From 697f619fc87aa9bf5b6c8c756f7ea54e950d5cd5 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Mon, 13 Apr 2009 14:39:54 -0700 Subject: filemap: fix kernel-doc warnings Fix filemap.c kernel-doc warnings: Warning(mm/filemap.c:575): No description found for parameter 'page' Warning(mm/filemap.c:575): No description found for parameter 'waiter' Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/filemap.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'mm') diff --git a/mm/filemap.c b/mm/filemap.c index 2e2d38ebda4..8bd498040f3 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -567,8 +567,8 @@ EXPORT_SYMBOL(wait_on_page_bit); /** * add_page_wait_queue - Add an arbitrary waiter to a page's wait queue - * @page - Page defining the wait queue of interest - * @waiter - Waiter to add to the queue + * @page: Page defining the wait queue of interest + * @waiter: Waiter to add to the queue * * Add an arbitrary @waiter to the wait queue for the nominated @page. */ -- cgit v1.2.3 From 5a52edded382c2f436721d5a044ed16c290c5750 Mon Sep 17 00:00:00 2001 From: David Howells Date: Mon, 13 Apr 2009 14:40:01 -0700 Subject: mm: point the UNEVICTABLE_LRU config option at the documentation Point the UNEVICTABLE_LRU config option at the documentation describing the option. Signed-off-by: David Howells Cc: Lee Schermerhorn Cc: Rik van Riel Cc: KOSAKI Motohiro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/Kconfig | 2 ++ 1 file changed, 2 insertions(+) (limited to 'mm') diff --git a/mm/Kconfig b/mm/Kconfig index b53427ad30a..57971d2ab84 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -213,6 +213,8 @@ config UNEVICTABLE_LRU will use one page flag and increase the code size a little, say Y unless you know what you are doing. + See Documentation/vm/unevictable-lru.txt for more information. + config HAVE_MLOCK bool default y if MMU=y -- cgit v1.2.3 From 9de100d001564f58c3fb2ec1bd03e540ac0aa357 Mon Sep 17 00:00:00 2001 From: Andy Grover Date: Mon, 13 Apr 2009 14:40:05 -0700 Subject: mm: document get_user_pages_fast() While better than get_user_pages(), the usage of gupf(), especially the return values and the fact that it can potentially only partially pin the range, warranted some documentation. Signed-off-by: Andy Grover Cc: Ingo Molnar Cc: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/util.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) (limited to 'mm') diff --git a/mm/util.c b/mm/util.c index 2599e83eea1..55bef160b9f 100644 --- a/mm/util.c +++ b/mm/util.c @@ -223,6 +223,22 @@ void arch_pick_mmap_layout(struct mm_struct *mm) } #endif +/** + * get_user_pages_fast() - pin user pages in memory + * @start: starting user address + * @nr_pages: number of pages from start to pin + * @write: whether pages will be written to + * @pages: array that receives pointers to the pages pinned. + * Should be at least nr_pages long. + * + * Attempt to pin user pages in memory without taking mm->mmap_sem. + * If not successful, it will fall back to taking the lock and + * calling get_user_pages(). + * + * Returns number of pages pinned. This may be fewer than the number + * requested. If nr_pages is 0 or negative, returns 0. If no pages + * were pinned, returns -errno. + */ int __attribute__((weak)) get_user_pages_fast(unsigned long start, int nr_pages, int write, struct page **pages) { -- cgit v1.2.3 From a8031cb00e286600ea08bd00a6812dbfec412376 Mon Sep 17 00:00:00 2001 From: KAMEZAWA Hiroyuki Date: Mon, 13 Apr 2009 14:40:08 -0700 Subject: memcg: remove warning when CONFIG_DEBUG_VM=n mm/memcontrol.c:318: warning: `mem_cgroup_is_obsolete' defined but not used [akpm@linux-foundation.org: simplify as suggested by Balbir] Signed-off-by: KAMEZAWA Hiroyuki Reviewed-by: Balbir Singh Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/memcontrol.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'mm') diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 2fc6d6c4823..e44fb0fbb80 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -932,7 +932,7 @@ static int __mem_cgroup_try_charge(struct mm_struct *mm, if (unlikely(!mem)) return 0; - VM_BUG_ON(mem_cgroup_is_obsolete(mem)); + VM_BUG_ON(!mem || mem_cgroup_is_obsolete(mem)); while (1) { int ret; -- cgit v1.2.3 From 61609d01cbb3ab865c8cccaf85e6837c47096480 Mon Sep 17 00:00:00 2001 From: Yuri Tikhonov Date: Mon, 13 Apr 2009 14:40:11 -0700 Subject: shmem: fix division by zero Fix a division by zero which we have in shmem_truncate_range() and shmem_unuse_inode() when using big PAGE_SIZE values (e.g. 256kB on ppc44x). With 256kB PAGE_SIZE, the ENTRIES_PER_PAGEPAGE constant becomes too large (0x1.0000.0000) on a 32-bit kernel, so this patch just changes its type from 'unsigned long' to 'unsigned long long'. Hugh: reverted its unsigned long longs in shmem_truncate_range() and shmem_getpage(): the pagecache index cannot be more than an unsigned long, so the divisions by zero occurred in unreached code. It's a pity we need any ULL arithmetic here, but I found no pretty way to avoid it. Signed-off-by: Yuri Tikhonov Signed-off-by: Hugh Dickins Cc: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/shmem.