[syzbot] [xfs?] possible deadlock in xfs_icwalk_ag (3)

[syzbot <syzbot+789028412a4af61a2b61@syzkaller.appspotmail.com>]

Hello,

syzbot found the following issue on:

HEAD commit:    6e64f4580381 Merge tag 'input-for-v6.17-rc0' of git://git...
git tree:       upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=170e0ea2580000
kernel config:  https://syzkaller.appspot.com/x/.config?x=ff0ac94f5fb505cf
dashboard link: https://syzkaller.appspot.com/bug?extid=789028412a4af61a2b61
compiler:       Debian clang version 20.1.7 (++20250616065708+6146a88f6049-1~exp1~20250616065826.132), Debian LLD 20.1.7

Unfortunately, I don't have any reproducer for this issue yet.

Downloadable assets:
disk image (non-bootable): https://storage.googleapis.com/syzbot-assets/d900f083ada3/non_bootable_disk-6e64f458.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/6b0d7c92b652/vmlinux-6e64f458.xz
kernel image: https://storage.googleapis.com/syzbot-assets/541b13915f7e/bzImage-6e64f458.xz

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+789028412a4af61a2b61@syzkaller.appspotmail.com

loop0: detected capacity change from 0 to 32768
XFS (loop0): DAX unsupported by block device. Turning off DAX.
XFS (loop0): Mounting V5 Filesystem c496e05e-540d-4c72-b591-04d79d8b4eeb
XFS (loop0): Ending clean mount
XFS (loop0): Quotacheck needed: Please wait.
XFS (loop0): Quotacheck: Done.
============================================
WARNING: possible recursive locking detected
6.16.0-syzkaller-11952-g6e64f4580381 #0 Not tainted
--------------------------------------------
syz.0.0/5359 is trying to acquire lock:
ffff88805250f758 (&xfs_nondir_ilock_class){++++}-{4:4}, at: xfs_reclaim_inode fs/xfs/xfs_icache.c:1042 [inline]
ffff88805250f758 (&xfs_nondir_ilock_class){++++}-{4:4}, at: xfs_icwalk_process_inode fs/xfs/xfs_icache.c:1734 [inline]
ffff88805250f758 (&xfs_nondir_ilock_class){++++}-{4:4}, at: xfs_icwalk_ag+0x12c5/0x1ab0 fs/xfs/xfs_icache.c:1816

but task is already holding lock:
ffff8880525327d8 (&xfs_nondir_ilock_class){++++}-{4:4}, at: xfs_bmap_punch_delalloc_range+0x26d/0x7c0 fs/xfs/xfs_bmap_util.c:452

other info that might help us debug this:
 Possible unsafe locking scenario:

       CPU0
       ----
  lock(&xfs_nondir_ilock_class);
  lock(&xfs_nondir_ilock_class);

 *** DEADLOCK ***

 May be due to missing lock nesting notation

4 locks held by syz.0.0/5359:
 #0: ffff8880525329f0 (&sb->s_type->i_mutex_key#20){+.+.}-{4:4}, at: xfs_ilock+0xfe/0x390 fs/xfs/xfs_inode.c:149
 #1: ffff888052532b90 (mapping.invalidate_lock#3){+.+.}-{4:4}, at: filemap_invalidate_lock include/linux/fs.h:924 [inline]
 #1: ffff888052532b90 (mapping.invalidate_lock#3){+.+.}-{4:4}, at: xfs_buffered_write_iomap_end+0x2b6/0x4c0 fs/xfs/xfs_iomap.c:1993
 #2: ffff8880525327d8 (&xfs_nondir_ilock_class){++++}-{4:4}, at: xfs_bmap_punch_delalloc_range+0x26d/0x7c0 fs/xfs/xfs_bmap_util.c:452
 #3: ffff888043ac40e0 (&type->s_umount_key#50){.+.+}-{4:4}, at: super_trylock_shared fs/super.c:563 [inline]
 #3: ffff888043ac40e0 (&type->s_umount_key#50){.+.+}-{4:4}, at: super_cache_scan+0x91/0x4b0 fs/super.c:197

