Hello All,
I'm trying the top-of-trunk drm-2.6 trees (both drm-next and drm-radeon-testing) with my Radeon HD 3200 GPU over HDMI. The primary application is mythtv which uses DRM syncing for the frame syncronisation. Now, with the exact same userland software I noticed the introduction of sync gliches in the May-timeframe. The drm-radeon-testing on May 9 was still ok, but both drm-next and drm-radeon-testing at the end of May showed that glitch: every couple of seconds there's a very visual hickup, especially in scroll texts.
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
What I've tried so far:
* I tried bisecting, but I failed to spot the offending commits somehow. My last hope is this list of experts - for me the git logs look sane, but you guy surely have more insight.
* I tried turning on drm debug=1, but did not see anything suspicious, especially with relation to the r600 irq handling changes.
* Currently I'm in the process of setting up systemtap to profile the scheduling, but I'm still on the learning curve :)
I'd be most grateful for any tips and things to try to isolate this issue.
Thanks Marius
On Mit, 2010-06-02 at 08:07 +0200, Marius Gröger wrote:
Hello All,
I'm trying the top-of-trunk drm-2.6 trees (both drm-next and drm-radeon-testing) with my Radeon HD 3200 GPU over HDMI. The primary application is mythtv which uses DRM syncing for the frame syncronisation. Now, with the exact same userland software I noticed the introduction of sync gliches in the May-timeframe. The drm-radeon-testing on May 9 was still ok, but both drm-next and drm-radeon-testing at the end of May showed that glitch: every couple of seconds there's a very visual hickup, especially in scroll texts.
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
What I've tried so far:
- I tried bisecting, but I failed to spot the offending commits somehow.
My last hope is this list of experts - for me the git logs look sane, but you guy surely have more insight.
- I tried turning on drm debug=1, but did not see anything suspicious,
especially with relation to the r600 irq handling changes.
- Currently I'm in the process of setting up systemtap to profile the
scheduling, but I'm still on the learning curve :)
I'd be most grateful for any tips and things to try to isolate this issue.
If you have dynamic PM enabled, does disabling that help?
Hi Michel,
Michel Dänzer schrieb:
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
Hm, ok... AFAIK MythTV tries several sync mechanisms with DRM being preferred. I tried with OpenGL syncing but the issue remains.
If you have dynamic PM enabled, does disabling that help?
No I don't have that enabled.
Any other ideas?
Thanks Marius
Hi All,
Michel Dänzer schrieb:
On Mit, 2010-06-02 at 08:07 +0200, Marius Gröger wrote:
Hello All,
I'm trying the top-of-trunk drm-2.6 trees (both drm-next and drm-radeon-testing) with my Radeon HD 3200 GPU over HDMI. The primary application is mythtv which uses DRM syncing for the frame syncronisation. Now, with the exact same userland software I noticed the introduction of sync gliches in the May-timeframe. The drm-radeon-testing on May 9 was still ok, but both drm-next and drm-radeon-testing at the end of May showed that glitch: every couple of seconds there's a very visual hickup, especially in scroll texts.
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
What about that new dri2 vsync stuff which was mentioned related to [Bug 28383]? Could the changes done for that in any way alter the timing? BTW I measured the glitches I'm experiencing and the appear to be to happen in intervals of 10 seconds. Again, all I'm changing is the kernel, and even the kernel config is the same. I'd be most grateful for any clues/hints/tips I could follow to resolve this regression.
If you have dynamic PM enabled, does disabling that help?
I checked again and there's method=profile and profile=default. Afaict this is not using dynpm, right?
Marius
2010/6/4 Marius Gröger marius.groeger@googlemail.com:
Hi All,
Michel Dänzer schrieb:
On Mit, 2010-06-02 at 08:07 +0200, Marius Gröger wrote:
Hello All,
I'm trying the top-of-trunk drm-2.6 trees (both drm-next and drm-radeon-testing) with my Radeon HD 3200 GPU over HDMI. The primary application is mythtv which uses DRM syncing for the frame syncronisation. Now, with the exact same userland software I noticed the introduction of sync gliches in the May-timeframe. The drm-radeon-testing on May 9 was still ok, but both drm-next and drm-radeon-testing at the end of May showed that glitch: every couple of seconds there's a very visual hickup, especially in scroll texts.
