Re: Graphic corruption with i915 (2.6.39.1)

[Leandro Lucarella <luca@llucax.com.ar>]

Chris Wilson, el 16 de junio a las 17:39 me escribiste:
> On Thu, 16 Jun 2011 13:18:03 -0300, Leandro Lucarella <luca@llucax.com.ar> wrote:
> > cannot access /sys/kernel/debug/dri/0/i915_error_state: No such file or directory
> > 
> > The debug directory is not present, should I boot or compile with some
> > particular option to enable it?
> 
> You need to compile with CONFIG_DEBUGFS and at runtime
> "mount -tdebugfs debug /sys/kernel/debug".
> 
> The /sys/kernel/debug/dri/0/i915_error_state will only be populated
> after a hang is detected.

OK, I'm having another i915 "episode" and now I have the debugfs
mounted. Here is the contents of
/sys/kernel/debug/dri/0/i915_error_state:
http://llucax.com.ar/tmp/i915_error_state.txt (1.4MiB)

Here is the error message shown in dmesg:

[1259708.264742] Out of memory: Kill process 16640 (firefox-bin) score 504 or sacrifice child
[1259708.264748] Killed process 13128 (plugin-containe) total-vm:450880kB, anon-rss:19588kB, file-rss:780kB
[1259726.000008] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
[1259726.000014] [drm] capturing error event; look for more information in /debug/dri/0/i915_error_state
[1259726.001254] [drm:i915_wait_request] *ERROR* i915_wait_request returns -11 (awaiting 28517162 at 28517161, next 28517164)
[1259726.503009] [drm:i915_reset] *ERROR* Failed to reset chip.

Any ideas? Thanks!

-- 
Leandro Lucarella (AKA luca)                     http://llucax.com.ar/
----------------------------------------------------------------------
GPG Key: 5F5A8D05 (F8CD F9A7 BF00 5431 4145  104C 949E BFB6 5F5A 8D05)
----------------------------------------------------------------------
The average person laughs 13 times a day