Bug 119833

Summary:	panic after swsusp reading image on x86-64
Product:	[openSUSE] SUSE LINUX 10.0	Reporter:	Serguei Leontiev <lse>
Component:	Kernel	Assignee:	Pavel Machek <pavel>
Status:	RESOLVED FIXED	QA Contact:	E-mail List <qa-bugs>
Severity:	Normal
Priority:	P5 - None	CC:	aj, behlert, leo
Version:	RC 1
Target Milestone:	---
Hardware:	x86-64
OS:	SUSE Other
Whiteboard:
Found By:	Other	Services Priority:
Business Priority:		Blocker:	---
Marketing QA Status:	---	IT Deployment:	---
Attachments:	"screenshot" kernel panic Serial console capture before crash and next reboot (stack trace - "screensot" only)

Description Serguei Leontiev 2005-10-03 05:19:50 UTC

I have kernel panic when resume from suspend2disk with minimal configuration.
Crash statistic: 10% Ok, 30% panic & dump stack trace, 60% crash & reboot.

suspend2ram and resume more and more stability, but need video=normal.

I have trouble to capture kernel stack trace. Serial console disable after
`swsusp' done. May be I needed additional configuration for dump stack trace to
serial console?

Comment 1 Serguei Leontiev 2005-10-03 05:21:49 UTC

Created attachment 51307 [details]
"screenshot" kernel panic

Comment 2 Serguei Leontiev 2005-10-03 05:26:06 UTC

Created attachment 51308 [details]
Serial console capture before crash and next reboot (stack trace - "screensot" only)

Comment 3 Serguei Leontiev 2005-10-03 05:30:20 UTC

P.S.

I try linux-2.6.13-8, linux-2.6.13-9, also.

May be serial console speed incorrect? I try 115K and 38K.

Comment 4 Pavel Machek 2005-10-04 09:43:41 UTC

There are some known problems with pagetables on x86-64. See Rafael's patches on
linux-kernel. (Alternatively, try 32-bit kernel, it does not suffer from those
problems).

Comment 5 Serguei Leontiev 2005-10-05 23:16:07 UTC

Pavel, Thanks.

I see <http://bugzilla.kernel.org/show_bug.cgi?id=4959>.

With patch "Another (better?) fix"
<http://bugzilla.kernel.org/attachment.cgi?id=6110&action=view> applied to SuSE
kernel-default-2.6.13.2-20050930150828.x86_64.rpm system more and more
stability. Now for 10 Ok from suspend2disk/resume attempt.

This patch seems need and good. Thanks.

Comment 6 Forgotten User ZhJd0F0L3x 2005-10-06 08:23:53 UTC

Pavel, can we get this into (at least) the STABLE a.k.a. 10.1Alpha kernel? I
know we will get it from mainline sooner or later, but i have no idea when the
update for 10.1 will happen and the earlier and more tested Rafaels patches are,
the better IMO.

Comment 7 Pavel Machek 2005-10-06 10:50:04 UTC

Andi, you had some problems with Rafaels patch, so I guess this is your call.

Comment 8 Andreas Kleen 2005-10-12 09:58:10 UTC

Backporting the final patch I submitted to Linus to 10.0 would be ok.

Comment 9 Pavel Machek 2005-11-22 23:39:17 UTC

*** Bug 98912 has been marked as a duplicate of this bug. ***

Comment 10 Pavel Machek 2006-01-11 10:18:36 UTC

It should be fixed in latest mainline, and I think it is fixed in 2.6.15, too (=> we should have it fixed already, right?)