Comment 10 for bug 708998

@Travis, actually apport *should* fire off if the GPU can't be restarted. However, there definitely are cases in the wild where for some reason the kernel doesn't fire off the event. NFI why though. Definitely would be nice to fix.

In any case, if apport doesn't get activated automatically, you could ssh in and I think you may be able to manually run the script /usr/share/apport/apport-gpu-error-intel.py, or even just collect the stuff directly (the output of 'intel_gpu_dump' and 'dmesg', and the /sys/kernel/debug/dri/0/i915_error_state file if it exists).