Merge remote-tracking branch 'qatar/master'

* qatar/master:
  x86: h264_intrapred: Fix C function names in comments
  x86: SPLATD: port to cpuflags

Merged-by: Michael Niedermayer <michaelni@gmx.at>

5 files changed, 38 insertions, 53 deletions

diff --git a/libavcodec/x86/dsputil.asm b/libavcodec/x86/dsputil.asm
index 7966294a8d..c51043db20 100644
--- a/libavcodec/x86/dsputil.asm
+++ b/libavcodec/x86/dsputil.asm
@@ -1124,11 +1124,9 @@ cglobal vector_clip_int32%5, 5,5,%1, dst, src, min, max, len
 %endmacro
 
 INIT_MMX mmx
-%define SPLATD SPLATD_MMX
 %define CLIPD CLIPD_MMX
 VECTOR_CLIP_INT32 0, 1, 0, 0
 INIT_XMM sse2
-%define SPLATD SPLATD_SSE2
 VECTOR_CLIP_INT32 6, 1, 0, 0, _int
 %define CLIPD CLIPD_SSE2
 VECTOR_CLIP_INT32 6, 2, 0, 1
diff --git a/libavcodec/x86/fmtconvert.asm b/libavcodec/x86/fmtconvert.asm
index 4f2379b2e2..1bd13fc28a 100644
--- a/libavcodec/x86/fmtconvert.asm
+++ b/libavcodec/x86/fmtconvert.asm
@@ -72,12 +72,9 @@ cglobal int32_to_float_fmul_scalar, 4, 4, %1, dst, src, mul, len
 %endmacro
 
 INIT_XMM sse
-%define SPLATD SPLATD_SSE
 INT32_TO_FLOAT_FMUL_SCALAR 5
 INIT_XMM sse2
-%define SPLATD SPLATD_SSE2
 INT32_TO_FLOAT_FMUL_SCALAR 3
-%undef SPLATD
 
 
 ;------------------------------------------------------------------------------
diff --git a/libavcodec/x86/h264_intrapred.asm b/libavcodec/x86/h264_intrapred.asm
index 4fd8976429..5c0dff429a 100644
--- a/libavcodec/x86/h264_intrapred.asm
+++ b/libavcodec/x86/h264_intrapred.asm
@@ -49,7 +49,7 @@ cextern pw_17
 cextern pw_32
 
 ;-----------------------------------------------------------------------------
-; void pred16x16_vertical(uint8_t *src, int stride)
+; void pred16x16_vertical_8(uint8_t *src, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmx
@@ -85,7 +85,7 @@ cglobal pred16x16_vertical_8, 2,3
     REP_RET
 
 ;-----------------------------------------------------------------------------
-; void pred16x16_horizontal(uint8_t *src, int stride)
+; void pred16x16_horizontal_8(uint8_t *src, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro PRED16x16_H 0
@@ -126,7 +126,7 @@ INIT_XMM ssse3
 PRED16x16_H
 
 ;-----------------------------------------------------------------------------
-; void pred16x16_dc(uint8_t *src, int stride)
+; void pred16x16_dc_8(uint8_t *src, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro PRED16x16_DC 0
@@ -188,7 +188,7 @@ INIT_XMM ssse3
 PRED16x16_DC
 
 ;-----------------------------------------------------------------------------
-; void pred16x16_tm_vp8(uint8_t *src, int stride)
+; void pred16x16_tm_vp8_8(uint8_t *src, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro PRED16x16_TM 0
@@ -269,7 +269,7 @@ cglobal pred16x16_tm_vp8_8, 2,6,6
     REP_RET
 
 ;-----------------------------------------------------------------------------
-; void pred16x16_plane(uint8_t *src, int stride)
+; void pred16x16_plane_*_8(uint8_t *src, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro H264_PRED16x16_PLANE 1
@@ -550,7 +550,7 @@ H264_PRED16x16_PLANE rv40
 H264_PRED16x16_PLANE svq3
 
 ;-----------------------------------------------------------------------------
-; void pred8x8_plane(uint8_t *src, int stride)
+; void pred8x8_plane_8(uint8_t *src, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro H264_PRED8x8_PLANE 0
@@ -724,7 +724,7 @@ INIT_XMM ssse3
 H264_PRED8x8_PLANE
 
 ;-----------------------------------------------------------------------------
-; void pred8x8_vertical(uint8_t *src, int stride)
+; void pred8x8_vertical_8(uint8_t *src, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmx
@@ -741,7 +741,7 @@ cglobal pred8x8_vertical_8, 2,2
     RET
 
 ;-----------------------------------------------------------------------------
-; void pred8x8_horizontal(uint8_t *src, int stride)
+; void pred8x8_horizontal_8(uint8_t *src, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro PRED8x8_H 0
@@ -769,7 +769,7 @@ INIT_MMX ssse3
 PRED8x8_H
 
