falkTX
/
FFmpeg
mirror of https://github.com/falkTX/FFmpeg.git
1
0
Fork 0
Code Issues 0 Releases 338 Wiki Activity
Browse Source

qpeldsp: Mark source pointer in qpel_mc_func function pointer const

tags/n2.4
Diego Biurrun 11 years ago
parent
060e4a9e06
commit
7fb993d338
15 changed files with 451 additions and 392 deletions
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. +64
    -64
      libavcodec/arm/h264qpel_init_arm.c
  2. +4
    -2
      libavcodec/arm/rv40dsp_init_arm.c
  3. +21
    -21
      libavcodec/cavsdsp.c
  4. +37
    -37
      libavcodec/h264qpel_template.c
  5. +16
    -16
      libavcodec/ppc/h264qpel.c
  6. +14
    -4
      libavcodec/ppc/h264qpel_template.c
  7. +52
    -50
      libavcodec/qpeldsp.c
  8. +16
    -12
      libavcodec/qpeldsp.h
  9. +20
    -20
      libavcodec/rv30dsp.c
  10. +20
    -20
      libavcodec/rv40dsp.c
  11. +9
    -9
      libavcodec/wmv2dsp.c
  12. +19
    -19
      libavcodec/x86/cavsdsp.c
  13. +61
    -61
      libavcodec/x86/h264_qpel.c
  14. +93
    -52
      libavcodec/x86/qpeldsp_init.c
  15. +5
    -5
      libavcodec/x86/rv40dsp_init.c

+ 64
- 64
libavcodec/arm/h264qpel_init_arm.c View File

@@ -26,73 +26,73 @@
#include "libavutil/arm/cpu.h"
#include "libavcodec/h264qpel.h"

void ff_put_h264_qpel16_mc00_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc10_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc20_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc30_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc01_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc11_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc21_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc31_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc02_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc12_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc22_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc32_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc03_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc13_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc23_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc33_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel16_mc00_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc10_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc20_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc30_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc01_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc11_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc21_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc31_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc02_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc12_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc22_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc32_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc03_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc13_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc23_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel16_mc33_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);

void ff_put_h264_qpel8_mc00_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc10_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc20_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc30_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc01_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc11_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc21_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc31_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc02_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc12_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc22_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc32_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc03_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc13_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc23_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc33_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_put_h264_qpel8_mc00_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc10_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc20_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc30_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc01_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc11_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc21_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc31_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc02_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc12_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc22_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc32_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc03_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc13_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc23_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc33_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);

void ff_avg_h264_qpel16_mc00_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc10_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc20_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc30_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc01_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc11_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc21_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc31_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc02_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc12_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc22_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc32_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc03_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc13_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc23_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc33_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel16_mc00_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc10_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc20_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc30_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc01_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc11_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc21_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc31_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc02_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc12_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc22_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc32_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc03_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc13_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc23_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc33_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);

void ff_avg_h264_qpel8_mc00_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc10_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc20_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc30_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc01_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc11_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc21_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc31_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc02_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc12_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc22_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc32_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc03_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc13_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc23_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc33_neon(uint8_t *, uint8_t *, ptrdiff_t);
void ff_avg_h264_qpel8_mc00_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc10_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc20_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc30_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc01_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc11_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc21_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc31_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc02_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc12_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc22_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc32_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc03_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc13_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc23_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc33_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);

av_cold void ff_h264qpel_init_arm(H264QpelContext *c, int bit_depth)
{


+ 4
- 2
libavcodec/arm/rv40dsp_init_arm.c View File

@@ -26,8 +26,10 @@
#include "libavutil/arm/cpu.h"

#define DECL_QPEL3(type, w, pos) \
void ff_##type##_rv40_qpel##w##_mc##pos##_neon(uint8_t *dst, uint8_t *src,\
ptrdiff_t stride)
void ff_ ## type ## _rv40_qpel ## w ## _mc ## pos ## _neon(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride)

#define DECL_QPEL2(w, pos) \
DECL_QPEL3(put, w, pos); \
DECL_QPEL3(avg, w, pos)


+ 21
- 21
libavcodec/cavsdsp.c View File

@@ -261,7 +261,7 @@ static void cavs_idct8_add_c(uint8_t *dst, int16_t *block, int stride) {
****************************************************************************/

#define CAVS_SUBPIX(OPNAME, OP, NAME, A, B, C, D, E, F) \
static void OPNAME ## cavs_filt8_h_ ## NAME(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## cavs_filt8_h_ ## NAME(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
const int h=8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
int i;\
@@ -280,7 +280,7 @@ static void OPNAME ## cavs_filt8_h_ ## NAME(uint8_t *dst, uint8_t *src, int dstS
}\
}\
\
static void OPNAME ## cavs_filt8_v_ ## NAME(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## cavs_filt8_v_ ## NAME(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
const int w=8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
int i;\
@@ -312,7 +312,7 @@ static void OPNAME ## cavs_filt8_v_ ## NAME(uint8_t *dst, uint8_t *src, int dst
}\
}\
\
static void OPNAME ## cavs_filt16_v_ ## NAME(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## cavs_filt16_v_ ## NAME(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## cavs_filt8_v_ ## NAME(dst , src , dstStride, srcStride);\
OPNAME ## cavs_filt8_v_ ## NAME(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\
@@ -321,7 +321,7 @@ static void OPNAME ## cavs_filt16_v_ ## NAME(uint8_t *dst, uint8_t *src, int dst
OPNAME ## cavs_filt8_v_ ## NAME(dst+8, src+8, dstStride, srcStride);\
}\
\
static void OPNAME ## cavs_filt16_h_ ## NAME(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## cavs_filt16_h_ ## NAME(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## cavs_filt8_h_ ## NAME(dst , src , dstStride, srcStride);\
OPNAME ## cavs_filt8_h_ ## NAME(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\
@@ -331,7 +331,7 @@ static void OPNAME ## cavs_filt16_h_ ## NAME(uint8_t *dst, uint8_t *src, int dst
}\

#define CAVS_SUBPIX_HV(OPNAME, OP, NAME, AH, BH, CH, DH, EH, FH, AV, BV, CV, DV, EV, FV, FULL) \
static void OPNAME ## cavs_filt8_hv_ ## NAME(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int srcStride){\
static void OPNAME ## cavs_filt8_hv_ ## NAME(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dstStride, int srcStride){\
int16_t temp[8*(8+5)];\
int16_t *tmp = temp;\
const int h=8;\
@@ -412,7 +412,7 @@ static void OPNAME ## cavs_filt8_hv_ ## NAME(uint8_t *dst, uint8_t *src1, uint8_
}\
}\
\
static void OPNAME ## cavs_filt16_hv_ ## NAME(uint8_t *dst, uint8_t *src1, uint8_t *src2, int dstStride, int srcStride){ \
static void OPNAME ## cavs_filt16_hv_ ## NAME(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int dstStride, int srcStride){ \
OPNAME ## cavs_filt8_hv_ ## NAME(dst , src1, src2 , dstStride, srcStride); \
OPNAME ## cavs_filt8_hv_ ## NAME(dst+8, src1+8, src2+8, dstStride, srcStride); \
src1 += 8*srcStride;\
@@ -423,77 +423,77 @@ static void OPNAME ## cavs_filt16_hv_ ## NAME(uint8_t *dst, uint8_t *src1, uint8
}\

#define CAVS_MC(OPNAME, SIZE) \
static void OPNAME ## cavs_qpel ## SIZE ## _mc10_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc10_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _h_qpel_l(dst, src, stride, stride);\
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc20_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc20_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _h_hpel(dst, src, stride, stride);\
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc30_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc30_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _h_qpel_r(dst, src, stride, stride);\
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc01_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc01_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _v_qpel_l(dst, src, stride, stride);\
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc02_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc02_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _v_hpel(dst, src, stride, stride);\
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc03_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc03_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _v_qpel_r(dst, src, stride, stride);\
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc22_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc22_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _hv_jj(dst, src, NULL, stride, stride); \
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc11_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc11_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _hv_egpr(dst, src, src, stride, stride); \
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc13_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc13_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _hv_egpr(dst, src, src+stride, stride, stride); \
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc31_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc31_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _hv_egpr(dst, src, src+1, stride, stride); \
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc33_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc33_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _hv_egpr(dst, src, src+stride+1,stride, stride); \
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc21_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc21_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _hv_ff(dst, src, src+stride+1,stride, stride); \
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc12_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc12_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _hv_ii(dst, src, src+stride+1,stride, stride); \
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc32_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc32_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _hv_kk(dst, src, src+stride+1,stride, stride); \
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc23_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc23_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_filt ## SIZE ## _hv_qq(dst, src, src+stride+1,stride, stride); \
}\


+ 37
- 37
libavcodec/h264qpel_template.c View File

@@ -75,12 +75,12 @@ static inline void FUNC(copy_block16)(uint8_t *dst, const uint8_t *src, int dstS
}

