On Sat, 2016-02-06 at 17:10 -0500, Ilia Mirkin wrote:

It would be have been nice to have some reasoning in here,
the scenario its fixing etc.

> Signed-off-by: Ilia Mirkin <[email protected]>
> Cc: [email protected]
> ---
>  src/compiler/glsl/linker.cpp | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/src/compiler/glsl/linker.cpp
> b/src/compiler/glsl/linker.cpp
> index 4776ffa..f1ac53a 100644
> --- a/src/compiler/glsl/linker.cpp
> +++ b/src/compiler/glsl/linker.cpp
> @@ -2125,6 +2125,7 @@ link_intrastage_shaders(void *mem_ctx,
>  
>        if (ok) {
>           memcpy(linking_shaders, shader_list, num_shaders *
> sizeof(gl_shader *));
> +         _mesa_glsl_initialize_builtin_functions();

Doesn't this defeat the purpose of glReleaseShaderCompiler() as its
currently implemented in mesa? Although thinking about it I can't think
of any better alternative.

If you add something to the commit message describing why we would need
to make sure this is initialize here i.e the problem scenario.

Reviewed-by: Timothy Arceri <[email protected]>


>           linking_shaders[num_shaders] =
> _mesa_glsl_get_builtin_function_shader();
>  
>           ok = link_function_calls(prog, linked, linking_shaders,
> num_shaders + 1);
_______________________________________________
mesa-dev mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/mesa-dev

Reply via email to