On Tue, Jul 14, 2015 at 11:40 AM, Siarhei Siamashka
<[email protected]> wrote:
> Acked-by: Siarhei Siamashka <[email protected]>
>
> Just one question about the code. There are some places where both big
> and little endian variants are identical but we still have an ifdef:
>
>> +static force_inline vector unsigned int
>> +expand_pixel_32_1x128 (uint32_t data)
>> +{
>> + vector unsigned int vdata;
>> +
>> + vdata = unpack_32_1x128 (data);
>> +
>> +#ifdef WORDS_BIGENDIAN
>> + return vec_perm (vdata, vdata,
>> + (vector unsigned char)AVV (
>> + 0x08, 0x09, 0x0A, 0x0B, 0x0C, 0x0D, 0x0E, 0x0F,
>> + 0x08, 0x09, 0x0A, 0x0B, 0x0C, 0x0D, 0x0E, 0x0F));
>> +#else
>> + return vec_perm (vdata, vdata,
>> + (vector unsigned char)AVV (
>> + 0x08, 0x09, 0x0A, 0x0B, 0x0C, 0x0D, 0x0E, 0x0F,
>> + 0x08, 0x09, 0x0A, 0x0B, 0x0C, 0x0D, 0x0E, 0x0F));
>> +#endif
>
> Is this intentional?
No, and it is of course erroneous. The reason it slipped under my
radar is because this function is not used at all.
Thanks for catching that :)
I removed it from the original patch.
Oded
_______________________________________________
Pixman mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/pixman
