Philipp Klaus Krause wrote:
> Brian Paul schrieb:
>
>> Call it an oversight, or defensive programming. :)
>
> Well, I'd say it's redundant and confusing, so I propose to remove the
> code (the comments about minimum size can stay, since they're still
> true). Patch attached (warning: Not tested).
Brian Paul schrieb:
>
> Call it an oversight, or defensive programming. :)
Well, I'd say it's redundant and confusing, so I propose to remove the
code (the comments about minimum size can stay, since they're still
true). Patch attached (warning: Not tested).
Philipp
diff --git a/src/mesa/main/t
Roland Scheidegger wrote:
> On 22.11.2008 02:00, Brian Paul wrote:
>> Philipp Klaus Krause wrote:
>>> In _mesa_compressed_texture_size(), texcompress.c there's lots of code
>>> similar to the following:
>>>
>>> case MESA_FORMAT_RGBA_DXT1:
>>> /* round up width, height to next multiple of 4 */
On 22.11.2008 02:00, Brian Paul wrote:
> Philipp Klaus Krause wrote:
>> In _mesa_compressed_texture_size(), texcompress.c there's lots of code
>> similar to the following:
>>
>> case MESA_FORMAT_RGBA_DXT1:
>> /* round up width, height to next multiple of 4 */
>> width = (width + 3) & ~3
Philipp Klaus Krause wrote:
> In _mesa_compressed_texture_size(), texcompress.c there's lots of code
> similar to the following:
>
> case MESA_FORMAT_RGBA_DXT1:
> /* round up width, height to next multiple of 4 */
> width = (width + 3) & ~3;
> height = (height + 3) & ~3;
>
In _mesa_compressed_texture_size(), texcompress.c there's lots of code
similar to the following:
case MESA_FORMAT_RGBA_DXT1:
/* round up width, height to next multiple of 4 */
width = (width + 3) & ~3;
height = (height + 3) & ~3;
/* 8 bytes per 4x4 tile of RGB[A] texels */