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
What about that new dri2 vsync stuff which was mentioned related to [Bug 28383]? Could the changes done for that in any way alter the timing? BTW I measured the glitches I'm experiencing and the appear to be to happen in intervals of 10 seconds. Again, all I'm changing is the kernel, and even the kernel config is the same. I'd be most grateful for any clues/hints/tips I could follow to resolve this regression.
If you have dynamic PM enabled, does disabling that help?
I checked again and there's method=profile and profile=default. Afaict this is not using dynpm, right?
Correct.
Alex
Alex Deucher schrieb:
2010/6/4 Marius Gröger marius.groeger@googlemail.com:
Hi All,
Michel Dänzer schrieb:
On Mit, 2010-06-02 at 08:07 +0200, Marius Gröger wrote:
Hello All,
I'm trying the top-of-trunk drm-2.6 trees (both drm-next and drm-radeon-testing) with my Radeon HD 3200 GPU over HDMI. The primary application is mythtv which uses DRM syncing for the frame syncronisation. Now, with the exact same userland software I noticed the introduction of sync gliches in the May-timeframe. The drm-radeon-testing on May 9 was still ok, but both drm-next and drm-radeon-testing at the end of May showed that glitch: every couple of seconds there's a very visual hickup, especially in scroll texts.
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
What about that new dri2 vsync stuff which was mentioned related to [Bug 28383]? Could the changes done for that in any way alter the timing? BTW I measured the glitches I'm experiencing and the appear to be to happen in intervals of 10 seconds. Again, all I'm changing is the kernel, and even the kernel config is the same. I'd be most grateful for any clues/hints/tips I could follow to resolve this regression.
If you have dynamic PM enabled, does disabling that help?
I checked again and there's method=profile and profile=default. Afaict this is not using dynpm, right?
Correct.
Ok so with dynpm more or less ruled out, what could have such a visible impact on the latencies? For instance, are we now more dependent (or less) on some kind of interrupt or deferred processing than 6 weeks ago?
Btw, I have HDMI audio pretty much ruled out as well.
Thanks Marius
2010/6/4 Marius Gröger marius.groeger@googlemail.com:
Alex Deucher schrieb:
2010/6/4 Marius Gröger marius.groeger@googlemail.com:
Hi All,
Michel Dänzer schrieb:
On Mit, 2010-06-02 at 08:07 +0200, Marius Gröger wrote:
Hello All,
I'm trying the top-of-trunk drm-2.6 trees (both drm-next and drm-radeon-testing) with my Radeon HD 3200 GPU over HDMI. The primary application is mythtv which uses DRM syncing for the frame syncronisation. Now, with the exact same userland software I noticed the introduction of sync gliches in the May-timeframe. The drm-radeon-testing on May 9 was still ok, but both drm-next and drm-radeon-testing at the end of May showed that glitch: every couple of seconds there's a very visual hickup, especially in scroll texts.
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
What about that new dri2 vsync stuff which was mentioned related to [Bug 28383]? Could the changes done for that in any way alter the timing? BTW I measured the glitches I'm experiencing and the appear to be to happen in intervals of 10 seconds. Again, all I'm changing is the kernel, and even the kernel config is the same. I'd be most grateful for any clues/hints/tips I could follow to resolve this regression.
If you have dynamic PM enabled, does disabling that help?
I checked again and there's method=profile and profile=default. Afaict this is not using dynpm, right?
Correct.
Ok so with dynpm more or less ruled out, what could have such a visible impact on the latencies? For instance, are we now more dependent (or less) on some kind of interrupt or deferred processing than 6 weeks ago?
Btw, I have HDMI audio pretty much ruled out as well.
Any chance you can bisect the problematic commit?
Alex
Am Fri, 4 Jun 2010 11:17:12 -0400 schrieb Alex Deucher alexdeucher@gmail.com:
2010/6/4 Marius Gröger marius.groeger@googlemail.com:
Alex Deucher schrieb:
2010/6/4 Marius Gröger marius.groeger@googlemail.com:
Hi All,
Michel Dänzer schrieb:
On Mit, 2010-06-02 at 08:07 +0200, Marius Gröger wrote:
Hello All,
I'm trying the top-of-trunk drm-2.6 trees (both drm-next and drm-radeon-testing) with my Radeon HD 3200 GPU over HDMI. The primary application is mythtv which uses DRM syncing for the frame syncronisation. Now, with the exact same userland software I noticed the introduction of sync gliches in the May-timeframe. The drm-radeon-testing on May 9 was still ok, but both drm-next and drm-radeon-testing at the end of May showed that glitch: every couple of seconds there's a very visual hickup, especially in scroll texts.