#define H264_LOWPASS(OPNAME, OP, OP2) \
static av_unused void FUNC(OPNAME ## h264_qpel2_h_lowpass)(uint8_t *_dst, uint8_t *_src, int dstStride, int srcStride){\
static av_unused void FUNC(OPNAME ## h264_qpel2_h_lowpass)(uint8_t *_dst, const uint8_t *_src, int dstStride, int srcStride){\
const int h=2;\
INIT_CLIP\
int i;\
pixel *dst = (pixel*)_dst;\
pixel *src = (pixel*)_src;\
const pixel *src = (const pixel*)_src;\
dstStride /= sizeof(pixel);\
srcStride /= sizeof(pixel);\
for(i=0; i<h; i++)\
@@ -92,12 +92,12 @@ static av_unused void FUNC(OPNAME ## h264_qpel2_h_lowpass)(uint8_t *_dst, uint8_
}\
}\
\
static av_unused void FUNC(OPNAME ## h264_qpel2_v_lowpass)(uint8_t *_dst, uint8_t *_src, int dstStride, int srcStride){\
static av_unused void FUNC(OPNAME ## h264_qpel2_v_lowpass)(uint8_t *_dst, const uint8_t *_src, int dstStride, int srcStride){\
const int w=2;\
INIT_CLIP\
int i;\
pixel *dst = (pixel*)_dst;\
pixel *src = (pixel*)_src;\
const pixel *src = (const pixel*)_src;\
dstStride /= sizeof(pixel);\
srcStride /= sizeof(pixel);\
for(i=0; i<w; i++)\
@@ -116,14 +116,14 @@ static av_unused void FUNC(OPNAME ## h264_qpel2_v_lowpass)(uint8_t *_dst, uint8_
}\
}\
\
static av_unused void FUNC(OPNAME ## h264_qpel2_hv_lowpass)(uint8_t *_dst, int16_t *tmp, uint8_t *_src, int dstStride, int tmpStride, int srcStride){\
static av_unused void FUNC(OPNAME ## h264_qpel2_hv_lowpass)(uint8_t *_dst, int16_t *tmp, const uint8_t *_src, int dstStride, int tmpStride, int srcStride){\
const int h=2;\
const int w=2;\
const int pad = (BIT_DEPTH > 9) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\
INIT_CLIP\
int i;\
pixel *dst = (pixel*)_dst;\
pixel *src = (pixel*)_src;\
const pixel *src = (const pixel*)_src;\
dstStride /= sizeof(pixel);\
srcStride /= sizeof(pixel);\
src -= 2*srcStride;\
@@ -150,12 +150,12 @@ static av_unused void FUNC(OPNAME ## h264_qpel2_hv_lowpass)(uint8_t *_dst, int16
tmp++;\
}\
}\
static void FUNC(OPNAME ## h264_qpel4_h_lowpass)(uint8_t *_dst, uint8_t *_src, int dstStride, int srcStride){\
static void FUNC(OPNAME ## h264_qpel4_h_lowpass)(uint8_t *_dst, const uint8_t *_src, int dstStride, int srcStride){\
const int h=4;\
INIT_CLIP\
int i;\
pixel *dst = (pixel*)_dst;\
pixel *src = (pixel*)_src;\
const pixel *src = (const pixel*)_src;\
dstStride /= sizeof(pixel);\
srcStride /= sizeof(pixel);\
for(i=0; i<h; i++)\
@@ -169,12 +169,12 @@ static void FUNC(OPNAME ## h264_qpel4_h_lowpass)(uint8_t *_dst, uint8_t *_src, i
}\
}\
\
static void FUNC(OPNAME ## h264_qpel4_v_lowpass)(uint8_t *_dst, uint8_t *_src, int dstStride, int srcStride){\
static void FUNC(OPNAME ## h264_qpel4_v_lowpass)(uint8_t *_dst, const uint8_t *_src, int dstStride, int srcStride){\
const int w=4;\
INIT_CLIP\
int i;\
pixel *dst = (pixel*)_dst;\
pixel *src = (pixel*)_src;\
const pixel *src = (const pixel*)_src;\
dstStride /= sizeof(pixel);\
srcStride /= sizeof(pixel);\
for(i=0; i<w; i++)\
@@ -197,14 +197,14 @@ static void FUNC(OPNAME ## h264_qpel4_v_lowpass)(uint8_t *_dst, uint8_t *_src, i
}\
}\
\
static void FUNC(OPNAME ## h264_qpel4_hv_lowpass)(uint8_t *_dst, int16_t *tmp, uint8_t *_src, int dstStride, int tmpStride, int srcStride){\
static void FUNC(OPNAME ## h264_qpel4_hv_lowpass)(uint8_t *_dst, int16_t *tmp, const uint8_t *_src, int dstStride, int tmpStride, int srcStride){\
const int h=4;\
const int w=4;\
const int pad = (BIT_DEPTH > 9) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\
INIT_CLIP\
int i;\
pixel *dst = (pixel*)_dst;\
pixel *src = (pixel*)_src;\
const pixel *src = (const pixel*)_src;\
dstStride /= sizeof(pixel);\
srcStride /= sizeof(pixel);\
src -= 2*srcStride;\
@@ -238,12 +238,12 @@ static void FUNC(OPNAME ## h264_qpel4_hv_lowpass)(uint8_t *_dst, int16_t *tmp, u
}\
}\
\
static void FUNC(OPNAME ## h264_qpel8_h_lowpass)(uint8_t *_dst, uint8_t *_src, int dstStride, int srcStride){\
static void FUNC(OPNAME ## h264_qpel8_h_lowpass)(uint8_t *_dst, const uint8_t *_src, int dstStride, int srcStride){\
const int h=8;\
INIT_CLIP\
int i;\
pixel *dst = (pixel*)_dst;\
pixel *src = (pixel*)_src;\
const pixel *src = (const pixel*)_src;\
dstStride /= sizeof(pixel);\
srcStride /= sizeof(pixel);\
for(i=0; i<h; i++)\
@@ -261,12 +261,12 @@ static void FUNC(OPNAME ## h264_qpel8_h_lowpass)(uint8_t *_dst, uint8_t *_src, i
}\
}\
\
static void FUNC(OPNAME ## h264_qpel8_v_lowpass)(uint8_t *_dst, uint8_t *_src, int dstStride, int srcStride){\
static void FUNC(OPNAME ## h264_qpel8_v_lowpass)(uint8_t *_dst, const uint8_t *_src, int dstStride, int srcStride){\
const int w=8;\
INIT_CLIP\
int i;\
pixel *dst = (pixel*)_dst;\
pixel *src = (pixel*)_src;\
const pixel *src = (const pixel*)_src;\
dstStride /= sizeof(pixel);\
srcStride /= sizeof(pixel);\
for(i=0; i<w; i++)\
@@ -297,14 +297,14 @@ static void FUNC(OPNAME ## h264_qpel8_v_lowpass)(uint8_t *_dst, uint8_t *_src, i
}\
}\
\
static void FUNC(OPNAME ## h264_qpel8_hv_lowpass)(uint8_t *_dst, int16_t *tmp, uint8_t *_src, int dstStride, int tmpStride, int srcStride){\
static void FUNC(OPNAME ## h264_qpel8_hv_lowpass)(uint8_t *_dst, int16_t *tmp, const uint8_t *_src, int dstStride, int tmpStride, int srcStride){\
const int h=8;\
const int w=8;\
const int pad = (BIT_DEPTH > 9) ? (-10 * ((1<<BIT_DEPTH)-1)) : 0;\
INIT_CLIP\
int i;\
pixel *dst = (pixel*)_dst;\
pixel *src = (pixel*)_src;\
const pixel *src = (const pixel*)_src;\
dstStride /= sizeof(pixel);\
srcStride /= sizeof(pixel);\
src -= 2*srcStride;\
@@ -350,7 +350,7 @@ static void FUNC(OPNAME ## h264_qpel8_hv_lowpass)(uint8_t *_dst, int16_t *tmp, u
}\
}\
\
static void FUNC(OPNAME ## h264_qpel16_v_lowpass)(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void FUNC(OPNAME ## h264_qpel16_v_lowpass)(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
FUNC(OPNAME ## h264_qpel8_v_lowpass)(dst , src , dstStride, srcStride);\
FUNC(OPNAME ## h264_qpel8_v_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\
src += 8*srcStride;\
@@ -359,7 +359,7 @@ static void FUNC(OPNAME ## h264_qpel16_v_lowpass)(uint8_t *dst, uint8_t *src, in
FUNC(OPNAME ## h264_qpel8_v_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\
}\
\
static void FUNC(OPNAME ## h264_qpel16_h_lowpass)(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void FUNC(OPNAME ## h264_qpel16_h_lowpass)(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
FUNC(OPNAME ## h264_qpel8_h_lowpass)(dst , src , dstStride, srcStride);\
FUNC(OPNAME ## h264_qpel8_h_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\
src += 8*srcStride;\
@@ -368,7 +368,7 @@ static void FUNC(OPNAME ## h264_qpel16_h_lowpass)(uint8_t *dst, uint8_t *src, in
FUNC(OPNAME ## h264_qpel8_h_lowpass)(dst+8*sizeof(pixel), src+8*sizeof(pixel), dstStride, srcStride);\
}\
\
static void FUNC(OPNAME ## h264_qpel16_hv_lowpass)(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride){\
static void FUNC(OPNAME ## h264_qpel16_hv_lowpass)(uint8_t *dst, int16_t *tmp, const uint8_t *src, int dstStride, int tmpStride, int srcStride){\
FUNC(OPNAME ## h264_qpel8_hv_lowpass)(dst , tmp , src , dstStride, tmpStride, srcStride);\
FUNC(OPNAME ## h264_qpel8_hv_lowpass)(dst+8*sizeof(pixel), tmp+8, src+8*sizeof(pixel), dstStride, tmpStride, srcStride);\
src += 8*srcStride;\
@@ -378,31 +378,31 @@ static void FUNC(OPNAME ## h264_qpel16_hv_lowpass)(uint8_t *dst, int16_t *tmp, u
}\

#define H264_MC(OPNAME, SIZE) \
static av_unused void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc00)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static av_unused void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc00)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
FUNCC(OPNAME ## pixels ## SIZE)(dst, src, stride, SIZE);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc10)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc10)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t half[SIZE*SIZE*sizeof(pixel)];\
FUNC(put_h264_qpel ## SIZE ## _h_lowpass)(half, src, SIZE*sizeof(pixel), stride);\
FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, src, half, stride, stride, SIZE*sizeof(pixel), SIZE);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc20)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc20)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
FUNC(OPNAME ## h264_qpel ## SIZE ## _h_lowpass)(dst, src, stride, stride);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc30)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc30)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t half[SIZE*SIZE*sizeof(pixel)];\
FUNC(put_h264_qpel ## SIZE ## _h_lowpass)(half, src, SIZE*sizeof(pixel), stride);\
FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, src+sizeof(pixel), half, stride, stride, SIZE*sizeof(pixel), SIZE);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc01)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc01)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
@@ -412,7 +412,7 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc01)(uint8_t *dst, uint8_t *s
FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, full_mid, half, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc02)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc02)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
@@ -420,7 +420,7 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc02)(uint8_t *dst, uint8_t *s
FUNC(OPNAME ## h264_qpel ## SIZE ## _v_lowpass)(dst, full_mid, stride, SIZE*sizeof(pixel));\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc03)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc03)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
@@ -430,7 +430,7 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc03)(uint8_t *dst, uint8_t *s
FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, full_mid+SIZE*sizeof(pixel), half, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc11)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc11)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
@@ -442,7 +442,7 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc11)(uint8_t *dst, uint8_t *s
FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc31)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc31)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
@@ -454,7 +454,7 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc31)(uint8_t *dst, uint8_t *s
FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc13)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc13)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
@@ -466,7 +466,7 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc13)(uint8_t *dst, uint8_t *s
FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc33)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc33)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
@@ -478,13 +478,13 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc33)(uint8_t *dst, uint8_t *s
FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc22)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc22)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
FUNC(OPNAME ## h264_qpel ## SIZE ## _hv_lowpass)(dst, tmp, src, stride, SIZE*sizeof(pixel), stride);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc21)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc21)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\
@@ -494,7 +494,7 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc21)(uint8_t *dst, uint8_t *s
FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfHV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc23)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc23)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
int16_t tmp[SIZE*(SIZE+5)*sizeof(pixel)];\
uint8_t halfH[SIZE*SIZE*sizeof(pixel)];\
@@ -504,7 +504,7 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc23)(uint8_t *dst, uint8_t *s
FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfH, halfHV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc12)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc12)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\
@@ -517,7 +517,7 @@ static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc12)(uint8_t *dst, uint8_t *s
FUNC(OPNAME ## pixels ## SIZE ## _l2)(dst, halfV, halfHV, stride, SIZE*sizeof(pixel), SIZE*sizeof(pixel), SIZE);\
}\
\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc32)(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void FUNCC(OPNAME ## h264_qpel ## SIZE ## _mc32)(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)*sizeof(pixel)];\
uint8_t * const full_mid= full + SIZE*2*sizeof(pixel);\


+ 16
- 16
libavcodec/ppc/h264qpel.c View File

@@ -66,50 +66,50 @@
#undef PREFIX_h264_qpel16_hv_lowpass_num

#define H264_MC(OPNAME, SIZE, CODETYPE) \
static void OPNAME ## h264_qpel ## SIZE ## _mc00_ ## CODETYPE (uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc00_ ## CODETYPE (uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
ff_ ## OPNAME ## pixels ## SIZE ## _ ## CODETYPE(dst, src, stride, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc10_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc10_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{ \
DECLARE_ALIGNED(16, uint8_t, half)[SIZE*SIZE];\
put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(half, src, SIZE, stride);\
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, src, half, stride, stride, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc20_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc20_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(dst, src, stride, stride);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc30_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc30_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(16, uint8_t, half)[SIZE*SIZE];\
put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(half, src, SIZE, stride);\
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, src+1, half, stride, stride, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc01_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc01_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(16, uint8_t, half)[SIZE*SIZE];\
put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(half, src, SIZE, stride);\
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, src, half, stride, stride, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc02_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc02_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(dst, src, stride, stride);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc03_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc03_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(16, uint8_t, half)[SIZE*SIZE];\
put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(half, src, SIZE, stride);\
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, src+stride, half, stride, stride, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc11_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc11_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(16, uint8_t, halfH)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfV)[SIZE*SIZE];\
@@ -118,7 +118,7 @@ static void OPNAME ## h264_qpel ## SIZE ## _mc11_ ## CODETYPE(uint8_t *dst, uint
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfH, halfV, stride, SIZE, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc31_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc31_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(16, uint8_t, halfH)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfV)[SIZE*SIZE];\
@@ -127,7 +127,7 @@ static void OPNAME ## h264_qpel ## SIZE ## _mc31_ ## CODETYPE(uint8_t *dst, uint
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfH, halfV, stride, SIZE, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc13_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc13_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(16, uint8_t, halfH)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfV)[SIZE*SIZE];\
@@ -136,7 +136,7 @@ static void OPNAME ## h264_qpel ## SIZE ## _mc13_ ## CODETYPE(uint8_t *dst, uint
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfH, halfV, stride, SIZE, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc33_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc33_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(16, uint8_t, halfH)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfV)[SIZE*SIZE];\
@@ -145,13 +145,13 @@ static void OPNAME ## h264_qpel ## SIZE ## _mc33_ ## CODETYPE(uint8_t *dst, uint
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfH, halfV, stride, SIZE, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc22_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc22_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(16, int16_t, tmp)[SIZE*(SIZE+8)];\
OPNAME ## h264_qpel ## SIZE ## _hv_lowpass_ ## CODETYPE(dst, tmp, src, stride, SIZE, stride);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc21_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc21_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(16, uint8_t, halfH)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfHV)[SIZE*SIZE];\
@@ -161,7 +161,7 @@ static void OPNAME ## h264_qpel ## SIZE ## _mc21_ ## CODETYPE(uint8_t *dst, uint
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfH, halfHV, stride, SIZE, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc23_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc23_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(16, uint8_t, halfH)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfHV)[SIZE*SIZE];\
@@ -171,7 +171,7 @@ static void OPNAME ## h264_qpel ## SIZE ## _mc23_ ## CODETYPE(uint8_t *dst, uint
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfH, halfHV, stride, SIZE, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc12_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc12_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(16, uint8_t, halfV)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfHV)[SIZE*SIZE];\
@@ -181,7 +181,7 @@ static void OPNAME ## h264_qpel ## SIZE ## _mc12_ ## CODETYPE(uint8_t *dst, uint
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, halfV, halfHV, stride, SIZE, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc32_ ## CODETYPE(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc32_ ## CODETYPE(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(16, uint8_t, halfV)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfHV)[SIZE*SIZE];\


