[Piglit] [PATCH] arb_separate_shader_objects: test mixed explicit and non-explicit locations

Timothy Arceri t_arceri at yahoo.com.au
Thu Nov 19 07:36:37 PST 2015


On Thu, 2015-11-19 at 14:21 +0100, Gregory Hainaut wrote:
> Hello Timothy,
> 
> > There seems to be 3 separate issues so I think we should have a separate
> > tests
> > (or at least cleary labebled subtests) for each of them.
> 
> In this case, I propose that we keep your new test. It has a better
> coverage than mine. Beside it will make my first test easier.
> 
> > I also think you will need to rework your test a bit because with my fixes
> > for
> > outputs it will max out the varying limits, this is consistent with the
> > issue
> > you were having on the Nvidia blob.
> Ok I will remove it.
> 
> Note: as far as I understand, spec allow to optimize the interface of
> a multiple-stage separated program. Extract from issue 8a and 8b of
> the spec. So both implementations were valid.
> "We allow the linker to perform internal optimizations on the
>         interfaces between stages of a multi-stage program object."

Yes you are right. In that case you can leave it as is, however we should
again make this case clear by having a separate piglit test or breaking it
into a subtest.

I also have modified my Mesa patches to allow for this, I'll send out V2
tomorrow after some additional testing.


> 
> Best regards,
> Gregory


More information about the Piglit mailing list