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
What about that new dri2 vsync stuff which was mentioned related to [Bug 28383]? Could the changes done for that in any way alter the timing? BTW I measured the glitches I'm experiencing and the appear to be to happen in intervals of 10 seconds. Again, all I'm changing is the kernel, and even the kernel config is the same. I'd be most grateful for any clues/hints/tips I could follow to resolve this regression.
If you have dynamic PM enabled, does disabling that help?
I checked again and there's method=profile and profile=default. Afaict this is not using dynpm, right?
Correct.
Ok so with dynpm more or less ruled out, what could have such a visible impact on the latencies? For instance, are we now more dependent (or less) on some kind of interrupt or deferred processing than 6 weeks ago?
Btw, I have HDMI audio pretty much ruled out as well.
Any chance you can bisect the problematic commit?
As I said, I already tried bisecting but failed. Perhaps I can try again and replay at least part of the log... But since we're talking about more than 120 commits I kinda was hoping to get some clues here first. Even with a tailored .config building/rebooting/testing takes ages. Well I suppose I needn't tell *you* guys... ;-)
Thanks Marius
2010/6/4 Marius Gröger marius.groeger@googlemail.com:
Am Fri, 4 Jun 2010 11:17:12 -0400 schrieb Alex Deucher alexdeucher@gmail.com:
2010/6/4 Marius Gröger marius.groeger@googlemail.com:
Alex Deucher schrieb:
2010/6/4 Marius Gröger marius.groeger@googlemail.com:
Hi All,
Michel Dänzer schrieb:
On Mit, 2010-06-02 at 08:07 +0200, Marius Gröger wrote: > Hello All, > > I'm trying the top-of-trunk drm-2.6 trees (both drm-next and > drm-radeon-testing) with my Radeon HD 3200 GPU over HDMI. The > primary application is mythtv which uses DRM syncing for the > frame syncronisation. Now, with the exact same userland > software I noticed the introduction of sync gliches in the > May-timeframe. The drm-radeon-testing on May 9 was still ok, > but both drm-next and drm-radeon-testing at the end of May > showed that glitch: every couple of seconds there's a very > visual hickup, especially in scroll texts. > > Apologies for such an unspecific description, and for what > almost seems like a support request for MythTV. I wouldn't post > here if I were not 100% sure it must be related with the recent > drm changes. Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
What about that new dri2 vsync stuff which was mentioned related to [Bug 28383]? Could the changes done for that in any way alter the timing? BTW I measured the glitches I'm experiencing and the appear to be to happen in intervals of 10 seconds. Again, all I'm changing is the kernel, and even the kernel config is the same. I'd be most grateful for any clues/hints/tips I could follow to resolve this regression.
If you have dynamic PM enabled, does disabling that help?
I checked again and there's method=profile and profile=default. Afaict this is not using dynpm, right?
Correct.
Ok so with dynpm more or less ruled out, what could have such a visible impact on the latencies? For instance, are we now more dependent (or less) on some kind of interrupt or deferred processing than 6 weeks ago?
Btw, I have HDMI audio pretty much ruled out as well.
Any chance you can bisect the problematic commit?
As I said, I already tried bisecting but failed. Perhaps I can try again and replay at least part of the log... But since we're talking about more than 120 commits I kinda was hoping to get some clues here first. Even with a tailored .config building/rebooting/testing takes ages. Well I suppose I needn't tell *you* guys... ;-)
for vsync stuff, It's probably one of these: http://git.kernel.org/?p=linux/kernel/git/airlied/drm-2.6.git;a=commitdiff;h... http://git.kernel.org/?p=linux/kernel/git/airlied/drm-2.6.git;a=commitdiff;h...
Alex
Thanks Marius
Am 04.06.2010 17:17, schrieb Alex Deucher:
2010/6/4 Marius Grögermarius.groeger@googlemail.com:
Alex Deucher schrieb:
2010/6/4 Marius Grögermarius.groeger@googlemail.com:
Hi All,
Michel Dänzer schrieb:
On Mit, 2010-06-02 at 08:07 +0200, Marius Gröger wrote:
Hello All,
I'm trying the top-of-trunk drm-2.6 trees (both drm-next and drm-radeon-testing) with my Radeon HD 3200 GPU over HDMI. The primary application is mythtv which uses DRM syncing for the frame syncronisation. Now, with the exact same userland software I noticed the introduction of sync gliches in the May-timeframe. The drm-radeon-testing on May 9 was still ok, but both drm-next and drm-radeon-testing at the end of May showed that glitch: every couple of seconds there's a very visual hickup, especially in scroll texts.
