[LU-1467] ASSERTION(lock->l_req_mode == lock->l_granted_mode) Created: 01/Jun/12  Updated: 29/Oct/12  Resolved: 11/Jun/12

Status: Resolved
Project: Lustre
Component/s: None
Affects Version/s: None
Fix Version/s: Lustre 2.3.0

Type: Bug Priority: Minor
Reporter: Christopher Morrone Assignee: Yang Sheng
Resolution: Fixed Votes: 0
Labels: None
Environment:

https://github.com/chaos/lustre/tree/2.1.1-4chaos


Issue Links:
Duplicate
is duplicated by LU-1523 ASSERTION(lock->l_req_mode == lock->l... Closed
Severity: 3
Rank (Obsolete): 4597

 Description   

We've hit this assertion a number of times on OSS nodes in production. Running 2.1.1-4chaos on those servers.


2012-05-24 16:23:09 LustreError: 32583:0:(ost_handler.c:1666:ost_prolong_lock_one()) ASSERTION(lock->l_req_mode == lock->l_granted_mode) failed
2012-05-24 16:23:09 LustreError: 32583:0:(ost_handler.c:1666:ost_prolong_lock_one()) LBUG
2012-05-24 16:23:09 May 24 16:23:09 Pid: 32583, comm: ll_ost_io_389
2012-05-24 16:23:09 sumom9 kernel: L
2012-05-24 16:23:09 ustreError: 3258Call Trace:
2012-05-24 16:23:09 3:0:(ost_handler [<ffffffffa03be895>] libcfs_debug_dumpstack+0x55/0x80 [libcfs]
2012-05-24 16:23:09 .c:1666:ost_prol [<ffffffffa03bedc7>] lbug_with_loc+0x47/0xb0 [libcfs]
2012-05-24 16:23:09 ong_lock_one()) [<ffffffffa03c842d>] libcfs_assertion_failed+0x2d/0x30 [libcfs]
2012-05-24 16:23:09 ASSERTION(lock-> [<ffffffffa0aab697>] ost_prolong_lock_one+0xd7/0x100 [ost]
2012-05-24 16:23:09 l_req_mode == lo [<ffffffffa0aabbcc>] ost_rw_hpreq_check+0x1fc/0x450 [ost]
2012-05-24 16:23:09 ck->l_granted_mo [<ffffffffa03bf6ce>] ? cfs_mem_cache_free+0xe/0x10 [libcfs]
2012-05-24 16:23:09 de) failed
2012-05-24 16:23:09 May 2 [<ffffffffa0aabe2e>] ost_rw_hpreq_fini+0xe/0x10 [ost]
2012-05-24 16:23:09 4 16:23:09 sumom [<ffffffffa063df68>] ptlrpc_server_finish_request+0x58/0x150 [ptlrpc]
2012-05-24 16:23:09 9 kernel: Lustre [<ffffffffa06435fa>] ptlrpc_main+0x97a/0x1690 [ptlrpc]
2012-05-24 16:23:09 Error: 32583:0 [<ffffffffa0642c80>] ? ptlrpc_main+0x0/0x1690 [ptlrpc]
2012-05-24 16:23:09 ost_handler.c:16 [<ffffffff8100c14a>] child_rip+0xa/0x20
2012-05-24 16:23:09 66:ost_prolong_l [<ffffffffa0642c80>] ? ptlrpc_main+0x0/0x1690 [ptlrpc]
2012-05-24 16:23:09 ock_one()) LBUG
2012-05-24 16:23:09 [<ffffffffa0642c80>] ? ptlrpc_main+0x0/0x1690 [ptlrpc]
2012-05-24 16:23:09 [<ffffffff8100c140>] ? child_rip+0x0/0x20
2012-05-24 16:23:09
2012-05-24 16:23:09 Kernel panic - not syncing: LBUG
2012-05-24 16:23:09 Pid: 32583, comm: ll_ost_io_389 Not tainted 2.6.32-220.7.1.7chaos.ch5.x86_64 #1
2012-05-24 16:23:09 Call Trace:
2012-05-24 16:23:09 May 24 16:23:09 [<ffffffff814ee818>] ? panic+0x78/0x143
2012-05-24 16:23:09 sumom9 kernel: K [<ffffffffa03bee1b>] ? lbug_with_loc+0x9b/0xb0 [libcfs]
2012-05-24 16:23:09 ernel panic - no [<ffffffffa03c842d>] ? libcfs_assertion_failed+0x2d/0x30 [libcfs]
2012-05-24 16:23:09 t syncing: LBUG
2012-05-24 16:23:09 [<ffffffffa0aab697>] ? ost_prolong_lock_one+0xd7/0x100 [ost]
2012-05-24 16:23:09 [<ffffffffa0aabbcc>] ? ost_rw_hpreq_check+0x1fc/0x450 [ost]
2012-05-24 16:23:09 [<ffffffffa03bf6ce>] ? cfs_mem_cache_free+0xe/0x10 [libcfs]
2012-05-24 16:23:09 [<ffffffffa0aabe2e>] ? ost_rw_hpreq_fini+0xe/0x10 [ost]
2012-05-24 16:23:09 [<ffffffffa063df68>] ? ptlrpc_server_finish_request+0x58/0x150 [ptlrpc]
2012-05-24 16:23:09 [<ffffffffa06435fa>] ? ptlrpc_main+0x97a/0x1690 [ptlrpc]
2012-05-24 16:23:09 [<ffffffffa0642c80>] ? ptlrpc_main+0x0/0x1690 [ptlrpc]
2012-05-24 16:23:09 [<ffffffff8100c14a>] ? child_rip+0xa/0x20
2012-05-24 16:23:09 [<ffffffffa0642c80>] ? ptlrpc_main+0x0/0x1690 [ptlrpc]
2012-05-24 16:23:09 [<ffffffffa0642c80>] ? ptlrpc_main+0x0/0x1690 [ptlrpc]
2012-05-24 16:23:09 [<ffffffff8100c140>] ? child_rip+0x0/0x20



 Comments   
Comment by Jinshan Xiong (Inactive) [ 01/Jun/12 ]

This must be an issue imported by LU-874. The root cause of this problem looks obvious - the lock was canceled while hpreq_check() is being called.

Comment by Peter Jones [ 03/Jun/12 ]

Niu will look into this one

Comment by Peter Jones [ 04/Jun/12 ]

Change of plan - YangSheng will handle this one

Comment by Yang Sheng [ 06/Jun/12 ]

Patch upload to: http://review.whamcloud.com/3042

Comment by Yang Sheng [ 11/Jun/12 ]

Patch landed. Close bug.

Generated at Sat Feb 10 01:16:53 UTC 2024 using Jira 9.4.14#940014-sha1:734e6822bbf0d45eff9af51f82432957f73aa32c.