Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(2144)

Issues Search
    My Issues | Starred     Open | Closed | All

Unified Diff: source/libvpx/vp8/common/x86/loopfilter_mmx.asm

Issue 7671004: Update libvpx snapshot to v0.9.7-p1 (Cayuga). (Closed) Base URL: svn://chrome-svn/chrome/trunk/deps/third_party/libvpx/
Patch Set: '' Created 9 years, 4 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View side-by-side diff with in-line comments
Download patch
« no previous file with comments | « source/libvpx/vp8/common/x86/iwalsh_sse2.asm ('k') | source/libvpx/vp8/common/x86/loopfilter_sse2.asm » ('j') | no next file with comments »
Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
Index: source/libvpx/vp8/common/x86/loopfilter_mmx.asm
===================================================================
--- source/libvpx/vp8/common/x86/loopfilter_mmx.asm (revision 96967)
+++ source/libvpx/vp8/common/x86/loopfilter_mmx.asm (working copy)
@@ -16,7 +16,7 @@
;(
; unsigned char *src_ptr,
; int src_pixel_step,
-; const char *flimit,
+; const char *blimit,
; const char *limit,
; const char *thresh,
; int count
@@ -122,12 +122,10 @@
paddusb mm5, mm5 ; abs(p0-q0)*2
paddusb mm5, mm2 ; abs (p0 - q0) *2 + abs(p1-q1)/2
- mov rdx, arg(2) ;flimit ; get flimit
- movq mm2, [rdx] ; flimit mm2
- paddb mm2, mm2 ; flimit*2 (less than 255)
- paddb mm7, mm2 ; flimit * 2 + limit (less than 255)
+ mov rdx, arg(2) ;blimit ; get blimit
+ movq mm7, [rdx] ; blimit
- psubusb mm5, mm7 ; abs (p0 - q0) *2 + abs(p1-q1)/2 > flimit * 2 + limit
+ psubusb mm5, mm7 ; abs (p0 - q0) *2 + abs(p1-q1)/2 > blimit
por mm1, mm5
pxor mm5, mm5
pcmpeqb mm1, mm5 ; mask mm1
@@ -230,7 +228,7 @@
;(
; unsigned char *src_ptr,
; int src_pixel_step,
-; const char *flimit,
+; const char *blimit,
; const char *limit,
; const char *thresh,
; int count
@@ -406,9 +404,9 @@
pand mm5, [GLOBAL(tfe)] ; set lsb of each byte to zero
psrlw mm5, 1 ; abs(p1-q1)/2
- mov rdx, arg(2) ;flimit ;
+ mov rdx, arg(2) ;blimit ;
- movq mm2, [rdx] ;flimit mm2
+ movq mm4, [rdx] ;blimit
movq mm1, mm3 ; mm1=mm3=p0
movq mm7, mm6 ; mm7=mm6=q0
@@ -419,10 +417,7 @@
paddusb mm1, mm1 ; abs(q0-p0)*2
paddusb mm1, mm5 ; abs (p0 - q0) *2 + abs(p1-q1)/2
- paddb mm2, mm2 ; flimit*2 (less than 255)
- paddb mm4, mm2 ; flimit * 2 + limit (less than 255)
-
- psubusb mm1, mm4 ; abs (p0 - q0) *2 + abs(p1-q1)/2 > flimit * 2 + limit
+ psubusb mm1, mm4 ; abs (p0 - q0) *2 + abs(p1-q1)/2 > blimit
por mm1, mm0; ; mask
pxor mm0, mm0
@@ -603,7 +598,7 @@
;(
; unsigned char *src_ptr,
; int src_pixel_step,
-; const char *flimit,
+; const char *blimit,
; const char *limit,
; const char *thresh,
; int count
@@ -719,17 +714,15 @@
paddusb mm5, mm5 ; abs(p0-q0)*2
paddusb mm5, mm2 ; abs (p0 - q0) *2 + abs(p1-q1)/2
- mov rdx, arg(2) ;flimit ; get flimit
- movq mm2, [rdx] ; flimit mm2
- paddb mm2, mm2 ; flimit*2 (less than 255)
- paddb mm7, mm2 ; flimit * 2 + limit (less than 255)
+ mov rdx, arg(2) ;blimit ; get blimit
+ movq mm7, [rdx] ; blimit
- psubusb mm5, mm7 ; abs (p0 - q0) *2 + abs(p1-q1)/2 > flimit * 2 + limit
+ psubusb mm5, mm7 ; abs (p0 - q0) *2 + abs(p1-q1)/2 > blimit
por mm1, mm5
pxor mm5, mm5
pcmpeqb mm1, mm5 ; mask mm1
- ; mm1 = mask, mm0=q0, mm7 = flimit, t0 = abs(q0-q1) t1 = abs(p1-p0)
+ ; mm1 = mask, mm0=q0, mm7 = blimit, t0 = abs(q0-q1) t1 = abs(p1-p0)
