[Mesa-dev] [PATCH] i965/fs: Don't compute_to_mrf() in the optimization loop.

[Matt Turner]

On Thu, Oct 30, 2014 at 9:55 AM, Ian Romanick <idr at freedesktop.org> wrote:
> On 10/29/2014 02:31 PM, Matt Turner wrote:
>> ... or on Gen >= 7 at all. We use load_payload to gather results for the
>> FB write(s) now, so we never write to MRFs directly. It's still called
>> after lower_load_payload() since that will generate MOVs to MRFs on
>> platforms with MRFs.
>>
>> No differences in shader-db on Haswell (Gen 7.5).
>
> But did you try it on a GEN that has MRFs?

I did not, and my assertion that everything uses load_payload seems to
be false. Jason said Gen5 did but Gen4 didn't but I don't even see
that Gen5 does.

I'll drop the second hunk of the patch and just disable the pass for Gen >= 7.