On Wed, Aug 1, 2012 at 20:22:30 +0200, Julien Danjou wrote: > On Wed, Aug 01 2012, Julien Cristau wrote: > > > Actually, this is the problem, looks like the gpu is getting stuck. > > Please attach the error state to the report per > > http://intellinuxgraphics.org/i915_error_state.html > > I've rebooted since then, so it's unlikely it will match. Will it > matter? > > However, I've doubt it works: > > # dmesg | grep i915_hangcheck > [26777.930037] [drm:i915_hangcheck_ring_idle] *ERROR* Hangcheck timer > elapsed... render ring idle [waiting on 4043152, at 4043164], missed IRQ? > [26779.703745] [drm:i915_hangcheck_ring_idle] *ERROR* Hangcheck timer > elapsed... render ring idle [waiting on 4043376, at 4043388], missed IRQ? > [87899.133450] [drm:i915_hangcheck_ring_idle] *ERROR* Hangcheck timer > elapsed... render ring idle [waiting on 7816051, at 7816062], missed IRQ? > [87905.201593] [drm:i915_hangcheck_ring_idle] *ERROR* Hangcheck timer > elapsed... render ring idle [waiting on 7819423, at 7819434], missed IRQ? > > dex ~ % cat /sys/kernel/debug/dri/0/i915_error_state > no error state collected > OK. Sad :( Probably worth mentioning in the upstream report though, and trying a newer kernel (3.4 or 3.5).
> Or is it because debugfs was not mounted at the time the hang happened? > Seems unlikely, but who knows. > That doesn't matter. Cheers, Julien
signature.asc
Description: Digital signature