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'mm') diff --git a/mm/shmem.c b/mm/shmem.c index d94d2e9146b..28024ce8b8f 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -66,7 +66,7 @@ static struct vfsmount *shm_mnt; #include #define ENTRIES_PER_PAGE (PAGE_CACHE_SIZE/sizeof(unsigned long)) -#define ENTRIES_PER_PAGEPAGE (ENTRIES_PER_PAGE*ENTRIES_PER_PAGE) +#define ENTRIES_PER_PAGEPAGE ((unsigned long long)ENTRIES_PER_PAGE*ENTRIES_PER_PAGE) #define BLOCKS_PER_PAGE (PAGE_CACHE_SIZE/512) #define SHMEM_MAX_INDEX (SHMEM_NR_DIRECT + (ENTRIES_PER_PAGEPAGE/2) * (ENTRIES_PER_PAGE+1)) -- cgit v1.2.3 From caefba1740d8016e6dfe8fda84f85bdcb8f8c85d Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Mon, 13 Apr 2009 14:40:12 -0700 Subject: shmem: respect MAX_LFS_FILESIZE SHMEM_MAX_BYTES was derived from the maximum size of its triple-indirect swap vector, forgetting to take the MAX_LFS_FILESIZE limit into account. Never mind 256kB pages, even 8kB pages on 32-bit kernels allowed files to grow slightly bigger than that supposed maximum. Fix this by using the min of both (at build time not run time). And it happens that this calculation is good as far as 8MB pages on 32-bit or 16MB pages on 64-bit: though SHMSWP_MAX_INDEX gets truncated before that, it's truncated to such large numbers that we don't need to care. [akpm@linux-foundation.org: it needs pagemap.h] [akpm@linux-foundation.org: fix sparc64 min() warnings] Signed-off-by: Hugh Dickins Cc: Yuri Tikhonov Cc: Paul Mackerras Cc: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/shmem.c | 25 ++++++++++++++++++++----- 1 file changed, 20 insertions(+), 5 deletions(-) (limited to 'mm') diff --git a/mm/shmem.c b/mm/shmem.c index 28024ce8b8f..f9cb20ebb99 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -24,6 +24,7 @@ #include #include #include +#include #include #include #include @@ -43,7 +44,6 @@ static struct vfsmount *shm_mnt; #include #include #include -#include #include #include #include @@ -65,13 +65,28 @@ static struct vfsmount *shm_mnt; #include #include +/* + * The maximum size of a shmem/tmpfs file is limited by the maximum size of + * its triple-indirect swap vector - see illustration at shmem_swp_entry(). + * + * With 4kB page size, maximum file size is just over 2TB on a 32-bit kernel, + * but one eighth of that on a 64-bit kernel. With 8kB page size, maximum + * file size is just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel, + * MAX_LFS_FILESIZE being then more restrictive than swap vector layout. + * + * We use / and * instead of shifts in the definitions below, so that the swap + * vector can be tested with small even values (e.g. 20) for ENTRIES_PER_PAGE. + */ #define ENTRIES_PER_PAGE (PAGE_CACHE_SIZE/sizeof(unsigned long)) #define ENTRIES_PER_PAGEPAGE ((unsigned long long)ENTRIES_PER_PAGE*ENTRIES_PER_PAGE) -#define BLOCKS_PER_PAGE (PAGE_CACHE_SIZE/512) -#define SHMEM_MAX_INDEX (SHMEM_NR_DIRECT + (ENTRIES_PER_PAGEPAGE/2) * (ENTRIES_PER_PAGE+1)) -#define SHMEM_MAX_BYTES ((unsigned long long)SHMEM_MAX_INDEX << PAGE_CACHE_SHIFT) +#define SHMSWP_MAX_INDEX (SHMEM_NR_DIRECT + (ENTRIES_PER_PAGEPAGE/2) * (ENTRIES_PER_PAGE+1)) +#define SHMSWP_MAX_BYTES (SHMSWP_MAX_INDEX << PAGE_CACHE_SHIFT) +#define SHMEM_MAX_BYTES min_t(unsigned long long, SHMSWP_MAX_BYTES, MAX_LFS_FILESIZE) +#define SHMEM_MAX_INDEX ((unsigned long)((SHMEM_MAX_BYTES+1) >> PAGE_CACHE_SHIFT)) + +#define BLOCKS_PER_PAGE (PAGE_CACHE_SIZE/512) #define VM_ACCT(size) (PAGE_CACHE_ALIGN(size) >> PAGE_SHIFT) /* info->flags needs VM_flags to handle pagein/truncate races efficiently */ @@ -2581,7 +2596,7 @@ int shmem_unuse(swp_entry_t entry, struct page *page) #define shmem_get_inode(sb, mode, dev, flags) ramfs_get_inode(sb, mode, dev) #define shmem_acct_size(flags, size) 0 #define shmem_unacct_size(flags, size) do {} while (0) -#define SHMEM_MAX_BYTES LLONG_MAX +#define SHMEM_MAX_BYTES MAX_LFS_FILESIZE #endif /* CONFIG_SHMEM */ -- cgit v1.2.3 From f69955855eac55a048d26a1618f50dfaa160a006 Mon Sep 17 00:00:00 2001 From: Chris Mason Date: Wed, 15 Apr 2009 13:22:37 -0400 Subject: Export filemap_write_and_wait_range This wasn't exported before and is useful (used by the experimental ext3 data=guarded code) Signed-off-by: Chris Mason Acked-by: Theodore Tso Acked-by: Jan Kara Signed-off-by: Linus Torvalds --- mm/filemap.c | 1 + 1 file changed, 1 insertion(+) (limited to 'mm') diff --git a/mm/filemap.c b/mm/filemap.c index 8bd498040f3..379ff0bcbf6 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -441,6 +441,7 @@ int filemap_write_and_wait_range(struct address_space *mapping, } return err; } +EXPORT_SYMBOL(filemap_write_and_wait_range); /** * add_to_page_cache_locked - add a locked page to the pagecache -- cgit v1.2.3 From 05fa199d45c54a9bda7aa3ae6537253d6f097aa9 Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Thu, 16 Apr 2009 21:58:12 +0100 Subject: mm: pass correct mm when growing stack Tetsuo Handa reports seeing the WARN_ON(current->mm == NULL) in security_vm_enough_memory(), when do_execve() is touching the target mm's stack, to set up its args and environment. Yes, a UMH_NO_WAIT or UMH_WAIT_PROC call_usermodehelper() spawns an mm-less kernel thread to do the exec. And in any case, that vm_enough_memory check when growing stack ought to be done on the target mm, not on the execer's mm (though apart from the warning, it only makes a slight tweak to OVERCOMMIT_NEVER behaviour). Reported-by: Tetsuo Handa Signed-off-by: Hugh Dickins Cc: stable@kernel.org Signed-off-by: Linus Torvalds --- mm/mmap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'mm') diff --git a/mm/mmap.c b/mm/mmap.c index 4a3841186c1..3303d1ba8e8 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -1575,7 +1575,7 @@ static int acct_stack_growth(struct vm_area_struct *vma, unsigned long size, uns * Overcommit.. This must be the final test, as it will * update security statistics. */ - if (security_vm_enough_memory(grow)) + if (security_vm_enough_memory_mm(mm, grow)) return -ENOMEM; /* Ok, everything looks good - let it rip */ -- cgit v1.2.3 From a21e25536169432cf9174d631972bc1cd4c75062 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Sat, 18 Apr 2009 17:23:41 +0200 Subject: PM/Hibernate: Fix memory shrinking Commit d979677c4c0 ("mm: shrink_all_memory(): use sc.nr_reclaimed") broke the memory shrinking used by hibernation, becuse it did not update shrink_all_zones() in accordance with the other changes it made. Fix this by making shrink_all_zones() update sc->nr_reclaimed instead of overwriting its value. This fixes http://bugzilla.kernel.org/show_bug.cgi?id=13058 Reported-and-tested-by: Alan Jenkins Signed-off-by: Rafael J. Wysocki Signed-off-by: Linus Torvalds --- mm/vmscan.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'mm') diff --git a/mm/vmscan.c b/mm/vmscan.c index 39fdfb14eea..99155b7b812 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -2088,13 +2088,13 @@ static void shrink_all_zones(unsigned long nr_pages, int prio, nr_reclaimed += shrink_list(l, nr_to_scan, zone, sc, prio); if (nr_reclaimed >= nr_pages) { - sc->nr_reclaimed = nr_reclaimed; + sc->nr_reclaimed += nr_reclaimed; return; } } } } - sc->nr_reclaimed = nr_reclaimed; + sc->nr_reclaimed += nr_reclaimed; } /* @@ -2115,6 +2115,7 @@ unsigned long shrink_all_memory(unsigned long nr_pages) .may_unmap = 0, .may_writepage = 1, .isolate_pages = isolate_pages_global, + .nr_reclaimed = 0, }; current->reclaim_state = &reclaim_state; -- cgit v1.2.3 From 2e2e425989080cc534fc0fca154cae515f971cf5 Mon Sep 17 00:00:00 2001 From: KOSAKI Motohiro Date: Tue, 21 Apr 2009 12:24:57 -0700 Subject: vmscan,memcg: reintroduce sc->may_swap Commit a6dc60f8975ad96d162915e07703a4439c80dcf0 ("vmscan: rename sc.may_swap to may_unmap") removed the may_swap flag, but memcg had used it as a flag for "we need to use swap?", as the name indicate. And in the current implementation, memcg cannot reclaim mapped file caches when mem+swap hits the limit. re-introduce may_swap flag and handle it at get_scan_ratio(). This patch doesn't influence any scan_control users other than memcg. Signed-off-by: KOSAKI Motohiro Signed-off-by: Daisuke Nishimura Acked-by: Johannes Weiner Cc: Balbir Singh Cc: KAMEZAWA Hiroyuki Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/vmscan.