Bug 118205

Summary:	(4G/4G?) kernel broken Savage/IX-MV (MobileSavage) X-server
Product:	[Fedora] Fedora	Reporter:	Norbert Roma <romas>
Component:	kernel	Assignee:	Dave Jones <davej>
Status:	CLOSED CURRENTRELEASE	QA Contact:
Severity:	medium	Docs Contact:
Priority:	medium
Version:	rawhide	CC:	archer, byte, davej, mingo, pfrields, wtogami
Target Milestone:	---
Target Release:	---
Hardware:	i686
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2005-01-11 05:00:18 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:
Bug Blocks:	114961, 123268, 136451

Description Norbert Roma 2004-03-13 14:31:45 UTC

Description of problem:
Updating from 2.6.3-1.118 to 2.6.3-2* (tried 240, 242, 246, 253)
breaks the X-server.  X comes up, but the image is severly distorted.
 Also, X-server complains about not being able to use bios.

Version-Release number of selected component (if applicable):
see above for kernel versions; the video controller is Savage/IX-MV,
MobileSavage, using XFree86 4.3.0-63.
The LCD on the laptop is 1400x1050 (I also tried 1280x1024 for
completeness); I tested 16 and 32 bpp

How reproducible:
Every time

Steps to Reproduce:
1.Boot any of the affected kernels into runlevel 5
2.
3.
  
Actual results:
Garbled image

Expected results:
Working X-server

Additional info:

Comment 1 Nitin Dahyabhai 2004-03-13 20:55:01 UTC

I get the same problem, only no image at all.

Comment 2 Warren Togami 2004-04-11 02:21:16 UTC

Just a hunch, please try the latest rawhide i686 and i586 kernels and
report if the behavior is different in any way.

Comment 3 Ling Li 2004-04-14 23:14:07 UTC

The same here, with a IBM Thinkpad T23. Severl kernels before kernel
2.6.5-1.308 didn't work (no image at all); 2.6.5-1.308 works well;
2.6.5-1.315 and 2.6.5-1.322 don't, again. SELINUX is on (permissive)
all the time.

Comment 4 Warren Togami 2004-04-15 00:04:51 UTC

Is behavior any different between the i586 and i686 kernels of the
same version?

Comment 5 Ling Li 2004-04-15 00:26:57 UTC

Your are right. I tried 2.6.5-1.322.i586 and X works! Wrong
optimization in i686?

Comment 6 Mike A. Harris 2004-04-15 00:34:28 UTC

4G/4G split is a suspect

Comment 7 Warren Togami 2004-04-15 01:30:51 UTC

http://people.redhat.com/wtogami/temp/
Just to be 100% sure that we have isolated this problem to 4G/4G,
please test with this kernel.  This is i686 with zero changes except
4G/4G disabled.

Comment 8 Ling Li 2004-04-15 03:17:48 UTC

The one with 4G/4G disabled works. (I didn't have the hardware to test
the smp kernel but the up one works.) By the way, two unknown symbols 

/lib/modules/2.6.5-1.322.disabled4G/unsupported/fs/reiserfs/reiserfs.ko
needs unknown symbol sleep_on
/lib/modules/2.6.5-1.322.disabled4G/kernel/drivers/net/typhoon.ko
needs unknown symbol direct_csum_partial_copy_generic

Comment 9 David Carrigan 2004-04-15 05:37:50 UTC

*** Bug 119462 has been marked as a duplicate of this bug. ***

Comment 10 Colin Charles 2004-05-06 05:45:29 UTC

This used to affect me as I have a Thinkpad T20, and now, sleep works.
Kernel 2.6.5-1.327 from the standard test3 cd works.

I've just updated to the latest rawhide tree available today, and even
kernel 2.6.5-1.350 works like a charm. Looks like the bug has been fixed.

Comment 11 Ling Li 2004-05-10 08:29:25 UTC

I can confirm that kernel-smp 2.6.5-1.350.i686 & 351.i686 also work
for IBM Thinkpad T23. Display is back to normal.