Bug 712260

Summary:	migrate: don't account swapcache as shmem
Product:	Red Hat Enterprise Linux 6	Reporter:	Andrea Arcangeli <aarcange>
Component:	kernel	Assignee:	Andrea Arcangeli <aarcange>
Status:	CLOSED ERRATA	QA Contact:	Caspar Zhang <czhang>
Severity:	unspecified	Docs Contact:
Priority:	low
Version:	6.2	CC:	czhang, juzhang, kzhang, qcai, tburke
Target Milestone:	rc
Target Release:	---
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	kernel-2.6.32-169.el6	Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2011-12-06 13:38:34 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:

Description Andrea Arcangeli 2011-06-10 01:27:35 UTC

Description of problem: src nr_shmem underflow during migration and another zone has it boosted by mistake


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


How reproducible: easy


Steps to Reproduce:
1. allocate order > 0 pages with THP, hugetlbfs or jumbo frames triggering a lot of compaction that in turn will run a lot of migrate_pages (the bug is in the migration code) and push the system into swap while the hugepage allocations  are running (swap must be present as the bug triggers only with lots of swapcache allocated)
2. run egrep 'Node|nr_shmem' /proc/zoneinfo
3.
  
Actual results: nr_shmem should be correct (with CONFIG_SMP=y we'll only see some zone increasing nr_shmem too much, we won't see the actual overflow because on SMP the nr_shmem will be returned 0 even if it's negative if casted to "signed long", with CONFIG_SMP=n a negative nr_shmem value should be visible for some zone)


Expected results: the nr_shmem statistics should be accurate for each zone


Additional info:

This seems only an aesthetical problem, not resulting in any runtime error. I'm not aware of any application reading nr_shmem from /proc/zoneinfo. nr_shmem in /proc/vmstat and /proc/meminfo should not be affected, only the _per-zone_ statistics are affected. So it's a minor bug.

Comment 2 RHEL Program Management 2011-06-13 09:29:46 UTC

This request was evaluated by Red Hat Product Management for inclusion
in a Red Hat Enterprise Linux maintenance release. Product Management has 
requested further review of this request by Red Hat Engineering, for potential
inclusion in a Red Hat Enterprise Linux Update release for currently deployed 
products. This request is not yet committed for inclusion in an Update release.

Comment 4 Aristeu Rozanski 2011-07-18 15:27:59 UTC

Patch(es) available on kernel-2.6.32-169.el6

Comment 8 errata-xmlrpc 2011-12-06 13:38:34 UTC

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

http://rhn.redhat.com/errata/RHSA-2011-1530.html