From 7a6610139a1e1d9297dd1c5d178022eac36839cb Mon Sep 17 00:00:00 2001 From: Jan Kiszka Date: Tue, 2 Nov 2010 08:05:51 +0100 Subject: intel-iommu: Fix use after release during device attach Obtain the new pgd pointer before releasing the page containing this value. Cc: stable@kernel.org Signed-off-by: Jan Kiszka Reviewed-by: Sheng Yang Signed-off-by: David Woodhouse diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index 4789f8e..35463dd 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -3627,9 +3627,9 @@ static int intel_iommu_attach_device(struct iommu_domain *domain, pte = dmar_domain->pgd; if (dma_pte_present(pte)) { - free_pgtable_page(dmar_domain->pgd); dmar_domain->pgd = (struct dma_pte *) phys_to_virt(dma_pte_addr(pte)); + free_pgtable_page(pte); } dmar_domain->agaw--; } -- cgit v0.10.2 From a97590e56d0d58e1dd262353f7cbd84e81d8e600 Mon Sep 17 00:00:00 2001 From: Alex Williamson Date: Fri, 4 Mar 2011 14:52:16 -0700 Subject: intel-iommu: Unlink domain from iommu When we remove a device, we unlink the iommu from the domain, but we never do the reverse unlinking of the domain from the iommu. This means that we never clear iommu->domain_ids, eventually leading to resource exhaustion if we repeatedly bind and unbind a device to a driver. Also free empty domains to avoid a resource leak. Signed-off-by: Alex Williamson Acked-by: Donald Dutile Signed-off-by: David Woodhouse Cc: stable@kernel.org diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index 35463dd..292f223 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -3260,9 +3260,15 @@ static int device_notifier(struct notifier_block *nb, if (!domain) return 0; - if (action == BUS_NOTIFY_UNBOUND_DRIVER && !iommu_pass_through) + if (action == BUS_NOTIFY_UNBOUND_DRIVER && !iommu_pass_through) { domain_remove_one_dev_info(domain, pdev); + if (!(domain->flags & DOMAIN_FLAG_VIRTUAL_MACHINE) && + !(domain->flags & DOMAIN_FLAG_STATIC_IDENTITY) && + list_empty(&domain->devices)) + domain_exit(domain); + } + return 0; } @@ -3411,6 +3417,11 @@ static void domain_remove_one_dev_info(struct dmar_domain *domain, domain->iommu_count--; domain_update_iommu_cap(domain); spin_unlock_irqrestore(&domain->iommu_lock, tmp_flags); + + spin_lock_irqsave(&iommu->lock, tmp_flags); + clear_bit(domain->id, iommu->domain_ids); + iommu->domains[domain->id] = NULL; + spin_unlock_irqrestore(&iommu->lock, tmp_flags); } spin_unlock_irqrestore(&device_domain_lock, flags); -- cgit v0.10.2 From 2fe9723df8e45fd247782adea244a5e653c30bf4 Mon Sep 17 00:00:00 2001 From: Alex Williamson Date: Fri, 4 Mar 2011 14:52:30 -0700 Subject: intel-iommu: Fix get_domain_for_dev() error path If we run out of domain_ids and fail iommu_attach_domain(), we fall into domain_exit() without having setup enough of the domain structure for this to do anything useful. In fact, it typically runs off into the weeds walking the bogus domain->devices list. Just free the domain. Signed-off-by: Alex Williamson Acked-by: Donald Dutile Signed-off-by: David Woodhouse Cc: stable@kernel.org diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index 292f223..5dc5d3e 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -1835,7 +1835,7 @@ static struct dmar_domain *get_domain_for_dev(struct pci_dev *pdev, int gaw) ret = iommu_attach_domain(domain, iommu); if (ret) { - domain_exit(domain); + free_domain_mem(domain); goto error; } -- cgit v0.10.2 From 523334ba508a8baaf5fc9f15fbad9ed04f334f48 Mon Sep 17 00:00:00 2001 From: Yongqiang Yang Date: Thu, 24 Mar 2011 08:48:39 +0800 Subject: ext3: Fix writepage credits computation for ordered mode Original computation forgets to count writes of indirect block themselves (it only counts with blocks necessary for their allocation) in ordered mode. Acked-by: Amir Goldstein Signed-off-by:Yongqiang Yang Signed-off-by: Jan Kara diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index ae94f6d..7f5db46 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -3294,7 +3294,7 @@ static int ext3_writepage_trans_blocks(struct inode *inode) if (ext3_should_journal_data(inode)) ret = 3 * (bpp + indirects) + 2; else - ret = 2 * (bpp + indirects) + 2; + ret = 2 * (bpp + indirects) + indirects + 2; #ifdef CONFIG_QUOTA /* We know that structure was already allocated during dquot_initialize so -- cgit v0.10.2 From 954032d2527f2fce7355ba70709b5e143d6b686f Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Thu, 24 Mar 2011 22:51:14 -0400 Subject: nfsd: fix auth_domain reference leak on nlm operations This was noticed by users who performed more than 2^32 lock operations and hence made this counter overflow (eventually leading to use-after-free's). Setting rq_client to NULL here means that it won't later get auth_domain_put() when it should be. Appears to have been introduced in 2.5.42 by "[PATCH] kNFSd: Move auth domain lookup into svcauth" which moved most of the rq_client handling to common svcauth code, but left behind this one line. Cc: Neil Brown Cc: stable@kernel.org Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/lockd.c b/fs/nfsd/lockd.c index 0c6d816..7c831a2 100644 --- a/fs/nfsd/lockd.c +++ b/fs/nfsd/lockd.c @@ -38,7 +38,6 @@ nlm_fopen(struct svc_rqst *rqstp, struct nfs_fh *f, struct file **filp) exp_readlock(); nfserr = nfsd_open(rqstp, &fh, S_IFREG, NFSD_MAY_LOCK, filp); fh_put(&fh); - rqstp->rq_client = NULL; exp_readunlock(); /* We return nlm error codes as nlm doesn't know * about nfsd, but nfsd does know about nlm.. -- cgit v0.10.2 From 49a341f2bb0e8eb2f2877956a758ea40cbd47b26 Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Mon, 28 Mar 2011 14:24:57 +0200 Subject: netfilter: ipset: list:set timeout variant fixes - the timeout value was actually not set - the garbage collector was broken The variant is fixed, the tests to the ipset testsuite are added. Signed-off-by: Jozsef Kadlecsik Signed-off-by: Patrick McHardy diff --git a/net/netfilter/ipset/ip_set_list_set.c b/net/netfilter/ipset/ip_set_list_set.c index a47c329..f4a46c0 100644 --- a/net/netfilter/ipset/ip_set_list_set.c +++ b/net/netfilter/ipset/ip_set_list_set.c @@ -43,14 +43,19 @@ struct list_set { static inline struct set_elem * list_set_elem(const struct list_set *map, u32 id) { - return (struct set_elem *)((char *)map->members + id * map->dsize); + return (struct set_elem *)((void *)map->members + id * map->dsize); +} + +static inline struct set_telem * +list_set_telem(const struct list_set *map, u32 id) +{ + return (struct set_telem *)((void *)map->members + id * map->dsize); } static inline bool list_set_timeout(const struct list_set *map, u32 id) { - const struct set_telem *elem = - (const struct set_telem *) list_set_elem(map, id); + const struct set_telem *elem = list_set_telem(map, id); return ip_set_timeout_test(elem->timeout); } @@ -58,19 +63,11 @@ list_set_timeout(const struct list_set *map, u32 id) static inline bool list_set_expired(const struct list_set *map, u32 id) { - const struct set_telem *elem = - (const struct set_telem *) list_set_elem(map, id); + const struct set_telem *elem = list_set_telem(map, id); return ip_set_timeout_expired(elem->timeout); } -static inline int -list_set_exist(const struct set_telem *elem) -{ - return elem->id != IPSET_INVALID_ID && - !ip_set_timeout_expired(elem->timeout); -} - /* Set list without and with timeout */ static int @@ -146,11 +143,11 @@ list_elem_tadd(struct list_set *map, u32 i, ip_set_id_t id, struct set_telem *e; for (; i < map->size; i++) { - e = (struct set_telem *)list_set_elem(map, i); + e = list_set_telem(map, i); swap(e->id, id); + swap(e->timeout, timeout); if (e->id == IPSET_INVALID_ID) break; - swap(e->timeout, timeout); } } @@ -164,7 +161,7 @@ list_set_add(struct list_set *map, u32 i, ip_set_id_t id, /* Last element replaced: e.g. add new,before,last */ ip_set_put_byindex(e->id); if (with_timeout(map->timeout)) - list_elem_tadd(map, i, id, timeout); + list_elem_tadd(map, i, id, ip_set_timeout_set(timeout)); else list_elem_add(map, i, id); @@ -172,11 +169,11 @@ list_set_add(struct list_set *map, u32 i, ip_set_id_t id, } static int -list_set_del(struct list_set *map, ip_set_id_t id, u32 i) +list_set_del(struct list_set *map, u32 i) { struct set_elem *a = list_set_elem(map, i), *b; - ip_set_put_byindex(id); + ip_set_put_byindex(a->id); for (; i < map->size - 1; i++) { b = list_set_elem(map, i + 1); @@ -308,11 +305,11 @@ list_set_uadt(struct ip_set *set, struct nlattr *tb[], (before == 0 || (before > 0 && next_id_eq(map, i, refid)))) - ret = list_set_del(map, id, i); + ret = list_set_del(map, i); else if (before < 0 && elem->id == refid && next_id_eq(map, i, id)) - ret = list_set_del(map, id, i + 1); + ret = list_set_del(map, i + 1); } break; default: @@ -460,17 +457,15 @@ list_set_gc(unsigned long ul_set) struct list_set *map = set->data; struct set_telem *e; u32 i; - - /* We run parallel with other readers (test element) - * but adding/deleting new entries is locked out */ - read_lock_bh(&set->lock); - for (i = map->size - 1; i >= 0; i--) { - e = (struct set_telem *) list_set_elem(map, i); - if (e->id != IPSET_INVALID_ID && - list_set_expired(map, i)) - list_set_del(map, e->id, i); + + /* nfnl_lock should be called */ + write_lock_bh(&set->lock); + for (i = 0; i < map->size; i++) { + e = list_set_telem(map, i); + if (e->id != IPSET_INVALID_ID && list_set_expired(map, i)) + list_set_del(map, i); } - read_unlock_bh(&set->lock); + write_unlock_bh(&set->lock); map->gc.expires = jiffies + IPSET_GC_PERIOD(map->timeout) * HZ; add_timer(&map->gc); -- cgit v0.10.2 From 6604271c5bc658a6067ed0c3deba4d89e0e50382 Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Tue, 29 Mar 2011 15:00:43 +0200 Subject: netfilter: ipset: References are protected by rwlock instead of mutex The timeout variant of the list:set type must reference the member sets. However, its garbage collector runs at timer interrupt so the mutex protection of the references is a no go. Therefore the reference protection is converted to rwlock. Signed-off-by: Jozsef Kadlecsik Signed-off-by: Patrick McHardy diff --git a/include/linux/netfilter/ipset/ip_set.h b/include/linux/netfilter/ipset/ip_set.h index ec333d8..5a262e3 100644 --- a/include/linux/netfilter/ipset/ip_set.h +++ b/include/linux/netfilter/ipset/ip_set.h @@ -293,7 +293,7 @@ struct ip_set { /* Lock protecting the set data */ rwlock_t lock; /* References to the set */ - atomic_t ref; + u32 ref; /* The core set type */ struct ip_set_type *type; /* The type variant doing the real job */ diff --git a/include/linux/netfilter/ipset/ip_set_ahash.h b/include/linux/netfilter/ipset/ip_set_ahash.h index ec9d9be..a0196ac 100644 --- a/include/linux/netfilter/ipset/ip_set_ahash.h +++ b/include/linux/netfilter/ipset/ip_set_ahash.h @@ -515,8 +515,7 @@ type_pf_head(struct ip_set *set, struct sk_buff *skb) if (h->netmask != HOST_MASK) NLA_PUT_U8(skb, IPSET_ATTR_NETMASK, h->netmask); #endif - NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, - htonl(atomic_read(&set->ref) - 1)); + NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)); NLA_PUT_NET32(skb, IPSET_ATTR_MEMSIZE, htonl(memsize)); if (with_timeout(h->timeout)) NLA_PUT_NET32(skb, IPSET_ATTR_TIMEOUT, htonl(h->timeout)); diff --git a/net/netfilter/ipset/ip_set_bitmap_ip.c b/net/netfilter/ipset/ip_set_bitmap_ip.c index bca9699..a113ff0 100644 --- a/net/netfilter/ipset/ip_set_bitmap_ip.c +++ b/net/netfilter/ipset/ip_set_bitmap_ip.c @@ -338,8 +338,7 @@ bitmap_ip_head(struct ip_set *set, struct sk_buff *skb) NLA_PUT_IPADDR4(skb, IPSET_ATTR_IP_TO, htonl(map->last_ip)); if (map->netmask != 32) NLA_PUT_U8(skb, IPSET_ATTR_NETMASK, map->netmask); - NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, - htonl(atomic_read(&set->ref) - 1)); + NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)); NLA_PUT_NET32(skb, IPSET_ATTR_MEMSIZE, htonl(sizeof(*map) + map->memsize)); if (with_timeout(map->timeout)) diff --git a/net/netfilter/ipset/ip_set_bitmap_ipmac.c b/net/netfilter/ipset/ip_set_bitmap_ipmac.c index 5e79017..00a3324 100644 --- a/net/netfilter/ipset/ip_set_bitmap_ipmac.c +++ b/net/netfilter/ipset/ip_set_bitmap_ipmac.c @@ -434,8 +434,7 @@ bitmap_ipmac_head(struct ip_set *set, struct sk_buff *skb) goto nla_put_failure; NLA_PUT_IPADDR4(skb, IPSET_ATTR_IP, htonl(map->first_ip)); NLA_PUT_IPADDR4(skb, IPSET_ATTR_IP_TO, htonl(map->last_ip)); - NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, - htonl(atomic_read(&set->ref) - 1)); + NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)); NLA_PUT_NET32(skb, IPSET_ATTR_MEMSIZE, htonl(sizeof(*map) + (map->last_ip - map->first_ip + 1) * map->dsize)); diff --git a/net/netfilter/ipset/ip_set_bitmap_port.c b/net/netfilter/ipset/ip_set_bitmap_port.c index 165f09b..6b38eb8 100644 --- a/net/netfilter/ipset/ip_set_bitmap_port.c +++ b/net/netfilter/ipset/ip_set_bitmap_port.c @@ -320,8 +320,7 @@ bitmap_port_head(struct ip_set *set, struct sk_buff *skb) goto nla_put_failure; NLA_PUT_NET16(skb, IPSET_ATTR_PORT, htons(map->first_port)); NLA_PUT_NET16(skb, IPSET_ATTR_PORT_TO, htons(map->last_port)); - NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, - htonl(atomic_read(&set->ref) - 1)); + NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)); NLA_PUT_NET32(skb, IPSET_ATTR_MEMSIZE, htonl(sizeof(*map) + map->memsize)); if (with_timeout(map->timeout)) diff --git a/net/netfilter/ipset/ip_set_core.c b/net/netfilter/ipset/ip_set_core.c index d6b4823..e88ac3c 100644 --- a/net/netfilter/ipset/ip_set_core.c +++ b/net/netfilter/ipset/ip_set_core.c @@ -26,6 +26,7 @@ static LIST_HEAD(ip_set_type_list); /* all registered set types */ static DEFINE_MUTEX(ip_set_type_mutex); /* protects ip_set_type_list */ +static DEFINE_RWLOCK(ip_set_ref_lock); /* protects the set refs */ static struct ip_set **ip_set_list; /* all individual sets */ static ip_set_id_t ip_set_max = CONFIG_IP_SET_MAX; /* max number of sets */ @@ -301,13 +302,18 @@ EXPORT_SYMBOL_GPL(ip_set_get_ipaddr6); static inline void __ip_set_get(ip_set_id_t index) { - atomic_inc(&ip_set_list[index]->ref); + write_lock_bh(&ip_set_ref_lock); + ip_set_list[index]->ref++; + write_unlock_bh(&ip_set_ref_lock); } static inline void __ip_set_put(ip_set_id_t index) { - atomic_dec(&ip_set_list[index]->ref); + write_lock_bh(&ip_set_ref_lock); + BUG_ON(ip_set_list[index]->ref == 0); + ip_set_list[index]->ref--; + write_unlock_bh(&ip_set_ref_lock); } /* @@ -324,7 +330,7 @@ ip_set_test(ip_set_id_t index, const struct sk_buff *skb, struct ip_set *set = ip_set_list[index]; int ret = 0; - BUG_ON(set == NULL || atomic_read(&set->ref) == 0); + BUG_ON(set == NULL); pr_debug("set %s, index %u\n", set->name, index); if (dim < set->type->dimension || @@ -356,7 +362,7 @@ ip_set_add(ip_set_id_t index, const struct sk_buff *skb, struct ip_set *set = ip_set_list[index]; int ret; - BUG_ON(set == NULL || atomic_read(&set->ref) == 0); + BUG_ON(set == NULL); pr_debug("set %s, index %u\n", set->name, index); if (dim < set->type->dimension || @@ -378,7 +384,7 @@ ip_set_del(ip_set_id_t index, const struct sk_buff *skb, struct ip_set *set = ip_set_list[index]; int ret = 0; - BUG_ON(set == NULL || atomic_read(&set->ref) == 0); + BUG_ON(set == NULL); pr_debug("set %s, index %u\n", set->name, index); if (dim < set->type->dimension || @@ -397,7 +403,6 @@ EXPORT_SYMBOL_GPL(ip_set_del); * Find set by name, reference it once. The reference makes sure the * thing pointed to, does not go away under our feet. * - * The nfnl mutex must already be activated. */ ip_set_id_t ip_set_get_byname(const char *name, struct ip_set **set) @@ -423,15 +428,12 @@ EXPORT_SYMBOL_GPL(ip_set_get_byname); * reference count by 1. The caller shall not assume the index * to be valid, after calling this function. * - * The nfnl mutex must already be activated. */ void ip_set_put_byindex(ip_set_id_t index) { - if (ip_set_list[index] != NULL) { - BUG_ON(atomic_read(&ip_set_list[index]->ref) == 0); + if (ip_set_list[index] != NULL) __ip_set_put(index); - } } EXPORT_SYMBOL_GPL(ip_set_put_byindex); @@ -441,7 +443,6 @@ EXPORT_SYMBOL_GPL(ip_set_put_byindex); * can't be destroyed. The set cannot be renamed due to * the referencing either. * - * The nfnl mutex must already be activated. */ const char * ip_set_name_byindex(ip_set_id_t index) @@ -449,7 +450,7 @@ ip_set_name_byindex(ip_set_id_t index) const struct ip_set *set = ip_set_list[index]; BUG_ON(set == NULL); - BUG_ON(atomic_read(&set->ref) == 0); + BUG_ON(set->ref == 0); /* Referenced, so it's safe */ return set->name; @@ -515,10 +516,7 @@ void ip_set_nfnl_put(ip_set_id_t index) { nfnl_lock(); - if (ip_set_list[index] != NULL) { - BUG_ON(atomic_read(&ip_set_list[index]->ref) == 0); - __ip_set_put(index); - } + ip_set_put_byindex(index); nfnl_unlock(); } EXPORT_SYMBOL_GPL(ip_set_nfnl_put); @@ -526,7 +524,7 @@ EXPORT_SYMBOL_GPL(ip_set_nfnl_put); /* * Communication protocol with userspace over netlink. * - * We already locked by nfnl_lock. + * The commands are serialized by the nfnl mutex. */ static inline bool @@ -657,7 +655,6 @@ ip_set_create(struct sock *ctnl, struct sk_buff *skb, return -ENOMEM; rwlock_init(&set->lock); strlcpy(set->name, name, IPSET_MAXNAMELEN); - atomic_set(&set->ref, 0); set->family = family; /* @@ -690,8 +687,8 @@ ip_set_create(struct sock *ctnl, struct sk_buff *skb, /* * Here, we have a valid, constructed set and we are protected - * by nfnl_lock. Find the first free index in ip_set_list and - * check clashing. + * by the nfnl mutex. Find the first free index in ip_set_list + * and check clashing. */ if ((ret = find_free_id(set->name, &index, &clash)) != 0) { /* If this is the same set and requested, ignore error */ @@ -751,31 +748,51 @@ ip_set_destroy(struct sock *ctnl, struct sk_buff *skb, const struct nlattr * const attr[]) { ip_set_id_t i; + int ret = 0; if (unlikely(protocol_failed(attr))) return -IPSET_ERR_PROTOCOL; - /* References are protected by the nfnl mutex */ + /* Commands are serialized and references are + * protected by the ip_set_ref_lock. + * External systems (i.e. xt_set) must call + * ip_set_put|get_nfnl_* functions, that way we + * can safely check references here. + * + * list:set timer can only decrement the reference + * counter, so if it's already zero, we can proceed + * without holding the lock. + */ + read_lock_bh(&ip_set_ref_lock); if (!attr[IPSET_ATTR_SETNAME]) { for (i = 0; i < ip_set_max; i++) { - if (ip_set_list[i] != NULL && - (atomic_read(&ip_set_list[i]->ref))) - return -IPSET_ERR_BUSY; + if (ip_set_list[i] != NULL && ip_set_list[i]->ref) { + ret = IPSET_ERR_BUSY; + goto out; + } } + read_unlock_bh(&ip_set_ref_lock); for (i = 0; i < ip_set_max; i++) { if (ip_set_list[i] != NULL) ip_set_destroy_set(i); } } else { i = find_set_id(nla_data(attr[IPSET_ATTR_SETNAME])); - if (i == IPSET_INVALID_ID) - return -ENOENT; - else if (atomic_read(&ip_set_list[i]->ref)) - return -IPSET_ERR_BUSY; + if (i == IPSET_INVALID_ID) { + ret = -ENOENT; + goto out; + } else if (ip_set_list[i]->ref) { + ret = -IPSET_ERR_BUSY; + goto out; + } + read_unlock_bh(&ip_set_ref_lock); ip_set_destroy_set(i); } return 0; +out: + read_unlock_bh(&ip_set_ref_lock); + return ret; } /* Flush sets */ @@ -834,6 +851,7 @@ ip_set_rename(struct sock *ctnl, struct sk_buff *skb, struct ip_set *set; const char *name2; ip_set_id_t i; + int ret = 0; if (unlikely(protocol_failed(attr) || attr[IPSET_ATTR_SETNAME] == NULL || @@ -843,25 +861,33 @@ ip_set_rename(struct sock *ctnl, struct sk_buff *skb, set = find_set(nla_data(attr[IPSET_ATTR_SETNAME])); if (set == NULL) return -ENOENT; - if (atomic_read(&set->ref) != 0) - return -IPSET_ERR_REFERENCED; + + read_lock_bh(&ip_set_ref_lock); + if (set->ref != 0) { + ret = -IPSET_ERR_REFERENCED; + goto out; + } name2 = nla_data(attr[IPSET_ATTR_SETNAME2]); for (i = 0; i < ip_set_max; i++) { if (ip_set_list[i] != NULL && - STREQ(ip_set_list[i]->name, name2)) - return -IPSET_ERR_EXIST_SETNAME2; + STREQ(ip_set_list[i]->name, name2)) { + ret = -IPSET_ERR_EXIST_SETNAME2; + goto out; + } } strncpy(set->name, name2, IPSET_MAXNAMELEN); - return 0; +out: + read_unlock_bh(&ip_set_ref_lock); + return ret; } /* Swap two sets so that name/index points to the other. * References and set names are also swapped. * - * We are protected by the nfnl mutex and references are - * manipulated only by holding the mutex. The kernel interfaces + * The commands are serialized by the nfnl mutex and references are + * protected by the ip_set_ref_lock. The kernel interfaces * do not hold the mutex but the pointer settings are atomic * so the ip_set_list always contains valid pointers to the sets. */ @@ -874,7 +900,6 @@ ip_set_swap(struct sock *ctnl, struct sk_buff *skb, struct ip_set *from, *to; ip_set_id_t from_id, to_id; char from_name[IPSET_MAXNAMELEN]; - u32 from_ref; if (unlikely(protocol_failed(attr) || attr[IPSET_ATTR_SETNAME] == NULL || @@ -899,17 +924,15 @@ ip_set_swap(struct sock *ctnl, struct sk_buff *skb, from->type->family == to->type->family)) return -IPSET_ERR_TYPE_MISMATCH; - /* No magic here: ref munging protected by the nfnl_lock */ strncpy(from_name, from->name, IPSET_MAXNAMELEN); - from_ref = atomic_read(&from->ref); - strncpy(from->name, to->name, IPSET_MAXNAMELEN); - atomic_set(&from->ref, atomic_read(&to->ref)); strncpy(to->name, from_name, IPSET_MAXNAMELEN); - atomic_set(&to->ref, from_ref); + write_lock_bh(&ip_set_ref_lock); + swap(from->ref, to->ref); ip_set_list[from_id] = to; ip_set_list[to_id] = from; + write_unlock_bh(&ip_set_ref_lock); return 0; } @@ -926,7 +949,7 @@ ip_set_dump_done(struct netlink_callback *cb) { if (cb->args[2]) { pr_debug("release set %s\n", ip_set_list[cb->args[1]]->name); - __ip_set_put((ip_set_id_t) cb->args[1]); + ip_set_put_byindex((ip_set_id_t) cb->args[1]); } return 0; } @@ -1068,7 +1091,7 @@ release_refcount: /* If there was an error or set is done, release set */ if (ret || !cb->args[2]) { pr_debug("release set %s\n", ip_set_list[index]->name); - __ip_set_put(index); + ip_set_put_byindex(index); } /* If we dump all sets, continue with dumping last ones */ diff --git a/net/netfilter/ipset/ip_set_list_set.c b/net/netfilter/ipset/ip_set_list_set.c index f4a46c0..e9159e9 100644 --- a/net/netfilter/ipset/ip_set_list_set.c +++ b/net/netfilter/ipset/ip_set_list_set.c @@ -366,8 +366,7 @@ list_set_head(struct ip_set *set, struct sk_buff *skb) NLA_PUT_NET32(skb, IPSET_ATTR_SIZE, htonl(map->size)); if (with_timeout(map->timeout)) NLA_PUT_NET32(skb, IPSET_ATTR_TIMEOUT, htonl(map->timeout)); - NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, - htonl(atomic_read(&set->ref) - 1)); + NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)); NLA_PUT_NET32(skb, IPSET_ATTR_MEMSIZE, htonl(sizeof(*map) + map->size * map->dsize)); ipset_nest_end(skb, nested); @@ -457,8 +456,7 @@ list_set_gc(unsigned long ul_set) struct list_set *map = set->data; struct set_telem *e; u32 i; - - /* nfnl_lock should be called */ + write_lock_bh(&set->lock); for (i = 0; i < map->size; i++) { e = list_set_telem(map, i); -- cgit v0.10.2 From efd6947945fe5ebcf759e762f51c4e1d867edc49 Mon Sep 17 00:00:00 2001 From: Antonio Ospite Date: Fri, 18 Mar 2011 12:47:33 +0100 Subject: ASoC: zylonite: set .codec_dai_name in initializer Fix the initialization of .codec_dai_name in zylonite_dai initializer, do not mix it with the initialization of .codec_name which is set already a few lines above. Signed-off-by: Antonio Ospite Acked-by: Eric Miao Acked-by: Liam Girdwood Signed-off-by: Mark Brown diff --git a/sound/soc/pxa/zylonite.c b/sound/soc/pxa/zylonite.c index ac57726..b644575 100644 --- a/sound/soc/pxa/zylonite.c +++ b/sound/soc/pxa/zylonite.c @@ -167,7 +167,7 @@ static struct snd_soc_dai_link zylonite_dai[] = { .codec_name = "wm9713-codec", .platform_name = "pxa-pcm-audio", .cpu_dai_name = "pxa2xx-ac97", - .codec_name = "wm9713-hifi", + .codec_dai_name = "wm9713-hifi", .init = zylonite_wm9713_init, }, { @@ -176,7 +176,7 @@ static struct snd_soc_dai_link zylonite_dai[] = { .codec_name = "wm9713-codec", .platform_name = "pxa-pcm-audio", .cpu_dai_name = "pxa2xx-ac97-aux", - .codec_name = "wm9713-aux", + .codec_dai_name = "wm9713-aux", }, { .name = "WM9713 Voice", @@ -184,7 +184,7 @@ static struct snd_soc_dai_link zylonite_dai[] = { .codec_name = "wm9713-codec", .platform_name = "pxa-pcm-audio", .cpu_dai_name = "pxa-ssp-dai.2", - .codec_name = "wm9713-voice", + .codec_dai_name = "wm9713-voice", .ops = &zylonite_voice_ops, }, }; -- cgit v0.10.2 From f6d5b33125c4fa63c16f7f54c533338c9695d82c Mon Sep 17 00:00:00 2001 From: John Stultz Date: Tue, 29 Mar 2011 18:00:27 -0700 Subject: RTC: Fix early irqs caused by calling rtc_set_alarm too early When we register an rtc device at boot, we read the alarm value in hardware and set the rtc device's aie_timer to that value. The initial method to do this was to simply call rtc_set_alarm() with the value read from hardware. However, this may cause problems as rtc_set_alarm may enable interupts, and the RTC alarm might fire, which can cause invalid pointer dereferencing since the RTC registration is not complete. This patch solves the issue by initializing the rtc_device.aie_timer y hand via rtc_initialize_alarm(). This avoids any calls to the RTC hardware which might enable interrupts too early. CC: Thomas Gleixner CC: Alessandro Zummo Reported-by: Konrad Rzeszutek Wilk Tested-by: Konrad Rzeszutek Wilk Signed-off-by: John Stultz diff --git a/drivers/rtc/class.c b/drivers/rtc/class.c index 09b4437..3901386 100644 --- a/drivers/rtc/class.c +++ b/drivers/rtc/class.c @@ -171,7 +171,7 @@ struct rtc_device *rtc_device_register(const char *name, struct device *dev, err = __rtc_read_alarm(rtc, &alrm); if (!err && !rtc_valid_tm(&alrm.time)) - rtc_set_alarm(rtc, &alrm); + rtc_initialize_alarm(rtc, &alrm); strlcpy(rtc->name, name, RTC_DEVICE_NAME_SIZE); dev_set_name(&rtc->dev, "rtc%d", id); diff --git a/drivers/rtc/interface.c b/drivers/rtc/interface.c index 8ec6b06..b2fea80 100644 --- a/drivers/rtc/interface.c +++ b/drivers/rtc/interface.c @@ -375,6 +375,32 @@ int rtc_set_alarm(struct rtc_device *rtc, struct rtc_wkalrm *alarm) } EXPORT_SYMBOL_GPL(rtc_set_alarm); +/* Called once per device from rtc_device_register */ +int rtc_initialize_alarm(struct rtc_device *rtc, struct rtc_wkalrm *alarm) +{ + int err; + + err = rtc_valid_tm(&alarm->time); + if (err != 0) + return err; + + err = mutex_lock_interruptible(&rtc->ops_lock); + if (err) + return err; + + rtc->aie_timer.node.expires = rtc_tm_to_ktime(alarm->time); + rtc->aie_timer.period = ktime_set(0, 0); + if (alarm->enabled) { + rtc->aie_timer.enabled = 1; + timerqueue_add(&rtc->timerqueue, &rtc->aie_timer.node); + } + mutex_unlock(&rtc->ops_lock); + return err; +} +EXPORT_SYMBOL_GPL(rtc_initialize_alarm); + + + int rtc_alarm_irq_enable(struct rtc_device *rtc, unsigned int enabled) { int err = mutex_lock_interruptible(&rtc->ops_lock); diff --git a/include/linux/rtc.h b/include/linux/rtc.h index 2ca7e8a..877ece4 100644 --- a/include/linux/rtc.h +++ b/include/linux/rtc.h @@ -228,6 +228,8 @@ extern int rtc_read_alarm(struct rtc_device *rtc, struct rtc_wkalrm *alrm); extern int rtc_set_alarm(struct rtc_device *rtc, struct rtc_wkalrm *alrm); +extern int rtc_initialize_alarm(struct rtc_device *rtc, + struct rtc_wkalrm *alrm); extern void rtc_update_irq(struct rtc_device *rtc, unsigned long num, unsigned long events); -- cgit v0.10.2 From a54aba87bb8e90f9e39bcfe151717b86abbbdd79 Mon Sep 17 00:00:00 2001 From: Vasily Khoruzhick Date: Thu, 24 Mar 2011 22:09:33 +0200 Subject: RTC: Fix s3c compile error due to missing s3c_rtc_setpie s3c_rtc_setpie was removed, and it resulted in compiler error: drivers/rtc/rtc-s3c.c: In function s3c_rtc_release drivers/rtc/rtc-s3c.c:339:2: error: implicit declaration of function s3c_rtc_setpie Fix it by removing s3c_rtc_release calls. [jstultz: An identical fix was also sent in by Jiri Pinkava ] CC: Thomas Gleixner CC: Alessandro Zummo Signed-off-by: Vasily Khoruzhick Signed-off-by: John Stultz diff --git a/drivers/rtc/rtc-s3c.c b/drivers/rtc/rtc-s3c.c index 7149649..b3466c4 100644 --- a/drivers/rtc/rtc-s3c.c +++ b/drivers/rtc/rtc-s3c.c @@ -336,7 +336,6 @@ static void s3c_rtc_release(struct device *dev) /* do not clear AIE here, it may be needed for wake */ - s3c_rtc_setpie(dev, 0); free_irq(s3c_rtc_alarmno, rtc_dev); free_irq(s3c_rtc_tickno, rtc_dev); } @@ -408,7 +407,6 @@ static int __devexit s3c_rtc_remove(struct platform_device *dev) platform_set_drvdata(dev, NULL); rtc_device_unregister(rtc); - s3c_rtc_setpie(&dev->dev, 0); s3c_rtc_setaie(&dev->dev, 0); clk_disable(rtc_clk); -- cgit v0.10.2 From 8c122b96866580c99e44f3f07ac93a993d964ec3 Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Fri, 18 Mar 2011 04:26:24 -0400 Subject: RTC: add missing "return 0" in new alarm func for rtc-bfin.c The new bfin_rtc_alarm_irq_enable function forgot to add a "return 0" to the end leading to the build warning: drivers/rtc/rtc-bfin.c: In function 'bfin_rtc_alarm_irq_enable': drivers/rtc/rtc-bfin.c:253: warning: control reaches end of non-void function CC: stable@kernel.org CC: Thomas Gleixner CC: Alessandro Zummo Signed-off-by: Mike Frysinger Signed-off-by: John Stultz diff --git a/drivers/rtc/rtc-bfin.c b/drivers/rtc/rtc-bfin.c index ca9cff8..f249644 100644 --- a/drivers/rtc/rtc-bfin.c +++ b/drivers/rtc/rtc-bfin.c @@ -250,6 +250,8 @@ static int bfin_rtc_alarm_irq_enable(struct device *dev, unsigned int enabled) bfin_rtc_int_set_alarm(rtc); else bfin_rtc_int_clear(~(RTC_ISTAT_ALARM | RTC_ISTAT_ALARM_DAY)); + + return 0; } static int bfin_rtc_read_time(struct device *dev, struct rtc_time *tm) -- cgit v0.10.2 From 361ae8cb57ec35ace2fa5f9695dba14eca9893e0 Mon Sep 17 00:00:00 2001 From: Nicolas Kaiser Date: Wed, 30 Mar 2011 21:14:26 +0200 Subject: xen: events: fix error checks in bind_*_to_irqhandler() Checking 'irq < 0' doesn't work when 'irq' is unsigned. The assigned bind_evtchn_to_irq() and bind_virq_to_irq() return int, so using int appears appropriate. Signed-off-by: Nicolas Kaiser Signed-off-by: Konrad Rzeszutek Wilk diff --git a/drivers/xen/events.c b/drivers/xen/events.c index 036343b..589f0a5 100644 --- a/drivers/xen/events.c +++ b/drivers/xen/events.c @@ -912,8 +912,7 @@ int bind_evtchn_to_irqhandler(unsigned int evtchn, unsigned long irqflags, const char *devname, void *dev_id) { - unsigned int irq; - int retval; + int irq, retval; irq = bind_evtchn_to_irq(evtchn); if (irq < 0) @@ -955,8 +954,7 @@ int bind_virq_to_irqhandler(unsigned int virq, unsigned int cpu, irq_handler_t handler, unsigned long irqflags, const char *devname, void *dev_id) { - unsigned int irq; - int retval; + int irq, retval; irq = bind_virq_to_irq(virq, cpu); if (irq < 0) -- cgit v0.10.2 From c8309ef6a4c52919d44bbc9743d7ea05ae8f4c7f Mon Sep 17 00:00:00 2001 From: Colin Cross Date: Wed, 30 Mar 2011 00:24:43 -0700 Subject: ARM: tegra: gpio: Fix unused variable warnings Change b0f18edaf6ee4e6fac89cae63a90bd38ad2a3418 (arm: tegra: Remove unused bogus irq enable/disable magic) introduces warnings: arch/arm/mach-tegra/gpio.c: In function 'tegra_gpio_resume': arch/arm/mach-tegra/gpio.c:260: warning: unused variable 'i' arch/arm/mach-tegra/gpio.c: In function 'tegra_gpio_suspend': arch/arm/mach-tegra/gpio.c:283: warning: unused variable 'i' Fix them, and fix a coding style issue on the same lines. Signed-off-by: Colin Cross Acked-by: Erik Gilling diff --git a/arch/arm/mach-tegra/gpio.c b/arch/arm/mach-tegra/gpio.c index 76a3f65..65a1aba 100644 --- a/arch/arm/mach-tegra/gpio.c +++ b/arch/arm/mach-tegra/gpio.c @@ -257,7 +257,8 @@ static void tegra_gpio_irq_handler(unsigned int irq, struct irq_desc *desc) void tegra_gpio_resume(void) { unsigned long flags; - int b, p, i; + int b; + int p; local_irq_save(flags); @@ -280,7 +281,8 @@ void tegra_gpio_resume(void) void tegra_gpio_suspend(void) { unsigned long flags; - int b, p, i; + int b; + int p; local_irq_save(flags); for (b = 0; b < ARRAY_SIZE(tegra_gpio_banks); b++) { -- cgit v0.10.2 From 272df502bcd1fbb1b95facc92bd861d604be8871 Mon Sep 17 00:00:00 2001 From: Ben Dooks Date: Thu, 31 Mar 2011 07:18:46 +0900 Subject: gpio/pca953x: fix error handling path in probe() call If the device fails to respond, then the error path tries to remove an interrupt that never got registered, which causes an backtrace from the interrupt handling code. Fix this by ensuring that the cleanup path has two labels and use the correct path as needed. fixes the following error: WARNING: at kernel/irq/manage.c:908 __free_irq+0x80/0x160() Trying to free already-free IRQ 0 Signed-off-by: Ben Dooks Signed-off-by: Mark Brown Signed-off-by: Grant Likely diff --git a/drivers/gpio/pca953x.c b/drivers/gpio/pca953x.c index 583e925..7630ab7 100644 --- a/drivers/gpio/pca953x.c +++ b/drivers/gpio/pca953x.c @@ -558,7 +558,7 @@ static int __devinit pca953x_probe(struct i2c_client *client, ret = gpiochip_add(&chip->gpio_chip); if (ret) - goto out_failed; + goto out_failed_irq; if (pdata->setup) { ret = pdata->setup(client, chip->gpio_chip.base, @@ -570,8 +570,9 @@ static int __devinit pca953x_probe(struct i2c_client *client, i2c_set_clientdata(client, chip); return 0; -out_failed: +out_failed_irq: pca953x_irq_teardown(chip); +out_failed: kfree(chip->dyn_pdata); kfree(chip); return ret; -- cgit v0.10.2 From 89b3600ccfb01aed6873bc499442fc0bed00bbdd Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Tue, 29 Mar 2011 07:09:20 +0000 Subject: xfs: fix unreferenced var error in xfs_buf.c Signed-off-by: Dave Chinner Signed-off-by: Alex Elder diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index 596bb2c..d917146 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -657,8 +657,6 @@ xfs_buf_readahead( xfs_off_t ioff, size_t isize) { - struct backing_dev_info *bdi; - if (bdi_read_congested(target->bt_bdi)) return; -- cgit v0.10.2 From 62f0988ee5280ac03f787e3abd70bd91366e3778 Mon Sep 17 00:00:00 2001 From: David Brown Date: Tue, 29 Mar 2011 11:48:45 -0700 Subject: msm: Remove extraneous ffa device check The qsd8x50 board file contains a few references to machine_is_... macros that are otherwise unused, and contain no machine definition. The recent purge of unused machine definitions breaks the compilation of this target. Since the machine cannot ever be used, just remove the bogus checks. Signed-off-by: David Brown diff --git a/arch/arm/mach-msm/board-qsd8x50.c b/arch/arm/mach-msm/board-qsd8x50.c index 7f56861..6a96911 100644 --- a/arch/arm/mach-msm/board-qsd8x50.c +++ b/arch/arm/mach-msm/board-qsd8x50.c @@ -160,10 +160,7 @@ static struct msm_mmc_platform_data qsd8x50_sdc1_data = { static void __init qsd8x50_init_mmc(void) { - if (machine_is_qsd8x50_ffa() || machine_is_qsd8x50a_ffa()) - vreg_mmc = vreg_get(NULL, "gp6"); - else - vreg_mmc = vreg_get(NULL, "gp5"); + vreg_mmc = vreg_get(NULL, "gp5"); if (IS_ERR(vreg_mmc)) { pr_err("vreg get for vreg_mmc failed (%ld)\n", -- cgit v0.10.2 From 893b66c39da812e7dd0d7b32aa0633e5d90d950c Mon Sep 17 00:00:00 2001 From: David Brown Date: Wed, 30 Mar 2011 11:26:57 -0700 Subject: msm: timer: fix missing return value Change af90f10d38 "ARM: 6759/1: smp: Select local timers vs broadcast timer support runtime" missed a return statement, causing a compile warning: arch/arm/mach-msm/timer.c:272: warning: 'return' with no value, in function returning non-void Trivially return 0 for success when running on cpu 0 (to match the comment and previous behavior). Signed-off-by: David Brown diff --git a/arch/arm/mach-msm/timer.c b/arch/arm/mach-msm/timer.c index 56f920c..38b95e9 100644 --- a/arch/arm/mach-msm/timer.c +++ b/arch/arm/mach-msm/timer.c @@ -269,7 +269,7 @@ int __cpuinit local_timer_setup(struct clock_event_device *evt) /* Use existing clock_event for cpu 0 */ if (!smp_processor_id()) - return; + return 0; writel(DGT_CLK_CTL_DIV_4, MSM_TMR_BASE + DGT_CLK_CTL); -- cgit v0.10.2 From ba43861277f1858472de4adfc0b28a047484da83 Mon Sep 17 00:00:00 2001 From: Peter Tyser Date: Thu, 24 Mar 2011 18:17:14 -0500 Subject: gpio/ml_ioh_gpio: Fix output value of ioh_gpio_direction_output() The ioh_gpio_direction_output() function was missing a write to set the desired output value. The function would properly set the GPIO direction, but not the output value. The value would have to manually be set with a follow up call to ioh_gpio_set(). Add the missing write so that ioh_gpio_direction_output() sets both the GPIO direction and value. Signed-off-by: Peter Tyser Tested-by: Tomoya MORINAGA Signed-off-by: Grant Likely diff --git a/drivers/gpio/ml_ioh_gpio.c b/drivers/gpio/ml_ioh_gpio.c index 7f6f01a..0a775f7 100644 --- a/drivers/gpio/ml_ioh_gpio.c +++ b/drivers/gpio/ml_ioh_gpio.c @@ -116,6 +116,7 @@ static int ioh_gpio_direction_output(struct gpio_chip *gpio, unsigned nr, reg_val |= (1 << nr); else reg_val &= ~(1 << nr); + iowrite32(reg_val, &chip->reg->regs[chip->ch].po); mutex_unlock(&chip->lock); -- cgit v0.10.2 From 88aab9341a315d81118be6b41c45e4fe32b94bc1 Mon Sep 17 00:00:00 2001 From: Peter Tyser Date: Fri, 25 Mar 2011 10:04:00 -0500 Subject: gpio/pch_gpio: Fix output value of pch_gpio_direction_output() The pch_gpio_direction_output() function was missing a write to set the desired output value. The function would properly set the GPIO direction, but not the output value. The value would have to manually be set with a follow up call to pch_gpio_set(). Add the missing write so that pch_gpio_direction_output() sets both the GPIO direction and value. Signed-off-by: Peter Tyser Tested-by: Tomoya MORINAGA Signed-off-by: Grant Likely diff --git a/drivers/gpio/pch_gpio.c b/drivers/gpio/pch_gpio.c index 2c6af87..f970a5f 100644 --- a/drivers/gpio/pch_gpio.c +++ b/drivers/gpio/pch_gpio.c @@ -105,6 +105,7 @@ static int pch_gpio_direction_output(struct gpio_chip *gpio, unsigned nr, reg_val |= (1 << nr); else reg_val &= ~(1 << nr); + iowrite32(reg_val, &chip->reg->po); mutex_unlock(&chip->lock); -- cgit v0.10.2 From b03f24567ce7caf2420b8be4c6eb74c191d59a91 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Thu, 31 Mar 2011 18:36:52 +0200 Subject: quota: Don't write quota info in dquot_commit() There's no reason to write quota info in dquot_commit(). The writing is a relict from the old days when we didn't have dquot_acquire() and dquot_release() and thus dquot_commit() could have created / removed quota structures from the file. These days dquot_commit() only updates usage counters / limits in quota structure and thus there's no need to write quota info. This also fixes an issue with journaling filesystem which didn't reserve enough space in the transaction for write of quota info (it could have been dirty at the time of dquot_commit() because of a race with other operation changing it). CC: stable@kernel.org Reported-and-tested-by: Lukas Czerner Signed-off-by: Jan Kara diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index a2a622e..b59ee61 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -442,7 +442,7 @@ EXPORT_SYMBOL(dquot_acquire); */ int dquot_commit(struct dquot *dquot) { - int ret = 0, ret2 = 0; + int ret = 0; struct quota_info *dqopt = sb_dqopt(dquot->dq_sb); mutex_lock(&dqopt->dqio_mutex); @@ -454,15 +454,10 @@ int dquot_commit(struct dquot *dquot) spin_unlock(&dq_list_lock); /* Inactive dquot can be only if there was error during read/init * => we have better not writing it */ - if (test_bit(DQ_ACTIVE_B, &dquot->dq_flags)) { + if (test_bit(DQ_ACTIVE_B, &dquot->dq_flags)) ret = dqopt->ops[dquot->dq_type]->commit_dqblk(dquot); - if (info_dirty(&dqopt->info[dquot->dq_type])) { - ret2 = dqopt->ops[dquot->dq_type]->write_file_info( - dquot->dq_sb, dquot->dq_type); - } - if (ret >= 0) - ret = ret2; - } + else + ret = -EIO; out_sem: mutex_unlock(&dqopt->dqio_mutex); return ret; -- cgit v0.10.2 From 2130ad32a03f8d2d772cbbd94eff5358d8ab9e33 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Tue, 29 Mar 2011 12:35:23 -0700 Subject: mtd: mtdswap: fix printk format warning The number is actually a sizeof(), so using %zu for it builds cleanly on 32-bit or 64-bit. drivers/mtd/mtdswap.c:1456: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'unsigned int' Signed-off-by: Randy Dunlap Signed-off-by: Artem Bityutskiy diff --git a/drivers/mtd/mtdswap.c b/drivers/mtd/mtdswap.c index 237913c..fed215c 100644 --- a/drivers/mtd/mtdswap.c +++ b/drivers/mtd/mtdswap.c @@ -1452,7 +1452,7 @@ static void mtdswap_add_mtd(struct mtd_blktrans_ops *tr, struct mtd_info *mtd) oinfo = mtd->ecclayout; if (!mtd->oobsize || !oinfo || oinfo->oobavail < MTDSWAP_OOBSIZE) { printk(KERN_ERR "%s: Not enough free bytes in OOB, " - "%d available, %lu needed.\n", + "%d available, %zu needed.\n", MTDSWAP_PREFIX, oinfo->oobavail, MTDSWAP_OOBSIZE); return; } -- cgit v0.10.2 From 042bc9c0c434b61241ceda1d332e3076e2571859 Mon Sep 17 00:00:00 2001 From: Nicolas Ferre Date: Wed, 30 Mar 2011 16:26:40 +0800 Subject: mtd: atmel_nand: trivial: change DMA usage information trace Signed-off-by: Nicolas Ferre Signed-off-by: Artem Bityutskiy diff --git a/drivers/mtd/nand/atmel_nand.c b/drivers/mtd/nand/atmel_nand.c index 6fae04b..7e88a80 100644 --- a/drivers/mtd/nand/atmel_nand.c +++ b/drivers/mtd/nand/atmel_nand.c @@ -611,7 +611,8 @@ static int __init atmel_nand_probe(struct platform_device *pdev) } } if (use_dma) - dev_info(host->dev, "Using DMA for NAND access.\n"); + dev_info(host->dev, "Using %s for DMA transfers.\n", + dma_chan_name(host->dma_chan)); else dev_info(host->dev, "No DMA support for NAND access.\n"); -- cgit v0.10.2 From cb457a4ddd4eeac9c81792a1e6a5d59b1b44abe1 Mon Sep 17 00:00:00 2001 From: Hong Xu Date: Wed, 30 Mar 2011 16:26:41 +0800 Subject: mtd: atmel_nand: fix support for CPUs that do not support DMA access use_dma was always "1" even if the CPU does not support DMA Tested on AT91SAM9261EK by Jean-Christophe PLAGNIOL-VILLARD Reported-by: Jean-Christophe PLAGNIOL-VILLARD Signed-off-by: Hong Xu Signed-off-by: Artem Bityutskiy diff --git a/drivers/mtd/nand/atmel_nand.c b/drivers/mtd/nand/atmel_nand.c index 7e88a80..f254fa6 100644 --- a/drivers/mtd/nand/atmel_nand.c +++ b/drivers/mtd/nand/atmel_nand.c @@ -599,7 +599,10 @@ static int __init atmel_nand_probe(struct platform_device *pdev) nand_chip->options |= NAND_USE_FLASH_BBT; } - if (cpu_has_dma() && use_dma) { + if (!cpu_has_dma()) + use_dma = 0; + + if (use_dma) { dma_cap_mask_t mask; dma_cap_zero(mask); -- cgit v0.10.2 From 9d51567e47a0d84f6b6e7751a4ea4f710cdcbbdf Mon Sep 17 00:00:00 2001 From: Nicolas Ferre Date: Fri, 1 Apr 2011 16:40:44 +0200 Subject: mtd: atmel_nand: modify test case for using DMA operations We have better performances not using DMA for oob operations. Modify size test so that it is using DMA for size greater than oobsize. Signed-off-by: Nicolas Ferre Signed-off-by: Artem Bityutskiy diff --git a/drivers/mtd/nand/atmel_nand.c b/drivers/mtd/nand/atmel_nand.c index f254fa6..e9fdbe4 100644 --- a/drivers/mtd/nand/atmel_nand.c +++ b/drivers/mtd/nand/atmel_nand.c @@ -280,7 +280,8 @@ static void atmel_read_buf(struct mtd_info *mtd, u8 *buf, int len) struct nand_chip *chip = mtd->priv; struct atmel_nand_host *host = chip->priv; - if (use_dma && len >= mtd->oobsize) + if (use_dma && len > mtd->oobsize) + /* only use DMA for bigger than oob size: better performances */ if (atmel_nand_dma_op(mtd, buf, len, 1) == 0) return; @@ -295,7 +296,8 @@ static void atmel_write_buf(struct mtd_info *mtd, const u8 *buf, int len) struct nand_chip *chip = mtd->priv; struct atmel_nand_host *host = chip->priv; - if (use_dma && len >= mtd->oobsize) + if (use_dma && len > mtd->oobsize) + /* only use DMA for bigger than oob size: better performances */ if (atmel_nand_dma_op(mtd, (void *)buf, len, 0) == 0) return; -- cgit v0.10.2 From 80b4f81a49809774f1b123c456fb179e472dbd0e Mon Sep 17 00:00:00 2001 From: Hong Xu Date: Thu, 31 Mar 2011 18:33:15 +0800 Subject: mtd: atmel_nand: use CPU I/O when buffer is in vmalloc(ed) region The previous way of dealing with vmalloc(ed) region by walking though the pages can not work well actually. We just fall back to CPU I/O when the buffer address is higher than `high_memory'. Signed-off-by: Nicolas Ferre Signed-off-by: Hong Xu Signed-off-by: Artem Bityutskiy diff --git a/drivers/mtd/nand/atmel_nand.c b/drivers/mtd/nand/atmel_nand.c index e9fdbe4..950646a 100644 --- a/drivers/mtd/nand/atmel_nand.c +++ b/drivers/mtd/nand/atmel_nand.c @@ -209,22 +209,8 @@ static int atmel_nand_dma_op(struct mtd_info *mtd, void *buf, int len, int err = -EIO; enum dma_data_direction dir = is_read ? DMA_FROM_DEVICE : DMA_TO_DEVICE; - if (buf >= high_memory) { - struct page *pg; - - if (((size_t)buf & PAGE_MASK) != - ((size_t)(buf + len - 1) & PAGE_MASK)) { - dev_warn(host->dev, "Buffer not fit in one page\n"); - goto err_buf; - } - - pg = vmalloc_to_page(buf); - if (pg == 0) { - dev_err(host->dev, "Failed to vmalloc_to_page\n"); - goto err_buf; - } - p = page_address(pg) + ((size_t)buf & ~PAGE_MASK); - } + if (buf >= high_memory) + goto err_buf; dma_dev = host->dma_chan->device; -- cgit v0.10.2 From bda1a730ab303609897a91c554542d28496d3684 Mon Sep 17 00:00:00 2001 From: Corentin Chary Date: Wed, 30 Mar 2011 16:32:31 +0200 Subject: asus-laptop: remove removed features from feature-removal-schedule.txt commit 3b81cf9d558c57406b4ed9b0d2639113d1d428b6 removed those features. Signed-off-by: Corentin Chary Signed-off-by: Matthew Garrett diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index 274b32d..492e81d 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -387,26 +387,6 @@ Who: Tejun Heo ---------------------------- -What: Support for lcd_switch and display_get in asus-laptop driver -When: March 2010 -Why: These two features use non-standard interfaces. There are the - only features that really need multiple path to guess what's - the right method name on a specific laptop. - - Removing them will allow to remove a lot of code an significantly - clean the drivers. - - This will affect the backlight code which won't be able to know - if the backlight is on or off. The platform display file will also be - write only (like the one in eeepc-laptop). - - This should'nt affect a lot of user because they usually know - when their display is on or off. - -Who: Corentin Chary - ----------------------------- - What: sysfs-class-rfkill state file When: Feb 2014 Files: net/rfkill/core.c -- cgit v0.10.2 From 58a9f397ac9288403d600b22b8472c0d164ee8cb Mon Sep 17 00:00:00 2001 From: Corentin Chary Date: Wed, 30 Mar 2011 16:32:32 +0200 Subject: asus-wmi: swap input name and phys Signed-off-by: Corentin Chary Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/asus-wmi.c b/drivers/platform/x86/asus-wmi.c index efc776c..832a3fd7 100644 --- a/drivers/platform/x86/asus-wmi.c +++ b/drivers/platform/x86/asus-wmi.c @@ -201,8 +201,8 @@ static int asus_wmi_input_init(struct asus_wmi *asus) if (!asus->inputdev) return -ENOMEM; - asus->inputdev->name = asus->driver->input_phys; - asus->inputdev->phys = asus->driver->input_name; + asus->inputdev->name = asus->driver->input_name; + asus->inputdev->phys = asus->driver->input_phys; asus->inputdev->id.bustype = BUS_HOST; asus->inputdev->dev.parent = &asus->platform_device->dev; -- cgit v0.10.2 From 9481472559b3ecd366390786628c2fb211d1477b Mon Sep 17 00:00:00 2001 From: Corentin Chary Date: Wed, 30 Mar 2011 16:32:33 +0200 Subject: eeepc-wmi: add keys found on EeePC 1215T Signed-off-by: Corentin Chary Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/eeepc-wmi.c b/drivers/platform/x86/eeepc-wmi.c index 0ddc434..649dcad 100644 --- a/drivers/platform/x86/eeepc-wmi.c +++ b/drivers/platform/x86/eeepc-wmi.c @@ -67,9 +67,11 @@ static const struct key_entry eeepc_wmi_keymap[] = { { KE_KEY, 0x82, { KEY_CAMERA } }, { KE_KEY, 0x83, { KEY_CAMERA_ZOOMIN } }, { KE_KEY, 0x88, { KEY_WLAN } }, + { KE_KEY, 0xbd, { KEY_CAMERA } }, { KE_KEY, 0xcc, { KEY_SWITCHVIDEOMODE } }, { KE_KEY, 0xe0, { KEY_PROG1 } }, /* Task Manager */ { KE_KEY, 0xe1, { KEY_F14 } }, /* Change Resolution */ + { KE_KEY, 0xe8, { KEY_SCREENLOCK } }, { KE_KEY, 0xe9, { KEY_BRIGHTNESS_ZERO } }, { KE_KEY, 0xeb, { KEY_CAMERA_ZOOMOUT } }, { KE_KEY, 0xec, { KEY_CAMERA_UP } }, -- cgit v0.10.2 From a0bcaee99df69b6d4d4dfd30bd40ac24b5ad3e22 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Wed, 30 Mar 2011 15:20:22 -0700 Subject: msi-laptop: fix config-dependent build error The msi-laptop driver uses input_*() and sparse_keymap_*() interfaces. It should depend on the INPUT subsystem being present and select INPUT_SPARSEKMAP so that those interfaces are present. ERROR: "input_free_device" [drivers/platform/x86/msi-laptop.ko] undefined! ERROR: "input_register_device" [drivers/platform/x86/msi-laptop.ko] undefined! ERROR: "sparse_keymap_setup" [drivers/platform/x86/msi-laptop.ko] undefined! ERROR: "input_allocate_device" [drivers/platform/x86/msi-laptop.ko] undefined! ERROR: "sparse_keymap_report_event" [drivers/platform/x86/msi-laptop.ko] undefined! ERROR: "input_unregister_device" [drivers/platform/x86/msi-laptop.ko] undefined! ERROR: "sparse_keymap_free" [drivers/platform/x86/msi-laptop.ko] undefined! Signed-off-by: Randy Dunlap Cc: Matthew Garrett Cc: "Lee, Chun-Yi" Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/Kconfig b/drivers/platform/x86/Kconfig index 2ee442c..0485e39 100644 --- a/drivers/platform/x86/Kconfig +++ b/drivers/platform/x86/Kconfig @@ -187,7 +187,8 @@ config MSI_LAPTOP depends on ACPI depends on BACKLIGHT_CLASS_DEVICE depends on RFKILL - depends on SERIO_I8042 + depends on INPUT && SERIO_I8042 + select INPUT_SPARSEKMAP ---help--- This is a driver for laptops built by MSI (MICRO-STAR INTERNATIONAL): -- cgit v0.10.2 From 906c3b616dcf8e64b11d7d665d62f5e9940f4d46 Mon Sep 17 00:00:00 2001 From: Nicolas Kaiser Date: Wed, 30 Mar 2011 15:59:11 +0200 Subject: arm: tegra: fix error check in tegra2_clocks.c Checking 'rate < 0' doesn't work because 'rate' is unsigned. Signed-off-by: Nicolas Kaiser Signed-off-by: Colin Cross diff --git a/arch/arm/mach-tegra/tegra2_clocks.c b/arch/arm/mach-tegra/tegra2_clocks.c index 6d7c4ee..4459470 100644 --- a/arch/arm/mach-tegra/tegra2_clocks.c +++ b/arch/arm/mach-tegra/tegra2_clocks.c @@ -1362,14 +1362,15 @@ static int tegra_clk_shared_bus_set_rate(struct clk *c, unsigned long rate) { unsigned long flags; int ret; + long new_rate = rate; - rate = clk_round_rate(c->parent, rate); - if (rate < 0) - return rate; + new_rate = clk_round_rate(c->parent, new_rate); + if (new_rate < 0) + return new_rate; spin_lock_irqsave(&c->parent->spinlock, flags); - c->u.shared_bus_user.rate = rate; + c->u.shared_bus_user.rate = new_rate; ret = tegra_clk_shared_bus_update(c->parent); spin_unlock_irqrestore(&c->parent->spinlock, flags); -- cgit v0.10.2 From 29ea23ff905d07d8559bac69cca46f4bbf20038c Mon Sep 17 00:00:00 2001 From: Russell King Date: Sat, 2 Apr 2011 10:08:55 +0100 Subject: ARM: Make consolidated PM sleep code depend on PM_SLEEP CONFIG_PM is now set whenever we support either runtime PM in addition to suspend and hibernate. This causes build errors when runtime PM is enabled on a platform, but the CPU does not have the appropriate support for suspend. So, switch this code to use CONFIG_PM_SLEEP rather than CONFIG_PM to allow runtime PM to be enabled without causing build errors. Signed-off-by: Russell King diff --git a/arch/arm/kernel/Makefile b/arch/arm/kernel/Makefile index 74554f1..8d95446 100644 --- a/arch/arm/kernel/Makefile +++ b/arch/arm/kernel/Makefile @@ -29,7 +29,7 @@ obj-$(CONFIG_MODULES) += armksyms.o module.o obj-$(CONFIG_ARTHUR) += arthur.o obj-$(CONFIG_ISA_DMA) += dma-isa.o obj-$(CONFIG_PCI) += bios32.o isa.o -obj-$(CONFIG_PM) += sleep.o +obj-$(CONFIG_PM_SLEEP) += sleep.o obj-$(CONFIG_HAVE_SCHED_CLOCK) += sched_clock.o obj-$(CONFIG_SMP) += smp.o smp_tlb.o obj-$(CONFIG_HAVE_ARM_SCU) += smp_scu.o diff --git a/arch/arm/mm/proc-arm920.S b/arch/arm/mm/proc-arm920.S index 219980e..394b623 100644 --- a/arch/arm/mm/proc-arm920.S +++ b/arch/arm/mm/proc-arm920.S @@ -390,7 +390,7 @@ ENTRY(cpu_arm920_set_pte_ext) /* Suspend/resume support: taken from arch/arm/plat-s3c24xx/sleep.S */ .globl cpu_arm920_suspend_size .equ cpu_arm920_suspend_size, 4 * 3 -#ifdef CONFIG_PM +#ifdef CONFIG_PM_SLEEP ENTRY(cpu_arm920_do_suspend) stmfd sp!, {r4 - r7, lr} mrc p15, 0, r4, c13, c0, 0 @ PID diff --git a/arch/arm/mm/proc-arm926.S b/arch/arm/mm/proc-arm926.S index 6a4bdb2..0ed85d9 100644 --- a/arch/arm/mm/proc-arm926.S +++ b/arch/arm/mm/proc-arm926.S @@ -404,7 +404,7 @@ ENTRY(cpu_arm926_set_pte_ext) /* Suspend/resume support: taken from arch/arm/plat-s3c24xx/sleep.S */ .globl cpu_arm926_suspend_size .equ cpu_arm926_suspend_size, 4 * 3 -#ifdef CONFIG_PM +#ifdef CONFIG_PM_SLEEP ENTRY(cpu_arm926_do_suspend) stmfd sp!, {r4 - r7, lr} mrc p15, 0, r4, c13, c0, 0 @ PID diff --git a/arch/arm/mm/proc-sa1100.S b/arch/arm/mm/proc-sa1100.S index 74483d1..184a9c9 100644 --- a/arch/arm/mm/proc-sa1100.S +++ b/arch/arm/mm/proc-sa1100.S @@ -171,7 +171,7 @@ ENTRY(cpu_sa1100_set_pte_ext) .globl cpu_sa1100_suspend_size .equ cpu_sa1100_suspend_size, 4*4 -#ifdef CONFIG_PM +#ifdef CONFIG_PM_SLEEP ENTRY(cpu_sa1100_do_suspend) stmfd sp!, {r4 - r7, lr} mrc p15, 0, r4, c3, c0, 0 @ domain ID diff --git a/arch/arm/mm/proc-v6.S b/arch/arm/mm/proc-v6.S index 832b6bd..3874412 100644 --- a/arch/arm/mm/proc-v6.S +++ b/arch/arm/mm/proc-v6.S @@ -124,7 +124,7 @@ ENTRY(cpu_v6_set_pte_ext) /* Suspend/resume support: taken from arch/arm/mach-s3c64xx/sleep.S */ .globl cpu_v6_suspend_size .equ cpu_v6_suspend_size, 4 * 8 -#ifdef CONFIG_PM +#ifdef CONFIG_PM_SLEEP ENTRY(cpu_v6_do_suspend) stmfd sp!, {r4 - r11, lr} mrc p15, 0, r4, c13, c0, 0 @ FCSE/PID diff --git a/arch/arm/mm/proc-v7.S b/arch/arm/mm/proc-v7.S index 262fa88..713cea1 100644 --- a/arch/arm/mm/proc-v7.S +++ b/arch/arm/mm/proc-v7.S @@ -211,7 +211,7 @@ cpu_v7_name: /* Suspend/resume support: derived from arch/arm/mach-s5pv210/sleep.S */ .globl cpu_v7_suspend_size .equ cpu_v7_suspend_size, 4 * 8 -#ifdef CONFIG_PM +#ifdef CONFIG_PM_SLEEP ENTRY(cpu_v7_do_suspend) stmfd sp!, {r4 - r11, lr} mrc p15, 0, r4, c13, c0, 0 @ FCSE/PID diff --git a/arch/arm/mm/proc-xsc3.S b/arch/arm/mm/proc-xsc3.S index 63d8b20..5962136 100644 --- a/arch/arm/mm/proc-xsc3.S +++ b/arch/arm/mm/proc-xsc3.S @@ -417,7 +417,7 @@ ENTRY(cpu_xsc3_set_pte_ext) .globl cpu_xsc3_suspend_size .equ cpu_xsc3_suspend_size, 4 * 8 -#ifdef CONFIG_PM +#ifdef CONFIG_PM_SLEEP ENTRY(cpu_xsc3_do_suspend) stmfd sp!, {r4 - r10, lr} mrc p14, 0, r4, c6, c0, 0 @ clock configuration, for turbo mode diff --git a/arch/arm/mm/proc-xscale.S b/arch/arm/mm/proc-xscale.S index 086038c..ce233bc 100644 --- a/arch/arm/mm/proc-xscale.S +++ b/arch/arm/mm/proc-xscale.S @@ -518,7 +518,7 @@ ENTRY(cpu_xscale_set_pte_ext) .globl cpu_xscale_suspend_size .equ cpu_xscale_suspend_size, 4 * 7 -#ifdef CONFIG_PM +#ifdef CONFIG_PM_SLEEP ENTRY(cpu_xscale_do_suspend) stmfd sp!, {r4 - r10, lr} mrc p14, 0, r4, c6, c0, 0 @ clock configuration, for turbo mode -- cgit v0.10.2 From 6a7861825f79f09213ef81b3c468f6f2e86f408e Mon Sep 17 00:00:00 2001 From: Russell King Date: Sat, 2 Apr 2011 10:15:28 +0100 Subject: ARM: Only allow PM_SLEEP with CPUs which support suspend Offering CONFIG_PM_SLEEP for CPUs which do not support suspend leads to build errors, so only set CONFIG_ARCH_SUSPEND_POSSIBLE if we have a CPU selected which supports suspend. Signed-off-by: Russell King diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 5b9f78b..9954c9b 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -2012,6 +2012,8 @@ source "kernel/power/Kconfig" config ARCH_SUSPEND_POSSIBLE depends on !ARCH_S5P64X0 && !ARCH_S5P6442 + depends on CPU_ARM920T || CPU_ARM926T || CPU_SA1100 || \ + CPU_V6 || CPU_V6K || CPU_V7 || CPU_XSC3 || CPU_XSCALE def_bool y endmenu -- cgit v0.10.2 From 97c96c076b8ef86533f2427ea145a9cd0ec5564f Mon Sep 17 00:00:00 2001 From: Vasily Khoruzhick Date: Sat, 2 Apr 2011 10:54:47 +0300 Subject: ASoC: PXA: Fix oops in __pxa2xx_pcm_prepare pxa2xx_pcm_hw_free frees dma channel and sets prtd->dma_ch to -1, but does not set prtd->params to NULL, so if pxa2xx_pcm_hw_params will be called immediately, it leaves prtd->dma_ch initialized with -1, and it results in oops in __pxa2xx_pcm_prepare. This bug is triggered via SDL. This patch adds check for prtd->dma_ch to __pxa2xx_pcm_prepare and cleans prtd->params, so now it works properly. Signed-off-by: Vasily Khoruzhick Signed-off-by: Mark Brown diff --git a/sound/arm/pxa2xx-pcm-lib.c b/sound/arm/pxa2xx-pcm-lib.c index 8808b82..76e0d56 100644 --- a/sound/arm/pxa2xx-pcm-lib.c +++ b/sound/arm/pxa2xx-pcm-lib.c @@ -140,6 +140,9 @@ int __pxa2xx_pcm_prepare(struct snd_pcm_substream *substream) if (!prtd || !prtd->params) return 0; + if (prtd->dma_ch == -1) + return -EINVAL; + DCSR(prtd->dma_ch) &= ~DCSR_RUN; DCSR(prtd->dma_ch) = 0; DCMD(prtd->dma_ch) = 0; diff --git a/sound/soc/pxa/pxa2xx-pcm.c b/sound/soc/pxa/pxa2xx-pcm.c index 02fb664..2ce0b2d 100644 --- a/sound/soc/pxa/pxa2xx-pcm.c +++ b/sound/soc/pxa/pxa2xx-pcm.c @@ -65,6 +65,7 @@ static int pxa2xx_pcm_hw_free(struct snd_pcm_substream *substream) if (prtd->dma_ch >= 0) { pxa_free_dma(prtd->dma_ch); prtd->dma_ch = -1; + prtd->params = NULL; } return 0; -- cgit v0.10.2 From 00b317a41c5428b13eb7e5b4bbc691b1aa7afa80 Mon Sep 17 00:00:00 2001 From: Stephen Warren Date: Fri, 1 Apr 2011 14:50:44 -0600 Subject: ASoC: format_register_str: Don't clip register values wordsize is used as the textual width of a register address. regsize is used as the textual width of a register value. The assignments to these values were swapped. In the case of WM8903, which has 8-bit register addresses and 16-bit register values, this caused the register values to be clipped to 2 digits instead of the full 4. Signed-off-by: Stephen Warren Signed-off-by: Mark Brown diff --git a/sound/soc/soc-core.c b/sound/soc/soc-core.c index 4dda589..b76b74d 100644 --- a/sound/soc/soc-core.c +++ b/sound/soc/soc-core.c @@ -92,8 +92,8 @@ static int min_bytes_needed(unsigned long val) static int format_register_str(struct snd_soc_codec *codec, unsigned int reg, char *buf, size_t len) { - int wordsize = codec->driver->reg_word_size * 2; - int regsize = min_bytes_needed(codec->driver->reg_cache_size) * 2; + int wordsize = min_bytes_needed(codec->driver->reg_cache_size) * 2; + int regsize = codec->driver->reg_word_size * 2; int ret; char tmpbuf[len + 1]; char regbuf[regsize + 1]; @@ -132,8 +132,8 @@ static ssize_t soc_codec_reg_show(struct snd_soc_codec *codec, char *buf, size_t total = 0; loff_t p = 0; - wordsize = codec->driver->reg_word_size * 2; - regsize = min_bytes_needed(codec->driver->reg_cache_size) * 2; + wordsize = min_bytes_needed(codec->driver->reg_cache_size) * 2; + regsize = codec->driver->reg_word_size * 2; len = wordsize + regsize + 2 + 1; -- cgit v0.10.2 From 65a94b1417badaee2ed923fd07e23b651b25c729 Mon Sep 17 00:00:00 2001 From: Jan Glauber Date: Mon, 4 Apr 2011 09:43:29 +0200 Subject: [S390] oprofile s390: prevent stack corruption Prevent stack corruption by memcpy which copies more bytes then available at the destination. While at it use the new test_facility to test for the facility bit. Signed-off-by: Jan Glauber Signed-off-by: Martin Schwidefsky diff --git a/arch/s390/oprofile/hwsampler.c b/arch/s390/oprofile/hwsampler.c index 3d48f4d..4952872 100644 --- a/arch/s390/oprofile/hwsampler.c +++ b/arch/s390/oprofile/hwsampler.c @@ -517,12 +517,8 @@ stop_exit: static int check_hardware_prerequisites(void) { - unsigned long long facility_bits[2]; - - memcpy(facility_bits, S390_lowcore.stfle_fac_list, 32); - if (!(facility_bits[1] & (1ULL << 59))) + if (!test_facility(68)) return -EOPNOTSUPP; - return 0; } /* -- cgit v0.10.2 From 034e9e966cd86e5d8062cf0b3daba4bde187986c Mon Sep 17 00:00:00 2001 From: Michael Holzheu Date: Mon, 4 Apr 2011 09:43:30 +0200 Subject: [S390] Fix parameter passing for smp_switch_to_cpu() After the execution has been switched to the destination CPU, the target function is called with the wrong parameter. According to the C calling convention on s390, the first parameter should be loaded into register 2. Currently in smp_restart_cpu() it is stored in register 3. To fix this, we load the parameter into the correct register 2. Signed-off-by: Michael Holzheu Signed-off-by: Martin Schwidefsky diff --git a/arch/s390/kernel/switch_cpu.S b/arch/s390/kernel/switch_cpu.S index 469f11b..20530dd 100644 --- a/arch/s390/kernel/switch_cpu.S +++ b/arch/s390/kernel/switch_cpu.S @@ -46,7 +46,9 @@ smp_restart_cpu: ltr %r4,%r4 /* New stack ? */ jz 1f lr %r15,%r4 -1: basr %r14,%r2 +1: lr %r14,%r2 /* r14: Function to call */ + lr %r2,%r3 /* r2 : Parameter for function*/ + basr %r14,%r14 /* Call function */ .gprregs_addr: .long .gprregs diff --git a/arch/s390/kernel/switch_cpu64.S b/arch/s390/kernel/switch_cpu64.S index d94aacc..5be3f43 100644 --- a/arch/s390/kernel/switch_cpu64.S +++ b/arch/s390/kernel/switch_cpu64.S @@ -42,7 +42,9 @@ smp_restart_cpu: ltgr %r4,%r4 /* New stack ? */ jz 1f lgr %r15,%r4 -1: basr %r14,%r2 +1: lgr %r14,%r2 /* r14: Function to call */ + lgr %r2,%r3 /* r2 : Parameter for function*/ + basr %r14,%r14 /* Call function */ .section .data,"aw",@progbits .gprregs: -- cgit v0.10.2 From aa5c8df398266a141fb5ff0e77cbc7904a3e0648 Mon Sep 17 00:00:00 2001 From: Sebastian Ott Date: Mon, 4 Apr 2011 09:43:31 +0200 Subject: [S390] qdio: fix init sequence Reorder the initialization sequence of the qdio module to avoid writing to an uninitialized debug feature entry. Also reorder the exit function to restore a consistent cleanup path. Signed-off-by: Sebastian Ott Signed-off-by: Martin Schwidefsky diff --git a/drivers/s390/cio/qdio_main.c b/drivers/s390/cio/qdio_main.c index 479c665..c532ba9 100644 --- a/drivers/s390/cio/qdio_main.c +++ b/drivers/s390/cio/qdio_main.c @@ -1649,26 +1649,26 @@ static int __init init_QDIO(void) { int rc; - rc = qdio_setup_init(); + rc = qdio_debug_init(); if (rc) return rc; + rc = qdio_setup_init(); + if (rc) + goto out_debug; rc = tiqdio_allocate_memory(); if (rc) goto out_cache; - rc = qdio_debug_init(); - if (rc) - goto out_ti; rc = tiqdio_register_thinints(); if (rc) - goto out_debug; + goto out_ti; return 0; -out_debug: - qdio_debug_exit(); out_ti: tiqdio_free_memory(); out_cache: qdio_setup_exit(); +out_debug: + qdio_debug_exit(); return rc; } @@ -1676,8 +1676,8 @@ static void __exit exit_QDIO(void) { tiqdio_unregister_thinints(); tiqdio_free_memory(); - qdio_debug_exit(); qdio_setup_exit(); + qdio_debug_exit(); } module_init(init_QDIO); -- cgit v0.10.2 From a2fc8485f838ccd2ce5db690f81ac086489a9e7e Mon Sep 17 00:00:00 2001 From: Peter Oberparleiter Date: Mon, 4 Apr 2011 09:43:32 +0200 Subject: [S390] cio: prevent purging of CCW devices in the online state The cio_ignore purge function is intended to only remove CCW devices which are in the offline state. There is a time frame after the purge function finished where a CCW device is scheduled for removal but still accessible. When the device is set online during this time frame, it may first appear online before it is then removed. Fix this by preventing that CCW devices can be set online while there is work (such as removal triggered by the purge function) for it pending. Also ensure that the purge function does not schedule devices for removal which are in the process of being set online. Signed-off-by: Peter Oberparleiter Signed-off-by: Martin Schwidefsky diff --git a/drivers/s390/cio/device.c b/drivers/s390/cio/device.c index df14c51..8e04c00 100644 --- a/drivers/s390/cio/device.c +++ b/drivers/s390/cio/device.c @@ -541,15 +541,24 @@ static ssize_t online_store (struct device *dev, struct device_attribute *attr, int force, ret; unsigned long i; - if (!dev_fsm_final_state(cdev) && - cdev->private->state != DEV_STATE_DISCONNECTED) - return -EAGAIN; + /* Prevent conflict between multiple on-/offline processing requests. */ if (atomic_cmpxchg(&cdev->private->onoff, 0, 1) != 0) return -EAGAIN; + /* Prevent conflict between internal I/Os and on-/offline processing. */ + if (!dev_fsm_final_state(cdev) && + cdev->private->state != DEV_STATE_DISCONNECTED) { + ret = -EAGAIN; + goto out_onoff; + } + /* Prevent conflict between pending work and on-/offline processing.*/ + if (work_pending(&cdev->private->todo_work)) { + ret = -EAGAIN; + goto out_onoff; + } if (cdev->drv && !try_module_get(cdev->drv->driver.owner)) { - atomic_set(&cdev->private->onoff, 0); - return -EINVAL; + ret = -EINVAL; + goto out_onoff; } if (!strncmp(buf, "force\n", count)) { force = 1; @@ -574,6 +583,7 @@ static ssize_t online_store (struct device *dev, struct device_attribute *attr, out: if (cdev->drv) module_put(cdev->drv->driver.owner); +out_onoff: atomic_set(&cdev->private->onoff, 0); return (ret < 0) ? ret : count; } @@ -1311,10 +1321,12 @@ static int purge_fn(struct device *dev, void *data) spin_lock_irq(cdev->ccwlock); if (is_blacklisted(id->ssid, id->devno) && - (cdev->private->state == DEV_STATE_OFFLINE)) { + (cdev->private->state == DEV_STATE_OFFLINE) && + (atomic_cmpxchg(&cdev->private->onoff, 0, 1) == 0)) { CIO_MSG_EVENT(3, "ccw: purging 0.%x.%04x\n", id->ssid, id->devno); ccw_device_sched_todo(cdev, CDEV_TODO_UNREG); + atomic_set(&cdev->private->onoff, 0); } spin_unlock_irq(cdev->ccwlock); /* Abort loop in case of pending signal. */ -- cgit v0.10.2 From 8838101183bba239f100b0cfe31d9ebbfc2f1bd4 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Mon, 4 Apr 2011 09:43:33 +0200 Subject: [S390] compile fix for latest binutils The latest binutils won't accept the stfl instruction with march=g5 which is the correct behaviour. Unfortunately head.S is assembled with -march=g5 even if the target cpu is z900 or later. To get 31-bit kernels compiled again the easiest fix is to use the .insn notation for the stfl instruction in head.S. Signed-off-by: Martin Schwidefsky diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S index 7061398..fb317bf 100644 --- a/arch/s390/kernel/head.S +++ b/arch/s390/kernel/head.S @@ -460,7 +460,7 @@ startup: #ifndef CONFIG_MARCH_G5 # check capabilities against MARCH_{G5,Z900,Z990,Z9_109,Z10} xc __LC_STFL_FAC_LIST(8),__LC_STFL_FAC_LIST - stfl __LC_STFL_FAC_LIST # store facility list + .insn s,0xb2b10000,__LC_STFL_FAC_LIST # store facility list tm __LC_STFL_FAC_LIST,0x01 # stfle available ? jz 0f la %r0,0 -- cgit v0.10.2 From 512d06b5b64fb422d90f199b1be188082729edf9 Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Mon, 4 Apr 2011 15:18:45 +0200 Subject: netfilter: ipset: list:set timeout variant fixes - the timeout value was actually not set - the garbage collector was broken The variant is fixed, the tests to the ipset testsuite are added. Signed-off-by: Jozsef Kadlecsik Signed-off-by: Patrick McHardy diff --git a/net/netfilter/ipset/ip_set_list_set.c b/net/netfilter/ipset/ip_set_list_set.c index a47c329..f4a46c0 100644 --- a/net/netfilter/ipset/ip_set_list_set.c +++ b/net/netfilter/ipset/ip_set_list_set.c @@ -43,14 +43,19 @@ struct list_set { static inline struct set_elem * list_set_elem(const struct list_set *map, u32 id) { - return (struct set_elem *)((char *)map->members + id * map->dsize); + return (struct set_elem *)((void *)map->members + id * map->dsize); +} + +static inline struct set_telem * +list_set_telem(const struct list_set *map, u32 id) +{ + return (struct set_telem *)((void *)map->members + id * map->dsize); } static inline bool list_set_timeout(const struct list_set *map, u32 id) { - const struct set_telem *elem = - (const struct set_telem *) list_set_elem(map, id); + const struct set_telem *elem = list_set_telem(map, id); return ip_set_timeout_test(elem->timeout); } @@ -58,19 +63,11 @@ list_set_timeout(const struct list_set *map, u32 id) static inline bool list_set_expired(const struct list_set *map, u32 id) { - const struct set_telem *elem = - (const struct set_telem *) list_set_elem(map, id); + const struct set_telem *elem = list_set_telem(map, id); return ip_set_timeout_expired(elem->timeout); } -static inline int -list_set_exist(const struct set_telem *elem) -{ - return elem->id != IPSET_INVALID_ID && - !ip_set_timeout_expired(elem->timeout); -} - /* Set list without and with timeout */ static int @@ -146,11 +143,11 @@ list_elem_tadd(struct list_set *map, u32 i, ip_set_id_t id, struct set_telem *e; for (; i < map->size; i++) { - e = (struct set_telem *)list_set_elem(map, i); + e = list_set_telem(map, i); swap(e->id, id); + swap(e->timeout, timeout); if (e->id == IPSET_INVALID_ID) break; - swap(e->timeout, timeout); } } @@ -164,7 +161,7 @@ list_set_add(struct list_set *map, u32 i, ip_set_id_t id, /* Last element replaced: e.g. add new,before,last */ ip_set_put_byindex(e->id); if (with_timeout(map->timeout)) - list_elem_tadd(map, i, id, timeout); + list_elem_tadd(map, i, id, ip_set_timeout_set(timeout)); else list_elem_add(map, i, id); @@ -172,11 +169,11 @@ list_set_add(struct list_set *map, u32 i, ip_set_id_t id, } static int -list_set_del(struct list_set *map, ip_set_id_t id, u32 i) +list_set_del(struct list_set *map, u32 i) { struct set_elem *a = list_set_elem(map, i), *b; - ip_set_put_byindex(id); + ip_set_put_byindex(a->id); for (; i < map->size - 1; i++) { b = list_set_elem(map, i + 1); @@ -308,11 +305,11 @@ list_set_uadt(struct ip_set *set, struct nlattr *tb[], (before == 0 || (before > 0 && next_id_eq(map, i, refid)))) - ret = list_set_del(map, id, i); + ret = list_set_del(map, i); else if (before < 0 && elem->id == refid && next_id_eq(map, i, id)) - ret = list_set_del(map, id, i + 1); + ret = list_set_del(map, i + 1); } break; default: @@ -460,17 +457,15 @@ list_set_gc(unsigned long ul_set) struct list_set *map = set->data; struct set_telem *e; u32 i; - - /* We run parallel with other readers (test element) - * but adding/deleting new entries is locked out */ - read_lock_bh(&set->lock); - for (i = map->size - 1; i >= 0; i--) { - e = (struct set_telem *) list_set_elem(map, i); - if (e->id != IPSET_INVALID_ID && - list_set_expired(map, i)) - list_set_del(map, e->id, i); + + /* nfnl_lock should be called */ + write_lock_bh(&set->lock); + for (i = 0; i < map->size; i++) { + e = list_set_telem(map, i); + if (e->id != IPSET_INVALID_ID && list_set_expired(map, i)) + list_set_del(map, i); } - read_unlock_bh(&set->lock); + write_unlock_bh(&set->lock); map->gc.expires = jiffies + IPSET_GC_PERIOD(map->timeout) * HZ; add_timer(&map->gc); -- cgit v0.10.2 From 2f9f28b212a2bd4948c8ceaaec33ce0123632129 Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Mon, 4 Apr 2011 15:19:25 +0200 Subject: netfilter: ipset: references are protected by rwlock instead of mutex The timeout variant of the list:set type must reference the member sets. However, its garbage collector runs at timer interrupt so the mutex protection of the references is a no go. Therefore the reference protection is converted to rwlock. Signed-off-by: Jozsef Kadlecsik Signed-off-by: Patrick McHardy diff --git a/include/linux/netfilter/ipset/ip_set.h b/include/linux/netfilter/ipset/ip_set.h index ec333d8..5a262e3 100644 --- a/include/linux/netfilter/ipset/ip_set.h +++ b/include/linux/netfilter/ipset/ip_set.h @@ -293,7 +293,7 @@ struct ip_set { /* Lock protecting the set data */ rwlock_t lock; /* References to the set */ - atomic_t ref; + u32 ref; /* The core set type */ struct ip_set_type *type; /* The type variant doing the real job */ diff --git a/include/linux/netfilter/ipset/ip_set_ahash.h b/include/linux/netfilter/ipset/ip_set_ahash.h index ec9d9be..a0196ac 100644 --- a/include/linux/netfilter/ipset/ip_set_ahash.h +++ b/include/linux/netfilter/ipset/ip_set_ahash.h @@ -515,8 +515,7 @@ type_pf_head(struct ip_set *set, struct sk_buff *skb) if (h->netmask != HOST_MASK) NLA_PUT_U8(skb, IPSET_ATTR_NETMASK, h->netmask); #endif - NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, - htonl(atomic_read(&set->ref) - 1)); + NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)); NLA_PUT_NET32(skb, IPSET_ATTR_MEMSIZE, htonl(memsize)); if (with_timeout(h->timeout)) NLA_PUT_NET32(skb, IPSET_ATTR_TIMEOUT, htonl(h->timeout)); diff --git a/net/netfilter/ipset/ip_set_bitmap_ip.c b/net/netfilter/ipset/ip_set_bitmap_ip.c index bca9699..a113ff0 100644 --- a/net/netfilter/ipset/ip_set_bitmap_ip.c +++ b/net/netfilter/ipset/ip_set_bitmap_ip.c @@ -338,8 +338,7 @@ bitmap_ip_head(struct ip_set *set, struct sk_buff *skb) NLA_PUT_IPADDR4(skb, IPSET_ATTR_IP_TO, htonl(map->last_ip)); if (map->netmask != 32) NLA_PUT_U8(skb, IPSET_ATTR_NETMASK, map->netmask); - NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, - htonl(atomic_read(&set->ref) - 1)); + NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)); NLA_PUT_NET32(skb, IPSET_ATTR_MEMSIZE, htonl(sizeof(*map) + map->memsize)); if (with_timeout(map->timeout)) diff --git a/net/netfilter/ipset/ip_set_bitmap_ipmac.c b/net/netfilter/ipset/ip_set_bitmap_ipmac.c index 5e79017..00a3324 100644 --- a/net/netfilter/ipset/ip_set_bitmap_ipmac.c +++ b/net/netfilter/ipset/ip_set_bitmap_ipmac.c @@ -434,8 +434,7 @@ bitmap_ipmac_head(struct ip_set *set, struct sk_buff *skb) goto nla_put_failure; NLA_PUT_IPADDR4(skb, IPSET_ATTR_IP, htonl(map->first_ip)); NLA_PUT_IPADDR4(skb, IPSET_ATTR_IP_TO, htonl(map->last_ip)); - NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, - htonl(atomic_read(&set->ref) - 1)); + NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)); NLA_PUT_NET32(skb, IPSET_ATTR_MEMSIZE, htonl(sizeof(*map) + (map->last_ip - map->first_ip + 1) * map->dsize)); diff --git a/net/netfilter/ipset/ip_set_bitmap_port.c b/net/netfilter/ipset/ip_set_bitmap_port.c index 165f09b..6b38eb8 100644 --- a/net/netfilter/ipset/ip_set_bitmap_port.c +++ b/net/netfilter/ipset/ip_set_bitmap_port.c @@ -320,8 +320,7 @@ bitmap_port_head(struct ip_set *set, struct sk_buff *skb) goto nla_put_failure; NLA_PUT_NET16(skb, IPSET_ATTR_PORT, htons(map->first_port)); NLA_PUT_NET16(skb, IPSET_ATTR_PORT_TO, htons(map->last_port)); - NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, - htonl(atomic_read(&set->ref) - 1)); + NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)); NLA_PUT_NET32(skb, IPSET_ATTR_MEMSIZE, htonl(sizeof(*map) + map->memsize)); if (with_timeout(map->timeout)) diff --git a/net/netfilter/ipset/ip_set_core.c b/net/netfilter/ipset/ip_set_core.c index d6b4823..e88ac3c 100644 --- a/net/netfilter/ipset/ip_set_core.c +++ b/net/netfilter/ipset/ip_set_core.c @@ -26,6 +26,7 @@ static LIST_HEAD(ip_set_type_list); /* all registered set types */ static DEFINE_MUTEX(ip_set_type_mutex); /* protects ip_set_type_list */ +static DEFINE_RWLOCK(ip_set_ref_lock); /* protects the set refs */ static struct ip_set **ip_set_list; /* all individual sets */ static ip_set_id_t ip_set_max = CONFIG_IP_SET_MAX; /* max number of sets */ @@ -301,13 +302,18 @@ EXPORT_SYMBOL_GPL(ip_set_get_ipaddr6); static inline void __ip_set_get(ip_set_id_t index) { - atomic_inc(&ip_set_list[index]->ref); + write_lock_bh(&ip_set_ref_lock); + ip_set_list[index]->ref++; + write_unlock_bh(&ip_set_ref_lock); } static inline void __ip_set_put(ip_set_id_t index) { - atomic_dec(&ip_set_list[index]->ref); + write_lock_bh(&ip_set_ref_lock); + BUG_ON(ip_set_list[index]->ref == 0); + ip_set_list[index]->ref--; + write_unlock_bh(&ip_set_ref_lock); } /* @@ -324,7 +330,7 @@ ip_set_test(ip_set_id_t index, const struct sk_buff *skb, struct ip_set *set = ip_set_list[index]; int ret = 0; - BUG_ON(set == NULL || atomic_read(&set->ref) == 0); + BUG_ON(set == NULL); pr_debug("set %s, index %u\n", set->name, index); if (dim < set->type->dimension || @@ -356,7 +362,7 @@ ip_set_add(ip_set_id_t index, const struct sk_buff *skb, struct ip_set *set = ip_set_list[index]; int ret; - BUG_ON(set == NULL || atomic_read(&set->ref) == 0); + BUG_ON(set == NULL); pr_debug("set %s, index %u\n", set->name, index); if (dim < set->type->dimension || @@ -378,7 +384,7 @@ ip_set_del(ip_set_id_t index, const struct sk_buff *skb, struct ip_set *set = ip_set_list[index]; int ret = 0; - BUG_ON(set == NULL || atomic_read(&set->ref) == 0); + BUG_ON(set == NULL); pr_debug("set %s, index %u\n", set->name, index); if (dim < set->type->dimension || @@ -397,7 +403,6 @@ EXPORT_SYMBOL_GPL(ip_set_del); * Find set by name, reference it once. The reference makes sure the * thing pointed to, does not go away under our feet. * - * The nfnl mutex must already be activated. */ ip_set_id_t ip_set_get_byname(const char *name, struct ip_set **set) @@ -423,15 +428,12 @@ EXPORT_SYMBOL_GPL(ip_set_get_byname); * reference count by 1. The caller shall not assume the index * to be valid, after calling this function. * - * The nfnl mutex must already be activated. */ void ip_set_put_byindex(ip_set_id_t index) { - if (ip_set_list[index] != NULL) { - BUG_ON(atomic_read(&ip_set_list[index]->ref) == 0); + if (ip_set_list[index] != NULL) __ip_set_put(index); - } } EXPORT_SYMBOL_GPL(ip_set_put_byindex); @@ -441,7 +443,6 @@ EXPORT_SYMBOL_GPL(ip_set_put_byindex); * can't be destroyed. The set cannot be renamed due to * the referencing either. * - * The nfnl mutex must already be activated. */ const char * ip_set_name_byindex(ip_set_id_t index) @@ -449,7 +450,7 @@ ip_set_name_byindex(ip_set_id_t index) const struct ip_set *set = ip_set_list[index]; BUG_ON(set == NULL); - BUG_ON(atomic_read(&set->ref) == 0); + BUG_ON(set->ref == 0); /* Referenced, so it's safe */ return set->name; @@ -515,10 +516,7 @@ void ip_set_nfnl_put(ip_set_id_t index) { nfnl_lock(); - if (ip_set_list[index] != NULL) { - BUG_ON(atomic_read(&ip_set_list[index]->ref) == 0); - __ip_set_put(index); - } + ip_set_put_byindex(index); nfnl_unlock(); } EXPORT_SYMBOL_GPL(ip_set_nfnl_put); @@ -526,7 +524,7 @@ EXPORT_SYMBOL_GPL(ip_set_nfnl_put); /* * Communication protocol with userspace over netlink. * - * We already locked by nfnl_lock. + * The commands are serialized by the nfnl mutex. */ static inline bool @@ -657,7 +655,6 @@ ip_set_create(struct sock *ctnl, struct sk_buff *skb, return -ENOMEM; rwlock_init(&set->lock); strlcpy(set->name, name, IPSET_MAXNAMELEN); - atomic_set(&set->ref, 0); set->family = family; /* @@ -690,8 +687,8 @@ ip_set_create(struct sock *ctnl, struct sk_buff *skb, /* * Here, we have a valid, constructed set and we are protected - * by nfnl_lock. Find the first free index in ip_set_list and - * check clashing. + * by the nfnl mutex. Find the first free index in ip_set_list + * and check clashing. */ if ((ret = find_free_id(set->name, &index, &clash)) != 0) { /* If this is the same set and requested, ignore error */ @@ -751,31 +748,51 @@ ip_set_destroy(struct sock *ctnl, struct sk_buff *skb, const struct nlattr * const attr[]) { ip_set_id_t i; + int ret = 0; if (unlikely(protocol_failed(attr))) return -IPSET_ERR_PROTOCOL; - /* References are protected by the nfnl mutex */ + /* Commands are serialized and references are + * protected by the ip_set_ref_lock. + * External systems (i.e. xt_set) must call + * ip_set_put|get_nfnl_* functions, that way we + * can safely check references here. + * + * list:set timer can only decrement the reference + * counter, so if it's already zero, we can proceed + * without holding the lock. + */ + read_lock_bh(&ip_set_ref_lock); if (!attr[IPSET_ATTR_SETNAME]) { for (i = 0; i < ip_set_max; i++) { - if (ip_set_list[i] != NULL && - (atomic_read(&ip_set_list[i]->ref))) - return -IPSET_ERR_BUSY; + if (ip_set_list[i] != NULL && ip_set_list[i]->ref) { + ret = IPSET_ERR_BUSY; + goto out; + } } + read_unlock_bh(&ip_set_ref_lock); for (i = 0; i < ip_set_max; i++) { if (ip_set_list[i] != NULL) ip_set_destroy_set(i); } } else { i = find_set_id(nla_data(attr[IPSET_ATTR_SETNAME])); - if (i == IPSET_INVALID_ID) - return -ENOENT; - else if (atomic_read(&ip_set_list[i]->ref)) - return -IPSET_ERR_BUSY; + if (i == IPSET_INVALID_ID) { + ret = -ENOENT; + goto out; + } else if (ip_set_list[i]->ref) { + ret = -IPSET_ERR_BUSY; + goto out; + } + read_unlock_bh(&ip_set_ref_lock); ip_set_destroy_set(i); } return 0; +out: + read_unlock_bh(&ip_set_ref_lock); + return ret; } /* Flush sets */ @@ -834,6 +851,7 @@ ip_set_rename(struct sock *ctnl, struct sk_buff *skb, struct ip_set *set; const char *name2; ip_set_id_t i; + int ret = 0; if (unlikely(protocol_failed(attr) || attr[IPSET_ATTR_SETNAME] == NULL || @@ -843,25 +861,33 @@ ip_set_rename(struct sock *ctnl, struct sk_buff *skb, set = find_set(nla_data(attr[IPSET_ATTR_SETNAME])); if (set == NULL) return -ENOENT; - if (atomic_read(&set->ref) != 0) - return -IPSET_ERR_REFERENCED; + + read_lock_bh(&ip_set_ref_lock); + if (set->ref != 0) { + ret = -IPSET_ERR_REFERENCED; + goto out; + } name2 = nla_data(attr[IPSET_ATTR_SETNAME2]); for (i = 0; i < ip_set_max; i++) { if (ip_set_list[i] != NULL && - STREQ(ip_set_list[i]->name, name2)) - return -IPSET_ERR_EXIST_SETNAME2; + STREQ(ip_set_list[i]->name, name2)) { + ret = -IPSET_ERR_EXIST_SETNAME2; + goto out; + } } strncpy(set->name, name2, IPSET_MAXNAMELEN); - return 0; +out: + read_unlock_bh(&ip_set_ref_lock); + return ret; } /* Swap two sets so that name/index points to the other. * References and set names are also swapped. * - * We are protected by the nfnl mutex and references are - * manipulated only by holding the mutex. The kernel interfaces + * The commands are serialized by the nfnl mutex and references are + * protected by the ip_set_ref_lock. The kernel interfaces * do not hold the mutex but the pointer settings are atomic * so the ip_set_list always contains valid pointers to the sets. */ @@ -874,7 +900,6 @@ ip_set_swap(struct sock *ctnl, struct sk_buff *skb, struct ip_set *from, *to; ip_set_id_t from_id, to_id; char from_name[IPSET_MAXNAMELEN]; - u32 from_ref; if (unlikely(protocol_failed(attr) || attr[IPSET_ATTR_SETNAME] == NULL || @@ -899,17 +924,15 @@ ip_set_swap(struct sock *ctnl, struct sk_buff *skb, from->type->family == to->type->family)) return -IPSET_ERR_TYPE_MISMATCH; - /* No magic here: ref munging protected by the nfnl_lock */ strncpy(from_name, from->name, IPSET_MAXNAMELEN); - from_ref = atomic_read(&from->ref); - strncpy(from->name, to->name, IPSET_MAXNAMELEN); - atomic_set(&from->ref, atomic_read(&to->ref)); strncpy(to->name, from_name, IPSET_MAXNAMELEN); - atomic_set(&to->ref, from_ref); + write_lock_bh(&ip_set_ref_lock); + swap(from->ref, to->ref); ip_set_list[from_id] = to; ip_set_list[to_id] = from; + write_unlock_bh(&ip_set_ref_lock); return 0; } @@ -926,7 +949,7 @@ ip_set_dump_done(struct netlink_callback *cb) { if (cb->args[2]) { pr_debug("release set %s\n", ip_set_list[cb->args[1]]->name); - __ip_set_put((ip_set_id_t) cb->args[1]); + ip_set_put_byindex((ip_set_id_t) cb->args[1]); } return 0; } @@ -1068,7 +1091,7 @@ release_refcount: /* If there was an error or set is done, release set */ if (ret || !cb->args[2]) { pr_debug("release set %s\n", ip_set_list[index]->name); - __ip_set_put(index); + ip_set_put_byindex(index); } /* If we dump all sets, continue with dumping last ones */ diff --git a/net/netfilter/ipset/ip_set_list_set.c b/net/netfilter/ipset/ip_set_list_set.c index f4a46c0..e9159e9 100644 --- a/net/netfilter/ipset/ip_set_list_set.c +++ b/net/netfilter/ipset/ip_set_list_set.c @@ -366,8 +366,7 @@ list_set_head(struct ip_set *set, struct sk_buff *skb) NLA_PUT_NET32(skb, IPSET_ATTR_SIZE, htonl(map->size)); if (with_timeout(map->timeout)) NLA_PUT_NET32(skb, IPSET_ATTR_TIMEOUT, htonl(map->timeout)); - NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, - htonl(atomic_read(&set->ref) - 1)); + NLA_PUT_NET32(skb, IPSET_ATTR_REFERENCES, htonl(set->ref - 1)); NLA_PUT_NET32(skb, IPSET_ATTR_MEMSIZE, htonl(sizeof(*map) + map->size * map->dsize)); ipset_nest_end(skb, nested); @@ -457,8 +456,7 @@ list_set_gc(unsigned long ul_set) struct list_set *map = set->data; struct set_telem *e; u32 i; - - /* nfnl_lock should be called */ + write_lock_bh(&set->lock); for (i = 0; i < map->size; i++) { e = list_set_telem(map, i); -- cgit v0.10.2 From b4232a22776aa5d063f890d21ca69870dbbe431b Mon Sep 17 00:00:00 2001 From: David Sterba Date: Mon, 4 Apr 2011 15:21:02 +0200 Subject: netfilter: h323: bug in parsing of ASN1 SEQOF field Static analyzer of clang found a dead store which appears to be a bug in reading count of items in SEQOF field, only the lower byte of word is stored. This may lead to corrupted read and communication shutdown. The bug has been in the module since it's first inclusion into linux kernel. [Patrick: the bug is real, but without practical consequence since the largest amount of sequence-of members we parse is 30.] Signed-off-by: David Sterba Signed-off-by: Patrick McHardy diff --git a/net/netfilter/nf_conntrack_h323_asn1.c b/net/netfilter/nf_conntrack_h323_asn1.c index 8678823..bcd5ed6 100644 --- a/net/netfilter/nf_conntrack_h323_asn1.c +++ b/net/netfilter/nf_conntrack_h323_asn1.c @@ -631,7 +631,7 @@ static int decode_seqof(bitstr_t *bs, const struct field_t *f, CHECK_BOUND(bs, 2); count = *bs->cur++; count <<= 8; - count = *bs->cur++; + count += *bs->cur++; break; case SEMI: BYTE_ALIGN(bs); -- cgit v0.10.2 From a09d19779f3ffac6e16821accc2c1cc4df1b643a Mon Sep 17 00:00:00 2001 From: Hans Schillstrom Date: Mon, 4 Apr 2011 15:25:18 +0200 Subject: IPVS: fix NULL ptr dereference in ip_vs_ctl.c ip_vs_genl_dump_daemons() ipvsadm -ln --daemon will trigger a Null pointer exception because ip_vs_genl_dump_daemons() uses skb_net() instead of skb_sknet(). To prevent others from NULL ptr a check is made in ip_vs.h skb_net(). Signed-off-by: Hans Schillstrom Signed-off-by: Simon Horman Signed-off-by: Patrick McHardy diff --git a/include/net/ip_vs.h b/include/net/ip_vs.h index 30b49ed..4d1b71a 100644 --- a/include/net/ip_vs.h +++ b/include/net/ip_vs.h @@ -52,7 +52,7 @@ static inline struct net *skb_net(const struct sk_buff *skb) */ if (likely(skb->dev && skb->dev->nd_net)) return dev_net(skb->dev); - if (skb_dst(skb)->dev) + if (skb_dst(skb) && skb_dst(skb)->dev) return dev_net(skb_dst(skb)->dev); WARN(skb->sk, "Maybe skb_sknet should be used in %s() at line:%d\n", __func__, __LINE__); diff --git a/net/netfilter/ipvs/ip_vs_ctl.c b/net/netfilter/ipvs/ip_vs_ctl.c index 33733c8..ae47090 100644 --- a/net/netfilter/ipvs/ip_vs_ctl.c +++ b/net/netfilter/ipvs/ip_vs_ctl.c @@ -3120,7 +3120,7 @@ nla_put_failure: static int ip_vs_genl_dump_daemons(struct sk_buff *skb, struct netlink_callback *cb) { - struct net *net = skb_net(skb); + struct net *net = skb_sknet(skb); struct netns_ipvs *ipvs = net_ipvs(net); mutex_lock(&__ip_vs_mutex); -- cgit v0.10.2 From 31ad3dd64e689bc79dd819f8f134b9b025240eb8 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Mon, 4 Apr 2011 16:56:29 +0200 Subject: netfilter: af_info: add network namespace parameter to route hook This is required to eventually replace the rt6_lookup call in xt_addrtype.c with nf_afinfo->route(). Signed-off-by: Florian Westphal Acked-by: David S. Miller Signed-off-by: Patrick McHardy diff --git a/include/linux/netfilter.h b/include/linux/netfilter.h index eeec00a..20ed452 100644 --- a/include/linux/netfilter.h +++ b/include/linux/netfilter.h @@ -270,7 +270,8 @@ struct nf_afinfo { unsigned int dataoff, unsigned int len, u_int8_t protocol); - int (*route)(struct dst_entry **dst, struct flowi *fl); + int (*route)(struct net *net, struct dst_entry **dst, + struct flowi *fl); void (*saveroute)(const struct sk_buff *skb, struct nf_queue_entry *entry); int (*reroute)(struct sk_buff *skb, diff --git a/net/ipv4/netfilter.c b/net/ipv4/netfilter.c index f3c0b54..f1035f0 100644 --- a/net/ipv4/netfilter.c +++ b/net/ipv4/netfilter.c @@ -221,9 +221,10 @@ static __sum16 nf_ip_checksum_partial(struct sk_buff *skb, unsigned int hook, return csum; } -static int nf_ip_route(struct dst_entry **dst, struct flowi *fl) +static int nf_ip_route(struct net *net, struct dst_entry **dst, + struct flowi *fl) { - struct rtable *rt = ip_route_output_key(&init_net, &fl->u.ip4); + struct rtable *rt = ip_route_output_key(net, &fl->u.ip4); if (IS_ERR(rt)) return PTR_ERR(rt); *dst = &rt->dst; diff --git a/net/ipv6/netfilter.c b/net/ipv6/netfilter.c index 39aaca2..e008b9b 100644 --- a/net/ipv6/netfilter.c +++ b/net/ipv6/netfilter.c @@ -90,9 +90,10 @@ static int nf_ip6_reroute(struct sk_buff *skb, return 0; } -static int nf_ip6_route(struct dst_entry **dst, struct flowi *fl) +static int nf_ip6_route(struct net *net, struct dst_entry **dst, + struct flowi *fl) { - *dst = ip6_route_output(&init_net, NULL, &fl->u.ip6); + *dst = ip6_route_output(net, NULL, &fl->u.ip6); return (*dst)->error; } diff --git a/net/netfilter/nf_conntrack_h323_main.c b/net/netfilter/nf_conntrack_h323_main.c index 533a183..39a4538 100644 --- a/net/netfilter/nf_conntrack_h323_main.c +++ b/net/netfilter/nf_conntrack_h323_main.c @@ -731,9 +731,9 @@ static int callforward_do_filter(const union nf_inet_addr *src, memset(&fl2, 0, sizeof(fl2)); fl2.daddr = dst->ip; - if (!afinfo->route((struct dst_entry **)&rt1, + if (!afinfo->route(&init_net, (struct dst_entry **)&rt1, flowi4_to_flowi(&fl1))) { - if (!afinfo->route((struct dst_entry **)&rt2, + if (!afinfo->route(&init_net, (struct dst_entry **)&rt2, flowi4_to_flowi(&fl2))) { if (rt1->rt_gateway == rt2->rt_gateway && rt1->dst.dev == rt2->dst.dev) @@ -755,9 +755,9 @@ static int callforward_do_filter(const union nf_inet_addr *src, memset(&fl2, 0, sizeof(fl2)); ipv6_addr_copy(&fl2.daddr, &dst->in6); - if (!afinfo->route((struct dst_entry **)&rt1, + if (!afinfo->route(&init_net, (struct dst_entry **)&rt1, flowi6_to_flowi(&fl1))) { - if (!afinfo->route((struct dst_entry **)&rt2, + if (!afinfo->route(&init_net, (struct dst_entry **)&rt2, flowi6_to_flowi(&fl2))) { if (!memcmp(&rt1->rt6i_gateway, &rt2->rt6i_gateway, sizeof(rt1->rt6i_gateway)) && diff --git a/net/netfilter/xt_TCPMSS.c b/net/netfilter/xt_TCPMSS.c index 6e6b46c..8690125 100644 --- a/net/netfilter/xt_TCPMSS.c +++ b/net/netfilter/xt_TCPMSS.c @@ -166,7 +166,7 @@ static u_int32_t tcpmss_reverse_mtu(const struct sk_buff *skb, rcu_read_lock(); ai = nf_get_afinfo(family); if (ai != NULL) - ai->route((struct dst_entry **)&rt, &fl); + ai->route(&init_net, (struct dst_entry **)&rt, &fl); rcu_read_unlock(); if (rt != NULL) { -- cgit v0.10.2 From 0fae2e7740aca7e384c5f337f458897e7e337d58 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Mon, 4 Apr 2011 17:00:54 +0200 Subject: netfilter: af_info: add 'strict' parameter to limit lookup to .oif ipv6 fib lookup can set RT6_LOOKUP_F_IFACE flag to restrict search to an interface, but this flag cannot be set via struct flowi. Also, it cannot be set via ip6_route_output: this function uses the passed sock struct to determine if this flag is required (by testing for nonzero sk_bound_dev_if). Work around this by passing in an artificial struct sk in case 'strict' argument is true. This is required to replace the rt6_lookup call in xt_addrtype.c with nf_afinfo->route(). Signed-off-by: Florian Westphal Acked-by: David S. Miller Signed-off-by: Patrick McHardy diff --git a/include/linux/netfilter.h b/include/linux/netfilter.h index 20ed452..7fa95df 100644 --- a/include/linux/netfilter.h +++ b/include/linux/netfilter.h @@ -271,7 +271,7 @@ struct nf_afinfo { unsigned int len, u_int8_t protocol); int (*route)(struct net *net, struct dst_entry **dst, - struct flowi *fl); + struct flowi *fl, bool strict); void (*saveroute)(const struct sk_buff *skb, struct nf_queue_entry *entry); int (*reroute)(struct sk_buff *skb, diff --git a/net/ipv4/netfilter.c b/net/ipv4/netfilter.c index f1035f0..4614bab 100644 --- a/net/ipv4/netfilter.c +++ b/net/ipv4/netfilter.c @@ -222,7 +222,7 @@ static __sum16 nf_ip_checksum_partial(struct sk_buff *skb, unsigned int hook, } static int nf_ip_route(struct net *net, struct dst_entry **dst, - struct flowi *fl) + struct flowi *fl, bool strict __always_unused) { struct rtable *rt = ip_route_output_key(net, &fl->u.ip4); if (IS_ERR(rt)) diff --git a/net/ipv6/netfilter.c b/net/ipv6/netfilter.c index e008b9b..28bc1f6 100644 --- a/net/ipv6/netfilter.c +++ b/net/ipv6/netfilter.c @@ -91,9 +91,17 @@ static int nf_ip6_reroute(struct sk_buff *skb, } static int nf_ip6_route(struct net *net, struct dst_entry **dst, - struct flowi *fl) + struct flowi *fl, bool strict) { - *dst = ip6_route_output(net, NULL, &fl->u.ip6); + static const struct ipv6_pinfo fake_pinfo; + static const struct inet_sock fake_sk = { + /* makes ip6_route_output set RT6_LOOKUP_F_IFACE: */ + .sk.sk_bound_dev_if = 1, + .pinet6 = (struct ipv6_pinfo *) &fake_pinfo, + }; + const void *sk = strict ? &fake_sk : NULL; + + *dst = ip6_route_output(net, sk, &fl->u.ip6); return (*dst)->error; } diff --git a/net/netfilter/nf_conntrack_h323_main.c b/net/netfilter/nf_conntrack_h323_main.c index 39a4538..18b2ce5 100644 --- a/net/netfilter/nf_conntrack_h323_main.c +++ b/net/netfilter/nf_conntrack_h323_main.c @@ -732,9 +732,9 @@ static int callforward_do_filter(const union nf_inet_addr *src, memset(&fl2, 0, sizeof(fl2)); fl2.daddr = dst->ip; if (!afinfo->route(&init_net, (struct dst_entry **)&rt1, - flowi4_to_flowi(&fl1))) { + flowi4_to_flowi(&fl1), false)) { if (!afinfo->route(&init_net, (struct dst_entry **)&rt2, - flowi4_to_flowi(&fl2))) { + flowi4_to_flowi(&fl2), false)) { if (rt1->rt_gateway == rt2->rt_gateway && rt1->dst.dev == rt2->dst.dev) ret = 1; @@ -756,9 +756,9 @@ static int callforward_do_filter(const union nf_inet_addr *src, memset(&fl2, 0, sizeof(fl2)); ipv6_addr_copy(&fl2.daddr, &dst->in6); if (!afinfo->route(&init_net, (struct dst_entry **)&rt1, - flowi6_to_flowi(&fl1))) { + flowi6_to_flowi(&fl1), false)) { if (!afinfo->route(&init_net, (struct dst_entry **)&rt2, - flowi6_to_flowi(&fl2))) { + flowi6_to_flowi(&fl2), false)) { if (!memcmp(&rt1->rt6i_gateway, &rt2->rt6i_gateway, sizeof(rt1->rt6i_gateway)) && rt1->dst.dev == rt2->dst.dev) diff --git a/net/netfilter/xt_TCPMSS.c b/net/netfilter/xt_TCPMSS.c index 8690125..9e63b43 100644 --- a/net/netfilter/xt_TCPMSS.c +++ b/net/netfilter/xt_TCPMSS.c @@ -166,7 +166,7 @@ static u_int32_t tcpmss_reverse_mtu(const struct sk_buff *skb, rcu_read_lock(); ai = nf_get_afinfo(family); if (ai != NULL) - ai->route(&init_net, (struct dst_entry **)&rt, &fl); + ai->route(&init_net, (struct dst_entry **)&rt, &fl, false); rcu_read_unlock(); if (rt != NULL) { -- cgit v0.10.2 From b7225041e93f81e7e38fcdf27fc82044e7695efd Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Mon, 4 Apr 2011 17:01:43 +0200 Subject: netfilter: xt_addrtype: replace rt6_lookup with nf_afinfo->route This avoids pulling in the ipv6 module when using (ipv4-only) iptables -m addrtype. Signed-off-by: Florian Westphal Acked-by: David S. Miller Signed-off-by: Patrick McHardy diff --git a/net/netfilter/Kconfig b/net/netfilter/Kconfig index c3f988a..32bff6d 100644 --- a/net/netfilter/Kconfig +++ b/net/netfilter/Kconfig @@ -652,7 +652,6 @@ comment "Xtables matches" config NETFILTER_XT_MATCH_ADDRTYPE tristate '"addrtype" address type match support' depends on NETFILTER_ADVANCED - depends on (IPV6 || IPV6=n) ---help--- This option allows you to match what routing thinks of an address, eg. UNICAST, LOCAL, BROADCAST, ... diff --git a/net/netfilter/xt_addrtype.c b/net/netfilter/xt_addrtype.c index 2220b85..b77d383 100644 --- a/net/netfilter/xt_addrtype.c +++ b/net/netfilter/xt_addrtype.c @@ -32,11 +32,32 @@ MODULE_ALIAS("ipt_addrtype"); MODULE_ALIAS("ip6t_addrtype"); #if defined(CONFIG_IP6_NF_IPTABLES) || defined(CONFIG_IP6_NF_IPTABLES_MODULE) -static u32 xt_addrtype_rt6_to_type(const struct rt6_info *rt) +static u32 match_lookup_rt6(struct net *net, const struct net_device *dev, + const struct in6_addr *addr) { + const struct nf_afinfo *afinfo; + struct flowi6 flow; + struct rt6_info *rt; u32 ret; + int route_err; - if (!rt) + memset(&flow, 0, sizeof(flow)); + ipv6_addr_copy(&flow.daddr, addr); + if (dev) + flow.flowi6_oif = dev->ifindex; + + rcu_read_lock(); + + afinfo = nf_get_afinfo(NFPROTO_IPV6); + if (afinfo != NULL) + route_err = afinfo->route(net, (struct dst_entry **)&rt, + flowi6_to_flowi(&flow), !!dev); + else + route_err = 1; + + rcu_read_unlock(); + + if (route_err) return XT_ADDRTYPE_UNREACHABLE; if (rt->rt6i_flags & RTF_REJECT) @@ -48,6 +69,9 @@ static u32 xt_addrtype_rt6_to_type(const struct rt6_info *rt) ret |= XT_ADDRTYPE_LOCAL; if (rt->rt6i_flags & RTF_ANYCAST) ret |= XT_ADDRTYPE_ANYCAST; + + + dst_release(&rt->dst); return ret; } @@ -65,18 +89,8 @@ static bool match_type6(struct net *net, const struct net_device *dev, return false; if ((XT_ADDRTYPE_LOCAL | XT_ADDRTYPE_ANYCAST | - XT_ADDRTYPE_UNREACHABLE) & mask) { - struct rt6_info *rt; - u32 type; - int ifindex = dev ? dev->ifindex : 0; - - rt = rt6_lookup(net, addr, NULL, ifindex, !!dev); - - type = xt_addrtype_rt6_to_type(rt); - - dst_release(&rt->dst); - return !!(mask & type); - } + XT_ADDRTYPE_UNREACHABLE) & mask) + return !!(mask & match_lookup_rt6(net, dev, addr)); return true; } -- cgit v0.10.2 From 96120d86fe302c006259baee9061eea9e1b9e486 Mon Sep 17 00:00:00 2001 From: Florian Westphal Date: Mon, 4 Apr 2011 17:06:21 +0200 Subject: netfilter: xt_conntrack: fix inverted conntrack direction test --ctdir ORIGINAL matches REPLY packets, and vv: userspace sets "invert_flags &= ~XT_CONNTRACK_DIRECTION" in ORIGINAL case. Thus: (CTINFO2DIR(ctinfo) == IP_CT_DIR_ORIGINAL) ^ !!(info->invert_flags & XT_CONNTRACK_DIRECTION)) yields "1 ^ 0", which is true -> returns false. Reproducer: iptables -I OUTPUT 1 -p tcp --syn -m conntrack --ctdir ORIGINAL Signed-off-by: Florian Westphal Signed-off-by: Patrick McHardy diff --git a/net/netfilter/xt_conntrack.c b/net/netfilter/xt_conntrack.c index 2c0086a..481a86f 100644 --- a/net/netfilter/xt_conntrack.c +++ b/net/netfilter/xt_conntrack.c @@ -195,7 +195,7 @@ conntrack_mt(const struct sk_buff *skb, struct xt_action_param *par, return info->match_flags & XT_CONNTRACK_STATE; if ((info->match_flags & XT_CONNTRACK_DIRECTION) && (CTINFO2DIR(ctinfo) == IP_CT_DIR_ORIGINAL) ^ - !!(info->invert_flags & XT_CONNTRACK_DIRECTION)) + !(info->invert_flags & XT_CONNTRACK_DIRECTION)) return false; if (info->match_flags & XT_CONNTRACK_ORIGSRC) -- cgit v0.10.2 From fef34861381eefe266c822fc28bdedc897a8265f Mon Sep 17 00:00:00 2001 From: Mattia Dongili Date: Sat, 2 Apr 2011 19:00:44 +0900 Subject: sony-laptop: fix early NULL pointer dereference The SNC acpi driver could get early notifications before it fully initializes and that could lead to dereferencing the sony_nc_handles structure pointer that is still NULL at that stage. Make sure we return early from the handle lookup function in these cases. Signed-off-by: Mattia Dongili Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index e642f5f..33c3b8c 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -808,6 +808,11 @@ static int sony_nc_handles_cleanup(struct platform_device *pd) static int sony_find_snc_handle(int handle) { int i; + + /* not initialized yet, return early */ + if (!handles) + return -1; + for (i = 0; i < 0x10; i++) { if (handles->cap[i] == handle) { dprintk("found handle 0x%.4x (offset: 0x%.2x)\n", -- cgit v0.10.2 From 65d7ac038e34702feedad91d0bf597b1f0a4174a Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Sat, 2 Apr 2011 21:23:36 +0200 Subject: platform-drivers: x86: pmic: Restore the dropped buslock/unlock When I added the buslock/unlock mechanism to the pmic code in order to get rid of the horrible work queue stuff, stupid me missed to add the new callbacks to the irq_chip. In consequence Andrew removed the unused functions, but I missed that. Add them back and hook them up proper. Signed-off-by: Thomas Gleixner Cc: Matthew Garrett Cc: Andrew Morton Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/intel_pmic_gpio.c b/drivers/platform/x86/intel_pmic_gpio.c index d653104..7bbb538 100644 --- a/drivers/platform/x86/intel_pmic_gpio.c +++ b/drivers/platform/x86/intel_pmic_gpio.c @@ -74,6 +74,19 @@ struct pmic_gpio { u32 trigger_type; }; +static void pmic_program_irqtype(int gpio, int type) +{ + if (type & IRQ_TYPE_EDGE_RISING) + intel_scu_ipc_update_register(GPIO0 + gpio, 0x20, 0x20); + else + intel_scu_ipc_update_register(GPIO0 + gpio, 0x00, 0x20); + + if (type & IRQ_TYPE_EDGE_FALLING) + intel_scu_ipc_update_register(GPIO0 + gpio, 0x10, 0x10); + else + intel_scu_ipc_update_register(GPIO0 + gpio, 0x00, 0x10); +}; + static int pmic_gpio_direction_input(struct gpio_chip *chip, unsigned offset) { if (offset > 8) { @@ -166,16 +179,38 @@ static int pmic_gpio_to_irq(struct gpio_chip *chip, unsigned offset) return pg->irq_base + offset; } +static void pmic_bus_lock(struct irq_data *data) +{ + struct pmic_gpio *pg = irq_data_get_irq_chip_data(data); + + mutex_lock(&pg->buslock); +} + +static void pmic_bus_sync_unlock(struct irq_data *data) +{ + struct pmic_gpio *pg = irq_data_get_irq_chip_data(data); + + if (pg->update_type) { + unsigned int gpio = pg->update_type & ~GPIO_UPDATE_TYPE; + + pmic_program_irqtype(gpio, pg->trigger_type); + pg->update_type = 0; + } + mutex_unlock(&pg->buslock); +} + /* the gpiointr register is read-clear, so just do nothing. */ static void pmic_irq_unmask(struct irq_data *data) { } static void pmic_irq_mask(struct irq_data *data) { } static struct irq_chip pmic_irqchip = { - .name = "PMIC-GPIO", - .irq_mask = pmic_irq_mask, - .irq_unmask = pmic_irq_unmask, - .irq_set_type = pmic_irq_type, + .name = "PMIC-GPIO", + .irq_mask = pmic_irq_mask, + .irq_unmask = pmic_irq_unmask, + .irq_set_type = pmic_irq_type, + .irq_bus_lock = pmic_irq_buslock, + .irq_bus_sync_unlock = pmic_bus_sync_unlock, }; static irqreturn_t pmic_irq_handler(int irq, void *data) -- cgit v0.10.2 From 50f689af019b19f9b9a39be782c21b6f52b1615a Mon Sep 17 00:00:00 2001 From: Zhu Yanhai Date: Mon, 4 Apr 2011 12:58:12 -0400 Subject: jbd2: move bdget out of critical section bdget() should not be called when we hold spinlocks since it might sleep. Reviewed-by: Jan Kara Signed-off-by: Zhu Yanhai Signed-off-by: "Theodore Ts'o" diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 90407b8..33dd3ef 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -2413,10 +2413,12 @@ const char *jbd2_dev_to_name(dev_t device) new_dev = kmalloc(sizeof(struct devname_cache), GFP_KERNEL); if (!new_dev) return "NODEV-ALLOCFAILURE"; /* Something non-NULL */ + bd = bdget(device); spin_lock(&devname_cache_lock); if (devcache[i]) { if (devcache[i]->device == device) { kfree(new_dev); + bdput(bd); ret = devcache[i]->devname; spin_unlock(&devname_cache_lock); return ret; @@ -2425,7 +2427,6 @@ const char *jbd2_dev_to_name(dev_t device) } devcache[i] = new_dev; devcache[i]->device = device; - bd = bdget(device); if (bd) { bdevname(bd, devcache[i]->devname); bdput(bd); -- cgit v0.10.2 From d88885d0923ae27b01dfcec644f94829b1e46bea Mon Sep 17 00:00:00 2001 From: Konrad Rzeszutek Wilk Date: Mon, 4 Apr 2011 14:48:20 -0400 Subject: xen/debug: Don't be so verbose with WARN on 1-1 mapping errors. There are valid situations in which this error is not a warning. Mainly when QEMU maps a guest memory and uses the VM_IO flag to set the MFNs. For right now make the WARN be WARN_ONCE. In the future we will: 1). Remove the VM_IO code handling.. 2). .. which will also remove this debug facility. Signed-off-by: Konrad Rzeszutek Wilk diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index c82df6c..a991b57 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -565,13 +565,13 @@ pte_t xen_make_pte_debug(pteval_t pte) if (io_page && (xen_initial_domain() || addr >= ISA_END_ADDRESS)) { other_addr = pfn_to_mfn(addr >> PAGE_SHIFT) << PAGE_SHIFT; - WARN(addr != other_addr, + WARN_ONCE(addr != other_addr, "0x%lx is using VM_IO, but it is 0x%lx!\n", (unsigned long)addr, (unsigned long)other_addr); } else { pteval_t iomap_set = (_pte.pte & PTE_FLAGS_MASK) & _PAGE_IOMAP; other_addr = (_pte.pte & PTE_PFN_MASK); - WARN((addr == other_addr) && (!io_page) && (!iomap_set), + WARN_ONCE((addr == other_addr) && (!io_page) && (!iomap_set), "0x%lx is missing VM_IO (and wasn't fixed)!\n", (unsigned long)addr); } -- cgit v0.10.2 From c85ce65ecac078ab1a1835c87c4a6319cf74660a Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Wed, 30 Mar 2011 14:02:46 -0400 Subject: b43: allocate receive buffers big enough for max frame len + offset Otherwise, skb_put inside of dma_rx can fail... https://bugzilla.kernel.org/show_bug.cgi?id=32042 Signed-off-by: John W. Linville Acked-by: Larry Finger Cc: stable@kernel.org diff --git a/drivers/net/wireless/b43/dma.c b/drivers/net/wireless/b43/dma.c index 3d5566e..ff0f5ba 100644 --- a/drivers/net/wireless/b43/dma.c +++ b/drivers/net/wireless/b43/dma.c @@ -1536,7 +1536,7 @@ static void dma_rx(struct b43_dmaring *ring, int *slot) dmaaddr = meta->dmaaddr; goto drop_recycle_buffer; } - if (unlikely(len > ring->rx_buffersize)) { + if (unlikely(len + ring->frameoffset > ring->rx_buffersize)) { /* The data did not fit into one descriptor buffer * and is split over multiple buffers. * This should never happen, as we try to allocate buffers diff --git a/drivers/net/wireless/b43/dma.h b/drivers/net/wireless/b43/dma.h index a01c210..e8a80a1 100644 --- a/drivers/net/wireless/b43/dma.h +++ b/drivers/net/wireless/b43/dma.h @@ -163,7 +163,7 @@ struct b43_dmadesc_generic { /* DMA engine tuning knobs */ #define B43_TXRING_SLOTS 256 #define B43_RXRING_SLOTS 64 -#define B43_DMA0_RX_BUFFERSIZE IEEE80211_MAX_FRAME_LEN +#define B43_DMA0_RX_BUFFERSIZE (B43_DMA0_RX_FRAMEOFFSET + IEEE80211_MAX_FRAME_LEN) /* Pointer poison */ #define B43_DMA_PTR_POISON ((void *)ERR_PTR(-ENOMEM)) -- cgit v0.10.2 From 5245e3a9f77a8dd4ea92d7bdab2e96629f1feda4 Mon Sep 17 00:00:00 2001 From: Luciano Coelho Date: Wed, 30 Mar 2011 21:31:39 +0300 Subject: wl12xx: fix module author's email address in the spi and sdio modules The MODULE_AUTHOR() macro in the main module (wl12xx) has been updated to reflect one of the author's new email address, but the wl12xx_spi and wl12xx_sdio modules haven't been updated. This patches updates them. Signed-off-by: Luciano Coelho Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/wl12xx/sdio.c b/drivers/net/wireless/wl12xx/sdio.c index 5b9dbea..b1c7d03 100644 --- a/drivers/net/wireless/wl12xx/sdio.c +++ b/drivers/net/wireless/wl12xx/sdio.c @@ -340,7 +340,7 @@ module_init(wl1271_init); module_exit(wl1271_exit); MODULE_LICENSE("GPL"); -MODULE_AUTHOR("Luciano Coelho "); +MODULE_AUTHOR("Luciano Coelho "); MODULE_AUTHOR("Juuso Oikarinen "); MODULE_FIRMWARE(WL1271_FW_NAME); MODULE_FIRMWARE(WL1271_AP_FW_NAME); diff --git a/drivers/net/wireless/wl12xx/spi.c b/drivers/net/wireless/wl12xx/spi.c index 18cf017..ffc745b 100644 --- a/drivers/net/wireless/wl12xx/spi.c +++ b/drivers/net/wireless/wl12xx/spi.c @@ -487,7 +487,7 @@ module_init(wl1271_init); module_exit(wl1271_exit); MODULE_LICENSE("GPL"); -MODULE_AUTHOR("Luciano Coelho "); +MODULE_AUTHOR("Luciano Coelho "); MODULE_AUTHOR("Juuso Oikarinen "); MODULE_FIRMWARE(WL1271_FW_NAME); MODULE_FIRMWARE(WL1271_AP_FW_NAME); -- cgit v0.10.2 From 8f06ca2c83689f4d352a34ef1c484f40c25b41ed Mon Sep 17 00:00:00 2001 From: Senthil Balasubramanian Date: Fri, 1 Apr 2011 17:16:33 +0530 Subject: ath9k: Fix phy info print message with AR9485 chipset. The phy information print during driver init time doesn't show the numeric part of the chip name properly for AR9485. This patch addresses this issue by adding the string to the respective array. Signed-off-by: Senthil Balasubramanian Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/ath/ath9k/hw.c b/drivers/net/wireless/ath/ath9k/hw.c index 338b075..1ec9bcd 100644 --- a/drivers/net/wireless/ath/ath9k/hw.c +++ b/drivers/net/wireless/ath/ath9k/hw.c @@ -2546,6 +2546,7 @@ static struct { { AR_SREV_VERSION_9287, "9287" }, { AR_SREV_VERSION_9271, "9271" }, { AR_SREV_VERSION_9300, "9300" }, + { AR_SREV_VERSION_9485, "9485" }, }; /* For devices with external radios */ -- cgit v0.10.2 From 5312c3f60ba49073081b2dc421f9f3c22dd43d99 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Fri, 1 Apr 2011 13:52:34 +0200 Subject: mac80211: fix comment regarding aggregation buf_size The description for buf_size was misleading and just said you couldn't TX larger aggregates, but of course you can't TX aggregates in a way that would exceed the window either, which is possible even if the aggregates are shorter than that. Expand the description, thanks to Emmanuel for explaining this to me. Cc: Emmanuel Grumbach Signed-off-by: Johannes Berg Signed-off-by: John W. Linville diff --git a/include/net/mac80211.h b/include/net/mac80211.h index cefe1b3..965f1b1 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -1753,8 +1753,19 @@ enum ieee80211_ampdu_mlme_action { * that TX/RX_STOP can pass NULL for this parameter. * The @buf_size parameter is only valid when the action is set to * %IEEE80211_AMPDU_TX_OPERATIONAL and indicates the peer's reorder - * buffer size (number of subframes) for this session -- aggregates - * containing more subframes than this may not be transmitted to the peer. + * buffer size (number of subframes) for this session -- the driver + * may neither send aggregates containing more subframes than this + * nor send aggregates in a way that lost frames would exceed the + * buffer size. If just limiting the aggregate size, this would be + * possible with a buf_size of 8: + * - TX: 1.....7 + * - RX: 2....7 (lost frame #1) + * - TX: 8..1... + * which is invalid since #1 was now re-transmitted well past the + * buffer size of 8. Correct ways to retransmit #1 would be: + * - TX: 1 or 18 or 81 + * Even "189" would be wrong since 1 could be lost again. + * * Returns a negative error code on failure. * The callback can sleep. * -- cgit v0.10.2 From fcf8bd3ba5362682f945a3f838070ac5e10ff871 Mon Sep 17 00:00:00 2001 From: Helmut Schaa Date: Fri, 1 Apr 2011 15:46:05 +0200 Subject: mac80211: Fix duplicate frames on cooked monitor Cleaning the ieee80211_rx_data.flags field here is wrong, instead the flags should be valid accross processing the frame on different interfaces. Fix this by removing the incorrect flags=0 assignment. Introduced in commit 554891e63a29af35cc6bb403ef34e319518114d0 (mac80211: move packet flags into packet). Signed-off-by: Helmut Schaa Signed-off-by: John W. Linville diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index aa5cc37..2afeac9 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -2541,7 +2541,6 @@ static void ieee80211_rx_handlers(struct ieee80211_rx_data *rx) * same TID from the same station */ rx->skb = skb; - rx->flags = 0; CALL_RXH(ieee80211_rx_h_decrypt) CALL_RXH(ieee80211_rx_h_check_more_data) @@ -2612,6 +2611,7 @@ void ieee80211_release_reorder_timeout(struct sta_info *sta, int tid) .sdata = sta->sdata, .local = sta->local, .queue = tid, + .flags = 0, }; struct tid_ampdu_rx *tid_agg_rx; -- cgit v0.10.2 From 2fc713b20469b2779fa89f582d2a9696b10031e7 Mon Sep 17 00:00:00 2001 From: Jussi Kivilinna Date: Fri, 1 Apr 2011 19:34:08 +0300 Subject: zd1211rw: remove URB_SHORT_NOT_OK flag in zd_usb_iowrite16v_async() Patch removes the bogus flag introduced by upstream commit eefdbec1ea8b7093d2c09d1825f68438701723cf. Old code had buffer length check that new code tried to handle with URB_SHORT_NOT_OK flag. With USB debugging enabled bogus flag caused usb_submit_urb fail. Remove URB_SHORT_NOT_OK flag and add buffer length check to urb completion handler. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=32092 Reported-by: Jonathan Callen Signed-off-by: Jussi Kivilinna Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/zd1211rw/zd_usb.c b/drivers/net/wireless/zd1211rw/zd_usb.c index 58236e6..9b1a26a 100644 --- a/drivers/net/wireless/zd1211rw/zd_usb.c +++ b/drivers/net/wireless/zd1211rw/zd_usb.c @@ -1671,6 +1671,10 @@ static void iowrite16v_urb_complete(struct urb *urb) if (urb->status && !usb->cmd_error) usb->cmd_error = urb->status; + + if (!usb->cmd_error && + urb->actual_length != urb->transfer_buffer_length) + usb->cmd_error = -EIO; } static int zd_submit_waiting_urb(struct zd_usb *usb, bool last) @@ -1805,7 +1809,7 @@ int zd_usb_iowrite16v_async(struct zd_usb *usb, const struct zd_ioreq16 *ioreqs, usb_fill_int_urb(urb, udev, usb_sndintpipe(udev, EP_REGS_OUT), req, req_len, iowrite16v_urb_complete, usb, ep->desc.bInterval); - urb->transfer_flags |= URB_FREE_BUFFER | URB_SHORT_NOT_OK; + urb->transfer_flags |= URB_FREE_BUFFER; /* Submit previous URB */ r = zd_submit_waiting_urb(usb, false); -- cgit v0.10.2 From 023535732f4db01af4921f20f058bc4561d9add7 Mon Sep 17 00:00:00 2001 From: Jussi Kivilinna Date: Fri, 1 Apr 2011 19:34:16 +0300 Subject: zd1211rw: reset rx idle timer from tasklet 2.6.38 added WARN_ON(in_irq) in del_timer_sync that triggers on zd1211rw when reseting rx idle timer in urb completion handler. Move timer reseting to tasklet. Signed-off-by: Jussi Kivilinna Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/zd1211rw/zd_usb.c b/drivers/net/wireless/zd1211rw/zd_usb.c index 9b1a26a..ab607bb 100644 --- a/drivers/net/wireless/zd1211rw/zd_usb.c +++ b/drivers/net/wireless/zd1211rw/zd_usb.c @@ -643,7 +643,7 @@ static void rx_urb_complete(struct urb *urb) usb = urb->context; rx = &usb->rx; - zd_usb_reset_rx_idle_timer(usb); + tasklet_schedule(&rx->reset_timer_tasklet); if (length%rx->usb_packet_size > rx->usb_packet_size-4) { /* If there is an old first fragment, we don't care. */ @@ -812,6 +812,7 @@ void zd_usb_disable_rx(struct zd_usb *usb) __zd_usb_disable_rx(usb); mutex_unlock(&rx->setup_mutex); + tasklet_kill(&rx->reset_timer_tasklet); cancel_delayed_work_sync(&rx->idle_work); } @@ -1106,6 +1107,13 @@ static void zd_rx_idle_timer_handler(struct work_struct *work) zd_usb_reset_rx(usb); } +static void zd_usb_reset_rx_idle_timer_tasklet(unsigned long param) +{ + struct zd_usb *usb = (struct zd_usb *)param; + + zd_usb_reset_rx_idle_timer(usb); +} + void zd_usb_reset_rx_idle_timer(struct zd_usb *usb) { struct zd_usb_rx *rx = &usb->rx; @@ -1127,6 +1135,7 @@ static inline void init_usb_interrupt(struct zd_usb *usb) static inline void init_usb_rx(struct zd_usb *usb) { struct zd_usb_rx *rx = &usb->rx; + spin_lock_init(&rx->lock); mutex_init(&rx->setup_mutex); if (interface_to_usbdev(usb->intf)->speed == USB_SPEED_HIGH) { @@ -1136,11 +1145,14 @@ static inline void init_usb_rx(struct zd_usb *usb) } ZD_ASSERT(rx->fragment_length == 0); INIT_DELAYED_WORK(&rx->idle_work, zd_rx_idle_timer_handler); + rx->reset_timer_tasklet.func = zd_usb_reset_rx_idle_timer_tasklet; + rx->reset_timer_tasklet.data = (unsigned long)usb; } static inline void init_usb_tx(struct zd_usb *usb) { struct zd_usb_tx *tx = &usb->tx; + spin_lock_init(&tx->lock); atomic_set(&tx->enabled, 0); tx->stopped = 0; diff --git a/drivers/net/wireless/zd1211rw/zd_usb.h b/drivers/net/wireless/zd1211rw/zd_usb.h index b3df2c8..325d0f9 100644 --- a/drivers/net/wireless/zd1211rw/zd_usb.h +++ b/drivers/net/wireless/zd1211rw/zd_usb.h @@ -183,6 +183,7 @@ struct zd_usb_rx { spinlock_t lock; struct mutex setup_mutex; struct delayed_work idle_work; + struct tasklet_struct reset_timer_tasklet; u8 fragment[2 * USB_MAX_RX_SIZE]; unsigned int fragment_length; unsigned int usb_packet_size; -- cgit v0.10.2 From 09b661b33268698d3b453dceb78cda129ad899b4 Mon Sep 17 00:00:00 2001 From: Luciano Coelho Date: Fri, 1 Apr 2011 19:42:02 +0300 Subject: wl12xx: fix potential buffer overflow in testmode nvs push We were allocating the size of the NVS file struct and not checking whether the length of the buffer passed was correct before copying it into the allocated memory. This is a security hole because buffer overflows can occur if the userspace passes a bigger file than what is expected. With this patch, we check if the size of the data passed from userspace matches the size required. This bug was introduced in 2.6.36. Cc: stable@kernel.org Reported-by: Ido Yariv Signed-off-by: Luciano Coelho Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/wl12xx/testmode.c b/drivers/net/wireless/wl12xx/testmode.c index e64403b..6ec06a4 100644 --- a/drivers/net/wireless/wl12xx/testmode.c +++ b/drivers/net/wireless/wl12xx/testmode.c @@ -204,7 +204,10 @@ static int wl1271_tm_cmd_nvs_push(struct wl1271 *wl, struct nlattr *tb[]) kfree(wl->nvs); - wl->nvs = kzalloc(sizeof(struct wl1271_nvs_file), GFP_KERNEL); + if (len != sizeof(struct wl1271_nvs_file)) + return -EINVAL; + + wl->nvs = kzalloc(len, GFP_KERNEL); if (!wl->nvs) { wl1271_error("could not allocate memory for the nvs file"); ret = -ENOMEM; -- cgit v0.10.2 From 220107610c7c2c9703e09eb363e8ab31025b9315 Mon Sep 17 00:00:00 2001 From: Christian Lamparter Date: Sat, 2 Apr 2011 11:31:29 +0200 Subject: p54usb: IDs for two new devices Cc: Reported-by: Mark Davis [via p54/devices wiki] Signed-off-by: Christian Lamparter Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/p54/p54usb.c b/drivers/net/wireless/p54/p54usb.c index 9b344a9..e183587 100644 --- a/drivers/net/wireless/p54/p54usb.c +++ b/drivers/net/wireless/p54/p54usb.c @@ -56,6 +56,7 @@ static struct usb_device_id p54u_table[] __devinitdata = { {USB_DEVICE(0x0846, 0x4210)}, /* Netgear WG121 the second ? */ {USB_DEVICE(0x0846, 0x4220)}, /* Netgear WG111 */ {USB_DEVICE(0x09aa, 0x1000)}, /* Spinnaker Proto board */ + {USB_DEVICE(0x0bf8, 0x1007)}, /* Fujitsu E-5400 USB */ {USB_DEVICE(0x0cde, 0x0006)}, /* Medion 40900, Roper Europe */ {USB_DEVICE(0x0db0, 0x6826)}, /* MSI UB54G (MS-6826) */ {USB_DEVICE(0x107b, 0x55f2)}, /* Gateway WGU-210 (Gemtek) */ @@ -68,6 +69,7 @@ static struct usb_device_id p54u_table[] __devinitdata = { {USB_DEVICE(0x1915, 0x2235)}, /* Linksys WUSB54G Portable OEM */ {USB_DEVICE(0x2001, 0x3701)}, /* DLink DWL-G120 Spinnaker */ {USB_DEVICE(0x2001, 0x3703)}, /* DLink DWL-G122 */ + {USB_DEVICE(0x2001, 0x3762)}, /* Conceptronic C54U */ {USB_DEVICE(0x5041, 0x2234)}, /* Linksys WUSB54G */ {USB_DEVICE(0x5041, 0x2235)}, /* Linksys WUSB54G Portable */ -- cgit v0.10.2 From 2e3e66e3bbcd5f999dc66d843a3efc2314e40e7b Mon Sep 17 00:00:00 2001 From: Larry Finger Date: Sat, 2 Apr 2011 18:10:22 -0500 Subject: rtlwifi: Fix some warnings/bugs Some compiler/architecture combinations generate some warnings that are not seen on my main system. Two of the "warnings" about unitialized variables are really bugs. Signed-off-by: Larry Finger Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/rtlwifi/efuse.c b/drivers/net/wireless/rtlwifi/efuse.c index f74a870..590f14f 100644 --- a/drivers/net/wireless/rtlwifi/efuse.c +++ b/drivers/net/wireless/rtlwifi/efuse.c @@ -685,7 +685,7 @@ static int efuse_pg_packet_read(struct ieee80211_hw *hw, u8 offset, u8 *data) u8 efuse_data, word_cnts = 0; u16 efuse_addr = 0; - u8 hworden; + u8 hworden = 0; u8 tmpdata[8]; if (data == NULL) diff --git a/drivers/net/wireless/rtlwifi/rtl8192c/fw_common.c b/drivers/net/wireless/rtlwifi/rtl8192c/fw_common.c index 5ef9137..28a6ce3 100644 --- a/drivers/net/wireless/rtlwifi/rtl8192c/fw_common.c +++ b/drivers/net/wireless/rtlwifi/rtl8192c/fw_common.c @@ -303,7 +303,7 @@ static void _rtl92c_fill_h2c_command(struct ieee80211_hw *hw, u16 box_reg, box_extreg; u8 u1b_tmp; bool isfw_read = false; - u8 buf_index; + u8 buf_index = 0; bool bwrite_sucess = false; u8 wait_h2c_limmit = 100; u8 wait_writeh2c_limmit = 100; diff --git a/drivers/net/wireless/rtlwifi/usb.c b/drivers/net/wireless/rtlwifi/usb.c index a4b2613..f5d8573 100644 --- a/drivers/net/wireless/rtlwifi/usb.c +++ b/drivers/net/wireless/rtlwifi/usb.c @@ -246,7 +246,7 @@ static void _rtl_usb_io_handler_init(struct device *dev, static void _rtl_usb_io_handler_release(struct ieee80211_hw *hw) { - struct rtl_priv *rtlpriv = rtl_priv(hw); + struct rtl_priv __maybe_unused *rtlpriv = rtl_priv(hw); mutex_destroy(&rtlpriv->io.bb_mutex); } -- cgit v0.10.2 From 37f4ee0b6b39640828dac5937a482c20e8ac794f Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Mon, 4 Apr 2011 13:50:32 +0200 Subject: rt2x00: fix cancelling uninitialized work {rx,tx}done_work's are only initialized for usb devices. Signed-off-by: Stanislaw Gruszka Cc: stable@kernel.org Acked-by: Ivo van Doorn Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/rt2x00/rt2x00dev.c b/drivers/net/wireless/rt2x00/rt2x00dev.c index 9de9dbe..84eb6ad 100644 --- a/drivers/net/wireless/rt2x00/rt2x00dev.c +++ b/drivers/net/wireless/rt2x00/rt2x00dev.c @@ -1062,8 +1062,10 @@ void rt2x00lib_remove_dev(struct rt2x00_dev *rt2x00dev) * Stop all work. */ cancel_work_sync(&rt2x00dev->intf_work); - cancel_work_sync(&rt2x00dev->rxdone_work); - cancel_work_sync(&rt2x00dev->txdone_work); + if (rt2x00_is_usb(rt2x00dev)) { + cancel_work_sync(&rt2x00dev->rxdone_work); + cancel_work_sync(&rt2x00dev->txdone_work); + } destroy_workqueue(rt2x00dev->workqueue); /* -- cgit v0.10.2 From 3d7dc7e8c1566acb0fc55df228b2ed91f5638e9d Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Mon, 4 Apr 2011 13:18:44 -0400 Subject: iwlwifi: accept EEPROM version 0x423 for iwl6000 A number of these devices have appeared "in the wild", and apparently the Windows driver is perfectly happy to support this EEPROM version. Signed-off-by: John W. Linville Cc: stable@kernel.org Acked-by: Wey-Yi Guy diff --git a/drivers/net/wireless/iwlwifi/iwl-eeprom.h b/drivers/net/wireless/iwlwifi/iwl-eeprom.h index 98aa8af..20b6646 100644 --- a/drivers/net/wireless/iwlwifi/iwl-eeprom.h +++ b/drivers/net/wireless/iwlwifi/iwl-eeprom.h @@ -241,7 +241,7 @@ struct iwl_eeprom_enhanced_txpwr { /* 6x00 Specific */ #define EEPROM_6000_TX_POWER_VERSION (4) -#define EEPROM_6000_EEPROM_VERSION (0x434) +#define EEPROM_6000_EEPROM_VERSION (0x423) /* 6x50 Specific */ #define EEPROM_6050_TX_POWER_VERSION (4) -- cgit v0.10.2 From 21f976975cbecbdaf23ceeacc1cab2b1c05a028e Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Mon, 4 Apr 2011 15:33:39 -0400 Subject: ext4: remove unnecessary [cm]time update of quota file It is not necessary to update [cm]time of quota file on each quota file write and it wastes journal space and IO throughput with inode writes. So just remove the updating from ext4_quota_write() and only update times when quotas are being turned off. Userspace cannot get anything reliable from quota files while they are used by the kernel anyway. Signed-off-by: Jan Kara Signed-off-by: "Theodore Ts'o" diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h index e25e99b..d0f5353 100644 --- a/fs/ext4/ext4_jbd2.h +++ b/fs/ext4/ext4_jbd2.h @@ -86,8 +86,8 @@ #ifdef CONFIG_QUOTA /* Amount of blocks needed for quota update - we know that the structure was - * allocated so we need to update only inode+data */ -#define EXT4_QUOTA_TRANS_BLOCKS(sb) (test_opt(sb, QUOTA) ? 2 : 0) + * allocated so we need to update only data block */ +#define EXT4_QUOTA_TRANS_BLOCKS(sb) (test_opt(sb, QUOTA) ? 1 : 0) /* Amount of blocks needed for quota insert/delete - we do some block writes * but inode, sb and group updates are done only once */ #define EXT4_QUOTA_INIT_BLOCKS(sb) (test_opt(sb, QUOTA) ? (DQUOT_INIT_ALLOC*\ diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 22546ad..35ff9fe 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -4614,11 +4614,24 @@ static int ext4_quota_on(struct super_block *sb, int type, int format_id, static int ext4_quota_off(struct super_block *sb, int type) { + struct inode *inode = sb_dqopt(sb)->files[type]; + handle_t *handle; + /* Force all delayed allocation blocks to be allocated. * Caller already holds s_umount sem */ if (test_opt(sb, DELALLOC)) sync_filesystem(sb); + /* Update modification times of quota files when userspace can + * start looking at them */ + handle = ext4_journal_start(inode, 1); + if (IS_ERR(handle)) + goto out; + inode->i_mtime = inode->i_ctime = CURRENT_TIME; + ext4_mark_inode_dirty(handle, inode); + ext4_journal_stop(handle); + +out: return dquot_quota_off(sb, type); } @@ -4714,9 +4727,8 @@ out: if (inode->i_size < off + len) { i_size_write(inode, off + len); EXT4_I(inode)->i_disksize = inode->i_size; + ext4_mark_inode_dirty(handle, inode); } - inode->i_mtime = inode->i_ctime = CURRENT_TIME; - ext4_mark_inode_dirty(handle, inode); mutex_unlock(&inode->i_mutex); return len; } -- cgit v0.10.2 From 5b41395fcc0265fc9f193aef9df39ce49d64677c Mon Sep 17 00:00:00 2001 From: Yongqiang Yang Date: Mon, 4 Apr 2011 15:40:24 -0400 Subject: ext4: fix credits computing for indirect mapped files When writing a contiguous set of blocks, two indirect blocks could be needed depending on how the blocks are aligned, so we need to increase the number of credits needed by one. [ Also fixed a another bug which could further underestimate the number of journal credits needed by 1; the code was using integer division instead of DIV_ROUND_UP() -- tytso] Signed-off-by: Yongqiang Yang Signed-off-by: "Theodore Ts'o" Cc: stable@kernel.org diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 1a86282..7d11e02 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -5398,13 +5398,12 @@ static int ext4_indirect_trans_blocks(struct inode *inode, int nrblocks, /* if nrblocks are contiguous */ if (chunk) { /* - * With N contiguous data blocks, it need at most - * N/EXT4_ADDR_PER_BLOCK(inode->i_sb) indirect blocks - * 2 dindirect blocks - * 1 tindirect block + * With N contiguous data blocks, we need at most + * N/EXT4_ADDR_PER_BLOCK(inode->i_sb) + 1 indirect blocks, + * 2 dindirect blocks, and 1 tindirect block */ - indirects = nrblocks / EXT4_ADDR_PER_BLOCK(inode->i_sb); - return indirects + 3; + return DIV_ROUND_UP(nrblocks, + EXT4_ADDR_PER_BLOCK(inode->i_sb)) + 4; } /* * if nrblocks are not contiguous, worse case, each block touch -- cgit v0.10.2 From 46e4690bbd9a4f8d9e7c4f34e34b48f703ad47e0 Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Mon, 4 Apr 2011 16:00:49 -0400 Subject: ext4: fix a double free in ext4_register_li_request In ext4_register_li_request, we malloc a ext4_li_request and inserts it into ext4_li_info->li_request_list. In case of any error later, we free it in the end. But if we have some error in ext4_run_lazyinit_thread, the whole li_request_list will be dropped and freed in it. So we will double free this ext4_li_request. This patch just sets elr to NULL after it is inserted to the list so that the latter kfree won't double free it. Signed-off-by: Tao Ma Reviewed-by: Lukas Czerner Signed-off-by: "Theodore Ts'o" Cc: stable@kernel.org diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 35ff9fe..35bd020 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -2975,6 +2975,12 @@ static int ext4_register_li_request(struct super_block *sb, mutex_unlock(&ext4_li_info->li_list_mtx); sbi->s_li_request = elr; + /* + * set elr to NULL here since it has been inserted to + * the request_list and the removal and free of it is + * handled by ext4_clear_request_list from now on. + */ + elr = NULL; if (!(ext4_li_info->li_state & EXT4_LAZYINIT_RUNNING)) { ret = ext4_run_lazyinit_thread(); -- cgit v0.10.2 From ac5c15fa585a89bcebca92dfdbb862dd507534fc Mon Sep 17 00:00:00 2001 From: Roy Spliet Date: Wed, 9 Feb 2011 14:56:42 +0100 Subject: drm/nouveau: correct memtiming table parsing for nv4x In line with envytools, verified on 4 or 5 BIOS'es. Signed-off-by: Ben Skeggs Signed-off-by: Roy Spliet diff --git a/drivers/gpu/drm/nouveau/nouveau_mem.c b/drivers/gpu/drm/nouveau/nouveau_mem.c index 2683377..78f467f 100644 --- a/drivers/gpu/drm/nouveau/nouveau_mem.c +++ b/drivers/gpu/drm/nouveau/nouveau_mem.c @@ -552,6 +552,7 @@ nouveau_mem_timing_init(struct drm_device *dev) u8 tRC; /* Byte 9 */ u8 tUNK_10, tUNK_11, tUNK_12, tUNK_13, tUNK_14; u8 tUNK_18, tUNK_19, tUNK_20, tUNK_21; + u8 magic_number = 0; /* Yeah... sorry*/ u8 *mem = NULL, *entry; int i, recordlen, entries; @@ -596,6 +597,12 @@ nouveau_mem_timing_init(struct drm_device *dev) if (!memtimings->timing) return; + /* Get "some number" from the timing reg for NV_40 + * Used in calculations later */ + if(dev_priv->card_type == NV_40) { + magic_number = (nv_rd32(dev,0x100228) & 0x0f000000) >> 24; + } + entry = mem + mem[1]; for (i = 0; i < entries; i++, entry += recordlen) { struct nouveau_pm_memtiming *timing = &pm->memtimings.timing[i]; @@ -635,36 +642,51 @@ nouveau_mem_timing_init(struct drm_device *dev) /* XXX: I don't trust the -1's and +1's... they must come * from somewhere! */ - timing->reg_100224 = ((tUNK_0 + tUNK_19 + 1) << 24 | + timing->reg_100224 = (tUNK_0 + tUNK_19 + 1 + magic_number) << 24 | tUNK_18 << 16 | - (tUNK_1 + tUNK_19 + 1) << 8 | - (tUNK_2 - 1)); + (tUNK_1 + tUNK_19 + 1 + magic_number) << 8; + if(dev_priv->chipset == 0xa8) { + timing->reg_100224 |= (tUNK_2 - 1); + } else { + timing->reg_100224 |= (tUNK_2 + 2 - magic_number); + } timing->reg_100228 = (tUNK_12 << 16 | tUNK_11 << 8 | tUNK_10); - if(recordlen > 19) { - timing->reg_100228 += (tUNK_19 - 1) << 24; - }/* I cannot back-up this else-statement right now - else { - timing->reg_100228 += tUNK_12 << 24; - }*/ - - /* XXX: reg_10022c */ - timing->reg_10022c = tUNK_2 - 1; - - timing->reg_100230 = (tUNK_20 << 24 | tUNK_21 << 16 | - tUNK_13 << 8 | tUNK_13); - - /* XXX: +6? */ - timing->reg_100234 = (tRAS << 24 | (tUNK_19 + 6) << 8 | tRC); - timing->reg_100234 += max(tUNK_10,tUNK_11) << 16; - - /* XXX; reg_100238, reg_10023c - * reg: 0x00?????? - * reg_10023c: - * 0 for pre-NV50 cards - * 0x????0202 for NV50+ cards (empirical evidence) */ - if(dev_priv->card_type >= NV_50) { + if(dev_priv->chipset >= 0xa3 && dev_priv->chipset < 0xaa) { + timing->reg_100228 |= (tUNK_19 - 1) << 24; + } + + if(dev_priv->card_type == NV_40) { + /* NV40: don't know what the rest of the regs are.. + * And don't need to know either */ + timing->reg_100228 |= 0x20200000 | magic_number << 24; + } else if(dev_priv->card_type >= NV_50) { + /* XXX: reg_10022c */ + timing->reg_10022c = tUNK_2 - 1; + + timing->reg_100230 = (tUNK_20 << 24 | tUNK_21 << 16 | + tUNK_13 << 8 | tUNK_13); + + timing->reg_100234 = (tRAS << 24 | tRC); + timing->reg_100234 += max(tUNK_10,tUNK_11) << 16; + + if(dev_priv->chipset < 0xa3) { + timing->reg_100234 |= (tUNK_2 + 2) << 8; + } else { + /* XXX: +6? */ + timing->reg_100234 |= (tUNK_19 + 6) << 8; + } + + /* XXX; reg_100238, reg_10023c + * reg_100238: 0x00?????? + * reg_10023c: 0x!!??0202 for NV50+ cards (empirical evidence) */ timing->reg_10023c = 0x202; + if(dev_priv->chipset < 0xa3) { + timing->reg_10023c |= 0x4000000 | (tUNK_2 - 1) << 16; + } else { + /* currently unknown + * 10023c seen as 06xxxxxx, 0bxxxxxx or 0fxxxxxx */ + } } NV_DEBUG(dev, "Entry %d: 220: %08x %08x %08x %08x\n", i, @@ -675,7 +697,7 @@ nouveau_mem_timing_init(struct drm_device *dev) timing->reg_100238, timing->reg_10023c); } - memtimings->nr_timing = entries; + memtimings->nr_timing = entries; memtimings->supported = true; } -- cgit v0.10.2 From b251d1a488b8a742b621f4cd28f7e57a7bacc434 Mon Sep 17 00:00:00 2001 From: Emil Velikov Date: Fri, 18 Mar 2011 20:19:53 +0000 Subject: nv30: Fix parsing of perf table Perf tables v 1.2 and 1.3 (seen on Geforce FX/ 5) are not long enough to store the voltage label/id v2 - Remove comment from the code Signed-off-by: Emil Velikov Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nouveau_perf.c b/drivers/gpu/drm/nouveau/nouveau_perf.c index ac62a1b..670e3cb 100644 --- a/drivers/gpu/drm/nouveau/nouveau_perf.c +++ b/drivers/gpu/drm/nouveau/nouveau_perf.c @@ -134,7 +134,7 @@ nouveau_perf_init(struct drm_device *dev) case 0x13: case 0x15: perflvl->fanspeed = entry[55]; - perflvl->voltage = entry[56]; + perflvl->voltage = (recordlen > 56) ? entry[56] : 0; perflvl->core = ROM32(entry[1]) * 10; perflvl->memory = ROM32(entry[5]) * 20; break; -- cgit v0.10.2 From fb522ee1aa57f2bb57ebd1111524601076d9ad0a Mon Sep 17 00:00:00 2001 From: Marcin Slusarz Date: Thu, 10 Mar 2011 22:43:25 +0100 Subject: drm/nouveau: fix oops on unload with disabled LVDS panel Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=35135 BUG: unable to handle kernel NULL pointer dereference at 000002d8 IP: [] nv04_dfp_restore+0x7f/0xd0 [nouveau] (...) Call Trace: [] nv04_display_destroy+0xa8/0x140 [nouveau] [] nouveau_unload+0x2a/0x160 [nouveau] [] drm_put_dev+0xbb/0x1b0 [drm] [] nouveau_pci_remove+0x15/0x20 [nouveau] [] pci_device_remove+0x44/0xf0 [] __device_release_driver+0x51/0xb0 [] driver_detach+0x8f/0xa0 [] bus_remove_driver+0x63/0xa0 [] driver_unregister+0x49/0x80 [] ? sysfs_remove_file+0x14/0x20 [] pci_unregister_driver+0x32/0x90 [] ? __stop_machine+0x5a/0x70 [] drm_exit+0x83/0x90 [drm] [] nouveau_exit+0x1b/0x8be [nouveau] [] sys_delete_module+0x13b/0x1f0 [] ? do_munmap+0x1fe/0x280 [] ? arch_unmap_area_topdown+0x0/0x20 [] syscall_call+0x7/0xb Reported-by: Francesco Marella Tested-by: Francesco Marella Signed-off-by: Marcin Slusarz [ currojerez@riseup.net: No need to spam the logs in that case, an unbound LVDS encoder is not an error. ] Signed-off-by: Francisco Jerez Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nv04_dfp.c b/drivers/gpu/drm/nouveau/nv04_dfp.c index c82db37..12098bf 100644 --- a/drivers/gpu/drm/nouveau/nv04_dfp.c +++ b/drivers/gpu/drm/nouveau/nv04_dfp.c @@ -581,12 +581,13 @@ static void nv04_dfp_restore(struct drm_encoder *encoder) int head = nv_encoder->restore.head; if (nv_encoder->dcb->type == OUTPUT_LVDS) { - struct drm_display_mode *native_mode = nouveau_encoder_connector_get(nv_encoder)->native_mode; - if (native_mode) - call_lvds_script(dev, nv_encoder->dcb, head, LVDS_PANEL_ON, - native_mode->clock); - else - NV_ERROR(dev, "Not restoring LVDS without native mode\n"); + struct nouveau_connector *connector = + nouveau_encoder_connector_get(nv_encoder); + + if (connector && connector->native_mode) + call_lvds_script(dev, nv_encoder->dcb, head, + LVDS_PANEL_ON, + connector->native_mode->clock); } else if (nv_encoder->dcb->type == OUTPUT_TMDS) { int clock = nouveau_hw_pllvals_to_clk -- cgit v0.10.2 From ec64a40868469d308ad1e040c6ead2373af334dd Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Mon, 21 Mar 2011 21:31:21 +1000 Subject: drm/nouveau: implement init table opcode 0x5c Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nouveau_bios.c b/drivers/gpu/drm/nouveau/nouveau_bios.c index 8314a49..eca191a 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bios.c +++ b/drivers/gpu/drm/nouveau/nouveau_bios.c @@ -269,7 +269,7 @@ struct init_tbl_entry { int (*handler)(struct nvbios *, uint16_t, struct init_exec *); }; -static int parse_init_table(struct nvbios *, unsigned int, struct init_exec *); +static int parse_init_table(struct nvbios *, uint16_t, struct init_exec *); #define MACRO_INDEX_SIZE 2 #define MACRO_SIZE 8 @@ -2011,6 +2011,27 @@ init_sub_direct(struct nvbios *bios, uint16_t offset, struct init_exec *iexec) } static int +init_jump(struct nvbios *bios, uint16_t offset, struct init_exec *iexec) +{ + /* + * INIT_JUMP opcode: 0x5C ('\') + * + * offset (8 bit): opcode + * offset + 1 (16 bit): offset (in bios) + * + * Continue execution of init table from 'offset' + */ + + uint16_t jmp_offset = ROM16(bios->data[offset + 1]); + + if (!iexec->execute) + return 3; + + BIOSLOG(bios, "0x%04X: Jump to 0x%04X\n", offset, jmp_offset); + return jmp_offset - offset; +} + +static int init_i2c_if(struct nvbios *bios, uint16_t offset, struct init_exec *iexec) { /* @@ -3659,6 +3680,7 @@ static struct init_tbl_entry itbl_entry[] = { { "INIT_ZM_REG_SEQUENCE" , 0x58, init_zm_reg_sequence }, /* INIT_INDIRECT_REG (0x5A, 7, 0, 0) removed due to no example of use */ { "INIT_SUB_DIRECT" , 0x5B, init_sub_direct }, + { "INIT_JUMP" , 0x5C, init_jump }, { "INIT_I2C_IF" , 0x5E, init_i2c_if }, { "INIT_COPY_NV_REG" , 0x5F, init_copy_nv_reg }, { "INIT_ZM_INDEX_IO" , 0x62, init_zm_index_io }, @@ -3700,8 +3722,7 @@ static struct init_tbl_entry itbl_entry[] = { #define MAX_TABLE_OPS 1000 static int -parse_init_table(struct nvbios *bios, unsigned int offset, - struct init_exec *iexec) +parse_init_table(struct nvbios *bios, uint16_t offset, struct init_exec *iexec) { /* * Parses all commands in an init table. -- cgit v0.10.2 From 59197c026257ab730c7fc8819a0e2b26e125a534 Mon Sep 17 00:00:00 2001 From: David Dillow Date: Mon, 21 Mar 2011 21:41:47 +1000 Subject: drm/nv50-nvc0: work around an evo channel hang that some people see Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nv50_evo.c b/drivers/gpu/drm/nouveau/nv50_evo.c index a2cfaa6..c8e83c1 100644 --- a/drivers/gpu/drm/nouveau/nv50_evo.c +++ b/drivers/gpu/drm/nouveau/nv50_evo.c @@ -186,6 +186,7 @@ nv50_evo_channel_init(struct nouveau_channel *evo) nv_mask(dev, 0x610028, 0x00000000, 0x00010001 << id); evo->dma.max = (4096/4) - 2; + evo->dma.max &= ~7; evo->dma.put = 0; evo->dma.cur = evo->dma.put; evo->dma.free = evo->dma.max - evo->dma.cur; -- cgit v0.10.2 From c0929b499f834210561fe5e8c48bcad4f2130d25 Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Mon, 21 Mar 2011 11:42:51 +1000 Subject: drm/nouveau: quirk for XFX GT-240X-YA Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nouveau_bios.c b/drivers/gpu/drm/nouveau/nouveau_bios.c index eca191a..90aef64 100644 --- a/drivers/gpu/drm/nouveau/nouveau_bios.c +++ b/drivers/gpu/drm/nouveau/nouveau_bios.c @@ -6354,6 +6354,32 @@ apply_dcb_encoder_quirks(struct drm_device *dev, int idx, u32 *conn, u32 *conf) } } + /* XFX GT-240X-YA + * + * So many things wrong here, replace the entire encoder table.. + */ + if (nv_match_device(dev, 0x0ca3, 0x1682, 0x3003)) { + if (idx == 0) { + *conn = 0x02001300; /* VGA, connector 1 */ + *conf = 0x00000028; + } else + if (idx == 1) { + *conn = 0x01010312; /* DVI, connector 0 */ + *conf = 0x00020030; + } else + if (idx == 2) { + *conn = 0x01010310; /* VGA, connector 0 */ + *conf = 0x00000028; + } else + if (idx == 3) { + *conn = 0x02022362; /* HDMI, connector 2 */ + *conf = 0x00020010; + } else { + *conn = 0x0000000e; /* EOL */ + *conf = 0x00000000; + } + } + return true; } -- cgit v0.10.2 From 2b4cebe4e165b0ef30a138e4cf602538dea15583 Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Tue, 29 Mar 2011 09:56:14 +1000 Subject: drm/nv50: use "nv86" tlb flush method on everything except 0x50/0xac It has been reported that this greatly improves (and possibly fixes completely) the stability of NVA3+ chipsets. In traces of my NVA8, NVIDIA now appear to be doing this too. The most recent traces of 0x50 and 0xac I could find don't show NVIDIA checking PGRAPH status on these flushes, so for now, we won't either. Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h index fff180a..59ed5be 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drv.h +++ b/drivers/gpu/drm/nouveau/nouveau_drv.h @@ -1190,7 +1190,7 @@ extern int nv50_graph_load_context(struct nouveau_channel *); extern int nv50_graph_unload_context(struct drm_device *); extern int nv50_grctx_init(struct nouveau_grctx *); extern void nv50_graph_tlb_flush(struct drm_device *dev); -extern void nv86_graph_tlb_flush(struct drm_device *dev); +extern void nv84_graph_tlb_flush(struct drm_device *dev); extern struct nouveau_enum nv50_data_error_names[]; /* nvc0_graph.c */ diff --git a/drivers/gpu/drm/nouveau/nouveau_state.c b/drivers/gpu/drm/nouveau/nouveau_state.c index 4fcbd09..1e7e006 100644 --- a/drivers/gpu/drm/nouveau/nouveau_state.c +++ b/drivers/gpu/drm/nouveau/nouveau_state.c @@ -376,15 +376,11 @@ static int nouveau_init_engine_ptrs(struct drm_device *dev) engine->graph.destroy_context = nv50_graph_destroy_context; engine->graph.load_context = nv50_graph_load_context; engine->graph.unload_context = nv50_graph_unload_context; - if (dev_priv->chipset != 0x86) + if (dev_priv->chipset == 0x50 || + dev_priv->chipset == 0xac) engine->graph.tlb_flush = nv50_graph_tlb_flush; - else { - /* from what i can see nvidia do this on every - * pre-NVA3 board except NVAC, but, we've only - * ever seen problems on NV86 - */ - engine->graph.tlb_flush = nv86_graph_tlb_flush; - } + else + engine->graph.tlb_flush = nv84_graph_tlb_flush; engine->fifo.channels = 128; engine->fifo.init = nv50_fifo_init; engine->fifo.takedown = nv50_fifo_takedown; diff --git a/drivers/gpu/drm/nouveau/nv50_graph.c b/drivers/gpu/drm/nouveau/nv50_graph.c index 8675b00..b02a5b1 100644 --- a/drivers/gpu/drm/nouveau/nv50_graph.c +++ b/drivers/gpu/drm/nouveau/nv50_graph.c @@ -503,7 +503,7 @@ nv50_graph_tlb_flush(struct drm_device *dev) } void -nv86_graph_tlb_flush(struct drm_device *dev) +nv84_graph_tlb_flush(struct drm_device *dev) { struct drm_nouveau_private *dev_priv = dev->dev_private; struct nouveau_timer_engine *ptimer = &dev_priv->engine.timer; -- cgit v0.10.2 From e61e51f134a3299c2c37051f69638cc9e9fd88c1 Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Wed, 30 Mar 2011 11:56:16 +1000 Subject: drm/nv50-nvc0: remove some code that doesn't belong here Not sure how this snuck in... Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nv50_crtc.c b/drivers/gpu/drm/nouveau/nv50_crtc.c index 2b99840..a19ccaa 100644 --- a/drivers/gpu/drm/nouveau/nv50_crtc.c +++ b/drivers/gpu/drm/nouveau/nv50_crtc.c @@ -469,9 +469,6 @@ nv50_crtc_wait_complete(struct drm_crtc *crtc) start = ptimer->read(dev); do { - nv_wr32(dev, 0x61002c, 0x370); - nv_wr32(dev, 0x000140, 1); - if (nv_ro32(disp->ntfy, 0x000)) return 0; } while (ptimer->read(dev) - start < 2000000000ULL); -- cgit v0.10.2 From a719726f4ceaf14842c80fd7e58dec40cb7022a2 Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Wed, 30 Mar 2011 13:57:44 +1000 Subject: drm/nvc0: improve vm flush function Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nvc0_vm.c b/drivers/gpu/drm/nouveau/nvc0_vm.c index 69af0ba..a0a2a02 100644 --- a/drivers/gpu/drm/nouveau/nvc0_vm.c +++ b/drivers/gpu/drm/nouveau/nvc0_vm.c @@ -104,20 +104,26 @@ nvc0_vm_flush(struct nouveau_vm *vm) struct nouveau_instmem_engine *pinstmem = &dev_priv->engine.instmem; struct drm_device *dev = vm->dev; struct nouveau_vm_pgd *vpgd; - u32 r100c80, engine; + u32 engine = (dev_priv->chan_vm == vm) ? 1 : 5; pinstmem->flush(vm->dev); - if (vm == dev_priv->chan_vm) - engine = 1; - else - engine = 5; - + spin_lock(&dev_priv->ramin_lock); list_for_each_entry(vpgd, &vm->pgd_list, head) { - r100c80 = nv_rd32(dev, 0x100c80); + /* looks like maybe a "free flush slots" counter, the + * faster you write to 0x100cbc to more it decreases + */ + if (!nv_wait_ne(dev, 0x100c80, 0x00ff0000, 0x00000000)) { + NV_ERROR(dev, "vm timeout 0: 0x%08x %d\n", + nv_rd32(dev, 0x100c80), engine); + } nv_wr32(dev, 0x100cb8, vpgd->obj->vinst >> 8); nv_wr32(dev, 0x100cbc, 0x80000000 | engine); - if (!nv_wait(dev, 0x100c80, 0xffffffff, r100c80)) - NV_ERROR(dev, "vm flush timeout eng %d\n", engine); + /* wait for flush to be queued? */ + if (!nv_wait(dev, 0x100c80, 0x00008000, 0x00008000)) { + NV_ERROR(dev, "vm timeout 1: 0x%08x %d\n", + nv_rd32(dev, 0x100c80), engine); + } } + spin_unlock(&dev_priv->ramin_lock); } -- cgit v0.10.2 From b2cb1292b1c7c73abbdc0e07ef3aab056fc2615f Mon Sep 17 00:00:00 2001 From: David Henningsson Date: Tue, 5 Apr 2011 07:55:24 +0200 Subject: ALSA: HDA: Fix dock mic for Lenovo X220-tablet Without the "thinkpad" quirk, the dock mic in Lenovo X220 tablet edition won't work. BugLink: http://bugs.launchpad.net/bugs/751033 Cc: stable@kernel.org Tested-by: James Ferguson Signed-off-by: David Henningsson Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_conexant.c b/sound/pci/hda/patch_conexant.c index 69e3386..ad97d93 100644 --- a/sound/pci/hda/patch_conexant.c +++ b/sound/pci/hda/patch_conexant.c @@ -3035,6 +3035,7 @@ static struct snd_pci_quirk cxt5066_cfg_tbl[] = { SND_PCI_QUIRK(0x17aa, 0x21c6, "Thinkpad Edge 13", CXT5066_ASUS), SND_PCI_QUIRK(0x17aa, 0x215e, "Lenovo Thinkpad", CXT5066_THINKPAD), SND_PCI_QUIRK(0x17aa, 0x21da, "Lenovo X220", CXT5066_THINKPAD), + SND_PCI_QUIRK(0x17aa, 0x21db, "Lenovo X220-tablet", CXT5066_THINKPAD), SND_PCI_QUIRK(0x17aa, 0x38af, "Lenovo G560", CXT5066_ASUS), SND_PCI_QUIRK_VENDOR(0x17aa, "Lenovo", CXT5066_IDEAPAD), /* Fallback for Lenovos without dock mic */ {} -- cgit v0.10.2 From 6cba611e600ded15f642552ce6b5f7ee243bacf0 Mon Sep 17 00:00:00 2001 From: Zhang Huan Date: Tue, 5 Apr 2011 19:16:20 -0400 Subject: jbd2: fix potential memory leak on transaction commit There is potential memory leak of journal head in function jbd2_journal_commit_transaction. The problem is that JBD2 will not reclaim the journal head of commit record if error occurs or journal is abotred. I use the following script to reproduce this issue, on a RHEL6 system. I found it very easy to reproduce with async commit enabled. mount /dev/sdb /mnt -o journal_checksum,journal_async_commit touch /mnt/xxx echo offline > /sys/block/sdb/device/state sync umount /mnt rmmod ext4 rmmod jbd2 Removal of the jbd2 module will make slab complaining that "cache `jbd2_journal_head': can't free all objects". Signed-off-by: Zhang Huan Signed-off-by: "Theodore Ts'o" diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index fa36d76..b98e4c1 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -105,6 +105,8 @@ static int journal_submit_commit_record(journal_t *journal, int ret; struct timespec now = current_kernel_time(); + *cbh = NULL; + if (is_journal_aborted(journal)) return 0; @@ -806,7 +808,7 @@ wait_for_iobuf: if (err) __jbd2_journal_abort_hard(journal); } - if (!err && !is_journal_aborted(journal)) + if (cbh) err = journal_wait_on_commit_record(journal, cbh); if (JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT) && -- cgit v0.10.2 From 0449641130f5652b344ef6fa39fa019d7e94660a Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Tue, 5 Apr 2011 19:55:28 -0400 Subject: ext4: init timer earlier to avoid a kernel panic in __save_error_info During mount, when we fail to open journal inode or root inode, the __save_error_info will mod_timer. But actually s_err_report isn't initialized yet and the kernel oops. The detailed information can be found https://bugzilla.kernel.org/show_bug.cgi?id=32082. The best way is to check whether the timer s_err_report is initialized or not. But it seems that in include/linux/timer.h, we can't find a good function to check the status of this timer, so this patch just move the initializtion of s_err_report earlier so that we can avoid the kernel panic. The corresponding del_timer is also added in the error path. Reported-by: Sami Liedes Signed-off-by: Tao Ma Signed-off-by: "Theodore Ts'o" diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 35bd020..551cb8e 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -3391,6 +3391,10 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) get_random_bytes(&sbi->s_next_generation, sizeof(u32)); spin_lock_init(&sbi->s_next_gen_lock); + init_timer(&sbi->s_err_report); + sbi->s_err_report.function = print_daily_error_info; + sbi->s_err_report.data = (unsigned long) sb; + err = percpu_counter_init(&sbi->s_freeblocks_counter, ext4_count_free_blocks(sb)); if (!err) { @@ -3652,9 +3656,6 @@ no_journal: "Opts: %s%s%s", descr, sbi->s_es->s_mount_opts, *sbi->s_es->s_mount_opts ? "; " : "", orig_data); - init_timer(&sbi->s_err_report); - sbi->s_err_report.function = print_daily_error_info; - sbi->s_err_report.data = (unsigned long) sb; if (es->s_error_count) mod_timer(&sbi->s_err_report, jiffies + 300*HZ); /* 5 minutes */ @@ -3678,6 +3679,7 @@ failed_mount_wq: sbi->s_journal = NULL; } failed_mount3: + del_timer(&sbi->s_err_report); if (sbi->s_flex_groups) { if (is_vmalloc_addr(sbi->s_flex_groups)) vfree(sbi->s_flex_groups); -- cgit v0.10.2 From 49c039f071d36586ba32da75996ef339e4ab8405 Mon Sep 17 00:00:00 2001 From: Tarek Soliman Date: Mon, 4 Apr 2011 09:23:53 -0500 Subject: ALSA: usb-audio: define another USB ID for a buggy USB MIDI cable There are many USB MIDI cables out there that have buggy firmware that reports it can do more than 4 bytes in a packet when they can only properly handle 4 This patch adds the ID of yet another one of those cables Signed-off-by: Tarek Soliman Signed-off-by: Takashi Iwai diff --git a/sound/usb/midi.c b/sound/usb/midi.c index b4b39c0..f928910 100644 --- a/sound/usb/midi.c +++ b/sound/usb/midi.c @@ -1301,6 +1301,7 @@ static int snd_usbmidi_out_endpoint_create(struct snd_usb_midi* umidi, case USB_ID(0x15ca, 0x0101): /* Textech USB Midi Cable */ case USB_ID(0x15ca, 0x1806): /* Textech USB Midi Cable */ case USB_ID(0x1a86, 0x752d): /* QinHeng CH345 "USB2.0-MIDI" */ + case USB_ID(0xfc08, 0x0101): /* Unknown vendor Cable */ ep->max_transfer = 4; break; /* -- cgit v0.10.2 From 61f4237d5b005767a76f4f3694e68e6f78f392d9 Mon Sep 17 00:00:00 2001 From: Jeremy Fitzhardinge Date: Sat, 18 Sep 2010 22:25:30 -0700 Subject: xen: just completely disable XSAVE Some (old) versions of Xen just kill the domain if it tries to set any unknown bits in CR4, so we can't reliably probe for OSXSAVE in CR4. Since Xen doesn't support XSAVE for guests at the moment, and no such support is being worked on, there's no downside in just unconditionally masking XSAVE support. Signed-off-by: Jeremy Fitzhardinge Signed-off-by: Konrad Rzeszutek Wilk diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index 49dbd78..66272a2 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -250,23 +250,7 @@ static __init void xen_init_cpuid_mask(void) ~((1 << X86_FEATURE_APIC) | /* disable local APIC */ (1 << X86_FEATURE_ACPI)); /* disable ACPI */ - ax = 1; - cx = 0; - xen_cpuid(&ax, &bx, &cx, &dx); - - /* cpuid claims we support xsave; try enabling it to see what happens */ - if (cx & (1 << (X86_FEATURE_XSAVE % 32))) { - unsigned long cr4; - - set_in_cr4(X86_CR4_OSXSAVE); - - cr4 = read_cr4(); - - if ((cr4 & X86_CR4_OSXSAVE) == 0) - cpuid_leaf1_ecx_mask &= ~(1 << (X86_FEATURE_XSAVE % 32)); - - clear_in_cr4(X86_CR4_OSXSAVE); - } + cpuid_leaf1_ecx_mask &= ~(1 << (X86_FEATURE_XSAVE % 32)); /* disable XSAVE */ } static void xen_set_debugreg(int reg, unsigned long val) -- cgit v0.10.2 From 947ccf9c3c30307b774af3666ee74fcd9f47f646 Mon Sep 17 00:00:00 2001 From: Shan Haitao Date: Tue, 9 Nov 2010 11:43:36 -0800 Subject: xen: Allow PV-OPS kernel to detect whether XSAVE is supported Xen fails to mask XSAVE from the cpuid feature, despite not historically supporting guest use of XSAVE. However, now that XSAVE support has been added to Xen, we need to reliably detect its presence. The most reliable way to do this is to look at the OSXSAVE feature in cpuid which is set iff the OS (Xen, in this case), has set CR4.OSXSAVE. [ Cleaned up conditional a bit. - Jeremy ] Signed-off-by: Shan Haitao Signed-off-by: Jeremy Fitzhardinge Signed-off-by: Konrad Rzeszutek Wilk diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c index 66272a2..e3c6a06 100644 --- a/arch/x86/xen/enlighten.c +++ b/arch/x86/xen/enlighten.c @@ -238,6 +238,7 @@ static void xen_cpuid(unsigned int *ax, unsigned int *bx, static __init void xen_init_cpuid_mask(void) { unsigned int ax, bx, cx, dx; + unsigned int xsave_mask; cpuid_leaf1_edx_mask = ~((1 << X86_FEATURE_MCE) | /* disable MCE */ @@ -249,8 +250,16 @@ static __init void xen_init_cpuid_mask(void) cpuid_leaf1_edx_mask &= ~((1 << X86_FEATURE_APIC) | /* disable local APIC */ (1 << X86_FEATURE_ACPI)); /* disable ACPI */ + ax = 1; + xen_cpuid(&ax, &bx, &cx, &dx); - cpuid_leaf1_ecx_mask &= ~(1 << (X86_FEATURE_XSAVE % 32)); /* disable XSAVE */ + xsave_mask = + (1 << (X86_FEATURE_XSAVE % 32)) | + (1 << (X86_FEATURE_OSXSAVE % 32)); + + /* Xen will set CR4.OSXSAVE if supported and not disabled by force */ + if ((cx & xsave_mask) != xsave_mask) + cpuid_leaf1_ecx_mask &= ~xsave_mask; /* disable XSAVE & OSXSAVE */ } static void xen_set_debugreg(int reg, unsigned long val) -- cgit v0.10.2 From 77f38e0eeac290827f41fd2215ab82546b8f73b8 Mon Sep 17 00:00:00 2001 From: Sage Weil Date: Wed, 6 Apr 2011 09:09:16 -0700 Subject: libceph: fix linger request requeueing Fix the request transition from linger -> normal request. The key is to preserve r_osd and requeue on the same OSD. Reregister as a normal request, add the request to the proper queues, then unregister the linger. Fix the unregister helper to avoid clearing r_osd (and also simplify the parallel check in __unregister_request()). Reported-by: Henry Chang Signed-off-by: Sage Weil diff --git a/net/ceph/osd_client.c b/net/ceph/osd_client.c index 3b91d65..9204de4 100644 --- a/net/ceph/osd_client.c +++ b/net/ceph/osd_client.c @@ -579,9 +579,15 @@ static void __kick_osd_requests(struct ceph_osd_client *osdc, list_for_each_entry_safe(req, nreq, &osd->o_linger_requests, r_linger_osd) { - __unregister_linger_request(osdc, req); + /* + * reregister request prior to unregistering linger so + * that r_osd is preserved. + */ + BUG_ON(!list_empty(&req->r_req_lru_item)); __register_request(osdc, req); - list_move(&req->r_req_lru_item, &osdc->req_unsent); + list_add(&req->r_req_lru_item, &osdc->req_unsent); + list_add(&req->r_osd_item, &req->r_osd->o_requests); + __unregister_linger_request(osdc, req); dout("requeued lingering %p tid %llu osd%d\n", req, req->r_tid, osd->o_osd); } @@ -798,7 +804,7 @@ static void __register_request(struct ceph_osd_client *osdc, req->r_request->hdr.tid = cpu_to_le64(req->r_tid); INIT_LIST_HEAD(&req->r_req_lru_item); - dout("register_request %p tid %lld\n", req, req->r_tid); + dout("__register_request %p tid %lld\n", req, req->r_tid); __insert_request(osdc, req); ceph_osdc_get_request(req); osdc->num_requests++; -- cgit v0.10.2 From 0867659fa3c245bf203d837a82e0f6ea5079c2c5 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 6 Apr 2011 10:13:32 -0700 Subject: Revert "net/sunrpc: Use static const char arrays" This reverts commit 411b5e05617593efebc06241dbc56f42150f2abe. Olga Kornievskaia reports: Problem: linux client mounting linux server using rc4-hmac-md5 enctype. gssd fails with create a context after receiving a reply from the server. Diagnose: putting printout statements in the server kernel and kerberos libraries revealed that client and server derived different integrity keys. Server kernel code was at fault due the the commit [aglo@skydive linux-pnfs]$ git show 411b5e05617593efebc06241dbc56f42150f2abe Trond: The problem is that since it relies on virt_to_page(), you cannot call sg_set_buf() for data in the const section. Reported-by: Olga Kornievskaia Signed-off-by: Trond Myklebust Cc: stable@kernel.org [2.6.36+] diff --git a/net/sunrpc/auth_gss/gss_krb5_mech.c b/net/sunrpc/auth_gss/gss_krb5_mech.c index 9022f0a..0a9a2ec 100644 --- a/net/sunrpc/auth_gss/gss_krb5_mech.c +++ b/net/sunrpc/auth_gss/gss_krb5_mech.c @@ -427,7 +427,7 @@ static int context_derive_keys_rc4(struct krb5_ctx *ctx) { struct crypto_hash *hmac; - static const char sigkeyconstant[] = "signaturekey"; + char sigkeyconstant[] = "signaturekey"; int slen = strlen(sigkeyconstant) + 1; /* include null terminator */ struct hash_desc desc; struct scatterlist sg[1]; -- cgit v0.10.2 From 35a67edf3570e9b44c34965dd52a6b7f845fb40f Mon Sep 17 00:00:00 2001 From: Enric Balletbo i Serra Date: Tue, 5 Apr 2011 06:52:49 +0000 Subject: smsc911x: fix mac_lock acquision before calling smsc911x_mac_read When SMSC911X_SAVE_MAC_ADDRESS flag is enabled the driver calls smsc911x_mac_read and smsc911x_mac_read function without acquiring mac_lock spinlock This patch fixes following warning smsc911x: Driver version 2008-10-21. ------------[ cut here ]------------ WARNING: at drivers/net/smsc911x.c:261 smsc911x_mac_read+0x24/0x220() Modules linked in: [] (unwind_backtrace+0x0/0xe0) from [] (warn_slowpath_common+0x4c/0x64) [] (warn_slowpath_common+0x4c/0x64) from [] (warn_slowpath_null+0x18/0x1c) [] (warn_slowpath_null+0x18/0x1c) from [] (smsc911x_mac_read+0x24/0x220) [] (smsc911x_mac_read+0x24/0x220) from [] (smsc911x_read_mac_address+0x18/0x6c) [] (smsc911x_read_mac_address+0x18/0x6c) from [] (smsc911x_drv_probe+0x498/0x1788) [] (smsc911x_drv_probe+0x498/0x1788) from [] (platform_drv_probe+0x14/0x18) [] (platform_drv_probe+0x14/0x18) from [] (driver_probe_device+0xc8/0x184) [] (driver_probe_device+0xc8/0x184) from [] (__driver_attach+0x68/0x8c) [] (__driver_attach+0x68/0x8c) from [] (bus_for_each_dev+0x48/0x74) [] (bus_for_each_dev+0x48/0x74) from [] (bus_add_driver+0x9c/0x228) [] (bus_add_driver+0x9c/0x228) from [] (driver_register+0xa0/0x124) [] (driver_register+0xa0/0x124) from [] (do_one_initcall+0x94/0x168) [] (do_one_initcall+0x94/0x168) from [] (kernel_init+0xa0/0x154) [] (kernel_init+0xa0/0x154) from [] (kernel_thread_exit+0x0/0x8) ---[ end trace 2c931a35b7885770 ]--- Signed-off-by: Enric Balletbo i Serra Signed-off-by: David S. Miller diff --git a/drivers/net/smsc911x.c b/drivers/net/smsc911x.c index 1566259..4747846 100644 --- a/drivers/net/smsc911x.c +++ b/drivers/net/smsc911x.c @@ -1818,6 +1818,7 @@ static int __devinit smsc911x_init(struct net_device *dev) SMSC_TRACE(PROBE, "PHY will be autodetected."); spin_lock_init(&pdata->dev_lock); + spin_lock_init(&pdata->mac_lock); if (pdata->ioaddr == 0) { SMSC_WARNING(PROBE, "pdata->ioaddr: 0x00000000"); @@ -1895,8 +1896,11 @@ static int __devinit smsc911x_init(struct net_device *dev) /* workaround for platforms without an eeprom, where the mac address * is stored elsewhere and set by the bootloader. This saves the * mac address before resetting the device */ - if (pdata->config.flags & SMSC911X_SAVE_MAC_ADDRESS) + if (pdata->config.flags & SMSC911X_SAVE_MAC_ADDRESS) { + spin_lock_irq(&pdata->mac_lock); smsc911x_read_mac_address(dev); + spin_unlock_irq(&pdata->mac_lock); + } /* Reset the LAN911x */ if (smsc911x_soft_reset(pdata)) @@ -2059,8 +2063,6 @@ static int __devinit smsc911x_drv_probe(struct platform_device *pdev) SMSC_TRACE(PROBE, "Network interface: \"%s\"", dev->name); } - spin_lock_init(&pdata->mac_lock); - retval = smsc911x_mii_init(pdev, dev); if (retval) { SMSC_WARNING(PROBE, -- cgit v0.10.2 From 34206f267120c839a479d0237db907fa062e7b0f Mon Sep 17 00:00:00 2001 From: Enric Balletbo i Serra Date: Tue, 5 Apr 2011 07:08:41 +0000 Subject: can: mcp251x: Allow pass IRQ flags through platform data. When an interrupt occurs, the INT pin is driven low by the MCP251x controller (falling edge) but in some cases the INT pin can be connected to the MPU through a transistor or level translator which inverts this signal. In this case interrupt should be configured in rising edge. This patch adds support to pass the IRQ flags via mcp251x_platform_data. Signed-off-by: Enric Balletbo i Serra Acked-by: Wolfgang Grandegger Acked-by: Marc Kleine-Budde Signed-off-by: David S. Miller diff --git a/drivers/net/can/mcp251x.c b/drivers/net/can/mcp251x.c index 7513c45..330140e 100644 --- a/drivers/net/can/mcp251x.c +++ b/drivers/net/can/mcp251x.c @@ -931,7 +931,8 @@ static int mcp251x_open(struct net_device *net) priv->tx_len = 0; ret = request_threaded_irq(spi->irq, NULL, mcp251x_can_ist, - IRQF_TRIGGER_FALLING, DEVICE_NAME, priv); + pdata->irq_flags ? pdata->irq_flags : IRQF_TRIGGER_FALLING, + DEVICE_NAME, priv); if (ret) { dev_err(&spi->dev, "failed to acquire irq %d\n", spi->irq); if (pdata->transceiver_enable) diff --git a/include/linux/can/platform/mcp251x.h b/include/linux/can/platform/mcp251x.h index 8e20540..089fe43 100644 --- a/include/linux/can/platform/mcp251x.h +++ b/include/linux/can/platform/mcp251x.h @@ -12,6 +12,7 @@ /** * struct mcp251x_platform_data - MCP251X SPI CAN controller platform data * @oscillator_frequency: - oscillator frequency in Hz + * @irq_flags: - IRQF configuration flags * @board_specific_setup: - called before probing the chip (power,reset) * @transceiver_enable: - called to power on/off the transceiver * @power_enable: - called to power on/off the mcp *and* the @@ -24,6 +25,7 @@ struct mcp251x_platform_data { unsigned long oscillator_frequency; + unsigned long irq_flags; int (*board_specific_setup)(struct spi_device *spi); int (*transceiver_enable)(int enable); int (*power_enable) (int enable); -- cgit v0.10.2 From 1e58148670908e9fd180761c0eed3839b6778f01 Mon Sep 17 00:00:00 2001 From: Rasesh Mody Date: Mon, 4 Apr 2011 08:29:59 +0000 Subject: bna: Fix for handling firmware heartbeat failure This patch contains a fix for gracefully handling firmware heartbeat failure instead of forcing panic. Signed-off-by: Debashis Dutt Signed-off-by: Rasesh Mody Signed-off-by: David S. Miller diff --git a/drivers/net/bna/bfa_ioc.c b/drivers/net/bna/bfa_ioc.c index 34933cb..e3de0b8 100644 --- a/drivers/net/bna/bfa_ioc.c +++ b/drivers/net/bna/bfa_ioc.c @@ -2219,13 +2219,9 @@ bfa_nw_ioc_get_mac(struct bfa_ioc *ioc) static void bfa_ioc_recover(struct bfa_ioc *ioc) { - u16 bdf; - - bdf = (ioc->pcidev.pci_slot << 8 | ioc->pcidev.pci_func << 3 | - ioc->pcidev.device_id); - - pr_crit("Firmware heartbeat failure at %d", bdf); - BUG_ON(1); + pr_crit("Heart Beat of IOC has failed\n"); + bfa_ioc_stats(ioc, ioc_hbfails); + bfa_fsm_send_event(ioc, IOC_E_HBFAIL); } static void -- cgit v0.10.2 From 2d5d4154650459b61a8e7097d186a89d89dab8ed Mon Sep 17 00:00:00 2001 From: Ajit Khaparde Date: Wed, 6 Apr 2011 05:53:13 +0000 Subject: be2net: Fix a potential crash during shutdown. adapter could remain uninitialized if probe fails for some reason. A null pointer access could cause a crash if be_shutdown is called after that. Signed-off-by: Ajit Khaparde Signed-off-by: David S. Miller diff --git a/drivers/net/benet/be_main.c b/drivers/net/benet/be_main.c index a71163f..6e8e211 100644 --- a/drivers/net/benet/be_main.c +++ b/drivers/net/benet/be_main.c @@ -3141,12 +3141,14 @@ static int be_resume(struct pci_dev *pdev) static void be_shutdown(struct pci_dev *pdev) { struct be_adapter *adapter = pci_get_drvdata(pdev); - struct net_device *netdev = adapter->netdev; - if (netif_running(netdev)) + if (!adapter) + return; + + if (netif_running(adapter->netdev)) cancel_delayed_work_sync(&adapter->work); - netif_device_detach(netdev); + netif_device_detach(adapter->netdev); be_cmd_reset_function(adapter); -- cgit v0.10.2 From 47482f132a689af168fae3055ff1899dfd032d3a Mon Sep 17 00:00:00 2001 From: Neil Horman Date: Wed, 6 Apr 2011 13:07:09 -0700 Subject: ipv6: Enable RFS sk_rxhash tracking for ipv6 sockets (v2) properly record sk_rxhash in ipv6 sockets (v2) Noticed while working on another project that flows to sockets which I had open on a test systems weren't getting steered properly when I had RFS enabled. Looking more closely I found that: 1) The affected sockets were all ipv6 2) They weren't getting steered because sk->sk_rxhash was never set from the incomming skbs on that socket. This was occuring because there are several points in the IPv4 tcp and udp code which save the rxhash value when a new connection is established. Those calls to sock_rps_save_rxhash were never added to the corresponding ipv6 code paths. This patch adds those calls. Tested by myself to properly enable RFS functionalty on ipv6. Change notes: v2: Filtered UDP to only arm RFS on bound sockets (Eric Dumazet) Signed-off-by: Neil Horman Signed-off-by: David S. Miller diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index 56fa125..4f49e5d 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1622,6 +1622,7 @@ static int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb) opt_skb = skb_clone(skb, GFP_ATOMIC); if (sk->sk_state == TCP_ESTABLISHED) { /* Fast path */ + sock_rps_save_rxhash(sk, skb->rxhash); if (tcp_rcv_established(sk, skb, tcp_hdr(skb), skb->len)) goto reset; if (opt_skb) @@ -1649,7 +1650,8 @@ static int tcp_v6_do_rcv(struct sock *sk, struct sk_buff *skb) __kfree_skb(opt_skb); return 0; } - } + } else + sock_rps_save_rxhash(sk, skb->rxhash); if (tcp_rcv_state_process(sk, skb, tcp_hdr(skb), skb->len)) goto reset; diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index d7037c0..15c3774 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -505,6 +505,9 @@ int udpv6_queue_rcv_skb(struct sock * sk, struct sk_buff *skb) int rc; int is_udplite = IS_UDPLITE(sk); + if (!ipv6_addr_any(&inet6_sk(sk)->daddr)) + sock_rps_save_rxhash(sk, skb->rxhash); + if (!xfrm6_policy_check(sk, XFRM_POLICY_IN, skb)) goto drop; -- cgit v0.10.2 From 418875900e3de4831c84f86ae4756690dac5be77 Mon Sep 17 00:00:00 2001 From: Bryan Schumaker Date: Wed, 6 Apr 2011 14:33:28 -0400 Subject: NFS: Fix a signed vs. unsigned secinfo bug rpc_authflavor_t is cast from an unsigned int, but the initial code tried to use it as a signed int. I fix this by passing an rpc_authflavor_t pointer around, and returning signed integers from functions. Signed-off-by: Bryan Schumaker Reported-by: Dan Carpenter Signed-off-by: Trond Myklebust diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c index 9166fcb..89fc160 100644 --- a/fs/nfs/namespace.c +++ b/fs/nfs/namespace.c @@ -148,67 +148,64 @@ static rpc_authflavor_t nfs_find_best_sec(struct nfs4_secinfo_flavors *flavors, return pseudoflavor; } -static rpc_authflavor_t nfs_negotiate_security(const struct dentry *parent, const struct dentry *dentry) +static int nfs_negotiate_security(const struct dentry *parent, + const struct dentry *dentry, + rpc_authflavor_t *flavor) { - int status = 0; struct page *page; struct nfs4_secinfo_flavors *flavors; int (*secinfo)(struct inode *, const struct qstr *, struct nfs4_secinfo_flavors *); - rpc_authflavor_t flavor = RPC_AUTH_UNIX; + int ret = -EPERM; secinfo = NFS_PROTO(parent->d_inode)->secinfo; if (secinfo != NULL) { page = alloc_page(GFP_KERNEL); if (!page) { - status = -ENOMEM; + ret = -ENOMEM; goto out; } flavors = page_address(page); - status = secinfo(parent->d_inode, &dentry->d_name, flavors); - flavor = nfs_find_best_sec(flavors, dentry->d_inode); + ret = secinfo(parent->d_inode, &dentry->d_name, flavors); + *flavor = nfs_find_best_sec(flavors, dentry->d_inode); put_page(page); } - return flavor; - out: - status = -ENOMEM; - return status; + return ret; } -static rpc_authflavor_t nfs_lookup_with_sec(struct nfs_server *server, struct dentry *parent, - struct dentry *dentry, struct path *path, - struct nfs_fh *fh, struct nfs_fattr *fattr) +static int nfs_lookup_with_sec(struct nfs_server *server, struct dentry *parent, + struct dentry *dentry, struct path *path, + struct nfs_fh *fh, struct nfs_fattr *fattr, + rpc_authflavor_t *flavor) { - rpc_authflavor_t flavor; struct rpc_clnt *clone; struct rpc_auth *auth; int err; - flavor = nfs_negotiate_security(parent, path->dentry); - if (flavor < 0) + err = nfs_negotiate_security(parent, path->dentry, flavor); + if (err < 0) goto out; clone = rpc_clone_client(server->client); - auth = rpcauth_create(flavor, clone); + auth = rpcauth_create(*flavor, clone); if (!auth) { - flavor = -EIO; + err = -EIO; goto out_shutdown; } err = server->nfs_client->rpc_ops->lookup(clone, parent->d_inode, &path->dentry->d_name, fh, fattr); - if (err < 0) - flavor = err; out_shutdown: rpc_shutdown_client(clone); out: - return flavor; + return err; } #else /* CONFIG_NFS_V4 */ -static inline rpc_authflavor_t nfs_lookup_with_sec(struct nfs_server *server, - struct dentry *parent, struct dentry *dentry, - struct path *path, struct nfs_fh *fh, - struct nfs_fattr *fattr) +static inline int nfs_lookup_with_sec(struct nfs_server *server, + struct dentry *parent, struct dentry *dentry, + struct path *path, struct nfs_fh *fh, + struct nfs_fattr *fattr, + rpc_authflavor_t *flavor) { return -EPERM; } @@ -234,7 +231,7 @@ struct vfsmount *nfs_d_automount(struct path *path) struct nfs_fh *fh = NULL; struct nfs_fattr *fattr = NULL; int err; - rpc_authflavor_t flavor = 1; + rpc_authflavor_t flavor = RPC_AUTH_UNIX; dprintk("--> nfs_d_automount()\n"); @@ -255,13 +252,8 @@ struct vfsmount *nfs_d_automount(struct path *path) err = server->nfs_client->rpc_ops->lookup(server->client, parent->d_inode, &path->dentry->d_name, fh, fattr); - if (err == -EPERM) { - flavor = nfs_lookup_with_sec(server, parent, path->dentry, path, fh, fattr); - if (flavor < 0) - err = flavor; - else - err = 0; - } + if (err == -EPERM && NFS_PROTO(parent->d_inode)->secinfo != NULL) + err = nfs_lookup_with_sec(server, parent, path->dentry, path, fh, fattr, &flavor); dput(parent); if (err != 0) { mnt = ERR_PTR(err); -- cgit v0.10.2 From ec80bfcb68a0c46443991991d459a0cde773cdea Mon Sep 17 00:00:00 2001 From: Peter Korsgaard Date: Tue, 5 Apr 2011 03:03:56 +0000 Subject: dsa/mv88e6131: add support for mv88e6085 switch The mv88e6085 is identical to the mv88e6095, except that all ports are 10/100 Mb/s, so use the existing setup code except for the cpu/dsa speed selection in _setup_port(). Signed-off-by: Peter Korsgaard Acked-by: Lennert Buytenhek Signed-off-by: David S. Miller diff --git a/net/dsa/mv88e6131.c b/net/dsa/mv88e6131.c index bb2b41b..a8e4f8c 100644 --- a/net/dsa/mv88e6131.c +++ b/net/dsa/mv88e6131.c @@ -14,6 +14,13 @@ #include "dsa_priv.h" #include "mv88e6xxx.h" +/* + * Switch product IDs + */ +#define ID_6085 0x04a0 +#define ID_6095 0x0950 +#define ID_6131 0x1060 + static char *mv88e6131_probe(struct mii_bus *bus, int sw_addr) { int ret; @@ -21,9 +28,11 @@ static char *mv88e6131_probe(struct mii_bus *bus, int sw_addr) ret = __mv88e6xxx_reg_read(bus, sw_addr, REG_PORT(0), 0x03); if (ret >= 0) { ret &= 0xfff0; - if (ret == 0x0950) + if (ret == ID_6085) + return "Marvell 88E6085"; + if (ret == ID_6095) return "Marvell 88E6095/88E6095F"; - if (ret == 0x1060) + if (ret == ID_6131) return "Marvell 88E6131"; } @@ -164,6 +173,7 @@ static int mv88e6131_setup_global(struct dsa_switch *ds) static int mv88e6131_setup_port(struct dsa_switch *ds, int p) { + struct mv88e6xxx_priv_state *ps = (void *)(ds + 1); int addr = REG_PORT(p); u16 val; @@ -171,10 +181,13 @@ static int mv88e6131_setup_port(struct dsa_switch *ds, int p) * MAC Forcing register: don't force link, speed, duplex * or flow control state to any particular values on physical * ports, but force the CPU port and all DSA ports to 1000 Mb/s - * full duplex. + * (100 Mb/s on 6085) full duplex. */ if (dsa_is_cpu_port(ds, p) || ds->dsa_port_mask & (1 << p)) - REG_WRITE(addr, 0x01, 0x003e); + if (ps->id == ID_6085) + REG_WRITE(addr, 0x01, 0x003d); /* 100 Mb/s */ + else + REG_WRITE(addr, 0x01, 0x003e); /* 1000 Mb/s */ else REG_WRITE(addr, 0x01, 0x0003); @@ -286,6 +299,8 @@ static int mv88e6131_setup(struct dsa_switch *ds) mv88e6xxx_ppu_state_init(ds); mutex_init(&ps->stats_mutex); + ps->id = REG_READ(REG_PORT(0), 0x03) & 0xfff0; + ret = mv88e6131_switch_reset(ds); if (ret < 0) return ret; diff --git a/net/dsa/mv88e6xxx.h b/net/dsa/mv88e6xxx.h index eb0e0aa..61156ca2 100644 --- a/net/dsa/mv88e6xxx.h +++ b/net/dsa/mv88e6xxx.h @@ -39,6 +39,8 @@ struct mv88e6xxx_priv_state { * Hold this mutex over snapshot + dump sequences. */ struct mutex stats_mutex; + + int id; /* switch product id */ }; struct mv88e6xxx_hw_stat { -- cgit v0.10.2 From ae07b0b221b6ab2edf9e3abd518aec6cd3f1ba66 Mon Sep 17 00:00:00 2001 From: Ulrich Weber Date: Wed, 6 Apr 2011 14:04:49 -0700 Subject: pppoe: drop PPPOX_ZOMBIEs in pppoe_flush_dev otherwise we loop forever if a PPPoE socket was set to PPPOX_ZOMBIE state by a PADT message when the ethernet device is going down afterwards. Signed-off-by: Ulrich Weber Signed-off-by: David S. Miller diff --git a/drivers/net/pppoe.c b/drivers/net/pppoe.c index 78c0e3c..71b1d8f 100644 --- a/drivers/net/pppoe.c +++ b/drivers/net/pppoe.c @@ -317,7 +317,7 @@ static void pppoe_flush_dev(struct net_device *dev) lock_sock(sk); if (po->pppoe_dev == dev && - sk->sk_state & (PPPOX_CONNECTED | PPPOX_BOUND)) { + sk->sk_state & (PPPOX_CONNECTED | PPPOX_BOUND | PPPOX_ZOMBIE)) { pppox_unbind_sock(sk); sk->sk_state = PPPOX_ZOMBIE; sk->sk_state_change(sk); -- cgit v0.10.2 From ecd6210765f93bd2c578d370d45bb49c710f3e89 Mon Sep 17 00:00:00 2001 From: Padmanabh Ratnakar Date: Sun, 3 Apr 2011 01:54:11 +0000 Subject: be2net: Rename some struct members for clarity Renamed msix_vec_idx to eq_idx in be_eq_obj struct. Renamed msix_vec_next_idx to eq_next_idx in be_adapter structure. These members are used in INTX mode also. Signed-off-by: Sarveswara Rao Mygapula Signed-off-by: Padmanabh Ratnakar Signed-off-by: David S. Miller diff --git a/drivers/net/benet/be.h b/drivers/net/benet/be.h index f803c58..66823ed 100644 --- a/drivers/net/benet/be.h +++ b/drivers/net/benet/be.h @@ -154,7 +154,7 @@ struct be_eq_obj { u16 min_eqd; /* in usecs */ u16 max_eqd; /* in usecs */ u16 cur_eqd; /* in usecs */ - u8 msix_vec_idx; + u8 eq_idx; struct napi_struct napi; }; @@ -291,7 +291,7 @@ struct be_adapter { u32 num_rx_qs; u32 big_page_size; /* Compounded page size shared by rx wrbs */ - u8 msix_vec_next_idx; + u8 eq_next_idx; struct be_drv_stats drv_stats; struct vlan_group *vlan_grp; diff --git a/drivers/net/benet/be_main.c b/drivers/net/benet/be_main.c index 6e8e211..1a4808c 100644 --- a/drivers/net/benet/be_main.c +++ b/drivers/net/benet/be_main.c @@ -1497,7 +1497,7 @@ static int be_tx_queues_create(struct be_adapter *adapter) if (be_cmd_eq_create(adapter, eq, adapter->tx_eq.cur_eqd)) goto tx_eq_free; - adapter->tx_eq.msix_vec_idx = adapter->msix_vec_next_idx++; + adapter->tx_eq.eq_idx = adapter->eq_next_idx++; /* Alloc TX eth compl queue */ @@ -1590,7 +1590,7 @@ static int be_rx_queues_create(struct be_adapter *adapter) if (rc) goto err; - rxo->rx_eq.msix_vec_idx = adapter->msix_vec_next_idx++; + rxo->rx_eq.eq_idx = adapter->eq_next_idx++; /* CQ */ cq = &rxo->cq; @@ -1666,11 +1666,11 @@ static irqreturn_t be_intx(int irq, void *dev) if (!isr) return IRQ_NONE; - if ((1 << adapter->tx_eq.msix_vec_idx & isr)) + if ((1 << adapter->tx_eq.eq_idx & isr)) event_handle(adapter, &adapter->tx_eq); for_all_rx_queues(adapter, rxo, i) { - if ((1 << rxo->rx_eq.msix_vec_idx & isr)) + if ((1 << rxo->rx_eq.eq_idx & isr)) event_handle(adapter, &rxo->rx_eq); } } @@ -1951,7 +1951,7 @@ static void be_sriov_disable(struct be_adapter *adapter) static inline int be_msix_vec_get(struct be_adapter *adapter, struct be_eq_obj *eq_obj) { - return adapter->msix_entries[eq_obj->msix_vec_idx].vector; + return adapter->msix_entries[eq_obj->eq_idx].vector; } static int be_request_irq(struct be_adapter *adapter, -- cgit v0.10.2 From 1f5db833758a9650ad002058e40760aaa11d732e Mon Sep 17 00:00:00 2001 From: Padmanabh Ratnakar Date: Sun, 3 Apr 2011 01:54:39 +0000 Subject: be2net: Fix suspend/resume operation eq_next_idx is not getting reset to zero during suspend. This causes resume to fail. Added the fix. Signed-off-by: Sarveswara Rao Mygapula Signed-off-by: Padmanabh Ratnakar Signed-off-by: David S. Miller diff --git a/drivers/net/benet/be_main.c b/drivers/net/benet/be_main.c index 1a4808c..88d4c80 100644 --- a/drivers/net/benet/be_main.c +++ b/drivers/net/benet/be_main.c @@ -2345,6 +2345,7 @@ static int be_clear(struct be_adapter *adapter) be_mcc_queues_destroy(adapter); be_rx_queues_destroy(adapter); be_tx_queues_destroy(adapter); + adapter->eq_next_idx = 0; if (be_physfn(adapter) && adapter->sriov_enabled) for (vf = 0; vf < num_vfs; vf++) -- cgit v0.10.2 From 9b57e1a79e2142df8412ab223c90ba6f47c6b3a3 Mon Sep 17 00:00:00 2001 From: Ian Campbell Date: Sun, 3 Apr 2011 23:12:23 +0000 Subject: MAINTAINERS: add entry for Xen network backend Signed-off-by: Ian Campbell Signed-off-by: David S. Miller diff --git a/MAINTAINERS b/MAINTAINERS index 6b4b9cd..649600c 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -6916,6 +6916,13 @@ T: git git://git.kernel.org/pub/scm/linux/kernel/git/mjg59/platform-drivers-x86. S: Maintained F: drivers/platform/x86 +XEN NETWORK BACKEND DRIVER +M: Ian Campbell +L: xen-devel@lists.xensource.com (moderated for non-subscribers) +L: netdev@vger.kernel.org +S: Supported +F: drivers/net/xen-netback/* + XEN PCI SUBSYSTEM M: Konrad Rzeszutek Wilk L: xen-devel@lists.xensource.com (moderated for non-subscribers) -- cgit v0.10.2 From 8f74c0661c42104b3e3d2c032bc61efde15360ad Mon Sep 17 00:00:00 2001 From: Shubhrajyoti D Date: Wed, 6 Apr 2011 15:31:22 -0700 Subject: Input: twl4030_keypad - avoid potential NULL-pointer dereference Signed-off-by: Shubhrajyoti D Acked-by: Axel Lin Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/keyboard/twl4030_keypad.c b/drivers/input/keyboard/twl4030_keypad.c index 09bef79..cc06c4b 100644 --- a/drivers/input/keyboard/twl4030_keypad.c +++ b/drivers/input/keyboard/twl4030_keypad.c @@ -338,7 +338,7 @@ static int __devinit twl4030_kp_probe(struct platform_device *pdev) u8 reg; int error; - if (!pdata || !pdata->rows || !pdata->cols || + if (!pdata || !pdata->rows || !pdata->cols || !pdata->keymap_data || pdata->rows > TWL4030_MAX_ROWS || pdata->cols > TWL4030_MAX_COLS) { dev_err(&pdev->dev, "Invalid platform_data\n"); return -EINVAL; -- cgit v0.10.2 From 908433833cac977563e9bef8b206990d846876b4 Mon Sep 17 00:00:00 2001 From: Christoph Fritz Date: Wed, 6 Apr 2011 15:33:16 -0700 Subject: Input: h3600_ts - fix error handling at connect In case of an error in h3600ts_connect(), deconstruct in correct order and with the right calls. Signed-off-by: Christoph Fritz Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/touchscreen/h3600_ts_input.c b/drivers/input/touchscreen/h3600_ts_input.c index efa0688..45f93d0 100644 --- a/drivers/input/touchscreen/h3600_ts_input.c +++ b/drivers/input/touchscreen/h3600_ts_input.c @@ -399,31 +399,34 @@ static int h3600ts_connect(struct serio *serio, struct serio_driver *drv) IRQF_SHARED | IRQF_DISABLED, "h3600_action", &ts->dev)) { printk(KERN_ERR "h3600ts.c: Could not allocate Action Button IRQ!\n"); err = -EBUSY; - goto fail2; + goto fail1; } if (request_irq(IRQ_GPIO_BITSY_NPOWER_BUTTON, npower_button_handler, IRQF_SHARED | IRQF_DISABLED, "h3600_suspend", &ts->dev)) { printk(KERN_ERR "h3600ts.c: Could not allocate Power Button IRQ!\n"); err = -EBUSY; - goto fail3; + goto fail2; } serio_set_drvdata(serio, ts); err = serio_open(serio, drv); if (err) - return err; + goto fail3; //h3600_flite_control(1, 25); /* default brightness */ - input_register_device(ts->dev); + err = input_register_device(ts->dev); + if (err) + goto fail4; return 0; -fail3: free_irq(IRQ_GPIO_BITSY_NPOWER_BUTTON, ts->dev); +fail4: serio_close(serio); +fail3: serio_set_drvdata(serio, NULL); + free_irq(IRQ_GPIO_BITSY_NPOWER_BUTTON, ts->dev); fail2: free_irq(IRQ_GPIO_BITSY_ACTION_BUTTON, ts->dev); -fail1: serio_set_drvdata(serio, NULL); - input_free_device(input_dev); +fail1: input_free_device(input_dev); kfree(ts); return err; } -- cgit v0.10.2 From 67c1b8c6aa354aad14aad85d36508fd73d1c6361 Mon Sep 17 00:00:00 2001 From: Feng Tang Date: Thu, 7 Apr 2011 09:39:44 +0800 Subject: RTC: rtc-mrst: follow on to the change of rtc_device_register() commit f44f7f96a20 (RTC: Initialize kernel state from RTC) will call rtc_read_alarm() inside rtc_device_register(), so rtc-mrst driver need to call dev_set_drvdata() before rtc_device_register() get called. Cc: Alessandro Zummo Cc: John Stultz Cc: Thomas Gleixner Signed-off-by: Feng Tang Signed-off-by: John Stultz diff --git a/drivers/rtc/rtc-mrst.c b/drivers/rtc/rtc-mrst.c index b86bc32..d2c397c 100644 --- a/drivers/rtc/rtc-mrst.c +++ b/drivers/rtc/rtc-mrst.c @@ -342,6 +342,8 @@ vrtc_mrst_do_probe(struct device *dev, struct resource *iomem, int rtc_irq) mrst_rtc.irq = rtc_irq; mrst_rtc.iomem = iomem; + mrst_rtc.dev = dev; + dev_set_drvdata(dev, &mrst_rtc); mrst_rtc.rtc = rtc_device_register(driver_name, dev, &mrst_rtc_ops, THIS_MODULE); @@ -350,8 +352,6 @@ vrtc_mrst_do_probe(struct device *dev, struct resource *iomem, int rtc_irq) goto cleanup0; } - mrst_rtc.dev = dev; - dev_set_drvdata(dev, &mrst_rtc); rename_region(iomem, dev_name(&mrst_rtc.rtc->dev)); spin_lock_irq(&rtc_lock); @@ -376,9 +376,10 @@ vrtc_mrst_do_probe(struct device *dev, struct resource *iomem, int rtc_irq) return 0; cleanup1: - mrst_rtc.dev = NULL; rtc_device_unregister(mrst_rtc.rtc); cleanup0: + dev_set_drvdata(dev, NULL); + mrst_rtc.dev = NULL; release_region(iomem->start, iomem->end + 1 - iomem->start); dev_err(dev, "rtc-mrst: unable to initialise\n"); return retval; -- cgit v0.10.2 From 1f348522844bb1f6e7b10d50b9e8aa89a2511b09 Mon Sep 17 00:00:00 2001 From: Aaron Plattner Date: Wed, 6 Apr 2011 17:19:04 -0700 Subject: ALSA: hda - HDMI: Fix MCP7x audio infoframe checksums The MCP7x hardware computes the audio infoframe channel count automatically, but requires the audio driver to set the audio infoframe checksum manually via the Nv_VERB_SET_Info_Frame_Checksum control verb. When audio starts playing, nvhdmi_8ch_7x_pcm_prepare sets the checksum to (0x71 - chan - chanmask). For example, for 2ch audio, chan == 1 and chanmask == 0 so the checksum is set to 0x70. When audio playback finishes and the device is closed, nvhdmi_8ch_7x_pcm_close resets the channel formats, causing the channel count to revert to 8ch. Since the checksum is not reset, the hardware starts generating audio infoframes with invalid checksums. This causes some displays to blank the video. Fix this by updating the checksum and channel mask when the device is closed and also when it is first initialized. In addition, make sure that the channel mask is appropriate for an 8ch infoframe by setting it to 0x13 (FL FR LFE FC RL RR RLC RRC). Signed-off-by: Aaron Plattner Acked-by: Stephen Warren Cc: Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_hdmi.c b/sound/pci/hda/patch_hdmi.c index 251773e..715615a 100644 --- a/sound/pci/hda/patch_hdmi.c +++ b/sound/pci/hda/patch_hdmi.c @@ -1280,6 +1280,39 @@ static int simple_playback_pcm_prepare(struct hda_pcm_stream *hinfo, stream_tag, format, substream); } +static void nvhdmi_8ch_7x_set_info_frame_parameters(struct hda_codec *codec, + int channels) +{ + unsigned int chanmask; + int chan = channels ? (channels - 1) : 1; + + switch (channels) { + default: + case 0: + case 2: + chanmask = 0x00; + break; + case 4: + chanmask = 0x08; + break; + case 6: + chanmask = 0x0b; + break; + case 8: + chanmask = 0x13; + break; + } + + /* Set the audio infoframe channel allocation and checksum fields. The + * channel count is computed implicitly by the hardware. */ + snd_hda_codec_write(codec, 0x1, 0, + Nv_VERB_SET_Channel_Allocation, chanmask); + + snd_hda_codec_write(codec, 0x1, 0, + Nv_VERB_SET_Info_Frame_Checksum, + (0x71 - chan - chanmask)); +} + static int nvhdmi_8ch_7x_pcm_close(struct hda_pcm_stream *hinfo, struct hda_codec *codec, struct snd_pcm_substream *substream) @@ -1298,6 +1331,10 @@ static int nvhdmi_8ch_7x_pcm_close(struct hda_pcm_stream *hinfo, AC_VERB_SET_STREAM_FORMAT, 0); } + /* The audio hardware sends a channel count of 0x7 (8ch) when all the + * streams are disabled. */ + nvhdmi_8ch_7x_set_info_frame_parameters(codec, 8); + return snd_hda_multi_out_dig_close(codec, &spec->multiout); } @@ -1308,37 +1345,16 @@ static int nvhdmi_8ch_7x_pcm_prepare(struct hda_pcm_stream *hinfo, struct snd_pcm_substream *substream) { int chs; - unsigned int dataDCC1, dataDCC2, chan, chanmask, channel_id; + unsigned int dataDCC1, dataDCC2, channel_id; int i; mutex_lock(&codec->spdif_mutex); chs = substream->runtime->channels; - chan = chs ? (chs - 1) : 1; - switch (chs) { - default: - case 0: - case 2: - chanmask = 0x00; - break; - case 4: - chanmask = 0x08; - break; - case 6: - chanmask = 0x0b; - break; - case 8: - chanmask = 0x13; - break; - } dataDCC1 = AC_DIG1_ENABLE | AC_DIG1_COPYRIGHT; dataDCC2 = 0x2; - /* set the Audio InforFrame Channel Allocation */ - snd_hda_codec_write(codec, 0x1, 0, - Nv_VERB_SET_Channel_Allocation, chanmask); - /* turn off SPDIF once; otherwise the IEC958 bits won't be updated */ if (codec->spdif_status_reset && (codec->spdif_ctls & AC_DIG1_ENABLE)) snd_hda_codec_write(codec, @@ -1413,10 +1429,7 @@ static int nvhdmi_8ch_7x_pcm_prepare(struct hda_pcm_stream *hinfo, } } - /* set the Audio Info Frame Checksum */ - snd_hda_codec_write(codec, 0x1, 0, - Nv_VERB_SET_Info_Frame_Checksum, - (0x71 - chan - chanmask)); + nvhdmi_8ch_7x_set_info_frame_parameters(codec, chs); mutex_unlock(&codec->spdif_mutex); return 0; @@ -1512,6 +1525,11 @@ static int patch_nvhdmi_8ch_7x(struct hda_codec *codec) spec->multiout.max_channels = 8; spec->pcm_playback = &nvhdmi_pcm_playback_8ch_7x; codec->patch_ops = nvhdmi_patch_ops_8ch_7x; + + /* Initialize the audio infoframe channel mask and checksum to something + * valid */ + nvhdmi_8ch_7x_set_info_frame_parameters(codec, 8); + return 0; } -- cgit v0.10.2 From 262ac22d21ee2bf3e1655b2e5e45cc94b356e62f Mon Sep 17 00:00:00 2001 From: David Henningsson Date: Thu, 7 Apr 2011 11:43:00 +0200 Subject: ALSA: HDA: Fix single internal mic on ALC275 (Sony Vaio VPCSB1C5E) In cases where there is only one internal mic connected to ADC 0x11, alc275_setup_dual_adc won't handle the case, so we need to add the ADC node to the array of candidates. Cc: stable@kernel.org BugLink: http://bugs.launchpad.net/bugs/752792 Reported-by: Vincenzo Pii Signed-off-by: David Henningsson Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 12c6f45..d9f1ef7 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -14124,7 +14124,7 @@ static hda_nid_t alc269vb_capsrc_nids[1] = { }; static hda_nid_t alc269_adc_candidates[] = { - 0x08, 0x09, 0x07, + 0x08, 0x09, 0x07, 0x11, }; #define alc269_modes alc260_modes -- cgit v0.10.2 From cd9abc7a22806f2340b853058a634b0e32a9f47c Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Thu, 7 Apr 2011 14:55:57 +0200 Subject: ALSA: hda - Don't query connections for widgets have no connections Fixes the kernel warnings with IDT codecs like hda_codec: connection list not available for 0x1e Signed-off-by: Takashi Iwai diff --git a/sound/pci/hda/patch_sigmatel.c b/sound/pci/hda/patch_sigmatel.c index 05fcd60..a2f57e3 100644 --- a/sound/pci/hda/patch_sigmatel.c +++ b/sound/pci/hda/patch_sigmatel.c @@ -3408,6 +3408,9 @@ static int get_connection_index(struct hda_codec *codec, hda_nid_t mux, hda_nid_t conn[HDA_MAX_NUM_INPUTS]; int i, nums; + if (!(get_wcaps(codec, mux) & AC_WCAP_CONN_LIST)) + return -1; + nums = snd_hda_get_connections(codec, mux, conn, ARRAY_SIZE(conn)); for (i = 0; i < nums; i++) if (conn[i] == nid) -- cgit v0.10.2 From 850a28ecd8044ef36b2c7699d2e3736a410b4d0a Mon Sep 17 00:00:00 2001 From: Vasily Khoruzhick Date: Wed, 6 Apr 2011 17:49:15 +0300 Subject: spi: Fix race condition in stop_queue() There's a race condition in stop_queue() in some drivers - if drv_data->queue is empty, but drv_data->busy is still set (or opposite situation) stop_queue will return -EBUSY. So fix loop condition to check that both drv_data->queue is empty and drv_data->busy is not set. This patch affects following drivers: pxa2xx_spi spi_bfin5xx amba-pl022 dw_spi Signed-off-by: Vasily Khoruzhick Acked-by: Eric Miao Acked-by: Mike Frysinger Signed-off-by: Grant Likely diff --git a/drivers/spi/amba-pl022.c b/drivers/spi/amba-pl022.c index 5a4e0af..4b8357b 100644 --- a/drivers/spi/amba-pl022.c +++ b/drivers/spi/amba-pl022.c @@ -1555,7 +1555,7 @@ static int stop_queue(struct pl022 *pl022) * A wait_queue on the pl022->busy could be used, but then the common * execution path (pump_messages) would be required to call wake_up or * friends on every SPI message. Do this instead */ - while (!list_empty(&pl022->queue) && pl022->busy && limit--) { + while ((!list_empty(&pl022->queue) || pl022->busy) && limit--) { spin_unlock_irqrestore(&pl022->queue_lock, flags); msleep(10); spin_lock_irqsave(&pl022->queue_lock, flags); diff --git a/drivers/spi/dw_spi.c b/drivers/spi/dw_spi.c index 9a61964..d040deb 100644 --- a/drivers/spi/dw_spi.c +++ b/drivers/spi/dw_spi.c @@ -821,7 +821,7 @@ static int stop_queue(struct dw_spi *dws) spin_lock_irqsave(&dws->lock, flags); dws->run = QUEUE_STOPPED; - while (!list_empty(&dws->queue) && dws->busy && limit--) { + while ((!list_empty(&dws->queue) || dws->busy) && limit--) { spin_unlock_irqrestore(&dws->lock, flags); msleep(10); spin_lock_irqsave(&dws->lock, flags); diff --git a/drivers/spi/pxa2xx_spi.c b/drivers/spi/pxa2xx_spi.c index a429b01..3aa7820 100644 --- a/drivers/spi/pxa2xx_spi.c +++ b/drivers/spi/pxa2xx_spi.c @@ -1493,7 +1493,7 @@ static int stop_queue(struct driver_data *drv_data) * execution path (pump_messages) would be required to call wake_up or * friends on every SPI message. Do this instead */ drv_data->run = QUEUE_STOPPED; - while (!list_empty(&drv_data->queue) && drv_data->busy && limit--) { + while ((!list_empty(&drv_data->queue) || drv_data->busy) && limit--) { spin_unlock_irqrestore(&drv_data->lock, flags); msleep(10); spin_lock_irqsave(&drv_data->lock, flags); diff --git a/drivers/spi/spi_bfin5xx.c b/drivers/spi/spi_bfin5xx.c index a284624..f3a1c52 100644 --- a/drivers/spi/spi_bfin5xx.c +++ b/drivers/spi/spi_bfin5xx.c @@ -1284,7 +1284,7 @@ static inline int bfin_spi_stop_queue(struct bfin_spi_master_data *drv_data) * friends on every SPI message. Do this instead */ drv_data->running = false; - while (!list_empty(&drv_data->queue) && drv_data->busy && limit--) { + while ((!list_empty(&drv_data->queue) || drv_data->busy) && limit--) { spin_unlock_irqrestore(&drv_data->lock, flags); msleep(10); spin_lock_irqsave(&drv_data->lock, flags); -- cgit v0.10.2 From 2845fd858c55c8a05674a071384a12a19cc17dbf Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Tue, 5 Apr 2011 08:55:05 -0700 Subject: iwlagn: override 5300 EEPROM # of chains At least EEPROM version 0x11A has the wrong number of chains programmed into it for some reason, so we need to override in the driver. Signed-off-by: Johannes Berg Signed-off-by: Wey-Yi Guy Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/iwlwifi/iwl-5000.c b/drivers/net/wireless/iwlwifi/iwl-5000.c index 3ea31b6..22e045b 100644 --- a/drivers/net/wireless/iwlwifi/iwl-5000.c +++ b/drivers/net/wireless/iwlwifi/iwl-5000.c @@ -530,6 +530,9 @@ static struct iwl_ht_params iwl5000_ht_params = { struct iwl_cfg iwl5300_agn_cfg = { .name = "Intel(R) Ultimate N WiFi Link 5300 AGN", IWL_DEVICE_5000, + /* at least EEPROM 0x11A has wrong info */ + .valid_tx_ant = ANT_ABC, /* .cfg overwrite */ + .valid_rx_ant = ANT_ABC, /* .cfg overwrite */ .ht_params = &iwl5000_ht_params, }; -- cgit v0.10.2 From bf3ca7f752d8f5009c9a83db56035566f3e313de Mon Sep 17 00:00:00 2001 From: Brian Cavagnolo Date: Wed, 6 Apr 2011 14:18:46 +0530 Subject: mwl8k: do not free unrequested irq When the mwl8k driver attempts and fails to switch from sta to ap firmware (or vice-versa) in the mwl8k_add_interface routine, the mwl8k_stop routine will be called. This routine must not attempt to free the irq if it was not requested. Signed-off-by: Brian Cavagnolo Signed-off-by: Nishant Sarmukadam Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/mwl8k.c b/drivers/net/wireless/mwl8k.c index 3695227..c1ceb4b 100644 --- a/drivers/net/wireless/mwl8k.c +++ b/drivers/net/wireless/mwl8k.c @@ -137,6 +137,7 @@ struct mwl8k_tx_queue { struct mwl8k_priv { struct ieee80211_hw *hw; struct pci_dev *pdev; + int irq; struct mwl8k_device_info *device_info; @@ -3761,9 +3762,11 @@ static int mwl8k_start(struct ieee80211_hw *hw) rc = request_irq(priv->pdev->irq, mwl8k_interrupt, IRQF_SHARED, MWL8K_NAME, hw); if (rc) { + priv->irq = -1; wiphy_err(hw->wiphy, "failed to register IRQ handler\n"); return -EIO; } + priv->irq = priv->pdev->irq; /* Enable TX reclaim and RX tasklets. */ tasklet_enable(&priv->poll_tx_task); @@ -3800,6 +3803,7 @@ static int mwl8k_start(struct ieee80211_hw *hw) if (rc) { iowrite32(0, priv->regs + MWL8K_HIU_A2H_INTERRUPT_MASK); free_irq(priv->pdev->irq, hw); + priv->irq = -1; tasklet_disable(&priv->poll_tx_task); tasklet_disable(&priv->poll_rx_task); } @@ -3818,7 +3822,10 @@ static void mwl8k_stop(struct ieee80211_hw *hw) /* Disable interrupts */ iowrite32(0, priv->regs + MWL8K_HIU_A2H_INTERRUPT_MASK); - free_irq(priv->pdev->irq, hw); + if (priv->irq != -1) { + free_irq(priv->pdev->irq, hw); + priv->irq = -1; + } /* Stop finalize join worker */ cancel_work_sync(&priv->finalize_join_worker); -- cgit v0.10.2 From bd39a274fb7b43374c797bafdb7f506598f36f77 Mon Sep 17 00:00:00 2001 From: Christian Lamparter Date: Wed, 6 Apr 2011 20:40:31 +0200 Subject: ath: add missing regdomain pair 0x5c mapping Joe Culler reported a problem with his AR9170 device: > ath: EEPROM regdomain: 0x5c > ath: EEPROM indicates we should expect a direct regpair map > ath: invalid regulatory domain/country code 0x5c > ath: Invalid EEPROM contents It turned out that the regdomain 'APL7_FCCA' was not mapped yet. According to Luis R. Rodriguez [Atheros' engineer] APL7 maps to FCC_CTL and FCCA maps to FCC_CTL as well, so the attached patch should be correct. Cc: Reported-by: Joe Culler Acked-by: Luis R. Rodriguez Signed-off-by: Christian Lamparter Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/ath/regd_common.h b/drivers/net/wireless/ath/regd_common.h index 248c670..5c2cfe6 100644 --- a/drivers/net/wireless/ath/regd_common.h +++ b/drivers/net/wireless/ath/regd_common.h @@ -195,6 +195,7 @@ static struct reg_dmn_pair_mapping regDomainPairs[] = { {APL9_WORLD, CTL_ETSI, CTL_ETSI}, {APL3_FCCA, CTL_FCC, CTL_FCC}, + {APL7_FCCA, CTL_FCC, CTL_FCC}, {APL1_ETSIC, CTL_FCC, CTL_ETSI}, {APL2_ETSIC, CTL_FCC, CTL_ETSI}, {APL2_APLD, CTL_FCC, NO_CTL}, -- cgit v0.10.2 From 37adb89fadd65ce47f7e5cfd564938a76b351948 Mon Sep 17 00:00:00 2001 From: Bryan Schumaker Date: Thu, 7 Apr 2011 16:02:20 -0400 Subject: NFS: Change initial mount authflavor only when server returns NFS4ERR_WRONGSEC When attempting an initial mount, we should only attempt other authflavors if AUTH_UNIX receives a NFS4ERR_WRONGSEC error. This allows other errors to be passed back to userspace programs. Signed-off-by: Bryan Schumaker Signed-off-by: Trond Myklebust diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index dfd1e6d..9bf41ea 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -2204,8 +2204,6 @@ static int nfs4_lookup_root_sec(struct nfs_server *server, struct nfs_fh *fhandl goto out; } ret = nfs4_lookup_root(server, fhandle, info); - if (ret < 0) - ret = -EAGAIN; out: return ret; } @@ -2226,7 +2224,7 @@ static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle, for (i = 0; i < len; i++) { status = nfs4_lookup_root_sec(server, fhandle, info, flav_array[i]); - if (status == 0) + if (status != -EPERM) break; } if (status == 0) -- cgit v0.10.2 From 1b86a58f9d7ce4fe2377687f378fbfb53bdc9b6c Mon Sep 17 00:00:00 2001 From: OGAWA Hirofumi Date: Thu, 7 Apr 2011 14:04:08 -0700 Subject: ipv4: Fix "Set rt->rt_iif more sanely on output routes." Commit 1018b5c01636c7c6bda31a719bda34fc631db29a ("Set rt->rt_iif more sanely on output routes.") breaks rt_is_{output,input}_route. This became the cause to return "IP_PKTINFO's ->ipi_ifindex == 0". To fix it, this does: 1) Add "int rt_route_iif;" to struct rtable 2) For input routes, always set rt_route_iif to same value as rt_iif 3) For output routes, always set rt_route_iif to zero. Set rt_iif as it is done currently. 4) Change rt_is_{output,input}_route() to test rt_route_iif Signed-off-by: OGAWA Hirofumi Signed-off-by: David S. Miller diff --git a/include/net/route.h b/include/net/route.h index f88429c..8fce062 100644 --- a/include/net/route.h +++ b/include/net/route.h @@ -64,6 +64,7 @@ struct rtable { __be32 rt_dst; /* Path destination */ __be32 rt_src; /* Path source */ + int rt_route_iif; int rt_iif; int rt_oif; __u32 rt_mark; @@ -80,12 +81,12 @@ struct rtable { static inline bool rt_is_input_route(struct rtable *rt) { - return rt->rt_iif != 0; + return rt->rt_route_iif != 0; } static inline bool rt_is_output_route(struct rtable *rt) { - return rt->rt_iif == 0; + return rt->rt_route_iif == 0; } struct ip_rt_acct { diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 4b0c811..1628be5 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -1891,6 +1891,7 @@ static int ip_route_input_mc(struct sk_buff *skb, __be32 daddr, __be32 saddr, #ifdef CONFIG_IP_ROUTE_CLASSID rth->dst.tclassid = itag; #endif + rth->rt_route_iif = dev->ifindex; rth->rt_iif = dev->ifindex; rth->dst.dev = init_net.loopback_dev; dev_hold(rth->dst.dev); @@ -2026,6 +2027,7 @@ static int __mkroute_input(struct sk_buff *skb, rth->rt_key_src = saddr; rth->rt_src = saddr; rth->rt_gateway = daddr; + rth->rt_route_iif = in_dev->dev->ifindex; rth->rt_iif = in_dev->dev->ifindex; rth->dst.dev = (out_dev)->dev; dev_hold(rth->dst.dev); @@ -2202,6 +2204,7 @@ local_input: #ifdef CONFIG_IP_ROUTE_CLASSID rth->dst.tclassid = itag; #endif + rth->rt_route_iif = dev->ifindex; rth->rt_iif = dev->ifindex; rth->dst.dev = net->loopback_dev; dev_hold(rth->dst.dev); @@ -2401,7 +2404,8 @@ static struct rtable *__mkroute_output(const struct fib_result *res, rth->rt_mark = oldflp4->flowi4_mark; rth->rt_dst = fl4->daddr; rth->rt_src = fl4->saddr; - rth->rt_iif = 0; + rth->rt_route_iif = 0; + rth->rt_iif = oldflp4->flowi4_oif ? : dev_out->ifindex; /* get references to the devices that are to be hold by the routing cache entry */ rth->dst.dev = dev_out; @@ -2716,6 +2720,7 @@ struct dst_entry *ipv4_blackhole_route(struct net *net, struct dst_entry *dst_or rt->rt_key_dst = ort->rt_key_dst; rt->rt_key_src = ort->rt_key_src; rt->rt_tos = ort->rt_tos; + rt->rt_route_iif = ort->rt_route_iif; rt->rt_iif = ort->rt_iif; rt->rt_oif = ort->rt_oif; rt->rt_mark = ort->rt_mark; @@ -2725,7 +2730,6 @@ struct dst_entry *ipv4_blackhole_route(struct net *net, struct dst_entry *dst_or rt->rt_type = ort->rt_type; rt->rt_dst = ort->rt_dst; rt->rt_src = ort->rt_src; - rt->rt_iif = ort->rt_iif; rt->rt_gateway = ort->rt_gateway; rt->rt_spec_dst = ort->rt_spec_dst; rt->peer = ort->peer; diff --git a/net/ipv4/xfrm4_policy.c b/net/ipv4/xfrm4_policy.c index 13e0e7f..d20a05e 100644 --- a/net/ipv4/xfrm4_policy.c +++ b/net/ipv4/xfrm4_policy.c @@ -74,6 +74,7 @@ static int xfrm4_fill_dst(struct xfrm_dst *xdst, struct net_device *dev, rt->rt_key_dst = fl4->daddr; rt->rt_key_src = fl4->saddr; rt->rt_tos = fl4->flowi4_tos; + rt->rt_route_iif = fl4->flowi4_iif; rt->rt_iif = fl4->flowi4_iif; rt->rt_oif = fl4->flowi4_oif; rt->rt_mark = fl4->flowi4_mark; -- cgit v0.10.2 From e828776a8abe6b9bae7ed9638710bff7642c568a Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Fri, 8 Apr 2011 12:45:07 +1000 Subject: xfs: fix extent format buffer allocation size When formatting an inode item, we have to allocate a separate buffer to hold extents when there are delayed allocation extents on the inode and it is in extent format. The allocation size is derived from the in-core data fork representation, which accounts for delayed allocation extents, while the on-disk representation does not contain any delalloc extents. As a result of this mismatch, the allocated buffer can be far larger than needed to hold the real extent list which, due to the fact the inode is in extent format, is limited to the size of the literal area of the inode. However, we can have thousands of delalloc extents, resulting in an allocation size orders of magnitude larger than is needed to hold all the real extents. Fix this by limiting the size of the buffer being allocated to the size of the literal area of the inodes in the filesystem (i.e. the maximum size an inode fork can grow to). Signed-off-by: Dave Chinner Reviewed-by: Alex Elder diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c index 46cc401..576fdfe 100644 --- a/fs/xfs/xfs_inode_item.c +++ b/fs/xfs/xfs_inode_item.c @@ -198,6 +198,41 @@ xfs_inode_item_size( } /* + * xfs_inode_item_format_extents - convert in-core extents to on-disk form + * + * For either the data or attr fork in extent format, we need to endian convert + * the in-core extent as we place them into the on-disk inode. In this case, we + * need to do this conversion before we write the extents into the log. Because + * we don't have the disk inode to write into here, we allocate a buffer and + * format the extents into it via xfs_iextents_copy(). We free the buffer in + * the unlock routine after the copy for the log has been made. + * + * In the case of the data fork, the in-core and on-disk fork sizes can be + * different due to delayed allocation extents. We only log on-disk extents + * here, so always use the physical fork size to determine the size of the + * buffer we need to allocate. + */ +STATIC void +xfs_inode_item_format_extents( + struct xfs_inode *ip, + struct xfs_log_iovec *vecp, + int whichfork, + int type) +{ + xfs_bmbt_rec_t *ext_buffer; + + ext_buffer = kmem_alloc(XFS_IFORK_SIZE(ip, whichfork), KM_SLEEP); + if (whichfork == XFS_DATA_FORK) + ip->i_itemp->ili_extents_buf = ext_buffer; + else + ip->i_itemp->ili_aextents_buf = ext_buffer; + + vecp->i_addr = ext_buffer; + vecp->i_len = xfs_iextents_copy(ip, ext_buffer, whichfork); + vecp->i_type = type; +} + +/* * This is called to fill in the vector of log iovecs for the * given inode log item. It fills the first item with an inode * log format structure, the second with the on-disk inode structure, @@ -213,7 +248,6 @@ xfs_inode_item_format( struct xfs_inode *ip = iip->ili_inode; uint nvecs; size_t data_bytes; - xfs_bmbt_rec_t *ext_buffer; xfs_mount_t *mp; vecp->i_addr = &iip->ili_format; @@ -320,22 +354,8 @@ xfs_inode_item_format( } else #endif { - /* - * There are delayed allocation extents - * in the inode, or we need to convert - * the extents to on disk format. - * Use xfs_iextents_copy() - * to copy only the real extents into - * a separate buffer. We'll free the - * buffer in the unlock routine. - */ - ext_buffer = kmem_alloc(ip->i_df.if_bytes, - KM_SLEEP); - iip->ili_extents_buf = ext_buffer; - vecp->i_addr = ext_buffer; - vecp->i_len = xfs_iextents_copy(ip, ext_buffer, - XFS_DATA_FORK); - vecp->i_type = XLOG_REG_TYPE_IEXT; + xfs_inode_item_format_extents(ip, vecp, + XFS_DATA_FORK, XLOG_REG_TYPE_IEXT); } ASSERT(vecp->i_len <= ip->i_df.if_bytes); iip->ili_format.ilf_dsize = vecp->i_len; @@ -445,19 +465,12 @@ xfs_inode_item_format( */ vecp->i_addr = ip->i_afp->if_u1.if_extents; vecp->i_len = ip->i_afp->if_bytes; + vecp->i_type = XLOG_REG_TYPE_IATTR_EXT; #else ASSERT(iip->ili_aextents_buf == NULL); - /* - * Need to endian flip before logging - */ - ext_buffer = kmem_alloc(ip->i_afp->if_bytes, - KM_SLEEP); - iip->ili_aextents_buf = ext_buffer; - vecp->i_addr = ext_buffer; - vecp->i_len = xfs_iextents_copy(ip, ext_buffer, - XFS_ATTR_FORK); + xfs_inode_item_format_extents(ip, vecp, + XFS_ATTR_FORK, XLOG_REG_TYPE_IATTR_EXT); #endif - vecp->i_type = XLOG_REG_TYPE_IATTR_EXT; iip->ili_format.ilf_asize = vecp->i_len; vecp++; nvecs++; -- cgit v0.10.2 From c6d09b666de11eb272326a6eb6cd3246da571014 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Fri, 8 Apr 2011 12:45:07 +1000 Subject: xfs: introduce a xfssyncd workqueue All of the work xfssyncd does is background functionality. There is no need for a thread per filesystem to do this work - it can al be managed by a global workqueue now they manage concurrency effectively. Introduce a new gglobal xfssyncd workqueue, and convert the periodic work to use this new functionality. To do this, use a delayed work construct to schedule the next running of the periodic sync work for the filesystem. When the sync work is complete, queue a new delayed work for the next running of the sync work. For laptop mode, we wait on completion for the sync works, so ensure that the sync work queuing interface can flush and wait for work to complete to enable the work queue infrastructure to replace the current sequence number and wakeup that is used. Because the sync work does non-trivial amounts of work, mark the new work queue as CPU intensive. Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Reviewed-by: Alex Elder diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 1ba5c45..c71b6ed 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -1191,22 +1191,12 @@ xfs_fs_sync_fs( return -error; if (laptop_mode) { - int prev_sync_seq = mp->m_sync_seq; - /* * The disk must be active because we're syncing. * We schedule xfssyncd now (now that the disk is * active) instead of later (when it might not be). */ - wake_up_process(mp->m_sync_task); - /* - * We have to wait for the sync iteration to complete. - * If we don't, the disk activity caused by the sync - * will come after the sync is completed, and that - * triggers another sync from laptop mode. - */ - wait_event(mp->m_wait_single_sync_task, - mp->m_sync_seq != prev_sync_seq); + flush_delayed_work_sync(&mp->m_sync_work); } return 0; @@ -1492,7 +1482,6 @@ xfs_fs_fill_super( atomic_set(&mp->m_active_trans, 0); INIT_LIST_HEAD(&mp->m_sync_list); spin_lock_init(&mp->m_sync_lock); - init_waitqueue_head(&mp->m_wait_single_sync_task); mp->m_super = sb; sb->s_fs_info = mp; @@ -1833,13 +1822,27 @@ init_xfs_fs(void) if (error) goto out_cleanup_procfs; + /* + * max_active is set to 8 to give enough concurency to allow + * multiple work operations on each CPU to run. This allows multiple + * filesystems to be running sync work concurrently, and scales with + * the number of CPUs in the system. + */ + xfs_syncd_wq = alloc_workqueue("xfssyncd", WQ_CPU_INTENSIVE, 8); + if (!xfs_syncd_wq) { + error = -ENOMEM; + goto out_sysctl_unregister; + } + vfs_initquota(); error = register_filesystem(&xfs_fs_type); if (error) - goto out_sysctl_unregister; + goto out_destroy_xfs_syncd; return 0; + out_destroy_xfs_syncd: + destroy_workqueue(xfs_syncd_wq); out_sysctl_unregister: xfs_sysctl_unregister(); out_cleanup_procfs: @@ -1861,6 +1864,7 @@ exit_xfs_fs(void) { vfs_exitquota(); unregister_filesystem(&xfs_fs_type); + destroy_workqueue(xfs_syncd_wq); xfs_sysctl_unregister(); xfs_cleanup_procfs(); xfs_buf_terminate(); diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c index 594cd82..4a582d8 100644 --- a/fs/xfs/linux-2.6/xfs_sync.c +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -39,6 +39,8 @@ #include #include +struct workqueue_struct *xfs_syncd_wq; /* sync workqueue */ + /* * The inode lookup is done in batches to keep the amount of lock traffic and * radix tree lookups to a minimum. The batch size is a trade off between @@ -489,32 +491,6 @@ xfs_flush_inodes( xfs_log_force(ip->i_mount, XFS_LOG_SYNC); } -/* - * Every sync period we need to unpin all items, reclaim inodes and sync - * disk quotas. We might need to cover the log to indicate that the - * filesystem is idle and not frozen. - */ -STATIC void -xfs_sync_worker( - struct xfs_mount *mp, - void *unused) -{ - int error; - - if (!(mp->m_flags & XFS_MOUNT_RDONLY)) { - /* dgc: errors ignored here */ - if (mp->m_super->s_frozen == SB_UNFROZEN && - xfs_log_need_covered(mp)) - error = xfs_fs_log_dummy(mp); - else - xfs_log_force(mp, 0); - xfs_reclaim_inodes(mp, 0); - error = xfs_qm_sync(mp, SYNC_TRYLOCK); - } - mp->m_sync_seq++; - wake_up(&mp->m_wait_single_sync_task); -} - STATIC int xfssyncd( void *arg) @@ -528,34 +504,19 @@ xfssyncd( timeleft = xfs_syncd_centisecs * msecs_to_jiffies(10); for (;;) { if (list_empty(&mp->m_sync_list)) - timeleft = schedule_timeout_interruptible(timeleft); + schedule_timeout_interruptible(timeleft); /* swsusp */ try_to_freeze(); if (kthread_should_stop() && list_empty(&mp->m_sync_list)) break; spin_lock(&mp->m_sync_lock); - /* - * We can get woken by laptop mode, to do a sync - - * that's the (only!) case where the list would be - * empty with time remaining. - */ - if (!timeleft || list_empty(&mp->m_sync_list)) { - if (!timeleft) - timeleft = xfs_syncd_centisecs * - msecs_to_jiffies(10); - INIT_LIST_HEAD(&mp->m_sync_work.w_list); - list_add_tail(&mp->m_sync_work.w_list, - &mp->m_sync_list); - } list_splice_init(&mp->m_sync_list, &tmp); spin_unlock(&mp->m_sync_lock); list_for_each_entry_safe(work, n, &tmp, w_list) { (*work->w_syncer)(mp, work->w_data); list_del(&work->w_list); - if (work == &mp->m_sync_work) - continue; if (work->w_completion) complete(work->w_completion); kmem_free(work); @@ -565,13 +526,49 @@ xfssyncd( return 0; } +static void +xfs_syncd_queue_sync( + struct xfs_mount *mp) +{ + queue_delayed_work(xfs_syncd_wq, &mp->m_sync_work, + msecs_to_jiffies(xfs_syncd_centisecs * 10)); +} + +/* + * Every sync period we need to unpin all items, reclaim inodes and sync + * disk quotas. We might need to cover the log to indicate that the + * filesystem is idle and not frozen. + */ +STATIC void +xfs_sync_worker( + struct work_struct *work) +{ + struct xfs_mount *mp = container_of(to_delayed_work(work), + struct xfs_mount, m_sync_work); + int error; + + if (!(mp->m_flags & XFS_MOUNT_RDONLY)) { + /* dgc: errors ignored here */ + if (mp->m_super->s_frozen == SB_UNFROZEN && + xfs_log_need_covered(mp)) + error = xfs_fs_log_dummy(mp); + else + xfs_log_force(mp, 0); + xfs_reclaim_inodes(mp, 0); + error = xfs_qm_sync(mp, SYNC_TRYLOCK); + } + + /* queue us up again */ + xfs_syncd_queue_sync(mp); +} + int xfs_syncd_init( struct xfs_mount *mp) { - mp->m_sync_work.w_syncer = xfs_sync_worker; - mp->m_sync_work.w_mount = mp; - mp->m_sync_work.w_completion = NULL; + INIT_DELAYED_WORK(&mp->m_sync_work, xfs_sync_worker); + xfs_syncd_queue_sync(mp); + mp->m_sync_task = kthread_run(xfssyncd, mp, "xfssyncd/%s", mp->m_fsname); if (IS_ERR(mp->m_sync_task)) return -PTR_ERR(mp->m_sync_task); @@ -582,6 +579,7 @@ void xfs_syncd_stop( struct xfs_mount *mp) { + cancel_delayed_work_sync(&mp->m_sync_work); kthread_stop(mp->m_sync_task); } diff --git a/fs/xfs/linux-2.6/xfs_sync.h b/fs/xfs/linux-2.6/xfs_sync.h index 32ba662..e3a6ad2 100644 --- a/fs/xfs/linux-2.6/xfs_sync.h +++ b/fs/xfs/linux-2.6/xfs_sync.h @@ -32,6 +32,8 @@ typedef struct xfs_sync_work { #define SYNC_WAIT 0x0001 /* wait for i/o to complete */ #define SYNC_TRYLOCK 0x0002 /* only try to lock inodes */ +extern struct workqueue_struct *xfs_syncd_wq; /* sync workqueue */ + int xfs_syncd_init(struct xfs_mount *mp); void xfs_syncd_stop(struct xfs_mount *mp); diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index a62e897..2c11e62 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -203,12 +203,10 @@ typedef struct xfs_mount { struct mutex m_icsb_mutex; /* balancer sync lock */ #endif struct xfs_mru_cache *m_filestream; /* per-mount filestream data */ + struct delayed_work m_sync_work; /* background sync work */ struct task_struct *m_sync_task; /* generalised sync thread */ - xfs_sync_work_t m_sync_work; /* work item for VFS_SYNC */ struct list_head m_sync_list; /* sync thread work item list */ spinlock_t m_sync_lock; /* work item list lock */ - int m_sync_seq; /* sync thread generation no. */ - wait_queue_head_t m_wait_single_sync_task; __int64_t m_update_flags; /* sb flags we need to update on the next remount,rw */ struct shrinker m_inode_shrink; /* inode reclaim shrinker */ -- cgit v0.10.2 From 89e4cb550a492cfca038a555fcc1bdac58822ec3 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Fri, 8 Apr 2011 12:45:07 +1000 Subject: xfs: convert ENOSPC inode flushing to use new syncd workqueue On of the problems with the current inode flush at ENOSPC is that we queue a flush per ENOSPC event, regardless of how many are already queued. Thi can result in hundreds of queued flushes, most of which simply burn CPU scanned and do no real work. This simply slows down allocation at ENOSPC. We really only need one active flush at a time, and we can easily implement that via the new xfs_syncd_wq. All we need to do is queue a flush if one is not already active, then block waiting for the currently active flush to complete. The result is that we only ever have a single ENOSPC inode flush active at a time and this greatly reduces the overhead of ENOSPC processing. On my 2p test machine, this results in tests exercising ENOSPC conditions running significantly faster - 042 halves execution time, 083 drops from 60s to 5s, etc - while not introducing test regressions. This allows us to remove the old xfssyncd threads and infrastructure as they are no longer used. Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Reviewed-by: Alex Elder diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index c71b6ed..ee0e981 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -1480,8 +1480,6 @@ xfs_fs_fill_super( spin_lock_init(&mp->m_sb_lock); mutex_init(&mp->m_growlock); atomic_set(&mp->m_active_trans, 0); - INIT_LIST_HEAD(&mp->m_sync_list); - spin_lock_init(&mp->m_sync_lock); mp->m_super = sb; sb->s_fs_info = mp; diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c index 4a582d8..af32759 100644 --- a/fs/xfs/linux-2.6/xfs_sync.c +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -433,99 +433,6 @@ xfs_quiesce_attr( xfs_unmountfs_writesb(mp); } -/* - * Enqueue a work item to be picked up by the vfs xfssyncd thread. - * Doing this has two advantages: - * - It saves on stack space, which is tight in certain situations - * - It can be used (with care) as a mechanism to avoid deadlocks. - * Flushing while allocating in a full filesystem requires both. - */ -STATIC void -xfs_syncd_queue_work( - struct xfs_mount *mp, - void *data, - void (*syncer)(struct xfs_mount *, void *), - struct completion *completion) -{ - struct xfs_sync_work *work; - - work = kmem_alloc(sizeof(struct xfs_sync_work), KM_SLEEP); - INIT_LIST_HEAD(&work->w_list); - work->w_syncer = syncer; - work->w_data = data; - work->w_mount = mp; - work->w_completion = completion; - spin_lock(&mp->m_sync_lock); - list_add_tail(&work->w_list, &mp->m_sync_list); - spin_unlock(&mp->m_sync_lock); - wake_up_process(mp->m_sync_task); -} - -/* - * Flush delayed allocate data, attempting to free up reserved space - * from existing allocations. At this point a new allocation attempt - * has failed with ENOSPC and we are in the process of scratching our - * heads, looking about for more room... - */ -STATIC void -xfs_flush_inodes_work( - struct xfs_mount *mp, - void *arg) -{ - struct inode *inode = arg; - xfs_sync_data(mp, SYNC_TRYLOCK); - xfs_sync_data(mp, SYNC_TRYLOCK | SYNC_WAIT); - iput(inode); -} - -void -xfs_flush_inodes( - xfs_inode_t *ip) -{ - struct inode *inode = VFS_I(ip); - DECLARE_COMPLETION_ONSTACK(completion); - - igrab(inode); - xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_inodes_work, &completion); - wait_for_completion(&completion); - xfs_log_force(ip->i_mount, XFS_LOG_SYNC); -} - -STATIC int -xfssyncd( - void *arg) -{ - struct xfs_mount *mp = arg; - long timeleft; - xfs_sync_work_t *work, *n; - LIST_HEAD (tmp); - - set_freezable(); - timeleft = xfs_syncd_centisecs * msecs_to_jiffies(10); - for (;;) { - if (list_empty(&mp->m_sync_list)) - schedule_timeout_interruptible(timeleft); - /* swsusp */ - try_to_freeze(); - if (kthread_should_stop() && list_empty(&mp->m_sync_list)) - break; - - spin_lock(&mp->m_sync_lock); - list_splice_init(&mp->m_sync_list, &tmp); - spin_unlock(&mp->m_sync_lock); - - list_for_each_entry_safe(work, n, &tmp, w_list) { - (*work->w_syncer)(mp, work->w_data); - list_del(&work->w_list); - if (work->w_completion) - complete(work->w_completion); - kmem_free(work); - } - } - - return 0; -} - static void xfs_syncd_queue_sync( struct xfs_mount *mp) @@ -562,16 +469,47 @@ xfs_sync_worker( xfs_syncd_queue_sync(mp); } +/* + * Flush delayed allocate data, attempting to free up reserved space + * from existing allocations. At this point a new allocation attempt + * has failed with ENOSPC and we are in the process of scratching our + * heads, looking about for more room. + * + * Queue a new data flush if there isn't one already in progress and + * wait for completion of the flush. This means that we only ever have one + * inode flush in progress no matter how many ENOSPC events are occurring and + * so will prevent the system from bogging down due to every concurrent + * ENOSPC event scanning all the active inodes in the system for writeback. + */ +void +xfs_flush_inodes( + struct xfs_inode *ip) +{ + struct xfs_mount *mp = ip->i_mount; + + queue_work(xfs_syncd_wq, &mp->m_flush_work); + flush_work_sync(&mp->m_flush_work); +} + +STATIC void +xfs_flush_worker( + struct work_struct *work) +{ + struct xfs_mount *mp = container_of(work, + struct xfs_mount, m_flush_work); + + xfs_sync_data(mp, SYNC_TRYLOCK); + xfs_sync_data(mp, SYNC_TRYLOCK | SYNC_WAIT); +} + int xfs_syncd_init( struct xfs_mount *mp) { + INIT_WORK(&mp->m_flush_work, xfs_flush_worker); INIT_DELAYED_WORK(&mp->m_sync_work, xfs_sync_worker); xfs_syncd_queue_sync(mp); - mp->m_sync_task = kthread_run(xfssyncd, mp, "xfssyncd/%s", mp->m_fsname); - if (IS_ERR(mp->m_sync_task)) - return -PTR_ERR(mp->m_sync_task); return 0; } @@ -580,7 +518,7 @@ xfs_syncd_stop( struct xfs_mount *mp) { cancel_delayed_work_sync(&mp->m_sync_work); - kthread_stop(mp->m_sync_task); + cancel_work_sync(&mp->m_flush_work); } void diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index 2c11e62..a0ad90e 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -204,9 +204,7 @@ typedef struct xfs_mount { #endif struct xfs_mru_cache *m_filestream; /* per-mount filestream data */ struct delayed_work m_sync_work; /* background sync work */ - struct task_struct *m_sync_task; /* generalised sync thread */ - struct list_head m_sync_list; /* sync thread work item list */ - spinlock_t m_sync_lock; /* work item list lock */ + struct work_struct m_flush_work; /* background inode flush */ __int64_t m_update_flags; /* sb flags we need to update on the next remount,rw */ struct shrinker m_inode_shrink; /* inode reclaim shrinker */ -- cgit v0.10.2 From a7b339f1b8698667eada006e717cdb4523be2ed5 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Fri, 8 Apr 2011 12:45:07 +1000 Subject: xfs: introduce background inode reclaim work Background inode reclaim needs to run more frequently that the XFS syncd work is run as 30s is too long between optimal reclaim runs. Add a new periodic work item to the xfs syncd workqueue to run a fast, non-blocking inode reclaim scan. Background inode reclaim is kicked by the act of marking inodes for reclaim. When an AG is first marked as having reclaimable inodes, the background reclaim work is kicked. It will continue to run periodically untill it detects that there are no more reclaimable inodes. It will be kicked again when the first inode is queued for reclaim. To ensure shrinker based inode reclaim throttles to the inode cleaning and reclaim rate but still reclaim inodes efficiently, make it kick the background inode reclaim so that when we are low on memory we are trying to reclaim inodes as efficiently as possible. This kick shoul d not be necessary, but it will protect against failures to kick the background reclaim when inodes are first dirtied. To provide the rate throttling, make the shrinker pass do synchronous inode reclaim so that it blocks on inodes under IO. This means that the shrinker will reclaim inodes rather than just skipping over them, but it does not adversely affect the rate of reclaim because most dirty inodes are already under IO due to the background reclaim work the shrinker kicked. These two modifications solve one of the two OOM killer invocations Chris Mason reported recently when running a stress testing script. The particular workload trigger for the OOM killer invocation is where there are more threads than CPUs all unlinking files in an extremely memory constrained environment. Unlike other solutions, this one does not have a performance impact on performance when memory is not constrained or the number of concurrent threads operating is <= to the number of CPUs. Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Reviewed-by: Alex Elder diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c index af32759..debe282 100644 --- a/fs/xfs/linux-2.6/xfs_sync.c +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -461,7 +461,6 @@ xfs_sync_worker( error = xfs_fs_log_dummy(mp); else xfs_log_force(mp, 0); - xfs_reclaim_inodes(mp, 0); error = xfs_qm_sync(mp, SYNC_TRYLOCK); } @@ -470,6 +469,52 @@ xfs_sync_worker( } /* + * Queue a new inode reclaim pass if there are reclaimable inodes and there + * isn't a reclaim pass already in progress. By default it runs every 5s based + * on the xfs syncd work default of 30s. Perhaps this should have it's own + * tunable, but that can be done if this method proves to be ineffective or too + * aggressive. + */ +static void +xfs_syncd_queue_reclaim( + struct xfs_mount *mp) +{ + + /* + * We can have inodes enter reclaim after we've shut down the syncd + * workqueue during unmount, so don't allow reclaim work to be queued + * during unmount. + */ + if (!(mp->m_super->s_flags & MS_ACTIVE)) + return; + + rcu_read_lock(); + if (radix_tree_tagged(&mp->m_perag_tree, XFS_ICI_RECLAIM_TAG)) { + queue_delayed_work(xfs_syncd_wq, &mp->m_reclaim_work, + msecs_to_jiffies(xfs_syncd_centisecs / 6 * 10)); + } + rcu_read_unlock(); +} + +/* + * This is a fast pass over the inode cache to try to get reclaim moving on as + * many inodes as possible in a short period of time. It kicks itself every few + * seconds, as well as being kicked by the inode cache shrinker when memory + * goes low. It scans as quickly as possible avoiding locked inodes or those + * already being flushed, and once done schedules a future pass. + */ +STATIC void +xfs_reclaim_worker( + struct work_struct *work) +{ + struct xfs_mount *mp = container_of(to_delayed_work(work), + struct xfs_mount, m_reclaim_work); + + xfs_reclaim_inodes(mp, SYNC_TRYLOCK); + xfs_syncd_queue_reclaim(mp); +} + +/* * Flush delayed allocate data, attempting to free up reserved space * from existing allocations. At this point a new allocation attempt * has failed with ENOSPC and we are in the process of scratching our @@ -508,7 +553,10 @@ xfs_syncd_init( { INIT_WORK(&mp->m_flush_work, xfs_flush_worker); INIT_DELAYED_WORK(&mp->m_sync_work, xfs_sync_worker); + INIT_DELAYED_WORK(&mp->m_reclaim_work, xfs_reclaim_worker); + xfs_syncd_queue_sync(mp); + xfs_syncd_queue_reclaim(mp); return 0; } @@ -518,6 +566,7 @@ xfs_syncd_stop( struct xfs_mount *mp) { cancel_delayed_work_sync(&mp->m_sync_work); + cancel_delayed_work_sync(&mp->m_reclaim_work); cancel_work_sync(&mp->m_flush_work); } @@ -537,6 +586,10 @@ __xfs_inode_set_reclaim_tag( XFS_INO_TO_AGNO(ip->i_mount, ip->i_ino), XFS_ICI_RECLAIM_TAG); spin_unlock(&ip->i_mount->m_perag_lock); + + /* schedule periodic background inode reclaim */ + xfs_syncd_queue_reclaim(ip->i_mount); + trace_xfs_perag_set_reclaim(ip->i_mount, pag->pag_agno, -1, _RET_IP_); } @@ -953,7 +1006,13 @@ xfs_reclaim_inodes( } /* - * Shrinker infrastructure. + * Inode cache shrinker. + * + * When called we make sure that there is a background (fast) inode reclaim in + * progress, while we will throttle the speed of reclaim via doiing synchronous + * reclaim of inodes. That means if we come across dirty inodes, we wait for + * them to be cleaned, which we hope will not be very long due to the + * background walker having already kicked the IO off on those dirty inodes. */ static int xfs_reclaim_inode_shrink( @@ -968,10 +1027,14 @@ xfs_reclaim_inode_shrink( mp = container_of(shrink, struct xfs_mount, m_inode_shrink); if (nr_to_scan) { + /* kick background reclaimer */ + xfs_syncd_queue_reclaim(mp); + if (!(gfp_mask & __GFP_FS)) return -1; - xfs_reclaim_inodes_ag(mp, SYNC_TRYLOCK, &nr_to_scan); + xfs_reclaim_inodes_ag(mp, SYNC_TRYLOCK | SYNC_WAIT, + &nr_to_scan); /* terminate if we don't exhaust the scan */ if (nr_to_scan > 0) return -1; diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index a0ad90e..19af0ab 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -204,6 +204,7 @@ typedef struct xfs_mount { #endif struct xfs_mru_cache *m_filestream; /* per-mount filestream data */ struct delayed_work m_sync_work; /* background sync work */ + struct delayed_work m_reclaim_work; /* background inode reclaim */ struct work_struct m_flush_work; /* background inode flush */ __int64_t m_update_flags; /* sb flags we need to update on the next remount,rw */ -- cgit v0.10.2 From 0bf6a5bd4b55b466964ead6fa566d8f346a828ee Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Fri, 8 Apr 2011 12:45:07 +1000 Subject: xfs: convert the xfsaild threads to a workqueue Similar to the xfssyncd, the per-filesystem xfsaild threads can be converted to a global workqueue and run periodically by delayed works. This makes sense for the AIL pushing because it uses variable timeouts depending on the work that needs to be done. By removing the xfsaild, we simplify the AIL pushing code and remove the need to spread the code to implement the threading and pushing across multiple files. Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Reviewed-by: Alex Elder diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index ee0e981..67d5b2c 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -816,75 +816,6 @@ xfs_setup_devices( return 0; } -/* - * XFS AIL push thread support - */ -void -xfsaild_wakeup( - struct xfs_ail *ailp, - xfs_lsn_t threshold_lsn) -{ - /* only ever move the target forwards */ - if (XFS_LSN_CMP(threshold_lsn, ailp->xa_target) > 0) { - ailp->xa_target = threshold_lsn; - wake_up_process(ailp->xa_task); - } -} - -STATIC int -xfsaild( - void *data) -{ - struct xfs_ail *ailp = data; - xfs_lsn_t last_pushed_lsn = 0; - long tout = 0; /* milliseconds */ - - while (!kthread_should_stop()) { - /* - * for short sleeps indicating congestion, don't allow us to - * get woken early. Otherwise all we do is bang on the AIL lock - * without making progress. - */ - if (tout && tout <= 20) - __set_current_state(TASK_KILLABLE); - else - __set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(tout ? - msecs_to_jiffies(tout) : MAX_SCHEDULE_TIMEOUT); - - /* swsusp */ - try_to_freeze(); - - ASSERT(ailp->xa_mount->m_log); - if (XFS_FORCED_SHUTDOWN(ailp->xa_mount)) - continue; - - tout = xfsaild_push(ailp, &last_pushed_lsn); - } - - return 0; -} /* xfsaild */ - -int -xfsaild_start( - struct xfs_ail *ailp) -{ - ailp->xa_target = 0; - ailp->xa_task = kthread_run(xfsaild, ailp, "xfsaild/%s", - ailp->xa_mount->m_fsname); - if (IS_ERR(ailp->xa_task)) - return -PTR_ERR(ailp->xa_task); - return 0; -} - -void -xfsaild_stop( - struct xfs_ail *ailp) -{ - kthread_stop(ailp->xa_task); -} - - /* Catch misguided souls that try to use this interface on XFS */ STATIC struct inode * xfs_fs_alloc_inode( @@ -1786,6 +1717,38 @@ xfs_destroy_zones(void) } STATIC int __init +xfs_init_workqueues(void) +{ + /* + * max_active is set to 8 to give enough concurency to allow + * multiple work operations on each CPU to run. This allows multiple + * filesystems to be running sync work concurrently, and scales with + * the number of CPUs in the system. + */ + xfs_syncd_wq = alloc_workqueue("xfssyncd", WQ_CPU_INTENSIVE, 8); + if (!xfs_syncd_wq) + goto out; + + xfs_ail_wq = alloc_workqueue("xfsail", WQ_CPU_INTENSIVE, 8); + if (!xfs_ail_wq) + goto out_destroy_syncd; + + return 0; + +out_destroy_syncd: + destroy_workqueue(xfs_syncd_wq); +out: + return -ENOMEM; +} + +STATIC void __exit +xfs_destroy_workqueues(void) +{ + destroy_workqueue(xfs_ail_wq); + destroy_workqueue(xfs_syncd_wq); +} + +STATIC int __init init_xfs_fs(void) { int error; @@ -1800,10 +1763,14 @@ init_xfs_fs(void) if (error) goto out; - error = xfs_mru_cache_init(); + error = xfs_init_workqueues(); if (error) goto out_destroy_zones; + error = xfs_mru_cache_init(); + if (error) + goto out_destroy_wq; + error = xfs_filestream_init(); if (error) goto out_mru_cache_uninit; @@ -1820,27 +1787,17 @@ init_xfs_fs(void) if (error) goto out_cleanup_procfs; - /* - * max_active is set to 8 to give enough concurency to allow - * multiple work operations on each CPU to run. This allows multiple - * filesystems to be running sync work concurrently, and scales with - * the number of CPUs in the system. - */ - xfs_syncd_wq = alloc_workqueue("xfssyncd", WQ_CPU_INTENSIVE, 8); - if (!xfs_syncd_wq) { - error = -ENOMEM; + error = xfs_init_workqueues(); + if (error) goto out_sysctl_unregister; - } vfs_initquota(); error = register_filesystem(&xfs_fs_type); if (error) - goto out_destroy_xfs_syncd; + goto out_sysctl_unregister; return 0; - out_destroy_xfs_syncd: - destroy_workqueue(xfs_syncd_wq); out_sysctl_unregister: xfs_sysctl_unregister(); out_cleanup_procfs: @@ -1851,6 +1808,8 @@ init_xfs_fs(void) xfs_filestream_uninit(); out_mru_cache_uninit: xfs_mru_cache_uninit(); + out_destroy_wq: + xfs_destroy_workqueues(); out_destroy_zones: xfs_destroy_zones(); out: @@ -1862,12 +1821,12 @@ exit_xfs_fs(void) { vfs_exitquota(); unregister_filesystem(&xfs_fs_type); - destroy_workqueue(xfs_syncd_wq); xfs_sysctl_unregister(); xfs_cleanup_procfs(); xfs_buf_terminate(); xfs_filestream_uninit(); xfs_mru_cache_uninit(); + xfs_destroy_workqueues(); xfs_destroy_zones(); } diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c index 12aff95..cb3aeac 100644 --- a/fs/xfs/xfs_trans_ail.c +++ b/fs/xfs/xfs_trans_ail.c @@ -28,6 +28,8 @@ #include "xfs_trans_priv.h" #include "xfs_error.h" +struct workqueue_struct *xfs_ail_wq; /* AIL workqueue */ + STATIC void xfs_ail_splice(struct xfs_ail *, struct list_head *, xfs_lsn_t); STATIC void xfs_ail_delete(struct xfs_ail *, xfs_log_item_t *); STATIC xfs_log_item_t * xfs_ail_min(struct xfs_ail *); @@ -69,36 +71,6 @@ xfs_trans_ail_tail( } /* - * xfs_trans_push_ail - * - * This routine is called to move the tail of the AIL forward. It does this by - * trying to flush items in the AIL whose lsns are below the given - * threshold_lsn. - * - * the push is run asynchronously in a separate thread, so we return the tail - * of the log right now instead of the tail after the push. This means we will - * either continue right away, or we will sleep waiting on the async thread to - * do its work. - * - * We do this unlocked - we only need to know whether there is anything in the - * AIL at the time we are called. We don't need to access the contents of - * any of the objects, so the lock is not needed. - */ -void -xfs_trans_ail_push( - struct xfs_ail *ailp, - xfs_lsn_t threshold_lsn) -{ - xfs_log_item_t *lip; - - lip = xfs_ail_min(ailp); - if (lip && !XFS_FORCED_SHUTDOWN(ailp->xa_mount)) { - if (XFS_LSN_CMP(threshold_lsn, ailp->xa_target) > 0) - xfsaild_wakeup(ailp, threshold_lsn); - } -} - -/* * AIL traversal cursor initialisation. * * The cursor keeps track of where our current traversal is up @@ -236,16 +208,16 @@ out: } /* - * xfsaild_push does the work of pushing on the AIL. Returning a timeout of - * zero indicates that the caller should sleep until woken. + * xfs_ail_worker does the work of pushing on the AIL. It will requeue itself + * to run at a later time if there is more work to do to complete the push. */ -long -xfsaild_push( - struct xfs_ail *ailp, - xfs_lsn_t *last_lsn) +STATIC void +xfs_ail_worker( + struct work_struct *work) { - long tout = 0; - xfs_lsn_t last_pushed_lsn = *last_lsn; + struct xfs_ail *ailp = container_of(to_delayed_work(work), + struct xfs_ail, xa_work); + long tout; xfs_lsn_t target = ailp->xa_target; xfs_lsn_t lsn; xfs_log_item_t *lip; @@ -256,15 +228,15 @@ xfsaild_push( spin_lock(&ailp->xa_lock); xfs_trans_ail_cursor_init(ailp, cur); - lip = xfs_trans_ail_cursor_first(ailp, cur, *last_lsn); + lip = xfs_trans_ail_cursor_first(ailp, cur, ailp->xa_last_pushed_lsn); if (!lip || XFS_FORCED_SHUTDOWN(mp)) { /* * AIL is empty or our push has reached the end. */ xfs_trans_ail_cursor_done(ailp, cur); spin_unlock(&ailp->xa_lock); - *last_lsn = 0; - return tout; + ailp->xa_last_pushed_lsn = 0; + return; } XFS_STATS_INC(xs_push_ail); @@ -301,13 +273,13 @@ xfsaild_push( case XFS_ITEM_SUCCESS: XFS_STATS_INC(xs_push_ail_success); IOP_PUSH(lip); - last_pushed_lsn = lsn; + ailp->xa_last_pushed_lsn = lsn; break; case XFS_ITEM_PUSHBUF: XFS_STATS_INC(xs_push_ail_pushbuf); IOP_PUSHBUF(lip); - last_pushed_lsn = lsn; + ailp->xa_last_pushed_lsn = lsn; push_xfsbufd = 1; break; @@ -319,7 +291,7 @@ xfsaild_push( case XFS_ITEM_LOCKED: XFS_STATS_INC(xs_push_ail_locked); - last_pushed_lsn = lsn; + ailp->xa_last_pushed_lsn = lsn; stuck++; break; @@ -374,9 +346,23 @@ xfsaild_push( wake_up_process(mp->m_ddev_targp->bt_task); } + /* assume we have more work to do in a short while */ + tout = 10; if (!count) { /* We're past our target or empty, so idle */ - last_pushed_lsn = 0; + ailp->xa_last_pushed_lsn = 0; + + /* + * Check for an updated push target before clearing the + * XFS_AIL_PUSHING_BIT. If the target changed, we've got more + * work to do. Wait a bit longer before starting that work. + */ + smp_rmb(); + if (ailp->xa_target == target) { + clear_bit(XFS_AIL_PUSHING_BIT, &ailp->xa_flags); + return; + } + tout = 50; } else if (XFS_LSN_CMP(lsn, target) >= 0) { /* * We reached the target so wait a bit longer for I/O to @@ -384,7 +370,7 @@ xfsaild_push( * start the next scan from the start of the AIL. */ tout = 50; - last_pushed_lsn = 0; + ailp->xa_last_pushed_lsn = 0; } else if ((stuck * 100) / count > 90) { /* * Either there is a lot of contention on the AIL or we @@ -396,14 +382,48 @@ xfsaild_push( * continuing from where we were. */ tout = 20; - } else { - /* more to do, but wait a short while before continuing */ - tout = 10; } - *last_lsn = last_pushed_lsn; - return tout; + + /* There is more to do, requeue us. */ + queue_delayed_work(xfs_syncd_wq, &ailp->xa_work, + msecs_to_jiffies(tout)); } +/* + * This routine is called to move the tail of the AIL forward. It does this by + * trying to flush items in the AIL whose lsns are below the given + * threshold_lsn. + * + * The push is run asynchronously in a workqueue, which means the caller needs + * to handle waiting on the async flush for space to become available. + * We don't want to interrupt any push that is in progress, hence we only queue + * work if we set the pushing bit approriately. + * + * We do this unlocked - we only need to know whether there is anything in the + * AIL at the time we are called. We don't need to access the contents of + * any of the objects, so the lock is not needed. + */ +void +xfs_trans_ail_push( + struct xfs_ail *ailp, + xfs_lsn_t threshold_lsn) +{ + xfs_log_item_t *lip; + + lip = xfs_ail_min(ailp); + if (!lip || XFS_FORCED_SHUTDOWN(ailp->xa_mount) || + XFS_LSN_CMP(threshold_lsn, ailp->xa_target) <= 0) + return; + + /* + * Ensure that the new target is noticed in push code before it clears + * the XFS_AIL_PUSHING_BIT. + */ + smp_wmb(); + ailp->xa_target = threshold_lsn; + if (!test_and_set_bit(XFS_AIL_PUSHING_BIT, &ailp->xa_flags)) + queue_delayed_work(xfs_syncd_wq, &ailp->xa_work, 0); +} /* * This is to be called when an item is unlocked that may have @@ -615,7 +635,6 @@ xfs_trans_ail_init( xfs_mount_t *mp) { struct xfs_ail *ailp; - int error; ailp = kmem_zalloc(sizeof(struct xfs_ail), KM_MAYFAIL); if (!ailp) @@ -624,15 +643,9 @@ xfs_trans_ail_init( ailp->xa_mount = mp; INIT_LIST_HEAD(&ailp->xa_ail); spin_lock_init(&ailp->xa_lock); - error = xfsaild_start(ailp); - if (error) - goto out_free_ailp; + INIT_DELAYED_WORK(&ailp->xa_work, xfs_ail_worker); mp->m_ail = ailp; return 0; - -out_free_ailp: - kmem_free(ailp); - return error; } void @@ -641,7 +654,7 @@ xfs_trans_ail_destroy( { struct xfs_ail *ailp = mp->m_ail; - xfsaild_stop(ailp); + cancel_delayed_work_sync(&ailp->xa_work); kmem_free(ailp); } diff --git a/fs/xfs/xfs_trans_priv.h b/fs/xfs/xfs_trans_priv.h index 35162c2..6ebd322 100644 --- a/fs/xfs/xfs_trans_priv.h +++ b/fs/xfs/xfs_trans_priv.h @@ -65,16 +65,22 @@ struct xfs_ail_cursor { struct xfs_ail { struct xfs_mount *xa_mount; struct list_head xa_ail; - uint xa_gen; - struct task_struct *xa_task; xfs_lsn_t xa_target; struct xfs_ail_cursor xa_cursors; spinlock_t xa_lock; + struct delayed_work xa_work; + xfs_lsn_t xa_last_pushed_lsn; + unsigned long xa_flags; }; +#define XFS_AIL_PUSHING_BIT 0 + /* * From xfs_trans_ail.c */ + +extern struct workqueue_struct *xfs_ail_wq; /* AIL workqueue */ + void xfs_trans_ail_update_bulk(struct xfs_ail *ailp, struct xfs_log_item **log_items, int nr_items, xfs_lsn_t lsn) __releases(ailp->xa_lock); @@ -112,11 +118,6 @@ struct xfs_log_item *xfs_trans_ail_cursor_next(struct xfs_ail *ailp, void xfs_trans_ail_cursor_done(struct xfs_ail *ailp, struct xfs_ail_cursor *cur); -long xfsaild_push(struct xfs_ail *, xfs_lsn_t *); -void xfsaild_wakeup(struct xfs_ail *, xfs_lsn_t); -int xfsaild_start(struct xfs_ail *); -void xfsaild_stop(struct xfs_ail *); - #if BITS_PER_LONG != 64 static inline void xfs_trans_ail_copy_lsn( -- cgit v0.10.2 From cd4a3c503c185f5f0a20f04f90da0a6966dd03bd Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Fri, 8 Apr 2011 12:45:07 +1000 Subject: xfs: clean up code layout in xfs_trans_ail.c This patch rearranges the location of functions in xfs_trans_ail.c to remove the need for forward declarations of those functions in preparation for adding new functions without the need for forward declarations. Signed-off-by: Dave Chinner Reviewed-by: Alex Elder diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c index cb3aeac..8012bfb 100644 --- a/fs/xfs/xfs_trans_ail.c +++ b/fs/xfs/xfs_trans_ail.c @@ -30,41 +30,100 @@ struct workqueue_struct *xfs_ail_wq; /* AIL workqueue */ -STATIC void xfs_ail_splice(struct xfs_ail *, struct list_head *, xfs_lsn_t); -STATIC void xfs_ail_delete(struct xfs_ail *, xfs_log_item_t *); -STATIC xfs_log_item_t * xfs_ail_min(struct xfs_ail *); -STATIC xfs_log_item_t * xfs_ail_next(struct xfs_ail *, xfs_log_item_t *); - #ifdef DEBUG -STATIC void xfs_ail_check(struct xfs_ail *, xfs_log_item_t *); -#else +/* + * Check that the list is sorted as it should be. + */ +STATIC void +xfs_ail_check( + struct xfs_ail *ailp, + xfs_log_item_t *lip) +{ + xfs_log_item_t *prev_lip; + + if (list_empty(&ailp->xa_ail)) + return; + + /* + * Check the next and previous entries are valid. + */ + ASSERT((lip->li_flags & XFS_LI_IN_AIL) != 0); + prev_lip = list_entry(lip->li_ail.prev, xfs_log_item_t, li_ail); + if (&prev_lip->li_ail != &ailp->xa_ail) + ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) <= 0); + + prev_lip = list_entry(lip->li_ail.next, xfs_log_item_t, li_ail); + if (&prev_lip->li_ail != &ailp->xa_ail) + ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) >= 0); + + +#ifdef XFS_TRANS_DEBUG + /* + * Walk the list checking lsn ordering, and that every entry has the + * XFS_LI_IN_AIL flag set. This is really expensive, so only do it + * when specifically debugging the transaction subsystem. + */ + prev_lip = list_entry(&ailp->xa_ail, xfs_log_item_t, li_ail); + list_for_each_entry(lip, &ailp->xa_ail, li_ail) { + if (&prev_lip->li_ail != &ailp->xa_ail) + ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) <= 0); + ASSERT((lip->li_flags & XFS_LI_IN_AIL) != 0); + prev_lip = lip; + } +#endif /* XFS_TRANS_DEBUG */ +} +#else /* !DEBUG */ #define xfs_ail_check(a,l) #endif /* DEBUG */ +/* + * Return a pointer to the first item in the AIL. If the AIL is empty, then + * return NULL. + */ +static xfs_log_item_t * +xfs_ail_min( + struct xfs_ail *ailp) +{ + if (list_empty(&ailp->xa_ail)) + return NULL; + + return list_first_entry(&ailp->xa_ail, xfs_log_item_t, li_ail); +} + +/* + * Return a pointer to the item which follows the given item in the AIL. If + * the given item is the last item in the list, then return NULL. + */ +static xfs_log_item_t * +xfs_ail_next( + struct xfs_ail *ailp, + xfs_log_item_t *lip) +{ + if (lip->li_ail.next == &ailp->xa_ail) + return NULL; + + return list_first_entry(&lip->li_ail, xfs_log_item_t, li_ail); +} /* - * This is called by the log manager code to determine the LSN - * of the tail of the log. This is exactly the LSN of the first - * item in the AIL. If the AIL is empty, then this function - * returns 0. + * This is called by the log manager code to determine the LSN of the tail of + * the log. This is exactly the LSN of the first item in the AIL. If the AIL + * is empty, then this function returns 0. * - * We need the AIL lock in order to get a coherent read of the - * lsn of the last item in the AIL. + * We need the AIL lock in order to get a coherent read of the lsn of the last + * item in the AIL. */ xfs_lsn_t xfs_trans_ail_tail( struct xfs_ail *ailp) { - xfs_lsn_t lsn; + xfs_lsn_t lsn = 0; xfs_log_item_t *lip; spin_lock(&ailp->xa_lock); lip = xfs_ail_min(ailp); - if (lip == NULL) { - lsn = (xfs_lsn_t)0; - } else { + if (lip) lsn = lip->li_lsn; - } spin_unlock(&ailp->xa_lock); return lsn; @@ -208,6 +267,47 @@ out: } /* + * splice the log item list into the AIL at the given LSN. + */ +static void +xfs_ail_splice( + struct xfs_ail *ailp, + struct list_head *list, + xfs_lsn_t lsn) +{ + xfs_log_item_t *next_lip; + + /* If the list is empty, just insert the item. */ + if (list_empty(&ailp->xa_ail)) { + list_splice(list, &ailp->xa_ail); + return; + } + + list_for_each_entry_reverse(next_lip, &ailp->xa_ail, li_ail) { + if (XFS_LSN_CMP(next_lip->li_lsn, lsn) <= 0) + break; + } + + ASSERT(&next_lip->li_ail == &ailp->xa_ail || + XFS_LSN_CMP(next_lip->li_lsn, lsn) <= 0); + + list_splice_init(list, &next_lip->li_ail); +} + +/* + * Delete the given item from the AIL. Return a pointer to the item. + */ +static void +xfs_ail_delete( + struct xfs_ail *ailp, + xfs_log_item_t *lip) +{ + xfs_ail_check(ailp, lip); + list_del(&lip->li_ail); + xfs_trans_ail_cursor_clear(ailp, lip); +} + +/* * xfs_ail_worker does the work of pushing on the AIL. It will requeue itself * to run at a later time if there is more work to do to complete the push. */ @@ -657,121 +757,3 @@ xfs_trans_ail_destroy( cancel_delayed_work_sync(&ailp->xa_work); kmem_free(ailp); } - -/* - * splice the log item list into the AIL at the given LSN. - */ -STATIC void -xfs_ail_splice( - struct xfs_ail *ailp, - struct list_head *list, - xfs_lsn_t lsn) -{ - xfs_log_item_t *next_lip; - - /* - * If the list is empty, just insert the item. - */ - if (list_empty(&ailp->xa_ail)) { - list_splice(list, &ailp->xa_ail); - return; - } - - list_for_each_entry_reverse(next_lip, &ailp->xa_ail, li_ail) { - if (XFS_LSN_CMP(next_lip->li_lsn, lsn) <= 0) - break; - } - - ASSERT((&next_lip->li_ail == &ailp->xa_ail) || - (XFS_LSN_CMP(next_lip->li_lsn, lsn) <= 0)); - - list_splice_init(list, &next_lip->li_ail); - return; -} - -/* - * Delete the given item from the AIL. Return a pointer to the item. - */ -STATIC void -xfs_ail_delete( - struct xfs_ail *ailp, - xfs_log_item_t *lip) -{ - xfs_ail_check(ailp, lip); - list_del(&lip->li_ail); - xfs_trans_ail_cursor_clear(ailp, lip); -} - -/* - * Return a pointer to the first item in the AIL. - * If the AIL is empty, then return NULL. - */ -STATIC xfs_log_item_t * -xfs_ail_min( - struct xfs_ail *ailp) -{ - if (list_empty(&ailp->xa_ail)) - return NULL; - - return list_first_entry(&ailp->xa_ail, xfs_log_item_t, li_ail); -} - -/* - * Return a pointer to the item which follows - * the given item in the AIL. If the given item - * is the last item in the list, then return NULL. - */ -STATIC xfs_log_item_t * -xfs_ail_next( - struct xfs_ail *ailp, - xfs_log_item_t *lip) -{ - if (lip->li_ail.next == &ailp->xa_ail) - return NULL; - - return list_first_entry(&lip->li_ail, xfs_log_item_t, li_ail); -} - -#ifdef DEBUG -/* - * Check that the list is sorted as it should be. - */ -STATIC void -xfs_ail_check( - struct xfs_ail *ailp, - xfs_log_item_t *lip) -{ - xfs_log_item_t *prev_lip; - - if (list_empty(&ailp->xa_ail)) - return; - - /* - * Check the next and previous entries are valid. - */ - ASSERT((lip->li_flags & XFS_LI_IN_AIL) != 0); - prev_lip = list_entry(lip->li_ail.prev, xfs_log_item_t, li_ail); - if (&prev_lip->li_ail != &ailp->xa_ail) - ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) <= 0); - - prev_lip = list_entry(lip->li_ail.next, xfs_log_item_t, li_ail); - if (&prev_lip->li_ail != &ailp->xa_ail) - ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) >= 0); - - -#ifdef XFS_TRANS_DEBUG - /* - * Walk the list checking lsn ordering, and that every entry has the - * XFS_LI_IN_AIL flag set. This is really expensive, so only do it - * when specifically debugging the transaction subsystem. - */ - prev_lip = list_entry(&ailp->xa_ail, xfs_log_item_t, li_ail); - list_for_each_entry(lip, &ailp->xa_ail, li_ail) { - if (&prev_lip->li_ail != &ailp->xa_ail) - ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) <= 0); - ASSERT((lip->li_flags & XFS_LI_IN_AIL) != 0); - prev_lip = lip; - } -#endif /* XFS_TRANS_DEBUG */ -} -#endif /* DEBUG */ -- cgit v0.10.2 From fd074841cfe01b006465fb9388091012585e8dfb Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Fri, 8 Apr 2011 12:45:07 +1000 Subject: xfs: push the AIL from memory reclaim and periodic sync When we are short on memory, we want to expedite the cleaning of dirty objects. Hence when we run short on memory, we need to kick the AIL flushing into action to clean as many dirty objects as quickly as possible. To implement this, sample the lsn of the log item at the head of the AIL and use that as the push target for the AIL flush. Further, we keep items in the AIL that are dirty that are not tracked any other way, so we can get objects sitting in the AIL that don't get written back until the AIL is pushed. Hence to get the filesystem to the idle state, we might need to push the AIL to flush out any remaining dirty objects sitting in the AIL. This requires the same push mechanism as the reclaim push. This patch also renames xfs_trans_ail_tail() to xfs_ail_min_lsn() to match the new xfs_ail_max_lsn() function introduced in this patch. Similarly for xfs_trans_ail_push -> xfs_ail_push. Signed-off-by: Dave Chinner Reviewed-by: Alex Elder diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c index debe282..9ad9560 100644 --- a/fs/xfs/linux-2.6/xfs_sync.c +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -22,6 +22,7 @@ #include "xfs_log.h" #include "xfs_inum.h" #include "xfs_trans.h" +#include "xfs_trans_priv.h" #include "xfs_sb.h" #include "xfs_ag.h" #include "xfs_mount.h" @@ -462,6 +463,9 @@ xfs_sync_worker( else xfs_log_force(mp, 0); error = xfs_qm_sync(mp, SYNC_TRYLOCK); + + /* start pushing all the metadata that is currently dirty */ + xfs_ail_push_all(mp->m_ail); } /* queue us up again */ @@ -1027,8 +1031,9 @@ xfs_reclaim_inode_shrink( mp = container_of(shrink, struct xfs_mount, m_inode_shrink); if (nr_to_scan) { - /* kick background reclaimer */ + /* kick background reclaimer and push the AIL */ xfs_syncd_queue_reclaim(mp); + xfs_ail_push_all(mp->m_ail); if (!(gfp_mask & __GFP_FS)) return -1; diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 25efa9b..2464316 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -761,7 +761,7 @@ xfs_log_need_covered(xfs_mount_t *mp) break; case XLOG_STATE_COVER_NEED: case XLOG_STATE_COVER_NEED2: - if (!xfs_trans_ail_tail(log->l_ailp) && + if (!xfs_ail_min_lsn(log->l_ailp) && xlog_iclogs_empty(log)) { if (log->l_covered_state == XLOG_STATE_COVER_NEED) log->l_covered_state = XLOG_STATE_COVER_DONE; @@ -801,7 +801,7 @@ xlog_assign_tail_lsn( xfs_lsn_t tail_lsn; struct log *log = mp->m_log; - tail_lsn = xfs_trans_ail_tail(mp->m_ail); + tail_lsn = xfs_ail_min_lsn(mp->m_ail); if (!tail_lsn) tail_lsn = atomic64_read(&log->l_last_sync_lsn); @@ -1239,7 +1239,7 @@ xlog_grant_push_ail( * the filesystem is shutting down. */ if (!XLOG_FORCED_SHUTDOWN(log)) - xfs_trans_ail_push(log->l_ailp, threshold_lsn); + xfs_ail_push(log->l_ailp, threshold_lsn); } /* diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c index 8012bfb..acdb92f 100644 --- a/fs/xfs/xfs_trans_ail.c +++ b/fs/xfs/xfs_trans_ail.c @@ -90,6 +90,20 @@ xfs_ail_min( return list_first_entry(&ailp->xa_ail, xfs_log_item_t, li_ail); } + /* + * Return a pointer to the last item in the AIL. If the AIL is empty, then + * return NULL. + */ +static xfs_log_item_t * +xfs_ail_max( + struct xfs_ail *ailp) +{ + if (list_empty(&ailp->xa_ail)) + return NULL; + + return list_entry(ailp->xa_ail.prev, xfs_log_item_t, li_ail); +} + /* * Return a pointer to the item which follows the given item in the AIL. If * the given item is the last item in the list, then return NULL. @@ -114,7 +128,7 @@ xfs_ail_next( * item in the AIL. */ xfs_lsn_t -xfs_trans_ail_tail( +xfs_ail_min_lsn( struct xfs_ail *ailp) { xfs_lsn_t lsn = 0; @@ -130,6 +144,25 @@ xfs_trans_ail_tail( } /* + * Return the maximum lsn held in the AIL, or zero if the AIL is empty. + */ +static xfs_lsn_t +xfs_ail_max_lsn( + struct xfs_ail *ailp) +{ + xfs_lsn_t lsn = 0; + xfs_log_item_t *lip; + + spin_lock(&ailp->xa_lock); + lip = xfs_ail_max(ailp); + if (lip) + lsn = lip->li_lsn; + spin_unlock(&ailp->xa_lock); + + return lsn; +} + +/* * AIL traversal cursor initialisation. * * The cursor keeps track of where our current traversal is up @@ -504,7 +537,7 @@ xfs_ail_worker( * any of the objects, so the lock is not needed. */ void -xfs_trans_ail_push( +xfs_ail_push( struct xfs_ail *ailp, xfs_lsn_t threshold_lsn) { @@ -526,6 +559,19 @@ xfs_trans_ail_push( } /* + * Push out all items in the AIL immediately + */ +void +xfs_ail_push_all( + struct xfs_ail *ailp) +{ + xfs_lsn_t threshold_lsn = xfs_ail_max_lsn(ailp); + + if (threshold_lsn) + xfs_ail_push(ailp, threshold_lsn); +} + +/* * This is to be called when an item is unlocked that may have * been in the AIL. It will wake up the first member of the AIL * wait list if this item's unlocking might allow it to progress. diff --git a/fs/xfs/xfs_trans_priv.h b/fs/xfs/xfs_trans_priv.h index 6ebd322..6b164e9 100644 --- a/fs/xfs/xfs_trans_priv.h +++ b/fs/xfs/xfs_trans_priv.h @@ -104,12 +104,13 @@ xfs_trans_ail_delete( xfs_trans_ail_delete_bulk(ailp, &lip, 1); } -void xfs_trans_ail_push(struct xfs_ail *, xfs_lsn_t); +void xfs_ail_push(struct xfs_ail *, xfs_lsn_t); +void xfs_ail_push_all(struct xfs_ail *); +xfs_lsn_t xfs_ail_min_lsn(struct xfs_ail *ailp); + void xfs_trans_unlocked_item(struct xfs_ail *, xfs_log_item_t *); -xfs_lsn_t xfs_trans_ail_tail(struct xfs_ail *ailp); - struct xfs_log_item *xfs_trans_ail_cursor_first(struct xfs_ail *ailp, struct xfs_ail_cursor *cur, xfs_lsn_t lsn); -- cgit v0.10.2 From be65b18a10e62321c5ba09a1dc0f70babeb0eba1 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Fri, 8 Apr 2011 12:45:07 +1000 Subject: xfs: catch bad block numbers freeing extents. A fuzzed filesystem crashed a kernel when freeing an extent with a block number beyond the end of the filesystem. Convert all the debug asserts in xfs_free_extent() to active checks so that we catch bad extents and return that the filesytsem is corrupted rather than crashing. Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Reviewed-by: Alex Elder diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c index 4bc3c64..27d64d7 100644 --- a/fs/xfs/xfs_alloc.c +++ b/fs/xfs/xfs_alloc.c @@ -2395,17 +2395,33 @@ xfs_free_extent( memset(&args, 0, sizeof(xfs_alloc_arg_t)); args.tp = tp; args.mp = tp->t_mountp; + + /* + * validate that the block number is legal - the enables us to detect + * and handle a silent filesystem corruption rather than crashing. + */ args.agno = XFS_FSB_TO_AGNO(args.mp, bno); - ASSERT(args.agno < args.mp->m_sb.sb_agcount); + if (args.agno >= args.mp->m_sb.sb_agcount) + return EFSCORRUPTED; + args.agbno = XFS_FSB_TO_AGBNO(args.mp, bno); + if (args.agbno >= args.mp->m_sb.sb_agblocks) + return EFSCORRUPTED; + args.pag = xfs_perag_get(args.mp, args.agno); - if ((error = xfs_alloc_fix_freelist(&args, XFS_ALLOC_FLAG_FREEING))) + ASSERT(args.pag); + + error = xfs_alloc_fix_freelist(&args, XFS_ALLOC_FLAG_FREEING); + if (error) goto error0; -#ifdef DEBUG - ASSERT(args.agbp != NULL); - ASSERT((args.agbno + len) <= - be32_to_cpu(XFS_BUF_TO_AGF(args.agbp)->agf_length)); -#endif + + /* validate the extent size is legal now we have the agf locked */ + if (args.agbno + len > + be32_to_cpu(XFS_BUF_TO_AGF(args.agbp)->agf_length)) { + error = EFSCORRUPTED; + goto error0; + } + error = xfs_free_ag_extent(tp, args.agbp, args.agno, args.agbno, len, 0); error0: xfs_perag_put(args.pag); -- cgit v0.10.2 From da8a1a4a4dfc1ead12c343b992fc8300a22d33d0 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Fri, 8 Apr 2011 12:45:07 +1000 Subject: xfs: convert log tail checking to a warning On the Power platform, the log tail debug checks fire excessively causing the system to panic early in testing. The debug checks are known to be racy, though on x86_64 there is no evidence that they trigger at all. We want to keep the checks active on debug systems to alert us to problems with log space accounting, but we need to reduce the impact of a racy check on testing on the Power platform. As a result, convert the ASSERT conditions to warnings, and allow them to fire only once per filesystem mount. This will prevent false positives from interfering with testing, whilst still providing us with the indication that they may be a problem with log space accounting should that occur. Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Reviewed-by: Alex Elder diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 2464316..b612ce4 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -3407,6 +3407,17 @@ xlog_verify_dest_ptr( xfs_emerg(log->l_mp, "%s: invalid ptr", __func__); } +/* + * Check to make sure the grant write head didn't just over lap the tail. If + * the cycles are the same, we can't be overlapping. Otherwise, make sure that + * the cycles differ by exactly one and check the byte count. + * + * This check is run unlocked, so can give false positives. Rather than assert + * on failures, use a warn-once flag and a panic tag to allow the admin to + * determine if they want to panic the machine when such an error occurs. For + * debug kernels this will have the same effect as using an assert but, unlinke + * an assert, it can be turned off at runtime. + */ STATIC void xlog_verify_grant_tail( struct log *log) @@ -3414,17 +3425,22 @@ xlog_verify_grant_tail( int tail_cycle, tail_blocks; int cycle, space; - /* - * Check to make sure the grant write head didn't just over lap the - * tail. If the cycles are the same, we can't be overlapping. - * Otherwise, make sure that the cycles differ by exactly one and - * check the byte count. - */ xlog_crack_grant_head(&log->l_grant_write_head, &cycle, &space); xlog_crack_atomic_lsn(&log->l_tail_lsn, &tail_cycle, &tail_blocks); if (tail_cycle != cycle) { - ASSERT(cycle - 1 == tail_cycle); - ASSERT(space <= BBTOB(tail_blocks)); + if (cycle - 1 != tail_cycle && + !(log->l_flags & XLOG_TAIL_WARN)) { + xfs_alert_tag(log->l_mp, XFS_PTAG_LOGRES, + "%s: cycle - 1 != tail_cycle", __func__); + log->l_flags |= XLOG_TAIL_WARN; + } + + if (space > BBTOB(tail_blocks) && + !(log->l_flags & XLOG_TAIL_WARN)) { + xfs_alert_tag(log->l_mp, XFS_PTAG_LOGRES, + "%s: space > BBTOB(tail_blocks)", __func__); + log->l_flags |= XLOG_TAIL_WARN; + } } } diff --git a/fs/xfs/xfs_log_priv.h b/fs/xfs/xfs_log_priv.h index 15dbf1f..bc988d4 100644 --- a/fs/xfs/xfs_log_priv.h +++ b/fs/xfs/xfs_log_priv.h @@ -144,6 +144,7 @@ static inline uint xlog_get_client_id(__be32 i) #define XLOG_RECOVERY_NEEDED 0x4 /* log was recovered */ #define XLOG_IO_ERROR 0x8 /* log hit an I/O error, and being shutdown */ +#define XLOG_TAIL_WARN 0x10 /* log tail verify warning issued */ #ifdef __KERNEL__ /* -- cgit v0.10.2 From ab6dc30da55d262c54ae12e47e4a2e4372e39fbf Mon Sep 17 00:00:00 2001 From: Yevgeny Petrilin Date: Wed, 6 Apr 2011 23:24:42 +0000 Subject: mlx4: Sensing link type at device initialization When bringing the port up, performing a SENSE_PORT command To try and check to which physical link type (IB or Ethernet) the physical port is connected. In case there is no valid link partner, the port will come up as its supported default. Signed-off-by: Yevgeny Petrilin Signed-off-by: David S. Miller diff --git a/drivers/net/mlx4/main.c b/drivers/net/mlx4/main.c index 62fa7ee..3814fc9 100644 --- a/drivers/net/mlx4/main.c +++ b/drivers/net/mlx4/main.c @@ -944,6 +944,10 @@ static int mlx4_setup_hca(struct mlx4_dev *dev) } for (port = 1; port <= dev->caps.num_ports; port++) { + enum mlx4_port_type port_type = 0; + mlx4_SENSE_PORT(dev, port, &port_type); + if (port_type) + dev->caps.port_type[port] = port_type; ib_port_default_caps = 0; err = mlx4_get_port_ib_caps(dev, port, &ib_port_default_caps); if (err) @@ -958,6 +962,7 @@ static int mlx4_setup_hca(struct mlx4_dev *dev) goto err_mcg_table_free; } } + mlx4_set_port_mask(dev); return 0; diff --git a/drivers/net/mlx4/mlx4.h b/drivers/net/mlx4/mlx4.h index c1e0e5f..dd7d745 100644 --- a/drivers/net/mlx4/mlx4.h +++ b/drivers/net/mlx4/mlx4.h @@ -431,6 +431,8 @@ void mlx4_srq_event(struct mlx4_dev *dev, u32 srqn, int event_type); void mlx4_handle_catas_err(struct mlx4_dev *dev); +int mlx4_SENSE_PORT(struct mlx4_dev *dev, int port, + enum mlx4_port_type *type); void mlx4_do_sense_ports(struct mlx4_dev *dev, enum mlx4_port_type *stype, enum mlx4_port_type *defaults); diff --git a/drivers/net/mlx4/sense.c b/drivers/net/mlx4/sense.c index 015fbe7..e2337a7 100644 --- a/drivers/net/mlx4/sense.c +++ b/drivers/net/mlx4/sense.c @@ -38,8 +38,8 @@ #include "mlx4.h" -static int mlx4_SENSE_PORT(struct mlx4_dev *dev, int port, - enum mlx4_port_type *type) +int mlx4_SENSE_PORT(struct mlx4_dev *dev, int port, + enum mlx4_port_type *type) { u64 out_param; int err = 0; -- cgit v0.10.2 From 60b1809f7e2ea2620aa548e1c866066756bdd1d2 Mon Sep 17 00:00:00 2001 From: Yevgeny Petrilin Date: Wed, 6 Apr 2011 23:25:45 +0000 Subject: mlx4_en: Restoring RX buffer pointer in case of failure If not done, second attempt to open the RX ring would cause memory corruption. Signed-off-by: Yevgeny Petrilin Signed-off-by: David S. Miller diff --git a/drivers/net/mlx4/en_rx.c b/drivers/net/mlx4/en_rx.c index 05998ee..98d4ce0 100644 --- a/drivers/net/mlx4/en_rx.c +++ b/drivers/net/mlx4/en_rx.c @@ -345,6 +345,8 @@ int mlx4_en_activate_rx_rings(struct mlx4_en_priv *priv) err = mlx4_en_init_allocator(priv, ring); if (err) { en_err(priv, "Failed initializing ring allocator\n"); + if (ring->stride <= TXBB_SIZE) + ring->buf -= TXBB_SIZE; ring_ind--; goto err_allocator; } @@ -369,6 +371,8 @@ err_buffers: ring_ind = priv->rx_ring_num - 1; err_allocator: while (ring_ind >= 0) { + if (priv->rx_ring[ring_ind].stride <= TXBB_SIZE) + priv->rx_ring[ring_ind].buf -= TXBB_SIZE; mlx4_en_destroy_allocator(priv, &priv->rx_ring[ring_ind]); ring_ind--; } -- cgit v0.10.2 From ecb697c16c1718ae97bb73ce41a5d5ac2aed29ec Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Mon, 4 Apr 2011 12:55:44 +0000 Subject: xfs: fix variable set but not used warnings GCC 4.6 now warnings about variables set but not used. Fix the trivially fixable warnings of this sort. Signed-off-by: Christoph Hellwig Signed-off-by: Alex Elder diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index d917146..2eef165 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -293,7 +293,6 @@ xfs_buf_allocate_memory( size_t nbytes, offset; gfp_t gfp_mask = xb_to_gfp(flags); unsigned short page_count, i; - pgoff_t first; xfs_off_t end; int error; @@ -333,7 +332,6 @@ use_alloc_page: return error; offset = bp->b_offset; - first = bp->b_file_offset >> PAGE_SHIFT; bp->b_flags |= _XBF_PAGES; for (i = 0; i < bp->b_page_count; i++) { diff --git a/fs/xfs/quota/xfs_qm.c b/fs/xfs/quota/xfs_qm.c index 254ee06..69228aa 100644 --- a/fs/xfs/quota/xfs_qm.c +++ b/fs/xfs/quota/xfs_qm.c @@ -461,12 +461,10 @@ xfs_qm_dqflush_all( struct xfs_quotainfo *q = mp->m_quotainfo; int recl; struct xfs_dquot *dqp; - int niters; int error; if (!q) return 0; - niters = 0; again: mutex_lock(&q->qi_dqlist_lock); list_for_each_entry(dqp, &q->qi_dqlist, q_mplist) { @@ -1314,14 +1312,9 @@ xfs_qm_dqiter_bufs( { xfs_buf_t *bp; int error; - int notcommitted; - int incr; int type; ASSERT(blkcnt > 0); - notcommitted = 0; - incr = (blkcnt > XFS_QM_MAX_DQCLUSTER_LOGSZ) ? - XFS_QM_MAX_DQCLUSTER_LOGSZ : blkcnt; type = flags & XFS_QMOPT_UQUOTA ? XFS_DQ_USER : (flags & XFS_QMOPT_PQUOTA ? XFS_DQ_PROJ : XFS_DQ_GROUP); error = 0; diff --git a/fs/xfs/quota/xfs_qm.h b/fs/xfs/quota/xfs_qm.h index c9446f1..567b29b 100644 --- a/fs/xfs/quota/xfs_qm.h +++ b/fs/xfs/quota/xfs_qm.h @@ -65,11 +65,6 @@ extern kmem_zone_t *qm_dqtrxzone; * block in the dquot/xqm code. */ #define XFS_DQUOT_CLUSTER_SIZE_FSB (xfs_filblks_t)1 -/* - * When doing a quotacheck, we log dquot clusters of this many FSBs at most - * in a single transaction. We don't want to ask for too huge a log reservation. - */ -#define XFS_QM_MAX_DQCLUSTER_LOGSZ 3 typedef xfs_dqhash_t xfs_dqlist_t; diff --git a/fs/xfs/quota/xfs_qm_syscalls.c b/fs/xfs/quota/xfs_qm_syscalls.c index c82f067..c79859e 100644 --- a/fs/xfs/quota/xfs_qm_syscalls.c +++ b/fs/xfs/quota/xfs_qm_syscalls.c @@ -313,14 +313,12 @@ xfs_qm_scall_quotaon( { int error; uint qf; - uint accflags; __int64_t sbflags; flags &= (XFS_ALL_QUOTA_ACCT | XFS_ALL_QUOTA_ENFD); /* * Switching on quota accounting must be done at mount time. */ - accflags = flags & XFS_ALL_QUOTA_ACCT; flags &= ~(XFS_ALL_QUOTA_ACCT); sbflags = 0; diff --git a/fs/xfs/xfs_itable.c b/fs/xfs/xfs_itable.c index dc1882a..751e94f 100644 --- a/fs/xfs/xfs_itable.c +++ b/fs/xfs/xfs_itable.c @@ -204,7 +204,6 @@ xfs_bulkstat( xfs_agi_t *agi; /* agi header data */ xfs_agino_t agino; /* inode # in allocation group */ xfs_agnumber_t agno; /* allocation group number */ - xfs_daddr_t bno; /* inode cluster start daddr */ int chunkidx; /* current index into inode chunk */ int clustidx; /* current index into inode cluster */ xfs_btree_cur_t *cur; /* btree cursor for ialloc btree */ @@ -463,7 +462,6 @@ xfs_bulkstat( mp->m_sb.sb_inopblog); } ino = XFS_AGINO_TO_INO(mp, agno, agino); - bno = XFS_AGB_TO_DADDR(mp, agno, agbno); /* * Skip if this inode is free. */ -- cgit v0.10.2 From 957935dcd8e11d6f789b4ed769b376040e15565b Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Sat, 2 Apr 2011 18:13:40 +0000 Subject: xfs: fix xfs_debug warnings For a CONFIG_XFS_DEBUG=n build gcc complains about statements with no effect in xfs_debug: fs/xfs/quota/xfs_qm_syscalls.c: In function 'xfs_qm_scall_trunc_qfiles': fs/xfs/quota/xfs_qm_syscalls.c:291:3: warning: statement with no effect The reason for that is that the various new xfs message functions have a return value which is never used, and in case of the non-debug build xfs_debug the macro evaluates to a plain 0 which produces the above warnings. This can be fixed by turning xfs_debug into an inline function instead of a macro, but in addition to that I've also changed all the message helpers to return void as we never use their return values. Signed-off-by: Christoph Hellwig Reviewed-by: Dave Chinner Signed-off-by: Alex Elder diff --git a/fs/xfs/linux-2.6/xfs_message.c b/fs/xfs/linux-2.6/xfs_message.c index 508e06f..3ca7956 100644 --- a/fs/xfs/linux-2.6/xfs_message.c +++ b/fs/xfs/linux-2.6/xfs_message.c @@ -28,53 +28,47 @@ /* * XFS logging functions */ -static int +static void __xfs_printk( const char *level, const struct xfs_mount *mp, struct va_format *vaf) { if (mp && mp->m_fsname) - return printk("%sXFS (%s): %pV\n", level, mp->m_fsname, vaf); - return printk("%sXFS: %pV\n", level, vaf); + printk("%sXFS (%s): %pV\n", level, mp->m_fsname, vaf); + printk("%sXFS: %pV\n", level, vaf); } -int xfs_printk( +void xfs_printk( const char *level, const struct xfs_mount *mp, const char *fmt, ...) { struct va_format vaf; va_list args; - int r; va_start(args, fmt); vaf.fmt = fmt; vaf.va = &args; - r = __xfs_printk(level, mp, &vaf); + __xfs_printk(level, mp, &vaf); va_end(args); - - return r; } #define define_xfs_printk_level(func, kern_level) \ -int func(const struct xfs_mount *mp, const char *fmt, ...) \ +void func(const struct xfs_mount *mp, const char *fmt, ...) \ { \ struct va_format vaf; \ va_list args; \ - int r; \ \ va_start(args, fmt); \ \ vaf.fmt = fmt; \ vaf.va = &args; \ \ - r = __xfs_printk(kern_level, mp, &vaf); \ + __xfs_printk(kern_level, mp, &vaf); \ va_end(args); \ - \ - return r; \ } \ define_xfs_printk_level(xfs_emerg, KERN_EMERG); @@ -88,7 +82,7 @@ define_xfs_printk_level(xfs_info, KERN_INFO); define_xfs_printk_level(xfs_debug, KERN_DEBUG); #endif -int +void xfs_alert_tag( const struct xfs_mount *mp, int panic_tag, @@ -97,7 +91,6 @@ xfs_alert_tag( struct va_format vaf; va_list args; int do_panic = 0; - int r; if (xfs_panic_mask && (xfs_panic_mask & panic_tag)) { xfs_printk(KERN_ALERT, mp, @@ -110,12 +103,10 @@ xfs_alert_tag( vaf.fmt = fmt; vaf.va = &args; - r = __xfs_printk(KERN_ALERT, mp, &vaf); + __xfs_printk(KERN_ALERT, mp, &vaf); va_end(args); BUG_ON(do_panic); - - return r; } void diff --git a/fs/xfs/linux-2.6/xfs_message.h b/fs/xfs/linux-2.6/xfs_message.h index e77ffa1..f1b3fc1 100644 --- a/fs/xfs/linux-2.6/xfs_message.h +++ b/fs/xfs/linux-2.6/xfs_message.h @@ -3,32 +3,34 @@ struct xfs_mount; -extern int xfs_printk(const char *level, const struct xfs_mount *mp, +extern void xfs_printk(const char *level, const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 3, 4))); -extern int xfs_emerg(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_emerg(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); -extern int xfs_alert(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_alert(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); -extern int xfs_alert_tag(const struct xfs_mount *mp, int tag, +extern void xfs_alert_tag(const struct xfs_mount *mp, int tag, const char *fmt, ...) __attribute__ ((format (printf, 3, 4))); -extern int xfs_crit(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_crit(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); -extern int xfs_err(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_err(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); -extern int xfs_warn(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_warn(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); -extern int xfs_notice(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_notice(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); -extern int xfs_info(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_info(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); #ifdef DEBUG -extern int xfs_debug(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_debug(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); #else -#define xfs_debug(mp, fmt, ...) (0) +static inline void xfs_debug(const struct xfs_mount *mp, const char *fmt, ...) +{ +} #endif extern void assfail(char *expr, char *f, int l); -- cgit v0.10.2 From a1b7ea5d58c53c13f082110e535d98bc4e8e5cfe Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 30 Mar 2011 11:05:09 +0000 Subject: xfs: use proper interfaces for on-stack plugging Add proper blk_start_plug/blk_finish_plug pairs for the two places where we issue buffer I/O, and remove the blk_flush_plug in xfs_buf_lock and xfs_buf_iowait, given that context switches already flush the per-process plugging lists. Signed-off-by: Christoph Hellwig Signed-off-by: Alex Elder diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index 2eef165..478ca15 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -915,8 +915,6 @@ xfs_buf_lock( if (atomic_read(&bp->b_pin_count) && (bp->b_flags & XBF_STALE)) xfs_log_force(bp->b_target->bt_mount, 0); - if (atomic_read(&bp->b_io_remaining)) - blk_flush_plug(current); down(&bp->b_sema); XB_SET_OWNER(bp); @@ -1305,8 +1303,6 @@ xfs_buf_iowait( { trace_xfs_buf_iowait(bp, _RET_IP_); - if (atomic_read(&bp->b_io_remaining)) - blk_flush_plug(current); wait_for_completion(&bp->b_iowait); trace_xfs_buf_iowait_done(bp, _RET_IP_); @@ -1743,8 +1739,8 @@ xfsbufd( do { long age = xfs_buf_age_centisecs * msecs_to_jiffies(10); long tout = xfs_buf_timer_centisecs * msecs_to_jiffies(10); - int count = 0; struct list_head tmp; + struct blk_plug plug; if (unlikely(freezing(current))) { set_bit(XBT_FORCE_SLEEP, &target->bt_flags); @@ -1760,16 +1756,15 @@ xfsbufd( xfs_buf_delwri_split(target, &tmp, age); list_sort(NULL, &tmp, xfs_buf_cmp); + + blk_start_plug(&plug); while (!list_empty(&tmp)) { struct xfs_buf *bp; bp = list_first_entry(&tmp, struct xfs_buf, b_list); list_del_init(&bp->b_list); xfs_bdstrat_cb(bp); - count++; } - if (count) - blk_flush_plug(current); - + blk_finish_plug(&plug); } while (!kthread_should_stop()); return 0; @@ -1789,6 +1784,7 @@ xfs_flush_buftarg( int pincount = 0; LIST_HEAD(tmp_list); LIST_HEAD(wait_list); + struct blk_plug plug; xfs_buf_runall_queues(xfsconvertd_workqueue); xfs_buf_runall_queues(xfsdatad_workqueue); @@ -1803,6 +1799,8 @@ xfs_flush_buftarg( * we do that after issuing all the IO. */ list_sort(NULL, &tmp_list, xfs_buf_cmp); + + blk_start_plug(&plug); while (!list_empty(&tmp_list)) { bp = list_first_entry(&tmp_list, struct xfs_buf, b_list); ASSERT(target == bp->b_target); @@ -1813,10 +1811,10 @@ xfs_flush_buftarg( } xfs_bdstrat_cb(bp); } + blk_finish_plug(&plug); if (wait) { - /* Expedite and wait for IO to complete. */ - blk_flush_plug(current); + /* Wait for IO to complete. */ while (!list_empty(&wait_list)) { bp = list_first_entry(&wait_list, struct xfs_buf, b_list); -- cgit v0.10.2 From be1a12a0dfed06cf1e62e35bf91620dc610a451a Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 6 Apr 2011 13:05:22 -0400 Subject: Btrfs: deal with the case that we run out of space in the cache Currently we don't handle running out of space in the cache, so to fix this we keep track of how far in the cache we are. Then we only dirty the pages if we successfully modify all of them, otherwise if we have an error or run out of space we can just drop them and not worry about the vm writing them out. Thanks, Tested-by Johannes Hirte Signed-off-by: Josef Bacik diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 3458b57..0d00a07 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -2576,6 +2576,11 @@ int btrfs_drop_extents(struct btrfs_trans_handle *trans, struct inode *inode, int btrfs_mark_extent_written(struct btrfs_trans_handle *trans, struct inode *inode, u64 start, u64 end); int btrfs_release_file(struct inode *inode, struct file *file); +void btrfs_drop_pages(struct page **pages, size_t num_pages); +int btrfs_dirty_pages(struct btrfs_root *root, struct inode *inode, + struct page **pages, size_t num_pages, + loff_t pos, size_t write_bytes, + struct extent_state **cached); /* tree-defrag.c */ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans, diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index e621ea5..75899a0 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -104,7 +104,7 @@ static noinline int btrfs_copy_from_user(loff_t pos, int num_pages, /* * unlocks pages after btrfs_file_write is done with them */ -static noinline void btrfs_drop_pages(struct page **pages, size_t num_pages) +void btrfs_drop_pages(struct page **pages, size_t num_pages) { size_t i; for (i = 0; i < num_pages; i++) { @@ -127,16 +127,13 @@ static noinline void btrfs_drop_pages(struct page **pages, size_t num_pages) * this also makes the decision about creating an inline extent vs * doing real data extents, marking pages dirty and delalloc as required. */ -static noinline int dirty_and_release_pages(struct btrfs_root *root, - struct file *file, - struct page **pages, - size_t num_pages, - loff_t pos, - size_t write_bytes) +int btrfs_dirty_pages(struct btrfs_root *root, struct inode *inode, + struct page **pages, size_t num_pages, + loff_t pos, size_t write_bytes, + struct extent_state **cached) { int err = 0; int i; - struct inode *inode = fdentry(file)->d_inode; u64 num_bytes; u64 start_pos; u64 end_of_last_block; @@ -149,7 +146,7 @@ static noinline int dirty_and_release_pages(struct btrfs_root *root, end_of_last_block = start_pos + num_bytes - 1; err = btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block, - NULL); + cached); if (err) return err; @@ -992,9 +989,9 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file, } if (copied > 0) { - ret = dirty_and_release_pages(root, file, pages, - dirty_pages, pos, - copied); + ret = btrfs_dirty_pages(root, inode, pages, + dirty_pages, pos, copied, + NULL); if (ret) { btrfs_delalloc_release_space(inode, dirty_pages << PAGE_CACHE_SHIFT); diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index f561c95..a3f420d 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -508,6 +508,7 @@ int btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode; struct rb_node *node; struct list_head *pos, *n; + struct page **pages; struct page *page; struct extent_state *cached_state = NULL; struct btrfs_free_cluster *cluster = NULL; @@ -517,13 +518,13 @@ int btrfs_write_out_cache(struct btrfs_root *root, u64 start, end, len; u64 bytes = 0; u32 *crc, *checksums; - pgoff_t index = 0, last_index = 0; unsigned long first_page_offset; - int num_checksums; + int index = 0, num_pages = 0; int entries = 0; int bitmaps = 0; int ret = 0; bool next_page = false; + bool out_of_space = false; root = root->fs_info->tree_root; @@ -551,24 +552,31 @@ int btrfs_write_out_cache(struct btrfs_root *root, return 0; } - last_index = (i_size_read(inode) - 1) >> PAGE_CACHE_SHIFT; + num_pages = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> + PAGE_CACHE_SHIFT; filemap_write_and_wait(inode->i_mapping); btrfs_wait_ordered_range(inode, inode->i_size & ~(root->sectorsize - 1), (u64)-1); /* We need a checksum per page. */ - num_checksums = i_size_read(inode) / PAGE_CACHE_SIZE; - crc = checksums = kzalloc(sizeof(u32) * num_checksums, GFP_NOFS); + crc = checksums = kzalloc(sizeof(u32) * num_pages, GFP_NOFS); if (!crc) { iput(inode); return 0; } + pages = kzalloc(sizeof(struct page *) * num_pages, GFP_NOFS); + if (!pages) { + kfree(crc); + iput(inode); + return 0; + } + /* Since the first page has all of our checksums and our generation we * need to calculate the offset into the page that we can start writing * our entries. */ - first_page_offset = (sizeof(u32) * num_checksums) + sizeof(u64); + first_page_offset = (sizeof(u32) * num_pages) + sizeof(u64); /* Get the cluster for this block_group if it exists */ if (!list_empty(&block_group->cluster_list)) @@ -590,20 +598,18 @@ int btrfs_write_out_cache(struct btrfs_root *root, * after find_get_page at this point. Just putting this here so people * know and don't freak out. */ - while (index <= last_index) { + while (index < num_pages) { page = grab_cache_page(inode->i_mapping, index); if (!page) { - pgoff_t i = 0; + int i; - while (i < index) { - page = find_get_page(inode->i_mapping, i); - unlock_page(page); - page_cache_release(page); - page_cache_release(page); - i++; + for (i = 0; i < num_pages; i++) { + unlock_page(pages[i]); + page_cache_release(pages[i]); } goto out_free; } + pages[index] = page; index++; } @@ -631,7 +637,12 @@ int btrfs_write_out_cache(struct btrfs_root *root, offset = start_offset; } - page = find_get_page(inode->i_mapping, index); + if (index >= num_pages) { + out_of_space = true; + break; + } + + page = pages[index]; addr = kmap(page); entry = addr + start_offset; @@ -708,23 +719,6 @@ int btrfs_write_out_cache(struct btrfs_root *root, bytes += PAGE_CACHE_SIZE; - ClearPageChecked(page); - set_page_extent_mapped(page); - SetPageUptodate(page); - set_page_dirty(page); - - /* - * We need to release our reference we got for grab_cache_page, - * except for the first page which will hold our checksums, we - * do that below. - */ - if (index != 0) { - unlock_page(page); - page_cache_release(page); - } - - page_cache_release(page); - index++; } while (node || next_page); @@ -734,6 +728,10 @@ int btrfs_write_out_cache(struct btrfs_root *root, struct btrfs_free_space *entry = list_entry(pos, struct btrfs_free_space, list); + if (index >= num_pages) { + out_of_space = true; + break; + } page = find_get_page(inode->i_mapping, index); addr = kmap(page); @@ -745,64 +743,58 @@ int btrfs_write_out_cache(struct btrfs_root *root, crc++; bytes += PAGE_CACHE_SIZE; - ClearPageChecked(page); - set_page_extent_mapped(page); - SetPageUptodate(page); - set_page_dirty(page); - unlock_page(page); - page_cache_release(page); - page_cache_release(page); list_del_init(&entry->list); index++; } + if (out_of_space) { + btrfs_drop_pages(pages, num_pages); + unlock_extent_cached(&BTRFS_I(inode)->io_tree, 0, + i_size_read(inode) - 1, &cached_state, + GFP_NOFS); + ret = 0; + goto out_free; + } + /* Zero out the rest of the pages just to make sure */ - while (index <= last_index) { + while (index < num_pages) { void *addr; - page = find_get_page(inode->i_mapping, index); - + page = pages[index]; addr = kmap(page); memset(addr, 0, PAGE_CACHE_SIZE); kunmap(page); - ClearPageChecked(page); - set_page_extent_mapped(page); - SetPageUptodate(page); - set_page_dirty(page); - unlock_page(page); - page_cache_release(page); - page_cache_release(page); bytes += PAGE_CACHE_SIZE; index++; } - btrfs_set_extent_delalloc(inode, 0, bytes - 1, &cached_state); - /* Write the checksums and trans id to the first page */ { void *addr; u64 *gen; - page = find_get_page(inode->i_mapping, 0); + page = pages[0]; addr = kmap(page); - memcpy(addr, checksums, sizeof(u32) * num_checksums); - gen = addr + (sizeof(u32) * num_checksums); + memcpy(addr, checksums, sizeof(u32) * num_pages); + gen = addr + (sizeof(u32) * num_pages); *gen = trans->transid; kunmap(page); - ClearPageChecked(page); - set_page_extent_mapped(page); - SetPageUptodate(page); - set_page_dirty(page); - unlock_page(page); - page_cache_release(page); - page_cache_release(page); } - BTRFS_I(inode)->generation = trans->transid; + ret = btrfs_dirty_pages(root, inode, pages, num_pages, 0, + bytes, &cached_state); + btrfs_drop_pages(pages, num_pages); unlock_extent_cached(&BTRFS_I(inode)->io_tree, 0, i_size_read(inode) - 1, &cached_state, GFP_NOFS); + if (ret) { + ret = 0; + goto out_free; + } + + BTRFS_I(inode)->generation = trans->transid; + filemap_write_and_wait(inode->i_mapping); key.objectid = BTRFS_FREE_SPACE_OBJECTID; @@ -853,6 +845,7 @@ out_free: BTRFS_I(inode)->generation = 0; } kfree(checksums); + kfree(pages); btrfs_update_inode(trans, root, inode); iput(inode); return ret; -- cgit v0.10.2 From 06d5a5899d6d3ac401d2359b5eac6d2a3a0fe331 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 5 Apr 2011 11:57:27 -0400 Subject: Btrfs: only retry transaction reservation once I saw a lockup where we kept getting into this start transaction->commit transaction loop because of enospce. The fact is if we fail to make our reservation, we've tried _everything_ several times, so we only need to try and commit the transaction once, and if that doesn't work then we really are out of space and need to just exit. Thanks, Signed-off-by: Josef Bacik diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 5b158da..4583008 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -181,6 +181,7 @@ static struct btrfs_trans_handle *start_transaction(struct btrfs_root *root, { struct btrfs_trans_handle *h; struct btrfs_transaction *cur_trans; + int retries = 0; int ret; if (root->fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) @@ -224,10 +225,18 @@ again: if (num_items > 0) { ret = btrfs_trans_reserve_metadata(h, root, num_items); - if (ret == -EAGAIN) { + if (ret == -EAGAIN && !retries) { + retries++; btrfs_commit_transaction(h, root); goto again; + } else if (ret == -EAGAIN) { + /* + * We have already retried and got EAGAIN, so really we + * don't have space, so set ret to -ENOSPC. + */ + ret = -ENOSPC; } + if (ret < 0) { btrfs_end_transaction(h, root); return ERR_PTR(ret); -- cgit v0.10.2 From 12ddb96cb6752218d8a1aeb696ec9b0ca7adb42f Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 5 Apr 2011 13:02:27 -0400 Subject: Btrfs: map the inode item when doing fill_inode_item Instead of calling kmap_atomic for every thing we set in the inode item, map the entire inode item at the start and unmap it at the end. This makes a sequential dd of 400mb O_DIRECT something like 1% faster. Thanks, Signed-off-by: Josef Bacik diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index cc60228..da26802 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -2589,6 +2589,13 @@ static void fill_inode_item(struct btrfs_trans_handle *trans, struct btrfs_inode_item *item, struct inode *inode) { + if (!leaf->map_token) + map_private_extent_buffer(leaf, (unsigned long)item, + sizeof(struct btrfs_inode_item), + &leaf->map_token, &leaf->kaddr, + &leaf->map_start, &leaf->map_len, + KM_USER1); + btrfs_set_inode_uid(leaf, item, inode->i_uid); btrfs_set_inode_gid(leaf, item, inode->i_gid); btrfs_set_inode_size(leaf, item, BTRFS_I(inode)->disk_i_size); @@ -2617,6 +2624,11 @@ static void fill_inode_item(struct btrfs_trans_handle *trans, btrfs_set_inode_rdev(leaf, item, inode->i_rdev); btrfs_set_inode_flags(leaf, item, BTRFS_I(inode)->flags); btrfs_set_inode_block_group(leaf, item, BTRFS_I(inode)->block_group); + + if (leaf->map_token) { + unmap_extent_buffer(leaf, leaf->map_token, KM_USER1); + leaf->map_token = NULL; + } } /* -- cgit v0.10.2 From 1ef30be142d2cc60e2687ef267de864cf31be995 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 5 Apr 2011 19:25:36 -0400 Subject: Btrfs: do not call btrfs_update_inode in endio if nothing changed In the DIO code we often don't update the i_disk_size because the i_size isn't updated until after the DIO is completed, so basically we are allocating a path, doing a search, and updating the inode item for no reason since nothing changed. btrfs_ordered_update_i_size will return 1 if it didn't update i_disk_size, so only run btrfs_update_inode if btrfs_ordered_update_i_size returns 0. Thanks, Signed-off-by: Josef Bacik diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index da26802..4a238d6 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1769,9 +1769,12 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end) add_pending_csums(trans, inode, ordered_extent->file_offset, &ordered_extent->list); - btrfs_ordered_update_i_size(inode, 0, ordered_extent); - ret = btrfs_update_inode(trans, root, inode); - BUG_ON(ret); + ret = btrfs_ordered_update_i_size(inode, 0, ordered_extent); + if (!ret) { + ret = btrfs_update_inode(trans, root, inode); + BUG_ON(ret); + } + ret = 0; out: if (nolock) { if (trans) @@ -5865,8 +5868,10 @@ again: } add_pending_csums(trans, inode, ordered->file_offset, &ordered->list); - btrfs_ordered_update_i_size(inode, 0, ordered); - btrfs_update_inode(trans, root, inode); + ret = btrfs_ordered_update_i_size(inode, 0, ordered); + if (!ret) + btrfs_update_inode(trans, root, inode); + ret = 0; out_unlock: unlock_extent_cached(&BTRFS_I(inode)->io_tree, ordered->file_offset, ordered->file_offset + ordered->len - 1, -- cgit v0.10.2 From 02f57c7aedef1a537f4b16db7061cdd8efa3bb4e Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 6 Apr 2011 14:25:44 -0400 Subject: Btrfs: don't split dio bios if we don't have to We have been unconditionally allocating a new bio and re-adding all pages from our original bio to the new bio. This is needed if our original bio is larger than our stripe size, but if it is smaller than the stripe size then there is no need to do this. So check the map length and if we are under that then go ahead and submit the original bio. Thanks, Signed-off-by: Josef Bacik diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 4a238d6..149c77f 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6006,13 +6006,6 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, int ret = 0; int write = rw & REQ_WRITE; - bio = btrfs_dio_bio_alloc(orig_bio->bi_bdev, start_sector, GFP_NOFS); - if (!bio) - return -ENOMEM; - bio->bi_private = dip; - bio->bi_end_io = btrfs_end_dio_bio; - atomic_inc(&dip->pending_bios); - map_length = orig_bio->bi_size; ret = btrfs_map_block(map_tree, READ, start_sector << 9, &map_length, NULL, 0); @@ -6021,6 +6014,18 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, return -EIO; } + if (map_length >= orig_bio->bi_size) { + bio = orig_bio; + goto submit; + } + + bio = btrfs_dio_bio_alloc(orig_bio->bi_bdev, start_sector, GFP_NOFS); + if (!bio) + return -ENOMEM; + bio->bi_private = dip; + bio->bi_end_io = btrfs_end_dio_bio; + atomic_inc(&dip->pending_bios); + while (bvec <= (orig_bio->bi_io_vec + orig_bio->bi_vcnt - 1)) { if (unlikely(map_length < submit_len + bvec->bv_len || bio_add_page(bio, bvec->bv_page, bvec->bv_len, @@ -6071,6 +6076,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, } } +submit: ret = __btrfs_submit_dio_bio(bio, inode, rw, file_offset, skip_sum, csums); if (!ret) -- cgit v0.10.2 From 1ae399382512b3e4d6c923e53da9e45935577040 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 6 Apr 2011 14:41:34 -0400 Subject: Btrfs: do not use async submit for small DIO io's When looking at our DIO performance Chris said that for small IO's doing the async submit stuff tends to be more overhead than it's worth. With this on top of my other fixes I get about a 17-20% speedup doing a sequential dd with 4k IO's. Basically if we don't have to split the bio for the map length it's small enough to be directly submitted, otherwise go back to the async submit. Thanks, Signed-off-by: Josef Bacik diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 149c77f..2bb76c6 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -5957,7 +5957,7 @@ static struct bio *btrfs_dio_bio_alloc(struct block_device *bdev, static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode, int rw, u64 file_offset, int skip_sum, - u32 *csums) + u32 *csums, int async_submit) { int write = rw & REQ_WRITE; struct btrfs_root *root = BTRFS_I(inode)->root; @@ -5968,13 +5968,24 @@ static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode, if (ret) goto err; - if (write && !skip_sum) { + if (skip_sum) + goto map; + + if (write && async_submit) { ret = btrfs_wq_submit_bio(root->fs_info, inode, rw, bio, 0, 0, file_offset, __btrfs_submit_bio_start_direct_io, __btrfs_submit_bio_done); goto err; + } else if (write) { + /* + * If we aren't doing async submit, calculate the csum of the + * bio now. + */ + ret = btrfs_csum_one_bio(root, inode, bio, file_offset, 1); + if (ret) + goto err; } else if (!skip_sum) { ret = btrfs_lookup_bio_sums_dio(root, inode, bio, file_offset, csums); @@ -5982,7 +5993,8 @@ static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode, goto err; } - ret = btrfs_map_bio(root, rw, bio, 0, 1); +map: + ret = btrfs_map_bio(root, rw, bio, 0, async_submit); err: bio_put(bio); return ret; @@ -6004,6 +6016,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, int nr_pages = 0; u32 *csums = dip->csums; int ret = 0; + int async_submit = 0; int write = rw & REQ_WRITE; map_length = orig_bio->bi_size; @@ -6019,6 +6032,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, goto submit; } + async_submit = 1; bio = btrfs_dio_bio_alloc(orig_bio->bi_bdev, start_sector, GFP_NOFS); if (!bio) return -ENOMEM; @@ -6039,7 +6053,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, atomic_inc(&dip->pending_bios); ret = __btrfs_submit_dio_bio(bio, inode, rw, file_offset, skip_sum, - csums); + csums, async_submit); if (ret) { bio_put(bio); atomic_dec(&dip->pending_bios); @@ -6078,7 +6092,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, submit: ret = __btrfs_submit_dio_bio(bio, inode, rw, file_offset, skip_sum, - csums); + csums, async_submit); if (!ret) return 0; -- cgit v0.10.2 From 16d299ac7446b5a75c5683a9ae11d7907d444c86 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 6 Apr 2011 14:53:07 -0400 Subject: Btrfs: reuse the extent_map we found when calling btrfs_get_extent In btrfs_get_block_direct we call btrfs_get_extent to lookup the extent for the range that we are looking for. If we don't find an extent, btrfs_get_extent will insert a extent_map for that area and mark it as a hole. So it does the job of allocating a new extent map and inserting it into the io tree. But if we're creating a new extent we free it up and redo all of that work. So instead pass the em to btrfs_new_extent_direct(), and if it will work just allocate the disk space and set it up properly and bypass the freeing/allocating of a new extent map and the expensive operation of inserting the thing into the io_tree. Thanks, Signed-off-by: Josef Bacik diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 2bb76c6..24310c9 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -5445,17 +5445,30 @@ out: } static struct extent_map *btrfs_new_extent_direct(struct inode *inode, + struct extent_map *em, u64 start, u64 len) { struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_trans_handle *trans; - struct extent_map *em; struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree; struct btrfs_key ins; u64 alloc_hint; int ret; + bool insert = false; - btrfs_drop_extent_cache(inode, start, start + len - 1, 0); + /* + * Ok if the extent map we looked up is a hole and is for the exact + * range we want, there is no reason to allocate a new one, however if + * it is not right then we need to free this one and drop the cache for + * our range. + */ + if (em->block_start != EXTENT_MAP_HOLE || em->start != start || + em->len != len) { + free_extent_map(em); + em = NULL; + insert = true; + btrfs_drop_extent_cache(inode, start, start + len - 1, 0); + } trans = btrfs_join_transaction(root, 0); if (IS_ERR(trans)) @@ -5471,10 +5484,12 @@ static struct extent_map *btrfs_new_extent_direct(struct inode *inode, goto out; } - em = alloc_extent_map(GFP_NOFS); if (!em) { - em = ERR_PTR(-ENOMEM); - goto out; + em = alloc_extent_map(GFP_NOFS); + if (!em) { + em = ERR_PTR(-ENOMEM); + goto out; + } } em->start = start; @@ -5484,9 +5499,15 @@ static struct extent_map *btrfs_new_extent_direct(struct inode *inode, em->block_start = ins.objectid; em->block_len = ins.offset; em->bdev = root->fs_info->fs_devices->latest_bdev; + + /* + * We need to do this because if we're using the original em we searched + * for, we could have EXTENT_FLAG_VACANCY set, and we don't want that. + */ + em->flags = 0; set_bit(EXTENT_FLAG_PINNED, &em->flags); - while (1) { + while (insert) { write_lock(&em_tree->lock); ret = add_extent_mapping(em_tree, em); write_unlock(&em_tree->lock); @@ -5704,8 +5725,7 @@ must_cow: * it above */ len = bh_result->b_size; - free_extent_map(em); - em = btrfs_new_extent_direct(inode, start, len); + em = btrfs_new_extent_direct(inode, em, start, len); if (IS_ERR(em)) return PTR_ERR(em); len = min(len, em->len - (start - em->start)); -- cgit v0.10.2 From 93a54bc4c28a125978cddbe2db9e347391e3522d Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Wed, 6 Apr 2011 15:11:44 -0400 Subject: Btrfs: check for duplicate iov_base's when doing dio reads Apparently it is ok to submit a read to an IDE device with the same target page for different offsets. This is what Windows does under qemu. The problem is under DIO we expect them to be different buffers for checksumming reasons, and so this sort of thing will result in checksum errors, when in reality the file is fine. So when reading, check to make sure that all iov bases are different, and if they aren't fall back to buffered mode, since that will work out right. Thanks, Signed-off-by: Josef Bacik diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 24310c9..00d59c6 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6207,6 +6207,7 @@ static ssize_t check_direct_IO(struct btrfs_root *root, int rw, struct kiocb *io unsigned long nr_segs) { int seg; + int i; size_t size; unsigned long addr; unsigned blocksize_mask = root->sectorsize - 1; @@ -6221,8 +6222,22 @@ static ssize_t check_direct_IO(struct btrfs_root *root, int rw, struct kiocb *io addr = (unsigned long)iov[seg].iov_base; size = iov[seg].iov_len; end += size; - if ((addr & blocksize_mask) || (size & blocksize_mask)) + if ((addr & blocksize_mask) || (size & blocksize_mask)) goto out; + + /* If this is a write we don't need to check anymore */ + if (rw & WRITE) + continue; + + /* + * Check to make sure we don't have duplicate iov_base's in this + * iovec, if so return EINVAL, otherwise we'll get csum errors + * when reading back. + */ + for (i = seg + 1; i < nr_segs; i++) { + if (iov[seg].iov_base == iov[i].iov_base) + goto out; + } } retval = 0; out: -- cgit v0.10.2 From 96f372c95d32f76fa2b0e035e0a6269234bfda09 Mon Sep 17 00:00:00 2001 From: Felix Fietkau Date: Thu, 7 Apr 2011 19:07:17 +0200 Subject: ath9k: fix missing ath9k_ps_wakeup/ath9k_ps_restore calls These missing chip wakeups mainly cause crashes on AR5416 cards in MIPS boards, but have also been reported to cause radio stability issues on AR9285. Signed-off-by: Felix Fietkau Cc: stable@kernel.org Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c index dddb85d..17d04ff 100644 --- a/drivers/net/wireless/ath/ath9k/main.c +++ b/drivers/net/wireless/ath/ath9k/main.c @@ -1376,7 +1376,6 @@ static void ath9k_calculate_summary_state(struct ieee80211_hw *hw, ath9k_calculate_iter_data(hw, vif, &iter_data); - ath9k_ps_wakeup(sc); /* Set BSSID mask. */ memcpy(common->bssidmask, iter_data.mask, ETH_ALEN); ath_hw_setbssidmask(common); @@ -1411,7 +1410,6 @@ static void ath9k_calculate_summary_state(struct ieee80211_hw *hw, } ath9k_hw_set_interrupts(ah, ah->imask); - ath9k_ps_restore(sc); /* Set up ANI */ if ((iter_data.naps + iter_data.nadhocs) > 0) { @@ -1457,6 +1455,7 @@ static int ath9k_add_interface(struct ieee80211_hw *hw, struct ath_vif *avp = (void *)vif->drv_priv; int ret = 0; + ath9k_ps_wakeup(sc); mutex_lock(&sc->mutex); switch (vif->type) { @@ -1503,6 +1502,7 @@ static int ath9k_add_interface(struct ieee80211_hw *hw, ath9k_do_vif_add_setup(hw, vif); out: mutex_unlock(&sc->mutex); + ath9k_ps_restore(sc); return ret; } @@ -1517,6 +1517,7 @@ static int ath9k_change_interface(struct ieee80211_hw *hw, ath_dbg(common, ATH_DBG_CONFIG, "Change Interface\n"); mutex_lock(&sc->mutex); + ath9k_ps_wakeup(sc); /* See if new interface type is valid. */ if ((new_type == NL80211_IFTYPE_ADHOC) && @@ -1546,6 +1547,7 @@ static int ath9k_change_interface(struct ieee80211_hw *hw, ath9k_do_vif_add_setup(hw, vif); out: + ath9k_ps_restore(sc); mutex_unlock(&sc->mutex); return ret; } @@ -1558,6 +1560,7 @@ static void ath9k_remove_interface(struct ieee80211_hw *hw, ath_dbg(common, ATH_DBG_CONFIG, "Detach Interface\n"); + ath9k_ps_wakeup(sc); mutex_lock(&sc->mutex); sc->nvifs--; @@ -1569,6 +1572,7 @@ static void ath9k_remove_interface(struct ieee80211_hw *hw, ath9k_calculate_summary_state(hw, NULL); mutex_unlock(&sc->mutex); + ath9k_ps_restore(sc); } static void ath9k_enable_ps(struct ath_softc *sc) @@ -1809,6 +1813,7 @@ static int ath9k_conf_tx(struct ieee80211_hw *hw, u16 queue, txq = sc->tx.txq_map[queue]; + ath9k_ps_wakeup(sc); mutex_lock(&sc->mutex); memset(&qi, 0, sizeof(struct ath9k_tx_queue_info)); @@ -1832,6 +1837,7 @@ static int ath9k_conf_tx(struct ieee80211_hw *hw, u16 queue, ath_beaconq_config(sc); mutex_unlock(&sc->mutex); + ath9k_ps_restore(sc); return ret; } @@ -1894,6 +1900,7 @@ static void ath9k_bss_info_changed(struct ieee80211_hw *hw, int slottime; int error; + ath9k_ps_wakeup(sc); mutex_lock(&sc->mutex); if (changed & BSS_CHANGED_BSSID) { @@ -1994,6 +2001,7 @@ static void ath9k_bss_info_changed(struct ieee80211_hw *hw, } mutex_unlock(&sc->mutex); + ath9k_ps_restore(sc); } static u64 ath9k_get_tsf(struct ieee80211_hw *hw) -- cgit v0.10.2 From a6756da9eace8b4af73e9dea43f1fc2889224c94 Mon Sep 17 00:00:00 2001 From: Jason Conti Date: Thu, 7 Apr 2011 21:09:57 +0200 Subject: p54: Initialize extra_len in p54_tx_80211 This patch fixes a very serious off-by-one bug in the driver, which could leave the device in an unresponsive state. The problem was that the extra_len variable [used to reserve extra scratch buffer space for the firmware] was left uninitialized. Because p54_assign_address later needs the value to reserve additional space, the resulting frame could be to big for the small device's memory window and everything would immediately come to a grinding halt. Reference: https://bugs.launchpad.net/bugs/722185 Cc: Acked-by: Christian Lamparter Signed-off-by: Jason Conti Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/p54/txrx.c b/drivers/net/wireless/p54/txrx.c index 7834c26..042842e 100644 --- a/drivers/net/wireless/p54/txrx.c +++ b/drivers/net/wireless/p54/txrx.c @@ -703,7 +703,7 @@ void p54_tx_80211(struct ieee80211_hw *dev, struct sk_buff *skb) struct p54_tx_info *p54info; struct p54_hdr *hdr; struct p54_tx_data *txhdr; - unsigned int padding, len, extra_len; + unsigned int padding, len, extra_len = 0; int i, j, ridx; u16 hdr_flags = 0, aid = 0; u8 rate, queue = 0, crypt_offset = 0; -- cgit v0.10.2 From f9fa0bc1fabe1d861e46d80ecbe7e85da359195c Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Fri, 8 Apr 2011 10:53:46 -0700 Subject: signal.c: fix erroneous syscall kernel-doc Fix erroneous syscall kernel-doc comments in kernel/signal.c. Reported-by: Matt Fleming Signed-off-by: Randy Dunlap Signed-off-by: Linus Torvalds diff --git a/kernel/signal.c b/kernel/signal.c index 29e233f..7165af5 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -2711,8 +2711,8 @@ out: /** * sys_rt_sigaction - alter an action taken by a process * @sig: signal to be sent - * @act: the thread group ID of the thread - * @oact: the PID of the thread + * @act: new sigaction + * @oact: used to save the previous sigaction * @sigsetsize: size of sigset_t type */ SYSCALL_DEFINE4(rt_sigaction, int, sig, -- cgit v0.10.2 From 0a0883c8433c743dad0a4d9ebe2717558f2c209e Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Tue, 5 Apr 2011 17:57:23 +1000 Subject: i915: restore only the mode of this driver on lastclose This has always used a big hammer, but that hammer is probably too big, I'm also not sure its necessary but at least this should be safe. Should fix: https://bugzilla.kernel.org/show_bug.cgi?id=23592 Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c index 7273037..12876f2 100644 --- a/drivers/gpu/drm/i915/i915_dma.c +++ b/drivers/gpu/drm/i915/i915_dma.c @@ -2207,7 +2207,7 @@ void i915_driver_lastclose(struct drm_device * dev) drm_i915_private_t *dev_priv = dev->dev_private; if (!dev_priv || drm_core_check_feature(dev, DRIVER_MODESET)) { - drm_fb_helper_restore(); + intel_fb_restore_mode(dev); vga_switcheroo_process_delayed_switch(); return; } diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index f5b0d83..1d20712 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -338,4 +338,5 @@ extern int intel_overlay_attrs(struct drm_device *dev, void *data, struct drm_file *file_priv); extern void intel_fb_output_poll_changed(struct drm_device *dev); +extern void intel_fb_restore_mode(struct drm_device *dev); #endif /* __INTEL_DRV_H__ */ diff --git a/drivers/gpu/drm/i915/intel_fb.c b/drivers/gpu/drm/i915/intel_fb.c index 5127827..96a45c4 100644 --- a/drivers/gpu/drm/i915/intel_fb.c +++ b/drivers/gpu/drm/i915/intel_fb.c @@ -264,3 +264,19 @@ void intel_fb_output_poll_changed(struct drm_device *dev) drm_i915_private_t *dev_priv = dev->dev_private; drm_fb_helper_hotplug_event(&dev_priv->fbdev->helper); } + +void intel_fb_restore_mode(struct drm_device *dev) +{ + drm_i915_private_t *dev_priv = dev->dev_private; + int ret, i; + + if (!dev_priv->fbdev) + return; + + for (i = 0; i < dev_priv->fbdev->helper.crtc_count; i++) { + struct drm_mode_set *mode_set = &dev_priv->fbdev->helper.crtc_info[i].mode_set; + ret = drm_crtc_helper_set_config(mode_set); + if (ret) + DRM_DEBUG("failed to restore crtc mode\n"); + } +} -- cgit v0.10.2 From dc66b325f161bb651493c7d96ad44876b629cf6a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Michel=20D=C3=A4nzer?= Date: Thu, 7 Apr 2011 16:17:47 +0200 Subject: radeon: Fix KMS CP writeback on big endian machines. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This is necessary even with PCI(e) GART, and it makes writeback work even with AGP on my PowerBook. Might still be unreliable with older revisions of UniNorth and other AGP bridges though. Signed-off-by: Michel Dänzer Reviewed-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c index 9e59868..bbcd1dd 100644 --- a/drivers/gpu/drm/radeon/radeon_fence.c +++ b/drivers/gpu/drm/radeon/radeon_fence.c @@ -79,7 +79,7 @@ static bool radeon_fence_poll_locked(struct radeon_device *rdev) scratch_index = R600_WB_EVENT_OFFSET + rdev->fence_drv.scratch_reg - rdev->scratch.reg_base; else scratch_index = RADEON_WB_SCRATCH_OFFSET + rdev->fence_drv.scratch_reg - rdev->scratch.reg_base; - seq = rdev->wb.wb[scratch_index/4]; + seq = le32_to_cpu(rdev->wb.wb[scratch_index/4]); } else seq = RREG32(rdev->fence_drv.scratch_reg); if (seq != rdev->fence_drv.last_seq) { diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c index bbc9cd8..c6776e4 100644 --- a/drivers/gpu/drm/radeon/radeon_ring.c +++ b/drivers/gpu/drm/radeon/radeon_ring.c @@ -248,7 +248,7 @@ void radeon_ib_pool_fini(struct radeon_device *rdev) void radeon_ring_free_size(struct radeon_device *rdev) { if (rdev->wb.enabled) - rdev->cp.rptr = rdev->wb.wb[RADEON_WB_CP_RPTR_OFFSET/4]; + rdev->cp.rptr = le32_to_cpu(rdev->wb.wb[RADEON_WB_CP_RPTR_OFFSET/4]); else { if (rdev->family >= CHIP_R600) rdev->cp.rptr = RREG32(R600_CP_RB_RPTR); -- cgit v0.10.2 From 88a2b75cceba1fdbc1fdb70c7c8cd2b40da3d95d Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Michel=20D=C3=A4nzer?= Date: Thu, 7 Apr 2011 16:20:49 +0200 Subject: drm/radeon: Fix KMS legacy backlight support if CONFIG_BACKLIGHT_CLASS_DEVICE=m. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Michel Dänzer Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_legacy_encoders.c b/drivers/gpu/drm/radeon/radeon_legacy_encoders.c index 5b54268..2f46e0c 100644 --- a/drivers/gpu/drm/radeon/radeon_legacy_encoders.c +++ b/drivers/gpu/drm/radeon/radeon_legacy_encoders.c @@ -269,7 +269,7 @@ static const struct drm_encoder_helper_funcs radeon_legacy_lvds_helper_funcs = { .disable = radeon_legacy_encoder_disable, }; -#ifdef CONFIG_BACKLIGHT_CLASS_DEVICE +#if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) || defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE) #define MAX_RADEON_LEVEL 0xFF -- cgit v0.10.2 From 9bb09fa1b5b07459279301ac6220d575f307597b Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Thu, 7 Apr 2011 10:31:25 -0400 Subject: drm/radeon/kms: pll tweaks for rv6xx Prefer minm over maxp. Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=35994 Signed-off-by: Alex Deucher Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index b41ec59..9d516a8 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -531,6 +531,12 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, pll->flags |= RADEON_PLL_PREFER_HIGH_FB_DIV; else pll->flags |= RADEON_PLL_PREFER_LOW_REF_DIV; + + if ((rdev->family == CHIP_R600) || + (rdev->family == CHIP_RV610) || + (rdev->family == CHIP_RV630) || + (rdev->family == CHIP_RV670)) + pll->flags |= RADEON_PLL_PREFER_MINM_OVER_MAXP; } else { pll->flags |= RADEON_PLL_LEGACY; -- cgit v0.10.2 From d4864d604a8d29a0d6a13a8d882be6e21c498c22 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Wed, 6 Apr 2011 13:44:10 -0400 Subject: drm/radeon/kms: make radeon i2c put/get bytes less noisy Switch some errors to debug output. These are generally harmless and tend to confuse users. Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_i2c.c b/drivers/gpu/drm/radeon/radeon_i2c.c index ded2a45..ccbabf7 100644 --- a/drivers/gpu/drm/radeon/radeon_i2c.c +++ b/drivers/gpu/drm/radeon/radeon_i2c.c @@ -1062,7 +1062,7 @@ void radeon_i2c_get_byte(struct radeon_i2c_chan *i2c_bus, *val = in_buf[0]; DRM_DEBUG("val = 0x%02x\n", *val); } else { - DRM_ERROR("i2c 0x%02x 0x%02x read failed\n", + DRM_DEBUG("i2c 0x%02x 0x%02x read failed\n", addr, *val); } } @@ -1084,7 +1084,7 @@ void radeon_i2c_put_byte(struct radeon_i2c_chan *i2c_bus, out_buf[1] = val; if (i2c_transfer(&i2c_bus->adapter, &msg, 1) != 1) - DRM_ERROR("i2c 0x%02x 0x%02x write failed\n", + DRM_DEBUG("i2c 0x%02x 0x%02x write failed\n", addr, val); } -- cgit v0.10.2 From 23fcf2ec93fb8573a653408316af599939ff9a8e Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Mon, 28 Mar 2011 15:15:09 +0800 Subject: nfsd4: fix oops on lock failure Lock stateid's can have access_bmap 0 if they were only partially initialized (due to a failed lock request); handle that case in free_generic_stateid. ------------[ cut here ]------------ kernel BUG at fs/nfsd/nfs4state.c:380! invalid opcode: 0000 [#1] SMP last sysfs file: /sys/kernel/mm/ksm/run Modules linked in: nfs fscache md4 nls_utf8 cifs ip6table_filter ip6_tables ebtable_nat ebtables ipt_MASQUERADE iptable_nat nf_nat bridge stp llc nfsd lockd nfs_acl auth_rpcgss sunrpc ipv6 ppdev parport_pc parport pcnet32 mii pcspkr microcode i2c_piix4 BusLogic floppy [last unloaded: mperf] Pid: 1468, comm: nfsd Not tainted 2.6.38+ #120 VMware, Inc. VMware Virtual Platform/440BX Desktop Reference Platform EIP: 0060:[] EFLAGS: 00010297 CPU: 0 EIP is at nfs4_access_to_omode+0x1c/0x29 [nfsd] EAX: ffffffff EBX: dd758120 ECX: 00000000 EDX: 00000004 ESI: dd758120 EDI: ddfe657c EBP: dd54dde0 ESP: dd54dde0 DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 Process nfsd (pid: 1468, ti=dd54c000 task=ddc92580 task.ti=dd54c000) Stack: dd54ddf0 e24f19ca 00000000 ddfe6560 dd54de08 e24f1a5d dd758130 deee3a20 ddfe6560 31270000 dd54df1c e24f52fd 0000000f dd758090 e2505dd0 0be304cf dbb51d68 0000000e ddfe657c ddcd8020 dd758130 dd758128 dd7580d8 dd54de68 Call Trace: [] free_generic_stateid+0x1c/0x3e [nfsd] [] release_lockowner+0x71/0x8a [nfsd] [] nfsd4_lock+0x617/0x66c [nfsd] [] ? nfsd_setuser+0x199/0x1bb [nfsd] [] ? nfsd_setuser_and_check_port+0x65/0x81 [nfsd] [] ? _cond_resched+0x8/0x1c [] ? slab_pre_alloc_hook.clone.33+0x23/0x27 [] ? kmem_cache_alloc+0x1a/0xd2 [] ? __call_rcu+0xd7/0xdd [] ? fh_verify+0x401/0x452 [nfsd] [] ? nfsd4_encode_operation+0x52/0x117 [nfsd] [] ? nfsd4_putfh+0x33/0x3b [nfsd] [] ? nfsd4_delegreturn+0xd4/0xd4 [nfsd] [] nfsd4_proc_compound+0x1ea/0x33e [nfsd] [] nfsd_dispatch+0xd1/0x1a5 [nfsd] [] svc_process_common+0x282/0x46f [sunrpc] [] svc_process+0xdc/0xfa [sunrpc] [] nfsd+0xd6/0x115 [nfsd] [] ? nfsd_shutdown+0x24/0x24 [nfsd] [] kthread+0x62/0x67 [] ? kthread_worker_fn+0x114/0x114 [] kernel_thread_helper+0x6/0x10 Code: eb 05 b8 00 00 27 4f 8d 65 f4 5b 5e 5f 5d c3 83 e0 03 55 83 f8 02 89 e5 74 17 83 f8 03 74 05 48 75 09 eb 09 b8 02 00 00 00 eb 0b <0f> 0b 31 c0 eb 05 b8 01 00 00 00 5d c3 55 89 e5 57 56 89 d6 8d EIP: [] nfs4_access_to_omode+0x1c/0x29 [nfsd] SS:ESP 0068:dd54dde0 ---[ end trace 2b0bf6c6557cb284 ]--- The trace route is: -> nfsd4_lock() -> if (lock->lk_is_new) { -> alloc_init_lock_stateid() 3739: stp->st_access_bmap = 0; ->if (status && lock->lk_is_new && lock_sop) -> release_lockowner() -> free_generic_stateid() -> nfs4_access_bmap_to_omode() -> nfs4_access_to_omode() 380: BUG(); ***** This problem was introduced by 0997b173609b9229ece28941c118a2a9b278796e. Reported-by: Mi Jinlong Tested-by: Mi Jinlong Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index fbde6f7..8e3c407 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -397,10 +397,13 @@ static void unhash_generic_stateid(struct nfs4_stateid *stp) static void free_generic_stateid(struct nfs4_stateid *stp) { - int oflag = nfs4_access_bmap_to_omode(stp); + int oflag; - nfs4_file_put_access(stp->st_file, oflag); - put_nfs4_file(stp->st_file); + if (stp->st_access_bmap) { + oflag = nfs4_access_bmap_to_omode(stp); + nfs4_file_put_access(stp->st_file, oflag); + put_nfs4_file(stp->st_file); + } kmem_cache_free(stateid_slab, stp); } -- cgit v0.10.2 From e89c0d7090c54d7b11b9b091e495a1ae345dd3ff Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Tue, 5 Apr 2011 13:57:53 +0100 Subject: ARM: 6864/1: hw_breakpoint: clear DBGVCR out of reset The DBGVCR, used for configuring vector catch debug events, is UNKNOWN out of reset on ARMv7. When enabling monitor mode, this must be zeroed to avoid UNPREDICTABLE behaviour. This patch adds the zeroing code to the debug reset path. Cc: stable Reported-by: Stepan Moskovchenko Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/kernel/hw_breakpoint.c b/arch/arm/kernel/hw_breakpoint.c index 8dbc126..87acc25 100644 --- a/arch/arm/kernel/hw_breakpoint.c +++ b/arch/arm/kernel/hw_breakpoint.c @@ -868,6 +868,13 @@ static void reset_ctrl_regs(void *info) */ asm volatile("mcr p14, 0, %0, c1, c0, 4" : : "r" (0)); isb(); + + /* + * Clear any configured vector-catch events before + * enabling monitor mode. + */ + asm volatile("mcr p14, 0, %0, c0, c7, 0" : : "r" (0)); + isb(); } if (enable_monitor_mode()) -- cgit v0.10.2 From 6759788b944139793bffa889761cc3d8d703fdc0 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Tue, 5 Apr 2011 14:01:24 +0100 Subject: ARM: 6865/1: perf: ensure pass through zero is counted on overflow Commit a737823d ("ARM: perf: ensure overflows aren't missed due to IRQ latency") changed the way that event deltas are calculated on overflow so that we don't miss events when the new count value overtakes the previous one. Unfortunately, we forget to count the event that passes through zero so we end up being off by 1. This patch adds on the correction. Reported-by: Chris Moore Signed-off-by: Will Deacon Signed-off-by: Russell King diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c index 69cfee0..979da39 100644 --- a/arch/arm/kernel/perf_event.c +++ b/arch/arm/kernel/perf_event.c @@ -221,7 +221,7 @@ again: prev_raw_count &= armpmu->max_period; if (overflow) - delta = armpmu->max_period - prev_raw_count + new_raw_count; + delta = armpmu->max_period - prev_raw_count + new_raw_count + 1; else delta = new_raw_count - prev_raw_count; -- cgit v0.10.2 From 974508262e94b567f9d5b7ba1eef9fc493561f63 Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Wed, 6 Apr 2011 16:14:25 +0100 Subject: ARM: 6866/1: Do not restrict HIGHPTE to !OUTER_CACHE The HIGHPTE config option depends on !OUTER_CACHE. However, there is no set_pte_ext() function that does outer cache maintenance by physical address, hence no need for such restriction. Signed-off-by: Catalin Marinas Signed-off-by: Russell King diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 9954c9b..e8cbecc 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -1540,7 +1540,6 @@ config HIGHMEM config HIGHPTE bool "Allocate 2nd-level pagetables from highmem" depends on HIGHMEM - depends on !OUTER_CACHE config HW_PERF_EVENTS bool "Enable hardware performance counter support for perf events" -- cgit v0.10.2 From 2e82669acf03e5bf2080f5d3ef005168e67d8a51 Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Wed, 6 Apr 2011 16:16:29 +0100 Subject: ARM: 6867/1: Introduce THREAD_NOTIFY_COPY for copy_thread() hooks This patch adds THREAD_NOTIFY_COPY for calling registered handlers during the copy_thread() function call. It also changes the VFP handler to use a switch statement rather than if..else and ignore this event. Signed-off-by: Catalin Marinas Signed-off-by: Russell King diff --git a/arch/arm/include/asm/thread_notify.h b/arch/arm/include/asm/thread_notify.h index c4391ba..1dc9806 100644 --- a/arch/arm/include/asm/thread_notify.h +++ b/arch/arm/include/asm/thread_notify.h @@ -43,6 +43,7 @@ static inline void thread_notify(unsigned long rc, struct thread_info *thread) #define THREAD_NOTIFY_FLUSH 0 #define THREAD_NOTIFY_EXIT 1 #define THREAD_NOTIFY_SWITCH 2 +#define THREAD_NOTIFY_COPY 3 #endif #endif diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c index 94bbedb..5e1e541 100644 --- a/arch/arm/kernel/process.c +++ b/arch/arm/kernel/process.c @@ -372,6 +372,8 @@ copy_thread(unsigned long clone_flags, unsigned long stack_start, if (clone_flags & CLONE_SETTLS) thread->tp_value = regs->ARM_r3; + thread_notify(THREAD_NOTIFY_COPY, thread); + return 0; } diff --git a/arch/arm/vfp/vfpmodule.c b/arch/arm/vfp/vfpmodule.c index bbf3da0..1c88bbd 100644 --- a/arch/arm/vfp/vfpmodule.c +++ b/arch/arm/vfp/vfpmodule.c @@ -104,12 +104,17 @@ static void vfp_thread_exit(struct thread_info *thread) static int vfp_notifier(struct notifier_block *self, unsigned long cmd, void *v) { struct thread_info *thread = v; + u32 fpexc; +#ifdef CONFIG_SMP + unsigned int cpu; +#endif - if (likely(cmd == THREAD_NOTIFY_SWITCH)) { - u32 fpexc = fmrx(FPEXC); + switch (cmd) { + case THREAD_NOTIFY_SWITCH: + fpexc = fmrx(FPEXC); #ifdef CONFIG_SMP - unsigned int cpu = thread->cpu; + cpu = thread->cpu; /* * On SMP, if VFP is enabled, save the old state in @@ -134,13 +139,16 @@ static int vfp_notifier(struct notifier_block *self, unsigned long cmd, void *v) * old state. */ fmxr(FPEXC, fpexc & ~FPEXC_EN); - return NOTIFY_DONE; - } + break; - if (cmd == THREAD_NOTIFY_FLUSH) + case THREAD_NOTIFY_FLUSH: vfp_thread_flush(thread); - else + break; + + case THREAD_NOTIFY_EXIT: vfp_thread_exit(thread); + break; + } return NOTIFY_DONE; } -- cgit v0.10.2 From c98c09773d80db93cae349f0496fef109feab54d Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Wed, 6 Apr 2011 16:17:17 +0100 Subject: ARM: 6868/1: Preserve the VFP state during fork VFP registers d16-d31 are callee saved registers and must be preserved during function calls, including fork(). The VFP configuration should also be preserved. The patch copies the full VFP state to the child process. Reported-by: Paul Wright Signed-off-by: Catalin Marinas Signed-off-by: Russell King diff --git a/arch/arm/vfp/vfpmodule.c b/arch/arm/vfp/vfpmodule.c index 1c88bbd..f746950 100644 --- a/arch/arm/vfp/vfpmodule.c +++ b/arch/arm/vfp/vfpmodule.c @@ -78,6 +78,14 @@ static void vfp_thread_exit(struct thread_info *thread) put_cpu(); } +static void vfp_thread_copy(struct thread_info *thread) +{ + struct thread_info *parent = current_thread_info(); + + vfp_sync_hwstate(parent); + thread->vfpstate = parent->vfpstate; +} + /* * When this function is called with the following 'cmd's, the following * is true while this function is being run: @@ -148,6 +156,10 @@ static int vfp_notifier(struct notifier_block *self, unsigned long cmd, void *v) case THREAD_NOTIFY_EXIT: vfp_thread_exit(thread); break; + + case THREAD_NOTIFY_COPY: + vfp_thread_copy(thread); + break; } return NOTIFY_DONE; -- cgit v0.10.2 From aec995900fbc8cffa9f0f9e797ef07a0beb2b079 Mon Sep 17 00:00:00 2001 From: "Justin P. Mattock" Date: Fri, 8 Apr 2011 16:46:12 +0100 Subject: ARM: 6872/1: arch:common:Makefile Remove unused config in the Makefile. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The patch below removes an unused config variable found by using a kernel cleanup script. Signed-off-by: Justin P. Mattock Acked-by: Uwe Kleine-König Acked-by: Jean-Christophe PLAGNIOL-VILLARD Signed-off-by: Russell King diff --git a/arch/arm/common/Makefile b/arch/arm/common/Makefile index e7521bca..6ea9b6f 100644 --- a/arch/arm/common/Makefile +++ b/arch/arm/common/Makefile @@ -16,5 +16,4 @@ obj-$(CONFIG_SHARP_SCOOP) += scoop.o obj-$(CONFIG_ARCH_IXP2000) += uengine.o obj-$(CONFIG_ARCH_IXP23XX) += uengine.o obj-$(CONFIG_PCI_HOST_ITE8152) += it8152.o -obj-$(CONFIG_COMMON_CLKDEV) += clkdev.o obj-$(CONFIG_ARM_TIMER_SP804) += timer-sp.o -- cgit v0.10.2 From 88edaa415966af965bb7eb7056d8b58145462c8e Mon Sep 17 00:00:00 2001 From: Steve Glendinning Date: Sun, 10 Apr 2011 18:59:27 -0700 Subject: net: Add support for SMSC LAN9530, LAN9730 and LAN89530 This patch adds support for SMSC's LAN9530, LAN9730 and LAN89530 USB ethernet controllers to the existing smsc95xx driver by adding their new USB VID/PID pairs. Signed-off-by: Steve Glendinning Signed-off-by: David S. Miller diff --git a/drivers/net/usb/smsc95xx.c b/drivers/net/usb/smsc95xx.c index 727874d..47a6c87 100644 --- a/drivers/net/usb/smsc95xx.c +++ b/drivers/net/usb/smsc95xx.c @@ -1313,6 +1313,21 @@ static const struct usb_device_id products[] = { USB_DEVICE(0x0424, 0x9909), .driver_info = (unsigned long) &smsc95xx_info, }, + { + /* SMSC LAN9530 USB Ethernet Device */ + USB_DEVICE(0x0424, 0x9530), + .driver_info = (unsigned long) &smsc95xx_info, + }, + { + /* SMSC LAN9730 USB Ethernet Device */ + USB_DEVICE(0x0424, 0x9730), + .driver_info = (unsigned long) &smsc95xx_info, + }, + { + /* SMSC LAN89530 USB Ethernet Device */ + USB_DEVICE(0x0424, 0x9E08), + .driver_info = (unsigned long) &smsc95xx_info, + }, { }, /* END */ }; MODULE_DEVICE_TABLE(usb, products); -- cgit v0.10.2 From 0893ed458b4b1d7c7667ca7ffb8b11febe7e7e6c Mon Sep 17 00:00:00 2001 From: Curt Wohlgemuth Date: Sun, 10 Apr 2011 22:05:31 -0400 Subject: ext4: sync the directory inode in ext4_sync_parent() ext4 has taken the stance that, in the absence of a journal, when an fsync/fdatasync of an inode is done, the parent directory should be sync'ed if this inode entry is new. ext4_sync_parent(), which implements this, does indeed sync the dirent pages for parent directories, but it does not sync the directory *inode*. This patch fixes this. Also now return error status from ext4_sync_parent(). I tested this using a power fail test, which panics a machine running a file server getting requests from a client. Without this patch, on about every other test run, the server is missing many, many files that had been synced. With this patch, on > 6 runs, I see zero files being lost. Google-Bug-Id: 4179519 Signed-off-by: Curt Wohlgemuth Signed-off-by: "Theodore Ts'o" diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c index 7f74019..b1f9b5f 100644 --- a/fs/ext4/fsync.c +++ b/fs/ext4/fsync.c @@ -125,9 +125,11 @@ extern int ext4_flush_completed_IO(struct inode *inode) * the parent directory's parent as well, and so on recursively, if * they are also freshly created. */ -static void ext4_sync_parent(struct inode *inode) +static int ext4_sync_parent(struct inode *inode) { + struct writeback_control wbc; struct dentry *dentry = NULL; + int ret = 0; while (inode && ext4_test_inode_state(inode, EXT4_STATE_NEWENTRY)) { ext4_clear_inode_state(inode, EXT4_STATE_NEWENTRY); @@ -136,8 +138,17 @@ static void ext4_sync_parent(struct inode *inode) if (!dentry || !dentry->d_parent || !dentry->d_parent->d_inode) break; inode = dentry->d_parent->d_inode; - sync_mapping_buffers(inode->i_mapping); + ret = sync_mapping_buffers(inode->i_mapping); + if (ret) + break; + memset(&wbc, 0, sizeof(wbc)); + wbc.sync_mode = WB_SYNC_ALL; + wbc.nr_to_write = 0; /* only write out the inode */ + ret = sync_inode(inode, &wbc); + if (ret) + break; } + return ret; } /* @@ -176,7 +187,7 @@ int ext4_sync_file(struct file *file, int datasync) if (!journal) { ret = generic_file_fsync(file, datasync); if (!ret && !list_empty(&inode->i_dentry)) - ext4_sync_parent(inode); + ret = ext4_sync_parent(inode); goto out; } -- cgit v0.10.2 From be4f27d324e8ddd57cc0d4d604fe85ee0425cba9 Mon Sep 17 00:00:00 2001 From: Yongqiang Yang Date: Sun, 10 Apr 2011 22:06:07 -0400 Subject: ext4: allow an active handle to be started when freezing ext4_journal_start_sb() should not prevent an active handle from being started due to s_frozen. Otherwise, deadlock is easy to happen, below is a situation. ================================================ freeze | truncate ================================================ | ext4_ext_truncate() freeze_super() | starts a handle sets s_frozen | | ext4_ext_truncate() | holds i_data_sem ext4_freeze() | waits for updates | | ext4_free_blocks() | calls dquot_free_block() | | dquot_free_blocks() | calls ext4_dirty_inode() | | ext4_dirty_inode() | trys to start an active | handle | | block due to s_frozen ================================================ Signed-off-by: Yongqiang Yang Signed-off-by: "Theodore Ts'o" Reported-by: Amir Goldstein Reviewed-by: Jan Kara Reviewed-by: Andreas Dilger diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 551cb8e..7b636ce 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -242,27 +242,44 @@ static void ext4_put_nojournal(handle_t *handle) * journal_end calls result in the superblock being marked dirty, so * that sync() will call the filesystem's write_super callback if * appropriate. + * + * To avoid j_barrier hold in userspace when a user calls freeze(), + * ext4 prevents a new handle from being started by s_frozen, which + * is in an upper layer. */ handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks) { journal_t *journal; + handle_t *handle; if (sb->s_flags & MS_RDONLY) return ERR_PTR(-EROFS); - vfs_check_frozen(sb, SB_FREEZE_TRANS); - /* Special case here: if the journal has aborted behind our - * backs (eg. EIO in the commit thread), then we still need to - * take the FS itself readonly cleanly. */ journal = EXT4_SB(sb)->s_journal; - if (journal) { - if (is_journal_aborted(journal)) { - ext4_abort(sb, "Detected aborted journal"); - return ERR_PTR(-EROFS); - } - return jbd2_journal_start(journal, nblocks); + handle = ext4_journal_current_handle(); + + /* + * If a handle has been started, it should be allowed to + * finish, otherwise deadlock could happen between freeze + * and others(e.g. truncate) due to the restart of the + * journal handle if the filesystem is forzen and active + * handles are not stopped. + */ + if (!handle) + vfs_check_frozen(sb, SB_FREEZE_TRANS); + + if (!journal) + return ext4_get_nojournal(); + /* + * Special case here: if the journal has aborted behind our + * backs (eg. EIO in the commit thread), then we still need to + * take the FS itself readonly cleanly. + */ + if (is_journal_aborted(journal)) { + ext4_abort(sb, "Detected aborted journal"); + return ERR_PTR(-EROFS); } - return ext4_get_nojournal(); + return jbd2_journal_start(journal, nblocks); } /* @@ -4146,6 +4163,11 @@ static int ext4_sync_fs(struct super_block *sb, int wait) /* * LVM calls this function before a (read-only) snapshot is created. This * gives us a chance to flush the journal completely and mark the fs clean. + * + * Note that only this function cannot bring a filesystem to be in a clean + * state independently, because ext4 prevents a new handle from being started + * by @sb->s_frozen, which stays in an upper layer. It thus needs help from + * the upper layer. */ static int ext4_freeze(struct super_block *sb) { -- cgit v0.10.2 From f80da1e70f1ffec3825aa0a1d0801f4896e002b6 Mon Sep 17 00:00:00 2001 From: Kazuya Mio Date: Sun, 10 Apr 2011 22:06:36 -0400 Subject: ext4: Allow indirect-block file to grow the file size to max file size We can create 4402345721856 byte file with indirect block mapping. However, if we grow an indirect-block file to the size with ftruncate(), we can see an ext4 warning. The following patch fixes this problem. How to reproduce: # dd if=/dev/zero of=/mnt/mp1/hoge bs=1 count=0 seek=4402345721856 0+0 records in 0+0 records out 0 bytes (0 B) copied, 0.000221428 s, 0.0 kB/s # tail -n 1 /var/log/messages Nov 25 15:10:27 test kernel: EXT4-fs warning (device sda8): ext4_block_to_path:345: block 1074791436 > max in inode 12 Signed-off-by: Kazuya Mio Signed-off-by: "Theodore Ts'o" diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 7d11e02..5560f78 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -4429,8 +4429,8 @@ void ext4_truncate(struct inode *inode) Indirect chain[4]; Indirect *partial; __le32 nr = 0; - int n; - ext4_lblk_t last_block; + int n = 0; + ext4_lblk_t last_block, max_block; unsigned blocksize = inode->i_sb->s_blocksize; trace_ext4_truncate_enter(inode); @@ -4455,14 +4455,18 @@ void ext4_truncate(struct inode *inode) last_block = (inode->i_size + blocksize-1) >> EXT4_BLOCK_SIZE_BITS(inode->i_sb); + max_block = (EXT4_SB(inode->i_sb)->s_bitmap_maxbytes + blocksize-1) + >> EXT4_BLOCK_SIZE_BITS(inode->i_sb); if (inode->i_size & (blocksize - 1)) if (ext4_block_truncate_page(handle, mapping, inode->i_size)) goto out_stop; - n = ext4_block_to_path(inode, last_block, offsets, NULL); - if (n == 0) - goto out_stop; /* error */ + if (last_block != max_block) { + n = ext4_block_to_path(inode, last_block, offsets, NULL); + if (n == 0) + goto out_stop; /* error */ + } /* * OK. This truncate is going to happen. We add the inode to the @@ -4493,7 +4497,13 @@ void ext4_truncate(struct inode *inode) */ ei->i_disksize = inode->i_size; - if (n == 1) { /* direct blocks */ + if (last_block == max_block) { + /* + * It is unnecessary to free any data blocks if last_block is + * equal to the indirect block limit. + */ + goto out_unlock; + } else if (n == 1) { /* direct blocks */ ext4_free_data(handle, inode, NULL, i_data+offsets[0], i_data + EXT4_NDIR_BLOCKS); goto do_indirects; @@ -4553,6 +4563,7 @@ do_indirects: ; } +out_unlock: up_write(&ei->i_data_sem); inode->i_mtime = inode->i_ctime = ext4_current_time(inode); ext4_mark_inode_dirty(handle, inode); -- cgit v0.10.2 From c8205636029fc869278c55b7336053b3e7ae3ef4 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sun, 10 Apr 2011 22:30:07 -0400 Subject: ext4: fix data corruption regression by reverting commit 6de9843dab3f Revert commit 6de9843dab3f2a1d4d66d80aa9e5782f80977d20, since it caused a data corruption regression with BitTorrent downloads. Thanks to Damien for discovering and bisecting to find the problem commit. https://bugzilla.kernel.org/show_bug.cgi?id=32972 Reported-by: Damien Grassart Signed-off-by: "Theodore Ts'o" diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 5560f78..9c8cf81 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2502,6 +2502,7 @@ static int ext4_da_get_block_prep(struct inode *inode, sector_t iblock, * for partial write. */ set_buffer_new(bh); + set_buffer_mapped(bh); } return 0; } -- cgit v0.10.2 From 8faa7cf828bca1745a4ed599876567f5afc47544 Mon Sep 17 00:00:00 2001 From: "Ira W. Snyder" Date: Thu, 7 Apr 2011 10:33:03 -0700 Subject: dt/fsldma: fix build warning caused by of_platform_device changes Commit 000061245a6797d542854106463b6b20fbdcb12e, "dt/powerpc: Eliminate users of of_platform_{,un}register_driver" forgot to convert the type of structure passed into platform_device_register() when it was converted from of_platform_device_register. Fix it. Signed-off-by: Ira W. Snyder Signed-off-by: Grant Likely diff --git a/drivers/dma/fsldma.c b/drivers/dma/fsldma.c index 6b39675..8a78154 100644 --- a/drivers/dma/fsldma.c +++ b/drivers/dma/fsldma.c @@ -1448,7 +1448,7 @@ static const struct of_device_id fsldma_of_ids[] = { {} }; -static struct of_platform_driver fsldma_of_driver = { +static struct platform_driver fsldma_of_driver = { .driver = { .name = "fsl-elo-dma", .owner = THIS_MODULE, -- cgit v0.10.2 From b0432d8f162c7d5d9537b4cb749d44076b76a783 Mon Sep 17 00:00:00 2001 From: Ken Chen Date: Thu, 7 Apr 2011 17:23:22 -0700 Subject: sched: Fix sched-domain avg_load calculation In function find_busiest_group(), the sched-domain avg_load isn't calculated at all if there is a group imbalance within the domain. This will cause erroneous imbalance calculation. The reason is that calculate_imbalance() sees sds->avg_load = 0 and it will dump entire sds->max_load into imbalance variable, which is used later on to migrate entire load from busiest CPU to the puller CPU. This has two really bad effect: 1. stampede of task migration, and they won't be able to break out of the bad state because of positive feedback loop: large load delta -> heavier load migration -> larger imbalance and the cycle goes on. 2. severe imbalance in CPU queue depth. This causes really long scheduling latency blip which affects badly on application that has tight latency requirement. The fix is to have kernel calculate domain avg_load in both cases. This will ensure that imbalance calculation is always sensible and the target is usually half way between busiest and puller CPU. Signed-off-by: Ken Chen Signed-off-by: Peter Zijlstra Cc: Link: http://lkml.kernel.org/r/20110408002322.3A0D812217F@elm.corp.google.com Signed-off-by: Ingo Molnar diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 7f00772..60f9d40 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -3127,6 +3127,8 @@ find_busiest_group(struct sched_domain *sd, int this_cpu, if (!sds.busiest || sds.busiest_nr_running == 0) goto out_balanced; + sds.avg_load = (SCHED_LOAD_SCALE * sds.total_load) / sds.total_pwr; + /* * If the busiest group is imbalanced the below checks don't * work because they assumes all things are equal, which typically @@ -3151,7 +3153,6 @@ find_busiest_group(struct sched_domain *sd, int this_cpu, * Don't pull any tasks if this group is already above the domain * average load. */ - sds.avg_load = (SCHED_LOAD_SCALE * sds.total_load) / sds.total_pwr; if (sds.this_load >= sds.avg_load) goto out_balanced; -- cgit v0.10.2 From b30aef17f71cf9e24b10c11cbb5e5f0ebe8a85ab Mon Sep 17 00:00:00 2001 From: Ken Chen Date: Fri, 8 Apr 2011 12:20:16 -0700 Subject: sched: Fix erroneous all_pinned logic The scheduler load balancer has specific code to deal with cases of unbalanced system due to lots of unmovable tasks (for example because of hard CPU affinity). In those situation, it excludes the busiest CPU that has pinned tasks for load balance consideration such that it can perform second 2nd load balance pass on the rest of the system. This all works as designed if there is only one cgroup in the system. However, when we have multiple cgroups, this logic has false positives and triggers multiple load balance passes despite there are actually no pinned tasks at all. The reason it has false positives is that the all pinned logic is deep in the lowest function of can_migrate_task() and is too low level: load_balance_fair() iterates each task group and calls balance_tasks() to migrate target load. Along the way, balance_tasks() will also set a all_pinned variable. Given that task-groups are iterated, this all_pinned variable is essentially the status of last group in the scanning process. Task group can have number of reasons that no load being migrated, none due to cpu affinity. However, this status bit is being propagated back up to the higher level load_balance(), which incorrectly think that no tasks were moved. It kick off the all pinned logic and start multiple passes attempt to move load onto puller CPU. To fix this, move the all_pinned aggregation up at the iterator level. This ensures that the status is aggregated over all task-groups, not just last one in the list. Signed-off-by: Ken Chen Cc: stable@kernel.org Signed-off-by: Peter Zijlstra Link: http://lkml.kernel.org/r/BANLkTi=ernzNawaR5tJZEsV_QVnfxqXmsQ@mail.gmail.com Signed-off-by: Ingo Molnar diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 60f9d40..6fa833a 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -2104,21 +2104,20 @@ balance_tasks(struct rq *this_rq, int this_cpu, struct rq *busiest, enum cpu_idle_type idle, int *all_pinned, int *this_best_prio, struct cfs_rq *busiest_cfs_rq) { - int loops = 0, pulled = 0, pinned = 0; + int loops = 0, pulled = 0; long rem_load_move = max_load_move; struct task_struct *p, *n; if (max_load_move == 0) goto out; - pinned = 1; - list_for_each_entry_safe(p, n, &busiest_cfs_rq->tasks, se.group_node) { if (loops++ > sysctl_sched_nr_migrate) break; if ((p->se.load.weight >> 1) > rem_load_move || - !can_migrate_task(p, busiest, this_cpu, sd, idle, &pinned)) + !can_migrate_task(p, busiest, this_cpu, sd, idle, + all_pinned)) continue; pull_task(busiest, p, this_rq, this_cpu); @@ -2153,9 +2152,6 @@ out: */ schedstat_add(sd, lb_gained[idle], pulled); - if (all_pinned) - *all_pinned = pinned; - return max_load_move - rem_load_move; } @@ -3341,6 +3337,7 @@ redo: * still unbalanced. ld_moved simply stays zero, so it is * correctly treated as an imbalance. */ + all_pinned = 1; local_irq_save(flags); double_rq_lock(this_rq, busiest); ld_moved = move_tasks(this_rq, this_cpu, busiest, -- cgit v0.10.2 From 109b81296c63228578d4760794d8dd46e02eddfb Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 11 Apr 2011 14:13:10 +0200 Subject: block: splice plug list to local context If the request_fn ends up blocking, we could be re-entering the plug flush. Since the list is protected by explicitly not allowing schedule events, this isn't a terribly good idea. Additionally, it can cause us to recurse. As request_fn called by __blk_run_queue is allowed to 'schedule()' (after dropping the queue lock of course), it is possible to get a recursive call: schedule -> blk_flush_plug -> __blk_finish_plug -> flush_plug_list -> __blk_run_queue -> request_fn -> schedule We must make sure that the second schedule does not call into blk_flush_plug again. So instead of leaving the list of requests on blk_plug->list, move them to a separate list leaving blk_plug->list empty. Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 90f22cc..eeaca09 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2673,19 +2673,24 @@ static void flush_plug_list(struct blk_plug *plug) struct request_queue *q; unsigned long flags; struct request *rq; + LIST_HEAD(list); BUG_ON(plug->magic != PLUG_MAGIC); if (list_empty(&plug->list)) return; - if (plug->should_sort) - list_sort(NULL, &plug->list, plug_rq_cmp); + list_splice_init(&plug->list, &list); + + if (plug->should_sort) { + list_sort(NULL, &list, plug_rq_cmp); + plug->should_sort = 0; + } q = NULL; local_irq_save(flags); - while (!list_empty(&plug->list)) { - rq = list_entry_rq(plug->list.next); + while (!list_empty(&list)) { + rq = list_entry_rq(list.next); list_del_init(&rq->queuelist); BUG_ON(!(rq->cmd_flags & REQ_ON_PLUG)); BUG_ON(!rq->q); @@ -2713,7 +2718,6 @@ static void flush_plug_list(struct blk_plug *plug) spin_unlock(q->queue_lock); } - BUG_ON(!list_empty(&plug->list)); local_irq_restore(flags); } -- cgit v0.10.2 From 30d746c68025ee69ac17219aacc9b1614d951f01 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Thu, 7 Apr 2011 14:13:15 +0200 Subject: x86/ce4100: Add reg property to bridges without the reg property Ben's new code won't find the PCI & ISA bridge and the devices won't get the DT-node attached. Signed-off-by: Sebastian Andrzej Siewior Acked-by: Thomas Gleixner Cc: davem@davemloft.net Cc: monstr@monstr.eu Cc: Benjamin Herrenschmidt Link: http://lkml.kernel.org/r/20110407121315.GA9204@linutronix.de Signed-off-by: Ingo Molnar diff --git a/arch/x86/platform/ce4100/falconfalls.dts b/arch/x86/platform/ce4100/falconfalls.dts index dc701ea..2d6d226 100644 --- a/arch/x86/platform/ce4100/falconfalls.dts +++ b/arch/x86/platform/ce4100/falconfalls.dts @@ -74,6 +74,7 @@ compatible = "intel,ce4100-pci", "pci"; device_type = "pci"; bus-range = <1 1>; + reg = <0x0800 0x0 0x0 0x0 0x0>; ranges = <0x2000000 0 0xdffe0000 0x2000000 0 0xdffe0000 0 0x1000>; interrupt-parent = <&ioapic2>; @@ -412,6 +413,7 @@ #address-cells = <2>; #size-cells = <1>; compatible = "isa"; + reg = <0xf800 0x0 0x0 0x0 0x0>; ranges = <1 0 0 0 0 0x100>; rtc@70 { -- cgit v0.10.2 From b42282e5a05018ecdc0d63a4ad530b0999785912 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 11 Apr 2011 10:53:11 -0700 Subject: pci: fix PCI bus allocation alignment handling In commit 13583b16592a ("PCI: refactor io size calculation code") Ram had a thinko in the refactorization of the code: the end result used the variable 'align' for the bus alignment, but the original code used 'min_align'. Since then, another use of that 'align' variable got introduced by commit c8adf9a3e873 ("PCI: pre-allocate additional resources to devices only after successful allocation of essential resources.") Fix both of those uses to use 'min_align' as they should. Daniel Hellstrom Acked-by: Ram Pai Acked-by: Jesse Barnes Signed-off-by: Linus Torvalds diff --git a/drivers/pci/setup-bus.c b/drivers/pci/setup-bus.c index 89d0a6a..ebf51ad 100644 --- a/drivers/pci/setup-bus.c +++ b/drivers/pci/setup-bus.c @@ -676,10 +676,10 @@ static int pbus_size_mem(struct pci_bus *bus, unsigned long mask, min_align = align1 >> 1; align += aligns[order]; } - size0 = calculate_memsize(size, min_size, 0, resource_size(b_res), align); + size0 = calculate_memsize(size, min_size, 0, resource_size(b_res), min_align); size1 = !add_size ? size : calculate_memsize(size, min_size+add_size, 0, - resource_size(b_res), align); + resource_size(b_res), min_align); if (!size0 && !size1) { if (b_res->start || b_res->end) dev_info(&bus->self->dev, "disabling bridge window " -- cgit v0.10.2 From 4a39e781682828c1b81a839e7d343fd91945a8d5 Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Mon, 11 Apr 2011 13:02:06 -0400 Subject: iwlegacy: make iwl3945 and iwl4965 select IWLWIFI_LEGACY Otherwise, IWLWIFI_LEGACY has to be selected independently before the drivers are made available. Reported-by: Dave Airlie Signed-off-by: John W. Linville Cc: Stanislaw Gruszka Cc: Wey-Yi Guy diff --git a/drivers/net/wireless/iwlegacy/Kconfig b/drivers/net/wireless/iwlegacy/Kconfig index 2a45dd4..aef65cd 100644 --- a/drivers/net/wireless/iwlegacy/Kconfig +++ b/drivers/net/wireless/iwlegacy/Kconfig @@ -1,6 +1,5 @@ config IWLWIFI_LEGACY - tristate "Intel Wireless Wifi legacy devices" - depends on PCI && MAC80211 + tristate select FW_LOADER select NEW_LEDS select LEDS_CLASS @@ -65,7 +64,8 @@ endmenu config IWL4965 tristate "Intel Wireless WiFi 4965AGN (iwl4965)" - depends on IWLWIFI_LEGACY + depends on PCI && MAC80211 + select IWLWIFI_LEGACY ---help--- This option enables support for @@ -92,7 +92,8 @@ config IWL4965 config IWL3945 tristate "Intel PRO/Wireless 3945ABG/BG Network Connection (iwl3945)" - depends on IWLWIFI_LEGACY + depends on PCI && MAC80211 + select IWLWIFI_LEGACY ---help--- Select to build the driver supporting the: -- cgit v0.10.2 From bded18c2dd09eee870f4446652dbce493a6dece1 Mon Sep 17 00:00:00 2001 From: Giuseppe CAVALLARO Date: Sun, 10 Apr 2011 23:16:44 +0000 Subject: stmmac: fixed dma lib build when turn-on the debug option This patch fixes a compilation error when build the dwmac_lib with the DEBUG option enabled. Signed-off-by: Giuseppe Cavallaro Signed-off-by: David S. Miller diff --git a/drivers/net/stmmac/dwmac_lib.c b/drivers/net/stmmac/dwmac_lib.c index d65fab1..e250935 100644 --- a/drivers/net/stmmac/dwmac_lib.c +++ b/drivers/net/stmmac/dwmac_lib.c @@ -26,9 +26,9 @@ #undef DWMAC_DMA_DEBUG #ifdef DWMAC_DMA_DEBUG -#define DBG(fmt, args...) printk(fmt, ## args) +#define DWMAC_LIB_DBG(fmt, args...) printk(fmt, ## args) #else -#define DBG(fmt, args...) do { } while (0) +#define DWMAC_LIB_DBG(fmt, args...) do { } while (0) #endif /* CSR1 enables the transmit DMA to check for new descriptor */ @@ -152,7 +152,7 @@ int dwmac_dma_interrupt(void __iomem *ioaddr, /* read the status register (CSR5) */ u32 intr_status = readl(ioaddr + DMA_STATUS); - DBG(INFO, "%s: [CSR5: 0x%08x]\n", __func__, intr_status); + DWMAC_LIB_DBG(KERN_INFO "%s: [CSR5: 0x%08x]\n", __func__, intr_status); #ifdef DWMAC_DMA_DEBUG /* It displays the DMA process states (CSR5 register) */ show_tx_process_state(intr_status); @@ -160,43 +160,43 @@ int dwmac_dma_interrupt(void __iomem *ioaddr, #endif /* ABNORMAL interrupts */ if (unlikely(intr_status & DMA_STATUS_AIS)) { - DBG(INFO, "CSR5[15] DMA ABNORMAL IRQ: "); + DWMAC_LIB_DBG(KERN_INFO "CSR5[15] DMA ABNORMAL IRQ: "); if (unlikely(intr_status & DMA_STATUS_UNF)) { - DBG(INFO, "transmit underflow\n"); + DWMAC_LIB_DBG(KERN_INFO "transmit underflow\n"); ret = tx_hard_error_bump_tc; x->tx_undeflow_irq++; } if (unlikely(intr_status & DMA_STATUS_TJT)) { - DBG(INFO, "transmit jabber\n"); + DWMAC_LIB_DBG(KERN_INFO "transmit jabber\n"); x->tx_jabber_irq++; } if (unlikely(intr_status & DMA_STATUS_OVF)) { - DBG(INFO, "recv overflow\n"); + DWMAC_LIB_DBG(KERN_INFO "recv overflow\n"); x->rx_overflow_irq++; } if (unlikely(intr_status & DMA_STATUS_RU)) { - DBG(INFO, "receive buffer unavailable\n"); + DWMAC_LIB_DBG(KERN_INFO "receive buffer unavailable\n"); x->rx_buf_unav_irq++; } if (unlikely(intr_status & DMA_STATUS_RPS)) { - DBG(INFO, "receive process stopped\n"); + DWMAC_LIB_DBG(KERN_INFO "receive process stopped\n"); x->rx_process_stopped_irq++; } if (unlikely(intr_status & DMA_STATUS_RWT)) { - DBG(INFO, "receive watchdog\n"); + DWMAC_LIB_DBG(KERN_INFO "receive watchdog\n"); x->rx_watchdog_irq++; } if (unlikely(intr_status & DMA_STATUS_ETI)) { - DBG(INFO, "transmit early interrupt\n"); + DWMAC_LIB_DBG(KERN_INFO "transmit early interrupt\n"); x->tx_early_irq++; } if (unlikely(intr_status & DMA_STATUS_TPS)) { - DBG(INFO, "transmit process stopped\n"); + DWMAC_LIB_DBG(KERN_INFO "transmit process stopped\n"); x->tx_process_stopped_irq++; ret = tx_hard_error; } if (unlikely(intr_status & DMA_STATUS_FBI)) { - DBG(INFO, "fatal bus error\n"); + DWMAC_LIB_DBG(KERN_INFO "fatal bus error\n"); x->fatal_bus_error_irq++; ret = tx_hard_error; } @@ -215,7 +215,7 @@ int dwmac_dma_interrupt(void __iomem *ioaddr, /* Clear the interrupt by writing a logic 1 to the CSR5[15-0] */ writel((intr_status & 0x1ffff), ioaddr + DMA_STATUS); - DBG(INFO, "\n\n"); + DWMAC_LIB_DBG(KERN_INFO "\n\n"); return ret; } -- cgit v0.10.2 From f66ffe285939559d2a6f630a36f676d7c056b99d Mon Sep 17 00:00:00 2001 From: Giuseppe CAVALLARO Date: Sun, 10 Apr 2011 23:16:45 +0000 Subject: stmmac: fix open funct when exit on error This patch reviews the open function and fixes some errors when exit with an error state. It also moves the request_irq after core is initialized when interrupts are properly masked. Signed-off-by: Shiraz Hashim Hacked-by: Giuseppe Cavallaro Signed-off-by: David S. Miller diff --git a/drivers/net/stmmac/stmmac_main.c b/drivers/net/stmmac/stmmac_main.c index 0e5f031..38f7b61 100644 --- a/drivers/net/stmmac/stmmac_main.c +++ b/drivers/net/stmmac/stmmac_main.c @@ -781,21 +781,6 @@ static int stmmac_open(struct net_device *dev) stmmac_verify_args(); - ret = stmmac_init_phy(dev); - if (unlikely(ret)) { - pr_err("%s: Cannot attach to PHY (error: %d)\n", __func__, ret); - return ret; - } - - /* Request the IRQ lines */ - ret = request_irq(dev->irq, stmmac_interrupt, - IRQF_SHARED, dev->name, dev); - if (unlikely(ret < 0)) { - pr_err("%s: ERROR: allocating the IRQ %d (error: %d)\n", - __func__, dev->irq, ret); - return ret; - } - #ifdef CONFIG_STMMAC_TIMER priv->tm = kzalloc(sizeof(struct stmmac_timer *), GFP_KERNEL); if (unlikely(priv->tm == NULL)) { @@ -814,6 +799,11 @@ static int stmmac_open(struct net_device *dev) } else priv->tm->enable = 1; #endif + ret = stmmac_init_phy(dev); + if (unlikely(ret)) { + pr_err("%s: Cannot attach to PHY (error: %d)\n", __func__, ret); + goto open_error; + } /* Create and initialize the TX/RX descriptors chains. */ priv->dma_tx_size = STMMAC_ALIGN(dma_txsize); @@ -822,12 +812,11 @@ static int stmmac_open(struct net_device *dev) init_dma_desc_rings(dev); /* DMA initialization and SW reset */ - if (unlikely(priv->hw->dma->init(priv->ioaddr, priv->plat->pbl, - priv->dma_tx_phy, - priv->dma_rx_phy) < 0)) { - + ret = priv->hw->dma->init(priv->ioaddr, priv->plat->pbl, + priv->dma_tx_phy, priv->dma_rx_phy); + if (ret < 0) { pr_err("%s: DMA initialization failed\n", __func__); - return -1; + goto open_error; } /* Copy the MAC addr into the HW */ @@ -848,6 +837,15 @@ static int stmmac_open(struct net_device *dev) writel(0xffffffff, priv->ioaddr + MMC_HIGH_INTR_MASK); writel(0xffffffff, priv->ioaddr + MMC_LOW_INTR_MASK); + /* Request the IRQ lines */ + ret = request_irq(dev->irq, stmmac_interrupt, + IRQF_SHARED, dev->name, dev); + if (unlikely(ret < 0)) { + pr_err("%s: ERROR: allocating the IRQ %d (error: %d)\n", + __func__, dev->irq, ret); + goto open_error; + } + /* Enable the MAC Rx/Tx */ stmmac_enable_mac(priv->ioaddr); @@ -878,7 +876,17 @@ static int stmmac_open(struct net_device *dev) napi_enable(&priv->napi); skb_queue_head_init(&priv->rx_recycle); netif_start_queue(dev); + return 0; + +open_error: +#ifdef CONFIG_STMMAC_TIMER + kfree(priv->tm); +#endif + if (priv->phydev) + phy_disconnect(priv->phydev); + + return ret; } /** -- cgit v0.10.2 From a08070acd43a76069ab20000cd02474369506add Mon Sep 17 00:00:00 2001 From: Giuseppe CAVALLARO Date: Sun, 10 Apr 2011 23:16:46 +0000 Subject: stmmac: fix Transmit Underflow error On some old MAC chips without COE sometime the Transmit Underflow error is issued. The driver aborted all the transmission process and initialized it from scratch. This breaks the network activity as raised by Nachiketa on a SPEAr board. The patch is to fix this rare underflow event. The driver will only clear the interrupt and the Tx DMA will go out the Suspend state as soon as the descriptor is fetched again. The driver will continue to bump-up the DMA FIFO threshold that, indeed, helped somebody to prevent this kind of error in the past as well. Reported-by: Nachiketa Prachanda Signed-off-by: Giuseppe Cavallaro Signed-off-by: David S. Miller diff --git a/drivers/net/stmmac/stmmac_main.c b/drivers/net/stmmac/stmmac_main.c index 38f7b61..cc973fc 100644 --- a/drivers/net/stmmac/stmmac_main.c +++ b/drivers/net/stmmac/stmmac_main.c @@ -750,7 +750,6 @@ static void stmmac_dma_interrupt(struct stmmac_priv *priv) priv->hw->dma->dma_mode(priv->ioaddr, tc, SF_DMA_MODE); priv->xstats.threshold = tc; } - stmmac_tx_err(priv); } else if (unlikely(status == tx_hard_error)) stmmac_tx_err(priv); } -- cgit v0.10.2 From 38dbaf0afb518e462de7afca552acad048237a73 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Peter=20Pan=28=E6=BD=98=E5=8D=AB=E5=B9=B3=29?= Date: Fri, 8 Apr 2011 03:40:19 +0000 Subject: bonding:set save_load to 0 when initializing MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit It is unnecessary to set save_load to 1 here, as the tx_hashtbl is just kzalloced. Signed-off-by: Weiping Pan(潘卫平) Signed-off-by: Jay Vosburgh Signed-off-by: David S. Miller diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c index 9bc5de3..ab69e5a 100644 --- a/drivers/net/bonding/bond_alb.c +++ b/drivers/net/bonding/bond_alb.c @@ -176,7 +176,7 @@ static int tlb_initialize(struct bonding *bond) bond_info->tx_hashtbl = new_hashtbl; for (i = 0; i < TLB_HASH_TABLE_SIZE; i++) { - tlb_init_table_entry(&bond_info->tx_hashtbl[i], 1); + tlb_init_table_entry(&bond_info->tx_hashtbl[i], 0); } _unlock_tx_hashtbl(bond); -- cgit v0.10.2 From 3b647568454acedb8f09f72931bfe92b73f812c7 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Peter=20Pan=28=E6=BD=98=E5=8D=AB=E5=B9=B3=29?= Date: Sun, 10 Apr 2011 22:17:24 +0000 Subject: bonding:delete unused alb_timer MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Now bonding-alb uses delayed_work instead of timer_list. Signed-off-by: Weiping Pan(潘卫平) Signed-off-by: David S. Miller diff --git a/drivers/net/bonding/bond_alb.h b/drivers/net/bonding/bond_alb.h index 86861f0..f2a1d0a 100644 --- a/drivers/net/bonding/bond_alb.h +++ b/drivers/net/bonding/bond_alb.h @@ -122,7 +122,6 @@ struct tlb_slave_info { }; struct alb_bond_info { - struct timer_list alb_timer; struct tlb_client_info *tx_hashtbl; /* Dynamically allocated */ spinlock_t tx_hashtbl_lock; u32 unbalanced_load; -- cgit v0.10.2 From 26f007b85a1bb73e1974ebd371cf5cd1e80c0a00 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Peter=20Pan=28=E6=BD=98=E5=8D=AB=E5=B9=B3=29?= Date: Sun, 10 Apr 2011 22:17:25 +0000 Subject: bonding:delete unused rlb_interval_counter MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Now, alb_bond_info uses rx_ntt,rlb_update_delay_counter and rlb_update_retry_counter to decide when to call rlb_update_rx_clients(). Signed-off-by: Weiping Pan(潘卫平) Signed-off-by: David S. Miller diff --git a/drivers/net/bonding/bond_alb.h b/drivers/net/bonding/bond_alb.h index f2a1d0a..01ed1fb 100644 --- a/drivers/net/bonding/bond_alb.h +++ b/drivers/net/bonding/bond_alb.h @@ -139,7 +139,6 @@ struct alb_bond_info { struct slave *next_rx_slave;/* next slave to be assigned * to a new rx client for */ - u32 rlb_interval_counter; u8 primary_is_promisc; /* boolean */ u32 rlb_promisc_timeout_counter;/* counts primary * promiscuity time -- cgit v0.10.2 From 9814290ad0314fa0ce7991eae34ac44956e40b12 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Peter=20Pan=28=E6=BD=98=E5=8D=AB=E5=B9=B3=29?= Date: Mon, 11 Apr 2011 00:15:57 +0000 Subject: net: fix tranmitted/tranmitting typo MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit replace tranmitted with transmitted. replace tranmitting with transmitting. Signed-off-by: Weiping Pan(潘卫平) Signed-off-by: David S. Miller diff --git a/drivers/net/bonding/bond_alb.h b/drivers/net/bonding/bond_alb.h index 01ed1fb..8ca7158 100644 --- a/drivers/net/bonding/bond_alb.h +++ b/drivers/net/bonding/bond_alb.h @@ -75,7 +75,7 @@ struct tlb_client_info { * gave this entry index. */ u32 tx_bytes; /* Each Client accumulates the BytesTx that - * were tranmitted to it, and after each + * were transmitted to it, and after each * CallBack the LoadHistory is divided * by the balance interval */ diff --git a/drivers/net/tokenring/3c359.c b/drivers/net/tokenring/3c359.c index 8a3b191..ff32bef 100644 --- a/drivers/net/tokenring/3c359.c +++ b/drivers/net/tokenring/3c359.c @@ -1251,7 +1251,7 @@ static netdev_tx_t xl_xmit(struct sk_buff *skb, struct net_device *dev) /* * The NIC has told us that a packet has been downloaded onto the card, we must * find out which packet it has done, clear the skb and information for the packet - * then advance around the ring for all tranmitted packets + * then advance around the ring for all transmitted packets */ static void xl_dn_comp(struct net_device *dev) @@ -1568,7 +1568,7 @@ static void xl_arb_cmd(struct net_device *dev) if (lan_status_diff & LSC_SOFT_ERR) printk(KERN_WARNING "%s: Adapter transmitted Soft Error Report Mac Frame\n",dev->name); if (lan_status_diff & LSC_TRAN_BCN) - printk(KERN_INFO "%s: We are tranmitting the beacon, aaah\n",dev->name); + printk(KERN_INFO "%s: We are transmitting the beacon, aaah\n",dev->name); if (lan_status_diff & LSC_SS) printk(KERN_INFO "%s: Single Station on the ring\n", dev->name); if (lan_status_diff & LSC_RING_REC) diff --git a/drivers/net/tokenring/lanstreamer.c b/drivers/net/tokenring/lanstreamer.c index 5bd1407..9354ca9 100644 --- a/drivers/net/tokenring/lanstreamer.c +++ b/drivers/net/tokenring/lanstreamer.c @@ -1675,7 +1675,7 @@ drop_frame: if (lan_status_diff & LSC_SOFT_ERR) printk(KERN_WARNING "%s: Adapter transmitted Soft Error Report Mac Frame\n", dev->name); if (lan_status_diff & LSC_TRAN_BCN) - printk(KERN_INFO "%s: We are tranmitting the beacon, aaah\n", dev->name); + printk(KERN_INFO "%s: We are transmitting the beacon, aaah\n", dev->name); if (lan_status_diff & LSC_SS) printk(KERN_INFO "%s: Single Station on the ring\n", dev->name); if (lan_status_diff & LSC_RING_REC) diff --git a/drivers/net/tokenring/olympic.c b/drivers/net/tokenring/olympic.c index 3d2fbe6..2684003 100644 --- a/drivers/net/tokenring/olympic.c +++ b/drivers/net/tokenring/olympic.c @@ -1500,7 +1500,7 @@ drop_frame: if (lan_status_diff & LSC_SOFT_ERR) printk(KERN_WARNING "%s: Adapter transmitted Soft Error Report Mac Frame\n",dev->name); if (lan_status_diff & LSC_TRAN_BCN) - printk(KERN_INFO "%s: We are tranmitting the beacon, aaah\n",dev->name); + printk(KERN_INFO "%s: We are transmitting the beacon, aaah\n",dev->name); if (lan_status_diff & LSC_SS) printk(KERN_INFO "%s: Single Station on the ring\n", dev->name); if (lan_status_diff & LSC_RING_REC) -- cgit v0.10.2 From 77c8e2c01542649f7a02fef8eb3b3d0e7fed6bbd Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Peter=20Pan=28=E6=BD=98=E5=8D=AB=E5=B9=B3=29?= Date: Mon, 11 Apr 2011 00:16:32 +0000 Subject: bonding:fix two typos MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit replace relpy with reply. replace premanent with permanent. Signed-off-by: Weiping Pan(潘卫平) Signed-off-by: David S. Miller diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c index ab69e5a..ba71582 100644 --- a/drivers/net/bonding/bond_alb.c +++ b/drivers/net/bonding/bond_alb.c @@ -701,7 +701,7 @@ static struct slave *rlb_arp_xmit(struct sk_buff *skb, struct bonding *bond) */ rlb_choose_channel(skb, bond); - /* The ARP relpy packets must be delayed so that + /* The ARP reply packets must be delayed so that * they can cancel out the influence of the ARP request. */ bond->alb_info.rlb_update_delay_counter = RLB_UPDATE_DELAY; @@ -1042,7 +1042,7 @@ static void alb_change_hw_addr_on_detach(struct bonding *bond, struct slave *sla * * If the permanent hw address of @slave is @bond's hw address, we need to * find a different hw address to give @slave, that isn't in use by any other - * slave in the bond. This address must be, of course, one of the premanent + * slave in the bond. This address must be, of course, one of the permanent * addresses of the other slaves. * * We go over the slave list, and for each slave there we compare its -- cgit v0.10.2 From c968bdf6912cad6d0fc63d7037cc1c870604a808 Mon Sep 17 00:00:00 2001 From: amit salecha Date: Mon, 11 Apr 2011 02:10:22 +0000 Subject: netxen: limit skb frags for non tso packet Machines are getting deadlock in four node cluster environment. All nodes are accessing (find /gfs2 -depth -print|cpio -ocv > /dev/null) 200 GB storage on a GFS2 filesystem. This result in memory fragmentation and driver receives 18 frags for 1448 byte packets. For non tso packet, fw drops the tx request, if it has >14 frags. Fixing it by pulling extra frags. Cc: stable@kernel.org Signed-off-by: Amit Kumar Salecha Signed-off-by: David S. Miller diff --git a/drivers/net/netxen/netxen_nic.h b/drivers/net/netxen/netxen_nic.h index d7299f1..679dc85 100644 --- a/drivers/net/netxen/netxen_nic.h +++ b/drivers/net/netxen/netxen_nic.h @@ -174,7 +174,7 @@ #define MAX_NUM_CARDS 4 -#define MAX_BUFFERS_PER_CMD 32 +#define NETXEN_MAX_FRAGS_PER_TX 14 #define MAX_TSO_HEADER_DESC 2 #define MGMT_CMD_DESC_RESV 4 #define TX_STOP_THRESH ((MAX_SKB_FRAGS >> 2) + MAX_TSO_HEADER_DESC \ @@ -558,7 +558,7 @@ struct netxen_recv_crb { */ struct netxen_cmd_buffer { struct sk_buff *skb; - struct netxen_skb_frag frag_array[MAX_BUFFERS_PER_CMD + 1]; + struct netxen_skb_frag frag_array[MAX_SKB_FRAGS + 1]; u32 frag_count; }; diff --git a/drivers/net/netxen/netxen_nic_main.c b/drivers/net/netxen/netxen_nic_main.c index 83348dc..e8a4b66 100644 --- a/drivers/net/netxen/netxen_nic_main.c +++ b/drivers/net/netxen/netxen_nic_main.c @@ -1844,6 +1844,8 @@ netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev) struct cmd_desc_type0 *hwdesc, *first_desc; struct pci_dev *pdev; int i, k; + int delta = 0; + struct skb_frag_struct *frag; u32 producer; int frag_count, no_of_desc; @@ -1851,6 +1853,21 @@ netxen_nic_xmit_frame(struct sk_buff *skb, struct net_device *netdev) frag_count = skb_shinfo(skb)->nr_frags + 1; + /* 14 frags supported for normal packet and + * 32 frags supported for TSO packet + */ + if (!skb_is_gso(skb) && frag_count > NETXEN_MAX_FRAGS_PER_TX) { + + for (i = 0; i < (frag_count - NETXEN_MAX_FRAGS_PER_TX); i++) { + frag = &skb_shinfo(skb)->frags[i]; + delta += frag->size; + } + + if (!__pskb_pull_tail(skb, delta)) + goto drop_packet; + + frag_count = 1 + skb_shinfo(skb)->nr_frags; + } /* 4 fragments per cmd des */ no_of_desc = (frag_count + 3) >> 2; -- cgit v0.10.2 From 0c184ed9032c58b21f0d90de28c796874b73d6a1 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Sjur=20Br=C3=A6ndeland?= Date: Mon, 11 Apr 2011 10:11:29 +0000 Subject: caif: Bugfix use for_each_safe when removing list nodes. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: Sjur Brændeland Signed-off-by: David S. Miller diff --git a/net/caif/cfmuxl.c b/net/caif/cfmuxl.c index 46f34b2..24f1ffa 100644 --- a/net/caif/cfmuxl.c +++ b/net/caif/cfmuxl.c @@ -244,9 +244,9 @@ static void cfmuxl_ctrlcmd(struct cflayer *layr, enum caif_ctrlcmd ctrl, int phyid) { struct cfmuxl *muxl = container_obj(layr); - struct list_head *node; + struct list_head *node, *next; struct cflayer *layer; - list_for_each(node, &muxl->srvl_list) { + list_for_each_safe(node, next, &muxl->srvl_list) { layer = list_entry(node, struct cflayer, node); if (cfsrvl_phyid_match(layer, phyid)) layer->ctrlcmd(layer, ctrl, phyid); -- cgit v0.10.2 From 4a9f65f6304a00f6473e83b19c1e83caa1e42530 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Sjur=20Br=C3=A6ndeland?= Date: Mon, 11 Apr 2011 10:11:30 +0000 Subject: caif: performance bugfix - allow radio stack to prioritize packets. MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In the CAIF Payload message the Packet Type indication must be set to UNCLASSIFIED in order to allow packet prioritization in the modem's network stack. Otherwise TCP-Ack is not prioritized in the modems transmit queue. Signed-off-by: Sjur Brændeland Signed-off-by: David S. Miller diff --git a/net/caif/cfdgml.c b/net/caif/cfdgml.c index 27dab26..054fdb5 100644 --- a/net/caif/cfdgml.c +++ b/net/caif/cfdgml.c @@ -13,6 +13,7 @@ #include #include + #define container_obj(layr) ((struct cfsrvl *) layr) #define DGM_CMD_BIT 0x80 @@ -83,6 +84,7 @@ static int cfdgml_receive(struct cflayer *layr, struct cfpkt *pkt) static int cfdgml_transmit(struct cflayer *layr, struct cfpkt *pkt) { + u8 packet_type; u32 zero = 0; struct caif_payload_info *info; struct cfsrvl *service = container_obj(layr); @@ -94,7 +96,9 @@ static int cfdgml_transmit(struct cflayer *layr, struct cfpkt *pkt) if (cfpkt_getlen(pkt) > DGM_MTU) return -EMSGSIZE; - cfpkt_add_head(pkt, &zero, 4); + cfpkt_add_head(pkt, &zero, 3); + packet_type = 0x08; /* B9 set - UNCLASSIFIED */ + cfpkt_add_head(pkt, &packet_type, 1); /* Add info for MUX-layer to route the packet out. */ info = cfpkt_info(pkt); -- cgit v0.10.2 From 1f112cee07b314e244ee9e71d9c1e6950dc13327 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Mon, 11 Apr 2011 22:54:42 +0200 Subject: PM / Hibernate: Introduce CONFIG_HIBERNATE_CALLBACKS Xen save/restore is going to use hibernate device callbacks for quiescing devices and putting them back to normal operations and it would need to select CONFIG_HIBERNATION for this purpose. However, that also would cause the hibernate interfaces for user space to be enabled, which might confuse user space, because the Xen kernels don't support hibernation. Moreover, it would be wasteful, as it would make the Xen kernels include a substantial amount of code that they would never use. To address this issue introduce new power management Kconfig option CONFIG_HIBERNATE_CALLBACKS, such that it will only select the code that is necessary for the hibernate device callbacks to work and make CONFIG_HIBERNATION select it. Then, Xen save/restore will be able to select CONFIG_HIBERNATE_CALLBACKS without dragging the entire hibernate code along with it. Signed-off-by: Rafael J. Wysocki Tested-by: Shriram Rajagopalan diff --git a/arch/powerpc/kernel/ibmebus.c b/arch/powerpc/kernel/ibmebus.c index c00d4ca..28581f1 100644 --- a/arch/powerpc/kernel/ibmebus.c +++ b/arch/powerpc/kernel/ibmebus.c @@ -527,7 +527,7 @@ static int ibmebus_bus_pm_resume_noirq(struct device *dev) #endif /* !CONFIG_SUSPEND */ -#ifdef CONFIG_HIBERNATION +#ifdef CONFIG_HIBERNATE_CALLBACKS static int ibmebus_bus_pm_freeze(struct device *dev) { @@ -665,7 +665,7 @@ static int ibmebus_bus_pm_restore_noirq(struct device *dev) return ret; } -#else /* !CONFIG_HIBERNATION */ +#else /* !CONFIG_HIBERNATE_CALLBACKS */ #define ibmebus_bus_pm_freeze NULL #define ibmebus_bus_pm_thaw NULL @@ -676,7 +676,7 @@ static int ibmebus_bus_pm_restore_noirq(struct device *dev) #define ibmebus_bus_pm_poweroff_noirq NULL #define ibmebus_bus_pm_restore_noirq NULL -#endif /* !CONFIG_HIBERNATION */ +#endif /* !CONFIG_HIBERNATE_CALLBACKS */ static struct dev_pm_ops ibmebus_bus_dev_pm_ops = { .prepare = ibmebus_bus_pm_prepare, diff --git a/drivers/amba/bus.c b/drivers/amba/bus.c index 8210405..7025593 100644 --- a/drivers/amba/bus.c +++ b/drivers/amba/bus.c @@ -214,7 +214,7 @@ static int amba_pm_resume_noirq(struct device *dev) #endif /* !CONFIG_SUSPEND */ -#ifdef CONFIG_HIBERNATION +#ifdef CONFIG_HIBERNATE_CALLBACKS static int amba_pm_freeze(struct device *dev) { @@ -352,7 +352,7 @@ static int amba_pm_restore_noirq(struct device *dev) return ret; } -#else /* !CONFIG_HIBERNATION */ +#else /* !CONFIG_HIBERNATE_CALLBACKS */ #define amba_pm_freeze NULL #define amba_pm_thaw NULL @@ -363,7 +363,7 @@ static int amba_pm_restore_noirq(struct device *dev) #define amba_pm_poweroff_noirq NULL #define amba_pm_restore_noirq NULL -#endif /* !CONFIG_HIBERNATION */ +#endif /* !CONFIG_HIBERNATE_CALLBACKS */ #ifdef CONFIG_PM diff --git a/drivers/base/platform.c b/drivers/base/platform.c index f051cff..9279231 100644 --- a/drivers/base/platform.c +++ b/drivers/base/platform.c @@ -771,7 +771,7 @@ int __weak platform_pm_resume_noirq(struct device *dev) #endif /* !CONFIG_SUSPEND */ -#ifdef CONFIG_HIBERNATION +#ifdef CONFIG_HIBERNATE_CALLBACKS static int platform_pm_freeze(struct device *dev) { @@ -909,7 +909,7 @@ static int platform_pm_restore_noirq(struct device *dev) return ret; } -#else /* !CONFIG_HIBERNATION */ +#else /* !CONFIG_HIBERNATE_CALLBACKS */ #define platform_pm_freeze NULL #define platform_pm_thaw NULL @@ -920,7 +920,7 @@ static int platform_pm_restore_noirq(struct device *dev) #define platform_pm_poweroff_noirq NULL #define platform_pm_restore_noirq NULL -#endif /* !CONFIG_HIBERNATION */ +#endif /* !CONFIG_HIBERNATE_CALLBACKS */ #ifdef CONFIG_PM_RUNTIME diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index 052dc53..fbc5b6e 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -233,7 +233,7 @@ static int pm_op(struct device *dev, } break; #endif /* CONFIG_SUSPEND */ -#ifdef CONFIG_HIBERNATION +#ifdef CONFIG_HIBERNATE_CALLBACKS case PM_EVENT_FREEZE: case PM_EVENT_QUIESCE: if (ops->freeze) { @@ -260,7 +260,7 @@ static int pm_op(struct device *dev, suspend_report_result(ops->restore, error); } break; -#endif /* CONFIG_HIBERNATION */ +#endif /* CONFIG_HIBERNATE_CALLBACKS */ default: error = -EINVAL; } @@ -308,7 +308,7 @@ static int pm_noirq_op(struct device *dev, } break; #endif /* CONFIG_SUSPEND */ -#ifdef CONFIG_HIBERNATION +#ifdef CONFIG_HIBERNATE_CALLBACKS case PM_EVENT_FREEZE: case PM_EVENT_QUIESCE: if (ops->freeze_noirq) { @@ -335,7 +335,7 @@ static int pm_noirq_op(struct device *dev, suspend_report_result(ops->restore_noirq, error); } break; -#endif /* CONFIG_HIBERNATION */ +#endif /* CONFIG_HIBERNATE_CALLBACKS */ default: error = -EINVAL; } diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c index d86ea8b..135df16 100644 --- a/drivers/pci/pci-driver.c +++ b/drivers/pci/pci-driver.c @@ -781,7 +781,7 @@ static int pci_pm_resume(struct device *dev) #endif /* !CONFIG_SUSPEND */ -#ifdef CONFIG_HIBERNATION +#ifdef CONFIG_HIBERNATE_CALLBACKS static int pci_pm_freeze(struct device *dev) { @@ -970,7 +970,7 @@ static int pci_pm_restore(struct device *dev) return error; } -#else /* !CONFIG_HIBERNATION */ +#else /* !CONFIG_HIBERNATE_CALLBACKS */ #define pci_pm_freeze NULL #define pci_pm_freeze_noirq NULL @@ -981,7 +981,7 @@ static int pci_pm_restore(struct device *dev) #define pci_pm_restore NULL #define pci_pm_restore_noirq NULL -#endif /* !CONFIG_HIBERNATION */ +#endif /* !CONFIG_HIBERNATE_CALLBACKS */ #ifdef CONFIG_PM_RUNTIME diff --git a/drivers/xen/manage.c b/drivers/xen/manage.c index 95143dd..1ac9412 100644 --- a/drivers/xen/manage.c +++ b/drivers/xen/manage.c @@ -61,7 +61,7 @@ static void xen_post_suspend(int cancelled) xen_mm_unpin_all(); } -#ifdef CONFIG_HIBERNATION +#ifdef CONFIG_HIBERNATE_CALLBACKS static int xen_suspend(void *data) { struct suspend_info *si = data; @@ -173,7 +173,7 @@ out: #endif shutting_down = SHUTDOWN_INVALID; } -#endif /* CONFIG_HIBERNATION */ +#endif /* CONFIG_HIBERNATE_CALLBACKS */ struct shutdown_handler { const char *command; @@ -202,7 +202,7 @@ static void shutdown_handler(struct xenbus_watch *watch, { "poweroff", do_poweroff }, { "halt", do_poweroff }, { "reboot", do_reboot }, -#ifdef CONFIG_HIBERNATION +#ifdef CONFIG_HIBERNATE_CALLBACKS { "suspend", do_suspend }, #endif {NULL, NULL}, diff --git a/include/linux/suspend.h b/include/linux/suspend.h index 5a89e36..083ffea 100644 --- a/include/linux/suspend.h +++ b/include/linux/suspend.h @@ -249,6 +249,8 @@ extern void hibernation_set_ops(const struct platform_hibernation_ops *ops); extern int hibernate(void); extern bool system_entering_hibernation(void); #else /* CONFIG_HIBERNATION */ +static inline void register_nosave_region(unsigned long b, unsigned long e) {} +static inline void register_nosave_region_late(unsigned long b, unsigned long e) {} static inline int swsusp_page_is_forbidden(struct page *p) { return 0; } static inline void swsusp_set_page_free(struct page *p) {} static inline void swsusp_unset_page_free(struct page *p) {} @@ -297,14 +299,7 @@ static inline bool pm_wakeup_pending(void) { return false; } extern struct mutex pm_mutex; -#ifndef CONFIG_HIBERNATION -static inline void register_nosave_region(unsigned long b, unsigned long e) -{ -} -static inline void register_nosave_region_late(unsigned long b, unsigned long e) -{ -} - +#ifndef CONFIG_HIBERNATE_CALLBACKS static inline void lock_system_sleep(void) {} static inline void unlock_system_sleep(void) {} diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig index 4603f08..0497914 100644 --- a/kernel/power/Kconfig +++ b/kernel/power/Kconfig @@ -18,9 +18,13 @@ config SUSPEND_FREEZER Turning OFF this setting is NOT recommended! If in doubt, say Y. +config HIBERNATE_CALLBACKS + bool + config HIBERNATION bool "Hibernation (aka 'suspend to disk')" depends on SWAP && ARCH_HIBERNATION_POSSIBLE + select HIBERNATE_CALLBACKS select LZO_COMPRESS select LZO_DECOMPRESS ---help--- @@ -85,7 +89,7 @@ config PM_STD_PARTITION config PM_SLEEP def_bool y - depends on SUSPEND || HIBERNATION || XEN_SAVE_RESTORE + depends on SUSPEND || HIBERNATE_CALLBACKS || XEN_SAVE_RESTORE config PM_SLEEP_SMP def_bool y -- cgit v0.10.2 From d419e4c0f7584ffc5c72d9aeeaac485cc756ebcf Mon Sep 17 00:00:00 2001 From: Shriram Rajagopalan Date: Mon, 11 Apr 2011 22:54:48 +0200 Subject: fix XEN_SAVE_RESTORE Kconfig dependencies Make XEN_SAVE_RESTORE select HIBERNATE_CALLBACKS. Remove XEN_SAVE_RESTORE dependency from PM_SLEEP. Signed-off-by: Shriram Rajagopalan Acked-by: Ian Campbell Signed-off-by: Rafael J. Wysocki diff --git a/arch/x86/xen/Kconfig b/arch/x86/xen/Kconfig index 1c7121b..5cc821c 100644 --- a/arch/x86/xen/Kconfig +++ b/arch/x86/xen/Kconfig @@ -39,6 +39,7 @@ config XEN_MAX_DOMAIN_MEMORY config XEN_SAVE_RESTORE bool depends on XEN + select HIBERNATE_CALLBACKS default y config XEN_DEBUG_FS diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig index 0497914..6de9a8f 100644 --- a/kernel/power/Kconfig +++ b/kernel/power/Kconfig @@ -89,7 +89,7 @@ config PM_STD_PARTITION config PM_SLEEP def_bool y - depends on SUSPEND || HIBERNATE_CALLBACKS || XEN_SAVE_RESTORE + depends on SUSPEND || HIBERNATE_CALLBACKS config PM_SLEEP_SMP def_bool y -- cgit v0.10.2 From 39411f81eec7dc01677b14dda97684c0ce23ac1b Mon Sep 17 00:00:00 2001 From: "Luck, Tony" Date: Mon, 11 Apr 2011 12:06:12 -0700 Subject: xfs_destroy_workqueues() should not be tagged with__exit ia64 throws away .exit sections for the built-in CONFIG case, so routines that are used in other circumstances should not be tagged as __exit. Signed-off-by: Tony Luck Reviewed-by: Christoph Hellwig Signed-off-by: Alex Elder Signed-off-by: Linus Torvalds diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 67d5b2c..b38e58d 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -1741,7 +1741,7 @@ out: return -ENOMEM; } -STATIC void __exit +STATIC void xfs_destroy_workqueues(void) { destroy_workqueue(xfs_ail_wq); -- cgit v0.10.2 From a6360dd37e1a144ed11e6548371bade559a1e4df Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 11 Apr 2011 17:21:51 -0700 Subject: Linux 2.6.39-rc3 diff --git a/Makefile b/Makefile index 8392b64..322e733 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 2 PATCHLEVEL = 6 SUBLEVEL = 39 -EXTRAVERSION = -rc2 +EXTRAVERSION = -rc3 NAME = Flesh-Eating Bats with Fangs # *DOCUMENTATION* -- cgit v0.10.2 From 8fb27640d0e2b43c5584bf0087431b7b8d3c319a Mon Sep 17 00:00:00 2001 From: Yoshinori Sano Date: Sat, 9 Apr 2011 02:30:07 +0000 Subject: Btrfs: fix memory leaks in btrfs_new_inode() This patch fixes memory leaks in btrfs_new_inode(). Signed-off-by: Yoshinori Sano Signed-off-by: Chris Mason diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index cc60228..2d1208f 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -4526,14 +4526,17 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, BUG_ON(!path); inode = new_inode(root->fs_info->sb); - if (!inode) + if (!inode) { + btrfs_free_path(path); return ERR_PTR(-ENOMEM); + } if (dir) { trace_btrfs_inode_request(dir); ret = btrfs_set_inode_index(dir, index); if (ret) { + btrfs_free_path(path); iput(inode); return ERR_PTR(ret); } -- cgit v0.10.2 From 3387206f26e1b48703e810175b98611a4fd8e8ea Mon Sep 17 00:00:00 2001 From: Sergei Trofimovich Date: Mon, 11 Apr 2011 21:52:52 +0000 Subject: btrfs: properly handle overlapping areas in memmove_extent_buffer Fix data corruption caused by memcpy() usage on overlapping data. I've observed it first when found out usermode linux crash on btrfs. ?all chain is the following: ------------[ cut here ]------------ WARNING: at /home/slyfox/linux-2.6/fs/btrfs/extent_io.c:3900 memcpy_extent_buffer+0x1a5/0x219() Call Trace: 6fa39a58: [<601b495e>] _raw_spin_unlock_irqrestore+0x18/0x1c 6fa39a68: [<60029ad9>] warn_slowpath_common+0x59/0x70 6fa39aa8: [<60029b05>] warn_slowpath_null+0x15/0x17 6fa39ab8: [<600efc97>] memcpy_extent_buffer+0x1a5/0x219 6fa39b48: [<600efd9f>] memmove_extent_buffer+0x94/0x208 6fa39bc8: [<600becbf>] btrfs_del_items+0x214/0x473 6fa39c78: [<600ce1b0>] btrfs_delete_one_dir_name+0x7c/0xda 6fa39cc8: [<600dad6b>] __btrfs_unlink_inode+0xad/0x25d 6fa39d08: [<600d7864>] btrfs_start_transaction+0xe/0x10 6fa39d48: [<600dc9ff>] btrfs_unlink_inode+0x1b/0x3b 6fa39d78: [<600e04bc>] btrfs_unlink+0x70/0xef 6fa39dc8: [<6007f0d0>] vfs_unlink+0x58/0xa3 6fa39df8: [<60080278>] do_unlinkat+0xd4/0x162 6fa39e48: [<600517db>] call_rcu_sched+0xe/0x10 6fa39e58: [<600452a8>] __put_cred+0x58/0x5a 6fa39e78: [<6007446c>] sys_faccessat+0x154/0x166 6fa39ed8: [<60080317>] sys_unlink+0x11/0x13 6fa39ee8: [<60016b80>] handle_syscall+0x58/0x70 6fa39f08: [<60021377>] userspace+0x2d4/0x381 6fa39fc8: [<60014507>] fork_handler+0x62/0x69 ---[ end trace 70b0ca2ef0266b93 ]--- http://www.mail-archive.com/linux-btrfs@vger.kernel.org/msg09302.html Signed-off-by: Sergei Trofimovich Reviewed-by: Josef Bacik Signed-off-by: Chris Mason diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 77c65a0..864e049 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -3885,6 +3885,12 @@ static void move_pages(struct page *dst_page, struct page *src_page, kunmap_atomic(dst_kaddr, KM_USER0); } +static inline bool areas_overlap(unsigned long src, unsigned long dst, unsigned long len) +{ + unsigned long distance = (src > dst) ? src - dst : dst - src; + return distance < len; +} + static void copy_pages(struct page *dst_page, struct page *src_page, unsigned long dst_off, unsigned long src_off, unsigned long len) @@ -3892,10 +3898,12 @@ static void copy_pages(struct page *dst_page, struct page *src_page, char *dst_kaddr = kmap_atomic(dst_page, KM_USER0); char *src_kaddr; - if (dst_page != src_page) + if (dst_page != src_page) { src_kaddr = kmap_atomic(src_page, KM_USER1); - else + } else { src_kaddr = dst_kaddr; + BUG_ON(areas_overlap(src_off, dst_off, len)); + } memcpy(dst_kaddr + dst_off, src_kaddr + src_off, len); kunmap_atomic(dst_kaddr, KM_USER0); @@ -3970,7 +3978,7 @@ void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset, "len %lu len %lu\n", dst_offset, len, dst->len); BUG_ON(1); } - if (dst_offset < src_offset) { + if (!areas_overlap(src_offset, dst_offset, len)) { memcpy_extent_buffer(dst, dst_offset, src_offset, len); return; } -- cgit v0.10.2 From a1b75f7d961955e697ec377f90115e3517df98f9 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 8 Apr 2011 15:51:18 +0000 Subject: Btrfs: check for duplicate iov_base's when doing dio reads Apparently it is ok to submit a read to an IDE device with the same target page for different offsets. This is what Windows does under qemu. The problem is under DIO we expect them to be different buffers for checksumming reasons, and so this sort of thing will result in checksum errors, when in reality the file is fine. So when reading, check to make sure that all iov bases are different, and if they aren't fall back to buffered mode, since that will work out right. Thanks, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 2d1208f..edafc288 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -6153,6 +6153,7 @@ static ssize_t check_direct_IO(struct btrfs_root *root, int rw, struct kiocb *io unsigned long nr_segs) { int seg; + int i; size_t size; unsigned long addr; unsigned blocksize_mask = root->sectorsize - 1; @@ -6167,8 +6168,22 @@ static ssize_t check_direct_IO(struct btrfs_root *root, int rw, struct kiocb *io addr = (unsigned long)iov[seg].iov_base; size = iov[seg].iov_len; end += size; - if ((addr & blocksize_mask) || (size & blocksize_mask)) + if ((addr & blocksize_mask) || (size & blocksize_mask)) goto out; + + /* If this is a write we don't need to check anymore */ + if (rw & WRITE) + continue; + + /* + * Check to make sure we don't have duplicate iov_base's in this + * iovec, if so return EINVAL, otherwise we'll get csum errors + * when reading back. + */ + for (i = seg + 1; i < nr_segs; i++) { + if (iov[seg].iov_base == iov[i].iov_base) + goto out; + } } retval = 0; out: -- cgit v0.10.2 From 13f2696f1da9700d401db0ac2bc27ebc17068b22 Mon Sep 17 00:00:00 2001 From: Daniel J Blueman Date: Mon, 11 Apr 2011 15:56:31 +0000 Subject: fix user annotation in ioctl.c Fix address space annotation correct in ioctl.c. Signed-off-by: Daniel J Blueman BTRFS_BLOCK_GROUP_SYSTEM, @@ -2387,7 +2387,7 @@ long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg) up_read(&info->groups_sem); } - user_dest = (struct btrfs_ioctl_space_info *) + user_dest = (struct btrfs_ioctl_space_info __user *) (arg + sizeof(struct btrfs_ioctl_space_args)); if (copy_to_user(user_dest, dest_orig, alloc_size)) Reviewed-by: Josef Bacik Signed-off-by: Chris Mason diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index f9c93a9..f580a3a 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -2287,7 +2287,7 @@ long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg) struct btrfs_ioctl_space_info space; struct btrfs_ioctl_space_info *dest; struct btrfs_ioctl_space_info *dest_orig; - struct btrfs_ioctl_space_info *user_dest; + struct btrfs_ioctl_space_info __user *user_dest; struct btrfs_space_info *info; u64 types[] = {BTRFS_BLOCK_GROUP_DATA, BTRFS_BLOCK_GROUP_SYSTEM, -- cgit v0.10.2 From e15d0542426f063dc53b4c51bdfc11e0bbe4d298 Mon Sep 17 00:00:00 2001 From: Xin Zhong Date: Wed, 6 Apr 2011 07:33:51 +0000 Subject: Btrfs: fix subvolume mount by name problem when default mount subvolume is set We create two subvolumes (meego_root and meego_home) in btrfs root directory. And set meego_root as default mount subvolume. After we remount btrfs, meego_root is mounted to top directory by default. Then when we try to mount meego_home (subvol=meego_home) to a subdirectory, it failed. The problem is when default mount subvolume is set to meego_root, we search meego_home in meego_root but can not find it. So the solution is to add a new mount option (subvolrootid) to specify subvol id of root and search subvol name in it. For our case, now we can use "-o subvolrootid=0,subvol=meego_home) to mount meego_home. Detail information can be found in meego bugzilla: https://bugs.meego.com/show_bug.cgi?id=15055 Signed-off-by: Zhong, Xin Signed-off-by: Chris Mason diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 58e7de9..0ac712e 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -159,7 +159,7 @@ enum { Opt_compress_type, Opt_compress_force, Opt_compress_force_type, Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_discard, Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed, - Opt_enospc_debug, Opt_err, + Opt_enospc_debug, Opt_subvolrootid, Opt_err, }; static match_table_t tokens = { @@ -189,6 +189,7 @@ static match_table_t tokens = { {Opt_clear_cache, "clear_cache"}, {Opt_user_subvol_rm_allowed, "user_subvol_rm_allowed"}, {Opt_enospc_debug, "enospc_debug"}, + {Opt_subvolrootid, "subvolrootid=%d"}, {Opt_err, NULL}, }; @@ -232,6 +233,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) break; case Opt_subvol: case Opt_subvolid: + case Opt_subvolrootid: case Opt_device: /* * These are parsed by btrfs_parse_early_options @@ -388,7 +390,7 @@ out: */ static int btrfs_parse_early_options(const char *options, fmode_t flags, void *holder, char **subvol_name, u64 *subvol_objectid, - struct btrfs_fs_devices **fs_devices) + u64 *subvol_rootid, struct btrfs_fs_devices **fs_devices) { substring_t args[MAX_OPT_ARGS]; char *opts, *orig, *p; @@ -429,6 +431,18 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags, *subvol_objectid = intarg; } break; + case Opt_subvolrootid: + intarg = 0; + error = match_int(&args[0], &intarg); + if (!error) { + /* we want the original fs_tree */ + if (!intarg) + *subvol_rootid = + BTRFS_FS_TREE_OBJECTID; + else + *subvol_rootid = intarg; + } + break; case Opt_device: error = btrfs_scan_one_device(match_strdup(&args[0]), flags, holder, fs_devices); @@ -736,6 +750,7 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, fmode_t mode = FMODE_READ; char *subvol_name = NULL; u64 subvol_objectid = 0; + u64 subvol_rootid = 0; int error = 0; if (!(flags & MS_RDONLY)) @@ -743,7 +758,7 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, error = btrfs_parse_early_options(data, mode, fs_type, &subvol_name, &subvol_objectid, - &fs_devices); + &subvol_rootid, &fs_devices); if (error) return ERR_PTR(error); @@ -807,15 +822,17 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, s->s_flags |= MS_ACTIVE; } - root = get_default_root(s, subvol_objectid); - if (IS_ERR(root)) { - error = PTR_ERR(root); - deactivate_locked_super(s); - goto error_free_subvol_name; - } /* if they gave us a subvolume name bind mount into that */ if (strcmp(subvol_name, ".")) { struct dentry *new_root; + + root = get_default_root(s, subvol_rootid); + if (IS_ERR(root)) { + error = PTR_ERR(root); + deactivate_locked_super(s); + goto error_free_subvol_name; + } + mutex_lock(&root->d_inode->i_mutex); new_root = lookup_one_len(subvol_name, root, strlen(subvol_name)); @@ -836,6 +853,13 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, } dput(root); root = new_root; + } else { + root = get_default_root(s, subvol_objectid); + if (IS_ERR(root)) { + error = PTR_ERR(root); + deactivate_locked_super(s); + goto error_free_subvol_name; + } } kfree(subvol_name); -- cgit v0.10.2 From 7797069305d13252fd66cf722aa8f2cbeb3c95cd Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 5 Apr 2011 16:23:47 -0700 Subject: cifs: check for private_data before trying to put it cifs_close doesn't check that the filp->private_data is non-NULL before trying to put it. That can cause an oops in certain error conditions that can occur on open or lookup before the private_data is set. Reported-by: Ben Greear CC: Stable Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/file.c b/fs/cifs/file.c index c27d236..5ae061c 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -575,8 +575,10 @@ reopen_error_exit: int cifs_close(struct inode *inode, struct file *file) { - cifsFileInfo_put(file->private_data); - file->private_data = NULL; + if (file->private_data != NULL) { + cifsFileInfo_put(file->private_data); + file->private_data = NULL; + } /* return code from the ->release op is always ignored */ return 0; -- cgit v0.10.2 From bdf1b03e093bdbc571f404e751c7b0e2dca412ea Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 22 Feb 2011 20:17:19 -0500 Subject: cifs: replace /proc/fs/cifs/Experimental with a module parm This flag currently only affects whether we allow "zero-copy" writes with signing enabled. Typically we map pages in the pagecache directly into the write request. If signing is enabled however and the contents of the page change after the signature is calculated but before the write is sent then the signature will be wrong. Servers typically respond to this by closing down the socket. Still, this can provide a performance benefit so the "Experimental" flag was overloaded to allow this. That's really not a good place for this option however since it's not clear what that flag does. Move that flag instead to a new module parameter that better describes its purpose. That's also better since it can be set at module insertion time by configuring modprobe.d. Reviewed-by: Suresh Jayaraman Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/README b/fs/cifs/README index fe16835..74ab165f 100644 --- a/fs/cifs/README +++ b/fs/cifs/README @@ -685,22 +685,6 @@ LinuxExtensionsEnabled If set to one then the client will attempt to support and want to map the uid and gid fields to values supplied at mount (rather than the actual values, then set this to zero. (default 1) -Experimental When set to 1 used to enable certain experimental - features (currently enables multipage writes - when signing is enabled, the multipage write - performance enhancement was disabled when - signing turned on in case buffer was modified - just before it was sent, also this flag will - be used to use the new experimental directory change - notification code). When set to 2 enables - an additional experimental feature, "raw ntlmssp" - session establishment support (which allows - specifying "sec=ntlmssp" on mount). The Linux cifs - module will use ntlmv2 authentication encapsulated - in "raw ntlmssp" (not using SPNEGO) when - "sec=ntlmssp" is specified on mount. - This support also requires building cifs with - the CONFIG_CIFS_EXPERIMENTAL configuration flag. These experimental features and tracing can be enabled by changing flags in /proc/fs/cifs (after the cifs module has been installed or built into the diff --git a/fs/cifs/cifs_debug.c b/fs/cifs/cifs_debug.c index 65829d3..30d01bc 100644 --- a/fs/cifs/cifs_debug.c +++ b/fs/cifs/cifs_debug.c @@ -423,7 +423,6 @@ static const struct file_operations cifs_lookup_cache_proc_fops; static const struct file_operations traceSMB_proc_fops; static const struct file_operations cifs_multiuser_mount_proc_fops; static const struct file_operations cifs_security_flags_proc_fops; -static const struct file_operations cifs_experimental_proc_fops; static const struct file_operations cifs_linux_ext_proc_fops; void @@ -441,8 +440,6 @@ cifs_proc_init(void) proc_create("cifsFYI", 0, proc_fs_cifs, &cifsFYI_proc_fops); proc_create("traceSMB", 0, proc_fs_cifs, &traceSMB_proc_fops); proc_create("OplockEnabled", 0, proc_fs_cifs, &cifs_oplock_proc_fops); - proc_create("Experimental", 0, proc_fs_cifs, - &cifs_experimental_proc_fops); proc_create("LinuxExtensionsEnabled", 0, proc_fs_cifs, &cifs_linux_ext_proc_fops); proc_create("MultiuserMount", 0, proc_fs_cifs, @@ -469,7 +466,6 @@ cifs_proc_clean(void) remove_proc_entry("OplockEnabled", proc_fs_cifs); remove_proc_entry("SecurityFlags", proc_fs_cifs); remove_proc_entry("LinuxExtensionsEnabled", proc_fs_cifs); - remove_proc_entry("Experimental", proc_fs_cifs); remove_proc_entry("LookupCacheEnabled", proc_fs_cifs); remove_proc_entry("fs/cifs", NULL); } @@ -550,45 +546,6 @@ static const struct file_operations cifs_oplock_proc_fops = { .write = cifs_oplock_proc_write, }; -static int cifs_experimental_proc_show(struct seq_file *m, void *v) -{ - seq_printf(m, "%d\n", experimEnabled); - return 0; -} - -static int cifs_experimental_proc_open(struct inode *inode, struct file *file) -{ - return single_open(file, cifs_experimental_proc_show, NULL); -} - -static ssize_t cifs_experimental_proc_write(struct file *file, - const char __user *buffer, size_t count, loff_t *ppos) -{ - char c; - int rc; - - rc = get_user(c, buffer); - if (rc) - return rc; - if (c == '0' || c == 'n' || c == 'N') - experimEnabled = 0; - else if (c == '1' || c == 'y' || c == 'Y') - experimEnabled = 1; - else if (c == '2') - experimEnabled = 2; - - return count; -} - -static const struct file_operations cifs_experimental_proc_fops = { - .owner = THIS_MODULE, - .open = cifs_experimental_proc_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, - .write = cifs_experimental_proc_write, -}; - static int cifs_linux_ext_proc_show(struct seq_file *m, void *v) { seq_printf(m, "%d\n", linuxExtEnabled); diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index f297013..41c78e8 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -53,7 +53,6 @@ int cifsFYI = 0; int cifsERROR = 1; int traceSMB = 0; unsigned int oplockEnabled = 1; -unsigned int experimEnabled = 0; unsigned int linuxExtEnabled = 1; unsigned int lookupCacheEnabled = 1; unsigned int multiuser_mount = 0; @@ -82,6 +81,10 @@ module_param(echo_retries, ushort, 0644); MODULE_PARM_DESC(echo_retries, "Number of echo attempts before giving up and " "reconnecting server. Default: 5. 