Commit 2904d052 authored by Michael Niedermayer's avatar Michael Niedermayer

Merge commit '7fb993d3'

* commit '7fb993d3':
  qpeldsp: Mark source pointer in qpel_mc_func function pointer const

Conflicts:
	libavcodec/h264qpel_template.c
	libavcodec/x86/cavsdsp.c
	libavcodec/x86/rv40dsp_init.c
Merged-by: 's avatarMichael Niedermayer <michaelni@gmx.at>
parents 937eaba2 7fb993d3
...@@ -26,73 +26,73 @@ ...@@ -26,73 +26,73 @@
#include "libavutil/arm/cpu.h" #include "libavutil/arm/cpu.h"
#include "libavcodec/h264qpel.h" #include "libavcodec/h264qpel.h"
void ff_put_h264_qpel16_mc00_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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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_mc00_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_put_h264_qpel8_mc10_neon(uint8_t *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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_mc00_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel16_mc10_neon(uint8_t *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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_mc00_neon(uint8_t *dst, const uint8_t *src, ptrdiff_t stride);
void ff_avg_h264_qpel8_mc10_neon(uint8_t *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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 *, uint8_t *, ptrdiff_t); 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) av_cold void ff_h264qpel_init_arm(H264QpelContext *c, int bit_depth)
{ {
......
...@@ -26,8 +26,10 @@ ...@@ -26,8 +26,10 @@
#include "libavutil/arm/cpu.h" #include "libavutil/arm/cpu.h"
#define DECL_QPEL3(type, w, pos) \ #define DECL_QPEL3(type, w, pos) \
void ff_##type##_rv40_qpel##w##_mc##pos##_neon(uint8_t *dst, uint8_t *src,\ void ff_ ## type ## _rv40_qpel ## w ## _mc ## pos ## _neon(uint8_t *dst, \
ptrdiff_t stride) const uint8_t *src, \
ptrdiff_t stride)
#define DECL_QPEL2(w, pos) \ #define DECL_QPEL2(w, pos) \
DECL_QPEL3(put, w, pos); \ DECL_QPEL3(put, w, pos); \
DECL_QPEL3(avg, w, pos) DECL_QPEL3(avg, w, pos)
......
...@@ -261,7 +261,7 @@ static void cavs_idct8_add_c(uint8_t *dst, int16_t *block, int stride) { ...@@ -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) \ #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 int h=8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\ const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
int i;\ int i;\
...@@ -280,7 +280,7 @@ static void OPNAME ## cavs_filt8_h_ ## NAME(uint8_t *dst, uint8_t *src, int dstS ...@@ -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 int w=8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\ const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
int i;\ int i;\
...@@ -312,7 +312,7 @@ static void OPNAME ## cavs_filt8_v_ ## NAME(uint8_t *dst, uint8_t *src, int dst ...@@ -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 , src , dstStride, srcStride);\
OPNAME ## cavs_filt8_v_ ## NAME(dst+8, src+8, dstStride, srcStride);\ OPNAME ## cavs_filt8_v_ ## NAME(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\ src += 8*srcStride;\
...@@ -321,7 +321,7 @@ static void OPNAME ## cavs_filt16_v_ ## NAME(uint8_t *dst, uint8_t *src, int dst ...@@ -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);\ 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 , src , dstStride, srcStride);\
OPNAME ## cavs_filt8_h_ ## NAME(dst+8, src+8, dstStride, srcStride);\ OPNAME ## cavs_filt8_h_ ## NAME(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\ src += 8*srcStride;\
...@@ -331,7 +331,7 @@ static void OPNAME ## cavs_filt16_h_ ## NAME(uint8_t *dst, uint8_t *src, int dst ...@@ -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) \ #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 temp[8*(8+5)];\
int16_t *tmp = temp;\ int16_t *tmp = temp;\
const int h=8;\ const int h=8;\
...@@ -412,7 +412,7 @@ static void OPNAME ## cavs_filt8_hv_ ## NAME(uint8_t *dst, uint8_t *src1, uint8_ ...@@ -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 , src1, src2 , dstStride, srcStride); \
OPNAME ## cavs_filt8_hv_ ## NAME(dst+8, src1+8, src2+8, dstStride, srcStride); \ OPNAME ## cavs_filt8_hv_ ## NAME(dst+8, src1+8, src2+8, dstStride, srcStride); \
src1 += 8*srcStride;\ src1 += 8*srcStride;\
...@@ -423,77 +423,77 @@ static void OPNAME ## cavs_filt16_hv_ ## NAME(uint8_t *dst, uint8_t *src1, uint8 ...@@ -423,77 +423,77 @@ static void OPNAME ## cavs_filt16_hv_ ## NAME(uint8_t *dst, uint8_t *src1, uint8
}\ }\
#define CAVS_MC(OPNAME, SIZE) \ #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);\ 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);\ 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);\ 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);\ 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);\ 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);\ 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); \ 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); \ 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); \ 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); \ 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); \ 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); \ 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); \ 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); \ 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); \ OPNAME ## cavs_filt ## SIZE ## _hv_qq(dst, src, src+stride+1,stride, stride); \
}\ }\
......
This diff is collapsed.
...@@ -66,50 +66,50 @@ ...@@ -66,50 +66,50 @@
#undef PREFIX_h264_qpel16_hv_lowpass_num #undef PREFIX_h264_qpel16_hv_lowpass_num
#define H264_MC(OPNAME, SIZE, CODETYPE) \ #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);\ 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];\ DECLARE_ALIGNED(16, uint8_t, half)[SIZE*SIZE];\
