|
Bugzilla – Full Text Bug Listing |
| Summary: | fglrx breaks suspend to disk on Acer TM-8202WLMi | ||
|---|---|---|---|
| Product: | [openSUSE] openSUSE 10.2 | Reporter: | Anders Biström <suse> |
| Component: | X11 3rd Party | Assignee: | Jammy Zhou <atilinuxnovellbugs> |
| Status: | RESOLVED FIXED | QA Contact: | Stefan Dirsch <sndirsch> |
| Severity: | Normal | ||
| Priority: | P5 - None | CC: | behlert |
| Version: | Final | ||
| Target Milestone: | --- | ||
| Hardware: | i686 | ||
| OS: | Other | ||
| Whiteboard: | |||
| Found By: | Other | Services Priority: | |
| Business Priority: | Blocker: | --- | |
| Marketing QA Status: | --- | IT Deployment: | --- |
| Attachments: |
pm-suspend.log of a hibernate.
Xorg-log, Ati drivers v 8.35.5. xorg.conf |
||
|
Description
Anders Biström
2007-01-16 15:43:26 UTC
Created attachment 113150 [details]
pm-suspend.log of a hibernate.
s2ram does nothing with suspend to disk. What does "resumes immediately" mean? Does write the image to disk (you see the percent counting while it writes) but then reboots instead of shutting down, read back the image and restore your desktop? Does it just "do nothing" in the meaning of "it does not write anything to disk, just lock my screen and return"? The log looks perfectly fine, but reproducing this without fglrx and vmware modules loaded would of course add some credibility to the bugreport :-) As I mentioned, it is similar to the other bug.. It writes to disk, then when it is supposed to shutdown, instead it resumes immediatly, no shutdown, no reboot - just resuming like when booting up from hibernate, but w/o the reboot. No reboot here between those (from attached log above): Tue Jan 16 17:25:51 EET 2007: finished hibernate hooks. Tue Jan 16 17:26:24 EET 2007: running thaw hooks. Sorry about vmware and fglrx modules.. I know it isn't worth much, but opensuse 10.1 did work with these modules loaded. Xorg will not start at proper resolution w/o the fglrx module. I'll attach a new log w/o the modules as soon as I have time (in a few hours probably). Ok, did some trials. - Seems VMWare modules are ok and done properly. - ATI proprietary drivers (version 8.33.6) seems to have issue with hibernation in this version again. Big surprise, sigh. These tests were done from runlevel 3. With fglrx modprobed I'm out of luck regaring hibernation - otoh without fglrx I'm out of luck using Xorg.. Yet another reason to not like Ati. Sorry for causing noise on bugzilla over something out of the hands of all the hardworking ppl. Thank you all. On that note, it would be good for bugzilla to have a new resolution category - instead of "Invalid" or "Wontfix" there could be "Proprietary". In this way once every 6 months a list of bugs filed because of some company could be sent to that company to encourage fixing stuff. well, maybe we can actually assign this bug somehow over to ATI. Let's try it :-) Could you attach /etc/X11/xorg.conf and /var/log/Xorg.0.log? Thanks. Created attachment 128576 [details]
Xorg-log, Ati drivers v 8.35.5.
Attachment of most recent Xorg log.
Since using Xgl, Xorg is on :93
Updated to Atis newest drivers today (8.35.5).
Now suspend to disk fails because it is unable to unload fglrx.
Created attachment 128578 [details]
xorg.conf
xorg.conf as left after running aticonfig --initial --input=/etc/X11/xorg.xonf
To add, lines in pm-suspend.log that seems to cause immediate thaw after trying to suspend: --------------------------------------------- trying to unload: fglrx FATAL: Module fglrx is in use. # could not unload 'fglrx', usage count was 54 --------------------------------------------- This is with fglrx version: 8.35.5 Before I updated (fglrx 8.34.8), I experienced same problem but higher usage count (64 if I remember right). well, you cannot unload fglrx AFAIK, but this failure to unload will not stop the suspend process. You cannot unload the kernel module as long as the Xserver is running. The same applies to any other drm module (radeon, i915, etc.). Could you verify, if this issue is still reproducable with release 8.44? Driver download: http://ati.amd.com/support/drivers/linux/linux-radeon.html Installation instructions: http://www.suse.de/~sndirsch/ati-installer-HOWTO.html#manual Unfortunatly I think I have to leave 8.44 and wait for the next version since 8.44 seems to have an issue that affects me. "Connecting a display device that supports 1680x1050 to a system running Linux may result in a maximum display resolution of 1280x1024 only being available" My system is at 1680x1050, furthermore it seems many have this issue even at 1440x900. Sorry for late reply, been very busy. The resolution problem is an unrelated known regression. It is my understanding that the initial issue could be retested with 8.44 nevertheless. Will take some time that 8.45/Catalyst 8.1, which fixes the resolution issue, will be released. Hopefully still this month. So could you please 8.44 give a try to verify if the initial issue is still present? Thanks. 8.45/Catalyst 8.1 has been released. Ok, teste with Ati drivers version 8.452.1. At least suspend to disk (hibernate) seems to work properly now. Well done! This time I didn't have XGL on though, just plain Xorg - tried to move to AIGLX but it locks up the machine. Sorry for late reply, have just had way too much work. Thanks for testing! |