Noticed this in the logs for cifs/102 with current for-next. Ideas?
2014.571858] run fstests cifs/102 at 2022-10-02 15:12:19
[ 2015.934812] ======================================================
[ 2015.935001] WARNING: possible circular locking dependency detected
[ 2015.935225] 6.0.0-rc7 #1 Not tainted
[ 2015.935307] ------------------------------------------------------
[ 2015.935307] cifsd/12418 is trying to acquire lock:
[ 2015.935307] ffffffffc0babaf8 (&cifs_tcp_ses_lock){+.+.}-{2:2}, at:
smb2_find_smb_tcon+0x24/0xd0 [cifs]
[ 2015.936656]
but task is already holding lock:
[ 2015.936771] ffff91f54fd02268 (&tcp_ses->mid_lock){+.+.}-{2:2}, at:
release_mid+0x29/0x3b0 [cifs]
[ 2015.936771]
which lock already depends on the new lock.
[ 2015.937709]
the existing dependency chain (in reverse order) is:
[ 2015.938412]
-> #1 (&tcp_ses->mid_lock){+.+.}-{2:2}:
[ 2015.938675] _raw_spin_lock+0x2f/0x40
[ 2015.938939] cifs_debug_data_proc_show+0x60d/0x980 [cifs]
[ 2015.938939] seq_read_iter+0xdf/0x470
[ 2015.939367] proc_reg_read_iter+0x46/0x70
[ 2015.939632] vfs_read+0x230/0x2d0
[ 2015.939806] ksys_read+0x88/0xc0
[ 2015.939806] do_syscall_64+0x3a/0x90
[ 2015.939806] entry_SYSCALL_64_after_hwframe+0x63/0xcd
[ 2015.939806]
-> #0 (&cifs_tcp_ses_lock){+.+.}-{2:2}:
[ 2015.939806] __lock_acquire+0x11d9/0x1690
[ 2015.939806] lock_acquire+0x254/0x300
[ 2015.939806] _raw_spin_lock+0x2f/0x40
[ 2015.941189] smb2_find_smb_tcon+0x24/0xd0 [cifs]
[ 2015.941189] smb2_handle_cancelled_mid+0x42/0x90 [cifs]
[ 2015.941633] release_mid+0x1f7/0x3b0 [cifs]
[ 2015.942363] cifs_demultiplex_thread+0x830/0xd40 [cifs]
[ 2015.942892] kthread+0xe8/0x110
[ 2015.943423] ret_from_fork+0x22/0x30
[ 2015.943423]
other info that might help us debug this:
[ 2015.943423] Possible unsafe locking scenario:
[ 2015.943423] CPU0 CPU1
[ 2015.945327] CIFS: VFS: \\win16.vm.test\Share Close unmatched open for MID:211
[ 2015.945142] ---- ----
[ 2015.945142] lock(&tcp_ses->mid_lock);
[ 2015.945142] lock(&cifs_tcp_ses_lock);
[ 2015.945142] lock(&tcp_ses->mid_lock);
[ 2015.945142] lock(&cifs_tcp_ses_lock);
[ 2015.945142]
*** DEADLOCK ***
[ 2015.945142] 1 lock held by cifsd/12418:
[ 2015.945142] #0: ffff91f54fd02268 (&tcp_ses->mid_lock){+.+.}-{2:2},
at: release_mid+0x29/0x3b0 [cifs]
[ 2015.949720]
stack backtrace:
[ 2015.950654] CPU: 6 PID: 12418 Comm: cifsd Not tainted 6.0.0-rc7 #1
[ 2015.951605] Hardware name: Red Hat KVM, BIOS 0.5.1 01/01/2011
[ 2015.951774] Call Trace:
[ 2015.951774] <TASK>
[ 2015.952771] dump_stack_lvl+0x55/0x71
[ 2015.952771] check_noncircular+0xeb/0x100
[ 2015.953646] ? __lock_acquire+0x490/0x1690
[ 2015.954175] ? cpu_attach_domain+0x2e7/0x9d0
[ 2015.954651] ? __lock_acquire+0x11d9/0x1690
[ 2015.955159] __lock_acquire+0x11d9/0x1690
[ 2015.955159] lock_acquire+0x254/0x300
[ 2015.955688] ? smb2_find_smb_tcon+0x24/0xd0 [cifs]
[ 2015.956203] CIFS: VFS: \\win16.vm.test\Share Close unmatched open for MID:212
[ 2015.955688] _raw_spin_lock+0x2f/0x40
[ 2015.957008] ? smb2_find_smb_tcon+0x24/0xd0 [cifs]
[ 2015.957008] smb2_find_smb_tcon+0x24/0xd0 [cifs]
[ 2015.957008] smb2_handle_cancelled_mid+0x42/0x90 [cifs]
[ 2015.957008] ? _raw_spin_lock+0x2f/0x40
[ 2015.959650] release_mid+0x1f7/0x3b0 [cifs]
[ 2015.959650] cifs_demultiplex_thread+0x830/0xd40 [cifs]
[ 2015.959650] ? cifs_handle_standard+0x1b0/0x1b0 [cifs]
[ 2015.960655] kthread+0xe8/0x110
[ 2015.960655] ? kthread_complete_and_exit+0x20/0x20
[ 2015.960655] ret_from_fork+0x22/0x30
[ 2015.960655] </TASK>
[ 2015.976931] CIFS: VFS: \\win16.vm.test\Share Close unmatched open for MID:213
--
Thanks,
Steve
