Re: [PATCH] ocfs2: reflink deadlock when clone file to the same directory simultaneously

From: Gang He
Date: Wed Aug 18 2021 - 21:50:04 EST

On 2021/8/18 19:20, Joseph Qi wrote:

On 8/18/21 5:20 PM, Gang He wrote:

On 2021/8/13 17:54, Joseph Qi wrote:

On 8/9/21 6:08 PM, Gang He wrote:
Hi Joseph and All,

The deadlock is caused by self-locking on one node.
There is three node cluster (mounted to /mnt/shared), the user run reflink command to clone the file to the same directory repeatedly,
  reflink "/mnt/shared/test" \
  "/mnt/shared/.snapshots/test.`date +%m%d%H%M%S`.`hostname`"

After a while, the reflink process on each node is hung, the file system cannot be listed.
The problematic reflink command process is blocked by itself, e.g. the reflink process is hung at ghe-sle15sp2-nd2,
kernel: task:reflink         state:D stack:    0 pid:16992 ppid:  4530
kernel: Call Trace:
kernel:  __schedule+0x2fd/0x750
kernel:  ? try_to_wake_up+0x17b/0x4e0
kernel:  schedule+0x2f/0xa0
kernel:  schedule_timeout+0x1cc/0x310
kernel:  ? __wake_up_common+0x74/0x120
kernel:  wait_for_completion+0xba/0x140
kernel:  ? wake_up_q+0xa0/0xa0
kernel:  __ocfs2_cluster_lock.isra.41+0x3b5/0x820 [ocfs2]
kernel:  ? ocfs2_inode_lock_full_nested+0x1fc/0x960 [ocfs2]
kernel:  ocfs2_inode_lock_full_nested+0x1fc/0x960 [ocfs2]
kernel:  ocfs2_init_security_and_acl+0xbe/0x1d0 [ocfs2]
kernel:  ocfs2_reflink+0x436/0x4c0 [ocfs2]
kernel:  ? ocfs2_reflink_ioctl+0x2ca/0x360 [ocfs2]
kernel:  ocfs2_reflink_ioctl+0x2ca/0x360 [ocfs2]
kernel:  ocfs2_ioctl+0x25e/0x670 [ocfs2]
kernel:  do_vfs_ioctl+0xa0/0x680
kernel:  ksys_ioctl+0x70/0x80

In fact, the destination directory(.snapshots) inode dlm lock was acquired by ghe-sle15sp2-nd2, next there is bast message from other nodes to ask ghe-sle15sp2-nd2 downconvert lock, but the operation failed, the kernel message is printed like,
kernel: (ocfs2dc-AA35DD9,2560,3):ocfs2_downconvert_lock:3660 ERROR: DLM error -16 while calling ocfs2_dlm_lock on resource M0000000000000000046e0200000000
kernel: (ocfs2dc-AA35DD9,2560,3):ocfs2_unblock_lock:3904 ERROR: status = -16
kernel: (ocfs2dc-AA35DD9,2560,3):ocfs2_process_blocked_lock:4303 ERROR: status = -16

Then, the reflink process tries to acquire this directory inode dlm lock, the process is blocked, the dlm lock resource in memory looks like

     l_name = "M0000000000000000046e0200000000",
     l_ro_holders = 0,
     l_ex_holders = 0,
     l_level = 5 '\005',
     l_requested = 0 '\000',
     l_blocking = 5 '\005',
     l_type = 0 '\000',
     l_action = 0 '\000',
     l_unlock_action = 0 '\000',
     l_pending_gen = 645948,

So far, I do not know what makes dlm lock function failed, it also looks we do not handle this failure case in dlmglue layer, but I always reproduce this hang with my test script, e.g.

   while ((loop++)) ; do
         for i in `seq 1 100`; do
           reflink "/mnt/shared/test" "/mnt/shared/.snapshots /test.${loop}.${i}.`date +%m%d%H%M%S`.`hostname`"
         usleep 500000
         rm -f /mnt/shared/.snapshots/testnode1.qcow2.*.`hostname`

My patch changes multiple acquiring dest directory inode dlm lock during in ocfs2_reflink function, it avoids the hang issue happen again.The code change also can improve reflink performance in this case.


'status = -16' implies DLM_CANCELGRANT.
Do you use stack user instead of o2cb? If yes, can you try o2cb with
your reproducer?

I setup o2cb based ocfs2 clusters with sle15sp2 and oracleLinux8u4.
After two day testing with the same script, I did not encounter dlm_lock downconvert failure, the hang issue did not happen.
After my patch was applied, there was not any side effect, the reflink performance was doubled in the case.

Do you mean the hang only happens on stack user?
Why? since o2cb based dlm_lock did not return error -16 when downcovert dlm lock during the whole testing.
But pmck based dlm_lock retuned error -16 during the testing, then we did not handle this error further in dlmglue layer, next encounter the hang issue when dlm_lock acquire the lock. Maybe there is a race condition when using dlm_lock/dlm_unlock(cancel) in dlmglue layer.
Anyway, the problem belongs to ocfs2 own parts.

IMO, we'd better directly fix the stack user hang first, and then optimize
reflink performance.
Before we fix the potential problem in dlmglue layer, I suggest to accept this patch first. Since the dlmglue layer code is kind of compliant, we need to carefully modify this part of the code. But this patch can workaround dlm_lock failure(-16) problem for pcmk stack, meanwhile, it also can improve the performance.