====================================================== WARNING: possible circular locking dependency detected 6.0.0-rc3-next-20220829 #1 Not tainted ------------------------------------------------------ syz-executor.0/4473 is trying to acquire lock: ffff888010024170 (&journal->j_barrier){+.+.}-{3:3}, at: jbd2_journal_lock_updates+0x15e/0x310 but task is already holding lock: ffff888010022bd0 (&sbi->s_writepages_rwsem){++++}-{0:0}, at: ext4_change_inode_journal_flag+0x177/0x530 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #3 (&sbi->s_writepages_rwsem){++++}-{0:0}: percpu_down_write+0x4d/0x3d0 ext4_ind_migrate+0x237/0x830 ext4_fileattr_set+0x1434/0x18a0 vfs_fileattr_set+0x77c/0xb80 do_vfs_ioctl+0xfc2/0x1610 __x64_sys_ioctl+0x10c/0x210 do_syscall_64+0x3b/0x90 entry_SYSCALL_64_after_hwframe+0x63/0xcd -> #2 (&sb->s_type->i_mutex_key#6){++++}-{3:3}: down_read+0x98/0x450 ext4_bmap+0x4e/0x470 bmap+0xac/0x120 jbd2_journal_bmap+0xa8/0x180 jbd2_journal_flush+0x853/0xc00 __ext4_ioctl+0x9e9/0x4090 __x64_sys_ioctl+0x19a/0x210 do_syscall_64+0x3b/0x90 entry_SYSCALL_64_after_hwframe+0x63/0xcd -> #1 (&journal->j_checkpoint_mutex){+.+.}-{3:3}: mutex_lock_io_nested+0x148/0x1310 jbd2_journal_flush+0x19a/0xc00 __ext4_ioctl+0x9e9/0x4090 __x64_sys_ioctl+0x19a/0x210 do_syscall_64+0x3b/0x90 entry_SYSCALL_64_after_hwframe+0x63/0xcd -> #0 (&journal->j_barrier){+.+.}-{3:3}: __lock_acquire+0x2a02/0x5e70 lock_acquire+0x1a2/0x530 __mutex_lock+0x136/0x14d0 jbd2_journal_lock_updates+0x15e/0x310 ext4_change_inode_journal_flag+0x17f/0x530 ext4_fileattr_set+0x140d/0x18a0 vfs_fileattr_set+0x77c/0xb80 do_vfs_ioctl+0xfc2/0x1610 __x64_sys_ioctl+0x10c/0x210 do_syscall_64+0x3b/0x90 entry_SYSCALL_64_after_hwframe+0x63/0xcd other info that might help us debug this: Chain exists of: &journal->j_barrier --> &sb->s_type->i_mutex_key#6 --> &sbi->s_writepages_rwsem Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&sbi->s_writepages_rwsem); lock(&sb->s_type->i_mutex_key#6); lock(&sbi->s_writepages_rwsem); lock(&journal->j_barrier); *** DEADLOCK *** 4 locks held by syz-executor.0/4473: #0: ffff888010020438 (sb_writers#3){.+.+}-{0:0}, at: do_vfs_ioctl+0xf87/0x1610 #1: ffff88801e372bc0 (&sb->s_type->i_mutex_key#6){++++}-{3:3}, at: vfs_fileattr_set+0x148/0xb80 #2: ffff88801e372d60 (mapping.invalidate_lock){++++}-{3:3}, at: ext4_change_inode_journal_flag+0x11e/0x530 #3: ffff888010022bd0 (&sbi->s_writepages_rwsem){++++}-{0:0}, at: ext4_change_inode_journal_flag+0x177/0x530 stack backtrace: CPU: 1 PID: 4473 Comm: syz-executor.0 Not tainted 6.0.0-rc3-next-20220829 #1 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.14.0-0-g155821a1990b-prebuilt.qemu.org 04/01/2014 Call Trace: dump_stack_lvl+0x8b/0xb3 check_noncircular+0x263/0x2e0 __lock_acquire+0x2a02/0x5e70 lock_acquire+0x1a2/0x530 __mutex_lock+0x136/0x14d0 jbd2_journal_lock_updates+0x15e/0x310 ext4_change_inode_journal_flag+0x17f/0x530 ext4_fileattr_set+0x140d/0x18a0 vfs_fileattr_set+0x77c/0xb80 do_vfs_ioctl+0xfc2/0x1610 __x64_sys_ioctl+0x10c/0x210 do_syscall_64+0x3b/0x90 FAT-fs (loop0): Unrecognized mount option "01777777777777777777777" or missing value entry_SYSCALL_64_after_hwframe+0x63/0xcd RIP: 0033:0x7f0cc56fcb19 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 bc ff ff ff f7 d8 64 89 01 48 RSP: 002b:00007f0cc2c51188 EFLAGS: 00000246 ORIG_RAX: 0000000000000010 RAX: ffffffffffffffda RBX: 00007f0cc5810020 RCX: 00007f0cc56fcb19 RDX: 0000000020000080 RSI: 0000000040086602 RDI: 0000000000000006 RBP: 00007f0cc5756f6d R08: 0000000000000000 R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000 R13: 00007ffc96e09e4f R14: 00007f0cc2c51300 R15: 0000000000022000 FAT-fs (loop0): Unrecognized mount option "01777777777777777777777" or missing value tmpfs: Bad value for 'mpol' FAT-fs (loop0): Unrecognized mount option "01777777777777777777777" or missing value FAT-fs (loop0): Unrecognized mount option "01777777777777777777777" or missing value mmap: syz-executor.6 (4638) uses deprecated remap_file_pages() syscall. See Documentation/mm/remap_file_pages.rst. Process accounting resumed Process accounting resumed blktrace: Concurrent blktraces are not allowed on sg0 audit: type=1400 audit(1662032658.266:10): avc: denied { block_suspend } for pid=4687 comm="syz-executor.4" capability=36 scontext=system_u:system_r:kernel_t:s0 tcontext=system_u:system_r:kernel_t:s0 tclass=capability2 permissive=1 loop4: detected capacity change from 0 to 264192 FAT-fs (loop4): bogus number of reserved sectors FAT-fs (loop4): Can't find a valid FAT filesystem loop4: detected capacity change from 0 to 264192 FAT-fs (loop4): Invalid FSINFO signature: 0x00000000, 0x00000000 (sector = 1) ======================================================= WARNING: The mand mount option has been deprecated and and is ignored by this kernel. Remove the mand option from the mount to silence this warning. ======================================================= loop4: detected capacity change from 0 to 264192 FAT-fs (loop4): Invalid FSINFO signature: 0x00000000, 0x00000000 (sector = 1) audit: type=1400 audit(1662032659.012:11): avc: denied { read } for pid=4739 comm="syz-executor.3" scontext=system_u:system_r:kernel_t:s0 tcontext=system_u:system_r:kernel_t:s0 tclass=perf_event permissive=1 loop4: detected capacity change from 0 to 264192 FAT-fs (loop4): bogus number of reserved sectors FAT-fs (loop4): Can't find a valid FAT filesystem loop2: detected capacity change from 0 to 4096 EXT4-fs (loop2): mounted filesystem without journal. Quota mode: none. EXT4-fs (loop2): unmounting filesystem.