+ 14
- 4
libavcodec/ppc/h264qpel_template.c View File

@@ -28,7 +28,10 @@

/* this code assume stride % 16 == 0 */
#ifdef PREFIX_h264_qpel16_h_lowpass_altivec
static void PREFIX_h264_qpel16_h_lowpass_altivec(uint8_t * dst, uint8_t * src, int dstStride, int srcStride) {
static void PREFIX_h264_qpel16_h_lowpass_altivec(uint8_t *dst,
const uint8_t *src,
int dstStride, int srcStride)
{
register int i;

LOAD_ZERO;
@@ -168,7 +171,10 @@ static void PREFIX_h264_qpel16_h_lowpass_altivec(uint8_t * dst, uint8_t * src, i

/* this code assume stride % 16 == 0 */
#ifdef PREFIX_h264_qpel16_v_lowpass_altivec
static void PREFIX_h264_qpel16_v_lowpass_altivec(uint8_t * dst, uint8_t * src, int dstStride, int srcStride) {
static void PREFIX_h264_qpel16_v_lowpass_altivec(uint8_t *dst,
const uint8_t *src,
int dstStride, int srcStride)
{
register int i;

LOAD_ZERO;
@@ -178,7 +184,7 @@ static void PREFIX_h264_qpel16_v_lowpass_altivec(uint8_t * dst, uint8_t * src, i
const vec_s16 v5ss = vec_splat_s16(5);
const vec_s16 v16ss = vec_sl(vec_splat_s16(1),vec_splat_u16(4));

uint8_t *srcbis = src - (srcStride * 2);
const uint8_t *srcbis = src - (srcStride * 2);

const vec_u8 srcM2a = vec_ld(0, srcbis);
const vec_u8 srcM2b = vec_ld(16, srcbis);
@@ -275,7 +281,11 @@ static void PREFIX_h264_qpel16_v_lowpass_altivec(uint8_t * dst, uint8_t * src, i

/* this code assume stride % 16 == 0 *and* tmp is properly aligned */
#ifdef PREFIX_h264_qpel16_hv_lowpass_altivec
static void PREFIX_h264_qpel16_hv_lowpass_altivec(uint8_t * dst, int16_t * tmp, uint8_t * src, int dstStride, int tmpStride, int srcStride) {
static void PREFIX_h264_qpel16_hv_lowpass_altivec(uint8_t *dst, int16_t *tmp,
const uint8_t *src,
int dstStride, int tmpStride,
int srcStride)
{
register int i;
LOAD_ZERO;
const vec_u8 permM2 = vec_lvsl(-2, src);


+ 52
- 50
libavcodec/qpeldsp.c View File

@@ -37,7 +37,7 @@
#include "qpel_template.c"

#define QPEL_MC(r, OPNAME, RND, OP) \
static void OPNAME ## mpeg4_qpel8_h_lowpass(uint8_t *dst, uint8_t *src, \
static void OPNAME ## mpeg4_qpel8_h_lowpass(uint8_t *dst, const uint8_t *src, \
int dstStride, int srcStride, \
int h) \
{ \
@@ -58,7 +58,7 @@ static void OPNAME ## mpeg4_qpel8_h_lowpass(uint8_t *dst, uint8_t *src, \
} \
} \
\
static void OPNAME ## mpeg4_qpel8_v_lowpass(uint8_t *dst, uint8_t *src, \
static void OPNAME ## mpeg4_qpel8_v_lowpass(uint8_t *dst, const uint8_t *src, \
int dstStride, int srcStride) \
{ \
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP; \
@@ -88,7 +88,8 @@ static void OPNAME ## mpeg4_qpel8_v_lowpass(uint8_t *dst, uint8_t *src, \
} \
} \
\
static void OPNAME ## mpeg4_qpel16_h_lowpass(uint8_t *dst, uint8_t *src, \
static void OPNAME ## mpeg4_qpel16_h_lowpass(uint8_t *dst, \
const uint8_t *src, \
int dstStride, int srcStride, \
int h) \
{ \
@@ -117,7 +118,8 @@ static void OPNAME ## mpeg4_qpel16_h_lowpass(uint8_t *dst, uint8_t *src, \
} \
} \
\
static void OPNAME ## mpeg4_qpel16_v_lowpass(uint8_t *dst, uint8_t *src, \
static void OPNAME ## mpeg4_qpel16_v_lowpass(uint8_t *dst, \
const uint8_t *src, \
int dstStride, int srcStride) \
{ \
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP; \
@@ -163,7 +165,7 @@ static void OPNAME ## mpeg4_qpel16_v_lowpass(uint8_t *dst, uint8_t *src, \
} \
} \
\
static void OPNAME ## qpel8_mc10_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc10_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t half[64]; \
@@ -172,13 +174,13 @@ static void OPNAME ## qpel8_mc10_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels8_l2_8(dst, src, half, stride, stride, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc20_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc20_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
OPNAME ## mpeg4_qpel8_h_lowpass(dst, src, stride, stride, 8); \
} \
\
static void OPNAME ## qpel8_mc30_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc30_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t half[64]; \
@@ -187,7 +189,7 @@ static void OPNAME ## qpel8_mc30_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels8_l2_8(dst, src + 1, half, stride, stride, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc01_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc01_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -198,7 +200,7 @@ static void OPNAME ## qpel8_mc01_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels8_l2_8(dst, full, half, stride, 16, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc02_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc02_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -207,7 +209,7 @@ static void OPNAME ## qpel8_mc02_c(uint8_t *dst, uint8_t *src, \
OPNAME ## mpeg4_qpel8_v_lowpass(dst, full, stride, 16); \
} \
\
static void OPNAME ## qpel8_mc03_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc03_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -218,7 +220,7 @@ static void OPNAME ## qpel8_mc03_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels8_l2_8(dst, full + 16, half, stride, 16, 8, 8); \
} \
\
void ff_ ## OPNAME ## qpel8_mc11_old_c(uint8_t *dst, uint8_t *src, \
void ff_ ## OPNAME ## qpel8_mc11_old_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -234,7 +236,7 @@ void ff_ ## OPNAME ## qpel8_mc11_old_c(uint8_t *dst, uint8_t *src, \
stride, 16, 8, 8, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc11_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc11_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -248,7 +250,7 @@ static void OPNAME ## qpel8_mc11_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels8_l2_8(dst, halfH, halfHV, stride, 8, 8, 8); \
} \
\
void ff_ ## OPNAME ## qpel8_mc31_old_c(uint8_t *dst, uint8_t *src, \
void ff_ ## OPNAME ## qpel8_mc31_old_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -264,7 +266,7 @@ void ff_ ## OPNAME ## qpel8_mc31_old_c(uint8_t *dst, uint8_t *src, \
stride, 16, 8, 8, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc31_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc31_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -278,7 +280,7 @@ static void OPNAME ## qpel8_mc31_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels8_l2_8(dst, halfH, halfHV, stride, 8, 8, 8); \
} \
\
void ff_ ## OPNAME ## qpel8_mc13_old_c(uint8_t *dst, uint8_t *src, \
void ff_ ## OPNAME ## qpel8_mc13_old_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -294,7 +296,7 @@ void ff_ ## OPNAME ## qpel8_mc13_old_c(uint8_t *dst, uint8_t *src, \
stride, 16, 8, 8, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc13_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc13_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -308,7 +310,7 @@ static void OPNAME ## qpel8_mc13_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels8_l2_8(dst, halfH + 8, halfHV, stride, 8, 8, 8); \
} \
\
void ff_ ## OPNAME ## qpel8_mc33_old_c(uint8_t *dst, uint8_t *src, \
void ff_ ## OPNAME ## qpel8_mc33_old_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -324,7 +326,7 @@ void ff_ ## OPNAME ## qpel8_mc33_old_c(uint8_t *dst, uint8_t *src, \
stride, 16, 8, 8, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc33_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc33_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -338,7 +340,7 @@ static void OPNAME ## qpel8_mc33_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels8_l2_8(dst, halfH + 8, halfHV, stride, 8, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc21_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc21_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t halfH[72]; \
@@ -349,7 +351,7 @@ static void OPNAME ## qpel8_mc21_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels8_l2_8(dst, halfH, halfHV, stride, 8, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc23_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc23_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t halfH[72]; \
@@ -360,7 +362,7 @@ static void OPNAME ## qpel8_mc23_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels8_l2_8(dst, halfH + 8, halfHV, stride, 8, 8, 8); \
} \
\
void ff_ ## OPNAME ## qpel8_mc12_old_c(uint8_t *dst, uint8_t *src, \
void ff_ ## OPNAME ## qpel8_mc12_old_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -375,7 +377,7 @@ void ff_ ## OPNAME ## qpel8_mc12_old_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels8_l2_8(dst, halfV, halfHV, stride, 8, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc12_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc12_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -387,7 +389,7 @@ static void OPNAME ## qpel8_mc12_c(uint8_t *dst, uint8_t *src, \
OPNAME ## mpeg4_qpel8_v_lowpass(dst, halfH, stride, 8); \
} \
\
void ff_ ## OPNAME ## qpel8_mc32_old_c(uint8_t *dst, uint8_t *src, \
void ff_ ## OPNAME ## qpel8_mc32_old_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -402,7 +404,7 @@ void ff_ ## OPNAME ## qpel8_mc32_old_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels8_l2_8(dst, halfV, halfHV, stride, 8, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc32_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc32_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[16 * 9]; \
@@ -414,7 +416,7 @@ static void OPNAME ## qpel8_mc32_c(uint8_t *dst, uint8_t *src, \
OPNAME ## mpeg4_qpel8_v_lowpass(dst, halfH, stride, 8); \
} \
\
static void OPNAME ## qpel8_mc22_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc22_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t halfH[72]; \
@@ -423,7 +425,7 @@ static void OPNAME ## qpel8_mc22_c(uint8_t *dst, uint8_t *src, \
OPNAME ## mpeg4_qpel8_v_lowpass(dst, halfH, stride, 8); \
} \
\
static void OPNAME ## qpel16_mc10_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc10_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t half[256]; \
@@ -432,13 +434,13 @@ static void OPNAME ## qpel16_mc10_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels16_l2_8(dst, src, half, stride, stride, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc20_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc20_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
OPNAME ## mpeg4_qpel16_h_lowpass(dst, src, stride, stride, 16); \
} \
\
static void OPNAME ## qpel16_mc30_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc30_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t half[256]; \
@@ -447,7 +449,7 @@ static void OPNAME ## qpel16_mc30_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels16_l2_8(dst, src + 1, half, stride, stride, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc01_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc01_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -458,7 +460,7 @@ static void OPNAME ## qpel16_mc01_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels16_l2_8(dst, full, half, stride, 24, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc02_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc02_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -467,7 +469,7 @@ static void OPNAME ## qpel16_mc02_c(uint8_t *dst, uint8_t *src, \
OPNAME ## mpeg4_qpel16_v_lowpass(dst, full, stride, 24); \
} \
\
static void OPNAME ## qpel16_mc03_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc03_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -478,7 +480,7 @@ static void OPNAME ## qpel16_mc03_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels16_l2_8(dst, full + 24, half, stride, 24, 16, 16); \
} \
\
void ff_ ## OPNAME ## qpel16_mc11_old_c(uint8_t *dst, uint8_t *src, \
void ff_ ## OPNAME ## qpel16_mc11_old_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -494,7 +496,7 @@ void ff_ ## OPNAME ## qpel16_mc11_old_c(uint8_t *dst, uint8_t *src, \
stride, 24, 16, 16, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc11_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc11_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -508,7 +510,7 @@ static void OPNAME ## qpel16_mc11_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels16_l2_8(dst, halfH, halfHV, stride, 16, 16, 16); \
} \
\
void ff_ ## OPNAME ## qpel16_mc31_old_c(uint8_t *dst, uint8_t *src, \
void ff_ ## OPNAME ## qpel16_mc31_old_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -524,7 +526,7 @@ void ff_ ## OPNAME ## qpel16_mc31_old_c(uint8_t *dst, uint8_t *src, \
stride, 24, 16, 16, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc31_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc31_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -538,7 +540,7 @@ static void OPNAME ## qpel16_mc31_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels16_l2_8(dst, halfH, halfHV, stride, 16, 16, 16); \
} \
\
void ff_ ## OPNAME ## qpel16_mc13_old_c(uint8_t *dst, uint8_t *src, \
void ff_ ## OPNAME ## qpel16_mc13_old_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -554,7 +556,7 @@ void ff_ ## OPNAME ## qpel16_mc13_old_c(uint8_t *dst, uint8_t *src, \
stride, 24, 16, 16, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc13_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc13_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -568,7 +570,7 @@ static void OPNAME ## qpel16_mc13_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels16_l2_8(dst, halfH + 16, halfHV, stride, 16, 16, 16); \
} \
\
void ff_ ## OPNAME ## qpel16_mc33_old_c(uint8_t *dst, uint8_t *src, \
void ff_ ## OPNAME ## qpel16_mc33_old_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -584,7 +586,7 @@ void ff_ ## OPNAME ## qpel16_mc33_old_c(uint8_t *dst, uint8_t *src, \
stride, 24, 16, 16, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc33_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc33_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -598,7 +600,7 @@ static void OPNAME ## qpel16_mc33_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels16_l2_8(dst, halfH + 16, halfHV, stride, 16, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc21_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc21_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t halfH[272]; \
@@ -609,7 +611,7 @@ static void OPNAME ## qpel16_mc21_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels16_l2_8(dst, halfH, halfHV, stride, 16, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc23_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc23_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t halfH[272]; \
@@ -620,7 +622,7 @@ static void OPNAME ## qpel16_mc23_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels16_l2_8(dst, halfH + 16, halfHV, stride, 16, 16, 16); \
} \
\
void ff_ ## OPNAME ## qpel16_mc12_old_c(uint8_t *dst, uint8_t *src, \
void ff_ ## OPNAME ## qpel16_mc12_old_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -635,7 +637,7 @@ void ff_ ## OPNAME ## qpel16_mc12_old_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels16_l2_8(dst, halfV, halfHV, stride, 16, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc12_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc12_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -647,7 +649,7 @@ static void OPNAME ## qpel16_mc12_c(uint8_t *dst, uint8_t *src, \
OPNAME ## mpeg4_qpel16_v_lowpass(dst, halfH, stride, 16); \
} \
\
void ff_ ## OPNAME ## qpel16_mc32_old_c(uint8_t *dst, uint8_t *src, \
void ff_ ## OPNAME ## qpel16_mc32_old_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -662,7 +664,7 @@ void ff_ ## OPNAME ## qpel16_mc32_old_c(uint8_t *dst, uint8_t *src, \
OPNAME ## pixels16_l2_8(dst, halfV, halfHV, stride, 16, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc32_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc32_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t full[24 * 17]; \
@@ -674,7 +676,7 @@ static void OPNAME ## qpel16_mc32_c(uint8_t *dst, uint8_t *src, \
OPNAME ## mpeg4_qpel16_v_lowpass(dst, halfH, stride, 16); \
} \
\
static void OPNAME ## qpel16_mc22_c(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc22_c(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint8_t halfH[272]; \
@@ -695,22 +697,22 @@ QPEL_MC(0, avg_, _, op_avg)
#undef op_put
#undef op_put_no_rnd

