[FFmpeg-devel] Patch: Inline asm fixes for Intel compiler on Windows
Reimar.Doeffinger at gmx.de
Sun Apr 6 07:59:59 CEST 2014
On 06.04.2014, at 06:29, Matt Oliver <protogonoi at gmail.com> wrote:
> On 6 April 2014 00:47, Reimar Döffinger <Reimar.Doeffinger at gmx.de> wrote:
> The way it is currently (as of last patch) is 5.4% faster. However changing
> it to a uint8 (and using = as suggested) is actually only 0.97% faster.
This is confusing, I thought the speedup was from removing the lea?
Changing the type would be on top.
So you mean that after the first patch, changing the type causes a ca. 4% slowdown?
> The function could be changed to return a uint8 but on inspection only vp5
> will benefit from this (it has a lot of if(rac_get_prob) that can work
> directly on uint8)
That should be wrong. All the ifs should be using the _branchy variant which does not use the asm code at all.
More information about the ffmpeg-devel