[Mesa-dev] [RFC PATCH 00/16] A new IR for Mesa

[Michel Dänzer]

On 28.08.2014 19:58, Henri Verbeet wrote:
> On 28 August 2014 05:21, Michel Dänzer <michel at daenzer.net> wrote:
>>> Sure, it's not impossible, but is that really the kind of process you
>>> want users to go through when bisecting a regression?
>>
>> I appreciate your theoretical concern, but in practice, people don't seem to
>> have trouble bisecting radeonsi regressions in general.
>>
> I suspect you may be getting some selection bias there.
> As far as Wine users are concerned, we certainly seem to have more
> r600g users than radeonsi ones.

That's hardly surprising, considering the respective periods of 
availability of the hardware and drivers.

> For Wine developers that comparison is even worse; as far as I'm aware
> none of the regular developers regularly develop on radeonsi. I've seen
> a couple more casual developers try, but I suspect they essentially gave
> up once they realized how much work would be required to make the Wine
> tests pass on radeonsi.

What kind of work are you referring to?

 From a piglit perspective, radeonsi has been on par with r600g for a 
while, possibly even slightly better now. Please file bug reports for 
Wine test failures.


> Perhaps more concretely, I think the r600-sb backend works at least as
> well as the r600-llvm one,

SB currently works better overall for graphics, which is why we decided 
to make it the default.

> and not for lack of effort put into the latter.

Sounds like you're overestimating the effort put into the LLVM R600 
backend for pre-SI graphics.


-- 
Earthling Michel Dänzer            |                  http://www.amd.com
Libre software enthusiast          |                Mesa and X developer