Summary: | x11-driver-video-radeonhd-1.2.5-1mdv2010.0 hangs / freezes after switching to virtual console and back. | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Product: | xorg | Reporter: | Shlomi Fish <shlomif> | ||||||||
Component: | Driver/radeonhd | Assignee: | Luc Verhaegen <lverhaegen> | ||||||||
Status: | RESOLVED WONTFIX | QA Contact: | Xorg Project Team <xorg-team> | ||||||||
Severity: | critical | ||||||||||
Priority: | high | CC: | mat, shlomif | ||||||||
Version: | 7.3 (2007.09) | ||||||||||
Hardware: | x86 (IA32) | ||||||||||
OS: | Linux (All) | ||||||||||
Whiteboard: | |||||||||||
i915 platform: | i915 features: | ||||||||||
Attachments: |
|
Description
Shlomi Fish
2009-09-26 11:54:53 UTC
(In reply to comment #0) Here is the bug in the Mandriva database: https://qa.mandriva.com/show_bug.cgi?id=54078 Regards, -- Shlomi Fish (In reply to comment #0) I can second that, but I believe it has nothing to with radeonhd, because when enabling the DRI option in xorg.conf, both radeonhd and radeon show the same behavior. No problem without DRI. My card is a HD 2600 Pro and I'm running kernel 2.6.31.1, 32-bit; still have to try latest libdrm though. My guess: try updating libdrm, Mesa, and use the 2.6.31 kernel (at least). radeonhd used to not enable DRI by default, which it does now. (In reply to comment #3) > My guess: try updating libdrm, Mesa, and use the 2.6.31 kernel (at least). > > radeonhd used to not enable DRI by default, which it does now. > What are the minimum versions of libdrm and Mesa that I need? I'm already using 2.6.31.2-desktop-0.rc1.1mnb . Regards, -- Shlomi Fish (In reply to comment #3) > My guess: try updating libdrm, Mesa, and use the 2.6.31 kernel (at least). > > radeonhd used to not enable DRI by default, which it does now. > This bug is still present in a close-to-latest Cooker with: * kernel 2.6.31.5-desktop-1mnb * xorg-x11-7.3-6mdv2010.0 * x11-driver-video-radeonhd-1.2.5-1mdv2010.0 * mesa-7.6-0.1mdv2010.0 * And I'm on IceWM from CVS head. This really makes it harder to test stuff in other fresh Unix user accounts. Why should we suffer from such a bug for over a month? (seems longer). Mr. Hopf - why haven't you replied to my question until now? Regards, -- Shlomi Fish What's the GPU in question? Can you ssh in and get at Xorg.0.log? dmesg? (In reply to comment #6) > What's the GPU in question? > I'm using an ATI Radeon HD 2600 Pro. > Can you ssh in and get at Xorg.0.log? dmesg? > Yes, I can - ssh is still working. I'm now trying to disable DRI. Regards, -- Shlomi Fish (In reply to comment #5) > Why should we suffer from such a bug for over a month? (seems longer). Because nobody has time and/or expertise to work on it? This is open source software - if something bothers you you can work on it yourself! I can only work on radeonhd if I have spare time, so do the other developers. > Mr. Hopf - why haven't you replied to my question until now? If I don't know the answer... BTW - there is a long standing R5xx DRI race, which apparently has been fixed by Egbert. You might want to retest git master. (In reply to comment #7) > (In reply to comment #6) > > What's the GPU in question? > > > > I'm using an ATI Radeon HD 2600 Pro. > > > Can you ssh in and get at Xorg.0.log? dmesg? > > > > Yes, I can - ssh is still working. > > I'm now trying to disable DRI. > With DRI disabled, everything works fine. I can switch to a virtual console and back, start new X sessions, etc. I'm a happy cat. I didn't seem to have noticed any slowdown in my XFCE session with DRI disabled (I'm using XFCE with many KDE and Gtk+ apps), but the KDE 4.x desktop itself may be very sluggish (not sure - did not try). What are the implications of DRI disabled? Regards, -- Shlomi Fish As the 2600 is a R6xx card, you will have no hardware acceleration when DRI is disabled. Sounds worse than it is (shadowfb is pretty good), however, video might be slow. And of course 3D applications won't work. BTW - the R5xx fix in master I mentioned earlier doesn't help in your case. (In reply to comment #10) > As the 2600 is a R6xx card, you will have no hardware acceleration when DRI is > disabled. Sounds worse than it is (shadowfb is pretty good), however, video > might be slow. Thanks! > And of course 3D applications won't work. > > BTW - the R5xx fix in master I mentioned earlier doesn't help in your case. > OK. Please let me know when I can re-enable DRI. And please give more specific instructions with exact versions / dates / etc. of stuff to try to upload (and hopefully make it a stable version that I can download without messing with git). This is still a serious bug. Regards, -- Shlomi Fish Any news? Created attachment 32088 [details]
My Xorg.0.log after switching back to X. (from an ssh section)
This is my Xorg.0.log file after switching back to X from a virtual console. I'm using the radeonhd driver from git master, kernel 2.6.32-desktop-1mnb , mesa-7.6.1-0.rc3.2mdv2010.1 , x11-server-xorg-1.7.3-1mdv2010.1 , xorg-x11-7.5-1mdv2010.1 - all from Mandriva Cooker.
I'll attach the dmesg soon.
Created attachment 32089 [details]
The dmesg file. See the previous attachment.
Does this issue occur with the preferred ati driver (xf86-vide-ati)? If so, please move this to the Driver/Radeon component. Development of radeonhd has pretty much halted and development focus is on the ati driver. Please see http://www.x.org/wiki/radeonhd If the issue does not exist in the ati driver (or if there is no response to this message), this bug will be closed as WONTFIX unless someone contributes a patch. It indeed did not happen with the radeon driver (and we ended up decomissioning the machine with the ATI Radeon HD 2600 Pro card, anyhow). So I'm closing this bug as WONTFIX, because the video-ati driver is OK. |
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.