[PATCH v3 3/3] d3dcompiler: Store some builtin types in the hlsl_ctx structure.

[Matteo Bruni]

On Sat, May 30, 2020 at 5:22 PM Zebediah Figura <zfigura at codeweavers.com> wrote:
>
>
>
> On 5/29/20 4:32 PM, Matteo Bruni wrote:
> > On Fri, May 22, 2020 at 10:46 PM Zebediah Figura <z.figura12 at gmail.com> wrote:
> >>
> >> Signed-off-by: Zebediah Figura <zfigura at codeweavers.com>
> >> ---
> >> v3: remove some earlier variable names...
> >>
> >>  dlls/d3dcompiler_43/d3dcompiler_private.h |   8 ++
> >>  dlls/d3dcompiler_43/hlsl.y                | 132 +++++++++++-----------
> >>  2 files changed, 76 insertions(+), 64 deletions(-)
> >>
> >> diff --git a/dlls/d3dcompiler_43/hlsl.y b/dlls/d3dcompiler_43/hlsl.y
> >> index 058a627dc27..4d89c61ef44 100644
> >> --- a/dlls/d3dcompiler_43/hlsl.y
> >> +++ b/dlls/d3dcompiler_43/hlsl.y
> >> @@ -234,12 +243,23 @@ static void declare_predefined_types(struct hlsl_scope *scope)
> >>                          sprintf(name, "%s", names[bt]);
> >>                          type = new_hlsl_type(d3dcompiler_strdup(name), HLSL_CLASS_SCALAR, bt, x, y);
> >>                          add_type_to_scope(scope, type);
> >> +                        hlsl_ctx.builtin_types.scalar[bt] = type;
> >>                      }
> >>                  }
> >>              }
> >>          }
> >>      }
> >>
> >> +    for (bt = 0; bt <= HLSL_SAMPLER_DIM_MAX; ++bt)
> >> +    {
> >> +        type = new_hlsl_type(d3dcompiler_strdup(sampler_names[bt]), HLSL_CLASS_OBJECT, HLSL_TYPE_SAMPLER, 1, 1);
> >> +        type->sampler_dim = bt;
> >> +        hlsl_ctx.builtin_types.sampler[bt] = type;
> >> +    }
> >> +
> >> +    hlsl_ctx.builtin_types.Void = new_hlsl_type(d3dcompiler_strdup("void"), HLSL_CLASS_OBJECT, HLSL_TYPE_VOID, 1, 1);
> >> +    add_type_to_scope(scope, hlsl_ctx.builtin_types.Void);
> >> +
> >
> > This is a sneaky tiny bugfix, right?
> >
>
> In that void wasn't added to the scope before? Maybe, except that "void"
> is uniquely a keyword for whatever reason, so I'm not sure what
> difference it'd make...
>
> ...and I guess in that case there's no point adding it to the scope, is
> there?

Probably not. It might have been intentional, even.