 ;-----------------------------------------------------------------------------
-; void pred8x8_top_dc_mmxext(uint8_t *src, int stride)
+; void pred8x8_top_dc_8_mmxext(uint8_t *src, int stride)
 ;-----------------------------------------------------------------------------
 INIT_MMX mmxext
 cglobal pred8x8_top_dc_8, 2,5
@@ -803,7 +803,7 @@ cglobal pred8x8_top_dc_8, 2,5
     RET
 
 ;-----------------------------------------------------------------------------
-; void pred8x8_dc_mmxext(uint8_t *src, int stride)
+; void pred8x8_dc_8_mmxext(uint8_t *src, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
@@ -864,7 +864,7 @@ cglobal pred8x8_dc_8, 2,5
     RET
 
 ;-----------------------------------------------------------------------------
-; void pred8x8_dc_rv40(uint8_t *src, int stride)
+; void pred8x8_dc_rv40_8(uint8_t *src, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
@@ -901,7 +901,7 @@ cglobal pred8x8_dc_rv40_8, 2,7
     REP_RET
 
 ;-----------------------------------------------------------------------------
-; void pred8x8_tm_vp8(uint8_t *src, int stride)
+; void pred8x8_tm_vp8_8(uint8_t *src, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro PRED8x8_TM 0
@@ -1013,7 +1013,7 @@ cglobal pred8x8_tm_vp8_8, 2,3,6
 %endmacro
 
 ;-----------------------------------------------------------------------------
-; void pred8x8l_top_dc(uint8_t *src, int has_topleft, int has_topright, int stride)
+; void pred8x8l_top_dc_8(uint8_t *src, int has_topleft, int has_topright, int stride)
 ;-----------------------------------------------------------------------------
 %macro PRED8x8L_TOP_DC 0
 cglobal pred8x8l_top_dc_8, 4,4
@@ -1068,7 +1068,7 @@ INIT_MMX ssse3
 PRED8x8L_TOP_DC
 
 ;-----------------------------------------------------------------------------
-;void pred8x8l_dc(uint8_t *src, int has_topleft, int has_topright, int stride)
+;void pred8x8l_dc_8(uint8_t *src, int has_topleft, int has_topright, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro PRED8x8L_DC 0
@@ -1171,7 +1171,7 @@ INIT_MMX ssse3
 PRED8x8L_DC
 
 ;-----------------------------------------------------------------------------
-; void pred8x8l_horizontal(uint8_t *src, int has_topleft, int has_topright, int stride)
+; void pred8x8l_horizontal_8(uint8_t *src, int has_topleft, int has_topright, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro PRED8x8L_HORIZONTAL 0
@@ -1242,7 +1242,7 @@ INIT_MMX ssse3
 PRED8x8L_HORIZONTAL
 
 ;-----------------------------------------------------------------------------
-; void pred8x8l_vertical(uint8_t *src, int has_topleft, int has_topright, int stride)
+; void pred8x8l_vertical_8(uint8_t *src, int has_topleft, int has_topright, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro PRED8x8L_VERTICAL 0
@@ -1292,7 +1292,7 @@ INIT_MMX ssse3
 PRED8x8L_VERTICAL
 
 ;-----------------------------------------------------------------------------
-;void pred8x8l_down_left(uint8_t *src, int has_topleft, int has_topright, int stride)
+;void pred8x8l_down_left_8(uint8_t *src, int has_topleft, int has_topright, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
@@ -1492,7 +1492,7 @@ INIT_MMX ssse3
 PRED8x8L_DOWN_LEFT
 
 ;-----------------------------------------------------------------------------
-;void pred8x8l_down_right_mmxext(uint8_t *src, int has_topleft, int has_topright, int stride)
+;void pred8x8l_down_right_8_mmxext(uint8_t *src, int has_topleft, int has_topright, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
@@ -1743,7 +1743,7 @@ INIT_MMX ssse3
 PRED8x8L_DOWN_RIGHT
 
 ;-----------------------------------------------------------------------------
-; void pred8x8l_vertical_right(uint8_t *src, int has_topleft, int has_topright, int stride)
+; void pred8x8l_vertical_right_8(uint8_t *src, int has_topleft, int has_topright, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
@@ -1970,7 +1970,7 @@ INIT_MMX ssse3
 PRED8x8L_VERTICAL_RIGHT
 
 ;-----------------------------------------------------------------------------
-;void pred8x8l_vertical_left(uint8_t *src, int has_topleft, int has_topright, int stride)
+;void pred8x8l_vertical_left_8(uint8_t *src, int has_topleft, int has_topright, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro PRED8x8L_VERTICAL_LEFT 0
@@ -2059,7 +2059,7 @@ INIT_MMX ssse3
 PRED8x8L_VERTICAL_LEFT
 
 ;-----------------------------------------------------------------------------
-; void pred8x8l_horizontal_up(uint8_t *src, int has_topleft, int has_topright, int stride)
+; void pred8x8l_horizontal_up_8(uint8_t *src, int has_topleft, int has_topright, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro PRED8x8L_HORIZONTAL_UP 0
@@ -2146,7 +2146,7 @@ INIT_MMX ssse3
 PRED8x8L_HORIZONTAL_UP
 
