On 2014-02-05 12:18:34 +0200, Martin Storsjö wrote:
> From: "Ronald S. Bultje" <[email protected]>
> 
> ---
>  libavcodec/arm/vp8dsp_neon.S | 165 
> +++++++++++++++++++++++--------------------
>  1 file changed, 87 insertions(+), 78 deletions(-)

ok in the sense that the code does the right thing afterwards. It should
be possible to avoid pushing lr to the stack by reording rsb and vdup and
loading mx/my directly to r12. Not sure which variant would be faster on
which CPUs though.

Janne
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to