[cc linux-xfs]
On Fri, May 22, 2020 at 08:21:50AM +1000, Dave Airlie wrote:
> Hi,
>
> Just updated a rawhide VM to the Fedora 5.7.0-rc5 kernel, did some
> package building,
>
> got the below trace, not sure if it's known and fixed or unknown.
It's a known false-positive. An inode can't simultaneously be getting
reclaimed due to zero refcount /and/ be the target of a getxattr call.
Unfortunately, lockdep can't tell the difference, and it seems a little
strange to set NOFS on the allocation (which increases the chances of a
runtime error) just to quiet that down.
--D
> Dave.
>
> 725.862536] ======================================================
> [ 725.862564] WARNING: possible circular locking dependency detected
> [ 725.862591] 5.7.0-0.rc5.20200515git1ae7efb38854.1.fc33.x86_64 #1 Not tainted
> [ 725.862612] ------------------------------------------------------
> [ 725.862630] kswapd0/159 is trying to acquire lock:
> [ 725.862645] ffff9b38d01a4470 (&xfs_nondir_ilock_class){++++}-{3:3},
> at: xfs_ilock+0xde/0x2c0 [xfs]
> [ 725.862718]
> but task is already holding lock:
> [ 725.862735] ffffffffbbb8bd00 (fs_reclaim){+.+.}-{0:0}, at:
> __fs_reclaim_acquire+0x5/0x30
> [ 725.862762]
> which lock already depends on the new lock.
>
> [ 725.862785]
> the existing dependency chain (in reverse order) is:
> [ 725.862806]
> -> #1 (fs_reclaim){+.+.}-{0:0}:
> [ 725.862824] fs_reclaim_acquire+0x34/0x40
> [ 725.862839] __kmalloc+0x4f/0x270
> [ 725.862878] kmem_alloc+0x93/0x1d0 [xfs]
> [ 725.862914] kmem_alloc_large+0x4c/0x130 [xfs]
> [ 725.862945] xfs_attr_copy_value+0x74/0xa0 [xfs]
> [ 725.862984] xfs_attr_get+0x9d/0xc0 [xfs]
> [ 725.863021] xfs_get_acl+0xb6/0x200 [xfs]
> [ 725.863036] get_acl+0x81/0x160
> [ 725.863052] posix_acl_xattr_get+0x3f/0xd0
> [ 725.863067] vfs_getxattr+0x148/0x170
> [ 725.863081] getxattr+0xa7/0x240
> [ 725.863093] path_getxattr+0x52/0x80
> [ 725.863111] do_syscall_64+0x5c/0xa0
> [ 725.863133] entry_SYSCALL_64_after_hwframe+0x49/0xb3
> [ 725.863149]
> -> #0 (&xfs_nondir_ilock_class){++++}-{3:3}:
> [ 725.863177] __lock_acquire+0x1257/0x20d0
> [ 725.863193] lock_acquire+0xb0/0x310
> [ 725.863207] down_write_nested+0x49/0x120
> [ 725.863242] xfs_ilock+0xde/0x2c0 [xfs]
> [ 725.863277] xfs_reclaim_inode+0x3f/0x400 [xfs]
> [ 725.863312] xfs_reclaim_inodes_ag+0x20b/0x410 [xfs]
> [ 725.863351] xfs_reclaim_inodes_nr+0x31/0x40 [xfs]
> [ 725.863368] super_cache_scan+0x190/0x1e0
> [ 725.863383] do_shrink_slab+0x184/0x420
> [ 725.863397] shrink_slab+0x182/0x290
> [ 725.863409] shrink_node+0x174/0x680
> [ 725.863927] balance_pgdat+0x2d0/0x5f0
> [ 725.864389] kswapd+0x21f/0x510
> [ 725.864836] kthread+0x131/0x150
> [ 725.865277] ret_from_fork+0x3a/0x50
> [ 725.865707]
> other info that might help us debug this:
>
> [ 725.866953] Possible unsafe locking scenario:
>
> [ 725.867764] CPU0 CPU1
> [ 725.868161] ---- ----
> [ 725.868531] lock(fs_reclaim);
> [ 725.868896] lock(&xfs_nondir_ilock_class);
> [ 725.869276] lock(fs_reclaim);
> [ 725.869633] lock(&xfs_nondir_ilock_class);
> [ 725.869996]
> *** DEADLOCK ***
>
> [ 725.871061] 4 locks held by kswapd0/159:
> [ 725.871406] #0: ffffffffbbb8bd00 (fs_reclaim){+.+.}-{0:0}, at:
> __fs_reclaim_acquire+0x5/0x30
> [ 725.871779] #1: ffffffffbbb7cef8 (shrinker_rwsem){++++}-{3:3}, at:
> shrink_slab+0x115/0x290
> [ 725.872167] #2: ffff9b39f07a50e8
> (&type->s_umount_key#56){++++}-{3:3}, at: super_cache_scan+0x38/0x1e0
> [ 725.872560] #3: ffff9b39f077f258
> (&pag->pag_ici_reclaim_lock){+.+.}-{3:3}, at:
> xfs_reclaim_inodes_ag+0x82/0x410 [xfs]
> [ 725.873013]
> stack backtrace:
> [ 725.873811] CPU: 3 PID: 159 Comm: kswapd0 Not tainted
> 5.7.0-0.rc5.20200515git1ae7efb38854.1.fc33.x86_64 #1
> [ 725.874249] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996),
> BIOS ?-20180724_192412-buildhw-07.phx2.fedoraproject.org-1.fc29
> 04/01/2014
> [ 725.875158] Call Trace:
> [ 725.875625] dump_stack+0x8b/0xc8
> [ 725.876090] check_noncircular+0x134/0x150
> [ 725.876547] __lock_acquire+0x1257/0x20d0
> [ 725.877019] lock_acquire+0xb0/0x310
> [ 725.877517] ? xfs_ilock+0xde/0x2c0 [xfs]
> [ 725.877988] down_write_nested+0x49/0x120
> [ 725.878473] ? xfs_ilock+0xde/0x2c0 [xfs]
> [ 725.878955] ? xfs_reclaim_inode+0x3f/0x400 [xfs]
> [ 725.879448] xfs_ilock+0xde/0x2c0 [xfs]
> [ 725.879925] xfs_reclaim_inode+0x3f/0x400 [xfs]
> [ 725.880414] xfs_reclaim_inodes_ag+0x20b/0x410 [xfs]
> [ 725.880876] ? sched_clock_cpu+0xc/0xb0
> [ 725.881343] ? mark_held_locks+0x2d/0x80
> [ 725.881798] ? _raw_spin_unlock_irqrestore+0x46/0x60
> [ 725.882268] ? lockdep_hardirqs_on+0x11e/0x1b0
> [ 725.882734] ? try_to_wake_up+0x249/0x820
> [ 725.883234] xfs_reclaim_inodes_nr+0x31/0x40 [xfs]
> [ 725.883700] super_cache_scan+0x190/0x1e0
> [ 725.884180] do_shrink_slab+0x184/0x420
> [ 725.884653] shrink_slab+0x182/0x290
> [ 725.885129] shrink_node+0x174/0x680
> [ 725.885596] balance_pgdat+0x2d0/0x5f0
> [ 725.886074] kswapd+0x21f/0x510
> [ 725.886540] ? finish_wait+0x90/0x90
> [ 725.887013] ? balance_pgdat+0x5f0/0x5f0
> [ 725.887477] kthread+0x131/0x150
> [ 725.887937] ? __kthread_bind_mask+0x60/0x60
> [ 725.888410] ret_from_fork+0x3a/0x50
