Summary: | XVIDEO performance regression with radeon | ||||||
---|---|---|---|---|---|---|---|
Product: | xorg | Reporter: | Bruce Cole <bacole9> | ||||
Component: | Driver/Radeon | Assignee: | Xorg Project Team <xorg-team> | ||||
Status: | RESOLVED DUPLICATE | QA Contact: | |||||
Severity: | normal | ||||||
Priority: | high | ||||||
Version: | 7.0.0 | ||||||
Hardware: | x86 (IA32) | ||||||
OS: | Linux (All) | ||||||
Whiteboard: | |||||||
i915 platform: | i915 features: | ||||||
Attachments: |
|
Description
Bruce Cole
2006-04-02 05:02:14 UTC
Created attachment 5157 [details]
X server log
This is probably due to bug 5801. Enabling the DRI might also improve performance. "Enabling the DRI might also improve performance" My X log reported DRI issues under 6.8.2 as well: (WW) RADEON(0): Direct rendering not yet supported on Radeon 9500 and newer cards and (II) RADEON(0): Direct rendering disabled yet xvideo performance was fine. I'm not interested in 3D, but this sounds just like the problem reported here: http://lists.debian.org/debian-amd64/2006/02/msg00330.html Is there a workaround? Note also xine-check reports: [ good ] found xvinfo: X-Video Extension version 2.2 [ good ] your Xv extension supports YV12 overlays (improves MPEG performance) [ good ] your Xv extension supports YUY2 overlays [ good ] Xv ports: YUY2 UYVY YV12 I420 I didn't say that the DRI not being enabled is the problem, just that enabling it might improve performance enough to mitigate the missing MTRR support. Bug 5801 has a patch, please try that. (In reply to comment #4) > I didn't say that the DRI not being enabled is the problem, just that enabling > it might improve performance enough to mitigate the missing MTRR support. > > Bug 5801 has a patch, please try that. Alright, thanks. Any build snapshot? It's been a few years since I've compiled X and I don't have all the requisite packages installed on my desktop at the moment. MTRR support should be fixed in xserver HEAD, so you could try a DRI snapshot from http://dri.freedesktop.org/snapshots/ . A common snapshot may be enough, but if that doesn't run, try an r300 snapshot as well. (In reply to comment #6) > MTRR support should be fixed in xserver HEAD, so you could try a DRI snapshot > from http://dri.freedesktop.org/snapshots/ . A common snapshot may be enough, > but if that doesn't run, try an r300 snapshot as well. (In reply to comment #6) > MTRR support should be fixed in xserver HEAD, so you could try a DRI snapshot > from http://dri.freedesktop.org/snapshots/ . A common snapshot may be enough, > but if that doesn't run, try an r300 snapshot as well. Thanks; looks like the snapshots haven't built in a couple days. I actually built from cvs and found that the mainline code does in fact restore xvideo performance to X 6.8.2 levels. Thank you. Assuming this is because of bug 5801, I've marked this a dup. Now if I could only figure out autoconf switches to match fedora's distribution (for install paths and such) , I could probably drop-replace the X server binary to patch my distro... *** This bug has been marked as a duplicate of 5801 *** |
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.