| Summary: | X segfault in miCopyRegion | ||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Product: | xorg | Reporter: | Bryce Harrington <bryce> | ||||||||||||||||||||||
| Component: | Server/Acceleration/EXA | Assignee: | Xorg Project Team <xorg-team> | ||||||||||||||||||||||
| Status: | RESOLVED FIXED | QA Contact: | Xorg Project Team <xorg-team> | ||||||||||||||||||||||
| Severity: | major | ||||||||||||||||||||||||
| Priority: | high | CC: | gary.pajer, jerrylamos, konstantin, wenzhuo | ||||||||||||||||||||||
| Version: | 7.4 (2008.09) | Keywords: | regression | ||||||||||||||||||||||
| Hardware: | x86 (IA32) | ||||||||||||||||||||||||
| OS: | Linux (All) | ||||||||||||||||||||||||
| Whiteboard: | |||||||||||||||||||||||||
| i915 platform: | i915 features: | ||||||||||||||||||||||||
| Attachments: |
|
||||||||||||||||||||||||
|
Description
Bryce Harrington
2010-03-30 14:25:57 UTC
00:00.0 Host bridge [0600]: Intel Corporation 82855PM Processor to I/O Controller [8086:3340] (rev 03) Subsystem: IBM Device [1014:0529] 01:00.0 VGA compatible controller [0300]: ATI Technologies Inc Radeon Mobility M7 LW [Radeon Mobility 7500] [1002:4c57] Subsystem: IBM Device [1014:0530] Created attachment 34552 [details]
Xorg.0.log
Created attachment 34553 [details]
dmesg.txt
Created attachment 34554 [details]
XorgLogOld.txt
Created attachment 34555 [details]
XorgLog.txt
Created attachment 34556 [details]
dmesg
Created attachment 34557 [details]
dmesg #2
Created attachment 34558 [details]
CurrentDmesg.txt
Created attachment 34559 [details]
XorgLogOld.txt
A gdb backtrace with debugging symbols for the radeon driver (and preferably the X server as well) would be helpful. The reporters say this is all they can get on this bug: . Thread 1 (Thread 7382): #0 0x002b5fb6 in ?? () from /lib/tls/i686/cmov/libc.so.6 No symbol table info available. #1 0x00000000 in ?? () No symbol table info available. There's a coredump on the downstream bug 561433 if you're interested. (In reply to comment #11) > The reporters say this is all they can get on this bug: > > . > Thread 1 (Thread 7382): > #0 0x002b5fb6 in ?? () from /lib/tls/i686/cmov/libc.so.6 > No symbol table info available. > #1 0x00000000 in ?? () > No symbol table info available. It doesn't look like that this is caused by same bug. This looks like stack overflow. But in any case valgrind is the best tool for debugging stack overflows. Running xserver in valgrind and collecting the output to a file would be best way to collect info. something like: valgrind X <X params> > /tmp/valgrind.crash.report.txt > something like:
> valgrind X <X params> &> /tmp/valgrind.crash.report.txt
& mark is important missing character from original command.
Created attachment 35173 [details]
valgrind report
I am one of the downstream bug reporter. I just collected valgrind report using the following command:
sudo ls && sudo valgrind Xorg :1.0 &> /tmp/valgrind-X.log & sleep 15; export DISPLAY=:1.0; /etc/X11/Xsession
Looks like it might be related to bug 27510. Has Ubuntu backported EXA changes from 1.8 to 1.7? If not, I can help with backporting the fix from that bug for testing. Ubuntu has not backported anything significant for EXA from 1.8 so far, so a backport of what fixes you think should be included would be helpful for this bug. Created attachment 35200 [details] [review] Backported fix Does this fix the problem? Anxiously awaiting test packages from Bryce... > Anxiously awaiting test packages from Bryce... Here you go: https://edge.launchpad.net/~bryceharrington/+archive/purple/+packages I've been running the updated X server without crash for about 5 hours already. The problem can no longer be reproduced using my method, i.e. typing in the search bar of Firefox. Fantastic! Fix pushed to server-1.7-nominations Git branch, thanks for testing. I just noticed that when the updated X server is running, the Linux kernel issues a lot of error messages like these: [ 4955.429783] [drm:radeon_fence_wait] *ERROR* last signaled fence(0x0001FEE8) [ 4955.621867] [drm:radeon_fence_wait] *ERROR* fence(f69361a0:0x0001FEE9) 40ms timeout [ 4955.621880] [drm:radeon_fence_wait] *ERROR* last signaled fence(0x0001FEE9) [ 4955.644982] [drm:radeon_fence_wait] *ERROR* fence(f1980c60:0x0001FEEA) 52ms timeout (In reply to comment #23) > I just noticed that when the updated X server is running, the Linux kernel > issues a lot of error messages like these: And those only appear with the patch? Is there any noticeable negative effect other than the messages? I just checked the timestamps of the error messages. They were generated when the random screensaver was running (at dinner time). - In order to give the updated X server more test, I purposely enabled the random screensaver. - I don't know if the linux kernel issues the error messages too with the previous version of the X server. It's probably not a side effect of the fix. (In reply to comment #24) > And those only appear with the patch? Is there any noticeable negative effect > other than the messages? I downgraded to the previous version and confirmed that the Linux kernel also issues these error messages when the random Gnome-screensaver is running on the previous version. I am not sure which specific screensaver is causing this. Anyway, it's a separate problem than this issue. No noticeable negative side effect. Thanks! *** Bug 27869 has been marked as a duplicate of this bug. *** *** Bug 28262 has been marked as a duplicate of this bug. *** |
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.