I can reproduce this WARNING if I run rsync -rtS -@2 --no-compress
--no-i-r --info=progress
2 /mnt/f2fs/ /mnt/sd/
where /mnt/f2fs/ is emmc card with f2fs and /mnt/sd/ external SD card.
And in same time start create many files in /mnt/f2fs/
[ 1264.858858] WARNING: possible circular locking dependency detected
[ 1264.865047] 5.4.31 #1 Not tainted
[ 1264.868365] ------------------------------------------------------
[ 1264.874549] kswapd0/86 is trying to acquire lock:
[ 1264.879258] e93fab88 (sb_internal){.+.+}, at:
f2fs_evict_inode+0x234/0x514 [f2fs]
[ 1264.887012]
[ 1264.887012] but task is already holding lock:
[ 1264.892854] c0c270f8 (fs_reclaim){+.+.}, at: __fs_reclaim_acquire+0x0/0x40
[ 1264.899750]
[ 1264.899750] which lock already depends on the new lock.
[ 1264.899750]
[ 1264.907931]
[ 1264.907931] the existing dependency chain (in reverse order) is:
[ 1264.915416]
[ 1264.915416] -> #2 (fs_reclaim){+.+.}:
[ 1264.920570] __alloc_pages_nodemask+0x1c8/0x1384
[ 1264.925717] pagecache_get_page+0x98/0x358
[ 1264.930552] f2fs_move_inline_dirents+0x44/0x838 [f2fs]
[ 1264.936499] f2fs_try_convert_inline_dir+0x10c/0x21c [f2fs]
[ 1264.942786] f2fs_rename2+0x3e8/0x1138 [f2fs]
[ 1264.947678] vfs_rename+0x210/0x7d4
[ 1264.951696] do_renameat2+0x470/0x4ec
[ 1264.955887] sys_rename+0x24/0x2c
[ 1264.959730] ret_fast_syscall+0x0/0x28
[ 1264.964005] 0xb691d704
[ 1264.966977]
[ 1264.966977] -> #1 (&sbi->cp_rwsem){++++}:
[ 1264.972670] f2fs_truncate_blocks+0x2f8/0x6d8 [f2fs]
[ 1264.978352] f2fs_truncate+0x14c/0x2a4 [f2fs]
[ 1264.983424] f2fs_evict_inode+0x2e0/0x514 [f2fs]
[ 1264.988574] evict+0x9c/0x164
[ 1264.992070] do_unlinkat+0x190/0x2a8
[ 1264.996172] ret_fast_syscall+0x0/0x28
[ 1265.000445] 0xb691d73c
[ 1265.003415]
[ 1265.003415] -> #0 (sb_internal){.+.+}:
[ 1265.008659] lock_acquire+0xd8/0x1f8
[ 1265.012764] __sb_start_write+0x110/0x210
[ 1265.017495] f2fs_evict_inode+0x234/0x514 [f2fs]
[ 1265.022644] evict+0x9c/0x164
[ 1265.026142] dispose_list+0x38/0x4c
[ 1265.030162] prune_icache_sb+0x5c/0x7c
[ 1265.034438] super_cache_scan+0x118/0x170
[ 1265.038980] shrink_slab.constprop.24+0x2b8/0x608
[ 1265.044215] shrink_node+0x42c/0x958
[ 1265.048316] kswapd+0x398/0xa04
[ 1265.051986] kthread+0x148/0x150
[ 1265.055742] ret_from_fork+0x14/0x20
[ 1265.059842] 0x0
[ 1265.062205]
[ 1265.062205] other info that might help us debug this:
[ 1265.062205]
[ 1265.070214] Chain exists of:
[ 1265.070214] sb_internal --> &sbi->cp_rwsem --> fs_reclaim
[ 1265.070214]
[ 1265.080138] Possible unsafe locking scenario:
[ 1265.080138]
[ 1265.086062] CPU0 CPU1
[ 1265.090595] ---- ----
[ 1265.095128] lock(fs_reclaim);
[ 1265.098276] lock(&sbi->cp_rwsem);
[ 1265.104287] lock(fs_reclaim);
[ 1265.109950] lock(sb_internal);
[ 1265.113183]
[ 1265.113183] *** DEADLOCK ***
[ 1265.113183]
[ 1265.119108] 3 locks held by kswapd0/86:
[ 1265.122948] #0: c0c270f8 (fs_reclaim){+.+.}, at:
__fs_reclaim_acquire+0x0/0x40
[ 1265.130273] #1: c0c1ead0 (shrinker_rwsem){++++}, at:
shrink_slab.constprop.24+0x38/0x608
[ 1265.138467] #2: e93fa87c (&type->s_umount_key#29){++++}, at:
trylock_super+0x14/0x58
[ 1265.146312]
[ 1265.146312] stack backtrace:
[ 1265.150679] CPU: 0 PID: 86 Comm: kswapd0 Not tainted 5.4.31 #1
[ 1265.156518] Hardware name: Freescale i.MX6 Quad/DualLite (Device Tree)
[ 1265.163063] [<c01116f8>] (unwind_backtrace) from [<c010c3fc>]
(show_stack+0x10/0x14)
[ 1265.170820] [<c010c3fc>] (show_stack) from [<c07dc8b0>]
(dump_stack+0xb4/0xe0)
[ 1265.178059] [<c07dc8b0>] (dump_stack) from [<c01747b4>]
(check_noncircular+0x1e4/0x1fc)
[ 1265.186076] [<c01747b4>] (check_noncircular) from [<c0176dbc>]
(__lock_acquire+0x15a8/0x25c4)
[ 1265.194613] [<c0176dbc>] (__lock_acquire) from [<c017867c>]
(lock_acquire+0xd8/0x1f8)
[ 1265.202452] [<c017867c>] (lock_acquire) from [<c0269e4c>]
(__sb_start_write+0x110/0x210)
[ 1265.210755] [<c0269e4c>] (__sb_start_write) from [<bf2cd8c0>]
(f2fs_evict_inode+0x234/0x514 [f2fs])
[ 1265.220013] [<bf2cd8c0>] (f2fs_evict_inode [f2fs]) from
[<c0286060>] (evict+0x9c/0x164)
[ 1265.228033] [<c0286060>] (evict) from [<c0286160>] (dispose_list+0x38/0x4c)
[ 1265.235005] [<c0286160>] (dispose_list) from [<c02871a0>]
(prune_icache_sb+0x5c/0x7c)
[ 1265.242847] [<c02871a0>] (prune_icache_sb) from [<c026a0bc>]
(super_cache_scan+0x118/0x170)
[ 1265.251214] [<c026a0bc>] (super_cache_scan) from [<c0215478>]
(shrink_slab.constprop.24+0x2b8/0x608)
[ 1265.260359] [<c0215478>] (shrink_slab.constprop.24) from
[<c02193dc>] (shrink_node+0x42c/0x958)
[ 1265.269065] [<c02193dc>] (shrink_node) from [<c021a4f8>] (kswapd+0x398/0xa04)
[ 1265.276210] [<c021a4f8>] (kswapd) from [<c01461e4>] (kthread+0x148/0x150)
[ 1265.283009] [<c01461e4>] (kthread) from [<c01010b4>]
(ret_from_fork+0x14/0x20)
[ 1265.290236] Exception stack(0xe8421fb0 to 0xe8421ff8)
[ 1265.295297] 1fa0: 00000000
00000000 00000000 00000000
[ 1265.303482] 1fc0: 00000000 00000000 00000000 00000000 00000000
00000000 00000000 00000000
[ 1265.311665] 1fe0: 00000000 00000000 00000000 00000000 00000013 00000000
_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