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
[...]
Any chance you can bisect the problematic commit?
I did a second attempt at bisecting and now I'm confident it is this commit which broke my real-time performance in mythtv:
commit eb1f8e4f3be898df808e2dfc131099f5831d491d Author: Dave Airlie airlied@redhat.com Date: Fri May 7 06:42:51 2010 +0000
drm/fbdev: rework output polling to be back in the core. (v4)
Having found the commit is the good news. The bad news is that this commit is rather large, so again I'd be depended on the experts around here to isolate what's going on.
Dave, would you mind giving a hand as well as the author of this commit?
Thanks Marius
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
Tho not the case for MythTV on a embedded device requiring apps to use GLX wuld be to heavy. I would agree libdrm should be used.
Am 06.06.2010 18:47, schrieb James Simmons:
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
Tho not the case for MythTV on a embedded device requiring apps to use GLX wuld be to heavy. I would agree libdrm should be used.
I have found the cause for the glitch and just filed http://bugs.freedesktop.org/show_bug.cgi?id=28411.
Regards Marius
On Mon, Jun 7, 2010 at 3:52 AM, Marius Gröger marius.groeger@googlemail.com wrote:
Am 06.06.2010 18:47, schrieb James Simmons:
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
Tho not the case for MythTV on a embedded device requiring apps to use GLX wuld be to heavy. I would agree libdrm should be used.
I have found the cause for the glitch and just filed http://bugs.freedesktop.org/show_bug.cgi?id=28411.
Okay I can see the problem, now I have to think of a good solution,
I probably need to break down the locking a bit further which is a bit messy.
Dave.
Am 07.06.2010 00:15, wrote Dave Airlie:
On Mon, Jun 7, 2010 at 3:52 AM, Marius Gröger marius.groeger@googlemail.com wrote:
Am 06.06.2010 18:47, schrieb James Simmons:
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
Tho not the case for MythTV on a embedded device requiring apps to use GLX wuld be to heavy. I would agree libdrm should be used.
I have found the cause for the glitch and just filed http://bugs.freedesktop.org/show_bug.cgi?id=28411.
Okay I can see the problem, now I have to think of a good solution,
I probably need to break down the locking a bit further which is a bit messy.
Is there any news on this issue? I didn't notice any commits to d-r-t or d-n WRT this, but then again those repos aren't updated very frequently anyway (how come?)
Thanks T& Regards, Marius
On Mon, Aug 2, 2010 at 8:12 PM, Marius Gröger marius.groeger@googlemail.com wrote:
Am 07.06.2010 00:15, wrote Dave Airlie:
On Mon, Jun 7, 2010 at 3:52 AM, Marius Gröger marius.groeger@googlemail.com wrote:
Am 06.06.2010 18:47, schrieb James Simmons:
Apologies for such an unspecific description, and for what almost seems like a support request for MythTV. I wouldn't post here if I were not 100% sure it must be related with the recent drm changes.
Note that the DRM APIs are intended for use by userspace components of graphics drivers / API libraries, not applications directly. MythTV shouldn't use the DRM directly for synchronization but rather use GLX synchronization APIs.
Tho not the case for MythTV on a embedded device requiring apps to use GLX wuld be to heavy. I would agree libdrm should be used.
I have found the cause for the glitch and just filed http://bugs.freedesktop.org/show_bug.cgi?id=28411.
Okay I can see the problem, now I have to think of a good solution,
I probably need to break down the locking a bit further which is a bit messy.
Is there any news on this issue? I didn't notice any commits to d-r-t or d-n WRT this, but then again those repos aren't updated very frequently anyway (how come?)
Should be fixed in 2.6.35 final, 29508eb66bfacdef324d2199eeaea31e0cdfaa29 drm/radeon/kms: drop taking lock around crtc lookup.
Generally regression fixes go via the drm-fixes tree, the other trees get rebased/updated as patches arrive, we didn't really have much for d-r-t this cycle.
Dave.
dri-devel@lists.freedesktop.org
-
Alex Deucher -
Dave Airlie -
James Simmons -
Marius Gröger -
Michel Dänzer