summaryrefslogtreecommitdiff
path: root/libavcodec/h264pred_template.c
diff options
context:
space:
mode:
Diffstat (limited to 'libavcodec/h264pred_template.c')
-rw-r--r--libavcodec/h264pred_template.c97
1 files changed, 49 insertions, 48 deletions
diff --git a/libavcodec/h264pred_template.c b/libavcodec/h264pred_template.c
index e15c76c849..8d8d62e0b6 100644
--- a/libavcodec/h264pred_template.c
+++ b/libavcodec/h264pred_template.c
@@ -2,20 +2,20 @@
* H.26L/H.264/AVC/JVT/14496-10/... encoder/decoder
* Copyright (c) 2003-2011 Michael Niedermayer <michaelni@gmx.at>
*
- * This file is part of Libav.
+ * This file is part of FFmpeg.
*
- * Libav is free software; you can redistribute it and/or
+ * FFmpeg is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
- * Libav is distributed in the hope that it will be useful,
+ * FFmpeg is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
- * License along with Libav; if not, write to the Free Software
+ * License along with FFmpeg; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
@@ -33,7 +33,7 @@ static void FUNCC(pred4x4_vertical)(uint8_t *_src, const uint8_t *topright,
ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const pixel4 a= AV_RN4PA(src-stride);
AV_WN4PA(src+0*stride, a);
@@ -46,7 +46,7 @@ static void FUNCC(pred4x4_horizontal)(uint8_t *_src, const uint8_t *topright,
ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
AV_WN4PA(src+0*stride, PIXEL_SPLAT_X4(src[-1+0*stride]));
AV_WN4PA(src+1*stride, PIXEL_SPLAT_X4(src[-1+1*stride]));
AV_WN4PA(src+2*stride, PIXEL_SPLAT_X4(src[-1+2*stride]));
@@ -57,7 +57,7 @@ static void FUNCC(pred4x4_dc)(uint8_t *_src, const uint8_t *topright,
ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const int dc= ( src[-stride] + src[1-stride] + src[2-stride] + src[3-stride]
+ src[-1+0*stride] + src[-1+1*stride] + src[-1+2*stride] + src[-1+3*stride] + 4) >>3;
const pixel4 a = PIXEL_SPLAT_X4(dc);
@@ -72,7 +72,7 @@ static void FUNCC(pred4x4_left_dc)(uint8_t *_src, const uint8_t *topright,
ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const int dc= ( src[-1+0*stride] + src[-1+1*stride] + src[-1+2*stride] + src[-1+3*stride] + 2) >>2;
const pixel4 a = PIXEL_SPLAT_X4(dc);
@@ -86,7 +86,7 @@ static void FUNCC(pred4x4_top_dc)(uint8_t *_src, const uint8_t *topright,
ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const int dc= ( src[-stride] + src[1-stride] + src[2-stride] + src[3-stride] + 2) >>2;
const pixel4 a = PIXEL_SPLAT_X4(dc);
@@ -100,7 +100,7 @@ static void FUNCC(pred4x4_128_dc)(uint8_t *_src, const uint8_t *topright,
ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const pixel4 a = PIXEL_SPLAT_X4(1<<(BIT_DEPTH-1));
AV_WN4PA(src+0*stride, a);
@@ -113,7 +113,7 @@ static void FUNCC(pred4x4_127_dc)(uint8_t *_src, const uint8_t *topright,
ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const pixel4 a = PIXEL_SPLAT_X4((1<<(BIT_DEPTH-1))-1);
AV_WN4PA(src+0*stride, a);
@@ -126,7 +126,7 @@ static void FUNCC(pred4x4_129_dc)(uint8_t *_src, const uint8_t *topright,
ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const pixel4 a = PIXEL_SPLAT_X4((1<<(BIT_DEPTH-1))+1);
AV_WN4PA(src+0*stride, a);
@@ -164,7 +164,7 @@ static void FUNCC(pred4x4_down_right)(uint8_t *_src, const uint8_t *topright,
ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const int lt= src[-1-1*stride];
LOAD_TOP_EDGE
LOAD_LEFT_EDGE
@@ -192,7 +192,7 @@ static void FUNCC(pred4x4_down_left)(uint8_t *_src, const uint8_t *_topright,
{
pixel *src = (pixel*)_src;
const pixel *topright = (const pixel*)_topright;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
LOAD_TOP_EDGE
LOAD_TOP_RIGHT_EDGE
// LOAD_LEFT_EDGE
@@ -220,7 +220,7 @@ static void FUNCC(pred4x4_vertical_right)(uint8_t *_src,
ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const int lt= src[-1-1*stride];
LOAD_TOP_EDGE
LOAD_LEFT_EDGE
@@ -249,7 +249,7 @@ static void FUNCC(pred4x4_vertical_left)(uint8_t *_src,
{
pixel *src = (pixel*)_src;
const pixel *topright = (const pixel*)_topright;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
LOAD_TOP_EDGE
LOAD_TOP_RIGHT_EDGE
@@ -275,7 +275,7 @@ static void FUNCC(pred4x4_horizontal_up)(uint8_t *_src, const uint8_t *topright,
ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
LOAD_LEFT_EDGE
src[0+0*stride]=(l0 + l1 + 1)>>1;
@@ -301,7 +301,7 @@ static void FUNCC(pred4x4_horizontal_down)(uint8_t *_src,
ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const int lt= src[-1-1*stride];
LOAD_TOP_EDGE
LOAD_LEFT_EDGE
@@ -328,7 +328,7 @@ static void FUNCC(pred16x16_vertical)(uint8_t *_src, ptrdiff_t _stride)
{
int i;
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const pixel4 a = AV_RN4PA(((pixel4*)(src-stride))+0);
const pixel4 b = AV_RN4PA(((pixel4*)(src-stride))+1);
const pixel4 c = AV_RN4PA(((pixel4*)(src-stride))+2);
@@ -346,7 +346,7 @@ static void FUNCC(pred16x16_horizontal)(uint8_t *_src, ptrdiff_t stride)
{
int i;
pixel *src = (pixel*)_src;
- stride /= sizeof(pixel);
+ stride >>= sizeof(pixel)-1;
for(i=0; i<16; i++){
const pixel4 a = PIXEL_SPLAT_X4(src[-1+i*stride]);
@@ -372,7 +372,7 @@ static void FUNCC(pred16x16_dc)(uint8_t *_src, ptrdiff_t stride)
int i, dc=0;
pixel *src = (pixel*)_src;
pixel4 dcsplat;
- stride /= sizeof(pixel);
+ stride >>= sizeof(pixel)-1;
for(i=0;i<16; i++){
dc+= src[-1+i*stride];
@@ -391,7 +391,7 @@ static void FUNCC(pred16x16_left_dc)(uint8_t *_src, ptrdiff_t stride)
int i, dc=0;
pixel *src = (pixel*)_src;
pixel4 dcsplat;
- stride /= sizeof(pixel);
+ stride >>= sizeof(pixel)-1;
for(i=0;i<16; i++){
dc+= src[-1+i*stride];
@@ -406,7 +406,7 @@ static void FUNCC(pred16x16_top_dc)(uint8_t *_src, ptrdiff_t stride)
int i, dc=0;
pixel *src = (pixel*)_src;
pixel4 dcsplat;
- stride /= sizeof(pixel);
+ stride >>= sizeof(pixel)-1;
for(i=0;i<16; i++){
dc+= src[i-stride];
@@ -421,7 +421,7 @@ static void FUNCC(pred16x16_##n##_dc)(uint8_t *_src, ptrdiff_t stride)\
{\
int i;\
pixel *src = (pixel*)_src;\
- stride /= sizeof(pixel);\
+ stride >>= sizeof(pixel)-1;\
PREDICT_16x16_DC(PIXEL_SPLAT_X4(v));\
}
@@ -438,7 +438,7 @@ static inline void FUNCC(pred16x16_plane_compat)(uint8_t *_src,
int a;
INIT_CLIP
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const pixel * const src0 = src +7-stride;
const pixel * src1 = src +8*stride-1;