0 means " "never reconnect."); +bool sign_zero_copy; /* globals init to false automatically */ +module_param(sign_zero_copy, bool, 0644); +MODULE_PARM_DESC(sign_zero_copy, "Don't copy pages on write with signing " + "enabled. Default: N"); extern mempool_t *cifs_sm_req_poolp; extern mempool_t *cifs_req_poolp; extern mempool_t *cifs_mid_poolp; diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 17afb0f..10e4afe 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -817,7 +817,6 @@ GLOBAL_EXTERN unsigned int multiuser_mount; /* if enabled allows new sessions have the uid/password or Kerberos credential or equivalent for current user */ GLOBAL_EXTERN unsigned int oplockEnabled; -GLOBAL_EXTERN unsigned int experimEnabled; GLOBAL_EXTERN unsigned int lookupCacheEnabled; GLOBAL_EXTERN unsigned int global_secflags; /* if on, session setup sent with more secure ntlmssp2 challenge/resp */ @@ -827,6 +826,7 @@ GLOBAL_EXTERN unsigned int CIFSMaxBufSize; /* max size not including hdr */ GLOBAL_EXTERN unsigned int cifs_min_rcv; /* min size of big ntwrk buf pool */ GLOBAL_EXTERN unsigned int cifs_min_small; /* min size of small buf pool */ GLOBAL_EXTERN unsigned int cifs_max_pending; /* MAX requests at once to server*/ +GLOBAL_EXTERN bool sign_zero_copy; /* don't copy written pages with signing */ /* reconnect after this many failed echo attempts */ GLOBAL_EXTERN unsigned short echo_retries; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index 5ae061c..e2d7b6b 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -981,7 +981,7 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file, if (rc != 0) break; } - if (experimEnabled || (pTcon->ses->server && + if (sign_zero_copy || (pTcon->ses->server && ((pTcon->ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) == 0))) { @@ -1242,7 +1242,7 @@ static int cifs_writepages(struct address_space *mapping, } tcon = tlink_tcon(open_file->tlink); - if (!experimEnabled && tcon->ses->server->secMode & + if (!sign_zero_copy && tcon->ses->server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) { cifsFileInfo_put(open_file); kfree(iov); -- cgit v0.10.2 From 8727c8a85f3951ef0eef36a665f5dceebb4c495d Mon Sep 17 00:00:00 2001 From: Steve French Date: Fri, 25 Feb 2011 01:11:56 -0600 Subject: Allow user names longer than 32 bytes We artificially limited the user name to 32 bytes, but modern servers handle larger. Set the maximum length to a reasonable 256, and make the user name string dynamically allocated rather than a fixed size in session structure. Also clean up old checkpatch warning. Signed-off-by: Steve French diff --git a/fs/cifs/cifs_spnego.c b/fs/cifs/cifs_spnego.c index 4dfba82..33d2213 100644 --- a/fs/cifs/cifs_spnego.c +++ b/fs/cifs/cifs_spnego.c @@ -113,7 +113,7 @@ cifs_get_spnego_key(struct cifsSesInfo *sesInfo) MAX_MECH_STR_LEN + UID_KEY_LEN + (sizeof(uid_t) * 2) + CREDUID_KEY_LEN + (sizeof(uid_t) * 2) + - USER_KEY_LEN + strlen(sesInfo->userName) + + USER_KEY_LEN + strlen(sesInfo->user_name) + PID_KEY_LEN + (sizeof(pid_t) * 2) + 1; spnego_key = ERR_PTR(-ENOMEM); @@ -153,7 +153,7 @@ cifs_get_spnego_key(struct cifsSesInfo *sesInfo) sprintf(dp, ";creduid=0x%x", sesInfo->cred_uid); dp = description + strlen(description); - sprintf(dp, ";user=%s", sesInfo->userName); + sprintf(dp, ";user=%s", sesInfo->user_name); dp = description + strlen(description); sprintf(dp, ";pid=0x%x", current->pid); diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index a51585f..e307a28 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -469,15 +469,15 @@ static int calc_ntlmv2_hash(struct cifsSesInfo *ses, char *ntlmv2_hash, return rc; } - /* convert ses->userName to unicode and uppercase */ - len = strlen(ses->userName); + /* convert ses->user_name to unicode and uppercase */ + len = strlen(ses->user_name); user = kmalloc(2 + (len * 2), GFP_KERNEL); if (user == NULL) { cERROR(1, "calc_ntlmv2_hash: user mem alloc failure\n"); rc = -ENOMEM; goto calc_exit_2; } - len = cifs_strtoUCS((__le16 *)user, ses->userName, len, nls_cp); + len = cifs_strtoUCS((__le16 *)user, ses->user_name, len, nls_cp); UniStrupr(user); crypto_shash_update(&ses->server->secmech.sdeschmacmd5->shash, diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 41c78e8..0e0cc60 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -412,8 +412,8 @@ cifs_show_options(struct seq_file *s, struct vfsmount *m) if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MULTIUSER) seq_printf(s, ",multiuser"); - else if (tcon->ses->userName) - seq_printf(s, ",username=%s", tcon->ses->userName); + else if (tcon->ses->user_name) + seq_printf(s, ",username=%s", tcon->ses->user_name); if (tcon->ses->domainName) seq_printf(s, ",domain=%s", tcon->ses->domainName); diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 10e4afe..94cd874 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -38,9 +38,8 @@ #define MAX_TREE_SIZE (2 + MAX_SERVER_SIZE + 1 + MAX_SHARE_SIZE + 1) #define MAX_SERVER_SIZE 15 #define MAX_SHARE_SIZE 64 /* used to be 20, this should still be enough */ -#define MAX_USERNAME_SIZE 32 /* 32 is to allow for 15 char names + null - termination then *2 for unicode versions */ -#define MAX_PASSWORD_SIZE 512 /* max for windows seems to be 256 wide chars */ +#define MAX_USERNAME_SIZE 256 /* reasonable maximum for current servers */ +#define MAX_PASSWORD_SIZE 512 /* max for windows seems to be 256 wide chars */ #define CIFS_MIN_RCV_POOL 4 @@ -274,7 +273,7 @@ struct cifsSesInfo { int capabilities; char serverName[SERVER_NAME_LEN_WITH_NULL * 2]; /* BB make bigger for TCP names - will ipv6 and sctp addresses fit? */ - char userName[MAX_USERNAME_SIZE + 1]; + char *user_name; char *domainName; char *password; struct session_key auth_key; diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 6e2b2ad..54436a3 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -881,7 +881,8 @@ cifs_parse_mount_options(char *options, const char *devname, /* null user, ie anonymous, authentication */ vol->nullauth = 1; } - if (strnlen(value, 200) < 200) { + if (strnlen(value, MAX_USERNAME_SIZE) < + MAX_USERNAME_SIZE) { vol->username = value; } else { printk(KERN_WARNING "CIFS: username too long\n"); @@ -1808,7 +1809,9 @@ cifs_find_smb_ses(struct TCP_Server_Info *server, struct smb_vol *vol) break; default: /* anything else takes username/password */ - if (strncmp(ses->userName, vol->username, + if (ses->user_name == NULL) + continue; + if (strncmp(ses->user_name, vol->username, MAX_USERNAME_SIZE)) continue; if (strlen(vol->username) != 0 && @@ -1906,9 +1909,11 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) else sprintf(ses->serverName, "%pI4", &addr->sin_addr); - if (volume_info->username) - strncpy(ses->userName, volume_info->username, - MAX_USERNAME_SIZE); + if (volume_info->username) { + ses->user_name = kstrdup(volume_info->username, GFP_KERNEL); + if (!ses->user_name) + goto get_ses_fail; + } /* volume_info->password freed at unmount */ if (volume_info->password) { diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index 2a930a7..7228179 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -100,6 +100,7 @@ sesInfoFree(struct cifsSesInfo *buf_to_free) memset(buf_to_free->password, 0, strlen(buf_to_free->password)); kfree(buf_to_free->password); } + kfree(buf_to_free->user_name); kfree(buf_to_free->domainName); kfree(buf_to_free); } diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c index 1676570..006485f 100644 --- a/fs/cifs/sess.c +++ b/fs/cifs/sess.c @@ -219,12 +219,12 @@ static void unicode_ssetup_strings(char **pbcc_area, struct cifsSesInfo *ses, bcc_ptr++; } */ /* copy user */ - if (ses->userName == NULL) { + if (ses->user_name == NULL) { /* null user mount */ *bcc_ptr = 0; *(bcc_ptr+1) = 0; } else { - bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, ses->userName, + bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, ses->user_name, MAX_USERNAME_SIZE, nls_cp); } bcc_ptr += 2 * bytes_ret; @@ -244,12 +244,11 @@ static void ascii_ssetup_strings(char **pbcc_area, struct cifsSesInfo *ses, /* copy user */ /* BB what about null user mounts - check that we do this BB */ /* copy user */ - if (ses->userName == NULL) { - /* BB what about null user mounts - check that we do this BB */ - } else { - strncpy(bcc_ptr, ses->userName, MAX_USERNAME_SIZE); - } - bcc_ptr += strnlen(ses->userName, MAX_USERNAME_SIZE); + if (ses->user_name != NULL) + strncpy(bcc_ptr, ses->user_name, MAX_USERNAME_SIZE); + /* else null user mount */ + + bcc_ptr += strnlen(ses->user_name, MAX_USERNAME_SIZE); *bcc_ptr = 0; bcc_ptr++; /* account for null termination */ @@ -523,14 +522,14 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, tmp += len; } - if (ses->userName == NULL) { + if (ses->user_name == NULL) { sec_blob->UserName.BufferOffset = cpu_to_le32(tmp - pbuffer); sec_blob->UserName.Length = 0; sec_blob->UserName.MaximumLength = 0; tmp += 2; } else { int len; - len = cifs_strtoUCS((__le16 *)tmp, ses->userName, + len = cifs_strtoUCS((__le16 *)tmp, ses->user_name, MAX_USERNAME_SIZE, nls_cp); len *= 2; /* unicode is 2 bytes each */ sec_blob->UserName.BufferOffset = cpu_to_le32(tmp - pbuffer); -- cgit v0.10.2 From 13c5a93e7005d7dae0b6d070d25203593e692d13 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 11 Apr 2011 15:45:29 -0400 Subject: Btrfs: avoid taking the trans_mutex in btrfs_end_transaction I've been working on making our O_DIRECT latency not suck and I noticed we were taking the trans_mutex in btrfs_end_transaction. So to do this we convert num_writers and use_count to atomic_t's and just decrement them in btrfs_end_transaction. Instead of deleting the transaction from the trans list in put_transaction we do that in btrfs_commit_transaction() since that's the only time it actually needs to be removed from the list. Thanks, Signed-off-by: Josef Bacik diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index a272bfd..ef6865c 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3136,7 +3136,7 @@ static int btrfs_cleanup_transaction(struct btrfs_root *root) btrfs_destroy_pinned_extent(root, root->fs_info->pinned_extents); - t->use_count = 0; + atomic_set(&t->use_count, 0); list_del_init(&t->list); memset(t, 0, sizeof(*t)); kmem_cache_free(btrfs_transaction_cachep, t); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 4583008..c571734 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -32,10 +32,8 @@ static noinline void put_transaction(struct btrfs_transaction *transaction) { - WARN_ON(transaction->use_count == 0); - transaction->use_count--; - if (transaction->use_count == 0) { - list_del_init(&transaction->list); + WARN_ON(atomic_read(&transaction->use_count) == 0); + if (atomic_dec_and_test(&transaction->use_count)) { memset(transaction, 0, sizeof(*transaction)); kmem_cache_free(btrfs_transaction_cachep, transaction); } @@ -60,14 +58,14 @@ static noinline int join_transaction(struct btrfs_root *root) if (!cur_trans) return -ENOMEM; root->fs_info->generation++; - cur_trans->num_writers = 1; + atomic_set(&cur_trans->num_writers, 1); cur_trans->num_joined = 0; cur_trans->transid = root->fs_info->generation; init_waitqueue_head(&cur_trans->writer_wait); init_waitqueue_head(&cur_trans->commit_wait); cur_trans->in_commit = 0; cur_trans->blocked = 0; - cur_trans->use_count = 1; + atomic_set(&cur_trans->use_count, 1); cur_trans->commit_done = 0; cur_trans->start_time = get_seconds(); @@ -88,7 +86,7 @@ static noinline int join_transaction(struct btrfs_root *root) root->fs_info->running_transaction = cur_trans; spin_unlock(&root->fs_info->new_trans_lock); } else { - cur_trans->num_writers++; + atomic_inc(&cur_trans->num_writers); cur_trans->num_joined++; } @@ -145,7 +143,7 @@ static void wait_current_trans(struct btrfs_root *root) cur_trans = root->fs_info->running_transaction; if (cur_trans && cur_trans->blocked) { DEFINE_WAIT(wait); - cur_trans->use_count++; + atomic_inc(&cur_trans->use_count); while (1) { prepare_to_wait(&root->fs_info->transaction_wait, &wait, TASK_UNINTERRUPTIBLE); @@ -205,7 +203,7 @@ again: } cur_trans = root->fs_info->running_transaction; - cur_trans->use_count++; + atomic_inc(&cur_trans->use_count); if (type != TRANS_JOIN_NOLOCK) mutex_unlock(&root->fs_info->trans_mutex); @@ -336,7 +334,7 @@ int btrfs_wait_for_commit(struct btrfs_root *root, u64 transid) goto out_unlock; /* nothing committing|committed */ } - cur_trans->use_count++; + atomic_inc(&cur_trans->use_count); mutex_unlock(&root->fs_info->trans_mutex); wait_for_commit(root, cur_trans); @@ -466,18 +464,14 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, wake_up_process(info->transaction_kthread); } - if (lock) - mutex_lock(&info->trans_mutex); WARN_ON(cur_trans != info->running_transaction); - WARN_ON(cur_trans->num_writers < 1); - cur_trans->num_writers--; + WARN_ON(atomic_read(&cur_trans->num_writers) < 1); + atomic_dec(&cur_trans->num_writers); smp_mb(); if (waitqueue_active(&cur_trans->writer_wait)) wake_up(&cur_trans->writer_wait); put_transaction(cur_trans); - if (lock) - mutex_unlock(&info->trans_mutex); if (current->journal_info == trans) current->journal_info = NULL; @@ -1187,7 +1181,7 @@ int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans, /* take transaction reference */ mutex_lock(&root->fs_info->trans_mutex); cur_trans = trans->transaction; - cur_trans->use_count++; + atomic_inc(&cur_trans->use_count); mutex_unlock(&root->fs_info->trans_mutex); btrfs_end_transaction(trans, root); @@ -1246,7 +1240,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, mutex_lock(&root->fs_info->trans_mutex); if (cur_trans->in_commit) { - cur_trans->use_count++; + atomic_inc(&cur_trans->use_count); mutex_unlock(&root->fs_info->trans_mutex); btrfs_end_transaction(trans, root); @@ -1268,7 +1262,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, prev_trans = list_entry(cur_trans->list.prev, struct btrfs_transaction, list); if (!prev_trans->commit_done) { - prev_trans->use_count++; + atomic_inc(&prev_trans->use_count); mutex_unlock(&root->fs_info->trans_mutex); wait_for_commit(root, prev_trans); @@ -1309,14 +1303,14 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, TASK_UNINTERRUPTIBLE); smp_mb(); - if (cur_trans->num_writers > 1) + if (atomic_read(&cur_trans->num_writers) > 1) schedule_timeout(MAX_SCHEDULE_TIMEOUT); else if (should_grow) schedule_timeout(1); mutex_lock(&root->fs_info->trans_mutex); finish_wait(&cur_trans->writer_wait, &wait); - } while (cur_trans->num_writers > 1 || + } while (atomic_read(&cur_trans->num_writers) > 1 || (should_grow && cur_trans->num_joined != joined)); ret = create_pending_snapshots(trans, root->fs_info); @@ -1403,6 +1397,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, wake_up(&cur_trans->commit_wait); + list_del_init(&cur_trans->list); put_transaction(cur_trans); put_transaction(cur_trans); diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 229a594..e441acc 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -27,11 +27,11 @@ struct btrfs_transaction { * total writers in this transaction, it must be zero before the * transaction can end */ - unsigned long num_writers; + atomic_t num_writers; unsigned long num_joined; int in_commit; - int use_count; + atomic_t use_count; int commit_done; int blocked; struct list_head list; -- cgit v0.10.2 From 507903b81840a70cc6a179d4eb03584ad50e8c5b Mon Sep 17 00:00:00 2001 From: Arne Jansen Date: Wed, 6 Apr 2011 10:02:20 +0000 Subject: btrfs: using cached extent_state in set/unlock combinations In several places the sequence (set_extent_uptodate, unlock_extent) is used. This leads to a duplicate lookup of the extent state. This patch lets set_extent_uptodate return a cached extent_state which can be passed to unlock_extent_cached. The occurences of the above sequences are updated to use the cache. Only end_bio_extent_readpage is updated that it first gets a cached state to pass it to the readpage_end_io_hook as the prototype requested and is later on being used for set/unlock. Signed-off-by: Arne Jansen Signed-off-by: Chris Mason diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 864e049..8dcfb77 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -690,6 +690,17 @@ static void cache_state(struct extent_state *state, } } +static void uncache_state(struct extent_state **cached_ptr) +{ + if (cached_ptr && (*cached_ptr)) { + struct extent_state *state = *cached_ptr; + if (state->state & (EXTENT_IOBITS | EXTENT_BOUNDARY)) { + *cached_ptr = NULL; + free_extent_state(state); + } + } +} + /* * set some bits on a range in the tree. This may require allocations or * sleeping, so the gfp mask is used to indicate what is allowed. @@ -940,10 +951,10 @@ static int clear_extent_new(struct extent_io_tree *tree, u64 start, u64 end, } int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end, - gfp_t mask) + struct extent_state **cached_state, gfp_t mask) { - return set_extent_bit(tree, start, end, EXTENT_UPTODATE, 0, NULL, - NULL, mask); + return set_extent_bit(tree, start, end, EXTENT_UPTODATE, 0, + NULL, cached_state, mask); } static int clear_extent_uptodate(struct extent_io_tree *tree, u64 start, @@ -1012,8 +1023,7 @@ int unlock_extent_cached(struct extent_io_tree *tree, u64 start, u64 end, mask); } -int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end, - gfp_t mask) +int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask) { return clear_extent_bit(tree, start, end, EXTENT_LOCKED, 1, 0, NULL, mask); @@ -1735,6 +1745,9 @@ static void end_bio_extent_readpage(struct bio *bio, int err) do { struct page *page = bvec->bv_page; + struct extent_state *cached = NULL; + struct extent_state *state; + tree = &BTRFS_I(page->mapping->host)->io_tree; start = ((u64)page->index << PAGE_CACHE_SHIFT) + @@ -1749,9 +1762,20 @@ static void end_bio_extent_readpage(struct bio *bio, int err) if (++bvec <= bvec_end) prefetchw(&bvec->bv_page->flags); + spin_lock(&tree->lock); + state = find_first_extent_bit_state(tree, start, 0); + if (state) { + /* + * take a reference on the state, unlock will drop + * the ref + */ + cache_state(state, &cached); + } + spin_unlock(&tree->lock); + if (uptodate && tree->ops && tree->ops->readpage_end_io_hook) { ret = tree->ops->readpage_end_io_hook(page, start, end, - NULL); + state); if (ret) uptodate = 0; } @@ -1764,15 +1788,16 @@ static void end_bio_extent_readpage(struct bio *bio, int err) test_bit(BIO_UPTODATE, &bio->bi_flags); if (err) uptodate = 0; + uncache_state(&cached); continue; } } if (uptodate) { - set_extent_uptodate(tree, start, end, + set_extent_uptodate(tree, start, end, &cached, GFP_ATOMIC); } - unlock_extent(tree, start, end, GFP_ATOMIC); + unlock_extent_cached(tree, start, end, &cached, GFP_ATOMIC); if (whole_page) { if (uptodate) { @@ -1811,6 +1836,7 @@ static void end_bio_extent_preparewrite(struct bio *bio, int err) do { struct page *page = bvec->bv_page; + struct extent_state *cached = NULL; tree = &BTRFS_I(page->mapping->host)->io_tree; start = ((u64)page->index << PAGE_CACHE_SHIFT) + @@ -1821,13 +1847,14 @@ static void end_bio_extent_preparewrite(struct bio *bio, int err) prefetchw(&bvec->bv_page->flags); if (uptodate) { - set_extent_uptodate(tree, start, end, GFP_ATOMIC); + set_extent_uptodate(tree, start, end, &cached, + GFP_ATOMIC); } else { ClearPageUptodate(page); SetPageError(page); } - unlock_extent(tree, start, end, GFP_ATOMIC); + unlock_extent_cached(tree, start, end, &cached, GFP_ATOMIC); } while (bvec >= bio->bi_io_vec); @@ -2016,14 +2043,17 @@ static int __extent_read_full_page(struct extent_io_tree *tree, while (cur <= end) { if (cur >= last_byte) { char *userpage; + struct extent_state *cached = NULL; + iosize = PAGE_CACHE_SIZE - page_offset; userpage = kmap_atomic(page, KM_USER0); memset(userpage + page_offset, 0, iosize); flush_dcache_page(page); kunmap_atomic(userpage, KM_USER0); set_extent_uptodate(tree, cur, cur + iosize - 1, - GFP_NOFS); - unlock_extent(tree, cur, cur + iosize - 1, GFP_NOFS); + &cached, GFP_NOFS); + unlock_extent_cached(tree, cur, cur + iosize - 1, + &cached, GFP_NOFS); break; } em = get_extent(inode, page, page_offset, cur, @@ -2063,14 +2093,17 @@ static int __extent_read_full_page(struct extent_io_tree *tree, /* we've found a hole, just zero and go on */ if (block_start == EXTENT_MAP_HOLE) { char *userpage; + struct extent_state *cached = NULL; + userpage = kmap_atomic(page, KM_USER0); memset(userpage + page_offset, 0, iosize); flush_dcache_page(page); kunmap_atomic(userpage, KM_USER0); set_extent_uptodate(tree, cur, cur + iosize - 1, - GFP_NOFS); - unlock_extent(tree, cur, cur + iosize - 1, GFP_NOFS); + &cached, GFP_NOFS); + unlock_extent_cached(tree, cur, cur + iosize - 1, + &cached, GFP_NOFS); cur = cur + iosize; page_offset += iosize; continue; @@ -2789,9 +2822,12 @@ int extent_prepare_write(struct extent_io_tree *tree, iocount++; block_start = block_start + iosize; } else { - set_extent_uptodate(tree, block_start, cur_end, + struct extent_state *cached = NULL; + + set_extent_uptodate(tree, block_start, cur_end, &cached, GFP_NOFS); - unlock_extent(tree, block_start, cur_end, GFP_NOFS); + unlock_extent_cached(tree, block_start, cur_end, + &cached, GFP_NOFS); block_start = cur_end + 1; } page_offset = block_start & (PAGE_CACHE_SIZE - 1); @@ -3457,7 +3493,7 @@ int set_extent_buffer_uptodate(struct extent_io_tree *tree, num_pages = num_extent_pages(eb->start, eb->len); set_extent_uptodate(tree, eb->start, eb->start + eb->len - 1, - GFP_NOFS); + NULL, GFP_NOFS); for (i = 0; i < num_pages; i++) { page = extent_buffer_page(eb, i); if ((i == 0 && (eb->start & (PAGE_CACHE_SIZE - 1))) || diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index f62c544..af2d717 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -208,7 +208,7 @@ int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits, int exclusive_bits, u64 *failed_start, struct extent_state **cached_state, gfp_t mask); int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end, - gfp_t mask); + struct extent_state **cached_state, gfp_t mask); int set_extent_new(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask); int set_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end, diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index edafc288..5a993e0 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -5226,7 +5226,7 @@ again: btrfs_mark_buffer_dirty(leaf); } set_extent_uptodate(io_tree, em->start, - extent_map_end(em) - 1, GFP_NOFS); + extent_map_end(em) - 1, NULL, GFP_NOFS); goto insert; } else { printk(KERN_ERR "btrfs unknown found_type %d\n", found_type); -- cgit v0.10.2 From 2e325d5973b99bb7421e689351ca74f349eee5ea Mon Sep 17 00:00:00 2001 From: Steve French Date: Tue, 8 Mar 2011 05:51:19 +0000 Subject: Max share size is too small Max share name was set to 64, and (at least for Windows) can be 80. Signed-off-by: Steve French diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 94cd874..1cff9aa 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -37,7 +37,7 @@ #define MAX_TREE_SIZE (2 + MAX_SERVER_SIZE + 1 + MAX_SHARE_SIZE + 1) #define MAX_SERVER_SIZE 15 -#define MAX_SHARE_SIZE 64 /* used to be 20, this should still be enough */ +#define MAX_SHARE_SIZE 80 #define MAX_USERNAME_SIZE 256 /* reasonable maximum for current servers */ #define MAX_PASSWORD_SIZE 512 /* max for windows seems to be 256 wide chars */ -- cgit v0.10.2 From 6da9791061ca3b7b3c7eb7350eb452443f40a0e5 Mon Sep 17 00:00:00 2001 From: Steve French Date: Sun, 13 Mar 2011 18:55:55 +0000 Subject: Elminate sparse __CHECK_ENDIAN__ warnings on port conversion Ports are __be16 not unsigned short int Eliminates the remaining fixable endian warnings: ~/cifs-2.6$ make modules C=1 M=fs/cifs CF=-D__CHECK_ENDIAN__ CHECK fs/cifs/connect.c fs/cifs/connect.c:2408:23: warning: incorrect type in assignment (different base types) fs/cifs/connect.c:2408:23: expected unsigned short *sport fs/cifs/connect.c:2408:23: got restricted __be16 * fs/cifs/connect.c:2410:23: warning: incorrect type in assignment (different base types) fs/cifs/connect.c:2410:23: expected unsigned short *sport fs/cifs/connect.c:2410:23: got restricted __be16 * fs/cifs/connect.c:2416:24: warning: incorrect type in assignment (different base types) fs/cifs/connect.c:2416:24: expected unsigned short [unsigned] [short] fs/cifs/connect.c:2416:24: got restricted __be16 [usertype] fs/cifs/connect.c:2423:24: warning: incorrect type in assignment (different base types) fs/cifs/connect.c:2423:24: expected unsigned short [unsigned] [short] fs/cifs/connect.c:2423:24: got restricted __be16 [usertype] fs/cifs/connect.c:2326:23: warning: incorrect type in assignment (different base types) fs/cifs/connect.c:2326:23: expected unsigned short [unsigned] sport fs/cifs/connect.c:2326:23: got restricted __be16 [usertype] sin6_port fs/cifs/connect.c:2330:23: warning: incorrect type in assignment (different base types) fs/cifs/connect.c:2330:23: expected unsigned short [unsigned] sport fs/cifs/connect.c:2330:23: got restricted __be16 [usertype] sin_port fs/cifs/connect.c:2394:22: warning: restricted __be16 degrades to integer Signed-off-by: Steve French diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 54436a3..94a05a6 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1473,7 +1473,7 @@ srcip_matches(struct sockaddr *srcaddr, struct sockaddr *rhs) static bool match_port(struct TCP_Server_Info *server, struct sockaddr *addr) { - unsigned short int port, *sport; + __be16 port, *sport; switch (addr->sa_family) { case AF_INET: @@ -2281,7 +2281,7 @@ static int generic_ip_connect(struct TCP_Server_Info *server) { int rc = 0; - unsigned short int sport; + __be16 sport; int slen, sfamily; struct socket *socket = server->ssocket; struct sockaddr *saddr; @@ -2366,7 +2366,7 @@ generic_ip_connect(struct TCP_Server_Info *server) static int ip_connect(struct TCP_Server_Info *server) { - unsigned short int *sport; + __be16 *sport; struct sockaddr_in6 *addr6 = (struct sockaddr_in6 *)&server->dstaddr; struct sockaddr_in *addr = (struct sockaddr_in *)&server->dstaddr; -- cgit v0.10.2 From 5443d130aa4990424a8e64984e64b50ec70661bb Mon Sep 17 00:00:00 2001 From: Steve French Date: Sun, 13 Mar 2011 05:08:25 +0000 Subject: various endian fixes to cifs make modules C=2 M=fs/cifs CF=-D__CHECK_ENDIAN__ Found for example: CHECK fs/cifs/cifssmb.c fs/cifs/cifssmb.c:728:22: warning: incorrect type in assignment (different base types) fs/cifs/cifssmb.c:728:22: expected unsigned short [unsigned] [usertype] Tid fs/cifs/cifssmb.c:728:22: got restricted __le16 [usertype] fs/cifs/cifssmb.c:1883:45: warning: incorrect type in assignment (different base types) fs/cifs/cifssmb.c:1883:45: expected long long [signed] [usertype] fl_start fs/cifs/cifssmb.c:1883:45: got restricted __le64 [usertype] start fs/cifs/cifssmb.c:1884:54: warning: restricted __le64 degrades to integer fs/cifs/cifssmb.c:1885:58: warning: restricted __le64 degrades to integer fs/cifs/cifssmb.c:1886:43: warning: incorrect type in assignment (different base types) fs/cifs/cifssmb.c:1886:43: expected unsigned int [unsigned] fl_pid fs/cifs/cifssmb.c:1886:43: got restricted __le32 [usertype] pid In checking new smb2 code for missing endian conversions, I noticed some endian errors had crept in over the last few releases into the cifs code (symlink, ntlmssp, posix lock, and also a less problematic warning in fscache). A followon patch will address a few smb2 endian problems. Reviewed-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/cache.c b/fs/cifs/cache.c index e654dfd..53d57a3 100644 --- a/fs/cifs/cache.c +++ b/fs/cifs/cache.c @@ -50,7 +50,7 @@ void cifs_fscache_unregister(void) */ struct cifs_server_key { uint16_t family; /* address family */ - uint16_t port; /* IP port */ + __be16 port; /* IP port */ union { struct in_addr ipv4_addr; struct in6_addr ipv6_addr; diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 2644a5d..3919ec8 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -729,7 +729,7 @@ CIFSSMBEcho(struct TCP_Server_Info *server) return rc; /* set up echo request */ - smb->hdr.Tid = cpu_to_le16(0xffff); + smb->hdr.Tid = 0xffff; smb->hdr.WordCount = 1; put_unaligned_le16(1, &smb->EchoCount); put_bcc_le(1, &smb->hdr); @@ -1884,10 +1884,10 @@ CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, __constant_cpu_to_le16(CIFS_WRLCK)) pLockData->fl_type = F_WRLCK; - pLockData->fl_start = parm_data->start; - pLockData->fl_end = parm_data->start + - parm_data->length - 1; - pLockData->fl_pid = parm_data->pid; + pLockData->fl_start = le64_to_cpu(parm_data->start); + pLockData->fl_end = pLockData->fl_start + + le64_to_cpu(parm_data->length) - 1; + pLockData->fl_pid = le32_to_cpu(parm_data->pid); } } diff --git a/fs/cifs/link.c b/fs/cifs/link.c index e8804d3..ce417a9 100644 --- a/fs/cifs/link.c +++ b/fs/cifs/link.c @@ -239,7 +239,7 @@ CIFSQueryMFSymLink(const int xid, struct cifsTconInfo *tcon, if (rc != 0) return rc; - if (file_info.EndOfFile != CIFS_MF_SYMLINK_FILE_SIZE) { + if (file_info.EndOfFile != cpu_to_le64(CIFS_MF_SYMLINK_FILE_SIZE)) { CIFSSMBClose(xid, tcon, netfid); /* it's not a symlink */ return -EINVAL; @@ -316,7 +316,7 @@ CIFSCheckMFSymlink(struct cifs_fattr *fattr, if (rc != 0) goto out; - if (file_info.EndOfFile != CIFS_MF_SYMLINK_FILE_SIZE) { + if (file_info.EndOfFile != cpu_to_le64(CIFS_MF_SYMLINK_FILE_SIZE)) { CIFSSMBClose(xid, pTcon, netfid); /* it's not a symlink */ goto out; diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c index 006485f..f6728eb 100644 --- a/fs/cifs/sess.c +++ b/fs/cifs/sess.c @@ -404,8 +404,8 @@ static int decode_ntlmssp_challenge(char *bcc_ptr, int blob_len, /* BB spec says that if AvId field of MsvAvTimestamp is populated then we must set the MIC field of the AUTHENTICATE_MESSAGE */ ses->ntlmssp->server_flags = le32_to_cpu(pblob->NegotiateFlags); - tioffset = cpu_to_le16(pblob->TargetInfoArray.BufferOffset); - tilen = cpu_to_le16(pblob->TargetInfoArray.Length); + tioffset = le32_to_cpu(pblob->TargetInfoArray.BufferOffset); + tilen = le16_to_cpu(pblob->TargetInfoArray.Length); if (tilen) { ses->auth_key.response = kmalloc(tilen, GFP_KERNEL); if (!ses->auth_key.response) { -- cgit v0.10.2 From 70945643722ffeac779d2529a348f99567fa5c33 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Mon, 14 Mar 2011 13:48:08 -0400 Subject: cifs: always do is_path_accessible check in cifs_mount Currently, we skip doing the is_path_accessible check in cifs_mount if there is no prefixpath. I have a report of at least one server however that allows a TREE_CONNECT to a share that has a DFS referral at its root. The reporter in this case was using a UNC that had no prefixpath, so the is_path_accessible check was not triggered and the box later hit a BUG() because we were chasing a DFS referral on the root dentry for the mount. This patch fixes this by removing the check for a zero-length prefixpath. That should make the is_path_accessible check be done in this situation and should allow the client to chase the DFS referral at mount time instead. Cc: stable@kernel.org Reported-and-Tested-by: Yogesh Sharma Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 94a05a6..5eacb89 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -2831,7 +2831,7 @@ try_mount_again: remote_path_check: /* check if a whole path (including prepath) is not remote */ - if (!rc && cifs_sb->prepathlen && tcon) { + if (!rc && tcon) { /* build_path_to_root works only when we have a valid tcon */ full_path = cifs_build_path_to_root(cifs_sb, tcon); if (full_path == NULL) { -- cgit v0.10.2 From 8679b0dba7cb98842cbe37f61ef05ef64106334c Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Wed, 16 Mar 2011 15:15:30 -0400 Subject: cifs: fix broken BCC check in is_valid_oplock_break The BCC is still __le16 at this point, and in any case we need to use the get_bcc_le macro to make sure we don't hit alignment problems. Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index 7228179..0c684ae 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -521,7 +521,7 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv) (struct smb_com_transaction_change_notify_rsp *)buf; struct file_notify_information *pnotify; __u32 data_offset = 0; - if (pSMBr->ByteCount > sizeof(struct file_notify_information)) { + if (get_bcc_le(buf) > sizeof(struct file_notify_information)) { data_offset = le32_to_cpu(pSMBr->DataOffset); pnotify = (struct file_notify_information *) -- cgit v0.10.2 From 2b6c26a0a62cc0bab0ad487533d5581d7c293fef Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Fri, 25 Mar 2011 16:25:57 -0400 Subject: cifs: set ra_pages in backing_dev_info Commit 522440ed made cifs set backing_dev_info on the mapping attached to new inodes. This change caused a fairly significant read performance regression, as cifs started doing page-sized reads exclusively. By virtue of the fact that they're allocated as part of cifs_sb_info by kzalloc, the ra_pages on cifs BDIs get set to 0, which prevents any readahead. This forces the normal read codepaths to use readpage instead of readpages causing a four-fold increase in the number of read calls with the default rsize. Fix it by setting ra_pages in the BDI to the same value as that in the default_backing_dev_info. Fixes https://bugzilla.kernel.org/show_bug.cgi?id=31662 Cc: stable@kernel.org Reported-and-Tested-by: Till Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 0e0cc60..e3352a1 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -130,6 +130,7 @@ cifs_read_super(struct super_block *sb, void *data, kfree(cifs_sb); return rc; } + cifs_sb->bdi.ra_pages = default_backing_dev_info.ra_pages; #ifdef CONFIG_CIFS_DFS_UPCALL /* copy mount params to sb for use in submounts */ -- cgit v0.10.2 From c0c7b905e911a1f1faf515a24e849d4ffcdd0a8a Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Thu, 31 Mar 2011 17:32:54 -0400 Subject: cifs: clean up length checks in check2ndT2 Thus spake David Howells: The code that follows this: remaining = total_data_size - data_in_this_rsp; if (remaining == 0) return 0; else if (remaining < 0) { generates better code if you drop the 'remaining' variable and compare the values directly. Clean it up per his recommendation... Reported-and-acked-by: David Howells Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 5eacb89..709fd9d 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -248,24 +248,24 @@ static int check2ndT2(struct smb_hdr *pSMB, unsigned int maxBufSize) total_data_size = get_unaligned_le16(&pSMBt->t2_rsp.TotalDataCount); data_in_this_rsp = get_unaligned_le16(&pSMBt->t2_rsp.DataCount); - remaining = total_data_size - data_in_this_rsp; - - if (remaining == 0) + if (total_data_size == data_in_this_rsp) return 0; - else if (remaining < 0) { + else if (total_data_size < data_in_this_rsp) { cFYI(1, "total data %d smaller than data in frame %d", total_data_size, data_in_this_rsp); return -EINVAL; - } else { - cFYI(1, "missing %d bytes from transact2, check next response", - remaining); - if (total_data_size > maxBufSize) { - cERROR(1, "TotalDataSize %d is over maximum buffer %d", - total_data_size, maxBufSize); - return -EINVAL; - } - return remaining; } + + remaining = total_data_size - data_in_this_rsp; + + cFYI(1, "missing %d bytes from transact2, check next response", + remaining); + if (total_data_size > maxBufSize) { + cERROR(1, "TotalDataSize %d is over maximum buffer %d", + total_data_size, maxBufSize); + return -EINVAL; + } + return remaining; } static int coalesce_t2(struct smb_hdr *psecond, struct smb_hdr *pTargetSMB) -- cgit v0.10.2 From 581ade4d1c025eb10421eda0d0c0a2f04447d7c5 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 5 Apr 2011 15:02:37 -0400 Subject: cifs: clean up various nits in unicode routines (try #2) Minor revision to the original patch. Don't abuse the __le16 variable on the stack by casting it to wchar_t and handing it off to char2uni. Declare an actual wchar_t on the stack instead. This fixes a valid sparse warning. Fix the spelling of UNI_ASTERISK. Eliminate the unneeded len_remaining variable in cifsConvertToUCS. Also, as David Howells points out. We were better off making cifsConvertToUCS *not* use put_unaligned_le16 since it means that we can't optimize the mapped characters at compile time. Switch them instead to use cpu_to_le16, and simply use put_unaligned to set them in the string. Reported-and-acked-by: David Howells Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/cifs_unicode.c b/fs/cifs/cifs_unicode.c index fc0fd4f..23d43cd 100644 --- a/fs/cifs/cifs_unicode.c +++ b/fs/cifs/cifs_unicode.c @@ -90,7 +90,7 @@ cifs_mapchar(char *target, const __u16 src_char, const struct nls_table *cp, case UNI_COLON: *target = ':'; break; - case UNI_ASTERIK: + case UNI_ASTERISK: *target = '*'; break; case UNI_QUESTION: @@ -264,40 +264,40 @@ cifs_strndup_from_ucs(const char *src, const int maxlen, const bool is_unicode, * names are little endian 16 bit Unicode on the wire */ int -cifsConvertToUCS(__le16 *target, const char *source, int maxlen, +cifsConvertToUCS(__le16 *target, const char *source, int srclen, const struct nls_table *cp, int mapChars) { int i, j, charlen; - int len_remaining = maxlen; char src_char; - __u16 temp; + __le16 dst_char; + wchar_t tmp; if (!mapChars) return cifs_strtoUCS(target, source, PATH_MAX, cp); - for (i = 0, j = 0; i < maxlen; j++) { + for (i = 0, j = 0; i < srclen; j++) { src_char = source[i]; switch (src_char) { case 0: - put_unaligned_le16(0, &target[j]); + put_unaligned(0, &target[j]); goto ctoUCS_out; case ':': - temp = UNI_COLON; + dst_char = cpu_to_le16(UNI_COLON); break; case '*': - temp = UNI_ASTERIK; + dst_char = cpu_to_le16(UNI_ASTERISK); break; case '?': - temp = UNI_QUESTION; + dst_char = cpu_to_le16(UNI_QUESTION); break; case '<': - temp = UNI_LESSTHAN; + dst_char = cpu_to_le16(UNI_LESSTHAN); break; case '>': - temp = UNI_GRTRTHAN; + dst_char = cpu_to_le16(UNI_GRTRTHAN); break; case '|': - temp = UNI_PIPE; + dst_char = cpu_to_le16(UNI_PIPE); break; /* * FIXME: We can not handle remapping backslash (UNI_SLASH) @@ -305,17 +305,17 @@ cifsConvertToUCS(__le16 *target, const char *source, int maxlen, * as they use backslash as separator. */ default: - charlen = cp->char2uni(source+i, len_remaining, - &temp); + charlen = cp->char2uni(source + i, srclen - i, &tmp); + dst_char = cpu_to_le16(tmp); + /* * if no match, use question mark, which at least in * some cases serves as wild card */ if (charlen < 1) { - temp = 0x003f; + dst_char = cpu_to_le16(0x003f); charlen = 1; } - len_remaining -= charlen; /* * character may take more than one byte in the source * string, but will take exactly two bytes in the @@ -324,9 +324,8 @@ cifsConvertToUCS(__le16 *target, const char *source, int maxlen, i += charlen; continue; } - put_unaligned_le16(temp, &target[j]); + put_unaligned(dst_char, &target[j]); i++; /* move to next char in source string */ - len_remaining--; } ctoUCS_out: diff --git a/fs/cifs/cifs_unicode.h b/fs/cifs/cifs_unicode.h index 7fe6b52..644dd882 100644 --- a/fs/cifs/cifs_unicode.h +++ b/fs/cifs/cifs_unicode.h @@ -44,7 +44,7 @@ * reserved symbols (along with \ and /), otherwise illegal to store * in filenames in NTFS */ -#define UNI_ASTERIK (__u16) ('*' + 0xF000) +#define UNI_ASTERISK (__u16) ('*' + 0xF000) #define UNI_QUESTION (__u16) ('?' + 0xF000) #define UNI_COLON (__u16) (':' + 0xF000) #define UNI_GRTRTHAN (__u16) ('>' + 0xF000) -- cgit v0.10.2 From 157c249114508aa71daa308a426e15d81a4eed00 Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Sat, 2 Apr 2011 07:34:30 -0400 Subject: cifs: wrap received signature check in srv_mutex While testing my patchset to fix asynchronous writes, I hit a bunch of signature problems when testing with signing on. The problem seems to be that signature checks on receive can be running at the same time as a process that is sending, or even that multiple receives can be checking signatures at the same time, clobbering the same data structures. While we're at it, clean up the comments over cifs_calculate_signature and add a note that the srv_mutex should be held when calling this function. This patch seems to fix the problems for me, but I'm not clear on whether it's the best approach. If it is, then this should probably go to stable too. Cc: stable@kernel.org Cc: Shirish Pargaonkar Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index e307a28..d1a016b 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -30,12 +30,13 @@ #include #include -/* Calculate and return the CIFS signature based on the mac key and SMB PDU */ -/* the 16 byte signature must be allocated by the caller */ -/* Note we only use the 1st eight bytes */ -/* Note that the smb header signature field on input contains the - sequence number before this function is called */ - +/* + * Calculate and return the CIFS signature based on the mac key and SMB PDU. + * The 16 byte signature must be allocated by the caller. Note we only use the + * 1st eight bytes and that the smb header signature field on input contains + * the sequence number before this function is called. Also, this function + * should be called with the server->srv_mutex held. + */ static int cifs_calculate_signature(const struct smb_hdr *cifs_pdu, struct TCP_Server_Info *server, char *signature) { @@ -209,8 +210,10 @@ int cifs_verify_signature(struct smb_hdr *cifs_pdu, cpu_to_le32(expected_sequence_number); cifs_pdu->Signature.Sequence.Reserved = 0; + mutex_lock(&server->srv_mutex); rc = cifs_calculate_signature(cifs_pdu, server, what_we_think_sig_should_be); + mutex_unlock(&server->srv_mutex); if (rc) return rc; -- cgit v0.10.2 From fd88ce9313e9f9d3b56eada7fc76a301828baefd Mon Sep 17 00:00:00 2001 From: Steve French Date: Tue, 12 Apr 2011 01:01:14 +0000 Subject: [CIFS] cifs: clarify the meaning of tcpStatus == CifsGood When the TCP_Server_Info is first allocated and connected, tcpStatus == CifsGood means that the NEGOTIATE_PROTOCOL request has completed and the socket is ready for other calls. cifs_reconnect however sets tcpStatus to CifsGood as soon as the socket is reconnected and the optional RFC1001 session setup is done. We have no clear way to tell the difference between these two states, and we need to know this in order to know whether we can send an echo or not. Resolve this by adding a new statusEnum value -- CifsNeedNegotiate. When the socket has been connected but has not yet had a NEGOTIATE_PROTOCOL request done, set it to this value. Once the NEGOTIATE is done, cifs_negotiate_protocol will set tcpStatus to CifsGood. This also fixes and cleans the logic in cifs_reconnect and cifs_reconnect_tcon. The old code checked for specific states when what it really wants to know is whether the state has actually changed from CifsNeedReconnect. Reported-and-Tested-by: JG Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 1cff9aa..ddb3599 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -91,7 +91,8 @@ enum statusEnum { CifsNew = 0, CifsGood, CifsExiting, - CifsNeedReconnect + CifsNeedReconnect, + CifsNeedNegotiate }; enum securityEnum { diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 3919ec8..df959ba 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -142,9 +142,9 @@ cifs_reconnect_tcon(struct cifsTconInfo *tcon, int smb_command) */ while (server->tcpStatus == CifsNeedReconnect) { wait_event_interruptible_timeout(server->response_q, - (server->tcpStatus == CifsGood), 10 * HZ); + (server->tcpStatus != CifsNeedReconnect), 10 * HZ); - /* is TCP session is reestablished now ?*/ + /* are we still trying to reconnect? */ if (server->tcpStatus != CifsNeedReconnect) break; diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 709fd9d..8cf4a63 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -199,8 +199,7 @@ cifs_reconnect(struct TCP_Server_Info *server) } spin_unlock(&GlobalMid_Lock); - while ((server->tcpStatus != CifsExiting) && - (server->tcpStatus != CifsGood)) { + while (server->tcpStatus == CifsNeedReconnect) { try_to_freeze(); /* we should try only the port we connected to before */ @@ -212,7 +211,7 @@ cifs_reconnect(struct TCP_Server_Info *server) atomic_inc(&tcpSesReconnectCount); spin_lock(&GlobalMid_Lock); if (server->tcpStatus != CifsExiting) - server->tcpStatus = CifsGood; + server->tcpStatus = CifsNeedNegotiate; spin_unlock(&GlobalMid_Lock); } } @@ -421,7 +420,7 @@ cifs_demultiplex_thread(struct TCP_Server_Info *server) pdu_length = 4; /* enough to get RFC1001 header */ incomplete_rcv: - if (echo_retries > 0 && + if (echo_retries > 0 && server->tcpStatus == CifsGood && time_after(jiffies, server->lstrp + (echo_retries * SMB_ECHO_INTERVAL))) { cERROR(1, "Server %s has not responded in %d seconds. " @@ -1766,6 +1765,7 @@ cifs_get_tcp_session(struct smb_vol *volume_info) module_put(THIS_MODULE); goto out_err_crypto_release; } + tcp_ses->tcpStatus = CifsNeedNegotiate; /* thread spawned, put it on the list */ spin_lock(&cifs_tcp_ses_lock); -- cgit v0.10.2 From d9b942013730c38ac83564d6669c6d0ecf6d754d Mon Sep 17 00:00:00 2001 From: Steve French Date: Tue, 12 Apr 2011 01:24:57 +0000 Subject: [CIFS] Warn on requesting default security (ntlm) on mount Warn once if default security (ntlm) requested. We will update the default to the stronger security mechanism (ntlmv2) in 2.6.41. Kerberos is also stronger than ntlm, but more servers support ntlmv2 and ntlmv2 does not require an upcall, so ntlmv2 is a better default. Reviewed-by: Jeff Layton CC: Suresh Jayaraman Reviewed-by: Shirish Pargaonkar Signed-off-by: Steve French diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 8cf4a63..db9d55b 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1854,6 +1854,8 @@ cifs_put_smb_ses(struct cifsSesInfo *ses) cifs_put_tcp_session(server); } +static bool warned_on_ntlm; /* globals init to false automatically */ + static struct cifsSesInfo * cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) { @@ -1928,6 +1930,15 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) } ses->cred_uid = volume_info->cred_uid; ses->linux_uid = volume_info->linux_uid; + + /* ntlmv2 is much stronger than ntlm security, and has been broadly + supported for many years, time to update default security mechanism */ + if ((volume_info->secFlg == 0) && warned_on_ntlm == false) { + warned_on_ntlm = true; + cERROR(1, "default security mechanism requested. The default " + "security mechanism will be upgraded from ntlm to " + "ntlmv2 in kernel release 2.6.41"); + } ses->overrideSecFlg = volume_info->secFlg; mutex_lock(&ses->session_mutex); -- cgit v0.10.2 From aa8673599f1d269b4e4d9b0c0f61fca57bc02699 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 11 Apr 2011 18:59:05 -0700 Subject: llc: Fix length check in llc_fixup_skb(). Fixes bugzilla #32872 The LLC stack pretends to support non-linear skbs but there is a direct use of skb_tail_pointer() in llc_fixup_skb(). Use pskb_may_pull() to see if data_size bytes remain and can be accessed linearly in the packet, instead of direct pointer checks. Signed-off-by: David S. Miller diff --git a/net/llc/llc_input.c b/net/llc/llc_input.c index 058f1e9..9032421 100644 --- a/net/llc/llc_input.c +++ b/net/llc/llc_input.c @@ -121,8 +121,7 @@ static inline int llc_fixup_skb(struct sk_buff *skb) s32 data_size = ntohs(pdulen) - llc_len; if (data_size < 0 || - ((skb_tail_pointer(skb) - - (u8 *)pdu) - llc_len) < data_size) + !pskb_may_pull(skb, data_size)) return 0; if (unlikely(pskb_trim_rcsum(skb, data_size))) return 0; -- cgit v0.10.2 From 9d90e49da57fe73a2f35334fdd2fb60dbf3933ed Mon Sep 17 00:00:00 2001 From: Jacob Pan Date: Fri, 8 Apr 2011 11:23:00 -0700 Subject: x86/mrst: Fix boot crash caused by incorrect pin to irq mapping Moorestown systems crash on boot because the secondary CPU clockevent (apbt1) will fail to request irq#1, which does not have ioapic chip in its irq_desc[] entry. Background: Moorestown platform does not have ISA bus nor legacy IRQs. It reuses the range of legacy IRQs for regular device interrupts. The routing information of early system device IRQs (timers) are obtained from firmware provided SFI tables. We reuse/fake MP configuration table to facilitate IRQ setup with IOAPIC. Maintaining a 1:1 mapping of IOAPIC pin (RTE entry) and IRQ# makes routing information clean and easy to understand on Moorestown. Though optional. This patch allows SFI timer and vRTC IRQ to be treated as ISA IRQ so that pin2irq mapping will be 1:1. Also fixed MP table type and use macros to clearly set MP IRQ entries. As a result, apbt timer and RTC interrupts on Moorestown are within legacy IRQ range: # cat /proc/interrupts CPU0 CPU1 0: 11249 0 IO-APIC-edge apbt0 1: 0 12271 IO-APIC-edge apbt1 8: 887 0 IO-APIC-fasteoi dw_spi 13: 0 0 IO-APIC-fasteoi INTEL_MID_DMAC2 14: 0 0 IO-APIC-fasteoi rtc0 Further discussion of this patch can be found at: https://lkml.org/lkml/2010/6/10/70 Suggested-by: "Eric W. Biederman" Signed-off-by: Jacob Pan Cc: Feng Tang Cc: Alan Cox Cc: Arjan van de Ven Link: http://lkml.kernel.org/r/1302286980-21139-1-git-send-email-jacob.jun.pan@linux.intel.com Signed-off-by: Ingo Molnar diff --git a/arch/x86/platform/mrst/mrst.c b/arch/x86/platform/mrst/mrst.c index 5c0207b..275dbc1 100644 --- a/arch/x86/platform/mrst/mrst.c +++ b/arch/x86/platform/mrst/mrst.c @@ -97,11 +97,11 @@ static int __init sfi_parse_mtmr(struct sfi_table_header *table) pentry->freq_hz, pentry->irq); if (!pentry->irq) continue; - mp_irq.type = MP_IOAPIC; + mp_irq.type = MP_INTSRC; mp_irq.irqtype = mp_INT; /* triggering mode edge bit 2-3, active high polarity bit 0-1 */ mp_irq.irqflag = 5; - mp_irq.srcbus = 0; + mp_irq.srcbus = MP_BUS_ISA; mp_irq.srcbusirq = pentry->irq; /* IRQ */ mp_irq.dstapic = MP_APIC_ALL; mp_irq.dstirq = pentry->irq; @@ -168,10 +168,10 @@ int __init sfi_parse_mrtc(struct sfi_table_header *table) for (totallen = 0; totallen < sfi_mrtc_num; totallen++, pentry++) { pr_debug("RTC[%d]: paddr = 0x%08x, irq = %d\n", totallen, (u32)pentry->phys_addr, pentry->irq); - mp_irq.type = MP_IOAPIC; + mp_irq.type = MP_INTSRC; mp_irq.irqtype = mp_INT; mp_irq.irqflag = 0xf; /* level trigger and active low */ - mp_irq.srcbus = 0; + mp_irq.srcbus = MP_BUS_ISA; mp_irq.srcbusirq = pentry->irq; /* IRQ */ mp_irq.dstapic = MP_APIC_ALL; mp_irq.dstirq = pentry->irq; @@ -282,7 +282,7 @@ void __init x86_mrst_early_setup(void) /* Avoid searching for BIOS MP tables */ x86_init.mpparse.find_smp_config = x86_init_noop; x86_init.mpparse.get_smp_config = x86_init_uint_noop; - + set_bit(MP_BUS_ISA, mp_bus_not_pci); } /* -- cgit v0.10.2 From c68b2081250a87ba9b9a173faa5a600cc684602e Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Mon, 11 Apr 2011 23:50:02 -0700 Subject: Input: twl4030_keypad - fix potential NULL dereference in twl4030_kp_probe() We should first check whether platform data is NULL or not, before dereferencing it to get the keymap. Signed-off-by: Axel Lin Reviewed-by: Felipe Balbi Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/keyboard/twl4030_keypad.c b/drivers/input/keyboard/twl4030_keypad.c index cc06c4b..a26922c 100644 --- a/drivers/input/keyboard/twl4030_keypad.c +++ b/drivers/input/keyboard/twl4030_keypad.c @@ -332,7 +332,7 @@ static int __devinit twl4030_kp_program(struct twl4030_keypad *kp) static int __devinit twl4030_kp_probe(struct platform_device *pdev) { struct twl4030_keypad_data *pdata = pdev->dev.platform_data; - const struct matrix_keymap_data *keymap_data = pdata->keymap_data; + const struct matrix_keymap_data *keymap_data; struct twl4030_keypad *kp; struct input_dev *input; u8 reg; @@ -344,6 +344,8 @@ static int __devinit twl4030_kp_probe(struct platform_device *pdev) return -EINVAL; } + keymap_data = pdata->keymap_data; + kp = kzalloc(sizeof(*kp), GFP_KERNEL); input = input_allocate_device(); if (!kp || !input) { -- cgit v0.10.2 From d9c97833179036408e53ef5f3f5c7eaf781769bc Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 12 Apr 2011 10:06:33 +0200 Subject: block: remove block_unplug_timer() trace point We no longer have an unplug timer running, so no point in keeping the trace point. Signed-off-by: Jens Axboe diff --git a/include/trace/events/block.h b/include/trace/events/block.h index 78f18ad..43a9853 100644 --- a/include/trace/events/block.h +++ b/include/trace/events/block.h @@ -419,20 +419,6 @@ DECLARE_EVENT_CLASS(block_unplug, ); /** - * block_unplug_timer - timed release of operations requests in queue to device driver - * @q: request queue to unplug - * - * Unplug the request queue @q because a timer expired and allow block - * operation requests to be sent to the device driver. - */ -DEFINE_EVENT(block_unplug, block_unplug_timer, - - TP_PROTO(struct request_queue *q), - - TP_ARGS(q) -); - -/** * block_unplug_io - release of operations requests in request queue * @q: request queue to unplug * diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 7aa40f8..824708c 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -863,19 +863,6 @@ static void blk_add_trace_unplug_io(void *ignore, struct request_queue *q) } } -static void blk_add_trace_unplug_timer(void *ignore, struct request_queue *q) -{ - struct blk_trace *bt = q->blk_trace; - - if (bt) { - unsigned int pdu = q->rq.count[READ] + q->rq.count[WRITE]; - __be64 rpdu = cpu_to_be64(pdu); - - __blk_add_trace(bt, 0, 0, 0, BLK_TA_UNPLUG_TIMER, 0, - sizeof(rpdu), &rpdu); - } -} - static void blk_add_trace_split(void *ignore, struct request_queue *q, struct bio *bio, unsigned int pdu) @@ -1015,8 +1002,6 @@ static void blk_register_tracepoints(void) WARN_ON(ret); ret = register_trace_block_plug(blk_add_trace_plug, NULL); WARN_ON(ret); - ret = register_trace_block_unplug_timer(blk_add_trace_unplug_timer, NULL); - WARN_ON(ret); ret = register_trace_block_unplug_io(blk_add_trace_unplug_io, NULL); WARN_ON(ret); ret = register_trace_block_split(blk_add_trace_split, NULL); @@ -1033,7 +1018,6 @@ static void blk_unregister_tracepoints(void) unregister_trace_block_bio_remap(blk_add_trace_bio_remap, NULL); unregister_trace_block_split(blk_add_trace_split, NULL); unregister_trace_block_unplug_io(blk_add_trace_unplug_io, NULL); - unregister_trace_block_unplug_timer(blk_add_trace_unplug_timer, NULL); unregister_trace_block_plug(blk_add_trace_plug, NULL); unregister_trace_block_sleeprq(blk_add_trace_sleeprq, NULL); unregister_trace_block_getrq(blk_add_trace_getrq, NULL); @@ -1348,7 +1332,6 @@ static const struct { [__BLK_TA_COMPLETE] = {{ "C", "complete" }, blk_log_with_error }, [__BLK_TA_PLUG] = {{ "P", "plug" }, blk_log_plug }, [__BLK_TA_UNPLUG_IO] = {{ "U", "unplug_io" }, blk_log_unplug }, - [__BLK_TA_UNPLUG_TIMER] = {{ "UT", "unplug_timer" }, blk_log_unplug }, [__BLK_TA_INSERT] = {{ "I", "insert" }, blk_log_generic }, [__BLK_TA_SPLIT] = {{ "X", "split" }, blk_log_split }, [__BLK_TA_BOUNCE] = {{ "B", "bounce" }, blk_log_generic }, -- cgit v0.10.2 From 94b5eb28b41cc79d9713696e0005ae167b5afd1b Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 12 Apr 2011 10:12:19 +0200 Subject: block: fixup block IO unplug trace call It was removed with the on-stack plugging, readd it and track the depth of requests added when flushing the plug. Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index eeaca09..d20ce1e 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2668,12 +2668,19 @@ static int plug_rq_cmp(void *priv, struct list_head *a, struct list_head *b) return !(rqa->q <= rqb->q); } +static void queue_unplugged(struct request_queue *q, unsigned int depth) +{ + trace_block_unplug_io(q, depth); + __blk_run_queue(q, false); +} + static void flush_plug_list(struct blk_plug *plug) { struct request_queue *q; unsigned long flags; struct request *rq; LIST_HEAD(list); + unsigned int depth; BUG_ON(plug->magic != PLUG_MAGIC); @@ -2688,6 +2695,7 @@ static void flush_plug_list(struct blk_plug *plug) } q = NULL; + depth = 0; local_irq_save(flags); while (!list_empty(&list)) { rq = list_entry_rq(list.next); @@ -2696,10 +2704,11 @@ static void flush_plug_list(struct blk_plug *plug) BUG_ON(!rq->q); if (rq->q != q) { if (q) { - __blk_run_queue(q, false); + queue_unplugged(q, depth); spin_unlock(q->queue_lock); } q = rq->q; + depth = 0; spin_lock(q->queue_lock); } rq->cmd_flags &= ~REQ_ON_PLUG; @@ -2711,10 +2720,12 @@ static void flush_plug_list(struct blk_plug *plug) __elv_add_request(q, rq, ELEVATOR_INSERT_FLUSH); else __elv_add_request(q, rq, ELEVATOR_INSERT_SORT_MERGE); + + depth++; } if (q) { - __blk_run_queue(q, false); + queue_unplugged(q, depth); spin_unlock(q->queue_lock); } diff --git a/include/trace/events/block.h b/include/trace/events/block.h index 43a9853..006e60b 100644 --- a/include/trace/events/block.h +++ b/include/trace/events/block.h @@ -401,9 +401,9 @@ TRACE_EVENT(block_plug, DECLARE_EVENT_CLASS(block_unplug, - TP_PROTO(struct request_queue *q), + TP_PROTO(struct request_queue *q, unsigned int depth), - TP_ARGS(q), + TP_ARGS(q, depth), TP_STRUCT__entry( __field( int, nr_rq ) @@ -411,7 +411,7 @@ DECLARE_EVENT_CLASS(block_unplug, ), TP_fast_assign( - __entry->nr_rq = q->rq.count[READ] + q->rq.count[WRITE]; + __entry->nr_rq = depth; memcpy(__entry->comm, current->comm, TASK_COMM_LEN); ), @@ -421,15 +421,16 @@ DECLARE_EVENT_CLASS(block_unplug, /** * block_unplug_io - release of operations requests in request queue * @q: request queue to unplug + * @depth: number of requests just added to the queue * * Unplug request queue @q because device driver is scheduled to work * on elements in the request queue. */ DEFINE_EVENT(block_unplug, block_unplug_io, - TP_PROTO(struct request_queue *q), + TP_PROTO(struct request_queue *q, unsigned int depth), - TP_ARGS(q) + TP_ARGS(q, depth) ); /** diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 824708c..3e3970d 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -850,13 +850,13 @@ static void blk_add_trace_plug(void *ignore, struct request_queue *q) __blk_add_trace(bt, 0, 0, 0, BLK_TA_PLUG, 0, 0, NULL); } -static void blk_add_trace_unplug_io(void *ignore, struct request_queue *q) +static void blk_add_trace_unplug_io(void *ignore, struct request_queue *q, + unsigned int depth) { struct blk_trace *bt = q->blk_trace; if (bt) { - unsigned int pdu = q->rq.count[READ] + q->rq.count[WRITE]; - __be64 rpdu = cpu_to_be64(pdu); + __be64 rpdu = cpu_to_be64(depth); __blk_add_trace(bt, 0, 0, 0, BLK_TA_UNPLUG_IO, 0, sizeof(rpdu), &rpdu); -- cgit v0.10.2 From 188112722cce083c8f1a7d0d84f55c2cd885920c Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 12 Apr 2011 10:11:24 +0200 Subject: block: add comment on why we save and disable interrupts in flush_plug_list() It's done at the top to avoid doing it for every queue we unplug. Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index d20ce1e..0c0ea10 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2696,6 +2696,11 @@ static void flush_plug_list(struct blk_plug *plug) q = NULL; depth = 0; + + /* + * Save and disable interrupts here, to avoid doing it for every + * queue lock we have to take. + */ local_irq_save(flags); while (!list_empty(&list)) { rq = list_entry_rq(list.next); -- cgit v0.10.2 From f75664570d8b75469cc468f23c2b27220984983b Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 12 Apr 2011 10:17:31 +0200 Subject: block: add callback function for unplug notification MD would like to know when a queue is unplugged, so it can flush it's bitmap writes. Add such a callback. Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 0c0ea10..76850fc 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2672,6 +2672,9 @@ static void queue_unplugged(struct request_queue *q, unsigned int depth) { trace_block_unplug_io(q, depth); __blk_run_queue(q, false); + + if (q->unplugged_fn) + q->unplugged_fn(q); } static void flush_plug_list(struct blk_plug *plug) diff --git a/block/blk-settings.c b/block/blk-settings.c index 1fa7692..eb94904 100644 --- a/block/blk-settings.c +++ b/block/blk-settings.c @@ -790,6 +790,22 @@ void blk_queue_flush(struct request_queue *q, unsigned int flush) } EXPORT_SYMBOL_GPL(blk_queue_flush); +/** + * blk_queue_unplugged - register a callback for an unplug event + * @q: the request queue for the device + * @fn: the function to call + * + * Some stacked drivers may need to know when IO is dispatched on an + * unplug event. By registrering a callback here, they will be notified + * when someone flushes their on-stack queue plug. The function will be + * called with the queue lock held. + */ +void blk_queue_unplugged(struct request_queue *q, unplugged_fn *fn) +{ + q->unplugged_fn = fn; +} +EXPORT_SYMBOL(blk_queue_unplugged); + static int __init blk_settings_init(void) { blk_max_low_pfn = max_low_pfn - 1; diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 32176cc..c07ffaf 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -196,6 +196,7 @@ typedef void (request_fn_proc) (struct request_queue *q); typedef int (make_request_fn) (struct request_queue *q, struct bio *bio); typedef int (prep_rq_fn) (struct request_queue *, struct request *); typedef void (unprep_rq_fn) (struct request_queue *, struct request *); +typedef void (unplugged_fn) (struct request_queue *); struct bio_vec; struct bvec_merge_data { @@ -283,6 +284,7 @@ struct request_queue rq_timed_out_fn *rq_timed_out_fn; dma_drain_needed_fn *dma_drain_needed; lld_busy_fn *lld_busy_fn; + unplugged_fn *unplugged_fn; /* * Dispatch queue sorting @@ -841,6 +843,7 @@ extern void blk_queue_dma_alignment(struct request_queue *, int); extern void blk_queue_update_dma_alignment(struct request_queue *, int); extern void blk_queue_softirq_done(struct request_queue *, softirq_done_fn *); extern void blk_queue_rq_timed_out(struct request_queue *, rq_timed_out_fn *); +extern void blk_queue_unplugged(struct request_queue *, unplugged_fn *); extern void blk_queue_rq_timeout(struct request_queue *, unsigned int); extern void blk_queue_flush(struct request_queue *q, unsigned int flush); extern struct backing_dev_info *blk_get_backing_dev_info(struct block_device *bdev); -- cgit v0.10.2 From dc6d36c9710d1fed42d1bbe7d8e4f742abd844c6 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 12 Apr 2011 10:28:28 +0200 Subject: block: readd plug trace event This was removed with the queue plug state. But we can easily readd by checking if this is the first request going to this queue. It's good information to have when tracing to see how effective the plugging is. Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 76850fc..52e756c 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -1311,7 +1311,15 @@ get_rq: plug = current->plug; if (plug) { - if (!plug->should_sort && !list_empty(&plug->list)) { + /* + * If this is the first request added after a plug, fire + * of a plug trace. If others have been added before, check + * if we have multiple devices in this plug. If so, make a + * note to sort the list before dispatch. + */ + if (list_empty(&plug->list)) + trace_block_plug(q); + else if (!plug->should_sort) { struct request *__rq; __rq = list_entry_rq(plug->list.prev); -- cgit v0.10.2 From cf82c798394cd443eed7d91f998b79a63f341e91 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 12 Apr 2011 10:30:53 +0200 Subject: block: kill queue_sync_plugs() The original use for this dates back to when we had to track write requests for serializing around barriers. That's not needed anymore, so kill it. Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 52e756c..c6eaa1f 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -198,19 +198,6 @@ void blk_dump_rq_flags(struct request *rq, char *msg) } EXPORT_SYMBOL(blk_dump_rq_flags); -/* - * Make sure that plugs that were pending when this function was entered, - * are now complete and requests pushed to the queue. -*/ -static inline void queue_sync_plugs(struct request_queue *q) -{ - /* - * If the current process is plugged and has barriers submitted, - * we will livelock if we don't unplug first. - */ - blk_flush_plug(current); -} - static void blk_delay_work(struct work_struct *work) { struct request_queue *q; @@ -298,7 +285,6 @@ void blk_sync_queue(struct request_queue *q) { del_timer_sync(&q->timeout); cancel_delayed_work_sync(&q->delay_work); - queue_sync_plugs(q); } EXPORT_SYMBOL(blk_sync_queue); -- cgit v0.10.2 From e710d7d5a9cab1041b7a3cf9e655b75d92786857 Mon Sep 17 00:00:00 2001 From: Samuel Ortiz Date: Fri, 8 Apr 2011 00:43:01 +0200 Subject: mfd: Fetch cell pointer from platform_device->mfd_cell In order for MFD drivers to fetch their cell pointer but also their platform data one, an mfd cell pointer is added to the platform_device structure. That allows all MFD sub devices drivers to be MFD agnostic, unless they really need to access their MFD cell data. Most of them don't, especially the ones for IPs used by both MFD and non MFD SoCs. Cc: Grant Likely Acked-by: Greg KH Signed-off-by: Samuel Ortiz diff --git a/drivers/base/platform.c b/drivers/base/platform.c index f051cff..6c3a2bd 100644 --- a/drivers/base/platform.c +++ b/drivers/base/platform.c @@ -149,6 +149,7 @@ static void platform_device_release(struct device *dev) of_device_node_put(&pa->pdev.dev); kfree(pa->pdev.dev.platform_data); + kfree(pa->pdev.mfd_cell); kfree(pa->pdev.resource); kfree(pa); } diff --git a/drivers/mfd/mfd-core.c b/drivers/mfd/mfd-core.c index d01574d..f4c8c84 100644 --- a/drivers/mfd/mfd-core.c +++ b/drivers/mfd/mfd-core.c @@ -55,6 +55,19 @@ int mfd_cell_disable(struct platform_device *pdev) } EXPORT_SYMBOL(mfd_cell_disable); +static int mfd_platform_add_cell(struct platform_device *pdev, + const struct mfd_cell *cell) +{ + if (!cell) + return 0; + + pdev->mfd_cell = kmemdup(cell, sizeof(*cell), GFP_KERNEL); + if (!pdev->mfd_cell) + return -ENOMEM; + + return 0; +} + static int mfd_add_device(struct device *parent, int id, const struct mfd_cell *cell, struct resource *mem_base, @@ -75,7 +88,7 @@ static int mfd_add_device(struct device *parent, int id, pdev->dev.parent = parent; - ret = platform_device_add_data(pdev, cell, sizeof(*cell)); + ret = mfd_platform_add_cell(pdev, cell); if (ret) goto fail_res; @@ -123,7 +136,6 @@ static int mfd_add_device(struct device *parent, int id, return 0; -/* platform_device_del(pdev); */ fail_res: kfree(res); fail_device: diff --git a/include/linux/mfd/core.h b/include/linux/mfd/core.h index ad1b19a..aef23309 100644 --- a/include/linux/mfd/core.h +++ b/include/linux/mfd/core.h @@ -86,16 +86,25 @@ extern int mfd_clone_cell(const char *cell, const char **clones, */ static inline const struct mfd_cell *mfd_get_cell(struct platform_device *pdev) { - return pdev->dev.platform_data; + return pdev->mfd_cell; } /* * Given a platform device that's been created by mfd_add_devices(), fetch * the .mfd_data entry from the mfd_cell that created it. + * Otherwise just return the platform_data pointer. + * This maintains compatibility with platform drivers whose devices aren't + * created by the mfd layer, and expect platform_data to contain what would've + * otherwise been in mfd_data. */ static inline void *mfd_get_data(struct platform_device *pdev) { - return mfd_get_cell(pdev)->mfd_data; + const struct mfd_cell *cell = mfd_get_cell(pdev); + + if (cell) + return cell->mfd_data; + else + return pdev->dev.platform_data; } extern int mfd_add_devices(struct device *parent, int id, diff --git a/include/linux/platform_device.h b/include/linux/platform_device.h index d96db98..744942c 100644 --- a/include/linux/platform_device.h +++ b/include/linux/platform_device.h @@ -14,6 +14,8 @@ #include #include +struct mfd_cell; + struct platform_device { const char * name; int id; @@ -23,6 +25,9 @@ struct platform_device { const struct platform_device_id *id_entry; + /* MFD cell pointer */ + struct mfd_cell *mfd_cell; + /* arch specific additions */ struct pdev_archdata archdata; }; -- cgit v0.10.2 From ef1fd2df85f5ea8ca8876a45d0ebb152d3a144d1 Mon Sep 17 00:00:00 2001 From: Prabhakar Kushwaha Date: Thu, 31 Mar 2011 12:31:09 +0530 Subject: powerpc/85xx: Don't add disabled PCIe devices PCIe nodes with the property status="disabled" are not usable and so avoid adding "disabled" PCIe bridge with the system. Signed-off-by: Prabhakar Kushwaha Signed-off-by: Kumar Gala diff --git a/arch/powerpc/sysdev/fsl_pci.c b/arch/powerpc/sysdev/fsl_pci.c index f8f7f28..68ca929 100644 --- a/arch/powerpc/sysdev/fsl_pci.c +++ b/arch/powerpc/sysdev/fsl_pci.c @@ -324,6 +324,11 @@ int __init fsl_add_bridge(struct device_node *dev, int is_primary) struct resource rsrc; const int *bus_range; + if (!of_device_is_available(dev)) { + pr_warning("%s: disabled\n", dev->full_name); + return -ENODEV; + } + pr_debug("Adding PCI host bridge %s\n", dev->full_name); /* Fetch host bridge registers address */ -- cgit v0.10.2 From 07d9fce24d871785dbd25458469032fea73f17b8 Mon Sep 17 00:00:00 2001 From: Prabhakar Kushwaha Date: Wed, 6 Apr 2011 12:56:29 +0530 Subject: powerpc: Check device status before adding serial device serial port nodes with the property status="disabled" are not usable and so avoid adding "disabled" port with the system. Signed-off-by: Prabhakar Kushwaha Signed-off-by: Kumar Gala diff --git a/arch/powerpc/kernel/legacy_serial.c b/arch/powerpc/kernel/legacy_serial.c index c834757..2b97b80 100644 --- a/arch/powerpc/kernel/legacy_serial.c +++ b/arch/powerpc/kernel/legacy_serial.c @@ -330,9 +330,11 @@ void __init find_legacy_serial_ports(void) if (!parent) continue; if (of_match_node(legacy_serial_parents, parent) != NULL) { - index = add_legacy_soc_port(np, np); - if (index >= 0 && np == stdout) - legacy_serial_console = index; + if (of_device_is_available(np)) { + index = add_legacy_soc_port(np, np); + if (index >= 0 && np == stdout) + legacy_serial_console = index; + } } of_node_put(parent); } -- cgit v0.10.2 From 11ed0db9f6c7811233632d2ab79c50c011b89902 Mon Sep 17 00:00:00 2001 From: Kumar Gala Date: Wed, 6 Apr 2011 00:11:06 -0500 Subject: powerpc/book3e: Fix CPU feature handling on 64-bit e5500 The CPU_FTRS_POSSIBLE and CPU_FTRS_ALWAYS defines did not encompass e5500 CPU features when built for 64-bit. This causes issues with cpu_has_feature() as it utilizes the POSSIBLE & ALWAYS defines as part of its check. Create a unique CPU_FTRS_E5500 (as its different from CPU_FTRS_E500MC), created a new group for 64-bit Book3e based CPUs and add CPU_FTRS_E5500 to that group. Signed-off-by: Kumar Gala diff --git a/arch/powerpc/include/asm/cputable.h b/arch/powerpc/include/asm/cputable.h index be3cdf9..f1fbf60 100644 --- a/arch/powerpc/include/asm/cputable.h +++ b/arch/powerpc/include/asm/cputable.h @@ -386,6 +386,9 @@ extern const char *powerpc_base_platform; CPU_FTR_MAYBE_CAN_NAP | CPU_FTR_NODSISRALIGN | \ CPU_FTR_L2CSR | CPU_FTR_LWSYNC | CPU_FTR_NOEXECUTE | \ CPU_FTR_DBELL) +#define CPU_FTRS_E5500 (CPU_FTR_USE_TB | CPU_FTR_NODSISRALIGN | \ + CPU_FTR_L2CSR | CPU_FTR_LWSYNC | CPU_FTR_NOEXECUTE | \ + CPU_FTR_DBELL | CPU_FTR_POPCNTB | CPU_FTR_POPCNTD) #define CPU_FTRS_GENERIC_32 (CPU_FTR_COMMON | CPU_FTR_NODSISRALIGN) /* 64-bit CPUs */ @@ -435,11 +438,15 @@ extern const char *powerpc_base_platform; #define CPU_FTRS_COMPATIBLE (CPU_FTR_USE_TB | CPU_FTR_PPCAS_ARCH_V2) #ifdef __powerpc64__ +#ifdef CONFIG_PPC_BOOK3E +#define CPU_FTRS_POSSIBLE (CPU_FTRS_E5500) +#else #define CPU_FTRS_POSSIBLE \ (CPU_FTRS_POWER3 | CPU_FTRS_RS64 | CPU_FTRS_POWER4 | \ CPU_FTRS_PPC970 | CPU_FTRS_POWER5 | CPU_FTRS_POWER6 | \ CPU_FTRS_POWER7 | CPU_FTRS_CELL | CPU_FTRS_PA6T | \ CPU_FTR_1T_SEGMENT | CPU_FTR_VSX) +#endif #else enum { CPU_FTRS_POSSIBLE = @@ -473,16 +480,21 @@ enum { #endif #ifdef CONFIG_E500 CPU_FTRS_E500 | CPU_FTRS_E500_2 | CPU_FTRS_E500MC | + CPU_FTRS_E5500 | #endif 0, }; #endif /* __powerpc64__ */ #ifdef __powerpc64__ +#ifdef CONFIG_PPC_BOOK3E +#define CPU_FTRS_ALWAYS (CPU_FTRS_E5500) +#else #define CPU_FTRS_ALWAYS \ (CPU_FTRS_POWER3 & CPU_FTRS_RS64 & CPU_FTRS_POWER4 & \ CPU_FTRS_PPC970 & CPU_FTRS_POWER5 & CPU_FTRS_POWER6 & \ CPU_FTRS_POWER7 & CPU_FTRS_CELL & CPU_FTRS_PA6T & CPU_FTRS_POSSIBLE) +#endif #else enum { CPU_FTRS_ALWAYS = @@ -513,6 +525,7 @@ enum { #endif #ifdef CONFIG_E500 CPU_FTRS_E500 & CPU_FTRS_E500_2 & CPU_FTRS_E500MC & + CPU_FTRS_E5500 & #endif CPU_FTRS_POSSIBLE, }; diff --git a/arch/powerpc/kernel/cputable.c b/arch/powerpc/kernel/cputable.c index c9b68d0..b9602ee 100644 --- a/arch/powerpc/kernel/cputable.c +++ b/arch/powerpc/kernel/cputable.c @@ -1973,7 +1973,7 @@ static struct cpu_spec __initdata cpu_specs[] = { .pvr_mask = 0xffff0000, .pvr_value = 0x80240000, .cpu_name = "e5500", - .cpu_features = CPU_FTRS_E500MC, + .cpu_features = CPU_FTRS_E5500, .cpu_user_features = COMMON_USER_BOOKE, .mmu_features = MMU_FTR_TYPE_FSL_E | MMU_FTR_BIG_PHYS | MMU_FTR_USE_TLBILX, -- cgit v0.10.2 From d51ad91535b75c043f074f093ef913fe20ff2b5e Mon Sep 17 00:00:00 2001 From: Scott Wood Date: Thu, 27 May 2010 17:35:12 -0500 Subject: powerpc/e500mc: Remove CPU_FTR_MAYBE_CAN_NAP/CPU_FTR_MAYBE_CAN_DOZE e500mc does not support the HID0/MSR mechanism that is used by e500_idle (and there are also issues with waking on certain types of interrupts). Further, even if napping is never actually enabled, just having CPU_FTR_CAN_NAP will cause machine_init() to overwrite the board's supplied ppc_md.power_save(). We drop CPU_FTR_MAYBE_CAN_DOZE becuase we should use 'wait' instead on e500mc. Signed-off-by: Scott Wood Signed-off-by: Kumar Gala diff --git a/arch/powerpc/include/asm/cputable.h b/arch/powerpc/include/asm/cputable.h index f1fbf60..1833d1a 100644 --- a/arch/powerpc/include/asm/cputable.h +++ b/arch/powerpc/include/asm/cputable.h @@ -382,8 +382,7 @@ extern const char *powerpc_base_platform; #define CPU_FTRS_E500_2 (CPU_FTR_MAYBE_CAN_DOZE | CPU_FTR_USE_TB | \ CPU_FTR_SPE_COMP | CPU_FTR_MAYBE_CAN_NAP | \ CPU_FTR_NODSISRALIGN | CPU_FTR_NOEXECUTE) -#define CPU_FTRS_E500MC (CPU_FTR_MAYBE_CAN_DOZE | CPU_FTR_USE_TB | \ - CPU_FTR_MAYBE_CAN_NAP | CPU_FTR_NODSISRALIGN | \ +#define CPU_FTRS_E500MC (CPU_FTR_USE_TB | CPU_FTR_NODSISRALIGN | \ CPU_FTR_L2CSR | CPU_FTR_LWSYNC | CPU_FTR_NOEXECUTE | \ CPU_FTR_DBELL) #define CPU_FTRS_E5500 (CPU_FTR_USE_TB | CPU_FTR_NODSISRALIGN | \ -- cgit v0.10.2 From e5462d16f76ad7a9156a82a97fbafba298da9ca6 Mon Sep 17 00:00:00 2001 From: Kumar Gala Date: Fri, 8 Apr 2011 04:20:54 -0500 Subject: powerpc/85xx: disable Suspend support if SMP enabled We currently dont have CPU Hotplug support working on 85xx so we need to disable Suspsend support as it will force enabling of CPU Hotplug. arch/powerpc/kernel/built-in.o: In function `cpu_die': arch/powerpc/kernel/smp.c:702: undefined reference to `start_secondary_resume' make: *** [.tmp_vmlinux1] Error 1 Signed-off-by: Kumar Gala diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index b6ff882..8f4d50b 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -209,7 +209,7 @@ config ARCH_HIBERNATION_POSSIBLE config ARCH_SUSPEND_POSSIBLE def_bool y depends on ADB_PMU || PPC_EFIKA || PPC_LITE5200 || PPC_83xx || \ - PPC_85xx || PPC_86xx || PPC_PSERIES || 44x || 40x + (PPC_85xx && !SMP) || PPC_86xx || PPC_PSERIES || 44x || 40x config PPC_DCR_NATIVE bool -- cgit v0.10.2 From f4af3c3d077a004762aaad052049c809fd8c6f0c Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 12 Apr 2011 14:58:51 +0200 Subject: block: move queue run on unplug to kblockd There are worries that we are now consuming a lot more stack in some cases, since we potentially call into IO dispatch from schedule() or io_schedule(). We can reduce this problem by moving the running of the queue to kblockd, like the old plugging scheme did as well. This may or may not be a good idea from a performance perspective, depending on how many tasks have queue plugs running at the same time. For even the slightly contended case, doing just a single queue run from kblockd instead of multiple runs directly from the unpluggers will be faster. Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index c6eaa1f..36b1a75 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2665,7 +2665,7 @@ static int plug_rq_cmp(void *priv, struct list_head *a, struct list_head *b) static void queue_unplugged(struct request_queue *q, unsigned int depth) { trace_block_unplug_io(q, depth); - __blk_run_queue(q, false); + __blk_run_queue(q, true); if (q->unplugged_fn) q->unplugged_fn(q); -- cgit v0.10.2 From ca83ce3d5b9ad321ee24f5870a77f0b21ac5a5de Mon Sep 17 00:00:00 2001 From: Jeff Layton Date: Tue, 12 Apr 2011 09:13:44 -0400 Subject: cifs: don't allow mmap'ed pages to be dirtied while under writeback (try #3) This is more or less the same patch as before, but with some merge conflicts fixed up. If a process has a dirty page mapped into its page tables, then it has the ability to change it while the client is trying to write the data out to the server. If that happens after the signature has been calculated then that signature will then be wrong, and the server will likely reset the TCP connection. This patch adds a page_mkwrite handler for CIFS that simply takes the page lock. Because the page lock is held over the life of writepage and writepages, this prevents the page from becoming writeable until the write call has completed. With this, we can also remove the "sign_zero_copy" module option and always inline the pages when writing. Signed-off-by: Jeff Layton Signed-off-by: Steve French diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index e3352a1..5c412b3 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -81,10 +81,6 @@ module_param(echo_retries, ushort, 0644); MODULE_PARM_DESC(echo_retries, "Number of echo attempts before giving up and " "reconnecting server. Default: 5. 0 means " "never reconnect."); -bool sign_zero_copy; /* globals init to false automatically */ -module_param(sign_zero_copy, bool, 0644); -MODULE_PARM_DESC(sign_zero_copy, "Don't copy pages on write with signing " - "enabled. Default: N"); extern mempool_t *cifs_sm_req_poolp; extern mempool_t *cifs_req_poolp; extern mempool_t *cifs_mid_poolp; diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index ddb3599..a5d1106 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -826,7 +826,6 @@ GLOBAL_EXTERN unsigned int CIFSMaxBufSize; /* max size not including hdr */ GLOBAL_EXTERN unsigned int cifs_min_rcv; /* min size of big ntwrk buf pool */ GLOBAL_EXTERN unsigned int cifs_min_small; /* min size of small buf pool */ GLOBAL_EXTERN unsigned int cifs_max_pending; /* MAX requests at once to server*/ -GLOBAL_EXTERN bool sign_zero_copy; /* don't copy written pages with signing */ /* reconnect after this many failed echo attempts */ GLOBAL_EXTERN unsigned short echo_retries; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index e2d7b6b..faf5952 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -972,6 +972,9 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file, total_written += bytes_written) { rc = -EAGAIN; while (rc == -EAGAIN) { + struct kvec iov[2]; + unsigned int len; + if (open_file->invalidHandle) { /* we could deadlock if we called filemap_fdatawait from here so tell @@ -981,31 +984,14 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file, if (rc != 0) break; } - if (sign_zero_copy || (pTcon->ses->server && - ((pTcon->ses->server->secMode & - (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) - == 0))) { - struct kvec iov[2]; - unsigned int len; - - len = min((size_t)cifs_sb->wsize, - write_size - total_written); - /* iov[0] is reserved for smb header */ - iov[1].iov_base = (char *)write_data + - total_written; - iov[1].iov_len = len; - rc = CIFSSMBWrite2(xid, pTcon, - open_file->netfid, len, - *poffset, &bytes_written, - iov, 1, 0); - } else - rc = CIFSSMBWrite(xid, pTcon, - open_file->netfid, - min_t(const int, cifs_sb->wsize, - write_size - total_written), - *poffset, &bytes_written, - write_data + total_written, - NULL, 0); + + len = min((size_t)cifs_sb->wsize, + write_size - total_written); + /* iov[0] is reserved for smb header */ + iov[1].iov_base = (char *)write_data + total_written; + iov[1].iov_len = len; + rc = CIFSSMBWrite2(xid, pTcon, open_file->netfid, len, + *poffset, &bytes_written, iov, 1, 0); } if (rc || (bytes_written == 0)) { if (total_written) @@ -1242,12 +1228,6 @@ static int cifs_writepages(struct address_space *mapping, } tcon = tlink_tcon(open_file->tlink); - if (!sign_zero_copy && tcon->ses->server->secMode & - (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) { - cifsFileInfo_put(open_file); - kfree(iov); - return generic_writepages(mapping, wbc); - } cifsFileInfo_put(open_file); xid = GetXid(); @@ -1982,6 +1962,24 @@ static ssize_t cifs_read(struct file *file, char *read_data, size_t read_size, return total_read; } +/* + * If the page is mmap'ed into a process' page tables, then we need to make + * sure that it doesn't change while being written back. + */ +static int +cifs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) +{ + struct page *page = vmf->page; + + lock_page(page); + return VM_FAULT_LOCKED; +} + +static struct vm_operations_struct cifs_file_vm_ops = { + .fault = filemap_fault, + .page_mkwrite = cifs_page_mkwrite, +}; + int cifs_file_strict_mmap(struct file *file, struct vm_area_struct *vma) { int rc, xid; @@ -1993,6 +1991,8 @@ int cifs_file_strict_mmap(struct file *file, struct vm_area_struct *vma) cifs_invalidate_mapping(inode); rc = generic_file_mmap(file, vma); + if (rc == 0) + vma->vm_ops = &cifs_file_vm_ops; FreeXid(xid); return rc; } @@ -2009,6 +2009,8 @@ int cifs_file_mmap(struct file *file, struct vm_area_struct *vma) return rc; } rc = generic_file_mmap(file, vma); + if (rc == 0) + vma->vm_ops = &cifs_file_vm_ops; FreeXid(xid); return rc; } -- cgit v0.10.2 From a6d710fefd1b2c209353a452d0f4c831b3af0da0 Mon Sep 17 00:00:00 2001 From: Marek Vasut Date: Tue, 22 Mar 2011 13:09:50 +0100 Subject: ARM: pxafb: Fix access to nonexistent member of pxafb_info In case CONFIG_FB_PXA_OVERLAY is not defined, the pxafb_freq_transition() function tests nonexistent member of pxafb_info (since the member is not part of the structure). Fix this by wraping the test in ifdef, even if I don't really like how the code looks now. The check doesn't have to happen if overlays are disabled at all as the check is always true then. Signed-off-by: Marek Vasut Acked-by: Vasily Khoruzhick Signed-off-by: Eric Miao diff --git a/drivers/video/pxafb.c b/drivers/video/pxafb.c index a2e5b51..0f4e8c9 100644 --- a/drivers/video/pxafb.c +++ b/drivers/video/pxafb.c @@ -1648,7 +1648,9 @@ pxafb_freq_transition(struct notifier_block *nb, unsigned long val, void *data) switch (val) { case CPUFREQ_PRECHANGE: - if (!fbi->overlay[0].usage && !fbi->overlay[1].usage) +#ifdef CONFIG_FB_PXA_OVERLAY + if (!(fbi->overlay[0].usage || fbi->overlay[1].usage)) +#endif set_ctrlr_state(fbi, C_DISABLE_CLKCHANGE); break; -- cgit v0.10.2 From af21cbb1ef6aa366fcb629ea4d4330300ba22de7 Mon Sep 17 00:00:00 2001 From: Dmitry Eremin-Solenikov Date: Fri, 1 Apr 2011 13:28:45 +0400 Subject: pcmcia: limit pxa2xx_balloon3 subdriver to balloon3 platform pxa2xx_balloon3 tries to register pxa2xx-pcmcia device not checking whether machine is really balloon3, thus messing multi-machine kernels. Fix it up. Signed-off-by: Dmitry Eremin-Solenikov Signed-off-by: Eric Miao diff --git a/drivers/pcmcia/pxa2xx_balloon3.c b/drivers/pcmcia/pxa2xx_balloon3.c index 453c54c..4c3e94c 100644 --- a/drivers/pcmcia/pxa2xx_balloon3.c +++ b/drivers/pcmcia/pxa2xx_balloon3.c @@ -25,6 +25,8 @@ #include +#include + #include "soc_common.h" /* @@ -127,6 +129,9 @@ static int __init balloon3_pcmcia_init(void) { int ret; + if (!machine_is_balloon3()) + return -ENODEV; + balloon3_pcmcia_device = platform_device_alloc("pxa2xx-pcmcia", -1); if (!balloon3_pcmcia_device) return -ENOMEM; -- cgit v0.10.2 From 735443fede138fbbad062ec258e4d49648a0a56d Mon Sep 17 00:00:00 2001 From: Dmitry Eremin-Solenikov Date: Fri, 1 Apr 2011 13:28:46 +0400 Subject: pcmcia: limit pxa2xx_trizeps4 subdriver to trizeps4 platform pxa2xx_trizeps4 tries to register pxa2xx-pcmcia device not checking whether machine is really trizeps4, thus messing multi-machine kernels. Fix it up. Signed-off-by: Dmitry Eremin-Solenikov Signed-off-by: Eric Miao diff --git a/drivers/pcmcia/pxa2xx_trizeps4.c b/drivers/pcmcia/pxa2xx_trizeps4.c index b7e5966..5bb4e1a 100644 --- a/drivers/pcmcia/pxa2xx_trizeps4.c +++ b/drivers/pcmcia/pxa2xx_trizeps4.c @@ -226,6 +226,9 @@ static int __init trizeps_pcmcia_init(void) { int ret; + if (!machine_is_trizeps4() && !machine_is_trizeps4wl()) + return -ENODEV; + trizeps_pcmcia_device = platform_device_alloc("pxa2xx-pcmcia", -1); if (!trizeps_pcmcia_device) return -ENOMEM; -- cgit v0.10.2 From 101650726e4a5c0eca3e2716aefeb3296bb177b9 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Fri, 8 Apr 2011 17:02:04 +0200 Subject: samsung-laptop: add support for N230 model Signed-off-by: Greg Kroah-Hartman [mmarek: cherry-picked from staging commit 0789b003] Signed-off-by: Michal Marek Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/samsung-laptop.c b/drivers/platform/x86/samsung-laptop.c index de434c6..11a3fe2 100644 --- a/drivers/platform/x86/samsung-laptop.c +++ b/drivers/platform/x86/samsung-laptop.c @@ -591,12 +591,12 @@ static struct dmi_system_id __initdata samsung_dmi_table[] = { .callback = dmi_check_cb, }, { - .ident = "N150/N210/N220", + .ident = "N150/N210/N220/N230", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "N150/N210/N220"), - DMI_MATCH(DMI_BOARD_NAME, "N150/N210/N220"), + DMI_MATCH(DMI_PRODUCT_NAME, "N150/N210/N220/N230"), + DMI_MATCH(DMI_BOARD_NAME, "N150/N210/N220/N230"), }, .callback = dmi_check_cb, }, -- cgit v0.10.2 From 3d536ed415a8f5d3f683597521df071af9f2866c Mon Sep 17 00:00:00 2001 From: Alberto Mardegan Date: Fri, 8 Apr 2011 17:02:03 +0200 Subject: samsung-laptop: Samsung R410P backlight driver Here's a trivial patch which adds support to the backlight device found in Samsung R410 Plus laptops. Signed-off-by: Alberto Mardegan Signed-off-by: Greg Kroah-Hartman [mmarek: cherry-picked from staging commit d542f180] Signed-off-by: Michal Marek Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/samsung-laptop.c b/drivers/platform/x86/samsung-laptop.c index 11a3fe2..39f30ff 100644 --- a/drivers/platform/x86/samsung-laptop.c +++ b/drivers/platform/x86/samsung-laptop.c @@ -571,6 +571,16 @@ static struct dmi_system_id __initdata samsung_dmi_table[] = { .callback = dmi_check_cb, }, { + .ident = "R410 Plus", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, + "SAMSUNG ELECTRONICS CO., LTD."), + DMI_MATCH(DMI_PRODUCT_NAME, "R410P"), + DMI_MATCH(DMI_BOARD_NAME, "R460"), + }, + .callback = dmi_check_cb, + }, + { .ident = "R518", .matches = { DMI_MATCH(DMI_SYS_VENDOR, -- cgit v0.10.2 From 9d1aea62e45d447e7fc05d4e7f4e90f633e3abfc Mon Sep 17 00:00:00 2001 From: Neil Turton Date: Mon, 4 Apr 2011 13:46:23 +0100 Subject: sfc: Stop the TX queues during loopback self-tests If the TX queues are running during loopback self tests, host traffic gets looped back which causes the test to fail. Avoid restarting the TX queues after the port reset so that any packets sent by the host get held back until after the tests have completed. [bwh: Also wake all TX queues at the end of self-tests.] Signed-off-by: Ben Hutchings diff --git a/drivers/net/sfc/efx.c b/drivers/net/sfc/efx.c index d890679..0dc800b 100644 --- a/drivers/net/sfc/efx.c +++ b/drivers/net/sfc/efx.c @@ -1436,7 +1436,7 @@ static void efx_start_all(struct efx_nic *efx) * restart the transmit interface early so the watchdog timer stops */ efx_start_port(efx); - if (efx_dev_registered(efx)) + if (efx_dev_registered(efx) && !efx->port_inhibited) netif_tx_wake_all_queues(efx->net_dev); efx_for_each_channel(channel, efx) diff --git a/drivers/net/sfc/selftest.c b/drivers/net/sfc/selftest.c index a0f49b3..8458084 100644 --- a/drivers/net/sfc/selftest.c +++ b/drivers/net/sfc/selftest.c @@ -770,6 +770,8 @@ int efx_selftest(struct efx_nic *efx, struct efx_self_tests *tests, __efx_reconfigure_port(efx); mutex_unlock(&efx->mac_lock); + netif_tx_wake_all_queues(efx->net_dev); + return rc_test; } diff --git a/drivers/net/sfc/tx.c b/drivers/net/sfc/tx.c index 1398019..d2c85df 100644 --- a/drivers/net/sfc/tx.c +++ b/drivers/net/sfc/tx.c @@ -435,7 +435,8 @@ void efx_xmit_done(struct efx_tx_queue *tx_queue, unsigned int index) * queue state. */ smp_mb(); if (unlikely(netif_tx_queue_stopped(tx_queue->core_txq)) && - likely(efx->port_enabled)) { + likely(efx->port_enabled) && + likely(!efx->port_inhibited)) { fill_level = tx_queue->insert_count - tx_queue->read_count; if (fill_level < EFX_TXQ_THRESHOLD(efx)) { EFX_BUG_ON_PARANOID(!efx_dev_registered(efx)); -- cgit v0.10.2 From 6932613060b77e2495843d0ce0ce8453d01961a5 Mon Sep 17 00:00:00 2001 From: Haojian Zhuang Date: Fri, 8 Apr 2011 20:15:39 +0800 Subject: ARM: pxa: always clear LPM bits for PXA168 MFPR Bit[9:7] should always be zero in PXA168. Signed-off-by: Haojian Zhuang Signed-off-by: Eric Miao diff --git a/arch/arm/mach-mmp/include/mach/mfp-pxa168.h b/arch/arm/mach-mmp/include/mach/mfp-pxa168.h index 4621067..713be15 100644 --- a/arch/arm/mach-mmp/include/mach/mfp-pxa168.h +++ b/arch/arm/mach-mmp/include/mach/mfp-pxa168.h @@ -8,6 +8,15 @@ #define MFP_DRIVE_MEDIUM (0x2 << 13) #define MFP_DRIVE_FAST (0x3 << 13) +#undef MFP_CFG +#undef MFP_CFG_DRV + +#define MFP_CFG(pin, af) \ + (MFP_LPM_INPUT | MFP_PIN(MFP_PIN_##pin) | MFP_##af | MFP_DRIVE_MEDIUM) + +#define MFP_CFG_DRV(pin, af, drv) \ + (MFP_LPM_INPUT | MFP_PIN(MFP_PIN_##pin) | MFP_##af | MFP_DRIVE_##drv) + /* GPIO */ #define GPIO0_GPIO MFP_CFG(GPIO0, AF5) #define GPIO1_GPIO MFP_CFG(GPIO1, AF5) -- cgit v0.10.2 From a0a4dcbca73a418477f439e812193e2591b46751 Mon Sep 17 00:00:00 2001 From: Haojian Zhuang Date: Fri, 8 Apr 2011 20:15:42 +0800 Subject: ARM: pxa: align NR_BUILTIN_GPIO with GPIO interrupt number Avoid to mismatch between NR_BUILTIN_GPIO and GPIO interrupt number Signed-off-by: Haojian Zhuang Signed-off-by: Eric Miao diff --git a/arch/arm/mach-pxa/include/mach/gpio.h b/arch/arm/mach-pxa/include/mach/gpio.h index b024a8b..48ef925 100644 --- a/arch/arm/mach-pxa/include/mach/gpio.h +++ b/arch/arm/mach-pxa/include/mach/gpio.h @@ -99,7 +99,7 @@ #define GAFR(x) GPIO_REG(0x54 + (((x) & 0x70) >> 2)) -#define NR_BUILTIN_GPIO 128 +#define NR_BUILTIN_GPIO PXA_GPIO_IRQ_NUM #define gpio_to_bank(gpio) ((gpio) >> 5) #define gpio_to_irq(gpio) IRQ_GPIO(gpio) -- cgit v0.10.2 From 83fd6c685bd8e83be1e29e2841bab94dd831e186 Mon Sep 17 00:00:00 2001 From: Haojian Zhuang Date: Fri, 8 Apr 2011 20:15:43 +0800 Subject: ARM: mmp: align NR_BUILTIN_GPIO with gpio interrupt number Avoid to mismatch between NR_BUILTIN_GPIO and gpio interrupt number. Signed-off-by: Haojian Zhuang Signed-off-by: Eric Miao diff --git a/arch/arm/mach-mmp/include/mach/gpio.h b/arch/arm/mach-mmp/include/mach/gpio.h index ee8b02e..7bfb827 100644 --- a/arch/arm/mach-mmp/include/mach/gpio.h +++ b/arch/arm/mach-mmp/include/mach/gpio.h @@ -10,7 +10,7 @@ #define BANK_OFF(n) (((n) < 3) ? (n) << 2 : 0x100 + (((n) - 3) << 2)) #define GPIO_REG(x) (*((volatile u32 *)(GPIO_REGS_VIRT + (x)))) -#define NR_BUILTIN_GPIO (192) +#define NR_BUILTIN_GPIO IRQ_GPIO_NUM #define gpio_to_bank(gpio) ((gpio) >> 5) #define gpio_to_irq(gpio) (IRQ_GPIO_START + (gpio)) -- cgit v0.10.2 From 993819c563695efc2782d312f55bec634c2d2da5 Mon Sep 17 00:00:00 2001 From: Michal Marek Date: Fri, 8 Apr 2011 17:02:06 +0200 Subject: staging: samsung-laptop has moved to platform/x86 Signed-off-by: Michal Marek Signed-off-by: Matthew Garrett diff --git a/drivers/staging/Kconfig b/drivers/staging/Kconfig index 18b43fc..9d5a033 100644 --- a/drivers/staging/Kconfig +++ b/drivers/staging/Kconfig @@ -133,8 +133,6 @@ source "drivers/staging/wlags49_h2/Kconfig" source "drivers/staging/wlags49_h25/Kconfig" -source "drivers/staging/samsung-laptop/Kconfig" - source "drivers/staging/sm7xx/Kconfig" source "drivers/staging/dt3155v4l/Kconfig" diff --git a/drivers/staging/Makefile b/drivers/staging/Makefile index cfd13cd..adebf8e 100644 --- a/drivers/staging/Makefile +++ b/drivers/staging/Makefile @@ -49,7 +49,6 @@ obj-$(CONFIG_XVMALLOC) += zram/ obj-$(CONFIG_ZCACHE) += zcache/ obj-$(CONFIG_WLAGS49_H2) += wlags49_h2/ obj-$(CONFIG_WLAGS49_H25) += wlags49_h25/ -obj-$(CONFIG_SAMSUNG_LAPTOP) += samsung-laptop/ obj-$(CONFIG_FB_SM7XX) += sm7xx/ obj-$(CONFIG_VIDEO_DT3155) += dt3155v4l/ obj-$(CONFIG_CRYSTALHD) += crystalhd/ diff --git a/drivers/staging/samsung-laptop/Kconfig b/drivers/staging/samsung-laptop/Kconfig deleted file mode 100644 index f27c608..0000000 --- a/drivers/staging/samsung-laptop/Kconfig +++ /dev/null @@ -1,10 +0,0 @@ -config SAMSUNG_LAPTOP - tristate "Samsung Laptop driver" - default n - depends on RFKILL && BACKLIGHT_CLASS_DEVICE && X86 - help - This module implements a driver for the N128 Samsung Laptop - providing control over the Wireless LED and the LCD backlight - - To compile this driver as a module, choose - M here: the module will be called samsung-laptop. diff --git a/drivers/staging/samsung-laptop/Makefile b/drivers/staging/samsung-laptop/Makefile deleted file mode 100644 index 3c6f420..0000000 --- a/drivers/staging/samsung-laptop/Makefile +++ /dev/null @@ -1 +0,0 @@ -obj-$(CONFIG_SAMSUNG_LAPTOP) += samsung-laptop.o diff --git a/drivers/staging/samsung-laptop/TODO b/drivers/staging/samsung-laptop/TODO deleted file mode 100644 index f7a6d58..0000000 --- a/drivers/staging/samsung-laptop/TODO +++ /dev/null @@ -1,5 +0,0 @@ -TODO: - - review from other developers - - figure out ACPI video issues - -Please send patches to Greg Kroah-Hartman diff --git a/drivers/staging/samsung-laptop/samsung-laptop.c b/drivers/staging/samsung-laptop/samsung-laptop.c deleted file mode 100644 index 2529446..0000000 --- a/drivers/staging/samsung-laptop/samsung-laptop.c +++ /dev/null @@ -1,843 +0,0 @@ -/* - * Samsung Laptop driver - * - * Copyright (C) 2009,2011 Greg Kroah-Hartman (gregkh@suse.de) - * Copyright (C) 2009,2011 Novell Inc. - * - * This program is free software; you can redistribute it and/or modify it - * under the terms of the GNU General Public License version 2 as published by - * the Free Software Foundation. - * - */ -#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -/* - * This driver is needed because a number of Samsung laptops do not hook - * their control settings through ACPI. So we have to poke around in the - * BIOS to do things like brightness values, and "special" key controls. - */ - -/* - * We have 0 - 8 as valid brightness levels. The specs say that level 0 should - * be reserved by the BIOS (which really doesn't make much sense), we tell - * userspace that the value is 0 - 7 and then just tell the hardware 1 - 8 - */ -#define MAX_BRIGHT 0x07 - - -#define SABI_IFACE_MAIN 0x00 -#define SABI_IFACE_SUB 0x02 -#define SABI_IFACE_COMPLETE 0x04 -#define SABI_IFACE_DATA 0x05 - -/* Structure to get data back to the calling function */ -struct sabi_retval { - u8 retval[20]; -}; - -struct sabi_header_offsets { - u8 port; - u8 re_mem; - u8 iface_func; - u8 en_mem; - u8 data_offset; - u8 data_segment; -}; - -struct sabi_commands { - /* - * Brightness is 0 - 8, as described above. - * Value 0 is for the BIOS to use - */ - u8 get_brightness; - u8 set_brightness; - - /* - * first byte: - * 0x00 - wireless is off - * 0x01 - wireless is on - * second byte: - * 0x02 - 3G is off - * 0x03 - 3G is on - * TODO, verify 3G is correct, that doesn't seem right... - */ - u8 get_wireless_button; - u8 set_wireless_button; - - /* 0 is off, 1 is on */ - u8 get_backlight; - u8 set_backlight; - - /* - * 0x80 or 0x00 - no action - * 0x81 - recovery key pressed - */ - u8 get_recovery_mode; - u8 set_recovery_mode; - - /* - * on seclinux: 0 is low, 1 is high, - * on swsmi: 0 is normal, 1 is silent, 2 is turbo - */ - u8 get_performance_level; - u8 set_performance_level; - - /* - * Tell the BIOS that Linux is running on this machine. - * 81 is on, 80 is off - */ - u8 set_linux; -}; - -struct sabi_performance_level { - const char *name; - u8 value; -}; - -struct sabi_config { - const char *test_string; - u16 main_function; - const struct sabi_header_offsets header_offsets; - const struct sabi_commands commands; - const struct sabi_performance_level performance_levels[4]; - u8 min_brightness; - u8 max_brightness; -}; - -static const struct sabi_config sabi_configs[] = { - { - .test_string = "SECLINUX", - - .main_function = 0x4c49, - - .header_offsets = { - .port = 0x00, - .re_mem = 0x02, - .iface_func = 0x03, - .en_mem = 0x04, - .data_offset = 0x05, - .data_segment = 0x07, - }, - - .commands = { - .get_brightness = 0x00, - .set_brightness = 0x01, - - .get_wireless_button = 0x02, - .set_wireless_button = 0x03, - - .get_backlight = 0x04, - .set_backlight = 0x05, - - .get_recovery_mode = 0x06, - .set_recovery_mode = 0x07, - - .get_performance_level = 0x08, - .set_performance_level = 0x09, - - .set_linux = 0x0a, - }, - - .performance_levels = { - { - .name = "silent", - .value = 0, - }, - { - .name = "normal", - .value = 1, - }, - { }, - }, - .min_brightness = 1, - .max_brightness = 8, - }, - { - .test_string = "SwSmi@", - - .main_function = 0x5843, - - .header_offsets = { - .port = 0x00, - .re_mem = 0x04, - .iface_func = 0x02, - .en_mem = 0x03, - .data_offset = 0x05, - .data_segment = 0x07, - }, - - .commands = { - .get_brightness = 0x10, - .set_brightness = 0x11, - - .get_wireless_button = 0x12, - .set_wireless_button = 0x13, - - .get_backlight = 0x2d, - .set_backlight = 0x2e, - - .get_recovery_mode = 0xff, - .set_recovery_mode = 0xff, - - .get_performance_level = 0x31, - .set_performance_level = 0x32, - - .set_linux = 0xff, - }, - - .performance_levels = { - { - .name = "normal", - .value = 0, - }, - { - .name = "silent", - .value = 1, - }, - { - .name = "overclock", - .value = 2, - }, - { }, - }, - .min_brightness = 0, - .max_brightness = 8, - }, - { }, -}; - -static const struct sabi_config *sabi_config; - -static void __iomem *sabi; -static void __iomem *sabi_iface; -static void __iomem *f0000_segment; -static struct backlight_device *backlight_device; -static struct mutex sabi_mutex; -static struct platform_device *sdev; -static struct rfkill *rfk; - -static int force; -module_param(force, bool, 0); -MODULE_PARM_DESC(force, - "Disable the DMI check and forces the driver to be loaded"); - -static int debug; -module_param(debug, bool, S_IRUGO | S_IWUSR); -MODULE_PARM_DESC(debug, "Debug enabled or not"); - -static int sabi_get_command(u8 command, struct sabi_retval *sretval) -{ - int retval = 0; - u16 port = readw(sabi + sabi_config->header_offsets.port); - u8 complete, iface_data; - - mutex_lock(&sabi_mutex); - - /* enable memory to be able to write to it */ - outb(readb(sabi + sabi_config->header_offsets.en_mem), port); - - /* write out the command */ - writew(sabi_config->main_function, sabi_iface + SABI_IFACE_MAIN); - writew(command, sabi_iface + SABI_IFACE_SUB); - writeb(0, sabi_iface + SABI_IFACE_COMPLETE); - outb(readb(sabi + sabi_config->header_offsets.iface_func), port); - - /* write protect memory to make it safe */ - outb(readb(sabi + sabi_config->header_offsets.re_mem), port); - - /* see if the command actually succeeded */ - complete = readb(sabi_iface + SABI_IFACE_COMPLETE); - iface_data = readb(sabi_iface + SABI_IFACE_DATA); - if (complete != 0xaa || iface_data == 0xff) { - pr_warn("SABI get command 0x%02x failed with completion flag 0x%02x and data 0x%02x\n", - command, complete, iface_data); - retval = -EINVAL; - goto exit; - } - /* - * Save off the data into a structure so the caller use it. - * Right now we only want the first 4 bytes, - * There are commands that need more, but not for the ones we - * currently care about. - */ - sretval->retval[0] = readb(sabi_iface + SABI_IFACE_DATA); - sretval->retval[1] = readb(sabi_iface + SABI_IFACE_DATA + 1); - sretval->retval[2] = readb(sabi_iface + SABI_IFACE_DATA + 2); - sretval->retval[3] = readb(sabi_iface + SABI_IFACE_DATA + 3); - -exit: - mutex_unlock(&sabi_mutex); - return retval; - -} - -static int sabi_set_command(u8 command, u8 data) -{ - int retval = 0; - u16 port = readw(sabi + sabi_config->header_offsets.port); - u8 complete, iface_data; - - mutex_lock(&sabi_mutex); - - /* enable memory to be able to write to it */ - outb(readb(sabi + sabi_config->header_offsets.en_mem), port); - - /* write out the command */ - writew(sabi_config->main_function, sabi_iface + SABI_IFACE_MAIN); - writew(command, sabi_iface + SABI_IFACE_SUB); - writeb(0, sabi_iface + SABI_IFACE_COMPLETE); - writeb(data, sabi_iface + SABI_IFACE_DATA); - outb(readb(sabi + sabi_config->header_offsets.iface_func), port); - - /* write protect memory to make it safe */ - outb(readb(sabi + sabi_config->header_offsets.re_mem), port); - - /* see if the command actually succeeded */ - complete = readb(sabi_iface + SABI_IFACE_COMPLETE); - iface_data = readb(sabi_iface + SABI_IFACE_DATA); - if (complete != 0xaa || iface_data == 0xff) { - pr_warn("SABI set command 0x%02x failed with completion flag 0x%02x and data 0x%02x\n", - command, complete, iface_data); - retval = -EINVAL; - } - - mutex_unlock(&sabi_mutex); - return retval; -} - -static void test_backlight(void) -{ - struct sabi_retval sretval; - - sabi_get_command(sabi_config->commands.get_backlight, &sretval); - printk(KERN_DEBUG "backlight = 0x%02x\n", sretval.retval[0]); - - sabi_set_command(sabi_config->commands.set_backlight, 0); - printk(KERN_DEBUG "backlight should be off\n"); - - sabi_get_command(sabi_config->commands.get_backlight, &sretval); - printk(KERN_DEBUG "backlight = 0x%02x\n", sretval.retval[0]); - - msleep(1000); - - sabi_set_command(sabi_config->commands.set_backlight, 1); - printk(KERN_DEBUG "backlight should be on\n"); - - sabi_get_command(sabi_config->commands.get_backlight, &sretval); - printk(KERN_DEBUG "backlight = 0x%02x\n", sretval.retval[0]); -} - -static void test_wireless(void) -{ - struct sabi_retval sretval; - - sabi_get_command(sabi_config->commands.get_wireless_button, &sretval); - printk(KERN_DEBUG "wireless led = 0x%02x\n", sretval.retval[0]); - - sabi_set_command(sabi_config->commands.set_wireless_button, 0); - printk(KERN_DEBUG "wireless led should be off\n"); - - sabi_get_command(sabi_config->commands.get_wireless_button, &sretval); - printk(KERN_DEBUG "wireless led = 0x%02x\n", sretval.retval[0]); - - msleep(1000); - - sabi_set_command(sabi_config->commands.set_wireless_button, 1); - printk(KERN_DEBUG "wireless led should be on\n"); - - sabi_get_command(sabi_config->commands.get_wireless_button, &sretval); - printk(KERN_DEBUG "wireless led = 0x%02x\n", sretval.retval[0]); -} - -static u8 read_brightness(void) -{ - struct sabi_retval sretval; - int user_brightness = 0; - int retval; - - retval = sabi_get_command(sabi_config->commands.get_brightness, - &sretval); - if (!retval) { - user_brightness = sretval.retval[0]; - if (user_brightness != 0) - user_brightness -= sabi_config->min_brightness; - } - return user_brightness; -} - -static void set_brightness(u8 user_brightness) -{ - u8 user_level = user_brightness - sabi_config->min_brightness; - - sabi_set_command(sabi_config->commands.set_brightness, user_level); -} - -static int get_brightness(struct backlight_device *bd) -{ - return (int)read_brightness(); -} - -static int update_status(struct backlight_device *bd) -{ - set_brightness(bd->props.brightness); - - if (bd->props.power == FB_BLANK_UNBLANK) - sabi_set_command(sabi_config->commands.set_backlight, 1); - else - sabi_set_command(sabi_config->commands.set_backlight, 0); - return 0; -} - -static const struct backlight_ops backlight_ops = { - .get_brightness = get_brightness, - .update_status = update_status, -}; - -static int rfkill_set(void *data, bool blocked) -{ - /* Do something with blocked...*/ - /* - * blocked == false is on - * blocked == true is off - */ - if (blocked) - sabi_set_command(sabi_config->commands.set_wireless_button, 0); - else - sabi_set_command(sabi_config->commands.set_wireless_button, 1); - - return 0; -} - -static struct rfkill_ops rfkill_ops = { - .set_block = rfkill_set, -}; - -static int init_wireless(struct platform_device *sdev) -{ - int retval; - - rfk = rfkill_alloc("samsung-wifi", &sdev->dev, RFKILL_TYPE_WLAN, - &rfkill_ops, NULL); - if (!rfk) - return -ENOMEM; - - retval = rfkill_register(rfk); - if (retval) { - rfkill_destroy(rfk); - return -ENODEV; - } - - return 0; -} - -static void destroy_wireless(void) -{ - rfkill_unregister(rfk); - rfkill_destroy(rfk); -} - -static ssize_t get_performance_level(struct device *dev, - struct device_attribute *attr, char *buf) -{ - struct sabi_retval sretval; - int retval; - int i; - - /* Read the state */ - retval = sabi_get_command(sabi_config->commands.get_performance_level, - &sretval); - if (retval) - return retval; - - /* The logic is backwards, yeah, lots of fun... */ - for (i = 0; sabi_config->performance_levels[i].name; ++i) { - if (sretval.retval[0] == sabi_config->performance_levels[i].value) - return sprintf(buf, "%s\n", sabi_config->performance_levels[i].name); - } - return sprintf(buf, "%s\n", "unknown"); -} - -static ssize_t set_performance_level(struct device *dev, - struct device_attribute *attr, const char *buf, - size_t count) -{ - if (count >= 1) { - int i; - for (i = 0; sabi_config->performance_levels[i].name; ++i) { - const struct sabi_performance_level *level = - &sabi_config->performance_levels[i]; - if (!strncasecmp(level->name, buf, strlen(level->name))) { - sabi_set_command(sabi_config->commands.set_performance_level, - level->value); - break; - } - } - if (!sabi_config->performance_levels[i].name) - return -EINVAL; - } - return count; -} -static DEVICE_ATTR(performance_level, S_IWUSR | S_IRUGO, - get_performance_level, set_performance_level); - - -static int __init dmi_check_cb(const struct dmi_system_id *id) -{ - pr_info("found laptop model '%s'\n", - id->ident); - return 0; -} - -static struct dmi_system_id __initdata samsung_dmi_table[] = { - { - .ident = "N128", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "N128"), - DMI_MATCH(DMI_BOARD_NAME, "N128"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "N130", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "N130"), - DMI_MATCH(DMI_BOARD_NAME, "N130"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "X125", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "X125"), - DMI_MATCH(DMI_BOARD_NAME, "X125"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "X120/X170", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "X120/X170"), - DMI_MATCH(DMI_BOARD_NAME, "X120/X170"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "NC10", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "NC10"), - DMI_MATCH(DMI_BOARD_NAME, "NC10"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "NP-Q45", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "SQ45S70S"), - DMI_MATCH(DMI_BOARD_NAME, "SQ45S70S"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "X360", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "X360"), - DMI_MATCH(DMI_BOARD_NAME, "X360"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "R410 Plus", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "R410P"), - DMI_MATCH(DMI_BOARD_NAME, "R460"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "R518", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "R518"), - DMI_MATCH(DMI_BOARD_NAME, "R518"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "R519/R719", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "R519/R719"), - DMI_MATCH(DMI_BOARD_NAME, "R519/R719"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "N150/N210/N220/N230", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "N150/N210/N220/N230"), - DMI_MATCH(DMI_BOARD_NAME, "N150/N210/N220/N230"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "N150P/N210P/N220P", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "N150P/N210P/N220P"), - DMI_MATCH(DMI_BOARD_NAME, "N150P/N210P/N220P"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "R530/R730", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "R530/R730"), - DMI_MATCH(DMI_BOARD_NAME, "R530/R730"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "NF110/NF210/NF310", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "NF110/NF210/NF310"), - DMI_MATCH(DMI_BOARD_NAME, "NF110/NF210/NF310"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "N145P/N250P/N260P", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "N145P/N250P/N260P"), - DMI_MATCH(DMI_BOARD_NAME, "N145P/N250P/N260P"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "R70/R71", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, - "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "R70/R71"), - DMI_MATCH(DMI_BOARD_NAME, "R70/R71"), - }, - .callback = dmi_check_cb, - }, - { - .ident = "P460", - .matches = { - DMI_MATCH(DMI_SYS_VENDOR, "SAMSUNG ELECTRONICS CO., LTD."), - DMI_MATCH(DMI_PRODUCT_NAME, "P460"), - DMI_MATCH(DMI_BOARD_NAME, "P460"), - }, - .callback = dmi_check_cb, - }, - { }, -}; -MODULE_DEVICE_TABLE(dmi, samsung_dmi_table); - -static int find_signature(void __iomem *memcheck, const char *testStr) -{ - int i = 0; - int loca; - - for (loca = 0; loca < 0xffff; loca++) { - char temp = readb(memcheck + loca); - - if (temp == testStr[i]) { - if (i == strlen(testStr)-1) - break; - ++i; - } else { - i = 0; - } - } - return loca; -} - -static int __init samsung_init(void) -{ - struct backlight_properties props; - struct sabi_retval sretval; - unsigned int ifaceP; - int i; - int loca; - int retval; - - mutex_init(&sabi_mutex); - - if (!force && !dmi_check_system(samsung_dmi_table)) - return -ENODEV; - - f0000_segment = ioremap_nocache(0xf0000, 0xffff); - if (!f0000_segment) { - pr_err("Can't map the segment at 0xf0000\n"); - return -EINVAL; - } - - /* Try to find one of the signatures in memory to find the header */ - for (i = 0; sabi_configs[i].test_string != 0; ++i) { - sabi_config = &sabi_configs[i]; - loca = find_signature(f0000_segment, sabi_config->test_string); - if (loca != 0xffff) - break; - } - - if (loca == 0xffff) { - pr_err("This computer does not support SABI\n"); - goto error_no_signature; - } - - /* point to the SMI port Number */ - loca += 1; - sabi = (f0000_segment + loca); - - if (debug) { - printk(KERN_DEBUG "This computer supports SABI==%x\n", - loca + 0xf0000 - 6); - printk(KERN_DEBUG "SABI header:\n"); - printk(KERN_DEBUG " SMI Port Number = 0x%04x\n", - readw(sabi + sabi_config->header_offsets.port)); - printk(KERN_DEBUG " SMI Interface Function = 0x%02x\n", - readb(sabi + sabi_config->header_offsets.iface_func)); - printk(KERN_DEBUG " SMI enable memory buffer = 0x%02x\n", - readb(sabi + sabi_config->header_offsets.en_mem)); - printk(KERN_DEBUG " SMI restore memory buffer = 0x%02x\n", - readb(sabi + sabi_config->header_offsets.re_mem)); - printk(KERN_DEBUG " SABI data offset = 0x%04x\n", - readw(sabi + sabi_config->header_offsets.data_offset)); - printk(KERN_DEBUG " SABI data segment = 0x%04x\n", - readw(sabi + sabi_config->header_offsets.data_segment)); - } - - /* Get a pointer to the SABI Interface */ - ifaceP = (readw(sabi + sabi_config->header_offsets.data_segment) & 0x0ffff) << 4; - ifaceP += readw(sabi + sabi_config->header_offsets.data_offset) & 0x0ffff; - sabi_iface = ioremap_nocache(ifaceP, 16); - if (!sabi_iface) { - pr_err("Can't remap %x\n", ifaceP); - goto exit; - } - if (debug) { - printk(KERN_DEBUG "ifaceP = 0x%08x\n", ifaceP); - printk(KERN_DEBUG "sabi_iface = %p\n", sabi_iface); - - test_backlight(); - test_wireless(); - - retval = sabi_get_command(sabi_config->commands.get_brightness, - &sretval); - printk(KERN_DEBUG "brightness = 0x%02x\n", sretval.retval[0]); - } - - /* Turn on "Linux" mode in the BIOS */ - if (sabi_config->commands.set_linux != 0xff) { - retval = sabi_set_command(sabi_config->commands.set_linux, - 0x81); - if (retval) { - pr_warn("Linux mode was not set!\n"); - goto error_no_platform; - } - } - - /* knock up a platform device to hang stuff off of */ - sdev = platform_device_register_simple("samsung", -1, NULL, 0); - if (IS_ERR(sdev)) - goto error_no_platform; - - /* create a backlight device to talk to this one */ - memset(&props, 0, sizeof(struct backlight_properties)); - props.type = BACKLIGHT_PLATFORM; - props.max_brightness = sabi_config->max_brightness; - backlight_device = backlight_device_register("samsung", &sdev->dev, - NULL, &backlight_ops, - &props); - if (IS_ERR(backlight_device)) - goto error_no_backlight; - - backlight_device->props.brightness = read_brightness(); - backlight_device->props.power = FB_BLANK_UNBLANK; - backlight_update_status(backlight_device); - - retval = init_wireless(sdev); - if (retval) - goto error_no_rfk; - - retval = device_create_file(&sdev->dev, &dev_attr_performance_level); - if (retval) - goto error_file_create; - -exit: - return 0; - -error_file_create: - destroy_wireless(); - -error_no_rfk: - backlight_device_unregister(backlight_device); - -error_no_backlight: - platform_device_unregister(sdev); - -error_no_platform: - iounmap(sabi_iface); - -error_no_signature: - iounmap(f0000_segment); - return -EINVAL; -} - -static void __exit samsung_exit(void) -{ - /* Turn off "Linux" mode in the BIOS */ - if (sabi_config->commands.set_linux != 0xff) - sabi_set_command(sabi_config->commands.set_linux, 0x80); - - device_remove_file(&sdev->dev, &dev_attr_performance_level); - backlight_device_unregister(backlight_device); - destroy_wireless(); - iounmap(sabi_iface); - iounmap(f0000_segment); - platform_device_unregister(sdev); -} - -module_init(samsung_init); -module_exit(samsung_exit); - -MODULE_AUTHOR("Greg Kroah-Hartman "); -MODULE_DESCRIPTION("Samsung Backlight driver"); -MODULE_LICENSE("GPL"); -- cgit v0.10.2 From 8713b04ab8da256493410bd4f25170d2e0836534 Mon Sep 17 00:00:00 2001 From: Michal Marek Date: Fri, 8 Apr 2011 17:02:05 +0200 Subject: samsung-laptop: set backlight type Cherry-picked from drivers/staging/samsung-laptop/samsung-laptop.c Signed-off-by: Michal Marek Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/samsung-laptop.c b/drivers/platform/x86/samsung-laptop.c index 39f30ff..d347116 100644 --- a/drivers/platform/x86/samsung-laptop.c +++ b/drivers/platform/x86/samsung-laptop.c @@ -781,6 +781,7 @@ static int __init samsung_init(void) /* create a backlight device to talk to this one */ memset(&props, 0, sizeof(struct backlight_properties)); + props.type = BACKLIGHT_PLATFORM; props.max_brightness = sabi_config->max_brightness; backlight_device = backlight_device_register("samsung", &sdev->dev, NULL, &backlight_ops, -- cgit v0.10.2 From 855b8bc9953fdf40095fdec9d91d49736ca7b17c Mon Sep 17 00:00:00 2001 From: Mattia Dongili Date: Tue, 5 Apr 2011 23:38:35 +0900 Subject: sony-laptop: only show the handles sysfs file in debug mode It makes no sense to expose this type of information to userspace unless the driver was explicitly loaded with the debug option. Signed-off-by: Mattia Dongili Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index 33c3b8c..33d9555 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -771,11 +771,6 @@ static int sony_nc_handles_setup(struct platform_device *pd) if (!handles) return -ENOMEM; - sysfs_attr_init(&handles->devattr.attr); - handles->devattr.attr.name = "handles"; - handles->devattr.attr.mode = S_IRUGO; - handles->devattr.show = sony_nc_handles_show; - for (i = 0; i < ARRAY_SIZE(handles->cap); i++) { if (!acpi_callsetfunc(sony_nc_acpi_handle, "SN00", i + 0x20, &result)) { @@ -785,11 +780,18 @@ static int sony_nc_handles_setup(struct platform_device *pd) } } - /* allow reading capabilities via sysfs */ - if (device_create_file(&pd->dev, &handles->devattr)) { - kfree(handles); - handles = NULL; - return -1; + if (debug) { + sysfs_attr_init(&handles->devattr.attr); + handles->devattr.attr.name = "handles"; + handles->devattr.attr.mode = S_IRUGO; + handles->devattr.show = sony_nc_handles_show; + + /* allow reading capabilities via sysfs */ + if (device_create_file(&pd->dev, &handles->devattr)) { + kfree(handles); + handles = NULL; + return -1; + } } return 0; @@ -798,7 +800,8 @@ static int sony_nc_handles_setup(struct platform_device *pd) static int sony_nc_handles_cleanup(struct platform_device *pd) { if (handles) { - device_remove_file(&pd->dev, &handles->devattr); + if (debug) + device_remove_file(&pd->dev, &handles->devattr); kfree(handles); handles = NULL; } -- cgit v0.10.2 From df410d522410e676602a14eb5957c4b3e1fa3902 Mon Sep 17 00:00:00 2001 From: Marco Chiappero Date: Tue, 5 Apr 2011 23:38:34 +0900 Subject: sony-laptop: keyboard backlight fixes Restore the original state on module removal, set the latest values on resume. When setting the keyboard backlight mode try to turn on/off backlight immediately. [malattia@linux.it: patch taken from a largely modified sony-laptop.c, ported and slightly modified to use defines already available.] Signed-off-by: Mattia Dongili Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/sony-laptop.c b/drivers/platform/x86/sony-laptop.c index 33d9555..8f709ae 100644 --- a/drivers/platform/x86/sony-laptop.c +++ b/drivers/platform/x86/sony-laptop.c @@ -138,6 +138,8 @@ MODULE_PARM_DESC(kbd_backlight_timeout, "1 for 30 seconds, 2 for 60 seconds and 3 to disable timeout " "(default: 0)"); +static void sony_nc_kbd_backlight_resume(void); + enum sony_nc_rfkill { SONY_WIFI, SONY_BLUETOOTH, @@ -1176,6 +1178,9 @@ static int sony_nc_resume(struct acpi_device *device) /* re-read rfkill state */ sony_nc_rfkill_update(); + /* restore kbd backlight states */ + sony_nc_kbd_backlight_resume(); + return 0; } @@ -1363,6 +1368,7 @@ out_no_enum: #define KBDBL_HANDLER 0x137 #define KBDBL_PRESENT 0xB00 #define SET_MODE 0xC00 +#define SET_STATE 0xD00 #define SET_TIMEOUT 0xE00 struct kbd_backlight { @@ -1385,6 +1391,10 @@ static ssize_t __sony_nc_kbd_backlight_mode_set(u8 value) (value << 0x10) | SET_MODE, &result)) return -EIO; + /* Try to turn the light on/off immediately */ + sony_call_snc_handle(KBDBL_HANDLER, (value << 0x10) | SET_STATE, + &result); + kbdbl_handle->mode = value; return 0; @@ -1466,7 +1476,7 @@ static int sony_nc_kbd_backlight_setup(struct platform_device *pd) { int result; - if (sony_call_snc_handle(0x137, KBDBL_PRESENT, &result)) + if (sony_call_snc_handle(KBDBL_HANDLER, KBDBL_PRESENT, &result)) return 0; if (!(result & 0x02)) return 0; @@ -1509,13 +1519,36 @@ outkzalloc: static int sony_nc_kbd_backlight_cleanup(struct platform_device *pd) { if (kbdbl_handle) { + int result; + device_remove_file(&pd->dev, &kbdbl_handle->mode_attr); device_remove_file(&pd->dev, &kbdbl_handle->timeout_attr); + + /* restore the default hw behaviour */ + sony_call_snc_handle(KBDBL_HANDLER, 0x1000 | SET_MODE, &result); + sony_call_snc_handle(KBDBL_HANDLER, SET_TIMEOUT, &result); + kfree(kbdbl_handle); } return 0; } +static void sony_nc_kbd_backlight_resume(void) +{ + int ignore = 0; + + if (!kbdbl_handle) + return; + + if (kbdbl_handle->mode == 0) + sony_call_snc_handle(KBDBL_HANDLER, SET_MODE, &ignore); + + if (kbdbl_handle->timeout != 0) + sony_call_snc_handle(KBDBL_HANDLER, + (kbdbl_handle->timeout << 0x10) | SET_TIMEOUT, + &ignore); +} + static void sony_nc_backlight_setup(void) { acpi_handle unused; -- cgit v0.10.2 From 08a0799d5736f1494ef35d386570d177447acbfb Mon Sep 17 00:00:00 2001 From: "Lee, Chun-Yi" Date: Wed, 6 Apr 2011 17:40:06 +0800 Subject: acer-wmi: Fix capitalisation of GUID in module alias MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit wmi:6AF4F258-B401-42Fd-BE91-3D4AC2D7C0D3 needs to be wmi:6AF4F258-B401-42FD-BE91-3D4AC2D7C0D3 in module alias for acer-wmi is automatically loaded. Cc: Pali Rohár Cc: Carlos Corbacho Cc: Matthew Garrett Signed-off-by: Lee, Chun-Yi Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/acer-wmi.c b/drivers/platform/x86/acer-wmi.c index 5ea6c34..ac4e7f83 100644 --- a/drivers/platform/x86/acer-wmi.c +++ b/drivers/platform/x86/acer-wmi.c @@ -89,7 +89,7 @@ MODULE_LICENSE("GPL"); #define ACERWMID_EVENT_GUID "676AA15E-6A47-4D9F-A2CC-1E6D18D14026" MODULE_ALIAS("wmi:67C3371D-95A3-4C37-BB61-DD47B491DAAB"); -MODULE_ALIAS("wmi:6AF4F258-B401-42Fd-BE91-3D4AC2D7C0D3"); +MODULE_ALIAS("wmi:6AF4F258-B401-42FD-BE91-3D4AC2D7C0D3"); MODULE_ALIAS("wmi:676AA15E-6A47-4D9F-A2CC-1E6D18D14026"); enum acer_wmi_event_ids { -- cgit v0.10.2 From b569ab3911aca64841bd819720d2b241aa09d713 Mon Sep 17 00:00:00 2001 From: Keith Packard Date: Thu, 31 Mar 2011 15:22:33 -0700 Subject: thinkpad-acpi fails to load with newer Thinkpad X201s BIOS The new BIOS has a slightly different EC version string. From a1541710300b083a1a9acff2890d721d15ede62b Mon Sep 17 00:00:00 2001 From: Keith Packard Date: Sun, 13 Mar 2011 23:46:22 -0700 Subject: [PATCH] thinkpad-acpi: Some BIOS versions don't end in WW, remove check My X201s BIOS version string is 6QET46V1 (1.16 ). The EC version string is 6QHT28WW-1.09. The driver was requiring that both of these have 'WW' in positions 6 and 7. I don't know what the significance of having 'V1' there instead is, but removing the test makes the driver load on my machine. Signed-off-by: Keith Packard Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/thinkpad_acpi.c b/drivers/platform/x86/thinkpad_acpi.c index a08561f..efb3b6b 100644 --- a/drivers/platform/x86/thinkpad_acpi.c +++ b/drivers/platform/x86/thinkpad_acpi.c @@ -8618,8 +8618,7 @@ static bool __pure __init tpacpi_is_valid_fw_id(const char* const s, tpacpi_is_fw_digit(s[1]) && s[2] == t && s[3] == 'T' && tpacpi_is_fw_digit(s[4]) && - tpacpi_is_fw_digit(s[5]) && - s[6] == 'W' && s[7] == 'W'; + tpacpi_is_fw_digit(s[5]); } /* returns 0 - probe ok, or < 0 - probe error. -- cgit v0.10.2 From d4fabcc8e8ecac21262b1a5b9684fe415b128bd2 Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Mon, 4 Apr 2011 14:22:11 +0100 Subject: sfc: Do not use efx_process_channel_now() in online self-test During self-tests we use efx_process_channel_now() to handle completion and other events synchronously. This disables interrupts and NAPI processing for the channel in question, but it may still be interrupted by another channel. A single socket may receive packets from multiple net devices or even multiple channels of the same net device, so this can result in deadlock on a socket lock. Receiving packets in process context will also result in incorrect classification by the network cgroup classifier. Therefore, we must only use efx_process_channel_now() in the offline loopback tests (which never deliver packets up the stack) and not for the online interrupt and event tests. For the interrupt test, there is no reason to process events. We only care that an interrupt is raised. For the event test, we want to know whether events have been received, and there may be many events ahead of the one we inject. Therefore remove efx_channel::magic_count and instead test whether efx_channel::eventq_read_ptr advances. This is currently an event queue index and might wrap around to exactly the same value, resulting in a false negative. Therefore move the masking to efx_event() and efx_nic_eventq_read_ack() so that it cannot wrap within the time of the test. The event test also tries to diagnose failures by checking whether an event was delivered without causing an interrupt. Add and use a helper function that only does this. Signed-off-by: Ben Hutchings diff --git a/drivers/net/sfc/efx.c b/drivers/net/sfc/efx.c index 0dc800b..a3c2aab 100644 --- a/drivers/net/sfc/efx.c +++ b/drivers/net/sfc/efx.c @@ -328,7 +328,8 @@ static int efx_poll(struct napi_struct *napi, int budget) * processing to finish, then directly poll (and ack ) the eventq. * Finally reenable NAPI and interrupts. * - * Since we are touching interrupts the caller should hold the suspend lock + * This is for use only during a loopback self-test. It must not + * deliver any packets up the stack as this can result in deadlock. */ void efx_process_channel_now(struct efx_channel *channel) { @@ -336,6 +337,7 @@ void efx_process_channel_now(struct efx_channel *channel) BUG_ON(channel->channel >= efx->n_channels); BUG_ON(!channel->enabled); + BUG_ON(!efx->loopback_selftest); /* Disable interrupts and wait for ISRs to complete */ efx_nic_disable_interrupts(efx); diff --git a/drivers/net/sfc/net_driver.h b/drivers/net/sfc/net_driver.h index 9ffa9a6..191a311 100644 --- a/drivers/net/sfc/net_driver.h +++ b/drivers/net/sfc/net_driver.h @@ -330,7 +330,6 @@ enum efx_rx_alloc_method { * @eventq_mask: Event queue pointer mask * @eventq_read_ptr: Event queue read pointer * @last_eventq_read_ptr: Last event queue read pointer value. - * @magic_count: Event queue test event count * @irq_count: Number of IRQs since last adaptive moderation decision * @irq_mod_score: IRQ moderation score * @rx_alloc_level: Watermark based heuristic counter for pushing descriptors @@ -360,7 +359,6 @@ struct efx_channel { unsigned int eventq_mask; unsigned int eventq_read_ptr; unsigned int last_eventq_read_ptr; - unsigned int magic_count; unsigned int irq_count; unsigned int irq_mod_score; diff --git a/drivers/net/sfc/nic.c b/drivers/net/sfc/nic.c index e839661..10f1cb7 100644 --- a/drivers/net/sfc/nic.c +++ b/drivers/net/sfc/nic.c @@ -84,7 +84,8 @@ static inline void efx_write_buf_tbl(struct efx_nic *efx, efx_qword_t *value, static inline efx_qword_t *efx_event(struct efx_channel *channel, unsigned int index) { - return ((efx_qword_t *) (channel->eventq.addr)) + index; + return ((efx_qword_t *) (channel->eventq.addr)) + + (index & channel->eventq_mask); } /* See if an event is present @@ -673,7 +674,8 @@ void efx_nic_eventq_read_ack(struct efx_channel *channel) efx_dword_t reg; struct efx_nic *efx = channel->efx; - EFX_POPULATE_DWORD_1(reg, FRF_AZ_EVQ_RPTR, channel->eventq_read_ptr); + EFX_POPULATE_DWORD_1(reg, FRF_AZ_EVQ_RPTR, + channel->eventq_read_ptr & channel->eventq_mask); efx_writed_table(efx, ®, efx->type->evq_rptr_tbl_base, channel->channel); } @@ -908,7 +910,7 @@ efx_handle_generated_event(struct efx_channel *channel, efx_qword_t *event) code = EFX_QWORD_FIELD(*event, FSF_AZ_DRV_GEN_EV_MAGIC); if (code == EFX_CHANNEL_MAGIC_TEST(channel)) - ++channel->magic_count; + ; /* ignore */ else if (code == EFX_CHANNEL_MAGIC_FILL(channel)) /* The queue must be empty, so we won't receive any rx * events, so efx_process_channel() won't refill the @@ -1015,8 +1017,7 @@ int efx_nic_process_eventq(struct efx_channel *channel, int budget) /* Clear this event by marking it all ones */ EFX_SET_QWORD(*p_event); - /* Increment read pointer */ - read_ptr = (read_ptr + 1) & channel->eventq_mask; + ++read_ptr; ev_code = EFX_QWORD_FIELD(event, FSF_AZ_EV_CODE); @@ -1060,6 +1061,13 @@ out: return spent; } +/* Check whether an event is present in the eventq at the current + * read pointer. Only useful for self-test. + */ +bool efx_nic_event_present(struct efx_channel *channel) +{ + return efx_event_present(efx_event(channel, channel->eventq_read_ptr)); +} /* Allocate buffer table entries for event queue */ int efx_nic_probe_eventq(struct efx_channel *channel) @@ -1165,7 +1173,7 @@ static void efx_poll_flush_events(struct efx_nic *efx) struct efx_tx_queue *tx_queue; struct efx_rx_queue *rx_queue; unsigned int read_ptr = channel->eventq_read_ptr; - unsigned int end_ptr = (read_ptr - 1) & channel->eventq_mask; + unsigned int end_ptr = read_ptr + channel->eventq_mask - 1; do { efx_qword_t *event = efx_event(channel, read_ptr); @@ -1205,7 +1213,7 @@ static void efx_poll_flush_events(struct efx_nic *efx) * it's ok to throw away every non-flush event */ EFX_SET_QWORD(*event); - read_ptr = (read_ptr + 1) & channel->eventq_mask; + ++read_ptr; } while (read_ptr != end_ptr); channel->eventq_read_ptr = read_ptr; diff --git a/drivers/net/sfc/nic.h b/drivers/net/sfc/nic.h index d9de1b6..a42db6e 100644 --- a/drivers/net/sfc/nic.h +++ b/drivers/net/sfc/nic.h @@ -184,6 +184,7 @@ extern void efx_nic_fini_eventq(struct efx_channel *channel); extern void efx_nic_remove_eventq(struct efx_channel *channel); extern int efx_nic_process_eventq(struct efx_channel *channel, int rx_quota); extern void efx_nic_eventq_read_ack(struct efx_channel *channel); +extern bool efx_nic_event_present(struct efx_channel *channel); /* MAC/PHY */ extern void falcon_drain_tx_fifo(struct efx_nic *efx); diff --git a/drivers/net/sfc/selftest.c b/drivers/net/sfc/selftest.c index 8458084..50ad3bc 100644 --- a/drivers/net/sfc/selftest.c +++ b/drivers/net/sfc/selftest.c @@ -131,8 +131,6 @@ static int efx_test_chip(struct efx_nic *efx, struct efx_self_tests *tests) static int efx_test_interrupts(struct efx_nic *efx, struct efx_self_tests *tests) { - struct efx_channel *channel; - netif_dbg(efx, drv, efx->net_dev, "testing interrupts\n"); tests->interrupt = -1; @@ -140,15 +138,6 @@ static int efx_test_interrupts(struct efx_nic *efx, efx->last_irq_cpu = -1; smp_wmb(); - /* ACK each interrupting event queue. Receiving an interrupt due to - * traffic before a test event is raised is considered a pass */ - efx_for_each_channel(channel, efx) { - if (channel->work_pending) - efx_process_channel_now(channel); - if (efx->last_irq_cpu >= 0) - goto success; - } - efx_nic_generate_interrupt(efx); /* Wait for arrival of test interrupt. */ @@ -173,13 +162,13 @@ static int efx_test_eventq_irq(struct efx_channel *channel, struct efx_self_tests *tests) { struct efx_nic *efx = channel->efx; - unsigned int magic_count, count; + unsigned int read_ptr, count; tests->eventq_dma[channel->channel] = -1; tests->eventq_int[channel->channel] = -1; tests->eventq_poll[channel->channel] = -1; - magic_count = channel->magic_count; + read_ptr = channel->eventq_read_ptr; channel->efx->last_irq_cpu = -1; smp_wmb(); @@ -190,10 +179,7 @@ static int efx_test_eventq_irq(struct efx_channel *channel, do { schedule_timeout_uninterruptible(HZ / 100); - if (channel->work_pending) - efx_process_channel_now(channel); - - if (channel->magic_count != magic_count) + if (ACCESS_ONCE(channel->eventq_read_ptr) != read_ptr) goto eventq_ok; } while (++count < 2); @@ -211,8 +197,7 @@ static int efx_test_eventq_irq(struct efx_channel *channel, } /* Check to see if event was received even if interrupt wasn't */ - efx_process_channel_now(channel); - if (channel->magic_count != magic_count) { + if (efx_nic_event_present(channel)) { netif_err(efx, drv, efx->net_dev, "channel %d event was generated, but " "failed to trigger an interrupt\n", channel->channel); -- cgit v0.10.2 From 8a226e00eeed8db843d4a580013a49ae3559bcd7 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Tue, 29 Mar 2011 09:45:57 -0700 Subject: PCI: pci-label: Fix build failure when CONFIG_NLS is set to 'm' by allmodconfig Create a kconfig option symbol for PCI_LABEL and enable it when DMI || ACPI are enabled. Signed-off-by: Randy Dunlap Signed-off-by: Jesse Barnes diff --git a/drivers/pci/Kconfig b/drivers/pci/Kconfig index c8ff646..0fa466a 100644 --- a/drivers/pci/Kconfig +++ b/drivers/pci/Kconfig @@ -88,4 +88,6 @@ config PCI_IOAPIC depends on HOTPLUG default y -select NLS if (DMI || ACPI) +config PCI_LABEL + def_bool y if (DMI || ACPI) + select NLS diff --git a/drivers/pci/Makefile b/drivers/pci/Makefile index 98d61c8..c85f744 100644 --- a/drivers/pci/Makefile +++ b/drivers/pci/Makefile @@ -56,10 +56,10 @@ obj-$(CONFIG_TILE) += setup-bus.o setup-irq.o # ACPI Related PCI FW Functions # ACPI _DSM provided firmware instance and string name # -obj-$(CONFIG_ACPI) += pci-acpi.o pci-label.o +obj-$(CONFIG_ACPI) += pci-acpi.o # SMBIOS provided firmware instance and labels -obj-$(CONFIG_DMI) += pci-label.o +obj-$(CONFIG_PCI_LABEL) += pci-label.o # Cardbus & CompactPCI use setup-bus obj-$(CONFIG_HOTPLUG) += setup-bus.o -- cgit v0.10.2 From d5627663f2088fa4be447fdcfd52bcb233448d85 Mon Sep 17 00:00:00 2001 From: Mathew McKernan Date: Tue, 12 Apr 2011 06:51:37 +0100 Subject: drm/i915/tv: Remember the detected TV type During detect() we would probe the connection bits to determine if there was a TV attached, and what video input type (Component, S-Video, Composite, etc) to use. However, we promptly discarded this vital bit of information and never propagated it to where it was used to determine the correct modes and setup the control registers. Fix it! This fixes a regression from 7b334fcb45b757ffb093696ca3de1b0c8b4a33f1. Reported-and-tested-by: Mathew McKernan Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=35977 Signed-off-by: Mathew McKernan Signed-off-by: Chris Wilson Cc: stable@kernel.org Acked-by: Paul Menzel Signed-off-by: Keith Packard diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c index 4256b8e..0be2265 100644 --- a/drivers/gpu/drm/i915/intel_tv.c +++ b/drivers/gpu/drm/i915/intel_tv.c @@ -1378,7 +1378,9 @@ intel_tv_detect(struct drm_connector *connector, bool force) if (type < 0) return connector_status_disconnected; + intel_tv->type = type; intel_tv_find_better_format(connector); + return connector_status_connected; } -- cgit v0.10.2 From 89ea42d716e1ee94f643ecdc516d90a4111ec135 Mon Sep 17 00:00:00 2001 From: Mathew McKernan Date: Tue, 12 Apr 2011 06:51:38 +0100 Subject: drm/i915/tv: Only poll for TV connections As a probe for a TV connection modifies the TV_CTL register, it causes a loss of sync and a regular glitch on the output. This is highly undesirable when using the TV, so only poll for TV connections and wait for an explicit query for detecting the disconnection event. Reported-by: Mathew McKernan Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=35977 Signed-off-by: Mathew McKernan Signed-off-by: Chris Wilson Signed-off-by: Keith Packard diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c index 0be2265..f5848cb 100644 --- a/drivers/gpu/drm/i915/intel_tv.c +++ b/drivers/gpu/drm/i915/intel_tv.c @@ -1672,8 +1672,7 @@ intel_tv_init(struct drm_device *dev) * * More recent chipsets favour HDMI rather than integrated S-Video. */ - connector->polled = - DRM_CONNECTOR_POLL_CONNECT | DRM_CONNECTOR_POLL_DISCONNECT; + connector->polled = DRM_CONNECTOR_POLL_CONNECT; drm_connector_init(dev, connector, &intel_tv_connector_funcs, DRM_MODE_CONNECTOR_SVIDEO); -- cgit v0.10.2 From ccacfec6cc22157971bd970486fd3fd4972b8e2e Mon Sep 17 00:00:00 2001 From: Sitsofe Wheeler Date: Tue, 12 Apr 2011 06:51:39 +0100 Subject: drm/i915/tv: Fix modeset flickering introduced in 7f58aabc3 The tidy ups in 7f58aabc369014fda3a4a33604ba0a1b63b941ac ("drm/i915: cleanup per-pipe reg usage") changed intel_crtc->plane to intel_crtc->pipe in intel_tv_mode_set(). This caused the screen to quickly turn off before returning whenever modesetting/mode probing took place on my 915GM EeePC 900 creating a flickering effect. This patch changes intel_crtc->pipe back to intel_crtc->plane which solves the problem for me. References: https://bugs.freedesktop.org/show_bug.cgi?id=35903 Signed-off-by: Sitsofe Wheeler Cc: Jesse Barnes Signed-off-by: Chris Wilson Humbly-acked-by: Jesse Barnes Signed-off-by: Keith Packard diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c index f5848cb..6b22c1d 100644 --- a/drivers/gpu/drm/i915/intel_tv.c +++ b/drivers/gpu/drm/i915/intel_tv.c @@ -1151,10 +1151,10 @@ intel_tv_mode_set(struct drm_encoder *encoder, struct drm_display_mode *mode, (video_levels->blank << TV_BLANK_LEVEL_SHIFT))); { int pipeconf_reg = PIPECONF(pipe); - int dspcntr_reg = DSPCNTR(pipe); + int dspcntr_reg = DSPCNTR(intel_crtc->plane); int pipeconf = I915_READ(pipeconf_reg); int dspcntr = I915_READ(dspcntr_reg); - int dspbase_reg = DSPADDR(pipe); + int dspbase_reg = DSPADDR(intel_crtc->plane); int xpos = 0x0, ypos = 0x0; unsigned int xsize, ysize; /* Pipe must be off here */ -- cgit v0.10.2 From f6e5b1603b8bb7131b6778d0d4e2e5dda120a379 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 12 Apr 2011 18:06:51 +0100 Subject: drm/i915: Sanitize the output registers after resume Similar to booting, we need to inspect the state left by the BIOS and remove any conflicting bits before we take over. The example reported by Seth Forshee is very similar to the bug we encountered with the state left by grub2, that the crtc pipe<->planning mapping was reversed from our expectations and so we failed to turn off the outputs when booting or, in this case, resuming. This may be in fact the same bug, but triggered at resume time. This patch rearranges the code we already have to clear up the conflicting state upon init and calls it from reset (which is called after we have lost control of the hardware, i.e. along both the boot and resume paths) instead. Reported-and-tested-by: Seth Forshee Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=35796 Signed-off-by: Chris Wilson Cc: stable@kernel.org Reviewed-by: Keith Packard Signed-off-by: Keith Packard diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 432fc04..4fc21e0 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -6215,36 +6215,6 @@ cleanup_work: return ret; } -static void intel_crtc_reset(struct drm_crtc *crtc) -{ - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); - - /* Reset flags back to the 'unknown' status so that they - * will be correctly set on the initial modeset. - */ - intel_crtc->dpms_mode = -1; -} - -static struct drm_crtc_helper_funcs intel_helper_funcs = { - .dpms = intel_crtc_dpms, - .mode_fixup = intel_crtc_mode_fixup, - .mode_set = intel_crtc_mode_set, - .mode_set_base = intel_pipe_set_base, - .mode_set_base_atomic = intel_pipe_set_base_atomic, - .load_lut = intel_crtc_load_lut, - .disable = intel_crtc_disable, -}; - -static const struct drm_crtc_funcs intel_crtc_funcs = { - .reset = intel_crtc_reset, - .cursor_set = intel_crtc_cursor_set, - .cursor_move = intel_crtc_cursor_move, - .gamma_set = intel_crtc_gamma_set, - .set_config = drm_crtc_helper_set_config, - .destroy = intel_crtc_destroy, - .page_flip = intel_crtc_page_flip, -}; - static void intel_sanitize_modesetting(struct drm_device *dev, int pipe, int plane) { @@ -6281,6 +6251,42 @@ static void intel_sanitize_modesetting(struct drm_device *dev, intel_disable_pipe(dev_priv, pipe); } +static void intel_crtc_reset(struct drm_crtc *crtc) +{ + struct drm_device *dev = crtc->dev; + struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + + /* Reset flags back to the 'unknown' status so that they + * will be correctly set on the initial modeset. + */ + intel_crtc->dpms_mode = -1; + + /* We need to fix up any BIOS configuration that conflicts with + * our expectations. + */ + intel_sanitize_modesetting(dev, intel_crtc->pipe, intel_crtc->plane); +} + +static struct drm_crtc_helper_funcs intel_helper_funcs = { + .dpms = intel_crtc_dpms, + .mode_fixup = intel_crtc_mode_fixup, + .mode_set = intel_crtc_mode_set, + .mode_set_base = intel_pipe_set_base, + .mode_set_base_atomic = intel_pipe_set_base_atomic, + .load_lut = intel_crtc_load_lut, + .disable = intel_crtc_disable, +}; + +static const struct drm_crtc_funcs intel_crtc_funcs = { + .reset = intel_crtc_reset, + .cursor_set = intel_crtc_cursor_set, + .cursor_move = intel_crtc_cursor_move, + .gamma_set = intel_crtc_gamma_set, + .set_config = drm_crtc_helper_set_config, + .destroy = intel_crtc_destroy, + .page_flip = intel_crtc_page_flip, +}; + static void intel_crtc_init(struct drm_device *dev, int pipe) { drm_i915_private_t *dev_priv = dev->dev_private; @@ -6330,8 +6336,6 @@ static void intel_crtc_init(struct drm_device *dev, int pipe) setup_timer(&intel_crtc->idle_timer, intel_crtc_idle_timer, (unsigned long)intel_crtc); - - intel_sanitize_modesetting(dev, intel_crtc->pipe, intel_crtc->plane); } int intel_get_pipe_from_crtc_id(struct drm_device *dev, void *data, -- cgit v0.10.2 From be85bccaa5aa5a11dcaf85f9e945ffefd253f631 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 12 Apr 2011 13:35:56 -0700 Subject: Revert "vfs: Export file system uuid via /proc//mountinfo" This reverts commit 93f1c20bc8cdb757be50566eff88d65c3b26881f. It turns out that libmount misparses it because it adds a '-' character in the uuid string, which libmount then incorrectly confuses with the separator string (" - ") at the end of all the optional arguments. Upstream libmount (in the util-linux tree) has been fixed, but until that fix actually percolates up to users, we'd better not expose this change in the kernel. Let's revisit this later (possibly by exposing the UUID without any '-' characters in it, avoiding the user-space bug). Reported-by: Dave Jones Cc: Aneesh Kumar K.V Cc: Al Viro Cc: Karel Zak Cc: Ram Pai Cc: Miklos Szeredi Cc: Eric Sandeen Signed-off-by: Linus Torvalds diff --git a/fs/namespace.c b/fs/namespace.c index 7dba2ed..d99bcf5 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1030,18 +1030,6 @@ const struct seq_operations mounts_op = { .show = show_vfsmnt }; -static int uuid_is_nil(u8 *uuid) -{ - int i; - u8 *cp = (u8 *)uuid; - - for (i = 0; i < 16; i++) { - if (*cp++) - return 0; - } - return 1; -} - static int show_mountinfo(struct seq_file *m, void *v) { struct proc_mounts *p = m->private; @@ -1085,10 +1073,6 @@ static int show_mountinfo(struct seq_file *m, void *v) if (IS_MNT_UNBINDABLE(mnt)) seq_puts(m, " unbindable"); - if (!uuid_is_nil(mnt->mnt_sb->s_uuid)) - /* print the uuid */ - seq_printf(m, " uuid:%pU", mnt->mnt_sb->s_uuid); - /* Filesystem specific data */ seq_puts(m, " - "); show_type(m, sb); diff --git a/include/linux/fs.h b/include/linux/fs.h index dbd860a..5b14843 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1403,7 +1403,6 @@ struct super_block { wait_queue_head_t s_wait_unfrozen; char s_id[32]; /* Informational name */ - u8 s_uuid[16]; /* UUID */ void *s_fs_info; /* Filesystem private info */ fmode_t s_mode; -- cgit v0.10.2 From f8e9881c2aef1e982e5abc25c046820cd0b7cf64 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Tue, 12 Apr 2011 13:39:14 -0700 Subject: bridge: reset IPCB in br_parse_ip_options MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit 462fb2af9788a82 (bridge : Sanitize skb before it enters the IP stack), missed one IPCB init before calling ip_options_compile() Thanks to Scot Doyle for his tests and bug reports. Reported-by: Scot Doyle Signed-off-by: Eric Dumazet Cc: Hiroaki SHIMODA Acked-by: Bandan Das Acked-by: Stephen Hemminger Cc: Jan Lübbe Signed-off-by: David S. Miller diff --git a/net/bridge/br_netfilter.c b/net/bridge/br_netfilter.c index 008ff6c..f3bc322 100644 --- a/net/bridge/br_netfilter.c +++ b/net/bridge/br_netfilter.c @@ -249,11 +249,9 @@ static int br_parse_ip_options(struct sk_buff *skb) goto drop; } - /* Zero out the CB buffer if no options present */ - if (iph->ihl == 5) { - memset(IPCB(skb), 0, sizeof(struct inet_skb_parm)); + memset(IPCB(skb), 0, sizeof(struct inet_skb_parm)); + if (iph->ihl == 5) return 0; - } opt->optlen = iph->ihl*4 - sizeof(struct iphdr); if (ip_options_compile(dev_net(dev), opt, skb)) -- cgit v0.10.2 From 5882da02e9d9089b7e8c739f3e774aaeeff8b7ba Mon Sep 17 00:00:00 2001 From: Felix Fietkau Date: Fri, 8 Apr 2011 20:13:18 +0200 Subject: ath9k_hw: fix stopping rx DMA during resets During PHY errors, the MAC can sometimes fail to enter an idle state on older hardware (before AR9380) after an rx stop has been requested. This typically shows up in the kernel log with messages like these: ath: Could not stop RX, we could be confusing the DMA engine when we start RX up ------------[ cut here ]------------ WARNING: at drivers/net/wireless/ath/ath9k/recv.c:504 ath_stoprecv+0xcc/0xf0 [ath9k]() Call Trace: [<8023f0e8>] dump_stack+0x8/0x34 [<80075050>] warn_slowpath_common+0x78/0xa4 [<80075094>] warn_slowpath_null+0x18/0x24 [<80d66d60>] ath_stoprecv+0xcc/0xf0 [ath9k] [<80d642cc>] ath_set_channel+0xbc/0x270 [ath9k] [<80d65254>] ath_radio_disable+0x4a4/0x7fc [ath9k] When this happens, the state that the MAC enters is easy to identify and does not result in bogus DMA traffic, however to ensure a working state after a channel change, the hardware should still be reset. This patch adds detection for this specific MAC state, after which the above warnings completely disappear in my tests. Signed-off-by: Felix Fietkau Cc: stable@kernel.org Cc: Kyungwan Nam Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/ath/ath9k/hw.c b/drivers/net/wireless/ath/ath9k/hw.c index 1ec9bcd..c95bc5c 100644 --- a/drivers/net/wireless/ath/ath9k/hw.c +++ b/drivers/net/wireless/ath/ath9k/hw.c @@ -1254,15 +1254,6 @@ int ath9k_hw_reset(struct ath_hw *ah, struct ath9k_channel *chan, ah->txchainmask = common->tx_chainmask; ah->rxchainmask = common->rx_chainmask; - if ((common->bus_ops->ath_bus_type != ATH_USB) && !ah->chip_fullsleep) { - ath9k_hw_abortpcurecv(ah); - if (!ath9k_hw_stopdmarecv(ah)) { - ath_dbg(common, ATH_DBG_XMIT, - "Failed to stop receive dma\n"); - bChannelChange = false; - } - } - if (!ath9k_hw_setpower(ah, ATH9K_PM_AWAKE)) return -EIO; diff --git a/drivers/net/wireless/ath/ath9k/mac.c b/drivers/net/wireless/ath/ath9k/mac.c index 562257a..edc1cbb 100644 --- a/drivers/net/wireless/ath/ath9k/mac.c +++ b/drivers/net/wireless/ath/ath9k/mac.c @@ -751,28 +751,47 @@ void ath9k_hw_abortpcurecv(struct ath_hw *ah) } EXPORT_SYMBOL(ath9k_hw_abortpcurecv); -bool ath9k_hw_stopdmarecv(struct ath_hw *ah) +bool ath9k_hw_stopdmarecv(struct ath_hw *ah, bool *reset) { #define AH_RX_STOP_DMA_TIMEOUT 10000 /* usec */ #define AH_RX_TIME_QUANTUM 100 /* usec */ struct ath_common *common = ath9k_hw_common(ah); + u32 mac_status, last_mac_status = 0; int i; + /* Enable access to the DMA observation bus */ + REG_WRITE(ah, AR_MACMISC, + ((AR_MACMISC_DMA_OBS_LINE_8 << AR_MACMISC_DMA_OBS_S) | + (AR_MACMISC_MISC_OBS_BUS_1 << + AR_MACMISC_MISC_OBS_BUS_MSB_S))); + REG_WRITE(ah, AR_CR, AR_CR_RXD); /* Wait for rx enable bit to go low */ for (i = AH_RX_STOP_DMA_TIMEOUT / AH_TIME_QUANTUM; i != 0; i--) { if ((REG_READ(ah, AR_CR) & AR_CR_RXE) == 0) break; + + if (!AR_SREV_9300_20_OR_LATER(ah)) { + mac_status = REG_READ(ah, AR_DMADBG_7) & 0x7f0; + if (mac_status == 0x1c0 && mac_status == last_mac_status) { + *reset = true; + break; + } + + last_mac_status = mac_status; + } + udelay(AH_TIME_QUANTUM); } if (i == 0) { ath_err(common, - "DMA failed to stop in %d ms AR_CR=0x%08x AR_DIAG_SW=0x%08x\n", + "DMA failed to stop in %d ms AR_CR=0x%08x AR_DIAG_SW=0x%08x DMADBG_7=0x%08x\n", AH_RX_STOP_DMA_TIMEOUT / 1000, REG_READ(ah, AR_CR), - REG_READ(ah, AR_DIAG_SW)); + REG_READ(ah, AR_DIAG_SW), + REG_READ(ah, AR_DMADBG_7)); return false; } else { return true; diff --git a/drivers/net/wireless/ath/ath9k/mac.h b/drivers/net/wireless/ath/ath9k/mac.h index b2b2ff8..c2a5938 100644 --- a/drivers/net/wireless/ath/ath9k/mac.h +++ b/drivers/net/wireless/ath/ath9k/mac.h @@ -695,7 +695,7 @@ bool ath9k_hw_setrxabort(struct ath_hw *ah, bool set); void ath9k_hw_putrxbuf(struct ath_hw *ah, u32 rxdp); void ath9k_hw_startpcureceive(struct ath_hw *ah, bool is_scanning); void ath9k_hw_abortpcurecv(struct ath_hw *ah); -bool ath9k_hw_stopdmarecv(struct ath_hw *ah); +bool ath9k_hw_stopdmarecv(struct ath_hw *ah, bool *reset); int ath9k_hw_beaconq_setup(struct ath_hw *ah); /* Interrupt Handling */ diff --git a/drivers/net/wireless/ath/ath9k/recv.c b/drivers/net/wireless/ath/ath9k/recv.c index a9c3f46..dcd19bc 100644 --- a/drivers/net/wireless/ath/ath9k/recv.c +++ b/drivers/net/wireless/ath/ath9k/recv.c @@ -486,12 +486,12 @@ start_recv: bool ath_stoprecv(struct ath_softc *sc) { struct ath_hw *ah = sc->sc_ah; - bool stopped; + bool stopped, reset = false; spin_lock_bh(&sc->rx.rxbuflock); ath9k_hw_abortpcurecv(ah); ath9k_hw_setrxfilter(ah, 0); - stopped = ath9k_hw_stopdmarecv(ah); + stopped = ath9k_hw_stopdmarecv(ah, &reset); if (sc->sc_ah->caps.hw_caps & ATH9K_HW_CAP_EDMA) ath_edma_stop_recv(sc); @@ -506,7 +506,7 @@ bool ath_stoprecv(struct ath_softc *sc) "confusing the DMA engine when we start RX up\n"); ATH_DBG_WARN_ON_ONCE(!stopped); } - return stopped; + return stopped || reset; } void ath_flushrecv(struct ath_softc *sc) -- cgit v0.10.2 From 50f6871218802be528961e24487f00c203a5c3c7 Mon Sep 17 00:00:00 2001 From: Sujith Manoharan Date: Mon, 11 Apr 2011 22:56:55 +0530 Subject: ath9k_htc: Fix ethtool reporting Pass the correct module name and device interface so that ethtool can display the proper values. The firmware version will be fixed later on when the FW can actually report a version. :) Reported-by: Richard Farina Signed-off-by: Sujith Manoharan Tested-by: Richard Farina Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/ath/ath9k/hif_usb.c b/drivers/net/wireless/ath/ath9k/hif_usb.c index f1b8af6..2d10239 100644 --- a/drivers/net/wireless/ath/ath9k/hif_usb.c +++ b/drivers/net/wireless/ath/ath9k/hif_usb.c @@ -1040,7 +1040,7 @@ static int ath9k_hif_usb_probe(struct usb_interface *interface, } ret = ath9k_htc_hw_init(hif_dev->htc_handle, - &hif_dev->udev->dev, hif_dev->device_id, + &interface->dev, hif_dev->device_id, hif_dev->udev->product, id->driver_info); if (ret) { ret = -EINVAL; @@ -1158,7 +1158,7 @@ fail_resume: #endif static struct usb_driver ath9k_hif_usb_driver = { - .name = "ath9k_hif_usb", + .name = KBUILD_MODNAME, .probe = ath9k_hif_usb_probe, .disconnect = ath9k_hif_usb_disconnect, #ifdef CONFIG_PM -- cgit v0.10.2 From 66944e1c5797562cebe2d1857d46dff60bf9a69e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 11 Apr 2011 22:39:40 +0000 Subject: inetpeer: reduce stack usage On 64bit arches, we use 752 bytes of stack when cleanup_once() is called from inet_getpeer(). Lets share the avl stack to save ~376 bytes. Before patch : # objdump -d net/ipv4/inetpeer.o | scripts/checkstack.pl 0x000006c3 unlink_from_pool [inetpeer.o]: 376 0x00000721 unlink_from_pool [inetpeer.o]: 376 0x00000cb1 inet_getpeer [inetpeer.o]: 376 0x00000e6d inet_getpeer [inetpeer.o]: 376 0x0004 inet_initpeers [inetpeer.o]: 112 # size net/ipv4/inetpeer.o text data bss dec hex filename 5320 432 21 5773 168d net/ipv4/inetpeer.o After patch : objdump -d net/ipv4/inetpeer.o | scripts/checkstack.pl 0x00000c11 inet_getpeer [inetpeer.o]: 376 0x00000dcd inet_getpeer [inetpeer.o]: 376 0x00000ab9 peer_check_expire [inetpeer.o]: 328 0x00000b7f peer_check_expire [inetpeer.o]: 328 0x0004 inet_initpeers [inetpeer.o]: 112 # size net/ipv4/inetpeer.o text data bss dec hex filename 5163 432 21 5616 15f0 net/ipv4/inetpeer.o Signed-off-by: Eric Dumazet Cc: Scot Doyle Cc: Stephen Hemminger Cc: Hiroaki SHIMODA Reviewed-by: Hiroaki SHIMODA Signed-off-by: David S. Miller diff --git a/net/ipv4/inetpeer.c b/net/ipv4/inetpeer.c index dd1b20e..9df4e63 100644 --- a/net/ipv4/inetpeer.c +++ b/net/ipv4/inetpeer.c @@ -354,7 +354,8 @@ static void inetpeer_free_rcu(struct rcu_head *head) } /* May be called with local BH enabled. */ -static void unlink_from_pool(struct inet_peer *p, struct inet_peer_base *base) +static void unlink_from_pool(struct inet_peer *p, struct inet_peer_base *base, + struct inet_peer __rcu **stack[PEER_MAXDEPTH]) { int do_free; @@ -368,7 +369,6 @@ static void unlink_from_pool(struct inet_peer *p, struct inet_peer_base *base) * We use refcnt=-1 to alert lockless readers this entry is deleted. */ if (atomic_cmpxchg(&p->refcnt, 1, -1) == 1) { - struct inet_peer __rcu **stack[PEER_MAXDEPTH]; struct inet_peer __rcu ***stackptr, ***delp; if (lookup(&p->daddr, stack, base) != p) BUG(); @@ -422,7 +422,7 @@ static struct inet_peer_base *peer_to_base(struct inet_peer *p) } /* May be called with local BH enabled. */ -static int cleanup_once(unsigned long ttl) +static int cleanup_once(unsigned long ttl, struct inet_peer __rcu **stack[PEER_MAXDEPTH]) { struct inet_peer *p = NULL; @@ -454,7 +454,7 @@ static int cleanup_once(unsigned long ttl) * happen because of entry limits in route cache. */ return -1; - unlink_from_pool(p, peer_to_base(p)); + unlink_from_pool(p, peer_to_base(p), stack); return 0; } @@ -524,7 +524,7 @@ struct inet_peer *inet_getpeer(struct inetpeer_addr *daddr, int create) if (base->total >= inet_peer_threshold) /* Remove one less-recently-used entry. */ - cleanup_once(0); + cleanup_once(0, stack); return p; } @@ -540,6 +540,7 @@ static void peer_check_expire(unsigned long dummy) { unsigned long now = jiffies; int ttl, total; + struct inet_peer __rcu **stack[PEER_MAXDEPTH]; total = compute_total(); if (total >= inet_peer_threshold) @@ -548,7 +549,7 @@ static void peer_check_expire(unsigned long dummy) ttl = inet_peer_maxttl - (inet_peer_maxttl - inet_peer_minttl) / HZ * total / inet_peer_threshold * HZ; - while (!cleanup_once(ttl)) { + while (!cleanup_once(ttl, stack)) { if (jiffies != now) break; } -- cgit v0.10.2 From 192910a6cca5e50e5bd6cbd1da0e7376c7adfe62 Mon Sep 17 00:00:00 2001 From: Joakim Tjernlund Date: Tue, 12 Apr 2011 13:59:33 -0700 Subject: net: Do not wrap sysctl igmp_max_memberships in IP_MULTICAST controlling igmp_max_membership is useful even when IP_MULTICAST is off. Quagga(an OSPF deamon) uses multicast addresses for all interfaces using a single socket and hits igmp_max_membership limit when there are 20 interfaces or more. Always export sysctl igmp_max_memberships in proc, just like igmp_max_msf Signed-off-by: Joakim Tjernlund Signed-off-by: David S. Miller diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c index 1a45665..321e6e8 100644 --- a/net/ipv4/sysctl_net_ipv4.c +++ b/net/ipv4/sysctl_net_ipv4.c @@ -311,7 +311,6 @@ static struct ctl_table ipv4_table[] = { .mode = 0644, .proc_handler = proc_do_large_bitmap, }, -#ifdef CONFIG_IP_MULTICAST { .procname = "igmp_max_memberships", .data = &sysctl_igmp_max_memberships, @@ -319,8 +318,6 @@ static struct ctl_table ipv4_table[] = { .mode = 0644, .proc_handler = proc_dointvec }, - -#endif { .procname = "igmp_max_msf", .data = &sysctl_igmp_max_msf, -- cgit v0.10.2 From 95042f9eb78a8d9a17455e2ef263f2f310ecef15 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 12 Apr 2011 14:15:51 -0700 Subject: vm: fix mlock() on stack guard page MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit 53a7706d5ed8 ("mlock: do not hold mmap_sem for extended periods of time") changed mlock() to care about the exact number of pages that __get_user_pages() had brought it. Before, it would only care about errors. And that doesn't work, because we also handled one page specially in __mlock_vma_pages_range(), namely the stack guard page. So when that case was handled, the number of pages that the function returned was off by one. In particular, it could be zero, and then the caller would end up not making any progress at all. Rather than try to fix up that off-by-one error for the mlock case specially, this just moves the logic to handle the stack guard page into__get_user_pages() itself, thus making all the counts come out right automatically. Reported-by: Robert Święcki Cc: Hugh Dickins Cc: Oleg Nesterov Cc: stable@kernel.org Signed-off-by: Linus Torvalds diff --git a/mm/memory.c b/mm/memory.c index 9da8cab..b623a24 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -1410,6 +1410,13 @@ no_page_table: return page; } +static inline int stack_guard_page(struct vm_area_struct *vma, unsigned long addr) +{ + return (vma->vm_flags & VM_GROWSDOWN) && + (vma->vm_start == addr) && + !vma_stack_continue(vma->vm_prev, addr); +} + /** * __get_user_pages() - pin user pages in memory * @tsk: task_struct of target task @@ -1488,7 +1495,6 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, vma = find_extend_vma(mm, start); if (!vma && in_gate_area(mm, start)) { unsigned long pg = start & PAGE_MASK; - struct vm_area_struct *gate_vma = get_gate_vma(mm); pgd_t *pgd; pud_t *pud; pmd_t *pmd; @@ -1513,10 +1519,11 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, pte_unmap(pte); return i ? : -EFAULT; } + vma = get_gate_vma(mm); if (pages) { struct page *page; - page = vm_normal_page(gate_vma, start, *pte); + page = vm_normal_page(vma, start, *pte); if (!page) { if (!