Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

For bugs related to Red Hat Enterprise Linux 5 product line. The current stable release is 5.10. For Red Hat Enterprise Linux 6 and above, please visit Red Hat JIRA https://issues.redhat.com/secure/CreateIssue!default.jspa?pid=12332745 to report new issues.

Bug 555266

Summary:	backport fix from kernel 2.6.32: dm: avoid _hash_lock deadlock
Product:	Red Hat Enterprise Linux 5	Reporter:	Eyal Gordon <eyal.gordon>
Component:	kernel	Assignee:	Mikuláš Patočka <mpatocka>
Status:	CLOSED DUPLICATE	QA Contact:	Red Hat Kernel QE team <kernel-qe>
Severity:	medium	Docs Contact:
Priority:	low
Version:	5.4	CC:	agk, jwilson, mbroz
Target Milestone:	rc
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2010-11-28 05:33:11 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Eyal Gordon 2010-01-14 08:28:14 UTC

Description of problem:

We hit a recreate of the following bug, which was fixed a few weeks ago in kernel 2.6.32 stable version:
http://repo.or.cz/w/linux-2.6.git/commit/6076905b5ef39e0ea58db32583c9e0036c05e47b

The scenario was:
1. logout of both iscsi paths that comprise a single multipath device.
2. We then saw that running 'dmsetup status' commands got stuck in D state 'infinitely'
3. We then ran 'echo t > /proc/sysrq-trigger' to dump all the stack traces in the system into /var/log/messages, and saw the following stack, which means it is a dup of the problem fixed by the above commit:
Dec 23 15:12:45 b55 kernel:  [<ffffffff8014b5d4>] __next_cpu+0x19/0x28
Dec 23 15:12:45 b55 kernel:  [<ffffffff8006475d>] __down_read+0x7a/0x92
Dec 23 15:12:45 b55 kernel:  [<ffffffff8812b740>] :dm_mod:dm_copy_name_and_uuid+0x2e/0x91
Dec 23 15:12:46 b55 kernel:  [<ffffffff8812e685>] :dm_mod:dm_send_uevents+0x64/0x123
Dec 23 15:12:46 b55 kernel:  [<ffffffff8812c7d7>] :dm_mod:dev_remove+0x0/0xb5
Dec 23 15:12:46 b55 kernel:  [<ffffffff88128678>] :dm_mod:event_callback+0x73/0x99
Dec 23 15:12:46 b55 kernel:  [<ffffffff88129d01>] :dm_mod:dm_table_event+0x46/0x53
Dec 23 15:12:46 b55 kernel:  [<ffffffff8812c7ad>] :dm_mod:__hash_remove+0x3e/0x68
Dec 23 15:12:46 b55 kernel:  [<ffffffff8812c865>] :dm_mod:dev_remove+0x8e/0xb5
Dec 23 15:12:46 b55 kernel:  [<ffffffff8812cd80>] :dm_mod:ctl_ioctl+0x210/0x25b
Dec 23 15:12:46 b55 kernel:  [<ffffffff800418c4>] do_ioctl+0x55/0x6b
Dec 23 15:12:46 b55 kernel:  [<ffffffff8002fab9>] vfs_ioctl+0x248/0x261
Dec 23 15:12:46 b55 kernel:  [<ffffffff8004bdaf>] sys_ioctl+0x59/0x78
Dec 23 15:12:46 b55 kernel:  [<ffffffff8005d28d>] tracesys+0xd5/0xe0

Please consider backporting the fix to RHEL 5.4, if not then perhaps to 5.6.
Thank you,
 Eyal

Version-Release number of selected component (if applicable):
RHEL 5.4

How reproducible:
Seen a few times, but we did not see it consistently.

Steps to Reproduce:
1.
2.
3.
  
Actual results:


Expected results:


Additional info:

Comment 1 Doug Ledford 2010-02-21 15:48:17 UTC

This is a dm issue, but was passed off to me.  Reassigning.

Comment 3 Mikuláš Patočka 2010-11-28 05:33:11 UTC

I already sent a backport of a patch in December 2009 and it was applied to RHEL 5. This is no longer an issue. I'm closing the bug.

*** This bug has been marked as a duplicate of bug 543270 ***