Bug 20447

Summary: HD3200 no output (HDMI)
Product: xorg Reporter: Pär Andersson <paran>
Component: Driver/RadeonAssignee: xf86-video-ati maintainers <xorg-driver-ati>
Status: RESOLVED INVALID QA Contact: Xorg Project Team <xorg-team>
Severity: normal    
Priority: medium    
Version: unspecified   
Hardware: x86-64 (AMD64)   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
Xorg.0.log
none
xrandr --verbose
none
avivotool regs all
none
avivotool regs all (radeonhd) none

Description Pär Andersson 2009-03-03 17:11:48 UTC
xf86-video-ati, master (d586a2c6f821c821a4a7708a3382acb63187534f)
Ubuntu 8.10, xserver-xorg 1:7.4~5ubuntu3

When I start X the monitor instantly enters sleep mode. By SSH-ing into the machine I see that the X-server is alive. xrandr show that the monitor is connected to HDMI-0 and using the correct resolution. Changing resolution using xrandr also works but the monitor stay off.

Attaching Xorg.0.log and output from "xrandr --verbose" and "avivotool regs all".


The computer have two ATI cards (HD3200 + HD3870). With radeon I can only use the one that is set as primary in BIOS, but that is a separate bug that I don't think affect this one that I am reporting now. :-)
Comment 1 Pär Andersson 2009-03-03 17:12:32 UTC
Created attachment 23491 [details]
Xorg.0.log
Comment 2 Pär Andersson 2009-03-03 17:12:59 UTC
Created attachment 23492 [details]
xrandr --verbose
Comment 3 Pär Andersson 2009-03-03 17:13:28 UTC
Created attachment 23493 [details]
avivotool regs all
Comment 4 Alex Deucher 2009-03-06 07:16:27 UTC
Are you connecting to an HDMI monitor or are you using an HDMI to DVI adapter?
Comment 5 Pär Andersson 2009-03-08 11:32:29 UTC
(In reply to comment #4)
> Are you connecting to an HDMI monitor or are you using an HDMI to DVI adapter?

HDMI to DVI adapter.
Comment 6 Pär Andersson 2009-03-15 15:18:57 UTC
I tried this card with -radeonhd (master) and latest drm kernel modules (r6xx-r7xx-support). Using this the output works fine.

I also verified that the bug is still present with latest -ati master (a55ced5ee20c07e743c7c0978803fd10589c1531).
Comment 7 Pär Andersson 2009-03-15 15:21:21 UTC
Created attachment 23882 [details]
avivotool regs all (radeonhd)

reg dump with working video output and running radeonhd.
Comment 8 Alex Deucher 2009-10-21 13:37:44 UTC
can you try with ati from git master?  Specifically commit 66b194a78c470cb3978f310828dd96c3f3e96944.
Comment 9 Alex Deucher 2009-12-09 10:09:25 UTC
Can you try with the latest code from xf86-video-ati git master?  I just committed some new PLL code that might help.
Comment 10 Pär Andersson 2009-12-14 08:30:30 UTC
(In reply to comment #9)
> Can you try with the latest code from xf86-video-ati git master?  I just
> committed some new PLL code that might help.

I just tried vanilla 2.6.32 together with latest -ati git packages
from Ubuntu xorg-edgers ppa.

X works, this blank screen-bug is definitely fixed!

On 2009-10-21 you asked me to try latest git, but I have not had time
to do that. This means that I unfortunately can't say if you fixed it
then or in your latest commits.


