From 381a80e6df396eaabef2c00f85974a4579ac1c70 Mon Sep 17 00:00:00 2001 From: Wu Fengguang Date: Wed, 6 May 2009 16:02:50 -0700 Subject: inotify: use GFP_NOFS in kernel_event() to work around a lockdep false-positive There is what we believe to be a false positive reported by lockdep. inotify_inode_queue_event() => take inotify_mutex => kernel_event() => kmalloc() => SLOB => alloc_pages_node() => page reclaim => slab reclaim => dcache reclaim => inotify_inode_is_dead => take inotify_mutex => deadlock The plan is to fix this via lockdep annotation, but that is proving to be quite involved. The patch flips the allocation over to GFP_NFS to shut the warning up, for the 2.6.30 release. Hopefully we will fix this for real in 2.6.31. I'll queue a patch in -mm to switch it back to GFP_KERNEL so we don't forget. ================================= [ INFO: inconsistent lock state ] 2.6.30-rc2-next-20090417 #203 --------------------------------- inconsistent {RECLAIM_FS-ON-W} -> {IN-RECLAIM_FS-W} usage. kswapd0/380 [HC0[0]:SC0[0]:HE1:SE1] takes: (&inode->inotify_mutex){+.+.?.}, at: [] inotify_inode_is_dead+0x35/0xb0 {RECLAIM_FS-ON-W} state was registered at: [] mark_held_locks+0x68/0x90 [] lockdep_trace_alloc+0xf5/0x100 [] __kmalloc_node+0x31/0x1e0 [] kernel_event+0xe2/0x190 [] inotify_dev_queue_event+0x126/0x230 [] inotify_inode_queue_event+0xc6/0x110 [] vfs_create+0xcd/0x140 [] do_filp_open+0x88d/0xa20 [] do_sys_open+0x98/0x140 [] sys_open+0x20/0x30 [] system_call_fastpath+0x16/0x1b [] 0xffffffffffffffff irq event stamp: 690455 hardirqs last enabled at (690455): [] _spin_unlock_irqrestore+0x44/0x80 hardirqs last disabled at (690454): [] _spin_lock_irqsave+0x32/0xa0 softirqs last enabled at (690178): [] __do_softirq+0x202/0x220 softirqs last disabled at (690157): [] call_softirq+0x1c/0x50 other info that might help us debug this: 2 locks held by kswapd0/380: #0: (shrinker_rwsem){++++..}, at: [] shrink_slab+0x37/0x180 #1: (&type->s_umount_key#17){++++..}, at: [] shrink_dcache_memory+0x11f/0x1e0 stack backtrace: Pid: 380, comm: kswapd0 Not tainted 2.6.30-rc2-next-20090417 #203 Call Trace: [] print_usage_bug+0x19f/0x200 [] ? save_stack_trace+0x2f/0x50 [] mark_lock+0x4bb/0x6d0 [] ? check_usage_forwards+0x0/0xc0 [] __lock_acquire+0xc62/0x1ae0 [] ? slob_free+0x10c/0x370 [] lock_acquire+0xe1/0x120 [] ? inotify_inode_is_dead+0x35/0xb0 [] mutex_lock_nested+0x63/0x420 [] ? inotify_inode_is_dead+0x35/0xb0 [] ? inotify_inode_is_dead+0x35/0xb0 [] ? sched_clock+0x9/0x10 [] ? lock_release_holdtime+0x35/0x1c0 [] inotify_inode_is_dead+0x35/0xb0 [] dentry_iput+0xbc/0xe0 [] d_kill+0x33/0x60 [] __shrink_dcache_sb+0x2d3/0x350 [] shrink_dcache_memory+0x15a/0x1e0 [] shrink_slab+0x125/0x180 [] kswapd+0x560/0x7a0 [] ? isolate_pages_global+0x0/0x2c0 [] ? autoremove_wake_function+0x0/0x40 [] ? trace_hardirqs_on+0xd/0x10 [] ? kswapd+0x0/0x7a0 [] kthread+0x5b/0xa0 [] child_rip+0xa/0x20 [] ? restore_args+0x0/0x30 [] ? kthread+0x0/0xa0 [] ? child_rip+0x0/0x20 [eparis@redhat.com: fix audit too] Cc: Al Viro Cc: Matt Mackall Cc: Christoph Lameter Signed-off-by: Wu Fengguang Signed-off-by: Eric Paris Cc: Peter Zijlstra Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/notify/inotify/inotify_user.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs/notify') diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index bed766e435b5..1634319e2404 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -220,7 +220,7 @@ static struct inotify_kernel_event * kernel_event(s32 wd, u32 mask, u32 cookie, rem = 0; } - kevent->name = kmalloc(len + rem, GFP_KERNEL); + kevent->name = kmalloc(len + rem, GFP_NOFS); if (unlikely(!kevent->name)) { kmem_cache_free(event_cachep, kevent); return NULL; -- cgit v1.2.3