void ff_put_pixels8x8_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
void ff_put_pixels8x8_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
put_pixels8_8_c(dst, src, stride, 8);
}

void ff_avg_pixels8x8_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
void ff_avg_pixels8x8_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
avg_pixels8_8_c(dst, src, stride, 8);
}

void ff_put_pixels16x16_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
void ff_put_pixels16x16_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
put_pixels16_8_c(dst, src, stride, 16);
}

void ff_avg_pixels16x16_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
void ff_avg_pixels16x16_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
avg_pixels16_8_c(dst, src, stride, 16);
}


+ 16
- 12
libavcodec/qpeldsp.h View File

@@ -29,22 +29,25 @@
#include <stddef.h>
#include <stdint.h>

void ff_put_pixels8x8_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride);
void ff_avg_pixels8x8_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride);
void ff_put_pixels16x16_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride);
void ff_avg_pixels16x16_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride);
void ff_put_pixels8x8_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_pixels8x8_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_pixels16x16_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_pixels16x16_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);

void ff_put_pixels8_l2_8(uint8_t *dst, const uint8_t *src1, const uint8_t *src2,
int dst_stride, int src_stride1, int src_stride2,
int h);

#define DEF_OLD_QPEL(name) \
void ff_put_ ## name(uint8_t *dst /* align width (8 or 16) */, \
uint8_t *src /* align 1 */, ptrdiff_t stride); \
void ff_put_no_rnd_ ## name(uint8_t *dst /* align width (8 or 16) */, \
uint8_t *src /* align 1 */, ptrdiff_t stride); \
void ff_avg_ ## name(uint8_t *dst /* align width (8 or 16) */, \
uint8_t *src /* align 1 */, ptrdiff_t stride);
#define DEF_OLD_QPEL(name) \
void ff_put_ ## name(uint8_t *dst /* align width (8 or 16) */, \
const uint8_t *src /* align 1 */, \
ptrdiff_t stride); \
void ff_put_no_rnd_ ## name(uint8_t *dst /* align width (8 or 16) */, \
const uint8_t *src /* align 1 */, \
ptrdiff_t stride); \
void ff_avg_ ## name(uint8_t *dst /* align width (8 or 16) */, \
const uint8_t *src /* align 1 */, \
ptrdiff_t stride);

DEF_OLD_QPEL(qpel16_mc11_old_c)
DEF_OLD_QPEL(qpel16_mc31_old_c)
@@ -60,7 +63,8 @@ DEF_OLD_QPEL(qpel8_mc13_old_c)
DEF_OLD_QPEL(qpel8_mc33_old_c)

typedef void (*qpel_mc_func)(uint8_t *dst /* align width (8 or 16) */,
uint8_t *src /* align 1 */, ptrdiff_t stride);
const uint8_t *src /* align 1 */,
ptrdiff_t stride);

/**
* quarterpel DSP context


+ 20
- 20
libavcodec/rv30dsp.c View File

@@ -31,7 +31,7 @@
#include "rv34dsp.h"

#define RV30_LOWPASS(OPNAME, OP) \
static av_unused void OPNAME ## rv30_tpel8_h_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, const int C1, const int C2){\
static av_unused void OPNAME ## rv30_tpel8_h_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, const int C1, const int C2){\
const int h = 8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
int i;\
@@ -50,7 +50,7 @@ static av_unused void OPNAME ## rv30_tpel8_h_lowpass(uint8_t *dst, uint8_t *src,
}\
}\
\
static void OPNAME ## rv30_tpel8_v_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, const int C1, const int C2){\
static void OPNAME ## rv30_tpel8_v_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, const int C1, const int C2){\
const int w = 8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
int i;\
@@ -80,7 +80,7 @@ static void OPNAME ## rv30_tpel8_v_lowpass(uint8_t *dst, uint8_t *src, int dstSt
}\
}\
\
static void OPNAME ## rv30_tpel8_hv_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## rv30_tpel8_hv_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
const int w = 8;\
const int h = 8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
@@ -99,7 +99,7 @@ static void OPNAME ## rv30_tpel8_hv_lowpass(uint8_t *dst, uint8_t *src, int dstS
}\
}\
\
static void OPNAME ## rv30_tpel8_hhv_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## rv30_tpel8_hhv_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
const int w = 8;\
const int h = 8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
@@ -118,7 +118,7 @@ static void OPNAME ## rv30_tpel8_hhv_lowpass(uint8_t *dst, uint8_t *src, int dst
}\
}\
\
static void OPNAME ## rv30_tpel8_hvv_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## rv30_tpel8_hvv_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
const int w = 8;\
const int h = 8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
@@ -137,7 +137,7 @@ static void OPNAME ## rv30_tpel8_hvv_lowpass(uint8_t *dst, uint8_t *src, int dst
}\
}\
\
static void OPNAME ## rv30_tpel8_hhvv_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## rv30_tpel8_hhvv_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
const int w = 8;\
const int h = 8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
@@ -155,7 +155,7 @@ static void OPNAME ## rv30_tpel8_hhvv_lowpass(uint8_t *dst, uint8_t *src, int ds
}\
}\
\
static void OPNAME ## rv30_tpel16_v_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, const int C1, const int C2){\
static void OPNAME ## rv30_tpel16_v_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, const int C1, const int C2){\
OPNAME ## rv30_tpel8_v_lowpass(dst , src , dstStride, srcStride, C1, C2);\
OPNAME ## rv30_tpel8_v_lowpass(dst+8, src+8, dstStride, srcStride, C1, C2);\
src += 8*srcStride;\
@@ -164,7 +164,7 @@ static void OPNAME ## rv30_tpel16_v_lowpass(uint8_t *dst, uint8_t *src, int dstS
OPNAME ## rv30_tpel8_v_lowpass(dst+8, src+8, dstStride, srcStride, C1, C2);\
}\
\
static void OPNAME ## rv30_tpel16_h_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, const int C1, const int C2){\
static void OPNAME ## rv30_tpel16_h_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, const int C1, const int C2){\
OPNAME ## rv30_tpel8_h_lowpass(dst , src , dstStride, srcStride, C1, C2);\
OPNAME ## rv30_tpel8_h_lowpass(dst+8, src+8, dstStride, srcStride, C1, C2);\
src += 8*srcStride;\
@@ -173,7 +173,7 @@ static void OPNAME ## rv30_tpel16_h_lowpass(uint8_t *dst, uint8_t *src, int dstS
OPNAME ## rv30_tpel8_h_lowpass(dst+8, src+8, dstStride, srcStride, C1, C2);\
}\
\
static void OPNAME ## rv30_tpel16_hv_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## rv30_tpel16_hv_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## rv30_tpel8_hv_lowpass(dst , src , dstStride, srcStride);\
OPNAME ## rv30_tpel8_hv_lowpass(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\
@@ -182,7 +182,7 @@ static void OPNAME ## rv30_tpel16_hv_lowpass(uint8_t *dst, uint8_t *src, int dst
OPNAME ## rv30_tpel8_hv_lowpass(dst+8, src+8, dstStride, srcStride);\
}\
\
static void OPNAME ## rv30_tpel16_hhv_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## rv30_tpel16_hhv_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## rv30_tpel8_hhv_lowpass(dst , src , dstStride, srcStride);\
OPNAME ## rv30_tpel8_hhv_lowpass(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\
@@ -191,7 +191,7 @@ static void OPNAME ## rv30_tpel16_hhv_lowpass(uint8_t *dst, uint8_t *src, int ds
OPNAME ## rv30_tpel8_hhv_lowpass(dst+8, src+8, dstStride, srcStride);\
}\
\
static void OPNAME ## rv30_tpel16_hvv_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## rv30_tpel16_hvv_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## rv30_tpel8_hvv_lowpass(dst , src , dstStride, srcStride);\
OPNAME ## rv30_tpel8_hvv_lowpass(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\
@@ -200,7 +200,7 @@ static void OPNAME ## rv30_tpel16_hvv_lowpass(uint8_t *dst, uint8_t *src, int ds
OPNAME ## rv30_tpel8_hvv_lowpass(dst+8, src+8, dstStride, srcStride);\
}\
\
static void OPNAME ## rv30_tpel16_hhvv_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## rv30_tpel16_hhvv_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## rv30_tpel8_hhvv_lowpass(dst , src , dstStride, srcStride);\
OPNAME ## rv30_tpel8_hhvv_lowpass(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\
@@ -211,42 +211,42 @@ static void OPNAME ## rv30_tpel16_hhvv_lowpass(uint8_t *dst, uint8_t *src, int d
\

#define RV30_MC(OPNAME, SIZE) \
static void OPNAME ## rv30_tpel ## SIZE ## _mc10_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv30_tpel ## SIZE ## _mc10_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## rv30_tpel ## SIZE ## _h_lowpass(dst, src, stride, stride, 12, 6);\
}\
\
static void OPNAME ## rv30_tpel ## SIZE ## _mc20_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv30_tpel ## SIZE ## _mc20_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## rv30_tpel ## SIZE ## _h_lowpass(dst, src, stride, stride, 6, 12);\
}\
\
static void OPNAME ## rv30_tpel ## SIZE ## _mc01_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv30_tpel ## SIZE ## _mc01_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## rv30_tpel ## SIZE ## _v_lowpass(dst, src, stride, stride, 12, 6);\
}\
\
static void OPNAME ## rv30_tpel ## SIZE ## _mc02_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv30_tpel ## SIZE ## _mc02_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## rv30_tpel ## SIZE ## _v_lowpass(dst, src, stride, stride, 6, 12);\
}\
\
static void OPNAME ## rv30_tpel ## SIZE ## _mc11_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv30_tpel ## SIZE ## _mc11_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## rv30_tpel ## SIZE ## _hv_lowpass(dst, src, stride, stride);\
}\
\
static void OPNAME ## rv30_tpel ## SIZE ## _mc12_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv30_tpel ## SIZE ## _mc12_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## rv30_tpel ## SIZE ## _hvv_lowpass(dst, src, stride, stride);\
}\
\
static void OPNAME ## rv30_tpel ## SIZE ## _mc21_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv30_tpel ## SIZE ## _mc21_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## rv30_tpel ## SIZE ## _hhv_lowpass(dst, src, stride, stride);\
}\
\
static void OPNAME ## rv30_tpel ## SIZE ## _mc22_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv30_tpel ## SIZE ## _mc22_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## rv30_tpel ## SIZE ## _hhvv_lowpass(dst, src, stride, stride);\
}\


+ 20
- 20
libavcodec/rv40dsp.c View File

@@ -34,7 +34,7 @@
#include "rv34dsp.h"

#define RV40_LOWPASS(OPNAME, OP) \
static av_unused void OPNAME ## rv40_qpel8_h_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride,\
static av_unused void OPNAME ## rv40_qpel8_h_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride,\
const int h, const int C1, const int C2, const int SHIFT){\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
int i;\
@@ -53,7 +53,7 @@ static av_unused void OPNAME ## rv40_qpel8_h_lowpass(uint8_t *dst, uint8_t *src,
}\
}\
\
static void OPNAME ## rv40_qpel8_v_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride,\
static void OPNAME ## rv40_qpel8_v_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride,\
const int w, const int C1, const int C2, const int SHIFT){\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
int i;\
@@ -85,7 +85,7 @@ static void OPNAME ## rv40_qpel8_v_lowpass(uint8_t *dst, uint8_t *src, int dstSt
}\
}\
\
static void OPNAME ## rv40_qpel16_v_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride,\
static void OPNAME ## rv40_qpel16_v_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride,\
const int w, const int C1, const int C2, const int SHIFT){\
OPNAME ## rv40_qpel8_v_lowpass(dst , src , dstStride, srcStride, 8, C1, C2, SHIFT);\
OPNAME ## rv40_qpel8_v_lowpass(dst+8, src+8, dstStride, srcStride, 8, C1, C2, SHIFT);\
@@ -95,7 +95,7 @@ static void OPNAME ## rv40_qpel16_v_lowpass(uint8_t *dst, uint8_t *src, int dstS
OPNAME ## rv40_qpel8_v_lowpass(dst+8, src+8, dstStride, srcStride, w-8, C1, C2, SHIFT);\
}\
\
static void OPNAME ## rv40_qpel16_h_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride,\
static void OPNAME ## rv40_qpel16_h_lowpass(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride,\
const int h, const int C1, const int C2, const int SHIFT){\
OPNAME ## rv40_qpel8_h_lowpass(dst , src , dstStride, srcStride, 8, C1, C2, SHIFT);\
OPNAME ## rv40_qpel8_h_lowpass(dst+8, src+8, dstStride, srcStride, 8, C1, C2, SHIFT);\
@@ -107,22 +107,22 @@ static void OPNAME ## rv40_qpel16_h_lowpass(uint8_t *dst, uint8_t *src, int dstS
\

#define RV40_MC(OPNAME, SIZE) \
static void OPNAME ## rv40_qpel ## SIZE ## _mc10_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv40_qpel ## SIZE ## _mc10_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## rv40_qpel ## SIZE ## _h_lowpass(dst, src, stride, stride, SIZE, 52, 20, 6);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc30_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv40_qpel ## SIZE ## _mc30_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## rv40_qpel ## SIZE ## _h_lowpass(dst, src, stride, stride, SIZE, 20, 52, 6);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc01_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv40_qpel ## SIZE ## _mc01_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, src, stride, stride, SIZE, 52, 20, 6);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc11_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv40_qpel ## SIZE ## _mc11_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\
@@ -130,7 +130,7 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc11_c(uint8_t *dst, uint8_t *src, p
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 52, 20, 6);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc21_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv40_qpel ## SIZE ## _mc21_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\
@@ -138,7 +138,7 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc21_c(uint8_t *dst, uint8_t *src, p
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 52, 20, 6);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc31_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv40_qpel ## SIZE ## _mc31_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\
@@ -146,7 +146,7 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc31_c(uint8_t *dst, uint8_t *src, p
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 52, 20, 6);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc12_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv40_qpel ## SIZE ## _mc12_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\
@@ -154,7 +154,7 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc12_c(uint8_t *dst, uint8_t *src, p
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 20, 5);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc22_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv40_qpel ## SIZE ## _mc22_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\
@@ -162,7 +162,7 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc22_c(uint8_t *dst, uint8_t *src, p
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 20, 5);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc32_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv40_qpel ## SIZE ## _mc32_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\
@@ -170,12 +170,12 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc32_c(uint8_t *dst, uint8_t *src, p
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 20, 5);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc03_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv40_qpel ## SIZE ## _mc03_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, src, stride, stride, SIZE, 20, 52, 6);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc13_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv40_qpel ## SIZE ## _mc13_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\
@@ -183,7 +183,7 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc13_c(uint8_t *dst, uint8_t *src, p
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 52, 6);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc23_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## rv40_qpel ## SIZE ## _mc23_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\
@@ -266,19 +266,19 @@ PIXOP2(put, op_put)
#undef op_avg
#undef op_put

static void put_rv40_qpel16_mc33_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
static void put_rv40_qpel16_mc33_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
put_pixels16_xy2_8_c(dst, src, stride, 16);
}
static void avg_rv40_qpel16_mc33_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
static void avg_rv40_qpel16_mc33_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
avg_pixels16_xy2_8_c(dst, src, stride, 16);
}
static void put_rv40_qpel8_mc33_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
static void put_rv40_qpel8_mc33_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
put_pixels8_xy2_8_c(dst, src, stride, 8);
}
static void avg_rv40_qpel8_mc33_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
static void avg_rv40_qpel8_mc33_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
avg_pixels8_xy2_8_c(dst, src, stride, 8);
}


+ 9
- 9
libavcodec/wmv2dsp.c View File

@@ -139,7 +139,7 @@ static void wmv2_idct_put_c(uint8_t *dest, int line_size, int16_t *block)
}
}

static void wmv2_mspel8_h_lowpass(uint8_t *dst, uint8_t *src,
static void wmv2_mspel8_h_lowpass(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride, int h)
{
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;
@@ -159,7 +159,7 @@ static void wmv2_mspel8_h_lowpass(uint8_t *dst, uint8_t *src,
}
}

static void wmv2_mspel8_v_lowpass(uint8_t *dst, uint8_t *src,
static void wmv2_mspel8_v_lowpass(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride, int w)
{
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;
@@ -190,7 +190,7 @@ static void wmv2_mspel8_v_lowpass(uint8_t *dst, uint8_t *src,
}
}

static void put_mspel8_mc10_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
static void put_mspel8_mc10_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
uint8_t half[64];

@@ -198,12 +198,12 @@ static void put_mspel8_mc10_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
ff_put_pixels8_l2_8(dst, src, half, stride, stride, 8, 8);
}

static void put_mspel8_mc20_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
static void put_mspel8_mc20_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
wmv2_mspel8_h_lowpass(dst, src, stride, stride, 8);
}

static void put_mspel8_mc30_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
static void put_mspel8_mc30_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
uint8_t half[64];

@@ -211,12 +211,12 @@ static void put_mspel8_mc30_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
ff_put_pixels8_l2_8(dst, src + 1, half, stride, stride, 8, 8);
}

static void put_mspel8_mc02_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
static void put_mspel8_mc02_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
wmv2_mspel8_v_lowpass(dst, src, stride, stride, 8);
}

static void put_mspel8_mc12_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
static void put_mspel8_mc12_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
uint8_t halfH[88];
uint8_t halfV[64];
@@ -228,7 +228,7 @@ static void put_mspel8_mc12_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
ff_put_pixels8_l2_8(dst, halfV, halfHV, stride, 8, 8, 8);
}

static void put_mspel8_mc32_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
static void put_mspel8_mc32_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
uint8_t halfH[88];
uint8_t halfV[64];
@@ -240,7 +240,7 @@ static void put_mspel8_mc32_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
ff_put_pixels8_l2_8(dst, halfV, halfHV, stride, 8, 8, 8);
}

static void put_mspel8_mc22_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride)
static void put_mspel8_mc22_c(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)
{
uint8_t halfH[88];



+ 19
- 19
libavcodec/x86/cavsdsp.c View File

@@ -334,7 +334,7 @@ static void cavs_idct8_add_mmx(uint8_t *dst, int16_t *block, int stride)
}

#define QPEL_CAVS(OPNAME, OP, MMX)\
static void OPNAME ## cavs_qpel8_h_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## cavs_qpel8_h_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
int h=8;\
__asm__ volatile(\
"pxor %%mm7, %%mm7 \n\t"\
@@ -381,43 +381,43 @@ static void OPNAME ## cavs_qpel8_h_ ## MMX(uint8_t *dst, uint8_t *src, int dstSt
);\
}\
\
static inline void OPNAME ## cavs_qpel8or16_v1_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h){\
static inline void OPNAME ## cavs_qpel8or16_v1_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h){\
QPEL_CAVSVNUM(QPEL_CAVSV1,OP,ff_pw_64,ff_pw_96,ff_pw_42) \
}\
\
static inline void OPNAME ## cavs_qpel8or16_v2_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h){\
static inline void OPNAME ## cavs_qpel8or16_v2_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h){\
QPEL_CAVSVNUM(QPEL_CAVSV2,OP,ff_pw_4,ff_pw_5,ff_pw_5) \
}\
\
static inline void OPNAME ## cavs_qpel8or16_v3_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h){\
static inline void OPNAME ## cavs_qpel8or16_v3_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h){\
QPEL_CAVSVNUM(QPEL_CAVSV3,OP,ff_pw_64,ff_pw_96,ff_pw_42) \
}\
\
static void OPNAME ## cavs_qpel8_v1_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## cavs_qpel8_v1_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## cavs_qpel8or16_v1_ ## MMX(dst , src , dstStride, srcStride, 8);\
}\
static void OPNAME ## cavs_qpel16_v1_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## cavs_qpel16_v1_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## cavs_qpel8or16_v1_ ## MMX(dst , src , dstStride, srcStride, 16);\
OPNAME ## cavs_qpel8or16_v1_ ## MMX(dst+8, src+8, dstStride, srcStride, 16);\
}\
\
static void OPNAME ## cavs_qpel8_v2_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## cavs_qpel8_v2_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## cavs_qpel8or16_v2_ ## MMX(dst , src , dstStride, srcStride, 8);\
}\
static void OPNAME ## cavs_qpel16_v2_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## cavs_qpel16_v2_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## cavs_qpel8or16_v2_ ## MMX(dst , src , dstStride, srcStride, 16);\
OPNAME ## cavs_qpel8or16_v2_ ## MMX(dst+8, src+8, dstStride, srcStride, 16);\
}\
\
static void OPNAME ## cavs_qpel8_v3_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## cavs_qpel8_v3_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## cavs_qpel8or16_v3_ ## MMX(dst , src , dstStride, srcStride, 8);\
}\
static void OPNAME ## cavs_qpel16_v3_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## cavs_qpel16_v3_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## cavs_qpel8or16_v3_ ## MMX(dst , src , dstStride, srcStride, 16);\
OPNAME ## cavs_qpel8or16_v3_ ## MMX(dst+8, src+8, dstStride, srcStride, 16);\
}\
\
static void OPNAME ## cavs_qpel16_h_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static void OPNAME ## cavs_qpel16_h_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
OPNAME ## cavs_qpel8_h_ ## MMX(dst , src , dstStride, srcStride);\
OPNAME ## cavs_qpel8_h_ ## MMX(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\
@@ -427,22 +427,22 @@ static void OPNAME ## cavs_qpel16_h_ ## MMX(uint8_t *dst, uint8_t *src, int dstS
}\

#define CAVS_MC(OPNAME, SIZE, MMX) \
static void OPNAME ## cavs_qpel ## SIZE ## _mc20_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc20_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_qpel ## SIZE ## _h_ ## MMX(dst, src, stride, stride);\
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc01_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc01_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_qpel ## SIZE ## _v1_ ## MMX(dst, src, stride, stride);\
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc02_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc02_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_qpel ## SIZE ## _v2_ ## MMX(dst, src, stride, stride);\
}\
\
static void OPNAME ## cavs_qpel ## SIZE ## _mc03_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## cavs_qpel ## SIZE ## _mc03_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
OPNAME ## cavs_qpel ## SIZE ## _v3_ ## MMX(dst, src, stride, stride);\
}\
@@ -460,25 +460,25 @@ static void OPNAME ## cavs_qpel ## SIZE ## _mc03_ ## MMX(uint8_t *dst, uint8_t *
#endif /* (HAVE_MMXEXT_INLINE || HAVE_AMD3DNOW_INLINE) */

#if HAVE_MMX_INLINE
static void put_cavs_qpel8_mc00_mmx(uint8_t *dst, uint8_t *src,
static void put_cavs_qpel8_mc00_mmx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
ff_put_pixels8_mmx(dst, src, stride, 8);
}

static void avg_cavs_qpel8_mc00_mmx(uint8_t *dst, uint8_t *src,
static void avg_cavs_qpel8_mc00_mmx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
ff_avg_pixels8_mmx(dst, src, stride, 8);
}

static void put_cavs_qpel16_mc00_mmx(uint8_t *dst, uint8_t *src,
static void put_cavs_qpel16_mc00_mmx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
ff_put_pixels16_mmx(dst, src, stride, 16);
}

static void avg_cavs_qpel16_mc00_mmx(uint8_t *dst, uint8_t *src,
static void avg_cavs_qpel16_mc00_mmx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
ff_avg_pixels16_mmx(dst, src, stride, 16);


+ 61
- 61
libavcodec/x86/h264_qpel.c View File

@@ -33,17 +33,17 @@ void ff_put_pixels4_mmxext(uint8_t *block, const uint8_t *pixels,
ptrdiff_t line_size, int h);
void ff_avg_pixels4_mmxext(uint8_t *block, const uint8_t *pixels,
ptrdiff_t line_size, int h);
void ff_put_pixels4_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
void ff_put_pixels4_l2_mmxext(uint8_t *dst, const uint8_t *src1, const uint8_t *src2,
int dstStride, int src1Stride, int h);
void ff_avg_pixels4_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
void ff_avg_pixels4_l2_mmxext(uint8_t *dst, const uint8_t *src1, const uint8_t *src2,
int dstStride, int src1Stride, int h);
void ff_put_pixels8_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
void ff_put_pixels8_l2_mmxext(uint8_t *dst, const uint8_t *src1, const uint8_t *src2,
int dstStride, int src1Stride, int h);
void ff_avg_pixels8_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
void ff_avg_pixels8_l2_mmxext(uint8_t *dst, const uint8_t *src1, const uint8_t *src2,
int dstStride, int src1Stride, int h);
void ff_put_pixels16_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
void ff_put_pixels16_l2_mmxext(uint8_t *dst, const uint8_t *src1, const uint8_t *src2,
int dstStride, int src1Stride, int h);
void ff_avg_pixels16_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
void ff_avg_pixels16_l2_mmxext(uint8_t *dst, const uint8_t *src1, const uint8_t *src2,
int dstStride, int src1Stride, int h);
#define ff_put_pixels8_l2_sse2 ff_put_pixels8_l2_mmxext
#define ff_avg_pixels8_l2_sse2 ff_avg_pixels8_l2_mmxext
@@ -54,29 +54,29 @@ CALL_2X_PIXELS(ff_avg_pixels16_mmxext, ff_avg_pixels8_mmxext, 8)
CALL_2X_PIXELS(ff_put_pixels16_mmxext, ff_put_pixels8_mmxext, 8)

#define DEF_QPEL(OPNAME)\
void ff_ ## OPNAME ## _h264_qpel4_h_lowpass_mmxext(uint8_t *dst, uint8_t *src, int dstStride, int srcStride);\
void ff_ ## OPNAME ## _h264_qpel8_h_lowpass_mmxext(uint8_t *dst, uint8_t *src, int dstStride, int srcStride);\
void ff_ ## OPNAME ## _h264_qpel8_h_lowpass_ssse3(uint8_t *dst, uint8_t *src, int dstStride, int srcStride);\
void ff_ ## OPNAME ## _h264_qpel4_h_lowpass_l2_mmxext(uint8_t *dst, uint8_t *src, uint8_t *src2, int dstStride, int src2Stride);\
void ff_ ## OPNAME ## _h264_qpel8_h_lowpass_l2_mmxext(uint8_t *dst, uint8_t *src, uint8_t *src2, int dstStride, int src2Stride);\
void ff_ ## OPNAME ## _h264_qpel8_h_lowpass_l2_ssse3(uint8_t *dst, uint8_t *src, uint8_t *src2, int dstStride, int src2Stride);\
void ff_ ## OPNAME ## _h264_qpel4_v_lowpass_mmxext(uint8_t *dst, uint8_t *src, int dstStride, int srcStride);\
void ff_ ## OPNAME ## _h264_qpel8or16_v_lowpass_op_mmxext(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h);\
void ff_ ## OPNAME ## _h264_qpel8or16_v_lowpass_sse2(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h);\
void ff_ ## OPNAME ## _h264_qpel4_hv_lowpass_v_mmxext(uint8_t *src, int16_t *tmp, int srcStride);\
void ff_ ## OPNAME ## _h264_qpel4_h_lowpass_mmxext(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride);\
void ff_ ## OPNAME ## _h264_qpel8_h_lowpass_mmxext(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride);\
void ff_ ## OPNAME ## _h264_qpel8_h_lowpass_ssse3(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride);\
void ff_ ## OPNAME ## _h264_qpel4_h_lowpass_l2_mmxext(uint8_t *dst, const uint8_t *src, const uint8_t *src2, int dstStride, int src2Stride);\
void ff_ ## OPNAME ## _h264_qpel8_h_lowpass_l2_mmxext(uint8_t *dst, const uint8_t *src, const uint8_t *src2, int dstStride, int src2Stride);\
void ff_ ## OPNAME ## _h264_qpel8_h_lowpass_l2_ssse3(uint8_t *dst, const uint8_t *src, const uint8_t *src2, int dstStride, int src2Stride);\
void ff_ ## OPNAME ## _h264_qpel4_v_lowpass_mmxext(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride);\
void ff_ ## OPNAME ## _h264_qpel8or16_v_lowpass_op_mmxext(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h);\
void ff_ ## OPNAME ## _h264_qpel8or16_v_lowpass_sse2(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h);\
void ff_ ## OPNAME ## _h264_qpel4_hv_lowpass_v_mmxext(const uint8_t *src, int16_t *tmp, int srcStride);\
void ff_ ## OPNAME ## _h264_qpel4_hv_lowpass_h_mmxext(int16_t *tmp, uint8_t *dst, int dstStride);\
void ff_ ## OPNAME ## _h264_qpel8or16_hv1_lowpass_op_mmxext(uint8_t *src, int16_t *tmp, int srcStride, int size);\
void ff_ ## OPNAME ## _h264_qpel8or16_hv1_lowpass_op_sse2(uint8_t *src, int16_t *tmp, int srcStride, int size);\
void ff_ ## OPNAME ## _h264_qpel8or16_hv1_lowpass_op_mmxext(const uint8_t *src, int16_t *tmp, int srcStride, int size);\
void ff_ ## OPNAME ## _h264_qpel8or16_hv1_lowpass_op_sse2(const uint8_t *src, int16_t *tmp, int srcStride, int size);\
void ff_ ## OPNAME ## _h264_qpel8or16_hv2_lowpass_op_mmxext(uint8_t *dst, int16_t *tmp, int dstStride, int unused, int h);\
void ff_ ## OPNAME ## _h264_qpel8or16_hv2_lowpass_ssse3(uint8_t *dst, int16_t *tmp, int dstStride, int tmpStride, int size);\
void ff_ ## OPNAME ## _pixels4_l2_shift5_mmxext(uint8_t *dst, int16_t *src16, uint8_t *src8, int dstStride, int src8Stride, int h);\
void ff_ ## OPNAME ## _pixels8_l2_shift5_mmxext(uint8_t *dst, int16_t *src16, uint8_t *src8, int dstStride, int src8Stride, int h);
void ff_ ## OPNAME ## _pixels4_l2_shift5_mmxext(uint8_t *dst, const int16_t *src16, const uint8_t *src8, int dstStride, int src8Stride, int h);\
void ff_ ## OPNAME ## _pixels8_l2_shift5_mmxext(uint8_t *dst, const int16_t *src16, const uint8_t *src8, int dstStride, int src8Stride, int h);

DEF_QPEL(avg)
DEF_QPEL(put)

#define QPEL_H264(OPNAME, OP, MMX)\
static av_always_inline void ff_ ## OPNAME ## h264_qpel4_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel4_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, const uint8_t *src, int dstStride, int tmpStride, int srcStride){\
int w=3;\
src -= 2*srcStride+2;\
while(w--){\
@@ -88,14 +88,14 @@ static av_always_inline void ff_ ## OPNAME ## h264_qpel4_hv_lowpass_ ## MMX(uint
ff_ ## OPNAME ## h264_qpel4_hv_lowpass_h_mmxext(tmp, dst, dstStride);\
}\
\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8or16_v_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8or16_v_lowpass_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride, int h){\
src -= 2*srcStride;\
ff_ ## OPNAME ## h264_qpel8or16_v_lowpass_op_mmxext(dst, src, dstStride, srcStride, h);\
src += 4;\
dst += 4;\
ff_ ## OPNAME ## h264_qpel8or16_v_lowpass_op_mmxext(dst, src, dstStride, srcStride, h);\
}\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8or16_hv1_lowpass_ ## MMX(int16_t *tmp, uint8_t *src, int tmpStride, int srcStride, int size){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8or16_hv1_lowpass_ ## MMX(int16_t *tmp, const uint8_t *src, int tmpStride, int srcStride, int size){\
int w = (size+8)>>2;\
src -= 2*srcStride+2;\
while(w--){\
@@ -113,15 +113,15 @@ static av_always_inline void ff_ ## OPNAME ## h264_qpel8or16_hv2_lowpass_ ## MMX
}while(w--);\
}\
\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8_v_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8_v_lowpass_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
ff_ ## OPNAME ## h264_qpel8or16_v_lowpass_ ## MMX(dst , src , dstStride, srcStride, 8);\
}\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_v_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_v_lowpass_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
ff_ ## OPNAME ## h264_qpel8or16_v_lowpass_ ## MMX(dst , src , dstStride, srcStride, 16);\
ff_ ## OPNAME ## h264_qpel8or16_v_lowpass_ ## MMX(dst+8, src+8, dstStride, srcStride, 16);\
}\
\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_h_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_h_lowpass_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
ff_ ## OPNAME ## h264_qpel8_h_lowpass_ ## MMX(dst , src , dstStride, srcStride);\
ff_ ## OPNAME ## h264_qpel8_h_lowpass_ ## MMX(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\
@@ -130,7 +130,7 @@ static av_always_inline void ff_ ## OPNAME ## h264_qpel16_h_lowpass_ ## MMX(uint
ff_ ## OPNAME ## h264_qpel8_h_lowpass_ ## MMX(dst+8, src+8, dstStride, srcStride);\
}\
\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_h_lowpass_l2_ ## MMX(uint8_t *dst, uint8_t *src, uint8_t *src2, int dstStride, int src2Stride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_h_lowpass_l2_ ## MMX(uint8_t *dst, const uint8_t *src, const uint8_t *src2, int dstStride, int src2Stride){\
ff_ ## OPNAME ## h264_qpel8_h_lowpass_l2_ ## MMX(dst , src , src2 , dstStride, src2Stride);\
ff_ ## OPNAME ## h264_qpel8_h_lowpass_l2_ ## MMX(dst+8, src+8, src2+8, dstStride, src2Stride);\
src += 8*dstStride;\
@@ -140,19 +140,19 @@ static av_always_inline void ff_ ## OPNAME ## h264_qpel16_h_lowpass_l2_ ## MMX(u
ff_ ## OPNAME ## h264_qpel8_h_lowpass_l2_ ## MMX(dst+8, src+8, src2+8, dstStride, src2Stride);\
}\
\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8or16_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride, int size){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8or16_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, const uint8_t *src, int dstStride, int tmpStride, int srcStride, int size){\
ff_put_h264_qpel8or16_hv1_lowpass_ ## MMX(tmp, src, tmpStride, srcStride, size);\
ff_ ## OPNAME ## h264_qpel8or16_hv2_lowpass_ ## MMX(dst, tmp, dstStride, tmpStride, size);\
}\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, const uint8_t *src, int dstStride, int tmpStride, int srcStride){\
ff_ ## OPNAME ## h264_qpel8or16_hv_lowpass_ ## MMX(dst , tmp , src , dstStride, tmpStride, srcStride, 8);\
}\
\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, const uint8_t *src, int dstStride, int tmpStride, int srcStride){\
ff_ ## OPNAME ## h264_qpel8or16_hv_lowpass_ ## MMX(dst , tmp , src , dstStride, tmpStride, srcStride, 16);\
}\
\
static av_always_inline void ff_ ## OPNAME ## pixels16_l2_shift5_ ## MMX(uint8_t *dst, int16_t *src16, uint8_t *src8, int dstStride, int src8Stride, int h)\
static av_always_inline void ff_ ## OPNAME ## pixels16_l2_shift5_ ## MMX(uint8_t *dst, const int16_t *src16, const uint8_t *src8, int dstStride, int src8Stride, int h)\
{\
ff_ ## OPNAME ## pixels8_l2_shift5_ ## MMX(dst , src16 , src8 , dstStride, src8Stride, h);\
ff_ ## OPNAME ## pixels8_l2_shift5_ ## MMX(dst+8, src16+8, src8+8, dstStride, src8Stride, h);\
@@ -162,12 +162,12 @@ static av_always_inline void ff_ ## OPNAME ## pixels16_l2_shift5_ ## MMX(uint8_t
#if ARCH_X86_64
#define QPEL_H264_H16_XMM(OPNAME, OP, MMX)\

void ff_avg_h264_qpel16_h_lowpass_l2_ssse3(uint8_t *dst, uint8_t *src, uint8_t *src2, int dstStride, int src2Stride);
void ff_put_h264_qpel16_h_lowpass_l2_ssse3(uint8_t *dst, uint8_t *src, uint8_t *src2, int dstStride, int src2Stride);
void ff_avg_h264_qpel16_h_lowpass_l2_ssse3(uint8_t *dst, const uint8_t *src, const uint8_t *src2, int dstStride, int src2Stride);
void ff_put_h264_qpel16_h_lowpass_l2_ssse3(uint8_t *dst, const uint8_t *src, const uint8_t *src2, int dstStride, int src2Stride);

#else // ARCH_X86_64
#define QPEL_H264_H16_XMM(OPNAME, OP, MMX)\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_h_lowpass_l2_ ## MMX(uint8_t *dst, uint8_t *src, uint8_t *src2, int dstStride, int src2Stride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_h_lowpass_l2_ ## MMX(uint8_t *dst, const uint8_t *src, const uint8_t *src2, int dstStride, int src2Stride){\
ff_ ## OPNAME ## h264_qpel8_h_lowpass_l2_ ## MMX(dst , src , src2 , dstStride, src2Stride);\
ff_ ## OPNAME ## h264_qpel8_h_lowpass_l2_ ## MMX(dst+8, src+8, src2+8, dstStride, src2Stride);\
src += 8*dstStride;\
@@ -180,7 +180,7 @@ static av_always_inline void ff_ ## OPNAME ## h264_qpel16_h_lowpass_l2_ ## MMX(u

#define QPEL_H264_H_XMM(OPNAME, OP, MMX)\
QPEL_H264_H16_XMM(OPNAME, OP, MMX)\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_h_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_h_lowpass_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
ff_ ## OPNAME ## h264_qpel8_h_lowpass_ ## MMX(dst , src , dstStride, srcStride);\
ff_ ## OPNAME ## h264_qpel8_h_lowpass_ ## MMX(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\
@@ -190,16 +190,16 @@ static av_always_inline void ff_ ## OPNAME ## h264_qpel16_h_lowpass_ ## MMX(uint
}\

#define QPEL_H264_V_XMM(OPNAME, OP, MMX)\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8_v_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8_v_lowpass_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
ff_ ## OPNAME ## h264_qpel8or16_v_lowpass_ ## MMX(dst , src , dstStride, srcStride, 8);\
}\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_v_lowpass_ ## MMX(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_v_lowpass_ ## MMX(uint8_t *dst, const uint8_t *src, int dstStride, int srcStride){\
ff_ ## OPNAME ## h264_qpel8or16_v_lowpass_ ## MMX(dst , src , dstStride, srcStride, 16);\
ff_ ## OPNAME ## h264_qpel8or16_v_lowpass_ ## MMX(dst+8, src+8, dstStride, srcStride, 16);\
}

static av_always_inline void put_h264_qpel8or16_hv1_lowpass_sse2(int16_t *tmp,
uint8_t *src,
const uint8_t *src,
int tmpStride,
int srcStride,
int size)
@@ -214,14 +214,14 @@ static av_always_inline void put_h264_qpel8or16_hv1_lowpass_sse2(int16_t *tmp,
}

#define QPEL_H264_HV_XMM(OPNAME, OP, MMX)\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8or16_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride, int size){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8or16_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, const uint8_t *src, int dstStride, int tmpStride, int srcStride, int size){\
put_h264_qpel8or16_hv1_lowpass_sse2(tmp, src, tmpStride, srcStride, size);\
ff_ ## OPNAME ## h264_qpel8or16_hv2_lowpass_ ## MMX(dst, tmp, dstStride, tmpStride, size);\
}\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel8_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, const uint8_t *src, int dstStride, int tmpStride, int srcStride){\
ff_ ## OPNAME ## h264_qpel8or16_hv_lowpass_ ## MMX(dst, tmp, src, dstStride, tmpStride, srcStride, 8);\
}\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, uint8_t *src, int dstStride, int tmpStride, int srcStride){\
static av_always_inline void ff_ ## OPNAME ## h264_qpel16_hv_lowpass_ ## MMX(uint8_t *dst, int16_t *tmp, const uint8_t *src, int dstStride, int tmpStride, int srcStride){\
ff_ ## OPNAME ## h264_qpel8or16_hv_lowpass_ ## MMX(dst, tmp, src, dstStride, tmpStride, srcStride, 16);\
}\

@@ -244,12 +244,12 @@ H264_MC_V(OPNAME, SIZE, MMX, ALIGN)\
H264_MC_H(OPNAME, SIZE, MMX, ALIGN)\
H264_MC_HV(OPNAME, SIZE, MMX, ALIGN)\

static void put_h264_qpel16_mc00_sse2 (uint8_t *dst, uint8_t *src,
static void put_h264_qpel16_mc00_sse2 (uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
ff_put_pixels16_sse2(dst, src, stride, 16);
}
static void avg_h264_qpel16_mc00_sse2 (uint8_t *dst, uint8_t *src,
static void avg_h264_qpel16_mc00_sse2 (uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
ff_avg_pixels16_sse2(dst, src, stride, 16);
@@ -258,41 +258,41 @@ static void avg_h264_qpel16_mc00_sse2 (uint8_t *dst, uint8_t *src,
#define avg_h264_qpel8_mc00_sse2 avg_h264_qpel8_mc00_mmxext

#define H264_MC_C(OPNAME, SIZE, MMX, ALIGN) \
static void OPNAME ## h264_qpel ## SIZE ## _mc00_ ## MMX (uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc00_ ## MMX (uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
ff_ ## OPNAME ## pixels ## SIZE ## _ ## MMX(dst, src, stride, SIZE);\
}\

#define H264_MC_H(OPNAME, SIZE, MMX, ALIGN) \
static void OPNAME ## h264_qpel ## SIZE ## _mc10_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc10_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
ff_ ## OPNAME ## h264_qpel ## SIZE ## _h_lowpass_l2_ ## MMX(dst, src, src, stride, stride);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc20_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc20_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
ff_ ## OPNAME ## h264_qpel ## SIZE ## _h_lowpass_ ## MMX(dst, src, stride, stride);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc30_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc30_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
ff_ ## OPNAME ## h264_qpel ## SIZE ## _h_lowpass_l2_ ## MMX(dst, src, src+1, stride, stride);\
}\

#define H264_MC_V(OPNAME, SIZE, MMX, ALIGN) \
static void OPNAME ## h264_qpel ## SIZE ## _mc01_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc01_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(ALIGN, uint8_t, temp)[SIZE*SIZE];\
ff_put_h264_qpel ## SIZE ## _v_lowpass_ ## MMX(temp, src, SIZE, stride);\
ff_ ## OPNAME ## pixels ## SIZE ## _l2_ ## MMX(dst, src, temp, stride, stride, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc02_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc02_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
ff_ ## OPNAME ## h264_qpel ## SIZE ## _v_lowpass_ ## MMX(dst, src, stride, stride);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc03_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc03_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(ALIGN, uint8_t, temp)[SIZE*SIZE];\
ff_put_h264_qpel ## SIZE ## _v_lowpass_ ## MMX(temp, src, SIZE, stride);\
@@ -300,41 +300,41 @@ static void OPNAME ## h264_qpel ## SIZE ## _mc03_ ## MMX(uint8_t *dst, uint8_t *
}\

#define H264_MC_HV(OPNAME, SIZE, MMX, ALIGN) \
static void OPNAME ## h264_qpel ## SIZE ## _mc11_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc11_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(ALIGN, uint8_t, temp)[SIZE*SIZE];\
ff_put_h264_qpel ## SIZE ## _v_lowpass_ ## MMX(temp, src, SIZE, stride);\
ff_ ## OPNAME ## h264_qpel ## SIZE ## _h_lowpass_l2_ ## MMX(dst, src, temp, stride, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc31_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc31_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(ALIGN, uint8_t, temp)[SIZE*SIZE];\
ff_put_h264_qpel ## SIZE ## _v_lowpass_ ## MMX(temp, src+1, SIZE, stride);\
ff_ ## OPNAME ## h264_qpel ## SIZE ## _h_lowpass_l2_ ## MMX(dst, src, temp, stride, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc13_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc13_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(ALIGN, uint8_t, temp)[SIZE*SIZE];\
ff_put_h264_qpel ## SIZE ## _v_lowpass_ ## MMX(temp, src, SIZE, stride);\
ff_ ## OPNAME ## h264_qpel ## SIZE ## _h_lowpass_l2_ ## MMX(dst, src+stride, temp, stride, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc33_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc33_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(ALIGN, uint8_t, temp)[SIZE*SIZE];\
ff_put_h264_qpel ## SIZE ## _v_lowpass_ ## MMX(temp, src+1, SIZE, stride);\
ff_ ## OPNAME ## h264_qpel ## SIZE ## _h_lowpass_l2_ ## MMX(dst, src+stride, temp, stride, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc22_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc22_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(ALIGN, uint16_t, temp)[SIZE*(SIZE<8?12:24)];\
ff_ ## OPNAME ## h264_qpel ## SIZE ## _hv_lowpass_ ## MMX(dst, temp, src, stride, SIZE, stride);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc21_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc21_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(ALIGN, uint8_t, temp)[SIZE*(SIZE<8?12:24)*2 + SIZE*SIZE];\
uint8_t * const halfHV= temp;\
@@ -344,7 +344,7 @@ static void OPNAME ## h264_qpel ## SIZE ## _mc21_ ## MMX(uint8_t *dst, uint8_t *
ff_ ## OPNAME ## h264_qpel ## SIZE ## _h_lowpass_l2_ ## MMX(dst, src, halfHV, stride, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc23_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc23_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(ALIGN, uint8_t, temp)[SIZE*(SIZE<8?12:24)*2 + SIZE*SIZE];\
uint8_t * const halfHV= temp;\
@@ -354,7 +354,7 @@ static void OPNAME ## h264_qpel ## SIZE ## _mc23_ ## MMX(uint8_t *dst, uint8_t *
ff_ ## OPNAME ## h264_qpel ## SIZE ## _h_lowpass_l2_ ## MMX(dst, src+stride, halfHV, stride, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc12_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc12_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(ALIGN, uint8_t, temp)[SIZE*(SIZE<8?12:24)*2 + SIZE*SIZE];\
uint8_t * const halfHV= temp;\
@@ -364,7 +364,7 @@ static void OPNAME ## h264_qpel ## SIZE ## _mc12_ ## MMX(uint8_t *dst, uint8_t *
ff_ ## OPNAME ## pixels ## SIZE ## _l2_shift5_mmxext(dst, halfV+2, halfHV, stride, SIZE, SIZE);\
}\
\
static void OPNAME ## h264_qpel ## SIZE ## _mc32_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride)\
static void OPNAME ## h264_qpel ## SIZE ## _mc32_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride)\
{\
DECLARE_ALIGNED(ALIGN, uint8_t, temp)[SIZE*(SIZE<8?12:24)*2 + SIZE*SIZE];\
uint8_t * const halfHV= temp;\
@@ -409,7 +409,7 @@ H264_MC_816(H264_MC_HV, ssse3)
//10bit
#define LUMA_MC_OP(OP, NUM, DEPTH, TYPE, OPT) \
void ff_ ## OP ## _h264_qpel ## NUM ## _ ## TYPE ## _ ## DEPTH ## _ ## OPT \
(uint8_t *dst, uint8_t *src, ptrdiff_t stride);
(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);

#define LUMA_MC_ALL(DEPTH, TYPE, OPT) \
LUMA_MC_OP(put, 4, DEPTH, TYPE, OPT) \
@@ -466,7 +466,7 @@ LUMA_MC_816(10, mc23, sse2)
LUMA_MC_816(10, mc33, sse2)

#define QPEL16_OPMC(OP, MC, MMX)\
void ff_ ## OP ## _h264_qpel16_ ## MC ## _10_ ## MMX(uint8_t *dst, uint8_t *src, ptrdiff_t stride){\
void ff_ ## OP ## _h264_qpel16_ ## MC ## _10_ ## MMX(uint8_t *dst, const uint8_t *src, ptrdiff_t stride){\
ff_ ## OP ## _h264_qpel8_ ## MC ## _10_ ## MMX(dst , src , stride);\
ff_ ## OP ## _h264_qpel8_ ## MC ## _10_ ## MMX(dst+16, src+16, stride);\
src += 8*stride;\


