[Mesa-dev] [PATCH 3/3] intel: Clean up confusion between logical and physical surface dimensions.

[Kenneth Graunke]

On 01/08/2013 02:27 PM, Paul Berry wrote:
[snip]
> diff --git a/src/mesa/drivers/dri/intel/intel_mipmap_tree.h b/src/mesa/drivers/dri/intel/intel_mipmap_tree.h
> index 8e84bef..195fa88 100644
> --- a/src/mesa/drivers/dri/intel/intel_mipmap_tree.h
> +++ b/src/mesa/drivers/dri/intel/intel_mipmap_tree.h
> @@ -215,22 +215,29 @@ struct intel_mipmap_tree
>      GLuint first_level;
>      GLuint last_level;
>
> -   GLuint width0, height0, depth0; /**< Level zero image dimensions */
> +   /**
> +    * Level zero image dimensions.  These dimensions correspond to the
> +    * physical layout of data in memory.  Accordingly, they account for the
> +    * extra width, height, and or depth that must be allocated in order to
> +    * accommodate multisample formats, and they account for the extra factor
> +    * of 6 in depth that must be allocated in order to accommodate cubemap
> +    * textures.
> +    */
> +   GLuint physical_width0, physical_height0, physical_depth0;
> +
>      GLuint cpp;
>      GLuint num_samples;
>      bool compressed;
>
>      /**
> -    * If num_samples > 0, then singlesample_width0 is the value that width0
> -    * would have if instead a singlesample miptree were created. Note that,
> -    * for non-interleaved msaa layouts, the two values are the same.
> -    *
> -    * If num_samples == 0, then singlesample_width0 is undefined.
> +    * Level zero image dimensions.  These dimsensions correspond to the

"dimensions" (typo here).

This does seem much clearer.  Thanks Paul!

For the series:
Reviewed-by: Kenneth Graunke <kenneth at whitecape.org>