserved segments percentage */ + + __le64 s_last_cno; /* Last checkpoint number */ +/*40*/ __le64 s_last_pseg; /* disk block addr pseg written last */ + __le64 s_last_seq; /* seq. number of seg written last */ +/*50*/ __le64 s_free_blocks_count; /* Free blocks count */ + + __le64 s_ctime; /* + * Creation time (execution time of + * newfs) + */ +/*60*/ __le64 s_mtime; /* Mount time */ + __le64 s_wtime; /* Write time */ +/*70*/ __le16 s_mnt_count; /* Mount count */ + __le16 s_max_mnt_count; /* Maximal mount count */ + __le16 s_state; /* File system state */ + __le16 s_errors; /* Behaviour when detecting errors */ + __le64 s_lastcheck; /* time of last check */ + +/*80*/ __le32 s_checkinterval; /* max. time between checks */ + __le32 s_creator_os; /* OS */ + __le16 s_def_resuid; /* Default uid for reserved blocks */ + __le16 s_def_resgid; /* Default gid for reserved blocks */ + __le32 s_first_ino; /* First non-reserved inode */ + +/*90*/ __le16 s_inode_size; /* Size of an inode */ + __le16 s_dat_entry_size; /* Size of a dat entry */ + __le16 s_checkpoint_size; /* Size of a checkpoint */ + __le16 s_segment_usage_size; /* Size of a segment usage */ + +/*98*/ __u8 s_uuid[16]; /* 128-bit uuid for volume */ +/*A8*/ char s_volume_name[80]; /* volume name */ + +/*F8*/ __le32 s_c_interval; /* Commit interval of segment */ + __le32 s_c_block_max; /* + * Threshold of data amount for + * the segment construction + */ +/*100*/ __le64 s_feature_compat; /* Compatible feature set */ + __le64 s_feature_compat_ro; /* Read-only compatible feature set */ + __le64 s_feature_incompat; /* Incompatible feature set */ + __u32 s_reserved[186]; /* padding to the end of the block */ +}; + +/* + * Codes for operating systems + */ +#define NILFS_OS_LINUX 0 +/* Codes from 1 to 4 are reserved to keep compatibility with ext2 creator-OS */ + +/* + * Revision levels + */ +#define NILFS_CURRENT_REV 2 /* current major revision */ +#define NILFS_MINOR_REV 0 /* minor revision */ +#define NILFS_MIN_SUPP_REV 2 /* minimum supported revision */ + +/* + * Feature set definitions + * + * If there is a bit set in the incompatible feature set that the kernel + * doesn't know about, it should refuse to mount the filesystem. + */ +#define NILFS_FEATURE_COMPAT_RO_BLOCK_COUNT 0x00000001ULL + +#define NILFS_FEATURE_COMPAT_SUPP 0ULL +#define NILFS_FEATURE_COMPAT_RO_SUPP NILFS_FEATURE_COMPAT_RO_BLOCK_COUNT +#define NILFS_FEATURE_INCOMPAT_SUPP 0ULL + +/* + * Bytes count of super_block for CRC-calculation + */ +#define NILFS_SB_BYTES \ + ((long)&((struct nilfs_super_block *)0)->s_reserved) + +/* + * Special inode number + */ +#define NILFS_ROOT_INO 2 /* Root file inode */ +#define NILFS_DAT_INO 3 /* DAT file */ +#define NILFS_CPFILE_INO 4 /* checkpoint file */ +#define NILFS_SUFILE_INO 5 /* segment usage file */ +#define NILFS_IFILE_INO 6 /* ifile */ +#define NILFS_ATIME_INO 7 /* Atime file (reserved) */ +#define NILFS_XATTR_INO 8 /* Xattribute file (reserved) */ +#define NILFS_SKETCH_INO 10 /* Sketch file */ +#define NILFS_USER_INO 11 /* Fisrt user's file inode number */ + +#define NILFS_SB_OFFSET_BYTES 1024 /* byte offset of nilfs superblock */ + +#define NILFS_SEG_MIN_BLOCKS 16 /* + * Minimum number of blocks in + * a full segment + */ +#define NILFS_PSEG_MIN_BLOCKS 2 /* + * Minimum number of blocks in + * a partial segment + */ +#define NILFS_MIN_NRSVSEGS 8 /* + * Minimum number of reserved + * segments + */ + +/* + * We call DAT, cpfile, and sufile root metadata files. Inodes of + * these files are written in super root block instead of ifile, and + * garbage collector doesn't keep any past versions of these files. + */ +#define NILFS_ROOT_METADATA_FILE(ino) \ + ((ino) >= NILFS_DAT_INO && (ino) <= NILFS_SUFILE_INO) + +/* + * bytes offset of secondary super block + */ +#define NILFS_SB2_OFFSET_BYTES(devsize) ((((devsize) >> 12) - 1) << 12) + +/* + * Maximal count of links to a file + */ +#define NILFS_LINK_MAX 32000 + +/* + * Structure of a directory entry + * (Same as ext2) + */ + +#define NILFS_NAME_LEN 255 + +/* + * Block size limitations + */ +#define NILFS_MIN_BLOCK_SIZE 1024 +#define NILFS_MAX_BLOCK_SIZE 65536 + +/* + * The new version of the directory entry. Since V0 structures are + * stored in intel byte order, and the name_len field could never be + * bigger than 255 chars, it's safe to reclaim the extra byte for the + * file_type field. + */ +struct nilfs_dir_entry { + __le64 inode; /* Inode number */ + __le16 rec_len; /* Directory entry length */ + __u8 name_len; /* Name length */ + __u8 file_type; /* Dir entry type (file, dir, etc) */ + char name[NILFS_NAME_LEN]; /* File name */ + char pad; +}; + +/* + * NILFS directory file types. Only the low 3 bits are used. The + * other bits are reserved for now. + */ +enum { + NILFS_FT_UNKNOWN, + NILFS_FT_REG_FILE, + NILFS_FT_DIR, + NILFS_FT_CHRDEV, + NILFS_FT_BLKDEV, + NILFS_FT_FIFO, + NILFS_FT_SOCK, + NILFS_FT_SYMLINK, + NILFS_FT_MAX +}; + +/* + * NILFS_DIR_PAD defines the directory entries boundaries + * + * NOTE: It must be a multiple of 8 + */ +#define NILFS_DIR_PAD 8 +#define NILFS_DIR_ROUND (NILFS_DIR_PAD - 1) +#define NILFS_DIR_REC_LEN(name_len) (((name_len) + 12 + NILFS_DIR_ROUND) & \ + ~NILFS_DIR_ROUND) +#define NILFS_MAX_REC_LEN ((1 << 16) - 1) + +/** + * struct nilfs_finfo - file information + * @fi_ino: inode number + * @fi_cno: checkpoint number + * @fi_nblocks: number of blocks (including intermediate blocks) + * @fi_ndatablk: number of file data blocks + */ +struct nilfs_finfo { + __le64 fi_ino; + __le64 fi_cno; + __le32 fi_nblocks; + __le32 fi_ndatablk; +}; + +/** + * struct nilfs_binfo_v - information on a data block (except DAT) + * @bi_vblocknr: virtual block number + * @bi_blkoff: block offset + */ +struct nilfs_binfo_v { + __le64 bi_vblocknr; + __le64 bi_blkoff; +}; + +/** + * struct nilfs_binfo_dat - information on a DAT node block + * @bi_blkoff: block offset + * @bi_level: level + * @bi_pad: padding + */ +struct nilfs_binfo_dat { + __le64 bi_blkoff; + __u8 bi_level; + __u8 bi_pad[7]; +}; + +/** + * union nilfs_binfo: block information + * @bi_v: nilfs_binfo_v structure + * @bi_dat: nilfs_binfo_dat structure + */ +union nilfs_binfo { + struct nilfs_binfo_v bi_v; + struct nilfs_binfo_dat bi_dat; +}; + +/** + * struct nilfs_segment_summary - segment summary header + * @ss_datasum: checksum of data + * @ss_sumsum: checksum of segment summary + * @ss_magic: magic number + * @ss_bytes: size of this structure in bytes + * @ss_flags: flags + * @ss_seq: sequence number + * @ss_create: creation timestamp + * @ss_next: next segment + * @ss_nblocks: number of blocks + * @ss_nfinfo: number of finfo structures + * @ss_sumbytes: total size of segment summary in bytes + * @ss_pad: padding + * @ss_cno: checkpoint number + */ +struct nilfs_segment_summary { + __le32 ss_datasum; + __le32 ss_sumsum; + __le32 ss_magic; + __le16 ss_bytes; + __le16 ss_flags; + __le64 ss_seq; + __le64 ss_create; + __le64 ss_next; + __le32 ss_nblocks; + __le32 ss_nfinfo; + __le32 ss_sumbytes; + __le32 ss_pad; + __le64 ss_cno; + /* array of finfo structures */ +}; + +#define NILFS_SEGSUM_MAGIC 0x1eaffa11 /* segment summary magic number */ + +/* + * Segment summary flags + */ +#define NILFS_SS_LOGBGN 0x0001 /* begins a logical segment */ +#define NILFS_SS_LOGEND 0x0002 /* ends a logical segment */ +#define NILFS_SS_SR 0x0004 /* has super root */ +#define NILFS_SS_SYNDT 0x0008 /* includes data only updates */ +#define NILFS_SS_GC 0x0010 /* segment written for cleaner operation */ + +/** + * struct nilfs_btree_node - header of B-tree node block + * @bn_flags: flags + * @bn_level: level + * @bn_nchildren: number of children + * @bn_pad: padding + */ +struct nilfs_btree_node { + __u8 bn_flags; + __u8 bn_level; + __le16 bn_nchildren; + __le32 bn_pad; +}; + +/* flags */ +#define NILFS_BTREE_NODE_ROOT 0x01 + +/* level */ +#define NILFS_BTREE_LEVEL_DATA 0 +#define NILFS_BTREE_LEVEL_NODE_MIN (NILFS_BTREE_LEVEL_DATA + 1) +#define NILFS_BTREE_LEVEL_MAX 14 /* Max level (exclusive) */ + +/** + * struct nilfs_direct_node - header of built-in bmap array + * @dn_flags: flags + * @dn_pad: padding + */ +struct nilfs_direct_node { + __u8 dn_flags; + __u8 pad[7]; +}; + +/** + * struct nilfs_palloc_group_desc - block group descriptor + * @pg_nfrees: number of free entries in block group + */ +struct nilfs_palloc_group_desc { + __le32 pg_nfrees; +}; + +/** + * struct nilfs_dat_entry - disk address translation entry + * @de_blocknr: block number + * @de_start: start checkpoint number + * @de_end: end checkpoint number + * @de_rsv: reserved for future use + */ +struct nilfs_dat_entry { + __le64 de_blocknr; + __le64 de_start; + __le64 de_end; + __le64 de_rsv; +}; + +#define NILFS_MIN_DAT_ENTRY_SIZE 32 + +/** + * struct nilfs_snapshot_list - snapshot list + * @ssl_next: next checkpoint number on snapshot list + * @ssl_prev: previous checkpoint number on snapshot list + */ +struct nilfs_snapshot_list { + __le64 ssl_next; + __le64 ssl_prev; +}; + +/** + * struct nilfs_checkpoint - checkpoint structure + * @cp_flags: flags + * @cp_checkpoints_count: checkpoints count in a block + * @cp_snapshot_list: snapshot list + * @cp_cno: checkpoint number + * @cp_create: creation timestamp + * @cp_nblk_inc: number of blocks incremented by this checkpoint + * @cp_inodes_count: inodes count + * @cp_blocks_count: blocks count + * @cp_ifile_inode: inode of ifile + */ +struct nilfs_checkpoint { + __le32 cp_flags; + __le32 cp_checkpoints_count; + struct nilfs_snapshot_list cp_snapshot_list; + __le64 cp_cno; + __le64 cp_create; + __le64 cp_nblk_inc; + __le64 cp_inodes_count; + __le64 cp_blocks_count; + + /* + * Do not change the byte offset of ifile inode. + * To keep the compatibility of the disk format, + * additional fields should be added behind cp_ifile_inode. + */ + struct nilfs_inode cp_ifile_inode; +}; + +#define NILFS_MIN_CHECKPOINT_SIZE (64 + NILFS_MIN_INODE_SIZE) + +/* checkpoint flags */ +enum { + NILFS_CHECKPOINT_SNAPSHOT, + NILFS_CHECKPOINT_INVALID, + NILFS_CHECKPOINT_SKETCH, + NILFS_CHECKPOINT_MINOR, +}; + +#define NILFS_CHECKPOINT_FNS(flag, name) \ +static inline void \ +nilfs_checkpoint_set_##name(struct nilfs_checkpoint *cp) \ +{ \ + cp->cp_flags = cpu_to_le32(le32_to_cpu(cp->cp_flags) | \ + (1UL << NILFS_CHECKPOINT_##flag)); \ +} \ +static inline void \ +nilfs_checkpoint_clear_##name(struct nilfs_checkpoint *cp) \ +{ \ + cp->cp_flags = cpu_to_le32(le32_to_cpu(cp->cp_flags) & \ + ~(1UL << NILFS_CHECKPOINT_##flag)); \ +} \ +static inline int \ +nilfs_checkpoint_##name(const struct nilfs_checkpoint *cp) \ +{ \ + return !!(le32_to_cpu(cp->cp_flags) & \ + (1UL << NILFS_CHECKPOINT_##flag)); \ +} + +NILFS_CHECKPOINT_FNS(SNAPSHOT, snapshot) +NILFS_CHECKPOINT_FNS(INVALID, invalid) +NILFS_CHECKPOINT_FNS(MINOR, minor) + +/** + * struct nilfs_cpfile_header - checkpoint file header + * @ch_ncheckpoints: number of checkpoints + * @ch_nsnapshots: number of snapshots + * @ch_snapshot_list: snapshot list + */ +struct nilfs_cpfile_header { + __le64 ch_ncheckpoints; + __le64 ch_nsnapshots; + struct nilfs_snapshot_list ch_snapshot_list; +}; + +#define NILFS_CPFILE_FIRST_CHECKPOINT_OFFSET \ + ((sizeof(struct nilfs_cpfile_header) + \ + sizeof(struct nilfs_checkpoint) - 1) / \ + sizeof(struct nilfs_checkpoint)) + +/** + * struct nilfs_segment_usage - segment usage + * @su_lastmod: last modified timestamp + * @su_nblocks: number of blocks in segment + * @su_flags: flags + */ +struct nilfs_segment_usage { + __le64 su_lastmod; + __le32 su_nblocks; + __le32 su_flags; +}; + +#define NILFS_MIN_SEGMENT_USAGE_SIZE 16 + +/* segment usage flag */ +enum { + NILFS_SEGMENT_USAGE_ACTIVE, + NILFS_SEGMENT_USAGE_DIRTY, + NILFS_SEGMENT_USAGE_ERROR, +}; + +#define NILFS_SEGMENT_USAGE_FNS(flag, name) \ +static inline void \ +nilfs_segment_usage_set_##name(struct nilfs_segment_usage *su) \ +{ \ + su->su_flags = cpu_to_le32(le32_to_cpu(su->su_flags) | \ + (1UL << NILFS_SEGMENT_USAGE_##flag));\ +} \ +static inline void \ +nilfs_segment_usage_clear_##name(struct nilfs_segment_usage *su) \ +{ \ + su->su_flags = \ + cpu_to_le32(le32_to_cpu(su->su_flags) & \ + ~(1UL << NILFS_SEGMENT_USAGE_##flag)); \ +} \ +static inline int \ +nilfs_segment_usage_##name(const struct nilfs_segment_usage *su) \ +{ \ + return !!(le32_to_cpu(su->su_flags) & \ + (1UL << NILFS_SEGMENT_USAGE_##flag)); \ +} + +NILFS_SEGMENT_USAGE_FNS(ACTIVE, active) +NILFS_SEGMENT_USAGE_FNS(DIRTY, dirty) +NILFS_SEGMENT_USAGE_FNS(ERROR, error) + +static inline void +nilfs_segment_usage_set_clean(struct nilfs_segment_usage *su) +{ + su->su_lastmod = cpu_to_le64(0); + su->su_nblocks = cpu_to_le32(0); + su->su_flags = cpu_to_le32(0); +} + +static inline int +nilfs_segment_usage_clean(const struct nilfs_segment_usage *su) +{ + return !le32_to_cpu(su->su_flags); +} + +/** + * struct nilfs_sufile_header - segment usage file header + * @sh_ncleansegs: number of clean segments + * @sh_ndirtysegs: number of dirty segments + * @sh_last_alloc: last allocated segment number + */ +struct nilfs_sufile_header { + __le64 sh_ncleansegs; + __le64 sh_ndirtysegs; + __le64 sh_last_alloc; + /* ... */ +}; + +#define NILFS_SUFILE_FIRST_SEGMENT_USAGE_OFFSET \ + ((sizeof(struct nilfs_sufile_header) + \ + sizeof(struct nilfs_segment_usage) - 1) / \ + sizeof(struct nilfs_segment_usage)) + +#endif /* _LINUX_NILFS2_ONDISK_H */ -- cgit v0.10.2 From 0a11b9aae49adf1f952427ef1a1d9e793dd6ffb6 Mon Sep 17 00:00:00 2001 From: Jeff Mahoney Date: Tue, 2 Aug 2016 14:05:33 -0700 Subject: reiserfs: fix "new_insert_key may be used uninitialized ..." new_insert_key only makes any sense when it's associated with a new_insert_ptr, which is initialized to NULL and changed to a buffer_head when we also initialize new_insert_key. We can key off of that to avoid the uninitialized warning. Link: http://lkml.kernel.org/r/5eca5ffb-2155-8df2-b4a2-f162f105efed@suse.com Signed-off-by: Jeff Mahoney Cc: Arnd Bergmann Cc: Jan Kara Cc: Linus Torvalds Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/reiserfs/ibalance.c b/fs/reiserfs/ibalance.c index b751eea..5db6f45 100644 --- a/fs/reiserfs/ibalance.c +++ b/fs/reiserfs/ibalance.c @@ -1153,8 +1153,9 @@ int balance_internal(struct tree_balance *tb, insert_ptr); } - memcpy(new_insert_key_addr, &new_insert_key, KEY_SIZE); insert_ptr[0] = new_insert_ptr; + if (new_insert_ptr) + memcpy(new_insert_key_addr, &new_insert_key, KEY_SIZE); return order; } -- cgit v0.10.2 From 7e7814180b334dff97ef8f56c7c40c277ad4531c Mon Sep 17 00:00:00 2001 From: Andy Lutomirski Date: Tue, 2 Aug 2016 14:05:36 -0700 Subject: signal: consolidate {TS,TLF}_RESTORE_SIGMASK code In general, there's no need for the "restore sigmask" flag to live in ti->flags. alpha, ia64, microblaze, powerpc, sh, sparc (64-bit only), tile, and x86 use essentially identical alternative implementations, placing the flag in ti->status. Replace those optimized implementations with an equally good common implementation that stores it in a bitfield in struct task_struct and drop the custom implementations. Additional architectures can opt in by removing their TIF_RESTORE_SIGMASK defines. Link: http://lkml.kernel.org/r/8a14321d64a28e40adfddc90e18a96c086a6d6f9.1468522723.git.luto@kernel.org Signed-off-by: Andy Lutomirski Tested-by: Michael Ellerman [powerpc] Cc: Richard Henderson Cc: Ivan Kokshaysky Cc: Matt Turner Cc: Tony Luck Cc: Fenghua Yu Cc: Michal Simek Cc: Benjamin Herrenschmidt Cc: Paul Mackerras Cc: Yoshinori Sato Cc: Rich Felker Cc: "David S. Miller" Cc: Chris Metcalf Cc: Peter Zijlstra Cc: Borislav Petkov Cc: Brian Gerst Cc: Dmitry Safonov Cc: Oleg Nesterov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/alpha/include/asm/thread_info.h b/arch/alpha/include/asm/thread_info.h index 32e920a..e9e90bf 100644 --- a/arch/alpha/include/asm/thread_info.h +++ b/arch/alpha/include/asm/thread_info.h @@ -86,33 +86,6 @@ register struct thread_info *__current_thread_info __asm__("$8"); #define TS_UAC_NOPRINT 0x0001 /* ! Preserve the following three */ #define TS_UAC_NOFIX 0x0002 /* ! flags as they match */ #define TS_UAC_SIGBUS 0x0004 /* ! userspace part of 'osf_sysinfo' */ -#define TS_RESTORE_SIGMASK 0x0008 /* restore signal mask in do_signal() */ - -#ifndef __ASSEMBLY__ -#define HAVE_SET_RESTORE_SIGMASK 1 -static inline void set_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - ti->status |= TS_RESTORE_SIGMASK; - WARN_ON(!test_bit(TIF_SIGPENDING, (unsigned long *)&ti->flags)); -} -static inline void clear_restore_sigmask(void) -{ - current_thread_info()->status &= ~TS_RESTORE_SIGMASK; -} -static inline bool test_restore_sigmask(void) -{ - return current_thread_info()->status & TS_RESTORE_SIGMASK; -} -static inline bool test_and_clear_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - if (!(ti->status & TS_RESTORE_SIGMASK)) - return false; - ti->status &= ~TS_RESTORE_SIGMASK; - return true; -} -#endif #define SET_UNALIGN_CTL(task,value) ({ \ __u32 status = task_thread_info(task)->status & ~UAC_BITMASK; \ diff --git a/arch/ia64/include/asm/thread_info.h b/arch/ia64/include/asm/thread_info.h index d1212b8..29bd597 100644 --- a/arch/ia64/include/asm/thread_info.h +++ b/arch/ia64/include/asm/thread_info.h @@ -121,32 +121,4 @@ struct thread_info { /* like TIF_ALLWORK_BITS but sans TIF_SYSCALL_TRACE or TIF_SYSCALL_AUDIT */ #define TIF_WORK_MASK (TIF_ALLWORK_MASK&~(_TIF_SYSCALL_TRACE|_TIF_SYSCALL_AUDIT)) -#define TS_RESTORE_SIGMASK 2 /* restore signal mask in do_signal() */ - -#ifndef __ASSEMBLY__ -#define HAVE_SET_RESTORE_SIGMASK 1 -static inline void set_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - ti->status |= TS_RESTORE_SIGMASK; - WARN_ON(!test_bit(TIF_SIGPENDING, &ti->flags)); -} -static inline void clear_restore_sigmask(void) -{ - current_thread_info()->status &= ~TS_RESTORE_SIGMASK; -} -static inline bool test_restore_sigmask(void) -{ - return current_thread_info()->status & TS_RESTORE_SIGMASK; -} -static inline bool test_and_clear_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - if (!(ti->status & TS_RESTORE_SIGMASK)) - return false; - ti->status &= ~TS_RESTORE_SIGMASK; - return true; -} -#endif /* !__ASSEMBLY__ */ - #endif /* _ASM_IA64_THREAD_INFO_H */ diff --git a/arch/microblaze/include/asm/thread_info.h b/arch/microblaze/include/asm/thread_info.h index 383f387..e7e8954 100644 --- a/arch/microblaze/include/asm/thread_info.h +++ b/arch/microblaze/include/asm/thread_info.h @@ -148,33 +148,6 @@ static inline struct thread_info *current_thread_info(void) */ /* FPU was used by this task this quantum (SMP) */ #define TS_USEDFPU 0x0001 -#define TS_RESTORE_SIGMASK 0x0002 - -#ifndef __ASSEMBLY__ -#define HAVE_SET_RESTORE_SIGMASK 1 -static inline void set_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - ti->status |= TS_RESTORE_SIGMASK; - WARN_ON(!test_bit(TIF_SIGPENDING, (unsigned long *)&ti->flags)); -} -static inline void clear_restore_sigmask(void) -{ - current_thread_info()->status &= ~TS_RESTORE_SIGMASK; -} -static inline bool test_restore_sigmask(void) -{ - return current_thread_info()->status & TS_RESTORE_SIGMASK; -} -static inline bool test_and_clear_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - if (!(ti->status & TS_RESTORE_SIGMASK)) - return false; - ti->status &= ~TS_RESTORE_SIGMASK; - return true; -} -#endif #endif /* __KERNEL__ */ #endif /* _ASM_MICROBLAZE_THREAD_INFO_H */ diff --git a/arch/powerpc/include/asm/thread_info.h b/arch/powerpc/include/asm/thread_info.h index b21bb1f..87e4b2d 100644 --- a/arch/powerpc/include/asm/thread_info.h +++ b/arch/powerpc/include/asm/thread_info.h @@ -138,40 +138,15 @@ static inline struct thread_info *current_thread_info(void) /* Don't move TLF_NAPPING without adjusting the code in entry_32.S */ #define TLF_NAPPING 0 /* idle thread enabled NAP mode */ #define TLF_SLEEPING 1 /* suspend code enabled SLEEP mode */ -#define TLF_RESTORE_SIGMASK 2 /* Restore signal mask in do_signal */ #define TLF_LAZY_MMU 3 /* tlb_batch is active */ #define TLF_RUNLATCH 4 /* Is the runlatch enabled? */ #define _TLF_NAPPING (1 << TLF_NAPPING) #define _TLF_SLEEPING (1 << TLF_SLEEPING) -#define _TLF_RESTORE_SIGMASK (1 << TLF_RESTORE_SIGMASK) #define _TLF_LAZY_MMU (1 << TLF_LAZY_MMU) #define _TLF_RUNLATCH (1 << TLF_RUNLATCH) #ifndef __ASSEMBLY__ -#define HAVE_SET_RESTORE_SIGMASK 1 -static inline void set_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - ti->local_flags |= _TLF_RESTORE_SIGMASK; - WARN_ON(!test_bit(TIF_SIGPENDING, &ti->flags)); -} -static inline void clear_restore_sigmask(void) -{ - current_thread_info()->local_flags &= ~_TLF_RESTORE_SIGMASK; -} -static inline bool test_restore_sigmask(void) -{ - return current_thread_info()->local_flags & _TLF_RESTORE_SIGMASK; -} -static inline bool test_and_clear_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - if (!(ti->local_flags & _TLF_RESTORE_SIGMASK)) - return false; - ti->local_flags &= ~_TLF_RESTORE_SIGMASK; - return true; -} static inline bool test_thread_local_flags(unsigned int flags) { diff --git a/arch/sh/include/asm/thread_info.h b/arch/sh/include/asm/thread_info.h index 2afa321..6c65dcd 100644 --- a/arch/sh/include/asm/thread_info.h +++ b/arch/sh/include/asm/thread_info.h @@ -151,19 +151,10 @@ extern void init_thread_xstate(void); * ever touches our thread-synchronous status, so we don't * have to worry about atomic accesses. */ -#define TS_RESTORE_SIGMASK 0x0001 /* restore signal mask in do_signal() */ #define TS_USEDFPU 0x0002 /* FPU used by this task this quantum */ #ifndef __ASSEMBLY__ -#define HAVE_SET_RESTORE_SIGMASK 1 -static inline void set_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - ti->status |= TS_RESTORE_SIGMASK; - WARN_ON(!test_bit(TIF_SIGPENDING, (unsigned long *)&ti->flags)); -} - #define TI_FLAG_FAULT_CODE_SHIFT 24 /* @@ -182,23 +173,6 @@ static inline unsigned int get_thread_fault_code(void) return ti->flags >> TI_FLAG_FAULT_CODE_SHIFT; } -static inline void clear_restore_sigmask(void) -{ - current_thread_info()->status &= ~TS_RESTORE_SIGMASK; -} -static inline bool test_restore_sigmask(void) -{ - return current_thread_info()->status & TS_RESTORE_SIGMASK; -} -static inline bool test_and_clear_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - if (!(ti->status & TS_RESTORE_SIGMASK)) - return false; - ti->status &= ~TS_RESTORE_SIGMASK; - return true; -} - #endif /* !__ASSEMBLY__ */ #endif /* __KERNEL__ */ diff --git a/arch/sparc/include/asm/thread_info_64.h b/arch/sparc/include/asm/thread_info_64.h index bde5982..3d7b925 100644 --- a/arch/sparc/include/asm/thread_info_64.h +++ b/arch/sparc/include/asm/thread_info_64.h @@ -222,32 +222,8 @@ register struct thread_info *current_thread_info_reg asm("g6"); * * Note that there are only 8 bits available. */ -#define TS_RESTORE_SIGMASK 0x0001 /* restore signal mask in do_signal() */ #ifndef __ASSEMBLY__ -#define HAVE_SET_RESTORE_SIGMASK 1 -static inline void set_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - ti->status |= TS_RESTORE_SIGMASK; - WARN_ON(!test_bit(TIF_SIGPENDING, &ti->flags)); -} -static inline void clear_restore_sigmask(void) -{ - current_thread_info()->status &= ~TS_RESTORE_SIGMASK; -} -static inline bool test_restore_sigmask(void) -{ - return current_thread_info()->status & TS_RESTORE_SIGMASK; -} -static inline bool test_and_clear_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - if (!(ti->status & TS_RESTORE_SIGMASK)) - return false; - ti->status &= ~TS_RESTORE_SIGMASK; - return true; -} #define thread32_stack_is_64bit(__SP) (((__SP) & 0x1) != 0) #define test_thread_64bit_stack(__SP) \ diff --git a/arch/tile/include/asm/thread_info.h b/arch/tile/include/asm/thread_info.h index c1467ac..b7659b8 100644 --- a/arch/tile/include/asm/thread_info.h +++ b/arch/tile/include/asm/thread_info.h @@ -166,32 +166,5 @@ extern void _cpu_idle(void); #ifdef __tilegx__ #define TS_COMPAT 0x0001 /* 32-bit compatibility mode */ #endif -#define TS_RESTORE_SIGMASK 0x0008 /* restore signal mask in do_signal */ - -#ifndef __ASSEMBLY__ -#define HAVE_SET_RESTORE_SIGMASK 1 -static inline void set_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - ti->status |= TS_RESTORE_SIGMASK; - WARN_ON(!test_bit(TIF_SIGPENDING, &ti->flags)); -} -static inline void clear_restore_sigmask(void) -{ - current_thread_info()->status &= ~TS_RESTORE_SIGMASK; -} -static inline bool test_restore_sigmask(void) -{ - return current_thread_info()->status & TS_RESTORE_SIGMASK; -} -static inline bool test_and_clear_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - if (!(ti->status & TS_RESTORE_SIGMASK)) - return false; - ti->status &= ~TS_RESTORE_SIGMASK; - return true; -} -#endif /* !__ASSEMBLY__ */ #endif /* _ASM_TILE_THREAD_INFO_H */ diff --git a/arch/x86/include/asm/thread_info.h b/arch/x86/include/asm/thread_info.h index 89bff04..b45ffdd 100644 --- a/arch/x86/include/asm/thread_info.h +++ b/arch/x86/include/asm/thread_info.h @@ -219,32 +219,8 @@ static inline unsigned long current_stack_pointer(void) * have to worry about atomic accesses. */ #define TS_COMPAT 0x0002 /* 32bit syscall active (64BIT)*/ -#define TS_RESTORE_SIGMASK 0x0008 /* restore signal mask in do_signal() */ #ifndef __ASSEMBLY__ -#define HAVE_SET_RESTORE_SIGMASK 1 -static inline void set_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - ti->status |= TS_RESTORE_SIGMASK; - WARN_ON(!test_bit(TIF_SIGPENDING, (unsigned long *)&ti->flags)); -} -static inline void clear_restore_sigmask(void) -{ - current_thread_info()->status &= ~TS_RESTORE_SIGMASK; -} -static inline bool test_restore_sigmask(void) -{ - return current_thread_info()->status & TS_RESTORE_SIGMASK; -} -static inline bool test_and_clear_restore_sigmask(void) -{ - struct thread_info *ti = current_thread_info(); - if (!