[vlc-devel] [PATCH 3/6] copy: remove need for cache memory in SSE routines

[Jean-Baptiste Kempf]

On 13 Jun, Jean-Yves Avenard wrote :
> SSE code used a 16-bytes aligned memory buffer to perform memory operations.
> Rewrite code to directory work on source and destination frame, regardless of memory alignment.
> 
> This speeds up NV12->YV12 conversion by over 100% by halving the number of memory writes and reads.

So, you remove the CopyFromUswc part, right?

Is that not too slow when using DxVA?
https://software.intel.com/en-us/articles/copying-accelerated-video-decode-frame-buffers


-- 
Jean-Baptiste Kempf
http://www.jbkempf.com/ - +33 672 704 734
Sent from my Electronic Device