| TITLE: possible deadlock in blkdev_reread_part |
| |
| [ 127.525803] ====================================================== |
| [ 127.532093] WARNING: possible circular locking dependency detected |
| [ 127.538376] 4.14.0-rc1+ #1 Not tainted |
| [ 127.542228] ------------------------------------------------------ |
| [ 127.548509] syz-executor0/22269 is trying to acquire lock: |
| [ 127.554094] (&bdev->bd_mutex){+.+.}, at: [<ffffffff8232bf0e>] blkdev_reread_part+0x1e/0x40 |
| [ 127.562560] |
| [ 127.562560] but task is already holding lock: |
| [ 127.568495] (&lo->lo_ctl_mutex#2){+.+.}, at: [<ffffffff83542c29>] lo_compat_ioctl+0x109/0x140 |
| [ 127.577221] |
| [ 127.577221] which lock already depends on the new lock. |
| [ 127.577221] |
| [ 127.585502] |
| [ 127.585502] the existing dependency chain (in reverse order) is: |
| [ 127.593087] |
| [ 127.593087] -> #1 (&lo->lo_ctl_mutex#2){+.+.}: |
| [ 127.599124] __lock_acquire+0x328f/0x4620 |
| [ 127.603759] lock_acquire+0x1d5/0x580 |
| [ 127.608048] __mutex_lock+0x16f/0x1870 |
| [ 127.612421] mutex_lock_nested+0x16/0x20 |
| [ 127.616972] lo_release+0x6b/0x180 |
| [ 127.621000] __blkdev_put+0x602/0x7c0 |
| [ 127.625288] blkdev_put+0x85/0x4f0 |
| [ 127.629314] blkdev_close+0x91/0xc0 |
| [ 127.633425] __fput+0x333/0x7f0 |
| [ 127.637192] ____fput+0x15/0x20 |
| [ 127.640960] task_work_run+0x199/0x270 |
| [ 127.645333] exit_to_usermode_loop+0x2a6/0x300 |
| [ 127.650404] syscall_return_slowpath+0x42f/0x500 |
| [ 127.655651] entry_SYSCALL_64_fastpath+0xbc/0xbe |
| [ 127.660888] |
| [ 127.660888] -> #0 (&bdev->bd_mutex){+.+.}: |
| [ 127.666578] check_prev_add+0x865/0x1520 |
| [ 127.671134] __lock_acquire+0x328f/0x4620 |
| [ 127.675778] lock_acquire+0x1d5/0x580 |
| [ 127.680067] __mutex_lock+0x16f/0x1870 |
| [ 127.684441] mutex_lock_nested+0x16/0x20 |
| [ 127.688991] blkdev_reread_part+0x1e/0x40 |
| [ 127.693629] loop_reread_partitions+0x12f/0x1a0 |
| [ 127.698783] loop_set_status+0x9ba/0xf60 |
| [ 127.703333] loop_set_status_compat+0x92/0xf0 |
| [ 127.708315] lo_compat_ioctl+0x114/0x140 |
| [ 127.712863] compat_blkdev_ioctl+0x3ba/0x1850 |
| [ 127.717848] compat_SyS_ioctl+0x1da/0x3300 |
| [ 127.722570] do_fast_syscall_32+0x3f2/0xeed |
| [ 127.727378] entry_SYSENTER_compat+0x51/0x60 |
| [ 127.732268] |
| [ 127.732268] other info that might help us debug this: |
| [ 127.732268] |
| [ 127.740375] Possible unsafe locking scenario: |
| [ 127.740375] |
| [ 127.746396] CPU0 CPU1 |
| [ 127.751028] ---- ---- |
| [ 127.755664] lock(&lo->lo_ctl_mutex#2); |
| [ 127.759694] lock(&bdev->bd_mutex); |
| [ 127.765892] lock(&lo->lo_ctl_mutex#2); |
| [ 127.772438] lock(&bdev->bd_mutex); |
| [ 127.776120] |
| [ 127.776120] *** DEADLOCK *** |
| [ 127.776120] |
| [ 127.782144] 1 lock held by syz-executor0/22269: |
| [ 127.786775] #0: (&lo->lo_ctl_mutex#2){+.+.}, at: [<ffffffff83542c29>] lo_compat_ioctl+0x109/0x140 |
| [ 127.795934] |
| [ 127.795934] stack backtrace: |
| [ 127.800405] CPU: 0 PID: 22269 Comm: syz-executor0 Not tainted 4.14.0-rc1+ #1 |
| [ 127.807556] Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 |
| [ 127.816876] Call Trace: |
| [ 127.819433] dump_stack+0x194/0x257 |
| [ 127.831436] print_circular_bug+0x503/0x710 |
| [ 127.844570] check_prev_add+0x865/0x1520 |
| [ 127.961665] lock_acquire+0x1d5/0x580 |
| ... |
| [ 128.182847] entry_SYSENTER_compat+0x51/0x60 |
| [ 128.187221] RIP: 0023:0xf7fd5c79 |
| [ 128.190551] RSP: 002b:00000000f77d105c EFLAGS: 00000296 ORIG_RAX: 0000000000000036 |
| [ 128.198227] RAX: ffffffffffffffda RBX: 0000000000000016 RCX: 0000000000004c02 |
| [ 128.205464] RDX: 00000000202e3000 RSI: 0000000000000000 RDI: 0000000000000000 |
| [ 128.212700] RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000 |
| [ 128.219935] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000 |