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) (limited to 'mm') diff --git a/mm/vmscan.c b/mm/vmscan.c index 99155b7b812..eac9577941f 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -63,6 +63,9 @@ struct scan_control { /* Can mapped pages be reclaimed? */ int may_unmap; + /* Can pages be swapped as part of reclaim? */ + int may_swap; + /* This context's SWAP_CLUSTER_MAX. If freeing memory for * suspend, we effectively ignore SWAP_CLUSTER_MAX. * In this context, it doesn't matter that we scan the @@ -1380,7 +1383,7 @@ static void get_scan_ratio(struct zone *zone, struct scan_control *sc, struct zone_reclaim_stat *reclaim_stat = get_reclaim_stat(zone, sc); /* If we have no swap space, do not bother scanning anon pages. */ - if (nr_swap_pages <= 0) { + if (!sc->may_swap || (nr_swap_pages <= 0)) { percent[0] = 0; percent[1] = 100; return; @@ -1697,6 +1700,7 @@ unsigned long try_to_free_pages(struct zonelist *zonelist, int order, .may_writepage = !laptop_mode, .swap_cluster_max = SWAP_CLUSTER_MAX, .may_unmap = 1, + .may_swap = 1, .swappiness = vm_swappiness, .order = order, .mem_cgroup = NULL, @@ -1717,6 +1721,7 @@ unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont, struct scan_control sc = { .may_writepage = !laptop_mode, .may_unmap = 1, + .may_swap = !noswap, .swap_cluster_max = SWAP_CLUSTER_MAX, .swappiness = swappiness, .order = 0, @@ -1726,9 +1731,6 @@ unsigned long try_to_free_mem_cgroup_pages(struct mem_cgroup *mem_cont, }; struct zonelist *zonelist; - if (noswap) - sc.may_unmap = 0; - sc.gfp_mask = (gfp_mask & GFP_RECLAIM_MASK) | (GFP_HIGHUSER_MOVABLE & ~GFP_RECLAIM_MASK); zonelist = NODE_DATA(numa_node_id())->node_zonelists; @@ -1767,6 +1769,7 @@ static unsigned long balance_pgdat(pg_data_t *pgdat, int order) struct scan_control sc = { .gfp_mask = GFP_KERNEL, .may_unmap = 1, + .may_swap = 1, .swap_cluster_max = SWAP_CLUSTER_MAX, .swappiness = vm_swappiness, .order = order, @@ -2298,6 +2301,7 @@ static int __zone_reclaim(struct zone *zone, gfp_t gfp_mask, unsigned int order) struct scan_control sc = { .may_writepage = !!(zone_reclaim_mode & RECLAIM_WRITE), .may_unmap = !!(zone_reclaim_mode & RECLAIM_SWAP), + .may_swap = 1, .swap_cluster_max = max_t(unsigned long, nr_pages, SWAP_CLUSTER_MAX), .gfp_mask = gfp_mask, -- cgit v1.2.3 From bc43f75cd9815833b27831600ccade672edb5e43 Mon Sep 17 00:00:00 2001 From: Johannes Weiner Date: Thu, 30 Apr 2009 15:08:08 -0700 Subject: mm: fix pageref leak in do_swap_page() By the time the memory cgroup code is notified about a swapin we already hold a reference on the fault page. If the cgroup callback fails make sure to unlock AND release the page reference which was taken by lookup_swap_cach(), or we leak the reference. Signed-off-by: Johannes Weiner Cc: Balbir Singh Reviewed-by: Minchan Kim Acked-by: KAMEZAWA Hiroyuki Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/memory.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'mm') diff --git a/mm/memory.c b/mm/memory.c index cf6873e91c6..6a4ef0fd071 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -2458,8 +2458,7 @@ static int do_swap_page(struct mm_struct *mm, struct vm_area_struct *vma, if (mem_cgroup_try_charge_swapin(mm, page, GFP_KERNEL, &ptr)) { ret = VM_FAULT_OOM; - unlock_page(page); - goto out; + goto out_page; } /* @@ -2521,6 +2520,7 @@ out: out_nomap: mem_cgroup_cancel_charge_swapin(ptr); pte_unmap_unlock(page_table, ptl); +out_page: unlock_page(page); page_cache_release(page); return ret; -- cgit v1.2.3 From c0bd3f63ce01a1757dbce6373122a05fbf99ced7 Mon Sep 17 00:00:00 2001 From: Daisuke Nishimura Date: Thu, 30 Apr 2009 15:08:11 -0700 Subject: memcg: fix try_get_mem_cgroup_from_swapcache() This is a bugfix for commit 3c776e64660028236313f0e54f3a9945764422df ("memcg: charge swapcache to proper memcg"). Used bit of swapcache is solid under page lock, but considering move_account, pc->mem_cgroup is not. We need lock_page_cgroup() anyway. Signed-off-by: Daisuke Nishimura Acked-by: KAMEZAWA Hiroyuki Cc: Balbir Singh Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/memcontrol.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'mm') diff --git a/mm/memcontrol.c b/mm/memcontrol.c index e44fb0fbb80..575203ae210 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -1024,9 +1024,7 @@ static struct mem_cgroup *try_get_mem_cgroup_from_swapcache(struct page *page) return NULL; pc = lookup_page_cgroup(page); - /* - * Used bit of swapcache is solid under page lock. - */ + lock_page_cgroup(pc); if (PageCgroupUsed(pc)) { mem = pc->mem_cgroup; if (mem && !css_tryget(&mem->css)) @@ -1040,6 +1038,7 @@ static struct mem_cgroup *try_get_mem_cgroup_from_swapcache(struct page *page) mem = NULL; rcu_read_unlock(); } + unlock_page_cgroup(pc); return mem; } -- cgit v1.2.3 From b827e496c893de0c0f142abfaeb8730a2fd6b37f Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Thu, 30 Apr 2009 