[Piglit] [PATCH 1/2] glslparsertest: Added some few more struct construction tests

[Timothy Arceri]

On Thu, 2016-08-11 at 17:49 -0700, Dylan Baker wrote:
> Quoting Timothy Arceri (2016-08-11 17:39:19)
> > 
> > On Thu, 2016-08-11 at 10:43 -0700, Dylan Baker wrote:
> > > 
> > > Quoting Timothy Arceri (2016-08-08 20:14:31)
> > > > 
> > > > 
> > > > On Fri, 2016-08-05 at 17:43 +0300, Andres Gomez wrote:
> > > > > 
> > > > > 
> > > > > This set focus on checking invalid structs construction.
> > > > > Specifically,
> > > > > adds a test to check that only implicit conversions are
> > > > > perform
> > > > > and
> > > > > vector conversions of the type vec3 to vec2 through dropping
> > > > > the
> > > > > z
> > > > > component are not allowed.
> > > > 
> > > > 
> > > > IMO we should try to avoid adding more tests to these
> > > > directories.
> > > > 
> > > > I would much rather these be added to tests/spec/glsl-
> > > > 1.20/compiler/constructor/
> > > > 
> > > > I would then name them struct-1.vert .. struct-N.vert
> > > > 
> > > > It's up to you if you want to move/rename the existing
> > > > contructor
> > > > tests
> > > > but eventually I think this is something we should do.
> > > 
> > > IIRC Brian had requested that the whole folder be moved in one go
> > > rather
> > > than piecemeal to ease tracking regressions (the same for
> > > shaders/
> > > and
> > > other non-sorted folders).
> > 
> > shaders/ contains c programs so I can see it being a bit more
> > tricky
> > but the tests in glslparsertest should only require a 'git mv'
> > where we
> > put it in the appropriate folder and rename it with a more
> > descritive
> > name.
> > 
> > I don't think there is any real problem with regression tracking
> > here
> > that moving in one go will solve.
> 
> I means you only have to know that commit xzy moves all of these
> tests,
> as opposed to tests 1-10 were moved a week ago, 11-30 were moved two
> months ago, etc...

I see. Well I'd be surprised if someone sat down and moved all 500~
tests in one go rather than multiple people chipping away at it. But I
guess we will see :)

> 
> > 
> > 
> > > 
> > >  And if we do it in one big go we can actually
> > > have the python framework rename the tests for us, making it
> > > seamless.
> > > 
> > > Dylan
> > > _______________________________________________
> > > Piglit mailing list
> > > Piglit at lists.freedesktop.org
> > > https://lists.freedesktop.org/mailman/listinfo/piglit
> _______________________________________________
> Piglit mailing list
> Piglit at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/piglit