; mm6 = p0,
; calculate high edge variance
@@ -922,7 +915,7 @@
;(
; unsigned char *src_ptr,
; int src_pixel_step,
-; const char *flimit,
+; const char *blimit,
; const char *limit,
; const char *thresh,
; int count
@@ -1108,9 +1101,9 @@
pand mm5, [GLOBAL(tfe)] ; set lsb of each byte to zero
psrlw mm5, 1 ; abs(p1-q1)/2
- mov rdx, arg(2) ;flimit ;
+ mov rdx, arg(2) ;blimit ;
- movq mm2, [rdx] ;flimit mm2
+ movq mm4, [rdx] ;blimit
movq mm1, mm3 ; mm1=mm3=p0
movq mm7, mm6 ; mm7=mm6=q0
@@ -1121,10 +1114,7 @@
paddusb mm1, mm1 ; abs(q0-p0)*2
paddusb mm1, mm5 ; abs (p0 - q0) *2 + abs(p1-q1)/2
- paddb mm2, mm2 ; flimit*2 (less than 255)
- paddb mm4, mm2 ; flimit * 2 + limit (less than 255)
-
- psubusb mm1, mm4 ; abs (p0 - q0) *2 + abs(p1-q1)/2 > flimit * 2 + limit
+ psubusb mm1, mm4 ; abs (p0 - q0) *2 + abs(p1-q1)/2 > blimit
por mm1, mm0; ; mask
pxor mm0, mm0
@@ -1392,16 +1382,13 @@
;(
; unsigned char *src_ptr,
; int src_pixel_step,
-; const char *flimit,
-; const char *limit,
-; const char *thresh,
-; int count
+; const char *blimit
;)
global sym(vp8_loop_filter_simple_horizontal_edge_mmx)
sym(vp8_loop_filter_simple_horizontal_edge_mmx):
push rbp
mov rbp, rsp
- SHADOW_ARGS_TO_STACK 6
+ SHADOW_ARGS_TO_STACK 3
GET_GOT rbx
push rsi
push rdi
@@ -1410,14 +1397,10 @@
mov rsi, arg(0) ;src_ptr
movsxd rax, dword ptr arg(1) ;src_pixel_step ; destination pitch?
- movsxd rcx, dword ptr arg(5) ;count
+ mov rcx, 2 ; count
nexts8_h:
- mov rdx, arg(3) ;limit
- movq mm7, [rdx]
- mov rdx, arg(2) ;flimit ; get flimit
+ mov rdx, arg(2) ;blimit ; get blimit
movq mm3, [rdx] ;
- paddb mm3, mm3 ; flimit*2 (less than 255)
- paddb mm3, mm7 ; flimit * 2 + limit (less than 255)
mov rdi, rsi ; rdi points to row +1 for indirect addressing
add rdi, rax
@@ -1445,7 +1428,7 @@
paddusb mm5, mm5 ; abs(p0-q0)*2
paddusb mm5, mm1 ; abs (p0 - q0) *2 + abs(p1-q1)/2
- psubusb mm5, mm3 ; abs(p0 - q0) *2 + abs(p1-q1)/2 > flimit * 2 + limit
+ psubusb mm5, mm3 ; abs(p0 - q0) *2 + abs(p1-q1)/2 > blimit
pxor mm3, mm3
pcmpeqb mm5, mm3
@@ -1515,16 +1498,13 @@
;(
; unsigned char *src_ptr,
; int src_pixel_step,
-; const char *flimit,
-; const char *limit,
-; const char *thresh,
-; int count
+; const char *blimit
;)
global sym(vp8_loop_filter_simple_vertical_edge_mmx)
sym(vp8_loop_filter_simple_vertical_edge_mmx):
push rbp
mov rbp, rsp
- SHADOW_ARGS_TO_STACK 6
+ SHADOW_ARGS_TO_STACK 3
GET_GOT rbx
push rsi
push rdi
@@ -1539,7 +1519,7 @@
movsxd rax, dword ptr arg(1) ;src_pixel_step ; destination pitch?
lea rsi, [rsi + rax*4- 2]; ;
- movsxd rcx, dword ptr arg(5) ;count
+ mov rcx, 2 ; count
nexts8_v:
lea rdi, [rsi + rax];
@@ -1602,14 +1582,10 @@
paddusb mm5, mm5 ; abs(p0-q0)*2
paddusb mm5, mm6 ; abs (p0 - q0) *2 + abs(p1-q1)/2
- mov rdx, arg(2) ;flimit ; get flimit
+ mov rdx, arg(2) ;blimit ; get blimit
movq mm7, [rdx]
- mov rdx, arg(3) ; get limit
- movq mm6, [rdx]
- paddb mm7, mm7 ; flimit*2 (less than 255)
- paddb mm7, mm6 ; flimit * 2 + limit (less than 255)
- psubusb mm5, mm7 ; abs(p0 - q0) *2 + abs(p1-q1)/2 > flimit * 2 + limit
+ psubusb mm5, mm7 ; abs(p0 - q0) *2 + abs(p1-q1)/2 > blimit
pxor mm7, mm7
pcmpeqb mm5, mm7 ; mm5 = mask
« no previous file with comments | « source/libvpx/vp8/common/x86/iwalsh_sse2.asm ('k') | source/libvpx/vp8/common/x86/loopfilter_sse2.asm » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698