15:08:16 -0700 Subject: mm: close page_mkwrite races Change page_mkwrite to allow implementations to return with the page locked, and also change it's callers (in page fault paths) to hold the lock until the page is marked dirty. This allows the filesystem to have full control of page dirtying events coming from the VM. Rather than simply hold the page locked over the page_mkwrite call, we call page_mkwrite with the page unlocked and allow callers to return with it locked, so filesystems can avoid LOR conditions with page lock. The problem with the current scheme is this: a filesystem that wants to associate some metadata with a page as long as the page is dirty, will perform this manipulation in its ->page_mkwrite. It currently then must return with the page unlocked and may not hold any other locks (according to existing page_mkwrite convention). In this window, the VM could write out the page, clearing page-dirty. The filesystem has no good way to detect that a dirty pte is about to be attached, so it will happily write out the page, at which point, the filesystem may manipulate the metadata to reflect that the page is no longer dirty. It is not always possible to perform the required metadata manipulation in ->set_page_dirty, because that function cannot block or fail. The filesystem may need to allocate some data structure, for example. And the VM cannot mark the pte dirty before page_mkwrite, because page_mkwrite is allowed to fail, so we must not allow any window where the page could be written to if page_mkwrite does fail. This solution of holding the page locked over the 3 critical operations (page_mkwrite, setting the pte dirty, and finally setting the page dirty) closes out races nicely, preventing page cleaning for writeout being initiated in that window. This provides the filesystem with a strong synchronisation against the VM here. - Sage needs this race closed for ceph filesystem. - Trond for NFS (http://bugzilla.kernel.org/show_bug.cgi?id=12913). - I need it for fsblock. - I suspect other filesystems may need it too (eg. btrfs). - I have converted buffer.c to the new locking. Even simple block allocation under dirty pages might be susceptible to i_size changing under partial page at the end of file (we also have a buffer.c-side problem here, but it cannot be fixed properly without this patch). - Other filesystems (eg. NFS, maybe btrfs) will need to change their page_mkwrite functions themselves. [ This also moves page_mkwrite another step closer to fault, which should eventually allow page_mkwrite to be moved into ->fault, and thus avoiding a filesystem calldown and page lock/unlock cycle in __do_fault. ] [akpm@linux-foundation.org: fix derefs of NULL ->mapping] Cc: Sage Weil Cc: Trond Myklebust Signed-off-by: Nick Piggin Cc: Valdis Kletnieks Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/memory.c | 108 ++++++++++++++++++++++++++++++++++++++++++------------------ 1 file changed, 76 insertions(+), 32 deletions(-) (limited to 'mm') diff --git a/mm/memory.c b/mm/memory.c index 6a4ef0fd071..4126dd16778 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -1971,6 +1971,15 @@ static int do_wp_page(struct mm_struct *mm, struct vm_area_struct *vma, ret = tmp; goto unwritable_page; } + if (unlikely(!(tmp & VM_FAULT_LOCKED))) { + lock_page(old_page); + if (!old_page->mapping) { + ret = 0; /* retry the fault */ + unlock_page(old_page); + goto unwritable_page; + } + } else + VM_BUG_ON(!PageLocked(old_page)); /* * Since we dropped the lock we need to revalidate @@ -1980,9 +1989,11 @@ static int do_wp_page(struct mm_struct *mm, struct vm_area_struct *vma, */ page_table = pte_offset_map_lock(mm, pmd, address, &ptl); - page_cache_release(old_page); - if (!pte_same(*page_table, orig_pte)) + if (!pte_same(*page_table, orig_pte)) { + unlock_page(old_page); + page_cache_release(old_page); goto unlock; + } page_mkwrite = 1; } @@ -2094,9 +2105,6 @@ gotten: unlock: pte_unmap_unlock(page_table, ptl); if (dirty_page) { - if (vma->vm_file) - file_update_time(vma->vm_file); - /* * Yes, Virginia, this is actually required to prevent a race * with clear_page_dirty_for_io() from clearing the page dirty @@ -2105,16 +2113,41 @@ unlock: * * do_no_page is protected similarly. */ - wait_on_page_locked(dirty_page); - set_page_dirty_balance(dirty_page, page_mkwrite); + if (!page_mkwrite) { + wait_on_page_locked(dirty_page); + set_page_dirty_balance(dirty_page, page_mkwrite); + } put_page(dirty_page); + if (page_mkwrite) { + struct address_space *mapping = dirty_page->mapping; + + set_page_dirty(dirty_page); + unlock_page(dirty_page); + page_cache_release(dirty_page); + if (mapping) { + /* + * Some device drivers do not set page.mapping + * but still dirty their pages + */ + balance_dirty_pages_ratelimited(mapping); + } + } + + /* file_update_time outside page_lock */ + if (vma->vm_file) + file_update_time(vma->vm_file); } return ret; oom_free_new: page_cache_release(new_page); oom: - if (old_page) + if (old_page) { + if (page_mkwrite) { + unlock_page(old_page); + page_cache_release(old_page); + } page_cache_release(old_page); + } return VM_FAULT_OOM; unwritable_page: @@ -2664,27 +2697,22 @@ static int __do_fault(struct mm_struct *mm, struct vm_area_struct *vma, int tmp; unlock_page(page); - vmf.flags |= FAULT_FLAG_MKWRITE; + vmf.flags = FAULT_FLAG_WRITE|FAULT_FLAG_MKWRITE; tmp = vma->vm_ops->page_mkwrite(vma, &vmf); if (unlikely(tmp & (VM_FAULT_ERROR | VM_FAULT_NOPAGE))) { ret = tmp; - anon = 1; /* no anon but release vmf.page */ - goto out_unlocked; - } - lock_page(page); - /* - * XXX: this is not quite right (racy vs - * invalidate) to unlock and relock the page - * like this, however a better fix requires - * reworking page_mkwrite locking API, which - * is better done later. - */ - if (!page->mapping) { - ret = 0; - anon = 1; /* no anon but release vmf.page */ - goto out; + goto unwritable_page; } + if (unlikely(!(tmp & VM_FAULT_LOCKED))) { + lock_page(page); + if (!page->mapping) { + ret = 0; /* retry the fault */ + unlock_page(page); + goto unwritable_page; + } + } else + VM_BUG_ON(!PageLocked(page)); page_mkwrite = 1; } } @@ -2736,19 +2764,35 @@ static int __do_fault(struct mm_struct *mm, struct vm_area_struct *vma, pte_unmap_unlock(page_table, ptl); out: - unlock_page(vmf.page); -out_unlocked: - if (anon) - page_cache_release(vmf.page); - else if (dirty_page) { - if (vma->vm_file) - file_update_time(vma->vm_file); + if (dirty_page) { + struct address_space *mapping = page->mapping; - set_page_dirty_balance(dirty_page, page_mkwrite); + if (set_page_dirty(dirty_page)) + page_mkwrite = 1; + unlock_page(dirty_page); put_page(dirty_page); + if (page_mkwrite && mapping) { + /* + * Some device drivers do not set page.mapping but still + * dirty their pages + */ + balance_dirty_pages_ratelimited(mapping); + } + + /* file_update_time outside page_lock */ + if (vma->vm_file) + file_update_time(vma->vm_file); + } else { + unlock_page(vmf.page); + if (anon) + page_cache_release(vmf.page); } return ret; + +unwritable_page: + page_cache_release(page); + return ret; } static int do_linear_fault(struct mm_struct *mm, struct vm_area_struct *vma, -- cgit v1.2.3 From ae3abae64f177586be55b04a7fb7047a34b21a3e Mon Sep 17 00:00:00 2001 From: Daisuke Nishimura Date: Thu, 30 Apr 2009 15:08:19 -0700 Subject: memcg: fix mem_cgroup_shrink_usage() Current mem_cgroup_shrink_usage() has two problems. 1. It doesn't call mem_cgroup_out_of_memory and doesn't update last_oom_jiffies, so pagefault_out_of_memory invokes global OOM. 2. Considering hierarchy, shrinking has to be done from the mem_over_limit, not from the memcg which the page would be charged to. mem_cgroup_try_charge_swapin() does all of these things properly, so we use it and call cancel_charge_swapin when it succeeded. The name of "shrink_usage" is not appropriate for this behavior, so we change it too. Signed-off-by: Daisuke Nishimura Acked-by: KAMEZAWA Hiroyuki Cc: Li Zefan Cc: Paul Menage Cc: Dhaval Giani Cc: Daisuke Nishimura Cc: YAMAMOTO Takashi Cc: KOSAKI Motohiro Cc: David Rientjes Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/memcontrol.c | 33 ++++++++++++--------------------- mm/shmem.c | 8 ++++++-- 2 files changed, 18 insertions(+), 23 deletions(-) (limited to 'mm') diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 575203ae210..01c2d8f1468 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -1617,37 +1617,28 @@ void mem_cgroup_end_migration(struct mem_cgroup *mem, } /* - * A call to try to shrink memory usage under specified resource controller. - * This is typically used for page reclaiming for shmem for reducing side - * effect of page allocation from shmem, which is used by some mem_cgroup. + * A call to try to shrink memory usage on charge failure at shmem's swapin. + * Calling hierarchical_reclaim is not enough because we should update + * last_oom_jiffies to prevent pagefault_out_of_memory from invoking global OOM. + * Moreover considering hierarchy, we should reclaim from the mem_over_limit, + * not from the memcg which this page would be charged to. + * try_charge_swapin does all of these works properly. */ -int mem_cgroup_shrink_usage(struct page *page, +int mem_cgroup_shmem_charge_fallback(struct page *page, struct mm_struct *mm, gfp_t gfp_mask) { struct mem_cgroup *mem = NULL; - int progress = 0; - int retry = MEM_CGROUP_RECLAIM_RETRIES; + int ret; if (mem_cgroup_disabled()) return 0; - if (page) - mem = try_get_mem_cgroup_from_swapcache(page); - if (!mem && mm) - mem = try_get_mem_cgroup_from_mm(mm); - if (unlikely(!mem)) - return 0; - do { - progress = mem_cgroup_hierarchical_reclaim(mem, - gfp_mask, true, false); - progress += mem_cgroup_check_under_limit(mem); - } while (!progress && --retry); + ret = mem_cgroup_try_charge_swapin(mm, page, gfp_mask, &mem); + if (!ret) + mem_cgroup_cancel_charge_swapin(mem); /* it does !mem check */ - css_put(&mem->css); - if (!retry) - return -ENOMEM; - return 0; + return ret; } static DEFINE_MUTEX(set_limit_mutex); diff --git a/mm/shmem.c b/mm/shmem.c index f9cb20ebb99..b25f95ce3db 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -1340,8 +1340,12 @@ repeat: shmem_swp_unmap(entry); spin_unlock(&info->lock); if (error == -ENOMEM) { - /* allow reclaim from this memory cgroup */ - error = mem_cgroup_shrink_usage(swappage, + /* + * reclaim from proper memory cgroup and + * call memcg's OOM if needed. + */ + error = mem_cgroup_shmem_charge_fallback( + swappage, current->mm, gfp); if (error) { -- cgit v1.2.3 From 00a62ce91e554198ef28234c91c36f850f5a3bc9 Mon Sep 17 00:00:00 2001 From: KOSAKI Motohiro Date: Thu, 30 Apr 2009 15:08:51 -0700 Subject: mm: fix Committed_AS underflow on large NR_CPUS environment The Committed_AS field can underflow in certain situations: > # while true; do cat /proc/meminfo | grep _AS; sleep 1; done | uniq -c > 1 Committed_AS: 18446744073709323392 kB > 11 Committed_AS: 18446744073709455488 kB > 6 Committed_AS: 35136 kB > 5 Committed_AS: 18446744073709454400 kB > 7 Committed_AS: 35904 kB > 3 Committed_AS: 18446744073709453248 kB > 2 Committed_AS: 34752 kB > 9 Committed_AS: 18446744073709453248 kB > 8 Committed_AS: 34752 kB > 3 Committed_AS: 18446744073709320960 kB > 7 Committed_AS: 18446744073709454080 kB > 3 Committed_AS: 18446744073709320960 kB > 5 Committed_AS: 18446744073709454080 kB > 6 Committed_AS: 18446744073709320960 kB Because NR_CPUS can be greater than 1000 and meminfo_proc_show() does not check for underflow. But NR_CPUS proportional isn't good calculation. In general, possibility of lock contention is proportional to the number of online cpus, not theorical maximum cpus (NR_CPUS). The current kernel has generic percpu-counter stuff. using it is right way. it makes code simplify and percpu_counter_read_positive() don't make underflow issue. Reported-by: Dave Hansen Signed-off-by: KOSAKI Motohiro Cc: Eric B Munson Cc: Mel Gorman Cc: Christoph Lameter Cc: [All kernel versions] Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/mmap.c | 12 ++++++------ mm/nommu.c | 13 +++++++------ mm/swap.c | 46 ---------------------------------------------- 3 files changed, 13 insertions(+), 58 deletions(-) (limited to 'mm') diff --git a/mm/mmap.c b/mm/mmap.c index 3303d1ba8e8..6b7b1a95944 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -85,7 +85,7 @@ EXPORT_SYMBOL(vm_get_page_prot); int sysctl_overcommit_memory = OVERCOMMIT_GUESS; /* heuristic overcommit */ int sysctl_overcommit_ratio = 50; /* default is 50% */ int sysctl_max_map_count __read_mostly = DEFAULT_MAX_MAP_COUNT; -atomic_long_t vm_committed_space = ATOMIC_LONG_INIT(0); +struct percpu_counter vm_committed_as; /* * Check that a process has enough memory to allocate a new virtual @@ -179,11 +179,7 @@ int __vm_enough_memory(struct mm_struct *mm, long pages, int cap_sys_admin) if (mm) allowed -= mm->total_vm / 32; - /* - * cast `allowed' as a signed long because vm_committed_space - * sometimes has a negative value - */ - if (atomic_long_read(&vm_committed_space) < (long)allowed) + if (percpu_counter_read_positive(&vm_committed_as) < allowed) return 0; error: vm_unacct_memory(pages); @@ -2481,4 +2477,8 @@ void mm_drop_all_locks(struct mm_struct *mm) */ void __init mmap_init(void) { + int ret; + + ret = percpu_counter_init(&vm_committed_as, 0); + VM_BUG_ON(ret); } diff --git a/mm/nommu.c b/mm/nommu.c index 72eda4aee2c..809998aa7b5 100644 --- a/mm/nommu.c +++ b/mm/nommu.c @@ -62,7 +62,7 @@ void *high_memory; struct page *mem_map; unsigned long max_mapnr; unsigned long num_physpages; -atomic_long_t vm_committed_space = ATOMIC_LONG_INIT(0); +struct percpu_counter vm_committed_as; int sysctl_overcommit_memory = OVERCOMMIT_GUESS; /* heuristic overcommit */ int