(gup_flags & FOLL_DUMP) && is_zero_pfn(pte_pfn(*pte))) @@ -1530,12 +1537,7 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, get_page(page); } pte_unmap(pte); - if (vmas) - vmas[i] = gate_vma; - i++; - start += PAGE_SIZE; - nr_pages--; - continue; + goto next_page; } if (!vma || @@ -1549,6 +1551,13 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, continue; } + /* + * If we don't actually want the page itself, + * and it's the stack guard page, just skip it. + */ + if (!pages && stack_guard_page(vma, start)) + goto next_page; + do { struct page *page; unsigned int foll_flags = gup_flags; @@ -1631,6 +1640,7 @@ int __get_user_pages(struct task_struct *tsk, struct mm_struct *mm, flush_anon_page(vma, page, start); flush_dcache_page(page); } +next_page: if (vmas) vmas[i] = vma; i++; diff --git a/mm/mlock.c b/mm/mlock.c index 2689a08c..6b55e3e 100644 --- a/mm/mlock.c +++ b/mm/mlock.c @@ -135,13 +135,6 @@ void munlock_vma_page(struct page *page) } } -static inline int stack_guard_page(struct vm_area_struct *vma, unsigned long addr) -{ - return (vma->vm_flags & VM_GROWSDOWN) && - (vma->vm_start == addr) && - !vma_stack_continue(vma->vm_prev, addr); -} - /** * __mlock_vma_pages_range() - mlock a range of pages in the vma. * @vma: target vma @@ -188,12 +181,6 @@ static long __mlock_vma_pages_range(struct vm_area_struct *vma, if (vma->vm_flags & VM_LOCKED) gup_flags |= FOLL_MLOCK; - /* We don't try to access the guard page of a stack vma */ - if (stack_guard_page(vma, start)) { - addr += PAGE_SIZE; - nr_pages--; - } - return __get_user_pages(current, mm, addr, nr_pages, gup_flags, NULL, NULL, nonblocking); } -- cgit v0.10.2 From 0d88f6e804c824454b5ed0d3034ed3dcf7467a87 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Tue, 12 Apr 2011 19:18:08 +1000 Subject: nfs: don't call __mark_inode_dirty while holding i_lock nfs_scan_commit() is called with the inode->i_lock held, but it then calls __mark_inode_dirty() while still holding the lock. This causes a deadlock. Push the inode->i_lock into nfs_scan_commit() so it can protect only the parts of the code it needs to and can be dropped before the call to __mark_inode_dirty() to avoid the deadlock. Signed-off-by: Dave Chinner Tested-by: Will Simoneau Signed-off-by: Linus Torvalds diff --git a/fs/nfs/write.c b/fs/nfs/write.c index af0c627..e4cbc11 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -542,11 +542,15 @@ nfs_scan_commit(struct inode *inode, struct list_head *dst, pgoff_t idx_start, u if (!nfs_need_commit(nfsi)) return 0; + spin_lock(&inode->i_lock); ret = nfs_scan_list(nfsi, dst, idx_start, npages, NFS_PAGE_TAG_COMMIT); if (ret > 0) nfsi->ncommit -= ret; + spin_unlock(&inode->i_lock); + if (nfs_need_commit(NFS_I(inode))) __mark_inode_dirty(inode, I_DIRTY_DATASYNC); + return ret; } #else @@ -1483,9 +1487,7 @@ int nfs_commit_inode(struct inode *inode, int how) res = nfs_commit_set_lock(NFS_I(inode), may_wait); if (res <= 0) goto out_mark_dirty; - spin_lock(&inode->i_lock); res = nfs_scan_commit(inode, &head, 0, 0); - spin_unlock(&inode->i_lock); if (res) { int error; -- cgit v0.10.2 From 0e08785845093ef4ed220463a739bc8d0db95de7 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Tue, 12 Apr 2011 05:39:51 +0000 Subject: connector: fix skb double free in cn_rx_skb() When a skb is delivered to a registered callback, cn_call_callback() incorrectly returns -ENODEV after freeing the skb, causing cn_rx_skb() to free the skb a second time. Reported-by: Eric B Munson Signed-off-by: Patrick McHardy Tested-by: Eric B Munson Signed-off-by: David S. Miller diff --git a/drivers/connector/connector.c b/drivers/connector/connector.c index d770058..219d88a 100644 --- a/drivers/connector/connector.c +++ b/drivers/connector/connector.c @@ -142,6 +142,7 @@ static int cn_call_callback(struct sk_buff *skb) cbq->callback(msg, nsp); kfree_skb(skb); cn_queue_release_callback(cbq); + err = 0; } return err; -- cgit v0.10.2 From efa2ad8918afc2f5e4ea533110758b26991f1937 Mon Sep 17 00:00:00 2001 From: Otavio Salvador Date: Tue, 12 Apr 2011 05:30:40 +0000 Subject: net/sis900: store MAC into perm_addr for SiS 900, 630E, 635 and 96x variants Signed-off-by: Otavio Salvador Signed-off-by: David S. Miller diff --git a/drivers/net/sis900.c b/drivers/net/sis900.c index cb317cd..484f795 100644 --- a/drivers/net/sis900.c +++ b/drivers/net/sis900.c @@ -240,7 +240,8 @@ static const struct ethtool_ops sis900_ethtool_ops; * @net_dev: the net device to get address for * * Older SiS900 and friends, use EEPROM to store MAC address. - * MAC address is read from read_eeprom() into @net_dev->dev_addr. + * MAC address is read from read_eeprom() into @net_dev->dev_addr and + * @net_dev->perm_addr. */ static int __devinit sis900_get_mac_addr(struct pci_dev * pci_dev, struct net_device *net_dev) @@ -261,6 +262,9 @@ static int __devinit sis900_get_mac_addr(struct pci_dev * pci_dev, struct net_de for (i = 0; i < 3; i++) ((u16 *)(net_dev->dev_addr))[i] = read_eeprom(ioaddr, i+EEPROMMACAddr); + /* Store MAC Address in perm_addr */ + memcpy(net_dev->perm_addr, net_dev->dev_addr, ETH_ALEN); + return 1; } @@ -271,7 +275,8 @@ static int __devinit sis900_get_mac_addr(struct pci_dev * pci_dev, struct net_de * * SiS630E model, use APC CMOS RAM to store MAC address. * APC CMOS RAM is accessed through ISA bridge. - * MAC address is read into @net_dev->dev_addr. + * MAC address is read into @net_dev->dev_addr and + * @net_dev->perm_addr. */ static int __devinit sis630e_get_mac_addr(struct pci_dev * pci_dev, @@ -296,6 +301,10 @@ static int __devinit sis630e_get_mac_addr(struct pci_dev * pci_dev, outb(0x09 + i, 0x70); ((u8 *)(net_dev->dev_addr))[i] = inb(0x71); } + + /* Store MAC Address in perm_addr */ + memcpy(net_dev->perm_addr, net_dev->dev_addr, ETH_ALEN); + pci_write_config_byte(isa_bridge, 0x48, reg & ~0x40); pci_dev_put(isa_bridge); @@ -310,7 +319,7 @@ static int __devinit sis630e_get_mac_addr(struct pci_dev * pci_dev, * * SiS635 model, set MAC Reload Bit to load Mac address from APC * to rfdr. rfdr is accessed through rfcr. MAC address is read into - * @net_dev->dev_addr. + * @net_dev->dev_addr and @net_dev->perm_addr. */ static int __devinit sis635_get_mac_addr(struct pci_dev * pci_dev, @@ -334,6 +343,9 @@ static int __devinit sis635_get_mac_addr(struct pci_dev * pci_dev, *( ((u16 *)net_dev->dev_addr) + i) = inw(ioaddr + rfdr); } + /* Store MAC Address in perm_addr */ + memcpy(net_dev->perm_addr, net_dev->dev_addr, ETH_ALEN); + /* enable packet filtering */ outl(rfcrSave | RFEN, rfcr + ioaddr); @@ -353,7 +365,7 @@ static int __devinit sis635_get_mac_addr(struct pci_dev * pci_dev, * EEDONE signal to refuse EEPROM access by LAN. * The EEPROM map of SiS962 or SiS963 is different to SiS900. * The signature field in SiS962 or SiS963 spec is meaningless. - * MAC address is read into @net_dev->dev_addr. + * MAC address is read into @net_dev->dev_addr and @net_dev->perm_addr. */ static int __devinit sis96x_get_mac_addr(struct pci_dev * pci_dev, @@ -372,6 +384,9 @@ static int __devinit sis96x_get_mac_addr(struct pci_dev * pci_dev, for (i = 0; i < 3; i++) ((u16 *)(net_dev->dev_addr))[i] = read_eeprom(ioaddr, i+EEPROMMACAddr); + /* Store MAC Address in perm_addr */ + memcpy(net_dev->perm_addr, net_dev->dev_addr, ETH_ALEN); + outl(EEDONE, ee_addr); return 1; } else { -- cgit v0.10.2 From 3d894a9c1959acaa6981283619b465f75f7e535e Mon Sep 17 00:00:00 2001 From: Otavio Salvador Date: Tue, 12 Apr 2011 05:30:41 +0000 Subject: net/natsami: store MAC into perm_addr Signed-off-by: Otavio Salvador Signed-off-by: David S. Miller diff --git a/drivers/net/natsemi.c b/drivers/net/natsemi.c index aa2813e..1074231 100644 --- a/drivers/net/natsemi.c +++ b/drivers/net/natsemi.c @@ -860,6 +860,9 @@ static int __devinit natsemi_probe1 (struct pci_dev *pdev, prev_eedata = eedata; } + /* Store MAC Address in perm_addr */ + memcpy(dev->perm_addr, dev->dev_addr, ETH_ALEN); + dev->base_addr = (unsigned long __force) ioaddr; dev->irq = irq; -- cgit v0.10.2 From 0bba01695b74fdd2f9286243bb39f88544d81401 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 12 Apr 2011 15:21:04 -0700 Subject: vfs: Re-introduce s_uuid in the superblock Gaah. When commit be85bccaa5aa reverted the export of file system uuid via /proc//mountinfo, it also unintentionally removed the s_uuid field in struct super_block. I didn't mean to do that, since filesystems have been taught to fill it in (and we want to keep it for future re-introduction in the mountinfo file). Stupid of me. This adds it back in. Signed-off-by: Linus Torvalds diff --git a/include/linux/fs.h b/include/linux/fs.h index 5b14843..dbd860a 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1403,6 +1403,7 @@ struct super_block { wait_queue_head_t s_wait_unfrozen; char s_id[32]; /* Informational name */ + u8 s_uuid[16]; /* UUID */ void *s_fs_info; /* Filesystem private info */ fmode_t s_mode; -- cgit v0.10.2 From 020318d0d2af51e0fd59ba654ede9b2171558720 Mon Sep 17 00:00:00 2001 From: Dave Jones Date: Tue, 12 Apr 2011 15:29:54 -0700 Subject: irda: fix locking unbalance in irda_sendmsg 5b40964eadea40509d353318d2c82e8b7bf5e8a5 ("irda: Remove BKL instances from af_irda.c") introduced a path where we have a locking unbalance. If we pass invalid flags, we unlock a socket we never locked, resulting in this... ===================================== [ BUG: bad unlock balance detected! ] ------------------------------------- trinity/20101 is trying to release lock (sk_lock-AF_IRDA) at: [] irda_sendmsg+0x207/0x21d [irda] but there are no more locks to release! other info that might help us debug this: no locks held by trinity/20101. stack backtrace: Pid: 20101, comm: trinity Not tainted 2.6.39-rc3+ #3 Call Trace: [] ? irda_sendmsg+0x207/0x21d [irda] [] print_unlock_inbalance_bug+0xc7/0xd2 [] ? irda_sendmsg+0x207/0x21d [irda] [] lock_release+0xcf/0x18e [] release_sock+0x2d/0x155 [] irda_sendmsg+0x207/0x21d [irda] [] __sock_sendmsg+0x69/0x75 [] sock_sendmsg+0xa1/0xb6 [] ? might_fault+0x5c/0xac [] ? lock_release+0x181/0x18e [] ? might_fault+0xa5/0xac [] ? might_fault+0x5c/0xac [] ? fcheck_files+0xb9/0xf0 [] ? copy_from_user+0x2f/0x31 [] ? verify_iovec+0x52/0xa6 [] sys_sendmsg+0x23a/0x2b8 [] ? lock_release+0x181/0x18e [] ? up_read+0x28/0x2c [] ? do_page_fault+0x360/0x3b4 [] ? trace_hardirqs_on_caller+0x10b/0x12f [] ? finish_task_switch+0xb2/0xe3 [] ? finish_task_switch+0x46/0xe3 [] ? trace_hardirqs_off_caller+0x33/0x90 [] ? retint_swapgs+0x13/0x1b [] ? trace_hardirqs_on_caller+0x10b/0x12f [] ? audit_syscall_entry+0x11c/0x148 [] ? trace_hardirqs_on_thunk+0x3a/0x3f [] system_call_fastpath+0x16/0x1b Signed-off-by: Dave Jones Signed-off-by: David S. Miller diff --git a/net/irda/af_irda.c b/net/irda/af_irda.c index c9890e2..cc61697 100644 --- a/net/irda/af_irda.c +++ b/net/irda/af_irda.c @@ -1297,8 +1297,7 @@ static int irda_sendmsg(struct kiocb *iocb, struct socket *sock, /* Note : socket.c set MSG_EOR on SEQPACKET sockets */ if (msg->msg_flags & ~(MSG_DONTWAIT | MSG_EOR | MSG_CMSG_COMPAT | MSG_NOSIGNAL)) { - err = -EINVAL; - goto out; + return -EINVAL; } lock_sock(sk); -- cgit v0.10.2 From bfac3693c426d280b026f6a1b77dc2294ea43fea Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 12 Apr 2011 15:33:23 -0700 Subject: ieee802154: Remove hacked CFLAGS in net/ieee802154/Makefile It adds -Wall (which the kernel carefully controls already) and of all things -DDEBUG (which should be set by other means if desired, please we have dynamic-debug these days). Kill this noise. Reported-by: Dave Jones Signed-off-by: David S. Miller diff --git a/net/ieee802154/Makefile b/net/ieee802154/Makefile index ce2d335..5761185 100644 --- a/net/ieee802154/Makefile +++ b/net/ieee802154/Makefile @@ -1,5 +1,3 @@ obj-$(CONFIG_IEEE802154) += ieee802154.o af_802154.o ieee802154-y := netlink.o nl-mac.o nl-phy.o nl_policy.o wpan-class.o af_802154-y := af_ieee802154.o raw.o dgram.o - -ccflags-y += -Wall -DDEBUG -- cgit v0.10.2 From fcfa060468a4edcf776f0c1211d826d5de1668c1 Mon Sep 17 00:00:00 2001 From: Neil Turton Date: Mon, 11 Apr 2011 11:42:43 +0100 Subject: sfc: Use rmb() to ensure reads occur in order Enabling write-combining may also enable read reordering. The BIU is only guaranteed to read from a 128-bit CSR or 64-bit SRAM word when the host reads from its lowest address; otherwise the BIU may use the latched value. Therefore we need to reinstate the read memory barriers after the first read operation for each CSR or SRAM word. Signed-off-by; Ben Hutchings diff --git a/drivers/net/sfc/io.h b/drivers/net/sfc/io.h index d9d8c2e..cc97880 100644 --- a/drivers/net/sfc/io.h +++ b/drivers/net/sfc/io.h @@ -152,6 +152,7 @@ static inline void efx_reado(struct efx_nic *efx, efx_oword_t *value, spin_lock_irqsave(&efx->biu_lock, flags); value->u32[0] = _efx_readd(efx, reg + 0); + rmb(); value->u32[1] = _efx_readd(efx, reg + 4); value->u32[2] = _efx_readd(efx, reg + 8); value->u32[3] = _efx_readd(efx, reg + 12); @@ -174,6 +175,7 @@ static inline void efx_sram_readq(struct efx_nic *efx, void __iomem *membase, value->u64[0] = (__force __le64)__raw_readq(membase + addr); #else value->u32[0] = (__force __le32)__raw_readl(membase + addr); + rmb(); value->u32[1] = (__force __le32)__raw_readl(membase + addr + 4); #endif spin_unlock_irqrestore(&efx->biu_lock, flags); -- cgit v0.10.2 From 92656d707e961c88b6f931038c7dc41790a52e1e Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 12 Apr 2011 13:32:13 -0400 Subject: drm/radeon/kms: clean up gart dummy page handling As per Konrad's original patch, the dummy page used by the gart code and allocated in radeon_gart_init() was not freed properly in radeon_gart_fini(). At the same time r6xx and newer allocated and freed the dummy page on their own. So to do Konrad's patch one better, just remove the allocation and freeing of the dummy page in the r6xx, 7xx, evergreen, and ni code and allocate and free in the gart_init/fini() functions for all asics. Cc: Jerome Glisse Signed-off-by: Alex Deucher Acked-by: Konrad Rzeszutek Wilk Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index 0b0cc74..cff7144 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -3036,9 +3036,6 @@ int evergreen_init(struct radeon_device *rdev) { int r; - r = radeon_dummy_page_init(rdev); - if (r) - return r; /* This don't do much */ r = radeon_gem_init(rdev); if (r) @@ -3150,7 +3147,6 @@ void evergreen_fini(struct radeon_device *rdev) radeon_atombios_fini(rdev); kfree(rdev->bios); rdev->bios = NULL; - radeon_dummy_page_fini(rdev); } static void evergreen_pcie_gen2_enable(struct radeon_device *rdev) diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c index be271c4..304fc76 100644 --- a/drivers/gpu/drm/radeon/r600.c +++ b/drivers/gpu/drm/radeon/r600.c @@ -2509,9 +2509,6 @@ int r600_init(struct radeon_device *rdev) { int r; - r = radeon_dummy_page_init(rdev); - if (r) - return r; if (r600_debugfs_mc_info_init(rdev)) { DRM_ERROR("Failed to register debugfs file for mc !\n"); } @@ -2625,7 +2622,6 @@ void r600_fini(struct radeon_device *rdev) radeon_atombios_fini(rdev); kfree(rdev->bios); rdev->bios = NULL; - radeon_dummy_page_fini(rdev); } diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c index f0534ef..8a955bb 100644 --- a/drivers/gpu/drm/radeon/radeon_gart.c +++ b/drivers/gpu/drm/radeon/radeon_gart.c @@ -285,4 +285,6 @@ void radeon_gart_fini(struct radeon_device *rdev) rdev->gart.pages = NULL; rdev->gart.pages_addr = NULL; rdev->gart.ttm_alloced = NULL; + + radeon_dummy_page_fini(rdev); } diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c index b974ac7..4faa2e5 100644 --- a/drivers/gpu/drm/radeon/rv770.c +++ b/drivers/gpu/drm/radeon/rv770.c @@ -1255,9 +1255,6 @@ int rv770_init(struct radeon_device *rdev) { int r; - r = radeon_dummy_page_init(rdev); - if (r) - return r; /* This don't do much */ r = radeon_gem_init(rdev); if (r) @@ -1372,7 +1369,6 @@ void rv770_fini(struct radeon_device *rdev) radeon_atombios_fini(rdev); kfree(rdev->bios); rdev->bios = NULL; - radeon_dummy_page_fini(rdev); } static void rv770_pcie_gen2_enable(struct radeon_device *rdev) -- cgit v0.10.2 From 71e16bfbd2b1c63d4d97cc5059694c9346aee340 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 12 Apr 2011 13:33:27 -0400 Subject: drm/radeon/kms: fix suspend on rv530 asics Apparently only rv515 asics need the workaround added in f24d86f1a49505cdea56728b853a5d0a3f8e3d11 (drm/radeon/kms: fix resume regression for some r5xx laptops). Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=34709 Signed-off-by: Alex Deucher Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/atom.c b/drivers/gpu/drm/radeon/atom.c index 258fa5e..d71d375 100644 --- a/drivers/gpu/drm/radeon/atom.c +++ b/drivers/gpu/drm/radeon/atom.c @@ -32,6 +32,7 @@ #include "atom.h" #include "atom-names.h" #include "atom-bits.h" +#include "radeon.h" #define ATOM_COND_ABOVE 0 #define ATOM_COND_ABOVEOREQUAL 1 @@ -101,7 +102,9 @@ static void debug_print_spaces(int n) static uint32_t atom_iio_execute(struct atom_context *ctx, int base, uint32_t index, uint32_t data) { + struct radeon_device *rdev = ctx->card->dev->dev_private; uint32_t temp = 0xCDCDCDCD; + while (1) switch (CU8(base)) { case ATOM_IIO_NOP: @@ -112,7 +115,8 @@ static uint32_t atom_iio_execute(struct atom_context *ctx, int base, base += 3; break; case ATOM_IIO_WRITE: - (void)ctx->card->ioreg_read(ctx->card, CU16(base + 1)); + if (rdev->family == CHIP_RV515) + (void)ctx->card->ioreg_read(ctx->card, CU16(base + 1)); ctx->card->ioreg_write(ctx->card, CU16(base + 1), temp); base += 3; break; -- cgit v0.10.2 From b4df8be1048eb6c8c5a8bd1646f5989fd34b549e Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 12 Apr 2011 13:40:18 -0400 Subject: drm/radeon/kms: fix pcie_p callbacks on btc and cayman btc and cayman asics use the same callback for pcie port registers. Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c index eb888ee..ca57619 100644 --- a/drivers/gpu/drm/radeon/radeon_asic.c +++ b/drivers/gpu/drm/radeon/radeon_asic.c @@ -94,7 +94,7 @@ static void radeon_register_accessor_init(struct radeon_device *rdev) rdev->mc_rreg = &rs600_mc_rreg; rdev->mc_wreg = &rs600_mc_wreg; } - if ((rdev->family >= CHIP_R600) && (rdev->family <= CHIP_HEMLOCK)) { + if (rdev->family >= CHIP_R600) { rdev->pciep_rreg = &r600_pciep_rreg; rdev->pciep_wreg = &r600_pciep_wreg; } -- cgit v0.10.2 From 8a83ec5ee824a6bd431b49bdb2428c8bf88f03eb Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 12 Apr 2011 14:49:23 -0400 Subject: drm/radeon/kms: add voltage type to atom set voltage function This is needed for setting voltages other than vddc. Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index cff7144..f244817 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -122,7 +122,7 @@ void evergreen_pm_misc(struct radeon_device *rdev) if ((voltage->type == VOLTAGE_SW) && voltage->voltage) { if (voltage->voltage != rdev->pm.current_vddc) { - radeon_atom_set_voltage(rdev, voltage->voltage); + radeon_atom_set_voltage(rdev, voltage->voltage, SET_VOLTAGE_TYPE_ASIC_VDDC); rdev->pm.current_vddc = voltage->voltage; DRM_DEBUG("Setting: v: %d\n", voltage->voltage); } diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c index 304fc76..15d5829 100644 --- a/drivers/gpu/drm/radeon/r600.c +++ b/drivers/gpu/drm/radeon/r600.c @@ -587,7 +587,7 @@ void r600_pm_misc(struct radeon_device *rdev) if ((voltage->type == VOLTAGE_SW) && voltage->voltage) { if (voltage->voltage != rdev->pm.current_vddc) { - radeon_atom_set_voltage(rdev, voltage->voltage); + radeon_atom_set_voltage(rdev, voltage->voltage, SET_VOLTAGE_TYPE_ASIC_VDDC); rdev->pm.current_vddc = voltage->voltage; DRM_DEBUG_DRIVER("Setting: v: %d\n", voltage->voltage); } diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 93f5365..9955740 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -177,7 +177,7 @@ void radeon_pm_suspend(struct radeon_device *rdev); void radeon_pm_resume(struct radeon_device *rdev); void radeon_combios_get_power_modes(struct radeon_device *rdev); void radeon_atombios_get_power_modes(struct radeon_device *rdev); -void radeon_atom_set_voltage(struct radeon_device *rdev, u16 level); +void radeon_atom_set_voltage(struct radeon_device *rdev, u16 voltage_level, u8 voltage_type); void rs690_pm_info(struct radeon_device *rdev); extern int rv6xx_get_temp(struct radeon_device *rdev); extern int rv770_get_temp(struct radeon_device *rdev); diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c index 99768d9..330a5b5 100644 --- a/drivers/gpu/drm/radeon/radeon_atombios.c +++ b/drivers/gpu/drm/radeon/radeon_atombios.c @@ -2577,25 +2577,25 @@ union set_voltage { struct _SET_VOLTAGE_PARAMETERS_V2 v2; }; -void radeon_atom_set_voltage(struct radeon_device *rdev, u16 level) +void radeon_atom_set_voltage(struct radeon_device *rdev, u16 voltage_level, u8 voltage_type) { union set_voltage args; int index = GetIndexIntoMasterTable(COMMAND, SetVoltage); - u8 frev, crev, volt_index = level; + u8 frev, crev, volt_index = voltage_level; if (!atom_parse_cmd_header(rdev->mode_info.atom_context, index, &frev, &crev)) return; switch (crev) { case 1: - args.v1.ucVoltageType = SET_VOLTAGE_TYPE_ASIC_VDDC; + args.v1.ucVoltageType = voltage_type; args.v1.ucVoltageMode = SET_ASIC_VOLTAGE_MODE_ALL_SOURCE; args.v1.ucVoltageIndex = volt_index; break; case 2: - args.v2.ucVoltageType = SET_VOLTAGE_TYPE_ASIC_VDDC; + args.v2.ucVoltageType = voltage_type; args.v2.ucVoltageMode = SET_ASIC_VOLTAGE_MODE_SET_VOLTAGE; - args.v2.usVoltageLevel = cpu_to_le16(level); + args.v2.usVoltageLevel = cpu_to_le16(voltage_level); break; default: DRM_ERROR("Unknown table version %d, %d\n", frev, crev); diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c index 08de669..4010571 100644 --- a/drivers/gpu/drm/radeon/radeon_pm.c +++ b/drivers/gpu/drm/radeon/radeon_pm.c @@ -23,6 +23,7 @@ #include "drmP.h" #include "radeon.h" #include "avivod.h" +#include "atom.h" #ifdef CONFIG_ACPI #include #endif @@ -535,7 +536,8 @@ void radeon_pm_resume(struct radeon_device *rdev) /* set up the default clocks if the MC ucode is loaded */ if (ASIC_IS_DCE5(rdev) && rdev->mc_fw) { if (rdev->pm.default_vddc) - radeon_atom_set_voltage(rdev, rdev->pm.default_vddc); + radeon_atom_set_voltage(rdev, rdev->pm.default_vddc, + SET_VOLTAGE_TYPE_ASIC_VDDC); if (rdev->pm.default_sclk) radeon_set_engine_clock(rdev, rdev->pm.default_sclk); if (rdev->pm.default_mclk) @@ -585,7 +587,8 @@ int radeon_pm_init(struct radeon_device *rdev) /* set up the default clocks if the MC ucode is loaded */ if (ASIC_IS_DCE5(rdev) && rdev->mc_fw) { if (rdev->pm.default_vddc) - radeon_atom_set_voltage(rdev, rdev->pm.default_vddc); + radeon_atom_set_voltage(rdev, rdev->pm.default_vddc, + SET_VOLTAGE_TYPE_ASIC_VDDC); if (rdev->pm.default_sclk) radeon_set_engine_clock(rdev, rdev->pm.default_sclk); if (rdev->pm.default_mclk) diff --git a/drivers/gpu/drm/radeon/rs600.c b/drivers/gpu/drm/radeon/rs600.c index 876cebc..6e3b11e 100644 --- a/drivers/gpu/drm/radeon/rs600.c +++ b/drivers/gpu/drm/radeon/rs600.c @@ -114,7 +114,7 @@ void rs600_pm_misc(struct radeon_device *rdev) udelay(voltage->delay); } } else if (voltage->type == VOLTAGE_VDDC) - radeon_atom_set_voltage(rdev, voltage->vddc_id); + radeon_atom_set_voltage(rdev, voltage->vddc_id, SET_VOLTAGE_TYPE_ASIC_VDDC); dyn_pwrmgt_sclk_length = RREG32_PLL(DYN_PWRMGT_SCLK_LENGTH); dyn_pwrmgt_sclk_length &= ~REDUCED_POWER_SCLK_HILEN(0xf); diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c index 4faa2e5..ef8a5bab 100644 --- a/drivers/gpu/drm/radeon/rv770.c +++ b/drivers/gpu/drm/radeon/rv770.c @@ -106,7 +106,7 @@ void rv770_pm_misc(struct radeon_device *rdev) if ((voltage->type == VOLTAGE_SW) && voltage->voltage) { if (voltage->voltage != rdev->pm.current_vddc) { - radeon_atom_set_voltage(rdev, voltage->voltage); + radeon_atom_set_voltage(rdev, voltage->voltage, SET_VOLTAGE_TYPE_ASIC_VDDC); rdev->pm.current_vddc = voltage->voltage; DRM_DEBUG("Setting: v: %d\n", voltage->voltage); } -- cgit v0.10.2 From 2feea49ae34a2fcea1035136b85f5eaca56f5cd0 Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 12 Apr 2011 14:49:24 -0400 Subject: drm/radeon/kms: properly program vddci on evergreen+ Change vddci as well as vddc when changing power modes on evergreen/ni. Also, properly set vddci on boot up for ni cards. The vbios only sets the limited clocks and voltages on boot until the mc ucode is loaded. This should fix stability problems on some btc cards. Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index f244817..3453910 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -120,11 +120,16 @@ void evergreen_pm_misc(struct radeon_device *rdev) struct radeon_power_state *ps = &rdev->pm.power_state[req_ps_idx]; struct radeon_voltage *voltage = &ps->clock_info[req_cm_idx].voltage; - if ((voltage->type == VOLTAGE_SW) && voltage->voltage) { - if (voltage->voltage != rdev->pm.current_vddc) { + if (voltage->type == VOLTAGE_SW) { + if (voltage->voltage && (voltage->voltage != rdev->pm.current_vddc)) { radeon_atom_set_voltage(rdev, voltage->voltage, SET_VOLTAGE_TYPE_ASIC_VDDC); rdev->pm.current_vddc = voltage->voltage; - DRM_DEBUG("Setting: v: %d\n", voltage->voltage); + DRM_DEBUG("Setting: vddc: %d\n", voltage->voltage); + } + if (voltage->vddci && (voltage->vddci != rdev->pm.current_vddci)) { + radeon_atom_set_voltage(rdev, voltage->vddci, SET_VOLTAGE_TYPE_ASIC_VDDCI); + rdev->pm.current_vddci = voltage->vddci; + DRM_DEBUG("Setting: vddci: %d\n", voltage->vddci); } } } diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 9955740..ba643b5 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -767,7 +767,9 @@ struct radeon_voltage { u8 vddci_id; /* index into vddci voltage table */ bool vddci_enabled; /* r6xx+ sw */ - u32 voltage; + u16 voltage; + /* evergreen+ vddci */ + u16 vddci; }; /* clock mode flags */ @@ -835,10 +837,12 @@ struct radeon_pm { int default_power_state_index; u32 current_sclk; u32 current_mclk; - u32 current_vddc; + u16 current_vddc; + u16 current_vddci; u32 default_sclk; u32 default_mclk; - u32 default_vddc; + u16 default_vddc; + u16 default_vddci; struct radeon_i2c_chan *i2c_bus; /* selected pm method */ enum radeon_pm_method pm_method; diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c index 330a5b5..f5d12fb 100644 --- a/drivers/gpu/drm/radeon/radeon_atombios.c +++ b/drivers/gpu/drm/radeon/radeon_atombios.c @@ -2176,24 +2176,27 @@ static void radeon_atombios_add_pplib_thermal_controller(struct radeon_device *r } } -static u16 radeon_atombios_get_default_vddc(struct radeon_device *rdev) +static void radeon_atombios_get_default_voltages(struct radeon_device *rdev, + u16 *vddc, u16 *vddci) { struct radeon_mode_info *mode_info = &rdev->mode_info; int index = GetIndexIntoMasterTable(DATA, FirmwareInfo); u8 frev, crev; u16 data_offset; union firmware_info *firmware_info; - u16 vddc = 0; + + *vddc = 0; + *vddci = 0; if (atom_parse_data_header(mode_info->atom_context, index, NULL, &frev, &crev, &data_offset)) { firmware_info = (union firmware_info *)(mode_info->atom_context->bios + data_offset); - vddc = le16_to_cpu(firmware_info->info_14.usBootUpVDDCVoltage); + *vddc = le16_to_cpu(firmware_info->info_14.usBootUpVDDCVoltage); + if ((frev == 2) && (crev >= 2)) + *vddci = le16_to_cpu(firmware_info->info_22.usBootUpVDDCIVoltage); } - - return vddc; } static void radeon_atombios_parse_pplib_non_clock_info(struct radeon_device *rdev, @@ -2203,7 +2206,9 @@ static void radeon_atombios_parse_pplib_non_clock_info(struct radeon_device *rde int j; u32 misc = le32_to_cpu(non_clock_info->ulCapsAndSettings); u32 misc2 = le16_to_cpu(non_clock_info->usClassification); - u16 vddc = radeon_atombios_get_default_vddc(rdev); + u16 vddc, vddci; + + radeon_atombios_get_default_voltages(rdev, &vddc, &vddci); rdev->pm.power_state[state_index].misc = misc; rdev->pm.power_state[state_index].misc2 = misc2; @@ -2244,6 +2249,7 @@ static void radeon_atombios_parse_pplib_non_clock_info(struct radeon_device *rde rdev->pm.default_sclk = rdev->pm.power_state[state_index].clock_info[0].sclk; rdev->pm.default_mclk = rdev->pm.power_state[state_index].clock_info[0].mclk; rdev->pm.default_vddc = rdev->pm.power_state[state_index].clock_info[0].voltage.voltage; + rdev->pm.default_vddci = rdev->pm.power_state[state_index].clock_info[0].voltage.vddci; } else { /* patch the table values with the default slck/mclk from firmware info */ for (j = 0; j < mode_index; j++) { @@ -2286,6 +2292,8 @@ static bool radeon_atombios_parse_pplib_clock_info(struct radeon_device *rdev, VOLTAGE_SW; rdev->pm.power_state[state_index].clock_info[mode_index].voltage.voltage = le16_to_cpu(clock_info->evergreen.usVDDC); + rdev->pm.power_state[state_index].clock_info[mode_index].voltage.vddci = + le16_to_cpu(clock_info->evergreen.usVDDCI); } else { sclk = le16_to_cpu(clock_info->r600.usEngineClockLow); sclk |= clock_info->r600.ucEngineClockHigh << 16; diff --git a/drivers/gpu/drm/radeon/radeon_pm.c b/drivers/gpu/drm/radeon/radeon_pm.c index 4010571..86eda1e 100644 --- a/drivers/gpu/drm/radeon/radeon_pm.c +++ b/drivers/gpu/drm/radeon/radeon_pm.c @@ -538,6 +538,9 @@ void radeon_pm_resume(struct radeon_device *rdev) if (rdev->pm.default_vddc) radeon_atom_set_voltage(rdev, rdev->pm.default_vddc, SET_VOLTAGE_TYPE_ASIC_VDDC); + if (rdev->pm.default_vddci) + radeon_atom_set_voltage(rdev, rdev->pm.default_vddci, + SET_VOLTAGE_TYPE_ASIC_VDDCI); if (rdev->pm.default_sclk) radeon_set_engine_clock(rdev, rdev->pm.default_sclk); if (rdev->pm.default_mclk) @@ -550,6 +553,7 @@ void radeon_pm_resume(struct radeon_device *rdev) rdev->pm.current_sclk = rdev->pm.default_sclk; rdev->pm.current_mclk = rdev->pm.default_mclk; rdev->pm.current_vddc = rdev->pm.power_state[rdev->pm.default_power_state_index].clock_info[0].voltage.voltage; + rdev->pm.current_vddci = rdev->pm.power_state[rdev->pm.default_power_state_index].clock_info[0].voltage.vddci; if (rdev->pm.pm_method == PM_METHOD_DYNPM && rdev->pm.dynpm_state == DYNPM_STATE_SUSPENDED) { rdev->pm.dynpm_state = DYNPM_STATE_ACTIVE; -- cgit v0.10.2 From cbf15bdbbdaec3b2e3b4f476803a6149707d1a1f Mon Sep 17 00:00:00 2001 From: Konstantin Khlebnikov Date: Tue, 12 Apr 2011 14:27:47 +0400 Subject: i915: select VIDEO_OUTPUT_CONTROL for ACPI_VIDEO fix Kconfig warning: (DRM_I915 && STUB_POULSBO) selects ACPI_VIDEO which has unmet direct dependencies (ACPI && X86 && BACKLIGHT_CLASS_DEVICE && VIDEO_OUTPUT_CONTROL && INPUT) Signed-off-by: Konstantin Khlebnikov Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig index a6feb78c..c58f691 100644 --- a/drivers/gpu/drm/Kconfig +++ b/drivers/gpu/drm/Kconfig @@ -96,6 +96,7 @@ config DRM_I915 # i915 depends on ACPI_VIDEO when ACPI is enabled # but for select to work, need to select ACPI_VIDEO's dependencies, ick select BACKLIGHT_CLASS_DEVICE if ACPI + select VIDEO_OUTPUT_CONTROL if ACPI select INPUT if ACPI select ACPI_VIDEO if ACPI select ACPI_BUTTON if ACPI diff --git a/drivers/gpu/stub/Kconfig b/drivers/gpu/stub/Kconfig index 70e60a4..4199179 100644 --- a/drivers/gpu/stub/Kconfig +++ b/drivers/gpu/stub/Kconfig @@ -5,6 +5,7 @@ config STUB_POULSBO # Poulsbo stub depends on ACPI_VIDEO when ACPI is enabled # but for select to work, need to select ACPI_VIDEO's dependencies, ick select BACKLIGHT_CLASS_DEVICE if ACPI + select VIDEO_OUTPUT_CONTROL if ACPI select INPUT if ACPI select ACPI_VIDEO if ACPI select THERMAL if ACPI -- cgit v0.10.2 From d87dfdbfc91c5e37288e7e8f7afdd992ba61a60d Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Wed, 13 Apr 2011 09:15:09 +1000 Subject: Revert "ttm: Utilize the DMA API for pages that have TTM_PAGE_FLAG_DMA32 set." This reverts commit 69a07f0b117a40fcc1a479358d8e1f41793617f2. We've tracked a number of problems back to this, and Thomas thinks we should redesign this for .40/41 anyways so I'm happy to revert it. Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc.c b/drivers/gpu/drm/ttm/ttm_page_alloc.c index 737a2a2e..9d9d929 100644 --- a/drivers/gpu/drm/ttm/ttm_page_alloc.c +++ b/drivers/gpu/drm/ttm/ttm_page_alloc.c @@ -683,22 +683,14 @@ int ttm_get_pages(struct list_head *pages, int flags, gfp_flags |= GFP_HIGHUSER; for (r = 0; r < count; ++r) { - if ((flags & TTM_PAGE_FLAG_DMA32) && dma_address) { - void *addr; - addr = dma_alloc_coherent(NULL, PAGE_SIZE, - &dma_address[r], - gfp_flags); - if (addr == NULL) - return -ENOMEM; - p = virt_to_page(addr); - } else - p = alloc_page(gfp_flags); + p = alloc_page(gfp_flags); if (!p) { printk(KERN_ERR TTM_PFX "Unable to allocate page."); return -ENOMEM; } + list_add(&p->lru, pages); } return 0; @@ -746,24 +738,12 @@ void ttm_put_pages(struct list_head *pages, unsigned page_count, int flags, unsigned long irq_flags; struct ttm_page_pool *pool = ttm_get_pool(flags, cstate); struct page *p, *tmp; - unsigned r; if (pool == NULL) { /* No pool for this memory type so free the pages */ - r = page_count-1; list_for_each_entry_safe(p, tmp, pages, lru) { - if ((flags & TTM_PAGE_FLAG_DMA32) && dma_address) { - void *addr = page_address(p); - WARN_ON(!addr || !dma_address[r]); - if (addr) - dma_free_coherent(NULL, PAGE_SIZE, - addr, - dma_address[r]); - dma_address[r] = 0; - } else - __free_page(p); - r--; + __free_page(p); } /* Make the pages list empty */ INIT_LIST_HEAD(pages); -- cgit v0.10.2 From 2582b6efceb43dce63b4a1090d289934067a972d Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Wed, 13 Apr 2011 09:20:24 +1000 Subject: Revert "i915: restore only the mode of this driver on lastclose" This reverts commit 0a0883c8433c743dad0a4d9ebe2717558f2c209e. this was in my tree by accident, I meant to rebase it out and didn't realise in time. Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c index 12876f2..7273037 100644 --- a/drivers/gpu/drm/i915/i915_dma.c +++ b/drivers/gpu/drm/i915/i915_dma.c @@ -2207,7 +2207,7 @@ void i915_driver_lastclose(struct drm_device * dev) drm_i915_private_t *dev_priv = dev->dev_private; if (!dev_priv || drm_core_check_feature(dev, DRIVER_MODESET)) { - intel_fb_restore_mode(dev); + drm_fb_helper_restore(); vga_switcheroo_process_delayed_switch(); return; } diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 1d20712..f5b0d83 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -338,5 +338,4 @@ extern int intel_overlay_attrs(struct drm_device *dev, void *data, struct drm_file *file_priv); extern void intel_fb_output_poll_changed(struct drm_device *dev); -extern void intel_fb_restore_mode(struct drm_device *dev); #endif /* __INTEL_DRV_H__ */ diff --git a/drivers/gpu/drm/i915/intel_fb.c b/drivers/gpu/drm/i915/intel_fb.c index 96a45c4..5127827 100644 --- a/drivers/gpu/drm/i915/intel_fb.c +++ b/drivers/gpu/drm/i915/intel_fb.c @@ -264,19 +264,3 @@ void intel_fb_output_poll_changed(struct drm_device *dev) drm_i915_private_t *dev_priv = dev->dev_private; drm_fb_helper_hotplug_event(&dev_priv->fbdev->helper); } - -void intel_fb_restore_mode(struct drm_device *dev) -{ - drm_i915_private_t *dev_priv = dev->dev_private; - int ret, i; - - if (!dev_priv->fbdev) - return; - - for (i = 0; i < dev_priv->fbdev->helper.crtc_count; i++) { - struct drm_mode_set *mode_set = &dev_priv->fbdev->helper.crtc_info[i].mode_set; - ret = drm_crtc_helper_set_config(mode_set); - if (ret) - DRM_DEBUG("failed to restore crtc mode\n"); - } -} -- cgit v0.10.2 From 109b36a2bb3eebf5c9994980e724958a5b2b62b6 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Tue, 12 Apr 2011 13:57:39 -0400 Subject: Btrfs: make uncache_state unconditional The extent_io code can take cached pointers into the extent state trees, and these can make lookups much faster in common operations. The caching only happens when specific bits are set that prevent merging and splitting of the extent state. A help function was added to uncache the state, and it was testing the same set of conditionals. This can leak in very strange corner cases where the lock bit goes away unexpectedly. The uncaching should be unconditional. Once we have a ref on the extent we should always give it up. Signed-off-by: Chris Mason diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 8dcfb77..1c462f89 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -694,10 +694,8 @@ static void uncache_state(struct extent_state **cached_ptr) { if (cached_ptr && (*cached_ptr)) { struct extent_state *state = *cached_ptr; - if (state->state & (EXTENT_IOBITS | EXTENT_BOUNDARY)) { - *cached_ptr = NULL; - free_extent_state(state); - } + *cached_ptr = NULL; + free_extent_state(state); } } @@ -1764,7 +1762,7 @@ static void end_bio_extent_readpage(struct bio *bio, int err) spin_lock(&tree->lock); state = find_first_extent_bit_state(tree, start, 0); - if (state) { + if (state && state->start == start) { /* * take a reference on the state, unlock will drop * the ref -- cgit v0.10.2 From 7db6a7fa09884b34d2a5d4e6e4ed58664a5f0cf8 Mon Sep 17 00:00:00 2001 From: Eric Miao Date: Tue, 12 Apr 2011 18:39:39 +0800 Subject: ARM: pxa: convert incorrect IRQ_TO_IRQ() to irq_to_gpio() This fixes the failure to register the IRQ_RTCAlrm alarm as a wakeup event. It is misinterpreted as a gpio irq not a PWER bitmask. Fixed this by converting the incorrect IRQ_TO_IRQ() to a correct version of irq_to_gpio(). Reported-by: Nick Bane Signed-off-by: Eric Miao diff --git a/arch/arm/mach-pxa/include/mach/gpio.h b/arch/arm/mach-pxa/include/mach/gpio.h index 48ef925..c463950 100644 --- a/arch/arm/mach-pxa/include/mach/gpio.h +++ b/arch/arm/mach-pxa/include/mach/gpio.h @@ -103,7 +103,20 @@ #define gpio_to_bank(gpio) ((gpio) >> 5) #define gpio_to_irq(gpio) IRQ_GPIO(gpio) -#define irq_to_gpio(irq) IRQ_TO_GPIO(irq) + +static inline int irq_to_gpio(unsigned int irq) +{ + int gpio; + + if (irq == IRQ_GPIO0 || irq == IRQ_GPIO1) + return irq - IRQ_GPIO0; + + gpio = irq - PXA_GPIO_IRQ_BASE; + if (gpio >= 2 && gpio < NR_BUILTIN_GPIO) + return gpio; + + return -1; +} #ifdef CONFIG_CPU_PXA26x /* GPIO86/87/88/89 on PXA26x have their direction bits in GPDR2 inverted, diff --git a/arch/arm/mach-pxa/include/mach/irqs.h b/arch/arm/mach-pxa/include/mach/irqs.h index a4285fc..0384024 100644 --- a/arch/arm/mach-pxa/include/mach/irqs.h +++ b/arch/arm/mach-pxa/include/mach/irqs.h @@ -93,9 +93,6 @@ #define GPIO_2_x_TO_IRQ(x) (PXA_GPIO_IRQ_BASE + (x)) #define IRQ_GPIO(x) (((x) < 2) ? (IRQ_GPIO0 + (x)) : GPIO_2_x_TO_IRQ(x)) -#define IRQ_TO_GPIO_2_x(i) ((i) - PXA_GPIO_IRQ_BASE) -#define IRQ_TO_GPIO(i) (((i) < IRQ_GPIO(2)) ? ((i) - IRQ_GPIO0) : IRQ_TO_GPIO_2_x(i)) - /* * The following interrupts are for board specific purposes. Since * the kernel can only run on one machine at a time, we can re-use diff --git a/arch/arm/mach-pxa/pxa25x.c b/arch/arm/mach-pxa/pxa25x.c index 6bde595..a4af8c5 100644 --- a/arch/arm/mach-pxa/pxa25x.c +++ b/arch/arm/mach-pxa/pxa25x.c @@ -285,7 +285,7 @@ static inline void pxa25x_init_pm(void) {} static int pxa25x_set_wake(struct irq_data *d, unsigned int on) { - int gpio = IRQ_TO_GPIO(d->irq); + int gpio = irq_to_gpio(d->irq); uint32_t mask = 0; if (gpio >= 0 && gpio < 85) diff --git a/arch/arm/mach-pxa/pxa27x.c b/arch/arm/mach-pxa/pxa27x.c index 1cb5d0f..909756e 100644 --- a/arch/arm/mach-pxa/pxa27x.c +++ b/arch/arm/mach-pxa/pxa27x.c @@ -345,7 +345,7 @@ static inline void pxa27x_init_pm(void) {} */ static int pxa27x_set_wake(struct irq_data *d, unsigned int on) { - int gpio = IRQ_TO_GPIO(d->irq); + int gpio = irq_to_gpio(d->irq); uint32_t mask; if (gpio >= 0 && gpio < 128) diff --git a/drivers/pcmcia/pxa2xx_trizeps4.c b/drivers/pcmcia/pxa2xx_trizeps4.c index 5bb4e1a..b829e65 100644 --- a/drivers/pcmcia/pxa2xx_trizeps4.c +++ b/drivers/pcmcia/pxa2xx_trizeps4.c @@ -69,15 +69,15 @@ static int trizeps_pcmcia_hw_init(struct soc_pcmcia_socket *skt) for (i = 0; i < ARRAY_SIZE(irqs); i++) { if (irqs[i].sock != skt->nr) continue; - if (gpio_request(IRQ_TO_GPIO(irqs[i].irq), irqs[i].str) < 0) { + if (gpio_request(irq_to_gpio(irqs[i].irq), irqs[i].str) < 0) { pr_err("%s: sock %d unable to request gpio %d\n", - __func__, skt->nr, IRQ_TO_GPIO(irqs[i].irq)); + __func__, skt->nr, irq_to_gpio(irqs[i].irq)); ret = -EBUSY; goto error; } - if (gpio_direction_input(IRQ_TO_GPIO(irqs[i].irq)) < 0) { + if (gpio_direction_input(irq_to_gpio(irqs[i].irq)) < 0) { pr_err("%s: sock %d unable to set input gpio %d\n", - __func__, skt->nr, IRQ_TO_GPIO(irqs[i].irq)); + __func__, skt->nr, irq_to_gpio(irqs[i].irq)); ret = -EINVAL; goto error; } @@ -86,7 +86,7 @@ static int trizeps_pcmcia_hw_init(struct soc_pcmcia_socket *skt) error: for (; i >= 0; i--) { - gpio_free(IRQ_TO_GPIO(irqs[i].irq)); + gpio_free(irq_to_gpio(irqs[i].irq)); } return (ret); } @@ -97,7 +97,7 @@ static void trizeps_pcmcia_hw_shutdown(struct soc_pcmcia_socket *skt) /* free allocated gpio's */ gpio_free(GPIO_PRDY); for (i = 0; i < ARRAY_SIZE(irqs); i++) - gpio_free(IRQ_TO_GPIO(irqs[i].irq)); + gpio_free(irq_to_gpio(irqs[i].irq)); } static unsigned long trizeps_pcmcia_status[2]; -- cgit v0.10.2 From 60d48c1e67dc8de0676453de18adba1768fb6fab Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Tue, 12 Apr 2011 22:24:45 +0200 Subject: m68k,m68knommu: Wire up name_to_handle_at, open_by_handle_at, clock_adjtime, syncfs Signed-off-by: Geert Uytterhoeven Acked-by: Greg Ungerer Signed-off-by: Linus Torvalds diff --git a/arch/m68k/include/asm/unistd.h b/arch/m68k/include/asm/unistd.h index 26d851d..29e1790 100644 --- a/arch/m68k/include/asm/unistd.h +++ b/arch/m68k/include/asm/unistd.h @@ -343,10 +343,14 @@ #define __NR_fanotify_init 337 #define __NR_fanotify_mark 338 #define __NR_prlimit64 339 +#define __NR_name_to_handle_at 340 +#define __NR_open_by_handle_at 341 +#define __NR_clock_adjtime 342 +#define __NR_syncfs 343 #ifdef __KERNEL__ -#define NR_syscalls 340 +#define NR_syscalls 344 #define __ARCH_WANT_IPC_PARSE_VERSION #define __ARCH_WANT_OLD_READDIR diff --git a/arch/m68k/kernel/entry_mm.S b/arch/m68k/kernel/entry_mm.S index 1559dea..1359ee6 100644 --- a/arch/m68k/kernel/entry_mm.S +++ b/arch/m68k/kernel/entry_mm.S @@ -750,4 +750,8 @@ sys_call_table: .long sys_fanotify_init .long sys_fanotify_mark .long sys_prlimit64 + .long sys_name_to_handle_at /* 340 */ + .long sys_open_by_handle_at + .long sys_clock_adjtime + .long sys_syncfs diff --git a/arch/m68k/kernel/syscalltable.S b/arch/m68k/kernel/syscalltable.S index 79b1ed1..9b8393d 100644 --- a/arch/m68k/kernel/syscalltable.S +++ b/arch/m68k/kernel/syscalltable.S @@ -358,6 +358,10 @@ ENTRY(sys_call_table) .long sys_fanotify_init .long sys_fanotify_mark .long sys_prlimit64 + .long sys_name_to_handle_at /* 340 */ + .long sys_open_by_handle_at + .long sys_clock_adjtime + .long sys_syncfs .rept NR_syscalls-(.-sys_call_table)/4 .long sys_ni_syscall -- cgit v0.10.2 From ea2d36883ca8e6caab23b6d15bfa80b1d1d81d2f Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Tue, 12 Apr 2011 14:38:37 +0000 Subject: net: Disable all TSO features when SG is disabled The feature flags NETIF_F_TSO and NETIF_F_TSO6 independently enable TSO for IPv4 and IPv6 respectively. However, the test in netdev_fix_features() and its predecessor functions was never updated to check for NETIF_F_TSO6, possibly because it was originally proposed that TSO for IPv6 would be dependent on both feature flags. Now that these feature flags can be changed independently from user-space and we depend on netdev_fix_features() to fix invalid feature combinations, it's important to disable them both if scatter-gather is disabled. Also disable NETIF_F_TSO_ECN so user-space sees all TSO features as disabled. Signed-off-by: Ben Hutchings Signed-off-by: David S. Miller diff --git a/net/core/dev.c b/net/core/dev.c index 956d3b0..6401fb5 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5203,9 +5203,9 @@ u32 netdev_fix_features(struct net_device *dev, u32 features) } /* TSO requires that SG is present as well. */ - if ((features & NETIF_F_TSO) && !(features & NETIF_F_SG)) { - netdev_info(dev, "Dropping NETIF_F_TSO since no SG feature.\n"); - features &= ~NETIF_F_TSO; + if ((features & NETIF_F_ALL_TSO) && !(features & NETIF_F_SG)) { + netdev_info(dev, "Dropping TSO features since no SG feature.\n"); + features &= ~NETIF_F_ALL_TSO; } /* Software GSO depends on SG. */ -- cgit v0.10.2 From 31d8b9e099e59f880aa65095951559896d4e20fa Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Tue, 12 Apr 2011 14:47:15 +0000 Subject: net: Disable NETIF_F_TSO_ECN when TSO is disabled NETIF_F_TSO_ECN has no effect when TSO is disabled; this just means that feature state will be accurately reported to user-space. Signed-off-by: Ben Hutchings Signed-off-by: David S. Miller diff --git a/net/core/dev.c b/net/core/dev.c index 6401fb5..c2ac599 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -5208,6 +5208,10 @@ u32 netdev_fix_features(struct net_device *dev, u32 features) features &= ~NETIF_F_ALL_TSO; } + /* TSO ECN requires that TSO is present as well. */ + if ((features & NETIF_F_ALL_TSO) == NETIF_F_TSO_ECN) + features &= ~NETIF_F_TSO_ECN; + /* Software GSO depends on SG. */ if ((features & NETIF_F_GSO) && !(features & NETIF_F_SG)) { netdev_info(dev, "Dropping NETIF_F_GSO since no SG feature.\n"); -- cgit v0.10.2 From 25f7bf7d0dfb460505cbe42676340e33100aca2e Mon Sep 17 00:00:00 2001 From: Vlad Yasevich Date: Tue, 12 Apr 2011 15:20:48 +0000 Subject: sctp: fix oops when updating retransmit path with DEBUG on commit fbdf501c9374966a56829ecca3a7f25d2b49a305 sctp: Do no select unconfirmed transports for retransmissions Introduced the initial falt. commit d598b166ced20d9b9281ea3527c0e18405ddb803 sctp: Make sure we always return valid retransmit path Solved the problem, but forgot to change the DEBUG statement. Thus it was still possible to dereference a NULL pointer. Signed-off-by: Wei Yongjun Signed-off-by: Vlad Yasevich Signed-off-by: David S. Miller diff --git a/net/sctp/associola.c b/net/sctp/associola.c index 0698cad..922fdd7 100644 --- a/net/sctp/associola.c +++ b/net/sctp/associola.c @@ -1323,6 +1323,8 @@ void sctp_assoc_update_retran_path(struct sctp_association *asoc) if (t) asoc->peer.retran_path = t; + else + t = asoc->peer.retran_path; SCTP_DEBUG_PRINTK_IPADDR("sctp_assoc_update_retran_path:association" " %p addr: ", -- cgit v0.10.2 From 9494c7c5774d64a84a269aad38c153c4dbff97e6 Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Tue, 12 Apr 2011 15:22:22 +0000 Subject: sctp: fix oops while removed transport still using as retran path Since we can not update retran path to unconfirmed transports, when we remove a peer, the retran path may not be update if the other transports are all unconfirmed, and we will still using the removed transport as the retran path. This may cause panic if retrasnmit happen. Signed-off-by: Wei Yongjun Signed-off-by: David S. Miller diff --git a/net/sctp/associola.c b/net/sctp/associola.c index 922fdd7..1a21c57 100644 --- a/net/sctp/associola.c +++ b/net/sctp/associola.c @@ -569,6 +569,8 @@ void sctp_assoc_rm_peer(struct sctp_association *asoc, sctp_assoc_set_primary(asoc, transport); if (asoc->peer.active_path == peer) asoc->peer.active_path = transport; + if (asoc->peer.retran_path == peer) + asoc->peer.retran_path = transport; if (asoc->peer.last_data_from == peer) asoc->peer.last_data_from = transport; -- cgit v0.10.2 From 2e6a00356a066d34cd00872b067589549169ad48 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Thu, 17 Mar 2011 15:17:59 +0800 Subject: Btrfs: Check if btrfs_next_leaf() returns error in btrfs_listxattr() btrfs_next_leaf() can return -errno, and we should propagate it to userspace. This also simplifies how we walk the btree path. Signed-off-by: Li Zefan diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c index e5d22f2..07b9bc3 100644 --- a/fs/btrfs/xattr.c +++ b/fs/btrfs/xattr.c @@ -180,11 +180,10 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size) struct btrfs_path *path; struct extent_buffer *leaf; struct btrfs_dir_item *di; - int ret = 0, slot, advance; + int ret = 0, slot; size_t total_size = 0, size_left = size; unsigned long name_ptr; size_t name_len; - u32 nritems; /* * ok we want all objects associated with this id. @@ -204,34 +203,24 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size) ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) goto err; - advance = 0; + while (1) { leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); slot = path->slots[0]; /* this is where we start walking through the path */ - if (advance || slot >= nritems) { + if (slot >= btrfs_header_nritems(leaf)) { /* * if we've reached the last slot in this leaf we need * to go to the next leaf and reset everything */ - if (slot >= nritems-1) { - ret = btrfs_next_leaf(root, path); - if (ret) - break; - leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); - slot = path->slots[0]; - } else { - /* - * just walking through the slots on this leaf - */ - slot++; - path->slots[0]++; - } + ret = btrfs_next_leaf(root, path); + if (ret < 0) + goto err; + else if (ret > 0) + break; + continue; } - advance = 1; btrfs_item_key_to_cpu(leaf, &found_key, slot); @@ -250,7 +239,7 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size) /* we are just looking for how big our buffer needs to be */ if (!size) - continue; + goto next; if (!buffer || (name_len + 1) > size_left) { ret = -ERANGE; @@ -263,6 +252,8 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size) size_left -= name_len + 1; buffer += name_len + 1; +next: + path->slots[0]++; } ret = total_size; -- cgit v0.10.2 From b9e03af0bcc11310f6be4a3951c9ee2c26465011 Mon Sep 17 00:00:00 2001 From: Li Zefan Date: Wed, 23 Mar 2011 10:43:58 +0800 Subject: Btrfs: Check if btrfs_next_leaf() returns error in btrfs_real_readdir() btrfs_next_leaf() can return -errno, and we should propagate it to userspace. This also simplifies how we walk the btree path. Signed-off-by: Li Zefan diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 55a6a0b..b9f7f52 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -4221,10 +4221,8 @@ static int btrfs_real_readdir(struct file *filp, void *dirent, struct btrfs_key found_key; struct btrfs_path *path; int ret; - u32 nritems; struct extent_buffer *leaf; int slot; - int advance; unsigned char d_type; int over = 0; u32 di_cur; @@ -4267,27 +4265,19 @@ static int btrfs_real_readdir(struct file *filp, void *dirent, ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) goto err; - advance = 0; while (1) { leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); slot = path->slots[0]; - if (advance || slot >= nritems) { - if (slot >= nritems - 1) { - ret = btrfs_next_leaf(root, path); - if (ret) - break; - leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); - slot = path->slots[0]; - } else { - slot++; - path->slots[0]++; - } + if (slot >= btrfs_header_nritems(leaf)) { + ret = btrfs_next_leaf(root, path); + if (ret < 0) + goto err; + else if (ret > 0) + break; + continue; } - advance = 1; item = btrfs_item_nr(leaf, slot); btrfs_item_key_to_cpu(leaf, &found_key, slot); @@ -4296,7 +4286,7 @@ static int btrfs_real_readdir(struct file *filp, void *dirent, if (btrfs_key_type(&found_key) != key_type) break; if (found_key.offset < filp->f_pos) - continue; + goto next; filp->f_pos = found_key.offset; @@ -4349,6 +4339,8 @@ skip: di_cur += di_len; di = (struct btrfs_dir_item *)((char *)di + di_len); } +next: + path->slots[0]++; } /* Reached end of directory/root. Bump pos past the last item. */ -- cgit v0.10.2 From 3153495d8ed6a9bb9f00aea42c18dc488a885dd6 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 13 Apr 2011 13:19:21 +0800 Subject: Btrfs: Fix incorrect inode nlink in btrfs_link() Link count of the inode is not decreased if btrfs_set_inode_index() fails. Signed-off-by: Miao Xie Singed-off-by: Li Zefan diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index b9f7f52..a4157cf 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -4846,9 +4846,6 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir, if (inode->i_nlink == ~0U) return -EMLINK; - btrfs_inc_nlink(inode); - inode->i_ctime = CURRENT_TIME; - err = btrfs_set_inode_index(dir, &index); if (err) goto fail; @@ -4864,6 +4861,9 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir, goto fail; } + btrfs_inc_nlink(inode); + inode->i_ctime = CURRENT_TIME; + btrfs_set_trans_block_group(trans, dir); ihold(inode); -- cgit v0.10.2 From 329c5056be8774255db04b01242a9ff4f02eb8ea Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 13 Apr 2011 14:07:59 +0800 Subject: Btrfs: Check validity before setting an acl Call posix_acl_valid() to check if an acl is valid or not. Signed-off-by: Miao Xie Signed-off-by: Li Zefan diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c index 9c94934..a892bc2 100644 --- a/fs/btrfs/acl.c +++ b/fs/btrfs/acl.c @@ -178,16 +178,17 @@ static int btrfs_xattr_acl_set(struct dentry *dentry, const char *name, if (value) { acl = posix_acl_from_xattr(value, size); - if (acl == NULL) { - value = NULL; - size = 0; + if (acl) { + ret = posix_acl_valid(acl); + if (ret) + goto out; } else if (IS_ERR(acl)) { return PTR_ERR(acl); } } ret = btrfs_set_acl(NULL, dentry->d_inode, acl, type); - +out: posix_acl_release(acl); return ret; -- cgit v0.10.2 From ba6a078b77e0dc1309d7e6e2ee034b92ab91f88c Mon Sep 17 00:00:00 2001 From: Jarod Wilson Date: Tue, 12 Apr 2011 23:13:08 -0700 Subject: Input: add KEY_IMAGES specifically for AL Image Browser Many media center remotes have buttons intended for jumping straight to one type of media browser or another -- commonly, images/photos/pictures, audio/music, television, and movies. At present, remotes with an images or photos or pictures button use any number of different keycodes which sort of maybe fit. I've seen at least KEY_MEDIA, KEY_CAMERA, KEY_GRAPHICSEDITOR and KEY_PRESENTATION. None of those seem quite right. In my mind, KEY_MEDIA should be something more like a media center application launcher (and I'd like to standardize on that for things like the windows media center button on the mce remotes). KEY_CAMERA is used in a lot of webcams, and typically means "take a picture now". KEY_GRAPHICSEDITOR implies an editor, not a browser. KEY_PRESENTATION might be the closest fit here, if you think "photo slide show", but it may well be more intended for "run application in full-screen presentation mode" or to launch something like magicpoint, I dunno. And thus, I'd like to have a KEY_IMAGES, which matches the HID Usage AL Image Browser, the meaning of which I think is crystal-clear. I believe AL Audio Browser is already covered by KEY_AUDIO, and AL Movie Browser by KEY_VIDEO, so I'm also adding appropriate comments next to those keys. Signed-off-by: Jarod Wilson Signed-off-by: Dmitry Torokhov diff --git a/include/linux/input.h b/include/linux/input.h index 056ae8a..0cc25e4 100644 --- a/include/linux/input.h +++ b/include/linux/input.h @@ -553,8 +553,8 @@ struct input_keymap_entry { #define KEY_DVD 0x185 /* Media Select DVD */ #define KEY_AUX 0x186 #define KEY_MP3 0x187 -#define KEY_AUDIO 0x188 -#define KEY_VIDEO 0x189 +#define KEY_AUDIO 0x188 /* AL Audio Browser */ +#define KEY_VIDEO 0x189 /* AL Movie Browser */ #define KEY_DIRECTORY 0x18a #define KEY_LIST 0x18b #define KEY_MEMO 0x18c /* Media Select Messages */ @@ -603,8 +603,9 @@ struct input_keymap_entry { #define KEY_FRAMEFORWARD 0x1b5 #define KEY_CONTEXT_MENU 0x1b6 /* GenDesc - system context menu */ #define KEY_MEDIA_REPEAT 0x1b7 /* Consumer - transport control */ -#define KEY_10CHANNELSUP 0x1b8 /* 10 channels up (10+) */ -#define KEY_10CHANNELSDOWN 0x1b9 /* 10 channels down (10-) */ +#define KEY_10CHANNELSUP 0x1b8 /* 10 channels up (10+) */ +#define KEY_10CHANNELSDOWN 0x1b9 /* 10 channels down (10-) */ +#define KEY_IMAGES 0x1ba /* AL Image Browser */ #define KEY_DEL_EOL 0x1c0 #define KEY_DEL_EOS 0x1c1 -- cgit v0.10.2 From b1e064b81e238d47cb56544b34c9baf473e09837 Mon Sep 17 00:00:00 2001 From: Chase Douglas Date: Tue, 12 Apr 2011 23:29:07 -0700 Subject: Input: document event types and codes and their intended use This commit adds the file Documentation/input/event-codes.txt. Acked-by: Henrik Rydberg Reviewed-by: Peter Hutterer Signed-off-by: Chase Douglas Signed-off-by: Chris Bagwell Signed-off-by: Dmitry Torokhov diff --git a/Documentation/input/event-codes.txt b/Documentation/input/event-codes.txt new file mode 100644 index 0000000..f13aee5 --- /dev/null +++ b/Documentation/input/event-codes.txt @@ -0,0 +1,256 @@ +The input protocol uses a map of types and codes to express input device values +to userspace. This document describes the types and codes and how and when they +may be used. + +A single hardware event generates multiple input events. Each input event +contains the new value of a single data item. A special event type, EV_SYN, is +used to separate input events into packets of input data changes occurring at +the same moment in time. In the following, the term "event" refers to a single +input event encompassing a type, code, and value. + +The input protocol is a stateful protocol. Events are emitted only when values +of event codes have changed. However, the state is maintained within the Linux +input subsystem; drivers do not need to maintain the state and may attempt to +emit unchanged values without harm. Userspace may obtain the current state of +event code values using the EVIOCG* ioctls defined in linux/input.h. The event +reports supported by a device are also provided by sysfs in +class/input/event*/device/capabilities/, and the properties of a device are +provided in class/input/event*/device/properties. + +Types: +========== +Types are groupings of codes under a logical input construct. Each type has a +set of applicable codes to be used in generating events. See the Codes section +for details on valid codes for each type. + +* EV_SYN: + - Used as markers to separate events. Events may be separated in time or in + space, such as with the multitouch protocol. + +* EV_KEY: + - Used to describe state changes of keyboards, buttons, or other key-like + devices. + +* EV_REL: + - Used to describe relative axis value changes, e.g. moving the mouse 5 units + to the left. + +* EV_ABS: + - Used to describe absolute axis value changes, e.g. describing the + coordinates of a touch on a touchscreen. + +* EV_MSC: + - Used to describe miscellaneous input data that do not fit into other types. + +* EV_SW: + - Used to describe binary state input switches. + +* EV_LED: + - Used to turn LEDs on devices on and off. + +* EV_SND: + - Used to output sound to devices. + +* EV_REP: + - Used for autorepeating devices. + +* EV_FF: + - Used to send force feedback commands to an input device. + +* EV_PWR: + - A special type for power button and switch input. + +* EV_FF_STATUS: + - Used to receive force feedback device status. + +Codes: +========== +Codes define the precise type of event. + +EV_SYN: +---------- +EV_SYN event values are undefined. Their usage is defined only by when they are +sent in the evdev event stream. + +* SYN_REPORT: + - Used to synchronize and separate events into packets of input data changes + occurring at the same moment in time. For example, motion of a mouse may set + the REL_X and REL_Y values for one motion, then emit a SYN_REPORT. The next + motion will emit more REL_X and REL_Y values and send another SYN_REPORT. + +* SYN_CONFIG: + - TBD + +* SYN_MT_REPORT: + - Used to synchronize and separate touch events. See the + multi-touch-protocol.txt document for more information. + +EV_KEY: +---------- +EV_KEY events take the form KEY_ or BTN_. For example, KEY_A is used +to represent the 'A' key on a keyboard. When a key is depressed, an event with +the key's code is emitted with value 1. When the key is released, an event is +emitted with value 0. Some hardware send events when a key is repeated. These +events have a value of 2. In general, KEY_ is used for keyboard keys, and +BTN_ is used for other types of momentary switch events. + +A few EV_KEY codes have special meanings: + +* BTN_TOOL_: + - These codes are used in conjunction with input trackpads, tablets, and + touchscreens. These devices may be used with fingers, pens, or other tools. + When an event occurs and a tool is used, the corresponding BTN_TOOL_ + code should be set to a value of 1. When the tool is no longer interacting + with the input device, the BTN_TOOL_ code should be reset to 0. All + trackpads, tablets, and touchscreens should use at least one BTN_TOOL_ + code when events are generated. + +* BTN_TOUCH: + BTN_TOUCH is used for touch contact. While an input tool is determined to be + within meaningful physical contact, the value of this property must be set + to 1. Meaningful physical contact may mean any contact, or it may mean + contact conditioned by an implementation defined property. For example, a + touchpad may set the value to 1 only when the touch pressure rises above a + certain value. BTN_TOUCH may be combined with BTN_TOOL_ codes. For + example, a pen tablet may set BTN_TOOL_PEN to 1 and BTN_TOUCH to 0 while the + pen is hovering over but not touching the tablet surface. + +Note: For appropriate function of the legacy mousedev emulation driver, +BTN_TOUCH must be the first evdev code emitted in a synchronization frame. + +Note: Historically a touch device with BTN_TOOL_FINGER and BTN_TOUCH was +interpreted as a touchpad by userspace, while a similar device without +BTN_TOOL_FINGER was interpreted as a touchscreen. For backwards compatibility +with current userspace it is recommended to follow this distinction. In the +future, this distinction will be deprecated and the device properties ioctl +EVIOCGPROP, defined in linux/input.h, will be used to convey the device type. + +* BTN_TOOL_FINGER, BTN_TOOL_DOUBLETAP, BTN_TOOL_TRIPLETAP, BTN_TOOL_QUADTAP: + - These codes denote one, two, three, and four finger interaction on a + trackpad or touchscreen. For example, if the user uses two fingers and moves + them on the touchpad in an effort to scroll content on screen, + BTN_TOOL_DOUBLETAP should be set to value 1 for the duration of the motion. + Note that all BTN_TOOL_ codes and the BTN_TOUCH code are orthogonal in + purpose. A trackpad event generated by finger touches should generate events + for one code from each group. At most only one of these BTN_TOOL_ + codes should have a value of 1 during any synchronization frame. + +Note: Historically some drivers emitted multiple of the finger count codes with +a value of 1 in the same synchronization frame. This usage is deprecated. + +Note: In multitouch drivers, the input_mt_report_finger_count() function should +be used to emit these codes. Please see multi-touch-protocol.txt for details. + +EV_REL: +---------- +EV_REL events describe relative changes in a property. For example, a mouse may +move to the left by a certain number of units, but its absolute position in +space is unknown. If the absolute position is known, EV_ABS codes should be used +instead of EV_REL codes. + +A few EV_REL codes have special meanings: + +* REL_WHEEL, REL_HWHEEL: + - These codes are used for vertical and horizontal scroll wheels, + respectively. + +EV_ABS: +---------- +EV_ABS events describe absolute changes in a property. For example, a touchpad +may emit coordinates for a touch location. + +A few EV_ABS codes have special meanings: + +* ABS_DISTANCE: + - Used to describe the distance of a tool from an interaction surface. This + event should only be emitted while the tool is hovering, meaning in close + proximity of the device and while the value of the BTN_TOUCH code is 0. If + the input device may be used freely in three dimensions, consider ABS_Z + instead. + +* ABS_MT_: + - Used to describe multitouch input events. Please see + multi-touch-protocol.txt for details. + +EV_SW: +---------- +EV_SW events describe stateful binary switches. For example, the SW_LID code is +used to denote when a laptop lid is closed. + +Upon binding to a device or resuming from suspend, a driver must report +the current switch state. This ensures that the device, kernel, and userspace +state is in sync. + +Upon resume, if the switch state is the same as before suspend, then the input +subsystem will filter out the duplicate switch state reports. The driver does +not need to keep the state of the switch at any time. + +EV_MSC: +---------- +EV_MSC events are used for input and output events that do not fall under other +categories. + +EV_LED: +---------- +EV_LED events are used for input and output to set and query the state of +various LEDs on devices. + +EV_REP: +---------- +EV_REP events are used for specifying autorepeating events. + +EV_SND: +---------- +EV_SND events are used for sending sound commands to simple sound output +devices. + +EV_FF: +---------- +EV_FF events are used to initialize a force feedback capable device and to cause +such device to feedback. + +EV_PWR: +---------- +EV_PWR events are a special type of event used specifically for power +mangement. Its usage is not well defined. To be addressed later. + +Guidelines: +========== +The guidelines below ensure proper single-touch and multi-finger functionality. +For multi-touch functionality, see the multi-touch-protocol.txt document for +more information. + +Mice: +---------- +REL_{X,Y} must be reported when the mouse moves. BTN_LEFT must be used to report +the primary button press. BTN_{MIDDLE,RIGHT,4,5,etc.} should be used to report +further buttons of the device. REL_WHEEL and REL_HWHEEL should be used to report +scroll wheel events where available. + +Touchscreens: +---------- +ABS_{X,Y} must be reported with the location of the touch. BTN_TOUCH must be +used to report when a touch is active on the screen. +BTN_{MOUSE,LEFT,MIDDLE,RIGHT} must not be reported as the result of touch +contact. BTN_TOOL_ events should be reported where possible. + +Trackpads: +---------- +Legacy trackpads that only provide relative position information must report +events like mice described above. + +Trackpads that provide absolute touch position must report ABS_{X,Y} for the +location of the touch. BTN_TOUCH should be used to report when a touch is active +on the trackpad. Where multi-finger support is available, BTN_TOOL_ should +be used to report the number of touches active on the trackpad. + +Tablets: +---------- +BTN_TOOL_ events must be reported when a stylus or other tool is active on +the tablet. ABS_{X,Y} must be reported with the location of the tool. BTN_TOUCH +should be used to report when the tool is in contact with the tablet. +BTN_{STYLUS,STYLUS2} should be used to report buttons on the tool itself. Any +button may be used for buttons on the tablet except BTN_{MOUSE,LEFT}. +BTN_{0,1,2,etc} are good generic codes for unlabeled buttons. Do not use +meaningful buttons, like BTN_FORWARD, unless the button is labeled for that +purpose on the device. -- cgit v0.10.2 From 9fb0f14e31b6101a0cc69a333b43541044f9b0a6 Mon Sep 17 00:00:00 2001 From: Jeff Brown Date: Tue, 12 Apr 2011 23:29:38 -0700 Subject: Input: evdev - indicate buffer overrun with SYN_DROPPED Add a new EV_SYN code, SYN_DROPPED, to inform the client when input events have been dropped from the evdev input buffer due to a buffer overrun. The client should use this event as a hint to reset its state or ignore all following events until the next packet begins. Signed-off-by: Jeff Brown [dtor@mail.ru: Implement Henrik's suggestion and drop old events in case of overflow.] Signed-off-by: Dmitry Torokhov diff --git a/Documentation/input/event-codes.txt b/Documentation/input/event-codes.txt index f13aee5..23fcb05 100644 --- a/Documentation/input/event-codes.txt +++ b/Documentation/input/event-codes.txt @@ -85,6 +85,12 @@ sent in the evdev event stream. - Used to synchronize and separate touch events. See the multi-touch-protocol.txt document for more information. +* SYN_DROPPED: + - Used to indicate buffer overrun in the evdev client's event queue. + Client should ignore all events up to and including next SYN_REPORT + event and query the device (using EVIOCG* ioctls) to obtain its + current state. + EV_KEY: ---------- EV_KEY events take the form KEY_ or BTN_. For example, KEY_A is used diff --git a/drivers/input/evdev.c b/drivers/input/evdev.c index 7f42d3a..88d8e4c 100644 --- a/drivers/input/evdev.c +++ b/drivers/input/evdev.c @@ -39,13 +39,13 @@ struct evdev { }; struct evdev_client { - int head; - int tail; + unsigned int head; + unsigned int tail; spinlock_t buffer_lock; /* protects access to buffer, head and tail */ struct fasync_struct *fasync; struct evdev *evdev; struct list_head node; - int bufsize; + unsigned int bufsize; struct input_event buffer[]; }; @@ -55,16 +55,25 @@ static DEFINE_MUTEX(evdev_table_mutex); static void evdev_pass_event(struct evdev_client *client, struct input_event *event) { - /* - * Interrupts are disabled, just acquire the lock. - * Make sure we don't leave with the client buffer - * "empty" by having client->head == client->tail. - */ + /* Interrupts are disabled, just acquire the lock. */ spin_lock(&client->buffer_lock); - do { - client->buffer[client->head++] = *event; - client->head &= client->bufsize - 1; - } while (client->head == client->tail); + + client->buffer[client->head++] = *event; + client->head &= client->bufsize - 1; + + if (unlikely(client->head == client->tail)) { + /* + * This effectively "drops" all unconsumed events, leaving + * EV_SYN/SYN_DROPPED plus the newest event in the queue. + */ + client->tail = (client->head - 2) & (client->bufsize - 1); + + client->buffer[client->tail].time = event->time; + client->buffer[client->tail].type = EV_SYN; + client->buffer[client->tail].code = SYN_DROPPED; + client->buffer[client->tail].value = 0; + } + spin_unlock(&client->buffer_lock); if (event->type == EV_SYN) diff --git a/include/linux/input.h b/include/linux/input.h index 0cc25e4..73a8c6e 100644 --- a/include/linux/input.h +++ b/include/linux/input.h @@ -167,6 +167,7 @@ struct input_keymap_entry { #define SYN_REPORT 0 #define SYN_CONFIG 1 #define SYN_MT_REPORT 2 +#define SYN_DROPPED 3 /* * Keys and buttons -- cgit v0.10.2 From 78530bf7f2559b317c04991b52217c1608d5a58d Mon Sep 17 00:00:00 2001 From: Artem Bityutskiy Date: Wed, 13 Apr 2011 10:31:52 +0300 Subject: UBIFS: fix oops when R/O file-system is fsync'ed This patch fixes severe UBIFS bug: UBIFS oopses when we 'fsync()' an file on R/O-mounter file-system. We (the UBIFS authors) incorrectly thought that VFS would not propagate 'fsync()' down to the file-system if it is read-only, but this is not the case. It is easy to exploit this bug using the following simple perl script: use strict; use File::Sync qw(fsync sync); die "File path is not specified" if not defined $ARGV[0]; my $path = $ARGV[0]; open FILE, "<", "$path" or die "Cannot open $path: $!"; fsync(\*FILE) or die "cannot fsync $path: $!"; close FILE or die "Cannot close $path: $!"; Thanks to Reuben Dowle for reporting about this issue. Signed-off-by: Artem Bityutskiy Reported-by: Reuben Dowle Cc: stable@kernel.org diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c index 28be1e6..b286db7 100644 --- a/fs/ubifs/file.c +++ b/fs/ubifs/file.c @@ -1312,6 +1312,9 @@ int ubifs_fsync(struct file *file, int datasync) dbg_gen("syncing inode %lu", inode->i_ino); + if (inode->i_sb->s_flags & MS_RDONLY) + return 0; + /* * VFS has already synchronized dirty pages for this inode. Synchronize * the inode unless this is a 'datasync()' call. -- cgit v0.10.2 From 100d4a9d20015315bf4215d11c2cf4b1f30c33b8 Mon Sep 17 00:00:00 2001 From: Mian Yousaf Kaukab Date: Tue, 15 Mar 2011 16:24:24 +0100 Subject: usb: musb: clear AUTOSET while clearing DMAENAB On the completion of tx dma, dma is disabled by clearing MUSB_TXCSR_DMAENAB in TXCSR. If MUSB_TXCSR_AUTOSET was set in txstate() it will remain set although it is not needed in PIO mode. Clear it as soon as it is not needed. Signed-off-by: Mian Yousaf Kaukab Signed-off-by: Felipe Balbi diff --git a/drivers/usb/musb/musb_gadget.c b/drivers/usb/musb/musb_gadget.c index 98519c5..9ecb057 100644 --- a/drivers/usb/musb/musb_gadget.c +++ b/drivers/usb/musb/musb_gadget.c @@ -535,7 +535,7 @@ void musb_g_tx(struct musb *musb, u8 epnum) is_dma = 1; csr |= MUSB_TXCSR_P_WZC_BITS; csr &= ~(MUSB_TXCSR_DMAENAB | MUSB_TXCSR_P_UNDERRUN | - MUSB_TXCSR_TXPKTRDY); + MUSB_TXCSR_TXPKTRDY | MUSB_TXCSR_AUTOSET); musb_writew(epio, MUSB_TXCSR, csr); /* Ensure writebuffer is empty. */ csr = musb_readw(epio, MUSB_TXCSR); -- cgit v0.10.2 From 8726606424738ca0341e1bb93ebac956d80f6d29 Mon Sep 17 00:00:00 2001 From: Mian Yousaf Kaukab Date: Tue, 15 Mar 2011 16:24:29 +0100 Subject: usb: musb: ux500: copy dma mask from platform device to musb device musb code checks dma mask before calling dma hooks. Signed-off-by: Mian Yousaf Kaukab Signed-off-by: Felipe Balbi diff --git a/drivers/usb/musb/ux500.c b/drivers/usb/musb/ux500.c index d6384e4..f7e04bf 100644 --- a/drivers/usb/musb/ux500.c +++ b/drivers/usb/musb/ux500.c @@ -93,6 +93,8 @@ static int __init ux500_probe(struct platform_device *pdev) } musb->dev.parent = &pdev->dev; + musb->dev.dma_mask = pdev->dev.dma_mask; + musb->dev.coherent_dma_mask = pdev->dev.coherent_dma_mask; glue->dev = &pdev->dev; glue->musb = musb; -- cgit v0.10.2 From ec63bf6c06b01ceeb6048a2b9fa9e73060259307 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sun, 20 Mar 2011 14:14:36 +0300 Subject: USB: musb: add missing unlock in cppi_interrupt() We should unlock before returning here. Signed-off-by: Dan Carpenter Signed-off-by: Felipe Balbi diff --git a/drivers/usb/musb/cppi_dma.c b/drivers/usb/musb/cppi_dma.c index de55a3c..6385eeb 100644 --- a/drivers/usb/musb/cppi_dma.c +++ b/drivers/usb/musb/cppi_dma.c @@ -1167,8 +1167,11 @@ irqreturn_t cppi_interrupt(int irq, void *dev_id) tx = musb_readl(tibase, DAVINCI_TXCPPI_MASKED_REG); rx = musb_readl(tibase, DAVINCI_RXCPPI_MASKED_REG); - if (!tx && !rx) + if (!tx && !rx) { + if (cppi->irq) + spin_unlock_irqrestore(&musb->lock, flags); return IRQ_NONE; + } DBG(4, "CPPI IRQ Tx%x Rx%x\n", tx, rx); -- cgit v0.10.2 From aca7f353219abfb7b8a1530fbba1b1acf0e30da4 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sun, 20 Mar 2011 14:15:24 +0300 Subject: USB: musb: using 0 instead of NULL Sparse complains (and rightly so): drivers/usb/musb/cppi_dma.c:1458:33: warning: Using plain integer as NULL pointer Signed-off-by: Dan Carpenter Signed-off-by: Felipe Balbi diff --git a/drivers/usb/musb/cppi_dma.c b/drivers/usb/musb/cppi_dma.c index 6385eeb..0db0f5e 100644 --- a/drivers/usb/musb/cppi_dma.c +++ b/drivers/usb/musb/cppi_dma.c @@ -1455,7 +1455,7 @@ static int cppi_channel_abort(struct dma_channel *channel) * compare mode by writing 1 to the tx_complete register. */ cppi_reset_tx(tx_ram, 1); - cppi_ch->head = 0; + cppi_ch->head = NULL; musb_writel(&tx_ram->tx_complete, 0, 1); cppi_dump_tx(5, cppi_ch, " (done teardown)"); -- cgit v0.10.2 From 2fbcf3fa43af809ebf4e4ad33c2f0a17e903385c Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sun, 20 Mar 2011 14:16:17 +0300 Subject: USB: musb: silence printk format warning MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Gcc gives the following warnings: drivers/usb/musb/cppi_dma.c: In function ‘cppi_next_tx_segment’: drivers/usb/musb/cppi_dma.c:600: warning: format ‘%x’ expects type ‘unsigned int’, but argument 8 has type ‘dma_addr_t’ drivers/usb/musb/cppi_dma.c: In function ‘cppi_next_rx_segment’: drivers/usb/musb/cppi_dma.c:822: warning: format ‘%x’ expects type ‘unsigned int’, but argument 9 has type ‘dma_addr_t’ drivers/usb/musb/cppi_dma.c: In function ‘cppi_rx_scan’: drivers/usb/musb/cppi_dma.c:1042: warning: format ‘%08x’ expects type ‘unsigned int’, but argument 4 has type ‘dma_addr_t’ drivers/usb/musb/cppi_dma.c:1114: warning: format ‘%08x’ expects type ‘unsigned int’, but argument 7 has type ‘dma_addr_t’ dma_addr_t is sometimes 32 bit and sometimes 64. We normally cast them to unsigned long long for printk(). Signed-off-by: Dan Carpenter Signed-off-by: Felipe Balbi diff --git a/drivers/usb/musb/cppi_dma.c b/drivers/usb/musb/cppi_dma.c index 0db0f5e..eb9161a 100644 --- a/drivers/usb/musb/cppi_dma.c +++ b/drivers/usb/musb/cppi_dma.c @@ -597,12 +597,12 @@ cppi_next_tx_segment(struct musb *musb, struct cppi_channel *tx) length = min(n_bds * maxpacket, length); } - DBG(4, "TX DMA%d, pktSz %d %s bds %d dma 0x%x len %u\n", + DBG(4, "TX DMA%d, pktSz %d %s bds %d dma 0x%llx len %u\n", tx->index, maxpacket, rndis ? "rndis" : "transparent", n_bds, - addr, length); + (unsigned long long)addr, length); cppi_rndis_update(tx, 0, musb->ctrl_base, rndis); @@ -820,7 +820,7 @@ cppi_next_rx_segment(struct musb *musb, struct cppi_channel *rx, int onepacket) length = min(n_bds * maxpacket, length); DBG(4, "RX DMA%d seg, maxp %d %s bds %d (cnt %d) " - "dma 0x%x len %u %u/%u\n", + "dma 0x%llx len %u %u/%u\n", rx->index, maxpacket, onepacket ? (is_rndis ? "rndis" : "onepacket") @@ -829,7 +829,8 @@ cppi_next_rx_segment(struct musb *musb, struct cppi_channel *rx, int onepacket) musb_readl(tibase, DAVINCI_RXCPPI_BUFCNT0_REG + (rx->index * 4)) & 0xffff, - addr, length, rx->channel.actual_len, rx->buf_len); + (unsigned long long)addr, length, + rx->channel.actual_len, rx->buf_len); /* only queue one segment at a time, since the hardware prevents * correct queue shutdown after unexpected short packets @@ -1039,9 +1040,9 @@ static bool cppi_rx_scan(struct cppi *cppi, unsigned ch) if (!completed && (bd->hw_options & CPPI_OWN_SET)) break; - DBG(5, "C/RXBD %08x: nxt %08x buf %08x " + DBG(5, "C/RXBD %llx: nxt %08x buf %08x " "off.len %08x opt.len %08x (%d)\n", - bd->dma, bd->hw_next, bd->hw_bufp, + (unsigned long long)bd->dma, bd->hw_next, bd->hw_bufp, bd->hw_off_len, bd->hw_options, rx->channel.actual_len); @@ -1111,11 +1112,12 @@ static bool cppi_rx_scan(struct cppi *cppi, unsigned ch) musb_ep_select(cppi->mregs, rx->index + 1); csr = musb_readw(regs, MUSB_RXCSR); if (csr & MUSB_RXCSR_DMAENAB) { - DBG(4, "list%d %p/%p, last %08x%s, csr %04x\n", + DBG(4, "list%d %p/%p, last %llx%s, csr %04x\n", rx->index, rx->head, rx->tail, rx->last_processed - ? rx->last_processed->dma + ? (unsigned long long) + rx->last_processed->dma : 0, completed ? ", completed" : "", csr); -- cgit v0.10.2 From 2e10f5e70f670d981f789075e3ebc394f5bb51e3 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sun, 20 Mar 2011 14:18:26 +0300 Subject: USB: musb: dereferencing an iomem pointer "tx_ram" points to io memory. We can't dereference it directly. Sparse complains about this: "drivers/usb/musb/cppi_dma.c:1205:25: warning: dereference of noderef expression" Signed-off-by: Dan Carpenter Signed-off-by: Felipe Balbi diff --git a/drivers/usb/musb/cppi_dma.c b/drivers/usb/musb/cppi_dma.c index eb9161a..ab434fb 100644 --- a/drivers/usb/musb/cppi_dma.c +++ b/drivers/usb/musb/cppi_dma.c @@ -1204,7 +1204,7 @@ irqreturn_t cppi_interrupt(int irq, void *dev_id) */ if (NULL == bd) { DBG(1, "null BD\n"); - tx_ram->tx_complete = 0; + musb_writel(&tx_ram->tx_complete, 0, 0); continue; } -- cgit v0.10.2 From 7a180e70cfc56e131bfe4796773df2acfc7d4180 Mon Sep 17 00:00:00 2001 From: Felipe Balbi Date: Tue, 22 Mar 2011 11:31:37 +0200 Subject: usb: musb: temporarily make it bool Due to the recent changes to musb's glue layers, we can't compile musb-hdrc as a module - compilation will break due to undefined symbol musb_debug. In order to fix that, we need a big re-work of the debug support on the MUSB driver. Because that would mean a lot of new code coming into the -rc series, it's best to defer that to next merge window and for now just disable module support for MUSB. Once we get the refactor of the debugging support done, we can simply revert this patch and things will go back to normal again. Cc: stable@kernel.org # v2.6.38 Signed-off-by: Felipe Balbi diff --git a/drivers/usb/musb/Kconfig b/drivers/usb/musb/Kconfig index 4cbb7e4..74073b3 100644 --- a/drivers/usb/musb/Kconfig +++ b/drivers/usb/musb/Kconfig @@ -14,7 +14,7 @@ config USB_MUSB_HDRC select TWL4030_USB if MACH_OMAP_3430SDP select TWL6030_USB if MACH_OMAP_4430SDP || MACH_OMAP4_PANDA select USB_OTG_UTILS - tristate 'Inventra Highspeed Dual Role Controller (TI, ADI, ...)' + bool 'Inventra Highspeed Dual Role Controller (TI, ADI, ...)' help Say Y here if your system has a dual role high speed USB controller based on the Mentor Graphics silicon IP. Then @@ -30,8 +30,8 @@ config USB_MUSB_HDRC If you do not know what this is, please say N. - To compile this driver as a module, choose M here; the - module will be called "musb-hdrc". +# To compile this driver as a module, choose M here; the +# module will be called "musb-hdrc". choice prompt "Platform Glue Layer" -- cgit v0.10.2 From 3d5ad13eac320292f64071ea7ded1b661edd9430 Mon Sep 17 00:00:00 2001 From: Felipe Balbi Date: Tue, 22 Mar 2011 11:38:49 +0200 Subject: usb: musb: gadget: check the correct list_head We are now using our own list_head, so we should be checking against that, not the gadget driver's list_head. Signed-off-by: Felipe Balbi diff --git a/drivers/usb/musb/musb_gadget.c b/drivers/usb/musb/musb_gadget.c index 9ecb057..6dfbf9f 100644 --- a/drivers/usb/musb/musb_gadget.c +++ b/drivers/usb/musb/musb_gadget.c @@ -1296,7 +1296,7 @@ static int musb_gadget_dequeue(struct usb_ep *ep, struct usb_request *request) } /* if the hardware doesn't have the request, easy ... */ - if (musb_ep->req_list.next != &request->list || musb_ep->busy) + if (musb_ep->req_list.next != &req->list || musb_ep->busy) musb_g_giveback(musb_ep, request, -ECONNRESET); /* ... else abort the dma transfer ... */ -- cgit v0.10.2 From 4f9edd2d7e8dd170d10780532cb76eb5890468b4 Mon Sep 17 00:00:00 2001 From: Hema HK Date: Tue, 22 Mar 2011 16:02:12 +0530 Subject: usb: musb: Fix the crash issue during reboot Below crash observed with commit 7acc6197b76edd0b932a7cbcc6cfad0a8a87f026 (usb: musb: Idle path retention and offmode support for OMAP3) during board reboot. The musb clock was disabled when musb_shutdown() was called by platform_drv_shutdown in which there are register accesses. call pm_runtime_get_sync() and pm_runtime_put_sync() in the musb_shutdown function. / # [ 172.368774] Unhandled fault: imprecise external abort (0x1406) at 0x400f0000 [ 172.376190] Internal error: : 1406 [#1] SMP [ 172.380554] last sysfs file: /sys/devices/platform/omap/omap_i2c.4/i2c-4/i2c-dev/i2c-4/dev [ 172.389221] Modules linked in: [ 172.392456] CPU: 0 Tainted: G W (2.6.38-06671-geddecbb #33) [ 172.399475] PC is at do_raw_spin_unlock+0x50/0xc0 [ 172.404418] LR is at _raw_spin_unlock_irqrestore+0x24/0x44 [ 172.410186] pc : [] lr : [] psr: 60000093 [ 172.410186] sp : ee993e40 ip : c0d00240 fp : bea9cf14 [ 172.422241] r10: 00000000 r9 : ee992000 r8 : c04b2fa8 [ 172.427703] r7 : 00000000 r6 : c0fa46c0 r5 : ef966124 r4 : ef966124 [ 172.434539] r3 : ef92cbc0 r2 : ef92cbc0 r1 : 00000000 r0 : ef966124 [ 172.441406] Flags: nZCv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user [ 172.448974] Control: 10c5387d Table: ae8d804a DAC: 00000015 [ 172.454986] Process init (pid: 1094, stack limit = 0xee9922f8) [ 172.461120] Stack: (0xee993e40 to 0xee994000) [ 172.465667] 3e40: a0000013 c085a7f8 ef966124 a0000013 c0fa46c0 c0761ab4 c0761a70 ef95c008 [ 172.474273] 3e60: ef95c014 c06e2fd0 c06e2fbc c06dea90 00000000 01234567 28121969 c04fccb4 [ 172.482849] 3e80: 00000000 c04fcd04 c0a302bc c04fce44 c0a02600 00000001 00000000 c085cd04 [ 172.491424] 3ea0: 00000000 00000002 c09ea000 c085afc0 ee993f24 00000000 00040001 00000445 [ 172.499999] 3ec0: a8eb9d34 00000027 00000000 00000000 00000000 c0a56a4c 00000000 00000000 [ 172.508575] 3ee0: 00000002 60000093 00000000 c0519aac 00000002 00000080 00000000 c0550420 [ 172.517150] 3f00: 00000000 00000002 ee970000 c0a56a3c c0a56a20 00000002 c0a56a3c 00000000 [ 172.525726] 3f20: c0a56a3c 0000000a c1580e00 c0a56a20 00000002 c0a56a3c c1580e00 c0a56a20 [ 172.534301] 3f40: ef92cbc0 c05173a0 00000001 ef92cbc0 c0576190 c04e3174 20000013 c0517324 [ 172.542877] 3f60: ef815c00 ee90b720 c04e3174 c0576190 00000001 ef92cbc0 c04b2f00 ffffffff [ 172.551483] 3f80: 00000058 c0517324 00000000 00000000 ffffffff 00000000 00000000 ffffffff [ 172.560058] 3fa0: 00000058 c04b2de0 00000000 00000000 fee1dead 28121969 01234567 00000000 [ 172.568634] 3fc0: 00000000 00000000 ffffffff 00000058 00000000 00000001 400aa000 bea9cf14 [ 172.577209] 3fe0: 000ea148 bea9c958 000aa750 40225728 60000010 fee1dead 00000000 00000000 [ 172.585784] [] (do_raw_spin_unlock+0x50/0xc0) from [] (_raw_spin_unlock_irqrestore+0x24/0x44) [ 172.596588] [] (_raw_spin_unlock_irqrestore+0x24/0x44) from [] (musb_shutdown+0x44/0x88) [ 172.606933] [] (musb_shutdown+0x44/0x88) from [] (platform_drv_shutdown+0x14/0x18) [ 172.616699] [] (platform_drv_shutdown+0x14/0x18) from [] (device_shutdown+0x74/0xb4) [ 172.626647] [] (device_shutdown+0x74/0xb4) from [] (kernel_restart_prepare+0x24/0x38) [ 172.636688] [] (kernel_restart_prepare+0x24/0x38) from [] (kernel_restart+0xc/0x48) [ 172.646545] [] (kernel_restart+0xc/0x48) from [] (sys_reboot+0xfc/0x1d8) [ 172.655426] [] (sys_reboot+0xfc/0x1d8) from [] (ret_fast_syscall+0x0/0x3c) [ 172.664459] Code: e3c3303f e594200c e593300c e1520003 (0a000002) [ 172.670867] ------------[ cut here ]------------ Signed-off-by: Hema HK Signed-off-by: Felipe Balbi diff --git a/drivers/usb/musb/musb_core.c b/drivers/usb/musb/musb_core.c index 630ae7f..f10ff00 100644 --- a/drivers/usb/musb/musb_core.c +++ b/drivers/usb/musb/musb_core.c @@ -1030,6 +1030,7 @@ static void musb_shutdown(struct platform_device *pdev) struct musb *musb = dev_to_musb(&pdev->dev); unsigned long flags; + pm_runtime_get_sync(musb->controller); spin_lock_irqsave(&musb->lock, flags); musb_platform_disable(musb); musb_generic_disable(musb); @@ -1040,6 +1041,7 @@ static void musb_shutdown(struct platform_device *pdev) musb_writeb(musb->mregs, MUSB_DEVCTL, 0); musb_platform_exit(musb); + pm_runtime_put(musb->controller); /* FIXME power down */ } -- cgit v0.10.2 From 132543074af3cf1e94e3608abf162880edbdcbb3 Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Wed, 30 Mar 2011 22:48:54 -0400 Subject: USB: musb: blackfin: work around anomaly 05000450 DMA mode 1 data corruption anomaly on Blackfin systems. This issue is specific to the Blackfin silicon as the bug appears to be related to the connection of the musb ip to the bus/dma fabric. Data corruption when using USB DMA mode 1. (Issue manager 17-01-0105) DMA mode 1 allows large size transfers to generate a single interrupt at the end of the entire transfer. The transfer is split up in packets of length specified in the Maximum Packet Size field for that endpoint. If the transfer size is not an integer multiple of the Maximum Packet Size, a short packet will be present at the end of the transfer. Under certain conditions this packet may be corrupted in the USB FIFO. Workaround: Use DMA mode 1 to transfer (n* Maximum Packet Size) and schedule DMA mode 0 to transfer the short packet. As an example if your transfer size is 33168 bytes and Maximum Packet Size equals 512, schedule [33168 - (33168 mod 512)] in DMA mode 1 and the remainder (33168 mod 512) in DMA mode 0. Signed-off-by: Mike Frysinger Signed-off-by: Felipe Balbi diff --git a/drivers/usb/musb/blackfin.c b/drivers/usb/musb/blackfin.c index 52312e8..8e2a1ff 100644 --- a/drivers/usb/musb/blackfin.c +++ b/drivers/usb/musb/blackfin.c @@ -21,6 +21,7 @@ #include #include "musb_core.h" +#include "musbhsdma.h" #include "blackfin.h" struct bfin_glue { @@ -332,6 +333,27 @@ static int bfin_musb_set_mode(struct musb *musb, u8 musb_mode) return -EIO; } +static int bfin_musb_adjust_channel_params(struct dma_channel *channel, + u16 packet_sz, u8 *mode, + dma_addr_t *dma_addr, u32 *len) +{ + struct musb_dma_channel *musb_channel = channel->private_data; + + /* + * Anomaly 05000450 might cause data corruption when using DMA + * MODE 1 transmits with short packet. So to work around this, + * we truncate all MODE 1 transfers down to a multiple of the + * max packet size, and then do the last short packet transfer + * (if there is any) using MODE 0. + */ + if (ANOMALY_05000450) { + if (musb_channel->transmit && *mode == 1) + *len = *len - (*len % packet_sz); + } + + return 0; +} + static void bfin_musb_reg_init(struct musb *musb) { if (ANOMALY_05000346) { @@ -430,6 +452,8 @@ static const struct musb_platform_ops bfin_ops = { .vbus_status = bfin_musb_vbus_status, .set_vbus = bfin_musb_set_vbus, + + .adjust_channel_params = bfin_musb_adjust_channel_params, }; static u64 bfin_dmamask = DMA_BIT_MASK(32); diff --git a/drivers/usb/musb/musb_core.h b/drivers/usb/musb/musb_core.h index 4bd9e21..0e053b5 100644 --- a/drivers/usb/musb/musb_core.h +++ b/drivers/usb/musb/musb_core.h @@ -261,6 +261,7 @@ enum musb_g_ep0_state { * @try_ilde: tries to idle the IP * @vbus_status: returns vbus status if possible * @set_vbus: forces vbus status + * @channel_program: pre check for standard dma channel_program func */ struct musb_platform_ops { int (*init)(struct musb *musb); @@ -274,6 +275,10 @@ struct musb_platform_ops { int (*vbus_status)(struct musb *musb); void (*set_vbus)(struct musb *musb, int on); + + int (*adjust_channel_params)(struct dma_channel *channel, + u16 packet_sz, u8 *mode, + dma_addr_t *dma_addr, u32 *len); }; /* diff --git a/drivers/usb/musb/musbhsdma.c b/drivers/usb/musb/musbhsdma.c index 0144a2d..d281792 100644 --- a/drivers/usb/musb/musbhsdma.c +++ b/drivers/usb/musb/musbhsdma.c @@ -169,6 +169,14 @@ static int dma_channel_program(struct dma_channel *channel, BUG_ON(channel->status == MUSB_DMA_STATUS_UNKNOWN || channel->status == MUSB_DMA_STATUS_BUSY); + /* Let targets check/tweak the arguments */ + if (musb->ops->adjust_channel_params) { + int ret = musb->ops->adjust_channel_params(channel, + packet_sz, &mode, &dma_addr, &len); + if (ret) + return ret; + } + /* * The DMA engine in RTL1.8 and above cannot handle * DMA addresses that are not aligned to a 4 byte boundary. -- cgit v0.10.2 From 1dcffad74183bb00e8129ba1c5bb2c9931d31bd7 Mon Sep 17 00:00:00 2001 From: Maksim Rayskiy Date: Tue, 12 Apr 2011 15:14:56 -0700 Subject: UBIFS: fix compilation warnings when compiling with gcc 4.5 When compiling UBIFS with CONFIG_UBIFS_FS_DEBUG not set, gcc-4.5.2 generates a slew of "warning: statement with no effect" on references to non-void functions defined as 0. To avoid these warnings, replace #defines with dummy inline functions. Artem: massage the patch a bit, also remove the duplicate 'dbg_check_lprops()' prototype. Signed-off-by: Maksim Rayskiy Acked-by: Mike Frysinger Signed-off-by: Artem Bityutskiy diff --git a/fs/ubifs/debug.h b/fs/ubifs/debug.h index 919f0de..e6493ca 100644 --- a/fs/ubifs/debug.h +++ b/fs/ubifs/debug.h @@ -23,6 +23,12 @@ #ifndef __UBIFS_DEBUG_H__ #define __UBIFS_DEBUG_H__ +/* Checking helper functions */ +typedef int (*dbg_leaf_callback)(struct ubifs_info *c, + struct ubifs_zbranch *zbr, void *priv); +typedef int (*dbg_znode_callback)(struct ubifs_info *c, + struct ubifs_znode *znode, void *priv); + #ifdef CONFIG_UBIFS_FS_DEBUG /** @@ -270,11 +276,6 @@ void dbg_dump_tnc(struct ubifs_info *c); void dbg_dump_index(struct ubifs_info *c); void dbg_dump_lpt_lebs(const struct ubifs_info *c); -/* Checking helper functions */ -typedef int (*dbg_leaf_callback)(struct ubifs_info *c, - struct ubifs_zbranch *zbr, void *priv); -typedef int (*dbg_znode_callback)(struct ubifs_info *c, - struct ubifs_znode *znode, void *priv); int dbg_walk_index(struct ubifs_info *c, dbg_leaf_callback leaf_cb, dbg_znode_callback znode_cb, void *priv); @@ -295,7 +296,6 @@ int dbg_check_idx_size(struct ubifs_info *c, long long idx_size); int dbg_check_filesystem(struct ubifs_info *c); void dbg_check_heap(struct ubifs_info *c, struct ubifs_lpt_heap *heap, int cat, int add_pos); -int dbg_check_lprops(struct ubifs_info *c); int dbg_check_lpt_nodes(struct ubifs_info *c, struct ubifs_cnode *cnode, int row, int col); int dbg_check_inode_size(struct ubifs_info *c, const struct inode *inode, @@ -401,58 +401,94 @@ void dbg_debugfs_exit_fs(struct ubifs_info *c); #define DBGKEY(key) ((char *)(key)) #define DBGKEY1(key) ((char *)(key)) -#define ubifs_debugging_init(c) 0 -#define ubifs_debugging_exit(c) ({}) - -#define dbg_ntype(type) "" -#define dbg_cstate(cmt_state) "" -#define dbg_jhead(jhead) "" -#define dbg_get_key_dump(c, key) ({}) -#define dbg_dump_inode(c, inode) ({}) -#define dbg_dump_node(c, node) ({}) -#define dbg_dump_lpt_node(c, node, lnum, offs) ({}) -#define dbg_dump_budget_req(req) ({}) -#define dbg_dump_lstats(lst) ({}) -#define dbg_dump_budg(c) ({}) -#define dbg_dump_lprop(c, lp) ({}) -#define dbg_dump_lprops(c) ({}) -#define dbg_dump_lpt_info(c) ({}) -#define dbg_dump_leb(c, lnum) ({}) -#define dbg_dump_znode(c, znode) ({}) -#define dbg_dump_heap(c, heap, cat) ({}) -#define dbg_dump_pnode(c, pnode, parent, iip) ({}) -#define dbg_dump_tnc(c) ({}) -#define dbg_dump_index(c) ({}) -#define dbg_dump_lpt_lebs(c) ({}) - -#define dbg_walk_index(c, leaf_cb, znode_cb, priv) 0 -#define dbg_old_index_check_init(c, zroot) 0 -#define dbg_save_space_info(c) ({}) -#define dbg_check_space_info(c) 0 -#define dbg_check_old_index(c, zroot) 0 -#define dbg_check_cats(c) 0 -#define dbg_check_ltab(c) 0 -#define dbg_chk_lpt_free_spc(c) 0 -#define dbg_chk_lpt_sz(c, action, len) 0 -#define dbg_check_synced_i_size(inode) 0 -#define dbg_check_dir_size(c, dir) 0 -#define dbg_check_tnc(c, x) 0 -#define dbg_check_idx_size(c, idx_size) 0 -#define dbg_check_filesystem(c) 0 -#define dbg_check_heap(c, heap, cat, add_pos) ({}) -#define dbg_check_lprops(c) 0 -#define dbg_check_lpt_nodes(c, cnode, row, col) 0 -#define dbg_check_inode_size(c, inode, size) 0 -#define dbg_check_data_nodes_order(c, head) 0 -#define dbg_check_nondata_nodes_order(c, head) 0 -#define dbg_force_in_the_gaps_enabled 0 -#define dbg_force_in_the_gaps() 0 -#define dbg_failure_mode 0 - -#define dbg_debugfs_init() 0 -#define dbg_debugfs_exit() -#define dbg_debugfs_init_fs(c) 0 -#define dbg_debugfs_exit_fs(c) 0 +static inline int ubifs_debugging_init(struct ubifs_info *c) { return 0; } +static inline void ubifs_debugging_exit(struct ubifs_info *c) { return; } +static inline const char *dbg_ntype(int type) { return ""; } +static inline const char *dbg_cstate(int cmt_state) { return ""; } +static inline const char *dbg_jhead(int jhead) { return ""; } +static inline const char * +dbg_get_key_dump(const struct ubifs_info *c, + const union ubifs_key *key) { return ""; } +static inline void dbg_dump_inode(const struct ubifs_info *c, + const struct inode *inode) { return; } +static inline void dbg_dump_node(const struct ubifs_info *c, + const void *node) { return; } +static inline void dbg_dump_lpt_node(const struct ubifs_info *c, + void *node, int lnum, + int offs) { return; } +static inline void +dbg_dump_budget_req(const struct ubifs_budget_req *req) { return; } +static inline void +dbg_dump_lstats(const struct ubifs_lp_stats *lst) { return; } +static inline void dbg_dump_budg(struct ubifs_info *c) { return; } +static inline void dbg_dump_lprop(const struct ubifs_info *c, + const struct ubifs_lprops *lp) { return; } +static inline void dbg_dump_lprops(struct ubifs_info *c) { return; } +static inline void dbg_dump_lpt_info(struct ubifs_info *c) { return; } +static inline void dbg_dump_leb(const struct ubifs_info *c, + int lnum) { return; } +static inline void +dbg_dump_znode(const struct ubifs_info *c, + const struct ubifs_znode *znode) { return; } +static inline void dbg_dump_heap(struct ubifs_info *c, + struct ubifs_lpt_heap *heap, + int cat) { return; } +static inline void dbg_dump_pnode(struct ubifs_info *c, + struct ubifs_pnode *pnode, + struct ubifs_nnode *parent, + int iip) { return; } +static inline void dbg_dump_tnc(struct ubifs_info *c) { return; } +static inline void dbg_dump_index(struct ubifs_info *c) { return; } +static inline void dbg_dump_lpt_lebs(const struct ubifs_info *c) { return; } + +static inline int dbg_walk_index(struct ubifs_info *c, + dbg_leaf_callback leaf_cb, + dbg_znode_callback znode_cb, + void *priv) { return 0; } +static inline void dbg_save_space_info(struct ubifs_info *c) { return; } +static inline int dbg_check_space_info(struct ubifs_info *c) { return 0; } +static inline int dbg_check_lprops(struct ubifs_info *c) { return 0; } +static inline int +dbg_old_index_check_init(struct ubifs_info *c, + struct ubifs_zbranch *zroot) { return 0; } +static inline int +dbg_check_old_index(struct ubifs_info *c, + struct ubifs_zbranch *zroot) { return 0; } +static inline int dbg_check_cats(struct ubifs_info *c) { return 0; } +static inline int dbg_check_ltab(struct ubifs_info *c) { return 0; } +static inline int dbg_chk_lpt_free_spc(struct ubifs_info *c) { return 0; } +static inline int dbg_chk_lpt_sz(struct ubifs_info *c, + int action, int len) { return 0; } +static inline int dbg_check_synced_i_size(struct inode *inode) { return 0; } +static inline int dbg_check_dir_size(struct ubifs_info *c, + const struct inode *dir) { return 0; } +static inline int dbg_check_tnc(struct ubifs_info *c, int extra) { return 0; } +static inline int dbg_check_idx_size(struct ubifs_info *c, + long long idx_size) { return 0; } +static inline int dbg_check_filesystem(struct ubifs_info *c) { return 0; } +static inline void dbg_check_heap(struct ubifs_info *c, + struct ubifs_lpt_heap *heap, + int cat, int add_pos) { return; } +static inline int dbg_check_lpt_nodes(struct ubifs_info *c, + struct ubifs_cnode *cnode, int row, int col) { return 0; } +static inline int dbg_check_inode_size(struct ubifs_info *c, + const struct inode *inode, + loff_t size) { return 0; } +static inline int +dbg_check_data_nodes_order(struct ubifs_info *c, + struct list_head *head) { return 0; } +static inline int +dbg_check_nondata_nodes_order(struct ubifs_info *c, + struct list_head *head) { return 0; } + +static inline int dbg_force_in_the_gaps(void) { return 0; } +#define dbg_force_in_the_gaps_enabled 0 +#define dbg_failure_mode 0 + +static inline int dbg_debugfs_init(void) { return 0; } +static inline void dbg_debugfs_exit(void) { return; } +static inline int dbg_debugfs_init_fs(struct ubifs_info *c) { return 0; } +static inline int dbg_debugfs_exit_fs(struct ubifs_info *c) { return 0; } #endif /* !CONFIG_UBIFS_FS_DEBUG */ #endif /* !__UBIFS_DEBUG_H__ */ -- cgit v0.10.2 From 0e8a835aa59d08d702af0fcfd296e2218b2e344b Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Wed, 13 Apr 2011 13:43:23 +0200 Subject: netfilter: ipset: bitmap:ip,mac type requires "src" for MAC Enforce that the second "src/dst" parameter of the set match and SET target must be "src", because we have access to the source MAC only in the packet. The previous behaviour, that the type required the second parameter but actually ignored the value was counter-intuitive and confusing. Signed-off-by: Jozsef Kadlecsik Signed-off-by: Patrick McHardy diff --git a/net/netfilter/ipset/ip_set_bitmap_ipmac.c b/net/netfilter/ipset/ip_set_bitmap_ipmac.c index 00a3324..a274300 100644 --- a/net/netfilter/ipset/ip_set_bitmap_ipmac.c +++ b/net/netfilter/ipset/ip_set_bitmap_ipmac.c @@ -343,6 +343,10 @@ bitmap_ipmac_kadt(struct ip_set *set, const struct sk_buff *skb, ipset_adtfn adtfn = set->variant->adt[adt]; struct ipmac data; + /* MAC can be src only */ + if (!(flags & IPSET_DIM_TWO_SRC)) + return 0; + data.id = ntohl(ip4addr(skb, flags & IPSET_DIM_ONE_SRC)); if (data.id < map->first_ip || data.id > map->last_ip) return -IPSET_ERR_BITMAP_RANGE; -- cgit v0.10.2 From eafbd3fde6fc5ada0d61307367e408813b04928a Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Wed, 13 Apr 2011 13:45:57 +0200 Subject: netfilter: ipset: set match and SET target fixes The SET target with --del-set did not work due to using wrongly the internal dimension of --add-set instead of --del-set. Also, the checkentries did not release the set references when returned an error. Bugs reported by Lennert Buytenhek. Signed-off-by: Jozsef Kadlecsik Signed-off-by: Patrick McHardy diff --git a/net/netfilter/xt_set.c b/net/netfilter/xt_set.c index 061d48c..b3babae 100644 --- a/net/netfilter/xt_set.c +++ b/net/netfilter/xt_set.c @@ -81,6 +81,7 @@ set_match_v0_checkentry(const struct xt_mtchk_param *par) if (info->match_set.u.flags[IPSET_DIM_MAX-1] != 0) { pr_warning("Protocol error: set match dimension " "is over the limit!\n"); + ip_set_nfnl_put(info->match_set.index); return -ERANGE; } @@ -135,6 +136,8 @@ set_target_v0_checkentry(const struct xt_tgchk_param *par) if (index == IPSET_INVALID_ID) { pr_warning("Cannot find del_set index %u as target\n", info->del_set.index); + if (info->add_set.index != IPSET_INVALID_ID) + ip_set_nfnl_put(info->add_set.index); return -ENOENT; } } @@ -142,6 +145,10 @@ set_target_v0_checkentry(const struct xt_tgchk_param *par) info->del_set.u.flags[IPSET_DIM_MAX-1] != 0) { pr_warning("Protocol error: SET target dimension " "is over the limit!\n"); + if (info->add_set.index != IPSET_INVALID_ID) + ip_set_nfnl_put(info->add_set.index); + if (info->del_set.index != IPSET_INVALID_ID) + ip_set_nfnl_put(info->del_set.index); return -ERANGE; } @@ -192,6 +199,7 @@ set_match_checkentry(const struct xt_mtchk_param *par) if (info->match_set.dim > IPSET_DIM_MAX) { pr_warning("Protocol error: set match dimension " "is over the limit!\n"); + ip_set_nfnl_put(info->match_set.index); return -ERANGE; } @@ -219,7 +227,7 @@ set_target(struct sk_buff *skb, const struct xt_action_param *par) if (info->del_set.index != IPSET_INVALID_ID) ip_set_del(info->del_set.index, skb, par->family, - info->add_set.dim, + info->del_set.dim, info->del_set.flags); return XT_CONTINUE; @@ -245,13 +253,19 @@ set_target_checkentry(const struct xt_tgchk_param *par) if (index == IPSET_INVALID_ID) { pr_warning("Cannot find del_set index %u as target\n", info->del_set.index); + if (info->add_set.index != IPSET_INVALID_ID) + ip_set_nfnl_put(info->add_set.index); return -ENOENT; } } if (info->add_set.dim > IPSET_DIM_MAX || - info->del_set.flags > IPSET_DIM_MAX) { + info->del_set.dim > IPSET_DIM_MAX) { pr_warning("Protocol error: SET target dimension " "is over the limit!\n"); + if (info->add_set.index != IPSET_INVALID_ID) + ip_set_nfnl_put(info->add_set.index); + if (info->del_set.index != IPSET_INVALID_ID) + ip_set_nfnl_put(info->del_set.index); return -ERANGE; } -- cgit v0.10.2 From 8faf9e3838c31c426e6203cb7dc6fc0f0bdb8b7e Mon Sep 17 00:00:00 2001 From: Wanlong Gao Date: Sun, 10 Apr 2011 14:14:43 +0800 Subject: fix the wrong argument of the functions definition The functions of eic_chip's memebers use the wrong argument . Signed-off-by: Wanlong Gao Signed-off-by: Hans-Christian Egtvedt diff --git a/arch/avr32/mach-at32ap/extint.c b/arch/avr32/mach-at32ap/extint.c index 47ba4b9..755ac59 100644 --- a/arch/avr32/mach-at32ap/extint.c +++ b/arch/avr32/mach-at32ap/extint.c @@ -61,34 +61,34 @@ struct eic { static struct eic *nmi_eic; static bool nmi_enabled; -static void eic_ack_irq(struct irq_chip *d) +static void eic_ack_irq(struct irq_data *d) { - struct eic *eic = irq_data_get_irq_chip_data(data); + struct eic *eic = irq_data_get_irq_chip_data(d); eic_writel(eic, ICR, 1 << (d->irq - eic->first_irq)); } -static void eic_mask_irq(struct irq_chip *d) +static void eic_mask_irq(struct irq_data *d) { - struct eic *eic = irq_data_get_irq_chip_data(data); + struct eic *eic = irq_data_get_irq_chip_data(d); eic_writel(eic, IDR, 1 << (d->irq - eic->first_irq)); } -static void eic_mask_ack_irq(struct irq_chip *d) +static void eic_mask_ack_irq(struct irq_data *d) { - struct eic *eic = irq_data_get_irq_chip_data(data); + struct eic *eic = irq_data_get_irq_chip_data(d); eic_writel(eic, ICR, 1 << (d->irq - eic->first_irq)); eic_writel(eic, IDR, 1 << (d->irq - eic->first_irq)); } -static void eic_unmask_irq(struct irq_chip *d) +static void eic_unmask_irq(struct irq_data *d) { - struct eic *eic = irq_data_get_irq_chip_data(data); + struct eic *eic = irq_data_get_irq_chip_data(d); eic_writel(eic, IER, 1 << (d->irq - eic->first_irq)); } -static int eic_set_irq_type(struct irq_chip *d, unsigned int flow_type) +static int eic_set_irq_type(struct irq_data *d, unsigned int flow_type) { - struct eic *eic = irq_data_get_irq_chip_data(data); + struct eic *eic = irq_data_get_irq_chip_data(d); unsigned int irq = d->irq; unsigned int i = irq - eic->first_irq; u32 mode, edge, level; -- cgit v0.10.2 From 024b3f2936c9a9393d2cf37b7c537b29fb894b62 Mon Sep 17 00:00:00 2001 From: Jean-Christophe PLAGNIOL-VILLARD Date: Sun, 10 Apr 2011 06:17:20 +0200 Subject: avr32: At32ap: pio fix typo "))" on gpio_irq_unmask prototype introduce in commit d75f1bfdbccb Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD Cc: Thomas Gleixner Cc: Hans-Christian Egtvedt Cc: Nicolas Ferre Cc: Patrice Vilchez Signed-off-by: Hans-Christian Egtvedt diff --git a/arch/avr32/mach-at32ap/pio.c b/arch/avr32/mach-at32ap/pio.c index f308e1d..2e0aa85 100644 --- a/arch/avr32/mach-at32ap/pio.c +++ b/arch/avr32/mach-at32ap/pio.c @@ -257,7 +257,7 @@ static void gpio_irq_mask(struct irq_data *d) pio_writel(pio, IDR, 1 << (gpio & 0x1f)); } -static void gpio_irq_unmask(struct irq_data *d)) +static void gpio_irq_unmask(struct irq_data *d) { unsigned gpio = irq_to_gpio(d->irq); struct pio_device *pio = &pio_dev[gpio >> 5]; -- cgit v0.10.2 From 51ef85d8f9ae24475a0cb1bd772258eafec91c69 Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Wed, 9 Mar 2011 00:32:36 +0000 Subject: avr32: Fix .size directive for cpu_enter_idle gas used to accept (and ignore?) .size directives which referred to undefined symbols, as this does. In binutils 2.21 these are treated as errors. Signed-off-by: Ben Hutchings Signed-off-by: Hans-Christian Egtvedt diff --git a/arch/avr32/mach-at32ap/pm-at32ap700x.S b/arch/avr32/mach-at32ap/pm-at32ap700x.S index 17503b0..f868f4c 100644 --- a/arch/avr32/mach-at32ap/pm-at32ap700x.S +++ b/arch/avr32/mach-at32ap/pm-at32ap700x.S @@ -53,7 +53,7 @@ cpu_enter_idle: st.w r8[TI_flags], r9 unmask_interrupts sleep CPU_SLEEP_IDLE - .size cpu_idle_sleep, . - cpu_idle_sleep + .size cpu_enter_idle, . - cpu_enter_idle /* * Common return path for PM functions that don't run from -- cgit v0.10.2 From 6e2ad51190cdb11b364377882134513f60dec6b9 Mon Sep 17 00:00:00 2001 From: Ole Henrik Jahren Date: Sun, 6 Mar 2011 20:42:39 +0100 Subject: avr32: fix deadlock when reading clock list in debugfs When writing out /sys/kernel/debug/at32ap_clk, clock list lock is being held while clk_get() is called. clk_get() attempts to take the same lock, which results in deadlock. Introduce and call lock free version, __clk_get(), instead. Signed-off-by: Ole Henrik Jahren Cc: Hans-Christian Egtvedt Signed-off-by: Hans-Christian Egtvedt diff --git a/arch/avr32/mach-at32ap/clock.c b/arch/avr32/mach-at32ap/clock.c index 442f08c..86925fd6 100644 --- a/arch/avr32/mach-at32ap/clock.c +++ b/arch/avr32/mach-at32ap/clock.c @@ -35,22 +35,30 @@ void at32_clk_register(struct clk *clk) spin_unlock(&clk_list_lock); } -struct clk *clk_get(struct device *dev, const char *id) +static struct clk *__clk_get(struct device *dev, const char *id) { struct clk *clk; - spin_lock(&clk_list_lock); - list_for_each_entry(clk, &at32_clock_list, list) { if (clk->dev == dev && strcmp(id, clk->name) == 0) { - spin_unlock(&clk_list_lock); return clk; } } - spin_unlock(&clk_list_lock); return ERR_PTR(-ENOENT); } + +struct clk *clk_get(struct device *dev, const char *id) +{ + struct clk *clk; + + spin_lock(&clk_list_lock); + clk = __clk_get(dev, id); + spin_unlock(&clk_list_lock); + + return clk; +} + EXPORT_SYMBOL(clk_get); void clk_put(struct clk *clk) @@ -257,15 +265,15 @@ static int clk_show(struct seq_file *s, void *unused) spin_lock(&clk_list_lock); /* show clock tree as derived from the three oscillators */ - clk = clk_get(NULL, "osc32k"); + clk = __clk_get(NULL, "osc32k"); dump_clock(clk, &r); clk_put(clk); - clk = clk_get(NULL, "osc0"); + clk = __clk_get(NULL, "osc0"); dump_clock(clk, &r); clk_put(clk); - clk = clk_get(NULL, "osc1"); + clk = __clk_get(NULL, "osc1"); dump_clock(clk, &r); clk_put(clk); -- cgit v0.10.2 From 9f0d15aac9987adaff18b85585fb7eaba266e112 Mon Sep 17 00:00:00 2001 From: Matt Fleming Date: Mon, 4 Apr 2011 15:58:04 +0100 Subject: avr32: init cannot ignore signals sent by force_sig_info() We can delete the code that checks to see if we're sending an ignored signal to init because force_sig_info() already handles this case. force_sig_info() will kill init even if the signal handler is SIG_DFL and the scenario described in the comment where init might "generate the same exception over and over again" cannot occur (force_sig_info() clears SIGNAL_UNKILLABLE to ensure that init will die). Also, the use of is_global_init() is not correct in the multhreaded case, as Oleg Nesterov explains, "is_global_init() is not right in theory, /sbin/init can be multithreaded. And, this doesn't cover the sub-namespace inits... I'd suggest to check SIGNAL_UNKILLABLE, but looking closer I think you can simply remove this code." It seems this code was copied from arch/powerpc in March 2007 in commit 623b0355d5b1 "[AVR32] Clean up exception handling code" but the code was deleted from arch/powerpc in November 2009 in commit a0592d42fe3e "powerpc: kill the obsolete code under is_global_init()" So catch up with powerpc and delete the bogus code. Signed-off-by: Matt Fleming Signed-off-by: Hans-Christian Egtvedt diff --git a/arch/avr32/kernel/traps.c b/arch/avr32/kernel/traps.c index b91b204..7aa2575 100644 --- a/arch/avr32/kernel/traps.c +++ b/arch/avr32/kernel/traps.c @@ -95,28 +95,6 @@ void _exception(long signr, struct pt_regs *regs, int code, info.si_code = code; info.si_addr = (void __user *)addr; force_sig_info(signr, &info, current); - - /* - * Init gets no signals that it doesn't have a handler for. - * That's all very well, but if it has caused a synchronous - * exception and we ignore the resulting signal, it will just - * generate the same exception over and over again and we get - * nowhere. Better to kill it and let the kernel panic. - */ - if (is_global_init(current)) { - __sighandler_t handler; - - spin_lock_irq(¤t->sighand->siglock); - handler = current->sighand->action[signr-1].sa.sa_handler; - spin_unlock_irq(¤t->sighand->siglock); - if (handler == SIG_DFL) { - /* init has generated a synchronous exception - and it doesn't have a handler for the signal */ - printk(KERN_CRIT "init has generated signal %ld " - "but has no handler for it\n", signr); - do_exit(signr); - } - } } asmlinkage void do_nmi(unsigned long ecr, struct pt_regs *regs) -- cgit v0.10.2 From c7d876321f4cf252bc70c1995bbc077a65b3af2a Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Wed, 9 Feb 2011 11:28:04 +0100 Subject: don't check platform_get_irq's return value against zero MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit platform_get_irq returns -ENXIO on failure, so !int_irq was probably always true. Better use (int)int_irq <= 0. Note that a return value of zero is still handled as error even though this could mean irq0. This is a followup to 305b3228f9ff4d59f49e6d34a7034d44ee8ce2f0 that changed the return value of platform_get_irq from 0 to -ENXIO on error. Acked-by: Hans-Christian Egtvedt Signed-off-by: Uwe Kleine-König Signed-off-by: Hans-Christian Egtvedt diff --git a/arch/avr32/mach-at32ap/extint.c b/arch/avr32/mach-at32ap/extint.c index 755ac59..fbc2aea 100644 --- a/arch/avr32/mach-at32ap/extint.c +++ b/arch/avr32/mach-at32ap/extint.c @@ -191,7 +191,7 @@ static int __init eic_probe(struct platform_device *pdev) regs = platform_get_resource(pdev, IORESOURCE_MEM, 0); int_irq = platform_get_irq(pdev, 0); - if (!regs || !int_irq) { + if (!regs || (int)int_irq <= 0) { dev_dbg(&pdev->dev, "missing regs and/or irq resource\n"); return -ENXIO; } -- cgit v0.10.2 From 24a1a47562b0fbb97321191dcc3a67b337b20f8f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Andreas=20Bie=C3=9Fmann?= Date: Wed, 13 Apr 2011 10:07:35 +0200 Subject: avr32: add ATAG_BOARDINFO MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The ATAG_BOARDINFO is intended to hand over the information bd->bi_board_number from u-boot to the kernel. This piece of information can be used to implement some kind of board identification while booting the kernel. Therefore it is placed in .initdata section and can be accessed via the new symbol board_number only while initializing the kernel. Signed-off-by: Andreas Bießmann Signed-off-by: Hans-Christian Egtvedt diff --git a/arch/avr32/include/asm/setup.h b/arch/avr32/include/asm/setup.h index ff5b7cf..160543d 100644 --- a/arch/avr32/include/asm/setup.h +++ b/arch/avr32/include/asm/setup.h @@ -94,6 +94,13 @@ struct tag_ethernet { #define ETH_INVALID_PHY 0xff +/* board information */ +#define ATAG_BOARDINFO 0x54410008 + +struct tag_boardinfo { + u32 board_number; +}; + struct tag { struct tag_header hdr; union { @@ -102,6 +109,7 @@ struct tag { struct tag_cmdline cmdline; struct tag_clock clock; struct tag_ethernet ethernet; + struct tag_boardinfo boardinfo; } u; }; @@ -128,6 +136,7 @@ extern struct tag *bootloader_tags; extern resource_size_t fbmem_start; extern resource_size_t fbmem_size; +extern u32 board_number; void setup_processor(void); diff --git a/arch/avr32/kernel/setup.c b/arch/avr32/kernel/setup.c index 5c70839..bb0974c 100644 --- a/arch/avr32/kernel/setup.c +++ b/arch/avr32/kernel/setup.c @@ -391,6 +391,21 @@ static int __init parse_tag_clock(struct tag *tag) __tagtable(ATAG_CLOCK, parse_tag_clock); /* + * The board_number correspond to the bd->bi_board_number in U-Boot. This + * parameter is only available during initialisation and can be used in some + * kind of board identification. + */ +u32 __initdata board_number; + +static int __init parse_tag_boardinfo(struct tag *tag) +{ + board_number = tag->u.boardinfo.board_number; + + return 0; +} +__tagtable(ATAG_BOARDINFO, parse_tag_boardinfo); + +/* * Scan the tag table for this tag, and call its parse function. The * tag table is built by the linker from all the __tagtable * declarations. -- cgit v0.10.2 From a626ca6a656450e9f4df91d0dda238fff23285f4 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Wed, 13 Apr 2011 08:07:28 -0700 Subject: vm: fix vm_pgoff wrap in stack expansion MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit 982134ba6261 ("mm: avoid wrapping vm_pgoff in mremap()") fixed the case of a expanding mapping causing vm_pgoff wrapping when you used mremap. But there was another case where we expand mappings hiding in plain sight: the automatic stack expansion. This fixes that case too. This one also found by Robert Święcki, using his nasty system call fuzzer tool. Good job. Reported-and-tested-by: Robert Święcki Cc: stable@kernel.org Signed-off-by: Linus Torvalds diff --git a/mm/mmap.c b/mm/mmap.c index 2ec8eb5..8c05e5b 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -1814,11 +1814,14 @@ static int expand_downwards(struct vm_area_struct *vma, size = vma->vm_end - address; grow = (vma->vm_start - address) >> PAGE_SHIFT; - error = acct_stack_growth(vma, size, grow); - if (!error) { - vma->vm_start = address; - vma->vm_pgoff -= grow; - perf_event_mmap(vma); + error = -ENOMEM; + if (grow <= vma->vm_pgoff) { + error = acct_stack_growth(vma, size, grow); + if (!error) { + vma->vm_start = address; + vma->vm_pgoff -= grow; + perf_event_mmap(vma); + } } } vma_unlock_anon_vma(vma); -- cgit v0.10.2 From 6631e635c65dc33cb798cc2f51d0ddd69ada6319 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Wed, 13 Apr 2011 08:08:20 -0700 Subject: block: don't flush plugged IO on forced preemtion scheduling We really only want to unplug the pending IO when the process actually goes to sleep. So move the test for flushing the plug up to the place where we actually deactivate the task - where we have properly checked for preemption and for the process really sleeping. Acked-by: Jens Axboe Acked-by: Peter Zijlstra Signed-off-by: Linus Torvalds diff --git a/kernel/sched.c b/kernel/sched.c index 4801363..a187c3f 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -4111,20 +4111,20 @@ need_resched: try_to_wake_up_local(to_wakeup); } deactivate_task(rq, prev, DEQUEUE_SLEEP); + + /* + * If we are going to sleep and we have plugged IO queued, make + * sure to submit it to avoid deadlocks. + */ + if (blk_needs_flush_plug(prev)) { + raw_spin_unlock(&rq->lock); + blk_flush_plug(prev); + raw_spin_lock(&rq->lock); + } } switch_count = &prev->nvcsw; } - /* - * If we are going to sleep and we have plugged IO queued, make - * sure to submit it to avoid deadlocks. - */ - if (prev->state != TASK_RUNNING && blk_needs_flush_plug(prev)) { - raw_spin_unlock(&rq->lock); - blk_flush_plug(prev); - raw_spin_lock(&rq->lock); - } - pre_schedule(rq, prev); if (unlikely(!rq->nr_running)) -- cgit v0.10.2 From 21a8d026e0721f90ae03084e96a215632c80fc13 Mon Sep 17 00:00:00 2001 From: Matthew Garrett Date: Wed, 13 Apr 2011 11:52:16 -0400 Subject: x86 platform drivers: Build fix for intel_pmic_gpio Fix an incorrect function name so the driver builds. Signed-off-by: Matthew Garrett diff --git a/drivers/platform/x86/intel_pmic_gpio.c b/drivers/platform/x86/intel_pmic_gpio.c index 7bbb538..464bb3f 100644 --- a/drivers/platform/x86/intel_pmic_gpio.c +++ b/drivers/platform/x86/intel_pmic_gpio.c @@ -209,7 +209,7 @@ static struct irq_chip pmic_irqchip = { .irq_mask = pmic_irq_mask, .irq_unmask = pmic_irq_unmask, .irq_set_type = pmic_irq_type, - .irq_bus_lock = pmic_irq_buslock, + .irq_bus_lock = pmic_bus_lock, .irq_bus_sync_unlock = pmic_bus_sync_unlock, }; -- cgit v0.10.2 From 5c72d064f7ead1126bed6faab0c2bfb7418036e2 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Wed, 13 Apr 2011 09:28:23 +0100 Subject: drm/i915: Initialise g4x watermarks for disabled pipes We were using uninitialised watermarks values for disabled pipes which were combined into a single WM register and so corrupting the values for the enabled pipe and upsetting the display hardware. Reported-by: Riccardo Magliocchetti Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=32612 Signed-off-by: Chris Wilson Signed-off-by: Keith Packard diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 4fc21e0..e522c70 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -3771,8 +3771,11 @@ static bool g4x_compute_wm0(struct drm_device *dev, int entries, tlb_miss; crtc = intel_get_crtc_for_plane(dev, plane); - if (crtc->fb == NULL || !crtc->enabled) + if (crtc->fb == NULL || !crtc->enabled) { + *cursor_wm = cursor->guard_size; + *plane_wm = display->guard_size; return false; + } htotal = crtc->mode.htotal; hdisplay = crtc->mode.hdisplay; -- cgit v0.10.2 From 7c9f6472d46d07000c197e1d24fa2a332606968e Mon Sep 17 00:00:00 2001 From: Anatolij Gustschin Date: Tue, 12 Apr 2011 23:49:28 +0000 Subject: net: can: mscan: fix build breakage in mpc5xxx_can Commit 74888760d40b3ac9054f9c5fa07b566c0676ba2d "dt/net: Eliminate users of of_platform_{,un}register_driver" broke building mscan driver. Fix it. Signed-off-by: Anatolij Gustschin Cc: Grant Likely Acked-by: Wolfgang Grandegger Acked-by: Marc Kleine-Budde Signed-off-by: David S. Miller diff --git a/drivers/net/can/mscan/mpc5xxx_can.c b/drivers/net/can/mscan/mpc5xxx_can.c index c0a1bc5..bd1d811 100644 --- a/drivers/net/can/mscan/mpc5xxx_can.c +++ b/drivers/net/can/mscan/mpc5xxx_can.c @@ -260,7 +260,7 @@ static int __devinit mpc5xxx_can_probe(struct platform_device *ofdev) if (!ofdev->dev.of_match) return -EINVAL; - data = (struct mpc5xxx_can_data *)of_dev->dev.of_match->data; + data = (struct mpc5xxx_can_data *)ofdev->dev.of_match->data; base = of_iomap(np, 0); if (!base) { -- cgit v0.10.2 From 91a403caf0f26c71ce4407fd235b2d6fb225fba9 Mon Sep 17 00:00:00 2001 From: Amit Kumar Salecha Date: Tue, 12 Apr 2011 17:05:55 +0000 Subject: qlcnic: limit skb frags for non tso packet Machines are getting deadlock in four node cluster environment. All nodes are accessing (find /gfs2 -depth -print|cpio -ocv > /dev/null) 200 GB storage on a GFS2 filesystem. This result in memory fragmentation and driver receives 18 frags for 1448 byte packets. For non tso packet, fw drops the tx request, if it has >14 frags. Fixing it by pulling extra frags. Cc: stable@kernel.org Signed-off-by: Amit Kumar Salecha Signed-off-by: David S. Miller diff --git a/drivers/net/qlcnic/qlcnic.h b/drivers/net/qlcnic/qlcnic.h index dc44564..b0dead0 100644 --- a/drivers/net/qlcnic/qlcnic.h +++ b/drivers/net/qlcnic/qlcnic.h @@ -99,6 +99,7 @@ #define TX_UDPV6_PKT 0x0c /* Tx defines */ +#define QLCNIC_MAX_FRAGS_PER_TX 14 #define MAX_TSO_HEADER_DESC 2 #define MGMT_CMD_DESC_RESV 4 #define TX_STOP_THRESH ((MAX_SKB_FRAGS >> 2) + MAX_TSO_HEADER_DESC \ diff --git a/drivers/net/qlcnic/qlcnic_main.c b/drivers/net/qlcnic/qlcnic_main.c index cd88c7e..cb1a1ef 100644 --- a/drivers/net/qlcnic/qlcnic_main.c +++ b/drivers/net/qlcnic/qlcnic_main.c @@ -2099,6 +2099,7 @@ qlcnic_xmit_frame(struct sk_buff *skb, struct net_device *netdev) struct cmd_desc_type0 *hwdesc, *first_desc; struct pci_dev *pdev; struct ethhdr *phdr; + int delta = 0; int i, k; u32 producer; @@ -2118,6 +2119,19 @@ qlcnic_xmit_frame(struct sk_buff *skb, struct net_device *netdev) } frag_count = skb_shinfo(skb)->nr_frags + 1; + /* 14 frags supported for normal packet and + * 32 frags supported for TSO packet + */ + if (!skb_is_gso(skb) && frag_count > QLCNIC_MAX_FRAGS_PER_TX) { + + for (i = 0; i < (frag_count - QLCNIC_MAX_FRAGS_PER_TX); i++) + delta += skb_shinfo(skb)->frags[i].size; + + if (!__pskb_pull_tail(skb, delta)) + goto drop_packet; + + frag_count = 1 + skb_shinfo(skb)->nr_frags; + } /* 4 fragments per cmd des */ no_of_desc = (frag_count + 3) >> 2; -- cgit v0.10.2 From 3e8c806a08c7beecd972e7ce15c570b9aba64baa Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Wed, 13 Apr 2011 12:01:14 -0700 Subject: Revert "tcp: disallow bind() to reuse addr/port" This reverts commit c191a836a908d1dd6b40c503741f91b914de3348. It causes known regressions for programs that expect to be able to use SO_REUSEADDR to shutdown a socket, then successfully rebind another socket to the same ID. Programs such as haproxy and amavisd expect this to work. This should fix kernel bugzilla 32832. Signed-off-by: David S. Miller diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 6c0b7f4..38f23e7 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -73,7 +73,7 @@ int inet_csk_bind_conflict(const struct sock *sk, !sk2->sk_bound_dev_if || sk->sk_bound_dev_if == sk2->sk_bound_dev_if)) { if (!reuse || !sk2->sk_reuse || - ((1 << sk2->sk_state) & (TCPF_LISTEN | TCPF_CLOSE))) { + sk2->sk_state == TCP_LISTEN) { const __be32 sk2_rcv_saddr = sk_rcv_saddr(sk2); if (!sk2_rcv_saddr || !sk_rcv_saddr(sk) || sk2_rcv_saddr == sk_rcv_saddr(sk)) @@ -122,8 +122,7 @@ again: (tb->num_owners < smallest_size || smallest_size == -1)) { smallest_size = tb->num_owners; smallest_rover = rover; - if (atomic_read(&hashinfo->bsockets) > (high - low) + 1 && - !inet_csk(sk)->icsk_af_ops->bind_conflict(sk, tb)) { + if (atomic_read(&hashinfo->bsockets) > (high - low) + 1) { spin_unlock(&head->lock); snum = smallest_rover; goto have_snum; diff --git a/net/ipv6/inet6_connection_sock.c b/net/ipv6/inet6_connection_sock.c index 1660546..f2c5b0f 100644 --- a/net/ipv6/inet6_connection_sock.c +++ b/net/ipv6/inet6_connection_sock.c @@ -44,7 +44,7 @@ int inet6_csk_bind_conflict(const struct sock *sk, !sk2->sk_bound_dev_if || sk->sk_bound_dev_if == sk2->sk_bound_dev_if) && (!sk->sk_reuse || !sk2->sk_reuse || - ((1 << sk2->sk_state) & (TCPF_LISTEN | TCPF_CLOSE))) && + sk2->sk_state == TCP_LISTEN) && ipv6_rcv_saddr_equal(sk, sk2)) break; } -- cgit v0.10.2 From 332704a51498a7e29aa92c19dc03f11f80b71bfe Mon Sep 17 00:00:00 2001 From: Stanislaw Gruszka Date: Wed, 13 Apr 2011 10:56:51 +0200 Subject: iwlegacy: fix tx_power initialization priv->tx_power_next is not initialized to max supported power, but instead default value is used, what cause errors like [ 58.597834] iwl3945 0000:03:00.0: Requested user TXPOWER 15 above upper limit 14. [ 58.597839] iwl3945 0000:03:00.0: Error setting Tx power (-22). if maximum tx power read from the eeprom is smaller than default. In consequence card is unable to initialize properly. Fix the problem and cleanup tx power initialization. Reported-and-tested-by: Robin Dong Signed-off-by: Stanislaw Gruszka Signed-off-by: John W. Linville diff --git a/drivers/net/wireless/iwlegacy/iwl-3945-hw.h b/drivers/net/wireless/iwlegacy/iwl-3945-hw.h index 779d3cb..5c3a68d 100644 --- a/drivers/net/wireless/iwlegacy/iwl-3945-hw.h +++ b/drivers/net/wireless/iwlegacy/iwl-3945-hw.h @@ -74,8 +74,6 @@ /* RSSI to dBm */ #define IWL39_RSSI_OFFSET 95 -#define IWL_DEFAULT_TX_POWER 0x0F - /* * EEPROM related constants, enums, and structures. */ diff --git a/drivers/net/wireless/iwlegacy/iwl-4965-hw.h b/drivers/net/wireless/iwlegacy/iwl-4965-hw.h index 08b189c..fc6fa28 100644 --- a/drivers/net/wireless/iwlegacy/iwl-4965-hw.h +++ b/drivers/net/wireless/iwlegacy/iwl-4965-hw.h @@ -804,9 +804,6 @@ struct iwl4965_scd_bc_tbl { #define IWL4965_DEFAULT_TX_RETRY 15 -/* Limit range of txpower output target to be between these values */ -#define IWL4965_TX_POWER_TARGET_POWER_MIN (0) /* 0 dBm: 1 milliwatt */ - /* EEPROM */ #define IWL4965_FIRST_AMPDU_QUEUE 10 diff --git a/drivers/net/wireless/iwlegacy/iwl-core.c b/drivers/net/wireless/iwlegacy/iwl-core.c index a209a0e..2b08efb 100644 --- a/drivers/net/wireless/iwlegacy/iwl-core.c +++ b/drivers/net/wireless/iwlegacy/iwl-core.c @@ -160,6 +160,7 @@ int iwl_legacy_init_geos(struct iwl_priv *priv) struct ieee80211_channel *geo_ch; struct ieee80211_rate *rates; int i = 0; + s8 max_tx_power = 0; if (priv->bands[IEEE80211_BAND_2GHZ].n_bitrates || priv->bands[IEEE80211_BAND_5GHZ].n_bitrates) { @@ -235,8 +236,8 @@ int iwl_legacy_init_geos(struct iwl_priv *priv) geo_ch->flags |= ch->ht40_extension_channel; - if (ch->max_power_avg > priv->tx_power_device_lmt) - priv->tx_power_device_lmt = ch->max_power_avg; + if (ch->max_power_avg > max_tx_power) + max_tx_power = ch->max_power_avg; } else { geo_ch->flags |= IEEE80211_CHAN_DISABLED; } @@ -249,6 +250,10 @@ int iwl_legacy_init_geos(struct iwl_priv *priv) geo_ch->flags); } + priv->tx_power_device_lmt = max_tx_power; + priv->tx_power_user_lmt = max_tx_power; + priv->tx_power_next = max_tx_power; + if ((priv->bands[IEEE80211_BAND_5GHZ].n_channels == 0) && priv->cfg->sku & IWL_SKU_A) { IWL_INFO(priv, "Incorrectly detected BG card as ABG. " @@ -1124,11 +1129,11 @@ int iwl_legacy_set_tx_power(struct iwl_priv *priv, s8 tx_power, bool force) if (!priv->cfg->ops->lib->send_tx_power) return -EOPNOTSUPP; - if (tx_power < IWL4965_TX_POWER_TARGET_POWER_MIN) { + /* 0 dBm mean 1 milliwatt */ + if (tx_power < 0) { IWL_WARN(priv, - "Requested user TXPOWER %d below lower limit %d.