stack backtrace:
CPU: 0 UID: 0 PID: 5359 Comm: syz.0.0 Not tainted 6.16.0-syzkaller-11952-g6e64f4580381 #0 PREEMPT(full) 
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2~bpo12+1 04/01/2014
Call Trace:
 <TASK>
 dump_stack_lvl+0x189/0x250 lib/dump_stack.c:120
 print_deadlock_bug+0x28b/0x2a0 kernel/locking/lockdep.c:3041
 check_deadlock kernel/locking/lockdep.c:3093 [inline]
 validate_chain+0x1a3f/0x2140 kernel/locking/lockdep.c:3895
 __lock_acquire+0xab9/0xd20 kernel/locking/lockdep.c:5237
 lock_acquire+0x120/0x360 kernel/locking/lockdep.c:5868
 down_write_nested+0x9d/0x200 kernel/locking/rwsem.c:1706
 xfs_reclaim_inode fs/xfs/xfs_icache.c:1042 [inline]
 xfs_icwalk_process_inode fs/xfs/xfs_icache.c:1734 [inline]
 xfs_icwalk_ag+0x12c5/0x1ab0 fs/xfs/xfs_icache.c:1816
 xfs_icwalk fs/xfs/xfs_icache.c:1864 [inline]
 xfs_reclaim_inodes_nr+0x1e3/0x260 fs/xfs/xfs_icache.c:1108
 super_cache_scan+0x41b/0x4b0 fs/super.c:228
 do_shrink_slab+0x6ef/0x1110 mm/shrinker.c:437
 shrink_slab+0xd74/0x10d0 mm/shrinker.c:664
 shrink_one+0x28a/0x7c0 mm/vmscan.c:4954
 shrink_many mm/vmscan.c:5015 [inline]
 lru_gen_shrink_node mm/vmscan.c:5093 [inline]
 shrink_node+0x314e/0x3760 mm/vmscan.c:6078
 shrink_zones mm/vmscan.c:6336 [inline]
 do_try_to_free_pages+0x668/0x1960 mm/vmscan.c:6398
 try_to_free_pages+0x8a2/0xdd0 mm/vmscan.c:6644
 __perform_reclaim mm/page_alloc.c:4310 [inline]
 __alloc_pages_direct_reclaim+0x144/0x300 mm/page_alloc.c:4332
 __alloc_pages_slowpath+0x5ff/0xce0 mm/page_alloc.c:4781
 __alloc_frozen_pages_noprof+0x319/0x370 mm/page_alloc.c:5161
 alloc_pages_mpol+0x232/0x4a0 mm/mempolicy.c:2416
 alloc_frozen_pages_noprof mm/mempolicy.c:2487 [inline]
 alloc_pages_noprof+0xa9/0x190 mm/mempolicy.c:2507
 stack_depot_save_flags+0x777/0x860 lib/stackdepot.c:677
 kasan_save_stack mm/kasan/common.c:48 [inline]
 kasan_save_track+0x4f/0x80 mm/kasan/common.c:68
 poison_kmalloc_redzone mm/kasan/common.c:388 [inline]
 __kasan_kmalloc+0x93/0xb0 mm/kasan/common.c:405
 kasan_kmalloc include/linux/kasan.h:260 [inline]
 __do_kmalloc_node mm/slub.c:4365 [inline]
 __kmalloc_node_track_caller_noprof+0x271/0x4e0 mm/slub.c:4384
 __do_krealloc mm/slub.c:4942 [inline]
 krealloc_noprof+0x124/0x340 mm/slub.c:4995
 xfs_iext_realloc_root fs/xfs/libxfs/xfs_iext_tree.c:613 [inline]
 xfs_iext_insert_raw+0x131/0x3260 fs/xfs/libxfs/xfs_iext_tree.c:647
 xfs_iext_insert+0x36/0x220 fs/xfs/libxfs/xfs_iext_tree.c:684
 xfs_bmap_del_extent_delay+0x105b/0x15b0 fs/xfs/libxfs/xfs_bmap.c:4787
 xfs_bmap_punch_delalloc_range+0x536/0x7c0 fs/xfs/xfs_bmap_util.c:483
 xfs_buffered_write_iomap_end+0x2d2/0x4c0 fs/xfs/xfs_iomap.c:1994
 iomap_iter+0x316/0xde0 fs/iomap/iter.c:79
 iomap_file_buffered_write+0x7fa/0x9b0 fs/iomap/buffered-io.c:1065
 xfs_file_buffered_write+0x209/0x8a0 fs/xfs/xfs_file.c:981
 aio_write+0x535/0x7a0 fs/aio.c:1634
 __io_submit_one fs/aio.c:-1 [inline]
 io_submit_one+0x78b/0x1310 fs/aio.c:2053
 __do_sys_io_submit fs/aio.c:2112 [inline]
 __se_sys_io_submit+0x185/0x2f0 fs/aio.c:2082
 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
 do_syscall_64+0xfa/0x3b0 arch/x86/entry/syscall_64.c:94
 entry_SYSCALL_64_after_hwframe+0x77/0x7f
RIP: 0033:0x7f17f498ebe9
Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 a8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f17f5846038 EFLAGS: 00000246 ORIG_RAX: 00000000000000d1
RAX: ffffffffffffffda RBX: 00007f17f4bb5fa0 RCX: 00007f17f498ebe9
RDX: 0000200000000540 RSI: 0000000000000008 RDI: 00007f17f5804000
RBP: 00007f17f4a11e19 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 00007f17f4bb6038 R14: 00007f17f4bb5fa0 R15: 00007ffe6a2d5798
 </TASK>
syz.0.0 (5359) used greatest stack depth: 19048 bytes left


---
This report is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzkaller@googlegroups.com.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.

If the report is already addressed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want to overwrite report's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the report is a duplicate of another one, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup