[x264-devel] x264: svn commit r717 (pengvado)

[Jindrich Makovicka]

On Sun,  6 Jan 2008 09:15:05 +0100 (CET)
Subversion daemon <svn at videolan.org> wrote:

> r717 | pengvado | 2008-01-06 09:15:04 +0100 (Sun, 06 Jan 2008) | 5
> lines Changed paths:
>    M /trunk/common/amd64/pixel-a.asm
>    M /trunk/common/i386/pixel-a.asm
>    M /trunk/common/i386/pixel.h
>    M /trunk/common/pixel.c
>    M /trunk/common/pixel.h
>    M /trunk/encoder/analyse.c
>    M /trunk/encoder/me.c
>    M /trunk/encoder/me.h
>    M /trunk/tools/checkasm.c
> 
> convert absolute difference of sums from mmx to sse2
> convert mv bits cost and ads threshold from C to sse2
> convert bytemask-to-list from C to scalar asm
> 1.6x faster me=esa (x86_64) or 1.3x faster (x86_32). (times consider
> only motion estimation. overall encode speedup may vary.)

I needed to force a long jump in one place to make it build with
nasm:

Index: pixel-a.asm
===================================================================
--- pixel-a.asm	(revision 717)
+++ pixel-a.asm	(working copy)
@@ -1593,7 +1593,7 @@
     mov     ebp, [ecx+esi]
     mov     edx, [ecx+esi+4]
     or      edx, ebp
-    jz .nexti
+    jz near .nexti
     xor     edx, edx
 %macro TEST 1
     mov     [ebx+eax*2], si

-- 
Jindrich Makovicka