(ti->status & TS_RESTORE_SIGMASK)) - return false; - ti->status &= ~TS_RESTORE_SIGMASK; - return true; -} static inline bool in_ia32_syscall(void) { diff --git a/include/linux/sched.h b/include/linux/sched.h index 553af29..62c68e5 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1547,6 +1547,9 @@ struct task_struct { /* unserialized, strictly 'current' */ unsigned in_execve:1; /* bit to tell LSMs we're in execve */ unsigned in_iowait:1; +#if !defined(TIF_RESTORE_SIGMASK) + unsigned restore_sigmask:1; +#endif #ifdef CONFIG_MEMCG unsigned memcg_may_oom:1; #ifndef CONFIG_SLOB @@ -2680,6 +2683,66 @@ extern void sigqueue_free(struct sigqueue *); extern int send_sigqueue(struct sigqueue *, struct task_struct *, int group); extern int do_sigaction(int, struct k_sigaction *, struct k_sigaction *); +#ifdef TIF_RESTORE_SIGMASK +/* + * Legacy restore_sigmask accessors. These are inefficient on + * SMP architectures because they require atomic operations. + */ + +/** + * set_restore_sigmask() - make sure saved_sigmask processing gets done + * + * This sets TIF_RESTORE_SIGMASK and ensures that the arch signal code + * will run before returning to user mode, to process the flag. For + * all callers, TIF_SIGPENDING is already set or it's no harm to set + * it. TIF_RESTORE_SIGMASK need not be in the set of bits that the + * arch code will notice on return to user mode, in case those bits + * are scarce. We set TIF_SIGPENDING here to ensure that the arch + * signal code always gets run when TIF_RESTORE_SIGMASK is set. + */ +static inline void set_restore_sigmask(void) +{ + set_thread_flag(TIF_RESTORE_SIGMASK); + WARN_ON(!test_thread_flag(TIF_SIGPENDING)); +} +static inline void clear_restore_sigmask(void) +{ + clear_thread_flag(TIF_RESTORE_SIGMASK); +} +static inline bool test_restore_sigmask(void) +{ + return test_thread_flag(TIF_RESTORE_SIGMASK); +} +static inline bool test_and_clear_restore_sigmask(void) +{ + return test_and_clear_thread_flag(TIF_RESTORE_SIGMASK); +} + +#else /* TIF_RESTORE_SIGMASK */ + +/* Higher-quality implementation, used if TIF_RESTORE_SIGMASK doesn't exist. */ +static inline void set_restore_sigmask(void) +{ + current->restore_sigmask = true; + WARN_ON(!test_thread_flag(TIF_SIGPENDING)); +} +static inline void clear_restore_sigmask(void) +{ + current->restore_sigmask = false; +} +static inline bool test_restore_sigmask(void) +{ + return current->restore_sigmask; +} +static inline bool test_and_clear_restore_sigmask(void) +{ + if (!current->restore_sigmask) + return false; + current->restore_sigmask = false; + return true; +} +#endif + static inline void restore_saved_sigmask(void) { if (test_and_clear_restore_sigmask()) diff --git a/include/linux/thread_info.h b/include/linux/thread_info.h index b4c2a48..352b154 100644 --- a/include/linux/thread_info.h +++ b/include/linux/thread_info.h @@ -105,47 +105,6 @@ static inline int test_ti_thread_flag(struct thread_info *ti, int flag) #define tif_need_resched() test_thread_flag(TIF_NEED_RESCHED) -#if defined TIF_RESTORE_SIGMASK && !defined HAVE_SET_RESTORE_SIGMASK -/* - * An arch can define its own version of set_restore_sigmask() to get the - * job done however works, with or without TIF_RESTORE_SIGMASK. - */ -#define HAVE_SET_RESTORE_SIGMASK 1 - -/** - * set_restore_sigmask() - make sure saved_sigmask processing gets done - * - * This sets TIF_RESTORE_SIGMASK and ensures that the arch signal code - * will run before returning to user mode, to process the flag. For - * all callers, TIF_SIGPENDING is already set or it's no harm to set - * it. TIF_RESTORE_SIGMASK need not be in the set of bits that the - * arch code will notice on return to user mode, in case those bits - * are scarce. We set TIF_SIGPENDING here to ensure that the arch - * signal code always gets run when TIF_RESTORE_SIGMASK is set. - */ -static inline void set_restore_sigmask(void) -{ - set_thread_flag(TIF_RESTORE_SIGMASK); - WARN_ON(!test_thread_flag(TIF_SIGPENDING)); -} -static inline void clear_restore_sigmask(void) -{ - clear_thread_flag(TIF_RESTORE_SIGMASK); -} -static inline bool test_restore_sigmask(void) -{ - return test_thread_flag(TIF_RESTORE_SIGMASK); -} -static inline bool test_and_clear_restore_sigmask(void) -{ - return test_and_clear_thread_flag(TIF_RESTORE_SIGMASK); -} -#endif /* TIF_RESTORE_SIGMASK && !HAVE_SET_RESTORE_SIGMASK */ - -#ifndef HAVE_SET_RESTORE_SIGMASK -#error "no set_restore_sigmask() provided and default one won't work" -#endif - #endif /* __KERNEL__ */ #endif /* _LINUX_THREAD_INFO_H */ -- cgit v0.10.2 From 627393d44860386e948bb63a8e5b53f2cc44d070 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Tue, 2 Aug 2016 14:05:40 -0700 Subject: kernel/exit.c: quieten greatest stack depth printk Many targets enable CONFIG_DEBUG_STACK_USAGE, and while the information is useful, it isn't worthy of pr_warn(). Reduce it to pr_info(). Link: http://lkml.kernel.org/r/1466982072-29836-1-git-send-email-anton@ozlabs.org Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/exit.c b/kernel/exit.c index 84ae830..2f974ae 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -715,7 +715,7 @@ static void check_stack_usage(void) spin_lock(&low_water_lock); if (free < lowest_to_date) { - pr_warn("%s (%d) used greatest stack depth: %lu bytes left\n", + pr_info("%s (%d) used greatest stack depth: %lu bytes left\n", current->comm, task_pid_nr(current), free); lowest_to_date = free; } -- cgit v0.10.2 From b06fb415331a7beb841f3d20d0fe60f6f0787dba Mon Sep 17 00:00:00 2001 From: Geliang Tang Date: Tue, 2 Aug 2016 14:05:42 -0700 Subject: cpumask: fix code comment Fix code comment for cpumask_parse(). Link: http://lkml.kernel.org/r/71aae2c60ae5dae0cf554199ce6aea8f88c69347.1465380581.git.geliangtang@gmail.com Signed-off-by: Geliang Tang Acked-by: Rusty Russell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/cpumask.h b/include/linux/cpumask.h index e828cf6..da7fbf1 100644 --- a/include/linux/cpumask.h +++ b/include/linux/cpumask.h @@ -579,7 +579,7 @@ static inline int cpumask_parselist_user(const char __user *buf, int len, } /** - * cpumask_parse - extract a cpumask from from a string + * cpumask_parse - extract a cpumask from a string * @buf: the buffer to extract from * @dstp: the cpumask to set. * -- cgit v0.10.2 From 4caf9615247aceab56e91df6c0e11892ea55f4f0 Mon Sep 17 00:00:00 2001 From: Minfei Huang Date: Tue, 2 Aug 2016 14:05:45 -0700 Subject: kexec: return error number directly This is a cleanup patch to make kexec more clear to return error number directly. The variable result is useless, because there is no other function's return value assignes to it. So remove it. Link: http://lkml.kernel.org/r/1464179273-57668-1-git-send-email-mnghuan@gmail.com Signed-off-by: Minfei Huang Cc: Dave Young Cc: Baoquan He Cc: Borislav Petkov Cc: Xunlei Pang Cc: Atsushi Kumagai Cc: Vivek Goyal Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/kexec_core.c b/kernel/kexec_core.c index 56b3ed0..23311c8 100644 --- a/kernel/kexec_core.c +++ b/kernel/kexec_core.c @@ -147,7 +147,7 @@ static struct page *kimage_alloc_page(struct kimage *image, int sanity_check_segment_list(struct kimage *image) { - int result, i; + int i; unsigned long nr_segments = image->nr_segments; /* @@ -163,16 +163,15 @@ int sanity_check_segment_list(struct kimage *image) * simply because addresses are changed to page size * granularity. */ - result = -EADDRNOTAVAIL; for (i = 0; i < nr_segments; i++) { unsigned long mstart, mend; mstart = image->segment[i].mem; mend = mstart + image->segment[i].memsz; if ((mstart & ~PAGE_MASK) || (mend & ~PAGE_MASK)) - return result; + return -EADDRNOTAVAIL; if (mend >= KEXEC_DESTINATION_MEMORY_LIMIT) - return result; + return -EADDRNOTAVAIL; } /* Verify our destination addresses do not overlap. @@ -180,7 +179,6 @@ int sanity_check_segment_list(struct kimage *image) * through very weird things can happen with no * easy explanation as one segment stops on another. */ - result = -EINVAL; for (i = 0; i < nr_segments; i++) { unsigned long mstart, mend; unsigned long j; @@ -194,7 +192,7 @@ int sanity_check_segment_list(struct kimage *image) pend = pstart + image->segment[j].memsz; /* Do the segments overlap ? */ if ((mend > pstart) && (mstart < pend)) - return result; + return -EINVAL; } } @@ -203,10 +201,9 @@ int sanity_check_segment_list(struct kimage *image) * and it is easier to check up front than to be surprised * later on. */ - result = -EINVAL; for (i = 0; i < nr_segments; i++) { if (image->segment[i].bufsz > image->segment[i].memsz) - return result; + return -EINVAL; } /* @@ -220,7 +217,6 @@ int sanity_check_segment_list(struct kimage *image) */ if (image->type == KEXEC_TYPE_CRASH) { - result = -EADDRNOTAVAIL; for (i = 0; i < nr_segments; i++) { unsigned long mstart, mend; @@ -229,7 +225,7 @@ int sanity_check_segment_list(struct kimage *image) /* Ensure we are within the crash kernel limits */ if ((mstart < crashk_res.start) || (mend > crashk_res.end)) - return result; + return -EADDRNOTAVAIL; } } -- cgit v0.10.2 From f7f0b7dc720f81b53afffb6779437086cdc3f62d Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 2 Aug 2016 14:05:48 -0700 Subject: ARM: kdump: advertise boot aliased crash kernel resource Advertise a resource which describes where the crash kernel is located in the boot view of RAM. This allows kexec-tools to have this vital information. Link: http://lkml.kernel.org/r/E1b8knz-0004H4-Bd@rmk-PC.armlinux.org.uk Signed-off-by: Russell King Cc: Baoquan He Cc: Keerthy Cc: Pratyush Anand Cc: Vitaly Andrianov Cc: Eric Biederman Cc: Dave Young Cc: Vivek Goyal Cc: Simon Horman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c index da2f6c3..6c8c888 100644 --- a/arch/arm/kernel/setup.c +++ b/arch/arm/kernel/setup.c @@ -1000,9 +1000,25 @@ static void __init reserve_crashkernel(void) (unsigned long)(crash_base >> 20), (unsigned long)(total_mem >> 20)); + /* The crashk resource must always be located in normal mem */ crashk_res.start = crash_base; crashk_res.end = crash_base + crash_size - 1; insert_resource(&iomem_resource, &crashk_res); + + if (arm_has_idmap_alias()) { + /* + * If we have a special RAM alias for use at boot, we + * need to advertise to kexec tools where the alias is. + */ + static struct resource crashk_boot_res = { + .name = "Crash kernel (boot alias)", + .flags = IORESOURCE_BUSY | IORESOURCE_MEM, + }; + + crashk_boot_res.start = phys_to_idmap(crash_base); + crashk_boot_res.end = crashk_boot_res.start + crash_size - 1; + insert_resource(&iomem_resource, &crashk_boot_res); + } } #else static inline void reserve_crashkernel(void) {} -- cgit v0.10.2 From 966fab00b0e19e0db3cb11d81bda5d0940176d5e Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 2 Aug 2016 14:05:51 -0700 Subject: ARM: kexec: advertise location of bootable RAM Advertise the location of bootable RAM to kexec-tools. kexec needs to know where it can place the kernel in RAM, and so be executable when the system needs to jump into it. Advertise these areas in /proc/iomem with a "System RAM (boot alias)" tag. Link: http://lkml.kernel.org/r/E1b8ko4-0004HA-GF@rmk-PC.armlinux.org.uk Signed-off-by: Russell King Reviewed-by: Pratyush Anand Cc: Baoquan He Cc: Keerthy Cc: Vitaly Andrianov Cc: Eric Biederman Cc: Dave Young Cc: Vivek Goyal Cc: Simon Horman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c index 6c8c888..df7f2a7 100644 --- a/arch/arm/kernel/setup.c +++ b/arch/arm/kernel/setup.c @@ -848,10 +848,29 @@ static void __init request_standard_resources(const struct machine_desc *mdesc) kernel_data.end = virt_to_phys(_end - 1); for_each_memblock(memory, region) { + phys_addr_t start = __pfn_to_phys(memblock_region_memory_base_pfn(region)); + phys_addr_t end = __pfn_to_phys(memblock_region_memory_end_pfn(region)) - 1; + unsigned long boot_alias_start; + + /* + * Some systems have a special memory alias which is only + * used for booting. We need to advertise this region to + * kexec-tools so they know where bootable RAM is located. + */ + boot_alias_start = phys_to_idmap(start); + if (arm_has_idmap_alias() && boot_alias_start != IDMAP_INVALID_ADDR) { + res = memblock_virt_alloc(sizeof(*res), 0); + res->name = "System RAM (boot alias)"; + res->start = boot_alias_start; + res->end = phys_to_idmap(end); + res->flags = IORESOURCE_MEM | IORESOURCE_BUSY; + request_resource(&iomem_resource, res); + } + res = memblock_virt_alloc(sizeof(*res), 0); res->name = "System RAM"; - res->start = __pfn_to_phys(memblock_region_memory_base_pfn(region)); - res->end = __pfn_to_phys(memblock_region_memory_end_pfn(region)) - 1; + res->start = start; + res->end = end; res->flags = IORESOURCE_SYSTEM_RAM | IORESOURCE_BUSY; request_resource(&iomem_resource, res); -- cgit v0.10.2 From dc5cccacf4272da4aba20a1fc0804d59d985ab32 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 2 Aug 2016 14:05:54 -0700 Subject: kexec: don't invoke OOM-killer for control page allocation If we are unable to find a suitable page when allocating the control page, do not invoke the OOM-killer: killing processes probably isn't going to help. Link: http://lkml.kernel.org/r/E1b8ko9-0004HG-R5@rmk-PC.armlinux.org.uk Signed-off-by: Russell King Reviewed-by: Pratyush Anand Acked-by: Baoquan He Cc: Keerthy Cc: Vitaly Andrianov Cc: Eric Biederman Cc: Dave Young Cc: Vivek Goyal Cc: Simon Horman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/kexec.h b/include/linux/kexec.h index e8acb2b..ce2fe19 100644 --- a/include/linux/kexec.h +++ b/include/linux/kexec.h @@ -41,7 +41,7 @@ #endif #ifndef KEXEC_CONTROL_MEMORY_GFP -#define KEXEC_CONTROL_MEMORY_GFP GFP_KERNEL +#define KEXEC_CONTROL_MEMORY_GFP (GFP_KERNEL | __GFP_NORETRY) #endif #ifndef KEXEC_CONTROL_PAGE_SIZE -- cgit v0.10.2 From 465d377701dfe6a08a9f361a3fd926dea7f89c74 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 2 Aug 2016 14:05:57 -0700 Subject: kexec: ensure user memory sizes do not wrap Ensure that user memory sizes do not wrap around when validating the user input, which can lead to the following input validation working incorrectly. [akpm@linux-foundation.org: fix it for kexec-return-error-number-directly.patch] Link: http://lkml.kernel.org/r/E1b8koF-0004HM-5x@rmk-PC.armlinux.org.uk Signed-off-by: Russell King Reviewed-by: Pratyush Anand Acked-by: Baoquan He Cc: Keerthy Cc: Vitaly Andrianov Cc: Eric Biederman Cc: Dave Young Cc: Vivek Goyal Cc: Simon Horman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/kexec_core.c b/kernel/kexec_core.c index 23311c8..5a83b2a 100644 --- a/kernel/kexec_core.c +++ b/kernel/kexec_core.c @@ -168,6 +168,8 @@ int sanity_check_segment_list(struct kimage *image) mstart = image->segment[i].mem; mend = mstart + image->segment[i].memsz; + if (mstart > mend) + return -EADDRNOTAVAIL; if ((mstart & ~PAGE_MASK) || (mend & ~PAGE_MASK)) return -EADDRNOTAVAIL; if (mend >= KEXEC_DESTINATION_MEMORY_LIMIT) -- cgit v0.10.2 From dae28018f56645b61f5beb84d5831346d3c5e457 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 2 Aug 2016 14:06:00 -0700 Subject: kdump: arrange for paddr_vmcoreinfo_note() to return phys_addr_t On PAE systems (eg, ARM LPAE) the vmcore note may be located above 4GB physical on 32-bit architectures, so we need a wider type than "unsigned long" here. Arrange for paddr_vmcoreinfo_note() to return a phys_addr_t, thereby allowing it to be located above 4GB. This makes no difference for kexec-tools, as they already assume a 64-bit type when reading from this file. Link: http://lkml.kernel.org/r/E1b8koK-0004HS-K9@rmk-PC.armlinux.org.uk Signed-off-by: Russell King Reviewed-by: Pratyush Anand Acked-by: Baoquan He Cc: Keerthy Cc: Vitaly Andrianov Cc: Eric Biederman Cc: Dave Young Cc: Vivek Goyal Cc: Simon Horman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/ia64/kernel/machine_kexec.c b/arch/ia64/kernel/machine_kexec.c index b72cd7a..599507b 100644 --- a/arch/ia64/kernel/machine_kexec.c +++ b/arch/ia64/kernel/machine_kexec.c @@ -163,7 +163,7 @@ void arch_crash_save_vmcoreinfo(void) #endif } -unsigned long paddr_vmcoreinfo_note(void) +phys_addr_t paddr_vmcoreinfo_note(void) { return ia64_tpa((unsigned long)(char *)&vmcoreinfo_note); } diff --git a/include/linux/kexec.h b/include/linux/kexec.h index ce2fe19..555227f 100644 --- a/include/linux/kexec.h +++ b/include/linux/kexec.h @@ -233,7 +233,7 @@ void crash_save_vmcoreinfo(void); void arch_crash_save_vmcoreinfo(void); __printf(1, 2) void vmcoreinfo_append_str(const char *fmt, ...); -unsigned long paddr_vmcoreinfo_note(void); +phys_addr_t paddr_vmcoreinfo_note(void); #define VMCOREINFO_OSRELEASE(value) \ vmcoreinfo_append_str("OSRELEASE=%s\n", value) diff --git a/kernel/kexec_core.c b/kernel/kexec_core.c index 5a83b2a..dab03f1 100644 --- a/kernel/kexec_core.c +++ b/kernel/kexec_core.c @@ -1372,7 +1372,7 @@ void vmcoreinfo_append_str(const char *fmt, ...) void __weak arch_crash_save_vmcoreinfo(void) {} -unsigned long __weak paddr_vmcoreinfo_note(void) +phys_addr_t __weak paddr_vmcoreinfo_note(void) { return __pa((unsigned long)(char *)&vmcoreinfo_note); } diff --git a/kernel/ksysfs.c b/kernel/ksysfs.c index 152da4a..9f1920d 100644 --- a/kernel/ksysfs.c +++ b/kernel/ksysfs.c @@ -128,8 +128,8 @@ KERNEL_ATTR_RW(kexec_crash_size); static ssize_t vmcoreinfo_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf) { - return sprintf(buf, "%lx %x\n", - paddr_vmcoreinfo_note(), + phys_addr_t vmcore_base = paddr_vmcoreinfo_note(); + return sprintf(buf, "%pa %x\n", &vmcore_base, (unsigned int)sizeof(vmcoreinfo_note)); } KERNEL_ATTR_RO(vmcoreinfo); -- cgit v0.10.2 From 43546d8669d62d75fa69ca9a45d2f586665f56bd Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 2 Aug 2016 14:06:04 -0700 Subject: kexec: allow architectures to override boot mapping kexec physical addresses are the boot-time view of the system. For certain ARM systems (such as Keystone 2), the boot view of the system does not match the kernel's view of the system: the boot view uses a special alias in the lower 4GB of the physical address space. To cater for these kinds of setups, we need to translate between the boot view physical addresses and the normal kernel view physical addresses. This patch extracts the current transation points into linux/kexec.h, and allows an architecture to override the functions. Due to the translations required, we unfortunately end up with six translation functions, which are reduced down to four that the architecture can override. [akpm@linux-foundation.org: kexec.h needs asm/io.h for phys_to_virt()] Link: http://lkml.kernel.org/r/E1b8koP-0004HZ-Vf@rmk-PC.armlinux.org.uk Signed-off-by: Russell King Cc: Keerthy Cc: Pratyush Anand Cc: Vitaly Andrianov Cc: Eric Biederman Cc: Dave Young Cc: Baoquan He Cc: Vivek Goyal Cc: Simon Horman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/kexec.h b/include/linux/kexec.h index 555227f..23e14a4 100644 --- a/include/linux/kexec.h +++ b/include/linux/kexec.h @@ -14,6 +14,8 @@ #if !defined(__ASSEMBLY__) +#include + #include #ifdef CONFIG_KEXEC_CORE @@ -318,6 +320,44 @@ int __weak arch_kexec_apply_relocations(const Elf_Ehdr *ehdr, Elf_Shdr *sechdrs, void arch_kexec_protect_crashkres(void); void arch_kexec_unprotect_crashkres(void); +#ifndef page_to_boot_pfn +static inline unsigned long page_to_boot_pfn(struct page *page) +{ + return page_to_pfn(page); +} +#endif + +#ifndef boot_pfn_to_page +static inline struct page *boot_pfn_to_page(unsigned long boot_pfn) +{ + return pfn_to_page(boot_pfn); +} +#endif + +#ifndef phys_to_boot_phys +static inline unsigned long phys_to_boot_phys(phys_addr_t phys) +{ + return phys; +} +#endif + +#ifndef boot_phys_to_phys +static inline phys_addr_t boot_phys_to_phys(unsigned long boot_phys) +{ + return boot_phys; +} +#endif + +static inline unsigned long virt_to_boot_phys(void *addr) +{ + return phys_to_boot_phys(__pa((unsigned long)addr)); +} + +static inline void *boot_phys_to_virt(unsigned long entry) +{ + return phys_to_virt(boot_phys_to_phys(entry)); +} + #else /* !CONFIG_KEXEC_CORE */ struct pt_regs; struct task_struct; diff --git a/kernel/kexec.c b/kernel/kexec.c index 4384672..980936a 100644 --- a/kernel/kexec.c +++ b/kernel/kexec.c @@ -48,7 +48,8 @@ static int kimage_alloc_init(struct kimage **rimage, unsigned long entry, if (kexec_on_panic) { /* Verify we have a valid entry point */ - if ((entry < crashk_res.start) || (entry > crashk_res.end)) + if ((entry < phys_to_boot_phys(crashk_res.start)) || + (entry > phys_to_boot_phys(crashk_res.end))) return -EADDRNOTAVAIL; } diff --git a/kernel/kexec_core.c b/kernel/kexec_core.c index dab03f1..73d4c5f 100644 --- a/kernel/kexec_core.c +++ b/kernel/kexec_core.c @@ -225,8 +225,8 @@ int sanity_check_segment_list(struct kimage *image) mstart = image->segment[i].mem; mend = mstart + image->segment[i].memsz - 1; /* Ensure we are within the crash kernel limits */ - if ((mstart < crashk_res.start) || - (mend > crashk_res.end)) + if ((mstart < phys_to_boot_phys(crashk_res.start)) || + (mend > phys_to_boot_phys(crashk_res.end))) return -EADDRNOTAVAIL; } } @@ -350,7 +350,7 @@ static struct page *kimage_alloc_normal_control_pages(struct kimage *image, pages = kimage_alloc_pages(KEXEC_CONTROL_MEMORY_GFP, order); if (!pages) break; - pfn = page_to_pfn(pages); + pfn = page_to_boot_pfn(pages); epfn = pfn + count; addr = pfn << PAGE_SHIFT; eaddr = epfn << PAGE_SHIFT; @@ -476,7 +476,7 @@ static int kimage_add_entry(struct kimage *image, kimage_entry_t entry) return -ENOMEM; ind_page = page_address(page); - *image->entry = virt_to_phys(ind_page) | IND_INDIRECTION; + *image->entry = virt_to_boot_phys(ind_page) | IND_INDIRECTION; image->entry = ind_page; image->last_entry = ind_page + ((PAGE_SIZE/sizeof(kimage_entry_t)) - 1); @@ -531,13 +531,13 @@ void kimage_terminate(struct kimage *image) #define for_each_kimage_entry(image, ptr, entry) \ for (ptr = &image->head; (entry = *ptr) && !(entry & IND_DONE); \ ptr = (entry & IND_INDIRECTION) ? \ - phys_to_virt((entry & PAGE_MASK)) : ptr + 1) + boot_phys_to_virt((entry & PAGE_MASK)) : ptr + 1) static void kimage_free_entry(kimage_entry_t entry) { struct page *page; - page = pfn_to_page(entry >> PAGE_SHIFT); + page = boot_pfn_to_page(entry >> PAGE_SHIFT); kimage_free_pages(page); } @@ -631,7 +631,7 @@ static struct page *kimage_alloc_page(struct kimage *image, * have a match. */ list_for_each_entry(page, &image->dest_pages, lru) { - addr = page_to_pfn(page) << PAGE_SHIFT; + addr = page_to_boot_pfn(page) << PAGE_SHIFT; if (addr == destination) { list_del(&page->lru); return page; @@ -646,12 +646,12 @@ static struct page *kimage_alloc_page(struct kimage *image, if (!page) return NULL; /* If the page cannot be used file it away */ - if (page_to_pfn(page) > + if (page_to_boot_pfn(page) > (KEXEC_SOURCE_MEMORY_LIMIT >> PAGE_SHIFT)) { list_add(&page->lru, &image->unusable_pages); continue; } - addr = page_to_pfn(page) << PAGE_SHIFT; + addr = page_to_boot_pfn(page) << PAGE_SHIFT; /* If it is the destination page we want use it */ if (addr == destination) @@ -674,7 +674,7 @@ static struct page *kimage_alloc_page(struct kimage *image, struct page *old_page; old_addr = *old & PAGE_MASK; - old_page = pfn_to_page(old_addr >> PAGE_SHIFT); + old_page = boot_pfn_to_page(old_addr >> PAGE_SHIFT); copy_highpage(page, old_page); *old = addr | (*old & ~PAGE_MASK); @@ -730,7 +730,7 @@ static int kimage_load_normal_segment(struct kimage *image, result = -ENOMEM; goto out; } - result = kimage_add_page(image, page_to_pfn(page) + result = kimage_add_page(image, page_to_boot_pfn(page) << PAGE_SHIFT); if (result < 0) goto out; @@ -791,7 +791,7 @@ static int kimage_load_crash_segment(struct kimage *image, char *ptr; size_t uchunk, mchunk; - page = pfn_to_page(maddr >> PAGE_SHIFT); + page = boot_pfn_to_page(maddr >> PAGE_SHIFT); if (!page) { result = -ENOMEM; goto out; @@ -919,7 +919,7 @@ void __weak crash_free_reserved_phys_range(unsigned long begin, unsigned long addr; for (addr = begin; addr < end; addr += PAGE_SIZE) - free_reserved_page(pfn_to_page(addr >> PAGE_SHIFT)); + free_reserved_page(boot_pfn_to_page(addr >> PAGE_SHIFT)); } int crash_shrink_memory(unsigned long new_size) -- cgit v0.10.2 From 51d5d12b8f3df2b770974ce4aa6196c6b7d485eb Mon Sep 17 00:00:00 2001 From: Vitaly Andrianov Date: Tue, 2 Aug 2016 14:06:07 -0700 Subject: ARM: keystone: dts: add psci command definition This commit adds definition for cpu_on, cpu_off and cpu_suspend commands. These definitions must match the corresponding PSCI definitions in boot monitor. Having those command and corresponding PSCI support in boot monitor allows run time CPU hot plugin. Link: http://lkml.kernel.org/r/E1b8koV-0004Hf-2j@rmk-PC.armlinux.org.uk Signed-off-by: Keerthy Signed-off-by: Vitaly Andrianov Signed-off-by: Russell King Cc: Pratyush Anand Cc: Eric Biederman Cc: Dave Young Cc: Baoquan He Cc: Vivek Goyal Cc: Simon Horman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/arm/boot/dts/keystone.dtsi b/arch/arm/boot/dts/keystone.dtsi index e23f46d..00cb314 100644 --- a/arch/arm/boot/dts/keystone.dtsi +++ b/arch/arm/boot/dts/keystone.dtsi @@ -70,6 +70,14 @@ cpu_on = <0x84000003>; }; + psci { + compatible = "arm,psci"; + method = "smc"; + cpu_suspend = <0x84000001>; + cpu_off = <0x84000002>; + cpu_on = <0x84000003>; + }; + soc { #address-cells = <1>; #size-cells = <1>; -- cgit v0.10.2 From 0719392a61a9dbc2c850bc7bd1a17efba953fcf5 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 2 Aug 2016 14:06:10 -0700 Subject: ARM: kexec: fix kexec for Keystone 2 Provide kexec with the boot view of memory by overriding the normal kexec translation functions added in a previous patch. We also need to fix a call to memblock in machine_kexec_prepare() so that we provide it with a running-view physical address rather than a boot- view physical address. Link: http://lkml.kernel.org/r/E1b8koa-0004Hl-Ey@rmk-PC.armlinux.org.uk Signed-off-by: Russell King Cc: Keerthy Cc: Pratyush Anand Cc: Vitaly Andrianov Cc: Eric Biederman Cc: Dave Young Cc: Baoquan He Cc: Vivek Goyal Cc: Simon Horman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/arm/include/asm/kexec.h b/arch/arm/include/asm/kexec.h index c2b9b4b..1869af6 100644 --- a/arch/arm/include/asm/kexec.h +++ b/arch/arm/include/asm/kexec.h @@ -53,6 +53,30 @@ static inline void crash_setup_regs(struct pt_regs *newregs, /* Function pointer to optional machine-specific reinitialization */ extern void (*kexec_reinit)(void); +static inline unsigned long phys_to_boot_phys(phys_addr_t phys) +{ + return phys_to_idmap(phys); +} +#define phys_to_boot_phys phys_to_boot_phys + +static inline phys_addr_t boot_phys_to_phys(unsigned long entry) +{ + return idmap_to_phys(entry); +} +#define boot_phys_to_phys boot_phys_to_phys + +static inline unsigned long page_to_boot_pfn(struct page *page) +{ + return page_to_pfn(page) + (arch_phys_to_idmap_offset >> PAGE_SHIFT); +} +#define page_to_boot_pfn page_to_boot_pfn + +static inline struct page *boot_pfn_to_page(unsigned long boot_pfn) +{ + return pfn_to_page(boot_pfn - (arch_phys_to_idmap_offset >> PAGE_SHIFT)); +} +#define boot_pfn_to_page boot_pfn_to_page + #endif /* __ASSEMBLY__ */ #endif /* CONFIG_KEXEC */ diff --git a/arch/arm/kernel/machine_kexec.c b/arch/arm/kernel/machine_kexec.c index 59fd0e2..b18c1ea 100644 --- a/arch/arm/kernel/machine_kexec.c +++ b/arch/arm/kernel/machine_kexec.c @@ -57,7 +57,7 @@ int machine_kexec_prepare(struct kimage *image) for (i = 0; i < image->nr_segments; i++) { current_segment = &image->segment[i]; - if (!memblock_is_region_memory(current_segment->mem, + if (!memblock_is_region_memory(idmap_to_phys(current_segment->mem), current_segment->memsz)) return -EINVAL; -- cgit v0.10.2 From b26e27ddfd2a986dc53e259aba572f3aac182eb8 Mon Sep 17 00:00:00 2001 From: Hidehiro Kawai Date: Tue, 2 Aug 2016 14:06:13 -0700 Subject: kexec: use core_param for crash_kexec_post_notifiers boot option crash_kexec_post_notifiers ia a boot option which controls whether the 1st kernel calls panic notifiers or not before booting the 2nd kernel. However, there is no need to limit it to being modifiable only at boot time. So, use core_param instead of early_param. Link: http://lkml.kernel.org/r/20160705113327.5864.43139.stgit@softrs Signed-off-by: Hidehiro Kawai Cc: Dave Young Cc: Baoquan He Cc: Vivek Goyal Cc: Eric Biederman Cc: Masami Hiramatsu Cc: Borislav Petkov Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/panic.c b/kernel/panic.c index 8aa7449..ca8cea1 100644 --- a/kernel/panic.c +++ b/kernel/panic.c @@ -108,6 +108,7 @@ void panic(const char *fmt, ...) long i, i_next = 0; int state = 0; int old_cpu, this_cpu; + bool _crash_kexec_post_notifiers = crash_kexec_post_notifiers; /* * Disable local interrupts. This will prevent panic_smp_self_stop @@ -160,7 +161,7 @@ void panic(const char *fmt, ...) * * Bypass the panic_cpu check and call __crash_kexec directly. */ - if (!crash_kexec_post_notifiers) { + if (!_crash_kexec_post_notifiers) { printk_nmi_flush_on_panic(); __crash_kexec(NULL); } @@ -191,7 +192,7 @@ void panic(const char *fmt, ...) * * Bypass the panic_cpu check and call __crash_kexec directly. */ - if (crash_kexec_post_notifiers) + if (_crash_kexec_post_notifiers) __crash_kexec(NULL); bust_spinlocks(0); @@ -571,13 +572,7 @@ EXPORT_SYMBOL(__stack_chk_fail); core_param(panic, panic_timeout, int, 0644); core_param(pause_on_oops, pause_on_oops, int, 0644); core_param(panic_on_warn, panic_on_warn, int, 0644); - -static int __init setup_crash_kexec_post_notifiers(char *s) -{ - crash_kexec_post_notifiers = true; - return 0; -} -early_param("crash_kexec_post_notifiers", setup_crash_kexec_post_notifiers); +core_param(crash_kexec_post_notifiers, crash_kexec_post_notifiers, bool, 0644); static int __init oops_setup(char *s) { -- cgit v0.10.2 From 21db79e8bb054d0351a6b1b464f1c9c47a2e6e8d Mon Sep 17 00:00:00 2001 From: Petr Tesarik Date: Tue, 2 Aug 2016 14:06:16 -0700 Subject: kexec: add a kexec_crash_loaded() function Provide a wrapper function to be used by kernel code to check whether a crash kernel is loaded. It returns the same value that can be seen in /sys/kernel/kexec_crash_loaded by userspace programs. I'm exporting the function, because it will be used by Xen, and it is possible to compile Xen modules separately to enable the use of PV drivers with unmodified bare-metal kernels. Link: http://lkml.kernel.org/r/20160713121955.14969.69080.stgit@hananiah.suse.cz Signed-off-by: Petr Tesarik Cc: Juergen Gross Cc: Josh Triplett Cc: Ingo Molnar Cc: Thomas Gleixner Cc: Eric Biederman Cc: "H. Peter Anvin" Cc: Boris Ostrovsky Cc: "Paul E. McKenney" Cc: Dave Young Cc: David Vrabel Cc: Vivek Goyal Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/kexec.h b/include/linux/kexec.h index 23e14a4..d743777 100644 --- a/include/linux/kexec.h +++ b/include/linux/kexec.h @@ -230,6 +230,7 @@ extern void *kexec_purgatory_get_symbol_addr(struct kimage *image, extern void __crash_kexec(struct pt_regs *); extern void crash_kexec(struct pt_regs *); int kexec_should_crash(struct task_struct *); +int kexec_crash_loaded(void); void crash_save_cpu(struct pt_regs *regs, int cpu); void crash_save_vmcoreinfo(void); void arch_crash_save_vmcoreinfo(void); @@ -364,6 +365,7 @@ struct task_struct; static inline void __crash_kexec(struct pt_regs *regs) { } static inline void crash_kexec(struct pt_regs *regs) { } static inline int kexec_should_crash(struct task_struct *p) { return 0; } +static inline int kexec_crash_loaded(void) { return 0; } #define kexec_in_progress false #endif /* CONFIG_KEXEC_CORE */ diff --git a/kernel/kexec_core.c b/kernel/kexec_core.c index 73d4c5f..7045340 100644 --- a/kernel/kexec_core.c +++ b/kernel/kexec_core.c @@ -95,6 +95,12 @@ int kexec_should_crash(struct task_struct *p) return 0; } +int kexec_crash_loaded(void) +{ + return !!kexec_crash_image; +} +EXPORT_SYMBOL_GPL(kexec_crash_loaded); + /* * When kexec transitions to the new kernel there is a one-to-one * mapping between physical and virtual addresses. On processors diff --git a/kernel/ksysfs.c b/kernel/ksysfs.c index 9f1920d..ee1bc1b 100644 --- a/kernel/ksysfs.c +++ b/kernel/ksysfs.c @@ -101,7 +101,7 @@ KERNEL_ATTR_RO(kexec_loaded); static ssize_t kexec_crash_loaded_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf) { - return sprintf(buf, "%d\n", !!kexec_crash_image); + return sprintf(buf, "%d\n", kexec_crash_loaded()); } KERNEL_ATTR_RO(kexec_crash_loaded); -- cgit v0.10.2 From c0253115968c35f3e1ee497282efb75ccf29fb98 Mon Sep 17 00:00:00 2001 From: Petr Tesarik Date: Tue, 2 Aug 2016 14:06:19 -0700 Subject: kexec: allow kdump with crash_kexec_post_notifiers If a crash kernel is loaded, do not crash the running domain. This is needed if the kernel is loaded with crash_kexec_post_notifiers, because panic notifiers are run before __crash_kexec() in that case, and this Xen hook prevents its being called later. [akpm@linux-foundation.org: build fix: unconditionally include kexec.h] Link: http://lkml.kernel.org/r/20160713122000.14969.99963.stgit@hananiah.suse.cz Signed-off-by: Petr Tesarik Cc: Juergen Gross Cc: Josh Triplett Cc: Ingo Molnar Cc: Thomas Gleixner Cc: Eric Biederman Cc: "H. Peter Anvin" Cc: Boris Ostrovsky Cc: "Paul E. McKenney" Cc: Dave Young Cc: David Vrabel Cc: Vivek Goyal Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index cd99305..8ffb089 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -34,9 +34,7 @@ #include #include -#ifdef CONFIG_KEXEC_CORE #include -#endif #include #include @@ -1334,7 +1332,8 @@ static void xen_crash_shutdown(struct pt_regs *regs) static int xen_panic_event(struct notifier_block *this, unsigned long event, void *ptr) { - xen_reboot(SHUTDOWN_crash); + if (!kexec_crash_loaded()) + xen_reboot(SHUTDOWN_crash); return NOTIFY_DONE; } -- cgit v0.10.2 From 1730f146604ea426e54938cdbcf87df1047ef0dc Mon Sep 17 00:00:00 2001 From: zhong jiang Date: Tue, 2 Aug 2016 14:06:22 -0700 Subject: kexec: add restriction on kexec_load() segment sizes I hit the following issue when run trinity in my system. The kernel is 3.4 version, but mainline has the same issue. The root cause is that the segment size is too large so the kerenl spends too long trying to allocate a page. Other cases will block until the test case quits. Also, OOM conditions will occur. Call Trace: __alloc_pages_nodemask+0x14c/0x8f0 alloc_pages_current+0xaf/0x120 kimage_alloc_pages+0x10/0x60 kimage_alloc_control_pages+0x5d/0x270 machine_kexec_prepare+0xe5/0x6c0 ? kimage_free_page_list+0x52/0x70 sys_kexec_load+0x141/0x600 ? vfs_write+0x100/0x180 system_call_fastpath+0x16/0x1b The patch changes sanity_check_segment_list() to verify that the usage by all segments does not exceed half of memory. [akpm@linux-foundation.org: fix for kexec-return-error-number-directly.patch, update comment] Link: http://lkml.kernel.org/r/1469625474-53904-1-git-send-email-zhongjiang@huawei.com Signed-off-by: zhong jiang Suggested-by: Eric W. Biederman Cc: Vivek Goyal Cc: Dave Young Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/kexec_core.c b/kernel/kexec_core.c index 7045340..5616755 100644 --- a/kernel/kexec_core.c +++ b/kernel/kexec_core.c @@ -146,6 +146,7 @@ EXPORT_SYMBOL_GPL(kexec_crash_loaded); * allocating pages whose destination address we do not care about. */ #define KIMAGE_NO_DEST (-1UL) +#define PAGE_COUNT(x) (((x) + PAGE_SIZE - 1) >> PAGE_SHIFT) static struct page *kimage_alloc_page(struct kimage *image, gfp_t gfp_mask, @@ -155,6 +156,7 @@ int sanity_check_segment_list(struct kimage *image) { int i; unsigned long nr_segments = image->nr_segments; + unsigned long total_pages = 0; /* * Verify we have good destination addresses. The caller is @@ -215,6 +217,21 @@ int sanity_check_segment_list(struct kimage *image) } /* + * Verify that no more than half of memory will be consumed. If the + * request from userspace is too large, a large amount of time will be + * wasted allocating pages, which can cause a soft lockup. + */ + for (i = 0; i < nr_segments; i++) { + if (PAGE_COUNT(image->segment[i].memsz) > totalram_pages / 2) + return -EINVAL; + + total_pages += PAGE_COUNT(image->segment[i].memsz); + } + + if (total_pages > totalram_pages / 2) + return -EINVAL; + + /* * Verify we have good destination addresses. Normally * the caller is responsible for making certain we don't * attempt to load the new image into invalid or reserved -- cgit v0.10.2 From b6e8d4aa1110306378af0f3472a6b85a1f039a16 Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:06:25 -0700 Subject: rapidio: add RapidIO channelized messaging driver Add channelized messaging driver to support native RapidIO messaging exchange between multiple senders/recipients on devices that use kernel RapidIO subsystem services. This device driver is the result of collaboration within the RapidIO.org Software Task Group (STG) between Texas Instruments, Prodrive Technologies, Nokia Networks, BAE and IDT. Additional input was received from other members of RapidIO.org. The objective was to create a character mode driver interface which exposes messaging capabilities of RapidIO endpoint devices (mports) directly to applications, in a manner that allows the numerous and varied RapidIO implementations to interoperate. This char mode device driver allows user-space applications to setup messaging communication channels using single shared RapidIO messaging mailbox. By default this driver uses RapidIO MBOX_1 (MBOX_0 is reserved for use by RIONET Ethernet emulation driver). [weiyj.lk@gmail.com: rapidio/rio_cm: fix return value check in riocm_init()] Link: http://lkml.kernel.org/r/1469198221-21970-1-git-send-email-alexandre.bounine@idt.com Link: http://lkml.kernel.org/r/1468952862-18056-1-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Tested-by: Barry Wood Cc: Matt Porter Cc: Aurelien Jacquiot Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/Documentation/rapidio/rio_cm.txt b/Documentation/rapidio/rio_cm.txt new file mode 100644 index 0000000..27aa401 --- /dev/null +++ b/Documentation/rapidio/rio_cm.txt @@ -0,0 +1,119 @@ +RapidIO subsystem Channelized Messaging character device driver (rio_cm.c) +========================================================================== + +Version History: +---------------- + 1.0.0 - Initial driver release. + +========================================================================== + +I. Overview + +This device driver is the result of collaboration within the RapidIO.org +Software Task Group (STG) between Texas Instruments, Prodrive Technologies, +Nokia Networks, BAE and IDT. Additional input was received from other members +of RapidIO.org. + +The objective was to create a character mode driver interface which exposes +messaging capabilities of RapidIO endpoint devices (mports) directly +to applications, in a manner that allows the numerous and varied RapidIO +implementations to interoperate. + +This driver (RIO_CM) provides to user-space applications shared access to +RapidIO mailbox messaging resources. + +RapidIO specification (Part 2) defines that endpoint devices may have up to four +messaging mailboxes in case of multi-packet message (up to 4KB) and +up to 64 mailboxes if single-packet messages (up to 256 B) are used. In addition +to protocol definition limitations, a particular hardware implementation can +have reduced number of messaging mailboxes. RapidIO aware applications must +therefore share the messaging resources of a RapidIO endpoint. + +Main purpose of this device driver is to provide RapidIO mailbox messaging +capability to large number of user-space processes by introducing socket-like +operations using a single messaging mailbox. This allows applications to +use the limited RapidIO messaging hardware resources efficiently. + +Most of device driver's operations are supported through 'ioctl' system calls. + +When loaded this device driver creates a single file system node named rio_cm +in /dev directory common for all registered RapidIO mport devices. + +Following ioctl commands are available to user-space applications: + +- RIO_CM_MPORT_GET_LIST : Returns to caller list of local mport devices that + support messaging operations (number of entries up to RIO_MAX_MPORTS). + Each list entry is combination of mport's index in the system and RapidIO + destination ID assigned to the port. +- RIO_CM_EP_GET_LIST_SIZE : Returns number of messaging capable remote endpoints + in a RapidIO network associated with the specified mport device. +- RIO_CM_EP_GET_LIST : Returns list of RapidIO destination IDs for messaging + capable remote endpoints (peers) available in a RapidIO network associated + with the specified mport device. +- RIO_CM_CHAN_CREATE : Creates RapidIO message exchange channel data structure + with channel ID assigned automatically or as requested by a caller. +- RIO_CM_CHAN_BIND : Binds the specified channel data structure to the specified + mport device. +- RIO_CM_CHAN_LISTEN : Enables listening for connection requests on the specified + channel. +- RIO_CM_CHAN_ACCEPT : Accepts a connection request from peer on the specified + channel. If wait timeout for this request is specified by a caller it is + a blocking call. If timeout set to 0 this is non-blocking call - ioctl + handler checks for a pending connection request and if one is not available + exits with -EGAIN error status immediately. +- RIO_CM_CHAN_CONNECT : Sends a connection request to a remote peer/channel. +- RIO_CM_CHAN_SEND : Sends a data message through the specified channel. + The handler for this request assumes that message buffer specified by + a caller includes the reserved space for a packet header required by + this driver. +- RIO_CM_CHAN_RECEIVE : Receives a data message through a connected channel. + If the channel does not have an incoming message ready to return this ioctl + handler will wait for new message until timeout specified by a caller + expires. If timeout value is set to 0, ioctl handler uses a default value + defined by MAX_SCHEDULE_TIMEOUT. +- RIO_CM_CHAN_CLOSE : Closes a specified channel and frees associated buffers. + If the specified channel is in the CONNECTED state, sends close notification + to the remote peer. + +The ioctl command codes and corresponding data structures intended for use by +user-space applications are defined in 'include/uapi/linux/rio_cm_cdev.h'. + +II. Hardware Compatibility + +This device driver uses standard interfaces defined by kernel RapidIO subsystem +and therefore it can be used with any mport device driver registered by RapidIO +subsystem with limitations set by available mport HW implementation of messaging +mailboxes. + +III. Module parameters + +- 'dbg_level' - This parameter allows to control amount of debug information + generated by this device driver. This parameter is formed by set of + bit masks that correspond to the specific functional block. + For mask definitions see 'drivers/rapidio/devices/rio_cm.c' + This parameter can be changed dynamically. + Use CONFIG_RAPIDIO_DEBUG=y to enable debug output at the top level. + +- 'cmbox' - Number of RapidIO mailbox to use (default value is 1). + This parameter allows to set messaging mailbox number that will be used + within entire RapidIO network. It can be used when default mailbox is + used by other device drivers or is not supported by some nodes in the + RapidIO network. + +- 'chstart' - Start channel number for dynamic assignment. Default value - 256. + Allows to exclude channel numbers below this parameter from dynamic + allocation to avoid conflicts with software components that use + reserved predefined channel numbers. + +IV. Known problems + + None. + +V. User-space Applications and API Library + +Messaging API library and applications that use this device driver are available +from RapidIO.org. + +VI. TODO List + +- Add support for system notification messages (reserved channel 0). diff --git a/drivers/rapidio/Kconfig b/drivers/rapidio/Kconfig index b5a10d3..d6d2f20 100644 --- a/drivers/rapidio/Kconfig +++ b/drivers/rapidio/Kconfig @@ -67,6 +67,15 @@ config RAPIDIO_ENUM_BASIC endchoice +config RAPIDIO_CHMAN + tristate "RapidIO Channelized Messaging driver" + depends on RAPIDIO + help + This option includes RapidIO channelized messaging driver which + provides socket-like interface to allow sharing of single RapidIO + messaging mailbox between multiple user-space applications. + See "Documentation/rapidio/rio_cm.txt" for driver description. + config RAPIDIO_MPORT_CDEV tristate "RapidIO /dev mport device driver" depends on RAPIDIO diff --git a/drivers/rapidio/Makefile b/drivers/rapidio/Makefile index 6271ada..74dcea4 100644 --- a/drivers/rapidio/Makefile +++ b/drivers/rapidio/Makefile @@ -5,6 +5,7 @@ obj-$(CONFIG_RAPIDIO) += rapidio.o rapidio-y := rio.o rio-access.o rio-driver.o rio-sysfs.o obj-$(CONFIG_RAPIDIO_ENUM_BASIC) += rio-scan.o +obj-$(CONFIG_RAPIDIO_CHMAN) += rio_cm.o obj-$(CONFIG_RAPIDIO) += switches/ obj-$(CONFIG_RAPIDIO) += devices/ diff --git a/drivers/rapidio/rio_cm.c b/drivers/rapidio/rio_cm.c new file mode 100644 index 0000000..cecc15a --- /dev/null +++ b/drivers/rapidio/rio_cm.c @@ -0,0 +1,2366 @@ +/* + * rio_cm - RapidIO Channelized Messaging Driver + * + * Copyright 2013-2016 Integrated Device Technology, Inc. + * Copyright (c) 2015, Prodrive Technologies + * Copyright (c) 2015, RapidIO Trade Association + * + * 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. + * + * THIS PROGRAM IS DISTRIBUTED IN THE HOPE THAT IT WILL BE USEFUL, + * BUT WITHOUT ANY WARRANTY; WITHOUT EVEN THE IMPLIED WARRANTY OF + * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. SEE THE + * GNU GENERAL PUBLIC LICENSE FOR MORE DETAILS. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#define DRV_NAME "rio_cm" +#define DRV_VERSION "1.0.0" +#define DRV_AUTHOR "Alexandre Bounine " +#define DRV_DESC "RapidIO Channelized Messaging Driver" +#define DEV_NAME "rio_cm" + +/* Debug output filtering masks */ +enum { + DBG_NONE = 0, + DBG_INIT = BIT(0), /* driver init */ + DBG_EXIT = BIT(1), /* driver exit */ + DBG_MPORT = BIT(2), /* mport add/remove */ + DBG_RDEV = BIT(3), /* RapidIO device add/remove */ + DBG_CHOP = BIT(4), /* channel operations */ + DBG_WAIT = BIT(5), /* waiting for events */ + DBG_TX = BIT(6), /* message TX */ + DBG_TX_EVENT = BIT(7), /* message TX event */ + DBG_RX_DATA = BIT(8), /* inbound data messages */ + DBG_RX_CMD = BIT(9), /* inbound REQ/ACK/NACK messages */ + DBG_ALL = ~0, +}; + +#ifdef DEBUG +#define riocm_debug(level, fmt, arg...) \ + do { \ + if (DBG_##level & dbg_level) \ + pr_debug(DRV_NAME ": %s " fmt "\n", \ + __func__, ##arg); \ + } while (0) +#else +#define riocm_debug(level, fmt, arg...) \ + no_printk(KERN_DEBUG pr_fmt(DRV_NAME fmt "\n"), ##arg) +#endif + +#define riocm_warn(fmt, arg...) \ + pr_warn(DRV_NAME ": %s WARNING " fmt "\n", __func__, ##arg) + +#define riocm_error(fmt, arg...) \ + pr_err(DRV_NAME ": %s ERROR " fmt "\n", __func__, ##arg) + + +static int cmbox = 1; +module_param(cmbox, int, S_IRUGO); +MODULE_PARM_DESC(cmbox, "RapidIO Mailbox number (default 1)"); + +static int chstart = 256; +module_param(chstart, int, S_IRUGO); +MODULE_PARM_DESC(chstart, + "Start channel number for dynamic allocation (default 256)"); + +#ifdef DEBUG +static u32 dbg_level = DBG_NONE; +module_param(dbg_level, uint, S_IWUSR | S_IRUGO); +MODULE_PARM_DESC(dbg_level, "Debugging output level (default 0 = none)"); +#endif + +MODULE_AUTHOR(DRV_AUTHOR); +MODULE_DESCRIPTION(DRV_DESC); +MODULE_LICENSE("GPL"); +MODULE_VERSION(DRV_VERSION); + +#define RIOCM_TX_RING_SIZE 128 +#define RIOCM_RX_RING_SIZE 128 +#define RIOCM_CONNECT_TO 3 /* connect response TO (in sec) */ + +#define RIOCM_MAX_CHNUM 0xffff /* Use full range of u16 field */ +#define RIOCM_CHNUM_AUTO 0 +#define RIOCM_MAX_EP_COUNT 0x10000 /* Max number of endpoints */ + +enum rio_cm_state { + RIO_CM_IDLE, + RIO_CM_CONNECT, + RIO_CM_CONNECTED, + RIO_CM_DISCONNECT, + RIO_CM_CHAN_BOUND, + RIO_CM_LISTEN, + RIO_CM_DESTROYING, +}; + +enum rio_cm_pkt_type { + RIO_CM_SYS = 0xaa, + RIO_CM_CHAN = 0x55, +}; + +enum rio_cm_chop { + CM_CONN_REQ, + CM_CONN_ACK, + CM_CONN_CLOSE, + CM_DATA_MSG, +}; + +struct rio_ch_base_bhdr { + u32 src_id; + u32 dst_id; +#define RIO_HDR_LETTER_MASK 0xffff0000 +#define RIO_HDR_MBOX_MASK 0x0000ffff + u8 src_mbox; + u8 dst_mbox; + u8 type; +} __attribute__((__packed__)); + +struct rio_ch_chan_hdr { + struct rio_ch_base_bhdr bhdr; + u8 ch_op; + u16 dst_ch; + u16 src_ch; + u16 msg_len; + u16 rsrvd; +} __attribute__((__packed__)); + +struct tx_req { + struct list_head node; + struct rio_dev *rdev; + void *buffer; + size_t len; +}; + +struct cm_dev { + struct list_head list; + struct rio_mport *mport; + void *rx_buf[RIOCM_RX_RING_SIZE]; + int rx_slots; + struct mutex rx_lock; + + void *tx_buf[RIOCM_TX_RING_SIZE]; + int tx_slot; + int tx_cnt; + int tx_ack_slot; + struct list_head tx_reqs; + spinlock_t tx_lock; + + struct list_head peers; + u32 npeers; + struct workqueue_struct *rx_wq; + struct work_struct rx_work; +}; + +struct chan_rx_ring { + void *buf[RIOCM_RX_RING_SIZE]; + int head; + int tail; + int count; + + /* Tracking RX buffers reported to upper level */ + void *inuse[RIOCM_RX_RING_SIZE]; + int inuse_cnt; +}; + +struct rio_channel { + u16 id; /* local channel ID */ + struct kref ref; /* channel refcount */ + struct file *filp; + struct cm_dev *cmdev; /* associated CM device object */ + struct rio_dev *rdev; /* remote RapidIO device */ + enum rio_cm_state state; + int error; + spinlock_t lock; + void *context; + u32 loc_destid; /* local destID */ + u32 rem_destid; /* remote destID */ + u16 rem_channel; /* remote channel ID */ + struct list_head accept_queue; + struct list_head ch_node; + struct completion comp; + struct completion comp_close; + struct chan_rx_ring rx_ring; +}; + +struct cm_peer { + struct list_head node; + struct rio_dev *rdev; +}; + +struct rio_cm_work { + struct work_struct work; + struct cm_dev *cm; + void *data; +}; + +struct conn_req { + struct list_head node; + u32 destid; /* requester destID */ + u16 chan; /* requester channel ID */ + struct cm_dev *cmdev; +}; + +/* + * A channel_dev structure represents a CM_CDEV + * @cdev Character device + * @dev Associated device object + */ +struct channel_dev { + struct cdev cdev; + struct device *dev; +}; + +static struct rio_channel *riocm_ch_alloc(u16 ch_num); +static void riocm_ch_free(struct kref *ref); +static int riocm_post_send(struct cm_dev *cm, struct rio_dev *rdev, + void *buffer, size_t len); +static int riocm_ch_close(struct rio_channel *ch); + +static DEFINE_SPINLOCK(idr_lock); +static DEFINE_IDR(ch_idr); + +static LIST_HEAD(cm_dev_list); +static DECLARE_RWSEM(rdev_sem); + +static struct class *dev_class; +static unsigned int dev_major; +static unsigned int dev_minor_base; +static dev_t dev_number; +static struct channel_dev riocm_cdev; + +#define is_msg_capable(src_ops, dst_ops) \ + ((src_ops & RIO_SRC_OPS_DATA_MSG) && \ + (dst_ops & RIO_DST_OPS_DATA_MSG)) +#define dev_cm_capable(dev) \ + is_msg_capable(dev->src_ops, dev->dst_ops) + +static int riocm_cmp(struct rio_channel *ch, enum rio_cm_state cmp) +{ + int ret; + + spin_lock_bh(&ch->lock); + ret = (ch->state == cmp); + spin_unlock_bh(&ch->lock); + return ret; +} + +static int riocm_cmp_exch(struct rio_channel *ch, + enum rio_cm_state cmp, enum rio_cm_state exch) +{ + int ret; + + spin_lock_bh(&ch->lock); + ret = (ch->state == cmp); + if (ret) + ch->state = exch; + spin_unlock_bh(&ch->lock); + return ret; +} + +static enum rio_cm_state riocm_exch(struct rio_channel *ch, + enum rio_cm_state exch) +{ + enum rio_cm_state old; + + spin_lock_bh(&ch->lock); + old = ch->state; + ch->state = exch; + spin_unlock_bh(&ch->lock); + return old; +} + +static struct rio_channel *riocm_get_channel(u16 nr) +{ + struct rio_channel *ch; + + spin_lock_bh(&idr_lock); + ch = idr_find(&ch_idr, nr); + if (ch) + kref_get(&ch->ref); + spin_unlock_bh(&idr_lock); + return ch; +} + +static void riocm_put_channel(struct rio_channel *ch) +{ + kref_put(&ch->ref, riocm_ch_free); +} + +static void *riocm_rx_get_msg(struct cm_dev *cm) +{ + void *msg; + int i; + + msg = rio_get_inb_message(cm->mport, cmbox); + if (msg) { + for (i = 0; i < RIOCM_RX_RING_SIZE; i++) { + if (cm->rx_buf[i] == msg) { + cm->rx_buf[i] = NULL; + cm->rx_slots++; + break; + } + } + + if (i == RIOCM_RX_RING_SIZE) + riocm_warn("no record for buffer 0x%p", msg); + } + + return msg; +} + +/* + * riocm_rx_fill - fills a ring of receive buffers for given cm device + * @cm: cm_dev object + * @nent: max number of entries to fill + * + * Returns: none + */ +static void riocm_rx_fill(struct cm_dev *cm, int nent) +{ + int i; + + if (cm->rx_slots == 0) + return; + + for (i = 0; i < RIOCM_RX_RING_SIZE && cm->rx_slots && nent; i++) { + if (cm->rx_buf[i] == NULL) { + cm->rx_buf[i] = kmalloc(RIO_MAX_MSG_SIZE, GFP_KERNEL); + if (cm->rx_buf[i] == NULL) + break; + rio_add_inb_buffer(cm->mport, cmbox, cm->rx_buf[i]); + cm->rx_slots--; + nent--; + } + } +} + +/* + * riocm_rx_free - frees all receive buffers associated with given cm device + * @cm: cm_dev object + * + * Returns: none + */ +static void riocm_rx_free(struct cm_dev *cm) +{ + int i; + + for (i = 0; i < RIOCM_RX_RING_SIZE; i++) { + if (cm->rx_buf[i] != NULL) { + kfree(cm->rx_buf[i]); + cm->rx_buf[i] = NULL; + } + } +} + +/* + * riocm_req_handler - connection request handler + * @cm: cm_dev object + * @req_data: pointer to the request packet + * + * Returns: 0 if success, or + * -EINVAL if channel is not in correct state, + * -ENODEV if cannot find a channel with specified ID, + * -ENOMEM if unable to allocate memory to store the request + */ +static int riocm_req_handler(struct cm_dev *cm, void *req_data) +{ + struct rio_channel *ch; + struct conn_req *req; + struct rio_ch_chan_hdr *hh = req_data; + u16 chnum; + + chnum = ntohs(hh->dst_ch); + + ch = riocm_get_channel(chnum); + + if (!ch) + return -ENODEV; + + if (ch->state != RIO_CM_LISTEN) { + riocm_debug(RX_CMD, "channel %d is not in listen state", chnum); + riocm_put_channel(ch); + return -EINVAL; + } + + req = kzalloc(sizeof(*req), GFP_KERNEL); + if (!req) { + riocm_put_channel(ch); + return -ENOMEM; + } + + req->destid = ntohl(hh->bhdr.src_id); + req->chan = ntohs(hh->src_ch); + req->cmdev = cm; + + spin_lock_bh(&ch->lock); + list_add_tail(&req->node, &ch->accept_queue); + spin_unlock_bh(&ch->lock); + complete(&ch->comp); + riocm_put_channel(ch); + + return 0; +} + +/* + * riocm_resp_handler - response to connection request handler + * @resp_data: pointer to the response packet + * + * Returns: 0 if success, or + * -EINVAL if channel is not in correct state, + * -ENODEV if cannot find a channel with specified ID, + */ +static int riocm_resp_handler(void *resp_data) +{ + struct rio_channel *ch; + struct rio_ch_chan_hdr *hh = resp_data; + u16 chnum; + + chnum = ntohs(hh->dst_ch); + ch = riocm_get_channel(chnum); + if (!ch) + return -ENODEV; + + if (ch->state != RIO_CM_CONNECT) { + riocm_put_channel(ch); + return -EINVAL; + } + + riocm_exch(ch, RIO_CM_CONNECTED); + ch->rem_channel = ntohs(hh->src_ch); + complete(&ch->comp); + riocm_put_channel(ch); + + return 0; +} + +/* + * riocm_close_handler - channel close request handler + * @req_data: pointer to the request packet + * + * Returns: 0 if success, or + * -ENODEV if cannot find a channel with specified ID, + * + error codes returned by riocm_ch_close. + */ +static int riocm_close_handler(void *data) +{ + struct rio_channel *ch; + struct rio_ch_chan_hdr *hh = data; + int ret; + + riocm_debug(RX_CMD, "for ch=%d", ntohs(hh->dst_ch)); + + spin_lock_bh(&idr_lock); + ch = idr_find(&ch_idr, ntohs(hh->dst_ch)); + if (!ch) { + spin_unlock_bh(&idr_lock); + return -ENODEV; + } + idr_remove(&ch_idr, ch->id); + spin_unlock_bh(&idr_lock); + + riocm_exch(ch, RIO_CM_DISCONNECT); + + ret = riocm_ch_close(ch); + if (ret) + riocm_debug(RX_CMD, "riocm_ch_close() returned %d", ret); + + return 0; +} + +/* + * rio_cm_handler - function that services request (non-data) packets + * @cm: cm_dev object + * @data: pointer to the packet + */ +static void rio_cm_handler(struct cm_dev *cm, void *data) +{ + struct rio_ch_chan_hdr *hdr; + + if (!rio_mport_is_running(cm->mport)) + goto out; + + hdr = data; + + riocm_debug(RX_CMD, "OP=%x for ch=%d from %d", + hdr->ch_op, ntohs(hdr->dst_ch), ntohs(hdr->src_ch)); + + switch (hdr->ch_op) { + case CM_CONN_REQ: + riocm_req_handler(cm, data); + break; + case CM_CONN_ACK: + riocm_resp_handler(data); + break; + case CM_CONN_CLOSE: + riocm_close_handler(data); + break; + default: + riocm_error("Invalid packet header"); + break; + } +out: + kfree(data); +} + +/* + * rio_rx_data_handler - received data packet handler + * @cm: cm_dev object + * @buf: data packet + * + * Returns: 0 if success, or + * -ENODEV if cannot find a channel with specified ID, + * -EIO if channel is not in CONNECTED state, + * -ENOMEM if channel RX queue is full (packet discarded) + */ +static int rio_rx_data_handler(struct cm_dev *cm, void *buf) +{ + struct rio_ch_chan_hdr *hdr; + struct rio_channel *ch; + + hdr = buf; + + riocm_debug(RX_DATA, "for ch=%d", ntohs(hdr->dst_ch)); + + ch = riocm_get_channel(ntohs(hdr->dst_ch)); + if (!ch) { + /* Discard data message for non-existing channel */ + kfree(buf); + return -ENODEV; + } + + /* Place pointer to the buffer into channel's RX queue */ + spin_lock(&ch->lock); + + if (ch->state != RIO_CM_CONNECTED) { + /* Channel is not ready to receive data, discard a packet */ + riocm_debug(RX_DATA, "ch=%d is in wrong state=%d", + ch->id, ch->state); + spin_unlock(&ch->lock); + kfree(buf); + riocm_put_channel(ch); + return -EIO; + } + + if (ch->rx_ring.count == RIOCM_RX_RING_SIZE) { + /* If RX ring is full, discard a packet */ + riocm_debug(RX_DATA, "ch=%d is full", ch->id); + spin_unlock(&ch->lock); + kfree(buf); + riocm_put_channel(ch); + return -ENOMEM; + } + + ch->rx_ring.buf[ch->rx_ring.head] = buf; + ch->rx_ring.head++; + ch->rx_ring.count++; + ch->rx_ring.head %= RIOCM_RX_RING_SIZE; + + complete(&ch->comp); + + spin_unlock(&ch->lock); + riocm_put_channel(ch); + + return 0; +} + +/* + * rio_ibmsg_handler - inbound message packet handler + */ +static void rio_ibmsg_handler(struct work_struct *work) +{ + struct cm_dev *cm = container_of(work, struct cm_dev, rx_work); + void *data; + struct rio_ch_chan_hdr *hdr; + + if (!rio_mport_is_running(cm->mport)) + return; + + while (1) { + mutex_lock(&cm->rx_lock); + data = riocm_rx_get_msg(cm); + if (data) + riocm_rx_fill(cm, 1); + mutex_unlock(&cm->rx_lock); + + if (data == NULL) + break; + + hdr = data; + + if (hdr->bhdr.type != RIO_CM_CHAN) { + /* For now simply discard packets other than channel */ + riocm_error("Unsupported TYPE code (0x%x). Msg dropped", + hdr->bhdr.type); + kfree(data); + continue; + } + + /* Process a channel message */ + if (hdr->ch_op == CM_DATA_MSG) + rio_rx_data_handler(cm, data); + else + rio_cm_handler(cm, data); + } +} + +static void riocm_inb_msg_event(struct rio_mport *mport, void *dev_id, + int mbox, int slot) +{ + struct cm_dev *cm = dev_id; + + if (rio_mport_is_running(cm->mport) && !work_pending(&cm->rx_work)) + queue_work(cm->rx_wq, &cm->rx_work); +} + +/* + * rio_txcq_handler - TX completion handler + * @cm: cm_dev object + * @slot: TX queue slot + * + * TX completion handler also ensures that pending request packets are placed + * into transmit queue as soon as a free slot becomes available. This is done + * to give higher priority to request packets during high intensity data flow. + */ +static void rio_txcq_handler(struct cm_dev *cm, int slot) +{ + int ack_slot; + + /* ATTN: Add TX completion notification if/when direct buffer + * transfer is implemented. At this moment only correct tracking + * of tx_count is important. + */ + riocm_debug(TX_EVENT, "for mport_%d slot %d tx_cnt %d", + cm->mport->id, slot, cm->tx_cnt); + + spin_lock(&cm->tx_lock); + ack_slot = cm->tx_ack_slot; + + if (ack_slot == slot) + riocm_debug(TX_EVENT, "slot == ack_slot"); + + while (cm->tx_cnt && ((ack_slot != slot) || + (cm->tx_cnt == RIOCM_TX_RING_SIZE))) { + + cm->tx_buf[ack_slot] = NULL; + ++ack_slot; + ack_slot &= (RIOCM_TX_RING_SIZE - 1); + cm->tx_cnt--; + } + + if (cm->tx_cnt < 0 || cm->tx_cnt > RIOCM_TX_RING_SIZE) + riocm_error("tx_cnt %d out of sync", cm->tx_cnt); + + WARN_ON((cm->tx_cnt < 0) || (cm->tx_cnt > RIOCM_TX_RING_SIZE)); + + cm->tx_ack_slot = ack_slot; + + /* + * If there are pending requests, insert them into transmit queue + */ + if (!list_empty(&cm->tx_reqs) && (cm->tx_cnt < RIOCM_TX_RING_SIZE)) { + struct tx_req *req, *_req; + int rc; + + list_for_each_entry_safe(req, _req, &cm->tx_reqs, node) { + list_del(&req->node); + cm->tx_buf[cm->tx_slot] = req->buffer; + rc = rio_add_outb_message(cm->mport, req->rdev, cmbox, + req->buffer, req->len); + kfree(req->buffer); + kfree(req); + + ++cm->tx_cnt; + ++cm->tx_slot; + cm->tx_slot &= (RIOCM_TX_RING_SIZE - 1); + if (cm->tx_cnt == RIOCM_TX_RING_SIZE) + break; + } + } + + spin_unlock(&cm->tx_lock); +} + +static void riocm_outb_msg_event(struct rio_mport *mport, void *dev_id, + int mbox, int slot) +{ + struct cm_dev *cm = dev_id; + + if (cm && rio_mport_is_running(cm->mport)) + rio_txcq_handler(cm, slot); +} + +static int riocm_queue_req(struct cm_dev *cm, struct rio_dev *rdev, + void *buffer, size_t len) +{ + unsigned long flags; + struct tx_req *treq; + + treq = kzalloc(sizeof(*treq), GFP_KERNEL); + if (treq == NULL) + return -ENOMEM; + + treq->rdev = rdev; + treq->buffer = buffer; + treq->len = len; + + spin_lock_irqsave(&cm->tx_lock, flags); + list_add_tail(&treq->node, &cm->tx_reqs); + spin_unlock_irqrestore(&cm->tx_lock, flags); + return 0; +} + +/* + * riocm_post_send - helper function that places packet into msg TX queue + * @cm: cm_dev object + * @rdev: target RapidIO device object (required by outbound msg interface) + * @buffer: pointer to a packet buffer to send + * @len: length of data to transfer + * @req: request priority flag + * + * Returns: 0 if success, or error code otherwise. + */ +static int riocm_post_send(struct cm_dev *cm, struct rio_dev *rdev, + void *buffer, size_t len) +{ + int rc; + unsigned long flags; + + spin_lock_irqsave(&cm->tx_lock, flags); + + if (cm->mport == NULL) { + rc = -ENODEV; + goto err_out; + } + + if (cm->tx_cnt == RIOCM_TX_RING_SIZE) { + riocm_debug(TX, "Tx Queue is full"); + rc = -EBUSY; + goto err_out; + } + + cm->tx_buf[cm->tx_slot] = buffer; + rc = rio_add_outb_message(cm->mport, rdev, cmbox, buffer, len); + + riocm_debug(TX, "Add buf@%p destid=%x tx_slot=%d tx_cnt=%d", + buffer, rdev->destid, cm->tx_slot, cm->tx_cnt); + + ++cm->tx_cnt; + ++cm->tx_slot; + cm->tx_slot &= (RIOCM_TX_RING_SIZE - 1); + +err_out: + spin_unlock_irqrestore(&cm->tx_lock, flags); + return rc; +} + +/* + * riocm_ch_send - sends a data packet to a remote device + * @ch_id: local channel ID + * @buf: pointer to a data buffer to send (including CM header) + * @len: length of data to transfer (including CM header) + * + * ATTN: ASSUMES THAT THE HEADER SPACE IS RESERVED PART OF THE DATA PACKET + * + * Returns: 0 if success, or + * -EINVAL if one or more input parameters is/are not valid, + * -ENODEV if cannot find a channel with specified ID, + * -EAGAIN if a channel is not in CONNECTED state, + * + error codes returned by HW send routine. + */ +static int riocm_ch_send(u16 ch_id, void *buf, int len) +{ + struct rio_channel *ch; + struct rio_ch_chan_hdr *hdr; + int ret; + + if (buf == NULL || ch_id == 0 || len == 0 || len > RIO_MAX_MSG_SIZE) + return -EINVAL; + + ch = riocm_get_channel(ch_id); + if (!ch) { + riocm_error("%s(%d) ch_%d not found", current->comm, + task_pid_nr(current), ch_id); + return -ENODEV; + } + + if (!riocm_cmp(ch, RIO_CM_CONNECTED)) { + ret = -EAGAIN; + goto err_out; + } + + /* + * Fill buffer header section with corresponding channel data + */ + hdr = buf; + + hdr->bhdr.src_id = htonl(ch->loc_destid); + hdr->bhdr.dst_id = htonl(ch->rem_destid); + hdr->bhdr.src_mbox = cmbox; + hdr->bhdr.dst_mbox = cmbox; + hdr->bhdr.type = RIO_CM_CHAN; + hdr->ch_op = CM_DATA_MSG; + hdr->dst_ch = htons(ch->rem_channel); + hdr->src_ch = htons(ch->id); + hdr->msg_len = htons((u16)len); + + /* ATTN: the function call below relies on the fact that underlying + * HW-specific add_outb_message() routine copies TX data into its own + * internal transfer buffer (true for all RIONET compatible mport + * drivers). Must be reviewed if mport driver uses the buffer directly. + */ + + ret = riocm_post_send(ch->cmdev, ch->rdev, buf, len); + if (ret) + riocm_debug(TX, "ch %d send_err=%d", ch->id, ret); +err_out: + riocm_put_channel(ch); + return ret; +} + +static int riocm_ch_free_rxbuf(struct rio_channel *ch, void *buf) +{ + int i, ret = -EINVAL; + + spin_lock_bh(&ch->lock); + + for (i = 0; i < RIOCM_RX_RING_SIZE; i++) { + if (ch->rx_ring.inuse[i] == buf) { + ch->rx_ring.inuse[i] = NULL; + ch->rx_ring.inuse_cnt--; + ret = 0; + break; + } + } + + spin_unlock_bh(&ch->lock); + + if (!ret) + kfree(buf); + + return ret; +} + +/* + * riocm_ch_receive - fetch a data packet received for the specified channel + * @ch: local channel ID + * @buf: pointer to a packet buffer + * @timeout: timeout to wait for incoming packet (in jiffies) + * + * Returns: 0 and valid buffer pointer if success, or NULL pointer and one of: + * -EAGAIN if a channel is not in CONNECTED state, + * -ENOMEM if in-use tracking queue is full, + * -ETIME if wait timeout expired, + * -EINTR if wait was interrupted. + */ +static int riocm_ch_receive(struct rio_channel *ch, void **buf, long timeout) +{ + void *rxmsg = NULL; + int i, ret = 0; + long wret; + + if (!riocm_cmp(ch, RIO_CM_CONNECTED)) { + ret = -EAGAIN; + goto out; + } + + if (ch->rx_ring.inuse_cnt == RIOCM_RX_RING_SIZE) { + /* If we do not have entries to track buffers given to upper + * layer, reject request. + */ + ret = -ENOMEM; + goto out; + } + + wret = wait_for_completion_interruptible_timeout(&ch->comp, timeout); + + riocm_debug(WAIT, "wait on %d returned %ld", ch->id, wret); + + if (!wret) + ret = -ETIME; + else if (wret == -ERESTARTSYS) + ret = -EINTR; + else + ret = riocm_cmp(ch, RIO_CM_CONNECTED) ? 0 : -ECONNRESET; + + if (ret) + goto out; + + spin_lock_bh(&ch->lock); + + rxmsg = ch->rx_ring.buf[ch->rx_ring.tail]; + ch->rx_ring.buf[ch->rx_ring.tail] = NULL; + ch->rx_ring.count--; + ch->rx_ring.tail++; + ch->rx_ring.tail %= RIOCM_RX_RING_SIZE; + ret = -ENOMEM; + + for (i = 0; i < RIOCM_RX_RING_SIZE; i++) { + if (ch->rx_ring.inuse[i] == NULL) { + ch->rx_ring.inuse[i] = rxmsg; + ch->rx_ring.inuse_cnt++; + ret = 0; + break; + } + } + + if (ret) { + /* We have no entry to store pending message: drop it */ + kfree(rxmsg); + rxmsg = NULL; + } + + spin_unlock_bh(&ch->lock); +out: + *buf = rxmsg; + return ret; +} + +/* + * riocm_ch_connect - sends a connect request to a remote device + * @loc_ch: local channel ID + * @cm: CM device to send connect request + * @peer: target RapidIO device + * @rem_ch: remote channel ID + * + * Returns: 0 if success, or + * -EINVAL if the channel is not in IDLE state, + * -EAGAIN if no connection request available immediately, + * -ETIME if ACK response timeout expired, + * -EINTR if wait for response was interrupted. + */ +static int riocm_ch_connect(u16 loc_ch, struct cm_dev *cm, + struct cm_peer *peer, u16 rem_ch) +{ + struct rio_channel *ch = NULL; + struct rio_ch_chan_hdr *hdr; + int ret; + long wret; + + ch = riocm_get_channel(loc_ch); + if (!ch) + return -ENODEV; + + if (!riocm_cmp_exch(ch, RIO_CM_IDLE, RIO_CM_CONNECT)) { + ret = -EINVAL; + goto conn_done; + } + + ch->cmdev = cm; + ch->rdev = peer->rdev; + ch->context = NULL; + ch->loc_destid = cm->mport->host_deviceid; + ch->rem_channel = rem_ch; + + /* + * Send connect request to the remote RapidIO device + */ + + hdr = kzalloc(sizeof(*hdr), GFP_KERNEL); + if (hdr == NULL) { + ret = -ENOMEM; + goto conn_done; + } + + hdr->bhdr.src_id = htonl(ch->loc_destid); + hdr->bhdr.dst_id = htonl(peer->rdev->destid); + hdr->bhdr.src_mbox = cmbox; + hdr->bhdr.dst_mbox = cmbox; + hdr->bhdr.type = RIO_CM_CHAN; + hdr->ch_op = CM_CONN_REQ; + hdr->dst_ch = htons(rem_ch); + hdr->src_ch = htons(loc_ch); + + /* ATTN: the function call below relies on the fact that underlying + * HW-specific add_outb_message() routine copies TX data into its + * internal transfer buffer. Must be reviewed if mport driver uses + * this buffer directly. + */ + ret = riocm_post_send(cm, peer->rdev, hdr, sizeof(*hdr)); + + if (ret != -EBUSY) { + kfree(hdr); + } else { + ret = riocm_queue_req(cm, peer->rdev, hdr, sizeof(*hdr)); + if (ret) + kfree(hdr); + } + + if (ret) { + riocm_cmp_exch(ch, RIO_CM_CONNECT, RIO_CM_IDLE); + goto conn_done; + } + + /* Wait for connect response from the remote device */ + wret = wait_for_completion_interruptible_timeout(&ch->comp, + RIOCM_CONNECT_TO * HZ); + riocm_debug(WAIT, "wait on %d returns %ld", ch->id, wret); + + if (!wret) + ret = -ETIME; + else if (wret == -ERESTARTSYS) + ret = -EINTR; + else + ret = riocm_cmp(ch, RIO_CM_CONNECTED) ? 0 : -1; + +conn_done: + riocm_put_channel(ch); + return ret; +} + +static int riocm_send_ack(struct rio_channel *ch) +{ + struct rio_ch_chan_hdr *hdr; + int ret; + + hdr = kzalloc(sizeof(*hdr), GFP_KERNEL); + if (hdr == NULL) + return -ENOMEM; + + hdr->bhdr.src_id = htonl(ch->loc_destid); + hdr->bhdr.dst_id = htonl(ch->rem_destid); + hdr->dst_ch = htons(ch->rem_channel); + hdr->src_ch = htons(ch->id); + hdr->bhdr.src_mbox = cmbox; + hdr->bhdr.dst_mbox = cmbox; + hdr->bhdr.type = RIO_CM_CHAN; + hdr->ch_op = CM_CONN_ACK; + + /* ATTN: the function call below relies on the fact that underlying + * add_outb_message() routine copies TX data into its internal transfer + * buffer. Review if switching to direct buffer version. + */ + ret = riocm_post_send(ch->cmdev, ch->rdev, hdr, sizeof(*hdr)); + + if (ret == -EBUSY && !riocm_queue_req(ch->cmdev, + ch->rdev, hdr, sizeof(*hdr))) + return 0; + kfree(hdr); + + if (ret) + riocm_error("send ACK to ch_%d on %s failed (ret=%d)", + ch->id, rio_name(ch->rdev), ret); + return ret; +} + +/* + * riocm_ch_accept - accept incoming connection request + * @ch_id: channel ID + * @new_ch_id: local mport device + * @timeout: wait timeout (if 0 non-blocking call, do not wait if connection + * request is not available). + * + * Returns: pointer to new channel struct if success, or error-valued pointer: + * -ENODEV - cannot find specified channel or mport, + * -EINVAL - the channel is not in IDLE state, + * -EAGAIN - no connection request available immediately (timeout=0), + * -ENOMEM - unable to allocate new channel, + * -ETIME - wait timeout expired, + * -EINTR - wait was interrupted. + */ +static struct rio_channel *riocm_ch_accept(u16 ch_id, u16 *new_ch_id, + long timeout) +{ + struct rio_channel *ch = NULL; + struct rio_channel *new_ch = NULL; + struct conn_req *req; + struct cm_peer *peer; + int found = 0; + int err = 0; + long wret; + + ch = riocm_get_channel(ch_id); + if (!ch) + return ERR_PTR(-EINVAL); + + if (!riocm_cmp(ch, RIO_CM_LISTEN)) { + err = -EINVAL; + goto err_put; + } + + /* Don't sleep if this is a non blocking call */ + if (!timeout) { + if (!try_wait_for_completion(&ch->comp)) { + err = -EAGAIN; + goto err_put; + } + } else { + riocm_debug(WAIT, "on %d", ch->id); + + wret = wait_for_completion_interruptible_timeout(&ch->comp, + timeout); + if (!wret) { + err = -ETIME; + goto err_put; + } else if (wret == -ERESTARTSYS) { + err = -EINTR; + goto err_put; + } + } + + spin_lock_bh(&ch->lock); + + if (ch->state != RIO_CM_LISTEN) { + err = -ECANCELED; + } else if (list_empty(&ch->accept_queue)) { + riocm_debug(WAIT, "on %d accept_queue is empty on completion", + ch->id); + err = -EIO; + } + + spin_unlock_bh(&ch->lock); + + if (err) { + riocm_debug(WAIT, "on %d returns %d", ch->id, err); + goto err_put; + } + + /* Create new channel for this connection */ + new_ch = riocm_ch_alloc(RIOCM_CHNUM_AUTO); + + if (IS_ERR(new_ch)) { + riocm_error("failed to get channel for new req (%ld)", + PTR_ERR(new_ch)); + err = -ENOMEM; + goto err_put; + } + + spin_lock_bh(&ch->lock); + + req = list_first_entry(&ch->accept_queue, struct conn_req, node); + list_del(&req->node); + new_ch->cmdev = ch->cmdev; + new_ch->loc_destid = ch->loc_destid; + new_ch->rem_destid = req->destid; + new_ch->rem_channel = req->chan; + + spin_unlock_bh(&ch->lock); + riocm_put_channel(ch); + kfree(req); + + down_read(&rdev_sem); + /* Find requester's device object */ + list_for_each_entry(peer, &new_ch->cmdev->peers, node) { + if (peer->rdev->destid == new_ch->rem_destid) { + riocm_debug(RX_CMD, "found matching device(%s)", + rio_name(peer->rdev)); + found = 1; + break; + } + } + up_read(&rdev_sem); + + if (!found) { + /* If peer device object not found, simply ignore the request */ + err = -ENODEV; + goto err_nodev; + } + + new_ch->rdev = peer->rdev; + new_ch->state = RIO_CM_CONNECTED; + spin_lock_init(&new_ch->lock); + + /* Acknowledge the connection request. */ + riocm_send_ack(new_ch); + + *new_ch_id = new_ch->id; + return new_ch; +err_put: + riocm_put_channel(ch); +err_nodev: + if (new_ch) { + spin_lock_bh(&idr_lock); + idr_remove(&ch_idr, new_ch->id); + spin_unlock_bh(&idr_lock); + riocm_put_channel(new_ch); + } + *new_ch_id = 0; + return ERR_PTR(err); +} + +/* + * riocm_ch_listen - puts a channel into LISTEN state + * @ch_id: channel ID + * + * Returns: 0 if success, or + * -EINVAL if the specified channel does not exists or + * is not in CHAN_BOUND state. + */ +static int riocm_ch_listen(u16 ch_id) +{ + struct rio_channel *ch = NULL; + int ret = 0; + + riocm_debug(CHOP, "(ch_%d)", ch_id); + + ch = riocm_get_channel(ch_id); + if (!ch || !riocm_cmp_exch(ch, RIO_CM_CHAN_BOUND, RIO_CM_LISTEN)) + ret = -EINVAL; + riocm_put_channel(ch); + return ret; +} + +/* + * riocm_ch_bind - associate a channel object and an mport device + * @ch_id: channel ID + * @mport_id: local mport device ID + * @context: pointer to the additional caller's context + * + * Returns: 0 if success, or + * -ENODEV if cannot find specified mport, + * -EINVAL if the specified channel does not exist or + * is not in IDLE state. + */ +static int riocm_ch_bind(u16 ch_id, u8 mport_id, void *context) +{ + struct rio_channel *ch = NULL; + struct cm_dev *cm; + int rc = -ENODEV; + + riocm_debug(CHOP, "ch_%d to mport_%d", ch_id, mport_id); + + /* Find matching cm_dev object */ + down_read(&rdev_sem); + list_for_each_entry(cm, &cm_dev_list, list) { + if ((cm->mport->id == mport_id) && + rio_mport_is_running(cm->mport)) { + rc = 0; + break; + } + } + + if (rc) + goto exit; + + ch = riocm_get_channel(ch_id); + if (!ch) { + rc = -EINVAL; + goto exit; + } + + spin_lock_bh(&ch->lock); + if (ch->state != RIO_CM_IDLE) { + spin_unlock_bh(&ch->lock); + rc = -EINVAL; + goto err_put; + } + + ch->cmdev = cm; + ch->loc_destid = cm->mport->host_deviceid; + ch->context = context; + ch->state = RIO_CM_CHAN_BOUND; + spin_unlock_bh(&ch->lock); +err_put: + riocm_put_channel(ch); +exit: + up_read(&rdev_sem); + return rc; +} + +/* + * riocm_ch_alloc - channel object allocation helper routine + * @ch_num: channel ID (1 ... RIOCM_MAX_CHNUM, 0 = automatic) + * + * Return value: pointer to newly created channel object, + * or error-valued pointer + */ +static struct rio_channel *riocm_ch_alloc(u16 ch_num) +{ + int id; + int start, end; + struct rio_channel *ch; + + ch = kzalloc(sizeof(*ch), GFP_KERNEL); + if (!ch) + return ERR_PTR(-ENOMEM); + + if (ch_num) { + /* If requested, try to obtain the specified channel ID */ + start = ch_num; + end = ch_num + 1; + } else { + /* Obtain channel ID from the dynamic allocation range */ + start = chstart; + end = RIOCM_MAX_CHNUM + 1; + } + + idr_preload(GFP_KERNEL); + spin_lock_bh(&idr_lock); + id = idr_alloc_cyclic(&ch_idr, ch, start, end, GFP_NOWAIT); + spin_unlock_bh(&idr_lock); + idr_preload_end(); + + if (id < 0) { + kfree(ch); + return ERR_PTR(id == -ENOSPC ? -EBUSY : id); + } + + ch->id = (u16)id; + ch->state = RIO_CM_IDLE; + spin_lock_init(&ch->lock); + INIT_LIST_HEAD(&ch->accept_queue); + INIT_LIST_HEAD(&ch->ch_node); + init_completion(&ch->comp); + init_completion(&ch->comp_close); + kref_init(&ch->ref); + ch->rx_ring.head = 0; + ch->rx_ring.tail = 0; + ch->rx_ring.count = 0; + ch->rx_ring.inuse_cnt = 0; + + return ch; +} + +/* + * riocm_ch_create - creates a new channel object and allocates ID for it + * @ch_num: channel ID (1 ... RIOCM_MAX_CHNUM, 0 = automatic) + * + * Allocates and initializes a new channel object. If the parameter ch_num > 0 + * and is within the valid range, riocm_ch_create tries to allocate the + * specified ID for the new channel. If ch_num = 0, channel ID will be assigned + * automatically from the range (chstart ... RIOCM_MAX_CHNUM). + * Module parameter 'chstart' defines start of an ID range available for dynamic + * allocation. Range below 'chstart' is reserved for pre-defined ID numbers. + * Available channel numbers are limited by 16-bit size of channel numbers used + * in the packet header. + * + * Return value: PTR to rio_channel structure if successful (with channel number + * updated via pointer) or error-valued pointer if error. + */ +static struct rio_channel *riocm_ch_create(u16 *ch_num) +{ + struct rio_channel *ch = NULL; + + ch = riocm_ch_alloc(*ch_num); + + if (IS_ERR(ch)) + riocm_debug(CHOP, "Failed to allocate channel %d (err=%ld)", + *ch_num, PTR_ERR(ch)); + else + *ch_num = ch->id; + + return ch; +} + +/* + * riocm_ch_free - channel object release routine + * @ref: pointer to a channel's kref structure + */ +static void riocm_ch_free(struct kref *ref) +{ + struct rio_channel *ch = container_of(ref, struct rio_channel, ref); + int i; + + riocm_debug(CHOP, "(ch_%d)", ch->id); + + if (ch->rx_ring.inuse_cnt) { + for (i = 0; + i < RIOCM_RX_RING_SIZE && ch->rx_ring.inuse_cnt; i++) { + if (ch->rx_ring.inuse[i] != NULL) { + kfree(ch->rx_ring.inuse[i]); + ch->rx_ring.inuse_cnt--; + } + } + } + + if (ch->rx_ring.count) + for (i = 0; i < RIOCM_RX_RING_SIZE && ch->rx_ring.count; i++) { + if (ch->rx_ring.buf[i] != NULL) { + kfree(ch->rx_ring.buf[i]); + ch->rx_ring.count--; + } + } + + complete(&ch->comp_close); +} + +static int riocm_send_close(struct rio_channel *ch) +{ + struct rio_ch_chan_hdr *hdr; + int ret; + + /* + * Send CH_CLOSE notification to the remote RapidIO device + */ + + hdr = kzalloc(sizeof(*hdr), GFP_KERNEL); + if (hdr == NULL) + return -ENOMEM; + + hdr->bhdr.src_id = htonl(ch->loc_destid); + hdr->bhdr.dst_id = htonl(ch->rem_destid); + hdr->bhdr.src_mbox = cmbox; + hdr->bhdr.dst_mbox = cmbox; + hdr->bhdr.type = RIO_CM_CHAN; + hdr->ch_op = CM_CONN_CLOSE; + hdr->dst_ch = htons(ch->rem_channel); + hdr->src_ch = htons(ch->id); + + /* ATTN: the function call below relies on the fact that underlying + * add_outb_message() routine copies TX data into its internal transfer + * buffer. Needs to be reviewed if switched to direct buffer mode. + */ + ret = riocm_post_send(ch->cmdev, ch->rdev, hdr, sizeof(*hdr)); + + if (ret == -EBUSY && !riocm_queue_req(ch->cmdev, ch->rdev, + hdr, sizeof(*hdr))) + return 0; + kfree(hdr); + + if (ret) + riocm_error("ch(%d) send CLOSE failed (ret=%d)", ch->id, ret); + + return ret; +} + +/* + * riocm_ch_close - closes a channel object with specified ID (by local request) + * @ch: channel to be closed + */ +static int riocm_ch_close(struct rio_channel *ch) +{ + unsigned long tmo = msecs_to_jiffies(3000); + enum rio_cm_state state; + long wret; + int ret = 0; + + riocm_debug(CHOP, "ch_%d by %s(%d)", + ch->id, current->comm, task_pid_nr(current)); + + state = riocm_exch(ch, RIO_CM_DESTROYING); + if (state == RIO_CM_CONNECTED) + riocm_send_close(ch); + + complete_all(&ch->comp); + + riocm_put_channel(ch); + wret = wait_for_completion_interruptible_timeout(&ch->comp_close, tmo); + + riocm_debug(WAIT, "wait on %d returns %ld", ch->id, wret); + + if (wret == 0) { + /* Timeout on wait occurred */ + riocm_debug(CHOP, "%s(%d) timed out waiting for ch %d", + current->comm, task_pid_nr(current), ch->id); + ret = -ETIMEDOUT; + } else if (wret == -ERESTARTSYS) { + /* Wait_for_completion was interrupted by a signal */ + riocm_debug(CHOP, "%s(%d) wait for ch %d was interrupted", + current->comm, task_pid_nr(current), ch->id); + ret = -EINTR; + } + + if (!ret) { + riocm_debug(CHOP, "ch_%d resources released", ch->id); + kfree(ch); + } else { + riocm_debug(CHOP, "failed to release ch_%d resources", ch->id); + } + + return ret; +} + +/* + * riocm_cdev_open() - Open character device + */ +static int riocm_cdev_open(struct inode *inode, struct file *filp) +{ + riocm_debug(INIT, "by %s(%d) filp=%p ", + current->comm, task_pid_nr(current), filp); + + if (list_empty(&cm_dev_list)) + return -ENODEV; + + return 0; +} + +/* + * riocm_cdev_release() - Release character device + */ +static int riocm_cdev_release(struct inode *inode, struct file *filp) +{ + struct rio_channel *ch, *_c; + unsigned int i; + LIST_HEAD(list); + + riocm_debug(EXIT, "by %s(%d) filp=%p", + current->comm, task_pid_nr(current), filp); + + /* Check if there are channels associated with this file descriptor */ + spin_lock_bh(&idr_lock); + idr_for_each_entry(&ch_idr, ch, i) { + if (ch && ch->filp == filp) { + riocm_debug(EXIT, "ch_%d not released by %s(%d)", + ch->id, current->comm, + task_pid_nr(current)); + idr_remove(&ch_idr, ch->id); + list_add(&ch->ch_node, &list); + } + } + spin_unlock_bh(&idr_lock); + + if (!list_empty(&list)) { + list_for_each_entry_safe(ch, _c, &list, ch_node) { + list_del(&ch->ch_node); + riocm_ch_close(ch); + } + } + + return 0; +} + +/* + * cm_ep_get_list_size() - Reports number of endpoints in the network + */ +static int cm_ep_get_list_size(void __user *arg) +{ + u32 __user *p = arg; + u32 mport_id; + u32 count = 0; + struct cm_dev *cm; + + if (get_user(mport_id, p)) + return -EFAULT; + if (mport_id >= RIO_MAX_MPORTS) + return -EINVAL; + + /* Find a matching cm_dev object */ + down_read(&rdev_sem); + list_for_each_entry(cm, &cm_dev_list, list) { + if (cm->mport->id == mport_id) { + count = cm->npeers; + up_read(&rdev_sem); + if (copy_to_user(arg, &count, sizeof(u32))) + return -EFAULT; + return 0; + } + } + up_read(&rdev_sem); + + return -ENODEV; +} + +/* + * cm_ep_get_list() - Returns list of attached endpoints + */ +static int cm_ep_get_list(void __user *arg) +{ + struct cm_dev *cm; + struct cm_peer *peer; + u32 info[2]; + void *buf; + u32 nent; + u32 *entry_ptr; + u32 i = 0; + int ret = 0; + + if (copy_from_user(&info, arg, sizeof(info))) + return -EFAULT; + + if (info[1] >= RIO_MAX_MPORTS || info[0] > RIOCM_MAX_EP_COUNT) + return -EINVAL; + + /* Find a matching cm_dev object */ + down_read(&rdev_sem); + list_for_each_entry(cm, &cm_dev_list, list) + if (cm->mport->id == (u8)info[1]) + goto found; + + up_read(&rdev_sem); + return -ENODEV; + +found: + nent = min(info[0], cm->npeers); + buf = kcalloc(nent + 2, sizeof(u32), GFP_KERNEL); + if (!buf) { + up_read(&rdev_sem); + return -ENOMEM; + } + + entry_ptr = (u32 *)((uintptr_t)buf + 2*sizeof(u32)); + + list_for_each_entry(peer, &cm->peers, node) { + *entry_ptr = (u32)peer->rdev->destid; + entry_ptr++; + if (++i == nent) + break; + } + up_read(&rdev_sem); + + ((u32 *)buf)[0] = i; /* report an updated number of entries */ + ((u32 *)buf)[1] = info[1]; /* put back an mport ID */ + if (copy_to_user(arg, buf, sizeof(u32) * (info[0] + 2))) + ret = -EFAULT; + + kfree(buf); + return ret; +} + +/* + * cm_mport_get_list() - Returns list of available local mport devices + */ +static int cm_mport_get_list(void __user *arg) +{ + int ret = 0; + u32 entries; + void *buf; + struct cm_dev *cm; + u32 *entry_ptr; + int count = 0; + + if (copy_from_user(&entries, arg, sizeof(entries))) + return -EFAULT; + if (entries == 0 || entries > RIO_MAX_MPORTS) + return -EINVAL; + buf = kcalloc(entries + 1, sizeof(u32), GFP_KERNEL); + if (!buf) + return -ENOMEM; + + /* Scan all registered cm_dev objects */ + entry_ptr = (u32 *)((uintptr_t)buf + sizeof(u32)); + down_read(&rdev_sem); + list_for_each_entry(cm, &cm_dev_list, list) { + if (count++ < entries) { + *entry_ptr = (cm->mport->id << 16) | + cm->mport->host_deviceid; + entry_ptr++; + } + } + up_read(&rdev_sem); + + *((u32 *)buf) = count; /* report a real number of entries */ + if (copy_to_user(arg, buf, sizeof(u32) * (count + 1))) + ret = -EFAULT; + + kfree(buf); + return ret; +} + +/* + * cm_chan_create() - Create a message exchange channel + */ +static int cm_chan_create(struct file *filp, void __user *arg) +{ + u16 __user *p = arg; + u16 ch_num; + struct rio_channel *ch; + + if (get_user(ch_num, p)) + return -EFAULT; + + riocm_debug(CHOP, "ch_%d requested by %s(%d)", + ch_num, current->comm, task_pid_nr(current)); + ch = riocm_ch_create(&ch_num); + if (IS_ERR(ch)) + return PTR_ERR(ch); + + ch->filp = filp; + riocm_debug(CHOP, "ch_%d created by %s(%d)", + ch_num, current->comm, task_pid_nr(current)); + return put_user(ch_num, p); +} + +/* + * cm_chan_close() - Close channel + * @filp: Pointer to file object + * @arg: Channel to close + */ +static int cm_chan_close(struct file *filp, void __user *arg) +{ + u16 __user *p = arg; + u16 ch_num; + struct rio_channel *ch; + + if (get_user(ch_num, p)) + return -EFAULT; + + riocm_debug(CHOP, "ch_%d by %s(%d)", + ch_num, current->comm, task_pid_nr(current)); + + spin_lock_bh(&idr_lock); + ch = idr_find(&ch_idr, ch_num); + if (!ch) { + spin_unlock_bh(&idr_lock); + return 0; + } + if (ch->filp != filp) { + spin_unlock_bh(&idr_lock); + return -EINVAL; + } + idr_remove(&ch_idr, ch->id); + spin_unlock_bh(&idr_lock); + + return riocm_ch_close(ch); +} + +/* + * cm_chan_bind() - Bind channel + * @arg: Channel number + */ +static int cm_chan_bind(void __user *arg) +{ + struct rio_cm_channel chan; + + if (copy_from_user(&chan, arg, sizeof(chan))) + return -EFAULT; + if (chan.mport_id >= RIO_MAX_MPORTS) + return -EINVAL; + + return riocm_ch_bind(chan.id, chan.mport_id, NULL); +} + +/* + * cm_chan_listen() - Listen on channel + * @arg: Channel number + */ +static int cm_chan_listen(void __user *arg) +{ + u16 __user *p = arg; + u16 ch_num; + + if (get_user(ch_num, p)) + return -EFAULT; + + return riocm_ch_listen(ch_num); +} + +/* + * cm_chan_accept() - Accept incoming connection + * @filp: Pointer to file object + * @arg: Channel number + */ +static int cm_chan_accept(struct file *filp, void __user *arg) +{ + struct rio_cm_accept param; + long accept_to; + struct rio_channel *ch; + + if (copy_from_user(¶m, arg, sizeof(param))) + return -EFAULT; + + riocm_debug(CHOP, "on ch_%d by %s(%d)", + param.ch_num, current->comm, task_pid_nr(current)); + + accept_to = param.wait_to ? + msecs_to_jiffies(param.wait_to) : 0; + + ch = riocm_ch_accept(param.ch_num, ¶m.ch_num, accept_to); + if (IS_ERR(ch)) + return PTR_ERR(ch); + ch->filp = filp; + + riocm_debug(CHOP, "new ch_%d for %s(%d)", + ch->id, current->comm, task_pid_nr(current)); + + if (copy_to_user(arg, ¶m, sizeof(param))) + return -EFAULT; + return 0; +} + +/* + * cm_chan_connect() - Connect on channel + * @arg: Channel information + */ +static int cm_chan_connect(void __user *arg) +{ + struct rio_cm_channel chan; + struct cm_dev *cm; + struct cm_peer *peer; + int ret = -ENODEV; + + if (copy_from_user(&chan, arg, sizeof(chan))) + return -EFAULT; + if (chan.mport_id >= RIO_MAX_MPORTS) + return -EINVAL; + + down_read(&rdev_sem); + + /* Find matching cm_dev object */ + list_for_each_entry(cm, &cm_dev_list, list) { + if (cm->mport->id == chan.mport_id) { + ret = 0; + break; + } + } + + if (ret) + goto err_out; + + if (chan.remote_destid >= RIO_ANY_DESTID(cm->mport->sys_size)) { + ret = -EINVAL; + goto err_out; + } + + /* Find corresponding RapidIO endpoint device object */ + ret = -ENODEV; + + list_for_each_entry(peer, &cm->peers, node) { + if (peer->rdev->destid == chan.remote_destid) { + ret = 0; + break; + } + } + + if (ret) + goto err_out; + + up_read(&rdev_sem); + + return riocm_ch_connect(chan.id, cm, peer, chan.remote_channel); +err_out: + up_read(&rdev_sem); + return ret; +} + +/* + * cm_chan_msg_send() - Send a message through channel + * @arg: Outbound message information + */ +static int cm_chan_msg_send(void __user *arg) +{ + struct rio_cm_msg msg; + void *buf; + int ret = 0; + + if (copy_from_user(&msg, arg, sizeof(msg))) + return -EFAULT; + if (msg.size > RIO_MAX_MSG_SIZE) + return -EINVAL; + + buf = kmalloc(msg.size, GFP_KERNEL); + if (!buf) + return -ENOMEM; + + if (copy_from_user(buf, (void __user *)(uintptr_t)msg.msg, msg.size)) { + ret = -EFAULT; + goto out; + } + + ret = riocm_ch_send(msg.ch_num, buf, msg.size); +out: + kfree(buf); + return ret; +} + +/* + * cm_chan_msg_rcv() - Receive a message through channel + * @arg: Inbound message information + */ +static int cm_chan_msg_rcv(void __user *arg) +{ + struct rio_cm_msg msg; + struct rio_channel *ch; + void *buf; + long rxto; + int ret = 0, msg_size; + + if (copy_from_user(&msg, arg, sizeof(msg))) + return -EFAULT; + + if (msg.ch_num == 0 || msg.size == 0) + return -EINVAL; + + ch = riocm_get_channel(msg.ch_num); + if (!ch) + return -ENODEV; + + rxto = msg.rxto ? msecs_to_jiffies(msg.rxto) : MAX_SCHEDULE_TIMEOUT; + + ret = riocm_ch_receive(ch, &buf, rxto); + if (ret) + goto out; + + msg_size = min(msg.size, (u16)(RIO_MAX_MSG_SIZE)); + + if (copy_to_user((void __user *)(uintptr_t)msg.msg, buf, msg_size)) + ret = -EFAULT; + + riocm_ch_free_rxbuf(ch, buf); +out: + riocm_put_channel(ch); + return ret; +} + +/* + * riocm_cdev_ioctl() - IOCTL requests handler + */ +static long +riocm_cdev_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) +{ + switch (cmd) { + case RIO_CM_EP_GET_LIST_SIZE: + return cm_ep_get_list_size((void __user *)arg); + case RIO_CM_EP_GET_LIST: + return cm_ep_get_list((void __user *)arg); + case RIO_CM_CHAN_CREATE: + return cm_chan_create(filp, (void __user *)arg); + case RIO_CM_CHAN_CLOSE: + return cm_chan_close(filp, (void __user *)arg); + case RIO_CM_CHAN_BIND: + return cm_chan_bind((void __user *)arg); + case RIO_CM_CHAN_LISTEN: + return cm_chan_listen((void __user *)arg); + case RIO_CM_CHAN_ACCEPT: + return cm_chan_accept(filp, (void __user *)arg); + case RIO_CM_CHAN_CONNECT: + return cm_chan_connect((void __user *)arg); + case RIO_CM_CHAN_SEND: + return cm_chan_msg_send((void __user *)arg); + case RIO_CM_CHAN_RECEIVE: + return cm_chan_msg_rcv((void __user *)arg); + case RIO_CM_MPORT_GET_LIST: + return cm_mport_get_list((void __user *)arg); + default: + break; + } + + return -EINVAL; +} + +static const struct file_operations riocm_cdev_fops = { + .owner = THIS_MODULE, + .open = riocm_cdev_open, + .release = riocm_cdev_release, + .unlocked_ioctl = riocm_cdev_ioctl, +}; + +/* + * riocm_add_dev - add new remote RapidIO device into channel management core + * @dev: device object associated with RapidIO device + * @sif: subsystem interface + * + * Adds the specified RapidIO device (if applicable) into peers list of + * the corresponding channel management device (cm_dev). + */ +static int riocm_add_dev(struct device *dev, struct subsys_interface *sif) +{ + struct cm_peer *peer; + struct rio_dev *rdev = to_rio_dev(dev); + struct cm_dev *cm; + + /* Check if the remote device has capabilities required to support CM */ + if (!dev_cm_capable(rdev)) + return 0; + + riocm_debug(RDEV, "(%s)", rio_name(rdev)); + + peer = kmalloc(sizeof(*peer), GFP_KERNEL); + if (!peer) + return -ENOMEM; + + /* Find a corresponding cm_dev object */ + down_write(&rdev_sem); + list_for_each_entry(cm, &cm_dev_list, list) { + if (cm->mport == rdev->net->hport) + goto found; + } + + up_write(&rdev_sem); + kfree(peer); + return -ENODEV; + +found: + peer->rdev = rdev; + list_add_tail(&peer->node, &cm->peers); + cm->npeers++; + + up_write(&rdev_sem); + return 0; +} + +/* + * riocm_remove_dev - remove remote RapidIO device from channel management core + * @dev: device object associated with RapidIO device + * @sif: subsystem interface + * + * Removes the specified RapidIO device (if applicable) from peers list of + * the corresponding channel management device (cm_dev). + */ +static void riocm_remove_dev(struct device *dev, struct subsys_interface *sif) +{ + struct rio_dev *rdev = to_rio_dev(dev); + struct cm_dev *cm; + struct cm_peer *peer; + struct rio_channel *ch, *_c; + unsigned int i; + bool found = false; + LIST_HEAD(list); + + /* Check if the remote device has capabilities required to support CM */ + if (!dev_cm_capable(rdev)) + return; + + riocm_debug(RDEV, "(%s)", rio_name(rdev)); + + /* Find matching cm_dev object */ + down_write(&rdev_sem); + list_for_each_entry(cm, &cm_dev_list, list) { + if (cm->mport == rdev->net->hport) { + found = true; + break; + } + } + + if (!found) { + up_write(&rdev_sem); + return; + } + + /* Remove remote device from the list of peers */ + found = false; + list_for_each_entry(peer, &cm->peers, node) { + if (peer->rdev == rdev) { + riocm_debug(RDEV, "removing peer %s", rio_name(rdev)); + found = true; + list_del(&peer->node); + cm->npeers--; + kfree(peer); + break; + } + } + + up_write(&rdev_sem); + + if (!found) + return; + + /* + * Release channels associated with this peer + */ + + spin_lock_bh(&idr_lock); + idr_for_each_entry(&ch_idr, ch, i) { + if (ch && ch->rdev == rdev) { + if (atomic_read(&rdev->state) != RIO_DEVICE_SHUTDOWN) + riocm_exch(ch, RIO_CM_DISCONNECT); + idr_remove(&ch_idr, ch->id); + list_add(&ch->ch_node, &list); + } + } + spin_unlock_bh(&idr_lock); + + if (!list_empty(&list)) { + list_for_each_entry_safe(ch, _c, &list, ch_node) { + list_del(&ch->ch_node); + riocm_ch_close(ch); + } + } +} + +/* + * riocm_cdev_add() - Create rio_cm char device + * @devno: device number assigned to device (MAJ + MIN) + */ +static int riocm_cdev_add(dev_t devno) +{ + int ret; + + cdev_init(&riocm_cdev.cdev, &riocm_cdev_fops); + riocm_cdev.cdev.owner = THIS_MODULE; + ret = cdev_add(&riocm_cdev.cdev, devno, 1); + if (ret < 0) { + riocm_error("Cannot register a device with error %d", ret); + return ret; + } + + riocm_cdev.dev = device_create(dev_class, NULL, devno, NULL, DEV_NAME); + if (IS_ERR(riocm_cdev.dev)) { + cdev_del(&riocm_cdev.cdev); + return PTR_ERR(riocm_cdev.dev); + } + + riocm_debug(MPORT, "Added %s cdev(%d:%d)", + DEV_NAME, MAJOR(devno), MINOR(devno)); + + return 0; +} + +/* + * riocm_add_mport - add new local mport device into channel management core + * @dev: device object associated with mport + * @class_intf: class interface + * + * When a new mport device is added, CM immediately reserves inbound and + * outbound RapidIO mailboxes that will be used. + */ +static int riocm_add_mport(struct device *dev, + struct class_interface *class_intf) +{ + int rc; + int i; + struct cm_dev *cm; + struct rio_mport *mport = to_rio_mport(dev); + + riocm_debug(MPORT, "add mport %s", mport->name); + + cm = kzalloc(sizeof(*cm), GFP_KERNEL); + if (!cm) + return -ENOMEM; + + cm->mport = mport; + + rc = rio_request_outb_mbox(mport, cm, cmbox, + RIOCM_TX_RING_SIZE, riocm_outb_msg_event); + if (rc) { + riocm_error("failed to allocate OBMBOX_%d on %s", + cmbox, mport->name); + kfree(cm); + return -ENODEV; + } + + rc = rio_request_inb_mbox(mport, cm, cmbox, + RIOCM_RX_RING_SIZE, riocm_inb_msg_event); + if (rc) { + riocm_error("failed to allocate IBMBOX_%d on %s", + cmbox, mport->name); + rio_release_outb_mbox(mport, cmbox); + kfree(cm); + return -ENODEV; + } + + /* + * Allocate and register inbound messaging buffers to be ready + * to receive channel and system management requests + */ + for (i = 0; i < RIOCM_RX_RING_SIZE; i++) + cm->rx_buf[i] = NULL; + + cm->rx_slots = RIOCM_RX_RING_SIZE; + mutex_init(&cm->rx_lock); + riocm_rx_fill(cm, RIOCM_RX_RING_SIZE); + cm->rx_wq = create_workqueue(DRV_NAME "/rxq"); + INIT_WORK(&cm->rx_work, rio_ibmsg_handler); + + cm->tx_slot = 0; + cm->tx_cnt = 0; + cm->tx_ack_slot = 0; + spin_lock_init(&cm->tx_lock); + + INIT_LIST_HEAD(&cm->peers); + cm->npeers = 0; + INIT_LIST_HEAD(&cm->tx_reqs); + + down_write(&rdev_sem); + list_add_tail(&cm->list, &cm_dev_list); + up_write(&rdev_sem); + + return 0; +} + +/* + * riocm_remove_mport - remove local mport device from channel management core + * @dev: device object associated with mport + * @class_intf: class interface + * + * Removes a local mport device from the list of registered devices that provide + * channel management services. Returns an error if the specified mport is not + * registered with the CM core. + */ +static void riocm_remove_mport(struct device *dev, + struct class_interface *class_intf) +{ + struct rio_mport *mport = to_rio_mport(dev); + struct cm_dev *cm; + struct cm_peer *peer, *temp; + struct rio_channel *ch, *_c; + unsigned int i; + bool found = false; + LIST_HEAD(list); + + riocm_debug(MPORT, "%s", mport->name); + + /* Find a matching cm_dev object */ + down_write(&rdev_sem); + list_for_each_entry(cm, &cm_dev_list, list) { + if (cm->mport == mport) { + list_del(&cm->list); + found = true; + break; + } + } + up_write(&rdev_sem); + if (!found) + return; + + flush_workqueue(cm->rx_wq); + destroy_workqueue(cm->rx_wq); + + /* Release channels bound to this mport */ + spin_lock_bh(&idr_lock); + idr_for_each_entry(&ch_idr, ch, i) { + if (ch->cmdev == cm) { + riocm_debug(RDEV, "%s drop ch_%d", + mport->name, ch->id); + idr_remove(&ch_idr, ch->id); + list_add(&ch->ch_node, &list); + } + } + spin_unlock_bh(&idr_lock); + + if (!list_empty(&list)) { + list_for_each_entry_safe(ch, _c, &list, ch_node) { + list_del(&ch->ch_node); + riocm_ch_close(ch); + } + } + + rio_release_inb_mbox(mport, cmbox); + rio_release_outb_mbox(mport, cmbox); + + /* Remove and free peer entries */ + if (!list_empty(&cm->peers)) + riocm_debug(RDEV, "ATTN: peer list not empty"); + list_for_each_entry_safe(peer, temp, &cm->peers, node) { + riocm_debug(RDEV, "removing peer %s", rio_name(peer->rdev)); + list_del(&peer->node); + kfree(peer); + } + + riocm_rx_free(cm); + kfree(cm); + riocm_debug(MPORT, "%s done", mport->name); +} + +static int rio_cm_shutdown(struct notifier_block *nb, unsigned long code, + void *unused) +{ + struct rio_channel *ch; + unsigned int i; + + riocm_debug(EXIT, "."); + + spin_lock_bh(&idr_lock); + idr_for_each_entry(&ch_idr, ch, i) { + riocm_debug(EXIT, "close ch %d", ch->id); + if (ch->state == RIO_CM_CONNECTED) + riocm_send_close(ch); + } + spin_unlock_bh(&idr_lock); + + return NOTIFY_DONE; +} + +/* + * riocm_interface handles addition/removal of remote RapidIO devices + */ +static struct subsys_interface riocm_interface = { + .name = "rio_cm", + .subsys = &rio_bus_type, + .add_dev = riocm_add_dev, + .remove_dev = riocm_remove_dev, +}; + +/* + * rio_mport_interface handles addition/removal local mport devices + */ +static struct class_interface rio_mport_interface __refdata = { + .class = &rio_mport_class, + .add_dev = riocm_add_mport, + .remove_dev = riocm_remove_mport, +}; + +static struct notifier_block rio_cm_notifier = { + .notifier_call = rio_cm_shutdown, +}; + +static int __init riocm_init(void) +{ + int ret; + + /* Create device class needed by udev */ + dev_class = class_create(THIS_MODULE, DRV_NAME); + if (IS_ERR(dev_class)) { + riocm_error("Cannot create " DRV_NAME " class"); + return PTR_ERR(dev_class); + } + + ret = alloc_chrdev_region(&dev_number, 0, 1, DRV_NAME); + if (ret) { + class_destroy(dev_class); + return ret; + } + + dev_major = MAJOR(dev_number); + dev_minor_base = MINOR(dev_number); + riocm_debug(INIT, "Registered class with %d major", dev_major); + + /* + * Register as rapidio_port class interface to get notifications about + * mport additions and removals. + */ + ret = class_interface_register(&rio_mport_interface); + if (ret) { + riocm_error("class_interface_register error: %d", ret); + goto err_reg; + } + + /* + * Register as RapidIO bus interface to get notifications about + * addition/removal of remote RapidIO devices. + */ + ret = subsys_interface_register(&riocm_interface); + if (ret) { + riocm_error("subsys_interface_register error: %d", ret); + goto err_cl; + } + + ret = register_reboot_notifier(&rio_cm_notifier); + if (ret) { + riocm_error("failed to register reboot notifier (err=%d)", ret); + goto err_sif; + } + + ret = riocm_cdev_add(dev_number); + if (ret) { + unregister_reboot_notifier(&rio_cm_notifier); + ret = -ENODEV; + goto err_sif; + } + + return 0; +err_sif: + subsys_interface_unregister(&riocm_interface); +err_cl: + class_interface_unregister(&rio_mport_interface); +err_reg: + unregister_chrdev_region(dev_number, 1); + class_destroy(dev_class); + return ret; +} + +static void __exit riocm_exit(void) +{ + riocm_debug(EXIT, "enter"); + unregister_reboot_notifier(&rio_cm_notifier); + subsys_interface_unregister(&riocm_interface); + class_interface_unregister(&rio_mport_interface); + idr_destroy(&ch_idr); + + device_unregister(riocm_cdev.dev); + cdev_del(&(riocm_cdev.cdev)); + + class_destroy(dev_class); + unregister_chrdev_region(dev_number, 1); +} + +late_initcall(riocm_init); +module_exit(riocm_exit); diff --git a/include/uapi/linux/Kbuild b/include/uapi/linux/Kbuild index 6d4e92c..c44747c 100644 --- a/include/uapi/linux/Kbuild +++ b/include/uapi/linux/Kbuild @@ -357,6 +357,7 @@ header-y += reiserfs_fs.h header-y += reiserfs_xattr.h header-y += resource.h header-y += rfkill.h +header-y += rio_cm_cdev.h header-y += rio_mport_cdev.h header-y += romfs_fs.h header-y += rose.h diff --git a/include/uapi/linux/rio_cm_cdev.h b/include/uapi/linux/rio_cm_cdev.h new file mode 100644 index 0000000..6edb900 --- /dev/null +++ b/include/uapi/linux/rio_cm_cdev.h @@ -0,0 +1,78 @@ +/* + * Copyright (c) 2015, Integrated Device Technology Inc. + * Copyright (c) 2015, Prodrive Technologies + * Copyright (c) 2015, RapidIO Trade Association + * All rights reserved. + * + * This software is available to you under a choice of one of two licenses. + * You may choose to be licensed under the terms of the GNU General Public + * License(GPL) Version 2, or the BSD-3 Clause license below: + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions are met: + * + * 1. Redistributions of source code must retain the above copyright notice, + * this list of conditions and the following disclaimer. + * + * 2. Redistributions in binary form must reproduce the above copyright notice, + * this list of conditions and the following disclaimer in the documentation + * and/or other materials provided with the distribution. + * + * 3. Neither the name of the copyright holder nor the names of its contributors + * may be used to endorse or promote products derived from this software without + * specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" + * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, + * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR + * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, + * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, + * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; + * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR + * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF + * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#ifndef _RIO_CM_CDEV_H_ +#define _RIO_CM_CDEV_H_ + +#include + +struct rio_cm_channel { + __u16 id; + __u16 remote_channel; + __u16 remote_destid; + __u8 mport_id; +}; + +struct rio_cm_msg { + __u16 ch_num; + __u16 size; + __u32 rxto; /* receive timeout in mSec. 0 = blocking */ + __u64 msg; +}; + +struct rio_cm_accept { + __u16 ch_num; + __u16 pad0; + __u32 wait_to; /* accept timeout in mSec. 0 = blocking */ +}; + +/* RapidIO Channelized Messaging Driver IOCTLs */ +#define RIO_CM_IOC_MAGIC 'c' + +#define RIO_CM_EP_GET_LIST_SIZE _IOWR(RIO_CM_IOC_MAGIC, 1, __u32) +#define RIO_CM_EP_GET_LIST _IOWR(RIO_CM_IOC_MAGIC, 2, __u32) +#define RIO_CM_CHAN_CREATE _IOWR(RIO_CM_IOC_MAGIC, 3, __u16) +#define RIO_CM_CHAN_CLOSE _IOW(RIO_CM_IOC_MAGIC, 4, __u16) +#define RIO_CM_CHAN_BIND _IOW(RIO_CM_IOC_MAGIC, 5, struct rio_cm_channel) +#define RIO_CM_CHAN_LISTEN _IOW(RIO_CM_IOC_MAGIC, 6, __u16) +#define RIO_CM_CHAN_ACCEPT _IOWR(RIO_CM_IOC_MAGIC, 7, struct rio_cm_accept) +#define RIO_CM_CHAN_CONNECT _IOW(RIO_CM_IOC_MAGIC, 8, struct rio_cm_channel) +#define RIO_CM_CHAN_SEND _IOW(RIO_CM_IOC_MAGIC, 9, struct rio_cm_msg) +#define RIO_CM_CHAN_RECEIVE _IOWR(RIO_CM_IOC_MAGIC, 10, struct rio_cm_msg) +#define RIO_CM_MPORT_GET_LIST _IOWR(RIO_CM_IOC_MAGIC, 11, __u32) + +#endif /* _RIO_CM_CDEV_H_ */ -- cgit v0.10.2 From ea87b8e1f72896d9852f534f77aeec5eaf377b32 Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Tue, 2 Aug 2016 14:06:28 -0700 Subject: rapidio: remove unnecessary 0x prefixes before %pa extension uses Patch series "RapidIO subsystem updates". This set of patches contains RapidIO subsystem fixes and updates that have been made since kernel v4.6. The most significant update brings changes related to the latest revision of RapidIO specification (rev.3.x) and introduction of next generation of RapidIO switches by IDT (RXS1632 and RXS2448). This patch (of 13): This is RapidIO part of the original patch submitted by Joe Perches. (see: https://lkml.org/lkml/2016/3/5/19) Since commit 3cab1e711297 ("lib/vsprintf: refactor duplicate code to special_hex_number()") %pa uses have been output with a 0x prefix. These 0x prefixes in the formats are unnecessary. Link: http://lkml.kernel.org/r/1469125134-16523-2-git-send-email-alexandre.bounine@idt.com Signed-off-by: Joe Perches Signed-off-by: Alexandre Bounine Cc: Matt Porter Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rapidio/devices/rio_mport_cdev.c b/drivers/rapidio/devices/rio_mport_cdev.c index e165b7c..de0c692 100644 --- a/drivers/rapidio/devices/rio_mport_cdev.c +++ b/drivers/rapidio/devices/rio_mport_cdev.c @@ -2242,7 +2242,7 @@ static void mport_mm_open(struct vm_area_struct *vma) { struct rio_mport_mapping *map = vma->vm_private_data; -rmcd_debug(MMAP, "0x%pad", &map->phys_addr); + rmcd_debug(MMAP, "%pad", &map->phys_addr); kref_get(&map->ref); } @@ -2250,7 +2250,7 @@ static void mport_mm_close(struct vm_area_struct *vma) { struct rio_mport_mapping *map = vma->vm_private_data; -rmcd_debug(MMAP, "0x%pad", &map->phys_addr); + rmcd_debug(MMAP, "%pad", &map->phys_addr); mutex_lock(&map->md->buf_mutex); kref_put(&map->ref, mport_release_mapping); mutex_unlock(&map->md->buf_mutex); diff --git a/drivers/rapidio/devices/tsi721.c b/drivers/rapidio/devices/tsi721.c index b5b4556..4c20e99 100644 --- a/drivers/rapidio/devices/tsi721.c +++ b/drivers/rapidio/devices/tsi721.c @@ -1101,7 +1101,7 @@ static int tsi721_rio_map_inb_mem(struct rio_mport *mport, dma_addr_t lstart, ibw_start = lstart & ~(ibw_size - 1); tsi_debug(IBW, &priv->pdev->dev, - "Direct (RIO_0x%llx -> PCIe_0x%pad), size=0x%x, ibw_start = 0x%llx", + "Direct (RIO_0x%llx -> PCIe_%pad), size=0x%x, ibw_start = 0x%llx", rstart, &lstart, size, ibw_start); while ((lstart + size) > (ibw_start + ibw_size)) { @@ -1120,7 +1120,7 @@ static int tsi721_rio_map_inb_mem(struct rio_mport *mport, dma_addr_t lstart, } else { tsi_debug(IBW, &priv->pdev->dev, - "Translated (RIO_0x%llx -> PCIe_0x%pad), size=0x%x", + "Translated (RIO_0x%llx -> PCIe_%pad), size=0x%x", rstart, &lstart, size); if (!is_power_of_2(size) || size < 0x1000 || @@ -1215,7 +1215,7 @@ static int tsi721_rio_map_inb_mem(struct rio_mport *mport, dma_addr_t lstart, priv->ibwin_cnt--; tsi_debug(IBW, &priv->pdev->dev, - "Configured IBWIN%d (RIO_0x%llx -> PCIe_0x%pad), size=0x%llx", + "Configured IBWIN%d (RIO_0x%llx -> PCIe_%pad), size=0x%llx", i, ibw_start, &loc_start, ibw_size); return 0; @@ -1237,7 +1237,7 @@ static void tsi721_rio_unmap_inb_mem(struct rio_mport *mport, int i; tsi_debug(IBW, &priv->pdev->dev, - "Unmap IBW mapped to PCIe_0x%pad", &lstart); + "Unmap IBW mapped to PCIe_%pad", &lstart); /* Search for matching active inbound translation window */ for (i = 0; i < TSI721_IBWIN_NUM; i++) { -- cgit v0.10.2 From cca446d4ce883ae28ff589dd1c9aef8d5148c7f7 Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:06:31 -0700 Subject: rapidio/documentation: fix mangled paragraph in mport_cdev Minor edits to correct parameter description. This patch is applicable to kernel versions starting from v4.6. Link: http://lkml.kernel.org/r/1469125134-16523-3-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Reported-by: Barry Wood Cc: Matt Porter Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/Documentation/rapidio/mport_cdev.txt b/Documentation/rapidio/mport_cdev.txt index 20c120d..6e491a6 100644 --- a/Documentation/rapidio/mport_cdev.txt +++ b/Documentation/rapidio/mport_cdev.txt @@ -82,8 +82,7 @@ III. Module parameters - 'dbg_level' - This parameter allows to control amount of debug information generated by this device driver. This parameter is formed by set of - This parameter can be changed bit masks that correspond to the specific - functional block. + bit masks that correspond to the specific functional blocks. For mask definitions see 'drivers/rapidio/devices/rio_mport_cdev.c' This parameter can be changed dynamically. Use CONFIG_RAPIDIO_DEBUG=y to enable debug output at the top level. -- cgit v0.10.2 From f8e3a68c05f0f09a0da947b9d447268d2d3f8780 Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:06:34 -0700 Subject: rapidio: fix return value description for dma_prep functions Update return value description for rio_dma_prep_... functions to include error-valued pointer that can be returned by HW mport device drivers. Return values from these functions must be checked using IS_ERR_OR_NULL macro. This patch is applicable to kernel versions starting from v4.6-rc1. Link: http://lkml.kernel.org/r/1469125134-16523-4-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Cc: Matt Porter Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rapidio/rio.c b/drivers/rapidio/rio.c index 0dcaa66..8408029 100644 --- a/drivers/rapidio/rio.c +++ b/drivers/rapidio/rio.c @@ -1848,7 +1848,9 @@ EXPORT_SYMBOL_GPL(rio_release_dma); * Initializes RapidIO capable DMA channel for the specified data transfer. * Uses DMA channel private extension to pass information related to remote * target RIO device. - * Returns pointer to DMA transaction descriptor or NULL if failed. + * + * Returns: pointer to DMA transaction descriptor if successful, + * error-valued pointer or NULL if failed. */ struct dma_async_tx_descriptor *rio_dma_prep_xfer(struct dma_chan *dchan, u16 destid, struct rio_dma_data *data, @@ -1883,7 +1885,9 @@ EXPORT_SYMBOL_GPL(rio_dma_prep_xfer); * Initializes RapidIO capable DMA channel for the specified data transfer. * Uses DMA channel private extension to pass information related to remote * target RIO device. - * Returns pointer to DMA transaction descriptor or NULL if failed. + * + * Returns: pointer to DMA transaction descriptor if successful, + * error-valued pointer or NULL if failed. */ struct dma_async_tx_descriptor *rio_dma_prep_slave_sg(struct rio_dev *rdev, struct dma_chan *dchan, struct rio_dma_data *data, -- cgit v0.10.2 From 4498c31adff99d243b34b0bf39363a35ea070928 Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:06:37 -0700 Subject: rapidio/tsi721_dma: add channel mask and queue size parameters Add module parameters to allow load time configuration of DMA channels. Depending on application, performance of DMA data transfers can benefit from adjusted sizes of buffer descriptor ring and/or transaction requests queue. Having HW DMA channel selector mask allows to define which channels (from seven available) are controlled by the mport device driver and reserve some of them for direct use by other drivers. Link: http://lkml.kernel.org/r/1469125134-16523-5-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Tested-by: Barry Wood Cc: Matt Porter Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/Documentation/rapidio/tsi721.txt b/Documentation/rapidio/tsi721.txt index 7c1c7bf..0e0e90be 100644 --- a/Documentation/rapidio/tsi721.txt +++ b/Documentation/rapidio/tsi721.txt @@ -25,6 +25,20 @@ fully compatible with RIONET driver (Ethernet over RapidIO messaging services). This parameter can be changed dynamically. Use CONFIG_RAPIDIO_DEBUG=y to enable debug output at the top level. +- 'dma_desc_per_channel' - This parameter defines number of hardware buffer + descriptors allocated for each registered Tsi721 DMA channel. + Its default value is 128. + +- 'dma_txqueue_sz' - DMA transactions queue size. Defines number of pending + transaction requests that can be accepted by each DMA channel. + Default value is 16. + +- 'dma_sel' - DMA channel selection mask. Bitmask that defines which hardware + DMA channels (0 ... 6) will be registered with DmaEngine core. + If bit is set to 1, the corresponding DMA channel will be registered. + DMA channels not selected by this mask will not be used by this device + driver. Default value is 0x7f (use all channels). + II. Known problems None. diff --git a/drivers/rapidio/devices/tsi721.h b/drivers/rapidio/devices/tsi721.h index 5456dbd..5941437 100644 --- a/drivers/rapidio/devices/tsi721.h +++ b/drivers/rapidio/devices/tsi721.h @@ -661,7 +661,7 @@ enum dma_rtype { */ #define TSI721_DMA_CHNUM TSI721_DMA_MAXCH -#define TSI721_DMACH_MAINT 0 /* DMA channel for maint requests */ +#define TSI721_DMACH_MAINT 7 /* DMA channel for maint requests */ #define TSI721_DMACH_MAINT_NBD 32 /* Number of BDs for maint requests */ #define TSI721_DMACH_DMA 1 /* DMA channel for data transfers */ diff --git a/drivers/rapidio/devices/tsi721_dma.c b/drivers/rapidio/devices/tsi721_dma.c index 155cae1..13c669b 100644 --- a/drivers/rapidio/devices/tsi721_dma.c +++ b/drivers/rapidio/devices/tsi721_dma.c @@ -36,18 +36,26 @@ #include "tsi721.h" -#define TSI721_DMA_TX_QUEUE_SZ 16 /* number of transaction descriptors */ - #ifdef CONFIG_PCI_MSI static irqreturn_t tsi721_bdma_msix(int irq, void *ptr); #endif static int tsi721_submit_sg(struct tsi721_tx_desc *desc); static unsigned int dma_desc_per_channel = 128; -module_param(dma_desc_per_channel, uint, S_IWUSR | S_IRUGO); +module_param(dma_desc_per_channel, uint, S_IRUGO); MODULE_PARM_DESC(dma_desc_per_channel, "Number of DMA descriptors per channel (default: 128)"); +static unsigned int dma_txqueue_sz = 16; +module_param(dma_txqueue_sz, uint, S_IRUGO); +MODULE_PARM_DESC(dma_txqueue_sz, + "DMA Transactions Queue Size (default: 16)"); + +static u8 dma_sel = 0x7f; +module_param(dma_sel, byte, S_IRUGO); +MODULE_PARM_DESC(dma_sel, + "DMA Channel Selection Mask (default: 0x7f = all)"); + static inline struct tsi721_bdma_chan *to_tsi721_chan(struct dma_chan *chan) { return container_of(chan, struct tsi721_bdma_chan, dchan); @@ -732,7 +740,7 @@ static int tsi721_alloc_chan_resources(struct dma_chan *dchan) tsi_debug(DMA, &dchan->dev->device, "DMAC%d", bdma_chan->id); if (bdma_chan->bd_base) - return TSI721_DMA_TX_QUEUE_SZ; + return dma_txqueue_sz; /* Initialize BDMA channel */ if (tsi721_bdma_ch_init(bdma_chan, dma_desc_per_channel)) { @@ -742,7 +750,7 @@ static int tsi721_alloc_chan_resources(struct dma_chan *dchan) } /* Allocate queue of transaction descriptors */ - desc = kcalloc(TSI721_DMA_TX_QUEUE_SZ, sizeof(struct tsi721_tx_desc), + desc = kcalloc(dma_txqueue_sz, sizeof(struct tsi721_tx_desc), GFP_ATOMIC); if (!desc) { tsi_err(&dchan->dev->device, @@ -754,7 +762,7 @@ static int tsi721_alloc_chan_resources(struct dma_chan *dchan) bdma_chan->tx_desc = desc; - for (i = 0; i < TSI721_DMA_TX_QUEUE_SZ; i++) { + for (i = 0; i < dma_txqueue_sz; i++) { dma_async_tx_descriptor_init(&desc[i].txd, dchan); desc[i].txd.tx_submit = tsi721_tx_submit; desc[i].txd.flags = DMA_CTRL_ACK; @@ -766,7 +774,7 @@ static int tsi721_alloc_chan_resources(struct dma_chan *dchan) bdma_chan->active = true; tsi721_bdma_interrupt_enable(bdma_chan, 1); - return TSI721_DMA_TX_QUEUE_SZ; + return dma_txqueue_sz; } static void tsi721_sync_dma_irq(struct tsi721_bdma_chan *bdma_chan) @@ -962,7 +970,7 @@ void tsi721_dma_stop_all(struct tsi721_device *priv) int i; for (i = 0; i < TSI721_DMA_MAXCH; i++) { - if (i != TSI721_DMACH_MAINT) + if ((i != TSI721_DMACH_MAINT) && (dma_sel & (1 << i))) tsi721_dma_stop(&priv->bdma[i]); } } @@ -979,7 +987,7 @@ int tsi721_register_dma(struct tsi721_device *priv) for (i = 0; i < TSI721_DMA_MAXCH; i++) { struct tsi721_bdma_chan *bdma_chan = &priv->bdma[i]; - if (i == TSI721_DMACH_MAINT) + if ((i == TSI721_DMACH_MAINT) || (dma_sel & (1 << i)) == 0) continue; bdma_chan->regs = priv->regs + TSI721_DMAC_BASE(i); -- cgit v0.10.2 From cb782cdd2ffffbf7fd17e4aefb20f4db5c67caeb Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:06:40 -0700 Subject: rapidio/tsi721: add PCIe MRRS override parameter Add PCIe Maximum Read Request Size (MRRS) adjustment parameter to allow users to override configuration register value set during PCIe bus initialization. Performance of Tsi721 device as PCIe bus master can be improved if MRRS is set to its maximum value (4096 bytes). Some platforms have limitations for supported MRRS and therefore the default value should be preserved, unless it is known that given platform supports full set of MRRS values defined by PCI Express specification. Link: http://lkml.kernel.org/r/1469125134-16523-6-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Cc: Matt Porter Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/Documentation/rapidio/tsi721.txt b/Documentation/rapidio/tsi721.txt index 0e0e90be..9c6ee38 100644 --- a/Documentation/rapidio/tsi721.txt +++ b/Documentation/rapidio/tsi721.txt @@ -39,6 +39,13 @@ fully compatible with RIONET driver (Ethernet over RapidIO messaging services). DMA channels not selected by this mask will not be used by this device driver. Default value is 0x7f (use all channels). +- 'pcie_mrrs' - override value for PCIe Maximum Read Request Size (MRRS). + This parameter gives an ability to override MRRS value set during PCIe + configuration process. Tsi721 supports read request sizes up to 4096B. + Value for this parameter must be set as defined by PCIe specification: + 0 = 128B, 1 = 256B, 2 = 512B, 3 = 1024B, 4 = 2048B and 5 = 4096B. + Default value is '-1' (= keep platform setting). + II. Known problems None. diff --git a/drivers/rapidio/devices/tsi721.c b/drivers/rapidio/devices/tsi721.c index 4c20e99..85098f8 100644 --- a/drivers/rapidio/devices/tsi721.c +++ b/drivers/rapidio/devices/tsi721.c @@ -37,11 +37,15 @@ #include "tsi721.h" #ifdef DEBUG -u32 dbg_level = DBG_INIT | DBG_EXIT; +u32 dbg_level; module_param(dbg_level, uint, S_IWUSR | S_IRUGO); MODULE_PARM_DESC(dbg_level, "Debugging output level (default 0 = none)"); #endif +static int pcie_mrrs = -1; +module_param(pcie_mrrs, int, S_IRUGO); +MODULE_PARM_DESC(pcie_mrrs, "PCIe MRRS override value (0...5)"); + static void tsi721_omsg_handler(struct tsi721_device *priv, int ch); static void tsi721_imsg_handler(struct tsi721_device *priv, int ch); @@ -2840,6 +2844,16 @@ static int tsi721_probe(struct pci_dev *pdev, pcie_capability_clear_and_set_word(pdev, PCI_EXP_DEVCTL, PCI_EXP_DEVCTL_RELAX_EN | PCI_EXP_DEVCTL_NOSNOOP_EN, 0); + /* Override PCIe Maximum Read Request Size setting if requested */ + if (pcie_mrrs >= 0) { + if (pcie_mrrs <= 5) + pcie_capability_clear_and_set_word(pdev, PCI_EXP_DEVCTL, + PCI_EXP_DEVCTL_READRQ, pcie_mrrs << 12); + else + tsi_info(&pdev->dev, + "Invalid MRRS override value %d", pcie_mrrs); + } + /* Adjust PCIe completion timeout. */ pcie_capability_clear_and_set_word(pdev, PCI_EXP_DEVCTL2, 0xf, 0x2); -- cgit v0.10.2 From e519685de3e44bb013d81f5ead04ac4b33c9b3a1 Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:06:43 -0700 Subject: rapidio/tsi721: add messaging mbox selector parameter Add module parameter to allow load time configuration of available RapidIO messaging mailboxes (MBOX1 - MBOX4). Having a messaging MBOX selector mask allows to define which MBOXes are controlled by the mport device driver and reserve some of them for direct use by other drivers. Link: http://lkml.kernel.org/r/1469125134-16523-7-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Tested-by: Barry Wood Cc: Matt Porter Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/Documentation/rapidio/tsi721.txt b/Documentation/rapidio/tsi721.txt index 9c6ee38..cd2a293 100644 --- a/Documentation/rapidio/tsi721.txt +++ b/Documentation/rapidio/tsi721.txt @@ -46,6 +46,11 @@ fully compatible with RIONET driver (Ethernet over RapidIO messaging services). 