\n", - tx_power, - IWL4965_TX_POWER_TARGET_POWER_MIN); + "Requested user TXPOWER %d below 1 mW.\n", + tx_power); return -EINVAL; } diff --git a/drivers/net/wireless/iwlegacy/iwl-eeprom.c b/drivers/net/wireless/iwlegacy/iwl-eeprom.c index 04c5648..cb346d1 100644 --- a/drivers/net/wireless/iwlegacy/iwl-eeprom.c +++ b/drivers/net/wireless/iwlegacy/iwl-eeprom.c @@ -471,13 +471,6 @@ int iwl_legacy_init_channel_map(struct iwl_priv *priv) flags & EEPROM_CHANNEL_RADAR)) ? "" : "not "); - /* Set the tx_power_user_lmt to the highest power - * supported by any channel */ - if (eeprom_ch_info[ch].max_power_avg > - priv->tx_power_user_lmt) - priv->tx_power_user_lmt = - eeprom_ch_info[ch].max_power_avg; - ch_info++; } } diff --git a/drivers/net/wireless/iwlegacy/iwl3945-base.c b/drivers/net/wireless/iwlegacy/iwl3945-base.c index 28eb3d8..cc7ebce 100644 --- a/drivers/net/wireless/iwlegacy/iwl3945-base.c +++ b/drivers/net/wireless/iwlegacy/iwl3945-base.c @@ -3825,10 +3825,6 @@ static int iwl3945_init_drv(struct iwl_priv *priv) priv->force_reset[IWL_FW_RESET].reset_duration = IWL_DELAY_NEXT_FORCE_FW_RELOAD; - - priv->tx_power_user_lmt = IWL_DEFAULT_TX_POWER; - priv->tx_power_next = IWL_DEFAULT_TX_POWER; - if (eeprom->version < EEPROM_3945_EEPROM_VERSION) { IWL_WARN(priv, "Unsupported EEPROM version: 0x%04X\n", eeprom->version); diff --git a/drivers/net/wireless/iwlegacy/iwl4965-base.c b/drivers/net/wireless/iwlegacy/iwl4965-base.c index 91b3d8b..d484c36 100644 --- a/drivers/net/wireless/iwlegacy/iwl4965-base.c +++ b/drivers/net/wireless/iwlegacy/iwl4965-base.c @@ -3140,12 +3140,6 @@ static int iwl4965_init_drv(struct iwl_priv *priv) iwl_legacy_init_scan_params(priv); - /* Set the tx_power_user_lmt to the lowest power level - * this value will get overwritten by channel max power avg - * from eeprom */ - priv->tx_power_user_lmt = IWL4965_TX_POWER_TARGET_POWER_MIN; - priv->tx_power_next = IWL4965_TX_POWER_TARGET_POWER_MIN; - ret = iwl_legacy_init_channel_map(priv); if (ret) { IWL_ERR(priv, "initializing regulatory failed: %d\n", ret); -- cgit v0.10.2 From 80656b67b3988f83edd86a280d9937124fe62050 Mon Sep 17 00:00:00 2001 From: Liu Yuan Date: Wed, 13 Apr 2011 22:14:54 +0200 Subject: block, blk-sysfs: Use the variable directly instead of a function call In the function blk_register_queue(), var _dev_ is already assigned by disk_to_dev().So use it directly instead of calling disk_to_dev() again. Signed-off-by: Liu Yuan Modified by me to delete an empty line in the same function while in there anyway. Signed-off-by: Jens Axboe diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index 261c75c..6d73512 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -498,7 +498,6 @@ int blk_register_queue(struct gendisk *disk) { int ret; struct device *dev = disk_to_dev(disk); - struct request_queue *q = disk->queue; if (WARN_ON(!q)) @@ -521,7 +520,7 @@ int blk_register_queue(struct gendisk *disk) if (ret) { kobject_uevent(&q->kobj, KOBJ_REMOVE); kobject_del(&q->kobj); - blk_trace_remove_sysfs(disk_to_dev(disk)); + blk_trace_remove_sysfs(dev); kobject_put(&dev->kobj); return ret; } -- cgit v0.10.2 From d06847fec256f4f902075ce5986e10f7c55fa250 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Tue, 15 Mar 2011 17:09:10 -0700 Subject: usb: fix ips1760-hcd printk format warning Fix printk format build warning and grammar typo on same line. drivers/usb/host/isp1760-hcd.c:300: warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'size_t' Signed-off-by: Randy Dunlap Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/host/isp1760-hcd.c b/drivers/usb/host/isp1760-hcd.c index f50e84a..795345a 100644 --- a/drivers/usb/host/isp1760-hcd.c +++ b/drivers/usb/host/isp1760-hcd.c @@ -295,7 +295,7 @@ static void alloc_mem(struct usb_hcd *hcd, struct isp1760_qtd *qtd) } dev_err(hcd->self.controller, - "%s: Can not allocate %lu bytes of memory\n" + "%s: Cannot allocate %zu bytes of memory\n" "Current memory map:\n", __func__, qtd->length); for (i = 0; i < BLOCKS; i++) { -- cgit v0.10.2 From 00cc7a5faf25b3ba5cf30fcffc62249bdd152006 Mon Sep 17 00:00:00 2001 From: Alexey Khoroshilov Date: Wed, 16 Mar 2011 21:54:05 +0200 Subject: USB: usb-gadget: unlock data->lock mutex on error path in ep_read() ep_read() acquires data->lock mutex in get_ready_ep() and releases it on all paths except for one: when usb_endpoint_xfer_isoc() failed. The patch adds mutex_unlock(&data->lock) at that path. Found by Linux Driver Verification project (linuxtesting.org). Signed-off-by: Alexey Khoroshilov Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/gadget/inode.c b/drivers/usb/gadget/inode.c index 3ed73f4..a01383f 100644 --- a/drivers/usb/gadget/inode.c +++ b/drivers/usb/gadget/inode.c @@ -386,8 +386,10 @@ ep_read (struct file *fd, char __user *buf, size_t len, loff_t *ptr) /* halt any endpoint by doing a "wrong direction" i/o call */ if (usb_endpoint_dir_in(&data->desc)) { - if (usb_endpoint_xfer_isoc(&data->desc)) + if (usb_endpoint_xfer_isoc(&data->desc)) { + mutex_unlock(&data->lock); return -EINVAL; + } DBG (data->dev, "%s halt\n", data->name); spin_lock_irq (&data->dev->lock); if (likely (data->ep != NULL)) -- cgit v0.10.2 From 0291303d37585ab9056bae9fac1d1e13ae1ce913 Mon Sep 17 00:00:00 2001 From: Michal Simek Date: Mon, 4 Apr 2011 11:35:05 +0200 Subject: usb: Fix Kconfig unmet dependencies for Microblaze EHCI Disable USB_ARCH_HAS_EHCI in arch Kconfig and enable it in usb Kconfig Warning log: warning: (MICROBLAZE) selects USB_ARCH_HAS_EHCI which has unmet direct dependencies (USB_SUPPORT) Signed-off-by: Michal Simek Signed-off-by: Greg Kroah-Hartman diff --git a/arch/microblaze/Kconfig b/arch/microblaze/Kconfig index 851b3bf..eccdefe 100644 --- a/arch/microblaze/Kconfig +++ b/arch/microblaze/Kconfig @@ -6,7 +6,6 @@ config MICROBLAZE select HAVE_FUNCTION_GRAPH_TRACER select HAVE_DYNAMIC_FTRACE select HAVE_FTRACE_MCOUNT_RECORD - select USB_ARCH_HAS_EHCI select ARCH_WANT_OPTIONAL_GPIOLIB select HAVE_OPROFILE select HAVE_ARCH_KGDB diff --git a/drivers/usb/Kconfig b/drivers/usb/Kconfig index 41b6e51..006489d 100644 --- a/drivers/usb/Kconfig +++ b/drivers/usb/Kconfig @@ -66,6 +66,7 @@ config USB_ARCH_HAS_EHCI default y if ARCH_VT8500 default y if PLAT_SPEAR default y if ARCH_MSM + default y if MICROBLAZE default PCI # ARM SA1111 chips have a non-PCI based "OHCI-compatible" USB host interface. -- cgit v0.10.2 From 2c2da1799ba776c4bd8d51ee46d9f00cb4cc6120 Mon Sep 17 00:00:00 2001 From: Yoshihiro Shimoda Date: Mon, 4 Apr 2011 13:09:22 +0900 Subject: usb: r8a66597-udc: fix spinlock usage Because the disconnect function in the composite driver will call spin_lock, this driver has to call spin_unlock before calling driver->disconnet(). Signed-off-by: Yoshihiro Shimoda Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/gadget/r8a66597-udc.c b/drivers/usb/gadget/r8a66597-udc.c index 0151185..6dcc1f6 100644 --- a/drivers/usb/gadget/r8a66597-udc.c +++ b/drivers/usb/gadget/r8a66597-udc.c @@ -1083,7 +1083,9 @@ static void irq_device_state(struct r8a66597 *r8a66597) if (dvsq == DS_DFLT) { /* bus reset */ + spin_unlock(&r8a66597->lock); r8a66597->driver->disconnect(&r8a66597->gadget); + spin_lock(&r8a66597->lock); r8a66597_update_usb_speed(r8a66597); } if (r8a66597->old_dvsq == DS_CNFG && dvsq != DS_CNFG) -- cgit v0.10.2 From 5a9443f08c83c294c5c806a689c1184b27cb26b3 Mon Sep 17 00:00:00 2001 From: Christian Simon Date: Mon, 28 Mar 2011 21:54:47 +0200 Subject: USB: ftdi_sio: Added IDs for CTI USB Serial Devices I added new ProdutIds for two devices from CTI GmbH Leipzig. Signed-off-by: Christian Simon Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/serial/ftdi_sio.c b/drivers/usb/serial/ftdi_sio.c index a973c7a..8a02407 100644 --- a/drivers/usb/serial/ftdi_sio.c +++ b/drivers/usb/serial/ftdi_sio.c @@ -151,6 +151,8 @@ static struct ftdi_sio_quirk ftdi_stmclite_quirk = { * /sys/bus/usb/ftdi_sio/new_id, then send patch/report! */ static struct usb_device_id id_table_combined [] = { + { USB_DEVICE(FTDI_VID, FTDI_CTI_MINI_PID) }, + { USB_DEVICE(FTDI_VID, FTDI_CTI_NANO_PID) }, { USB_DEVICE(FTDI_VID, FTDI_AMC232_PID) }, { USB_DEVICE(FTDI_VID, FTDI_CANUSB_PID) }, { USB_DEVICE(FTDI_VID, FTDI_CANDAPTER_PID) }, diff --git a/drivers/usb/serial/ftdi_sio_ids.h b/drivers/usb/serial/ftdi_sio_ids.h index c543e55..770b5dd 100644 --- a/drivers/usb/serial/ftdi_sio_ids.h +++ b/drivers/usb/serial/ftdi_sio_ids.h @@ -1141,3 +1141,12 @@ #define QIHARDWARE_VID 0x20B7 #define MILKYMISTONE_JTAGSERIAL_PID 0x0713 +/* + * CTI GmbH RS485 Converter http://www.cti-lean.com/ + */ +/* USB-485-Mini*/ +#define FTDI_CTI_MINI_PID 0xF608 +/* USB-Nano-485*/ +#define FTDI_CTI_NANO_PID 0xF60B + + -- cgit v0.10.2 From 505d1f69ec4f8697a74711fb3a01ed151fda3834 Mon Sep 17 00:00:00 2001 From: Yauheni Kaliuta Date: Tue, 5 Apr 2011 16:55:25 +0300 Subject: usb: gadget: eem: fix echo command processing During processing of bunch of eem frames if "echo" command is found skb is cloned and the cloned version should be used to send reply. Unfortunately, the data of the original skb were actually used and the cloned skb is never freed. Using the cloned skb and freeing the skb in the completion callback for usb request. Signed-off-by: Yauheni Kaliuta Reviewed-by: Felipe Balbi Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/gadget/f_eem.c b/drivers/usb/gadget/f_eem.c index 95dd466..b3c3042 100644 --- a/drivers/usb/gadget/f_eem.c +++ b/drivers/usb/gadget/f_eem.c @@ -314,6 +314,9 @@ eem_unbind(struct usb_configuration *c, struct usb_function *f) static void eem_cmd_complete(struct usb_ep *ep, struct usb_request *req) { + struct sk_buff *skb = (struct sk_buff *)req->context; + + dev_kfree_skb_any(skb); } /* @@ -428,10 +431,11 @@ static int eem_unwrap(struct gether *port, skb_trim(skb2, len); put_unaligned_le16(BIT(15) | BIT(11) | len, skb_push(skb2, 2)); - skb_copy_bits(skb, 0, req->buf, skb->len); - req->length = skb->len; + skb_copy_bits(skb2, 0, req->buf, skb2->len); + req->length = skb2->len; req->complete = eem_cmd_complete; req->zero = 1; + req->context = skb2; if (usb_ep_queue(port->in_ep, req, GFP_ATOMIC)) DBG(cdev, "echo response queue fail\n"); break; -- cgit v0.10.2 From 94ae4976e253757e9b03a44d27d41b20f1829d80 Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Tue, 5 Apr 2011 13:36:15 -0400 Subject: USB: EHCI: unlink unused QHs when the controller is stopped This patch (as1458) fixes a problem affecting ultra-reliable systems: When hardware failover of an EHCI controller occurs, the data structures do not get released correctly. This is because the routine responsible for removing unused QHs from the async schedule assumes the controller is running properly (the frame counter is used in determining how long the QH has been idle) -- but when a failover causes the controller to be electronically disconnected from the PCI bus, obviously it stops running. The solution is simple: Allow scan_async() to remove a QH from the async schedule if it has been idle for long enough _or_ if the controller is stopped. Signed-off-by: Alan Stern Reported-and-Tested-by: Dan Duval CC: Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/host/ehci-q.c b/drivers/usb/host/ehci-q.c index 98ded66..42abd0f 100644 --- a/drivers/usb/host/ehci-q.c +++ b/drivers/usb/host/ehci-q.c @@ -1247,24 +1247,27 @@ static void start_unlink_async (struct ehci_hcd *ehci, struct ehci_qh *qh) static void scan_async (struct ehci_hcd *ehci) { + bool stopped; struct ehci_qh *qh; enum ehci_timer_action action = TIMER_IO_WATCHDOG; ehci->stamp = ehci_readl(ehci, &ehci->regs->frame_index); timer_action_done (ehci, TIMER_ASYNC_SHRINK); rescan: + stopped = !HC_IS_RUNNING(ehci_to_hcd(ehci)->state); qh = ehci->async->qh_next.qh; if (likely (qh != NULL)) { do { /* clean any finished work for this qh */ - if (!list_empty (&qh->qtd_list) - && qh->stamp != ehci->stamp) { + if (!list_empty(&qh->qtd_list) && (stopped || + qh->stamp != ehci->stamp)) { int temp; /* unlinks could happen here; completion * reporting drops the lock. rescan using * the latest schedule, but don't rescan - * qhs we already finished (no looping). + * qhs we already finished (no looping) + * unless the controller is stopped. */ qh = qh_get (qh); qh->stamp = ehci->stamp; @@ -1285,9 +1288,9 @@ rescan: */ if (list_empty(&qh->qtd_list) && qh->qh_state == QH_STATE_LINKED) { - if (!ehci->reclaim - && ((ehci->stamp - qh->stamp) & 0x1fff) - >= (EHCI_SHRINK_FRAMES * 8)) + if (!ehci->reclaim && (stopped || + ((ehci->stamp - qh->stamp) & 0x1fff) + >= EHCI_SHRINK_FRAMES * 8)) start_unlink_async(ehci, qh); else action = TIMER_ASYNC_SHRINK; -- cgit v0.10.2 From 36a52c009a39049893c9a3a4091f8f8f48585f47 Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Tue, 5 Apr 2011 19:50:34 +0200 Subject: usb: musb: omap2430: fix build failure MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix build failure introduced by commit 7acc6197b76edd0b932a7cbcc6cfad0a8a87f026 (usb: musb: Idle path retention and offmode support for OMAP3) when building without gadget support. CC drivers/usb/musb/omap2430.o drivers/usb/musb/omap2430.c: In function ‘musb_otg_notifications’: drivers/usb/musb/omap2430.c:262: error: ‘struct musb’ has no member named ‘gadget_driver’ Signed-off-by: Johan Hovold Cc: Felipe Balbi Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/musb/omap2430.c b/drivers/usb/musb/omap2430.c index 25cb8b0..57a27fa 100644 --- a/drivers/usb/musb/omap2430.c +++ b/drivers/usb/musb/omap2430.c @@ -259,9 +259,10 @@ static int musb_otg_notifications(struct notifier_block *nb, case USB_EVENT_VBUS: DBG(4, "VBUS Connect\n"); +#ifdef CONFIG_USB_GADGET_MUSB_HDRC if (musb->gadget_driver) pm_runtime_get_sync(musb->controller); - +#endif otg_init(musb->xceiv); break; -- cgit v0.10.2 From 11a31d84129dc3133417d626643d714c9df5317e Mon Sep 17 00:00:00 2001 From: Johan Hovold Date: Fri, 8 Apr 2011 17:38:22 +0200 Subject: USB: ftdi_sio: add PID for OCT DK201 docking station Add PID 0x0103 for serial port of the OCT DK201 docking station. Reported-by: Jan Hoogenraad Signed-off-by: Johan Hovold Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/serial/ftdi_sio.c b/drivers/usb/serial/ftdi_sio.c index 8a02407..51b7801 100644 --- a/drivers/usb/serial/ftdi_sio.c +++ b/drivers/usb/serial/ftdi_sio.c @@ -527,6 +527,7 @@ static struct usb_device_id id_table_combined [] = { { USB_DEVICE(SEALEVEL_VID, SEALEVEL_2803_8_PID) }, { USB_DEVICE(IDTECH_VID, IDTECH_IDT1221U_PID) }, { USB_DEVICE(OCT_VID, OCT_US101_PID) }, + { USB_DEVICE(OCT_VID, OCT_DK201_PID) }, { USB_DEVICE(FTDI_VID, FTDI_HE_TIRA1_PID), .driver_info = (kernel_ulong_t)&ftdi_HE_TIRA1_quirk }, { USB_DEVICE(FTDI_VID, FTDI_USB_UIRT_PID), diff --git a/drivers/usb/serial/ftdi_sio_ids.h b/drivers/usb/serial/ftdi_sio_ids.h index 770b5dd..0bdba7a 100644 --- a/drivers/usb/serial/ftdi_sio_ids.h +++ b/drivers/usb/serial/ftdi_sio_ids.h @@ -572,6 +572,7 @@ /* Note: OCT US101 is also rebadged as Dick Smith Electronics (NZ) XH6381 */ /* Also rebadged as Dick Smith Electronics (Aus) XH6451 */ /* Also rebadged as SIIG Inc. model US2308 hardware version 1 */ +#define OCT_DK201_PID 0x0103 /* OCT DK201 USB docking station */ #define OCT_US101_PID 0x0421 /* OCT US101 USB to RS-232 */ /* -- cgit v0.10.2 From 0fdf65c3394358a3634ad1064ea4984a03a6aa07 Mon Sep 17 00:00:00 2001 From: Yoichi Yuasa Date: Mon, 11 Apr 2011 21:56:39 +0900 Subject: USB: ohci-au1xxx: fix warning "__BIG_ENDIAN" is not defined In file included from drivers/usb/host/ohci-hcd.c:1028:0: drivers/usb/host/ohci-au1xxx.c:36:7: warning: "__BIG_ENDIAN" is not defined Signed-off-by: Yoichi Yuasa Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/host/ohci-au1xxx.c b/drivers/usb/host/ohci-au1xxx.c index 17a6043..958d985f 100644 --- a/drivers/usb/host/ohci-au1xxx.c +++ b/drivers/usb/host/ohci-au1xxx.c @@ -33,7 +33,7 @@ #ifdef __LITTLE_ENDIAN #define USBH_ENABLE_INIT (USBH_ENABLE_CE | USBH_ENABLE_E | USBH_ENABLE_C) -#elif __BIG_ENDIAN +#elif defined(__BIG_ENDIAN) #define USBH_ENABLE_INIT (USBH_ENABLE_CE | USBH_ENABLE_E | USBH_ENABLE_C | \ USBH_ENABLE_BE) #else -- cgit v0.10.2 From 16a2f970f3b4beb8d21009b2c45b9b5ab56bb621 Mon Sep 17 00:00:00 2001 From: Sebastian Andrzej Siewior Date: Mon, 11 Apr 2011 20:44:30 +0200 Subject: usb/gadget: don't leak hs_descriptors We should free both descriptors. Signed-off-by: Sebastian Andrzej Siewior Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/gadget/f_audio.c b/drivers/usb/gadget/f_audio.c index 9abecfd..0111f8a 100644 --- a/drivers/usb/gadget/f_audio.c +++ b/drivers/usb/gadget/f_audio.c @@ -706,6 +706,7 @@ f_audio_unbind(struct usb_configuration *c, struct usb_function *f) struct f_audio *audio = func_to_audio(f); usb_free_descriptors(f->descriptors); + usb_free_descriptors(f->hs_descriptors); kfree(audio); } -- cgit v0.10.2 From 9ab7927bb845cf2549110b19c212fe44a2bfbacb Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Wed, 13 Apr 2011 08:38:16 +0200 Subject: USB host: Fix lockdep warning in AMD PLL quirk Booting latest kernel on my test machine produces a lockdep warning from the usb_amd_find_chipset_info() function: WARNING: at /data/lemmy/linux.trees.git/kernel/lockdep.c:2465 lockdep_trace_alloc+0x95/0xc2() Hardware name: Snook Modules linked in: Pid: 959, comm: work_for_cpu Not tainted 2.6.39-rc2+ #22 Call Trace: [] warn_slowpath_common+0x80/0x98 [] ? T.492+0x24/0x26 [] warn_slowpath_null+0x15/0x17 [] lockdep_trace_alloc+0x95/0xc2 [] slab_pre_alloc_hook+0x18/0x3b [] kmem_cache_alloc_trace+0x25/0xba [] T.492+0x24/0x26 [] pci_get_subsys+0x2e/0x73 [] pci_get_device+0x11/0x13 [] usb_amd_find_chipset_info+0x3f/0x18a ... It turns out that this function calls pci_get_device under a spin_lock with irqs disabled, but the pci_get_device function is only allowed in preemptible context. This patch fixes the warning by making all data-structure modifications on temporal storage and commiting this back into the visible structure at the end. While at it, this patch also moves the pci_dev_put calls out of the spinlocks because this function might sleep too. Signed-off-by: Joerg Roedel Acked-by: Alan Stern Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/host/pci-quirks.c b/drivers/usb/host/pci-quirks.c index 1d586d4..9b166d7 100644 --- a/drivers/usb/host/pci-quirks.c +++ b/drivers/usb/host/pci-quirks.c @@ -84,65 +84,92 @@ int usb_amd_find_chipset_info(void) { u8 rev = 0; unsigned long flags; + struct amd_chipset_info info; + int ret; spin_lock_irqsave(&amd_lock, flags); - amd_chipset.probe_count++; /* probe only once */ - if (amd_chipset.probe_count > 1) { + if (amd_chipset.probe_count > 0) { + amd_chipset.probe_count++; spin_unlock_irqrestore(&amd_lock, flags); return amd_chipset.probe_result; } + memset(&info, 0, sizeof(info)); + spin_unlock_irqrestore(&amd_lock, flags); - amd_chipset.smbus_dev = pci_get_device(PCI_VENDOR_ID_ATI, 0x4385, NULL); - if (amd_chipset.smbus_dev) { - rev = amd_chipset.smbus_dev->revision; + info.smbus_dev = pci_get_device(PCI_VENDOR_ID_ATI, 0x4385, NULL); + if (info.smbus_dev) { + rev = info.smbus_dev->revision; if (rev >= 0x40) - amd_chipset.sb_type = 1; + info.sb_type = 1; else if (rev >= 0x30 && rev <= 0x3b) - amd_chipset.sb_type = 3; + info.sb_type = 3; } else { - amd_chipset.smbus_dev = pci_get_device(PCI_VENDOR_ID_AMD, - 0x780b, NULL); - if (!amd_chipset.smbus_dev) { - spin_unlock_irqrestore(&amd_lock, flags); - return 0; + info.smbus_dev = pci_get_device(PCI_VENDOR_ID_AMD, + 0x780b, NULL); + if (!info.smbus_dev) { + ret = 0; + goto commit; } - rev = amd_chipset.smbus_dev->revision; + + rev = info.smbus_dev->revision; if (rev >= 0x11 && rev <= 0x18) - amd_chipset.sb_type = 2; + info.sb_type = 2; } - if (amd_chipset.sb_type == 0) { - if (amd_chipset.smbus_dev) { - pci_dev_put(amd_chipset.smbus_dev); - amd_chipset.smbus_dev = NULL; + if (info.sb_type == 0) { + if (info.smbus_dev) { + pci_dev_put(info.smbus_dev); + info.smbus_dev = NULL; } - spin_unlock_irqrestore(&amd_lock, flags); - return 0; + ret = 0; + goto commit; } - amd_chipset.nb_dev = pci_get_device(PCI_VENDOR_ID_AMD, 0x9601, NULL); - if (amd_chipset.nb_dev) { - amd_chipset.nb_type = 1; + info.nb_dev = pci_get_device(PCI_VENDOR_ID_AMD, 0x9601, NULL); + if (info.nb_dev) { + info.nb_type = 1; } else { - amd_chipset.nb_dev = pci_get_device(PCI_VENDOR_ID_AMD, - 0x1510, NULL); - if (amd_chipset.nb_dev) { - amd_chipset.nb_type = 2; - } else { - amd_chipset.nb_dev = pci_get_device(PCI_VENDOR_ID_AMD, - 0x9600, NULL); - if (amd_chipset.nb_dev) - amd_chipset.nb_type = 3; + info.nb_dev = pci_get_device(PCI_VENDOR_ID_AMD, 0x1510, NULL); + if (info.nb_dev) { + info.nb_type = 2; + } else { + info.nb_dev = pci_get_device(PCI_VENDOR_ID_AMD, + 0x9600, NULL); + if (info.nb_dev) + info.nb_type = 3; } } - amd_chipset.probe_result = 1; + ret = info.probe_result = 1; printk(KERN_DEBUG "QUIRK: Enable AMD PLL fix\n"); - spin_unlock_irqrestore(&amd_lock, flags); - return amd_chipset.probe_result; +commit: + + spin_lock_irqsave(&amd_lock, flags); + if (amd_chipset.probe_count > 0) { + /* race - someone else was faster - drop devices */ + + /* Mark that we where here */ + amd_chipset.probe_count++; + ret = amd_chipset.probe_result; + + spin_unlock_irqrestore(&amd_lock, flags); + + if (info.nb_dev) + pci_dev_put(info.nb_dev); + if (info.smbus_dev) + pci_dev_put(info.smbus_dev); + + } else { + /* no race - commit the result */ + info.probe_count++; + amd_chipset = info; + spin_unlock_irqrestore(&amd_lock, flags); + } + + return ret; } EXPORT_SYMBOL_GPL(usb_amd_find_chipset_info); @@ -284,6 +311,7 @@ EXPORT_SYMBOL_GPL(usb_amd_quirk_pll_enable); void usb_amd_dev_put(void) { + struct pci_dev *nb, *smbus; unsigned long flags; spin_lock_irqsave(&amd_lock, flags); @@ -294,20 +322,23 @@ void usb_amd_dev_put(void) return; } - if (amd_chipset.nb_dev) { - pci_dev_put(amd_chipset.nb_dev); - amd_chipset.nb_dev = NULL; - } - if (amd_chipset.smbus_dev) { - pci_dev_put(amd_chipset.smbus_dev); - amd_chipset.smbus_dev = NULL; - } + /* save them to pci_dev_put outside of spinlock */ + nb = amd_chipset.nb_dev; + smbus = amd_chipset.smbus_dev; + + amd_chipset.nb_dev = NULL; + amd_chipset.smbus_dev = NULL; amd_chipset.nb_type = 0; amd_chipset.sb_type = 0; amd_chipset.isoc_reqs = 0; amd_chipset.probe_result = 0; spin_unlock_irqrestore(&amd_lock, flags); + + if (nb) + pci_dev_put(nb); + if (smbus) + pci_dev_put(smbus); } EXPORT_SYMBOL_GPL(usb_amd_dev_put); -- cgit v0.10.2 From 485707116b3949ab41e1aba8d74f45a7551e3a65 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sun, 20 Mar 2011 14:09:50 +0300 Subject: usb: pch_udc: unlock on allocation failure There was an unlock missing on the error path. Also I did a small cleanup by changing ep->dev->lock for just dev->lock. They're the same lock, but dev->lock is shorter and that's how it is used for the spin_unlock_irqrestore() call. Signed-off-by: Dan Carpenter Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/gadget/pch_udc.c b/drivers/usb/gadget/pch_udc.c index 3e4b35e..68dbcc3 100644 --- a/drivers/usb/gadget/pch_udc.c +++ b/drivers/usb/gadget/pch_udc.c @@ -1608,7 +1608,7 @@ static int pch_udc_pcd_queue(struct usb_ep *usbep, struct usb_request *usbreq, return -EINVAL; if (!dev->driver || (dev->gadget.speed == USB_SPEED_UNKNOWN)) return -ESHUTDOWN; - spin_lock_irqsave(&ep->dev->lock, iflags); + spin_lock_irqsave(&dev->lock, iflags); /* map the buffer for dma */ if (usbreq->length && ((usbreq->dma == DMA_ADDR_INVALID) || !usbreq->dma)) { @@ -1625,8 +1625,10 @@ static int pch_udc_pcd_queue(struct usb_ep *usbep, struct usb_request *usbreq, DMA_FROM_DEVICE); } else { req->buf = kzalloc(usbreq->length, GFP_ATOMIC); - if (!req->buf) - return -ENOMEM; + if (!req->buf) { + retval = -ENOMEM; + goto probe_end; + } if (ep->in) { memcpy(req->buf, usbreq->buf, usbreq->length); req->dma = dma_map_single(&dev->pdev->dev, -- cgit v0.10.2 From 80f9df3e0093ad9f1eeefd2ff7fd27daaa518d25 Mon Sep 17 00:00:00 2001 From: "Marius B. Kotsbak" Date: Tue, 22 Mar 2011 00:01:53 +0100 Subject: USB: option: Added support for Samsung GT-B3730/GT-B3710 LTE USB modem. Bind only modem AT command endpoint to option. Signed-off-by: Marius B. Kotsbak Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/serial/option.c b/drivers/usb/serial/option.c index 75c7f45..d77ff04 100644 --- a/drivers/usb/serial/option.c +++ b/drivers/usb/serial/option.c @@ -407,6 +407,10 @@ static void option_instat_callback(struct urb *urb); /* ONDA MT825UP HSDPA 14.2 modem */ #define ONDA_MT825UP 0x000b +/* Samsung products */ +#define SAMSUNG_VENDOR_ID 0x04e8 +#define SAMSUNG_PRODUCT_GT_B3730 0x6889 + /* some devices interfaces need special handling due to a number of reasons */ enum option_blacklist_reason { OPTION_BLACKLIST_NONE = 0, @@ -968,6 +972,7 @@ static const struct usb_device_id option_ids[] = { { USB_DEVICE(OLIVETTI_VENDOR_ID, OLIVETTI_PRODUCT_OLICARD100) }, { USB_DEVICE(CELOT_VENDOR_ID, CELOT_PRODUCT_CT680M) }, /* CT-650 CDMA 450 1xEVDO modem */ { USB_DEVICE(ONDA_VENDOR_ID, ONDA_MT825UP) }, /* ONDA MT825UP modem */ + { USB_DEVICE_AND_INTERFACE_INFO(SAMSUNG_VENDOR_ID, SAMSUNG_PRODUCT_GT_B3730, USB_CLASS_CDC_DATA, 0x00, 0x00) }, /* Samsung GT-B3730/GT-B3710 LTE USB modem.*/ { } /* Terminating entry */ }; MODULE_DEVICE_TABLE(usb, option_ids); -- cgit v0.10.2 From c53c2fab40cf16e13af66f40bfd27200cda98d2f Mon Sep 17 00:00:00 2001 From: Paul Friedrich Date: Fri, 18 Mar 2011 11:13:55 +0100 Subject: USB: ftdi_sio: add ids for Hameg HO720 and HO730 usb serial: ftdi_sio: add two missing USB ID's for Hameg interfaces HO720 and HO730 Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/serial/ftdi_sio.c b/drivers/usb/serial/ftdi_sio.c index 51b7801..4de6ef0 100644 --- a/drivers/usb/serial/ftdi_sio.c +++ b/drivers/usb/serial/ftdi_sio.c @@ -790,6 +790,8 @@ static struct usb_device_id id_table_combined [] = { { USB_DEVICE(FTDI_VID, MARVELL_OPENRD_PID), .driver_info = (kernel_ulong_t)&ftdi_jtag_quirk }, { USB_DEVICE(FTDI_VID, HAMEG_HO820_PID) }, + { USB_DEVICE(FTDI_VID, HAMEG_HO720_PID) }, + { USB_DEVICE(FTDI_VID, HAMEG_HO730_PID) }, { USB_DEVICE(FTDI_VID, HAMEG_HO870_PID) }, { USB_DEVICE(FTDI_VID, MJSG_GENERIC_PID) }, { USB_DEVICE(FTDI_VID, MJSG_SR_RADIO_PID) }, diff --git a/drivers/usb/serial/ftdi_sio_ids.h b/drivers/usb/serial/ftdi_sio_ids.h index 0bdba7a..efffc23 100644 --- a/drivers/usb/serial/ftdi_sio_ids.h +++ b/drivers/usb/serial/ftdi_sio_ids.h @@ -300,6 +300,8 @@ * Hameg HO820 and HO870 interface (using VID 0x0403) */ #define HAMEG_HO820_PID 0xed74 +#define HAMEG_HO730_PID 0xed73 +#define HAMEG_HO720_PID 0xed72 #define HAMEG_HO870_PID 0xed71 /* -- cgit v0.10.2 From 10c9ab15d6aee153968d150c05b3ee3df89673de Mon Sep 17 00:00:00 2001 From: Steven Hardy Date: Mon, 4 Apr 2011 17:57:37 +0100 Subject: usb: Fix qcserial memory leak on rmmod qcprobe function allocates serial->private but this is never freed, this patch adds a new function qc_release() which frees serial->private, after calling usb_wwan_release Signed-off-by: Steven Hardy Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/serial/qcserial.c b/drivers/usb/serial/qcserial.c index 8858201..6e3b933 100644 --- a/drivers/usb/serial/qcserial.c +++ b/drivers/usb/serial/qcserial.c @@ -205,6 +205,18 @@ static int qcprobe(struct usb_serial *serial, const struct usb_device_id *id) return retval; } +static void qc_release(struct usb_serial *serial) +{ + struct usb_wwan_intf_private *priv = usb_get_serial_data(serial); + + dbg("%s", __func__); + + /* Call usb_wwan release & free the private data allocated in qcprobe */ + usb_wwan_release(serial); + usb_set_serial_data(serial, NULL); + kfree(priv); +} + static struct usb_serial_driver qcdevice = { .driver = { .owner = THIS_MODULE, @@ -222,7 +234,7 @@ static struct usb_serial_driver qcdevice = { .chars_in_buffer = usb_wwan_chars_in_buffer, .attach = usb_wwan_startup, .disconnect = usb_wwan_disconnect, - .release = usb_wwan_release, + .release = qc_release, #ifdef CONFIG_PM .suspend = usb_wwan_suspend, .resume = usb_wwan_resume, -- cgit v0.10.2 From 99ab3f9e4eaec35fd2d7159c31b71f17f7e613e3 Mon Sep 17 00:00:00 2001 From: Steven Hardy Date: Mon, 4 Apr 2011 17:59:55 +0100 Subject: usb: qcserial avoid pointing to freed memory Rework the qcprobe logic such that serial->private is not set when qcprobe exits with -ENODEV, otherwise serial->private will point to freed memory on -ENODEV Signed-off-by: Steven Hardy Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/serial/qcserial.c b/drivers/usb/serial/qcserial.c index 6e3b933..cd63864 100644 --- a/drivers/usb/serial/qcserial.c +++ b/drivers/usb/serial/qcserial.c @@ -111,7 +111,7 @@ static int qcprobe(struct usb_serial *serial, const struct usb_device_id *id) ifnum = intf->desc.bInterfaceNumber; dbg("This Interface = %d", ifnum); - data = serial->private = kzalloc(sizeof(struct usb_wwan_intf_private), + data = kzalloc(sizeof(struct usb_wwan_intf_private), GFP_KERNEL); if (!data) return -ENOMEM; @@ -134,8 +134,10 @@ static int qcprobe(struct usb_serial *serial, const struct usb_device_id *id) usb_endpoint_is_bulk_out(&intf->endpoint[1].desc)) { dbg("QDL port found"); - if (serial->interface->num_altsetting == 1) - return 0; + if (serial->interface->num_altsetting == 1) { + retval = 0; /* Success */ + break; + } retval = usb_set_interface(serial->dev, ifnum, 1); if (retval < 0) { @@ -145,7 +147,6 @@ static int qcprobe(struct usb_serial *serial, const struct usb_device_id *id) retval = -ENODEV; kfree(data); } - return retval; } break; @@ -177,7 +178,6 @@ static int qcprobe(struct usb_serial *serial, const struct usb_device_id *id) retval = -ENODEV; kfree(data); } - return retval; } else if (ifnum==3) { /* * NMEA (serial line 9600 8N1) @@ -199,9 +199,12 @@ static int qcprobe(struct usb_serial *serial, const struct usb_device_id *id) dev_err(&serial->dev->dev, "unknown number of interfaces: %d\n", nintf); kfree(data); - return -ENODEV; + retval = -ENODEV; } + /* Set serial->private if not returning -ENODEV */ + if (retval != -ENODEV) + usb_set_serial_data(serial, data); return retval; } -- cgit v0.10.2 From cb62d65f966146a39fdde548cb474dacf1d00fa5 Mon Sep 17 00:00:00 2001 From: Steven Hardy Date: Mon, 4 Apr 2011 18:02:25 +0100 Subject: usb: qcserial add missing errorpath kfrees There are two -ENODEV error paths in qcprobe where the allocated private data is not freed, this patch adds the two missing kfrees to avoid leaking memory on the error path Signed-off-by: Steven Hardy Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/serial/qcserial.c b/drivers/usb/serial/qcserial.c index cd63864..54a9dab 100644 --- a/drivers/usb/serial/qcserial.c +++ b/drivers/usb/serial/qcserial.c @@ -167,6 +167,7 @@ static int qcprobe(struct usb_serial *serial, const struct usb_device_id *id) "Could not set interface, error %d\n", retval); retval = -ENODEV; + kfree(data); } } else if (ifnum == 2) { dbg("Modem port found"); @@ -191,6 +192,7 @@ static int qcprobe(struct usb_serial *serial, const struct usb_device_id *id) "Could not set interface, error %d\n", retval); retval = -ENODEV; + kfree(data); } } break; -- cgit v0.10.2 From d834508e159fe8936f9e7fd941b1e2fe9a209d4b Mon Sep 17 00:00:00 2001 From: Valentin Longchamp Date: Wed, 23 Mar 2011 17:47:00 +0100 Subject: USB: fsl_qe_udc: send ZLP when zero flag and length % maxpacket == 0 The driver did not take the zero flag in the USB request. If the request length is the same as the endpoint's maxpacket, an additional ZLP with no data has to be transmitted. The method used here is inspired to what is done in fsl_udc_core.c (and pxa27x_udc.c and at91_udc.c) where this is supported. There already was a discussion about this topic with people from Keymile, and I propose here a better implementation: http://thread.gmane.org/gmane.linux.usb.general/38951 Signed-off-by: Valentin Longchamp Acked-by: Li Yang Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/gadget/fsl_qe_udc.c b/drivers/usb/gadget/fsl_qe_udc.c index aee7e3c..36613b3 100644 --- a/drivers/usb/gadget/fsl_qe_udc.c +++ b/drivers/usb/gadget/fsl_qe_udc.c @@ -1148,6 +1148,12 @@ static int qe_ep_tx(struct qe_ep *ep, struct qe_frame *frame) static int txcomplete(struct qe_ep *ep, unsigned char restart) { if (ep->tx_req != NULL) { + struct qe_req *req = ep->tx_req; + unsigned zlp = 0, last_len = 0; + + last_len = min_t(unsigned, req->req.length - ep->sent, + ep->ep.maxpacket); + if (!restart) { int asent = ep->last; ep->sent += asent; @@ -1156,9 +1162,18 @@ static int txcomplete(struct qe_ep *ep, unsigned char restart) ep->last = 0; } + /* zlp needed when req->re.zero is set */ + if (req->req.zero) { + if (last_len == 0 || + (req->req.length % ep->ep.maxpacket) != 0) + zlp = 0; + else + zlp = 1; + } else + zlp = 0; + /* a request already were transmitted completely */ - if ((ep->tx_req->req.length - ep->sent) <= 0) { - ep->tx_req->req.actual = (unsigned int)ep->sent; + if (((ep->tx_req->req.length - ep->sent) <= 0) && !zlp) { done(ep, ep->tx_req, 0); ep->tx_req = NULL; ep->last = 0; @@ -1191,6 +1206,7 @@ static int qe_usb_senddata(struct qe_ep *ep, struct qe_frame *frame) buf = (u8 *)ep->tx_req->req.buf + ep->sent; if (buf && size) { ep->last = size; + ep->tx_req->req.actual += size; frame_set_data(frame, buf); frame_set_length(frame, size); frame_set_status(frame, FRAME_OK); -- cgit v0.10.2 From 5808544690300071f09eef9ab83a0fb1f60cf1cd Mon Sep 17 00:00:00 2001 From: Richard Retanubun Date: Thu, 17 Mar 2011 17:39:28 -0400 Subject: USB: isp1760-hcd: move imask clear after pending work is done This patch moves the HcInterrupt register write to clear the pending interrupt to after the isr work is done, doing this removes glitches in the irq line. Signed-off-by: Richard Retanubun Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/host/isp1760-hcd.c b/drivers/usb/host/isp1760-hcd.c index 795345a..564b033 100644 --- a/drivers/usb/host/isp1760-hcd.c +++ b/drivers/usb/host/isp1760-hcd.c @@ -1676,13 +1676,15 @@ static irqreturn_t isp1760_irq(struct usb_hcd *hcd) if (unlikely(!imask)) goto leave; - reg_write32(hcd->regs, HC_INTERRUPT_REG, imask); if (imask & (HC_ATL_INT | HC_SOT_INT)) do_atl_int(hcd); if (imask & HC_INTL_INT) do_intl_int(hcd); + /* Clear interrupt mask on device after the work is done */ + reg_write32(hcd->regs, HC_INTERRUPT_REG, imask); + irqret = IRQ_HANDLED; leave: spin_unlock(&priv->lock); -- cgit v0.10.2 From 2868a2b1ba8f9c7f6c4170519ebb6c62934df70e Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Fri, 18 Mar 2011 21:29:01 -0700 Subject: USB: fix formatting of SuperSpeed endpoints in /proc/bus/usb/devices Isochronous and interrupt SuperSpeed endpoints use the same mechanisms for decoding bInterval values as HighSpeed ones so adjust the code accordingly. Also bandwidth reservation for SuperSpeed matches highspeed, not low/full speed. Signed-off-by: Dmitry Torokhov Cc: stable Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/core/devices.c b/drivers/usb/core/devices.c index a3d2e23..96fdfb8 100644 --- a/drivers/usb/core/devices.c +++ b/drivers/usb/core/devices.c @@ -221,7 +221,7 @@ static char *usb_dump_endpoint_descriptor(int speed, char *start, char *end, break; case USB_ENDPOINT_XFER_INT: type = "Int."; - if (speed == USB_SPEED_HIGH) + if (speed == USB_SPEED_HIGH || speed == USB_SPEED_SUPER) interval = 1 << (desc->bInterval - 1); else interval = desc->bInterval; @@ -229,7 +229,8 @@ static char *usb_dump_endpoint_descriptor(int speed, char *start, char *end, default: /* "can't happen" */ return start; } - interval *= (speed == USB_SPEED_HIGH) ? 125 : 1000; + interval *= (speed == USB_SPEED_HIGH || + speed == USB_SPEED_SUPER) ? 125 : 1000; if (interval % 1000) unit = 'u'; else { @@ -542,8 +543,9 @@ static ssize_t usb_device_dump(char __user **buffer, size_t *nbytes, if (level == 0) { int max; - /* high speed reserves 80%, full/low reserves 90% */ - if (usbdev->speed == USB_SPEED_HIGH) + /* super/high speed reserves 80%, full/low reserves 90% */ + if (usbdev->speed == USB_SPEED_HIGH || + usbdev->speed == USB_SPEED_SUPER) max = 800; else max = FRAME_TIME_MAX_USECS_ALLOC; -- cgit v0.10.2 From 5a6c2f3ff039154872ce597952f8b8900ea0d732 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Sun, 20 Mar 2011 02:15:17 -0700 Subject: USB: xhci - fix unsafe macro definitions Macro arguments used in expressions need to be enclosed in parenthesis to avoid unpleasant surprises. This should be queued for kernels back to 2.6.31 Signed-off-by: Dmitry Torokhov Signed-off-by: Sarah Sharp Cc: stable@kernel.org diff --git a/drivers/usb/host/xhci.h b/drivers/usb/host/xhci.h index 07e2630..91b1a23 100644 --- a/drivers/usb/host/xhci.h +++ b/drivers/usb/host/xhci.h @@ -232,7 +232,7 @@ struct xhci_op_regs { * notification type that matches a bit set in this bit field. */ #define DEV_NOTE_MASK (0xffff) -#define ENABLE_DEV_NOTE(x) (1 << x) +#define ENABLE_DEV_NOTE(x) (1 << (x)) /* Most of the device notification types should only be used for debug. * SW does need to pay attention to function wake notifications. */ @@ -601,11 +601,11 @@ struct xhci_ep_ctx { #define EP_STATE_STOPPED 3 #define EP_STATE_ERROR 4 /* Mult - Max number of burtst within an interval, in EP companion desc. */ -#define EP_MULT(p) ((p & 0x3) << 8) +#define EP_MULT(p) (((p) & 0x3) << 8) /* bits 10:14 are Max Primary Streams */ /* bit 15 is Linear Stream Array */ /* Interval - period between requests to an endpoint - 125u increments. */ -#define EP_INTERVAL(p) ((p & 0xff) << 16) +#define EP_INTERVAL(p) (((p) & 0xff) << 16) #define EP_INTERVAL_TO_UFRAMES(p) (1 << (((p) >> 16) & 0xff)) #define EP_MAXPSTREAMS_MASK (0x1f << 10) #define EP_MAXPSTREAMS(p) (((p) << 10) & EP_MAXPSTREAMS_MASK) -- cgit v0.10.2 From 22e0487047567252d5677ff35766cd884375efc2 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Thu, 17 Mar 2011 22:39:49 +0300 Subject: USB: xhci: unsigned char never equals -1 There were some places that compared port_speed == -1 where port_speed is a u8. This doesn't work unless we cast the -1 to u8. Some places did it correctly. Instead of using -1 directly, I've created a DUPLICATE_ENTRY define which does the cast and is more descriptive as well. Signed-off-by: Dan Carpenter Signed-off-by: Sarah Sharp diff --git a/drivers/usb/host/xhci-mem.c b/drivers/usb/host/xhci-mem.c index a003e79..ab7fc2b 100644 --- a/drivers/usb/host/xhci-mem.c +++ b/drivers/usb/host/xhci-mem.c @@ -846,7 +846,7 @@ static u32 xhci_find_real_port_number(struct xhci_hcd *xhci, * Skip ports that don't have known speeds, or have duplicate * Extended Capabilities port speed entries. */ - if (port_speed == 0 || port_speed == -1) + if (port_speed == 0 || port_speed == DUPLICATE_ENTRY) continue; /* @@ -1727,12 +1727,12 @@ static void xhci_add_in_port(struct xhci_hcd *xhci, unsigned int num_ports, * found a similar duplicate. */ if (xhci->port_array[i] != major_revision && - xhci->port_array[i] != (u8) -1) { + xhci->port_array[i] != DUPLICATE_ENTRY) { if (xhci->port_array[i] == 0x03) xhci->num_usb3_ports--; else xhci->num_usb2_ports--; - xhci->port_array[i] = (u8) -1; + xhci->port_array[i] = DUPLICATE_ENTRY; } /* FIXME: Should we disable the port? */ continue; @@ -1831,7 +1831,7 @@ static int xhci_setup_port_arrays(struct xhci_hcd *xhci, gfp_t flags) for (i = 0; i < num_ports; i++) { if (xhci->port_array[i] == 0x03 || xhci->port_array[i] == 0 || - xhci->port_array[i] == -1) + xhci->port_array[i] == DUPLICATE_ENTRY) continue; xhci->usb2_ports[port_index] = diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c index cfc1ad9..c6d1462 100644 --- a/drivers/usb/host/xhci-ring.c +++ b/drivers/usb/host/xhci-ring.c @@ -1209,7 +1209,7 @@ static unsigned int find_faked_portnum_from_hw_portnum(struct usb_hcd *hcd, * Skip ports that don't have known speeds, or have duplicate * Extended Capabilities port speed entries. */ - if (port_speed == 0 || port_speed == -1) + if (port_speed == 0 || port_speed == DUPLICATE_ENTRY) continue; /* @@ -1260,7 +1260,7 @@ static void handle_port_status(struct xhci_hcd *xhci, port_id); goto cleanup; } - if (major_revision == (u8) -1) { + if (major_revision == DUPLICATE_ENTRY) { xhci_warn(xhci, "Event for port %u duplicated in" "Extended Capabilities, ignoring.\n", port_id); diff --git a/drivers/usb/host/xhci.h b/drivers/usb/host/xhci.h index 91b1a23..bdb78f5 100644 --- a/drivers/usb/host/xhci.h +++ b/drivers/usb/host/xhci.h @@ -348,6 +348,9 @@ struct xhci_op_regs { /* Initiate a warm port reset - complete when PORT_WRC is '1' */ #define PORT_WR (1 << 31) +/* We mark duplicate entries with -1 */ +#define DUPLICATE_ENTRY ((u8)(-1)) + /* Port Power Management Status and Control - port_power_base bitmasks */ /* Inactivity timer value for transitions into U1, in microseconds. * Timeout can be up to 127us. 0xFF means an infinite timeout. -- cgit v0.10.2 From 575688e1e5f462c44ddd608ce3ec9f38b64c3c0d Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Sun, 20 Mar 2011 02:15:16 -0700 Subject: USB: xhci - remove excessive 'inline' markings Remove 'inline' markings from file-local functions and let compiler do its job and inline what makes sense for given architecture. Signed-off-by: Dmitry Torokhov Signed-off-by: Sarah Sharp diff --git a/drivers/usb/host/xhci-mem.c b/drivers/usb/host/xhci-mem.c index ab7fc2b..7f56c9b 100644 --- a/drivers/usb/host/xhci-mem.c +++ b/drivers/usb/host/xhci-mem.c @@ -982,7 +982,7 @@ int xhci_setup_addressable_virt_dev(struct xhci_hcd *xhci, struct usb_device *ud * The NAK interval is one NAK per 1 to 255 microframes, or no NAKs if interval * is set to 0. */ -static inline unsigned int xhci_get_endpoint_interval(struct usb_device *udev, +static unsigned int xhci_get_endpoint_interval(struct usb_device *udev, struct usb_host_endpoint *ep) { unsigned int interval = 0; @@ -1041,7 +1041,7 @@ static inline unsigned int xhci_get_endpoint_interval(struct usb_device *udev, * transaction opportunities per microframe", but that goes in the Max Burst * endpoint context field. */ -static inline u32 xhci_get_endpoint_mult(struct usb_device *udev, +static u32 xhci_get_endpoint_mult(struct usb_device *udev, struct usb_host_endpoint *ep) { if (udev->speed != USB_SPEED_SUPER || @@ -1050,7 +1050,7 @@ static inline u32 xhci_get_endpoint_mult(struct usb_device *udev, return ep->ss_ep_comp.bmAttributes; } -static inline u32 xhci_get_endpoint_type(struct usb_device *udev, +static u32 xhci_get_endpoint_type(struct usb_device *udev, struct usb_host_endpoint *ep) { int in; @@ -1084,7 +1084,7 @@ static inline u32 xhci_get_endpoint_type(struct usb_device *udev, * Basically, this is the maxpacket size, multiplied by the burst size * and mult size. */ -static inline u32 xhci_get_max_esit_payload(struct xhci_hcd *xhci, +static u32 xhci_get_max_esit_payload(struct xhci_hcd *xhci, struct usb_device *udev, struct usb_host_endpoint *ep) { diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c index c6d1462..0161eb0 100644 --- a/drivers/usb/host/xhci-ring.c +++ b/drivers/usb/host/xhci-ring.c @@ -93,7 +93,7 @@ dma_addr_t xhci_trb_virt_to_dma(struct xhci_segment *seg, /* Does this link TRB point to the first segment in a ring, * or was the previous TRB the last TRB on the last segment in the ERST? */ -static inline bool last_trb_on_last_seg(struct xhci_hcd *xhci, struct xhci_ring *ring, +static bool last_trb_on_last_seg(struct xhci_hcd *xhci, struct xhci_ring *ring, struct xhci_segment *seg, union xhci_trb *trb) { if (ring == xhci->event_ring) @@ -107,7 +107,7 @@ static inline bool last_trb_on_last_seg(struct xhci_hcd *xhci, struct xhci_ring * segment? I.e. would the updated event TRB pointer step off the end of the * event seg? */ -static inline int last_trb(struct xhci_hcd *xhci, struct xhci_ring *ring, +static int last_trb(struct xhci_hcd *xhci, struct xhci_ring *ring, struct xhci_segment *seg, union xhci_trb *trb) { if (ring == xhci->event_ring) @@ -116,7 +116,7 @@ static inline int last_trb(struct xhci_hcd *xhci, struct xhci_ring *ring, return (trb->link.control & TRB_TYPE_BITMASK) == TRB_TYPE(TRB_LINK); } -static inline int enqueue_is_link_trb(struct xhci_ring *ring) +static int enqueue_is_link_trb(struct xhci_ring *ring) { struct xhci_link_trb *link = &ring->enqueue->link; return ((link->control & TRB_TYPE_BITMASK) == TRB_TYPE(TRB_LINK)); @@ -592,7 +592,7 @@ void xhci_queue_new_dequeue_state(struct xhci_hcd *xhci, ep->ep_state |= SET_DEQ_PENDING; } -static inline void xhci_stop_watchdog_timer_in_irq(struct xhci_hcd *xhci, +static void xhci_stop_watchdog_timer_in_irq(struct xhci_hcd *xhci, struct xhci_virt_ep *ep) { ep->ep_state &= ~EP_HALT_PENDING; -- cgit v0.10.2 From 926008c9386dde09b015753b6681c502177baa30 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Wed, 23 Mar 2011 20:47:05 -0700 Subject: USB: xhci: simplify logic of skipping missed isoc TDs The logic of the handling Missed Service Error Events was pretty confusing as we were checking the same condition several times. In addition, it caused compiler warning since the compiler could not figure out that event_trb is actually unused in case we are skipping current TD. Fix that by rearranging "skip" condition checks, and factor out skip_isoc_td() so that it is called explicitly. Signed-off-by: Dmitry Torokhov Signed-off-by: Sarah Sharp diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c index 0161eb0..b69a0a1 100644 --- a/drivers/usb/host/xhci-ring.c +++ b/drivers/usb/host/xhci-ring.c @@ -1675,71 +1675,52 @@ static int process_isoc_td(struct xhci_hcd *xhci, struct xhci_td *td, struct urb_priv *urb_priv; int idx; int len = 0; - int skip_td = 0; union xhci_trb *cur_trb; struct xhci_segment *cur_seg; + struct usb_iso_packet_descriptor *frame; u32 trb_comp_code; + bool skip_td = false; ep_ring = xhci_dma_to_transfer_ring(ep, event->buffer); trb_comp_code = GET_COMP_CODE(event->transfer_len); urb_priv = td->urb->hcpriv; idx = urb_priv->td_cnt; + frame = &td->urb->iso_frame_desc[idx]; - if (ep->skip) { - /* The transfer is partly done */ - *status = -EXDEV; - td->urb->iso_frame_desc[idx].status = -EXDEV; - } else { - /* handle completion code */ - switch (trb_comp_code) { - case COMP_SUCCESS: - td->urb->iso_frame_desc[idx].status = 0; - xhci_dbg(xhci, "Successful isoc transfer!\n"); - break; - case COMP_SHORT_TX: - if (td->urb->transfer_flags & URB_SHORT_NOT_OK) - td->urb->iso_frame_desc[idx].status = - -EREMOTEIO; - else - td->urb->iso_frame_desc[idx].status = 0; - break; - case COMP_BW_OVER: - td->urb->iso_frame_desc[idx].status = -ECOMM; - skip_td = 1; - break; - case COMP_BUFF_OVER: - case COMP_BABBLE: - td->urb->iso_frame_desc[idx].status = -EOVERFLOW; - skip_td = 1; - break; - case COMP_STALL: - td->urb->iso_frame_desc[idx].status = -EPROTO; - skip_td = 1; - break; - case COMP_STOP: - case COMP_STOP_INVAL: - break; - default: - td->urb->iso_frame_desc[idx].status = -1; - break; - } - } - - /* calc actual length */ - if (ep->skip) { - td->urb->iso_frame_desc[idx].actual_length = 0; - /* Update ring dequeue pointer */ - while (ep_ring->dequeue != td->last_trb) - inc_deq(xhci, ep_ring, false); - inc_deq(xhci, ep_ring, false); - return finish_td(xhci, td, event_trb, event, ep, status, true); + /* handle completion code */ + switch (trb_comp_code) { + case COMP_SUCCESS: + frame->status = 0; + xhci_dbg(xhci, "Successful isoc transfer!\n"); + break; + case COMP_SHORT_TX: + frame->status = td->urb->transfer_flags & URB_SHORT_NOT_OK ? + -EREMOTEIO : 0; + break; + case COMP_BW_OVER: + frame->status = -ECOMM; + skip_td = true; + break; + case COMP_BUFF_OVER: + case COMP_BABBLE: + frame->status = -EOVERFLOW; + skip_td = true; + break; + case COMP_STALL: + frame->status = -EPROTO; + skip_td = true; + break; + case COMP_STOP: + case COMP_STOP_INVAL: + break; + default: + frame->status = -1; + break; } - if (trb_comp_code == COMP_SUCCESS || skip_td == 1) { - td->urb->iso_frame_desc[idx].actual_length = - td->urb->iso_frame_desc[idx].length; - td->urb->actual_length += - td->urb->iso_frame_desc[idx].length; + if (trb_comp_code == COMP_SUCCESS || skip_td) { + frame->actual_length = frame->length; + td->urb->actual_length += frame->length; } else { for (cur_trb = ep_ring->dequeue, cur_seg = ep_ring->deq_seg; cur_trb != event_trb; @@ -1755,7 +1736,7 @@ static int process_isoc_td(struct xhci_hcd *xhci, struct xhci_td *td, TRB_LEN(event->transfer_len); if (trb_comp_code != COMP_STOP_INVAL) { - td->urb->iso_frame_desc[idx].actual_length = len; + frame->actual_length = len; td->urb->actual_length += len; } } @@ -1766,6 +1747,35 @@ static int process_isoc_td(struct xhci_hcd *xhci, struct xhci_td *td, return finish_td(xhci, td, event_trb, event, ep, status, false); } +static int skip_isoc_td(struct xhci_hcd *xhci, struct xhci_td *td, + struct xhci_transfer_event *event, + struct xhci_virt_ep *ep, int *status) +{ + struct xhci_ring *ep_ring; + struct urb_priv *urb_priv; + struct usb_iso_packet_descriptor *frame; + int idx; + + ep_ring = xhci_dma_to_transfer_ring(ep, event->buffer); + urb_priv = td->urb->hcpriv; + idx = urb_priv->td_cnt; + frame = &td->urb->iso_frame_desc[idx]; + + /* The transfer is partly done */ + *status = -EXDEV; + frame->status = -EXDEV; + + /* calc actual length */ + frame->actual_length = 0; + + /* Update ring dequeue pointer */ + while (ep_ring->dequeue != td->last_trb) + inc_deq(xhci, ep_ring, false); + inc_deq(xhci, ep_ring, false); + + return finish_td(xhci, td, NULL, event, ep, status, true); +} + /* * Process bulk and interrupt tds, update urb status and actual_length. */ @@ -2024,36 +2034,42 @@ static int handle_tx_event(struct xhci_hcd *xhci, } td = list_entry(ep_ring->td_list.next, struct xhci_td, td_list); + /* Is this a TRB in the currently executing TD? */ event_seg = trb_in_td(ep_ring->deq_seg, ep_ring->dequeue, td->last_trb, event_dma); - if (event_seg && ep->skip) { + if (!event_seg) { + if (!ep->skip || + !usb_endpoint_xfer_isoc(&td->urb->ep->desc)) { + /* HC is busted, give up! */ + xhci_err(xhci, + "ERROR Transfer event TRB DMA ptr not " + "part of current TD\n"); + return -ESHUTDOWN; + } + + ret = skip_isoc_td(xhci, td, event, ep, &status); + goto cleanup; + } + + if (ep->skip) { xhci_dbg(xhci, "Found td. Clear skip flag.\n"); ep->skip = false; } - if (!event_seg && - (!ep->skip || !usb_endpoint_xfer_isoc(&td->urb->ep->desc))) { - /* HC is busted, give up! */ - xhci_err(xhci, "ERROR Transfer event TRB DMA ptr not " - "part of current TD\n"); - return -ESHUTDOWN; - } - if (event_seg) { - event_trb = &event_seg->trbs[(event_dma - - event_seg->dma) / sizeof(*event_trb)]; - /* - * No-op TRB should not trigger interrupts. - * If event_trb is a no-op TRB, it means the - * corresponding TD has been cancelled. Just ignore - * the TD. - */ - if ((event_trb->generic.field[3] & TRB_TYPE_BITMASK) - == TRB_TYPE(TRB_TR_NOOP)) { - xhci_dbg(xhci, "event_trb is a no-op TRB. " - "Skip it\n"); - goto cleanup; - } + event_trb = &event_seg->trbs[(event_dma - event_seg->dma) / + sizeof(*event_trb)]; + /* + * No-op TRB should not trigger interrupts. + * If event_trb is a no-op TRB, it means the + * corresponding TD has been cancelled. Just ignore + * the TD. + */ + if ((event_trb->generic.field[3] & TRB_TYPE_BITMASK) + == TRB_TYPE(TRB_TR_NOOP)) { + xhci_dbg(xhci, + "event_trb is a no-op TRB. Skip it\n"); + goto cleanup; } /* Now update the urb's actual_length and give back to -- cgit v0.10.2 From dfa49c4ad120a784ef1ff0717168aa79f55a483a Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Wed, 23 Mar 2011 22:41:23 -0700 Subject: USB: xhci - fix math in xhci_get_endpoint_interval() When parsing exponent-expressed intervals we subtract 1 from the value and then expect it to match with original + 1, which is highly unlikely, and we end with frequent spew: usb 3-4: ep 0x83 - rounding interval to 512 microframes Also, parsing interval for fullspeed isochronous endpoints was incorrect - according to USB spec they use exponent-based intervals (but xHCI spec claims frame-based intervals). I trust USB spec more, especially since USB core agrees with it. This should be queued for stable kernels back to 2.6.31. Reviewed-by: Micah Elizabeth Scott Signed-off-by: Dmitry Torokhov Signed-off-by: Sarah Sharp Cc: stable@kernel.org diff --git a/drivers/usb/host/xhci-mem.c b/drivers/usb/host/xhci-mem.c index 7f56c9b..627f343 100644 --- a/drivers/usb/host/xhci-mem.c +++ b/drivers/usb/host/xhci-mem.c @@ -974,6 +974,47 @@ int xhci_setup_addressable_virt_dev(struct xhci_hcd *xhci, struct usb_device *ud return 0; } +/* + * Convert interval expressed as 2^(bInterval - 1) == interval into + * straight exponent value 2^n == interval. + * + */ +static unsigned int xhci_parse_exponent_interval(struct usb_device *udev, + struct usb_host_endpoint *ep) +{ + unsigned int interval; + + interval = clamp_val(ep->desc.bInterval, 1, 16) - 1; + if (interval != ep->desc.bInterval - 1) + dev_warn(&udev->dev, + "ep %#x - rounding interval to %d microframes\n", + ep->desc.bEndpointAddress, + 1 << interval); + + return interval; +} + +/* + * Convert bInterval expressed in frames (in 1-255 range) to exponent of + * microframes, rounded down to nearest power of 2. + */ +static unsigned int xhci_parse_frame_interval(struct usb_device *udev, + struct usb_host_endpoint *ep) +{ + unsigned int interval; + + interval = fls(8 * ep->desc.bInterval) - 1; + interval = clamp_val(interval, 3, 10); + if ((1 << interval) != 8 * ep->desc.bInterval) + dev_warn(&udev->dev, + "ep %#x - rounding interval to %d microframes, ep desc says %d microframes\n", + ep->desc.bEndpointAddress, + 1 << interval, + 8 * ep->desc.bInterval); + + return interval; +} + /* Return the polling or NAK interval. * * The polling interval is expressed in "microframes". If xHCI's Interval field @@ -991,45 +1032,38 @@ static unsigned int xhci_get_endpoint_interval(struct usb_device *udev, case USB_SPEED_HIGH: /* Max NAK rate */ if (usb_endpoint_xfer_control(&ep->desc) || - usb_endpoint_xfer_bulk(&ep->desc)) + usb_endpoint_xfer_bulk(&ep->desc)) { interval = ep->desc.bInterval; + break; + } /* Fall through - SS and HS isoc/int have same decoding */ + case USB_SPEED_SUPER: if (usb_endpoint_xfer_int(&ep->desc) || - usb_endpoint_xfer_isoc(&ep->desc)) { - if (ep->desc.bInterval == 0) - interval = 0; - else - interval = ep->desc.bInterval - 1; - if (interval > 15) - interval = 15; - if (interval != ep->desc.bInterval + 1) - dev_warn(&udev->dev, "ep %#x - rounding interval to %d microframes\n", - ep->desc.bEndpointAddress, 1 << interval); + usb_endpoint_xfer_isoc(&ep->desc)) { + interval = xhci_parse_exponent_interval(udev, ep); } break; - /* Convert bInterval (in 1-255 frames) to microframes and round down to - * nearest power of 2. - */ + case USB_SPEED_FULL: + if (usb_endpoint_xfer_int(&ep->desc)) { + interval = xhci_parse_exponent_interval(udev, ep); + break; + } + /* + * Fall through for isochronous endpoint interval decoding + * since it uses the same rules as low speed interrupt + * endpoints. + */ + case USB_SPEED_LOW: if (usb_endpoint_xfer_int(&ep->desc) || - usb_endpoint_xfer_isoc(&ep->desc)) { - interval = fls(8*ep->desc.bInterval) - 1; - if (interval > 10) - interval = 10; - if (interval < 3) - interval = 3; - if ((1 << interval) != 8*ep->desc.bInterval) - dev_warn(&udev->dev, - "ep %#x - rounding interval" - " to %d microframes, " - "ep desc says %d microframes\n", - ep->desc.bEndpointAddress, - 1 << interval, - 8*ep->desc.bInterval); + usb_endpoint_xfer_isoc(&ep->desc)) { + + interval = xhci_parse_frame_interval(udev, ep); } break; + default: BUG(); } -- cgit v0.10.2 From 386139d7c8f22d4983ca89de35d339cc41bb0996 Mon Sep 17 00:00:00 2001 From: Sarah Sharp Date: Thu, 24 Mar 2011 08:02:58 -0700 Subject: xhci: Fix NULL pointer deref in handle_port_status() When we get a port status change event, we need to figure out what type of port it came from: a USB 3.0 port, or a USB 2.0/1.1 port. We can't know which usb_hcd to use until that point, so hcd will be NULL for part of the function. Unfortunately, if any of the sanity checks fail, we'll jump to the cleanup label before hcd is set to a valid pointer, and then we'll attempt to tell the USB core to kick the hcd, which is NULL. Skip kicking the roothub if the sanity checks fail. Signed-off-by: Sarah Sharp diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c index b69a0a1..b0b4cc3 100644 --- a/drivers/usb/host/xhci-ring.c +++ b/drivers/usb/host/xhci-ring.c @@ -1235,6 +1235,7 @@ static void handle_port_status(struct xhci_hcd *xhci, u8 major_revision; struct xhci_bus_state *bus_state; u32 __iomem **port_array; + bool bogus_port_status = false; /* Port status change events always have a successful completion code */ if (GET_COMP_CODE(event->generic.field[2]) != COMP_SUCCESS) { @@ -1247,6 +1248,7 @@ static void handle_port_status(struct xhci_hcd *xhci, max_ports = HCS_MAX_PORTS(xhci->hcs_params1); if ((port_id <= 0) || (port_id > max_ports)) { xhci_warn(xhci, "Invalid port id %d\n", port_id); + bogus_port_status = true; goto cleanup; } @@ -1258,12 +1260,14 @@ static void handle_port_status(struct xhci_hcd *xhci, xhci_warn(xhci, "Event for port %u not in " "Extended Capabilities, ignoring.\n", port_id); + bogus_port_status = true; goto cleanup; } if (major_revision == DUPLICATE_ENTRY) { xhci_warn(xhci, "Event for port %u duplicated in" "Extended Capabilities, ignoring.\n", port_id); + bogus_port_status = true; goto cleanup; } @@ -1335,6 +1339,13 @@ cleanup: /* Update event ring dequeue pointer before dropping the lock */ inc_deq(xhci, xhci->event_ring, true); + /* Don't make the USB core poll the roothub if we got a bad port status + * change event. Besides, at that point we can't tell which roothub + * (USB 2.0 or USB 3.0) to kick. + */ + if (bogus_port_status) + return; + spin_unlock(&xhci->lock); /* Pass this up to the core */ usb_hcd_poll_rh_status(hcd); -- cgit v0.10.2 From 943aee0c685d0563228d5a2ad9c8394ad0300fb5 Mon Sep 17 00:00:00 2001 From: Graf Yang Date: Thu, 7 Jan 2010 06:57:30 +0000 Subject: Blackfin: SMP: make all barriers handle cache issues When suspending/resuming, the common task freezing code will run in parallel and freeze processes on each core. This is because the code uses the non-smp version of memory barriers (as well it should). The Blackfin smp barrier logic at the moment contains the cache sync logic, but the non-smp barriers do not. This is incorrect as Rafel summarized: > ... > The existing memory barriers are SMP barriers too, but they are more > than _just_ SMP barriers. At least that's how it is _supposed_ to be > (eg. rmb() is supposed to be stronger than smp_rmb()). > ... > However, looking at the blackfin's definitions of SMP barriers I see > that it uses extra stuff that should _also_ be used in the definitions > of the mandatory barriers. > ... URL: http://lkml.org/lkml/2011/4/13/11 LKML-Reference: Signed-off-by: Graf Yang Signed-off-by: Mike Frysinger diff --git a/arch/blackfin/include/asm/system.h b/arch/blackfin/include/asm/system.h index 19e2c7c..44bd0cc 100644 --- a/arch/blackfin/include/asm/system.h +++ b/arch/blackfin/include/asm/system.h @@ -19,11 +19,11 @@ * Force strict CPU ordering. */ #define nop() __asm__ __volatile__ ("nop;\n\t" : : ) -#define mb() __asm__ __volatile__ ("" : : : "memory") -#define rmb() __asm__ __volatile__ ("" : : : "memory") -#define wmb() __asm__ __volatile__ ("" : : : "memory") -#define set_mb(var, value) do { (void) xchg(&var, value); } while (0) -#define read_barrier_depends() do { } while(0) +#define smp_mb() mb() +#define smp_rmb() rmb() +#define smp_wmb() wmb() +#define set_mb(var, value) do { var = value; mb(); } while (0) +#define smp_read_barrier_depends() read_barrier_depends() #ifdef CONFIG_SMP asmlinkage unsigned long __raw_xchg_1_asm(volatile void *ptr, unsigned long value); @@ -37,16 +37,16 @@ asmlinkage unsigned long __raw_cmpxchg_4_asm(volatile void *ptr, unsigned long new, unsigned long old); #ifdef __ARCH_SYNC_CORE_DCACHE -# define smp_mb() do { barrier(); smp_check_barrier(); smp_mark_barrier(); } while (0) -# define smp_rmb() do { barrier(); smp_check_barrier(); } while (0) -# define smp_wmb() do { barrier(); smp_mark_barrier(); } while (0) -#define smp_read_barrier_depends() do { barrier(); smp_check_barrier(); } while (0) - +/* Force Core data cache coherence */ +# define mb() do { barrier(); smp_check_barrier(); smp_mark_barrier(); } while (0) +# define rmb() do { barrier(); smp_check_barrier(); } while (0) +# define wmb() do { barrier(); smp_mark_barrier(); } while (0) +# define read_barrier_depends() do { barrier(); smp_check_barrier(); } while (0) #else -# define smp_mb() barrier() -# define smp_rmb() barrier() -# define smp_wmb() barrier() -#define smp_read_barrier_depends() barrier() +# define mb() barrier() +# define rmb() barrier() +# define wmb() barrier() +# define read_barrier_depends() do { } while (0) #endif static inline unsigned long __xchg(unsigned long x, volatile void *ptr, @@ -99,10 +99,10 @@ static inline unsigned long __cmpxchg(volatile void *ptr, unsigned long old, #else /* !CONFIG_SMP */ -#define smp_mb() barrier() -#define smp_rmb() barrier() -#define smp_wmb() barrier() -#define smp_read_barrier_depends() do { } while(0) +#define mb() barrier() +#define rmb() barrier() +#define wmb() barrier() +#define read_barrier_depends() do { } while (0) struct __xchg_dummy { unsigned long a[100]; -- cgit v0.10.2 From ce24ee468aabb7c499b910aa3c4ab3cb338326ed Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Mon, 4 Apr 2011 15:20:50 +0000 Subject: Blackfin: gptimers: fix thinko when disabling timers We only want to clear the run bit for this one timer, not all status bits. So don't read the whole reg and then write all the bits back out. Reported-by: Isabelle Leonardi Signed-off-by: Mike Frysinger diff --git a/arch/blackfin/kernel/gptimers.c b/arch/blackfin/kernel/gptimers.c index cdbe075..8b81dc0 100644 --- a/arch/blackfin/kernel/gptimers.c +++ b/arch/blackfin/kernel/gptimers.c @@ -268,7 +268,7 @@ void disable_gptimers(uint16_t mask) _disable_gptimers(mask); for (i = 0; i < MAX_BLACKFIN_GPTIMERS; ++i) if (mask & (1 << i)) - group_regs[BFIN_TIMER_OCTET(i)]->status |= trun_mask[i]; + group_regs[BFIN_TIMER_OCTET(i)]->status = trun_mask[i]; SSYNC(); } EXPORT_SYMBOL(disable_gptimers); -- cgit v0.10.2 From 0bf02ce605b8780223b10739ab7c533de9eb10cc Mon Sep 17 00:00:00 2001 From: Mike Frysinger Date: Mon, 4 Apr 2011 15:26:11 +0000 Subject: Blackfin: time-ts: ack gptimer sooner to avoid missing short ints If the period of a gptimer is fairly low, we might miss an interrupt by acking it too late (we end up acking the new int as well). Reported-by: Isabelle Leonardi Signed-off-by: Mike Frysinger diff --git a/arch/blackfin/kernel/time-ts.c b/arch/blackfin/kernel/time-ts.c index 8c9a43d..cdb4beb 100644 --- a/arch/blackfin/kernel/time-ts.c +++ b/arch/blackfin/kernel/time-ts.c @@ -206,8 +206,14 @@ irqreturn_t bfin_gptmr0_interrupt(int irq, void *dev_id) { struct clock_event_device *evt = dev_id; smp_mb(); - evt->event_handler(evt); + /* + * We want to ACK before we handle so that we can handle smaller timer + * intervals. This way if the timer expires again while we're handling + * things, we're more likely to see that 2nd int rather than swallowing + * it by ACKing the int at the end of this handler. + */ bfin_gptmr0_ack(); + evt->event_handler(evt); return IRQ_HANDLED; } -- cgit v0.10.2 From 8d50de9ee77b38a239dc5b1d6a63ad92a78f119d Mon Sep 17 00:00:00 2001 From: Sonic Zhang Date: Tue, 12 Apr 2011 08:16:04 +0000 Subject: Blackfin: SMP: fix cache flush loop The recent commit (10774912647781) wasn't entirely correct. While it fixed some issues, it introduced others. So pull in the fixes from the public cache flush functions, and document why we need to call things directly ourselves. Signed-off-by: Sonic Zhang Signed-off-by: Mike Frysinger diff --git a/arch/blackfin/mach-common/smp.c b/arch/blackfin/mach-common/smp.c index 6e17a26..8bce5ed 100644 --- a/arch/blackfin/mach-common/smp.c +++ b/arch/blackfin/mach-common/smp.c @@ -109,10 +109,23 @@ static void ipi_flush_icache(void *info) struct blackfin_flush_data *fdata = info; /* Invalidate the memory holding the bounds of the flushed region. */ - invalidate_dcache_range((unsigned long)fdata, - (unsigned long)fdata + sizeof(*fdata)); + blackfin_dcache_invalidate_range((unsigned long)fdata, + (unsigned long)fdata + sizeof(*fdata)); + + /* Make sure all write buffers in the data side of the core + * are flushed before trying to invalidate the icache. This + * needs to be after the data flush and before the icache + * flush so that the SSYNC does the right thing in preventing + * the instruction prefetcher from hitting things in cached + * memory at the wrong time -- it runs much further ahead than + * the pipeline. + */ + SSYNC(); - flush_icache_range(fdata->start, fdata->end); + /* ipi_flaush_icache is invoked by generic flush_icache_range, + * so call blackfin arch icache flush directly here. + */ + blackfin_icache_flush_range(fdata->start, fdata->end); } static void ipi_call_function(unsigned int cpu, struct ipi_message *msg) -- cgit v0.10.2 From 2dea75d96ade3c7cd2bfe73f99c7b3291dc3d03a Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Tue, 12 Apr 2011 23:06:28 -0700 Subject: USB: xhci - also free streams when resetting devices Currently, when resetting a device, xHCI driver disables all but one endpoints and frees their rings, but leaves alone any streams that might have been allocated. Later, when users try to free allocated streams, we oops in xhci_setup_no_streams_ep_input_ctx() because ep->ring is NULL. Let's free not only rings but also stream data as well, so that calling free_streams() on a device that was reset will be safe. This should be queued for stable trees back to 2.6.35. Reviewed-by: Micah Elizabeth Scott Signed-off-by: Dmitry Torokhov Signed-off-by: Sarah Sharp Cc: stable@kernel.org diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c index 196e018..48706c0 100644 --- a/drivers/usb/host/xhci.c +++ b/drivers/usb/host/xhci.c @@ -2386,10 +2386,18 @@ int xhci_discover_or_reset_device(struct usb_hcd *hcd, struct usb_device *udev) /* Everything but endpoint 0 is disabled, so free or cache the rings. */ last_freed_endpoint = 1; for (i = 1; i < 31; ++i) { - if (!virt_dev->eps[i].ring) - continue; - xhci_free_or_cache_endpoint_ring(xhci, virt_dev, i); - last_freed_endpoint = i; + struct xhci_virt_ep *ep = &virt_dev->eps[i]; + + if (ep->ep_state & EP_HAS_STREAMS) { + xhci_free_stream_info(xhci, ep->stream_info); + ep->stream_info = NULL; + ep->ep_state &= ~EP_HAS_STREAMS; + } + + if (ep->ring) { + xhci_free_or_cache_endpoint_ring(xhci, virt_dev, i); + last_freed_endpoint = i; + } } xhci_dbg(xhci, "Output context after successful reset device cmd:\n"); xhci_dbg_ctx(xhci, virt_dev->out_ctx, last_freed_endpoint); -- cgit v0.10.2 From b214f191d95ba4b5a35aebd69cd129cf7e3b1884 Mon Sep 17 00:00:00 2001 From: Matthew Wilcox Date: Tue, 28 Sep 2010 00:57:32 -0400 Subject: USB: Fix unplug of device with active streams If I unplug a device while the UAS driver is loaded, I get an oops in usb_free_streams(). This is because usb_unbind_interface() calls usb_disable_interface() which calls usb_disable_endpoint() which sets ep_out and ep_in to NULL. Then the UAS driver calls usb_pipe_endpoint() which returns a NULL pointer and passes an array of NULL pointers to usb_free_streams(). I think the correct fix for this is to check for the NULL pointer in usb_free_streams() rather than making the driver check for this situation. My original patch for this checked for dev->state == USB_STATE_NOTATTACHED, but the call to usb_disable_interface() is conditional, so not all drivers would want this check. Note from Sarah Sharp: This patch does avoid a potential dereference, but the real fix (which will be implemented later) is to set the .soft_unbind flag in the usb_driver structure for the UAS driver, and all drivers that allocate streams. The driver should free any streams when it is unbound from the interface. This avoids leaking stream rings in the xHCI driver when usb_disable_interface() is called. This should be queued for stable trees back to 2.6.35. Signed-off-by: Matthew Wilcox Signed-off-by: Sarah Sharp Cc: stable@kernel.org diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c index 8eed05d2..77a7fae 100644 --- a/drivers/usb/core/hcd.c +++ b/drivers/usb/core/hcd.c @@ -1908,7 +1908,7 @@ void usb_free_streams(struct usb_interface *interface, /* Streams only apply to bulk endpoints. */ for (i = 0; i < num_eps; i++) - if (!usb_endpoint_xfer_bulk(&eps[i]->desc)) + if (!eps[i] || !usb_endpoint_xfer_bulk(&eps[i]->desc)) return; hcd->driver->free_streams(hcd, dev, eps, num_eps, mem_flags); -- cgit v0.10.2 From a8f08d86dbf1b7bb5869cf1807d2fd40ec9d6d0a Mon Sep 17 00:00:00 2001 From: Andiry Xu Date: Thu, 31 Mar 2011 14:56:50 +0800 Subject: usbcore: Bug fix: system can't suspend with USB3.0 device connected to USB3.0 hub This patch clear PORT_POWER when suspend a USB3.0 device behind a USB3.0 external hub, so the system can suspend and resume. Note USB3.0 device may not work after system resume and this is a temporary workaround. The correct fix will be in future patches. Signed-off-by: Andiry Xu Signed-off-by: Sarah Sharp diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c index 8fb7549..93720bdc 100644 --- a/drivers/usb/core/hub.c +++ b/drivers/usb/core/hub.c @@ -2285,7 +2285,17 @@ int usb_port_suspend(struct usb_device *udev, pm_message_t msg) } /* see 7.1.7.6 */ - status = set_port_feature(hub->hdev, port1, USB_PORT_FEAT_SUSPEND); + /* Clear PORT_POWER if it's a USB3.0 device connected to USB 3.0 + * external hub. + * FIXME: this is a temporary workaround to make the system able + * to suspend/resume. + */ + if ((hub->hdev->parent != NULL) && hub_is_superspeed(hub->hdev)) + status = clear_port_feature(hub->hdev, port1, + USB_PORT_FEAT_POWER); + else + status = set_port_feature(hub->hdev, port1, + USB_PORT_FEAT_SUSPEND); if (status) { dev_dbg(hub->intfdev, "can't suspend port %d, status %d\n", port1, status); -- cgit v0.10.2 From fedd383e33f9ba9b91626f72c593ea327403bf59 Mon Sep 17 00:00:00 2001 From: Sarah Sharp Date: Tue, 12 Apr 2011 17:43:19 -0700 Subject: xhci: Tell USB core both roothubs lost power. On a resume, when the power is lost during hibernate, the USB core will call hub_reset_resume for the xHCI USB 2.0 roothub, but not for the USB 3.0 roothub: [ 164.748310] usb usb1: root hub lost power or was reset [ 164.748353] usb usb2: root hub lost power or was reset [ 164.748487] usb usb3: root hub lost power or was reset [ 164.748488] xhci_hcd 0000:01:00.0: Stop HCD ... [ 164.870039] hub 4-0:1.0: hub_resume ... [ 164.870054] hub 3-0:1.0: hub_reset_resume This causes issues later, because the USB core assumes the USB 3.0 hub attached to the USB 3.0 roothub is still active. It attempts to queue a control URB for the external hub, which fails because all the device slot contexts were released when the USB 3.0 roothub lost power: [ 164.980044] hub 4-1:1.0: hub_resume [ 164.980047] xhci_hcd 0000:01:00.0: Get port status returned 0x10101 [ 164.980049] xHCI xhci_urb_enqueue called with unaddressed device [ 164.980053] hub 3-0:1.0: port 1: status 0101 change 0001 [ 164.980056] hub 4-1:1.0: hub_port_status failed (err = -22) [ 164.980060] xhci_hcd 0000:01:00.0: `MEM_WRITE_DWORD(3'b000, 32'hffffc90008948440, 32'h202e1, 4'hf); [ 164.980062] xHCI xhci_urb_enqueue called with unaddressed device [ 164.980066] xhci_hcd 0000:01:00.0: clear port connect change, actual port 0 status = 0x2e1 [ 164.980069] hub 4-1:1.0: hub_port_status failed (err = -22) [ 164.980072] xhci_hcd 0000:01:00.0: get port status, actual port 1 status = 0x2a0 [ 164.980074] xHCI xhci_urb_enqueue called with unaddressed device [ 164.980077] xhci_hcd 0000:01:00.0: Get port status returned 0x100 [ 164.980079] hub 4-1:1.0: hub_port_status failed (err = -22) [ 164.980082] xHCI xhci_urb_enqueue called with unaddressed device [ 164.980085] hub 4-1:1.0: hub_port_status failed (err = -22) [ 164.980088] hub 4-1:1.0: port 4: status 0000 change 0000 [ 164.980091] xHCI xhci_urb_enqueue called with unaddressed device [ 164.980094] hub 4-1:1.0: activate --> -22 [ 164.980113] xHCI xhci_urb_enqueue called with unaddressed device [ 164.980117] hub 4-1:1.0: hub_port_status failed (err = -22) [ 164.980119] xHCI xhci_urb_enqueue called with unaddressed device [ 164.980123] hub 4-1:1.0: can't resume port 4, status -22 [ 164.980126] hub 4-1:1.0: port 4 status ffff.ffff after resume, -22 [ 164.980129] usb 4-1.4: can't resume, status -22 [ 164.980131] hub 4-1:1.0: logical disconnect on port 4 This causes issues when a USB 3.0 hard drive is attached to the external USB 3.0 hub when the system is hibernated: [ 6249.849653] sd 8:0:0:0: [sdb] Unhandled error code [ 6249.849659] sd 8:0:0:0: [sdb] Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK [ 6249.849663] sd 8:0:0:0: [sdb] CDB: Read(10): 28 00 00 00 2a 08 00 00 02 00 [ 6249.849671] end_request: I/O error, dev sdb, sector 10760 Make sure to inform the USB core that *both* xHCI roothubs lost power. Signed-off-by: Sarah Sharp diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c index 48706c0..c41358e 100644 --- a/drivers/usb/host/xhci.c +++ b/drivers/usb/host/xhci.c @@ -771,7 +771,9 @@ int xhci_resume(struct xhci_hcd *xhci, bool hibernated) /* If restore operation fails, re-initialize the HC during resume */ if ((temp & STS_SRE) || hibernated) { - usb_root_hub_lost_power(hcd->self.root_hub); + /* Let the USB core know _both_ roothubs lost power. */ + usb_root_hub_lost_power(xhci->main_hcd->self.root_hub); + usb_root_hub_lost_power(xhci->shared_hcd->self.root_hub); xhci_dbg(xhci, "Stop HCD\n"); xhci_halt(xhci); -- cgit v0.10.2 From c41136b05d3fb213a192f76a5688ff83687c1136 Mon Sep 17 00:00:00 2001 From: Andiry Xu Date: Tue, 22 Mar 2011 17:08:14 +0800 Subject: xHCI: Implement AMD PLL quirk This patch disable the optional PM feature inside the Hudson3 platform under the following conditions: 1. If an isochronous device is connected to xHCI port and is active; 2. Optional PM feature that powers down the internal Bus PLL when the link is in low power state is enabled. The PM feature needs to be disabled to eliminate PLL startup delays when the link comes out of low power state. The performance of DMA data transfer could be impacted if system delay were encountered and in addition to the PLL start up delays. Disabling the PM would leave room for unpredictable system delays in order to guarantee uninterrupted data transfer to isochronous audio or video stream devices that require time sensitive information. If data in an audio/video stream was interrupted then erratic audio or video performance may be encountered. AMD PLL quirk is already implemented in OHCI/EHCI driver. After moving the quirk code to pci-quirks.c and export them, xHCI driver can call it directly without having the quirk implementation in itself. Signed-off-by: Andiry Xu Signed-off-by: Sarah Sharp diff --git a/drivers/usb/host/xhci-pci.c b/drivers/usb/host/xhci-pci.c index ceea9f3..a10494c 100644 --- a/drivers/usb/host/xhci-pci.c +++ b/drivers/usb/host/xhci-pci.c @@ -114,6 +114,10 @@ static int xhci_pci_setup(struct usb_hcd *hcd) if (pdev->vendor == PCI_VENDOR_ID_NEC) xhci->quirks |= XHCI_NEC_HOST; + /* AMD PLL quirk */ + if (pdev->vendor == PCI_VENDOR_ID_AMD && usb_amd_find_chipset_info()) + xhci->quirks |= XHCI_AMD_PLL_FIX; + /* Make sure the HC is halted. */ retval = xhci_halt(xhci); if (retval) diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c index b0b4cc3..7437386 100644 --- a/drivers/usb/host/xhci-ring.c +++ b/drivers/usb/host/xhci-ring.c @@ -619,6 +619,13 @@ static void xhci_giveback_urb_in_irq(struct xhci_hcd *xhci, /* Only giveback urb when this is the last td in urb */ if (urb_priv->td_cnt == urb_priv->length) { + if (usb_pipetype(urb->pipe) == PIPE_ISOCHRONOUS) { + xhci_to_hcd(xhci)->self.bandwidth_isoc_reqs--; + if (xhci_to_hcd(xhci)->self.bandwidth_isoc_reqs == 0) { + if (xhci->quirks & XHCI_AMD_PLL_FIX) + usb_amd_quirk_pll_enable(); + } + } usb_hcd_unlink_urb_from_ep(hcd, urb); xhci_dbg(xhci, "Giveback %s URB %p\n", adjective, urb); @@ -1565,8 +1572,17 @@ td_cleanup: urb_priv->td_cnt++; /* Giveback the urb when all the tds are completed */ - if (urb_priv->td_cnt == urb_priv->length) + if (urb_priv->td_cnt == urb_priv->length) { ret = 1; + if (usb_pipetype(urb->pipe) == PIPE_ISOCHRONOUS) { + xhci_to_hcd(xhci)->self.bandwidth_isoc_reqs--; + if (xhci_to_hcd(xhci)->self.bandwidth_isoc_reqs + == 0) { + if (xhci->quirks & XHCI_AMD_PLL_FIX) + usb_amd_quirk_pll_enable(); + } + } + } } return ret; @@ -3153,6 +3169,12 @@ static int xhci_queue_isoc_tx(struct xhci_hcd *xhci, gfp_t mem_flags, } } + if (xhci_to_hcd(xhci)->self.bandwidth_isoc_reqs == 0) { + if (xhci->quirks & XHCI_AMD_PLL_FIX) + usb_amd_quirk_pll_disable(); + } + xhci_to_hcd(xhci)->self.bandwidth_isoc_reqs++; + giveback_first_trb(xhci, slot_id, ep_index, urb->stream_id, start_cycle, start_trb); return 0; diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c index c41358e..81b976e 100644 --- a/drivers/usb/host/xhci.c +++ b/drivers/usb/host/xhci.c @@ -550,6 +550,9 @@ void xhci_stop(struct usb_hcd *hcd) del_timer_sync(&xhci->event_ring_timer); #endif + if (xhci->quirks & XHCI_AMD_PLL_FIX) + usb_amd_dev_put(); + xhci_dbg(xhci, "// Disabling event ring interrupts\n"); temp = xhci_readl(xhci, &xhci->op_regs->status); xhci_writel(xhci, temp & ~STS_EINT, &xhci->op_regs->status); diff --git a/drivers/usb/host/xhci.h b/drivers/usb/host/xhci.h index bdb78f5..ba1be6b 100644 --- a/drivers/usb/host/xhci.h +++ b/drivers/usb/host/xhci.h @@ -30,6 +30,7 @@ /* Code sharing between pci-quirks and xhci hcd */ #include "xhci-ext-caps.h" +#include "pci-quirks.h" /* xHCI PCI Configuration Registers */ #define XHCI_SBRN_OFFSET (0x60) @@ -1279,6 +1280,7 @@ struct xhci_hcd { #define XHCI_LINK_TRB_QUIRK (1 << 0) #define XHCI_RESET_EP_QUIRK (1 << 1) #define XHCI_NEC_HOST (1 << 2) +#define XHCI_AMD_PLL_FIX (1 << 3) /* There are two roothubs to keep track of bus suspend info for */ struct xhci_bus_state bus_state[2]; /* Is each xHCI roothub port a USB 3.0, USB 2.0, or USB 1.1 port? */ -- cgit v0.10.2 From dc5b966214cdbc4000f1da03a231952a701ebe6f Mon Sep 17 00:00:00 2001 From: Abhilash Kesavan Date: Fri, 25 Mar 2011 17:00:48 +0900 Subject: ARM: S5P: Remove unused s3c_pm_check_resume_pin The s3c_pm_check_resume_pin() is not being used and can be safely removed to fix the build warning. Signed-off-by: Abhilash Kesavan Signed-off-by: Kukjin Kim diff --git a/arch/arm/plat-s5p/pm.c b/arch/arm/plat-s5p/pm.c index d592b63..d15dc47 100644 --- a/arch/arm/plat-s5p/pm.c +++ b/arch/arm/plat-s5p/pm.c @@ -19,17 +19,6 @@ #define PFX "s5p pm: " -/* s3c_pm_check_resume_pin - * - * check to see if the pin is configured correctly for sleep mode, and - * make any necessary adjustments if it is not -*/ - -static void s3c_pm_check_resume_pin(unsigned int pin, unsigned int irqoffs) -{ - /* nothing here yet */ -} - /* s3c_pm_configure_extint * * configure all external interrupt pins -- cgit v0.10.2 From 96cfb97dd4474da6f3c30b2bfbe2286e5554fc97 Mon Sep 17 00:00:00 2001 From: Abhilash Kesavan Date: Fri, 25 Mar 2011 17:45:19 +0900 Subject: ARM: SAMSUNG: Fix build failure in PM CRC check code This patch fixes build error that occurs on enabling the Samsung specific PM CRC check code. Missed removing this reference of s3c_sleep_save_phys during move to generic cpu suspend/resume support. Signed-off-by: Abhilash Kesavan Cc: Russell King Signed-off-by: Kukjin Kim diff --git a/arch/arm/plat-samsung/pm-check.c b/arch/arm/plat-samsung/pm-check.c index e4baf76..6b733fa 100644 --- a/arch/arm/plat-samsung/pm-check.c +++ b/arch/arm/plat-samsung/pm-check.c @@ -164,7 +164,6 @@ static inline int in_region(void *ptr, int size, void *what, size_t whatsz) */ static u32 *s3c_pm_runcheck(struct resource *res, u32 *val) { - void *save_at = phys_to_virt(s3c_sleep_save_phys); unsigned long addr; unsigned long left; void *stkpage; @@ -192,11 +191,6 @@ static u32 *s3c_pm_runcheck(struct resource *res, u32 *val) goto skip_check; } - if (in_region(ptr, left, save_at, 32*4 )) { - S3C_PMDBG("skipping %08lx, has save block in\n", addr); - goto skip_check; - } - /* calculate and check the checksum */ calc = crc32_le(~0, ptr, left); -- cgit v0.10.2 From baab7307c7af963fbd993149d50dd45232b9d04c Mon Sep 17 00:00:00 2001 From: Maurus Cuelenaere Date: Sat, 2 Apr 2011 10:50:22 +0900 Subject: ARM: SAMSUNG: Fix warning 's3c_pm_show_resume_irqs' defined but not used s3c_pm_show_resume_irqs() is used by some s3c_pm_arch_show_resume_irqs() implementations, which get included through mach/pm-core.h. Add __maybe_unused to silence warnings when it isn't used (e.g. on S3C64XX platforms). Signed-off-by: Maurus Cuelenaere Signed-off-by: Kukjin Kim diff --git a/arch/arm/plat-samsung/pm.c b/arch/arm/plat-samsung/pm.c index d5b58d3..5c0a440 100644 --- a/arch/arm/plat-samsung/pm.c +++ b/arch/arm/plat-samsung/pm.c @@ -214,8 +214,9 @@ void s3c_pm_do_restore_core(struct sleep_save *ptr, int count) * * print any IRQs asserted at resume time (ie, we woke from) */ -static void s3c_pm_show_resume_irqs(int start, unsigned long which, - unsigned long mask) +static void __maybe_unused s3c_pm_show_resume_irqs(int start, + unsigned long which, + unsigned long mask) { int i; -- cgit v0.10.2 From 38a2f37258f9e2ae3f6e4241e01088be8dfaf4e9 Mon Sep 17 00:00:00 2001 From: huajun li Date: Wed, 13 Apr 2011 15:43:32 +0000 Subject: usbnet: Fix up 'FLAG_POINTTOPOINT' and 'FLAG_MULTI_PACKET' overlaps. USB tethering does not work anymore since 2.6.39-rc2, but it's okay in -rc1. The root cause is the new added mask code 'FLAG_POINTTOPOINT' overlaps 'FLAG_MULTI_PACKET' in include/linux/usb/usbnet.h, this causes logic issue in rx_process(). This patch cleans up the overlap. Reported-and-Tested-by: Gottfried Haider Signed-off-by: Huajun Li Signed-off-by: David S. Miller diff --git a/include/linux/usb/usbnet.h b/include/linux/usb/usbnet.h index 3c7329b..0e18550 100644 --- a/include/linux/usb/usbnet.h +++ b/include/linux/usb/usbnet.h @@ -103,8 +103,8 @@ struct driver_info { * Indicates to usbnet, that USB driver accumulates multiple IP packets. * Affects statistic (counters) and short packet handling. */ -#define FLAG_MULTI_PACKET 0x1000 -#define FLAG_RX_ASSEMBLE 0x2000 /* rx packets may span >1 frames */ +#define FLAG_MULTI_PACKET 0x2000 +#define FLAG_RX_ASSEMBLE 0x4000 /* rx packets may span >1 frames */ /* init device ... can sleep, or cause probe() failure */ int (*bind)(struct usbnet *, struct usb_interface *); -- cgit v0.10.2 From b025a3f836d1e8785ae65b59282a4befef9892bb Mon Sep 17 00:00:00 2001 From: Stephen Boyd Date: Mon, 11 Apr 2011 19:06:26 +0100 Subject: ARM: 6876/1: Kconfig.debug: Remove unused CONFIG_DEBUG_ERRORS This config option isn't actually used anywhere and can be safely removed. The last user was traps.c before commit 082f47a ([ARM] always allow dump_stack() to produce a backtrace, 2007-07-05). Reviewed-by: Jesper Juhl Signed-off-by: Stephen Boyd Signed-off-by: Russell King diff --git a/arch/arm/Kconfig.debug b/arch/arm/Kconfig.debug index 494224a..03d01d7 100644 --- a/arch/arm/Kconfig.debug +++ b/arch/arm/Kconfig.debug @@ -63,17 +63,6 @@ config DEBUG_USER 8 - SIGSEGV faults 16 - SIGBUS faults -config DEBUG_ERRORS - bool "Verbose kernel error messages" - depends on DEBUG_KERNEL - help - This option controls verbose debugging information which can be - printed when the kernel detects an internal error. This debugging - information is useful to kernel hackers when tracking down problems, - but mostly meaningless to other people. It's safe to say Y unless - you are concerned with the code size or don't want to see these - messages. - config DEBUG_STACK_USAGE bool "Enable stack utilization instrumentation" depends on DEBUG_KERNEL -- cgit v0.10.2 From df5419a9a90ac4ea2d853d68cc788e32cfe71278 Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Wed, 13 Apr 2011 04:57:17 +0100 Subject: ARM: 6877/1: the ADDR_NO_RANDOMIZE personality flag should be honored with mmap() Signed-off-by: Nicolas Pitre Signed-off-by: Russell King diff --git a/arch/arm/mm/mmap.c b/arch/arm/mm/mmap.c index afe209e..74be05f 100644 --- a/arch/arm/mm/mmap.c +++ b/arch/arm/mm/mmap.c @@ -7,6 +7,7 @@ #include #include #include +#include #include #include #include @@ -82,7 +83,8 @@ arch_get_unmapped_area(struct file *filp, unsigned long addr, mm->cached_hole_size = 0; } /* 8 bits of randomness in 20 address space bits */ - if (current->flags & PF_RANDOMIZE) + if ((current->flags & PF_RANDOMIZE) && + !(current->personality & ADDR_NO_RANDOMIZE)) addr += (get_random_int() % (1 << 8)) << PAGE_SHIFT; full_search: -- cgit v0.10.2 From 5e143436d04465c937c1a242808a99c46393af3e Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Wed, 13 Apr 2011 04:59:36 +0100 Subject: ARM: 6878/1: fix personality flag propagation across an exec Our SET_PERSONALITY() implementation was overwriting all existing personality flags, including ADDR_NO_RANDOMIZE, making them unavailable to processes being exec'd after a call to personality() in user space. This prevents the gdb test suite from running successfully. Signed-off-by: Nicolas Pitre Signed-off-by: Russell King diff --git a/arch/arm/kernel/elf.c b/arch/arm/kernel/elf.c index d4a0da1..9b05c6a 100644 --- a/arch/arm/kernel/elf.c +++ b/arch/arm/kernel/elf.c @@ -40,15 +40,22 @@ EXPORT_SYMBOL(elf_check_arch); void elf_set_personality(const struct elf32_hdr *x) { unsigned int eflags = x->e_flags; - unsigned int personality = PER_LINUX_32BIT; + unsigned int personality = current->personality & ~PER_MASK; + + /* + * We only support Linux ELF executables, so always set the + * personality to LINUX. + */ + personality |= PER_LINUX; /* * APCS-26 is only valid for OABI executables */ - if ((eflags & EF_ARM_EABI_MASK) == EF_ARM_EABI_UNKNOWN) { - if (eflags & EF_ARM_APCS_26) - personality = PER_LINUX; - } + if ((eflags & EF_ARM_EABI_MASK) == EF_ARM_EABI_UNKNOWN && + (eflags & EF_ARM_APCS_26)) + personality &= ~ADDR_LIMIT_32BIT; + else + personality |= ADDR_LIMIT_32BIT; set_personality(personality); -- cgit v0.10.2 From 88b9ef452690233d200abf57a3fa2c0f3bd874c5 Mon Sep 17 00:00:00 2001 From: Nicolas Pitre Date: Wed, 13 Apr 2011 05:01:52 +0100 Subject: ARM: 6879/1: fix personality test wrt usage of domain handlers There are optional bits that may complement a personality ID. It is therefore wrong to simply test against the absolute current->personality value to determine the effective personality. The PER_LINUX_32BIT is itself just PER_LINUX with one of those optional bits set. Signed-off-by: Nicolas Pitre Signed-off-by: Russell King diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c index f0000e1..3b54ad1 100644 --- a/arch/arm/kernel/traps.c +++ b/arch/arm/kernel/traps.c @@ -410,8 +410,7 @@ static int bad_syscall(int n, struct pt_regs *regs) struct thread_info *thread = current_thread_info(); siginfo_t info; - if (current->personality != PER_LINUX && - current->personality != PER_LINUX_32BIT && + if ((current->personality & PER_MASK) != PER_LINUX && thread->exec_domain->handler) { thread->exec_domain->handler(n, regs); return regs->ARM_r0; -- cgit v0.10.2 From 753d8534cc190ed144caebc2ea49ab7a43dca662 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Thu, 14 Apr 2011 13:37:07 -0700 Subject: Revert "USB: isp1760-hcd: move imask clear after pending work is done" This reverts commit 5808544690300071f09eef9ab83a0fb1f60cf1cd. To quote Richard: I don't think this should be mainlined. It was a misunderstanding on my part. If you see all the other hdc drivers in the same location, they all do the same thing (i.e. clear the interrupt status first, then do the work) that "glitch" I think I saw was actually two back-to-back interrupts. Sebastian (the original author of isp1760) explained it to me a few days after my submission. sorry for the confusion Cc: Richard Retanubun Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/usb/host/isp1760-hcd.c b/drivers/usb/host/isp1760-hcd.c index 564b033..795345a 100644 --- a/drivers/usb/host/isp1760-hcd.c +++ b/drivers/usb/host/isp1760-hcd.c @@ -1676,15 +1676,13 @@ static irqreturn_t isp1760_irq(struct usb_hcd *hcd) if (unlikely(!imask)) goto leave; + reg_write32(hcd->regs, HC_INTERRUPT_REG, imask); if (imask & (HC_ATL_INT | HC_SOT_INT)) do_atl_int(hcd); if (imask & HC_INTL_INT) do_intl_int(hcd); - /* Clear interrupt mask on device after the work is done */ - reg_write32(hcd->regs, HC_INTERRUPT_REG, imask); - irqret = IRQ_HANDLED; leave: spin_unlock(&priv->lock); -- cgit v0.10.2 From 79ea6c8966b861f80e77a14b2acf581563f3c83e Mon Sep 17 00:00:00 2001 From: Rasesh Mody Date: Thu, 14 Apr 2011 08:05:18 +0000 Subject: bna: fix for clean fw re-initialization During a kernel crash, bna control path state machine and firmware do not get a notification and hence are not cleanly shutdown. The registers holding driver/IOC state information are not reset back to valid disabled/parking values. This causes subsequent driver initialization to hang during kdump kernel boot. This patch, during the initialization of first PCI function, resets corresponding register when unclean shutown is detect by reading chip registers. This will make sure that ioc/fw gets clean re-initialization. Signed-off-by: Debashis Dutt Signed-off-by: Rasesh Mody Signed-off-by: David S. Miller diff --git a/drivers/net/bna/bfa_ioc.c b/drivers/net/bna/bfa_ioc.c index e3de0b8..7581518 100644 --- a/drivers/net/bna/bfa_ioc.c +++ b/drivers/net/bna/bfa_ioc.c @@ -38,6 +38,8 @@ #define bfa_ioc_map_port(__ioc) ((__ioc)->ioc_hwif->ioc_map_port(__ioc)) #define bfa_ioc_notify_fail(__ioc) \ ((__ioc)->ioc_hwif->ioc_notify_fail(__ioc)) +#define bfa_ioc_sync_start(__ioc) \ + ((__ioc)->ioc_hwif->ioc_sync_start(__ioc)) #define bfa_ioc_sync_join(__ioc) \ ((__ioc)->ioc_hwif->ioc_sync_join(__ioc)) #define bfa_ioc_sync_leave(__ioc) \ @@ -602,7 +604,7 @@ bfa_iocpf_sm_fwcheck(struct bfa_iocpf *iocpf, enum iocpf_event event) switch (event) { case IOCPF_E_SEMLOCKED: if (bfa_ioc_firmware_lock(ioc)) { - if (bfa_ioc_sync_complete(ioc)) { + if (bfa_ioc_sync_start(ioc)) { iocpf->retry_count = 0; bfa_ioc_sync_join(ioc); bfa_fsm_set_state(iocpf, bfa_iocpf_sm_hwinit); @@ -1314,7 +1316,7 @@ bfa_nw_ioc_fwver_cmp(struct bfa_ioc *ioc, struct bfi_ioc_image_hdr *fwhdr) * execution context (driver/bios) must match. */ static bool -bfa_ioc_fwver_valid(struct bfa_ioc *ioc) +bfa_ioc_fwver_valid(struct bfa_ioc *ioc, u32 boot_env) { struct bfi_ioc_image_hdr fwhdr, *drv_fwhdr; @@ -1325,7 +1327,7 @@ bfa_ioc_fwver_valid(struct bfa_ioc *ioc) if (fwhdr.signature != drv_fwhdr->signature) return false; - if (fwhdr.exec != drv_fwhdr->exec) + if (swab32(fwhdr.param) != boot_env) return false; return bfa_nw_ioc_fwver_cmp(ioc, &fwhdr); @@ -1352,9 +1354,12 @@ bfa_ioc_hwinit(struct bfa_ioc *ioc, bool force) { enum bfi_ioc_state ioc_fwstate; bool fwvalid; + u32 boot_env; ioc_fwstate = readl(ioc->ioc_regs.ioc_fwstate); + boot_env = BFI_BOOT_LOADER_OS; + if (force) ioc_fwstate = BFI_IOC_UNINIT; @@ -1362,10 +1367,10 @@ bfa_ioc_hwinit(struct bfa_ioc *ioc, bool force) * check if firmware is valid */ fwvalid = (ioc_fwstate == BFI_IOC_UNINIT) ? - false : bfa_ioc_fwver_valid(ioc); + false : bfa_ioc_fwver_valid(ioc, boot_env); if (!fwvalid) { - bfa_ioc_boot(ioc, BFI_BOOT_TYPE_NORMAL, ioc->pcidev.device_id); + bfa_ioc_boot(ioc, BFI_BOOT_TYPE_NORMAL, boot_env); return; } @@ -1396,7 +1401,7 @@ bfa_ioc_hwinit(struct bfa_ioc *ioc, bool force) /** * Initialize the h/w for any other states. */ - bfa_ioc_boot(ioc, BFI_BOOT_TYPE_NORMAL, ioc->pcidev.device_id); + bfa_ioc_boot(ioc, BFI_BOOT_TYPE_NORMAL, boot_env); } void @@ -1506,7 +1511,7 @@ bfa_ioc_hb_stop(struct bfa_ioc *ioc) */ static void bfa_ioc_download_fw(struct bfa_ioc *ioc, u32 boot_type, - u32 boot_param) + u32 boot_env) { u32 *fwimg; u32 pgnum, pgoff; @@ -1558,10 +1563,10 @@ bfa_ioc_download_fw(struct bfa_ioc *ioc, u32 boot_type, /* * Set boot type and boot param at the end. */ - writel((swab32(swab32(boot_type))), ((ioc->ioc_regs.smem_page_start) + writel(boot_type, ((ioc->ioc_regs.smem_page_start) + (BFI_BOOT_TYPE_OFF))); - writel((swab32(swab32(boot_param))), ((ioc->ioc_regs.smem_page_start) - + (BFI_BOOT_PARAM_OFF))); + writel(boot_env, ((ioc->ioc_regs.smem_page_start) + + (BFI_BOOT_LOADER_OFF))); } static void @@ -1721,7 +1726,7 @@ bfa_ioc_pll_init(struct bfa_ioc *ioc) * as the entry vector. */ static void -bfa_ioc_boot(struct bfa_ioc *ioc, u32 boot_type, u32 boot_param) +bfa_ioc_boot(struct bfa_ioc *ioc, u32 boot_type, u32 boot_env) { void __iomem *rb; @@ -1734,7 +1739,7 @@ bfa_ioc_boot(struct bfa_ioc *ioc, u32 boot_type, u32 boot_param) * Initialize IOC state of all functions on a chip reset. */ rb = ioc->pcidev.pci_bar_kva; - if (boot_param == BFI_BOOT_TYPE_MEMTEST) { + if (boot_type == BFI_BOOT_TYPE_MEMTEST) { writel(BFI_IOC_MEMTEST, (rb + BFA_IOC0_STATE_REG)); writel(BFI_IOC_MEMTEST, (rb + BFA_IOC1_STATE_REG)); } else { @@ -1743,7 +1748,7 @@ bfa_ioc_boot(struct bfa_ioc *ioc, u32 boot_type, u32 boot_param) } bfa_ioc_msgflush(ioc); - bfa_ioc_download_fw(ioc, boot_type, boot_param); + bfa_ioc_download_fw(ioc, boot_type, boot_env); /** * Enable interrupts just before starting LPU diff --git a/drivers/net/bna/bfa_ioc.h b/drivers/net/bna/bfa_ioc.h index e4974bc..bd48abe 100644 --- a/drivers/net/bna/bfa_ioc.h +++ b/drivers/net/bna/bfa_ioc.h @@ -194,6 +194,7 @@ struct bfa_ioc_hwif { bool msix); void (*ioc_notify_fail) (struct bfa_ioc *ioc); void (*ioc_ownership_reset) (struct bfa_ioc *ioc); + bool (*ioc_sync_start) (struct bfa_ioc *ioc); void (*ioc_sync_join) (struct bfa_ioc *ioc); void (*ioc_sync_leave) (struct bfa_ioc *ioc); void (*ioc_sync_ack) (struct bfa_ioc *ioc); diff --git a/drivers/net/bna/bfa_ioc_ct.c b/drivers/net/bna/bfa_ioc_ct.c index 469997c..87aecdf 100644 --- a/drivers/net/bna/bfa_ioc_ct.c +++ b/drivers/net/bna/bfa_ioc_ct.c @@ -41,6 +41,7 @@ static void bfa_ioc_ct_map_port(struct bfa_ioc *ioc); static void bfa_ioc_ct_isr_mode_set(struct bfa_ioc *ioc, bool msix); static void bfa_ioc_ct_notify_fail(struct bfa_ioc *ioc); static void bfa_ioc_ct_ownership_reset(struct bfa_ioc *ioc); +static bool bfa_ioc_ct_sync_start(struct bfa_ioc *ioc); static void bfa_ioc_ct_sync_join(struct bfa_ioc *ioc); static void bfa_ioc_ct_sync_leave(struct bfa_ioc *ioc); static void bfa_ioc_ct_sync_ack(struct bfa_ioc *ioc); @@ -63,6 +64,7 @@ bfa_nw_ioc_set_ct_hwif(struct bfa_ioc *ioc) nw_hwif_ct.ioc_isr_mode_set = bfa_ioc_ct_isr_mode_set; nw_hwif_ct.ioc_notify_fail = bfa_ioc_ct_notify_fail; nw_hwif_ct.ioc_ownership_reset = bfa_ioc_ct_ownership_reset; + nw_hwif_ct.ioc_sync_start = bfa_ioc_ct_sync_start; nw_hwif_ct.ioc_sync_join = bfa_ioc_ct_sync_join; nw_hwif_ct.ioc_sync_leave = bfa_ioc_ct_sync_leave; nw_hwif_ct.ioc_sync_ack = bfa_ioc_ct_sync_ack; @@ -345,6 +347,32 @@ bfa_ioc_ct_ownership_reset(struct bfa_ioc *ioc) /** * Synchronized IOC failure processing routines */ +static bool +bfa_ioc_ct_sync_start(struct bfa_ioc *ioc) +{ + u32 r32 = readl(ioc->ioc_regs.ioc_fail_sync); + u32 sync_reqd = bfa_ioc_ct_get_sync_reqd(r32); + + /* + * Driver load time. If the sync required bit for this PCI fn + * is set, it is due to an unclean exit by the driver for this + * PCI fn in the previous incarnation. Whoever comes here first + * should clean it up, no matter which PCI fn. + */ + + if (sync_reqd & bfa_ioc_ct_sync_pos(ioc)) { + writel(0, ioc->ioc_regs.ioc_fail_sync); + writel(1, ioc->ioc_regs.ioc_usage_reg); + writel(BFI_IOC_UNINIT, ioc->ioc_regs.ioc_fwstate); + writel(BFI_IOC_UNINIT, ioc->ioc_regs.alt_ioc_fwstate); + return true; + } + + return bfa_ioc_ct_sync_complete(ioc); +} +/** + * Synchronized IOC failure processing routines + */ static void bfa_ioc_ct_sync_join(struct bfa_ioc *ioc) { diff --git a/drivers/net/bna/bfi.h b/drivers/net/bna/bfi.h index a973968..6050379 100644 --- a/drivers/net/bna/bfi.h +++ b/drivers/net/bna/bfi.h @@ -184,12 +184,14 @@ enum bfi_mclass { #define BFI_IOC_MSGLEN_MAX 32 /* 32 bytes */ #define BFI_BOOT_TYPE_OFF 8 -#define BFI_BOOT_PARAM_OFF 12 +#define BFI_BOOT_LOADER_OFF 12 -#define BFI_BOOT_TYPE_NORMAL 0 /* param is device id */ +#define BFI_BOOT_TYPE_NORMAL 0 #define BFI_BOOT_TYPE_FLASH 1 #define BFI_BOOT_TYPE_MEMTEST 2 +#define BFI_BOOT_LOADER_OS 0 + #define BFI_BOOT_MEMTEST_RES_ADDR 0x900 #define BFI_BOOT_MEMTEST_RES_SIG 0xA0A1A2A3 -- cgit v0.10.2 From 49b4947aae404c8c5758ffd1007d0d78c48b3643 Mon Sep 17 00:00:00 2001 From: Rasesh Mody Date: Thu, 14 Apr 2011 08:05:19 +0000 Subject: bna: fix memory leak during RX path cleanup The memory leak was caused by unintentional assignment of the Rx path destroy callback function pointer to NULL just after correct initialization. Signed-off-by: Debashis Dutt Signed-off-by: Rasesh Mody Signed-off-by: David S. Miller diff --git a/drivers/net/bna/bnad.c b/drivers/net/bna/bnad.c index 9f356d5..8e6ceab 100644 --- a/drivers/net/bna/bnad.c +++ b/drivers/net/bna/bnad.c @@ -1837,7 +1837,6 @@ bnad_setup_rx(struct bnad *bnad, uint rx_id) /* Initialize the Rx event handlers */ rx_cbfn.rcb_setup_cbfn = bnad_cb_rcb_setup; rx_cbfn.rcb_destroy_cbfn = bnad_cb_rcb_destroy; - rx_cbfn.rcb_destroy_cbfn = NULL; rx_cbfn.ccb_setup_cbfn = bnad_cb_ccb_setup; rx_cbfn.ccb_destroy_cbfn = bnad_cb_ccb_destroy; rx_cbfn.rx_cleanup_cbfn = bnad_cb_rx_cleanup; -- cgit v0.10.2 From 8e461123f28e6b17456225e70eb834b3b30d28bb Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Thu, 14 Apr 2011 11:19:50 -0400 Subject: drm/radeon/kms: fix bad shift in atom iio table parser Noticed by Patrick Lowry. Signed-off-by: Alex Deucher Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/atom.c b/drivers/gpu/drm/radeon/atom.c index d71d375..7bd7456 100644 --- a/drivers/gpu/drm/radeon/atom.c +++ b/drivers/gpu/drm/radeon/atom.c @@ -135,7 +135,7 @@ static uint32_t atom_iio_execute(struct atom_context *ctx, int base, case ATOM_IIO_MOVE_INDEX: temp &= ~((0xFFFFFFFF >> (32 - CU8(base + 1))) << - CU8(base + 2)); + CU8(base + 3)); temp |= ((index >> CU8(base + 2)) & (0xFFFFFFFF >> (32 - CU8(base + 1)))) << CU8(base + @@ -145,7 +145,7 @@ static uint32_t atom_iio_execute(struct atom_context *ctx, int base, case ATOM_IIO_MOVE_DATA: temp &= ~((0xFFFFFFFF >> (32 - CU8(base + 1))) << - CU8(base + 2)); + CU8(base + 3)); temp |= ((data >> CU8(base + 2)) & (0xFFFFFFFF >> (32 - CU8(base + 1)))) << CU8(base + @@ -155,7 +155,7 @@ static uint32_t atom_iio_execute(struct atom_context *ctx, int base, case ATOM_IIO_MOVE_ATTR: temp &= ~((0xFFFFFFFF >> (32 - CU8(base + 1))) << - CU8(base + 2)); + CU8(base + 3)); temp |= ((ctx-> io_attr >> CU8(base + 2)) & (0xFFFFFFFF >> (32 - -- cgit v0.10.2 From a70882aa3137fff9532b51ed5d6a92922e1c4c9c Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Thu, 14 Apr 2011 17:24:07 -0400 Subject: drm/radeon/kms: add connectors even if i2c fails Sometimes the i2c test in i2c_bit_add_bus fails if this happens we fail to register the i2c adapter and eventually fail to add the connector. If i2c fails, add the connector to the user can at least force it on. Note that some distros set i2c-algo-bit.bit_test to 1 by default which sometimes fails preventing the ddc i2c adapter from being added. The i2c adapter works even if the bit test fails, probably due to pre/post_xfer not getting called in the test_bit function. I have another patch to follow up on that. See: https://bugs.freedesktop.org/show_bug.cgi?id=36221 Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c index 2ef6d51..5f45fa1 100644 --- a/drivers/gpu/drm/radeon/radeon_connectors.c +++ b/drivers/gpu/drm/radeon/radeon_connectors.c @@ -1199,7 +1199,7 @@ radeon_add_atom_connector(struct drm_device *dev, if (router->ddc_valid || router->cd_valid) { radeon_connector->router_bus = radeon_i2c_lookup(rdev, &router->i2c_info); if (!radeon_connector->router_bus) - goto failed; + DRM_ERROR("Failed to assign router i2c bus! Check dmesg for i2c errors.\n"); } switch (connector_type) { case DRM_MODE_CONNECTOR_VGA: @@ -1208,7 +1208,7 @@ radeon_add_atom_connector(struct drm_device *dev, if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus); if (!radeon_connector->ddc_bus) - goto failed; + DRM_ERROR("VGA: Failed to assign ddc bus! Check dmesg for i2c errors.\n"); } radeon_connector->dac_load_detect = true; drm_connector_attach_property(&radeon_connector->base, @@ -1226,7 +1226,7 @@ radeon_add_atom_connector(struct drm_device *dev, if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus); if (!radeon_connector->ddc_bus) - goto failed; + DRM_ERROR("DVIA: Failed to assign ddc bus! Check dmesg for i2c errors.\n"); } radeon_connector->dac_load_detect = true; drm_connector_attach_property(&radeon_connector->base, @@ -1249,7 +1249,7 @@ radeon_add_atom_connector(struct drm_device *dev, if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus); if (!radeon_connector->ddc_bus) - goto failed; + DRM_ERROR("DVI: Failed to assign ddc bus! Check dmesg for i2c errors.\n"); } subpixel_order = SubPixelHorizontalRGB; drm_connector_attach_property(&radeon_connector->base, @@ -1290,7 +1290,7 @@ radeon_add_atom_connector(struct drm_device *dev, if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus); if (!radeon_connector->ddc_bus) - goto failed; + DRM_ERROR("HDMI: Failed to assign ddc bus! Check dmesg for i2c errors.\n"); } drm_connector_attach_property(&radeon_connector->base, rdev->mode_info.coherent_mode_property, @@ -1329,10 +1329,10 @@ radeon_add_atom_connector(struct drm_device *dev, else radeon_dig_connector->dp_i2c_bus = radeon_i2c_create_dp(dev, i2c_bus, "DP-auxch"); if (!radeon_dig_connector->dp_i2c_bus) - goto failed; + DRM_ERROR("DP: Failed to assign dp ddc bus! Check dmesg for i2c errors.\n"); radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus); if (!radeon_connector->ddc_bus) - goto failed; + DRM_ERROR("DP: Failed to assign ddc bus! Check dmesg for i2c errors.\n"); } subpixel_order = SubPixelHorizontalRGB; drm_connector_attach_property(&radeon_connector->base, @@ -1381,7 +1381,7 @@ radeon_add_atom_connector(struct drm_device *dev, if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus); if (!radeon_connector->ddc_bus) - goto failed; + DRM_ERROR("LVDS: Failed to assign ddc bus! Check dmesg for i2c errors.\n"); } drm_connector_attach_property(&radeon_connector->base, dev->mode_config.scaling_mode_property, @@ -1457,7 +1457,7 @@ radeon_add_legacy_connector(struct drm_device *dev, if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus); if (!radeon_connector->ddc_bus) - goto failed; + DRM_ERROR("VGA: Failed to assign ddc bus! Check dmesg for i2c errors.\n"); } radeon_connector->dac_load_detect = true; drm_connector_attach_property(&radeon_connector->base, @@ -1475,7 +1475,7 @@ radeon_add_legacy_connector(struct drm_device *dev, if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus); if (!radeon_connector->ddc_bus) - goto failed; + DRM_ERROR("DVIA: Failed to assign ddc bus! Check dmesg for i2c errors.\n"); } radeon_connector->dac_load_detect = true; drm_connector_attach_property(&radeon_connector->base, @@ -1493,7 +1493,7 @@ radeon_add_legacy_connector(struct drm_device *dev, if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus); if (!radeon_connector->ddc_bus) - goto failed; + DRM_ERROR("DVI: Failed to assign ddc bus! Check dmesg for i2c errors.\n"); } if (connector_type == DRM_MODE_CONNECTOR_DVII) { radeon_connector->dac_load_detect = true; @@ -1538,7 +1538,7 @@ radeon_add_legacy_connector(struct drm_device *dev, if (i2c_bus->valid) { radeon_connector->ddc_bus = radeon_i2c_lookup(rdev, i2c_bus); if (!radeon_connector->ddc_bus) - goto failed; + DRM_ERROR("LVDS: Failed to assign ddc bus! Check dmesg for i2c errors.\n"); } drm_connector_attach_property(&radeon_connector->base, dev->mode_config.scaling_mode_property, @@ -1567,9 +1567,4 @@ radeon_add_legacy_connector(struct drm_device *dev, radeon_legacy_backlight_init(radeon_encoder, connector); } } - return; - -failed: - drm_connector_cleanup(connector); - kfree(connector); } diff --git a/drivers/gpu/drm/radeon/radeon_i2c.c b/drivers/gpu/drm/radeon/radeon_i2c.c index ccbabf7..983cbac 100644 --- a/drivers/gpu/drm/radeon/radeon_i2c.c +++ b/drivers/gpu/drm/radeon/radeon_i2c.c @@ -1096,6 +1096,9 @@ void radeon_router_select_ddc_port(struct radeon_connector *radeon_connector) if (!radeon_connector->router.ddc_valid) return; + if (!radeon_connector->router_bus) + return; + radeon_i2c_get_byte(radeon_connector->router_bus, radeon_connector->router.i2c_addr, 0x3, &val); @@ -1121,6 +1124,9 @@ void radeon_router_select_cd_port(struct radeon_connector *radeon_connector) if (!radeon_connector->router.cd_valid) return; + if (!radeon_connector->router_bus) + return; + radeon_i2c_get_byte(radeon_connector->router_bus, radeon_connector->router.i2c_addr, 0x3, &val); -- cgit v0.10.2 From 67954fe95705a8ff80335964bd7e621d13fbc499 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 14 Apr 2011 15:21:52 -0700 Subject: memcg: fix mem_cgroup_rotate_reclaimable_page() commit 3f58a8294333 ("move memcg reclaimable page into tail of inactive list") added inline keyword twice in its prototype. CC arch/x86/kernel/asm-offsets.s In file included from include/linux/swap.h:8, from include/linux/suspend.h:4, from arch/x86/kernel/asm-offsets.c:12: include/linux/memcontrol.h:220: error: duplicate `inline' Signed-off-by: Eric Dumazet Reviewed-by: Minchan Kim Cc: Balbir Singh Cc: KAMEZAWA Hiroyuki Cc: KOSAKI Motohiro Cc: Johannes Weiner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h index 5a5ce70..5e9840f5 100644 --- a/include/linux/memcontrol.h +++ b/include/linux/memcontrol.h @@ -216,7 +216,7 @@ static inline void mem_cgroup_del_lru_list(struct page *page, int lru) return ; } -static inline inline void mem_cgroup_rotate_reclaimable_page(struct page *page) +static inline void mem_cgroup_rotate_reclaimable_page(struct page *page) { return ; } -- cgit v0.10.2 From 584208e6b4103d2cfb08a7889c9fa3540826e0d5 Mon Sep 17 00:00:00 2001 From: Daniel Kiper Date: Thu, 14 Apr 2011 15:21:53 -0700 Subject: mm: optimize pfn calculation in online_page() If CONFIG_FLATMEM is enabled pfn is calculated in online_page() more than once. It is possible to optimize that and use value established at beginning of that function. Signed-off-by: Daniel Kiper Acked-by: Dave Hansen Cc: KAMEZAWA Hiroyuki Cc: Wu Fengguang Cc: Mel Gorman Cc: Christoph Lameter Acked-by: David Rientjes Reviewed-by: Jesper Juhl Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index a2acaf8..9ca1d60 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -375,7 +375,7 @@ void online_page(struct page *page) #endif #ifdef CONFIG_FLATMEM - max_mapnr = max(page_to_pfn(page), max_mapnr); + max_mapnr = max(pfn, max_mapnr); #endif ClearPageReserved(page); -- cgit v0.10.2 From c344180c9e77145a9e7eab0050169c68afae04b2 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Thu, 14 Apr 2011 15:21:55 -0700 Subject: drivers/rtc/rtc-mc13xxx.c: fix unterminated platform_device_id table MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The platform_device_id table is supposed to be zero-terminated. Signed-off-by: Axel Lin Acked-by: Uwe Kleine-König Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/rtc/rtc-mc13xxx.c b/drivers/rtc/rtc-mc13xxx.c index c420064..c5ac037 100644 --- a/drivers/rtc/rtc-mc13xxx.c +++ b/drivers/rtc/rtc-mc13xxx.c @@ -401,6 +401,7 @@ const struct platform_device_id mc13xxx_rtc_idtable[] = { }, { .name = "mc13892-rtc", }, + { } }; static struct platform_driver mc13xxx_rtc_driver = { -- cgit v0.10.2 From c340b1d640001c8c9ecff74f68fd90422ae2448a Mon Sep 17 00:00:00 2001 From: Timo Warns Date: Thu, 14 Apr 2011 15:21:56 -0700 Subject: fs/partitions/ldm.c: fix oops caused by corrupted partition table The kernel automatically evaluates partition tables of storage devices. The code for evaluating LDM partitions (in fs/partitions/ldm.c) contains a bug that causes a kernel oops on certain corrupted LDM partitions. A kernel subsystem seems to crash, because, after the oops, the kernel no longer recognizes newly connected storage devices. The patch validates the value of vblk_size. [akpm@linux-foundation.org: coding-style fixes] Signed-off-by: Timo Warns Cc: Eugene Teo Cc: Harvey Harrison Cc: Richard Russon Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/partitions/ldm.c b/fs/partitions/ldm.c index b10e354..ce4f624 100644 --- a/fs/partitions/ldm.c +++ b/fs/partitions/ldm.c @@ -1299,6 +1299,11 @@ static bool ldm_frag_add (const u8 *data, int size, struct list_head *frags) BUG_ON (!data || !frags); + if (size < 2 * VBLK_SIZE_HEAD) { + ldm_error("Value of size is to small."); + return false; + } + group = get_unaligned_be32(data + 0x08); rec = get_unaligned_be16(data + 0x0C); num = get_unaligned_be16(data + 0x0E); @@ -1306,6 +1311,10 @@ static bool ldm_frag_add (const u8 *data, int size, struct list_head *frags) ldm_error ("A VBLK claims to have %d parts.", num); return false; } + if (rec >= num) { + ldm_error("REC value (%d) exceeds NUM value (%d)", rec, num); + return false; + } list_for_each (item, frags) { f = list_entry (item, struct frag, list); @@ -1334,10 +1343,9 @@ found: f->map |= (1 << rec); - if (num > 0) { - data += VBLK_SIZE_HEAD; - size -= VBLK_SIZE_HEAD; - } + data += VBLK_SIZE_HEAD; + size -= VBLK_SIZE_HEAD; + memcpy (f->data+rec*(size-VBLK_SIZE_HEAD)+VBLK_SIZE_HEAD, data, size); return true; -- cgit v0.10.2 From 9f6ae448bfc6cdf40279f43bb0b4fd159edc4e0a Mon Sep 17 00:00:00 2001 From: Paul Mundt Date: Thu, 14 Apr 2011 15:21:57 -0700 Subject: mm/page_alloc.c: silence build_all_zonelists() section mismatch The memory hotplug case involves calling to build_all_zonelists() which in turns calls in to setup_zone_pageset(). The latter is marked __meminit while build_all_zonelists() itself has no particular annotation. build_all_zonelists() is only handed a non-NULL pointer in the case of memory hotplug through an existing __meminit path, so the setup_zone_pageset() reference is always safe. The options as such are either to flag build_all_zonelists() as __ref (as per __build_all_zonelists()), or to simply discard the __meminit annotation from setup_zone_pageset(). Signed-off-by: Paul Mundt Acked-by: Mel Gorman Cc: KAMEZAWA Hiroyuki Acked-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 2747f5e5a..9f8a97b 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3176,7 +3176,7 @@ static __init_refok int __build_all_zonelists(void *data) * Called with zonelists_mutex held always * unless system_state == SYSTEM_BOOTING. */ -void build_all_zonelists(void *data) +void __ref build_all_zonelists(void *data) { set_zonelist_order(); -- cgit v0.10.2 From d3bc2367180f7ee6afe4ee6e886bfba3ad4eb290 Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Thu, 14 Apr 2011 15:21:58 -0700 Subject: vmstat: update comment regarding stat_threshold Signed-off-by: Christoph Lameter Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/vmstat.c b/mm/vmstat.c index 772b39b..8cb0f0a 100644 --- a/mm/vmstat.c +++ b/mm/vmstat.c @@ -321,9 +321,12 @@ static inline void mod_state(struct zone *zone, /* * The fetching of the stat_threshold is racy. We may apply * a counter threshold to the wrong the cpu if we get - * rescheduled while executing here. However, the following - * will apply the threshold again and therefore bring the - * counter under the threshold. + * rescheduled while executing here. However, the next + * counter update will apply the threshold again and + * therefore bring the counter under the threshold again. + * + * Most of the time the thresholds are the same anyways + * for all cpus in a zone. */ t = this_cpu_read(pcp->stat_threshold); -- cgit v0.10.2 From 592ce316395abc6b4e96c1ac198e5f347bb5d578 Mon Sep 17 00:00:00 2001 From: Antonio Ospite Date: Thu, 14 Apr 2011 15:21:59 -0700 Subject: leds/leds-regulator.c: fix handling of already enabled regulators Make the driver aware of the initial status of the regulator. The leds-regulator driver was ignoring the initial status of the regulator; this resulted in rdev->use_count being incremented to 2 after calling regulator_led_set_value() in the .probe method when a regulator was already enabled at insmod time, which made it impossible to ever disable the regulator. Signed-off-by: Antonio Ospite Cc: Richard Purdie Cc: Antonio Ospite Acked-by: Mark Brown Cc: Liam Girdwood Cc: Daniel Ribeiro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/leds/leds-regulator.c b/drivers/leds/leds-regulator.c index 3790816..8497f56 100644 --- a/drivers/leds/leds-regulator.c +++ b/drivers/leds/leds-regulator.c @@ -178,6 +178,10 @@ static int __devinit regulator_led_probe(struct platform_device *pdev) led->cdev.flags |= LED_CORE_SUSPENDRESUME; led->vcc = vcc; + /* to handle correctly an already enabled regulator */ + if (regulator_is_enabled(led->vcc)) + led->enabled = 1; + mutex_init(&led->mutex); INIT_WORK(&led->work, led_work); -- cgit v0.10.2 From 01eda2e0c0cf035308308a19581e4979285b51ec Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Thu, 14 Apr 2011 15:22:00 -0700 Subject: kstrtox: fix compile warnings in test Fix the following warnings: CC [M] lib/test-kstrtox.o lib/test-kstrtox.c: In function 'test_kstrtou64_ok': lib/test-kstrtox.c:318: warning: this decimal constant is unsigned only in ISO C90 ... Signed-off-by: Alexey Dobriyan Reported-by: Geert Uytterhoeven Acked-by: Geert Uytterhoeven Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/lib/test-kstrtox.c b/lib/test-kstrtox.c index 325c2f9..d55769d 100644 --- a/lib/test-kstrtox.c +++ b/lib/test-kstrtox.c @@ -315,12 +315,12 @@ static void __init test_kstrtou64_ok(void) {"65537", 10, 65537}, {"2147483646", 10, 2147483646}, {"2147483647", 10, 2147483647}, - {"2147483648", 10, 2147483648}, - {"2147483649", 10, 2147483649}, - {"4294967294", 10, 4294967294}, - {"4294967295", 10, 4294967295}, - {"4294967296", 10, 4294967296}, - {"4294967297", 10, 4294967297}, + {"2147483648", 10, 2147483648ULL}, + {"2147483649", 10, 2147483649ULL}, + {"4294967294", 10, 4294967294ULL}, + {"4294967295", 10, 4294967295ULL}, + {"4294967296", 10, 4294967296ULL}, + {"4294967297", 10, 4294967297ULL}, {"9223372036854775806", 10, 9223372036854775806ULL}, {"9223372036854775807", 10, 9223372036854775807ULL}, {"9223372036854775808", 10, 9223372036854775808ULL}, @@ -369,12 +369,12 @@ static void __init test_kstrtos64_ok(void) {"65537", 10, 65537}, {"2147483646", 10, 2147483646}, {"2147483647", 10, 2147483647}, - {"2147483648", 10, 2147483648}, - {"2147483649", 10, 2147483649}, - {"4294967294", 10, 4294967294}, - {"4294967295", 10, 4294967295}, - {"4294967296", 10, 4294967296}, - {"4294967297", 10, 4294967297}, + {"2147483648", 10, 2147483648LL}, + {"2147483649", 10, 2147483649LL}, + {"4294967294", 10, 4294967294LL}, + {"4294967295", 10, 4294967295LL}, + {"4294967296", 10, 4294967296LL}, + {"4294967297", 10, 4294967297LL}, {"9223372036854775806", 10, 9223372036854775806LL}, {"9223372036854775807", 10, 9223372036854775807LL}, }; @@ -418,10 +418,10 @@ static void __init test_kstrtou32_ok(void) {"65537", 10, 65537}, {"2147483646", 10, 2147483646}, {"2147483647", 10, 2147483647}, - {"2147483648", 10, 2147483648}, - {"2147483649", 10, 2147483649}, - {"4294967294", 10, 4294967294}, - {"4294967295", 10, 4294967295}, + {"2147483648", 10, 2147483648U}, + {"2147483649", 10, 2147483649U}, + {"4294967294", 10, 4294967294U}, + {"4294967295", 10, 4294967295U}, }; TEST_OK(kstrtou32, u32, "%u", test_u32_ok); } -- cgit v0.10.2 From 78be959e38567f0e020848179a5d64d2b064391a Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Thu, 14 Apr 2011 15:22:02 -0700 Subject: kstrtox: simpler code in _kstrtoull() Signed-off-by: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/lib/kstrtox.c b/lib/kstrtox.c index 05672e8..a235f3c 100644 --- a/lib/kstrtox.c +++ b/lib/kstrtox.c @@ -49,12 +49,9 @@ static int _kstrtoull(const char *s, unsigned int base, unsigned long long *res) val = *s - '0'; else if ('a' <= _tolower(*s) && _tolower(*s) <= 'f') val = _tolower(*s) - 'a' + 10; - else if (*s == '\n') { - if (*(s + 1) == '\0') - break; - else - return -EINVAL; - } else + else if (*s == '\n' && *(s + 1) == '\0') + break; + else return -EINVAL; if (val >= base) -- cgit v0.10.2 From d69ac131384aa735192b63fd6e0abbe42dec68dc Mon Sep 17 00:00:00 2001 From: Alexander Clouter Date: Thu, 14 Apr 2011 15:22:02 -0700 Subject: MAINTAINERS: add ARM/ts78xx-setup platform maintainer Signed-off-by: Alexander Clouter Cc: Russell King Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/MAINTAINERS b/MAINTAINERS index 649600c..555a7bf 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -877,6 +877,13 @@ F: arch/arm/mach-mv78xx0/ F: arch/arm/mach-orion5x/ F: arch/arm/plat-orion/ +ARM/Orion SoC/Technologic Systems TS-78xx platform support +M: Alexander Clouter +L: linux-arm-kernel@lists.infradead.org (moderated for non-subscribers) +W: http://www.digriz.org.uk/ts78xx/kernel +S: Maintained +F: arch/arm/mach-orion5x/ts78xx-* + ARM/MIOA701 MACHINE SUPPORT M: Robert Jarzmik L: linux-arm-kernel@lists.infradead.org (moderated for non-subscribers) -- cgit v0.10.2 From 61bc02bb252d438a3bb12a236a141d222b35da7b Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Thu, 14 Apr 2011 15:22:04 -0700 Subject: MAINTAINERS: update m68knommu patterns Commit 66d857b08b ("m68k: merge m68k and m68knommu arch directories") moved the files around. Signed-off-by: Joe Perches Acked-by: Greg Ungerer Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/MAINTAINERS b/MAINTAINERS index 555a7bf..e36170b 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -6255,7 +6255,8 @@ M: Greg Ungerer W: http://www.uclinux.org/ L: uclinux-dev@uclinux.org (subscribers-only) S: Maintained -F: arch/m68knommu/ +F: arch/m68k/*/*_no.* +F: arch/m68k/include/asm/*_no.* UCLINUX FOR RENESAS H8/300 (H8300) M: Yoshinori Sato -- cgit v0.10.2 From c897401bac2b099dd2ff673a9afe7193723d253c Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Thu, 14 Apr 2011 15:22:05 -0700 Subject: MAINTAINERS: update various tty patterns Commits 4a6514e6d0 ("tty: move obsolete and broken tty drivers to drivers/staging/tty/") and a6afd9f3e8 ("tty: move a number of tty drivers from drivers/char/ to drivers/tty/") moved files around. Update patterns and orphan some files that were moved to staging. Signed-off-by: Joe Perches Cc: Greg Kroah-Hartman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/MAINTAINERS b/MAINTAINERS index e36170b..e205f6f 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -184,10 +184,9 @@ F: Documentation/filesystems/9p.txt F: fs/9p/ A2232 SERIAL BOARD DRIVER -M: Enver Haase L: linux-m68k@lists.linux-m68k.org -S: Maintained -F: drivers/char/ser_a2232* +S: Orphan +F: drivers/staging/generic_serial/ser_a2232* AACRAID SCSI RAID DRIVER M: Adaptec OEM Raid Solutions @@ -1830,11 +1829,10 @@ S: Maintained F: drivers/platform/x86/compal-laptop.c COMPUTONE INTELLIPORT MULTIPORT CARD -M: "Michael H. Warfield" W: http://www.wittsend.com/computone.html -S: Maintained +S: Orphan F: Documentation/serial/computone.txt -F: drivers/char/ip2/ +F: drivers/staging/tty/ip2/ CONEXANT ACCESSRUNNER USB DRIVER M: Simon Arlott @@ -2017,7 +2015,7 @@ F: drivers/net/wan/cycx* CYCLADES ASYNC MUX DRIVER W: http://www.cyclades.com/ S: Orphan -F: drivers/char/cyclades.c +F: drivers/tty/cyclades.c F: include/linux/cyclades.h CYCLADES PC300 DRIVER @@ -2131,8 +2129,8 @@ L: Eng.Linux@digi.com W: http://www.digi.com S: Orphan F: Documentation/serial/digiepca.txt -F: drivers/char/epca* -F: drivers/char/digi* +F: drivers/staging/tty/epca* +F: drivers/staging/tty/digi* DIOLAN U2C-12 I2C DRIVER M: Guenter Roeck @@ -4199,7 +4197,7 @@ MOXA SMARTIO/INDUSTIO/INTELLIO SERIAL CARD M: Jiri Slaby S: Maintained F: Documentation/serial/moxa-smartio -F: drivers/char/mxser.* +F: drivers/tty/mxser.* MSI LAPTOP SUPPORT M: "Lee, Chun-Yi" @@ -4241,7 +4239,7 @@ F: sound/oss/msnd* MULTITECH MULTIPORT CARD (ISICOM) S: Orphan -F: drivers/char/isicom.c +F: drivers/tty/isicom.c F: include/linux/isicom.h MUSB MULTIPOINT HIGH SPEED DUAL-ROLE CONTROLLER @@ -5280,14 +5278,14 @@ F: drivers/memstick/host/r592.* RISCOM8 DRIVER S: Orphan F: Documentation/serial/riscom8.txt -F: drivers/char/riscom8* +F: drivers/staging/tty/riscom8* ROCKETPORT DRIVER P: Comtrol Corp. W: http://www.comtrol.com S: Maintained F: Documentation/serial/rocket.txt -F: drivers/char/rocket* +F: drivers/tty/rocket* ROSE NETWORK LAYER M: Ralf Baechle @@ -5923,10 +5921,9 @@ F: arch/arm/mach-spear6xx/spear600.c F: arch/arm/mach-spear6xx/spear600_evb.c SPECIALIX IO8+ MULTIPORT SERIAL CARD DRIVER -M: Roger Wolff -S: Supported +S: Orphan F: Documentation/serial/specialix.txt -F: drivers/char/specialix* +F: drivers/staging/tty/specialix* SPI SUBSYSTEM M: David Brownell -- cgit v0.10.2 From 81ab4201fb7d91d6b0cd9ad5b4b16776e4bed145 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Thu, 14 Apr 2011 15:22:06 -0700 Subject: mm: add VM counters for transparent hugepages I found it difficult to make sense of transparent huge pages without having any counters for its actions. Add some counters to vmstat for allocation of transparent hugepages and fallback to smaller pages. Optional patch, but useful for development and understanding the system. Contains improvements from Andrea Arcangeli and Johannes Weiner [akpm@linux-foundation.org: coding-style fixes] [hannes@cmpxchg.org: fix vmstat_text[] entries] Signed-off-by: Andi Kleen Acked-by: Andrea Arcangeli Reviewed-by: KAMEZAWA Hiroyuki Signed-off-by: Johannes Weiner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/include/linux/vmstat.h b/include/linux/vmstat.h index 461c011..2b3831b 100644 --- a/include/linux/vmstat.h +++ b/include/linux/vmstat.h @@ -58,6 +58,13 @@ enum vm_event_item { PGPGIN, PGPGOUT, PSWPIN, PSWPOUT, UNEVICTABLE_PGCLEARED, /* on COW, page truncate */ UNEVICTABLE_PGSTRANDED, /* unable to isolate on unlock */ UNEVICTABLE_MLOCKFREED, +#ifdef CONFIG_TRANSPARENT_HUGEPAGE + THP_FAULT_ALLOC, + THP_FAULT_FALLBACK, + THP_COLLAPSE_ALLOC, + THP_COLLAPSE_ALLOC_FAILED, + THP_SPLIT, +#endif NR_VM_EVENT_ITEMS }; diff --git a/mm/huge_memory.c b/mm/huge_memory.c index 0a619e0..1722683 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -680,8 +680,11 @@ int do_huge_pmd_anonymous_page(struct mm_struct *mm, struct vm_area_struct *vma, return VM_FAULT_OOM; page = alloc_hugepage_vma(transparent_hugepage_defrag(vma), vma, haddr, numa_node_id(), 0); - if (unlikely(!page)) + if (unlikely(!page)) { + count_vm_event(THP_FAULT_FALLBACK); goto out; + } + count_vm_event(THP_FAULT_ALLOC); if (unlikely(mem_cgroup_newpage_charge(page, mm, GFP_KERNEL))) { put_page(page); goto out; @@ -909,11 +912,13 @@ int do_huge_pmd_wp_page(struct mm_struct *mm, struct vm_area_struct *vma, new_page = NULL; if (unlikely(!new_page)) { + count_vm_event(THP_FAULT_FALLBACK); ret = do_huge_pmd_wp_page_fallback(mm, vma, address, pmd, orig_pmd, page, haddr); put_page(page); goto out; } + count_vm_event(THP_FAULT_ALLOC); if (unlikely(mem_cgroup_newpage_charge(new_page, mm, GFP_KERNEL))) { put_page(new_page); @@ -1390,6 +1395,7 @@ int split_huge_page(struct page *page) BUG_ON(!PageSwapBacked(page)); __split_huge_page(page, anon_vma); + count_vm_event(THP_SPLIT); BUG_ON(PageCompound(page)); out_unlock: @@ -1784,9 +1790,11 @@ static void collapse_huge_page(struct mm_struct *mm, node, __GFP_OTHER_NODE); if (unlikely(!new_page)) { up_read(&mm->mmap_sem); + count_vm_event(THP_COLLAPSE_ALLOC_FAILED); *hpage = ERR_PTR(-ENOMEM); return; } + count_vm_event(THP_COLLAPSE_ALLOC); if (unlikely(mem_cgroup_newpage_charge(new_page, mm, GFP_KERNEL))) { up_read(&mm->mmap_sem); put_page(new_page); @@ -2151,8 +2159,11 @@ static void khugepaged_do_scan(struct page **hpage) #ifndef CONFIG_NUMA if (!*hpage) { *hpage = alloc_hugepage(khugepaged_defrag()); - if (unlikely(!*hpage)) + if (unlikely(!*hpage)) { + count_vm_event(THP_COLLAPSE_ALLOC_FAILED); break; + } + count_vm_event(THP_COLLAPSE_ALLOC); } #else if (IS_ERR(*hpage)) @@ -2192,8 +2203,11 @@ static struct page *khugepaged_alloc_hugepage(void) do { hpage = alloc_hugepage(khugepaged_defrag()); - if (!hpage) + if (!hpage) { + count_vm_event(THP_COLLAPSE_ALLOC_FAILED); khugepaged_alloc_sleep(); + } else + count_vm_event(THP_COLLAPSE_ALLOC); } while (unlikely(!hpage) && likely(khugepaged_enabled())); return hpage; @@ -2210,8 +2224,11 @@ static void khugepaged_loop(void) while (likely(khugepaged_enabled())) { #ifndef CONFIG_NUMA hpage = khugepaged_alloc_hugepage(); - if (unlikely(!hpage)) + if (unlikely(!hpage)) { + count_vm_event(THP_COLLAPSE_ALLOC_FAILED); break; + } + count_vm_event(THP_COLLAPSE_ALLOC); #else if (IS_ERR(hpage)) { khugepaged_alloc_sleep(); diff --git a/mm/vmstat.c b/mm/vmstat.c index 8cb0f0a..897ea9e 100644 --- a/mm/vmstat.c +++ b/mm/vmstat.c @@ -948,7 +948,16 @@ static const char * const vmstat_text[] = { "unevictable_pgs_cleared", "unevictable_pgs_stranded", "unevictable_pgs_mlockfreed", + +#ifdef CONFIG_TRANSPARENT_HUGEPAGE + "thp_fault_alloc", + "thp_fault_fallback", + "thp_collapse_alloc", + "thp_collapse_alloc_failed", + "thp_split", #endif + +#endif /* CONFIG_VM_EVENTS_COUNTERS */ }; static void zoneinfo_show_print(struct seq_file *m, pg_data_t *pgdat, -- cgit v0.10.2 From d00ebeac5f24f290636f7a895dafc124b2930a08 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Thu, 14 Apr 2011 15:22:07 -0700 Subject: MAINTAINERS: update STABLE BRANCH info Drop Chris Wright from STABLE maintainers. He hasn't done STABLE release work for quite some time. Signed-off-by: Randy Dunlap Acked-by: Chris Wright Cc: Greg KH Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/MAINTAINERS b/MAINTAINERS index e205f6f..17c22c5 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -5968,7 +5968,6 @@ F: arch/alpha/kernel/srm_env.c STABLE BRANCH M: Greg Kroah-Hartman -M: Chris Wright L: stable@kernel.org S: Maintained -- cgit v0.10.2 From fc5da22ae35d4720be59af8787a8a6d5e4da9517 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Thu, 14 Apr 2011 15:22:07 -0700 Subject: tmpfs: fix off-by-one in max_blocks checks If you fill up a tmpfs, df was showing tmpfs 460800 - - - /tmp because of an off-by-one in the max_blocks checks. Fix it so df shows tmpfs 460800 460800 0 100% /tmp Signed-off-by: Hugh Dickins Cc: Tim Chen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/shmem.c b/mm/shmem.c index 58da7c1..8fa27e4 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -421,7 +421,8 @@ static swp_entry_t *shmem_swp_alloc(struct shmem_inode_info *info, unsigned long * a waste to allocate index if we cannot allocate data. */ if (sbinfo->max_blocks) { - if (percpu_counter_compare(&sbinfo->used_blocks, (sbinfo->max_blocks - 1)) > 0) + if (percpu_counter_compare(&sbinfo->used_blocks, + sbinfo->max_blocks - 1) >= 0) return ERR_PTR(-ENOSPC); percpu_counter_inc(&sbinfo->used_blocks); spin_lock(&inode->i_lock); @@ -1397,7 +1398,8 @@ repeat: shmem_swp_unmap(entry); sbinfo = SHMEM_SB(inode->i_sb); if (sbinfo->max_blocks) { - if ((percpu_counter_compare(&sbinfo->used_blocks, sbinfo->max_blocks) > 0) || + if (percpu_counter_compare(&sbinfo->used_blocks, + sbinfo->max_blocks) >= 0 || shmem_acct_block(info->flags)) { spin_unlock(&info->lock); error = -ENOSPC; -- cgit v0.10.2 From 5de1743e2434fcb24e3d944a20130029b8fe867a Mon Sep 17 00:00:00 2001 From: Wanlong Gao Date: Thu, 14 Apr 2011 15:22:08 -0700 Subject: drivers/misc/sgi-gru/grufile.c: fix the wrong members of gru_chip Fix the wrong members and the wrong function's definition, since the irq_chip had changed. Signed-off-by: Wanlong Gao Cc: Jack Steiner Cc: Thomas Gleixner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/drivers/misc/sgi-gru/grufile.c b/drivers/misc/sgi-gru/grufile.c index 20e4e93..ecafa4b 100644 --- a/drivers/misc/sgi-gru/grufile.c +++ b/drivers/misc/sgi-gru/grufile.c @@ -348,15 +348,15 @@ static unsigned long gru_chiplet_cpu_to_mmr(int chiplet, int cpu, int *corep) static int gru_irq_count[GRU_CHIPLETS_PER_BLADE]; -static void gru_noop(unsigned int irq) +static void gru_noop(struct irq_data *d) { } static struct irq_chip gru_chip[GRU_CHIPLETS_PER_BLADE] = { [0 ... GRU_CHIPLETS_PER_BLADE - 1] { - .mask = gru_noop, - .unmask = gru_noop, - .ack = gru_noop + .irq_mask = gru_noop, + .irq_unmask = gru_noop, + .irq_ack = gru_noop } }; -- cgit v0.10.2 From 4471a675dfc7ca676c165079e91c712b09dc9ce4 Mon Sep 17 00:00:00 2001 From: Jiri Kosina Date: Thu, 14 Apr 2011 15:22:09 -0700 Subject: brk: COMPAT_BRK: fix detection of randomized brk 5520e89 ("brk: fix min_brk lower bound computation for COMPAT_BRK") tried to get the whole logic of brk randomization for legacy (libc5-based) applications finally right. It turns out that the way to detect whether brk has actually been randomized in the end or not introduced by that patch still doesn't work for those binaries, as reported by Geert: : /sbin/init from my old m68k ramdisk exists prematurely. : : Before the patch: : : | brk(0x80005c8e) = 0x80006000 : : After the patch: : : | brk(0x80005c8e) = 0x80005c8e : : Old libc5 considers brk() to have failed if the return value is not : identical to the requested value. I don't like it, but currently see no better option than a bit flag in task_struct to catch the CONFIG_COMPAT_BRK && randomize_va_space == 2 case. Signed-off-by: Jiri Kosina Tested-by: Geert Uytterhoeven Reported-by: Geert Uytterhoeven Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index f34078d..303983f 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -941,9 +941,13 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) current->mm->start_stack = bprm->p; #ifdef arch_randomize_brk - if ((current->flags & PF_RANDOMIZE) && (randomize_va_space > 1)) + if ((current->flags & PF_RANDOMIZE) && (randomize_va_space > 1)) { current->mm->brk = current->mm->start_brk = arch_randomize_brk(current->mm); +#ifdef CONFIG_COMPAT_BRK + current->brk_randomized = 1; +#endif + } #endif if (current->personality & MMAP_PAGE_ZERO) { diff --git a/include/linux/sched.h b/include/linux/sched.h index 4ec2c02..18d63ce 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1254,6 +1254,9 @@ struct task_struct { #endif struct mm_struct *mm, *active_mm; +#ifdef CONFIG_COMPAT_BRK + unsigned brk_randomized:1; +#endif #if defined(SPLIT_RSS_COUNTING) struct task_rss_stat rss_stat; #endif diff --git a/mm/mmap.c b/mm/mmap.c index 8c05e5b..e27e0cf 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -259,7 +259,7 @@ SYSCALL_DEFINE1(brk, unsigned long, brk) * randomize_va_space to 2, which will still cause mm->start_brk * to be arbitrarily shifted */ - if (mm->start_brk > PAGE_ALIGN(mm->end_data)) + if (current->brk_randomized) min_brk = mm->start_brk; else min_brk = mm->end_data; -- cgit v0.10.2 From fe936dfc23fed3475b11067e8d9b70553eafcd9e Mon Sep 17 00:00:00 2001 From: Michael Ellerman Date: Thu, 14 Apr 2011 15:22:10 -0700 Subject: mm: check that we have the right vma in __access_remote_vm() In __access_remote_vm() we need to check that we have found the right vma, not the following vma before we try to access it. Otherwise we might call the vma's access routine with an address which does not fall inside the vma. It was discovered on a current kernel but with an unreleased driver, from memory it was strace leading to a kernel bad access, but it obviously depends on what the access implementation does. Looking at other access implementations I only see: $ git grep -A 5 vm_operations|grep access arch/powerpc/platforms/cell/spufs/file.c- .access = spufs_mem_mmap_access, arch/x86/pci/i386.c- .access = generic_access_phys, drivers/char/mem.c- .access = generic_access_phys fs/sysfs/bin.c- .access = bin_access, The spufs one looks like it might behave badly given the wrong vma, it assumes vma->vm_file->private_data is a spu_context, and looks like it would probably blow up pretty quickly if it wasn't. generic_access_phys() only uses the vma to check vm_flags and get the mm, and then walks page tables using the address. So it should bail on the vm_flags check, or at worst let you access some other VM_IO mapping. And bin_access() just proxies to another access implementation. Signed-off-by: Michael Ellerman Reviewed-by: KOSAKI Motohiro Cc: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/memory.c b/mm/memory.c index b623a24..ce22a25 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -3688,7 +3688,7 @@ static int __access_remote_vm(struct task_struct *tsk, struct mm_struct *mm, */ #ifdef CONFIG_HAVE_IOREMAP_PROT vma = find_vma(mm, addr); - if (!vma) + if (!vma || vma->vm_start > addr) break; if (vma->vm_ops && vma->vm_ops->access) ret = vma->vm_ops->access(vma, addr, buf, -- cgit v0.10.2 From 929bea7c714220fc76ce3f75bef9056477c28e74 Mon Sep 17 00:00:00 2001 From: KOSAKI Motohiro Date: Thu, 14 Apr 2011 15:22:12 -0700 Subject: vmscan: all_unreclaimable() use zone->all_unreclaimable as a name all_unreclaimable check in direct reclaim has been introduced at 2.6.19 by following commit. 