[Mesa-dev] [PATCH v2 09/11] intel/nir: Use nir_shrink_vec_array_vars

[Jason Ekstrand]

On Tue, Aug 21, 2018 at 6:00 PM Caio Marcelo de Oliveira Filho <
caio.oliveira at intel.com> wrote:

> > diff --git a/src/intel/compiler/brw_nir.c b/src/intel/compiler/brw_nir.c
> > index 96ad77c3906..5e9da9e1ef2 100644
> > --- a/src/intel/compiler/brw_nir.c
> > +++ b/src/intel/compiler/brw_nir.c
> > @@ -542,6 +542,7 @@ brw_nir_optimize(nir_shader *nir, const struct
> brw_compiler *compiler,
> >     do {
> >        progress = false;
> >        OPT(nir_split_array_vars, nir_var_local);
> > +      OPT(nir_shrink_vec_array_vars, nir_var_local);
>
> I noticed these new passes are being used only for the nir_va_local
> case.  Any reason not go ahead and call them with global too?
>

Because we inline functions and get rid of all global variables before we
get here.

--Jason
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/mesa-dev/attachments/20180821/5f16dca7/attachment.html>