Summary: | [legacy 945GM] crashes (100% reproducible) on X startup or immediately (within seconds) after | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | xorg | Reporter: | celejar | ||||||
Component: | Driver/intel | Assignee: | Carl Worth <cworth> | ||||||
Status: | RESOLVED WONTFIX | QA Contact: | Xorg Project Team <xorg-team> | ||||||
Severity: | critical | ||||||||
Priority: | medium | ||||||||
Version: | 7.5 (2009.10) | ||||||||
Hardware: | x86 (IA32) | ||||||||
OS: | Linux (All) | ||||||||
Whiteboard: | |||||||||
i915 platform: | i915 features: | ||||||||
Attachments: |
|
Created attachment 38268 [details]
log from a crashed system
Why did you disable KMS? (In reply to comment #2) > Why did you disable KMS? KMS used to be broken: https://bugs.freedesktop.org/show_bug.cgi?id=25681 I've since re-enabled it, but when I started having this new problem, one of my flailing-around troubleshooting steps was to disable it and see if that made a difference. I'm pretty sure that I've seen the crashing both before disabling as well as after re-enabling it, but if you suspect that I've messed up here, I can try to get a log with KMS enabled. > --- Comment #3 from celejar@gmail.com 2010-08-29 11:35:45 PDT ---
> I'm pretty sure that I've seen the crashing both before disabling
> as well as after re-enabling it, but if you suspect that I've messed up here, I
> can try to get a log with KMS enabled.
>
Yes, please.
(In reply to comment #4) > > --- Comment #3 from celejar@gmail.com 2010-08-29 11:35:45 PDT --- > > I'm pretty sure that I've seen the crashing both before disabling > > as well as after re-enabling it, but if you suspect that I've messed up here, I > > can try to get a log with KMS enabled. > > > Yes, please. I'm still pretty confused about how I can determine, in a given case, whether KMS is active. On my current system, my config shows: CONFIG_DRM_I915=m # CONFIG_DRM_I915_KMS is not set but running the code provided by Chris Wilson here: https://bugs.freedesktop.org/show_bug.cgi?id=25681#c9 shows: ~# a.out /dev/dri/card0: GEM enabled How can I be sure that I'm running KMS? I've commented out the line in /etc/modprobe.d/i915-kms.conf that had disabled it #options i915 modeset=0 and (I'm pretty sure) rebuilt my initramfs. I suppose I'll have to rebuild the kernel now and turn on DRM_I915_KMS? > --- Comment #5 from celejar@gmail.com 2010-08-29 12:29:24 PDT ---
> How can I be sure that I'm running KMS? I've commented out the line in
> /etc/modprobe.d/i915-kms.conf that had disabled it
>
> #options i915 modeset=0
>
> and (I'm pretty sure) rebuilt my initramfs. I suppose I'll have to rebuild the
> kernel now and turn on DRM_I915_KMS?
>
Change that line to 'options i915 modeset=1', or pass i915.modeset=1 on
the kernel command line in grub.
(In reply to comment #6) > > --- Comment #5 from celejar@gmail.com 2010-08-29 12:29:24 PDT --- > > How can I be sure that I'm running KMS? I've commented out the line in > > /etc/modprobe.d/i915-kms.conf that had disabled it > > > > #options i915 modeset=0 > > > > and (I'm pretty sure) rebuilt my initramfs. I suppose I'll have to rebuild the > > kernel now and turn on DRM_I915_KMS? > > > Change that line to 'options i915 modeset=1', or pass i915.modeset=1 on > the kernel command line in grub. Okay, so now I'm running X successfully, after having changed the 'options' line as per above. I suppose it's possible that I really had KMS disabled all along ... Do you want me to see if reverting to KMS disabled reenables the crashing? > --- Comment #7 from celejar@gmail.com 2010-08-30 06:16:41 PDT ---
> Do you want me to see if reverting to KMS disabled reenables the crashing?
>
Don't bother, I'm fairly sure it will crash..
(In reply to comment #8) > > --- Comment #7 from celejar@gmail.com 2010-08-30 06:16:41 PDT --- > > Do you want me to see if reverting to KMS disabled reenables the crashing? > > > Don't bother, I'm fairly sure it will crash.. Boy, testing bleeding edge X is not really that much fun :/ With KMS enabled, my system is unstable again, as it used to be here: https://bugs.freedesktop.org/show_bug.cgi?id=25681 I thought that that had gone away, but I may have simply mistakenly thought that KMS was enabled when really it wasn't. In any event, what I'm seeing now is X starting up fine, but then freezing (frozen screen, no response to mouse, keyboard). This seems to happen shortly after starting to watch a video (Flash), but that may be a coincidence. OTOH, I've gone for several hours doing just word processing, email and simple CLI stuff. I may be able to get ssh access into a crashed system; if so, I'll see if I can get i915_error_state. Anything else that would be useful? (In reply to comment #9) > (In reply to comment #8) > > > --- Comment #7 from celejar@gmail.com 2010-08-30 06:16:41 PDT --- > > > Do you want me to see if reverting to KMS disabled reenables the crashing? > > > > > Don't bother, I'm fairly sure it will crash.. > > Boy, testing bleeding edge X is not really that much fun :/ With KMS enabled, > my system is unstable again, as it used to be here: > > https://bugs.freedesktop.org/show_bug.cgi?id=25681 > > I thought that that had gone away, but I may have simply mistakenly thought > that KMS was enabled when really it wasn't. Update: I actually haven't had any more crashes for several days now, and at the time I was experiencing other flakiness that was unrelated to X (possible hardware issues). If I see any more X problems, I'll report back. I'm no longer even contemplating bringing back the broken code for broken systems... |
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.
Created attachment 38267 [details] dmesg (from a working system) Debian Sid, Celeron M (i686). Problem experienced with various kernels, from 2.6.32 through 2.6.36-rc2 (currently 2.6.36-rc2-lizzie-00373-g29cfcdd). Acer laptop (Aspire 3690) with Intel 82801G (ICH7) stuff. I'm not sure which X version started the problem, but I currently get it consistently with the Debian package xserver-xorg-video-intel version 2.12.0+legacy1-1 0 (from Sid), but not with 2.9.1-4 (from Testing). X either crashes the system immediately, or seconds after starting. I'm attaching an Xorg log from the latter case. I tried looking at the i915_error_state (after reboot), but there's no state collected. I currently don't have the option of ssh'ing in from another machine.