[Bug 89959] [SKL]igt/gem_concurrent_blit subcase sporadically causes system hang and *ERROR* blitter: timed out waiting for forcewake ack to clear.

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=89959

--- Comment #19 from Mika Kuoppala <mika.kuoppala at intel.com> ---
Fast way to reproduce is:

gem_concurrent_blit --run-subtest prw-blt-overwrite-source-read-rcs-forked

The forcewake ack to clear error messages are misleading as they result from
execlist interrupt handler accessing already dead gpu. 

These items are _less_ likely to be the root cause as I have disabled them
selectively without impact on occurrence.

- rc6
- dynamic page tables (reproducible with static pdps and aliasing ppgtt)
- execlist scheduling/pre-emption (reproducible with 1 slot entry submission)
- per ctx bb / pre-emption disabling (Arun's patches are already in)


So in the probable list remains:

- flush problem (tlb). Forced flushing per ring/bb didn't have impact.
- context save/restore corruption.
- missing workaround

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20150716/61c4bf64/attachment-0001.html>