[Mesa-dev] intel GPU freeze with master, with apitrace log

[tom fogal]

Hi all,

Our application has a render mode which is causing a Sandybridge-based
system to hang for a second with Mesa master.  In dmesg I see:

  [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU hung
  [drm:i915:do_wait_request] *ERROR* i915_do_wait_request returns -11 (awaiting 23347 at 23343, next 23348)

and on my terminal I get

  intel_do_flush_locked failed: Input/output error

A few seconds before X server does not seem to receive/process updates
(i.e. mouse movements).  Then the screen flickers once and after that
mouse movements are processed as normal.  I've made multiple traces,
and sometimes this appears to happen twice per run, other times it just
occurs once... might be a timing thing with how quickly I can close the
window after the first hang.

I've generated an apitrace (! really great tool!) log for it, but it's
a bit big (~3mb) to attach to a mailing list posting so I put it on the
web:

  http://www.sci.utah.edu/~tfogal/tmp/iv3d.trace

If I 'glretrace' that on the Intel system, things freeze for a second
just like when I run my application.  If I copy it to my nvidia-blob
system, the program displays a black window quickly and then ends; no
flickers / freezes / etc.

I'm a little stuck debugging... the GPU hang really screws up gdb.
Thoughts/ideas?

Thanks,

-tom