Bug 24940 - KMS: dual screen regression
Summary: KMS: dual screen regression
Status: RESOLVED FIXED
Alias: None
Product: xorg
Classification: Unclassified
Component: Driver/Radeon (show other bugs)
Version: git
Hardware: Other All
: medium normal
Assignee: xf86-video-ati maintainers
QA Contact: Xorg Project Team
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2009-11-05 07:58 UTC by Casey Jones
Modified: 2009-11-12 09:03 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
dmesg log (66.00 KB, text/plain)
2009-11-05 07:58 UTC, Casey Jones
no flags Details

Description Casey Jones 2009-11-05 07:58:44 UTC
Created attachment 30985 [details]
dmesg log

I did a bisect and found that the commit 8a9832e89ff1bc7a039c8f966f07101570be3d5e (disable D1VGA and D2VGA if enabled) has caused a regression for me.

That commit caused my secondary display to not display anything with KMS enabled, but the primary display works fine.  Xorg seemed to think that the display is there though, since I can move windows off screen; I just can't see them.

Attaching dmesg log.

Thanks.
Comment 1 Casey Jones 2009-11-05 08:06:43 UTC
Looks like I forgot to mention that this commit is in airlied's drm-next kernel branch.
Comment 2 Dave Airlie 2009-11-05 20:36:51 UTC
Wierd, can you retry with latest drm-next?

Comment 3 Casey Jones 2009-11-06 08:10:32 UTC
Pulled drm-next, and the problem is still there.

I looked a little more closely at the commit, and I found it's in rv515.c, but I have an rv770.  So I don't know why it affected me at all.  Maybe I'm just a really bad bisecter; I'll try again and see if I get the same result.
Comment 4 Casey Jones 2009-11-06 08:27:44 UTC
Yep, I got the same result.

Additionally from the current HEAD I reverted 8a9832 and it's working fine.

So maybe the real question is why a file for rv515 affecting my rv770?
Comment 5 Alex Deucher 2009-11-06 08:43:58 UTC
(In reply to comment #4)
> So maybe the real question is why a file for rv515 affecting my rv770?
> 

It's common functionality that affects all chips starting with r5xx.
Comment 6 Florian Scandella 2009-11-07 20:35:17 UTC
hmm .. have the same problem and i've bisected it to the same commit. radeon 4800 (r700 i think) here.

should have looked into bugzilla first :)
Comment 7 Florian Scandella 2009-11-11 18:23:55 UTC
i'm sorry to bug you, but as i read the pull request i wonder what's the plan with this known regression.

Comment 8 Casey Jones 2009-11-12 07:29:00 UTC
Fixed for me by 8ccb5075a99434f88b4445723970f1402e13c26c
It's currently in the drm-radeon-testing branch.

@Florian - This commit is not in the drm-linus branch, so Linus won't be pulling this commit this time.

I'm going to close this bug because it works for me, and it apparently works for Dave since he committed it after all.
Comment 9 Florian Scandella 2009-11-12 07:47:45 UTC
hmm .. according to http://git.kernel.org/?p=linux/kernel/git/airlied/drm-2.6.git;a=shortlog;h=refs/heads/drm-linus it is in drm-linus.

will test the fix soon ..
Comment 10 Casey Jones 2009-11-12 08:15:02 UTC
This is the commit, drm/radeon/kms: fix handling of d1/d2 vga
http://git.kernel.org/?p=linux/kernel/git/airlied/drm-2.6.git;a=commit;h=fba0a16eb5c55ae3bb9cf1306a84b83451389612

It was committed 9 hours ago, and the head from drm-linus is from 2 days ago.
Comment 11 Casey Jones 2009-11-12 08:16:26 UTC
(In reply to comment #10)
> This is the commit, drm/radeon/kms: fix handling of d1/d2 vga
> http://git.kernel.org/?p=linux/kernel/git/airlied/drm-2.6.git;a=commit;h=fba0a16eb5c55ae3bb9cf1306a84b83451389612
> 
> It was committed 9 hours ago, and the head from drm-linus is from 2 days ago.
> 

Woah, where did that url come from?

This is the right one  :P
http://git.kernel.org/?p=linux/kernel/git/airlied/drm-2.6.git;a=commit;h=8ccb5075a99434f88b4445723970f1402e13c26c
Comment 12 Florian Scandella 2009-11-12 09:03:36 UTC
sorry, i misread your comment, i thought you where talking about the offending commit.

FYI, it's working for me to.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.