0 = 128B, 1 = 256B, 2 = 512B, 3 = 1024B, 4 = 2048B and 5 = 4096B. Default value is '-1' (= keep platform setting). +- 'mbox_sel' - RIO messaging MBOX selection mask. This is a bitmask that defines + messaging MBOXes are managed by this device driver. Mask bits 0 - 3 + correspond to MBOX0 - MBOX3. MBOX is under driver's control if the + corresponding bit is set to '1'. Default value is 0x0f (= all). + II. Known problems None. diff --git a/drivers/rapidio/devices/tsi721.c b/drivers/rapidio/devices/tsi721.c index 85098f8..8e07cd5 100644 --- a/drivers/rapidio/devices/tsi721.c +++ b/drivers/rapidio/devices/tsi721.c @@ -46,6 +46,11 @@ static int pcie_mrrs = -1; module_param(pcie_mrrs, int, S_IRUGO); MODULE_PARM_DESC(pcie_mrrs, "PCIe MRRS override value (0...5)"); +static u8 mbox_sel = 0x0f; +module_param(mbox_sel, byte, S_IRUGO); +MODULE_PARM_DESC(mbox_sel, + "RIO Messaging MBOX Selection Mask (default: 0x0f = all)"); + static void tsi721_omsg_handler(struct tsi721_device *priv, int ch); static void tsi721_imsg_handler(struct tsi721_device *priv, int ch); @@ -1881,6 +1886,11 @@ static int tsi721_open_outb_mbox(struct rio_mport *mport, void *dev_id, goto out; } + if ((mbox_sel & (1 << mbox)) == 0) { + rc = -ENODEV; + goto out; + } + priv->omsg_ring[mbox].dev_id = dev_id; priv->omsg_ring[mbox].size = entries; priv->omsg_ring[mbox].sts_rdptr = 0; @@ -2165,6 +2175,11 @@ static int tsi721_open_inb_mbox(struct rio_mport *mport, void *dev_id, goto out; } + if ((mbox_sel & (1 << mbox)) == 0) { + rc = -ENODEV; + goto out; + } + /* Initialize IB Messaging Ring */ priv->imsg_ring[mbox].dev_id = dev_id; priv->imsg_ring[mbox].size = entries; -- cgit v0.10.2 From f5485eb0b6eb8a3e5841cfea34a930822f7252bc Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:06:46 -0700 Subject: rapidio/tsi721_dma: advance queue processing from transfer submit call Add advancing transfer queue immediately from transfer submit call. DMA performance improvement: This will start transfer without waiting for 'issue_pending' command if there is no DMA transfer in progress. Link: http://lkml.kernel.org/r/1469125134-16523-8-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Cc: Matt Porter Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rapidio/devices/tsi721_dma.c b/drivers/rapidio/devices/tsi721_dma.c index 13c669b..e2a4185 100644 --- a/drivers/rapidio/devices/tsi721_dma.c +++ b/drivers/rapidio/devices/tsi721_dma.c @@ -726,6 +726,7 @@ static dma_cookie_t tsi721_tx_submit(struct dma_async_tx_descriptor *txd) cookie = dma_cookie_assign(txd); desc->status = DMA_IN_PROGRESS; list_add_tail(&desc->desc_node, &bdma_chan->queue); + tsi721_advance_work(bdma_chan, NULL); spin_unlock_bh(&bdma_chan->lock); return cookie; -- cgit v0.10.2 From 06e1b2497ca4783f5f9997b09c77d93aeea69ec1 Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:06:49 -0700 Subject: rapidio: fix error handling in mbox request/release functions Add checking for error code returned by HW-specific mbox open routines. Ensure that resources are properly release if failed. This patch is applicable to kernel versions starting from v2.6.15. Link: http://lkml.kernel.org/r/1469125134-16523-9-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Cc: Matt Porter Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rapidio/rio.c b/drivers/rapidio/rio.c index 8408029..1cd3260 100644 --- a/drivers/rapidio/rio.c +++ b/drivers/rapidio/rio.c @@ -268,6 +268,12 @@ int rio_request_inb_mbox(struct rio_mport *mport, mport->inb_msg[mbox].mcback = minb; rc = mport->ops->open_inb_mbox(mport, dev_id, mbox, entries); + if (rc) { + mport->inb_msg[mbox].mcback = NULL; + mport->inb_msg[mbox].res = NULL; + release_resource(res); + kfree(res); + } } else rc = -ENOMEM; @@ -285,13 +291,22 @@ int rio_request_inb_mbox(struct rio_mport *mport, */ int rio_release_inb_mbox(struct rio_mport *mport, int mbox) { - if (mport->ops->close_inb_mbox) { - mport->ops->close_inb_mbox(mport, mbox); + int rc; - /* Release the mailbox resource */ - return release_resource(mport->inb_msg[mbox].res); - } else - return -ENOSYS; + if (!mport->ops->close_inb_mbox || !mport->inb_msg[mbox].res) + return -EINVAL; + + mport->ops->close_inb_mbox(mport, mbox); + mport->inb_msg[mbox].mcback = NULL; + + rc = release_resource(mport->inb_msg[mbox].res); + if (rc) + return rc; + + kfree(mport->inb_msg[mbox].res); + mport->inb_msg[mbox].res = NULL; + + return 0; } /** @@ -336,6 +351,12 @@ int rio_request_outb_mbox(struct rio_mport *mport, mport->outb_msg[mbox].mcback = moutb; rc = mport->ops->open_outb_mbox(mport, dev_id, mbox, entries); + if (rc) { + mport->outb_msg[mbox].mcback = NULL; + mport->outb_msg[mbox].res = NULL; + release_resource(res); + kfree(res); + } } else rc = -ENOMEM; @@ -353,13 +374,22 @@ int rio_request_outb_mbox(struct rio_mport *mport, */ int rio_release_outb_mbox(struct rio_mport *mport, int mbox) { - if (mport->ops->close_outb_mbox) { - mport->ops->close_outb_mbox(mport, mbox); + int rc; - /* Release the mailbox resource */ - return release_resource(mport->outb_msg[mbox].res); - } else - return -ENOSYS; + if (!mport->ops->close_outb_mbox || !mport->outb_msg[mbox].res) + return -EINVAL; + + mport->ops->close_outb_mbox(mport, mbox); + mport->outb_msg[mbox].mcback = NULL; + + rc = release_resource(mport->outb_msg[mbox].res); + if (rc) + return rc; + + kfree(mport->outb_msg[mbox].res); + mport->outb_msg[mbox].res = NULL; + + return 0; } /** -- cgit v0.10.2 From 60e377b5c1226d6737786947d0e915ab45d7f188 Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:06:52 -0700 Subject: rapidio/idt_gen2: fix locking warning Fix lockdep warning during device probing: move sysfs initialization out of code protected by a spin lock. Link: http://lkml.kernel.org/r/1469125134-16523-10-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Cc: Matt Porter Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rapidio/switches/idt_gen2.c b/drivers/rapidio/switches/idt_gen2.c index 9f7fe21..e67b923 100644 --- a/drivers/rapidio/switches/idt_gen2.c +++ b/drivers/rapidio/switches/idt_gen2.c @@ -436,10 +436,11 @@ static int idtg2_probe(struct rio_dev *rdev, const struct rio_device_id *id) RIO_STD_RTE_DEFAULT_PORT, IDT_NO_ROUTE); } + spin_unlock(&rdev->rswitch->lock); + /* Create device-specific sysfs attributes */ idtg2_sysfs(rdev, true); - spin_unlock(&rdev->rswitch->lock); return 0; } @@ -452,11 +453,9 @@ static void idtg2_remove(struct rio_dev *rdev) return; } rdev->rswitch->ops = NULL; - + spin_unlock(&rdev->rswitch->lock); /* Remove device-specific sysfs attributes */ idtg2_sysfs(rdev, false); - - spin_unlock(&rdev->rswitch->lock); } static struct rio_device_id idtg2_id_table[] = { -- cgit v0.10.2 From a057a52e94e15d89be8af557584e0144a496b6c6 Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:06:54 -0700 Subject: rapidio: change inbound window size type to u64 Current definition of map_inb() mport operations callback uses u32 type to specify required inbound window (IBW) size. This is limiting factor because existing hardware - tsi721 and fsl_rio, both support IBW size up to 16GB. Changing type of size parameter to u64 to allow IBW size configurations larger than 4GB. [alexandre.bounine@idt.com: remove compiler warning about size of constant] Link: http://lkml.kernel.org/r/20160802184856.2566-1-alexandre.bounine@idt.com Link: http://lkml.kernel.org/r/1469125134-16523-11-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Cc: Matt Porter Cc: Benjamin Herrenschmidt Cc: Michael Ellerman Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/powerpc/sysdev/fsl_rio.c b/arch/powerpc/sysdev/fsl_rio.c index f5bf38b9..386790c 100644 --- a/arch/powerpc/sysdev/fsl_rio.c +++ b/arch/powerpc/sysdev/fsl_rio.c @@ -289,7 +289,7 @@ static void fsl_rio_inbound_mem_init(struct rio_priv *priv) } int fsl_map_inb_mem(struct rio_mport *mport, dma_addr_t lstart, - u64 rstart, u32 size, u32 flags) + u64 rstart, u64 size, u32 flags) { struct rio_priv *priv = mport->priv; u32 base_size; @@ -298,7 +298,7 @@ int fsl_map_inb_mem(struct rio_mport *mport, dma_addr_t lstart, u32 riwar; int i; - if ((size & (size - 1)) != 0) + if ((size & (size - 1)) != 0 || size > 0x400000000ULL) return -EINVAL; base_size_log = ilog2(size); diff --git a/drivers/rapidio/devices/tsi721.c b/drivers/rapidio/devices/tsi721.c index 8e07cd5..53daf63 100644 --- a/drivers/rapidio/devices/tsi721.c +++ b/drivers/rapidio/devices/tsi721.c @@ -1090,7 +1090,7 @@ static void tsi721_init_pc2sr_mapping(struct tsi721_device *priv) * from rstart to lstart. */ static int tsi721_rio_map_inb_mem(struct rio_mport *mport, dma_addr_t lstart, - u64 rstart, u32 size, u32 flags) + u64 rstart, u64 size, u32 flags) { struct tsi721_device *priv = mport->priv; int i, avail = -1; @@ -1103,6 +1103,10 @@ static int tsi721_rio_map_inb_mem(struct rio_mport *mport, dma_addr_t lstart, struct tsi721_ib_win_mapping *map = NULL; int ret = -EBUSY; + /* Max IBW size supported by HW is 16GB */ + if (size > 0x400000000UL) + return -EINVAL; + if (direct) { /* Calculate minimal acceptable window size and base address */ @@ -1110,15 +1114,15 @@ static int tsi721_rio_map_inb_mem(struct rio_mport *mport, dma_addr_t lstart, ibw_start = lstart & ~(ibw_size - 1); tsi_debug(IBW, &priv->pdev->dev, - "Direct (RIO_0x%llx -> PCIe_%pad), size=0x%x, ibw_start = 0x%llx", + "Direct (RIO_0x%llx -> PCIe_%pad), size=0x%llx, ibw_start = 0x%llx", rstart, &lstart, size, ibw_start); while ((lstart + size) > (ibw_start + ibw_size)) { ibw_size *= 2; ibw_start = lstart & ~(ibw_size - 1); - if (ibw_size > 0x80000000) { /* Limit max size to 2GB */ + /* Check for crossing IBW max size 16GB */ + if (ibw_size > 0x400000000UL) return -EBUSY; - } } loc_start = ibw_start; @@ -1129,7 +1133,7 @@ static int tsi721_rio_map_inb_mem(struct rio_mport *mport, dma_addr_t lstart, } else { tsi_debug(IBW, &priv->pdev->dev, - "Translated (RIO_0x%llx -> PCIe_%pad), size=0x%x", + "Translated (RIO_0x%llx -> PCIe_%pad), size=0x%llx", rstart, &lstart, size); if (!is_power_of_2(size) || size < 0x1000 || diff --git a/include/linux/rio.h b/include/linux/rio.h index aa23238..f7ec35b 100644 --- a/include/linux/rio.h +++ b/include/linux/rio.h @@ -425,7 +425,7 @@ struct rio_ops { int (*add_inb_buffer)(struct rio_mport *mport, int mbox, void *buf); void *(*get_inb_message)(struct rio_mport *mport, int mbox); int (*map_inb)(struct rio_mport *mport, dma_addr_t lstart, - u64 rstart, u32 size, u32 flags); + u64 rstart, u64 size, u32 flags); void (*unmap_inb)(struct rio_mport *mport, dma_addr_t lstart); int (*query_mport)(struct rio_mport *mport, struct rio_mport_attr *attr); -- cgit v0.10.2 From 1ae842de1dd8051cbb65b396b6f029d07f992641 Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:06:57 -0700 Subject: rapidio: modify for rev.3 specification changes Implement changes made in RapidIO specification rev.3 to LP-Serial Physical Layer register definitions: - use per-port register offset calculations based on LP-Serial Extended Features Block (EFB) Register Map type (I or II) with different per-port offset step (0x20 vs 0x40 respectfully). - remove deprecated Parallel Physical layer definitions and related code. [alexandre.bounine@idt.com: fix DocBook warning for gen3 update] Link: http://lkml.kernel.org/r/1469191173-19338-1-git-send-email-alexandre.bounine@idt.com Link: http://lkml.kernel.org/r/1469125134-16523-12-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Tested-by: Barry Wood Cc: Matt Porter Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rapidio/devices/rio_mport_cdev.c b/drivers/rapidio/devices/rio_mport_cdev.c index de0c692..436dfe8 100644 --- a/drivers/rapidio/devices/rio_mport_cdev.c +++ b/drivers/rapidio/devices/rio_mport_cdev.c @@ -1813,7 +1813,7 @@ static int rio_mport_add_riodev(struct mport_cdev_priv *priv, if (rdev->pef & RIO_PEF_EXT_FEATURES) { rdev->efptr = rval & 0xffff; rdev->phys_efptr = rio_mport_get_physefb(mport, 0, destid, - hopcount); + hopcount, &rdev->phys_rmap); rdev->em_efptr = rio_mport_get_feature(mport, 0, destid, hopcount, RIO_EFB_ERR_MGMNT); diff --git a/drivers/rapidio/devices/tsi721.c b/drivers/rapidio/devices/tsi721.c index 53daf63..32f0f01 100644 --- a/drivers/rapidio/devices/tsi721.c +++ b/drivers/rapidio/devices/tsi721.c @@ -2555,11 +2555,11 @@ static int tsi721_query_mport(struct rio_mport *mport, struct tsi721_device *priv = mport->priv; u32 rval; - rval = ioread32(priv->regs + (0x100 + RIO_PORT_N_ERR_STS_CSR(0))); + rval = ioread32(priv->regs + 0x100 + RIO_PORT_N_ERR_STS_CSR(0, 0)); if (rval & RIO_PORT_N_ERR_STS_PORT_OK) { - rval = ioread32(priv->regs + (0x100 + RIO_PORT_N_CTL2_CSR(0))); + rval = ioread32(priv->regs + 0x100 + RIO_PORT_N_CTL2_CSR(0, 0)); attr->link_speed = (rval & RIO_PORT_N_CTL2_SEL_BAUD) >> 28; - rval = ioread32(priv->regs + (0x100 + RIO_PORT_N_CTL_CSR(0))); + rval = ioread32(priv->regs + 0x100 + RIO_PORT_N_CTL_CSR(0, 0)); attr->link_width = (rval & RIO_PORT_N_CTL_IPW) >> 27; } else attr->link_speed = RIO_LINK_DOWN; @@ -2673,9 +2673,9 @@ static int tsi721_setup_mport(struct tsi721_device *priv) mport->ops = &tsi721_rio_ops; mport->index = 0; mport->sys_size = 0; /* small system */ - mport->phy_type = RIO_PHY_SERIAL; mport->priv = (void *)priv; mport->phys_efptr = 0x100; + mport->phys_rmap = 1; mport->dev.parent = &pdev->dev; mport->dev.release = tsi721_mport_release; diff --git a/drivers/rapidio/rio-scan.c b/drivers/rapidio/rio-scan.c index a63a380..23429bd 100644 --- a/drivers/rapidio/rio-scan.c +++ b/drivers/rapidio/rio-scan.c @@ -49,15 +49,6 @@ struct rio_id_table { static int next_destid = 0; static int next_comptag = 1; -static int rio_mport_phys_table[] = { - RIO_EFB_PAR_EP_ID, - RIO_EFB_PAR_EP_REC_ID, - RIO_EFB_SER_EP_ID, - RIO_EFB_SER_EP_REC_ID, - -1, -}; - - /** * rio_destid_alloc - Allocate next available destID for given network * @net: RIO network @@ -380,10 +371,15 @@ static struct rio_dev *rio_setup_device(struct rio_net *net, if (rdev->pef & RIO_PEF_EXT_FEATURES) { rdev->efptr = result & 0xffff; rdev->phys_efptr = rio_mport_get_physefb(port, 0, destid, - hopcount); + hopcount, &rdev->phys_rmap); + pr_debug("RIO: %s Register Map %d device\n", + __func__, rdev->phys_rmap); rdev->em_efptr = rio_mport_get_feature(port, 0, destid, hopcount, RIO_EFB_ERR_MGMNT); + if (!rdev->em_efptr) + rdev->em_efptr = rio_mport_get_feature(port, 0, destid, + hopcount, RIO_EFB_ERR_MGMNT_HS); } rio_mport_read_config_32(port, destid, hopcount, RIO_SRC_OPS_CAR, @@ -445,7 +441,7 @@ static struct rio_dev *rio_setup_device(struct rio_net *net, rio_route_clr_table(rdev, RIO_GLOBAL_TABLE, 0); } else { if (do_enum) - /*Enable Input Output Port (transmitter reviever)*/ + /*Enable Input Output Port (transmitter receiver)*/ rio_enable_rx_tx_port(port, 0, destid, hopcount, 0); dev_set_name(&rdev->dev, "%02x:e:%04x", rdev->net->id, @@ -481,10 +477,8 @@ cleanup: /** * rio_sport_is_active- Tests if a switch port has an active connection. - * @port: Master port to send transaction - * @destid: Associated destination ID for switch - * @hopcount: Hopcount to reach switch - * @sport: Switch port number + * @rdev: RapidIO device object + * @sp: Switch port number * * Reads the port error status CSR for a particular switch port to * determine if the port has an active link. Returns @@ -492,31 +486,12 @@ cleanup: * inactive. */ static int -rio_sport_is_active(struct rio_mport *port, u16 destid, u8 hopcount, int sport) +rio_sport_is_active(struct rio_dev *rdev, int sp) { u32 result = 0; - u32 ext_ftr_ptr; - ext_ftr_ptr = rio_mport_get_efb(port, 0, destid, hopcount, 0); - - while (ext_ftr_ptr) { - rio_mport_read_config_32(port, destid, hopcount, - ext_ftr_ptr, &result); - result = RIO_GET_BLOCK_ID(result); - if ((result == RIO_EFB_SER_EP_FREE_ID) || - (result == RIO_EFB_SER_EP_FREE_ID_V13P) || - (result == RIO_EFB_SER_EP_FREC_ID)) - break; - - ext_ftr_ptr = rio_mport_get_efb(port, 0, destid, hopcount, - ext_ftr_ptr); - } - - if (ext_ftr_ptr) - rio_mport_read_config_32(port, destid, hopcount, - ext_ftr_ptr + - RIO_PORT_N_ERR_STS_CSR(sport), - &result); + rio_read_config_32(rdev, RIO_DEV_PORT_N_ERR_STS_CSR(rdev, sp), + &result); return result & RIO_PORT_N_ERR_STS_PORT_OK; } @@ -655,9 +630,7 @@ static int rio_enum_peer(struct rio_net *net, struct rio_mport *port, cur_destid = next_destid; - if (rio_sport_is_active - (port, RIO_ANY_DESTID(port->sys_size), hopcount, - port_num)) { + if (rio_sport_is_active(rdev, port_num)) { pr_debug( "RIO: scanning device on port %d\n", port_num); @@ -785,8 +758,7 @@ rio_disc_peer(struct rio_net *net, struct rio_mport *port, u16 destid, if (RIO_GET_PORT_NUM(rdev->swpinfo) == port_num) continue; - if (rio_sport_is_active - (port, destid, hopcount, port_num)) { + if (rio_sport_is_active(rdev, port_num)) { pr_debug( "RIO: scanning device on port %d\n", port_num); @@ -831,21 +803,11 @@ rio_disc_peer(struct rio_net *net, struct rio_mport *port, u16 destid, static int rio_mport_is_active(struct rio_mport *port) { u32 result = 0; - u32 ext_ftr_ptr; - int *entry = rio_mport_phys_table; - - do { - if ((ext_ftr_ptr = - rio_mport_get_feature(port, 1, 0, 0, *entry))) - break; - } while (*++entry >= 0); - - if (ext_ftr_ptr) - rio_local_read_config_32(port, - ext_ftr_ptr + - RIO_PORT_N_ERR_STS_CSR(port->index), - &result); + rio_local_read_config_32(port, + port->phys_efptr + + RIO_PORT_N_ERR_STS_CSR(port->index, port->phys_rmap), + &result); return result & RIO_PORT_N_ERR_STS_PORT_OK; } diff --git a/drivers/rapidio/rio.c b/drivers/rapidio/rio.c index 1cd3260..3704285 100644 --- a/drivers/rapidio/rio.c +++ b/drivers/rapidio/rio.c @@ -786,10 +786,11 @@ EXPORT_SYMBOL_GPL(rio_unmap_outb_region); * @local: Indicate a local master port or remote device access * @destid: Destination ID of the device * @hopcount: Number of switch hops to the device + * @rmap: pointer to location to store register map type info */ u32 rio_mport_get_physefb(struct rio_mport *port, int local, - u16 destid, u8 hopcount) + u16 destid, u8 hopcount, u32 *rmap) { u32 ext_ftr_ptr; u32 ftr_header; @@ -807,14 +808,21 @@ rio_mport_get_physefb(struct rio_mport *port, int local, ftr_header = RIO_GET_BLOCK_ID(ftr_header); switch (ftr_header) { - case RIO_EFB_SER_EP_ID_V13P: - case RIO_EFB_SER_EP_REC_ID_V13P: - case RIO_EFB_SER_EP_FREE_ID_V13P: case RIO_EFB_SER_EP_ID: case RIO_EFB_SER_EP_REC_ID: case RIO_EFB_SER_EP_FREE_ID: - case RIO_EFB_SER_EP_FREC_ID: + case RIO_EFB_SER_EP_M1_ID: + case RIO_EFB_SER_EP_SW_M1_ID: + case RIO_EFB_SER_EPF_M1_ID: + case RIO_EFB_SER_EPF_SW_M1_ID: + *rmap = 1; + return ext_ftr_ptr; + case RIO_EFB_SER_EP_M2_ID: + case RIO_EFB_SER_EP_SW_M2_ID: + case RIO_EFB_SER_EPF_M2_ID: + case RIO_EFB_SER_EPF_SW_M2_ID: + *rmap = 2; return ext_ftr_ptr; default: @@ -873,16 +881,16 @@ int rio_set_port_lockout(struct rio_dev *rdev, u32 pnum, int lock) u32 regval; rio_read_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_CTL_CSR(pnum), - ®val); + RIO_DEV_PORT_N_CTL_CSR(rdev, pnum), + ®val); if (lock) regval |= RIO_PORT_N_CTL_LOCKOUT; else regval &= ~RIO_PORT_N_CTL_LOCKOUT; rio_write_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_CTL_CSR(pnum), - regval); + RIO_DEV_PORT_N_CTL_CSR(rdev, pnum), + regval); return 0; } EXPORT_SYMBOL_GPL(rio_set_port_lockout); @@ -906,6 +914,7 @@ int rio_enable_rx_tx_port(struct rio_mport *port, #ifdef CONFIG_RAPIDIO_ENABLE_RX_TX_PORTS u32 regval; u32 ext_ftr_ptr; + u32 rmap; /* * enable rx input tx output port @@ -913,34 +922,29 @@ int rio_enable_rx_tx_port(struct rio_mport *port, pr_debug("rio_enable_rx_tx_port(local = %d, destid = %d, hopcount = " "%d, port_num = %d)\n", local, destid, hopcount, port_num); - ext_ftr_ptr = rio_mport_get_physefb(port, local, destid, hopcount); + ext_ftr_ptr = rio_mport_get_physefb(port, local, destid, + hopcount, &rmap); if (local) { - rio_local_read_config_32(port, ext_ftr_ptr + - RIO_PORT_N_CTL_CSR(0), + rio_local_read_config_32(port, + ext_ftr_ptr + RIO_PORT_N_CTL_CSR(0, rmap), ®val); } else { if (rio_mport_read_config_32(port, destid, hopcount, - ext_ftr_ptr + RIO_PORT_N_CTL_CSR(port_num), ®val) < 0) + ext_ftr_ptr + RIO_PORT_N_CTL_CSR(port_num, rmap), + ®val) < 0) return -EIO; } - if (regval & RIO_PORT_N_CTL_P_TYP_SER) { - /* serial */ - regval = regval | RIO_PORT_N_CTL_EN_RX_SER - | RIO_PORT_N_CTL_EN_TX_SER; - } else { - /* parallel */ - regval = regval | RIO_PORT_N_CTL_EN_RX_PAR - | RIO_PORT_N_CTL_EN_TX_PAR; - } + regval = regval | RIO_PORT_N_CTL_EN_RX | RIO_PORT_N_CTL_EN_TX; if (local) { - rio_local_write_config_32(port, ext_ftr_ptr + - RIO_PORT_N_CTL_CSR(0), regval); + rio_local_write_config_32(port, + ext_ftr_ptr + RIO_PORT_N_CTL_CSR(0, rmap), regval); } else { if (rio_mport_write_config_32(port, destid, hopcount, - ext_ftr_ptr + RIO_PORT_N_CTL_CSR(port_num), regval) < 0) + ext_ftr_ptr + RIO_PORT_N_CTL_CSR(port_num, rmap), + regval) < 0) return -EIO; } #endif @@ -1042,14 +1046,14 @@ rio_get_input_status(struct rio_dev *rdev, int pnum, u32 *lnkresp) /* Read from link maintenance response register * to clear valid bit */ rio_read_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_MNT_RSP_CSR(pnum), + RIO_DEV_PORT_N_MNT_RSP_CSR(rdev, pnum), ®val); udelay(50); } /* Issue Input-status command */ rio_write_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_MNT_REQ_CSR(pnum), + RIO_DEV_PORT_N_MNT_REQ_CSR(rdev, pnum), RIO_MNT_REQ_CMD_IS); /* Exit if the response is not expected */ @@ -1060,7 +1064,7 @@ rio_get_input_status(struct rio_dev *rdev, int pnum, u32 *lnkresp) while (checkcount--) { udelay(50); rio_read_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_MNT_RSP_CSR(pnum), + RIO_DEV_PORT_N_MNT_RSP_CSR(rdev, pnum), ®val); if (regval & RIO_PORT_N_MNT_RSP_RVAL) { *lnkresp = regval; @@ -1076,6 +1080,13 @@ rio_get_input_status(struct rio_dev *rdev, int pnum, u32 *lnkresp) * @rdev: Pointer to RIO device control structure * @pnum: Switch port number to clear errors * @err_status: port error status (if 0 reads register from device) + * + * TODO: Currently this routine is not compatible with recovery process + * specified for idt_gen3 RapidIO switch devices. It has to be reviewed + * to implement universal recovery process that is compatible full range + * off available devices. + * IDT gen3 switch driver now implements HW-specific error handler that + * issues soft port reset to the port to reset ERR_STOP bits and ackIDs. */ static int rio_clr_err_stopped(struct rio_dev *rdev, u32 pnum, u32 err_status) { @@ -1085,10 +1096,10 @@ static int rio_clr_err_stopped(struct rio_dev *rdev, u32 pnum, u32 err_status) if (err_status == 0) rio_read_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_ERR_STS_CSR(pnum), + RIO_DEV_PORT_N_ERR_STS_CSR(rdev, pnum), &err_status); - if (err_status & RIO_PORT_N_ERR_STS_PW_OUT_ES) { + if (err_status & RIO_PORT_N_ERR_STS_OUT_ES) { pr_debug("RIO_EM: servicing Output Error-Stopped state\n"); /* * Send a Link-Request/Input-Status control symbol @@ -1103,7 +1114,7 @@ static int rio_clr_err_stopped(struct rio_dev *rdev, u32 pnum, u32 err_status) far_ackid = (regval & RIO_PORT_N_MNT_RSP_ASTAT) >> 5; far_linkstat = regval & RIO_PORT_N_MNT_RSP_LSTAT; rio_read_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_ACK_STS_CSR(pnum), + RIO_DEV_PORT_N_ACK_STS_CSR(rdev, pnum), ®val); pr_debug("RIO_EM: SP%d_ACK_STS_CSR=0x%08x\n", pnum, regval); near_ackid = (regval & RIO_PORT_N_ACK_INBOUND) >> 24; @@ -1121,43 +1132,43 @@ static int rio_clr_err_stopped(struct rio_dev *rdev, u32 pnum, u32 err_status) * far inbound. */ rio_write_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_ACK_STS_CSR(pnum), + RIO_DEV_PORT_N_ACK_STS_CSR(rdev, pnum), (near_ackid << 24) | (far_ackid << 8) | far_ackid); /* Align far outstanding/outbound ackIDs with * near inbound. */ far_ackid++; - if (nextdev) - rio_write_config_32(nextdev, - nextdev->phys_efptr + - RIO_PORT_N_ACK_STS_CSR(RIO_GET_PORT_NUM(nextdev->swpinfo)), - (far_ackid << 24) | - (near_ackid << 8) | near_ackid); - else - pr_debug("RIO_EM: Invalid nextdev pointer (NULL)\n"); + if (!nextdev) { + pr_debug("RIO_EM: nextdev pointer == NULL\n"); + goto rd_err; + } + + rio_write_config_32(nextdev, + RIO_DEV_PORT_N_ACK_STS_CSR(nextdev, + RIO_GET_PORT_NUM(nextdev->swpinfo)), + (far_ackid << 24) | + (near_ackid << 8) | near_ackid); } rd_err: - rio_read_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_ERR_STS_CSR(pnum), - &err_status); + rio_read_config_32(rdev, RIO_DEV_PORT_N_ERR_STS_CSR(rdev, pnum), + &err_status); pr_debug("RIO_EM: SP%d_ERR_STS_CSR=0x%08x\n", pnum, err_status); } - if ((err_status & RIO_PORT_N_ERR_STS_PW_INP_ES) && nextdev) { + if ((err_status & RIO_PORT_N_ERR_STS_INP_ES) && nextdev) { pr_debug("RIO_EM: servicing Input Error-Stopped state\n"); rio_get_input_status(nextdev, RIO_GET_PORT_NUM(nextdev->swpinfo), NULL); udelay(50); - rio_read_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_ERR_STS_CSR(pnum), - &err_status); + rio_read_config_32(rdev, RIO_DEV_PORT_N_ERR_STS_CSR(rdev, pnum), + &err_status); pr_debug("RIO_EM: SP%d_ERR_STS_CSR=0x%08x\n", pnum, err_status); } - return (err_status & (RIO_PORT_N_ERR_STS_PW_OUT_ES | - RIO_PORT_N_ERR_STS_PW_INP_ES)) ? 1 : 0; + return (err_status & (RIO_PORT_N_ERR_STS_OUT_ES | + RIO_PORT_N_ERR_STS_INP_ES)) ? 1 : 0; } /** @@ -1257,9 +1268,8 @@ int rio_inb_pwrite_handler(struct rio_mport *mport, union rio_pw_msg *pw_msg) if (rdev->rswitch->ops && rdev->rswitch->ops->em_handle) rdev->rswitch->ops->em_handle(rdev, portnum); - rio_read_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_ERR_STS_CSR(portnum), - &err_status); + rio_read_config_32(rdev, RIO_DEV_PORT_N_ERR_STS_CSR(rdev, portnum), + &err_status); pr_debug("RIO_PW: SP%d_ERR_STS_CSR=0x%08x\n", portnum, err_status); if (err_status & RIO_PORT_N_ERR_STS_PORT_OK) { @@ -1276,8 +1286,8 @@ int rio_inb_pwrite_handler(struct rio_mport *mport, union rio_pw_msg *pw_msg) * Depending on the link partner state, two attempts * may be needed for successful recovery. */ - if (err_status & (RIO_PORT_N_ERR_STS_PW_OUT_ES | - RIO_PORT_N_ERR_STS_PW_INP_ES)) { + if (err_status & (RIO_PORT_N_ERR_STS_OUT_ES | + RIO_PORT_N_ERR_STS_INP_ES)) { if (rio_clr_err_stopped(rdev, portnum, err_status)) rio_clr_err_stopped(rdev, portnum, 0); } @@ -1287,10 +1297,18 @@ int rio_inb_pwrite_handler(struct rio_mport *mport, union rio_pw_msg *pw_msg) rdev->rswitch->port_ok &= ~(1 << portnum); rio_set_port_lockout(rdev, portnum, 1); + if (rdev->phys_rmap == 1) { rio_write_config_32(rdev, - rdev->phys_efptr + - RIO_PORT_N_ACK_STS_CSR(portnum), + RIO_DEV_PORT_N_ACK_STS_CSR(rdev, portnum), RIO_PORT_N_ACK_CLEAR); + } else { + rio_write_config_32(rdev, + RIO_DEV_PORT_N_OB_ACK_CSR(rdev, portnum), + RIO_PORT_N_OB_ACK_CLEAR); + rio_write_config_32(rdev, + RIO_DEV_PORT_N_IB_ACK_CSR(rdev, portnum), + 0); + } /* Schedule Extraction Service */ pr_debug("RIO_PW: Device Extraction on [%s]-P%d\n", @@ -1319,9 +1337,8 @@ int rio_inb_pwrite_handler(struct rio_mport *mport, union rio_pw_msg *pw_msg) } /* Clear remaining error bits and Port-Write Pending bit */ - rio_write_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_ERR_STS_CSR(portnum), - err_status); + rio_write_config_32(rdev, RIO_DEV_PORT_N_ERR_STS_CSR(rdev, portnum), + err_status); return 0; } @@ -1372,20 +1389,7 @@ EXPORT_SYMBOL_GPL(rio_mport_get_efb); * Tell if a device supports a given RapidIO capability. * Returns the offset of the requested extended feature * block within the device's RIO configuration space or - * 0 in case the device does not support it. Possible - * values for @ftr: - * - * %RIO_EFB_PAR_EP_ID LP/LVDS EP Devices - * - * %RIO_EFB_PAR_EP_REC_ID LP/LVDS EP Recovery Devices - * - * %RIO_EFB_PAR_EP_FREE_ID LP/LVDS EP Free Devices - * - * %RIO_EFB_SER_EP_ID LP/Serial EP Devices - * - * %RIO_EFB_SER_EP_REC_ID LP/Serial EP Recovery Devices - * - * %RIO_EFB_SER_EP_FREE_ID LP/Serial EP Free Devices + * 0 in case the device does not support it. */ u32 rio_mport_get_feature(struct rio_mport * port, int local, u16 destid, diff --git a/drivers/rapidio/rio.h b/drivers/rapidio/rio.h index 625d09a..9796b3f 100644 --- a/drivers/rapidio/rio.h +++ b/drivers/rapidio/rio.h @@ -22,7 +22,7 @@ extern u32 rio_mport_get_feature(struct rio_mport *mport, int local, u16 destid, u8 hopcount, int ftr); extern u32 rio_mport_get_physefb(struct rio_mport *port, int local, - u16 destid, u8 hopcount); + u16 destid, u8 hopcount, u32 *rmap); extern u32 rio_mport_get_efb(struct rio_mport *port, int local, u16 destid, u8 hopcount, u32 from); extern int rio_mport_chk_dev_access(struct rio_mport *mport, u16 destid, diff --git a/drivers/rapidio/switches/tsi57x.c b/drivers/rapidio/switches/tsi57x.c index 42c8b01..2700d15 100644 --- a/drivers/rapidio/switches/tsi57x.c +++ b/drivers/rapidio/switches/tsi57x.c @@ -175,12 +175,10 @@ tsi57x_em_init(struct rio_dev *rdev) /* Clear all pending interrupts */ rio_read_config_32(rdev, - rdev->phys_efptr + - RIO_PORT_N_ERR_STS_CSR(portnum), + RIO_DEV_PORT_N_ERR_STS_CSR(rdev, portnum), ®val); rio_write_config_32(rdev, - rdev->phys_efptr + - RIO_PORT_N_ERR_STS_CSR(portnum), + RIO_DEV_PORT_N_ERR_STS_CSR(rdev, portnum), regval & 0x07120214); rio_read_config_32(rdev, @@ -198,7 +196,7 @@ tsi57x_em_init(struct rio_dev *rdev) /* Skip next (odd) port if the current port is in x4 mode */ rio_read_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_CTL_CSR(portnum), + RIO_DEV_PORT_N_CTL_CSR(rdev, portnum), ®val); if ((regval & RIO_PORT_N_CTL_PWIDTH) == RIO_PORT_N_CTL_PWIDTH_4) portnum++; @@ -221,23 +219,23 @@ tsi57x_em_handler(struct rio_dev *rdev, u8 portnum) u32 regval; rio_read_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_ERR_STS_CSR(portnum), + RIO_DEV_PORT_N_ERR_STS_CSR(rdev, portnum), &err_status); if ((err_status & RIO_PORT_N_ERR_STS_PORT_OK) && - (err_status & (RIO_PORT_N_ERR_STS_PW_OUT_ES | - RIO_PORT_N_ERR_STS_PW_INP_ES))) { + (err_status & (RIO_PORT_N_ERR_STS_OUT_ES | + RIO_PORT_N_ERR_STS_INP_ES))) { /* Remove any queued packets by locking/unlocking port */ rio_read_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_CTL_CSR(portnum), + RIO_DEV_PORT_N_CTL_CSR(rdev, portnum), ®val); if (!