[Mesa-dev] [PATCH 1/4] st/mesa: handle indirect samplers in arrays/structs properly (v3)

[Timothy Arceri]

On Fri, 2016-02-05 at 14:35 +1000, Dave Airlie wrote:
> On 5 February 2016 at 14:28, Ilia Mirkin <imirkin at alum.mit.edu>
> wrote:
> > On Thu, Feb 4, 2016 at 11:25 PM, Dave Airlie <airlied at gmail.com>
> > wrote:
> > > > +
> > > > +   *base = this->shader_program-
> > > > >UniformStorage[location].opaque[shader].index;
> > > 
> > > self-review: this needs to use UniformRemapTable here, so it
> > > picks up
> > > the correct uniforms
> > > (fixes fs-texelFetch-2D).
> > 
> > So... _mesa_get_sampler_uniform_value does:
> > 
> >    get_sampler_name getname(sampler, shader_program);
> >    unsigned location;
> >    shader_program->UniformHash->get(location, getname.name)
> > return shader_program-
> > >UniformStorage[location].opaque[shader].index +
> >           getname.offset
> > 
> > Was it wrong before? Or are you getting the location incorrectly?
> 
> Looking up the value in the UniformHash with the name gives you the
> correct answer,

Looking up the name in UniformHash will give you the base location of
the array you are looking at so the offset is just the array offset.

> 
> pulling it from var->data.location means going via the
> UniformRemapTable.

var->data.location

Is the base location of the entire structure/array, if the offset is
calculated correctly I'm not sure why you would need UniformRemapTable

> 
> Dave.
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/mesa-dev