[Bug 77742] i965: Support instruction compaction on Gen < 6

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=77742

--- Comment #4 from Samuel Iglesias <siglesias at igalia.com> ---
Created attachment 98676
  --> https://bugs.freedesktop.org/attachment.cgi?id=98676&action=edit
Patch

This is the patch I had developed with my errors fixed... but I couldn't make
it work on Ironlake. As I said in the previous comment, I found that in
Ironlake it outputs these kernel messages when I run piglit tests (after some
thousands of them):

[...]
[757604.292530] [drm] stuck on render ring
[757604.292581] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[757604.292583] [drm] GPU hangs can indicate a bug anywhere in the entire gfx
stack, including userspace.
[757604.292585] [drm] Please file a _new_ bug report on bugs.freedesktop.org
against DRI -> DRM/Intel
[757604.292587] [drm] drm/i915 developers can then reassign to the right
component if it's not a kernel issue.
[757604.292588] [drm] The gpu crash dump is required to analyze gpu hangs, so
please always attach it.
[757606.269708] [drm] stuck on render ring
[757608.270874] [drm] stuck on render ring
[...]

I don't know if in G45 architecture this patch would work (the struct values
are copied from its docs) because I don't access to a testing machine that has
it.

It might be that Ironlake needs different values or that my patch is directly
wrong. If you want I can submit the GPU crash dump file if needed.

Thanks

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/intel-3d-bugs/attachments/20140508/87036980/attachment.html>