(regval & RIO_PORT_N_CTL_LOCKOUT)) { rio_write_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_CTL_CSR(portnum), + RIO_DEV_PORT_N_CTL_CSR(rdev, portnum), regval | RIO_PORT_N_CTL_LOCKOUT); udelay(50); rio_write_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_CTL_CSR(portnum), + RIO_DEV_PORT_N_CTL_CSR(rdev, portnum), regval); } @@ -245,7 +243,7 @@ tsi57x_em_handler(struct rio_dev *rdev, u8 portnum) * valid bit */ rio_read_config_32(rdev, - rdev->phys_efptr + RIO_PORT_N_MNT_RSP_CSR(portnum), + RIO_DEV_PORT_N_MNT_RSP_CSR(rdev, portnum), ®val); /* Send a Packet-Not-Accepted/Link-Request-Input-Status control @@ -259,8 +257,8 @@ tsi57x_em_handler(struct rio_dev *rdev, u8 portnum) while (checkcount--) { udelay(50); rio_read_config_32(rdev, - rdev->phys_efptr + - RIO_PORT_N_MNT_RSP_CSR(portnum), + RIO_DEV_PORT_N_MNT_RSP_CSR(rdev, + portnum), ®val); if (regval & RIO_PORT_N_MNT_RSP_RVAL) goto exit_es; diff --git a/include/linux/rio.h b/include/linux/rio.h index f7ec35b..37b95c4 100644 --- a/include/linux/rio.h +++ b/include/linux/rio.h @@ -163,6 +163,7 @@ enum rio_device_state { * @dst_ops: Destination operation capabilities * @comp_tag: RIO component tag * @phys_efptr: RIO device extended features pointer + * @phys_rmap: LP-Serial Register Map Type (1 or 2) * @em_efptr: RIO Error Management features pointer * @dma_mask: Mask of bits of RIO address this device implements * @driver: Driver claiming this device @@ -193,6 +194,7 @@ struct rio_dev { u32 dst_ops; u32 comp_tag; u32 phys_efptr; + u32 phys_rmap; u32 em_efptr; u64 dma_mask; struct rio_driver *driver; /* RIO driver claiming this device */ @@ -237,11 +239,6 @@ struct rio_dbell { void *dev_id; }; -enum rio_phy_type { - RIO_PHY_PARALLEL, - RIO_PHY_SERIAL, -}; - /** * struct rio_mport - RIO master port info * @dbells: List of doorbell events @@ -259,8 +256,8 @@ enum rio_phy_type { * @id: Port ID, unique among all ports * @index: Port index, unique among all port interfaces of the same type * @sys_size: RapidIO common transport system size - * @phy_type: RapidIO phy type * @phys_efptr: RIO port extended features pointer + * @phys_rmap: LP-Serial EFB Register Mapping type (1 or 2). * @name: Port name string * @dev: device structure associated with an mport * @priv: Master port private data @@ -289,8 +286,8 @@ struct rio_mport { * 0 - Small size. 256 devices. * 1 - Large size, 65536 devices. */ - enum rio_phy_type phy_type; /* RapidIO phy type */ u32 phys_efptr; + u32 phys_rmap; unsigned char name[RIO_MAX_MPORT_NAME]; struct device dev; void *priv; /* Master port private data */ diff --git a/include/linux/rio_regs.h b/include/linux/rio_regs.h index 1063ae3..40c04ef 100644 --- a/include/linux/rio_regs.h +++ b/include/linux/rio_regs.h @@ -42,9 +42,11 @@ #define RIO_PEF_INB_MBOX2 0x00200000 /* [II, <= 1.2] Mailbox 2 */ #define RIO_PEF_INB_MBOX3 0x00100000 /* [II, <= 1.2] Mailbox 3 */ #define RIO_PEF_INB_DOORBELL 0x00080000 /* [II, <= 1.2] Doorbells */ +#define RIO_PEF_DEV32 0x00001000 /* [III] PE supports Common TRansport Dev32 */ #define RIO_PEF_EXT_RT 0x00000200 /* [III, 1.3] Extended route table support */ #define RIO_PEF_STD_RT 0x00000100 /* [III, 1.3] Standard route table support */ -#define RIO_PEF_CTLS 0x00000010 /* [III] CTLS */ +#define RIO_PEF_CTLS 0x00000010 /* [III] Common Transport Large System (< rev.3) */ +#define RIO_PEF_DEV16 0x00000010 /* [III] PE Supports Common Transport Dev16 (rev.3) */ #define RIO_PEF_EXT_FEATURES 0x00000008 /* [I] EFT_PTR valid */ #define RIO_PEF_ADDR_66 0x00000004 /* [I] 66 bits */ #define RIO_PEF_ADDR_50 0x00000002 /* [I] 50 bits */ @@ -194,70 +196,101 @@ #define RIO_GET_BLOCK_ID(x) (x & RIO_EFB_ID_MASK) /* Extended Feature Block IDs */ -#define RIO_EFB_PAR_EP_ID 0x0001 /* [IV] LP/LVDS EP Devices */ -#define RIO_EFB_PAR_EP_REC_ID 0x0002 /* [IV] LP/LVDS EP Recovery Devices */ -#define RIO_EFB_PAR_EP_FREE_ID 0x0003 /* [IV] LP/LVDS EP Free Devices */ -#define RIO_EFB_SER_EP_ID_V13P 0x0001 /* [VI] LP/Serial EP Devices, RapidIO Spec ver 1.3 and above */ -#define RIO_EFB_SER_EP_REC_ID_V13P 0x0002 /* [VI] LP/Serial EP Recovery Devices, RapidIO Spec ver 1.3 and above */ -#define RIO_EFB_SER_EP_FREE_ID_V13P 0x0003 /* [VI] LP/Serial EP Free Devices, RapidIO Spec ver 1.3 and above */ -#define RIO_EFB_SER_EP_ID 0x0004 /* [VI] LP/Serial EP Devices */ -#define RIO_EFB_SER_EP_REC_ID 0x0005 /* [VI] LP/Serial EP Recovery Devices */ -#define RIO_EFB_SER_EP_FREE_ID 0x0006 /* [VI] LP/Serial EP Free Devices */ -#define RIO_EFB_SER_EP_FREC_ID 0x0009 /* [VI] LP/Serial EP Free Recovery Devices */ +#define RIO_EFB_SER_EP_M1_ID 0x0001 /* [VI] LP-Serial EP Devices, Map I */ +#define RIO_EFB_SER_EP_SW_M1_ID 0x0002 /* [VI] LP-Serial EP w SW Recovery Devices, Map I */ +#define RIO_EFB_SER_EPF_M1_ID 0x0003 /* [VI] LP-Serial EP Free Devices, Map I */ +#define RIO_EFB_SER_EP_ID 0x0004 /* [VI] LP-Serial EP Devices, RIO 1.2 */ +#define RIO_EFB_SER_EP_REC_ID 0x0005 /* [VI] LP-Serial EP w SW Recovery Devices, RIO 1.2 */ +#define RIO_EFB_SER_EP_FREE_ID 0x0006 /* [VI] LP-Serial EP Free Devices, RIO 1.2 */ #define RIO_EFB_ERR_MGMNT 0x0007 /* [VIII] Error Management Extensions */ +#define RIO_EFB_SER_EPF_SW_M1_ID 0x0009 /* [VI] LP-Serial EP Free w SW Recovery Devices, Map I */ +#define RIO_EFB_SW_ROUTING_TBL 0x000E /* [III] Switch Routing Table Block */ +#define RIO_EFB_SER_EP_M2_ID 0x0011 /* [VI] LP-Serial EP Devices, Map II */ +#define RIO_EFB_SER_EP_SW_M2_ID 0x0012 /* [VI] LP-Serial EP w SW Recovery Devices, Map II */ +#define RIO_EFB_SER_EPF_M2_ID 0x0013 /* [VI] LP-Serial EP Free Devices, Map II */ +#define RIO_EFB_ERR_MGMNT_HS 0x0017 /* [VIII] Error Management Extensions, Hot-Swap only */ +#define RIO_EFB_SER_EPF_SW_M2_ID 0x0019 /* [VI] LP-Serial EP Free w SW Recovery Devices, Map II */ /* - * Physical 8/16 LP-LVDS - * ID=0x0001, Generic End Point Devices - * ID=0x0002, Generic End Point Devices, software assisted recovery option - * ID=0x0003, Generic End Point Free Devices - * - * Physical LP-Serial - * ID=0x0004, Generic End Point Devices - * ID=0x0005, Generic End Point Devices, software assisted recovery option - * ID=0x0006, Generic End Point Free Devices + * Physical LP-Serial Registers Definitions + * Parameters in register macros: + * n - port number, m - Register Map Type (1 or 2) */ #define RIO_PORT_MNT_HEADER 0x0000 #define RIO_PORT_REQ_CTL_CSR 0x0020 -#define RIO_PORT_RSP_CTL_CSR 0x0024 /* 0x0001/0x0002 */ -#define RIO_PORT_LINKTO_CTL_CSR 0x0020 /* Serial */ -#define RIO_PORT_RSPTO_CTL_CSR 0x0024 /* Serial */ +#define RIO_PORT_RSP_CTL_CSR 0x0024 +#define RIO_PORT_LINKTO_CTL_CSR 0x0020 +#define RIO_PORT_RSPTO_CTL_CSR 0x0024 #define RIO_PORT_GEN_CTL_CSR 0x003c #define RIO_PORT_GEN_HOST 0x80000000 #define RIO_PORT_GEN_MASTER 0x40000000 #define RIO_PORT_GEN_DISCOVERED 0x20000000 -#define RIO_PORT_N_MNT_REQ_CSR(x) (0x0040 + x*0x20) /* 0x0002 */ +#define RIO_PORT_N_MNT_REQ_CSR(n, m) (0x40 + (n) * (0x20 * (m))) #define RIO_MNT_REQ_CMD_RD 0x03 /* Reset-device command */ #define RIO_MNT_REQ_CMD_IS 0x04 /* Input-status command */ -#define RIO_PORT_N_MNT_RSP_CSR(x) (0x0044 + x*0x20) /* 0x0002 */ +#define RIO_PORT_N_MNT_RSP_CSR(n, m) (0x44 + (n) * (0x20 * (m))) #define RIO_PORT_N_MNT_RSP_RVAL 0x80000000 /* Response Valid */ #define RIO_PORT_N_MNT_RSP_ASTAT 0x000007e0 /* ackID Status */ #define RIO_PORT_N_MNT_RSP_LSTAT 0x0000001f /* Link Status */ -#define RIO_PORT_N_ACK_STS_CSR(x) (0x0048 + x*0x20) /* 0x0002 */ +#define RIO_PORT_N_ACK_STS_CSR(n) (0x48 + (n) * 0x20) /* Only in RM-I */ #define RIO_PORT_N_ACK_CLEAR 0x80000000 #define RIO_PORT_N_ACK_INBOUND 0x3f000000 #define RIO_PORT_N_ACK_OUTSTAND 0x00003f00 #define RIO_PORT_N_ACK_OUTBOUND 0x0000003f -#define RIO_PORT_N_CTL2_CSR(x) (0x0054 + x*0x20) +#define RIO_PORT_N_CTL2_CSR(n, m) (0x54 + (n) * (0x20 * (m))) #define RIO_PORT_N_CTL2_SEL_BAUD 0xf0000000 -#define RIO_PORT_N_ERR_STS_CSR(x) (0x0058 + x*0x20) -#define RIO_PORT_N_ERR_STS_PW_OUT_ES 0x00010000 /* Output Error-stopped */ -#define RIO_PORT_N_ERR_STS_PW_INP_ES 0x00000100 /* Input Error-stopped */ +#define RIO_PORT_N_ERR_STS_CSR(n, m) (0x58 + (n) * (0x20 * (m))) +#define RIO_PORT_N_ERR_STS_OUT_ES 0x00010000 /* Output Error-stopped */ +#define RIO_PORT_N_ERR_STS_INP_ES 0x00000100 /* Input Error-stopped */ #define RIO_PORT_N_ERR_STS_PW_PEND 0x00000010 /* Port-Write Pending */ +#define RIO_PORT_N_ERR_STS_PORT_UA 0x00000008 /* Port Unavailable */ #define RIO_PORT_N_ERR_STS_PORT_ERR 0x00000004 #define RIO_PORT_N_ERR_STS_PORT_OK 0x00000002 #define RIO_PORT_N_ERR_STS_PORT_UNINIT 0x00000001 -#define RIO_PORT_N_CTL_CSR(x) (0x005c + x*0x20) +#define RIO_PORT_N_CTL_CSR(n, m) (0x5c + (n) * (0x20 * (m))) #define RIO_PORT_N_CTL_PWIDTH 0xc0000000 #define RIO_PORT_N_CTL_PWIDTH_1 0x00000000 #define RIO_PORT_N_CTL_PWIDTH_4 0x40000000 #define RIO_PORT_N_CTL_IPW 0x38000000 /* Initialized Port Width */ #define RIO_PORT_N_CTL_P_TYP_SER 0x00000001 #define RIO_PORT_N_CTL_LOCKOUT 0x00000002 -#define RIO_PORT_N_CTL_EN_RX_SER 0x00200000 -#define RIO_PORT_N_CTL_EN_TX_SER 0x00400000 -#define RIO_PORT_N_CTL_EN_RX_PAR 0x08000000 -#define RIO_PORT_N_CTL_EN_TX_PAR 0x40000000 +#define RIO_PORT_N_CTL_EN_RX 0x00200000 +#define RIO_PORT_N_CTL_EN_TX 0x00400000 +#define RIO_PORT_N_OB_ACK_CSR(n) (0x60 + (n) * 0x40) /* Only in RM-II */ +#define RIO_PORT_N_OB_ACK_CLEAR 0x80000000 +#define RIO_PORT_N_OB_ACK_OUTSTD 0x00fff000 +#define RIO_PORT_N_OB_ACK_OUTBND 0x00000fff +#define RIO_PORT_N_IB_ACK_CSR(n) (0x64 + (n) * 0x40) /* Only in RM-II */ +#define RIO_PORT_N_IB_ACK_INBND 0x00000fff + +/* + * Device-based helper macros for serial port register access. + * d - pointer to rapidio device object, n - port number + */ + +#define RIO_DEV_PORT_N_MNT_REQ_CSR(d, n) \ + (d->phys_efptr + RIO_PORT_N_MNT_REQ_CSR(n, d->phys_rmap)) + +#define RIO_DEV_PORT_N_MNT_RSP_CSR(d, n) \ + (d->phys_efptr + RIO_PORT_N_MNT_RSP_CSR(n, d->phys_rmap)) + +#define RIO_DEV_PORT_N_ACK_STS_CSR(d, n) \ + (d->phys_efptr + RIO_PORT_N_ACK_STS_CSR(n)) + +#define RIO_DEV_PORT_N_CTL2_CSR(d, n) \ + (d->phys_efptr + RIO_PORT_N_CTL2_CSR(n, d->phys_rmap)) + +#define RIO_DEV_PORT_N_ERR_STS_CSR(d, n) \ + (d->phys_efptr + RIO_PORT_N_ERR_STS_CSR(n, d->phys_rmap)) + +#define RIO_DEV_PORT_N_CTL_CSR(d, n) \ + (d->phys_efptr + RIO_PORT_N_CTL_CSR(n, d->phys_rmap)) + +#define RIO_DEV_PORT_N_OB_ACK_CSR(d, n) \ + (d->phys_efptr + RIO_PORT_N_OB_ACK_CSR(n)) + +#define RIO_DEV_PORT_N_IB_ACK_CSR(d, n) \ + (d->phys_efptr + RIO_PORT_N_IB_ACK_CSR(n)) /* * Error Management Extensions (RapidIO 1.3+, Part 8) @@ -268,6 +301,7 @@ /* General EM Registers (Common for all Ports) */ #define RIO_EM_EFB_HEADER 0x000 /* Error Management Extensions Block Header */ +#define RIO_EM_EMHS_CAR 0x004 /* EM Functionality CAR */ #define RIO_EM_LTL_ERR_DETECT 0x008 /* Logical/Transport Layer Error Detect CSR */ #define RIO_EM_LTL_ERR_EN 0x00c /* Logical/Transport Layer Error Enable CSR */ #define REM_LTL_ERR_ILLTRAN 0x08000000 /* Illegal Transaction decode */ @@ -278,15 +312,33 @@ #define RIO_EM_LTL_ADDR_CAP 0x014 /* Logical/Transport Layer Address Capture CSR */ #define RIO_EM_LTL_DEVID_CAP 0x018 /* Logical/Transport Layer Device ID Capture CSR */ #define RIO_EM_LTL_CTRL_CAP 0x01c /* Logical/Transport Layer Control Capture CSR */ +#define RIO_EM_LTL_DID32_CAP 0x020 /* Logical/Transport Layer Dev32 DestID Capture CSR */ +#define RIO_EM_LTL_SID32_CAP 0x024 /* Logical/Transport Layer Dev32 source ID Capture CSR */ #define RIO_EM_PW_TGT_DEVID 0x028 /* Port-write Target deviceID CSR */ +#define RIO_EM_PW_TGT_DEVID_D16M 0xff000000 /* Port-write Target DID16 MSB */ +#define RIO_EM_PW_TGT_DEVID_D8 0x00ff0000 /* Port-write Target DID16 LSB or DID8 */ +#define RIO_EM_PW_TGT_DEVID_DEV16 0x00008000 /* Port-write Target DID16 LSB or DID8 */ +#define RIO_EM_PW_TGT_DEVID_DEV32 0x00004000 /* Port-write Target DID16 LSB or DID8 */ #define RIO_EM_PKT_TTL 0x02c /* Packet Time-to-live CSR */ +#define RIO_EM_PKT_TTL_VAL 0xffff0000 /* Packet Time-to-live value */ +#define RIO_EM_PW_TGT32_DEVID 0x030 /* Port-write Dev32 Target deviceID CSR */ +#define RIO_EM_PW_TX_CTRL 0x034 /* Port-write Transmission Control CSR */ +#define RIO_EM_PW_TX_CTRL_PW_DIS 0x00000001 /* Port-write Transmission Disable bit */ /* Per-Port EM Registers */ #define RIO_EM_PN_ERR_DETECT(x) (0x040 + x*0x40) /* Port N Error Detect CSR */ #define REM_PED_IMPL_SPEC 0x80000000 +#define REM_PED_LINK_OK2U 0x40000000 /* Link OK to Uninit transition */ +#define REM_PED_LINK_UPDA 0x20000000 /* Link Uninit Packet Discard Active */ +#define REM_PED_LINK_U2OK 0x10000000 /* Link Uninit to OK transition */ #define REM_PED_LINK_TO 0x00000001 + #define RIO_EM_PN_ERRRATE_EN(x) (0x044 + x*0x40) /* Port N Error Rate Enable CSR */ +#define RIO_EM_PN_ERRRATE_EN_OK2U 0x40000000 /* Enable notification for OK2U */ +#define RIO_EM_PN_ERRRATE_EN_UPDA 0x20000000 /* Enable notification for UPDA */ +#define RIO_EM_PN_ERRRATE_EN_U2OK 0x10000000 /* Enable notification for U2OK */ + #define RIO_EM_PN_ATTRIB_CAP(x) (0x048 + x*0x40) /* Port N Attributes Capture CSR */ #define RIO_EM_PN_PKT_CAP_0(x) (0x04c + x*0x40) /* Port N Packet/Control Symbol Capture 0 CSR */ #define RIO_EM_PN_PKT_CAP_1(x) (0x050 + x*0x40) /* Port N Packet Capture 1 CSR */ @@ -294,5 +346,50 @@ #define RIO_EM_PN_PKT_CAP_3(x) (0x058 + x*0x40) /* Port N Packet Capture 3 CSR */ #define RIO_EM_PN_ERRRATE(x) (0x068 + x*0x40) /* Port N Error Rate CSR */ #define RIO_EM_PN_ERRRATE_TR(x) (0x06c + x*0x40) /* Port N Error Rate Threshold CSR */ +#define RIO_EM_PN_LINK_UDT(x) (0x070 + x*0x40) /* Port N Link Uninit Discard Timer CSR */ +#define RIO_EM_PN_LINK_UDT_TO 0xffffff00 /* Link Uninit Timeout value */ + +/* + * Switch Routing Table Register Block ID=0x000E (RapidIO 3.0+, part 3) + * Register offsets are defined from beginning of the block. + */ + +/* Broadcast Routing Table Control CSR */ +#define RIO_BC_RT_CTL_CSR 0x020 +#define RIO_RT_CTL_THREE_LVL 0x80000000 +#define RIO_RT_CTL_DEV32_RT_CTRL 0x40000000 +#define RIO_RT_CTL_MC_MASK_SZ 0x03000000 /* 3.0+ Part 11: Multicast */ + +/* Broadcast Level 0 Info CSR */ +#define RIO_BC_RT_LVL0_INFO_CSR 0x030 +#define RIO_RT_L0I_NUM_GR 0xff000000 +#define RIO_RT_L0I_GR_PTR 0x00fffc00 + +/* Broadcast Level 1 Info CSR */ +#define RIO_BC_RT_LVL1_INFO_CSR 0x034 +#define RIO_RT_L1I_NUM_GR 0xff000000 +#define RIO_RT_L1I_GR_PTR 0x00fffc00 + +/* Broadcast Level 2 Info CSR */ +#define RIO_BC_RT_LVL2_INFO_CSR 0x038 +#define RIO_RT_L2I_NUM_GR 0xff000000 +#define RIO_RT_L2I_GR_PTR 0x00fffc00 + +/* Per-Port Routing Table registers. + * Register fields defined in the broadcast section above are + * applicable to the corresponding registers below. + */ +#define RIO_SPx_RT_CTL_CSR(x) (0x040 + (0x20 * x)) +#define RIO_SPx_RT_LVL0_INFO_CSR(x) (0x50 + (0x20 * x)) +#define RIO_SPx_RT_LVL1_INFO_CSR(x) (0x54 + (0x20 * x)) +#define RIO_SPx_RT_LVL2_INFO_CSR(x) (0x58 + (0x20 * x)) + +/* Register Formats for Routing Table Group entry. + * Register offsets are calculated using GR_PTR field in the corresponding + * table Level_N and group/entry numbers (see RapidIO 3.0+ Part 3). + */ +#define RIO_RT_Ln_ENTRY_IMPL_DEF 0xf0000000 +#define RIO_RT_Ln_ENTRY_RTE_VAL 0x000003ff +#define RIO_RT_ENTRY_DROP_PKT 0x300 #endif /* LINUX_RIO_REGS_H */ -- cgit v0.10.2 From adff1649e6d66d9dda7631701eb98e8482edaff6 Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:07:00 -0700 Subject: powerpc/fsl_rio: apply changes for RIO spec rev 3 - Remove check for parallel PHY - Set LP-Serial Register Map type [akpm@linux-foundation.org: fix build] [alexandre.bounine@idt.com: fix build fix] Link: http://lkml.kernel.org/r/20160802184932.2755-1-alexandre.bounine@idt.com Link: http://lkml.kernel.org/r/1469125134-16523-13-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Cc: Matt Porter Cc: Benjamin Herrenschmidt Cc: Michael Ellerman Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/powerpc/sysdev/fsl_rio.c b/arch/powerpc/sysdev/fsl_rio.c index 386790c..984e816 100644 --- a/arch/powerpc/sysdev/fsl_rio.c +++ b/arch/powerpc/sysdev/fsl_rio.c @@ -643,19 +643,11 @@ int fsl_rio_setup(struct platform_device *dev) port->ops = ops; port->priv = priv; port->phys_efptr = 0x100; + port->phys_rmap = 1; priv->regs_win = rio_regs_win; - /* Probe the master port phy type */ ccsr = in_be32(priv->regs_win + RIO_CCSR + i*0x20); - port->phy_type = (ccsr & 1) ? RIO_PHY_SERIAL : RIO_PHY_PARALLEL; - if (port->phy_type == RIO_PHY_PARALLEL) { - dev_err(&dev->dev, "RIO: Parallel PHY type, unsupported port type!\n"); - release_resource(&port->iores); - kfree(priv); - kfree(port); - continue; - } - dev_info(&dev->dev, "RapidIO PHY type: Serial\n"); + /* Checking the port training status */ if (in_be32((priv->regs_win + RIO_ESCSR + i*0x20)) & 1) { dev_err(&dev->dev, "Port %d is not ready. " @@ -705,11 +697,9 @@ int fsl_rio_setup(struct platform_device *dev) ((i == 0) ? RIO_INB_ATMU_REGS_PORT1_OFFSET : RIO_INB_ATMU_REGS_PORT2_OFFSET)); - - /* Set to receive any dist ID for serial RapidIO controller. */ - if (port->phy_type == RIO_PHY_SERIAL) - out_be32((priv->regs_win - + RIO_ISR_AACR + i*0x80), RIO_ISR_AACR_AA); + /* Set to receive packets with any dest ID */ + out_be32((priv->regs_win + RIO_ISR_AACR + i*0x80), + RIO_ISR_AACR_AA); /* Configure maintenance transaction window */ out_be32(&priv->maint_atmu_regs->rowbar, -- cgit v0.10.2 From 0b9364b5cf11c6e504f4b77e24b15a0dc8a82df0 Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Tue, 2 Aug 2016 14:07:03 -0700 Subject: rapidio/switches: add driver for IDT gen3 switches Add RapidIO switch driver for IDT Gen3 switch devices: RXS1632 and RXS2448. [alexandre.bounine@idt.com: fixup for original driver patch] Link: http://lkml.kernel.org/r/1469137596-18241-1-git-send-email-alexandre.bounine@idt.com Link: http://lkml.kernel.org/r/1469125134-16523-14-git-send-email-alexandre.bounine@idt.com Signed-off-by: Alexandre Bounine Tested-by: Barry Wood Cc: Matt Porter Cc: Andre van Herk Cc: Barry Wood Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rapidio/switches/Kconfig b/drivers/rapidio/switches/Kconfig index 3458415..92767fd 100644 --- a/drivers/rapidio/switches/Kconfig +++ b/drivers/rapidio/switches/Kconfig @@ -22,3 +22,9 @@ config RAPIDIO_CPS_GEN2 default n ---help--- Includes support for ITD CPS Gen.2 serial RapidIO switches. + +config RAPIDIO_RXS_GEN3 + tristate "IDT RXS Gen.3 SRIO switch support" + default n + ---help--- + Includes support for ITD RXS Gen.3 serial RapidIO switches. diff --git a/drivers/rapidio/switches/Makefile b/drivers/rapidio/switches/Makefile index 051cc6b..6bdd54c 100644 --- a/drivers/rapidio/switches/Makefile +++ b/drivers/rapidio/switches/Makefile @@ -6,3 +6,4 @@ obj-$(CONFIG_RAPIDIO_TSI57X) += tsi57x.o obj-$(CONFIG_RAPIDIO_CPS_XX) += idtcps.o obj-$(CONFIG_RAPIDIO_TSI568) += tsi568.o obj-$(CONFIG_RAPIDIO_CPS_GEN2) += idt_gen2.o +obj-$(CONFIG_RAPIDIO_RXS_GEN3) += idt_gen3.o diff --git a/drivers/rapidio/switches/idt_gen3.c b/drivers/rapidio/switches/idt_gen3.c new file mode 100644 index 0000000..c5923a5 --- /dev/null +++ b/drivers/rapidio/switches/idt_gen3.c @@ -0,0 +1,382 @@ +/* + * IDT RXS Gen.3 Serial RapidIO switch family support + * + * Copyright 2016 Integrated Device Technology, Inc. + * + * 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 +#include +#include +#include + +#include +#include "../rio.h" + +#define RIO_EM_PW_STAT 0x40020 +#define RIO_PW_CTL 0x40204 +#define RIO_PW_CTL_PW_TMR 0xffffff00 +#define RIO_PW_ROUTE 0x40208 + +#define RIO_EM_DEV_INT_EN 0x40030 + +#define RIO_PLM_SPx_IMP_SPEC_CTL(x) (0x10100 + (x)*0x100) +#define RIO_PLM_SPx_IMP_SPEC_CTL_SOFT_RST 0x02000000 + +#define RIO_PLM_SPx_PW_EN(x) (0x10118 + (x)*0x100) +#define RIO_PLM_SPx_PW_EN_OK2U 0x40000000 +#define RIO_PLM_SPx_PW_EN_LINIT 0x10000000 + +#define RIO_BC_L2_Gn_ENTRYx_CSR(n, x) (0x31000 + (n)*0x400 + (x)*0x4) +#define RIO_SPx_L2_Gn_ENTRYy_CSR(x, n, y) \ + (0x51000 + (x)*0x2000 + (n)*0x400 + (y)*0x4) + +static int +idtg3_route_add_entry(struct rio_mport *mport, u16 destid, u8 hopcount, + u16 table, u16 route_destid, u8 route_port) +{ + u32 rval; + u32 entry = route_port; + int err = 0; + + pr_debug("RIO: %s t=0x%x did_%x to p_%x\n", + __func__, table, route_destid, entry); + + if (route_destid > 0xFF) + return -EINVAL; + + if (route_port == RIO_INVALID_ROUTE) + entry = RIO_RT_ENTRY_DROP_PKT; + + if (table == RIO_GLOBAL_TABLE) { + /* Use broadcast register to update all per-port tables */ + err = rio_mport_write_config_32(mport, destid, hopcount, + RIO_BC_L2_Gn_ENTRYx_CSR(0, route_destid), + entry); + return err; + } + + /* + * Verify that specified port/table number is valid + */ + err = rio_mport_read_config_32(mport, destid, hopcount, + RIO_SWP_INFO_CAR, &rval); + if (err) + return err; + + if (table >= RIO_GET_TOTAL_PORTS(rval)) + return -EINVAL; + + err = rio_mport_write_config_32(mport, destid, hopcount, + RIO_SPx_L2_Gn_ENTRYy_CSR(table, 0, route_destid), + entry); + return err; +} + +static int +idtg3_route_get_entry(struct rio_mport *mport, u16 destid, u8 hopcount, + u16 table, u16 route_destid, u8 *route_port) +{ + u32 rval; + int err; + + if (route_destid > 0xFF) + return -EINVAL; + + err = rio_mport_read_config_32(mport, destid, hopcount, + RIO_SWP_INFO_CAR, &rval); + if (err) + return err; + + /* + * This switch device does not have the dedicated global routing table. + * It is substituted by reading routing table of the ingress port of + * maintenance read requests. + */ + if (table == RIO_GLOBAL_TABLE) + table = RIO_GET_PORT_NUM(rval); + else if (table >= RIO_GET_TOTAL_PORTS(rval)) + return -EINVAL; + + err = rio_mport_read_config_32(mport, destid, hopcount, + RIO_SPx_L2_Gn_ENTRYy_CSR(table, 0, route_destid), + &rval); + if (err) + return err; + + if (rval == RIO_RT_ENTRY_DROP_PKT) + *route_port = RIO_INVALID_ROUTE; + else + *route_port = (u8)rval; + + return 0; +} + +static int +idtg3_route_clr_table(struct rio_mport *mport, u16 destid, u8 hopcount, + u16 table) +{ + u32 i; + u32 rval; + int err; + + if (table == RIO_GLOBAL_TABLE) { + for (i = 0; i <= 0xff; i++) { + err = rio_mport_write_config_32(mport, destid, hopcount, + RIO_BC_L2_Gn_ENTRYx_CSR(0, i), + RIO_RT_ENTRY_DROP_PKT); + if (err) + break; + } + + return err; + } + + err = rio_mport_read_config_32(mport, destid, hopcount, + RIO_SWP_INFO_CAR, &rval); + if (err) + return err; + + if (table >= RIO_GET_TOTAL_PORTS(rval)) + return -EINVAL; + + for (i = 0; i <= 0xff; i++) { + err = rio_mport_write_config_32(mport, destid, hopcount, + RIO_SPx_L2_Gn_ENTRYy_CSR(table, 0, i), + RIO_RT_ENTRY_DROP_PKT); + if (err) + break; + } + + return err; +} + +/* + * This routine performs device-specific initialization only. + * All standard EM configuration should be performed at upper level. + */ +static int +idtg3_em_init(struct rio_dev *rdev) +{ + int i, tmp; + u32 rval; + + pr_debug("RIO: %s [%d:%d]\n", __func__, rdev->destid, rdev->hopcount); + + /* Disable assertion of interrupt signal */ + rio_write_config_32(rdev, RIO_EM_DEV_INT_EN, 0); + + /* Disable port-write event notifications during initialization */ + rio_write_config_32(rdev, rdev->em_efptr + RIO_EM_PW_TX_CTRL, + RIO_EM_PW_TX_CTRL_PW_DIS); + + /* Configure Port-Write notifications for hot-swap events */ + tmp = RIO_GET_TOTAL_PORTS(rdev->swpinfo); + for (i = 0; i < tmp; i++) { + + rio_read_config_32(rdev, + RIO_DEV_PORT_N_ERR_STS_CSR(rdev, i), + &rval); + if (rval & RIO_PORT_N_ERR_STS_PORT_UA) + continue; + + /* Clear events signaled before enabling notification */ + rio_write_config_32(rdev, + rdev->em_efptr + RIO_EM_PN_ERR_DETECT(i), 0); + + /* Enable event notifications */ + rio_write_config_32(rdev, + rdev->em_efptr + RIO_EM_PN_ERRRATE_EN(i), + RIO_EM_PN_ERRRATE_EN_OK2U | RIO_EM_PN_ERRRATE_EN_U2OK); + /* Enable port-write generation on events */ + rio_write_config_32(rdev, RIO_PLM_SPx_PW_EN(i), + RIO_PLM_SPx_PW_EN_OK2U | RIO_PLM_SPx_PW_EN_LINIT); + + } + + /* Set Port-Write destination port */ + tmp = RIO_GET_PORT_NUM(rdev->swpinfo); + rio_write_config_32(rdev, RIO_PW_ROUTE, 1 << tmp); + + + /* Enable sending port-write event notifications */ + rio_write_config_32(rdev, rdev->em_efptr + RIO_EM_PW_TX_CTRL, 0); + + /* set TVAL = ~50us */ + rio_write_config_32(rdev, + rdev->phys_efptr + RIO_PORT_LINKTO_CTL_CSR, 0x8e << 8); + return 0; +} + + +/* + * idtg3_em_handler - device-specific error handler + * + * If the link is down (PORT_UNINIT) does nothing - this is considered + * as link partner removal from the port. + * + * If the link is up (PORT_OK) - situation is handled as *new* device insertion. + * In this case ERR_STOP bits are cleared by issuing soft reset command to the + * reporting port. Inbound and outbound ackIDs are cleared by the reset as well. + * This way the port is synchronized with freshly inserted device (assuming it + * was reset/powered-up on insertion). + * + * TODO: This is not sufficient in a situation when a link between two devices + * was down and up again (e.g. cable disconnect). For that situation full ackID + * realignment process has to be implemented. + */ +static int +idtg3_em_handler(struct rio_dev *rdev, u8 pnum) +{ + u32 err_status; + u32 rval; + + rio_read_config_32(rdev, + RIO_DEV_PORT_N_ERR_STS_CSR(rdev, pnum), + &err_status); + + /* Do nothing for device/link removal */ + if (err_status & RIO_PORT_N_ERR_STS_PORT_UNINIT) + return 0; + + /* When link is OK we have a device insertion. + * Request port soft reset to clear errors if they present. + * Inbound and outbound ackIDs will be 0 after reset. + */ + if (err_status & (RIO_PORT_N_ERR_STS_OUT_ES | + RIO_PORT_N_ERR_STS_INP_ES)) { + rio_read_config_32(rdev, RIO_PLM_SPx_IMP_SPEC_CTL(pnum), &rval); + rio_write_config_32(rdev, RIO_PLM_SPx_IMP_SPEC_CTL(pnum), + rval | RIO_PLM_SPx_IMP_SPEC_CTL_SOFT_RST); + udelay(10); + rio_write_config_32(rdev, RIO_PLM_SPx_IMP_SPEC_CTL(pnum), rval); + msleep(500); + } + + return 0; +} + +static struct rio_switch_ops idtg3_switch_ops = { + .owner = THIS_MODULE, + .add_entry = idtg3_route_add_entry, + .get_entry = idtg3_route_get_entry, + .clr_table = idtg3_route_clr_table, + .em_init = idtg3_em_init, + .em_handle = idtg3_em_handler, +}; + +static int idtg3_probe(struct rio_dev *rdev, const struct rio_device_id *id) +{ + pr_debug("RIO: %s for %s\n", __func__, rio_name(rdev)); + + spin_lock(&rdev->rswitch->lock); + + if (rdev->rswitch->ops) { + spin_unlock(&rdev->rswitch->lock); + return -EINVAL; + } + + rdev->rswitch->ops = &idtg3_switch_ops; + + if (rdev->do_enum) { + /* Disable hierarchical routing support: Existing fabric + * enumeration/discovery process (see rio-scan.c) uses 8-bit + * flat destination ID routing only. + */ + rio_write_config_32(rdev, 0x5000 + RIO_BC_RT_CTL_CSR, 0); + } + + spin_unlock(&rdev->rswitch->lock); + + return 0; +} + +static void idtg3_remove(struct rio_dev *rdev) +{ + pr_debug("RIO: %s for %s\n", __func__, rio_name(rdev)); + spin_lock(&rdev->rswitch->lock); + if (rdev->rswitch->ops == &idtg3_switch_ops) + rdev->rswitch->ops = NULL; + spin_unlock(&rdev->rswitch->lock); +} + +/* + * Gen3 switches repeat sending PW messages until a corresponding event flag + * is cleared. Use shutdown notification to disable generation of port-write + * messages if their destination node is shut down. + */ +static void idtg3_shutdown(struct rio_dev *rdev) +{ + int i; + u32 rval; + u16 destid; + + /* Currently the enumerator node acts also as PW handler */ + if (!rdev->do_enum) + return; + + pr_debug("RIO: %s(%s)\n", __func__, rio_name(rdev)); + + rio_read_config_32(rdev, RIO_PW_ROUTE, &rval); + i = RIO_GET_PORT_NUM(rdev->swpinfo); + + /* Check port-write destination port */ + if (!((1 << i) & rval)) + return; + + /* Disable sending port-write event notifications if PW destID + * matches to one of the enumerator node + */ + rio_read_config_32(rdev, rdev->em_efptr + RIO_EM_PW_TGT_DEVID, &rval); + + if (rval & RIO_EM_PW_TGT_DEVID_DEV16) + destid = rval >> 16; + else + destid = ((rval & RIO_EM_PW_TGT_DEVID_D8) >> 16); + + if (rdev->net->hport->host_deviceid == destid) { + rio_write_config_32(rdev, + rdev->em_efptr + RIO_EM_PW_TX_CTRL, 0); + pr_debug("RIO: %s(%s) PW transmission disabled\n", + __func__, rio_name(rdev)); + } +} + +static struct rio_device_id idtg3_id_table[] = { + {RIO_DEVICE(RIO_DID_IDTRXS1632, RIO_VID_IDT)}, + {RIO_DEVICE(RIO_DID_IDTRXS2448, RIO_VID_IDT)}, + { 0, } /* terminate list */ +}; + +static struct rio_driver idtg3_driver = { + .name = "idt_gen3", + .id_table = idtg3_id_table, + .probe = idtg3_probe, + .remove = idtg3_remove, + .shutdown = idtg3_shutdown, +}; + +static int __init idtg3_init(void) +{ + return rio_register_driver(&idtg3_driver); +} + +static void __exit idtg3_exit(void) +{ + pr_debug("RIO: %s\n", __func__); + rio_unregister_driver(&idtg3_driver); + pr_debug("RIO: %s done\n", __func__); +} + +device_initcall(idtg3_init); +module_exit(idtg3_exit); + +MODULE_DESCRIPTION("IDT RXS Gen.3 Serial RapidIO switch family driver"); +MODULE_AUTHOR("Integrated Device Technology, Inc."); +MODULE_LICENSE("GPL"); diff --git a/include/linux/rio_ids.h b/include/linux/rio_ids.h index 2543bc1..334c576 100644 --- a/include/linux/rio_ids.h +++ b/include/linux/rio_ids.h @@ -38,5 +38,7 @@ #define RIO_DID_IDTVPS1616 0x0377 #define RIO_DID_IDTSPS1616 0x0378 #define RIO_DID_TSI721 0x80ab +#define RIO_DID_IDTRXS1632 0x80e5 +#define RIO_DID_IDTRXS2448 0x80e6 #endif /* LINUX_RIO_IDS_H */ -- cgit v0.10.2 From 098f9fb0c962eb2fdba5f9d34f4cf7a938237184 Mon Sep 17 00:00:00 2001 From: "Andrew F. Davis" Date: Tue, 2 Aug 2016 14:07:06 -0700 Subject: w1: remove need for ida and use PLATFORM_DEVID_AUTO PLATFORM_DEVID_AUTO can be used to have the platform core assign a unique ID instead of manually creating one with IDA. Do this in all applicable drivers. Link: http://lkml.kernel.org/r/20160531204313.20979-1-afd@ti.com Signed-off-by: Andrew F. Davis Acked-by: Evgeniy Polyakov Cc: Greg KH Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/w1/slaves/w1_ds2760.c b/drivers/w1/slaves/w1_ds2760.c index d9079d4..59a81cf 100644 --- a/drivers/w1/slaves/w1_ds2760.c +++ b/drivers/w1/slaves/w1_ds2760.c @@ -121,25 +121,14 @@ static const struct attribute_group *w1_ds2760_groups[] = { NULL, }; -static DEFINE_IDA(bat_ida); - static int w1_ds2760_add_slave(struct w1_slave *sl) { int ret; - int id; struct platform_device *pdev; - id = ida_simple_get(&bat_ida, 0, 0, GFP_KERNEL); - if (id < 0) { - ret = id; - goto noid; - } - - pdev = platform_device_alloc("ds2760-battery", id); - if (!pdev) { - ret = -ENOMEM; - goto pdev_alloc_failed; - } + pdev = platform_device_alloc("ds2760-battery", PLATFORM_DEVID_AUTO); + if (!pdev) + return -ENOMEM; pdev->dev.parent = &sl->dev; ret = platform_device_add(pdev); @@ -148,24 +137,19 @@ static int w1_ds2760_add_slave(struct w1_slave *sl) dev_set_drvdata(&sl->dev, pdev); - goto success; + return 0; pdev_add_failed: platform_device_put(pdev); -pdev_alloc_failed: - ida_simple_remove(&bat_ida, id); -noid: -success: + return ret; } static void w1_ds2760_remove_slave(struct w1_slave *sl) { struct platform_device *pdev = dev_get_drvdata(&sl->dev); - int id = pdev->id; platform_device_unregister(pdev); - ida_simple_remove(&bat_ida, id); } static struct w1_family_ops w1_ds2760_fops = { @@ -182,14 +166,12 @@ static struct w1_family w1_ds2760_family = { static int __init w1_ds2760_init(void) { pr_info("1-Wire driver for the DS2760 battery monitor chip - (c) 2004-2005, Szabolcs Gyurko\n"); - ida_init(&bat_ida); return w1_register_family(&w1_ds2760_family); } static void __exit w1_ds2760_exit(void) { w1_unregister_family(&w1_ds2760_family); - ida_destroy(&bat_ida); } EXPORT_SYMBOL(w1_ds2760_read); diff --git a/drivers/w1/slaves/w1_ds2780.c b/drivers/w1/slaves/w1_ds2780.c index 50e85f7..e63eb86 100644 --- a/drivers/w1/slaves/w1_ds2780.c +++ b/drivers/w1/slaves/w1_ds2780.c @@ -113,25 +113,14 @@ static const struct attribute_group *w1_ds2780_groups[] = { NULL, }; -static DEFINE_IDA(bat_ida); - static int w1_ds2780_add_slave(struct w1_slave *sl) { int ret; - int id; struct platform_device *pdev; - id = ida_simple_get(&bat_ida, 0, 0, GFP_KERNEL); - if (id < 0) { - ret = id; - goto noid; - } - - pdev = platform_device_alloc("ds2780-battery", id); - if (!pdev) { - ret = -ENOMEM; - goto pdev_alloc_failed; - } + pdev = platform_device_alloc("ds2780-battery", PLATFORM_DEVID_AUTO); + if (!pdev) + return -ENOMEM; pdev->dev.parent = &sl->dev; ret = platform_device_add(pdev); @@ -144,19 +133,15 @@ static int w1_ds2780_add_slave(struct w1_slave *sl) pdev_add_failed: platform_device_put(pdev); -pdev_alloc_failed: - ida_simple_remove(&bat_ida, id); -noid: + return ret; } static void w1_ds2780_remove_slave(struct w1_slave *sl) { struct platform_device *pdev = dev_get_drvdata(&sl->dev); - int id = pdev->id; platform_device_unregister(pdev); - ida_simple_remove(&bat_ida, id); } static struct w1_family_ops w1_ds2780_fops = { @@ -172,14 +157,12 @@ static struct w1_family w1_ds2780_family = { static int __init w1_ds2780_init(void) { - ida_init(&bat_ida); return w1_register_family(&w1_ds2780_family); } static void __exit w1_ds2780_exit(void) { w1_unregister_family(&w1_ds2780_family); - ida_destroy(&bat_ida); } module_init(w1_ds2780_init); diff --git a/drivers/w1/slaves/w1_ds2781.c b/drivers/w1/slaves/w1_ds2781.c index 1eb98fb..99b0f4d 100644 --- a/drivers/w1/slaves/w1_ds2781.c +++ b/drivers/w1/slaves/w1_ds2781.c @@ -17,7 +17,6 @@ #include #include #include -#include #include "../w1.h" #include "../w1_int.h" @@ -111,25 +110,14 @@ static const struct attribute_group *w1_ds2781_groups[] = { NULL, }; -static DEFINE_IDA(bat_ida); - static int w1_ds2781_add_slave(struct w1_slave *sl) { int ret; - int id; struct platform_device *pdev; - id = ida_simple_get(&bat_ida, 0, 0, GFP_KERNEL); - if (id < 0) { - ret = id; - goto noid; - } - - pdev = platform_device_alloc("ds2781-battery", id); - if (!pdev) { - ret = -ENOMEM; - goto pdev_alloc_failed; - } + pdev = platform_device_alloc("ds2781-battery", PLATFORM_DEVID_AUTO); + if (!pdev) + return -ENOMEM; pdev->dev.parent = &sl->dev; ret = platform_device_add(pdev); @@ -142,19 +130,15 @@ static int w1_ds2781_add_slave(struct w1_slave *sl) pdev_add_failed: platform_device_put(pdev); -pdev_alloc_failed: - ida_simple_remove(&bat_ida, id); -noid: + return ret; } static void w1_ds2781_remove_slave(struct w1_slave *sl) { struct platform_device *pdev = dev_get_drvdata(&sl->dev); - int id = pdev->id; platform_device_unregister(pdev); - ida_simple_remove(&bat_ida, id); } static struct w1_family_ops w1_ds2781_fops = { @@ -170,14 +154,12 @@ static struct w1_family