 ;-----------------------------------------------------------------------------
-;void pred8x8l_horizontal_down(uint8_t *src, int has_topleft, int has_topright, int stride)
+;void pred8x8l_horizontal_down_8(uint8_t *src, int has_topleft, int has_topright, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
@@ -2394,7 +2394,7 @@ INIT_MMX ssse3
 PRED8x8L_HORIZONTAL_DOWN
 
 ;-----------------------------------------------------------------------------
-; void pred4x4_dc_mmxext(uint8_t *src, const uint8_t *topright, int stride)
+; void pred4x4_dc_8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
@@ -2424,7 +2424,7 @@ cglobal pred4x4_dc_8, 3,5
     RET
 
 ;-----------------------------------------------------------------------------
-; void pred4x4_tm_vp8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
+; void pred4x4_tm_vp8_8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
 ;-----------------------------------------------------------------------------
 
 %macro PRED4x4_TM 0
@@ -2505,7 +2505,7 @@ cglobal pred4x4_tm_vp8_8, 3,3
     RET
 
 ;-----------------------------------------------------------------------------
-; void pred4x4_vertical_vp8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
+; void pred4x4_vertical_vp8_8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
@@ -2525,7 +2525,7 @@ cglobal pred4x4_vertical_vp8_8, 3,3
     RET
 
 ;-----------------------------------------------------------------------------
-; void pred4x4_down_left_mmxext(uint8_t *src, const uint8_t *topright, int stride)
+; void pred4x4_down_left_8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
 ;-----------------------------------------------------------------------------
 INIT_MMX mmxext
 cglobal pred4x4_down_left_8, 3,3
@@ -2551,7 +2551,7 @@ cglobal pred4x4_down_left_8, 3,3
     RET
 
 ;-----------------------------------------------------------------------------
-; void pred4x4_vertical_left_mmxext(uint8_t *src, const uint8_t *topright, int stride)
+; void pred4x4_vertical_left_8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
@@ -2576,7 +2576,7 @@ cglobal pred4x4_vertical_left_8, 3,3
     RET
 
 ;-----------------------------------------------------------------------------
-; void pred4x4_horizontal_up_mmxext(uint8_t *src, const uint8_t *topright, int stride)
+; void pred4x4_horizontal_up_8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
@@ -2609,7 +2609,7 @@ cglobal pred4x4_horizontal_up_8, 3,3
     RET
 
 ;-----------------------------------------------------------------------------
-; void pred4x4_horizontal_down_mmxext(uint8_t *src, const uint8_t *topright, int stride)
+; void pred4x4_horizontal_down_8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
@@ -2644,7 +2644,7 @@ cglobal pred4x4_horizontal_down_8, 3,3
     RET
 
 ;-----------------------------------------------------------------------------
-; void pred4x4_vertical_right_mmxext(uint8_t *src, const uint8_t *topright, int stride)
+; void pred4x4_vertical_right_8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
@@ -2674,7 +2674,7 @@ cglobal pred4x4_vertical_right_8, 3,3
     RET
 
 ;-----------------------------------------------------------------------------
-; void pred4x4_down_right_mmxext(uint8_t *src, const uint8_t *topright, int stride)
+; void pred4x4_down_right_8_mmxext(uint8_t *src, const uint8_t *topright, int stride)
 ;-----------------------------------------------------------------------------
 
 INIT_MMX mmxext
diff --git a/libavutil/x86/x86util.asm b/libavutil/x86/x86util.asm
index dfc180e981..c11df90386 100644
--- a/libavutil/x86/x86util.asm
+++ b/libavutil/x86/x86util.asm
@@ -574,24 +574,14 @@
 %endif
 %endmacro
 
-%macro SPLATD 2-3 0
-%if mmsize == 16
-    pshufd %1, %2, (%3)*0x55
-%else
-    pshufw %1, %2, (%3)*0x11 + ((%3)+1)*0x44
-%endif
-%endmacro
-
-%macro SPLATD_MMX 1
+%macro SPLATD 1
+%if mmsize == 8
     punpckldq  %1, %1
-%endmacro
-
-%macro SPLATD_SSE 1
-    shufps  %1, %1, 0
-%endmacro
-
-%macro SPLATD_SSE2 1
+%elif cpuflag(sse2)
     pshufd  %1, %1, 0
+%elif cpuflag(sse)
+    shufps  %1, %1, 0
+%endif
 %endmacro
 
 %macro CLIPW 3 ;(dst, min, max)
diff --git a/libswscale/x86/output.asm b/libswscale/x86/output.asm
index d57af9b1a9..f9add355d1 100644
--- a/libswscale/x86/output.asm
+++ b/libswscale/x86/output.asm
@@ -187,7 +187,7 @@ cglobal yuv2planeX_%1, %3, 8, %2, filter, fltsize, src, dst, w, dither, offset
 %else ; %1 == 10/9/8
     punpcklwd       m5,  m3,  m4
     punpckhwd       m3,  m4
-    SPLATD          m0,  m0
+    SPLATD          m0
 
     pmaddwd         m5,  m0
     pmaddwd         m3,  m0