2006 Sep 25; commit 408d8544; oom: use unreclaimable info And it went through strange history. firstly, following commit broke the logic unintentionally. 2008 Apr 29; commit a41f24ea; page allocator: smarter retry of costly-order allocations Two years later, I've found obvious meaningless code fragment and restored original intention by following commit. 2010 Jun 04; commit bb21c7ce; vmscan: fix do_try_to_free_pages() return value when priority==0 But, the logic didn't works when 32bit highmem system goes hibernation and Minchan slightly changed the algorithm and fixed it . 2010 Sep 22: commit d1908362: vmscan: check all_unreclaimable in direct reclaim path But, recently, Andrey Vagin found the new corner case. Look, struct zone { .. int all_unreclaimable; .. unsigned long pages_scanned; .. } zone->all_unreclaimable and zone->pages_scanned are neigher atomic variables nor protected by lock. Therefore zones can become a state of zone->page_scanned=0 and zone->all_unreclaimable=1. In this case, current all_unreclaimable() return false even though zone->all_unreclaimabe=1. This resulted in the kernel hanging up when executing a loop of the form 1. fork 2. mmap 3. touch memory 4. read memory 5. munmmap as described in http://www.gossamer-threads.com/lists/linux/kernel/1348725#1348725 Is this ignorable minor issue? No. Unfortunately, x86 has very small dma zone and it become zone->all_unreclamble=1 easily. and if it become all_unreclaimable=1, it never restore all_unreclaimable=0. Why? if all_unreclaimable=1, vmscan only try DEF_PRIORITY reclaim and a-few-lru-pages>>DEF_PRIORITY always makes 0. that mean no page scan at all! Eventually, oom-killer never works on such systems. That said, we can't use zone->pages_scanned for this purpose. This patch restore all_unreclaimable() use zone->all_unreclaimable as old. and in addition, to add oom_killer_disabled check to avoid reintroduce the issue of commit d1908362 ("vmscan: check all_unreclaimable in direct reclaim path"). Reported-by: Andrey Vagin Signed-off-by: KOSAKI Motohiro Cc: Nick Piggin Reviewed-by: Minchan Kim Reviewed-by: KAMEZAWA Hiroyuki Acked-by: David Rientjes Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/vmscan.c b/mm/vmscan.c index c7f5a6d..f6b435c 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -41,6 +41,7 @@ #include #include #include +#include #include #include @@ -1988,17 +1989,12 @@ static bool zone_reclaimable(struct zone *zone) return zone->pages_scanned < zone_reclaimable_pages(zone) * 6; } -/* - * As hibernation is going on, kswapd is freezed so that it can't mark - * the zone into all_unreclaimable. It can't handle OOM during hibernation. - * So let's check zone's unreclaimable in direct reclaim as well as kswapd. - */ +/* All zones in zonelist are unreclaimable? */ static bool all_unreclaimable(struct zonelist *zonelist, struct scan_control *sc) { struct zoneref *z; struct zone *zone; - bool all_unreclaimable = true; for_each_zone_zonelist_nodemask(zone, z, zonelist, gfp_zone(sc->gfp_mask), sc->nodemask) { @@ -2006,13 +2002,11 @@ static bool all_unreclaimable(struct zonelist *zonelist, continue; if (!cpuset_zone_allowed_hardwall(zone, GFP_KERNEL)) continue; - if (zone_reclaimable(zone)) { - all_unreclaimable = false; - break; - } + if (!zone->all_unreclaimable) + return false; } - return all_unreclaimable; + return true; } /* @@ -2108,6 +2102,14 @@ out: if (sc->nr_reclaimed) return sc->nr_reclaimed; + /* + * As hibernation is going on, kswapd is freezed so that it can't mark + * the zone into all_unreclaimable. Thus bypassing all_unreclaimable + * check. + */ + if (oom_killer_disabled) + return 0; + /* top priority shrink_zones still had more to do? don't OOM, then */ if (scanning_global_lru(sc) && !all_unreclaimable(zonelist, sc)) return 1; -- cgit v0.10.2 From 341aea2bc48bf652777fb015cc2b3dfa9a451817 Mon Sep 17 00:00:00 2001 From: KOSAKI Motohiro Date: Thu, 14 Apr 2011 15:22:13 -0700 Subject: oom-kill: remove boost_dying_task_prio() This is an almost-revert of commit 93b43fa ("oom: give the dying task a higher priority"). That commit dramatically improved oom killer logic when a fork-bomb occurs. But I've found that it has nasty corner case. Now cpu cgroup has strange default RT runtime. It's 0! That said, if a process under cpu cgroup promote RT scheduling class, the process never run at all. If an admin inserts a !RT process into a cpu cgroup by setting rtruntime=0, usually it runs perfectly because a !RT task isn't affected by the rtruntime knob. But if it promotes an RT task via an explicit setscheduler() syscall or an OOM, the task can't run at all. In short, the oom killer doesn't work at all if admins are using cpu cgroup and don't touch the rtruntime knob. Eventually, kernel may hang up when oom kill occur. I and the original author Luis agreed to disable this logic. Signed-off-by: KOSAKI Motohiro Acked-by: Luis Claudio R. Goncalves Acked-by: KAMEZAWA Hiroyuki Reviewed-by: Minchan Kim Acked-by: David Rientjes Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/oom_kill.c b/mm/oom_kill.c index 6a819d1..83fb72c1 100644 --- a/mm/oom_kill.c +++ b/mm/oom_kill.c @@ -84,24 +84,6 @@ static bool has_intersects_mems_allowed(struct task_struct *tsk, #endif /* CONFIG_NUMA */ /* - * If this is a system OOM (not a memcg OOM) and the task selected to be - * killed is not already running at high (RT) priorities, speed up the - * recovery by boosting the dying task to the lowest FIFO priority. - * That helps with the recovery and avoids interfering with RT tasks. - */ -static void boost_dying_task_prio(struct task_struct *p, - struct mem_cgroup *mem) -{ - struct sched_param param = { .sched_priority = 1 }; - - if (mem) - return; - - if (!rt_task(p)) - sched_setscheduler_nocheck(p, SCHED_FIFO, ¶m); -} - -/* * The process p may have detached its own ->mm while exiting or through * use_mm(), but one or more of its subthreads may still have a valid * pointer. Return p, or any of its subthreads with a valid ->mm, with @@ -452,13 +434,6 @@ static int oom_kill_task(struct task_struct *p, struct mem_cgroup *mem) set_tsk_thread_flag(p, TIF_MEMDIE); force_sig(SIGKILL, p); - /* - * We give our sacrificial lamb high priority and access to - * all the memory it needs. That way it should be able to - * exit() and clear out its resources quickly... - */ - boost_dying_task_prio(p, mem); - return 0; } #undef K @@ -482,7 +457,6 @@ static int oom_kill_process(struct task_struct *p, gfp_t gfp_mask, int order, */ if (p->flags & PF_EXITING) { set_tsk_thread_flag(p, TIF_MEMDIE); - boost_dying_task_prio(p, mem); return 0; } @@ -556,7 +530,6 @@ void mem_cgroup_out_of_memory(struct mem_cgroup *mem, gfp_t gfp_mask) */ if (fatal_signal_pending(current)) { set_thread_flag(TIF_MEMDIE); - boost_dying_task_prio(current, NULL); return; } @@ -712,7 +685,6 @@ void out_of_memory(struct zonelist *zonelist, gfp_t gfp_mask, */ if (fatal_signal_pending(current)) { set_thread_flag(TIF_MEMDIE); - boost_dying_task_prio(current, NULL); return; } -- cgit v0.10.2 From 13209c2a52afa691ca19e7e6ebd22d4034bdfeed Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Thu, 14 Apr 2011 15:22:14 -0700 Subject: RapidIO: add IDT CPS-1432 switch definitions Signed-off-by: Alexandre Bounine Cc: Matt Porter Cc: Kumar Gala 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 095016a..ac2701b 100644 --- a/drivers/rapidio/switches/idt_gen2.c +++ b/drivers/rapidio/switches/idt_gen2.c @@ -418,3 +418,4 @@ DECLARE_RIO_SWITCH_INIT(RIO_VID_IDT, RIO_DID_IDTCPS1848, idtg2_switch_init); DECLARE_RIO_SWITCH_INIT(RIO_VID_IDT, RIO_DID_IDTCPS1616, idtg2_switch_init); DECLARE_RIO_SWITCH_INIT(RIO_VID_IDT, RIO_DID_IDTVPS1616, idtg2_switch_init); DECLARE_RIO_SWITCH_INIT(RIO_VID_IDT, RIO_DID_IDTSPS1616, idtg2_switch_init); +DECLARE_RIO_SWITCH_INIT(RIO_VID_IDT, RIO_DID_IDTCPS1432, idtg2_switch_init); diff --git a/include/linux/rio_ids.h b/include/linux/rio_ids.h index 7410d33..0cee015 100644 --- a/include/linux/rio_ids.h +++ b/include/linux/rio_ids.h @@ -35,6 +35,7 @@ #define RIO_DID_IDTCPS6Q 0x035f #define RIO_DID_IDTCPS10Q 0x035e #define RIO_DID_IDTCPS1848 0x0374 +#define RIO_DID_IDTCPS1432 0x0375 #define RIO_DID_IDTCPS1616 0x0379 #define RIO_DID_IDTVPS1616 0x0377 #define RIO_DID_IDTSPS1616 0x0378 -- cgit v0.10.2 From 59f9996555542f901f2d01ccab5c0612c8c5c480 Mon Sep 17 00:00:00 2001 From: Alexandre Bounine Date: Thu, 14 Apr 2011 15:22:14 -0700 Subject: RapidIO/mpc85xx: fix possible mport registration problems Fix a possible problem with mport registration left non-cleared after fsl_rio_setup() exits on link error. Abort mport initialization if registration failed. This patch is applicable to 2.6.39-rc1 only. The problem does not exist for earlier versions. Signed-off-by: Alexandre Bounine Cc: Kumar Gala Cc: Matt Porter Cc: Li Yang Cc: Thomas Moll 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 14232d5..4979853 100644 --- a/arch/powerpc/sysdev/fsl_rio.c +++ b/arch/powerpc/sysdev/fsl_rio.c @@ -1457,7 +1457,6 @@ int fsl_rio_setup(struct platform_device *dev) port->ops = ops; port->priv = priv; port->phys_efptr = 0x100; - rio_register_mport(port); priv->regs_win = ioremap(regs.start, regs.end - regs.start + 1); rio_regs_win = priv->regs_win; @@ -1504,6 +1503,9 @@ int fsl_rio_setup(struct platform_device *dev) dev_info(&dev->dev, "RapidIO Common Transport System size: %d\n", port->sys_size ? 65536 : 256); + if (rio_register_mport(port)) + goto err; + if (port->host_deviceid >= 0) out_be32(priv->regs_win + RIO_GCCSR, RIO_PORT_GEN_HOST | RIO_PORT_GEN_MASTER | RIO_PORT_GEN_DISCOVERED); diff --git a/drivers/rapidio/rio.c b/drivers/rapidio/rio.c index c29719c..86c9a09 100644 --- a/drivers/rapidio/rio.c +++ b/drivers/rapidio/rio.c @@ -1171,16 +1171,17 @@ static int rio_hdid_setup(char *str) __setup("riohdid=", rio_hdid_setup); -void rio_register_mport(struct rio_mport *port) +int rio_register_mport(struct rio_mport *port) { if (next_portid >= RIO_MAX_MPORTS) { pr_err("RIO: reached specified max number of mports\n"); - return; + return 1; } port->id = next_portid++; port->host_deviceid = rio_get_hdid(port->id); list_add_tail(&port->node, &rio_mports); + return 0; } EXPORT_SYMBOL_GPL(rio_local_get_device_id); diff --git a/include/linux/rio.h b/include/linux/rio.h index 4e37a7cf..4d50611 100644 --- a/include/linux/rio.h +++ b/include/linux/rio.h @@ -396,7 +396,7 @@ union rio_pw_msg { }; /* Architecture and hardware-specific functions */ -extern void rio_register_mport(struct rio_mport *); +extern int rio_register_mport(struct rio_mport *); extern int rio_open_inb_mbox(struct rio_mport *, void *, int, int); extern void rio_close_inb_mbox(struct rio_mport *, int); extern int rio_open_outb_mbox(struct rio_mport *, void *, int, int); -- cgit v0.10.2 From 6a534c9d265ebabca4c3ae6f8712fc5a27cd3999 Mon Sep 17 00:00:00 2001 From: "Hans J. Koch" Date: Thu, 14 Apr 2011 15:22:16 -0700 Subject: MAINTAINERS: change mail adress of Hans J. Koch My old mail address doesn't exist anymore. This patch changes all occurences in MAINTAINERS to my new address. Signed-off-by: Hans J. Koch Cc: Joe Perches Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/MAINTAINERS b/MAINTAINERS index 17c22c5..ec36003 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1069,7 +1069,7 @@ F: arch/arm/mach-shmobile/ F: drivers/sh/ ARM/TELECHIPS ARM ARCHITECTURE -M: "Hans J. Koch" +M: "Hans J. Koch" L: linux-arm-kernel@lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm/plat-tcc/ @@ -4082,7 +4082,7 @@ F: drivers/video/matrox/matroxfb_* F: include/linux/matroxfb.h MAX6650 HARDWARE MONITOR AND FAN CONTROLLER DRIVER -M: "Hans J. Koch" +M: "Hans J. Koch" L: lm-sensors@lm-sensors.org S: Maintained F: Documentation/hwmon/max6650 @@ -6622,7 +6622,7 @@ F: fs/hostfs/ F: fs/hppfs/ USERSPACE I/O (UIO) -M: "Hans J. Koch" +M: "Hans J. Koch" M: Greg Kroah-Hartman S: Maintained F: Documentation/DocBook/uio-howto.tmpl -- cgit v0.10.2 From ed5afeaf422202485bbebc7e911f13b2a6be2666 Mon Sep 17 00:00:00 2001 From: Jeff Mahoney Date: Thu, 14 Apr 2011 15:22:16 -0700 Subject: fs/fhandle.c: add for ia64 force_o_largefile() on ia64 is defined in and requires . Signed-off-by: Jeff Mahoney Cc: Aneesh Kumar K.V Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/fhandle.c b/fs/fhandle.c index bf93ad2..6b08864 100644 --- a/fs/fhandle.c +++ b/fs/fhandle.c @@ -7,6 +7,7 @@ #include #include #include +#include #include #include "internal.h" -- cgit v0.10.2 From 6d56dad3ae070511e59792a62f27b2394cc936bc Mon Sep 17 00:00:00 2001 From: Richard Weinberger Date: Thu, 14 Apr 2011 15:22:17 -0700 Subject: um: fix call tracer and bug handler Commit 1de1502c ("x86, um: now we can get rid of trivial uml headers") removed accidentally bug.h which broke UML's call tracer and bug handler. Without asm-generic/bug.h UML uses BUG() from arch/x86/ which makes use of ud2. UML cannot use ud2, it raises SIGILL in user mode. As UML has a different stack for handling signals the call trace will be cut off. Signed-off-by: Richard Weinberger Reported-by: Sergei Trofimovich Tested-by: Sergei Trofimovich Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/um/include/asm/bug.h b/arch/um/include/asm/bug.h new file mode 100644 index 0000000..9e33b86 --- /dev/null +++ b/arch/um/include/asm/bug.h @@ -0,0 +1,6 @@ +#ifndef __UM_BUG_H +#define __UM_BUG_H + +#include + +#endif -- cgit v0.10.2 From 084189a88754d40e1bb9bfbc278e70c33e571685 Mon Sep 17 00:00:00 2001 From: Richard Weinberger Date: Thu, 14 Apr 2011 15:22:18 -0700 Subject: um: disable CONFIG_CMPXCHG_LOCAL Commit 8a5ec0ba "Lockless (and preemptless) fastpaths for slub" makes use of this_cpu_cmpxchg_double() which needs this_cpu_cmpxchg16b_emu() on x86_64. Implementing cmpxchg16b emulation for UML would introduce too much complexity. So just disable it. Signed-off-by: Richard Weinberger Reported-by: Sergei Trofimovich Acked-by: Pekka Enberg Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/arch/um/Kconfig.x86 b/arch/um/Kconfig.x86 index 02fb017..a9da516 100644 --- a/arch/um/Kconfig.x86 +++ b/arch/um/Kconfig.x86 @@ -4,6 +4,10 @@ menu "UML-specific options" menu "Host processor type and features" +config CMPXCHG_LOCAL + bool + default n + source "arch/x86/Kconfig.cpu" endmenu -- cgit v0.10.2 From b836aec53e2bce71de1d5415313380688c851477 Mon Sep 17 00:00:00 2001 From: Bob Liu Date: Thu, 14 Apr 2011 15:22:20 -0700 Subject: ramfs: fix memleak on no-mmu arch On no-mmu arch, there is a memleak during shmem test. The cause of this memleak is ramfs_nommu_expand_for_mapping() added page refcount to 2 which makes iput() can't free that pages. The simple test file is like this: int main(void) { int i; key_t k = ftok("/etc", 42); for ( i=0; i<100; ++i) { int id = shmget(k, 10000, 0644|IPC_CREAT); if (id == -1) { printf("shmget error\n"); } if(shmctl(id, IPC_RMID, NULL ) == -1) { printf("shm rm error\n"); return -1; } } printf("run ok...\n"); return 0; } And the result: root:/> free total used free shared buffers Mem: 60320 17912 42408 0 0 -/+ buffers: 17912 42408 root:/> shmem run ok... root:/> free total used free shared buffers Mem: 60320 19096 41224 0 0 -/+ buffers: 19096 41224 root:/> shmem run ok... root:/> free total used free shared buffers Mem: 60320 20296 40024 0 0 -/+ buffers: 20296 40024 ... After this patch the test result is:(no memleak anymore) root:/> free total used free shared buffers Mem: 60320 16668 43652 0 0 -/+ buffers: 16668 43652 root:/> shmem run ok... root:/> free total used free shared buffers Mem: 60320 16668 43652 0 0 -/+ buffers: 16668 43652 Signed-off-by: Bob Liu Acked-by: Hugh Dickins Signed-off-by: David Howells Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/fs/ramfs/file-nommu.c b/fs/ramfs/file-nommu.c index 9eead2c..fbb0b47 100644 --- a/fs/ramfs/file-nommu.c +++ b/fs/ramfs/file-nommu.c @@ -112,6 +112,7 @@ int ramfs_nommu_expand_for_mapping(struct inode *inode, size_t newsize) SetPageDirty(page); unlock_page(page); + put_page(page); } return 0; -- cgit v0.10.2 From e27e6151b154ff6e5e8162efa291bc60196d29ea Mon Sep 17 00:00:00 2001 From: Ben Hutchings Date: Thu, 14 Apr 2011 15:22:21 -0700 Subject: mm/thp: use conventional format for boolean attributes The conventional format for boolean attributes in sysfs is numeric ("0" or "1" followed by new-line). Any boolean attribute can then be read and written using a generic function. Using the strings "yes [no]", "[yes] no" (read), "yes" and "no" (write) will frustrate this. [akpm@linux-foundation.org: use kstrtoul()] [akpm@linux-foundation.org: test_bit() doesn't return 1/0, per Neil] Signed-off-by: Ben Hutchings Cc: Andrea Arcangeli Cc: Mel Gorman Cc: Johannes Weiner Cc: Rik van Riel Cc: Hugh Dickins Tested-by: David Rientjes Cc: NeilBrown Cc: [2.6.38.x] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds diff --git a/mm/huge_memory.c b/mm/huge_memory.c index 1722683..470dcda 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -244,24 +244,28 @@ static ssize_t single_flag_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf, enum transparent_hugepage_flag flag) { - if (test_bit(flag, &transparent_hugepage_flags)) - return sprintf(buf, "[yes] no\n"); - else - return sprintf(buf, "yes [no]\n"); + return sprintf(buf, "%d\n", + !!test_bit(flag, &transparent_hugepage_flags)); } + static ssize_t single_flag_store(struct kobject *kobj, struct kobj_attribute *attr, const char *buf, size_t count, enum transparent_hugepage_flag flag) { - if (!memcmp("yes", buf, - min(sizeof("yes")-1, count))) { + unsigned long value; + int ret; + + ret = kstrtoul(buf, 10, &value); + if (ret < 0) + return ret; + if (value > 1) + return -EINVAL; + + if (value) set_bit(flag, &transparent_hugepage_flags); - } else if (!memcmp("no", buf, - min(sizeof("no")-1, count))) { + else clear_bit(flag, &transparent_hugepage_flags); - } else - return -EINVAL; return count; } -- cgit v0.10.2 From 12dfc843f43efe14d0cfc7a52753d971a0cc759d Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Thu, 14 Apr 2011 19:07:34 -0400 Subject: drm/radeon/kms: adjust evergreen display watermark setup This patch fixes two issues: - A disabled crtc does not use any lb, so return 0 for lb size. This makes the display priority calculation more exact. - Only use 1/2 and whole lb partitions. Using smaller partitions can cause underflow to one of the displays if you have multiple large displays on the same lb. Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=34534 Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index 3453910..43fd016 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -353,7 +353,7 @@ static u32 evergreen_line_buffer_adjust(struct radeon_device *rdev, struct drm_display_mode *mode, struct drm_display_mode *other_mode) { - u32 tmp = 0; + u32 tmp; /* * Line Buffer Setup * There are 3 line buffers, each one shared by 2 display controllers. @@ -363,64 +363,63 @@ static u32 evergreen_line_buffer_adjust(struct radeon_device *rdev, * first display controller * 0 - first half of lb (3840 * 2) * 1 - first 3/4 of lb (5760 * 2) - * 2 - whole lb (7680 * 2) + * 2 - whole lb (7680 * 2), other crtc must be disabled * 3 - first 1/4 of lb (1920 * 2) * second display controller * 4 - second half of lb (3840 * 2) * 5 - second 3/4 of lb (5760 * 2) - * 6 - whole lb (7680 * 2) + * 6 - whole lb (7680 * 2), other crtc must be disabled * 7 - last 1/4 of lb (1920 * 2) */ - if (mode && other_mode) { - if (mode->hdisplay > other_mode->hdisplay) { - if (mode->hdisplay > 2560) - tmp = 1; /* 3/4 */ - else - tmp = 0; /* 1/2 */ - } else if (other_mode->hdisplay > mode->hdisplay) { - if (other_mode->hdisplay > 2560) - tmp = 3; /* 1/4 */ - else - tmp = 0; /* 1/2 */ - } else + /* this can get tricky if we have two large displays on a paired group + * of crtcs. Ideally for multiple large displays we'd assign them to + * non-linked crtcs for maximum line buffer allocation. + */ + if (radeon_crtc->base.enabled && mode) { + if (other_mode) tmp = 0; /* 1/2 */ - } else if (mode) - tmp = 2; /* whole */ - else if (other_mode) - tmp = 3; /* 1/4 */ + else + tmp = 2; /* whole */ + } else + tmp = 0; /* second controller of the pair uses second half of the lb */ if (radeon_crtc->crtc_id % 2) tmp += 4; WREG32(DC_LB_MEMORY_SPLIT + radeon_crtc->crtc_offset, tmp); - switch (tmp) { - case 0: - case 4: - default: - if (ASIC_IS_DCE5(rdev)) - return 4096 * 2; - else - return 3840 * 2; - case 1: - case 5: - if (ASIC_IS_DCE5(rdev)) - return 6144 * 2; - else - return 5760 * 2; - case 2: - case 6: - if (ASIC_IS_DCE5(rdev)) - return 8192 * 2; - else - return 7680 * 2; - case 3: - case 7: - if (ASIC_IS_DCE5(rdev)) - return 2048 * 2; - else - return 1920 * 2; + if (radeon_crtc->base.enabled && mode) { + switch (tmp) { + case 0: + case 4: + default: + if (ASIC_IS_DCE5(rdev)) + return 4096 * 2; + else + return 3840 * 2; + case 1: + case 5: + if (ASIC_IS_DCE5(rdev)) + return 6144 * 2; + else + return 5760 * 2; + case 2: + case 6: + if (ASIC_IS_DCE5(rdev)) + return 8192 * 2; + else + return 7680 * 2; + case 3: + case 7: + if (ASIC_IS_DCE5(rdev)) + return 2048 * 2; + else + return 1920 * 2; + } } + + /* controller not enabled, so no lb used */ + return 0; } static u32 evergreen_get_number_of_dram_channels(struct radeon_device *rdev) -- cgit v0.10.2 From c65353daf137dd41f3ede3baf62d561fca076228 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 14 Apr 2011 05:55:37 +0000 Subject: ip: ip_options_compile() resilient to NULL skb route Scot Doyle demonstrated ip_options_compile() could be called with an skb without an attached route, using a setup involving a bridge, netfilter, and forged IP packets. Let's make ip_options_compile() and ip_options_rcv_srr() a bit more robust, instead of changing bridge/netfilter code. With help from Hiroaki SHIMODA. Reported-by: Scot Doyle Tested-by: Scot Doyle Signed-off-by: Eric Dumazet Cc: Stephen Hemminger Acked-by: Hiroaki SHIMODA Signed-off-by: David S. Miller diff --git a/net/ipv4/ip_options.c b/net/ipv4/ip_options.c index 28a736f..2391b24 100644 --- a/net/ipv4/ip_options.c +++ b/net/ipv4/ip_options.c @@ -329,7 +329,7 @@ int ip_options_compile(struct net *net, pp_ptr = optptr + 2; goto error; } - if (skb) { + if (rt) { memcpy(&optptr[optptr[2]-1], &rt->rt_spec_dst, 4); opt->is_changed = 1; } @@ -371,7 +371,7 @@ int ip_options_compile(struct net *net, goto error; } opt->ts = optptr - iph; - if (skb) { + if (rt) { memcpy(&optptr[optptr[2]-1], &rt->rt_spec_dst, 4); timeptr = (__be32*)&optptr[optptr[2]+3]; } @@ -603,7 +603,7 @@ int ip_options_rcv_srr(struct sk_buff *skb) unsigned long orefdst; int err; - if (!opt->srr) + if (!opt->srr || !rt) return 0; if (skb->pkt_type != PACKET_HOST) -- cgit v0.10.2 From 0d58a2824d777923b2438107053c6e073c9c5ec1 Mon Sep 17 00:00:00 2001 From: Russell King Date: Fri, 15 Apr 2011 13:26:40 +0100 Subject: ARM: Add new syscalls Add syscalls for name_to_handle_at, open_by_handle_at, clock_adjtime and syncfs. Signed-off-by: Russell King diff --git a/arch/arm/include/asm/unistd.h b/arch/arm/include/asm/unistd.h index c891eb7..87dbe3e 100644 --- a/arch/arm/include/asm/unistd.h +++ b/arch/arm/include/asm/unistd.h @@ -396,6 +396,10 @@ #define __NR_fanotify_init (__NR_SYSCALL_BASE+367) #define __NR_fanotify_mark (__NR_SYSCALL_BASE+368) #define __NR_prlimit64 (__NR_SYSCALL_BASE+369) +#define __NR_name_to_handle_at (__NR_SYSCALL_BASE+370) +#define __NR_open_by_handle_at (__NR_SYSCALL_BASE+371) +#define __NR_clock_adjtime (__NR_SYSCALL_BASE+372) +#define __NR_syncfs (__NR_SYSCALL_BASE+373) /* * The following SWIs are ARM private. diff --git a/arch/arm/kernel/calls.S b/arch/arm/kernel/calls.S index 5c26ecc..7fbf28c 100644 --- a/arch/arm/kernel/calls.S +++ b/arch/arm/kernel/calls.S @@ -379,6 +379,10 @@ CALL(sys_fanotify_init) CALL(sys_fanotify_mark) CALL(sys_prlimit64) +/* 370 */ CALL(sys_name_to_handle_at) + CALL(sys_open_by_handle_at) + CALL(sys_clock_adjtime) + CALL(sys_syncfs) #ifndef syscalls_counted .equ syscalls_padding, ((NR_syscalls + 3) & ~3) - NR_syscalls #define syscalls_counted -- cgit v0.10.2 From 88b996cd0652280cc9b9fc70008fda15f14175e1 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 15 Apr 2011 15:20:10 +0200 Subject: block: cleanup the block plug helper functions It's a bit of a mess currently. task->plug is being cleared and reset in __blk_finish_plug(), and blk_finish_plug() is testing for a NULL plug which cannot happen even from schedule() anymore since it uses blk_needs_flush_plug() to determine whether to call into this function at all. So get rid of some of the cruft. Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 36b1a75..b598fa7 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2671,7 +2671,7 @@ static void queue_unplugged(struct request_queue *q, unsigned int depth) q->unplugged_fn(q); } -static void flush_plug_list(struct blk_plug *plug) +void blk_flush_plug_list(struct blk_plug *plug) { struct request_queue *q; unsigned long flags; @@ -2733,28 +2733,16 @@ static void flush_plug_list(struct blk_plug *plug) local_irq_restore(flags); } - -static void __blk_finish_plug(struct task_struct *tsk, struct blk_plug *plug) -{ - flush_plug_list(plug); - - if (plug == tsk->plug) - tsk->plug = NULL; -} +EXPORT_SYMBOL(blk_flush_plug_list); void blk_finish_plug(struct blk_plug *plug) { - if (plug) - __blk_finish_plug(current, plug); -} -EXPORT_SYMBOL(blk_finish_plug); + blk_flush_plug_list(plug); -void __blk_flush_plug(struct task_struct *tsk, struct blk_plug *plug) -{ - __blk_finish_plug(tsk, plug); - tsk->plug = plug; + if (plug == current->plug) + current->plug = NULL; } -EXPORT_SYMBOL(__blk_flush_plug); +EXPORT_SYMBOL(blk_finish_plug); int __init blk_dev_init(void) { diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index c07ffaf..ffe48ff 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -865,14 +865,14 @@ struct blk_plug { extern void blk_start_plug(struct blk_plug *); extern void blk_finish_plug(struct blk_plug *); -extern void __blk_flush_plug(struct task_struct *, struct blk_plug *); +extern void blk_flush_plug_list(struct blk_plug *); static inline void blk_flush_plug(struct task_struct *tsk) { struct blk_plug *plug = tsk->plug; - if (unlikely(plug)) - __blk_flush_plug(tsk, plug); + if (plug) + blk_flush_plug_list(plug); } static inline bool blk_needs_flush_plug(struct task_struct *tsk) -- cgit v0.10.2 From f6603783f9f099bf7a83b3f6c689bbbf74f0e96e Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Fri, 15 Apr 2011 15:49:07 +0200 Subject: block: only force kblockd unplugging from the schedule() path For the explicit unplugging, we'd prefer to kick things off immediately and not pay the penalty of the latency to switch to kblockd. So let blk_finish_plug() do the run inline, while the implicit-on-schedule-out unplug will punt to kblockd. Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index b598fa7..3c81210 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2662,16 +2662,17 @@ static int plug_rq_cmp(void *priv, struct list_head *a, struct list_head *b) return !(rqa->q <= rqb->q); } -static void queue_unplugged(struct request_queue *q, unsigned int depth) +static void queue_unplugged(struct request_queue *q, unsigned int depth, + bool force_kblockd) { trace_block_unplug_io(q, depth); - __blk_run_queue(q, true); + __blk_run_queue(q, force_kblockd); if (q->unplugged_fn) q->unplugged_fn(q); } -void blk_flush_plug_list(struct blk_plug *plug) +void blk_flush_plug_list(struct blk_plug *plug, bool force_kblockd) { struct request_queue *q; unsigned long flags; @@ -2706,7 +2707,7 @@ void blk_flush_plug_list(struct blk_plug *plug) BUG_ON(!rq->q); if (rq->q != q) { if (q) { - queue_unplugged(q, depth); + queue_unplugged(q, depth, force_kblockd); spin_unlock(q->queue_lock); } q = rq->q; @@ -2727,7 +2728,7 @@ void blk_flush_plug_list(struct blk_plug *plug) } if (q) { - queue_unplugged(q, depth); + queue_unplugged(q, depth, force_kblockd); spin_unlock(q->queue_lock); } @@ -2737,7 +2738,7 @@ EXPORT_SYMBOL(blk_flush_plug_list); void blk_finish_plug(struct blk_plug *plug) { - blk_flush_plug_list(plug); + blk_flush_plug_list(plug, false); if (plug == current->plug) current->plug = NULL; diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index ffe48ff..1c76506 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -865,14 +865,14 @@ struct blk_plug { extern void blk_start_plug(struct blk_plug *); extern void blk_finish_plug(struct blk_plug *); -extern void blk_flush_plug_list(struct blk_plug *); +extern void blk_flush_plug_list(struct blk_plug *, bool); static inline void blk_flush_plug(struct task_struct *tsk) { struct blk_plug *plug = tsk->plug; if (plug) - blk_flush_plug_list(plug); + blk_flush_plug_list(plug, true); } static inline bool blk_needs_flush_plug(struct task_struct *tsk) -- cgit v0.10.2 From 7ebfa57f6d307b66bb88600145afccde31016ab5 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 15 Apr 2011 07:34:26 -0700 Subject: vfs: fix incorrect dentry_update_name_case() BUG_ON() test The case we should be verifying when updating the dentry name is that the _parent_ inode (the directory) semaphore is held, not the semaphore for the dentry itself. It's the directory locking that rename and readdir() etc all care about. The comment just above even says so - but then the BUG_ON() still checked the dentry inode itself. Very few people noticed, because this helper function really isn't used for very much, so you had to be using ncpfs to ever hit it. I think I should just remove the BUG_ON (the function really has just one user), but let's run with it fixed for a while before getting rid of it entirely. Reported-and-tested-by: Bongani Hlope Reported-and-tested-by: Bernd Feige Cc: Petr Vandrovec , Cc: Arnd Bergmann Cc: Christoph Hellwig Cc: Nick Piggin Signed-off-by: Linus Torvalds diff --git a/fs/dcache.c b/fs/dcache.c index ad25c4c..129a357 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -2131,7 +2131,7 @@ EXPORT_SYMBOL(d_rehash); */ void dentry_update_name_case(struct dentry *dentry, struct qstr *name) { - BUG_ON(!mutex_is_locked(&dentry->d_inode->i_mutex)); + BUG_ON(!mutex_is_locked(&dentry->d_parent->d_inode->i_mutex)); BUG_ON(dentry->d_name.len != name->len); /* d_lookup gives this */ spin_lock(&dentry->d_lock); -- cgit v0.10.2 From 0cd9c6494ee5c19aef085152bc37f3a4e774a9e1 Mon Sep 17 00:00:00 2001 From: Darren Hart Date: Thu, 14 Apr 2011 15:41:57 -0700 Subject: futex: Set FLAGS_HAS_TIMEOUT during futex_wait restart setup The FLAGS_HAS_TIMEOUT flag was not getting set, causing the restart_block to restart futex_wait() without a timeout after a signal. Commit b41277dc7a18ee332d in 2.6.38 introduced the regression by accidentally removing the the FLAGS_HAS_TIMEOUT assignment from futex_wait() during the setup of the restart block. Restore the originaly behavior. Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=32922 Reported-by: Tim Smith Reported-by: Torsten Hilbrich Signed-off-by: Darren Hart Signed-off-by: Eric Dumazet Cc: Peter Zijlstra Cc: John Kacur Cc: stable@kernel.org Link: http://lkml.kernel.org/r/%3Cdaac0eb3af607f72b9a4d3126b2ba8fb5ed3b883.1302820917.git.dvhart%40linux.intel.com%3E Signed-off-by: Thomas Gleixner diff --git a/kernel/futex.c b/kernel/futex.c index dfb924f..fe28dc2 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -1886,7 +1886,7 @@ retry: restart->futex.val = val; restart->futex.time = abs_time->tv64; restart->futex.bitset = bitset; - restart->futex.flags = flags; + restart->futex.flags = flags | FLAGS_HAS_TIMEOUT; ret = -ERESTART_RESTARTBLOCK; -- cgit v0.10.2 From 7d6b46707f2491a94f4bd3b4329d2d7f809e9368 Mon Sep 17 00:00:00 2001 From: KOSAKI Motohiro Date: Fri, 15 Apr 2011 20:39:01 +0900 Subject: x86, NUMA: Fix fakenuma boot failure Currently, numa=fake boot parameter is broken. If it's used, kernel may panic due to devide by zero error depending on CPU configuration Call Trace: [] find_busiest_group+0x38c/0xd30 [] ? local_clock+0x6f/0x80 [] load_balance+0xa3/0x600 [] idle_balance+0xf3/0x180 [] schedule+0x722/0x7d0 [] ? wait_for_common+0x128/0x190 [] schedule_timeout+0x265/0x320 [] ? lock_release_holdtime+0x35/0x1a0 [] ? wait_for_common+0x128/0x190 [] ? __lock_release+0x9c/0x1d0 [] ? _raw_spin_unlock_irq+0x30/0x40 [] ? _raw_spin_unlock_irq+0x30/0x40 [] wait_for_common+0x130/0x190 [] ? try_to_wake_up+0x510/0x510 [] wait_for_completion+0x1d/0x20 [] kthread_create_on_node+0xac/0x150 [] ? process_scheduled_works+0x40/0x40 [] ? wait_for_common+0x4f/0x190 [] __alloc_workqueue_key+0x1a3/0x590 [] cpuset_init_smp+0x6b/0x7b [] kernel_init+0xc3/0x182 [] kernel_thread_helper+0x4/0x10 [] ? retint_restore_args+0x13/0x13 [] ? start_kernel+0x400/0x400 [] ? gs_change+0x13/0x13 The divede by zero is caused by the following line, group->cpu_power==0: kernel/sched_fair.c::update_sg_lb_stats() /* Adjust by relative CPU power of the group */ sgs->avg_load = (sgs->group_load * SCHED_LOAD_SCALE) / group->cpu_power; This regression was caused by commit e23bba6044 ("x86-64, NUMA: Unify emulated distance mapping") because it changes cpu -> node mapping in the process of dropping fake_physnodes(). old) all cpus are assinged node 0 now) cpus are assigned round robin (the logic is implemented by numa_init_array()) Note: The change in behavior only happens if the system doesn't have neither ACPI SRAT table nor AMD northbridge NUMA information. Round robin assignment doesn't work because init_numa_sched_groups_power() assumes all logical cpus in the same physical cpu share the same node (then it only accounts for group_first_cpu()), and the simple round robin breaks the above assumption. Thus, this patch implements a reassignment of node-ids if buggy firmware or numa emulation makes wrong cpu node map. Tt enforce all logical cpus in the same physical cpu share the same node. Signed-off-by: KOSAKI Motohiro Acked-by: Tejun Heo Cc: Yinghai Lu Cc: Brian Gerst Cc: Cyrill Gorcunov Cc: Shaohui Zheng Cc: David Rientjes Cc: H. Peter Anvin Link: http://lkml.kernel.org/r/20110415203928.1303.A69D9226@jp.fujitsu.com Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index c2871d3..8ed8908 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -312,6 +312,26 @@ void __cpuinit smp_store_cpu_info(int id) identify_secondary_cpu(c); } +static void __cpuinit check_cpu_siblings_on_same_node(int cpu1, int cpu2) +{ + int node1 = early_cpu_to_node(cpu1); + int node2 = early_cpu_to_node(cpu2); + + /* + * Our CPU scheduler assumes all logical cpus in the same physical cpu + * share the same node. But, buggy ACPI or NUMA emulation might assign + * them to different node. Fix it. + */ + if (node1 != node2) { + pr_warning("CPU %d in node %d and CPU %d in node %d are in the same physical CPU. forcing same node %d\n", + cpu1, node1, cpu2, node2, node2); + + numa_remove_cpu(cpu1); + numa_set_node(cpu1, node2); + numa_add_cpu(cpu1); + } +} + static void __cpuinit link_thread_siblings(int cpu1, int cpu2) { cpumask_set_cpu(cpu1, cpu_sibling_mask(cpu2)); @@ -320,6 +340,7 @@ static void __cpuinit link_thread_siblings(int cpu1, int cpu2) cpumask_set_cpu(cpu2, cpu_core_mask(cpu1)); cpumask_set_cpu(cpu1, cpu_llc_shared_mask(cpu2)); cpumask_set_cpu(cpu2, cpu_llc_shared_mask(cpu1)); + check_cpu_siblings_on_same_node(cpu1, cpu2); } @@ -361,10 +382,12 @@ void __cpuinit set_cpu_sibling_map(int cpu) per_cpu(cpu_llc_id, cpu) == per_cpu(cpu_llc_id, i)) { cpumask_set_cpu(i, cpu_llc_shared_mask(cpu)); cpumask_set_cpu(cpu, cpu_llc_shared_mask(i)); + check_cpu_siblings_on_same_node(cpu, i); } if (c->phys_proc_id == cpu_data(i).phys_proc_id) { cpumask_set_cpu(i, cpu_core_mask(cpu)); cpumask_set_cpu(cpu, cpu_core_mask(i)); + check_cpu_siblings_on_same_node(cpu, i); /* * Does this new cpu bringup a new core? */ -- cgit v0.10.2 From 0e4f8f888845f9dca540ad175884244e5db5eea2 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Fri, 15 Apr 2011 16:05:44 -0400 Subject: Btrfs: don't force chunk allocation in find_free_extent find_free_extent likes to allocate in contiguous clusters, which makes writeback faster, especially on SSD storage. As the FS fragments, these clusters become harder to find and we have to decide between allocating a new chunk to make more clusters or giving up on the cluster to allocate from the free space we have. Right now it creates too many chunks, and you can end up with a whole FS that is mostly empty metadata chunks. This commit changes the allocation code to be more strict and only allocate new chunks when we've made good use of the chunks we already have. Signed-off-by: Chris Mason diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index f619c3c..2647948 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -33,6 +33,25 @@ #include "locking.h" #include "free-space-cache.h" +/* control flags for do_chunk_alloc's force field + * CHUNK_ALLOC_NO_FORCE means to only allocate a chunk + * if we really need one. + * + * CHUNK_ALLOC_FORCE means it must try to allocate one + * + * CHUNK_ALLOC_LIMITED means to only try and allocate one + * if we have very few chunks already allocated. This is + * used as part of the clustering code to help make sure + * we have a good pool of storage to cluster in, without + * filling the FS with empty chunks + * + */ +enum { + CHUNK_ALLOC_NO_FORCE = 0, + CHUNK_ALLOC_FORCE = 1, + CHUNK_ALLOC_LIMITED = 2, +}; + static int update_block_group(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 bytenr, u64 num_bytes, int alloc); @@ -3019,7 +3038,7 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags, found->bytes_readonly = 0; found->bytes_may_use = 0; found->full = 0; - found->force_alloc = 0; + found->force_alloc = CHUNK_ALLOC_NO_FORCE; *space_info = found; list_add_rcu(&found->list, &info->space_info); atomic_set(&found->caching_threads, 0); @@ -3150,7 +3169,7 @@ again: if (!data_sinfo->full && alloc_chunk) { u64 alloc_target; - data_sinfo->force_alloc = 1; + data_sinfo->force_alloc = CHUNK_ALLOC_FORCE; spin_unlock(&data_sinfo->lock); alloc: alloc_target = btrfs_get_alloc_profile(root, 1); @@ -3160,7 +3179,8 @@ alloc: ret = do_chunk_alloc(trans, root->fs_info->extent_root, bytes + 2 * 1024 * 1024, - alloc_target, 0); + alloc_target, + CHUNK_ALLOC_NO_FORCE); btrfs_end_transaction(trans, root); if (ret < 0) { if (ret != -ENOSPC) @@ -3239,31 +3259,56 @@ static void force_metadata_allocation(struct btrfs_fs_info *info) rcu_read_lock(); list_for_each_entry_rcu(found, head, list) { if (found->flags & BTRFS_BLOCK_GROUP_METADATA) - found->force_alloc = 1; + found->force_alloc = CHUNK_ALLOC_FORCE; } rcu_read_unlock(); } static int should_alloc_chunk(struct btrfs_root *root, - struct btrfs_space_info *sinfo, u64 alloc_bytes) + struct btrfs_space_info *sinfo, u64 alloc_bytes, + int force) { u64 num_bytes = sinfo->total_bytes - sinfo->bytes_readonly; + u64 num_allocated = sinfo->bytes_used + sinfo->bytes_reserved; u64 thresh; - if (sinfo->bytes_used + sinfo->bytes_reserved + - alloc_bytes + 256 * 1024 * 1024 < num_bytes) + if (force == CHUNK_ALLOC_FORCE) + return 1; + + /* + * in limited mode, we want to have some free space up to + * about 1% of the FS size. + */ + if (force == CHUNK_ALLOC_LIMITED) { + thresh = btrfs_super_total_bytes(&root->fs_info->super_copy); + thresh = max_t(u64, 64 * 1024 * 1024, + div_factor_fine(thresh, 1)); + + if (num_bytes - num_allocated < thresh) + return 1; + } + + /* + * we have two similar checks here, one based on percentage + * and once based on a hard number of 256MB. The idea + * is that if we have a good amount of free + * room, don't allocate a chunk. A good mount is + * less than 80% utilized of the chunks we have allocated, + * or more than 256MB free + */ + if (num_allocated + alloc_bytes + 256 * 1024 * 1024 < num_bytes) return 0; - if (sinfo->bytes_used + sinfo->bytes_reserved + - alloc_bytes < div_factor(num_bytes, 8)) + if (num_allocated + alloc_bytes < div_factor(num_bytes, 8)) return 0; thresh = btrfs_super_total_bytes(&root->fs_info->super_copy); + + /* 256MB or 5% of the FS */ thresh = max_t(u64, 256 * 1024 * 1024, div_factor_fine(thresh, 5)); if (num_bytes > thresh && sinfo->bytes_used < div_factor(num_bytes, 3)) return 0; - return 1; } @@ -3289,17 +3334,17 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans, spin_lock(&space_info->lock); if (space_info->force_alloc) - force = 1; + force = space_info->force_alloc; if (space_info->full) { spin_unlock(&space_info->lock); goto out; } - if (!force && !should_alloc_chunk(extent_root, space_info, - alloc_bytes)) { + if (!should_alloc_chunk(extent_root, space_info, alloc_bytes, force)) { spin_unlock(&space_info->lock); goto out; } + spin_unlock(&space_info->lock); /* @@ -3327,7 +3372,7 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans, space_info->full = 1; else ret = 1; - space_info->force_alloc = 0; + space_info->force_alloc = CHUNK_ALLOC_NO_FORCE; spin_unlock(&space_info->lock); out: mutex_unlock(&extent_root->fs_info->chunk_mutex); @@ -5303,11 +5348,13 @@ loop: if (allowed_chunk_alloc) { ret = do_chunk_alloc(trans, root, num_bytes + - 2 * 1024 * 1024, data, 1); + 2 * 1024 * 1024, data, + CHUNK_ALLOC_LIMITED); allowed_chunk_alloc = 0; done_chunk_alloc = 1; - } else if (!done_chunk_alloc) { - space_info->force_alloc = 1; + } else if (!done_chunk_alloc && + space_info->force_alloc == CHUNK_ALLOC_NO_FORCE) { + space_info->force_alloc = CHUNK_ALLOC_LIMITED; } if (loop < LOOP_NO_EMPTY_SIZE) { @@ -5393,7 +5440,8 @@ again: */ if (empty_size || root->ref_cows) ret = do_chunk_alloc(trans, root->fs_info->extent_root, - num_bytes + 2 * 1024 * 1024, data, 0); + num_bytes + 2 * 1024 * 1024, data, + CHUNK_ALLOC_NO_FORCE); WARN_ON(num_bytes < root->sectorsize); ret = find_free_extent(trans, root, num_bytes, empty_size, @@ -5405,7 +5453,7 @@ again: num_bytes = num_bytes & ~(root->sectorsize - 1); num_bytes = max(num_bytes, min_alloc_size); do_chunk_alloc(trans, root->fs_info->extent_root, - num_bytes, data, 1); + num_bytes, data, CHUNK_ALLOC_FORCE); goto again; } if (ret == -ENOSPC && btrfs_test_opt(root, ENOSPC_DEBUG)) { @@ -8109,13 +8157,15 @@ int btrfs_set_block_group_ro(struct btrfs_root *root, alloc_flags = update_block_group_flags(root, cache->flags); if (alloc_flags != cache->flags) - do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, 1); + do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, + CHUNK_ALLOC_FORCE); ret = set_block_group_ro(cache); if (!ret) goto out; alloc_flags = get_alloc_profile(root, cache->space_info->flags); - ret = do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, 1); + ret = do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, + CHUNK_ALLOC_FORCE); if (ret < 0) goto out; ret = set_block_group_ro(cache); @@ -8128,7 +8178,8 @@ int btrfs_force_chunk_alloc(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 type) { u64 alloc_flags = get_alloc_profile(root, type); - return do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, 1); + return do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, + CHUNK_ALLOC_FORCE); } /* -- cgit v0.10.2 From f2eda2c6cc138710ae775490397657e8877774b6 Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Wed, 23 Mar 2011 19:48:16 +0530 Subject: fs/9p: Fix revalidate to return correct value revalidate should return > 0 on success. Also return 0 on ENOENT to force do_revalidate to return NULL dentry; Signed-off-by: Aneesh Kumar K.V Signed-off-by: Venkateswararao Jujjuri Signed-off-by: Eric Van Hensbergen diff --git a/fs/9p/vfs_dentry.c b/fs/9p/vfs_dentry.c index b6a3b9f..e022890 100644 --- a/fs/9p/vfs_dentry.c +++ b/fs/9p/vfs_dentry.c @@ -126,7 +126,9 @@ static int v9fs_lookup_revalidate(struct dentry *dentry, struct nameidata *nd) retval = v9fs_refresh_inode_dotl(fid, inode); else retval = v9fs_refresh_inode(fid, inode); - if (retval <= 0) + if (retval == -ENOENT) + return 0; + if (retval < 0) return retval; } out_valid: -- cgit v0.10.2 From c2ed388021a60bb4a9449fddfef770c95875b052 Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Wed, 23 Mar 2011 15:11:27 +0530 Subject: fs/9p: Use write_inode for data sync on server Signed-off-by: Aneesh Kumar K.V Signed-off-by: Venkateswararao Jujjuri Signed-off-by: Eric Van Hensbergen diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index f3eed33..acdc265 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -307,6 +307,51 @@ static int v9fs_drop_inode(struct inode *inode) return 1; } +static int v9fs_write_inode(struct inode *inode, + struct writeback_control *wbc) +{ + int ret; + struct p9_wstat wstat; + struct v9fs_inode *v9inode; + /* + * send an fsync request to server irrespective of + * wbc->sync_mode. + */ + P9_DPRINTK(P9_DEBUG_VFS, "%s: inode %p\n", __func__, inode); + v9inode = V9FS_I(inode); + if (!v9inode->writeback_fid) + return 0; + v9fs_blank_wstat(&wstat); + + ret = p9_client_wstat(v9inode->writeback_fid, &wstat); + if (ret < 0) { + __mark_inode_dirty(inode, I_DIRTY_DATASYNC); + return ret; + } + return 0; +} + +static int v9fs_write_inode_dotl(struct inode *inode, + struct writeback_control *wbc) +{ + int ret; + struct v9fs_inode *v9inode; + /* + * send an fsync request to server irrespective of + * wbc->sync_mode. + */ + P9_DPRINTK(P9_DEBUG_VFS, "%s: inode %p\n", __func__, inode); + v9inode = V9FS_I(inode); + if (!v9inode->writeback_fid) + return 0; + ret = p9_client_fsync(v9inode->writeback_fid, 0); + if (ret < 0) { + __mark_inode_dirty(inode, I_DIRTY_DATASYNC); + return ret; + } + return 0; +} + static const struct super_operations v9fs_super_ops = { .alloc_inode = v9fs_alloc_inode, .destroy_inode = v9fs_destroy_inode, @@ -314,6 +359,7 @@ static const struct super_operations v9fs_super_ops = { .evict_inode = v9fs_evict_inode, .show_options = generic_show_options, .umount_begin = v9fs_umount_begin, + .write_inode = v9fs_write_inode, }; static const struct super_operations v9fs_super_ops_dotl = { @@ -325,6 +371,7 @@ static const struct super_operations v9fs_super_ops_dotl = { .evict_inode = v9fs_evict_inode, .show_options = generic_show_options, .umount_begin = v9fs_umount_begin, + .write_inode = v9fs_write_inode_dotl, }; struct file_system_type v9fs_fs_type = { -- cgit v0.10.2 From df5d8c80f1871d9e79af4b0f3656a9528a7d4bab Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Thu, 24 Mar 2011 20:38:35 +0530 Subject: 9p: revert tsyncfs related changes Now that we use write_inode to flush server cache related to fid, we don't need tsyncfs either fort dotl or dotu protocols. For dotu this helps to do a more efficient server flush. Signed-off-by: Aneesh Kumar K.V Signed-off-by: Venkateswararao Jujjuri Signed-off-by: Eric Van Hensbergen diff --git a/fs/9p/fid.c b/fs/9p/fid.c index 0ee5945..85b67ff 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -286,11 +286,9 @@ static struct p9_fid *v9fs_fid_clone_with_uid(struct dentry *dentry, uid_t uid) struct p9_fid *v9fs_writeback_fid(struct dentry *dentry) { - int err, flags; + int err; struct p9_fid *fid; - struct v9fs_session_info *v9ses; - v9ses = v9fs_dentry2v9ses(dentry); fid = v9fs_fid_clone_with_uid(dentry, 0); if (IS_ERR(fid)) goto error_out; @@ -299,17 +297,8 @@ struct p9_fid *v9fs_writeback_fid(struct dentry *dentry) * dirty pages. We always request for the open fid in read-write * mode so that a partial page write which result in page * read can work. - * - * we don't have a tsyncfs operation for older version - * of protocol. So make sure the write back fid is - * opened in O_SYNC mode. */ - if (!v9fs_proto_dotl(v9ses)) - flags = O_RDWR | O_SYNC; - else - flags = O_RDWR; - - err = p9_client_open(fid, flags); + err = p9_client_open(fid, O_RDWR); if (err < 0) { p9_client_clunk(fid); fid = ERR_PTR(err); diff --git a/fs/9p/v9fs.h b/fs/9p/v9fs.h index 9665c2b..e5ebedf 100644 --- a/fs/9p/v9fs.h +++ b/fs/9p/v9fs.h @@ -116,7 +116,6 @@ struct v9fs_session_info { struct list_head slist; /* list of sessions registered with v9fs */ struct backing_dev_info bdi; struct rw_semaphore rename_sem; - struct p9_fid *root_fid; /* Used for file system sync */ }; /* cache_validity flags */ diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index acdc265..feef6cd 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -154,6 +154,7 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, retval = PTR_ERR(inode); goto release_sb; } + root = d_alloc_root(inode); if (!root) { iput(inode); @@ -185,21 +186,10 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, p9stat_free(st); kfree(st); } - v9fs_fid_add(root, fid); retval = v9fs_get_acl(inode, fid); if (retval) goto release_sb; - /* - * Add the root fid to session info. This is used - * for file system sync. We want a cloned fid here - * so that we can do a sync_filesystem after a - * shrink_dcache_for_umount - */ - v9ses->root_fid = v9fs_fid_clone(root); - if (IS_ERR(v9ses->root_fid)) { - retval = PTR_ERR(v9ses->root_fid); - goto release_sb; - } + v9fs_fid_add(root, fid); P9_DPRINTK(P9_DEBUG_VFS, " simple set mount, return 0\n"); return dget(sb->s_root); @@ -210,11 +200,15 @@ close_session: v9fs_session_close(v9ses); kfree(v9ses); return ERR_PTR(retval); + release_sb: /* - * we will do the session_close and root dentry - * release in the below call. + * we will do the session_close and root dentry release + * in the below call. But we need to clunk fid, because we haven't + * attached the fid to dentry so it won't get clunked + * automatically. */ + p9_client_clunk(fid); deactivate_locked_super(sb); return ERR_PTR(retval); } @@ -232,7 +226,7 @@ static void v9fs_kill_super(struct super_block *s) P9_DPRINTK(P9_DEBUG_VFS, " %p\n", s); kill_anon_super(s); - p9_client_clunk(v9ses->root_fid); + v9fs_session_cancel(v9ses); v9fs_session_close(v9ses); kfree(v9ses); @@ -285,14 +279,6 @@ done: return res; } -static int v9fs_sync_fs(struct super_block *sb, int wait) -{ - struct v9fs_session_info *v9ses = sb->s_fs_info; - - P9_DPRINTK(P9_DEBUG_VFS, "v9fs_sync_fs: super_block %p\n", sb); - return p9_client_sync_fs(v9ses->root_fid); -} - static int v9fs_drop_inode(struct inode *inode) { struct v9fs_session_info *v9ses; @@ -365,7 +351,6 @@ static const struct super_operations v9fs_super_ops = { static const struct super_operations v9fs_super_ops_dotl = { .alloc_inode = v9fs_alloc_inode, .destroy_inode = v9fs_destroy_inode, - .sync_fs = v9fs_sync_fs, .statfs = v9fs_statfs, .drop_inode = v9fs_drop_inode, .evict_inode = v9fs_evict_inode, diff --git a/include/net/9p/9p.h b/include/net/9p/9p.h index cdf2e8a..d2df55b 100644 --- a/include/net/9p/9p.h +++ b/include/net/9p/9p.h @@ -139,8 +139,6 @@ do { \ */ enum p9_msg_t { - P9_TSYNCFS = 0, - P9_RSYNCFS, P9_TLERROR = 6, P9_RLERROR, P9_TSTATFS = 8, diff --git a/include/net/9p/client.h b/include/net/9p/client.h index 85c1413..59b5df5 100644 --- a/include/net/9p/client.h +++ b/include/net/9p/client.h @@ -230,7 +230,6 @@ int p9_client_create_dotl(struct p9_fid *ofid, char *name, u32 flags, u32 mode, gid_t gid, struct p9_qid *qid); int p9_client_clunk(struct p9_fid *fid); int p9_client_fsync(struct p9_fid *fid, int datasync); -int p9_client_sync_fs(struct p9_fid *fid); int p9_client_remove(struct p9_fid *fid); int p9_client_read(struct p9_fid *fid, char *data, char __user *udata, u64 offset, u32 count); diff --git a/net/9p/client.c b/net/9p/client.c index 48b8e08..d72aac7 100644 --- a/net/9p/client.c +++ b/net/9p/client.c @@ -1220,27 +1220,6 @@ error: } EXPORT_SYMBOL(p9_client_fsync); -int p9_client_sync_fs(struct p9_fid *fid) -{ - int err = 0; - struct p9_req_t *req; - struct p9_client *clnt; - - P9_DPRINTK(P9_DEBUG_9P, ">>> TSYNC_FS fid %d\n", fid->fid); - - clnt = fid->clnt; - req = p9_client_rpc(clnt, P9_TSYNCFS, "d", fid->fid); - if (IS_ERR(req)) { - err = PTR_ERR(req); - goto error; - } - P9_DPRINTK(P9_DEBUG_9P, "<<< RSYNCFS fid %d\n", fid->fid); - p9_free_req(clnt, req); -error: - return err; -} -EXPORT_SYMBOL(p9_client_sync_fs); - int p9_client_clunk(struct p9_fid *fid) { int err; -- cgit v0.10.2 From 936bb2d7034165fd6ef7afea54057da65c329a27 Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Thu, 24 Mar 2011 23:04:41 +0530 Subject: fs/9p: Fix error reported by coccicheck Signed-off-by: Aneesh Kumar K.V Signed-off-by: Venkateswararao Jujjuri Signed-off-by: Eric Van Hensbergen diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c index ffbb113..82a7c38 100644 --- a/fs/9p/vfs_inode_dotl.c +++ b/fs/9p/vfs_inode_dotl.c @@ -811,7 +811,7 @@ v9fs_vfs_follow_link_dotl(struct dentry *dentry, struct nameidata *nd) fid = v9fs_fid_lookup(dentry); if (IS_ERR(fid)) { __putname(link); - link = ERR_PTR(PTR_ERR(fid)); + link = ERR_CAST(fid); goto ndset; } retval = p9_client_readlink(fid, &target); -- cgit v0.10.2 From bd8c8ade6b6f109bc3dce14a8d12013f27f2a590 Mon Sep 17 00:00:00 2001 From: "Aneesh Kumar K.V" Date: Thu, 24 Mar 2011 23:14:46 +0530 Subject: 9p: Fix sparse error Signed-off-by: Aneesh Kumar K.V Signed-off-by: Venkateswararao Jujjuri Signed-off-by: Eric Van Hensbergen diff --git a/net/9p/protocol.c b/net/9p/protocol.c index 8a4084f..a7e8997 100644 --- a/net/9p/protocol.c +++ b/net/9p/protocol.c @@ -468,7 +468,8 @@ p9pdu_vwritef(struct p9_fcall *pdu, int proto_version, const char *fmt, case 'E':{ int32_t cnt = va_arg(ap, int32_t); const char *k = va_arg(ap, const void *); - const char *u = va_arg(ap, const void *); + const char __user *u = va_arg(ap, + const void __user *); errcode = p9pdu_writef(pdu, proto_version, "d", cnt); if (!errcode && pdu_write_urw(pdu, k, u, cnt)) diff --git a/net/9p/trans_common.c b/net/9p/trans_common.c index d47880e..e883172 100644 --- a/net/9p/trans_common.c +++ b/net/9p/trans_common.c @@ -66,7 +66,7 @@ p9_payload_gup(struct p9_req_t *req, size_t *pdata_off, int *pdata_len, uint32_t pdata_mapped_pages; struct trans_rpage_info *rpinfo; - *pdata_off = (size_t)req->tc->pubuf & (PAGE_SIZE-1); + *pdata_off = (__force size_t)req->tc->pubuf & (PAGE_SIZE-1); if (*pdata_off) first_page_bytes = min(((size_t)PAGE_SIZE - *pdata_off), diff --git a/net/9p/trans_virtio.c b/net/9p/trans_virtio.c index e8f046b..244e707 100644 --- a/net/9p/trans_virtio.c +++ b/net/9p/trans_virtio.c @@ -326,8 +326,11 @@ req_retry_pinned: outp = pack_sg_list_p(chan->sg, out, VIRTQUEUE_NUM, pdata_off, rpinfo->rp_data, pdata_len); } else { - char *pbuf = req->tc->pubuf ? req->tc->pubuf : - req->tc->pkbuf; + char *pbuf; + if (req->tc->pubuf) + pbuf = (__force char *) req->tc->pubuf; + else + pbuf = req->tc->pkbuf; outp = pack_sg_list(chan->sg, out, VIRTQUEUE_NUM, pbuf, req->tc->pbuf_size); } @@ -352,8 +355,12 @@ req_retry_pinned: in = pack_sg_list_p(chan->sg, out+inp, VIRTQUEUE_NUM, pdata_off, rpinfo->rp_data, pdata_len); } else { - char *pbuf = req->tc->pubuf ? req->tc->pubuf : - req->tc->pkbuf; + char *pbuf; + if (req->tc->pubuf) + pbuf = (__force char *) req->tc->pubuf; + else + pbuf = req->tc->pkbuf; + in = pack_sg_list(chan->sg, out+inp, VIRTQUEUE_NUM, pbuf, req->tc->pbuf_size); } -- cgit v0.10.2 From b76225e22ac98070325ee2ba89473c1e1360c4cb Mon Sep 17 00:00:00 2001 From: Harsh Prateek Bora Date: Thu, 31 Mar 2011 15:49:39 +0530 Subject: net/9p: nwname should be an unsigned int Signed-off-by: Harsh Prateek Bora Signed-off-by: Venkateswararao Jujjuri Signed-off-by: Eric VAn Hensbergen diff --git a/include/net/9p/client.h b/include/net/9p/client.h index 59b5df5..051a99f 100644 --- a/include/net/9p/client.h +++ b/include/net/9p/client.h @@ -218,8 +218,8 @@ void p9_client_disconnect(struct p9_client *clnt); void p9_client_begin_disconnect(struct p9_client *clnt); struct p9_fid *p9_client_attach(struct p9_client *clnt, struct p9_fid *afid, char *uname, u32 n_uname, char *aname); -struct p9_fid *p9_client_walk(struct p9_fid *oldfid, int nwname, char **wnames, - int clone); +struct p9_fid *p9_client_walk(struct p9_fid *oldfid, uint16_t nwname, + char **wnames, int clone); int p9_client_open(struct p9_fid *fid, int mode); int p9_client_fcreate(struct p9_fid *fid, char *name, u32 perm, int mode, char *extension); diff --git a/net/9p/client.c b/net/9p/client.c index d72aac7..7736774 100644 --- a/net/9p/client.c +++ b/net/9p/client.c @@ -929,15 +929,15 @@ error: } EXPORT_SYMBOL(p9_client_attach); -struct p9_fid *p9_client_walk(struct p9_fid *oldfid, int nwname, char **wnames, - int clone) +struct p9_fid *p9_client_walk(struct p9_fid *oldfid, uint16_t nwname, + char **wnames, int clone) { int err; struct p9_client *clnt; struct p9_fid *fid; struct p9_qid *wqids; struct p9_req_t *req; - int16_t nwqids, count; + uint16_t nwqids, count; err = 0; wqids = NULL; @@ -955,7 +955,7 @@ struct p9_fid *p9_client_walk(struct p9_fid *oldfid, int nwname, char **wnames, fid = oldfid; - P9_DPRINTK(P9_DEBUG_9P, ">>> TWALK fids %d,%d nwname %d wname[0] %s\n", + P9_DPRINTK(P9_DEBUG_9P, ">>> TWALK fids %d,%d nwname %ud wname[0] %s\n", oldfid->fid, fid->fid, nwname, wnames ? wnames[0] : NULL); req = p9_client_rpc(clnt, P9_TWALK, "ddT", oldfid->fid, fid->fid, diff --git a/net/9p/protocol.c b/net/9p/protocol.c index a7e8997..b58a501 100644 --- a/net/9p/protocol.c +++ b/net/9p/protocol.c @@ -265,7 +265,7 @@ p9pdu_vreadf(struct p9_fcall *pdu, int proto_version, const char *fmt, } break; case 'T':{ - int16_t *nwname = va_arg(ap, int16_t *); + uint16_t *nwname = va_arg(ap, uint16_t *); char ***wnames = va_arg(ap, char ***); errcode = p9pdu_readf(pdu, proto_version, @@ -496,7 +496,7 @@ p9pdu_vwritef(struct p9_fcall *pdu, int proto_version, const char *fmt, } break; case 'T':{ - int16_t nwname = va_arg(ap, int); + uint16_t nwname = va_arg(ap, int); const char **wnames = va_arg(ap, const char **); errcode = p9pdu_writef(pdu, proto_version, "w", -- cgit v0.10.2 From c1530019e311c91d14b24d8e74d233152d806e45 Mon Sep 17 00:00:00 2001 From: Tim Chen Date: Fri, 15 Apr 2011 11:39:29 -0700 Subject: vfs: Fix absolute RCU path walk failures due to uninitialized seq number During RCU walk in path_lookupat and path_openat, the rcu lookup frequently failed if looking up an absolute path, because when root directory was looked up, seq number was not properly set in nameidata. We dropped out of RCU walk in nameidata_drop_rcu due to mismatch in directory entry's seq number. We reverted to slow path walk that need to take references. With the following patch, I saw a 50% increase in an exim mail server benchmark throughput on a 4-socket Nehalem-EX system. Signed-off-by: Tim Chen Reviewed-by: Andi Kleen Cc: stable@kernel.org (v2.6.38) Signed-off-by: Linus Torvalds diff --git a/fs/namei.c b/fs/namei.c index e6cd611..54fc993 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -697,6 +697,7 @@ static __always_inline void set_root_rcu(struct nameidata *nd) do { seq = read_seqcount_begin(&fs->seq); nd->root = fs->root; + nd->seq = __read_seqcount_begin(&nd->root.dentry->d_seq); } while (read_seqcount_retry(&fs->seq, seq)); } } -- cgit v0.10.2 From 5bbc097d890409d8eff4e3f1d26f11a9d6b7c07e Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Fri, 15 Apr 2011 14:47:40 +0200 Subject: x86, amd: Disable GartTlbWlkErr when BIOS forgets it This patch disables GartTlbWlk errors on AMD Fam10h CPUs if the BIOS forgets to do is (or is just too old). Letting these errors enabled can cause a sync-flood on the CPU causing a reboot. The AMD BKDG recommends disabling GART TLB Wlk Error completely. This patch is the fix for https://bugzilla.kernel.org/show_bug.cgi?id=33012 on my machine. Signed-off-by: Joerg Roedel Link: http://lkml.kernel.org/r/20110415131152.GJ18463@8bytes.org Tested-by: Alexandre Demers Cc: Signed-off-by: H. Peter Anvin diff --git a/arch/x86/include/asm/msr-index.h b/arch/x86/include/asm/msr-index.h index fd5a1f3..3cce714 100644 --- a/arch/x86/include/asm/msr-index.h +++ b/arch/x86/include/asm/msr-index.h @@ -96,11 +96,15 @@ #define MSR_IA32_MC0_ADDR 0x00000402 #define MSR_IA32_MC0_MISC 0x00000403 +#define MSR_AMD64_MC0_MASK 0xc0010044 + #define MSR_IA32_MCx_CTL(x) (MSR_IA32_MC0_CTL + 4*(x)) #define MSR_IA32_MCx_STATUS(x) (MSR_IA32_MC0_STATUS + 4*(x)) #define MSR_IA32_MCx_ADDR(x) (MSR_IA32_MC0_ADDR + 4*(x)) #define MSR_IA32_MCx_MISC(x) (MSR_IA32_MC0_MISC + 4*(x)) +#define MSR_AMD64_MCx_MASK(x) (MSR_AMD64_MC0_MASK + (x)) + /* These are consecutive and not in the normal 4er MCE bank block */ #define MSR_IA32_MC0_CTL2 0x00000280 #define MSR_IA32_MCx_CTL2(x) (MSR_IA32_MC0_CTL2 + (x)) diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index 3ecece0..3532d3b 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -615,6 +615,25 @@ static void __cpuinit init_amd(struct cpuinfo_x86 *c) /* As a rule processors have APIC timer running in deep C states */ if (c->x86 >= 0xf && !cpu_has_amd_erratum(amd_erratum_400)) set_cpu_cap(c, X86_FEATURE_ARAT); + + /* + * Disable GART TLB Walk Errors on Fam10h. We do this here + * because this is always needed when GART is enabled, even in a + * kernel which has no MCE support built in. + */ + if (c->x86 == 0x10) { + /* + * BIOS should disable GartTlbWlk Errors themself. If + * it doesn't do it here as suggested by the BKDG. + * + * Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=33012 + */ + u64 mask; + + rdmsrl(MSR_AMD64_MCx_MASK(4), mask); + mask |= (1 << 10); + wrmsrl(MSR_AMD64_MCx_MASK(4), mask); + } } #ifdef CONFIG_X86_32 -- cgit v0.10.2 From 0d399205edf3a4c290e76ebb36e541593af4a1b4 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Sat, 16 Apr 2011 06:55:39 -0400 Subject: Btrfs end_bio_extent_readpage should look for locked bits A recent commit caches the extent state in end_bio_extent_readpage, but the search it does should look for locked extents. This fixes things to make it more effective. Signed-off-by: Chris Mason diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 1c462f89..5ae0bff 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -1761,7 +1761,7 @@ static void end_bio_extent_readpage(struct bio *bio, int err) prefetchw(&bvec->bv_page->flags); spin_lock(&tree->lock); - state = find_first_extent_bit_state(tree, start, 0); + state = find_first_extent_bit_state(tree, start, EXTENT_LOCKED); if (state && state->start == start) { /* * take a reference on the state, unlock will drop -- cgit v0.10.2 From 6d74119f1a3efad9dc7f79a16c201242324b731f Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 11 Apr 2011 20:20:11 -0400 Subject: Btrfs: avoid taking the chunk_mutex in do_chunk_alloc Everytime we try to allocate disk space we try and see if we can pre-emptively allocate a chunk, but in the common case we don't allocate anything, so there is no sense in taking the chunk_mutex at all. So instead if we are allocating a chunk, mark it in the space_info so we don't get two people trying to allocate at the same time. Thanks, Signed-off-by: Josef Bacik Reviewed-by: Liu Bo diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 0d00a07..2e61fe1 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -740,8 +740,10 @@ struct btrfs_space_info { */ unsigned long reservation_progress; - int full; /* indicates that we cannot allocate any more + int full:1; /* indicates that we cannot allocate any more chunks for this space */ + int chunk_alloc:1; /* set if we are allocating a chunk */ + int force_alloc; /* set if we need to force a chunk alloc for this space */ diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 2647948..31f33ba 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3039,6 +3039,7 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags, found->bytes_may_use = 0; found->full = 0; found->force_alloc = CHUNK_ALLOC_NO_FORCE; + found->chunk_alloc = 0; *space_info = found; list_add_rcu(&found->list, &info->space_info); atomic_set(&found->caching_threads, 0); @@ -3318,10 +3319,9 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans, { struct btrfs_space_info *space_info; struct btrfs_fs_info *fs_info = extent_root->fs_info; + int wait_for_alloc = 0; int ret = 0; - mutex_lock(&fs_info->chunk_mutex); - flags = btrfs_reduce_alloc_profile(extent_root, flags); space_info = __find_space_info(extent_root->fs_info, flags); @@ -3332,21 +3332,40 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans, } BUG_ON(!space_info); +again: spin_lock(&space_info->lock); if (space_info->force_alloc) force = space_info->force_alloc; if (space_info->full) { spin_unlock(&space_info->lock); - goto out; + return 0; } if (!should_alloc_chunk(extent_root, space_info, alloc_bytes, force)) { spin_unlock(&space_info->lock); - goto out; + return 0; + } else if (space_info->chunk_alloc) { + wait_for_alloc = 1; + } else { + space_info->chunk_alloc = 1; } spin_unlock(&space_info->lock); + mutex_lock(&fs_info->chunk_mutex); + + /* + * The chunk_mutex is held throughout the entirety of a chunk + * allocation, so once we've acquired the chunk_mutex we know that the + * other guy is done and we need to recheck and see if we should + * allocate. + */ + if (wait_for_alloc) { + mutex_unlock(&fs_info->chunk_mutex); + wait_for_alloc = 0; + goto again; + } + /* * If we have mixed data/metadata chunks we want to make sure we keep * allocating mixed chunks instead of individual chunks. @@ -3372,9 +3391,10 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans, space_info->full = 1; else ret = 1; + space_info->force_alloc = CHUNK_ALLOC_NO_FORCE; + space_info->chunk_alloc = 0; spin_unlock(&space_info->lock); -out: mutex_unlock(&extent_root->fs_info->chunk_mutex); return ret; } -- cgit v0.10.2 From a237c1c5bc5dc5c76a21be922dca4826f3eca8ca Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Sat, 16 Apr 2011 13:27:55 +0200 Subject: block: let io_schedule() flush the plug inline Linus correctly observes that the most important dispatch cases are now done from kblockd, this isn't ideal for latency reasons. The original reason for switching dispatches out-of-line was to avoid too deep a stack, so by _only_ letting the "accidental" flush directly in schedule() be guarded by offload to kblockd, we should be able to get the best of both worlds. So add a blk_schedule_flush_plug() that offloads to kblockd, and only use that from the schedule() path. Signed-off-by: Jens Axboe diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 1c76506..ec0357d 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -872,6 +872,14 @@ static inline void blk_flush_plug(struct task_struct *tsk) struct blk_plug *plug = tsk->plug; if (plug) + blk_flush_plug_list(plug, false); +} + +static inline void blk_schedule_flush_plug(struct task_struct *tsk) +{ + struct blk_plug *plug = tsk->plug; + + if (plug) blk_flush_plug_list(plug, true); } @@ -1317,6 +1325,11 @@ static inline void blk_flush_plug(struct task_struct *task) { } +static inline void blk_schedule_flush_plug(struct task_struct *task) +{ +} + + static inline bool blk_needs_flush_plug(struct task_struct *tsk) { return false; diff --git a/kernel/sched.c b/kernel/sched.c index a187c3f..312f8b9 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -4118,7 +4118,7 @@ need_resched: */ if (blk_needs_flush_plug(prev)) { raw_spin_unlock(&rq->lock); - blk_flush_plug(prev); + blk_schedule_flush_plug(prev); raw_spin_lock(&rq->lock); } } -- cgit v0.10.2 From 49cac01e1fa74174d72adb0e872504a7fefd7c01 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Sat, 16 Apr 2011 13:51:05 +0200 Subject: block: make unplug timer trace event correspond to the schedule() unplug It's a pretty close match to what we had before - the timer triggering would mean that nobody unplugged the plug in due time, in the new scheme this matches very closely what the schedule() unplug now is. It's essentially the difference between an explicit unplug (IO unplug) or an implicit unplug (timer unplug, we scheduled with pending IO queued). Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 3c81210..78b7b0c 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2662,17 +2662,23 @@ static int plug_rq_cmp(void *priv, struct list_head *a, struct list_head *b) return !(rqa->q <= rqb->q); } +/* + * If 'from_schedule' is true, then postpone the dispatch of requests + * until a safe kblockd context. We due this to avoid accidental big + * additional stack usage in driver dispatch, in places where the originally + * plugger did not intend it. + */ static void queue_unplugged(struct request_queue *q, unsigned int depth, - bool force_kblockd) + bool from_schedule) { - trace_block_unplug_io(q, depth); - __blk_run_queue(q, force_kblockd); + trace_block_unplug(q, depth, !from_schedule); + __blk_run_queue(q, from_schedule); if (q->unplugged_fn) q->unplugged_fn(q); } -void blk_flush_plug_list(struct blk_plug *plug, bool force_kblockd) +void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) { struct request_queue *q; unsigned long flags; @@ -2707,7 +2713,7 @@ void blk_flush_plug_list(struct blk_plug *plug, bool force_kblockd) BUG_ON(!rq->q); if (rq->q != q) { if (q) { - queue_unplugged(q, depth, force_kblockd); + queue_unplugged(q, depth, from_schedule); spin_unlock(q->queue_lock); } q = rq->q; @@ -2728,7 +2734,7 @@ void blk_flush_plug_list(struct blk_plug *plug, bool force_kblockd) } if (q) { - queue_unplugged(q, depth, force_kblockd); + queue_unplugged(q, depth, from_schedule); spin_unlock(q->queue_lock); } diff --git a/include/trace/events/block.h b/include/trace/events/block.h index 006e60b..bf36654 100644 --- a/include/trace/events/block.h +++ b/include/trace/events/block.h @@ -401,9 +401,9 @@ TRACE_EVENT(block_plug, DECLARE_EVENT_CLASS(block_unplug, - TP_PROTO(struct request_queue *q, unsigned int depth), + TP_PROTO(struct request_queue *q, unsigned int depth, bool explicit), - TP_ARGS(q, depth), + TP_ARGS(q, depth, explicit), TP_STRUCT__entry( __field( int, nr_rq ) @@ -419,18 +419,19 @@ DECLARE_EVENT_CLASS(block_unplug, ); /** - * block_unplug_io - release of operations requests in request queue + * block_unplug - release of operations requests in request queue * @q: request queue to unplug * @depth: number of requests just added to the queue + * @explicit: whether this was an explicit unplug, or one from schedule() * * Unplug request queue @q because device driver is scheduled to work * on elements in the request queue. */ -DEFINE_EVENT(block_unplug, block_unplug_io, +DEFINE_EVENT(block_unplug, block_unplug, - TP_PROTO(struct request_queue *q, unsigned int depth), + TP_PROTO(struct request_queue *q, unsigned int depth, bool explicit), - TP_ARGS(q, depth) + TP_ARGS(q, depth, explicit) ); /** diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c index 3e3970d..6957aa2 100644 --- a/kernel/trace/blktrace.c +++ b/kernel/trace/blktrace.c @@ -850,16 +850,21 @@ static void blk_add_trace_plug(void *ignore, struct request_queue *q) __blk_add_trace(bt, 0, 0, 0, BLK_TA_PLUG, 0, 0, NULL); } -static void blk_add_trace_unplug_io(void *ignore, struct request_queue *q, - unsigned int depth) +static void blk_add_trace_unplug(void *ignore, struct request_queue *q, + unsigned int depth, bool explicit) { struct blk_trace *bt = q->blk_trace; if (bt) { __be64 rpdu = cpu_to_be64(depth); + u32 what; - __blk_add_trace(bt, 0, 0, 0, BLK_TA_UNPLUG_IO, 0, - sizeof(rpdu), &rpdu); + if (explicit) + what = BLK_TA_UNPLUG_IO; + else + what = BLK_TA_UNPLUG_TIMER; + + __blk_add_trace(bt, 0, 0, 0, what, 0, sizeof(rpdu), &rpdu); } } @@ -1002,7 +1007,7 @@ static void blk_register_tracepoints(void) WARN_ON(ret); ret = register_trace_block_plug(blk_add_trace_plug, NULL); WARN_ON(ret); - ret = register_trace_block_unplug_io(blk_add_trace_unplug_io, NULL); + ret = register_trace_block_unplug(blk_add_trace_unplug, NULL); WARN_ON(ret); ret = register_trace_block_split(blk_add_trace_split, NULL); WARN_ON(ret); @@ -1017,7 +1022,7 @@ static void blk_unregister_tracepoints(void) unregister_trace_block_rq_remap(blk_add_trace_rq_remap, NULL); unregister_trace_block_bio_remap(blk_add_trace_bio_remap, NULL); unregister_trace_block_split(blk_add_trace_split, NULL); - unregister_trace_block_unplug_io(blk_add_trace_unplug_io, NULL); + unregister_trace_block_unplug(blk_add_trace_unplug, NULL); unregister_trace_block_plug(blk_add_trace_plug, NULL); unregister_trace_block_sleeprq(blk_add_trace_sleeprq, NULL); unregister_trace_block_getrq(blk_add_trace_getrq, NULL); @@ -1332,6 +1337,7 @@ static const struct { [__BLK_TA_COMPLETE] = {{ "C", "complete" }, blk_log_with_error }, [__BLK_TA_PLUG] = {{ "P", "plug" }, blk_log_plug }, [__BLK_TA_UNPLUG_IO] = {{ "U", "unplug_io" }, blk_log_unplug }, + [__BLK_TA_UNPLUG_TIMER] = {{ "UT", "unplug_timer" }, blk_log_unplug }, [__BLK_TA_INSERT] = {{ "I", "insert" }, blk_log_generic }, [__BLK_TA_SPLIT] = {{ "X", "split" }, blk_log_split }, [__BLK_TA_BOUNCE] = {{ "B", "bounce" }, blk_log_generic }, -- cgit v0.10.2 From e9569c1511d2590a27b46b94bafb7acece034e5c Mon Sep 17 00:00:00 2001 From: Jonathan Cameron Date: Thu, 14 Apr 2011 12:11:42 +0100 Subject: ARM: 6881/1: cputype.h uses __attribute_const__ which requires including kernel.h Issue manifests as: In file included from arch/arm/mach-pxa/include/mach/hardware.h:62, from arch/arm/mach-pxa/include/mach/gpio.h:28, from /home/jic23/src/kernel/temp-remove/arch/arm/include/asm/gpio.h:5, from include/linux/gpio.h:7, from drivers/staging/iio/gyro/adis16080_core.c:8: /home/jic23/src/kernel/temp-remove/arch/arm/include/asm/cputype.h:57: error: expected '=', ',', ';', 'asm' or '__attribute__' before 'read_cpuid_id' ... Signed-off-by: Jonathan Cameron Signed-off-by: Russell King diff --git a/arch/arm/include/asm/cputype.h b/arch/arm/include/asm/cputype.h index ed5bc9e..cd4458f 100644 --- a/arch/arm/include/asm/cputype.h +++ b/arch/arm/include/asm/cputype.h @@ -2,6 +2,7 @@ #define __ASM_ARM_CPUTYPE_H #include +#include #define CPUID_ID 0 #define CPUID_CACHETYPE 1 -- cgit v0.10.2 From a920ff41cb3d2b03da095c4fa1a11b71417ae2a4 Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Sun, 17 Apr 2011 10:20:19 +0200 Subject: i2c: Improve deprecation warnings When warning on the use of deprecated i2c_driver methods attach_adapter and detach_adapter, mention the name of the driver which needs to be updated. Signed-off-by: Jean Delvare Cc: Benjamin Herrenschmidt diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c index 70c30e6..9a58994 100644 --- a/drivers/i2c/i2c-core.c +++ b/drivers/i2c/i2c-core.c @@ -797,7 +797,8 @@ static int i2c_do_add_adapter(struct i2c_driver *driver, /* Let legacy drivers scan this bus for matching devices */ if (driver->attach_adapter) { - dev_warn(&adap->dev, "attach_adapter method is deprecated\n"); + dev_warn(&adap->dev, "%s: attach_adapter method is deprecated\n", + driver->driver.name); dev_warn(&adap->dev, "Please use another way to instantiate " "your i2c_client\n"); /* We ignore the return code; if it fails, too bad */ @@ -984,7 +985,8 @@ static int i2c_do_del_adapter(struct i2c_driver *driver, if (!driver->detach_adapter) return 0; - dev_warn(&adapter->dev, "detach_adapter method is deprecated\n"); + dev_warn(&adapter->dev, "%s: detach_adapter method is deprecated\n", + driver->driver.name); res = driver->detach_adapter(adapter); if (res) dev_err(&adapter->dev, "detach_adapter failed (%d) " -- cgit v0.10.2 From d3b3e15da14ded61c9654db05863b04a2435f4cc Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Sun, 17 Apr 2011 10:20:19 +0200 Subject: i2c-algo-bit: Call pre/post_xfer for bit_test Apparently some distros set i2c-algo-bit.bit_test to 1 by default. In some cases this causes i2c_bit_add_bus to fail and prevents the i2c bus from being added. In the radeon case, we fail to add the ddc i2c buses which prevents the driver from being able to detect attached monitors. The i2c bus works fine even if bit_test fails. This is likely due to gpio switching that is required and handled in the pre/post_xfer hooks, so call the pre/post_xfer hooks in the bit test as well. Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=36221 Signed-off-by: Alex Deucher Signed-off-by: Jean Delvare Cc: stable@kernel.org [.38 down to .34] diff --git a/drivers/i2c/algos/i2c-algo-bit.c b/drivers/i2c/algos/i2c-algo-bit.c index 38319a6..d6d5868 100644 --- a/drivers/i2c/algos/i2c-algo-bit.c +++ b/drivers/i2c/algos/i2c-algo-bit.c @@ -232,9 +232,17 @@ static int i2c_inb(struct i2c_adapter *i2c_adap) * Sanity check for the adapter hardware - check the reaction of * the bus lines only if it seems to be idle. */ -static int test_bus(struct i2c_algo_bit_data *adap, char *name) +static int test_bus(struct i2c_adapter *i2c_adap) { - int scl, sda; + struct i2c_algo_bit_data *adap = i2c_adap->algo_data; + const char *name = i2c_adap->name; + int scl, sda, ret; + + if (adap->pre_xfer) { + ret = adap->pre_xfer(i2c_adap); + if (ret < 0) + return -ENODEV; + } if (adap->getscl == NULL) pr_info("%s: Testing SDA only, SCL is not readable\n", name); @@ -297,11 +305,19 @@ static int test_bus(struct i2c_algo_bit_data *adap, char *name) "while pulling SCL high!\n", name); goto bailout; } + + if (adap->post_xfer) + adap->post_xfer(i2c_adap); + pr_info("%s: Test OK\n", name); return 0; bailout: sdahi(adap); sclhi(adap); + + if (adap->post_xfer) + adap->post_xfer(i2c_adap); + return -ENODEV; } @@ -607,7 +623,7 @@ static int __i2c_bit_add_bus(struct i2c_adapter *adap, int ret; if (bit_test) { - ret = test_bus(bit_adap, adap->name); + ret = test_bus(adap); if (ret < 0) return -ENODEV; } -- cgit v0.10.2 From fff3e5ade4455a4b42a19c95dd7a167a3cb7956a Mon Sep 17 00:00:00 2001 From: Milton Miller Date: Thu, 14 Apr 2011 10:30:08 -0500 Subject: fs: synchronize_rcu when unregister_filesystem success not failure While checking unregister_filesystem for saftey vs extra calls for "ext4: register ext2 and ext3 alias after ext4" I realized that the synchronize_rcu() was called on the error path but not on the success path. Cc: stable (2.6.38) Signed-off-by: Milton Miller [ This probably won't really make a difference since commit d863b50ab013 ("vfs: call rcu_barrier after ->kill_sb()"), but it's the right thing to do. - Linus ] Signed-off-by: Linus Torvalds diff --git a/fs/filesystems.c b/fs/filesystems.c index 751d6b2..0845f84 100644 --- a/fs/filesystems.c +++ b/fs/filesystems.c @@ -110,14 +110,13 @@ int unregister_filesystem(struct file_system_type * fs) *tmp = fs->next; fs->next = NULL; write_unlock(&file_systems_lock); + synchronize_rcu(); return 0; } tmp = &(*tmp)->next; } write_unlock(&file_systems_lock); - synchronize_rcu(); - return -EINVAL; } -- cgit v0.10.2 From 90fd30c914ec71eb8dc91259bf720b3641672696 Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Sun, 17 Apr 2011 13:05:23 -0700 Subject: alpha: Don't force -Werror. There are outstanding gcc 4.6 warnings that need to be cleaned up in the subdirectory. No sense forcing the issue immediately. Signed-off-by: Richard Henderson Signed-off-by: Linus Torvalds diff --git a/arch/alpha/kernel/Makefile b/arch/alpha/kernel/Makefile index 9bb7b858..7a6d908 100644 --- a/arch/alpha/kernel/Makefile +++ b/arch/alpha/kernel/Makefile @@ -4,7 +4,7 @@ extra-y := head.o vmlinux.lds asflags-y := $(KBUILD_CFLAGS) -ccflags-y := -Werror -Wno-sign-compare +ccflags-y := -Wno-sign-compare obj-y := entry.o traps.o process.o init_task.o osf_sys.o irq.o \ irq_alpha.o signal.o setup.o ptrace.o time.o \ -- cgit v0.10.2 From 280da4e4d306667b7faa95152b54f7ca4266ff1e Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Sun, 17 Apr 2011 13:05:24 -0700 Subject: alpha: Remove set but unused variables. This is a new warning in gcc 4.6. Several of these variables are used within #if 0 code, which probably ought to be removed. Most of the changes are legitimate cleanups. Signed-off-by: Richard Henderson Signed-off-by: Linus Torvalds diff --git a/arch/alpha/kernel/core_mcpcia.c b/arch/alpha/kernel/core_mcpcia.c index 381fec0..da7bcc3 100644 --- a/arch/alpha/kernel/core_mcpcia.c +++ b/arch/alpha/kernel/core_mcpcia.c @@ -88,7 +88,7 @@ conf_read(unsigned long addr, unsigned char type1, { unsigned long flags; unsigned long mid = MCPCIA_HOSE2MID(hose->index); - unsigned int stat0, value, temp, cpu; + unsigned int stat0, value, cpu; cpu = smp_processor_id(); @@ -101,7 +101,7 @@ conf_read(unsigned long addr, unsigned char type1, stat0 = *(vuip)MCPCIA_CAP_ERR(mid); *(vuip)MCPCIA_CAP_ERR(mid) = stat0; mb(); - temp = *(vuip)MCPCIA_CAP_ERR(mid); + *(vuip)MCPCIA_CAP_ERR(mid); DBG_CFG(("conf_read: MCPCIA_CAP_ERR(%d) was 0x%x\n", mid, stat0)); mb(); @@ -136,7 +136,7 @@ conf_write(unsigned long addr, unsigned int value, unsigned char type1, { unsigned long flags; unsigned long mid = MCPCIA_HOSE2MID(hose->index); - unsigned int stat0, temp, cpu; + unsigned int stat0, cpu; cpu = smp_processor_id(); @@ -145,7 +145,7 @@ conf_write(unsigned long addr, unsigned int value, unsigned char type1, /* Reset status register to avoid losing errors. */ stat0 = *(vuip)MCPCIA_CAP_ERR(mid); *(vuip)MCPCIA_CAP_ERR(mid) = stat0; mb(); - temp = *(vuip)MCPCIA_CAP_ERR(mid); + *(vuip)MCPCIA_CAP_ERR(mid); DBG_CFG(("conf_write: MCPCIA CAP_ERR(%d) was 0x%x\n", mid, stat0)); draina(); @@ -157,7 +157,7 @@ conf_write(unsigned long addr, unsigned int value, unsigned char type1, *((vuip)addr) = value; mb(); mb(); /* magic */ - temp = *(vuip)MCPCIA_CAP_ERR(mid); /* read to force the write */ + *(vuip)MCPCIA_CAP_ERR(mid); /* read to force the write */ mcheck_expected(cpu) = 0; mb(); @@ -572,12 +572,10 @@ mcpcia_print_system_area(unsigned long la_ptr) void mcpcia_machine_check(unsigned long vector, unsigned long la_ptr) { - struct el_common *mchk_header; struct el_MCPCIA_uncorrected_frame_mcheck *mchk_logout; unsigned int cpu = smp_processor_id(); int expected; - mchk_header = (struct el_common *)la_ptr; mchk_logout = (struct el_MCPCIA_uncorrected_frame_mcheck *)la_ptr; expected = mcheck_expected(cpu); diff --git a/arch/alpha/kernel/err_titan.c b/arch/alpha/kernel/err_titan.c index c3b3781..14b26c4 100644 --- a/arch/alpha/kernel/err_titan.c +++ b/arch/alpha/kernel/err_titan.c @@ -533,8 +533,6 @@ static struct el_subpacket_annotation el_titan_annotations[] = { static struct el_subpacket * el_process_regatta_subpacket(struct el_subpacket *header) { - int status; - if (header->class != EL_CLASS__REGATTA_FAMILY) { printk("%s ** Unexpected header CLASS %d TYPE %d, aborting\n", err_print_prefix, @@ -551,7 +549,7 @@ el_process_regatta_subpacket(struct el_subpacket *header) printk("%s ** Occurred on CPU %d:\n", err_print_prefix, (int)header->by_type.regatta_frame.cpuid); - status = privateer_process_logout_frame((struct el_common *) + privateer_process_logout_frame((struct el_common *) header->by_type.regatta_frame.data_start, 1); break; default: diff --git a/arch/alpha/kernel/setup.c b/arch/alpha/kernel/setup.c index d2634e4..edbddcb 100644 --- a/arch/alpha/kernel/setup.c +++ b/arch/alpha/kernel/setup.c @@ -1404,8 +1404,6 @@ determine_cpu_caches (unsigned int cpu_type) case PCA56_CPU: case PCA57_CPU: { - unsigned long cbox_config, size; - if (cpu_type == PCA56_CPU) { L1I = CSHAPE(16*1024, 6, 1); L1D = CSHAPE(8*1024, 5, 1); @@ -1415,10 +1413,12 @@ determine_cpu_caches (unsigned int cpu_type) } L3 = -1; +#if 0 + unsigned long cbox_config, size; + cbox_config = *(vulp) phys_to_virt (0xfffff00008UL); size = 512*1024 * (1 << ((cbox_config >> 12) & 3)); -#if 0 L2 = ((cbox_config >> 31) & 1 ? CSHAPE (size, 6, 1) : -1); #else L2 = external_cache_probe(512*1024, 6); diff --git a/arch/alpha/kernel/smc37c93x.c b/arch/alpha/kernel/smc37c93x.c index 3e6a289..6886b83 100644 --- a/arch/alpha/kernel/smc37c93x.c +++ b/arch/alpha/kernel/smc37c93x.c @@ -79,7 +79,6 @@ static unsigned long __init SMCConfigState(unsigned long baseAddr) { unsigned char devId; - unsigned char devRev; unsigned long configPort; unsigned long indexPort; @@ -100,7 +99,7 @@ static unsigned long __init SMCConfigState(unsigned long baseAddr) devId = inb(dataPort); if (devId == VALID_DEVICE_ID) { outb(DEVICE_REV, indexPort); - devRev = inb(dataPort); + /* unsigned char devRev = */ inb(dataPort); break; } else diff --git a/arch/alpha/kernel/sys_wildfire.c b/arch/alpha/kernel/sys_wildfire.c index d3cb28b..d92cdc7 100644 --- a/arch/alpha/kernel/sys_wildfire.c +++ b/arch/alpha/kernel/sys_wildfire.c @@ -156,7 +156,6 @@ static void __init wildfire_init_irq_per_pca(int qbbno, int pcano) { int i, irq_bias; - unsigned long io_bias; static struct irqaction isa_enable = { .handler = no_action, .name = "isa_enable", @@ -165,10 +164,12 @@ wildfire_init_irq_per_pca(int qbbno, int pcano) irq_bias = qbbno * (WILDFIRE_PCA_PER_QBB * WILDFIRE_IRQ_PER_PCA) + pcano * WILDFIRE_IRQ_PER_PCA; +#if 0 + unsigned long io_bias; + /* Only need the following for first PCI bus per PCA. */ io_bias = WILDFIRE_IO(qbbno, pcano<<1) - WILDFIRE_IO_BIAS; -#if 0 outb(0, DMA1_RESET_REG + io_bias); outb(0, DMA2_RESET_REG + io_bias); outb(DMA_MODE_CASCADE, DMA2_MODE_REG + io_bias); -- cgit v0.10.2 From 6181318897258749a637829c542b8448fdce346d Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Sun, 17 Apr 2011 13:05:25 -0700 Subject: alpha: Fix RTC interrupt setup. Following commit 091738a266fc ("genirq: Remove real old transition functions") we removed an automatic conversion of no_irq_chip to dummy_irq_chip. This change needs to be propagated back into the alpha backend. Signed-off-by: Richard Henderson Signed-off-by: Linus Torvalds diff --git a/arch/alpha/kernel/irq_alpha.c b/arch/alpha/kernel/irq_alpha.c index 1479dc6..51b7fbd 100644 --- a/arch/alpha/kernel/irq_alpha.c +++ b/arch/alpha/kernel/irq_alpha.c @@ -228,7 +228,7 @@ struct irqaction timer_irqaction = { void __init init_rtc_irq(void) { - irq_set_chip_and_handler_name(RTC_IRQ, &no_irq_chip, + irq_set_chip_and_handler_name(RTC_IRQ, &dummy_irq_chip, handle_simple_irq, "RTC"); setup_irq(RTC_IRQ, &timer_irqaction); } -- cgit v0.10.2 From a78eda5cd3055852231ad10cfd047e65cb44dfde Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Sun, 17 Apr 2011 13:05:26 -0700 Subject: alpha: Fix uninitialized value in read_persistent_clock. Signed-off-by: Richard Henderson Signed-off-by: Linus Torvalds diff --git a/arch/alpha/kernel/time.c b/arch/alpha/kernel/time.c index a58e84f..918e8e0 100644 --- a/arch/alpha/kernel/time.c +++ b/arch/alpha/kernel/time.c @@ -153,6 +153,7 @@ void read_persistent_clock(struct timespec *ts) year += 100; ts->tv_sec = mktime(year, mon, day, hour, min, sec); + ts->tv_nsec = 0; } -- cgit v0.10.2 From c8a75b345b38a9b63ffd077c0eb71707fa194fc6 Mon Sep 17 00:00:00 2001 From: Francois Romieu Date: Sun, 17 Apr 2011 17:46:40 -0700 Subject: r8169: add Realtek as maintainer. Per Hayes's request. Signed-off-by: Francois Romieu Signed-off-by: David S. Miller diff --git a/MAINTAINERS b/MAINTAINERS index 649600c..c85368d 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -151,6 +151,7 @@ S: Maintained F: drivers/net/hamradio/6pack.c 8169 10/100/1000 GIGABIT ETHERNET DRIVER +M: Realtek linux nic maintainers M: Francois Romieu L: netdev@vger.kernel.org S: Maintained -- cgit v0.10.2 From 70dda99c41fc8a153e09bdba9adb7805ba2a4bb3 Mon Sep 17 00:00:00 2001 From: Yaniv Rosner Date: Sun, 17 Apr 2011 17:50:01 -0700 Subject: bnx2x: Fix port identification problem This patch fixes port identification on optic devices when there's no link on the port. Signed-off-by: Yaniv Rosner Signed-off-by: Eilon Greenstein Signed-off-by: David S. Miller diff --git a/drivers/net/bnx2x/bnx2x_ethtool.c b/drivers/net/bnx2x/bnx2x_ethtool.c index f505015..89cb977 100644 --- a/drivers/net/bnx2x/bnx2x_ethtool.c +++ b/drivers/net/bnx2x/bnx2x_ethtool.c @@ -2114,19 +2114,18 @@ static int bnx2x_phys_id(struct net_device *dev, u32 data) for (i = 0; i < (data * 2); i++) { if ((i % 2) == 0) bnx2x_set_led(&bp->link_params, &bp->link_vars, - LED_MODE_OPER, SPEED_1000); + LED_MODE_ON, SPEED_1000); else bnx2x_set_led(&bp->link_params, &bp->link_vars, - LED_MODE_OFF, 0); + LED_MODE_FRONT_PANEL_OFF, 0); msleep_interruptible(500); if (signal_pending(current)) break; } - if (bp->link_vars.link_up) - bnx2x_set_led(&bp->link_params, &bp->link_vars, LED_MODE_OPER, - bp->link_vars.line_speed); + bnx2x_set_led(&bp->link_params, &bp->link_vars, + LED_MODE_OPER, bp->link_vars.line_speed); return 0; } -- cgit v0.10.2 From 7c7a81b53e581d727d069cc45df5510516faac31 Mon Sep 17 00:00:00 2001 From: Paul Gortmaker Date: Wed, 13 Apr 2011 06:30:08 +0000 Subject: powerpc/kexec: Fix regression causing compile failure on UP Recent commit b987812b3fcaf70fdf0037589e5d2f5f2453e6ce caused a compile failure on UP because a considerably large block of the file was included within CONFIG_SMP, hence making a stub function not exposed on UP builds when it needed to be. Relocate the stub to the #else /* ! CONFIG_SMP */ section and also annotate the relevant else/endif so that nobody else falls into the same trap I did. Reported-by: Michael Guntsche Signed-off-by: Paul Gortmaker Signed-off-by: Benjamin Herrenschmidt diff --git a/arch/powerpc/kernel/crash.c b/arch/powerpc/kernel/crash.c index 3d3d416..5b5e1f0 100644 --- a/arch/powerpc/kernel/crash.c +++ b/arch/powerpc/kernel/crash.c @@ -163,7 +163,7 @@ static void crash_kexec_prepare_cpus(int cpu) } /* wait for all the CPUs to hit real mode but timeout if they don't come in */ -#if defined(CONFIG_PPC_STD_MMU_64) && defined(CONFIG_SMP) +#ifdef CONFIG_PPC_STD_MMU_64 static void crash_kexec_wait_realmode(int cpu) { unsigned int msecs; @@ -188,9 +188,7 @@ static void crash_kexec_wait_realmode(int cpu) } mb(); } -#else -static inline void crash_kexec_wait_realmode(int cpu) {} -#endif +#endif /* CONFIG_PPC_STD_MMU_64 */ /* * This function will be called by secondary cpus or by kexec cpu @@ -235,7 +233,9 @@ void crash_kexec_secondary(struct pt_regs *regs) crash_ipi_callback(regs); } -#else +#else /* ! CONFIG_SMP */ +static inline void crash_kexec_wait_realmode(int cpu) {} + static void crash_kexec_prepare_cpus(int cpu) { /* @@ -255,7 +255,7 @@ void crash_kexec_secondary(struct pt_regs *regs) { cpus_in_sr = CPU_MASK_NONE; } -#endif +#endif /* CONFIG_SMP */ /* * Register a function to be called on shutdown. Only use this if you -- cgit v0.10.2 From 127493d5dc73589cbe00ea5ec8357cc2a4c0d82a Mon Sep 17 00:00:00 2001 From: Nishanth Aravamudan Date: Wed, 13 Apr 2011 19:45:59 +0000 Subject: powerpc/pseries: Use a kmem cache for DTL buffers PAPR specifies that DTL buffers can not cross AMS environments (aka CMO in the PAPR) and can not cross a memory entitlement granule boundary (4k). This is found in section 14.11.3.2 H_REGISTER_VPA of the PAPR. kmalloc does not guarantee an alignment of the allocation, though, beyond 8 bytes (at least in my understanding). Create a special kmem cache for DTL buffers with the alignment requirement. Signed-off-by: Benjamin Herrenschmidt diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c index 0007241..6c42cfd 100644 --- a/arch/powerpc/platforms/pseries/setup.c +++ b/arch/powerpc/platforms/pseries/setup.c @@ -287,14 +287,22 @@ static int alloc_dispatch_logs(void) int cpu, ret; struct paca_struct *pp; struct dtl_entry *dtl; + struct kmem_cache *dtl_cache; if (!firmware_has_feature(FW_FEATURE_SPLPAR)) return 0; + dtl_cache = kmem_cache_create("dtl", DISPATCH_LOG_BYTES, + DISPATCH_LOG_BYTES, 0, NULL); + if (!dtl_cache) { + pr_warn("Failed to create dispatch trace log buffer cache\n"); + pr_warn("Stolen time statistics will be unreliable\n"); + return 0; + } + for_each_possible_cpu(cpu) { pp = &paca[cpu]; - dtl = kmalloc_node(DISPATCH_LOG_BYTES, GFP_KERNEL, - cpu_to_node(cpu)); + dtl = kmem_cache_alloc(dtl_cache, GFP_KERNEL); if (!dtl) { pr_warn("Failed to allocate dispatch trace log for cpu %d\n", cpu); -- cgit v0.10.2 From 84ffae55af79d7b8834fd0c08d0d1ebf2c77f91e Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Thu, 7 Apr 2011 21:44:21 +0000 Subject: powerpc: Fix oops if scan_dispatch_log is called too early We currently enable interrupts before the dispatch log for the boot cpu is setup. If a timer interrupt comes in early enough we oops in scan_dispatch_log: Unable to handle kernel paging request for data at address 0x00000010 ... .scan_dispatch_log+0xb0/0x170 .account_system_vtime+0xa0/0x220 .irq_enter+0x88/0xc0 .do_IRQ+0x48/0x230 The patch below adds a check to scan_dispatch_log to ensure the dispatch log has been allocated. Signed-off-by: Anton Blanchard Cc: Signed-off-by: Benjamin Herrenschmidt diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c index 375480c..f33acfd 100644 --- a/arch/powerpc/kernel/time.c +++ b/arch/powerpc/kernel/time.c @@ -229,6 +229,9 @@ static u64 scan_dispatch_log(u64 stop_tb) u64 stolen = 0; u64 dtb; + if (!dtl) + return 0; + if (i == vpa->dtl_idx) return 0; while (i < vpa->dtl_idx) { -- cgit v0.10.2 From 09597cfe93d3cc2c6e064a3ead5956b882511560 Mon Sep 17 00:00:00 2001 From: Stefan Roese Date: Thu, 14 Apr 2011 23:49:53 +0000 Subject: powerpc: Don't write protect kernel text with CONFIG_DYNAMIC_FTRACE enabled This problem was noticed on an MPC855T platform. Ftrace did oops when trying to write to the kernel text segment. Many thanks to Joakim for finding the root cause of this problem. Signed-off-by: Stefan Roese Cc: Joakim Tjernlund Cc: Benjamin Herrenschmidt Cc: Steven Rostedt Signed-off-by: Benjamin Herrenschmidt diff --git a/arch/powerpc/include/asm/pte-common.h b/arch/powerpc/include/asm/pte-common.h index 811f04a..8d1569c 100644 --- a/arch/powerpc/include/asm/pte-common.h +++ b/arch/powerpc/include/asm/pte-common.h @@ -162,7 +162,7 @@ extern unsigned long bad_call_to_PMD_PAGE_SIZE(void); * on platforms where such control is possible. */ #if defined(CONFIG_KGDB) || defined(CONFIG_XMON) || defined(CONFIG_BDI_SWITCH) ||\ - defined(CONFIG_KPROBES) + defined(CONFIG_KPROBES) || defined(CONFIG_DYNAMIC_FTRACE) #define PAGE_KERNEL_TEXT PAGE_KERNEL_X #else #define PAGE_KERNEL_TEXT PAGE_KERNEL_ROX -- cgit v0.10.2 From 86c74ab317c1ef4d37325e0d7ca8a01a796b0bd7 Mon Sep 17 00:00:00 2001 From: Eric B Munson Date: Fri, 15 Apr 2011 08:12:30 +0000 Subject: powerpc/perf_event: Skip updating kernel counters if register value shrinks Because of speculative event roll back, it is possible for some event coutners to decrease between reads on POWER7. This causes a problem with the way that counters are updated. Delta calues are calculated in a 64 bit value and the top 32 bits are masked. If the register value has decreased, this leaves us with a very large positive value added to the kernel counters. This patch protects against this by skipping the update if the delta would be negative. This can lead to a lack of precision in the coutner values, but from my testing the value is typcially fewer than 10 samples at a time. Signed-off-by: Eric B Munson Cc: stable@kernel.org Signed-off-by: Benjamin Herrenschmidt diff --git a/arch/powerpc/kernel/perf_event.c b/arch/powerpc/kernel/perf_event.c index c4063b7..822f630 100644 --- a/arch/powerpc/kernel/perf_event.c +++ b/arch/powerpc/kernel/perf_event.c @@ -398,6 +398,25 @@ static int check_excludes(struct perf_event **ctrs, unsigned int cflags[], return 0; } +static u64 check_and_compute_delta(u64 prev, u64 val) +{ + u64 delta = (val - prev) & 0xfffffffful; + + /* + * POWER7 can roll back counter values, if the new value is smaller + * than the previous value it will cause the delta and the counter to + * have bogus values unless we rolled a counter over. If a coutner is + * rolled back, it will be smaller, but within 256, which is the maximum + * number of events to rollback at once. If we dectect a rollback + * return 0. This can lead to a small lack of precision in the + * counters. + */ + if (prev > val && (prev - val) < 256) + delta = 0; + + return delta; +} + static void power_pmu_read(struct perf_event *event) { s64 val, delta, prev; @@ -416,10 +435,11 @@ static void power_pmu_read(struct perf_event *event) prev = local64_read(&event->hw.prev_count); barrier(); val = read_pmc(event->hw.idx); + delta = check_and_compute_delta(prev, val); + if (!delta) + return; } while (local64_cmpxchg(&event->hw.prev_count, prev, val) != prev); - /* The counters are only 32 bits wide */ - delta = (val - prev) & 0xfffffffful; local64_add(delta, &event->count); local64_sub(delta, &event->hw.period_left); } @@ -449,8 +469,9 @@ static void freeze_limited_counters(struct cpu_hw_events *cpuhw, val = (event->hw.idx == 5) ? pmc5 : pmc6; prev = local64_read(&event->hw.prev_count); event->hw.idx = 0; - delta = (val - prev) & 0xfffffffful; - local64_add(delta, &event->count); + delta = check_and_compute_delta(prev, val); + if (delta) + local64_add(delta, &event->count); } } @@ -458,14 +479,16 @@ static void thaw_limited_counters(struct cpu_hw_events *cpuhw, unsigned long pmc5, unsigned long pmc6) { struct perf_event *event; - u64 val; + u64 val, prev; int i; for (i = 0; i < cpuhw->n_limited; ++i) { event = cpuhw->limited_counter[i]; event->hw.idx = cpuhw->limited_hwidx[i]; val = (event->hw.idx == 5) ? pmc5 : pmc6; - local64_set(&event->hw.prev_count, val); + prev = local64_read(&event->hw.prev_count); + if (check_and_compute_delta(prev, val)) + local64_set(&event->hw.prev_count, val); perf_event_update_userpage(event); } } @@ -1197,7 +1220,7 @@ static void record_and_restart(struct perf_event *event, unsigned long val, /* we don't have to worry about interrupts here */ prev = local64_read(&event->hw.prev_count); - delta = (val - prev) & 0xfffffffful; + delta = check_and_compute_delta(prev, val); local64_add(delta, &event->count); /* -- cgit v0.10.2 From 7b84b29b8c2711fe64e0dba4db22f02ce0f16015 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Mon, 18 Apr 2011 15:46:35 +1000 Subject: powerpc/powermac: Build fix with SMP and CPU hotplug Signed-off-by: Benjamin Herrenschmidt diff --git a/arch/powerpc/platforms/powermac/smp.c b/arch/powerpc/platforms/powermac/smp.c index a830c5e..bc5f0dc 100644 --- a/arch/powerpc/platforms/powermac/smp.c +++ b/arch/powerpc/platforms/powermac/smp.c @@ -842,6 +842,7 @@ static void __devinit smp_core99_setup_cpu(int cpu_nr) mpic_setup_this_cpu(); } +#ifdef CONFIG_PPC64 #ifdef CONFIG_HOTPLUG_CPU static int smp_core99_cpu_notify(struct notifier_block *self, unsigned long action, void *hcpu) @@ -879,7 +880,6 @@ static struct notifier_block __cpuinitdata smp_core99_cpu_nb = { static void __init smp_core99_bringup_done(void) { -#ifdef CONFIG_PPC64 extern void g5_phy_disable_cpu1(void); /* Close i2c bus if it was used for tb sync */ @@ -894,14 +894,14 @@ static void __init smp_core99_bringup_done(void) set_cpu_present(1, false); g5_phy_disable_cpu1(); } -#endif /* CONFIG_PPC64 */ - #ifdef CONFIG_HOTPLUG_CPU register_cpu_notifier(&smp_core99_cpu_nb); #endif + if (ppc_md.progress) ppc_md.progress("smp_core99_bringup_done", 0x349); } +#endif /* CONFIG_PPC64 */ #ifdef CONFIG_HOTPLUG_CPU @@ -975,7 +975,9 @@ static void pmac_cpu_die(void) struct smp_ops_t core99_smp_ops = { .message_pass = smp_mpic_message_pass, .probe = smp_core99_probe, +#ifdef CONFIG_PPC64 .bringup_done = smp_core99_bringup_done, +#endif .kick_cpu = smp_core99_kick_cpu, .setup_cpu = smp_core99_setup_cpu, .give_timebase = smp_core99_give_timebase, -- cgit v0.10.2 From 0553c891fabd287726b41076cfd03fe7e5ab596f Mon Sep 17 00:00:00 2001 From: Krishna Kumar Date: Thu, 14 Apr 2011 06:07:04 +0000 Subject: ip6_pol_route panic: Do not allow VLAN on loopback Several tests in the ipv6 routing code check IFF_LOOPBACK, and allowing stacking such as VLAN'ing on top of loopback results in a netdevice which reports IFF_LOOPBACK but really isn't the loopback device. Instead of spamming the ipv6 routing code with even more special tests, simply disallow VLAN over loopback. The result of this patch is: # modprobe 8021q # vconfig add lo 43 ERROR: trying to add VLAN #43 to IF -:lo:- error: Operation not supported Signed-off-by: Krishna Kumar Signed-off-by: David S. Miller diff --git a/drivers/net/loopback.c b/drivers/net/loopback.c index ea0dc45..d70fb76 100644 --- a/drivers/net/loopback.c +++ b/drivers/net/loopback.c @@ -173,7 +173,8 @@ static void loopback_setup(struct net_device *dev) | NETIF_F_RXCSUM | NETIF_F_HIGHDMA | NETIF_F_LLTX - | NETIF_F_NETNS_LOCAL; + | NETIF_F_NETNS_LOCAL + | NETIF_F_VLAN_CHALLENGED; dev->ethtool_ops = &loopback_ethtool_ops; dev->header_ops = ð_header_ops; dev->netdev_ops = &loopback_ops; -- cgit v0.10.2 From 048c9374a749a27f16493cea033fa4a8ff492356 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 18 Apr 2011 09:52:22 +0200 Subject: block: Enhance new plugging support to support general callbacks md/raid requires an unplug callback, but as it does not uses requests the current code cannot provide one. So allow arbitrary callbacks to be attached to the blk_plug. Signed-off-by: NeilBrown Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 78b7b0c..77edf05 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2638,6 +2638,7 @@ void blk_start_plug(struct blk_plug *plug) plug->magic = PLUG_MAGIC; INIT_LIST_HEAD(&plug->list); + INIT_LIST_HEAD(&plug->cb_list); plug->should_sort = 0; /* @@ -2678,6 +2679,24 @@ static void queue_unplugged(struct request_queue *q, unsigned int depth, q->unplugged_fn(q); } +static void flush_plug_callbacks(struct blk_plug *plug) +{ + LIST_HEAD(callbacks); + + if (list_empty(&plug->cb_list)) + return; + + list_splice_init(&plug->cb_list, &callbacks); + + while (!list_empty(&callbacks)) { + struct blk_plug_cb *cb = list_first_entry(&callbacks, + struct blk_plug_cb, + list); + list_del(&cb->list); + cb->callback(cb); + } +} + void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) { struct request_queue *q; @@ -2688,6 +2707,7 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) BUG_ON(plug->magic != PLUG_MAGIC); + flush_plug_callbacks(plug); if (list_empty(&plug->list)) return; diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index ec0357d..f3f7879 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -860,8 +860,13 @@ extern void blk_put_queue(struct request_queue *); struct blk_plug { unsigned long magic; struct list_head list; + struct list_head cb_list; unsigned int should_sort; }; +struct blk_plug_cb { + struct list_head list; + void (*callback)(struct blk_plug_cb *); +}; extern void blk_start_plug(struct blk_plug *); extern void blk_finish_plug(struct blk_plug *); @@ -887,7 +892,7 @@ static inline bool blk_needs_flush_plug(struct task_struct *tsk) { struct blk_plug *plug = tsk->plug; - return plug && !list_empty(&plug->list); + return plug && (!list_empty(&plug->list) || !list_empty(&plug->cb_list)); } /* -- cgit v0.10.2 From b4cb290e0a7d19235bd075c2ad4d60dbab0bac15 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 18 Apr 2011 09:54:05 +0200 Subject: Revert "block: add callback function for unplug notification" MD can't use this since it really requires us to be able to keep more than a single piece of state for the unplug. Commit 048c9374 added the required support for MD, so get rid of this now unused code. This reverts commit f75664570d8b75469cc468f23c2b27220984983b. Conflicts: block/blk-core.c Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 77edf05..09b2628 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2674,9 +2674,6 @@ static void queue_unplugged(struct request_queue *q, unsigned int depth, { trace_block_unplug(q, depth, !from_schedule); __blk_run_queue(q, from_schedule); - - if (q->unplugged_fn) - q->unplugged_fn(q); } static void flush_plug_callbacks(struct blk_plug *plug) diff --git a/block/blk-settings.c b/block/blk-settings.c index eb94904..1fa7692 100644 --- a/block/blk-settings.c +++ b/block/blk-settings.c @@ -790,22 +790,6 @@ void blk_queue_flush(struct request_queue *q, unsigned int flush) } EXPORT_SYMBOL_GPL(blk_queue_flush); -/** - * blk_queue_unplugged - register a callback for an unplug event - * @q: the request queue for the device - * @fn: the function to call - * - * Some stacked drivers may need to know when IO is dispatched on an - * unplug event. By registrering a callback here, they will be notified - * when someone flushes their on-stack queue plug. The function will be - * called with the queue lock held. - */ -void blk_queue_unplugged(struct request_queue *q, unplugged_fn *fn) -{ - q->unplugged_fn = fn; -} -EXPORT_SYMBOL(blk_queue_unplugged); - static int __init blk_settings_init(void) { blk_max_low_pfn = max_low_pfn - 1; diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index f3f7879..3448d89 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -196,7 +196,6 @@ typedef void (request_fn_proc) (struct request_queue *q); typedef int (make_request_fn) (struct request_queue *q, struct bio *bio); typedef int (prep_rq_fn) (struct request_queue *, struct request *); typedef void (unprep_rq_fn) (struct request_queue *, struct request *); -typedef void (unplugged_fn) (struct request_queue *); struct bio_vec; struct bvec_merge_data { @@ -284,7 +283,6 @@ struct request_queue rq_timed_out_fn *rq_timed_out_fn; dma_drain_needed_fn *dma_drain_needed; lld_busy_fn *lld_busy_fn; - unplugged_fn *unplugged_fn; /* * Dispatch queue sorting @@ -843,7 +841,6 @@ extern void blk_queue_dma_alignment(struct request_queue *, int); extern void blk_queue_update_dma_alignment(struct request_queue *, int); extern void blk_queue_softirq_done(struct request_queue *, softirq_done_fn *); extern void blk_queue_rq_timed_out(struct request_queue *, rq_timed_out_fn *); -extern void blk_queue_unplugged(struct request_queue *, unplugged_fn *); extern void blk_queue_rq_timeout(struct request_queue *, unsigned int); extern void blk_queue_flush(struct request_queue *q, unsigned int flush); extern struct backing_dev_info *blk_get_backing_dev_info(struct block_device *bdev); -- cgit v0.10.2 From 99e22598e9a8e0a996d69c8c0f6b7027cb57720a Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 18 Apr 2011 09:59:55 +0200 Subject: block: drop queue lock before calling __blk_run_queue() for kblockd punt If we know we are going to punt to kblockd, we can drop the queue lock before calling into __blk_run_queue() since it only does a safe bit test and a workqueue call. Since kblockd needs to grab this very lock as one of the first things it does, it's a good optimization to drop the lock before waking kblockd. Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 09b2628..5e41393 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -295,7 +295,8 @@ EXPORT_SYMBOL(blk_sync_queue); * * Description: * See @blk_run_queue. This variant must be called with the queue lock - * held and interrupts disabled. + * held and interrupts disabled. If force_kblockd is true, then it is + * safe to call this without holding the queue lock. * */ void __blk_run_queue(struct request_queue *q, bool force_kblockd) @@ -2671,9 +2672,23 @@ static int plug_rq_cmp(void *priv, struct list_head *a, struct list_head *b) */ static void queue_unplugged(struct request_queue *q, unsigned int depth, bool from_schedule) + __releases(q->queue_lock) { trace_block_unplug(q, depth, !from_schedule); - __blk_run_queue(q, from_schedule); + + /* + * If we are punting this to kblockd, then we can safely drop + * the queue_lock before waking kblockd (which needs to take + * this lock). + */ + if (from_schedule) { + spin_unlock(q->queue_lock); + __blk_run_queue(q, true); + } else { + __blk_run_queue(q, false); + spin_unlock(q->queue_lock); + } + } static void flush_plug_callbacks(struct blk_plug *plug) @@ -2729,10 +2744,11 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) BUG_ON(!(rq->cmd_flags & REQ_ON_PLUG)); BUG_ON(!rq->q); if (rq->q != q) { - if (q) { + /* + * This drops the queue lock + */ + if (q) queue_unplugged(q, depth, from_schedule); - spin_unlock(q->queue_lock); - } q = rq->q; depth = 0; spin_lock(q->queue_lock); @@ -2750,10 +2766,11 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) depth++; } - if (q) { + /* + * This drops the queue lock + */ + if (q) queue_unplugged(q, depth, from_schedule); - spin_unlock(q->queue_lock); - } local_irq_restore(flags); } -- cgit v0.10.2 From e1dfa0a29737142c32f00a3bac0f609dc85b4a82 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 18 Apr 2011 18:25:41 +1000 Subject: md: use new plugging interface for RAID IO. md/raid submits a lot of IO from the various raid threads. So adding start/finish plug calls to those so that some plugging happens. Signed-off-by: NeilBrown diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c index c2a21ae5..5449cd5 100644 --- a/drivers/md/raid1.c +++ b/drivers/md/raid1.c @@ -1516,9 +1516,11 @@ static void raid1d(mddev_t *mddev) conf_t *conf = mddev->private; struct list_head *head = &conf->retry_list; mdk_rdev_t *rdev; + struct blk_plug plug; md_check_recovery(mddev); - + + blk_start_plug(&plug); for (;;) { char b[BDEVNAME_SIZE]; @@ -1593,6 +1595,7 @@ static void raid1d(mddev_t *mddev) } cond_resched(); } + blk_finish_plug(&plug); } diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index 2da83d5..c8e5dac 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -1640,9 +1640,11 @@ static void raid10d(mddev_t *mddev) conf_t *conf = mddev->private; struct list_head *head = &conf->retry_list; mdk_rdev_t *rdev; + struct blk_plug plug; md_check_recovery(mddev); + blk_start_plug(&plug); for (;;) { char b[BDEVNAME_SIZE]; @@ -1716,6 +1718,7 @@ static void raid10d(mddev_t *mddev) } cond_resched(); } + blk_finish_plug(&plug); } diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index e867ee4..ce6960b 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -4478,11 +4478,13 @@ static void raid5d(mddev_t *mddev) struct stripe_head *sh; raid5_conf_t *conf = mddev->private; int handled; + struct blk_plug plug; pr_debug("+++ raid5d active\n"); md_check_recovery(mddev); + blk_start_plug(&plug); handled = 0; spin_lock_irq(&conf->device_lock); while (1) { @@ -4525,6 +4527,7 @@ static void raid5d(mddev_t *mddev) spin_unlock_irq(&conf->device_lock); async_tx_issue_pending_all(); + blk_finish_plug(&plug); pr_debug("--- raid5d inactive\n"); } -- cgit v0.10.2 From af1db72d8b340f97ad12b60175afdef43e6f0e60 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 18 Apr 2011 18:25:41 +1000 Subject: md/dm - remove remains of plug_fn callback. Now that unplugging is done differently, the unplug_fn callback is never called, so it can be completely discarded. Signed-off-by: NeilBrown diff --git a/drivers/md/dm-raid.c b/drivers/md/dm-raid.c index 5ef136c..e5d8904 100644 --- a/drivers/md/dm-raid.c +++ b/drivers/md/dm-raid.c @@ -390,13 +390,6 @@ static int raid_is_congested(struct dm_target_callbacks *cb, int bits) return md_raid5_congested(&rs->md, bits); } -static void raid_unplug(struct dm_target_callbacks *cb) -{ - struct raid_set *rs = container_of(cb, struct raid_set, callbacks); - - md_raid5_kick_device(rs->md.private); -} - /* * Construct a RAID4/5/6 mapping: * Args: @@ -487,7 +480,6 @@ static int raid_ctr(struct dm_target *ti, unsigned argc, char **argv) } rs->callbacks.congested_fn = raid_is_congested; - rs->callbacks.unplug_fn = raid_unplug; dm_table_add_target_callbacks(ti->table, &rs->callbacks); return 0; diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h index e276883..32a4423 100644 --- a/include/linux/device-mapper.h +++ b/include/linux/device-mapper.h @@ -197,7 +197,6 @@ struct dm_target { struct dm_target_callbacks { struct list_head list; int (*congested_fn) (struct dm_target_callbacks *, int); - void (*unplug_fn)(struct dm_target_callbacks *); }; int dm_register_target(struct target_type *t); -- cgit v0.10.2 From 482c083492ddaa32ef5864bae3d143dc8bcdf7d1 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 18 Apr 2011 18:25:42 +1000 Subject: md - remove old plugging code. md has some plugging infrastructure for RAID5 to use because the normal plugging infrastructure required a 'request_queue', and when called from dm, RAID5 doesn't have one of those available. This relied on the ->unplug_fn callback which doesn't exist any more. So remove all of that code, both in md and raid5. Subsequent patches with restore the plugging functionality. Signed-off-by: NeilBrown diff --git a/drivers/md/md.c b/drivers/md/md.c index b12b377..fb11170 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -445,50 +445,6 @@ void md_flush_request(mddev_t *mddev, struct bio *bio) } EXPORT_SYMBOL(md_flush_request); -/* Support for plugging. - * This mirrors the plugging support in request_queue, but does not - * require having a whole queue - */ -static void plugger_work(struct work_struct *work) -{ - struct plug_handle *plug = - container_of(work, struct plug_handle, unplug_work); - plug->unplug_fn(plug); -} -static void plugger_timeout(unsigned long data) -{ - struct plug_handle *plug = (void *)data; - kblockd_schedule_work(NULL, &plug->unplug_work); -} -void plugger_init(struct plug_handle *plug, - void (*unplug_fn)(struct plug_handle *)) -{ - plug->unplug_flag = 0; - plug->unplug_fn = unplug_fn; - init_timer(&plug->unplug_timer); - plug->unplug_timer.function = plugger_timeout; - plug->unplug_timer.data = (unsigned long)plug; - INIT_WORK(&plug->unplug_work, plugger_work); -} -EXPORT_SYMBOL_GPL(plugger_init); - -void plugger_set_plug(struct plug_handle *plug) -{ - if (!test_and_set_bit(PLUGGED_FLAG, &plug->unplug_flag)) - mod_timer(&plug->unplug_timer, jiffies + msecs_to_jiffies(3)+1); -} -EXPORT_SYMBOL_GPL(plugger_set_plug); - -int plugger_remove_plug(struct plug_handle *plug) -{ - if (test_and_clear_bit(PLUGGED_FLAG, &plug->unplug_flag)) { - del_timer(&plug->unplug_timer); - return 1; - } else - return 0; -} -EXPORT_SYMBOL_GPL(plugger_remove_plug); - static inline mddev_t *mddev_get(mddev_t *mddev) { @@ -4723,7 +4679,6 @@ static void md_clean(mddev_t *mddev) mddev->bitmap_info.chunksize = 0; mddev->bitmap_info.daemon_sleep = 0; mddev->bitmap_info.max_write_behind = 0; - mddev->plug = NULL; } static void __md_stop_writes(mddev_t *mddev) @@ -6688,12 +6643,6 @@ int md_allow_write(mddev_t *mddev) } EXPORT_SYMBOL_GPL(md_allow_write); -void md_unplug(mddev_t *mddev) -{ - if (mddev->plug) - mddev->plug->unplug_fn(mddev->plug); -} - #define SYNC_MARKS 10 #define SYNC_MARK_STEP (3*HZ) void md_do_sync(mddev_t *mddev) diff --git a/drivers/md/md.h b/drivers/md/md.h index 52b4073..fad9022 100644 --- a/drivers/md/md.h +++ b/drivers/md/md.h @@ -29,26 +29,6 @@ typedef struct mddev_s mddev_t; typedef struct mdk_rdev_s mdk_rdev_t; -/* generic plugging support - like that provided with request_queue, - * but does not require a request_queue - */ -struct plug_handle { - void (*unplug_fn)(struct plug_handle *); - struct timer_list unplug_timer; - struct work_struct unplug_work; - unsigned long unplug_flag; -}; -#define PLUGGED_FLAG 1 -void plugger_init(struct plug_handle *plug, - void (*unplug_fn)(struct plug_handle *)); -void plugger_set_plug(struct plug_handle *plug); -int plugger_remove_plug(struct plug_handle *plug); -static inline void plugger_flush(struct plug_handle *plug) -{ - del_timer_sync(&plug->unplug_timer); - cancel_work_sync(&plug->unplug_work); -} - /* * MD's 'extended' device */ @@ -336,7 +316,6 @@ struct mddev_s struct list_head all_mddevs; struct attribute_group *to_remove; - struct plug_handle *plug; /* if used by personality */ struct bio_set *bio_set; @@ -516,7 +495,6 @@ extern int md_integrity_register(mddev_t *mddev); extern void md_integrity_add_rdev(mdk_rdev_t *rdev, mddev_t *mddev); extern int strict_strtoul_scaled(const char *cp, unsigned long *res, int scale); extern void restore_bitmap_write_access(struct file *file); -extern void md_unplug(mddev_t *mddev); extern void mddev_init(mddev_t *mddev); extern int md_run(mddev_t *mddev); diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index ce6960b..a1755a6 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -199,14 +199,12 @@ static void __release_stripe(raid5_conf_t *conf, struct stripe_head *sh) BUG_ON(!list_empty(&sh->lru)); BUG_ON(atomic_read(&conf->active_stripes)==0); if (test_bit(STRIPE_HANDLE, &sh->state)) { - if (test_bit(STRIPE_DELAYED, &sh->state)) { + if (test_bit(STRIPE_DELAYED, &sh->state)) list_add_tail(&sh->lru, &conf->delayed_list); - plugger_set_plug(&conf->plug); - } else if (test_bit(STRIPE_BIT_DELAY, &sh->state) && - sh->bm_seq - conf->seq_write > 0) { + else if (test_bit(STRIPE_BIT_DELAY, &sh->state) && + sh->bm_seq - conf->seq_write > 0) list_add_tail(&sh->lru, &conf->bitmap_list); - plugger_set_plug(&conf->plug); - } else { + else { clear_bit(STRIPE_BIT_DELAY, &sh->state); list_add_tail(&sh->lru, &conf->handle_list); } @@ -461,7 +459,7 @@ get_active_stripe(raid5_conf_t *conf, sector_t sector, < (conf->max_nr_stripes *3/4) || !conf->inactive_blocked), conf->device_lock, - md_raid5_kick_device(conf)); + md_wakeup_thread(conf->mddev->thread)); conf->inactive_blocked = 0; } else init_stripe(sh, sector, previous); @@ -1470,7 +1468,7 @@ static int resize_stripes(raid5_conf_t *conf, int newsize) wait_event_lock_irq(conf->wait_for_stripe, !list_empty(&conf->inactive_list), conf->device_lock, - blk_flush_plug(current)); + ); osh = get_free_stripe(conf); spin_unlock_irq(&conf->device_lock); atomic_set(&nsh->count, 1); @@ -3623,8 +3621,7 @@ static void raid5_activate_delayed(raid5_conf_t *conf) atomic_inc(&conf->preread_active_stripes); list_add_tail(&sh->lru, &conf->hold_list); } - } else - plugger_set_plug(&conf->plug); + } } static void activate_bit_delay(raid5_conf_t *conf) @@ -3641,21 +3638,6 @@ static void activate_bit_delay(raid5_conf_t *conf) } } -void md_raid5_kick_device(raid5_conf_t *conf) -{ - blk_flush_plug(current); - raid5_activate_delayed(conf); - md_wakeup_thread(conf->mddev->thread); -} -EXPORT_SYMBOL_GPL(md_raid5_kick_device); - -static void raid5_unplug(struct plug_handle *plug) -{ - raid5_conf_t *conf = container_of(plug, raid5_conf_t, plug); - - md_raid5_kick_device(conf); -} - int md_raid5_congested(mddev_t *mddev, int bits) { raid5_conf_t *conf = mddev->private; @@ -4057,7 +4039,7 @@ static int make_request(mddev_t *mddev, struct bio * bi) * add failed due to overlap. Flush everything * and wait a while */ - md_raid5_kick_device(conf); + md_wakeup_thread(mddev->thread); release_stripe(sh); schedule(); goto retry; @@ -5144,8 +5126,6 @@ static int run(mddev_t *mddev) mdname(mddev)); md_set_array_sectors(mddev, raid5_size(mddev, 0, 0)); - plugger_init(&conf->plug, raid5_unplug); - mddev->plug = &conf->plug; if (mddev->queue) { int chunk_size; /* read-ahead size must cover two whole stripes, which @@ -5195,7 +5175,6 @@ static int stop(mddev_t *mddev) mddev->thread = NULL; if (mddev->queue) mddev->queue->backing_dev_info.congested_fn = NULL; - plugger_flush(&conf->plug); /* the unplug fn references 'conf'*/ free_conf(conf); mddev->private = NULL; mddev->to_remove = &raid5_attrs_group; diff --git a/drivers/md/raid5.h b/drivers/md/raid5.h index 8d563a4..3ca77a2 100644 --- a/drivers/md/raid5.h +++ b/drivers/md/raid5.h @@ -400,8 +400,6 @@ struct raid5_private_data { * Cleared when a sync completes. */ - struct plug_handle plug; - /* per cpu variables */ struct raid5_percpu { struct page *spare_page; /* Used when checking P/Q in raid6 */ -- cgit v0.10.2 From 97658cdd3af7d01461874c93b89afa4a2465e7c6 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 18 Apr 2011 18:25:42 +1000 Subject: md: provide generic support for handling unplug callbacks. When an md device adds a request to a queue, it can call mddev_check_plugged. If this succeeds then we know that the md thread will be woken up shortly, and ->plug_cnt will be non-zero until then, so some processing can be delayed. If it fails, then no unplug callback is expected and the make_request function needs to do whatever is required to make the request happen. Signed-off-by: NeilBrown diff --git a/drivers/md/md.c b/drivers/md/md.c index fb11170..6e853c6 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -445,6 +445,61 @@ void md_flush_request(mddev_t *mddev, struct bio *bio) } EXPORT_SYMBOL(md_flush_request); +/* Support for plugging. + * This mirrors the plugging support in request_queue, but does not + * require having a whole queue or request structures. + * We allocate an md_plug_cb for each md device and each thread it gets + * plugged on. This links tot the private plug_handle structure in the + * personality data where we keep a count of the number of outstanding + * plugs so other code can see if a plug is active. + */ +struct md_plug_cb { + struct blk_plug_cb cb; + mddev_t *mddev; +}; + +static void plugger_unplug(struct blk_plug_cb *cb) +{ + struct md_plug_cb *mdcb = container_of(cb, struct md_plug_cb, cb); + if (atomic_dec_and_test(&mdcb->mddev->plug_cnt)) + md_wakeup_thread(mdcb->mddev->thread); + kfree(mdcb); +} + +/* Check that an unplug wakeup will come shortly. + * If not, wakeup the md thread immediately + */ +int mddev_check_plugged(mddev_t *mddev) +{ + struct blk_plug *plug = current->plug; + struct md_plug_cb *mdcb; + + if (!plug) + return 0; + + list_for_each_entry(mdcb, &plug->cb_list, cb.list) { + if (mdcb->cb.callback == plugger_unplug && + mdcb->mddev == mddev) { + /* Already on the list, move to top */ + if (mdcb != list_first_entry(&plug->cb_list, + struct md_plug_cb, + cb.list)) + list_move(&mdcb->cb.list, &plug->cb_list); + return 1; + } + } + /* Not currently on the callback list */ + mdcb = kmalloc(sizeof(*mdcb), GFP_ATOMIC); + if (!mdcb) + return 0; + + mdcb->mddev = mddev; + mdcb->cb.callback = plugger_unplug; + atomic_inc(&mddev->plug_cnt); + list_add(&mdcb->cb.list, &plug->cb_list); + return 1; +} +EXPORT_SYMBOL_GPL(mddev_check_plugged); static inline mddev_t *mddev_get(mddev_t *mddev) { @@ -494,6 +549,7 @@ void mddev_init(mddev_t *mddev) atomic_set(&mddev->active, 1); atomic_set(&mddev->openers, 0); atomic_set(&mddev->active_io, 0); + atomic_set(&mddev->plug_cnt, 0); spin_lock_init(&mddev->write_lock); atomic_set(&mddev->flush_pending, 0); init_waitqueue_head(&mddev->sb_wait); diff --git a/drivers/md/md.h b/drivers/md/md.h index fad9022..0b1fd3f 100644 --- a/drivers/md/md.h +++ b/drivers/md/md.h @@ -179,6 +179,9 @@ struct mddev_s int delta_disks, new_level, new_layout; int new_chunk_sectors; + atomic_t plug_cnt; /* If device is expecting + * more bios soon. + */ struct mdk_thread_s *thread; /* management thread */ struct mdk_thread_s *sync_thread; /* doing resync or reconstruct */ sector_t curr_resync; /* last block scheduled */ @@ -508,4 +511,5 @@ extern struct bio *bio_clone_mddev(struct bio *bio, gfp_t gfp_mask, mddev_t *mddev); extern struct bio *bio_alloc_mddev(gfp_t gfp_mask, int nr_iovecs, mddev_t *mddev); +extern int mddev_check_plugged(mddev_t *mddev); #endif /* _MD_MD_H */ -- cgit v0.10.2 From 7c13edc87510f665da3094174e1fd633e06649f4 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 18 Apr 2011 18:25:43 +1000 Subject: md: incorporate new plugging into raid5. In raid5 plugging is used for 2 things: 1/ collecting writes that require a bitmap update 2/ collecting writes in the hope that we can create full stripes - or at least more-full. We now release these different sets of stripes when plug_cnt is zero. Also in make_request, we call mddev_check_plug to hopefully increase plug_cnt, and wake up the thread at the end if plugging wasn't achieved for some reason. Signed-off-by: NeilBrown diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index a1755a6..f301e6a 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -27,12 +27,12 @@ * * We group bitmap updates into batches. Each batch has a number. * We may write out several batches at once, but that isn't very important. - * conf->bm_write is the number of the last batch successfully written. - * conf->bm_flush is the number of the last batch that was closed to + * conf->seq_write is the number of the last batch successfully written. + * conf->seq_flush is the number of the last batch that was closed to * new additions. * When we discover that we will need to write to any block in a stripe * (in add_stripe_bio) we update the in-memory bitmap and record in sh->bm_seq - * the number of the batch it will be in. This is bm_flush+1. + * the number of the batch it will be in. This is seq_flush+1. * When we are ready to do a write, if that batch hasn't been written yet, * we plug the array and queue the stripe for later. * When an unplug happens, we increment bm_flush, thus closing the current @@ -459,7 +459,7 @@ get_active_stripe(raid5_conf_t *conf, sector_t sector, < (conf->max_nr_stripes *3/4) || !conf->inactive_blocked), conf->device_lock, - md_wakeup_thread(conf->mddev->thread)); + ); conf->inactive_blocked = 0; } else init_stripe(sh, sector, previous); @@ -3927,6 +3927,7 @@ static int make_request(mddev_t *mddev, struct bio * bi) struct stripe_head *sh; const int rw = bio_data_dir(bi); int remaining; + int plugged; if (unlikely(bi->bi_rw & REQ_FLUSH)) { md_flush_request(mddev, bi); @@ -3945,6 +3946,7 @@ static int make_request(mddev_t *mddev, struct bio * bi) bi->bi_next = NULL; bi->bi_phys_segments = 1; /* over-loaded to count active stripes */ + plugged = mddev_check_plugged(mddev); for (;logical_sector < last_sector; logical_sector += STRIPE_SECTORS) { DEFINE_WAIT(w); int disks, data_disks; @@ -4059,6 +4061,9 @@ static int make_request(mddev_t *mddev, struct bio * bi) } } + if (!plugged) + md_wakeup_thread(mddev->thread); + spin_lock_irq(&conf->device_lock); remaining = raid5_dec_bi_phys_segments(bi); spin_unlock_irq(&conf->device_lock); @@ -4472,14 +4477,18 @@ static void raid5d(mddev_t *mddev) while (1) { struct bio *bio; - if (conf->seq_flush != conf->seq_write) { - int seq = conf->seq_flush; + if (atomic_read(&mddev->plug_cnt) == 0 && + !list_empty(&conf->bitmap_list)) { + /* Now is a good time to flush some bitmap updates */ + conf->seq_flush++; spin_unlock_irq(&conf->device_lock); bitmap_unplug(mddev->bitmap); spin_lock_irq(&conf->device_lock); - conf->seq_write = seq; + conf->seq_write = conf->seq_flush; activate_bit_delay(conf); } + if (atomic_read(&mddev->plug_cnt) == 0) + raid5_activate_delayed(conf); while ((bio = remove_bio_from_retry(conf))) { int ok; -- cgit v0.10.2 From c3b328ac846bcf6b9a62c5563380a81ab723006d Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 18 Apr 2011 18:25:43 +1000 Subject: md: fix up raid1/raid10 unplugging. We just need to make sure that an unplug event wakes up the md thread, which is exactly what mddev_check_plugged does. Also remove some plug-related code that is no longer needed. Signed-off-by: NeilBrown diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c index 5449cd5..2b7a7ff 100644 --- a/drivers/md/raid1.c +++ b/drivers/md/raid1.c @@ -565,12 +565,6 @@ static void flush_pending_writes(conf_t *conf) spin_unlock_irq(&conf->device_lock); } -static void md_kick_device(mddev_t *mddev) -{ - blk_flush_plug(current); - md_wakeup_thread(mddev->thread); -} - /* Barriers.... * Sometimes we need to suspend IO while we do something else, * either some resync/recovery, or reconfigure the array. @@ -600,7 +594,7 @@ static void raise_barrier(conf_t *conf) /* Wait until no block IO is waiting */ wait_event_lock_irq(conf->wait_barrier, !conf->nr_waiting, - conf->resync_lock, md_kick_device(conf->mddev)); + conf->resync_lock, ); /* block any new IO from starting */ conf->barrier++; @@ -608,7 +602,7 @@ static void raise_barrier(conf_t *conf) /* Now wait for all pending IO to complete */ wait_event_lock_irq(conf->wait_barrier, !conf->nr_pending && conf->barrier < RESYNC_DEPTH, - conf->resync_lock, md_kick_device(conf->mddev)); + conf->resync_lock, ); spin_unlock_irq(&conf->resync_lock); } @@ -630,7 +624,7 @@ static void wait_barrier(conf_t *conf) conf->nr_waiting++; wait_event_lock_irq(conf->wait_barrier, !conf->barrier, conf->resync_lock, - md_kick_device(conf->mddev)); + ); conf->nr_waiting--; } conf->nr_pending++; @@ -666,8 +660,7 @@ static void freeze_array(conf_t *conf) wait_event_lock_irq(conf->wait_barrier, conf->nr_pending == conf->nr_queued+1, conf->resync_lock, - ({ flush_pending_writes(conf); - md_kick_device(conf->mddev); })); + flush_pending_writes(conf)); spin_unlock_irq(&conf->resync_lock); } static void unfreeze_array(conf_t *conf) @@ -729,6 +722,7 @@ static int make_request(mddev_t *mddev, struct bio * bio) const unsigned long do_sync = (bio->bi_rw & REQ_SYNC); const unsigned long do_flush_fua = (bio->bi_rw & (REQ_FLUSH | REQ_FUA)); mdk_rdev_t *blocked_rdev; + int plugged; /* * Register the new request and wait if the reconstruction @@ -820,6 +814,8 @@ static int make_request(mddev_t *mddev, struct bio * bio) * inc refcount on their rdev. Record them by setting * bios[x] to bio */ + plugged = mddev_check_plugged(mddev); + disks = conf->raid_disks; retry_write: blocked_rdev = NULL; @@ -925,7 +921,7 @@ static int make_request(mddev_t *mddev, struct bio * bio) /* In case raid1d snuck in to freeze_array */ wake_up(&conf->wait_barrier); - if (do_sync || !bitmap) + if (do_sync || !bitmap || !plugged) md_wakeup_thread(mddev->thread); return 0; @@ -1524,7 +1520,8 @@ static void raid1d(mddev_t *mddev) for (;;) { char b[BDEVNAME_SIZE]; - flush_pending_writes(conf); + if (atomic_read(&mddev->plug_cnt) == 0) + flush_pending_writes(conf); spin_lock_irqsave(&conf->device_lock, flags); if (list_empty(head)) { @@ -2042,7 +2039,6 @@ static int stop(mddev_t *mddev) md_unregister_thread(mddev->thread); mddev->thread = NULL; - blk_sync_queue(mddev->queue); /* the unplug fn references 'conf'*/ if (conf->r1bio_pool) mempool_destroy(conf->r1bio_pool); kfree(conf->mirrors); diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c index c8e5dac..8e94626 100644 --- a/drivers/md/raid10.c +++ b/drivers/md/raid10.c @@ -634,12 +634,6 @@ static void flush_pending_writes(conf_t *conf) spin_unlock_irq(&conf->device_lock); } -static void md_kick_device(mddev_t *mddev) -{ - blk_flush_plug(current); - md_wakeup_thread(mddev->thread); -} - /* Barriers.... * Sometimes we need to suspend IO while we do something else, * either some resync/recovery, or reconfigure the array. @@ -669,15 +663,15 @@ static void raise_barrier(conf_t *conf, int force) /* Wait until no block IO is waiting (unless 'force') */ wait_event_lock_irq(conf->wait_barrier, force || !conf->nr_waiting, - conf->resync_lock, md_kick_device(conf->mddev)); + conf->resync_lock, ); /* block any new IO from starting */ conf->barrier++; - /* No wait for all pending IO to complete */ + /* Now wait for all pending IO to complete */ wait_event_lock_irq(conf->wait_barrier, !conf->nr_pending && conf->barrier < RESYNC_DEPTH, - conf->resync_lock, md_kick_device(conf->mddev)); + conf->resync_lock, ); spin_unlock_irq(&conf->resync_lock); } @@ -698,7 +692,7 @@ static void wait_barrier(conf_t *conf) conf->nr_waiting++; wait_event_lock_irq(conf->wait_barrier, !conf->barrier, conf->resync_lock, - md_kick_device(conf->mddev)); + ); conf->nr_waiting--; } conf->nr_pending++; @@ -734,8 +728,8 @@ static void freeze_array(conf_t *conf) wait_event_lock_irq(conf->wait_barrier, conf->nr_pending == conf->nr_queued+1, conf->resync_lock, - ({ flush_pending_writes(conf); - md_kick_device(conf->mddev); })); + flush_pending_writes(conf)); + spin_unlock_irq(&conf->resync_lock); } @@ -762,6 +756,7 @@ static int make_request(mddev_t *mddev, struct bio * bio) const unsigned long do_fua = (bio->bi_rw & REQ_FUA); unsigned long flags; mdk_rdev_t *blocked_rdev; + int plugged; if (unlikely(bio->bi_rw & REQ_FLUSH)) { md_flush_request(mddev, bio); @@ -870,6 +865,8 @@ static int make_request(mddev_t *mddev, struct bio * bio) * inc refcount on their rdev. Record them by setting * bios[x] to bio */ + plugged = mddev_check_plugged(mddev); + raid10_find_phys(conf, r10_bio); retry_write: blocked_rdev = NULL; @@ -946,9 +943,8 @@ static int make_request(mddev_t *mddev, struct bio * bio) /* In case raid10d snuck in to freeze_array */ wake_up(&conf->wait_barrier); - if (do_sync || !mddev->bitmap) + if (do_sync || !mddev->bitmap || !plugged) md_wakeup_thread(mddev->thread); - return 0; } -- cgit v0.10.2 From 1791f881435fab951939ad700e947b66c062e083 Mon Sep 17 00:00:00 2001 From: Richard Cochran Date: Wed, 30 Mar 2011 15:24:21 +0200 Subject: posix clocks: Replace mutex with reader/writer semaphore A dynamic posix clock is protected from asynchronous removal by a mutex. However, using a mutex has the unwanted effect that a long running clock operation in one process will unnecessarily block other processes. For example, one process might call read() to get an external time stamp coming in at one pulse per second. A second process calling clock_gettime would have to wait for almost a whole second. This patch fixes the issue by using a reader/writer semaphore instead of a mutex. Signed-off-by: Richard Cochran Cc: John Stultz Link: http://lkml.kernel.org/r/%3C20110330132421.GA31771%40riccoc20.at.omicron.at%3E Signed-off-by: Thomas Gleixner diff --git a/include/linux/posix-clock.h b/include/linux/posix-clock.h index 369e19d..7f1183d 100644 --- a/include/linux/posix-clock.h +++ b/include/linux/posix-clock.h @@ -24,6 +24,7 @@ #include #include #include +#include struct posix_clock; @@ -104,7 +105,7 @@ struct posix_clock_operations { * @ops: Functional interface to the clock * @cdev: Character device instance for this clock * @kref: Reference count. - * @mutex: Protects the 'zombie' field from concurrent access. + * @rwsem: Protects the 'zombie' field from concurrent access. * @zombie: If 'zombie' is true, then the hardware has disappeared. * @release: A function to free the structure when the reference count reaches * zero. May be NULL if structure is statically allocated. @@ -117,7 +118,7 @@ struct posix_clock { struct posix_clock_operations ops; struct cdev cdev; struct kref kref; - struct mutex mutex; + struct rw_semaphore rwsem; bool zombie; void (*release)(struct posix_clock *clk); }; diff --git a/kernel/time/posix-clock.c b/kernel/time/posix-clock.c index 25028dd..c340ca6 100644 --- a/kernel/time/posix-clock.c +++ b/kernel/time/posix-clock.c @@ -19,7 +19,6 @@ */ #include #include -#include #include #include #include @@ -34,19 +33,19 @@ static struct posix_clock *get_posix_clock(struct file *fp) { struct posix_clock *clk = fp->private_data; - mutex_lock(&clk->mutex); + down_read(&clk->rwsem); if (!clk->zombie) return clk; - mutex_unlock(&clk->mutex); + up_read(&clk->rwsem); return NULL; } static void put_posix_clock(struct posix_clock *clk) { - mutex_unlock(&clk->mutex); + up_read(&clk->rwsem); } static ssize_t posix_clock_read(struct file *fp, char __user *buf, @@ -156,7 +155,7 @@ static int posix_clock_open(struct inode *inode, struct file *fp) struct posix_clock *clk = container_of(inode->i_cdev, struct posix_clock, cdev); - mutex_lock(&clk->mutex); + down_read(&clk->rwsem); if (clk->zombie) { err = -ENODEV; @@ -172,7 +171,7 @@ static int posix_clock_open(struct inode *inode, struct file *fp) fp->private_data = clk; } out: - mutex_unlock(&clk->mutex); + up_read(&clk->rwsem); return err; } @@ -211,25 +210,20 @@ int posix_clock_register(struct posix_clock *clk, dev_t devid) int err; kref_init(&clk->kref); - mutex_init(&clk->mutex); + init_rwsem(&clk->rwsem); cdev_init(&clk->cdev, &posix_clock_file_operations); clk->cdev.owner = clk->ops.owner; err = cdev_add(&clk->cdev, devid, 1); - if (err) - goto no_cdev; return err; -no_cdev: - mutex_destroy(&clk->mutex); - return err; } EXPORT_SYMBOL_GPL(posix_clock_register); static void delete_clock(struct kref *kref) { struct posix_clock *clk = container_of(kref, struct posix_clock, kref); - mutex_destroy(&clk->mutex); + if (clk->release) clk->release(clk); } @@ -238,9 +232,9 @@ void posix_clock_unregister(struct posix_clock *clk) { cdev_del(&clk->cdev); - mutex_lock(&clk->mutex); + down_write(&clk->rwsem); clk->zombie = true; - mutex_unlock(&clk->mutex); + up_write(&clk->rwsem); kref_put(&clk->kref, delete_clock); } -- cgit v0.10.2 From 2dd93c4f47d506c586b827d75646a4257aafa43e Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sun, 17 Apr 2011 10:02:58 +0800 Subject: RTC: rtc-omap: Fix a leak of the IRQ during init failure In omap_rtc_probe error path, free_irq() was using NULL rather than the driver data as the data pointer so free_irq() wouldn't have matched. Signed-off-by: Axel Lin Cc: "George G. Davis" Cc: Alessandro Zummo Cc: rtc-linux@googlegroups.com Link: http://lkml.kernel.org/r/%3C1303005778.2889.2.camel%40phoenix%3E Signed-off-by: Thomas Gleixner diff --git a/drivers/rtc/rtc-omap.c b/drivers/rtc/rtc-omap.c index de0dd7b..bcae8dd 100644 --- a/drivers/rtc/rtc-omap.c +++ b/drivers/rtc/rtc-omap.c @@ -394,7 +394,7 @@ static int __init omap_rtc_probe(struct platform_device *pdev) return 0; fail2: - free_irq(omap_rtc_timer, NULL); + free_irq(omap_rtc_timer, rtc); fail1: rtc_device_unregister(rtc); fail0: -- cgit v0.10.2 From 4521cc4ed5173f92714f6999a69910c3385fed68 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 18 Apr 2011 11:36:39 +0200 Subject: block: blk_delay_queue() should use kblockd workqueue Reported-by: Christoph Hellwig Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 5e41393..e2bacfa 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -220,7 +220,8 @@ static void blk_delay_work(struct work_struct *work) */ void blk_delay_queue(struct request_queue *q, unsigned long msecs) { - schedule_delayed_work(&q->delay_work, msecs_to_jiffies(msecs)); + queue_delayed_work(kblockd_workqueue, &q->delay_work, + msecs_to_jiffies(msecs)); } EXPORT_SYMBOL(blk_delay_queue); -- cgit v0.10.2 From 24ecfbe27f65563909b14492afda2f1c21f7c044 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Mon, 18 Apr 2011 11:41:33 +0200 Subject: block: add blk_run_queue_async Instead of overloading __blk_run_queue to force an offload to kblockd add a new blk_run_queue_async helper to do it explicitly. I've kept the blk_queue_stopped check for now, but I suspect it's not needed as the check we do when the workqueue items runs should be enough. Signed-off-by: Christoph Hellwig Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index e2bacfa..5fa3dd2 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -204,7 +204,7 @@ static void blk_delay_work(struct work_struct *work) q = container_of(work, struct request_queue, delay_work.work); spin_lock_irq(q->queue_lock); - __blk_run_queue(q, false); + __blk_run_queue(q); spin_unlock_irq(q->queue_lock); } @@ -239,7 +239,7 @@ void blk_start_queue(struct request_queue *q) WARN_ON(!irqs_disabled()); queue_flag_clear(QUEUE_FLAG_STOPPED, q); - __blk_run_queue(q, false); + __blk_run_queue(q); } EXPORT_SYMBOL(blk_start_queue); @@ -296,11 +296,9 @@ EXPORT_SYMBOL(blk_sync_queue); * * Description: * See @blk_run_queue. This variant must be called with the queue lock - * held and interrupts disabled. If force_kblockd is true, then it is - * safe to call this without holding the queue lock. - * + * held and interrupts disabled. */ -void __blk_run_queue(struct request_queue *q, bool force_kblockd) +void __blk_run_queue(struct request_queue *q) { if (unlikely(blk_queue_stopped(q))) return; @@ -309,7 +307,7 @@ void __blk_run_queue(struct request_queue *q, bool force_kblockd) * Only recurse once to avoid overrunning the stack, let the unplug * handling reinvoke the handler shortly if we already got there. */ - if (!force_kblockd && !queue_flag_test_and_set(QUEUE_FLAG_REENTER, q)) { + if (!queue_flag_test_and_set(QUEUE_FLAG_REENTER, q)) { q->request_fn(q); queue_flag_clear(QUEUE_FLAG_REENTER, q); } else @@ -318,6 +316,20 @@ void __blk_run_queue(struct request_queue *q, bool force_kblockd) EXPORT_SYMBOL(__blk_run_queue); /** + * blk_run_queue_async - run a single device queue in workqueue context + * @q: The queue to run + * + * Description: + * Tells kblockd to perform the equivalent of @blk_run_queue on behalf + * of us. + */ +void blk_run_queue_async(struct request_queue *q) +{ + if (likely(!blk_queue_stopped(q))) + queue_delayed_work(kblockd_workqueue, &q->delay_work, 0); +} + +/** * blk_run_queue - run a single device queue * @q: The queue to run * @@ -330,7 +342,7 @@ void blk_run_queue(struct request_queue *q) unsigned long flags; spin_lock_irqsave(q->queue_lock, flags); - __blk_run_queue(q, false); + __blk_run_queue(q); spin_unlock_irqrestore(q->queue_lock, flags); } EXPORT_SYMBOL(blk_run_queue); @@ -979,7 +991,7 @@ void blk_insert_request(struct request_queue *q, struct request *rq, blk_queue_end_tag(q, rq); add_acct_request(q, rq, where); - __blk_run_queue(q, false); + __blk_run_queue(q); spin_unlock_irqrestore(q->queue_lock, flags); } EXPORT_SYMBOL(blk_insert_request); @@ -1323,7 +1335,7 @@ get_rq: } else { spin_lock_irq(q->queue_lock); add_acct_request(q, req, where); - __blk_run_queue(q, false); + __blk_run_queue(q); out_unlock: spin_unlock_irq(q->queue_lock); } @@ -2684,9 +2696,9 @@ static void queue_unplugged(struct request_queue *q, unsigned int depth, */ if (from_schedule) { spin_unlock(q->queue_lock); - __blk_run_queue(q, true); + blk_run_queue_async(q); } else { - __blk_run_queue(q, false); + __blk_run_queue(q); spin_unlock(q->queue_lock); } diff --git a/block/blk-exec.c b/block/blk-exec.c index 7482b7f..81e3181 100644 --- a/block/blk-exec.c +++ b/block/blk-exec.c @@ -55,7 +55,7 @@ void blk_execute_rq_nowait(struct request_queue *q, struct gendisk *bd_disk, WARN_ON(irqs_disabled()); spin_lock_irq(q->queue_lock); __elv_add_request(q, rq, where); - __blk_run_queue(q, false); + __blk_run_queue(q); /* the queue is stopped so it won't be plugged+unplugged */ if (rq->cmd_type == REQ_TYPE_PM_RESUME) q->request_fn(q); diff --git a/block/blk-flush.c b/block/blk-flush.c index eba4a27..6c9b5e1 100644 --- a/block/blk-flush.c +++ b/block/blk-flush.c @@ -218,7 +218,7 @@ static void flush_end_io(struct request *flush_rq, int error) * request_fn may confuse the driver. Always use kblockd. */ if (queued) - __blk_run_queue(q, true); + blk_run_queue_async(q); } /** @@ -274,7 +274,7 @@ static void flush_data_end_io(struct request *rq, int error) * the comment in flush_end_io(). */ if (blk_flush_complete_seq(rq, REQ_FSEQ_DATA, error)) - __blk_run_queue(q, true); + blk_run_queue_async(q); } /** diff --git a/block/blk.h b/block/blk.h index 6126346..c9df8fc 100644 --- a/block/blk.h +++ b/block/blk.h @@ -22,6 +22,7 @@ void blk_rq_timed_out_timer(unsigned long data); void blk_delete_timer(struct request *); void blk_add_timer(struct request *); void __generic_unplug_device(struct request_queue *); +void blk_run_queue_async(struct request_queue *q); /* * Internal atomic flags for request handling diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index 3be881e..46b0a1d 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -3368,7 +3368,7 @@ cfq_rq_enqueued(struct cfq_data *cfqd, struct cfq_queue *cfqq, cfqd->busy_queues > 1) { cfq_del_timer(cfqd, cfqq); cfq_clear_cfqq_wait_request(cfqq); - __blk_run_queue(cfqd->queue, false); + __blk_run_queue(cfqd->queue); } else { cfq_blkiocg_update_idle_time_stats( &cfqq->cfqg->blkg); @@ -3383,7 +3383,7 @@ cfq_rq_enqueued(struct cfq_data *cfqd, struct cfq_queue *cfqq, * this new queue is RT and the current one is BE */ cfq_preempt_queue(cfqd, cfqq); - __blk_run_queue(cfqd->queue, false); + __blk_run_queue(cfqd->queue); } } @@ -3743,7 +3743,7 @@ static void cfq_kick_queue(struct work_struct *work) struct request_queue *q = cfqd->queue; spin_lock_irq(q->queue_lock); - __blk_run_queue(cfqd->queue, false); + __blk_run_queue(cfqd->queue); spin_unlock_irq(q->queue_lock); } diff --git a/block/elevator.c b/block/elevator.c index 0cdb4e7..6f6abc0 100644 --- a/block/elevator.c +++ b/block/elevator.c @@ -642,7 +642,7 @@ void elv_quiesce_start(struct request_queue *q) */ elv_drain_elevator(q); while (q->rq.elvpriv) { - __blk_run_queue(q, false); + __blk_run_queue(q); spin_unlock_irq(q->queue_lock); msleep(10); spin_lock_irq(q->queue_lock); @@ -695,7 +695,7 @@ void __elv_add_request(struct request_queue *q, struct request *rq, int where) * with anything. There's no point in delaying queue * processing. */ - __blk_run_queue(q, false); + __blk_run_queue(q); break; case ELEVATOR_INSERT_SORT_MERGE: diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 6d5c7ff..ab55c2f 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -443,7 +443,7 @@ static void scsi_run_queue(struct request_queue *q) &sdev->request_queue->queue_flags); if (flagset) queue_flag_set(QUEUE_FLAG_REENTER, sdev->request_queue); - __blk_run_queue(sdev->request_queue, false); + __blk_run_queue(sdev->request_queue); if (flagset) queue_flag_clear(QUEUE_FLAG_REENTER, sdev->request_queue); spin_unlock(sdev->request_queue->queue_lock); diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c index fdf3fa6..28c3350 100644 --- a/drivers/scsi/scsi_transport_fc.c +++ b/drivers/scsi/scsi_transport_fc.c @@ -3829,7 +3829,7 @@ fc_bsg_goose_queue(struct fc_rport *rport) !test_bit(QUEUE_FLAG_REENTER, &rport->rqst_q->queue_flags); if (flagset) queue_flag_set(QUEUE_FLAG_REENTER, rport->rqst_q); - __blk_run_queue(rport->rqst_q, false); + __blk_run_queue(rport->rqst_q); if (flagset) queue_flag_clear(QUEUE_FLAG_REENTER, rport->rqst_q); spin_unlock_irqrestore(rport->rqst_q->queue_lock, flags); diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index 3448d89..cbbfd98 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -697,7 +697,7 @@ extern void blk_start_queue(struct request_queue *q); extern void blk_stop_queue(struct request_queue *q); extern void blk_sync_queue(struct request_queue *q); extern void __blk_stop_queue(struct request_queue *q); -extern void __blk_run_queue(struct request_queue *q, bool force_kblockd); +extern void __blk_run_queue(struct request_queue *q); extern void blk_run_queue(struct request_queue *); extern int blk_rq_map_user(struct request_queue *, struct request *, struct rq_map_data *, void __user *, unsigned long, -- cgit v0.10.2 From f65647c29b14f5a32ff6f3237b0ef3b375ed5a79 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Mon, 18 Apr 2011 08:55:34 -0400 Subject: Btrfs: fix free space cache leak The free space caching code was recently reworked to cache all the pages it needed instead of using find_get_page everywhere. One loop was missed though, so it ended up leaking pages. This fixes it to use our page array instead of find_get_page. Signed-off-by: Chris Mason diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index a3f420d..11d2e9c 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -732,7 +732,7 @@ int btrfs_write_out_cache(struct btrfs_root *root, out_of_space = true; break; } - page = find_get_page(inode->i_mapping, index); + page = pages[index]; addr = kmap(page); memcpy(addr, entry->bitmap, PAGE_CACHE_SIZE); -- cgit v0.10.2 From deab72d3797e3d4340c7ddf968234b8c3d01d7a5 Mon Sep 17 00:00:00 2001 From: Bob Peterson Date: Wed, 16 Mar 2011 16:32:39 -0400 Subject: GFS2: write_end error path fails to unlock transaction lock I did an audit of gfs2's transaction glock for bugzilla bug 658619 and ran across this: In function gfs2_write_end, in the unlikely event that gfs2_meta_inode_buffer returns an error, the code may forget to unlock the transaction lock because the "failed" label appears after the call to function gfs2_trans_end. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index c71995b..0f5c4f9 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -884,8 +884,8 @@ static int gfs2_write_end(struct file *file, struct address_space *mapping, } brelse(dibh); - gfs2_trans_end(sdp); failed: + gfs2_trans_end(sdp); if (al) { gfs2_inplace_release(ip); gfs2_quota_unlock(ip); -- cgit v0.10.2 From 0ee532062fa7ff0795b3862c2d50efe32e552f9f Mon Sep 17 00:00:00 2001 From: Benjamin Marzinski Date: Thu, 17 Mar 2011 21:54:46 -0500 Subject: GFS2: directly write blocks past i_size GFS2 was relying on the writepage code to write out the zeroed data for fallocate. However, with FALLOC_FL_KEEP_SIZE set, this may be past i_size. If it is, it will be ignored. To work around this, gfs2 now calls write_dirty_buffer directly on the buffer_heads when FALLOC_FL_KEEP_SIZE is set, and it's writing past i_size. This version is just a cleanup of my last version Signed-off-by: Benjamin Marzinski Signed-off-by: Steven Whitehouse diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index b2682e0..e483108 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -617,18 +617,51 @@ static ssize_t gfs2_file_aio_write(struct kiocb *iocb, const struct iovec *iov, return generic_file_aio_write(iocb, iov, nr_segs, pos); } -static void empty_write_end(struct page *page, unsigned from, - unsigned to) +static int empty_write_end(struct page *page, unsigned from, + unsigned to, int mode) { - struct gfs2_inode *ip = GFS2_I(page->mapping->host); + struct inode *inode = page->mapping->host; + struct gfs2_inode *ip = GFS2_I(inode); + struct buffer_head *bh; + unsigned offset, blksize = 1 << inode->i_blkbits; + pgoff_t end_index = i_size_read(inode) >> PAGE_CACHE_SHIFT; zero_user(page, from, to-from); mark_page_accessed(page); - if (!gfs2_is_writeback(ip)) - gfs2_page_add_databufs(ip, page, from, to); + if (page->index < end_index || !(mode & FALLOC_FL_KEEP_SIZE)) { + if (!gfs2_is_writeback(ip)) + gfs2_page_add_databufs(ip, page, from, to); + + block_commit_write(page, from, to); + return 0; + } + + offset = 0; + bh = page_buffers(page); + while (offset < to) { + if (offset >= from) { + set_buffer_uptodate(bh); + mark_buffer_dirty(bh); + clear_buffer_new(bh); + write_dirty_buffer(bh, WRITE); + } + offset += blksize; + bh = bh->b_this_page; + } - block_commit_write(page, from, to); + offset = 0; + bh = page_buffers(page); + while (offset < to) { + if (offset >= from) { + wait_on_buffer(bh); + if (!buffer_uptodate(bh)) + return -EIO; + } + offset += blksize; + bh = bh->b_this_page; + } + return 0; } static int needs_empty_write(sector_t block, struct inode *inode) @@ -643,7 +676,8 @@ static int needs_empty_write(sector_t block, struct inode *inode) return !buffer_mapped(&bh_map); } -static int write_empty_blocks(struct page *page, unsigned from, unsigned to) +static int write_empty_blocks(struct page *page, unsigned from, unsigned to, + int mode) { struct inode *inode = page->mapping->host; unsigned start, end, next, blksize; @@ -668,7 +702,9 @@ static int write_empty_blocks(struct page *page, unsigned from, unsigned to) gfs2_block_map); if (unlikely(ret)) return ret; - empty_write_end(page, start, end); + ret = empty_write_end(page, start, end, mode); + if (unlikely(ret)) + return ret; end = 0; } start = next; @@ -682,7 +718,9 @@ static int write_empty_blocks(struct page *page, unsigned from, unsigned to) ret = __block_write_begin(page, start, end - start, gfs2_block_map); if (unlikely(ret)) return ret; - empty_write_end(page, start, end); + ret = empty_write_end(page, start, end, mode); + if (unlikely(ret)) + return ret; } return 0; @@ -731,7 +769,7 @@ static int fallocate_chunk(struct inode *inode, loff_t offset, loff_t len, if (curr == end) to = end_offset; - error = write_empty_blocks(page, from, to); + error = write_empty_blocks(page, from, to, mode); if (!error && offset + to > inode->i_size && !(mode & FALLOC_FL_KEEP_SIZE)) { i_size_write(inode, offset + to); -- cgit v0.10.2 From 001e8e8df4283dd4ef7a0297c012fce364c05cf1 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Wed, 30 Mar 2011 14:17:51 +0100 Subject: GFS2: Don't try to deallocate unlinked inodes when mounted ro This adds a couple of missing tests to avoid read-only nodes from attempting to deallocate unlinked inodes. Signed-off-by: Steven Whitehouse Reported-by: Michel Andre de la Porte diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c index 3754e3c..25eeb2b 100644 --- a/fs/gfs2/glops.c +++ b/fs/gfs2/glops.c @@ -385,6 +385,10 @@ static int trans_go_demote_ok(const struct gfs2_glock *gl) static void iopen_go_callback(struct gfs2_glock *gl) { struct gfs2_inode *ip = (struct gfs2_inode *)gl->gl_object; + struct gfs2_sbd *sdp = gl->gl_sbd; + + if (sdp->sd_vfs->s_flags & MS_RDONLY) + return; if (gl->gl_demote_state == LM_ST_UNLOCKED && gl->gl_state == LM_ST_SHARED && ip) { diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index a4e23d6..5b2cb81 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -1318,12 +1318,13 @@ static int gfs2_show_options(struct seq_file *s, struct vfsmount *mnt) static void gfs2_evict_inode(struct inode *inode) { - struct gfs2_sbd *sdp = inode->i_sb->s_fs_info; + struct super_block *sb = inode->i_sb; + struct gfs2_sbd *sdp = sb->s_fs_info; struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_holder gh; int error; - if (inode->i_nlink) + if (inode->i_nlink || (sb->s_flags & MS_RDONLY)) goto out; error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); -- cgit v0.10.2 From 44ad37d69b2cc421d5b5c7ad7fed16230685b092 Mon Sep 17 00:00:00 2001 From: Bob Peterson Date: Thu, 17 Mar 2011 16:19:58 -0400 Subject: GFS2: filesystem hang caused by incorrect lock order This patch fixes a deadlock in GFS2 where two processes are trying to reclaim an unlinked dinode: One holds the inode glock and calls gfs2_lookup_by_inum trying to look up the inode, which it can't, due to I_FREEING. The other has set I_FREEING from vfs and is at the beginning of gfs2_delete_inode waiting for the glock, which is held by the first. The solution is to add a new non_block parameter to the gfs2_iget function that causes it to return -ENOENT if the inode is being freed. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c index 5c356d0..f789c57 100644 --- a/fs/gfs2/dir.c +++ b/fs/gfs2/dir.c @@ -1506,7 +1506,7 @@ struct inode *gfs2_dir_search(struct inode *dir, const struct qstr *name) inode = gfs2_inode_lookup(dir->i_sb, be16_to_cpu(dent->de_type), be64_to_cpu(dent->de_inum.no_addr), - be64_to_cpu(dent->de_inum.no_formal_ino)); + be64_to_cpu(dent->de_inum.no_formal_ino), 0); brelse(bh); return inode; } diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index 97d54a2..9134dcb 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -40,37 +40,61 @@ struct gfs2_inum_range_host { u64 ir_length; }; +struct gfs2_skip_data { + u64 no_addr; + int skipped; + int non_block; +}; + static int iget_test(struct inode *inode, void *opaque) { struct gfs2_inode *ip = GFS2_I(inode); - u64 *no_addr = opaque; + struct gfs2_skip_data *data = opaque; - if (ip->i_no_addr == *no_addr) + if (ip->i_no_addr == data->no_addr) { + if (data->non_block && + inode->i_state & (I_FREEING|I_CLEAR|I_WILL_FREE)) { + data->skipped = 1; + return 0; + } return 1; - + } return 0; } static int iget_set(struct inode *inode, void *opaque) { struct gfs2_inode *ip = GFS2_I(inode); - u64 *no_addr = opaque; + struct gfs2_skip_data *data = opaque; - inode->i_ino = (unsigned long)*no_addr; - ip->i_no_addr = *no_addr; + if (data->skipped) + return -ENOENT; + inode->i_ino = (unsigned long)(data->no_addr); + ip->i_no_addr = data->no_addr; return 0; } struct inode *gfs2_ilookup(struct super_block *sb, u64 no_addr) { unsigned long hash = (unsigned long)no_addr; - return ilookup5(sb, hash, iget_test, &no_addr); + struct gfs2_skip_data data; + + data.no_addr = no_addr; + data.skipped = 0; + data.non_block = 0; + return ilookup5(sb, hash, iget_test, &data); } -static struct inode *gfs2_iget(struct super_block *sb, u64 no_addr) +static struct inode *gfs2_iget(struct super_block *sb, u64 no_addr, + int non_block) { + struct gfs2_skip_data data; unsigned long hash = (unsigned long)no_addr; - return iget5_locked(sb, hash, iget_test, iget_set, &no_addr); + + data.no_addr = no_addr; + data.skipped = 0; + data.non_block = non_block; + return iget5_locked(sb, hash, iget_test, iget_set, &data); } /** @@ -111,19 +135,20 @@ static void gfs2_set_iop(struct inode *inode) * @sb: The super block * @no_addr: The inode number * @type: The type of the inode + * non_block: Can we block on inodes that are being freed? * * Returns: A VFS inode, or an error */ struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned int type, - u64 no_addr, u64 no_formal_ino) + u64 no_addr, u64 no_formal_ino, int non_block) { struct inode *inode; struct gfs2_inode *ip; struct gfs2_glock *io_gl = NULL; int error; - inode = gfs2_iget(sb, no_addr); + inode = gfs2_iget(sb, no_addr, non_block); ip = GFS2_I(inode); if (!inode) @@ -185,11 +210,12 @@ struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr, { struct super_block *sb = sdp->sd_vfs; struct gfs2_holder i_gh; - struct inode *inode; + struct inode *inode = NULL; int error; + /* Must not read in block until block type is verified */ error = gfs2_glock_nq_num(sdp, no_addr, &gfs2_inode_glops, - LM_ST_SHARED, LM_FLAG_ANY, &i_gh); + LM_ST_EXCLUSIVE, GL_SKIP, &i_gh); if (error) return ERR_PTR(error); @@ -197,7 +223,7 @@ struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr, if (error) goto fail; - inode = gfs2_inode_lookup(sb, DT_UNKNOWN, no_addr, 0); + inode = gfs2_inode_lookup(sb, DT_UNKNOWN, no_addr, 0, 1); if (IS_ERR(inode)) goto fail; @@ -843,7 +869,7 @@ struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name, goto fail_gunlock2; inode = gfs2_inode_lookup(dir->i_sb, IF2DT(mode), inum.no_addr, - inum.no_formal_ino); + inum.no_formal_ino, 0); if (IS_ERR(inode)) goto fail_gunlock2; diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h index 3e00a66..099ca30 100644 --- a/fs/gfs2/inode.h +++ b/fs/gfs2/inode.h @@ -97,7 +97,8 @@ err: } extern struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned type, - u64 no_addr, u64 no_formal_ino); + u64 no_addr, u64 no_formal_ino, + int non_block); extern struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr, u64 *no_formal_ino, unsigned int blktype); diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 42ef243..d3c69eb 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -430,7 +430,7 @@ static int gfs2_lookup_root(struct super_block *sb, struct dentry **dptr, struct dentry *dentry; struct inode *inode; - inode = gfs2_inode_lookup(sb, DT_DIR, no_addr, 0); + inode = gfs2_inode_lookup(sb, DT_DIR, no_addr, 0, 0); if (IS_ERR(inode)) { fs_err(sdp, "can't read in %s inode: %ld\n", name, PTR_ERR(inode)); return PTR_ERR(inode); diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index cf930cd..6fcae84 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -945,7 +945,7 @@ static void try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked, u64 skip /* rgblk_search can return a block < goal, so we need to keep it marching forward. */ no_addr = block + rgd->rd_data0; - goal++; + goal = max(block + 1, goal + 1); if (*last_unlinked != NO_BLOCK && no_addr <= *last_unlinked) continue; if (no_addr == skip) @@ -971,7 +971,7 @@ static void try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked, u64 skip found++; /* Limit reclaim to sensible number of tasks */ - if (found > 2*NR_CPUS) + if (found > NR_CPUS) return; } diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index 5b2cb81..b9f28e6 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -1327,7 +1327,8 @@ static void gfs2_evict_inode(struct inode *inode) if (inode->i_nlink || (sb->s_flags & MS_RDONLY)) goto out; - error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); + /* Must not read inode block until block type has been verified */ + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_SKIP, &gh); if (unlikely(error)) { gfs2_glock_dq_uninit(&ip->i_iopen_gh); goto out; @@ -1337,6 +1338,12 @@ static void gfs2_evict_inode(struct inode *inode) if (error) goto out_truncate; + if (test_bit(GIF_INVALID, &ip->i_flags)) { + error = gfs2_inode_refresh(ip); + if (error) + goto out_truncate; + } + ip->i_iopen_gh.gh_flags |= GL_NOCACHE; gfs2_glock_dq_wait(&ip->i_iopen_gh); gfs2_holder_reinit(LM_ST_EXCLUSIVE, LM_FLAG_TRY_1CB | GL_NOCACHE, &ip->i_iopen_gh); -- cgit v0.10.2 From af289bfe15fc92ecfbf6d8312713815b33e452c0 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Mon, 18 Apr 2011 15:45:44 +0200 Subject: x86, gart: Convert spaces to tabs in enable_gart_translation Probably by copy&paste this function was indented by spaces. Convert this to tabs. Signed-off-by: Joerg Roedel Link: http://lkml.kernel.org/r/1303134346-5805-3-git-send-email-joerg.roedel@amd.com Signed-off-by: H. Peter Anvin diff --git a/arch/x86/include/asm/gart.h b/arch/x86/include/asm/gart.h index 43085bf..88c1ebe 100644 --- a/arch/x86/include/asm/gart.h +++ b/arch/x86/include/asm/gart.h @@ -75,17 +75,17 @@ static inline void enable_gart_translation(struct pci_dev *dev, u64 addr) { u32 tmp, ctl; - /* address of the mappings table */ - addr >>= 12; - tmp = (u32) addr<<4; - tmp &= ~0xf; - pci_write_config_dword(dev, AMD64_GARTTABLEBASE, tmp); - - /* Enable GART translation for this hammer. */ - pci_read_config_dword(dev, AMD64_GARTAPERTURECTL, &ctl); - ctl |= GARTEN; - ctl &= ~(DISGARTCPU | DISGARTIO); - pci_write_config_dword(dev, AMD64_GARTAPERTURECTL, ctl); + /* address of the mappings table */ + addr >>= 12; + tmp = (u32) addr<<4; + tmp &= ~0xf; + pci_write_config_dword(dev, AMD64_GARTTABLEBASE, tmp); + + /* Enable GART translation for this hammer. */ + pci_read_config_dword(dev, AMD64_GARTAPERTURECTL, &ctl); + ctl |= GARTEN; + ctl &= ~(DISGARTCPU | DISGARTIO); + pci_write_config_dword(dev, AMD64_GARTAPERTURECTL, ctl); } static inline int aperture_valid(u64 aper_base, u32 aper_size, u32 min_size) -- cgit v0.10.2 From c34151a742d84ae65db2088ea30495063f697fbe Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Mon, 18 Apr 2011 15:45:45 +0200 Subject: x86, gart: Set DISTLBWALKPRB bit always The DISTLBWALKPRB bit must be set for the GART because the gatt table is mapped UC. But the current code does not set the bit at boot when the BIOS setup the aperture correctly. Fix that by setting this bit when enabling the GART instead of the other places. Cc: Cc: Borislav Petkov Signed-off-by: Joerg Roedel Link: http://lkml.kernel.org/r/1303134346-5805-4-git-send-email-joerg.roedel@amd.com Signed-off-by: H. Peter Anvin diff --git a/arch/x86/include/asm/gart.h b/arch/x86/include/asm/gart.h index 88c1ebe..156cd5d 100644 --- a/arch/x86/include/asm/gart.h +++ b/arch/x86/include/asm/gart.h @@ -66,7 +66,7 @@ static inline void gart_set_size_and_enable(struct pci_dev *dev, u32 order) * Don't enable translation but enable GART IO and CPU accesses. * Also, set DISTLBWALKPRB since GART tables memory is UC. */ - ctl = DISTLBWALKPRB | order << 1; + ctl = order << 1; pci_write_config_dword(dev, AMD64_GARTAPERTURECTL, ctl); } @@ -83,7 +83,7 @@ static inline void enable_gart_translation(struct pci_dev *dev, u64 addr) /* Enable GART translation for this hammer. */ pci_read_config_dword(dev, AMD64_GARTAPERTURECTL, &ctl); - ctl |= GARTEN; + ctl |= GARTEN | DISTLBWALKPRB; ctl &= ~(DISGARTCPU | DISGARTIO); pci_write_config_dword(dev, AMD64_GARTAPERTURECTL, ctl); } diff --git a/arch/x86/kernel/aperture_64.c b/arch/x86/kernel/aperture_64.c index 86d1ad4..73fb469 100644 --- a/arch/x86/kernel/aperture_64.c +++ b/arch/x86/kernel/aperture_64.c @@ -499,7 +499,7 @@ out: * Don't enable translation yet but enable GART IO and CPU * accesses and set DISTLBWALKPRB since GART table memory is UC. */ - u32 ctl = DISTLBWALKPRB | aper_order << 1; + u32 ctl = aper_order << 1; bus = amd_nb_bus_dev_ranges[i].bus; dev_base = amd_nb_bus_dev_ranges[i].dev_base; -- cgit v0.10.2 From 665d3e2af83c8fbd149534db8f57d82fa6fa6753 Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Mon, 18 Apr 2011 15:45:46 +0200 Subject: x86, gart: Make sure GART does not map physmem above 1TB The GART can only map physical memory below 1TB. Make sure the gart driver in the kernel does not try to map memory above 1TB. Cc: Signed-off-by: Joerg Roedel Link: http://lkml.kernel.org/r/1303134346-5805-5-git-send-email-joerg.roedel@amd.com Signed-off-by: H. Peter Anvin diff --git a/arch/x86/kernel/pci-gart_64.c b/arch/x86/kernel/pci-gart_64.c index 82ada01..b117efd 100644 --- a/arch/x86/kernel/pci-gart_64.c +++ b/arch/x86/kernel/pci-gart_64.c @@ -81,6 +81,9 @@ static u32 gart_unmapped_entry; #define AGPEXTERN #endif +/* GART can only remap to physical addresses < 1TB */ +#define GART_MAX_PHYS_ADDR (1ULL << 40) + /* backdoor interface to AGP driver */ AGPEXTERN int agp_memory_reserved; AGPEXTERN __u32 *agp_gatt_table; @@ -212,9 +215,13 @@ static dma_addr_t dma_map_area(struct device *dev, dma_addr_t phys_mem, size_t size, int dir, unsigned long align_mask) { unsigned long npages = iommu_num_pages(phys_mem, size, PAGE_SIZE); - unsigned long iommu_page = alloc_iommu(dev, npages, align_mask); + unsigned long iommu_page; int i; + if (unlikely(phys_mem + size > GART_MAX_PHYS_ADDR)) + return bad_dma_addr; + + iommu_page = alloc_iommu(dev, npages, align_mask); if (iommu_page == -1) { if (!nonforced_iommu(dev, phys_mem, size)) return phys_mem; -- cgit v0.10.2 From 80b4895aa4578e9372d76cd4063f82d0c3994d77 Mon Sep 17 00:00:00 2001 From: Jeff Brown Date: Mon, 18 Apr 2011 10:08:02 -0700 Subject: Input: estimate number of events per packet Calculate a default based on the number of ABS axes, REL axes, and MT slots for the device during input device registration. Signed-off-by: Jeff Brown Reviewed-by: Henrik Rydberg Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/input.c b/drivers/input/input.c index d6e8bd8..ebbceed 100644 --- a/drivers/input/input.c +++ b/drivers/input/input.c @@ -1746,6 +1746,42 @@ void input_set_capability(struct input_dev *dev, unsigned int type, unsigned int } EXPORT_SYMBOL(input_set_capability); +static unsigned int input_estimate_events_per_packet(struct input_dev *dev) +{ + int mt_slots; + int i; + unsigned int events; + + if (dev->mtsize) { + mt_slots = dev->mtsize; + } else if (test_bit(ABS_MT_TRACKING_ID, dev->absbit)) { + mt_slots = dev->absinfo[ABS_MT_TRACKING_ID].maximum - + dev->absinfo[ABS_MT_TRACKING_ID].minimum + 1, + clamp(mt_slots, 2, 32); + } else if (test_bit(ABS_MT_POSITION_X, dev->absbit)) { + mt_slots = 2; + } else { + mt_slots = 0; + } + + events = mt_slots + 1; /* count SYN_MT_REPORT and SYN_REPORT */ + + for (i = 0; i < ABS_CNT; i++) { + if (test_bit(i, dev->absbit)) { + if (input_is_mt_axis(i)) + events += mt_slots; + else + events++; + } + } + + for (i = 0; i < REL_CNT; i++) + if (test_bit(i, dev->relbit)) + events++; + + return events; +} + #define INPUT_CLEANSE_BITMASK(dev, type, bits) \ do { \ if (!test_bit(EV_##type, dev->evbit)) \ @@ -1793,6 +1829,10 @@ int input_register_device(struct input_dev *dev) /* Make sure that bitmasks not mentioned in dev->evbit are clean. */ input_cleanse_bitmasks(dev); + if (!dev->hint_events_per_packet) + dev->hint_events_per_packet = + input_estimate_events_per_packet(dev); + /* * If delay and period are pre-set by the driver, then autorepeating * is handled by the driver itself and we don't do it in input.c. diff --git a/include/linux/input/mt.h b/include/linux/input/mt.h index b3ac06a..318bb82 100644 --- a/include/linux/input/mt.h +++ b/include/linux/input/mt.h @@ -48,6 +48,12 @@ static inline void input_mt_slot(struct input_dev *dev, int slot) input_event(dev, EV_ABS, ABS_MT_SLOT, slot); } +static inline bool input_is_mt_axis(int axis) +{ + return axis == ABS_MT_SLOT || + (axis >= ABS_MT_FIRST && axis <= ABS_MT_LAST); +} + void input_mt_report_slot_state(struct input_dev *dev, unsigned int tool_type, bool active); -- cgit v0.10.2 From c36b58e8a9112017c2bcc322cc98e71241814303 Mon Sep 17 00:00:00 2001 From: Igor Mammedov Date: Mon, 18 Apr 2011 10:17:17 -0700 Subject: Input: xen-kbdfront - fix mouse getting stuck after save/restore Mouse gets "stuck" after restore of PV guest but buttons are in working condition. If driver has been configured for ABS coordinates at start it will get XENKBD_TYPE_POS events and then suddenly after restore it'll start getting XENKBD_TYPE_MOTION events, that will be dropped later and they won't get into user-space. Regression was introduced by hunk 5 and 6 of 5ea5254aa0ad269cfbd2875c973ef25ab5b5e9db ("Input: xen-kbdfront - advertise either absolute or relative coordinates"). Driver on restore should ask xen for request-abs-pointer again if it is available. So restore parts that did it before 5ea5254. Acked-by: Olaf Hering Signed-off-by: Igor Mammedov [v1: Expanded the commit description] Signed-off-by: Konrad Rzeszutek Wilk Signed-off-by: Dmitry Torokhov diff --git a/drivers/input/misc/xen-kbdfront.c b/drivers/input/misc/xen-kbdfront.c index 7077f9b..62bae99 100644 --- a/drivers/input/misc/xen-kbdfront.c +++ b/drivers/input/misc/xen-kbdfront.c @@ -303,7 +303,7 @@ static void xenkbd_backend_changed(struct xenbus_device *dev, enum xenbus_state backend_state) { struct xenkbd_info *info = dev_get_drvdata(&dev->dev); - int val; + int ret, val; switch (backend_state) { case XenbusStateInitialising: @@ -316,6 +316,17 @@ static void xenkbd_backend_changed(struct xenbus_device *dev, case XenbusStateInitWait: InitWait: + ret = xenbus_scanf(XBT_NIL, info->xbdev->otherend, + "feature-abs-pointer", "%d", &val); + if (ret < 0) + val = 0; + if (val) { + ret = xenbus_printf(XBT_NIL, info->xbdev->nodename, + "request-abs-pointer", "1"); + if (ret) + pr_warning("xenkbd: can't request abs-pointer"); + } + xenbus_switch_state(dev, XenbusStateConnected); break; -- cgit v0.10.2 From 4ee63624fd927376b97ead3a8d00728d437bc8e8 Mon Sep 17 00:00:00 2001 From: "J. Bruce Fields" Date: Fri, 15 Apr 2011 18:08:26 -0400 Subject: nfsd4: fix struct file leak on delegation Introduced by acfdf5c383b38f7f4dddae41b97c97f1ae058f49. Cc: stable@kernel.org Reported-by: Gerhard Heift Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index aa309aa..c79a983 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -258,6 +258,7 @@ static void nfs4_put_deleg_lease(struct nfs4_file *fp) if (atomic_dec_and_test(&fp->fi_delegees)) { vfs_setlease(fp->fi_deleg_file, F_UNLCK, &fp->fi_lease); fp->fi_lease = NULL; + fput(fp->fi_deleg_file); fp->fi_deleg_file = NULL; } } -- cgit v0.10.2 From c78193e9c7bcbf25b8237ad0dec82f805c4ea69b Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 18 Apr 2011 10:35:30 -0700 Subject: next_pidmap: fix overflow condition MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit next_pidmap() just quietly accepted whatever 'last' pid that was passed in, which is not all that safe when one of the users is /proc. Admittedly the proc code should do some sanity checking on the range (and that will be the next commit), but that doesn't mean that the helper functions should just do that pidmap pointer arithmetic without checking the range of its arguments. So clamp 'last' to PID_MAX_LIMIT. The fact that we then do "last+1" doesn't really matter, the for-loop does check against the end of the pidmap array properly (it's only the actual pointer arithmetic overflow case we need to worry about, and going one bit beyond isn't going to overflow). [ Use PID_MAX_LIMIT rather than pid_max as per Eric Biederman ] Reported-by: Tavis Ormandy Analyzed-by: Robert Święcki Cc: Eric W. Biederman Cc: Pavel Emelyanov Signed-off-by: Linus Torvalds diff --git a/include/linux/pid.h b/include/linux/pid.h index 31afb7e..cdced84 100644 --- a/include/linux/pid.h +++ b/include/linux/pid.h @@ -117,7 +117,7 @@ extern struct pid *find_vpid(int nr); */ extern struct pid *find_get_pid(int nr); extern struct pid *find_ge_pid(int nr, struct pid_namespace *); -int next_pidmap(struct pid_namespace *pid_ns, int last); +int next_pidmap(struct pid_namespace *pid_ns, unsigned int last); extern struct pid *alloc_pid(struct pid_namespace *ns); extern void free_pid(struct pid *pid); diff --git a/kernel/pid.c b/kernel/pid.c index 02f2212..57a8346 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -217,11 +217,14 @@ static int alloc_pidmap(struct pid_namespace *pid_ns) return -1; } -int next_pidmap(struct pid_namespace *pid_ns, int last) +int next_pidmap(struct pid_namespace *pid_ns, unsigned int last) { int offset; struct pidmap *map, *end; + if (last >= PID_MAX_LIMIT) + return -1; + offset = (last + 1) & BITS_PER_PAGE_MASK; map = &pid_ns->pidmap[(last + 1)/BITS_PER_PAGE]; end = &pid_ns->pidmap[PIDMAP_ENTRIES]; -- cgit v0.10.2 From d8bdc59f215e62098bc5b4256fd9928bf27053a1 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 18 Apr 2011 10:36:54 -0700 Subject: proc: do proper range check on readdir offset Rather than pass in some random truncated offset to the pid-related functions, check that the offset is in range up-front. This is just cleanup, the previous commit fixed the real problem. Cc: stable@kernel.org Signed-off-by: Linus Torvalds diff --git a/fs/proc/base.c b/fs/proc/base.c index dd6628d..dfa5327 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -3124,11 +3124,16 @@ static int proc_pid_fill_cache(struct file *filp, void *dirent, filldir_t filldi /* for the /proc/ directory itself, after non-process stuff has been done */ int proc_pid_readdir(struct file * filp, void * dirent, filldir_t filldir) { - unsigned int nr = filp->f_pos - FIRST_PROCESS_ENTRY; - struct task_struct *reaper = get_proc_task(filp->f_path.dentry->d_inode); + unsigned int nr; + struct task_struct *reaper; struct tgid_iter iter; struct pid_namespace *ns; + if (filp->f_pos >= PID_MAX_LIMIT + TGID_OFFSET) + goto out_no_task; + nr = filp->f_pos - FIRST_PROCESS_ENTRY; + + reaper = get_proc_task(filp->f_path.dentry->d_inode); if (!reaper) goto out_no_task; -- cgit v0.10.2 From bd900d4580107c899d43b262fbbd995f11097a43 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Mon, 18 Apr 2011 22:06:57 +0200 Subject: block: kill blk_flush_plug_list() export With all drivers and file systems converted, we only have in-core use of this function. So remove the export. Reporteed-by: Christoph Hellwig Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 5fa3dd2..580eee5 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -2787,7 +2787,6 @@ void blk_flush_plug_list(struct blk_plug *plug, bool from_schedule) local_irq_restore(flags); } -EXPORT_SYMBOL(blk_flush_plug_list); void blk_finish_plug(struct blk_plug *plug) { -- cgit v0.10.2 From f0e615c3cb72b42191b558c130409335812621d8 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 18 Apr 2011 21:26:00 -0700 Subject: Linux 2.6.39-rc4 diff --git a/Makefile b/Makefile index 322e733..b967b96 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 2 PATCHLEVEL = 6 SUBLEVEL = 39 -EXTRAVERSION = -rc3 +EXTRAVERSION = -rc4 NAME = Flesh-Eating Bats with Fangs # *DOCUMENTATION* -- cgit v0.10.2 From 5f45c69589b7d2953584e6cd0b31e35dbe960ad0 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 19 Apr 2011 09:10:35 +0200 Subject: cfq-iosched: read_lock() does not always imply rcu_read_lock() For some configurations of CONFIG_PREEMPT that is not true. So get rid of __call_for_each_cic() and always uses the explicitly rcu_read_lock() protected call_for_each_cic() instead. This fixes a potential bug related to IO scheduler removal or online switching. Thanks to Paul McKenney for clarifying this. Signed-off-by: Jens Axboe diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c index 46b0a1d..5b52011 100644 --- a/block/cfq-iosched.c +++ b/block/cfq-iosched.c @@ -2582,28 +2582,20 @@ static void cfq_put_queue(struct cfq_queue *cfqq) } /* - * Must always be called with the rcu_read_lock() held + * Call func for each cic attached to this ioc. */ static void -__call_for_each_cic(struct io_context *ioc, - void (*func)(struct io_context *, struct cfq_io_context *)) +call_for_each_cic(struct io_context *ioc, + void (*func)(struct io_context *, struct cfq_io_context *)) { struct cfq_io_context *cic; struct hlist_node *n; + rcu_read_lock(); + hlist_for_each_entry_rcu(cic, n, &ioc->cic_list, cic_list) func(ioc, cic); -} -/* - * Call func for each cic attached to this ioc. - */ -static void -call_for_each_cic(struct io_context *ioc, - void (*func)(struct io_context *, struct cfq_io_context *)) -{ - rcu_read_lock(); - __call_for_each_cic(ioc, func); rcu_read_unlock(); } @@ -2664,7 +2656,7 @@ static void cfq_free_io_context(struct io_context *ioc) * should be ok to iterate over the known list, we will see all cic's * since no new ones are added. */ - __call_for_each_cic(ioc, cic_free_func); + call_for_each_cic(ioc, cic_free_func); } static void cfq_put_cooperator(struct cfq_queue *cfqq) -- cgit v0.10.2 From 83112e688f5f05dea1e63787db9a6c16b2887a1d Mon Sep 17 00:00:00 2001 From: Andre Przywara Date: Sat, 16 Apr 2011 02:27:53 +0200 Subject: perf, x86: Fix pre-defined cache-misses event for AMD family 15h cpus With AMD cpu family 15h a unit mask was introduced for the Data Cache Miss event (0x041/L1-dcache-load-misses). We need to enable bit 0 (first data cache miss or streaming store to a 64 B cache line) of this mask to proper count data cache misses. Now we set this bit for all families and models. In case a PMU does not implement a unit mask for event 0x041 the bit is ignored. Signed-off-by: Andre Przywara Signed-off-by: Robert Richter Signed-off-by: Peter Zijlstra Link: http://lkml.kernel.org/r/1302913676-14352-2-git-send-email-robert.richter@amd.com Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c index 461f62b..4e16138 100644 --- a/arch/x86/kernel/cpu/perf_event_amd.c +++ b/arch/x86/kernel/cpu/perf_event_amd.c @@ -8,7 +8,7 @@ static __initconst const u64 amd_hw_cache_event_ids [ C(L1D) ] = { [ C(OP_READ) ] = { [ C(RESULT_ACCESS) ] = 0x0040, /* Data Cache Accesses */ - [ C(RESULT_MISS) ] = 0x0041, /* Data Cache Misses */ + [ C(RESULT_MISS) ] = 0x0141, /* Data Cache Misses */ }, [ C(OP_WRITE) ] = { [ C(RESULT_ACCESS) ] = 0x0142, /* Data Cache Refills :system */ -- cgit v0.10.2 From 855357a21744e488cbee23a47d2b124035160a87 Mon Sep 17 00:00:00 2001 From: Robert Richter Date: Sat, 16 Apr 2011 02:27:54 +0200 Subject: perf, x86: Fix AMD family 15h FPU event constraints Depending on the unit mask settings some FPU events may be scheduled only on cpu counter #3. This patch fixes this. Signed-off-by: Robert Richter Signed-off-by: Peter Zijlstra Cc: Stephane Eranian Link: http://lkml.kernel.org/r/1302913676-14352-3-git-send-email-robert.richter@amd.com Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/cpu/perf_event_amd.c b/arch/x86/kernel/cpu/perf_event_amd.c index 4e16138..cf4e369 100644 --- a/arch/x86/kernel/cpu/perf_event_amd.c +++ b/arch/x86/kernel/cpu/perf_event_amd.c @@ -427,7 +427,9 @@ static __initconst const struct x86_pmu amd_pmu = { * * Exceptions: * + * 0x000 FP PERF_CTL[3], PERF_CTL[5:3] (*) * 0x003 FP PERF_CTL[3] + * 0x004 FP PERF_CTL[3], PERF_CTL[5:3] (*) * 0x00B FP PERF_CTL[3] * 0x00D FP PERF_CTL[3] * 0x023 DE PERF_CTL[2:0] @@ -448,6 +450,8 @@ static __initconst const struct x86_pmu amd_pmu = { * 0x0DF LS PERF_CTL[5:0] * 0x1D6 EX PERF_CTL[5:0] * 0x1D8 EX PERF_CTL[5:0] + * + * (*) depending on the umask all FPU counters may be used */ static struct event_constraint amd_f15_PMC0 = EVENT_CONSTRAINT(0, 0x01, 0); @@ -460,18 +464,28 @@ static struct event_constraint amd_f15_PMC53 = EVENT_CONSTRAINT(0, 0x38, 0); static struct event_constraint * amd_get_event_constraints_f15h(struct cpu_hw_events *cpuc, struct perf_event *event) { - unsigned int event_code = amd_get_event_code(&event->hw); + struct hw_perf_event *hwc = &event->hw; + unsigned int event_code = amd_get_event_code(hwc); switch (event_code & AMD_EVENT_TYPE_MASK) { case AMD_EVENT_FP: switch (event_code) { + case 0x000: + if (!(hwc->config & 0x0000F000ULL)) + break; + if (!(hwc->config & 0x00000F00ULL)) + break; + return &amd_f15_PMC3; + case 0x004: + if (hweight_long(hwc->config & ARCH_PERFMON_EVENTSEL_UMASK) <= 1) + break; + return &amd_f15_PMC3; case 0x003: case 0x00B: case 0x00D: return &amd_f15_PMC3; - default: - return &amd_f15_PMC53; } + return &amd_f15_PMC53; case AMD_EVENT_LS: case AMD_EVENT_DC: case AMD_EVENT_EX_LS: -- cgit v0.10.2 From c21e6beba8835d09bb80e34961430b13e60381c5 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 19 Apr 2011 13:32:46 +0200 Subject: block: get rid of QUEUE_FLAG_REENTER We are currently using this flag to check whether it's safe to call into ->request_fn(). If it is set, we punt to kblockd. But we get a lot of false positives and excessive punts to kblockd, which hurts performance. The only real abuser of this infrastructure is SCSI. So export the async queue run and convert SCSI over to use that. There's room for improvement in that SCSI need not always use the async call, but this fixes our performance issue and they can fix that up in due time. Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 580eee5..40725b9 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -303,15 +303,7 @@ void __blk_run_queue(struct request_queue *q) if (unlikely(blk_queue_stopped(q))) return; - /* - * Only recurse once to avoid overrunning the stack, let the unplug - * handling reinvoke the handler shortly if we already got there. - */ - if (!queue_flag_test_and_set(QUEUE_FLAG_REENTER, q)) { - q->request_fn(q); - queue_flag_clear(QUEUE_FLAG_REENTER, q); - } else - queue_delayed_work(kblockd_workqueue, &q->delay_work, 0); + q->request_fn(q); } EXPORT_SYMBOL(__blk_run_queue); @@ -328,6 +320,7 @@ void blk_run_queue_async(struct request_queue *q) if (likely(!blk_queue_stopped(q))) queue_delayed_work(kblockd_workqueue, &q->delay_work, 0); } +EXPORT_SYMBOL(blk_run_queue_async); /** * blk_run_queue - run a single device queue diff --git a/block/blk.h b/block/blk.h index c9df8fc..6126346 100644 --- a/block/blk.h +++ b/block/blk.h @@ -22,7 +22,6 @@ void blk_rq_timed_out_timer(unsigned long data); void blk_delete_timer(struct request *); void blk_add_timer(struct request *); void __generic_unplug_device(struct request_queue *); -void blk_run_queue_async(struct request_queue *q); /* * Internal atomic flags for request handling diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index ab55c2f..e9901b8 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -411,8 +411,6 @@ static void scsi_run_queue(struct request_queue *q) list_splice_init(&shost->starved_list, &starved_list); while (!list_empty(&starved_list)) { - int flagset; - /* * As long as shost is accepting commands and we have * starved queues, call blk_run_queue. scsi_request_fn @@ -435,20 +433,7 @@ static void scsi_run_queue(struct request_queue *q) continue; } - spin_unlock(shost->host_lock); - - spin_lock(sdev->request_queue->queue_lock); - flagset = test_bit(QUEUE_FLAG_REENTER, &q->queue_flags) && - !test_bit(QUEUE_FLAG_REENTER, - &sdev->request_queue->queue_flags); - if (flagset) - queue_flag_set(QUEUE_FLAG_REENTER, sdev->request_queue); - __blk_run_queue(sdev->request_queue); - if (flagset) - queue_flag_clear(QUEUE_FLAG_REENTER, sdev->request_queue); - spin_unlock(sdev->request_queue->queue_lock); - - spin_lock(shost->host_lock); + blk_run_queue_async(sdev->request_queue); } /* put any unprocessed entries back */ list_splice(&starved_list, &shost->starved_list); diff --git a/drivers/scsi/scsi_transport_fc.c b/drivers/scsi/scsi_transport_fc.c index 28c3350..815069d 100644 --- a/drivers/scsi/scsi_transport_fc.c +++ b/drivers/scsi/scsi_transport_fc.c @@ -3816,28 +3816,17 @@ fail_host_msg: static void fc_bsg_goose_queue(struct fc_rport *rport) { - int flagset; - unsigned long flags; - if (!rport->rqst_q) return; + /* + * This get/put dance makes no sense + */ get_device(&rport->dev); - - spin_lock_irqsave(rport->rqst_q->queue_lock, flags); - flagset = test_bit(QUEUE_FLAG_REENTER, &rport->rqst_q->queue_flags) && - !test_bit(QUEUE_FLAG_REENTER, &rport->rqst_q->queue_flags); - if (flagset) - queue_flag_set(QUEUE_FLAG_REENTER, rport->rqst_q); - __blk_run_queue(rport->rqst_q); - if (flagset) - queue_flag_clear(QUEUE_FLAG_REENTER, rport->rqst_q); - spin_unlock_irqrestore(rport->rqst_q->queue_lock, flags); - + blk_run_queue_async(rport->rqst_q); put_device(&rport->dev); } - /** * fc_bsg_rport_dispatch - process rport bsg requests and dispatch to LLDD * @q: rport request queue diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h index cbbfd98..2ad95fa 100644 --- a/include/linux/blkdev.h +++ b/include/linux/blkdev.h @@ -388,20 +388,19 @@ struct request_queue #define QUEUE_FLAG_SYNCFULL 3 /* read queue has been filled */ #define QUEUE_FLAG_ASYNCFULL 4 /* write queue has been filled */ #define QUEUE_FLAG_DEAD 5 /* queue being torn down */ -#define QUEUE_FLAG_REENTER 6 /* Re-entrancy avoidance */ -#define QUEUE_FLAG_ELVSWITCH 7 /* don't use elevator, just do FIFO */ -#define QUEUE_FLAG_BIDI 8 /* queue supports bidi requests */ -#define QUEUE_FLAG_NOMERGES 9 /* disable merge attempts */ -#define QUEUE_FLAG_SAME_COMP 10 /* force complete on same CPU */ -#define QUEUE_FLAG_FAIL_IO 11 /* fake timeout */ -#define QUEUE_FLAG_STACKABLE 12 /* supports request stacking */ -#define QUEUE_FLAG_NONROT 13 /* non-rotational device (SSD) */ +#define QUEUE_FLAG_ELVSWITCH 6 /* don't use elevator, just do FIFO */ +#define QUEUE_FLAG_BIDI 7 /* queue supports bidi requests */ +#define QUEUE_FLAG_NOMERGES 8 /* disable merge attempts */ +#define QUEUE_FLAG_SAME_COMP 9 /* force complete on same CPU */ +#define QUEUE_FLAG_FAIL_IO 10 /* fake timeout */ +#define QUEUE_FLAG_STACKABLE 11 /* supports request stacking */ +#define QUEUE_FLAG_NONROT 12 /* non-rotational device (SSD) */ #define QUEUE_FLAG_VIRT QUEUE_FLAG_NONROT /* paravirt device */ -#define QUEUE_FLAG_IO_STAT 15 /* do IO stats */ -#define QUEUE_FLAG_DISCARD 16 /* supports DISCARD */ -#define QUEUE_FLAG_NOXMERGES 17 /* No extended merges */ -#define QUEUE_FLAG_ADD_RANDOM 18 /* Contributes to random pool */ -#define QUEUE_FLAG_SECDISCARD 19 /* supports SECDISCARD */ +#define QUEUE_FLAG_IO_STAT 13 /* do IO stats */ +#define QUEUE_FLAG_DISCARD 14 /* supports DISCARD */ +#define QUEUE_FLAG_NOXMERGES 15 /* No extended merges */ +#define QUEUE_FLAG_ADD_RANDOM 16 /* Contributes to random pool */ +#define QUEUE_FLAG_SECDISCARD 17 /* supports SECDISCARD */ #define QUEUE_FLAG_DEFAULT ((1 << QUEUE_FLAG_IO_STAT) | \ (1 << QUEUE_FLAG_STACKABLE) | \ @@ -699,6 +698,7 @@ extern void blk_sync_queue(struct request_queue *q); extern void __blk_stop_queue(struct request_queue *q); extern void __blk_run_queue(struct request_queue *q); extern void blk_run_queue(struct request_queue *); +extern void blk_run_queue_async(struct request_queue *q); extern int blk_rq_map_user(struct request_queue *, struct request *, struct rq_map_data *, void __user *, unsigned long, gfp_t); -- cgit v0.10.2 From d350e6b6e819df0a383ff34465720bfaa0f91c79 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Tue, 19 Apr 2011 13:34:14 +0200 Subject: block: remove stale kerneldoc member from __blk_run_queue() We don't pass in a 'force_kblockd' anymore, get rid of the stsale comment. Reported-by: Mike Snitzer Signed-off-by: Jens Axboe diff --git a/block/blk-core.c b/block/blk-core.c index 40725b9..a2e58ee 100644 --- a/block/blk-core.c +++ b/block/blk-core.c @@ -292,7 +292,6 @@ EXPORT_SYMBOL(blk_sync_queue); /** * __blk_run_queue - run a single device queue * @q: The queue to run - * @force_kblockd: Don't run @q->request_fn directly. Use kblockd. * * Description: * See @blk_run_queue. This variant must be called with the queue lock -- cgit v0.10.2 From ed5302d3c25006a9edc7a7fbea97a30483f89ef7 Mon Sep 17 00:00:00 2001 From: Liu Yuan Date: Tue, 19 Apr 2011 13:47:58 +0200 Subject: block, blk-sysfs: Fix an err return path in blk_register_queue() We do not call blk_trace_remove_sysfs() in err return path if kobject_add() fails. This path fixes it. Cc: stable@kernel.org Signed-off-by: Liu Yuan Signed-off-by: Jens Axboe diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index 6d73512..5d696ad 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -508,8 +508,10 @@ int blk_register_queue(struct gendisk *disk) return ret; ret = kobject_add(&q->kobj, kobject_get(&dev->kobj), "%s", "queue"); - if (ret < 0) + if (ret < 0) { + blk_trace_remove_sysfs(dev); return ret; + } kobject_uevent(&q->kobj, KOBJ_ADD); -- cgit v0.10.2 From 60735b6362f29b52b5635a2dfa9ab5ad39948345 Mon Sep 17 00:00:00 2001 From: Tao Ma Date: Tue, 19 Apr 2011 13:50:40 +0200 Subject: block: Remove the extra check in queue_requests_store In queue_requests_store, the code looks like if (rl->count[BLK_RW_SYNC] >= q->nr_requests) { blk_set_queue_full(q, BLK_RW_SYNC); } else if (rl->count[BLK_RW_SYNC]+1 <= q->nr_requests) { blk_clear_queue_full(q, BLK_RW_SYNC); wake_up(&rl->wait[BLK_RW_SYNC]); } If we don't satify the situation of "if", we can get that rl->count[BLK_RW_SYNC} < q->nr_quests. It is the same as rl->count[BLK_RW_SYNC]+1 <= q->nr_requests. All the "else" should satisfy the "else if" check so it isn't needed actually. Signed-off-by: Tao Ma Signed-off-by: Jens Axboe diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c index 5d696ad..bd23631 100644 --- a/block/blk-sysfs.c +++ b/block/blk-sysfs.c @@ -66,14 +66,14 @@ queue_requests_store(struct request_queue *q, const char *page, size_t count) if (rl->count[BLK_RW_SYNC] >= q->nr_requests) { blk_set_queue_full(q, BLK_RW_SYNC); - } else if (rl->count[BLK_RW_SYNC]+1 <= q->nr_requests) { + } else { blk_clear_queue_full(q, BLK_RW_SYNC); wake_up(&rl->wait[BLK_RW_SYNC]); } if (rl->count[BLK_RW_ASYNC] >= q->nr_requests) { blk_set_queue_full(q, BLK_RW_ASYNC); - } else if (rl->count[BLK_RW_ASYNC]+1 <= q->nr_requests) { + } else { blk_clear_queue_full(q, BLK_RW_ASYNC); wake_up(&rl->wait[BLK_RW_ASYNC]); } -- cgit v0.10.2 From a8a8a0937e22a5fd55aeb22586724ba6bb70aadd Mon Sep 17 00:00:00 2001 From: Jozsef Kadlecsik Date: Tue, 19 Apr 2011 15:59:15 +0200 Subject: netfilter: ipset: Fix the order of listing of sets A restoreable saving of sets requires that list:set type of sets come last and the code part which should have taken into account the ordering was broken. The patch fixes the listing order. Signed-off-by: Jozsef Kadlecsik Signed-off-by: Patrick McHardy diff --git a/net/netfilter/ipset/ip_set_core.c b/net/netfilter/ipset/ip_set_core.c index e88ac3c..d87e03b 100644 --- a/net/netfilter/ipset/ip_set_core.c +++ b/net/netfilter/ipset/ip_set_core.c @@ -1022,8 +1022,9 @@ ip_set_dump_start(struct sk_buff *skb, struct netlink_callback *cb) if (cb->args[1] >= ip_set_max) goto out; - pr_debug("args[0]: %ld args[1]: %ld\n", cb->args[0], cb->args[1]); max = cb->args[0] == DUMP_ONE ? cb->args[1] + 1 : ip_set_max; +dump_last: + pr_debug("args[0]: %ld args[1]: %ld\n", cb->args[0], cb->args[1]); for (; cb->args[1] < max; cb->args[1]++) { index = (ip_set_id_t) cb->args[1]; set = ip_set_list[index]; @@ -1038,8 +1039,8 @@ ip_set_dump_start(struct sk_buff *skb, struct netlink_callback *cb) * so that lists (unions of sets) are dumped last. */ if (cb->args[0] != DUMP_ONE && - !((cb->args[0] == DUMP_ALL) ^ - (set->type->features & IPSET_DUMP_LAST))) + ((cb->args[0] == DUMP_ALL) == + !!(set->type->features & IPSET_DUMP_LAST))) continue; pr_debug("List set: %s\n", set->name); if (!cb->args[2]) { @@ -1083,6 +1084,12 @@ ip_set_dump_start(struct sk_buff *skb, struct netlink_callback *cb) goto release_refcount; } } + /* If we dump all sets, continue with dumping last ones */ + if (cb->args[0] == DUMP_ALL) { + cb->args[0] = DUMP_LAST; + cb->args[1] = 0; + goto dump_last; + } goto out; nla_put_failure: @@ -1093,11 +1100,6 @@ release_refcount: pr_debug("release set %s\n", ip_set_list[index]->name); ip_set_put_byindex(index); } - - /* If we dump all sets, continue with dumping last ones */ - if (cb->args[0] == DUMP_ALL && cb->args[1] >= max && !cb->args[2]) - cb->args[0] = DUMP_LAST; - out: if (nlh) { nlmsg_end(skb, nlh); -- cgit v0.10.2 From c3a2f0ad4917c678fcd828f16102518c33d8393c Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Sat, 2 Apr 2011 08:26:34 -0700 Subject: hwmon: Add submitting-patches checklist to documentation When writing hardware monitoring drivers, there are some common pitfalls which keep coming up in code reviews. This patch provides a document describing all those pitfalls and how to avoid them. Signed-off-by: Guenter Roeck Acked-by: Jean Delvare diff --git a/Documentation/hwmon/submitting-patches b/Documentation/hwmon/submitting-patches new file mode 100644 index 0000000..86f42e8 --- /dev/null +++ b/Documentation/hwmon/submitting-patches @@ -0,0 +1,109 @@ + How to Get Your Patch Accepted Into the Hwmon Subsystem + ------------------------------------------------------- + +This text is is a collection of suggestions for people writing patches or +drivers for the hwmon subsystem. Following these suggestions will greatly +increase the chances of your change being accepted. + + +1. General +---------- + +* It should be unnecessary to mention, but please read and follow + Documentation/SubmitChecklist + Documentation/SubmittingDrivers + Documentation/SubmittingPatches + Documentation/CodingStyle + +* If your patch generates checkpatch warnings, please refrain from explanations + such as "I don't like that coding style". Keep in mind that each unnecessary + warning helps hiding a real problem. If you don't like the kernel coding + style, don't write kernel drivers. + +* Please test your patch thoroughly. We are not your test group. + Sometimes a patch can not or not completely be tested because of missing + hardware. In such cases, you should test-build the code on at least one + architecture. If run-time testing was not achieved, it should be written + explicitly below the patch header. + +* If your patch (or the driver) is affected by configuration options such as + CONFIG_SMP or CONFIG_HOTPLUG, make sure it compiles for all configuration + variants. + + +2. Adding functionality to existing drivers +------------------------------------------- + +* Make sure the documentation in Documentation/hwmon/ is up to + date. + +* Make sure the information in Kconfig is up to date. + +* If the added functionality requires some cleanup or structural changes, split + your patch into a cleanup part and the actual addition. This makes it easier + to review your changes, and to bisect any resulting problems. + +* Never mix bug fixes, cleanup, and functional enhancements in a single patch. + + +3. New drivers +-------------- + +* Running your patch or driver file(s) through checkpatch does not mean its + formatting is clean. If unsure about formatting in your new driver, run it + through Lindent. Lindent is not perfect, and you may have to do some minor + cleanup, but it is a good start. + +* Consider adding yourself to MAINTAINERS. + +* Document the driver in Documentation/hwmon/. + +* Add the driver to Kconfig and Makefile in alphabetical order. + +* Make sure that all dependencies are listed in Kconfig. For new drivers, it + is most likely prudent to add a dependency on EXPERIMENTAL. + +* Avoid forward declarations if you can. Rearrange the code if necessary. + +* Avoid calculations in macros and macro-generated functions. While such macros + may save a line or so in the source, it obfuscates the code and makes code + review more difficult. It may also result in code which is more complicated + than necessary. Use inline functions or just regular functions instead. + +* If the driver has a detect function, make sure it is silent. Debug messages + and messages printed after a successful detection are acceptable, but it + must not print messages such as "Chip XXX not found/supported". + + Keep in mind that the detect function will run for all drivers supporting an + address if a chip is detected on that address. Unnecessary messages will just + pollute the kernel log and not provide any value. + +* Provide a detect function if and only if a chip can be detected reliably. + +* Avoid writing to chip registers in the detect function. If you have to write, + only do it after you have already gathered enough data to be certain that the + detection is going to be successful. + + Keep in mind that the chip might not be what your driver believes it is, and + writing to it might cause a bad misconfiguration. + +* Make sure there are no race conditions in the probe function. Specifically, + completely initialize your chip first, then create sysfs entries and register + with the hwmon subsystem. + +* Do not provide support for deprecated sysfs attributes. + +* Do not create non-standard attributes unless really needed. If you have to use + non-standard attributes, or you believe you do, discuss it on the mailing list + first. Either case, provide a detailed explanation why you need the + non-standard attribute(s). + Standard attributes are specified in Documentation/hwmon/sysfs-interface. + +* When deciding which sysfs attributes to support, look at the chip's + capabilities. While we do not expect your driver to support everything the + chip may offer, it should at least support all limits and alarms. + +* Last but not least, please check if a driver for your chip already exists + before starting to write a new driver. Especially for temperature sensors, + new chips are often variants of previously released chips. In some cases, + a presumably new chip may simply have been relabeled. -- cgit v0.10.2 From 3b17857dfbcb698c2dd0b4c38775ab04cf643d1a Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Mon, 14 Mar 2011 17:54:25 -0700 Subject: hwmon: (pmbus) Removed unused variable from struct pmbus_data struct pmbus_data included an unused variable named status_bits. Remove it. Signed-off-by: Guenter Roeck Reviewed-by: Tom Grennan diff --git a/drivers/hwmon/pmbus_core.c b/drivers/hwmon/pmbus_core.c index edfb92e..196ffaf 100644 --- a/drivers/hwmon/pmbus_core.c +++ b/drivers/hwmon/pmbus_core.c @@ -139,7 +139,6 @@ struct pmbus_data { * A single status register covers multiple attributes, * so we keep them all together. */ - u8 status_bits; u8 status[PB_NUM_STATUS_REG]; u8 currpage; -- cgit v0.10.2 From c1a76b47419ebb143559feaedfb35e4f285095b9 Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Mon, 18 Apr 2011 09:43:22 -0700 Subject: hwmon: (smm665) Fix spelling error in driver documentation tempererature may sound interesting, but temperature is still preferred. Signed-off-by: Guenter Roeck Acked-by: Jean Delvare diff --git a/Documentation/hwmon/smm665 b/Documentation/hwmon/smm665 index 3820fc9..59e3161 100644 --- a/Documentation/hwmon/smm665 +++ b/Documentation/hwmon/smm665 @@ -150,8 +150,8 @@ in8_crit_alarm Channel F critical alarm in9_crit_alarm AIN1 critical alarm in10_crit_alarm AIN2 critical alarm -temp1_input Chip tempererature -temp1_min Mimimum chip tempererature -temp1_max Maximum chip tempererature -temp1_crit Critical chip tempererature +temp1_input Chip temperature +temp1_min Mimimum chip temperature +temp1_max Maximum chip temperature +temp1_crit Critical chip temperature temp1_crit_alarm Temperature critical alarm -- cgit v0.10.2 From 180b3d889c85ce25d080997cc0c6cee4e46eed54 Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Mon, 18 Apr 2011 09:48:58 -0700 Subject: hwmon: (pmbus) Documentation updates Fix spelling, correct label name error, and add missing attribute to PMBus driver documentation. Signed-off-by: Guenter Roeck Reviewed-by: Tom Grennan diff --git a/Documentation/hwmon/pmbus b/Documentation/hwmon/pmbus index dc4933e..852d9eb 100644 --- a/Documentation/hwmon/pmbus +++ b/Documentation/hwmon/pmbus @@ -175,11 +175,13 @@ currX_crit Critical maximum current. From IIN_OC_FAULT_LIMIT or IOUT_OC_FAULT_LIMIT register. currX_alarm Current high alarm. From IIN_OC_WARNING or IOUT_OC_WARNING status. +currX_max_alarm Current high alarm. + From IIN_OC_WARN_LIMIT or IOUT_OC_WARN_LIMIT status. currX_lcrit_alarm Output current critical low alarm. From IOUT_UC_FAULT status. currX_crit_alarm Current critical high alarm. From IIN_OC_FAULT or IOUT_OC_FAULT status. -currX_label "iin" or "vinY" +currX_label "iin" or "ioutY" powerX_input Measured power. From READ_PIN or READ_POUT register. powerX_cap Output power cap. From POUT_MAX register. @@ -193,13 +195,13 @@ powerX_crit_alarm Output power critical high alarm. From POUT_OP_FAULT status. powerX_label "pin" or "poutY" -tempX_input Measured tempererature. +tempX_input Measured temperature. From READ_TEMPERATURE_X register. -tempX_min Mimimum tempererature. From UT_WARN_LIMIT register. -tempX_max Maximum tempererature. From OT_WARN_LIMIT register. -tempX_lcrit Critical low tempererature. +tempX_min Mimimum temperature. From UT_WARN_LIMIT register. +tempX_max Maximum temperature. From OT_WARN_LIMIT register. +tempX_lcrit Critical low temperature. From UT_FAULT_LIMIT register. -tempX_crit Critical high tempererature. +tempX_crit Critical high temperature. From OT_FAULT_LIMIT register. tempX_min_alarm Chip temperature low alarm. Set by comparing READ_TEMPERATURE_X with UT_WARN_LIMIT if -- cgit v0.10.2 From 2669d9f542b7eecb327d2bd3b7ee1d9eb46307b4 Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Mon, 18 Apr 2011 09:51:04 -0700 Subject: hwmon: (max8688) Add driver documentation MAX8688 has its own driver, thus should have its own documentation instead of being mentioned in the generic PMBus driver documentation. Signed-off-by: Guenter Roeck Reviewed-by: Tom Grennan diff --git a/Documentation/hwmon/max8688 b/Documentation/hwmon/max8688 new file mode 100644 index 0000000..0ddd3a4 --- /dev/null +++ b/Documentation/hwmon/max8688 @@ -0,0 +1,69 @@ +Kernel driver max8688 +===================== + +Supported chips: + * Maxim MAX8688 + Prefix: 'max8688' + Addresses scanned: - + Datasheet: http://datasheets.maxim-ic.com/en/ds/MAX8688.pdf + +Author: Guenter Roeck + + +Description +----------- + +This driver supports hardware montoring for Maxim MAX8688 Digital Power-Supply +Controller/Monitor with PMBus Interface. + +The driver is a client driver to the core PMBus driver. Please see +Documentation/hwmon/pmbus for details on PMBus client drivers. + + +Usage Notes +----------- + +This driver does not auto-detect devices. You will have to instantiate the +devices explicitly. Please see Documentation/i2c/instantiating-devices for +details. + + +Platform data support +--------------------- + +The driver supports standard PMBus driver platform data. + + +Sysfs entries +------------- + +The following attributes are supported. Limits are read-write; all other +attributes are read-only. + +in1_label "vout1" +in1_input Measured voltage. From READ_VOUT register. +in1_min Minumum Voltage. From VOUT_UV_WARN_LIMIT register. +in1_max Maximum voltage. From VOUT_OV_WARN_LIMIT register. +in1_lcrit Critical minumum Voltage. VOUT_UV_FAULT_LIMIT register. +in1_crit Critical maximum voltage. From VOUT_OV_FAULT_LIMIT register. +in1_min_alarm Voltage low alarm. From VOLTAGE_UV_WARNING status. +in1_max_alarm Voltage high alarm. From VOLTAGE_OV_WARNING status. +in1_lcrit_alarm Voltage critical low alarm. From VOLTAGE_UV_FAULT status. +in1_crit_alarm Voltage critical high alarm. From VOLTAGE_OV_FAULT status. + +curr1_label "iout1" +curr1_input Measured current. From READ_IOUT register. +curr1_max Maximum current. From IOUT_OC_WARN_LIMIT register. +curr1_crit Critical maximum current. From IOUT_OC_FAULT_LIMIT register. +curr1_max_alarm Current high alarm. From IOUT_OC_WARN_LIMIT register. +curr1_crit_alarm Current critical high alarm. From IOUT_OC_FAULT status. + +temp1_input Measured temperature. From READ_TEMPERATURE_1 register. +temp1_max Maximum temperature. From OT_WARN_LIMIT register. +temp1_crit Critical high temperature. From OT_FAULT_LIMIT register. +temp1_max_alarm Chip temperature high alarm. Set by comparing + READ_TEMPERATURE_1 with OT_WARN_LIMIT if TEMP_OT_WARNING + status is set. +temp1_crit_alarm Chip temperature critical high alarm. Set by comparing + READ_TEMPERATURE_1 with OT_FAULT_LIMIT if TEMP_OT_FAULT + status is set. diff --git a/Documentation/hwmon/pmbus b/Documentation/hwmon/pmbus index 852d9eb..0740a6c 100644 --- a/Documentation/hwmon/pmbus +++ b/Documentation/hwmon/pmbus @@ -28,11 +28,6 @@ Supported chips: Prefixes: 'max34441' Addresses scanned: - Datasheet: http://datasheets.maxim-ic.com/en/ds/MAX34441.pdf - * Maxim MAX8688 - Digital Power-Supply Controller/Monitor - Prefix: 'max8688' - Addresses scanned: - - Datasheet: http://datasheets.maxim-ic.com/en/ds/MAX8688.pdf * Generic PMBus devices Prefix: 'pmbus' Addresses scanned: - -- cgit v0.10.2 From 4af33f1726b03997f70180466769cf28ddcaef85 Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Mon, 18 Apr 2011 09:53:54 -0700 Subject: hwmon: (max16064) Add driver documentation MAX16064 has its own driver, thus should have its own documentation instead of being mentioned in the generic PMBus driver documentation. Signed-off-by: Guenter Roeck Reviewed-by: Tom Grennan diff --git a/Documentation/hwmon/max16064 b/Documentation/hwmon/max16064 new file mode 100644 index 0000000..4172899 --- /dev/null +++ b/Documentation/hwmon/max16064 @@ -0,0 +1,62 @@ +Kernel driver max16064 +====================== + +Supported chips: + * Maxim MAX16064 + Prefix: 'max16064' + Addresses scanned: - + Datasheet: http://datasheets.maxim-ic.com/en/ds/MAX16064.pdf + +Author: Guenter Roeck + + +Description +----------- + +This driver supports hardware montoring for Maxim MAX16064 Quad Power-Supply +Controller with Active-Voltage Output Control and PMBus Interface. + +The driver is a client driver to the core PMBus driver. +Please see Documentation/hwmon/pmbus for details on PMBus client drivers. + + +Usage Notes +----------- + +This driver does not auto-detect devices. You will have to instantiate the +devices explicitly. Please see Documentation/i2c/instantiating-devices for +details. + + +Platform data support +--------------------- + +The driver supports standard PMBus driver platform data. + + +Sysfs entries +------------- + +The following attributes are supported. Limits are read-write; all other +attributes are read-only. + +in[1-4]_label "vout[1-4]" +in[1-4]_input Measured voltage. From READ_VOUT register. +in[1-4]_min Minumum Voltage. From VOUT_UV_WARN_LIMIT register. +in[1-4]_max Maximum voltage. From VOUT_OV_WARN_LIMIT register. +in[1-4]_lcrit Critical minumum Voltage. VOUT_UV_FAULT_LIMIT register. +in[1-4]_crit Critical maximum voltage. From VOUT_OV_FAULT_LIMIT register. +in[1-4]_min_alarm Voltage low alarm. From VOLTAGE_UV_WARNING status. +in[1-4]_max_alarm Voltage high alarm. From VOLTAGE_OV_WARNING status. +in[1-4]_lcrit_alarm Voltage critical low alarm. From VOLTAGE_UV_FAULT status. +in[1-4]_crit_alarm Voltage critical high alarm. From VOLTAGE_OV_FAULT status. + +temp1_input Measured temperature. From READ_TEMPERATURE_1 register. +temp1_max Maximum temperature. From OT_WARN_LIMIT register. +temp1_crit Critical high temperature. From OT_FAULT_LIMIT register. +temp1_max_alarm Chip temperature high alarm. Set by comparing + READ_TEMPERATURE_1 with OT_WARN_LIMIT if TEMP_OT_WARNING + status is set. +temp1_crit_alarm Chip temperature critical high alarm. Set by comparing + READ_TEMPERATURE_1 with OT_FAULT_LIMIT if TEMP_OT_FAULT + status is set. diff --git a/Documentation/hwmon/pmbus b/Documentation/hwmon/pmbus index 0740a6c..e3c527a 100644 --- a/Documentation/hwmon/pmbus +++ b/Documentation/hwmon/pmbus @@ -13,11 +13,6 @@ Supported chips: Prefix: 'ltc2978' Addresses scanned: - Datasheet: http://cds.linear.com/docs/Datasheet/2978fa.pdf - * Maxim MAX16064 - Quad Power-Supply Controller - Prefix: 'max16064' - Addresses scanned: - - Datasheet: http://datasheets.maxim-ic.com/en/ds/MAX16064.pdf * Maxim MAX34440 PMBus 6-Channel Power-Supply Manager Prefixes: 'max34440' -- cgit v0.10.2 From e428d8d3bd164ad36fb545b2162bdfcb8100dc08 Mon Sep 17 00:00:00 2001 From: Guenter Roeck Date: Mon, 18 Apr 2011 09:55:59 -0700 Subject: hwmon: (max34440) Add driver documentation MAX34440 and MAX34441 have their own driver, thus there should be explicit documentation instead of mentioning the chips in the generic PMBus driver documentation. Signed-off-by: Guenter Roeck Reviewed-by: Tom Grennan diff --git a/Documentation/hwmon/max34440 b/Documentation/hwmon/max34440 new file mode 100644 index 0000000..6c525dd --- /dev/null +++ b/Documentation/hwmon/max34440 @@ -0,0 +1,79 @@ +Kernel driver max34440 +====================== + +Supported chips: + * Maxim MAX34440 + Prefixes: 'max34440' + Addresses scanned: - + Datasheet: http://datasheets.maxim-ic.com/en/ds/MAX34440.pdf + * Maxim MAX34441 + PMBus 5-Channel Power-Supply Manager and Intelligent Fan Controller + Prefixes: 'max34441' + Addresses scanned: - + Datasheet: http://datasheets.maxim-ic.com/en/ds/MAX34441.pdf + +Author: Guenter Roeck + + +Description +----------- + +This driver supports hardware montoring for Maxim MAX34440 PMBus 6-Channel +Power-Supply Manager and MAX34441 PMBus 5-Channel Power-Supply Manager +and Intelligent Fan Controller. + +The driver is a client driver to the core PMBus driver. Please see +Documentation/hwmon/pmbus for details on PMBus client drivers. + + +Usage Notes +----------- + +This driver does not auto-detect devices. You will have to instantiate the +devices explicitly. Please see Documentation/i2c/instantiating-devices for +details. + + +Platform data support +--------------------- + +The driver supports standard PMBus driver platform data. + + +Sysfs entries +------------- + +The following attributes are supported. Limits are read-write; all other +attributes are read-only. + +in[1-6]_label "vout[1-6]". +in[1-6]_input Measured voltage. From READ_VOUT register. +in[1-6]_min Minumum Voltage. From VOUT_UV_WARN_LIMIT register. +in[1-6]_max Maximum voltage. From VOUT_OV_WARN_LIMIT register. +in[1-6]_lcrit Critical minumum Voltage. VOUT_UV_FAULT_LIMIT register. +in[1-6]_crit Critical maximum voltage. From VOUT_OV_FAULT_LIMIT register. +in[1-6]_min_alarm Voltage low alarm. From VOLTAGE_UV_WARNING status. +in[1-6]_max_alarm Voltage high alarm. From VOLTAGE_OV_WARNING status. +in[1-6]_lcrit_alarm Voltage critical low alarm. From VOLTAGE_UV_FAULT status. +in[1-6]_crit_alarm Voltage critical high alarm. From VOLTAGE_OV_FAULT status. + +curr[1-6]_label "iout[1-6]". +curr[1-6]_input Measured current. From READ_IOUT register. +curr[1-6]_max Maximum current. From IOUT_OC_WARN_LIMIT register. +curr[1-6]_crit Critical maximum current. From IOUT_OC_FAULT_LIMIT register. +curr[1-6]_max_alarm Current high alarm. From IOUT_OC_WARNING status. +curr[1-6]_crit_alarm Current critical high alarm. From IOUT_OC_FAULT status. + + in6 and curr6 attributes only exist for MAX34440. + +temp[1-8]_input Measured temperatures. From READ_TEMPERATURE_1 register. + temp1 is the chip's internal temperature. temp2..temp5 + are remote I2C temperature sensors. For MAX34441, temp6 + is a remote thermal-diode sensor. For MAX34440, temp6..8 + are remote I2C temperature sensors. +temp[1-8]_max Maximum temperature. From OT_WARN_LIMIT register. +temp[1-8]_crit Critical high temperature. From OT_FAULT_LIMIT register. +temp[1-8]_max_alarm Temperature high alarm. +temp[1-8]_crit_alarm Temperature critical high alarm. + + temp7 and temp8 attributes only exist for MAX34440. diff --git a/Documentation/hwmon/pmbus b/Documentation/hwmon/pmbus index e3c527a..5e462fc 100644 --- a/Documentation/hwmon/pmbus +++ b/Documentation/hwmon/pmbus @@ -13,16 +13,6 @@ Supported chips: Prefix: 'ltc2978' Addresses scanned: - Datasheet: http://cds.linear.com/docs/Datasheet/2978fa.pdf - * Maxim MAX34440 - PMBus 6-Channel Power-Supply Manager - Prefixes: 'max34440' - Addresses scanned: - - Datasheet: http://datasheets.maxim-ic.com/en/ds/MAX34440.pdf - * Maxim MAX34441 - PMBus 5-Channel Power-Supply Manager and Intelligent Fan Controller - Prefixes: 'max34441' - Addresses scanned: - - Datasheet: http://datasheets.maxim-ic.com/en/ds/MAX34441.pdf * Generic PMBus devices Prefix: 'pmbus' Addresses scanned: - -- cgit v0.10.2 From 71460af58f8565110160283849db4d6bf7e1efa1 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 19 Apr 2011 10:54:44 -0700 Subject: Revert "[media] V4L: videobuf, don't use dma addr as physical" This reverts commit 35d9f510b67b10338161aba6229d4f55b4000f5b. Quoth Jiri Slaby: "It fixes mmap when IOMMU is used on x86 only, but breaks architectures like ARM or PPC where virt_to_phys(dma_alloc_coherent) doesn't work. We need there dma_mmap_coherent or similar (the trickery what snd_pcm_default_mmap does but in some saner way). But this cannot be done at this phase." Requested-by: Jiri Slaby Cc: Russell King - ARM Linux Cc: Janusz Krzysztofik Acked-by: Mauro Carvalho Chehab Signed-off-by: Linus Torvalds diff --git a/drivers/media/video/videobuf-dma-contig.c b/drivers/media/video/videobuf-dma-contig.c index c4742fc..c969111 100644 --- a/drivers/media/video/videobuf-dma-contig.c +++ b/drivers/media/video/videobuf-dma-contig.c @@ -300,7 +300,7 @@ static int __videobuf_mmap_mapper(struct videobuf_queue *q, vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); retval = remap_pfn_range(vma, vma->vm_start, - PFN_DOWN(virt_to_phys(mem->vaddr)), + mem->dma_handle >> PAGE_SHIFT, size, vma->vm_page_prot); if (retval) { dev_err(q->dev, "mmap: remap failed with error %d. ", retval); -- cgit v0.10.2 From a96e5b90804be8b540d30f4a1453fc87f95b3149 Mon Sep 17 00:00:00 2001 From: OGAWA Hirofumi Date: Mon, 18 Apr 2011 11:48:55 -0400 Subject: nfsd4: Fix filp leak 23fcf2ec93fb8573a653408316af599939ff9a8e (nfsd4: fix oops on lock failure) The above patch breaks free path for stp->st_file. If stp was inserted into sop->so_stateids, we have to free stp->st_file refcount. Because stp->st_file refcount itself is taken whether or not any refcounts are taken on the stp->st_file->fi_fds[]. Signed-off-by: OGAWA Hirofumi Cc: stable@kernel.org Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index c79a983..4cf04e1 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -403,8 +403,8 @@ static void free_generic_stateid(struct nfs4_stateid *stp) if (stp->st_access_bmap) { oflag = nfs4_access_bmap_to_omode(stp); nfs4_file_put_access(stp->st_file, oflag); - put_nfs4_file(stp->st_file); } + put_nfs4_file(stp->st_file); kmem_cache_free(stateid_slab, stp); } -- cgit v0.10.2 From 04eb34a43ce5168e05e2748bd46a62a09289cdde Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Wed, 6 Apr 2011 13:28:35 +1000 Subject: drm/nouveau: split ramin_lock into two locks, one hardirq safe Fixes a possible lock ordering reversal between context_switch_lock and ramin_lock. Signed-off-by: Ben Skeggs Reviewed-by: Francisco Jerez diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h index 856d56a..a76514a 100644 --- a/drivers/gpu/drm/nouveau/nouveau_drv.h +++ b/drivers/gpu/drm/nouveau/nouveau_drv.h @@ -682,6 +682,9 @@ struct drm_nouveau_private { /* For PFIFO and PGRAPH. */ spinlock_t context_switch_lock; + /* VM/PRAMIN flush, legacy PRAMIN aperture */ + spinlock_t vm_lock; + /* RAMIN configuration, RAMFC, RAMHT and RAMRO offsets */ struct nouveau_ramht *ramht; struct nouveau_gpuobj *ramfc; diff --git a/drivers/gpu/drm/nouveau/nouveau_object.c b/drivers/gpu/drm/nouveau/nouveau_object.c index 4f00c87..67a16e0 100644 --- a/drivers/gpu/drm/nouveau/nouveau_object.c +++ b/drivers/gpu/drm/nouveau/nouveau_object.c @@ -1039,19 +1039,20 @@ nv_ro32(struct nouveau_gpuobj *gpuobj, u32 offset) { struct drm_nouveau_private *dev_priv = gpuobj->dev->dev_private; struct drm_device *dev = gpuobj->dev; + unsigned long flags; if (gpuobj->pinst == ~0 || !dev_priv->ramin_available) { u64 ptr = gpuobj->vinst + offset; u32 base = ptr >> 16; u32 val; - spin_lock(&dev_priv->ramin_lock); + spin_lock_irqsave(&dev_priv->vm_lock, flags); if (dev_priv->ramin_base != base) { dev_priv->ramin_base = base; nv_wr32(dev, 0x001700, dev_priv->ramin_base); } val = nv_rd32(dev, 0x700000 + (ptr & 0xffff)); - spin_unlock(&dev_priv->ramin_lock); + spin_unlock_irqrestore(&dev_priv->vm_lock, flags); return val; } @@ -1063,18 +1064,19 @@ nv_wo32(struct nouveau_gpuobj *gpuobj, u32 offset, u32 val) { struct drm_nouveau_private *dev_priv = gpuobj->dev->dev_private; struct drm_device *dev = gpuobj->dev; + unsigned long flags; if (gpuobj->pinst == ~0 || !dev_priv->ramin_available) { u64 ptr = gpuobj->vinst + offset; u32 base = ptr >> 16; - spin_lock(&dev_priv->ramin_lock); + spin_lock_irqsave(&dev_priv->vm_lock, flags); if (dev_priv->ramin_base != base) { dev_priv->ramin_base = base; nv_wr32(dev, 0x001700, dev_priv->ramin_base); } nv_wr32(dev, 0x700000 + (ptr & 0xffff), val); - spin_unlock(&dev_priv->ramin_lock); + spin_unlock_irqrestore(&dev_priv->vm_lock, flags); return; } diff --git a/drivers/gpu/drm/nouveau/nouveau_state.c b/drivers/gpu/drm/nouveau/nouveau_state.c index 6e2b1a6..a30adec 100644 --- a/drivers/gpu/drm/nouveau/nouveau_state.c +++ b/drivers/gpu/drm/nouveau/nouveau_state.c @@ -608,6 +608,7 @@ nouveau_card_init(struct drm_device *dev) spin_lock_init(&dev_priv->channels.lock); spin_lock_init(&dev_priv->tile.lock); spin_lock_init(&dev_priv->context_switch_lock); + spin_lock_init(&dev_priv->vm_lock); /* Make the CRTCs and I2C buses accessible */ ret = engine->display.early_init(dev); diff --git a/drivers/gpu/drm/nouveau/nv50_instmem.c b/drivers/gpu/drm/nouveau/nv50_instmem.c index a6f8aa6..4f95a1e 100644 --- a/drivers/gpu/drm/nouveau/nv50_instmem.c +++ b/drivers/gpu/drm/nouveau/nv50_instmem.c @@ -404,23 +404,25 @@ void nv50_instmem_flush(struct drm_device *dev) { struct drm_nouveau_private *dev_priv = dev->dev_private; + unsigned long flags; - spin_lock(&dev_priv->ramin_lock); + spin_lock_irqsave(&dev_priv->vm_lock, flags); nv_wr32(dev, 0x00330c, 0x00000001); if (!nv_wait(dev, 0x00330c, 0x00000002, 0x00000000)) NV_ERROR(dev, "PRAMIN flush timeout\n"); - spin_unlock(&dev_priv->ramin_lock); + spin_unlock_irqrestore(&dev_priv->vm_lock, flags); } void nv84_instmem_flush(struct drm_device *dev) { struct drm_nouveau_private *dev_priv = dev->dev_private; + unsigned long flags; - spin_lock(&dev_priv->ramin_lock); + spin_lock_irqsave(&dev_priv->vm_lock, flags); nv_wr32(dev, 0x070000, 0x00000001); if (!nv_wait(dev, 0x070000, 0x00000002, 0x00000000)) NV_ERROR(dev, "PRAMIN flush timeout\n"); - spin_unlock(&dev_priv->ramin_lock); + spin_unlock_irqrestore(&dev_priv->vm_lock, flags); } diff --git a/drivers/gpu/drm/nouveau/nv50_vm.c b/drivers/gpu/drm/nouveau/nv50_vm.c index 4fd3432..6c26944 100644 --- a/drivers/gpu/drm/nouveau/nv50_vm.c +++ b/drivers/gpu/drm/nouveau/nv50_vm.c @@ -174,10 +174,11 @@ void nv50_vm_flush_engine(struct drm_device *dev, int engine) { struct drm_nouveau_private *dev_priv = dev->dev_private; + unsigned long flags; - spin_lock(&dev_priv->ramin_lock); + spin_lock_irqsave(&dev_priv->vm_lock, flags); nv_wr32(dev, 0x100c80, (engine << 16) | 1); if (!nv_wait(dev, 0x100c80, 0x00000001, 0x00000000)) NV_ERROR(dev, "vm flush timeout: engine %d\n", engine); - spin_unlock(&dev_priv->ramin_lock); + spin_unlock_irqrestore(&dev_priv->vm_lock, flags); } diff --git a/drivers/gpu/drm/nouveau/nvc0_vm.c b/drivers/gpu/drm/nouveau/nvc0_vm.c index a0a2a02..a179e6c 100644 --- a/drivers/gpu/drm/nouveau/nvc0_vm.c +++ b/drivers/gpu/drm/nouveau/nvc0_vm.c @@ -104,11 +104,12 @@ nvc0_vm_flush(struct nouveau_vm *vm) struct nouveau_instmem_engine *pinstmem = &dev_priv->engine.instmem; struct drm_device *dev = vm->dev; struct nouveau_vm_pgd *vpgd; + unsigned long flags; u32 engine = (dev_priv->chan_vm == vm) ? 1 : 5; pinstmem->flush(vm->dev); - spin_lock(&dev_priv->ramin_lock); + spin_lock_irqsave(&dev_priv->vm_lock, flags); list_for_each_entry(vpgd, &vm->pgd_list, head) { /* looks like maybe a "free flush slots" counter, the * faster you write to 0x100cbc to more it decreases @@ -125,5 +126,5 @@ nvc0_vm_flush(struct nouveau_vm *vm) nv_rd32(dev, 0x100c80), engine); } } - spin_unlock(&dev_priv->ramin_lock); + spin_unlock_irqrestore(&dev_priv->vm_lock, flags); } -- cgit v0.10.2 From 01d153326ecd81ed77395f1699a30af416cf77ab Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Fri, 8 Apr 2011 10:07:34 +1000 Subject: drm/nouveau: fix nv30 pcie boards Wasn't aware they even existed, apparently they do! They're actually AGP chips with a bridge as far as I can tell, which puts them in the same boat as nv40/nv45. Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nouveau_mem.c b/drivers/gpu/drm/nouveau/nouveau_mem.c index 78f467f..5045f8b 100644 --- a/drivers/gpu/drm/nouveau/nouveau_mem.c +++ b/drivers/gpu/drm/nouveau/nouveau_mem.c @@ -398,7 +398,7 @@ nouveau_mem_vram_init(struct drm_device *dev) dma_bits = 40; } else if (drm_pci_device_is_pcie(dev) && - dev_priv->chipset != 0x40 && + dev_priv->chipset > 0x40 && dev_priv->chipset != 0x45) { if (pci_dma_supported(dev->pdev, DMA_BIT_MASK(39))) dma_bits = 39; diff --git a/drivers/gpu/drm/nouveau/nouveau_sgdma.c b/drivers/gpu/drm/nouveau/nouveau_sgdma.c index a33fe40..fc29244 100644 --- a/drivers/gpu/drm/nouveau/nouveau_sgdma.c +++ b/drivers/gpu/drm/nouveau/nouveau_sgdma.c @@ -427,7 +427,7 @@ nouveau_sgdma_init(struct drm_device *dev) u32 aper_size, align; int ret; - if (dev_priv->card_type >= NV_50 || drm_pci_device_is_pcie(dev)) + if (dev_priv->card_type >= NV_40 && drm_pci_device_is_pcie(dev)) aper_size = 512 * 1024 * 1024; else aper_size = 64 * 1024 * 1024; @@ -457,7 +457,7 @@ nouveau_sgdma_init(struct drm_device *dev) dev_priv->gart_info.func = &nv50_sgdma_backend; } else if (drm_pci_device_is_pcie(dev) && - dev_priv->chipset != 0x40 && dev_priv->chipset != 0x45) { + dev_priv->chipset > 0x40 && dev_priv->chipset != 0x45) { if (nv44_graph_class(dev)) { dev_priv->gart_info.func = &nv44_sgdma_backend; align = 512 * 1024; -- cgit v0.10.2 From 8706398bf8841868d4e56fc924a5edcd3f156243 Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Mon, 11 Apr 2011 16:37:44 +1000 Subject: drm/nouveau: populate ttm_alloced with false, when it's not Caught with kmemcheck on unrelated business. Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nouveau_sgdma.c b/drivers/gpu/drm/nouveau/nouveau_sgdma.c index fc29244..4bce801 100644 --- a/drivers/gpu/drm/nouveau/nouveau_sgdma.c +++ b/drivers/gpu/drm/nouveau/nouveau_sgdma.c @@ -55,6 +55,7 @@ nouveau_sgdma_populate(struct ttm_backend *be, unsigned long num_pages, be->func->clear(be); return -EFAULT; } + nvbe->ttm_alloced[nvbe->nr_pages] = false; } nvbe->nr_pages++; -- cgit v0.10.2 From 11dea1a2144f24216551fbeddacbde0980ae8a55 Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Mon, 18 Apr 2011 09:12:25 +1000 Subject: drm/nouveau: fix pinning of notifier block Problem introduced with commit 6ba9a68317781537d6184d3fdb2d0f20c97da3a4 Reported-by: Bob Gleitsmann Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nouveau_notifier.c b/drivers/gpu/drm/nouveau/nouveau_notifier.c index 7ba3fc0..5b39718 100644 --- a/drivers/gpu/drm/nouveau/nouveau_notifier.c +++ b/drivers/gpu/drm/nouveau/nouveau_notifier.c @@ -35,19 +35,22 @@ nouveau_notifier_init_channel(struct nouveau_channel *chan) { struct drm_device *dev = chan->dev; struct nouveau_bo *ntfy = NULL; - uint32_t flags; + uint32_t flags, ttmpl; int ret; - if (nouveau_vram_notify) + if (nouveau_vram_notify) { flags = NOUVEAU_GEM_DOMAIN_VRAM; - else + ttmpl = TTM_PL_FLAG_VRAM; + } else { flags = NOUVEAU_GEM_DOMAIN_GART; + ttmpl = TTM_PL_FLAG_TT; + } ret = nouveau_gem_new(dev, NULL, PAGE_SIZE, 0, flags, 0, 0, &ntfy); if (ret) return ret; - ret = nouveau_bo_pin(ntfy, flags); + ret = nouveau_bo_pin(ntfy, ttmpl); if (ret) goto out_err; -- cgit v0.10.2 From a18d89ca026140eb8ac4459bf70a01c571dd9a32 Mon Sep 17 00:00:00 2001 From: Marcin Slusarz Date: Tue, 19 Apr 2011 23:50:48 +0200 Subject: drm/nouveau: fix notifier memory corruption bug nouveau_bo_wr32 expects offset to be in words, but we pass value in bytes, so after commit 73412c3854c877e5f37ad944ee8977addde4d35a ("drm/nouveau: allocate kernel's notifier object at end of block") we started to overwrite some memory after notifier buffer object (previously m2mf_ntfy was always 0, so it didn't matter it was a value in bytes). Reported-by: Dominik Brodowski Reported-by: Nigel Cunningham Signed-off-by: Marcin Slusarz Cc: Ben Skeggs Cc: Pekka Paalanen Cc: stable@kernel.org [2.6.38] Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nouveau_fbcon.c b/drivers/gpu/drm/nouveau/nouveau_fbcon.c index 889c445..39aee6d 100644 --- a/drivers/gpu/drm/nouveau/nouveau_fbcon.c +++ b/drivers/gpu/drm/nouveau/nouveau_fbcon.c @@ -181,13 +181,13 @@ nouveau_fbcon_sync(struct fb_info *info) OUT_RING (chan, 0); } - nouveau_bo_wr32(chan->notifier_bo, chan->m2mf_ntfy + 3, 0xffffffff); + nouveau_bo_wr32(chan->notifier_bo, chan->m2mf_ntfy/4 + 3, 0xffffffff); FIRE_RING(chan); mutex_unlock(&chan->mutex); ret = -EBUSY; for (i = 0; i < 100000; i++) { - if (!nouveau_bo_rd32(chan->notifier_bo, chan->m2mf_ntfy + 3)) { + if (!nouveau_bo_rd32(chan->notifier_bo, chan->m2mf_ntfy/4 + 3)) { ret = 0; break; } -- cgit v0.10.2 From e4ac93bf3c05bbe9fed1498a1461a8cdaf4b944d Mon Sep 17 00:00:00 2001 From: Marcin Slusarz Date: Tue, 19 Apr 2011 23:52:42 +0200 Subject: drm/nouveau: fix allocation of notifier object Commit 73412c3854c877e5f37ad944ee8977addde4d35a ("drm/nouveau: allocate kernel's notifier object at end of block") intended to align end of notifier block to page boundary, but start of block was miscalculated to be off by -16 bytes. Fix it. Signed-off-by: Marcin Slusarz Cc: Ben Skeggs Signed-off-by: Ben Skeggs diff --git a/drivers/gpu/drm/nouveau/nouveau_dma.c b/drivers/gpu/drm/nouveau/nouveau_dma.c index ce38e97..568caed 100644 --- a/drivers/gpu/drm/nouveau/nouveau_dma.c +++ b/drivers/gpu/drm/nouveau/nouveau_dma.c @@ -83,7 +83,7 @@ nouveau_dma_init(struct nouveau_channel *chan) return ret; /* NV_MEMORY_TO_MEMORY_FORMAT requires a notifier object */ - ret = nouveau_notifier_alloc(chan, NvNotify0, 32, 0xfd0, 0x1000, + ret = nouveau_notifier_alloc(chan, NvNotify0, 32, 0xfe0, 0x1000, &chan->m2mf_ntfy); if (ret) return ret; -- cgit v0.10.2 From 5785e53ffa73f77fb19e378c899027afc07272bc Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 19 Apr 2011 15:24:59 -0400 Subject: drm/radeon/kms: pll tweaks for r7xx Prefer min m to max p only on pre-r7xx asics. Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=36197 Signed-off-by: Alex Deucher Cc: stable@kernel.org Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c index 9d516a8..529a3a7 100644 --- a/drivers/gpu/drm/radeon/atombios_crtc.c +++ b/drivers/gpu/drm/radeon/atombios_crtc.c @@ -532,10 +532,7 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, else pll->flags |= RADEON_PLL_PREFER_LOW_REF_DIV; - if ((rdev->family == CHIP_R600) || - (rdev->family == CHIP_RV610) || - (rdev->family == CHIP_RV630) || - (rdev->family == CHIP_RV670)) + if (rdev->family < CHIP_RV770) pll->flags |= RADEON_PLL_PREFER_MINM_OVER_MAXP; } else { pll->flags |= RADEON_PLL_LEGACY; @@ -565,7 +562,6 @@ static u32 atombios_adjust_pll(struct drm_crtc *crtc, if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) { if (ss_enabled) { if (ss->refdiv) { - pll->flags |= RADEON_PLL_PREFER_MINM_OVER_MAXP; pll->flags |= RADEON_PLL_USE_REF_DIV; pll->reference_div = ss->refdiv; if (ASIC_IS_AVIVO(rdev)) -- cgit v0.10.2 From 204ae24dc71f794aaad4df48a0083dfedc711afb Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?C=C3=A9dric=20Cano?= Date: Tue, 19 Apr 2011 11:07:13 -0400 Subject: drm/radeon/kms: fix IH writeback on r6xx+ on big endian machines MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit agd5f: fix commit message. Signed-off-by: Cedric Cano Reviewed-by: Michel Dänzer Signed-off-by: Alex Deucher Signed-off-by: Dave Airlie diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c index 43fd016..e9bc135 100644 --- a/drivers/gpu/drm/radeon/evergreen.c +++ b/drivers/gpu/drm/radeon/evergreen.c @@ -2580,7 +2580,7 @@ static inline u32 evergreen_get_ih_wptr(struct radeon_device *rdev) u32 wptr, tmp; if (rdev->wb.enabled) - wptr = rdev->wb.wb[R600_WB_IH_WPTR_OFFSET/4]; + wptr = le32_to_cpu(rdev->wb.wb[R600_WB_IH_WPTR_OFFSET/4]); else wptr = RREG32(IH_RB_WPTR); diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c index 15d5829..6f27593 100644 --- a/drivers/gpu/drm/radeon/r600.c +++ b/drivers/gpu/drm/radeon/r600.c @@ -3231,7 +3231,7 @@ static inline u32 r600_get_ih_wptr(struct radeon_device *rdev) u32 wptr, tmp; if (rdev->wb.enabled) - wptr = rdev->wb.wb[R600_WB_IH_WPTR_OFFSET/4]; + wptr = le32_to_cpu(rdev->wb.wb[R600_WB_IH_WPTR_OFFSET/4]); else wptr = RREG32(IH_RB_WPTR); -- cgit v0.10.2 From 2b7da309ffe602d222558cee4d7e407b96e34b3a Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Wed, 20 Apr 2011 15:38:07 +1000 Subject: md/raid5: remove setting of ->queue_lock We previously needed to set ->queue_lock to match the raid5 device_lock so we could safely use queue_flag_* operations (e.g. for plugging). which test the ->queue_lock is in fact locked. However that need has completely gone away and is unlikely to come back to remove this now-pointless setting. Signed-off-by: NeilBrown diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index f301e6a..179bdfc 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -5151,7 +5151,6 @@ static int run(mddev_t *mddev) mddev->queue->backing_dev_info.congested_data = mddev; mddev->queue->backing_dev_info.congested_fn = raid5_congested; - mddev->queue->queue_lock = &conf->device_lock; chunk_size = mddev->chunk_sectors << 9; blk_queue_io_min(mddev->queue, chunk_size); -- cgit v0.10.2 From 3b71bd9337b404baab5c894e066be6b6bf51b1c3 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Wed, 20 Apr 2011 15:38:18 +1000 Subject: md: Fix dev_sectors on takeover from raid0 to raid4/5 A raid0 array doesn't set 'dev_sectors' as each device might contribute a different number of sectors. So when converting to a RAID4 or RAID5 we need to set dev_sectors as they need the number. We have already verified that in fact all devices do contribute the same number of sectors, so use that number. Signed-off-by: NeilBrown diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index 179bdfc..fd50011 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -5678,6 +5678,7 @@ static void raid5_quiesce(mddev_t *mddev, int state) static void *raid45_takeover_raid0(mddev_t *mddev, int level) { struct raid0_private_data *raid0_priv = mddev->private; + unsigned long long sectors; /* for raid0 takeover only one zone is supported */ if (raid0_priv->nr_strip_zones > 1) { @@ -5686,6 +5687,9 @@ static void *raid45_takeover_raid0(mddev_t *mddev, int level) return ERR_PTR(-EINVAL); } + sectors = raid0_priv->strip_zone[0].zone_end; + sector_div(sectors, raid0_priv->strip_zone[0].nb_dev); + mddev->dev_sectors = sectors; mddev->new_level = level; mddev->new_layout = ALGORITHM_PARITY_N; mddev->new_chunk_sectors = mddev->chunk_sectors; -- cgit v0.10.2 From fee68723cf6ae00082f70f3eff17fceab2a4f7d7 Mon Sep 17 00:00:00 2001 From: Krzysztof Wojcik Date: Wed, 20 Apr 2011 15:39:53 +1000 Subject: md: Cleanup after raid45->raid0 takeover Problem: After raid4->raid0 takeover operation, another takeover operation (e.g raid0->raid10) results "kernel oops". Root cause: Variables 'degraded' in mddev structure is not cleared on raid45->raid0 takeover. This patch reset this variable. Signed-off-by: Krzysztof Wojcik Signed-off-by: NeilBrown diff --git a/drivers/md/md.c b/drivers/md/md.c index 6e853c6..7d6f7f1 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -3170,6 +3170,7 @@ level_store(mddev_t *mddev, const char *buf, size_t len) mddev->layout = mddev->new_layout; mddev->chunk_sectors = mddev->new_chunk_sectors; mddev->delta_disks = 0; + mddev->degraded = 0; if (mddev->pers->sync_request == NULL) { /* this is now an array without redundancy, so * it must always be in_sync -- cgit v0.10.2 From 28a8397852011e323d16a1eecd4787d72b7b9a9e Mon Sep 17 00:00:00 2001 From: CoolCold Date: Wed, 20 Apr 2011 15:40:01 +1000 Subject: md: Update documentation for sync_min and sync_max entries linux/Documentation/md.txt is missing description for sync_min and sync_max entries. This patch adds description for sync_min and sync_max entries. Signed-off-by: Roman Ovchinnikov Signed-off-by: NeilBrown diff --git a/Documentation/md.txt b/Documentation/md.txt index a81c7b4..2366b1c 100644 --- a/Documentation/md.txt +++ b/Documentation/md.txt @@ -552,6 +552,16 @@ also have within the array where IO will be blocked. This is currently only supported for raid4/5/6. + sync_min + sync_max + The two values, given as numbers of sectors, indicate a range + withing the array where 'check'/'repair' will operate. Must be + a multiple of chunk_size. When it reaches "sync_max" it will + pause, rather than complete. + You can use 'select' or 'poll' on "sync_completed" to wait for + that number to reach sync_max. Then you can either increase + "sync_max", or can write 'idle' to "sync_action". + Each active md device may also have attributes specific to the personality module that manages it. -- cgit v0.10.2 From 24bdb0b62cc82120924762ae6bc85afc8c3f2b26 Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Tue, 12 Apr 2011 12:19:52 +0100 Subject: xen: do not create the extra e820 region at an addr lower than 4G Do not add the extra e820 region at a physical address lower than 4G because it breaks e820_end_of_low_ram_pfn(). It is OK for us to move the xen_extra_mem_start up and down because this is the index of the memory that can be ballooned in/out - it is memory not available to the kernel during bootup. Signed-off-by: Stefano Stabellini Signed-off-by: Konrad Rzeszutek Wilk diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c index fa0269a..90bac0a 100644 --- a/arch/x86/xen/setup.c +++ b/arch/x86/xen/setup.c @@ -227,7 +227,7 @@ char * __init xen_memory_setup(void) memcpy(map_raw, map, sizeof(map)); e820.nr_map = 0; - xen_extra_mem_start = mem_end; + xen_extra_mem_start = max((1ULL << 32), mem_end); for (i = 0; i < memmap.nr_entries; i++) { unsigned long long end; -- cgit v0.10.2 From ee176455e28469e2420032aab3db11ac2ae3eaa8 Mon Sep 17 00:00:00 2001 From: Stefano Stabellini Date: Tue, 19 Apr 2011 14:47:31 +0100 Subject: xen: mask_rw_pte: do not apply the early_ioremap checks on x86_32 The two "is_early_ioremap_ptep" checks in mask_rw_pte are only used on x86_64, in fact early_ioremap is not used at all to setup the initial pagetable on x86_32. Moreover on x86_32 the two checks are wrong because the range pgt_buf_start..pgt_buf_end initially should be mapped RW because the pages in the range are not pagetable pages yet and haven't been cleared yet. Afterwards considering the pgt_buf_start..pgt_buf_end is part of the initial mapping, xen_alloc_pte is capable of turning the ptes RO when they become pagetable pages. Fix the issue and improve the readability of the code providing two different implementation of mask_rw_pte for x86_32 and x86_64. Signed-off-by: Stefano Stabellini Signed-off-by: Konrad Rzeszutek Wilk diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index a991b57..aef7af9 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -1473,16 +1473,20 @@ static void xen_pgd_free(struct mm_struct *mm, pgd_t *pgd) #endif } +#ifdef CONFIG_X86_32 static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) { - unsigned long pfn = pte_pfn(pte); - -#ifdef CONFIG_X86_32 /* If there's an existing pte, then don't allow _PAGE_RW to be set */ if (pte_val_ma(*ptep) & _PAGE_PRESENT) pte = __pte_ma(((pte_val_ma(*ptep) & _PAGE_RW) | ~_PAGE_RW) & pte_val_ma(pte)); -#endif + + return pte; +} +#else /* CONFIG_X86_64 */ +static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) +{ + unsigned long pfn = pte_pfn(pte); /* * If the new pfn is within the range of the newly allocated @@ -1497,6 +1501,7 @@ static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) return pte; } +#endif /* CONFIG_X86_64 */ /* Init-time set_pte while constructing initial pagetables, which doesn't allow RO pagetable pages to be remapped RW */ -- cgit v0.10.2 From 1574dff8996ab1ed92c09012f8038b5566fce313 Mon Sep 17 00:00:00 2001 From: Sachin Prabhu Date: Wed, 20 Apr 2011 13:09:35 +0100 Subject: Open with O_CREAT flag set fails to open existing files on non writable directories An open on a NFS4 share using the O_CREAT flag on an existing file for which we have permissions to open but contained in a directory with no write permissions will fail with EACCES. A tcpdump shows that the client had set the open mode to UNCHECKED which indicates that the file should be created if it doesn't exist and encountering an existing flag is not an error. Since in this case the file exists and can be opened by the user, the NFS server is wrong in attempting to check create permissions on the parent directory. The patch adds a conditional statement to check for create permissions only if the file doesn't exist. Signed-off-by: Sachin S. Prabhu Signed-off-by: J. Bruce Fields diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 2e1cebd..129f3c9 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -1363,7 +1363,7 @@ nfsd_create_v3(struct svc_rqst *rqstp, struct svc_fh *fhp, goto out; if (!(iap->ia_valid & ATTR_MODE)) iap->ia_mode = 0; - err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_CREATE); + err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_EXEC); if (err) goto out; @@ -1385,6 +1385,13 @@ nfsd_create_v3(struct svc_rqst *rqstp, struct svc_fh *fhp, if (IS_ERR(dchild)) goto out_nfserr; + /* If file doesn't exist, check for permissions to create one */ + if (!dchild->d_inode) { + err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_CREATE); + if (err) + goto out; + } + err = fh_compose(resfhp, fhp->fh_export, dchild, fhp); if (err) goto out; -- cgit v0.10.2 From 3eff1268994f72266b660782e87f215720c29639 Mon Sep 17 00:00:00 2001 From: Dave Chinner Date: Tue, 12 Apr 2011 13:34:20 +0000 Subject: xfs: fix duplicate message output Commit 957935dc ("xfs: fix xfs_debug warnings" broke the logic in __xfs_printk(). Instead of only printing one of two possible output strings based on whether the fs has a name or not, it outputs both. Fix it to only output one message again. Signed-off-by: Dave Chinner Reviewed-by: Christoph Hellwig Signed-off-by: Alex Elder diff --git a/fs/xfs/linux-2.6/xfs_message.c b/fs/xfs/linux-2.6/xfs_message.c index 3ca7956..9f76cce 100644 --- a/fs/xfs/linux-2.6/xfs_message.c +++ b/fs/xfs/linux-2.6/xfs_message.c @@ -34,8 +34,10 @@ __xfs_printk( const struct xfs_mount *mp, struct va_format *vaf) { - if (mp && mp->m_fsname) + if (mp && mp->m_fsname) { printk("%sXFS (%s): %pV\n", level, mp->m_fsname, vaf); + return; + } printk("%sXFS: %pV\n", level, vaf); } -- cgit v0.10.2 From b522f02184b413955f3bc952e3776ce41edc6355 Mon Sep 17 00:00:00 2001 From: Vasiliy Kulikov Date: Thu, 14 Apr 2011 20:55:19 +0400 Subject: agp: fix OOM and buffer overflow page_count is copied from userspace. agp_allocate_memory() tries to check whether this number is too big, but doesn't take into account the wrap case. Also agp_create_user_memory() doesn't check whether alloc_size is calculated from num_agp_pages variable without overflow. This may lead to allocation of too small buffer with following buffer overflow. Another problem in agp code is not addressed in the patch - kernel memory exhaustion (AGPIOC_RESERVE and AGPIOC_ALLOCATE ioctls). It is not checked whether requested pid is a pid of the caller (no check in agpioc_reserve_wrap()). Each allocation is limited to 16KB, though, there is no per-process limit. This might lead to OOM situation, which is not even solved in case of the caller death by OOM killer - the memory is allocated for another (faked) process. Signed-off-by: Vasiliy Kulikov Signed-off-by: Dave Airlie diff --git a/drivers/char/agp/generic.c b/drivers/char/agp/generic.c index 012cba0..850a643 100644 --- a/drivers/char/agp/generic.c +++ b/drivers/char/agp/generic.c @@ -115,6 +115,9 @@ static struct agp_memory *agp_create_user_memory(unsigned long num_agp_pages) struct agp_memory *new; unsigned long alloc_size = num_agp_pages*sizeof(struct page *); + if (INT_MAX/sizeof(struct page *) < num_agp_pages) + return NULL; + new = kzalloc(sizeof(struct agp_memory), GFP_KERNEL); if (new == NULL) return NULL; @@ -234,11 +237,14 @@ struct agp_memory *agp_allocate_memory(struct agp_bridge_data *bridge, int scratch_pages; struct agp_memory *new; size_t i; + int cur_memory; if (!bridge) return NULL; - if ((atomic_read(&bridge->current_memory_agp) + page_count) > bridge->max_memory_agp) + cur_memory = atomic_read(&bridge->current_memory_agp); + if ((cur_memory + page_count > bridge->max_memory_agp) || + (cur_memory + page_count < page_count)) return NULL; if (type >= AGP_USER_TYPES) { -- cgit v0.10.2 From 194b3da873fd334ef183806db751473512af29ce Mon Sep 17 00:00:00 2001 From: Vasiliy Kulikov Date: Thu, 14 Apr 2011 20:55:16 +0400 Subject: agp: fix arbitrary kernel memory writes pg_start is copied from userspace on AGPIOC_BIND and AGPIOC_UNBIND ioctl cmds of agp_ioctl() and passed to agpioc_bind_wrap(). As said in the comment, (pg_start + mem->page_count) may wrap in case of AGPIOC_BIND, and it is not checked at all in case of AGPIOC_UNBIND. As a result, user with sufficient privileges (usually "video" group) may generate either local DoS or privilege escalation. Signed-off-by: Vasiliy Kulikov Signed-off-by: Dave Airlie diff --git a/drivers/char/agp/generic.c b/drivers/char/agp/generic.c index 850a643..b072648 100644 --- a/drivers/char/agp/generic.c +++ b/drivers/char/agp/generic.c @@ -1095,8 +1095,8 @@ int agp_generic_insert_memory(struct agp_memory * mem, off_t pg_start, int type) return -EINVAL; } - /* AK: could wrap */ - if ((pg_start + mem->page_count) > num_entries) + if (((pg_start + mem->page_count) > num_entries) || + ((pg_start + mem->page_count) < pg_start)) return -EINVAL; j = pg_start; @@ -1130,7 +1130,7 @@ int agp_generic_remove_memory(struct agp_memory *mem, off_t pg_start, int type) { size_t i; struct agp_bridge_data *bridge; - int mask_type; + int mask_type, num_entries; bridge = mem->bridge; if (!bridge) @@ -1142,6 +1142,11 @@ int agp_generic_remove_memory(struct agp_memory *mem, off_t pg_start, int type) if (type != mem->type) return -EINVAL; + num_entries = agp_num_entries(); + if (((pg_start + mem->page_count) > num_entries) || + ((pg_start + mem->page_count) < pg_start)) + return -EINVAL; + mask_type = bridge->driver->agp_type_to_mask_type(bridge, type); if (mask_type != 0) { /* The generic routines know nothing of memory types */ -- cgit v0.10.2 From 37f8527dbfd05af0f670aa02370d0c4cca7fbda6 Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Wed, 20 Apr 2011 19:19:10 -0700 Subject: Revert "x86, NUMA: Fix fakenuma boot failure" Andreas Herrmann reported that 7d6b46707f24 ("x86, NUMA: Fix fakenuma boot failure") causes certain physical NUMA topologies (for example AMD Magny-Cours) to move sibling cpus to a single node when in reality they are in separate domains. This may result in some nodes being completely void of cpus, which doesn't accurately represent the correct topology. The system will boot, but will have suboptimal NUMA performance. This commit was intended as a fix for NUMA emulation, but should not cause a regression for real NUMA machines as a side effect. ( There will be a separate fix for the numa-debug code, which will not affect physical topologies. ) Reported-by: Andreas Herrmann Signed-off-by: David Rientjes Acked-by: KOSAKI Motohiro Cc: Tejun Heo Cc: Linus Torvalds Link: http://lkml.kernel.org/r/alpine.DEB.2.00.1104201918110.12634@chino.kir.corp.google.com Signed-off-by: Ingo Molnar diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 8ed8908..c2871d3 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -312,26 +312,6 @@ void __cpuinit smp_store_cpu_info(int id) identify_secondary_cpu(c); } -static void __cpuinit check_cpu_siblings_on_same_node(int cpu1, int cpu2) -{ - int node1 = early_cpu_to_node(cpu1); - int node2 = early_cpu_to_node(cpu2); - - /* - * Our CPU scheduler assumes all logical cpus in the same physical cpu - * share the same node. But, buggy ACPI or NUMA emulation might assign - * them to different node. Fix it. - */ - if (node1 != node2) { - pr_warning("CPU %d in node %d and CPU %d in node %d are in the same physical CPU. forcing same node %d\n", - cpu1, node1, cpu2, node2, node2); - - numa_remove_cpu(cpu1); - numa_set_node(cpu1, node2); - numa_add_cpu(cpu1); - } -} - static void __cpuinit link_thread_siblings(int cpu1, int cpu2) { cpumask_set_cpu(cpu1, cpu_sibling_mask(cpu2)); @@ -340,7 +320,6 @@ static void __cpuinit link_thread_siblings(int cpu1, int cpu2) cpumask_set_cpu(cpu2, cpu_core_mask(cpu1)); cpumask_set_cpu(cpu1, cpu_llc_shared_mask(cpu2)); cpumask_set_cpu(cpu2, cpu_llc_shared_mask(cpu1)); - check_cpu_siblings_on_same_node(cpu1, cpu2); } @@ -382,12 +361,10 @@ void __cpuinit set_cpu_sibling_map(int cpu) per_cpu(cpu_llc_id, cpu) == per_cpu(cpu_llc_id, i)) { cpumask_set_cpu(i, cpu_llc_shared_mask(cpu)); cpumask_set_cpu(cpu, cpu_llc_shared_mask(i)); - check_cpu_siblings_on_same_node(cpu, i); } if (c->phys_proc_id == cpu_data(i).phys_proc_id) { cpumask_set_cpu(i, cpu_core_mask(cpu)); cpumask_set_cpu(cpu, cpu_core_mask(i)); - check_cpu_siblings_on_same_node(cpu, i); /* * Does this new cpu bringup a new core? */ -- cgit v0.10.2 From 7a6c6547825a2324faa76cff856db11d78de075e Mon Sep 17 00:00:00 2001 From: David Rientjes Date: Wed, 20 Apr 2011 19:19:13 -0700 Subject: x86, numa: Fix cpu nodemasks for NUMA emulation and CONFIG_DEBUG_PER_CPU_MAPS The cpu<->node mappings under CONFIG_DEBUG_PER_CPU_MAPS=y when NUMA emulation is enabled is currently broken because it does not iterate through every emulated node and bind cpus that have affinity to it. NUMA emulation should bind each cpu to every local node to accurately represent the true NUMA topology of the underlying machine. debug_cpumask_set_cpu() needs to be fixed at the same time so that the debugging information that it emits shows the new cpumask of the node being assigned when the cpu is being added or removed. It can now take responsibility of setting or clearing the cpu itself to remove the need for duplicate code. Also change its last parameter, "enable", to have the correct bool type since it can only be true or false. -v2: Fix the return statements, by Kosaki Motohiro Acked-and-Tested-by: KOSAKI Motohiro Signed-off-by: David Rientjes Cc: Andreas Herrmann Cc: Tejun Heo Cc: Linus Torvalds Link: http://lkml.kernel.org/r/alpine.DEB.2.00.1104201918470.12634@chino.kir.corp.google.com Signed-off-by: Ingo Molnar diff --git a/arch/x86/include/asm/numa.h b/arch/x86/include/asm/numa.h index 3d4dab4..a50fc9f 100644 --- a/arch/x86/include/asm/numa.h +++ b/arch/x86/include/asm/numa.h @@ -51,7 +51,7 @@ static inline void numa_remove_cpu(int cpu) { } #endif /* CONFIG_NUMA */ #ifdef CONFIG_DEBUG_PER_CPU_MAPS -struct cpumask __cpuinit *debug_cpumask_set_cpu(int cpu, int enable); +void debug_cpumask_set_cpu(int cpu, int node, bool enable); #endif #endif /* _ASM_X86_NUMA_H */ diff --git a/arch/x86/mm/numa.c b/arch/x86/mm/numa.c index 9559d36..745258d 100644 --- a/arch/x86/mm/numa.c +++ b/arch/x86/mm/numa.c @@ -213,53 +213,48 @@ int early_cpu_to_node(int cpu) return per_cpu(x86_cpu_to_node_map, cpu); } -struct cpumask __cpuinit *debug_cpumask_set_cpu(int cpu, int enable) +void debug_cpumask_set_cpu(int cpu, int node, bool enable) { - int node = early_cpu_to_node(cpu); struct cpumask *mask; char buf[64]; if (node == NUMA_NO_NODE) { /* early_cpu_to_node() already emits a warning and trace */ - return NULL; + return; } mask = node_to_cpumask_map[node]; if (!mask) { pr_err("node_to_cpumask_map[%i] NULL\n", node); dump_stack(); - return NULL; + return; } + if (enable) + cpumask_set_cpu(cpu, mask); + else + cpumask_clear_cpu(cpu, mask); + cpulist_scnprintf(buf, sizeof(buf), mask); printk(KERN_DEBUG "%s cpu %d node %d: mask now %s\n", enable ? "numa_add_cpu" : "numa_remove_cpu", cpu, node, buf); - return mask; + return; } # ifndef CONFIG_NUMA_EMU -static void __cpuinit numa_set_cpumask(int cpu, int enable) +static void __cpuinit numa_set_cpumask(int cpu, bool enable) { - struct cpumask *mask; - - mask = debug_cpumask_set_cpu(cpu, enable); - if (!mask) - return; - - if (enable) - cpumask_set_cpu(cpu, mask); - else - cpumask_clear_cpu(cpu, mask); + debug_cpumask_set_cpu(cpu, early_cpu_to_node(cpu), enable); } void __cpuinit numa_add_cpu(int cpu) { - numa_set_cpumask(cpu, 1); + numa_set_cpumask(cpu, true); } void __cpuinit numa_remove_cpu(int cpu) { - numa_set_cpumask(cpu, 0); + numa_set_cpumask(cpu, false); } # endif /* !CONFIG_NUMA_EMU */ diff --git a/arch/x86/mm/numa_emulation.c b/arch/x86/mm/numa_emulation.c index ad091e4..de84cc1 100644 --- a/arch/x86/mm/numa_emulation.c +++ b/arch/x86/mm/numa_emulation.c @@ -454,10 +454,9 @@ void __cpuinit numa_remove_cpu(int cpu) cpumask_clear_cpu(cpu, node_to_cpumask_map[i]); } #else /* !CONFIG_DEBUG_PER_CPU_MAPS */ -static void __cpuinit numa_set_cpumask(int cpu, int enable) +static void __cpuinit numa_set_cpumask(int cpu, bool enable) { - struct cpumask *mask; - int nid, physnid, i; + int nid, physnid; nid = early_cpu_to_node(cpu); if (nid == NUMA_NO_NODE) { @@ -467,28 +466,21 @@ static void __cpuinit numa_set_cpumask(int cpu, int enable) physnid = emu_nid_to_phys[nid]; - for_each_online_node(i) { + for_each_online_node(nid) { if (emu_nid_to_phys[nid] != physnid) continue; - mask = debug_cpumask_set_cpu(cpu, enable); - if (!mask) - return; - - if (enable) - cpumask_set_cpu(cpu, mask); - else - cpumask_clear_cpu(cpu, mask); + debug_cpumask_set_cpu(cpu, nid, enable); } } void __cpuinit numa_add_cpu(int cpu) { - numa_set_cpumask(cpu, 1); + numa_set_cpumask(cpu, true); } void __cpuinit numa_remove_cpu(int cpu) { - numa_set_cpumask(cpu, 0); + numa_set_cpumask(cpu, false); } #endif /* !CONFIG_DEBUG_PER_CPU_MAPS */ -- cgit v0.10.2 From 51a63e67da6056c13b5b597dcc9e1b3bd7ceaa55 Mon Sep 17 00:00:00 2001 From: Joseph Cihula Date: Mon, 21 Mar 2011 11:04:24 -0700 Subject: intel_iommu: disable all VT-d PMRs when TXT launched Intel VT-d Protected Memory Regions (PMRs) are supposed to be disabled, on each VT-d engine, after DMA remapping is enabled on the engines. This is because the behavior of having both enabled is not deterministic and because, if TXT has been used to launch the kernel, the PMRs may be programmed to cover memory regions that will be used for DMA. Under some circumstances (certain quirks detected, lack of multiple devices, etc.), the current code does not set up DMA remapping on some VT-d engines. In such cases it also skips disabling the PMRs. This causes failures when the kernel is launched with TXT (most often this occurs on the graphics engine and results in colored vertical bars on the display). This patch detects when the kernel has been launched with TXT and then disables the PMRs on all VT-d engines. In some cases where the reason that remapping is not being enabled is due to possible ACPI DMAR table errors, the VT-d engine addresses may not be correct and thus not able to be safely programmed even to disable PMRs. Because part of the TXT launch process is the verification of these addresses, it will always be safe to disable PMRs if the TXT launch has succeeded and hence only doing this in such cases. Signed-off-by: Joseph Cihula Signed-off-by: David Woodhouse diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c index 5dc5d3e..cdded1e 100644 --- a/drivers/pci/intel-iommu.c +++ b/drivers/pci/intel-iommu.c @@ -1299,7 +1299,7 @@ static void iommu_detach_domain(struct dmar_domain *domain, static struct iova_domain reserved_iova_list; static struct lock_class_key reserved_rbtree_key; -static void dmar_init_reserved_ranges(void) +static int dmar_init_reserved_ranges(void) { struct pci_dev *pdev = NULL; struct iova *iova; @@ -1313,8 +1313,10 @@ static void dmar_init_reserved_ranges(void) /* IOAPIC ranges shouldn't be accessed by DMA */ iova = reserve_iova(&reserved_iova_list, IOVA_PFN(IOAPIC_RANGE_START), IOVA_PFN(IOAPIC_RANGE_END)); - if (!iova) + if (!iova) { printk(KERN_ERR "Reserve IOAPIC range failed\n"); + return -ENODEV; + } /* Reserve all PCI MMIO to avoid peer-to-peer access */ for_each_pci_dev(pdev) { @@ -1327,11 +1329,13 @@ static void dmar_init_reserved_ranges(void) iova = reserve_iova(&reserved_iova_list, IOVA_PFN(r->start), IOVA_PFN(r->end)); - if (!iova) + if (!iova) { printk(KERN_ERR "Reserve iova failed\n"); + return -ENODEV; + } } } - + return 0; } static void domain_reserve_special_ranges(struct dmar_domain *domain) @@ -2213,7 +2217,7 @@ static int __init iommu_prepare_static_identity_mapping(int hw) return 0; } -int __init init_dmars(void) +static int __init init_dmars(int force_on) { struct dmar_drhd_unit *drhd; struct dmar_rmrr_unit *rmrr; @@ -2393,8 +2397,15 @@ int __init init_dmars(void) * enable translation */ for_each_drhd_unit(drhd) { - if (drhd->ignored) + if (drhd->ignored) { + /* + * we always have to disable PMRs or DMA may fail on + * this device + */ + if (force_on) + iommu_disable_protect_mem_regions(drhd->iommu); continue; + } iommu = drhd->iommu; iommu_flush_write_buffer(iommu); @@ -3303,12 +3314,21 @@ int __init intel_iommu_init(void) if (no_iommu || dmar_disabled) return -ENODEV; - iommu_init_mempool(); - dmar_init_reserved_ranges(); + if (iommu_init_mempool()) { + if (force_on) + panic("tboot: Failed to initialize iommu memory\n"); + return -ENODEV; + } + + if (dmar_init_reserved_ranges()) { + if (force_on) + panic("tboot: Failed to reserve iommu ranges\n"); + return -ENODEV; + } init_no_remapping_devices(); - ret = init_dmars(); + ret = init_dmars(force_on); if (ret) { if (force_on) panic("tboot: Failed to initialize DMARs\n"); -- cgit v0.10.2 From b3258ff1d6086bd2b9eeb556844a868ad7d49bc8 Mon Sep 17 00:00:00 2001 From: Amit Shah Date: Wed, 16 Mar 2011 19:12:10 +0530 Subject: virtio: Decrement avail idx on buffer detach When detaching a buffer from a vq, the avail.idx value should be decremented as well. This was noticed by hot-unplugging a virtio console port and then plugging in a new one on the same number (re-using the vqs which were just 'disowned'). qemu reported 'Guest moved used index from 0 to 256' when any IO was attempted on the new port. CC: stable@kernel.org Reported-by: juzhang Signed-off-by: Amit Shah Signed-off-by: Rusty Russell diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c index cc2f73e..b0043fb 100644 --- a/drivers/virtio/virtio_ring.c +++ b/drivers/virtio/virtio_ring.c @@ -371,6 +371,7 @@ void *virtqueue_detach_unused_buf(struct virtqueue *_vq) /* detach_buf clears data, so grab it now. */ buf = vq->data[i]; detach_buf(vq, i); + vq->vring.avail->idx--; END_USE(vq); return buf; } -- cgit v0.10.2 From 31a3ddda166cda86d2b5111e09ba4bda5239fae6 Mon Sep 17 00:00:00 2001 From: Amit Shah Date: Mon, 14 Mar 2011 17:45:02 +0530 Subject: virtio_pci: Prevent double-free of pci regions after device hot-unplug In the case where a virtio-console port is in use (opened by a program) and a virtio-console device is removed, the port is kept around but all the virtio-related state is assumed to be gone. When the port is finally released (close() called), we call device_destroy() on the port's device. This results in the parent device's structures to be freed as well. This includes the PCI regions for the virtio-console PCI device. Once this is done, however, virtio_pci_release_dev() kicks in, as the last ref to the virtio device is now gone, and attempts to do pci_iounmap(pci_dev, vp_dev->ioaddr); pci_release_regions(pci_dev); pci_disable_device(pci_dev); which results in a double-free warning. Move the code that releases regions, etc., to the virtio_pci_remove() function, and all that's now left in release_dev is the final freeing of the vp_dev. Signed-off-by: Amit Shah Signed-off-by: Rusty Russell diff --git a/drivers/virtio/virtio_pci.c b/drivers/virtio/virtio_pci.c index 4fb5b2b..4bcc8b8 100644 --- a/drivers/virtio/virtio_pci.c +++ b/drivers/virtio/virtio_pci.c @@ -590,15 +590,10 @@ static struct virtio_config_ops virtio_pci_config_ops = { static void virtio_pci_release_dev(struct device *_d) { - struct virtio_device *dev = container_of(_d, struct virtio_device, dev); + struct virtio_device *dev = container_of(_d, struct virtio_device, + dev); struct virtio_pci_device *vp_dev = to_vp_device(dev); - struct pci_dev *pci_dev = vp_dev->pci_dev; - vp_del_vqs(dev); - pci_set_drvdata(pci_dev, NULL); - pci_iounmap(pci_dev, vp_dev->ioaddr); - pci_release_regions(pci_dev); - pci_disable_device(pci_dev); kfree(vp_dev); } @@ -681,6 +676,12 @@ static void __devexit virtio_pci_remove(struct pci_dev *pci_dev) struct virtio_pci_device *vp_dev = pci_get_drvdata(pci_dev); unregister_virtio_device(&vp_dev->vdev); + + vp_del_vqs(&vp_dev->vdev); + pci_set_drvdata(pci_dev, NULL); + pci_iounmap(pci_dev, vp_dev->ioaddr); + pci_release_regions(pci_dev); + pci_disable_device(pci_dev); } #ifdef CONFIG_PM -- cgit v0.10.2 From afa2689e19073cd2e762d0f2c1358fab1ab9f18c Mon Sep 17 00:00:00 2001 From: Amit Shah Date: Mon, 14 Mar 2011 17:45:48 +0530 Subject: virtio: console: Enable call to hvc_remove() on console port remove This call was disabled as hot-unplugging one virtconsole port led to another virtconsole port freezing. Upon testing it again, this now works, so enable it. In addition, a bug was found in qemu wherein removing a port of one type caused the guest output from another port to stop working. I doubt it was just this bug that caused it (since disabling the hvc_remove() call did allow other ports to continue working), but since it's all solved now, we're fine with hot-unplugging of virtconsole ports. Signed-off-by: Amit Shah Signed-off-by: Rusty Russell diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c index 84b164d..838568a 100644 --- a/drivers/char/virtio_console.c +++ b/drivers/char/virtio_console.c @@ -1280,18 +1280,7 @@ static void unplug_port(struct port *port) spin_lock_irq(&pdrvdata_lock); list_del(&port->cons.list); spin_unlock_irq(&pdrvdata_lock); -#if 0 - /* - * hvc_remove() not called as removing one hvc port - * results in other hvc ports getting frozen. - * - * Once this is resolved in hvc, this functionality - * will be enabled. Till that is done, the -EPIPE - * return from get_chars() above will help - * hvc_console.c to clean up on ports we remove here. - */ hvc_remove(port->cons.hvc); -#endif } /* Remove unused data this port might have received. */ -- cgit v0.10.2 From df7e130384efd1c732aa08648dad46687fee3d96 Mon Sep 17 00:00:00 2001 From: Jan Kara Date: Wed, 20 Apr 2011 20:30:40 +0200 Subject: vfs: Pass setxattr(2) flags properly For some reason generic_setxattr() did not pass flags (XATTR_CREATE, XATTR_REPLACE) to the filesystem specific helper. This caused that setxattr(2) syscall just ignored these flags. Fix the bug by passing flags correctly. Signed-off-by: Jan Kara Acked-by: Christoph Hellwig Signed-off-by: Linus Torvalds diff --git a/fs/xattr.c b/fs/xattr.c index a19acdb..f1ef949 100644 --- a/fs/xattr.c +++ b/fs/xattr.c @@ -666,7 +666,7 @@ generic_setxattr(struct dentry *dentry, const char *name, const void *value, siz handler = xattr_resolve_name(dentry->d_sb->s_xattr, &name); if (!handler) return -EOPNOTSUPP; - return handler->set(dentry, name, value, size, 0, handler->flags); + return handler->set(dentry, name, value, size, flags, handler->flags); } /* -- cgit v0.10.2 From d76c8420c3cf8e468901b0bd58306637335c98ea Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Thu, 21 Apr 2011 09:07:26 -0700 Subject: raid5: fix build error, sector_t usage Change from unsigned long long to sector_t. This matches its source field. ERROR: "__udivdi3" [drivers/md/raid456.ko] undefined! Signed-off-by: Randy Dunlap Signed-off-by: Linus Torvalds diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c index fd50011..49bf5f8 100644 --- a/drivers/md/raid5.c +++ b/drivers/md/raid5.c @@ -5678,7 +5678,7 @@ static void raid5_quiesce(mddev_t *mddev, int state) static void *raid45_takeover_raid0(mddev_t *mddev, int level) { struct raid0_private_data *raid0_priv = mddev->private; - unsigned long long sectors; + sector_t sectors; /* for raid0 takeover only one zone is supported */ if (raid0_priv->nr_strip_zones > 1) { -- cgit v0.10.2 From 3aa72873ffdcc2f7919743efbbefc351ec73f5cb Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 21 Apr 2011 19:28:35 +0200 Subject: elevator: check for ELEVATOR_INSERT_SORT_MERGE in !elvpriv case too The sort insert is the one that goes to the IO scheduler. With the SORT_MERGE addition, we could bypass IO scheduler setup but still ask the IO scheduler to insert the request. This would cause an oops on switching IO schedulers through the sysfs interface, unless the disk just happened to be idle while it occured. Signed-off-by: Jens Axboe diff --git a/block/elevator.c b/block/elevator.c index 6f6abc0..45ca1e3 100644 --- a/block/elevator.c +++ b/block/elevator.c @@ -671,7 +671,8 @@ void __elv_add_request(struct request_queue *q, struct request *rq, int where) q->boundary_rq = rq; } } else if (!(rq->cmd_flags & REQ_ELVPRIV) && - where == ELEVATOR_INSERT_SORT) + (where == ELEVATOR_INSERT_SORT || + where == ELEVATOR_INSERT_SORT_MERGE)) where = ELEVATOR_INSERT_BACK; switch (where) { -- cgit v0.10.2 From 7c88a168da8003fd4d8fb6ae103c4ecf29cb1130 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Thu, 21 Apr 2011 19:43:58 +0200 Subject: block: don't propagate unlisted DISK_EVENTs to userland DISK_EVENT_MEDIA_CHANGE is used for both userland visible event and internal event for revalidation of removeable devices. Some legacy drivers don't implement proper event detection and continuously generate events under certain circumstances. For example, ide-cd generates media changed continuously if there's no media in the drive, which can lead to infinite loop of events jumping back and forth between the driver and userland event handler. This patch updates disk event infrastructure such that it never propagates events not listed in disk->events to userland. Those events are processed the same for internal purposes but uevent generation is suppressed. This also ensures that userland only gets events which are advertised in the @events sysfs node lowering risk of confusion. Signed-off-by: Tejun Heo Signed-off-by: Jens Axboe diff --git a/block/genhd.c b/block/genhd.c index b364bd0..2dd9887 100644 --- a/block/genhd.c +++ b/block/genhd.c @@ -1588,9 +1588,13 @@ static void disk_events_workfn(struct work_struct *work) spin_unlock_irq(&ev->lock); - /* tell userland about new events */ + /* + * Tell userland about new events. Only the events listed in + * @disk->events are reported. Unlisted events are processed the + * same internally but never get reported to userland. + */ for (i = 0; i < ARRAY_SIZE(disk_uevents); i++) - if (events & (1 << i)) + if (events & disk->events & (1 << i)) envp[nr_events++] = disk_uevents[i]; if (nr_events) -- cgit v0.10.2 From 7eec77a1816a7042591a6cbdb4820e9e7ebffe0e Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Thu, 21 Apr 2011 19:43:59 +0200 Subject: ide: unexport DISK_EVENT_MEDIA_CHANGE for ide-gd and ide-cd check_events() implementations in both ide-gd and ide-cd are inadequate for in-kernel event polling. Both generate media change events continuously when certain conditions are met causing infinite event loop between the driver and userland event handler. As disk event now supports suppression of unlisted events, simply de-listing DISK_EVENT_MEDIA_CHANGE from disk->events resolves the problem. Internal handling around media revalidation will behave the same while userland will fall back to userland event polling after detecting the device doesn't support disk events. Signed-off-by: Tejun Heo Reported-by: Jens Axboe Acked-by: "David S. Miller" Signed-off-by: Jens Axboe diff --git a/drivers/ide/ide-cd.c b/drivers/ide/ide-cd.c index fd1e117..a5ec5a7 100644 --- a/drivers/ide/ide-cd.c +++ b/drivers/ide/ide-cd.c @@ -1782,7 +1782,6 @@ static int ide_cd_probe(ide_drive_t *drive) ide_cd_read_toc(drive, &sense); g->fops = &idecd_ops; g->flags |= GENHD_FL_REMOVABLE; - g->events = DISK_EVENT_MEDIA_CHANGE; add_disk(g); return 0; diff --git a/drivers/ide/ide-cd_ioctl.c b/drivers/ide/ide-cd_ioctl.c index 2a6bc50..02caa7d 100644 --- a/drivers/ide/ide-cd_ioctl.c +++ b/drivers/ide/ide-cd_ioctl.c @@ -79,6 +79,12 @@ int ide_cdrom_drive_status(struct cdrom_device_info *cdi, int slot_nr) return CDS_DRIVE_NOT_READY; } +/* + * ide-cd always generates media changed event if media is missing, which + * makes it impossible to use for proper event reporting, so disk->events + * is cleared to 0 and the following function is used only to trigger + * revalidation and never propagated to userland. + */ unsigned int ide_cdrom_check_events_real(struct cdrom_device_info *cdi, unsigned int clearing, int slot_nr) { diff --git a/drivers/ide/ide-gd.c b/drivers/ide/ide-gd.c index c4ffd48..70ea876 100644 --- a/drivers/ide/ide-gd.c +++ b/drivers/ide/ide-gd.c @@ -298,6 +298,12 @@ static unsigned int ide_gd_check_events(struct gendisk *disk, return 0; } + /* + * The following is used to force revalidation on the first open on + * removeable devices, and never gets reported to userland as + * genhd->events is 0. This is intended as removeable ide disk + * can't really detect MEDIA_CHANGE events. + */ ret = drive->dev_flags & IDE_DFLAG_MEDIA_CHANGED; drive->dev_flags &= ~IDE_DFLAG_MEDIA_CHANGED; @@ -413,7 +419,6 @@ static int ide_gd_probe(ide_drive_t *drive) if (drive->dev_flags & IDE_DFLAG_REMOVABLE) g->flags = GENHD_FL_REMOVABLE; g->fops = &ide_gd_ops; - g->events = DISK_EVENT_MEDIA_CHANGE; add_disk(g); return 0; -- cgit v0.10.2