I see quite a bit of flickering pixels now, which goes away if I lower
the resolution from the panels native 1600x1200 down to
1280x1024. This is noticeable in text mode before X is
started, so I guess it is an unrelated kernel issue.
Comment 11 Alex Deucher 2009-12-14 08:45:17 UTC
(In reply to comment #10)
> I see quite a bit of flickering pixels now, which goes away if I lower
> the resolution from the panels native 1600x1200 down to
> 1280x1024. This is noticeable in text mode before X is
> started, so I guess it is an unrelated kernel issue.
> 

For non-KMS, does:
Option "DisplayPriority" "HIGH"
in the device section of your config help?
Comment 12 Pär Andersson 2009-12-14 10:44:32 UTC
(In reply to comment #11)
> For non-KMS, does:
> Option "DisplayPriority" "HIGH"
> in the device section of your config help?

If I boot without KMS (options radeon modeset=0) there is no flicker
while in text-mode as expected.

X still flickers, but significantly less than with KMS. No noticeable
difference when setting DisplayPriority=HIGH. However setting
NewPLL=FALSE make non-KMS flicker as much as KMS, maybe even more.

Would you like to see how the flicker look? I filmed the monitor a few
seconds in KMS and non-KMS, but the movies are 16M each so I probably
should not attach them to bugzilla. :)
Comment 13 Alex Deucher 2009-12-14 11:36:03 UTC
(In reply to comment #12)
> X still flickers, but significantly less than with KMS. No noticeable
> difference when setting DisplayPriority=HIGH. However setting
> NewPLL=FALSE make non-KMS flicker as much as KMS, maybe even more.

KMS and non-KMS should use the same pll code assuming you have a new enough drm.  If you update your drm, you should get the same behavior with both KMS and non-KMS.
Comment 14 Pär Andersson 2009-12-14 12:05:21 UTC
(In reply to comment #13)
> KMS and non-KMS should use the same pll code assuming you have a new enough
> drm.  If you update your drm, you should get the same behavior with both KMS
> and non-KMS.

My kernel is vanilla 2.6.32, is that new enough?

librdm is latest git, package from xorg-edgers:

libdrm (2.4.16+git20091211.edc77dd2-0ubuntu0sarvatt2) karmic; urgency=low

  * Checkout from git 20091211 (master branch) up to commit
    edc77dd291594e017ca0ee96a785412107ebff74
Comment 15 Alex Deucher 2009-12-14 12:13:15 UTC
(In reply to comment #14)
> (In reply to comment #13)
> > KMS and non-KMS should use the same pll code assuming you have a new enough
> > drm.  If you update your drm, you should get the same behavior with both KMS
> > and non-KMS.
> 
> My kernel is vanilla 2.6.32, is that new enough?

You need what will become 2.6.33.  Either Linus' or Dave's tree.
Comment 16 Pär Andersson 2009-12-15 02:38:54 UTC
(In reply to comment #15)
> (In reply to comment #14)
> > (In reply to comment #13)
> > > KMS and non-KMS should use the same pll code assuming you have a new enough
> > > drm.  If you update your drm, you should get the same behavior with both KMS
> > > and non-KMS.
> > 
> > My kernel is vanilla 2.6.32, is that new enough?
> 
> You need what will become 2.6.33.  Either Linus' or Dave's tree.

I tried latest from Linus master-branch, with all necessary firmwares. R600_rlc.bin and R700_rlc.bin was a bit hard to find :)

Non-KMS and KMS now behave the same way. Both have a low amount of flicker, like with 2.6.32 and non-KMS. DisplayPriority=HIGH still makes no difference.
Comment 17 Alex Deucher 2010-10-19 17:01:39 UTC
Is this still an issue with KMS on a newer kernel?
Comment 18 Christopher M. Penalver 2016-02-24 06:14:53 UTC
Pär Andersson, Karmic reached EOL on April 30, 2011. For more on this, please see https://wiki.ubuntu.com/Releases .

If this is reproducible in a supported release, it will help immensely if you filed a new report with Ubuntu by ensuring you have the package xdiagnose installed, and that you click the Yes button for attaching additional debugging information running the following from a terminal:
ubuntu-bug xorg

Also, please feel free to subscribe me to it.

For more on why this is helpful, please see https://wiki.ubuntu.com/ReportingBugs.

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.