w1_ds2781_family = { static int __init w1_ds2781_init(void) { - ida_init(&bat_ida); return w1_register_family(&w1_ds2781_family); } static void __exit w1_ds2781_exit(void) { w1_unregister_family(&w1_ds2781_family); - ida_destroy(&bat_ida); } module_init(w1_ds2781_init); -- cgit v0.10.2 From 939fc832290d548a02b6a309992b3c1ff7de1ff9 Mon Sep 17 00:00:00 2001 From: "Andrew F. Davis" Date: Tue, 2 Aug 2016 14:07:09 -0700 Subject: w1: add helper macro module_w1_family The helper macro module_w1_family can be used in module drivers that only register a w1 driver in their module init functions. Add this macro and use it in all applicable drivers. Link: http://lkml.kernel.org/r/20160531204313.20979-2-afd@ti.com Signed-off-by: Andrew F. Davis Acked-by: Evgeniy Polyakov Cc: Greg KH Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/w1/slaves/w1_ds2406.c b/drivers/w1/slaves/w1_ds2406.c index d488961..51f2f66 100644 --- a/drivers/w1/slaves/w1_ds2406.c +++ b/drivers/w1/slaves/w1_ds2406.c @@ -153,16 +153,4 @@ static struct w1_family w1_family_12 = { .fid = W1_FAMILY_DS2406, .fops = &w1_f12_fops, }; - -static int __init w1_f12_init(void) -{ - return w1_register_family(&w1_family_12); -} - -static void __exit w1_f12_exit(void) -{ - w1_unregister_family(&w1_family_12); -} - -module_init(w1_f12_init); -module_exit(w1_f12_exit); +module_w1_family(w1_family_12); diff --git a/drivers/w1/slaves/w1_ds2408.c b/drivers/w1/slaves/w1_ds2408.c index 7dfa0e1..aec5958 100644 --- a/drivers/w1/slaves/w1_ds2408.c +++ b/drivers/w1/slaves/w1_ds2408.c @@ -351,16 +351,4 @@ static struct w1_family w1_family_29 = { .fid = W1_FAMILY_DS2408, .fops = &w1_f29_fops, }; - -static int __init w1_f29_init(void) -{ - return w1_register_family(&w1_family_29); -} - -static void __exit w1_f29_exit(void) -{ - w1_unregister_family(&w1_family_29); -} - -module_init(w1_f29_init); -module_exit(w1_f29_exit); +module_w1_family(w1_family_29); diff --git a/drivers/w1/slaves/w1_ds2413.c b/drivers/w1/slaves/w1_ds2413.c index ee28fc1..f2e1c51 100644 --- a/drivers/w1/slaves/w1_ds2413.c +++ b/drivers/w1/slaves/w1_ds2413.c @@ -135,16 +135,4 @@ static struct w1_family w1_family_3a = { .fid = W1_FAMILY_DS2413, .fops = &w1_f3a_fops, }; - -static int __init w1_f3a_init(void) -{ - return w1_register_family(&w1_family_3a); -} - -static void __exit w1_f3a_exit(void) -{ - w1_unregister_family(&w1_family_3a); -} - -module_init(w1_f3a_init); -module_exit(w1_f3a_exit); +module_w1_family(w1_family_3a); diff --git a/drivers/w1/slaves/w1_ds2423.c b/drivers/w1/slaves/w1_ds2423.c index 7e41b7d..4ab54fd 100644 --- a/drivers/w1/slaves/w1_ds2423.c +++ b/drivers/w1/slaves/w1_ds2423.c @@ -138,19 +138,7 @@ static struct w1_family w1_family_1d = { .fid = W1_COUNTER_DS2423, .fops = &w1_f1d_fops, }; - -static int __init w1_f1d_init(void) -{ - return w1_register_family(&w1_family_1d); -} - -static void __exit w1_f1d_exit(void) -{ - w1_unregister_family(&w1_family_1d); -} - -module_init(w1_f1d_init); -module_exit(w1_f1d_exit); +module_w1_family(w1_family_1d); MODULE_LICENSE("GPL"); MODULE_AUTHOR("Mika Laitio "); diff --git a/drivers/w1/slaves/w1_ds2431.c b/drivers/w1/slaves/w1_ds2431.c index 9c4ff9d..80572cb 100644 --- a/drivers/w1/slaves/w1_ds2431.c +++ b/drivers/w1/slaves/w1_ds2431.c @@ -288,19 +288,7 @@ static struct w1_family w1_family_2d = { .fid = W1_EEPROM_DS2431, .fops = &w1_f2d_fops, }; - -static int __init w1_f2d_init(void) -{ - return w1_register_family(&w1_family_2d); -} - -static void __exit w1_f2d_fini(void) -{ - w1_unregister_family(&w1_family_2d); -} - -module_init(w1_f2d_init); -module_exit(w1_f2d_fini); +module_w1_family(w1_family_2d); MODULE_LICENSE("GPL"); MODULE_AUTHOR("Bernhard Weirich "); diff --git a/drivers/w1/slaves/w1_ds2433.c b/drivers/w1/slaves/w1_ds2433.c index 72319a9..6cf378c 100644 --- a/drivers/w1/slaves/w1_ds2433.c +++ b/drivers/w1/slaves/w1_ds2433.c @@ -305,16 +305,4 @@ static struct w1_family w1_family_23 = { .fid = W1_EEPROM_DS2433, .fops = &w1_f23_fops, }; - -static int __init w1_f23_init(void) -{ - return w1_register_family(&w1_family_23); -} - -static void __exit w1_f23_fini(void) -{ - w1_unregister_family(&w1_family_23); -} - -module_init(w1_f23_init); -module_exit(w1_f23_fini); +module_w1_family(w1_family_23); diff --git a/drivers/w1/slaves/w1_ds2760.c b/drivers/w1/slaves/w1_ds2760.c index 59a81cf..ffa37f77 100644 --- a/drivers/w1/slaves/w1_ds2760.c +++ b/drivers/w1/slaves/w1_ds2760.c @@ -162,26 +162,13 @@ static struct w1_family w1_ds2760_family = { .fid = W1_FAMILY_DS2760, .fops = &w1_ds2760_fops, }; - -static int __init w1_ds2760_init(void) -{ - pr_info("1-Wire driver for the DS2760 battery monitor chip - (c) 2004-2005, Szabolcs Gyurko\n"); - return w1_register_family(&w1_ds2760_family); -} - -static void __exit w1_ds2760_exit(void) -{ - w1_unregister_family(&w1_ds2760_family); -} +module_w1_family(w1_ds2760_family); EXPORT_SYMBOL(w1_ds2760_read); EXPORT_SYMBOL(w1_ds2760_write); EXPORT_SYMBOL(w1_ds2760_store_eeprom); EXPORT_SYMBOL(w1_ds2760_recall_eeprom); -module_init(w1_ds2760_init); -module_exit(w1_ds2760_exit); - MODULE_LICENSE("GPL"); MODULE_AUTHOR("Szabolcs Gyurko "); MODULE_DESCRIPTION("1-wire Driver Dallas 2760 battery monitor chip"); diff --git a/drivers/w1/slaves/w1_ds2780.c b/drivers/w1/slaves/w1_ds2780.c index e63eb86..f5c2aa4 100644 --- a/drivers/w1/slaves/w1_ds2780.c +++ b/drivers/w1/slaves/w1_ds2780.c @@ -154,19 +154,7 @@ static struct w1_family w1_ds2780_family = { .fid = W1_FAMILY_DS2780, .fops = &w1_ds2780_fops, }; - -static int __init w1_ds2780_init(void) -{ - return w1_register_family(&w1_ds2780_family); -} - -static void __exit w1_ds2780_exit(void) -{ - w1_unregister_family(&w1_ds2780_family); -} - -module_init(w1_ds2780_init); -module_exit(w1_ds2780_exit); +module_w1_family(w1_ds2780_family); MODULE_LICENSE("GPL"); MODULE_AUTHOR("Clifton Barnes "); diff --git a/drivers/w1/slaves/w1_ds2781.c b/drivers/w1/slaves/w1_ds2781.c index 99b0f4d..9c03e01 100644 --- a/drivers/w1/slaves/w1_ds2781.c +++ b/drivers/w1/slaves/w1_ds2781.c @@ -151,19 +151,7 @@ static struct w1_family w1_ds2781_family = { .fid = W1_FAMILY_DS2781, .fops = &w1_ds2781_fops, }; - -static int __init w1_ds2781_init(void) -{ - return w1_register_family(&w1_ds2781_family); -} - -static void __exit w1_ds2781_exit(void) -{ - w1_unregister_family(&w1_ds2781_family); -} - -module_init(w1_ds2781_init); -module_exit(w1_ds2781_exit); +module_w1_family(w1_ds2781_family); MODULE_LICENSE("GPL"); MODULE_AUTHOR("Renata Sayakhova "); diff --git a/drivers/w1/slaves/w1_ds28e04.c b/drivers/w1/slaves/w1_ds28e04.c index 365d6df..5e348d3 100644 --- a/drivers/w1/slaves/w1_ds28e04.c +++ b/drivers/w1/slaves/w1_ds28e04.c @@ -427,16 +427,4 @@ static struct w1_family w1_family_1C = { .fid = W1_FAMILY_DS28E04, .fops = &w1_f1C_fops, }; - -static int __init w1_f1C_init(void) -{ - return w1_register_family(&w1_family_1C); -} - -static void __exit w1_f1C_fini(void) -{ - w1_unregister_family(&w1_family_1C); -} - -module_init(w1_f1C_init); -module_exit(w1_f1C_fini); +module_w1_family(w1_family_1C); diff --git a/drivers/w1/w1_family.h b/drivers/w1/w1_family.h index ed5dcb8..10a7a076 100644 --- a/drivers/w1/w1_family.h +++ b/drivers/w1/w1_family.h @@ -88,4 +88,16 @@ struct w1_family * w1_family_registered(u8); void w1_unregister_family(struct w1_family *); int w1_register_family(struct w1_family *); +/** + * module_w1_driver() - Helper macro for registering a 1-Wire families + * @__w1_family: w1_family struct + * + * Helper macro for 1-Wire families which do not do anything special in module + * init/exit. This eliminates a lot of boilerplate. Each module may only + * use this macro once, and calling it replaces module_init() and module_exit() + */ +#define module_w1_family(__w1_family) \ + module_driver(__w1_family, w1_register_family, \ + w1_unregister_family) + #endif /* __W1_FAMILY_H */ -- cgit v0.10.2 From ecfaf0c42fc4306b5ec4bf6be01b66f8fe9a9733 Mon Sep 17 00:00:00 2001 From: "H. Nikolaus Schaller" Date: Tue, 2 Aug 2016 14:07:12 -0700 Subject: w1:omap_hdq: fix regression Commit e93762bbf681 ("w1: masters: omap_hdq: add support for 1-wire mode") added a statement to clear the hdq_irqstatus flags in hdq_read_byte(). If the hdq reading process is scheduled slowly or interrupts are disabled for a while the hardware read activity might already be finished on entry of hdq_read_byte(). And hdq_isr() already has set the hdq_irqstatus to 0x6 (can be seen in debug mode) denoting that both, the TXCOMPLETE and RXCOMPLETE interrupts occurred in parallel. This means there is no need to wait and the hdq_read_byte() can just read the byte from the hdq controller. By resetting hdq_irqstatus to 0 the read process is forced to be always waiting again (because the if statement always succeeds) but the hardware will not issue another RXCOMPLETE interrupt. This results in a false timeout. After such a situation the hdq bus hangs. Link: http://lkml.kernel.org/r/b724765f87ad276a69625bc19806c8c8844c4590.1469513669.git.hns@goldelico.com Signed-off-by: H. Nikolaus Schaller Cc: Evgeniy Polyakov Cc: Greg Kroah-Hartman Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/w1/masters/omap_hdq.c b/drivers/w1/masters/omap_hdq.c index a2eec97..bb09de6 100644 --- a/drivers/w1/masters/omap_hdq.c +++ b/drivers/w1/masters/omap_hdq.c @@ -390,8 +390,6 @@ static int hdq_read_byte(struct hdq_data *hdq_data, u8 *val) goto out; } - hdq_data->hdq_irqstatus = 0; - if (!(hdq_data->hdq_irqstatus & OMAP_HDQ_INT_STATUS_RXCOMPLETE)) { hdq_reg_merge(hdq_data, OMAP_HDQ_CTRL_STATUS, OMAP_HDQ_CTRL_STATUS_DIR | OMAP_HDQ_CTRL_STATUS_GO, -- cgit v0.10.2 From 841c06d71e25a4e5fe8f7ed4ba7ba4324397f910 Mon Sep 17 00:00:00 2001 From: Prarit Bhargava Date: Tue, 2 Aug 2016 14:07:15 -0700 Subject: init: allow blacklisting of module_init functions sprint_symbol_no_offset() returns the string "function_name [module_name]" where [module_name] is not printed for built in kernel functions. This means that the blacklisting code will fail when comparing module function names with the extended string. This patch adds the functionality to block a module's module_init() function by finding the space in the string and truncating the comparison to that length. Link: http://lkml.kernel.org/r/1466124387-20446-1-git-send-email-prarit@redhat.com Signed-off-by: Prarit Bhargava Cc: Thomas Gleixner Cc: Yang Shi Cc: Prarit Bhargava Cc: Ingo Molnar Cc: Mel Gorman Cc: Rasmus Villemoes Cc: Kees Cook Cc: Yaowei Bai Cc: Andrey Ryabinin Cc: Rusty Russell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/init/main.c b/init/main.c index e7345dc..a8a58e2 100644 --- a/init/main.c +++ b/init/main.c @@ -716,6 +716,12 @@ static bool __init_or_module initcall_blacklisted(initcall_t fn) addr = (unsigned long) dereference_function_descriptor(fn); sprint_symbol_no_offset(fn_name, addr); + /* + * fn will be "function_name [module_name]" where [module_name] is not + * displayed for built-in init functions. Strip off the [module_name]. + */ + strreplace(fn_name, ' ', '\0'); + list_for_each_entry(entry, &blacklisted_initcalls, next) { if (!strcmp(fn_name, entry->buf)) { pr_debug("initcall %s blacklisted\n", fn_name); -- cgit v0.10.2 From 59dbb2a06fc2bcb752b964e036884fe9bb9dbbe0 Mon Sep 17 00:00:00 2001 From: Akash Goel Date: Tue, 2 Aug 2016 14:07:18 -0700 Subject: relay: add global mode support for buffer-only channels Commit 20d8b67c06fa ("relay: add buffer-only channels; useful for early logging") added support to use channels with no associated files. This is useful when the exact location of relay file is not known or the the parent directory of relay file is not available, while creating the channel and the logging has to start right from the boot. But there was no provision to use global mode with buffer-only channels, which is added by this patch, without modifying the interface where initially there will be a dummy invocation of create_buf_file callback through which kernel client can convey the need of a global buffer. For the use case where drivers/kernel clients want a simple interface for the userspace, which enables them to capture data/logs from relay file inorder & without any post processing, support of Global buffer mode is warranted. Modules, like i915, using relay_open() in early init would have to later register their buffer-only relays, once debugfs is available, by calling relay_late_setup_files(). Hence relay_late_setup_files() symbol also needs to be exported. Link: http://lkml.kernel.org/r/1468404563-11653-1-git-send-email-akash.goel@intel.com Signed-off-by: Akash Goel Cc: Eduard - Gabriel Munteanu Cc: Tom Zanussi Cc: Chris Wilson Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/kernel/relay.c b/kernel/relay.c index 04d7cf3..d797502 100644 --- a/kernel/relay.c +++ b/kernel/relay.c @@ -451,6 +451,13 @@ static struct rchan_buf *relay_open_buf(struct rchan *chan, unsigned int cpu) if (!dentry) goto free_buf; relay_set_buf_dentry(buf, dentry); + } else { + /* Only retrieve global info, nothing more, nothing less */ + dentry = chan->cb->create_buf_file(NULL, NULL, + S_IRUSR, buf, + &chan->is_global); + if (WARN_ON(dentry)) + goto free_buf; } buf->cpu = cpu; @@ -562,6 +569,10 @@ static int relay_hotcpu_callback(struct notifier_block *nb, * attributes specified. The created channel buffer files * will be named base_filename0...base_filenameN-1. File * permissions will be %S_IRUSR. + * + * If opening a buffer (@parent = NULL) that you later wish to register + * in a filesystem, call relay_late_setup_files() once the @parent dentry + * is available. */ struct rchan *relay_open(const char *base_filename, struct dentry *parent, @@ -640,8 +651,12 @@ static void __relay_set_buf_dentry(void *info) * * Returns 0 if successful, non-zero otherwise. * - * Use to setup files for a previously buffer-only channel. - * Useful to do early tracing in kernel, before VFS is up, for example. + * Use to setup files for a previously buffer-only channel created + * by relay_open() with a NULL parent dentry. + * + * For example, this is useful for perfomring early tracing in kernel, + * before VFS is up and then exposing the early results once the dentry + * is available. */ int relay_late_setup_files(struct rchan *chan, const char *base_filename, @@ -666,6 +681,20 @@ int relay_late_setup_files(struct rchan *chan, } chan->has_base_filename = 1; chan->parent = parent; + + if (chan->is_global) { + err = -EINVAL; + if (!WARN_ON_ONCE(!chan->buf[0])) { + dentry = relay_create_buf_file(chan, chan->buf[0], 0); + if (dentry && !WARN_ON_ONCE(!chan->is_global)) { + relay_set_buf_dentry(chan->buf[0], dentry); + err = 0; + } + } + mutex_unlock(&relay_channels_mutex); + return err; + } + curr_cpu = get_cpu(); /* * The CPU hotplug notifier ran before us and created buffers with @@ -706,6 +735,7 @@ int relay_late_setup_files(struct rchan *chan, return err; } +EXPORT_SYMBOL_GPL(relay_late_setup_files); /** * relay_switch_subbuf - switch to a new sub-buffer -- cgit v0.10.2 From ac3339baffd724edfb188ef57d1345d9649ba9af Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Tue, 2 Aug 2016 14:07:21 -0700 Subject: init/Kconfig: ban CONFIG_LOCALVERSION_AUTO with allmodconfig Doing patches with allmodconfig kernel compiled and committing stuff into local tree have unfortunate consequence: kernel version changes (as it should) leading to recompiling and relinking of several files even if they weren't touched (or interesting at all). This and "git-whatever" figuring out current version slow down compilation for no good reason. But lets face it, "allmodconfig" kernels don't care about kernel version, they are simply compile check guinea pigs. Make LOCALVERSION_AUTO depend on !COMPILE_TEST, so it doesn't sneak into allmodconfig .config. Link: http://lkml.kernel.org/r/20160707214954.GC31678@p183.telecom.by Signed-off-by: Alexey Dobriyan Cc: Michal Marek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/init/Kconfig b/init/Kconfig index 8f08f49..380798f 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -81,6 +81,7 @@ config LOCALVERSION config LOCALVERSION_AUTO bool "Automatically append version information to the version string" default y + depends on !COMPILE_TEST help This will try to automatically determine if the current tree is a release tree by looking for git tags that belong to the current -- cgit v0.10.2 From 27eb6622ab67bad75814c9b7b08096cfb16be63a Mon Sep 17 00:00:00 2001 From: Rob Herring Date: Tue, 2 Aug 2016 14:07:24 -0700 Subject: config: add android config fragments Copy the config fragments from the AOSP common kernel android-4.4 branch. It is becoming possible to run mainline kernels with Android, but the kernel defconfigs don't work as-is and debugging missing config options is a pain. Adding the config fragments into the kernel tree, makes configuring a mainline kernel as simple as: make ARCH=arm multi_v7_defconfig android-base.config android-recommended.config The following non-upstream config options were removed: CONFIG_NETFILTER_XT_MATCH_QTAGUID CONFIG_NETFILTER_XT_MATCH_QUOTA2 CONFIG_NETFILTER_XT_MATCH_QUOTA2_LOG CONFIG_PPPOLAC CONFIG_PPPOPNS CONFIG_SECURITY_PERF_EVENTS_RESTRICT CONFIG_USB_CONFIGFS_F_MTP CONFIG_USB_CONFIGFS_F_PTP CONFIG_USB_CONFIGFS_F_ACC CONFIG_USB_CONFIGFS_F_AUDIO_SRC CONFIG_USB_CONFIGFS_UEVENT CONFIG_INPUT_KEYCHORD CONFIG_INPUT_KEYRESET Link: http://lkml.kernel.org/r/1466708235-28593-1-git-send-email-robh@kernel.org Signed-off-by: Rob Herring Cc: Amit Pundir Cc: John Stultz Cc: Dmitry Shmidt Cc: Rom Lemarchand Cc: Greg Kroah-Hartman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/MAINTAINERS b/MAINTAINERS index e9eacac..ce38536 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -778,6 +778,11 @@ W: http://ez.analog.com/community/linux-device-drivers S: Supported F: drivers/dma/dma-axi-dmac.c +ANDROID CONFIG FRAGMENTS +M: Rob Herring +S: Supported +F: kernel/configs/android* + ANDROID DRIVERS M: Greg Kroah-Hartman M: Arve Hjønnevåg diff --git a/kernel/configs/android-base.config b/kernel/configs/android-base.config new file mode 100644 index 0000000..9f748ed --- /dev/null +++ b/kernel/configs/android-base.config @@ -0,0 +1,152 @@ +# KEEP ALPHABETICALLY SORTED +# CONFIG_DEVKMEM is not set +# CONFIG_DEVMEM is not set +# CONFIG_INET_LRO is not set +# CONFIG_MODULES is not set +# CONFIG_OABI_COMPAT is not set +# CONFIG_SYSVIPC is not set +CONFIG_ANDROID=y +CONFIG_ANDROID_BINDER_IPC=y +CONFIG_ANDROID_LOW_MEMORY_KILLER=y +CONFIG_ARMV8_DEPRECATED=y +CONFIG_ASHMEM=y +CONFIG_AUDIT=y +CONFIG_BLK_DEV_DM=y +CONFIG_BLK_DEV_INITRD=y +CONFIG_CGROUPS=y +CONFIG_CGROUP_CPUACCT=y +CONFIG_CGROUP_DEBUG=y +CONFIG_CGROUP_FREEZER=y +CONFIG_CGROUP_SCHED=y +CONFIG_CP15_BARRIER_EMULATION=y +CONFIG_DM_CRYPT=y +CONFIG_DM_VERITY=y +CONFIG_DM_VERITY_FEC=y +CONFIG_EMBEDDED=y +CONFIG_FB=y +CONFIG_HIGH_RES_TIMERS=y +CONFIG_INET6_AH=y +CONFIG_INET6_ESP=y +CONFIG_INET6_IPCOMP=y +CONFIG_INET=y +CONFIG_INET_DIAG_DESTROY=y +CONFIG_INET_ESP=y +CONFIG_INET_XFRM_MODE_TUNNEL=y +CONFIG_IP6_NF_FILTER=y +CONFIG_IP6_NF_IPTABLES=y +CONFIG_IP6_NF_MANGLE=y +CONFIG_IP6_NF_RAW=y +CONFIG_IP6_NF_TARGET_REJECT=y +CONFIG_IPV6=y +CONFIG_IPV6_MIP6=y +CONFIG_IPV6_MULTIPLE_TABLES=y +CONFIG_IPV6_OPTIMISTIC_DAD=y +CONFIG_IPV6_PRIVACY=y +CONFIG_IPV6_ROUTER_PREF=y +CONFIG_IPV6_ROUTE_INFO=y +CONFIG_IP_ADVANCED_ROUTER=y +CONFIG_IP_MULTICAST=y +CONFIG_IP_MULTIPLE_TABLES=y +CONFIG_IP_NF_ARPFILTER=y +CONFIG_IP_NF_ARPTABLES=y +CONFIG_IP_NF_ARP_MANGLE=y +CONFIG_IP_NF_FILTER=y +CONFIG_IP_NF_IPTABLES=y +CONFIG_IP_NF_MANGLE=y +CONFIG_IP_NF_MATCH_AH=y +CONFIG_IP_NF_MATCH_ECN=y +CONFIG_IP_NF_MATCH_TTL=y +CONFIG_IP_NF_NAT=y +CONFIG_IP_NF_RAW=y +CONFIG_IP_NF_SECURITY=y +CONFIG_IP_NF_TARGET_MASQUERADE=y +CONFIG_IP_NF_TARGET_NETMAP=y +CONFIG_IP_NF_TARGET_REDIRECT=y +CONFIG_IP_NF_TARGET_REJECT=y +CONFIG_NET=y +CONFIG_NETDEVICES=y +CONFIG_NETFILTER=y +CONFIG_NETFILTER_TPROXY=y +CONFIG_NETFILTER_XT_MATCH_COMMENT=y +CONFIG_NETFILTER_XT_MATCH_CONNLIMIT=y +CONFIG_NETFILTER_XT_MATCH_CONNMARK=y +CONFIG_NETFILTER_XT_MATCH_CONNTRACK=y +CONFIG_NETFILTER_XT_MATCH_HASHLIMIT=y +CONFIG_NETFILTER_XT_MATCH_HELPER=y +CONFIG_NETFILTER_XT_MATCH_IPRANGE=y +CONFIG_NETFILTER_XT_MATCH_LENGTH=y +CONFIG_NETFILTER_XT_MATCH_LIMIT=y +CONFIG_NETFILTER_XT_MATCH_MAC=y +CONFIG_NETFILTER_XT_MATCH_MARK=y +CONFIG_NETFILTER_XT_MATCH_PKTTYPE=y +CONFIG_NETFILTER_XT_MATCH_POLICY=y +CONFIG_NETFILTER_XT_MATCH_QUOTA=y +CONFIG_NETFILTER_XT_MATCH_SOCKET=y +CONFIG_NETFILTER_XT_MATCH_STATE=y +CONFIG_NETFILTER_XT_MATCH_STATISTIC=y +CONFIG_NETFILTER_XT_MATCH_STRING=y +CONFIG_NETFILTER_XT_MATCH_TIME=y +CONFIG_NETFILTER_XT_MATCH_U32=y +CONFIG_NETFILTER_XT_TARGET_CLASSIFY=y +CONFIG_NETFILTER_XT_TARGET_CONNMARK=y +CONFIG_NETFILTER_XT_TARGET_CONNSECMARK=y +CONFIG_NETFILTER_XT_TARGET_IDLETIMER=y +CONFIG_NETFILTER_XT_TARGET_MARK=y +CONFIG_NETFILTER_XT_TARGET_NFLOG=y +CONFIG_NETFILTER_XT_TARGET_NFQUEUE=y +CONFIG_NETFILTER_XT_TARGET_SECMARK=y +CONFIG_NETFILTER_XT_TARGET_TCPMSS=y +CONFIG_NETFILTER_XT_TARGET_TPROXY=y +CONFIG_NETFILTER_XT_TARGET_TRACE=y +CONFIG_NET_CLS_ACT=y +CONFIG_NET_CLS_U32=y +CONFIG_NET_EMATCH=y +CONFIG_NET_EMATCH_U32=y +CONFIG_NET_KEY=y +CONFIG_NET_SCHED=y +CONFIG_NET_SCH_HTB=y +CONFIG_NF_CONNTRACK=y +CONFIG_NF_CONNTRACK_AMANDA=y +CONFIG_NF_CONNTRACK_EVENTS=y +CONFIG_NF_CONNTRACK_FTP=y +CONFIG_NF_CONNTRACK_H323=y +CONFIG_NF_CONNTRACK_IPV4=y +CONFIG_NF_CONNTRACK_IPV6=y +CONFIG_NF_CONNTRACK_IRC=y +CONFIG_NF_CONNTRACK_NETBIOS_NS=y +CONFIG_NF_CONNTRACK_PPTP=y +CONFIG_NF_CONNTRACK_SANE=y +CONFIG_NF_CONNTRACK_SECMARK=y +CONFIG_NF_CONNTRACK_TFTP=y +CONFIG_NF_CT_NETLINK=y +CONFIG_NF_CT_PROTO_DCCP=y +CONFIG_NF_CT_PROTO_SCTP=y +CONFIG_NF_CT_PROTO_UDPLITE=y +CONFIG_NF_NAT=y +CONFIG_NO_HZ=y +CONFIG_PACKET=y +CONFIG_PM_AUTOSLEEP=y +CONFIG_PM_WAKELOCKS=y +CONFIG_PPP=y +CONFIG_PPP_BSDCOMP=y +CONFIG_PPP_DEFLATE=y +CONFIG_PPP_MPPE=y +CONFIG_PREEMPT=y +CONFIG_QUOTA=y +CONFIG_RTC_CLASS=y +CONFIG_RT_GROUP_SCHED=y +CONFIG_SECURITY=y +CONFIG_SECURITY_NETWORK=y +CONFIG_SECURITY_SELINUX=y +CONFIG_SETEND_EMULATION=y +CONFIG_STAGING=y +CONFIG_SWP_EMULATION=y +CONFIG_SYNC=y +CONFIG_TUN=y +CONFIG_UNIX=y +CONFIG_USB_GADGET=y +CONFIG_USB_CONFIGFS=y +CONFIG_USB_CONFIGFS_F_FS=y +CONFIG_USB_CONFIGFS_F_MIDI=y +CONFIG_USB_OTG_WAKELOCK=y +CONFIG_XFRM_USER=y diff --git a/kernel/configs/android-recommended.config b/kernel/configs/android-recommended.config new file mode 100644 index 0000000..e3b953e --- /dev/null +++ b/kernel/configs/android-recommended.config @@ -0,0 +1,121 @@ +# KEEP ALPHABETICALLY SORTED +# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set +# CONFIG_INPUT_MOUSE is not set +# CONFIG_LEGACY_PTYS is not set +# CONFIG_NF_CONNTRACK_SIP is not set +# CONFIG_PM_WAKELOCKS_GC is not set +# CONFIG_VT is not set +CONFIG_BACKLIGHT_LCD_SUPPORT=y +CONFIG_BLK_DEV_LOOP=y +CONFIG_BLK_DEV_RAM=y +CONFIG_BLK_DEV_RAM_SIZE=8192 +CONFIG_COMPACTION=y +CONFIG_DEBUG_RODATA=y +CONFIG_DM_UEVENT=y +CONFIG_DRAGONRISE_FF=y +CONFIG_ENABLE_DEFAULT_TRACERS=y +CONFIG_EXT4_FS=y +CONFIG_EXT4_FS_SECURITY=y +CONFIG_FUSE_FS=y +CONFIG_GREENASIA_FF=y +CONFIG_HIDRAW=y +CONFIG_HID_A4TECH=y +CONFIG_HID_ACRUX=y +CONFIG_HID_ACRUX_FF=y +CONFIG_HID_APPLE=y +CONFIG_HID_BELKIN=y +CONFIG_HID_CHERRY=y +CONFIG_HID_CHICONY=y +CONFIG_HID_CYPRESS=y +CONFIG_HID_DRAGONRISE=y +CONFIG_HID_ELECOM=y +CONFIG_HID_EMS_FF=y +CONFIG_HID_EZKEY=y +CONFIG_HID_GREENASIA=y +CONFIG_HID_GYRATION=y +CONFIG_HID_HOLTEK=y +CONFIG_HID_KENSINGTON=y +CONFIG_HID_KEYTOUCH=y +CONFIG_HID_KYE=y +CONFIG_HID_LCPOWER=y +CONFIG_HID_LOGITECH=y +CONFIG_HID_LOGITECH_DJ=y +CONFIG_HID_MAGICMOUSE=y +CONFIG_HID_MICROSOFT=y +CONFIG_HID_MONTEREY=y +CONFIG_HID_MULTITOUCH=y +CONFIG_HID_NTRIG=y +CONFIG_HID_ORTEK=y +CONFIG_HID_PANTHERLORD=y +CONFIG_HID_PETALYNX=y +CONFIG_HID_PICOLCD=y +CONFIG_HID_PRIMAX=y +CONFIG_HID_PRODIKEYS=y +CONFIG_HID_ROCCAT=y +CONFIG_HID_SAITEK=y +CONFIG_HID_SAMSUNG=y +CONFIG_HID_SMARTJOYPLUS=y +CONFIG_HID_SONY=y +CONFIG_HID_SPEEDLINK=y +CONFIG_HID_SUNPLUS=y +CONFIG_HID_THRUSTMASTER=y +CONFIG_HID_TIVO=y +CONFIG_HID_TOPSEED=y +CONFIG_HID_TWINHAN=y +CONFIG_HID_UCLOGIC=y +CONFIG_HID_WACOM=y +CONFIG_HID_WALTOP=y +CONFIG_HID_WIIMOTE=y +CONFIG_HID_ZEROPLUS=y +CONFIG_HID_ZYDACRON=y +CONFIG_INPUT_EVDEV=y +CONFIG_INPUT_GPIO=y +CONFIG_INPUT_JOYSTICK=y +CONFIG_INPUT_MISC=y +CONFIG_INPUT_TABLET=y +CONFIG_INPUT_UINPUT=y +CONFIG_ION=y +CONFIG_JOYSTICK_XPAD=y +CONFIG_JOYSTICK_XPAD_FF=y +CONFIG_JOYSTICK_XPAD_LEDS=y +CONFIG_KALLSYMS_ALL=y +CONFIG_KSM=y +CONFIG_LOGIG940_FF=y +CONFIG_LOGIRUMBLEPAD2_FF=y +CONFIG_LOGITECH_FF=y +CONFIG_MD=y +CONFIG_MEDIA_SUPPORT=y +CONFIG_MSDOS_FS=y +CONFIG_PANIC_TIMEOUT=5 +CONFIG_PANTHERLORD_FF=y +CONFIG_PERF_EVENTS=y +CONFIG_PM_DEBUG=y +CONFIG_PM_RUNTIME=y +CONFIG_PM_WAKELOCKS_LIMIT=0 +CONFIG_POWER_SUPPLY=y +CONFIG_PSTORE=y +CONFIG_PSTORE_CONSOLE=y +CONFIG_PSTORE_RAM=y +CONFIG_SCHEDSTATS=y +CONFIG_SMARTJOYPLUS_FF=y +CONFIG_SND=y +CONFIG_SOUND=y +CONFIG_SUSPEND_TIME=y +CONFIG_TABLET_USB_ACECAD=y +CONFIG_TABLET_USB_AIPTEK=y +CONFIG_TABLET_USB_GTCO=y +CONFIG_TABLET_USB_HANWANG=y +CONFIG_TABLET_USB_KBTAB=y +CONFIG_TASKSTATS=y +CONFIG_TASK_DELAY_ACCT=y +CONFIG_TASK_IO_ACCOUNTING=y +CONFIG_TASK_XACCT=y +CONFIG_TIMER_STATS=y +CONFIG_TMPFS=y +CONFIG_TMPFS_POSIX_ACL=y +CONFIG_UHID=y +CONFIG_USB_ANNOUNCE_NEW_DEVICES=y +CONFIG_USB_EHCI_HCD=y +CONFIG_USB_HIDDEV=y +CONFIG_USB_USBNET=y +CONFIG_VFAT_FS=y -- cgit v0.10.2 From f1cb637e75b59a07450cf81ad68b04f3f46b03d7 Mon Sep 17 00:00:00 2001 From: Valdis Kletnieks Date: Tue, 2 Aug 2016 14:07:27 -0700 Subject: init/Kconfig: add clarification for out-of-tree modules It doesn't trim just symbols that are totally unused in-tree - it trims the symbols unused by any in-tree modules actually built. If you've done a 'make localmodconfig' and only build a hundred or so modules, it's pretty likely that your out-of-tree module will come up lacking something... Hopefully this will save the next guy from a Homer Simpson "D'oh!" moment. Link: http://lkml.kernel.org/r/10177.1469787292@turing-police.cc.vt.edu Signed-off-by: Valdis Kletnieks Cc: Michal Marek Cc: Rusty Russell Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/init/Kconfig b/init/Kconfig index 380798f..6988649 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -2080,7 +2080,7 @@ config TRIM_UNUSED_KSYMS (especially when using LTO) for optimizing the code and reducing binary size. This might have some security advantages as well. - If unsure say N. + If unsure, or if you need to build out-of-tree modules, say N. endif # MODULES -- cgit v0.10.2 From a4691deabf284a601149a067525759939cc563b2 Mon Sep 17 00:00:00 2001 From: Vegard Nossum Date: Tue, 2 Aug 2016 14:07:30 -0700 Subject: kcov: allow more fine-grained coverage instrumentation For more targeted fuzzing, it's better to disable kernel-wide instrumentation and instead enable it on a per-subsystem basis. This follows the pattern of UBSAN and allows you to compile in the kcov driver without instrumenting the whole kernel. To instrument a part of the kernel, you can use either # for a single file in the current directory KCOV_INSTRUMENT_filename.o := y or # for all the files in the current directory (excluding subdirectories) KCOV_INSTRUMENT := y or # (same as above) ccflags-y += $(CFLAGS_KCOV) or # for all the files in the current directory (including subdirectories) subdir-ccflags-y += $(CFLAGS_KCOV) Link: http://lkml.kernel.org/r/1464008380-11405-1-git-send-email-vegard.nossum@oracle.com Signed-off-by: Vegard Nossum Cc: Dmitry Vyukov Cc: Quentin Casasnovas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug index f07842e..cc02f28 100644 --- a/lib/Kconfig.debug +++ b/lib/Kconfig.debug @@ -719,6 +719,17 @@ config KCOV For more details, see Documentation/kcov.txt. +config KCOV_INSTRUMENT_ALL + bool "Instrument all code by default" + depends on KCOV + default y if KCOV + help + If you are doing generic system call fuzzing (like e.g. syzkaller), + then you will want to instrument the whole kernel and you should + say y here. If you are doing more targeted fuzzing (like e.g. + filesystem fuzzing with AFL) then you will want to enable coverage + for more specific subsets of files, and should say n here. + config DEBUG_SHIRQ bool "Debug shared IRQ handlers" depends on DEBUG_KERNEL diff --git a/scripts/Makefile.lib b/scripts/Makefile.lib index e7df0f5..76494e1 100644 --- a/scripts/Makefile.lib +++ b/scripts/Makefile.lib @@ -138,7 +138,7 @@ endif ifeq ($(CONFIG_KCOV),y) _c_flags += $(if $(patsubst n%,, \ - $(KCOV_INSTRUMENT_$(basetarget).o)$(KCOV_INSTRUMENT)y), \ + $(KCOV_INSTRUMENT_$(basetarget).o)$(KCOV_INSTRUMENT)$(CONFIG_KCOV_INSTRUMENT_ALL)), \ $(CFLAGS_KCOV)) endif -- cgit v0.10.2 From 3bd080e4d8f2351ee3e143f0ec9307cc95ae6639 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Tue, 2 Aug 2016 14:07:32 -0700 Subject: ipc: delete "nr_ipc_ns" Write-only variable. Link: http://lkml.kernel.org/r/20160708214356.GA6785@p183.telecom.by Signed-off-by: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/ipc_namespace.h b/include/linux/ipc_namespace.h index 1eee6bc..d10e54f 100644 --- a/include/linux/ipc_namespace.h +++ b/include/linux/ipc_namespace.h @@ -63,8 +63,6 @@ struct ipc_namespace { }; extern struct ipc_namespace init_ipc_ns; -extern atomic_t nr_ipc_ns; - extern spinlock_t mq_lock; #ifdef CONFIG_SYSVIPC diff --git a/ipc/msgutil.c b/ipc/msgutil.c index ed81aaf..a521999 100644 --- a/ipc/msgutil.c +++ b/ipc/msgutil.c @@ -37,8 +37,6 @@ struct ipc_namespace init_ipc_ns = { #endif }; -atomic_t nr_ipc_ns = ATOMIC_INIT(1); - struct msg_msgseg { struct msg_msgseg *next; /* the next part of the message follows immediately */ diff --git a/ipc/namespace.c b/ipc/namespace.c index 04cb07e..d87e6ba 100644 --- a/ipc/namespace.c +++ b/ipc/namespace.c @@ -43,7 +43,6 @@ static struct ipc_namespace *create_ipc_ns(struct user_namespace *user_ns, kfree(ns); return ERR_PTR(err); } - atomic_inc(&nr_ipc_ns); sem_init_ns(ns); msg_init_ns(ns); @@ -96,7 +95,6 @@ static void free_ipc_ns(struct ipc_namespace *ns) sem_exit_ns(ns); msg_exit_ns(ns); shm_exit_ns(ns); - atomic_dec(&nr_ipc_ns); put_user_ns(ns->user_ns); ns_free_inum(&ns->ns); -- cgit v0.10.2