+ 93
- 52
libavcodec/x86/qpeldsp_init.c View File

@@ -29,44 +29,53 @@
#include "libavcodec/qpeldsp.h"
#include "fpel.h"

void ff_put_pixels8_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
void ff_put_pixels8_l2_mmxext(uint8_t *dst,
const uint8_t *src1, const uint8_t *src2,
int dstStride, int src1Stride, int h);
void ff_put_no_rnd_pixels8_l2_mmxext(uint8_t *dst, uint8_t *src1,
uint8_t *src2, int dstStride,
int src1Stride, int h);
void ff_avg_pixels8_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
void ff_put_no_rnd_pixels8_l2_mmxext(uint8_t *dst,
const uint8_t *src1, const uint8_t *src2,
int dstStride, int src1Stride, int h);
void ff_avg_pixels8_l2_mmxext(uint8_t *dst,
const uint8_t *src1, const uint8_t *src2,
int dstStride, int src1Stride, int h);
void ff_put_pixels16_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
void ff_put_pixels16_l2_mmxext(uint8_t *dst,
const uint8_t *src1, const uint8_t *src2,
int dstStride, int src1Stride, int h);
void ff_avg_pixels16_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
void ff_avg_pixels16_l2_mmxext(uint8_t *dst,
const uint8_t *src1, const uint8_t *src2,
int dstStride, int src1Stride, int h);
void ff_put_no_rnd_pixels16_l2_mmxext(uint8_t *dst, uint8_t *src1, uint8_t *src2,
void ff_put_no_rnd_pixels16_l2_mmxext(uint8_t *dst,
const uint8_t *src1, const uint8_t *src2,
int dstStride, int src1Stride, int h);
void ff_put_mpeg4_qpel16_h_lowpass_mmxext(uint8_t *dst, uint8_t *src,
void ff_put_mpeg4_qpel16_h_lowpass_mmxext(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride, int h);
void ff_avg_mpeg4_qpel16_h_lowpass_mmxext(uint8_t *dst, uint8_t *src,
void ff_avg_mpeg4_qpel16_h_lowpass_mmxext(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride, int h);
void ff_put_no_rnd_mpeg4_qpel16_h_lowpass_mmxext(uint8_t *dst, uint8_t *src,
void ff_put_no_rnd_mpeg4_qpel16_h_lowpass_mmxext(uint8_t *dst,
const uint8_t *src,
int dstStride, int srcStride,
int h);
void ff_put_mpeg4_qpel8_h_lowpass_mmxext(uint8_t *dst, uint8_t *src,
void ff_put_mpeg4_qpel8_h_lowpass_mmxext(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride, int h);
void ff_avg_mpeg4_qpel8_h_lowpass_mmxext(uint8_t *dst, uint8_t *src,
void ff_avg_mpeg4_qpel8_h_lowpass_mmxext(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride, int h);
void ff_put_no_rnd_mpeg4_qpel8_h_lowpass_mmxext(uint8_t *dst, uint8_t *src,
void ff_put_no_rnd_mpeg4_qpel8_h_lowpass_mmxext(uint8_t *dst,
const uint8_t *src,
int dstStride, int srcStride,
int h);
void ff_put_mpeg4_qpel16_v_lowpass_mmxext(uint8_t *dst, uint8_t *src,
void ff_put_mpeg4_qpel16_v_lowpass_mmxext(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride);
void ff_avg_mpeg4_qpel16_v_lowpass_mmxext(uint8_t *dst, uint8_t *src,
void ff_avg_mpeg4_qpel16_v_lowpass_mmxext(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride);
void ff_put_no_rnd_mpeg4_qpel16_v_lowpass_mmxext(uint8_t *dst, uint8_t *src,
void ff_put_no_rnd_mpeg4_qpel16_v_lowpass_mmxext(uint8_t *dst,
const uint8_t *src,
int dstStride, int srcStride);
void ff_put_mpeg4_qpel8_v_lowpass_mmxext(uint8_t *dst, uint8_t *src,
void ff_put_mpeg4_qpel8_v_lowpass_mmxext(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride);
void ff_avg_mpeg4_qpel8_v_lowpass_mmxext(uint8_t *dst, uint8_t *src,
void ff_avg_mpeg4_qpel8_v_lowpass_mmxext(uint8_t *dst, const uint8_t *src,
int dstStride, int srcStride);
void ff_put_no_rnd_mpeg4_qpel8_v_lowpass_mmxext(uint8_t *dst, uint8_t *src,
void ff_put_no_rnd_mpeg4_qpel8_v_lowpass_mmxext(uint8_t *dst,
const uint8_t *src,
int dstStride, int srcStride);
#define ff_put_no_rnd_pixels16_mmxext ff_put_pixels16_mmxext
#define ff_put_no_rnd_pixels8_mmxext ff_put_pixels8_mmxext
@@ -77,13 +86,15 @@ CALL_2X_PIXELS(ff_avg_pixels16_mmxext, ff_avg_pixels8_mmxext, 8)
CALL_2X_PIXELS(ff_put_pixels16_mmxext, ff_put_pixels8_mmxext, 8)

#define QPEL_OP(OPNAME, RND, MMX) \
static void OPNAME ## qpel8_mc00_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc00_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
ff_ ## OPNAME ## pixels8_ ## MMX(dst, src, stride, 8); \
} \
\
static void OPNAME ## qpel8_mc10_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc10_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t temp[8]; \
@@ -94,14 +105,16 @@ static void OPNAME ## qpel8_mc10_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, stride, 8); \
} \
\
static void OPNAME ## qpel8_mc20_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc20_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
ff_ ## OPNAME ## mpeg4_qpel8_h_lowpass_ ## MMX(dst, src, stride, \
stride, 8); \
} \
\
static void OPNAME ## qpel8_mc30_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc30_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t temp[8]; \
@@ -112,7 +125,8 @@ static void OPNAME ## qpel8_mc30_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 8); \
} \
\
static void OPNAME ## qpel8_mc01_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc01_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t temp[8]; \
@@ -123,14 +137,16 @@ static void OPNAME ## qpel8_mc01_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, stride, 8); \
} \
\
static void OPNAME ## qpel8_mc02_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc02_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
ff_ ## OPNAME ## mpeg4_qpel8_v_lowpass_ ## MMX(dst, src, \
stride, stride); \
} \
\
static void OPNAME ## qpel8_mc03_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc03_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t temp[8]; \
@@ -141,7 +157,8 @@ static void OPNAME ## qpel8_mc03_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 8); \
} \
\
static void OPNAME ## qpel8_mc11_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc11_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[8 + 9]; \
@@ -156,7 +173,8 @@ static void OPNAME ## qpel8_mc11_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc31_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc31_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[8 + 9]; \
@@ -171,7 +189,8 @@ static void OPNAME ## qpel8_mc31_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc13_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc13_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[8 + 9]; \
@@ -186,7 +205,8 @@ static void OPNAME ## qpel8_mc13_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc33_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc33_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[8 + 9]; \
@@ -201,7 +221,8 @@ static void OPNAME ## qpel8_mc33_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc21_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc21_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[8 + 9]; \
@@ -214,7 +235,8 @@ static void OPNAME ## qpel8_mc21_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc23_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc23_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[8 + 9]; \
@@ -227,7 +249,8 @@ static void OPNAME ## qpel8_mc23_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 8, 8); \
} \
\
static void OPNAME ## qpel8_mc12_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc12_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[8 + 9]; \
@@ -240,7 +263,8 @@ static void OPNAME ## qpel8_mc12_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 8); \
} \
\
static void OPNAME ## qpel8_mc32_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc32_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[8 + 9]; \
@@ -253,7 +277,8 @@ static void OPNAME ## qpel8_mc32_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 8); \
} \
\
static void OPNAME ## qpel8_mc22_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel8_mc22_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[9]; \
@@ -264,13 +289,15 @@ static void OPNAME ## qpel8_mc22_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 8); \
} \
\
static void OPNAME ## qpel16_mc00_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc00_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
ff_ ## OPNAME ## pixels16_ ## MMX(dst, src, stride, 16); \
} \
\
static void OPNAME ## qpel16_mc10_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc10_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t temp[32]; \
@@ -281,14 +308,16 @@ static void OPNAME ## qpel16_mc10_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 16); \
} \
\
static void OPNAME ## qpel16_mc20_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc20_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
ff_ ## OPNAME ## mpeg4_qpel16_h_lowpass_ ## MMX(dst, src, \
stride, stride, 16);\
} \
\
static void OPNAME ## qpel16_mc30_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc30_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t temp[32]; \
@@ -299,7 +328,8 @@ static void OPNAME ## qpel16_mc30_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, stride, 16); \
} \
\
static void OPNAME ## qpel16_mc01_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc01_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t temp[32]; \
@@ -310,14 +340,16 @@ static void OPNAME ## qpel16_mc01_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 16); \
} \
\
static void OPNAME ## qpel16_mc02_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc02_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
ff_ ## OPNAME ## mpeg4_qpel16_v_lowpass_ ## MMX(dst, src, \
stride, stride); \
} \
\
static void OPNAME ## qpel16_mc03_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc03_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t temp[32]; \
@@ -328,7 +360,8 @@ static void OPNAME ## qpel16_mc03_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, stride, 16); \
} \
\
static void OPNAME ## qpel16_mc11_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc11_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[16 * 2 + 17 * 2]; \
@@ -344,7 +377,8 @@ static void OPNAME ## qpel16_mc11_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc31_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc31_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[16 * 2 + 17 * 2]; \
@@ -360,7 +394,8 @@ static void OPNAME ## qpel16_mc31_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc13_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc13_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[16 * 2 + 17 * 2]; \
@@ -376,7 +411,8 @@ static void OPNAME ## qpel16_mc13_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc33_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc33_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[16 * 2 + 17 * 2]; \
@@ -392,7 +428,8 @@ static void OPNAME ## qpel16_mc33_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc21_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc21_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[16 * 2 + 17 * 2]; \
@@ -406,7 +443,8 @@ static void OPNAME ## qpel16_mc21_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc23_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc23_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[16 * 2 + 17 * 2]; \
@@ -420,7 +458,8 @@ static void OPNAME ## qpel16_mc23_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 16, 16); \
} \
\
static void OPNAME ## qpel16_mc12_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc12_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[17 * 2]; \
@@ -433,7 +472,8 @@ static void OPNAME ## qpel16_mc12_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 16); \
} \
\
static void OPNAME ## qpel16_mc32_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc32_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[17 * 2]; \
@@ -446,7 +486,8 @@ static void OPNAME ## qpel16_mc32_ ## MMX(uint8_t *dst, uint8_t *src, \
stride, 16); \
} \
\
static void OPNAME ## qpel16_mc22_ ## MMX(uint8_t *dst, uint8_t *src, \
static void OPNAME ## qpel16_mc22_ ## MMX(uint8_t *dst, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
uint64_t half[17 * 2]; \


+ 5
- 5
libavcodec/x86/rv40dsp_init.c View File

@@ -70,7 +70,7 @@ DECLARE_WEIGHT(ssse3)
*/
#define QPEL_FUNC_DECL(OP, SIZE, PH, PV, OPT) \
static void OP ## rv40_qpel ##SIZE ##_mc ##PH ##PV ##OPT(uint8_t *dst, \
uint8_t *src, \
const uint8_t *src, \
ptrdiff_t stride) \
{ \
int i; \
@@ -189,22 +189,22 @@ QPEL_FUNCS_SET (OP, 3, 2, OPT)
#endif /* HAVE_YASM */

#if HAVE_MMX_INLINE
static void put_rv40_qpel8_mc33_mmx(uint8_t *dst, uint8_t *src,
static void put_rv40_qpel8_mc33_mmx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
ff_put_pixels8_xy2_mmx(dst, src, stride, 8);
}
static void put_rv40_qpel16_mc33_mmx(uint8_t *dst, uint8_t *src,
static void put_rv40_qpel16_mc33_mmx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
ff_put_pixels16_xy2_mmx(dst, src, stride, 16);
}
static void avg_rv40_qpel8_mc33_mmx(uint8_t *dst, uint8_t *src,
static void avg_rv40_qpel8_mc33_mmx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
ff_avg_pixels8_xy2_mmx(dst, src, stride, 8);
}
static void avg_rv40_qpel16_mc33_mmx(uint8_t *dst, uint8_t *src,
static void avg_rv40_qpel16_mc33_mmx(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride)
{
ff_avg_pixels16_xy2_mmx(dst, src, stride, 16);


Write Preview
Loading…
Cancel
Save