Bug 246114

Summary: GFS2: soft lockup in rgblk_search
Product: Red Hat Enterprise Linux 5 Reporter: Abhijith Das <adas>
Component: kernelAssignee: Don Zickus <dzickus>
Status: CLOSED ERRATA QA Contact: Dean Jansa <djansa>
Severity: medium Docs Contact:
Priority: medium    
Version: 5.0CC: rpeterso, swhiteho
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: RHBA-2007-0959 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2007-11-07 19:54:50 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:
Bug Depends On: 245832    
Bug Blocks: 204760    
Attachments:
Description Flags
patch to fix the problem--try #1 none

Description Abhijith Das 2007-06-28 16:33:44 UTC
Hit this soft lockup while running revolver on smoke cluster. The iteration
killed 3 nodes {kool, merit, salem}, winston hit this lockup.

I'm running the cvs RHEL5 kernel with GFS2 patches to rhkernel-list as of 06/27

 BUG: soft lockup detected on CPU#0!
 [<c014b4ec>] softlockup_tick+0x98/0xa6
 [<c012be40>] update_process_times+0x39/0x5c
 [<c0116adc>] smp_apic_timer_interrupt+0x5b/0x6b
 [<c0104a3b>] apic_timer_interrupt+0x1f/0x24
 [<e0701058>] rgblk_search+0x7e/0x145 [gfs2]
 [<e0701596>] try_rgrp_unlink+0x19/0x59 [gfs2]
 [<e0702127>] gfs2_inplace_reserve_i+0x244/0x57d [gfs2]
 [<c011f31f>] __cond_resched+0x16/0x34
 [<c0152ae3>] find_get_page+0x18/0x38
 [<e06fdf24>] gfs2_sharewrite_nopage+0x17c/0x296 [gfs2]
 [<e06fddeb>] gfs2_sharewrite_nopage+0x43/0x296 [gfs2]
 [<c02a7a1e>] net_rx_action+0x92/0x175
 [<c015e53e>] __handle_mm_fault+0x17d/0x8c8
 [<c0300269>] do_page_fault+0x23c/0x51c
 [<c030002d>] do_page_fault+0x0/0x51c
 [<c0104ac9>] error_code+0x39/0x40

Comment 1 Steve Whitehouse 2007-07-04 14:14:59 UTC
It would be well worth trying out the patch for bz #245832 to see if that fixes
this problem too.


Comment 2 Kiersten (Kerri) Anderson 2007-07-12 17:40:06 UTC
Is this one now verified that it is not #245832?

Comment 3 Robert Peterson 2007-07-12 21:53:34 UTC
Created attachment 159101 [details]
patch to fix the problem--try #1

This patch seems to fix the problem.  It was written by Steve Whitehouse
with some tweaking by me.

The code was looping in the relatively new section of code designed to
search for and reuse unlinked inodes.  In cases where it was finding an
appropriate inode to reuse, it was looping around and finding the same
block over and over because a "<=" check should have been a "<" when
comparing the goal block to the last unlinked block found.

Comment 4 Robert Peterson 2007-07-12 22:17:06 UTC
Fix was tested on the roth cluster using QE's revolver test.
I posted the patch to cluster-devel for inclusion in the upstream
kernel and to rhkernel-list for RHEL51.  I'm changing the status to
POST and reassigning to Don Zickus.

Comment 6 Don Zickus 2007-07-20 18:13:22 UTC
in 2.6.18-35.el5
You can download this test kernel from http://people.redhat.com/dzickus/el5

Comment 9 errata-xmlrpc 2007-11-07 19:54:50 UTC
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on the solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.

http://rhn.redhat.com/errata/RHBA-2007-0959.html