[Nouveau] [PATCH v3] nv110/exa: update sched codes

[Ilia Mirkin]

On Mon, Jun 19, 2017 at 11:22 AM, Aaryaman Vasishta
<jem456.vasishta at gmail.com> wrote:
> Sorry for the late response, been busy with some personal stuff + work...
>
> On Tue, Jun 13, 2017 at 6:52 AM, Ilia Mirkin <imirkin at alum.mit.edu> wrote:
>>
>> On Mon, Jun 12, 2017 at 5:46 PM, Samuel Pitoiset
>> <samuel.pitoiset at gmail.com> wrote:
>> >
>> >
>> > On 06/10/2017 09:14 AM, Aaryaman Vasishta wrote:
>> >>
>> >> See the 'wt' on the first fmul in exacanv110.fp, exacmnv110.fp and
>> >> exasanv110.fp. Any ideas on what could be causing the first fmul to
>> >> require
>> >> $r0 and/or $r1?
>> >
>> >
>> > 'tex nodep $r4 $r2 0x0 0x1 t2d 0xf'
>> >
>> > is actually:
>> >
>> > 'tex nodep $r4:$r7 $r2 0x0 0x1 t2d 0xf'
>
> That's definitely confusing, but quite interesting.
>>
>>
>> Actually more like:
>>
>> tex nodep $r4:$r5:$r6:$r7 $r2:$r3 ...
>
> Any idea why is it this way? The only way I could figure out so far is by
> trial and error, but is there any better way to detect such dependencies?

The t2d means "2 args for source" (with various additional modifiers
which could add arguments). 0xf means return all 4 components (it's a
component mask... e.g. 0x8 would be return only the alpha).

  -ilia