Bug 304811

Summary:	Radeon KMS: Garbled kwin shadows and pixmaps
Product:	Gentoo Linux	Reporter:	Tanktalus
Component:	[OLD] Core system	Assignee:	Gentoo Kernel Bug Wranglers and Kernel Maintainers <kernel>
Status:	RESOLVED TEST-REQUEST
Severity:	normal	CC:	kfm
Priority:	High
Version:	unspecified
Hardware:	All
OS:	Linux
URL:	http://bugzilla.kernel.org/show_bug.cgi?id=15186
Whiteboard:
Package list:		Runtime testing required:	---
Attachments:	Backfit of upstream patch

Description Tanktalus 2010-02-13 05:57:53 UTC

This is a duplicate of upstream bug 15186.  I've taken their patch to 2.6.33-rc's, and backfit it to 2.6.32 kernels.  I'm running this patch against 2.6.32-r3 right now, and it seems to apply against -r5.

The problem in question is garbled shadows, which I only started experiencing when I moved to KDE 4.4, and pixmaps, which I was experiencing with Qt4.6.1 and KDE 4.3.5.  This patch makes the desktop experience far better, so I'm submitting it back to the distro in hopes that it's accepted to the .32 gentoo-sources stream.

Reproducible: Always

Comment 1 Tanktalus 2010-02-13 05:59:19 UTC

Created attachment 219403 [details, diff]
Backfit of upstream patch

This is a backfit of the "proper patch" in the upstream bugzilla.  Hopefully this can be accepted to the distro.  If not, then at least it's somewhere that others can find it.

Comment 2 Tanktalus 2010-03-12 16:38:00 UTC

Now that 2.6.33 is in portage, the alternative solution to this patch is to direct anyone who has this problem to upgrade to 2.6.33, which I already have.  However, some may consider this to be a blocker for stabilising *any* of the 2.6.32 kernels, and thus may be considered for stabilising some 2.6.32 kernel before you're ready to stabilise a 2.6.33 kernel.

Comment 3 Markos Chandras (RETIRED) gentoo-dev

2010-05-28 16:40:12 UTC

I think this bug should close now unless there is interest to backport the patch to 2.6.23 series

Comment 4 kfm 2010-05-28 17:25:39 UTC

Tanktalus, perhaps you would be so kind as to submit your backported patch to stable@kernel.org, while CC'ing jglisse@redhat.com in his capacity as the author of the relevant commit. As 2.6.32.x will be maintained as a long-term stable branch, and this patch seems to meet the criteria for inclusion, it may well be accepted.

Comment 5 kfm 2010-05-28 17:27:27 UTC

I forgot to say that the patch should be generated against a vanilla 2.6.32.14 tree in the event of submission.

Comment 6 Mike Pagano gentoo-dev

2010-06-16 23:47:04 UTC

Is this still a problem with later kernels?

Comment 7 Markos Chandras (RETIRED) gentoo-dev

2010-06-29 20:46:16 UTC

Upstream claims that the problems is fixed in .33 

Comment #21 From Robert Schedel 2010-02-07 10:58:54 (-) [reply]
Applied in 2.6.33-rc7 mainline, regression resolved.

Mike how do you want us to handle this bug? Closed as TEST-REQUEST?

Comment 8 Markos Chandras (RETIRED) gentoo-dev

2010-06-30 13:32:31 UTC

Reopen if you still experience the same problem with 2.6.34 kernels

Thanks