const pixel * src2 = src1-2*stride; // == src+6*stride-1;
@@ -487,7 +487,7 @@ static void FUNCC(pred8x8_vertical)(uint8_t *_src, ptrdiff_t _stride)
{
int i;
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const pixel4 a= AV_RN4PA(((pixel4*)(src-stride))+0);
const pixel4 b= AV_RN4PA(((pixel4*)(src-stride))+1);
@@ -515,7 +515,7 @@ static void FUNCC(pred8x8_horizontal)(uint8_t *_src, ptrdiff_t stride)
{
int i;
pixel *src = (pixel*)_src;
- stride /= sizeof(pixel);
+ stride >>= sizeof(pixel)-1;
for(i=0; i<8; i++){
const pixel4 a = PIXEL_SPLAT_X4(src[-1+i*stride]);
@@ -542,7 +542,7 @@ static void FUNCC(pred8x8_##n##_dc)(uint8_t *_src, ptrdiff_t stride)\
int i;\
const pixel4 a = PIXEL_SPLAT_X4(v);\
pixel *src = (pixel*)_src;\
- stride /= sizeof(pixel);\
+ stride >>= sizeof(pixel)-1;\
for(i=0; i<8; i++){\
AV_WN4PA(((pixel4*)(src+i*stride))+0, a);\
AV_WN4PA(((pixel4*)(src+i*stride))+1, a);\
@@ -565,7 +565,7 @@ static void FUNCC(pred8x8_left_dc)(uint8_t *_src, ptrdiff_t stride)
int dc0, dc2;
pixel4 dc0splat, dc2splat;
pixel *src = (pixel*)_src;
- stride /= sizeof(pixel);
+ stride >>= sizeof(pixel)-1;
dc0=dc2=0;
for(i=0;i<4; i++){
@@ -597,7 +597,7 @@ static void FUNCC(pred8x8_top_dc)(uint8_t *_src, ptrdiff_t stride)
int dc0, dc1;
pixel4 dc0splat, dc1splat;
pixel *src = (pixel*)_src;
- stride /= sizeof(pixel);
+ stride >>= sizeof(pixel)-1;
dc0=dc1=0;
for(i=0;i<4; i++){
@@ -645,7 +645,7 @@ static void FUNCC(pred8x8_dc)(uint8_t *_src, ptrdiff_t stride)
int dc0, dc1, dc2;
pixel4 dc0splat, dc1splat, dc2splat, dc3splat;
pixel *src = (pixel*)_src;
- stride /= sizeof(pixel);
+ stride >>= sizeof(pixel)-1;
dc0=dc1=dc2=0;
for(i=0;i<4; i++){
@@ -711,6 +711,7 @@ static void FUNCC(pred8x16_dc)(uint8_t *_src, ptrdiff_t stride)
}
}
+//the following 4 function should not be optimized!
static void FUNC(pred8x8_mad_cow_dc_l0t)(uint8_t *src, ptrdiff_t stride)
{
FUNCC(pred8x8_top_dc)(src, stride);
@@ -769,7 +770,7 @@ static void FUNCC(pred8x8_plane)(uint8_t *_src, ptrdiff_t _stride)
int a;
INIT_CLIP
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
const pixel * const src0 = src +3-stride;
const pixel * src1 = src +4*stride-1;
const pixel * src2 = src1-2*stride; // == src+2*stride-1;
@@ -883,7 +884,7 @@ static void FUNCC(pred8x8l_128_dc)(uint8_t *_src, int has_topleft,
int has_topright, ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
PREDICT_8x8_DC(PIXEL_SPLAT_X4(1<<(BIT_DEPTH-1)));
}
@@ -891,7 +892,7 @@ static void FUNCC(pred8x8l_left_dc)(uint8_t *_src, int has_topleft,
int has_topright, ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
PREDICT_8x8_LOAD_LEFT;
const pixel4 dc = PIXEL_SPLAT_X4((l0+l1+l2+l3+l4+l5+l6+l7+4) >> 3);
@@ -901,7 +902,7 @@ static void FUNCC(pred8x8l_top_dc)(uint8_t *_src, int has_topleft,
int has_topright, ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
PREDICT_8x8_LOAD_TOP;
const pixel4 dc = PIXEL_SPLAT_X4((t0+t1+t2+t3+t4+t5+t6+t7+4) >> 3);
@@ -911,7 +912,7 @@ static void FUNCC(pred8x8l_dc)(uint8_t *_src, int has_topleft,
int has_topright, ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
PREDICT_8x8_LOAD_LEFT;
PREDICT_8x8_LOAD_TOP;
