[cairo] [RFC] altivec optimization

[Luca Barbato]

Owen Taylor wrote:
> On Wed, 2007-01-24 at 15:20 +0100, Luca Barbato wrote:
> 
> This may or may not be helpful, but in developing MMX code to do
> compositing (something of a precursor of Soeren's code in fbmmx.c), 
> I found it was easier to handle the four channels of a single pixel
> in parallel rather than trying to do pixels in in parallel.

Currently most of the functions process 4 pixels at time 2 channels per 
step.

> 
> Which gets rid of the problem of left-over pixels at the ends of
> lines.
> 

Unpacking a pixel in a 128bit vector doesn't sound good =/

Anyway here is a patch with the autotools bits and some code, please 
comment.


lu

-- 

Luca Barbato

Gentoo/linux Gentoo/PPC
http://dev.gentoo.org/~lu_zero

-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-Initial-support-for-VMX-Altivec.patch
Type: text/x-patch
Size: 37451 bytes
Desc: not available
Url : http://lists.freedesktop.org/archives/cairo/attachments/20070125/3114c7e3/0001-Initial-support-for-VMX-Altivec-0001.bin