[x265] [PATCH 0 of 8 ] asm:intra_pred4 modes 2-9 sse2

[chen]

Yes, it is from our SSE4 code, we have't time to modify it, but in new patch, we can do it to reduce future jobs.
 
replace r3 and r4 to dword register name can reduce 1 byte in code

At 2015-03-24 06:12:44,dave <dtyx265 at gmail.com> wrote:

r4 and r3 are also not accounted for as arguments in cglobal declarations.  They all use 3.  This is taken from sse4 code.
On 03/23/2015 01:58 PM, chen wrote:


Patches are right.

when you use r4d and r3d, you may get less code size

At 2015-03-24 04:30:28,dtyx265 at gmail.com wrote:
>The code is backported from the sse4 primitives




_______________________________________________
x265-devel mailing list
x265-devel at videolan.orghttps://mailman.videolan.org/listinfo/x265-devel

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.videolan.org/pipermail/x265-devel/attachments/20150324/9ed83480/attachment-0001.html>