sysctl_overcommit_ratio = 50; /* default is 50% */ int sysctl_max_map_count = DEFAULT_MAX_MAP_COUNT; @@ -463,6 +463,10 @@ SYSCALL_DEFINE1(brk, unsigned long, brk) */ void __init mmap_init(void) { + int ret; + + ret = percpu_counter_init(&vm_committed_as, 0); + VM_BUG_ON(ret); vm_region_jar = KMEM_CACHE(vm_region, SLAB_PANIC); } @@ -1847,12 +1851,9 @@ int __vm_enough_memory(struct mm_struct *mm, long pages, int cap_sys_admin) if (mm) allowed -= mm->total_vm / 32; - /* - * cast `allowed' as a signed long because vm_committed_space - * sometimes has a negative value - */ - if (atomic_long_read(&vm_committed_space) < (long)allowed) + if (percpu_counter_read_positive(&vm_committed_as) < allowed) return 0; + error: vm_unacct_memory(pages); diff --git a/mm/swap.c b/mm/swap.c index bede23ce64e..cb29ae5d33a 100644 --- a/mm/swap.c +++ b/mm/swap.c @@ -491,49 +491,6 @@ unsigned pagevec_lookup_tag(struct pagevec *pvec, struct address_space *mapping, EXPORT_SYMBOL(pagevec_lookup_tag); -#ifdef CONFIG_SMP -/* - * We tolerate a little inaccuracy to avoid ping-ponging the counter between - * CPUs - */ -#define ACCT_THRESHOLD max(16, NR_CPUS * 2) - -static DEFINE_PER_CPU(long, committed_space); - -void vm_acct_memory(long pages) -{ - long *local; - - preempt_disable(); - local = &__get_cpu_var(committed_space); - *local += pages; - if (*local > ACCT_THRESHOLD || *local < -ACCT_THRESHOLD) { - atomic_long_add(*local, &vm_committed_space); - *local = 0; - } - preempt_enable(); -} - -#ifdef CONFIG_HOTPLUG_CPU - -/* Drop the CPU's cached committed space back into the central pool. */ -static int cpu_swap_callback(struct notifier_block *nfb, - unsigned long action, - void *hcpu) -{ - long *committed; - - committed = &per_cpu(committed_space, (long)hcpu); - if (action == CPU_DEAD || action == CPU_DEAD_FROZEN) { - atomic_long_add(*committed, &vm_committed_space); - *committed = 0; - drain_cpu_pagevecs((long)hcpu); - } - return NOTIFY_OK; -} -#endif /* CONFIG_HOTPLUG_CPU */ -#endif /* CONFIG_SMP */ - /* * Perform any setup for the swap system */ @@ -554,7 +511,4 @@ void __init swap_setup(void) * Right now other parts of the system means that we * _really_ don't want to cluster much more */ -#ifdef CONFIG_HOTPLUG_CPU - hotcpu_notifier(cpu_swap_callback, 0); -#endif } -- cgit v1.2.3 From 8713e01295140f674a41f2199b0f7ca99dfb69d5 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Thu, 30 Apr 2009 15:08:55 -0700 Subject: vmscan: avoid multiplication overflow in shrink_zone() Local variable `scan' can overflow on zones which are larger than (2G * 4k) / 100 = 80GB. Making it 64-bit on 64-bit will fix that up. Cc: KOSAKI Motohiro Cc: Wu Fengguang Cc: Peter Zijlstra Cc: Rik van Riel Cc: Lee Schermerhorn Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/vmscan.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'mm') diff --git a/mm/vmscan.c b/mm/vmscan.c index eac9577941f..5fa3eda1f03 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -1471,7 +1471,7 @@ static void shrink_zone(int priority, struct zone *zone, for_each_evictable_lru(l) { int file = is_file_lru(l); - int scan; + unsigned long scan; scan = zone_nr_pages(zone, sc, l); if (priority) { -- cgit v1.2.3 From a425a638c858fd10370b573bde81df3ba500e271 Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Tue, 5 May 2009 16:37:17 +0100 Subject: Ignore madvise(MADV_WILLNEED) for hugetlbfs-backed regions madvise(MADV_WILLNEED) forces page cache readahead on a range of memory backed by a file. The assumption is made that the page required is order-0 and "normal" page cache. On hugetlbfs, this assumption is not true and order-0 pages are allocated and inserted into the hugetlbfs page cache. This leaks hugetlbfs page reservations and can cause BUGs to trigger related to corrupted page tables. This patch causes MADV_WILLNEED to be ignored for hugetlbfs-backed regions. Signed-off-by: Mel Gorman Cc: stable@kernel.org Signed-off-by: Linus Torvalds --- mm/madvise.c | 8 ++++++++ 1 file changed, 8 insertions(+) (limited to 'mm') diff --git a/mm/madvise.c b/mm/madvise.c index b9ce574827c..36d6ea2b634 100644 --- a/mm/madvise.c +++ b/mm/madvise.c @@ -112,6 +112,14 @@ static long madvise_willneed(struct vm_area_struct * vma, if (!file) return -EBADF; + /* + * Page cache readahead assumes page cache pages are order-0 which + * is not the case for hugetlbfs. Do not give a bad return value + * but ignore the advice. + */ + if (vma->vm_flags & VM_HUGETLB) + return 0; + if (file->f_mapping->a_ops->get_xip_mem) { /* no bad return value, but ignore advice */ return 0; -- cgit v1.2.3