@@ -923,7 +924,7 @@ static void FUNCC(pred8x8l_horizontal)(uint8_t *_src, int has_topleft,
int has_topright, ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
pixel4 a;
PREDICT_8x8_LOAD_LEFT;
@@ -938,7 +939,7 @@ static void FUNCC(pred8x8l_vertical)(uint8_t *_src, int has_topleft,
{
int y;
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
pixel4 a, b;
PREDICT_8x8_LOAD_TOP;
@@ -961,7 +962,7 @@ static void FUNCC(pred8x8l_down_left)(uint8_t *_src, int has_topleft,
int has_topright, ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
PREDICT_8x8_LOAD_TOP;
PREDICT_8x8_LOAD_TOPRIGHT;
SRC(0,0)= (t0 + 2*t1 + t2 + 2) >> 2;
@@ -984,7 +985,7 @@ static void FUNCC(pred8x8l_down_right)(uint8_t *_src, int has_topleft,
int has_topright, ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
PREDICT_8x8_LOAD_TOP;
PREDICT_8x8_LOAD_LEFT;
PREDICT_8x8_LOAD_TOPLEFT;
@@ -1008,7 +1009,7 @@ static void FUNCC(pred8x8l_vertical_right)(uint8_t *_src, int has_topleft,
int has_topright, ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
PREDICT_8x8_LOAD_TOP;
PREDICT_8x8_LOAD_LEFT;
PREDICT_8x8_LOAD_TOPLEFT;
@@ -1039,7 +1040,7 @@ static void FUNCC(pred8x8l_horizontal_down)(uint8_t *_src, int has_topleft,
int has_topright, ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
PREDICT_8x8_LOAD_TOP;
PREDICT_8x8_LOAD_LEFT;
PREDICT_8x8_LOAD_TOPLEFT;
@@ -1070,7 +1071,7 @@ static void FUNCC(pred8x8l_vertical_left)(uint8_t *_src, int has_topleft,
int has_topright, ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
PREDICT_8x8_LOAD_TOP;
PREDICT_8x8_LOAD_TOPRIGHT;
SRC(0,0)= (t0 + t1 + 1) >> 1;
@@ -1100,7 +1101,7 @@ static void FUNCC(pred8x8l_horizontal_up)(uint8_t *_src, int has_topleft,
int has_topright, ptrdiff_t _stride)
{
pixel *src = (pixel*)_src;
- int stride = _stride/sizeof(pixel);
+ int stride = _stride>>(sizeof(pixel)-1);
PREDICT_8x8_LOAD_LEFT;
SRC(0,0)= (l0 + l1 + 1) >> 1;
SRC(1,0)= (l0 + 2*l1 + l2 + 2) >> 2;
@@ -1137,7 +1138,7 @@ static void FUNCC(pred4x4_vertical_add)(uint8_t *_pix, int16_t *_block,
int i;
pixel *pix = (pixel*)_pix;
const dctcoef *block = (const dctcoef*)_block;
- stride /= sizeof(pixel);
+ stride >>= sizeof(pixel)-1;
pix -= stride;
for(i=0; i<4; i++){
pixel v = pix[0];
@@ -1158,7 +1159,7 @@ static void FUNCC(pred4x4_horizontal_add)(uint8_t *_pix, int16_t *_block,
int i;
pixel *pix = (pixel*)_pix;
const dctcoef *block = (const dctcoef*)_block;
- stride /= sizeof(pixel);
+ stride >>= sizeof(pixel)-1;
for(i=0; i<4; i++){
pixel v = pix[-1];
pix[0]= v += block[0];
@@ -1178,7 +1179,7 @@ static void FUNCC(pred8x8l_vertical_add)(uint8_t *_pix, int16_t *_block,
int i;
pixel *pix = (pixel*)_pix;
const dctcoef *block = (const dctcoef*)_block;
- stride /= sizeof(pixel);
+ stride >>= sizeof(pixel)-1;
pix -= stride;
for(i=0; i<8; i++){
pixel v = pix[0];
@@ -1203,7 +1204,7 @@ static void FUNCC(pred8x8l_horizontal_add)(uint8_t *_pix, int16_t *_block,
int i;
pixel *pix = (pixel*)_pix;
const dctcoef *block = (const dctcoef*)_block;
- stride /= sizeof(pixel);
+ stride >>= sizeof(pixel)-1;
for(i=0; i<8; i++){
pixel v = pix[-1];
pix[0]= v += block[0];
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-09 12:43:30 +0000