[Mesa-dev] [PATCH] glsl: Generate an error for array-of-array declarations

[Kenneth Graunke]

On 10/24/2011 12:29 PM, Ian Romanick wrote:
> From: Ian Romanick <ian.d.romanick at intel.com>
> 
> Other parts of the code already caught things like 'float x[4][2]'.
> However, nothing caught 'float [4] x[2]'.
> 
> Fixes piglit test array-multidimensional-new-syntax.vert.
> 
> Signed-off-by: Ian Romanick <ian.d.romanick at intel.com>
> ---
>  src/glsl/ast_to_hir.cpp |   12 ++++++++++++
>  1 files changed, 12 insertions(+), 0 deletions(-)
> 
> diff --git a/src/glsl/ast_to_hir.cpp b/src/glsl/ast_to_hir.cpp
> index 70afb67..b08b443 100644
> --- a/src/glsl/ast_to_hir.cpp
> +++ b/src/glsl/ast_to_hir.cpp
> @@ -2476,6 +2476,18 @@ ast_declarator_list::hir(exec_list *instructions,
>        }
>  
>        if (decl->is_array) {
> +	 /* From page 19 (page 25) of the GLSL 1.20 spec:
> +	  *
> +	  *     "Only one-dimensional arrays may be declared."
> +	  */
> +	 if (decl_type->is_array()) {
> +	    _mesa_glsl_error(& loc, state,
> +			     "invalid array of `%s' in declaration of `%s'",
> +			     decl_type->name,
> +			     decl->identifier);
> +	    continue;
> +	 }
> +
>  	 var_type = process_array_type(&loc, decl_type, decl->array_size,
>  				       state);
>        } else {

Is there some reason you didn't this down into process_array_type
itself?  It may be unnecessary, but it's probably better safe than
sorry, right?

Plus process_array_type has a lovely
   /* FINISHME: Reject delcarations of multidimensional arrays. */
that you could swap out for this code.