put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(half, src, SIZE, stride);\ put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(half, src, SIZE, stride);\
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, src, half, stride, stride, SIZE);\ 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);\ 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];\ DECLARE_ALIGNED(16, uint8_t, half)[SIZE*SIZE];\
put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(half, src, SIZE, stride);\ put_h264_qpel ## SIZE ## _h_lowpass_ ## CODETYPE(half, src, SIZE, stride);\
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, src+1, half, stride, stride, SIZE);\ 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];\ DECLARE_ALIGNED(16, uint8_t, half)[SIZE*SIZE];\
put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(half, src, SIZE, stride);\ put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(half, src, SIZE, stride);\
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, src, half, stride, stride, SIZE);\ 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);\ 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];\ DECLARE_ALIGNED(16, uint8_t, half)[SIZE*SIZE];\
put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(half, src, SIZE, stride);\ put_h264_qpel ## SIZE ## _v_lowpass_ ## CODETYPE(half, src, SIZE, stride);\
OPNAME ## pixels ## SIZE ## _l2_ ## CODETYPE(dst, src+stride, half, stride, stride, SIZE);\ 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, halfH)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfV)[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 ...@@ -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);\ 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, halfH)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfV)[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 ...@@ -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);\ 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, halfH)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfV)[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 ...@@ -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);\ 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, halfH)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfV)[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 ...@@ -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);\ 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)];\ DECLARE_ALIGNED(16, int16_t, tmp)[SIZE*(SIZE+8)];\
OPNAME ## h264_qpel ## SIZE ## _hv_lowpass_ ## CODETYPE(dst, tmp, src, stride, SIZE, stride);\ 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, halfH)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfHV)[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 ...@@ -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);\ 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, halfH)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfHV)[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 ...@@ -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);\ 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, halfV)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfHV)[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 ...@@ -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);\ 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, halfV)[SIZE*SIZE];\
DECLARE_ALIGNED(16, uint8_t, halfHV)[SIZE*SIZE];\ DECLARE_ALIGNED(16, uint8_t, halfHV)[SIZE*SIZE];\
......
...@@ -28,7 +28,10 @@ ...@@ -28,7 +28,10 @@
/* this code assume stride % 16 == 0 */ /* this code assume stride % 16 == 0 */
#ifdef PREFIX_h264_qpel16_h_lowpass_altivec #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; register int i;
LOAD_ZERO; LOAD_ZERO;
...@@ -168,7 +171,10 @@ static void PREFIX_h264_qpel16_h_lowpass_altivec(uint8_t * dst, uint8_t * src, i ...@@ -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 */ /* this code assume stride % 16 == 0 */
#ifdef PREFIX_h264_qpel16_v_lowpass_altivec #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; register int i;
LOAD_ZERO; LOAD_ZERO;
...@@ -178,7 +184,7 @@ static void PREFIX_h264_qpel16_v_lowpass_altivec(uint8_t * dst, uint8_t * src, i ...@@ -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 v5ss = vec_splat_s16(5);
const vec_s16 v16ss = vec_sl(vec_splat_s16(1),vec_splat_u16(4)); 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 srcM2a = vec_ld(0, srcbis);
const vec_u8 srcM2b = vec_ld(16, 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 ...@@ -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 */ /* this code assume stride % 16 == 0 *and* tmp is properly aligned */
#ifdef PREFIX_h264_qpel16_hv_lowpass_altivec #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; register int i;
LOAD_ZERO; LOAD_ZERO;
const vec_u8 permM2 = vec_lvsl(-2, src); const vec_u8 permM2 = vec_lvsl(-2, src);
......
This diff is collapsed.
...@@ -29,22 +29,25 @@ ...@@ -29,22 +29,25 @@
#include <stddef.h> #include <stddef.h>
#include <stdint.h> #include <stdint.h>
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);
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);
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);
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);
void ff_put_pixels8_l2_8(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, 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 dst_stride, int src_stride1, int src_stride2,
int h); int h);
#define DEF_OLD_QPEL(name) \ #define DEF_OLD_QPEL(name) \
void ff_put_ ## name(uint8_t *dst /* align width (8 or 16) */, \ void ff_put_ ## name(uint8_t *dst /* align width (8 or 16) */, \
uint8_t *src /* align 1 */, ptrdiff_t stride); \ const uint8_t *src /* align 1 */, \
void ff_put_no_rnd_ ## name(uint8_t *dst /* align width (8 or 16) */, \ ptrdiff_t stride); \
uint8_t *src /* align 1 */, ptrdiff_t stride); \ void ff_put_no_rnd_ ## name(uint8_t *dst /* align width (8 or 16) */, \
void ff_avg_ ## name(uint8_t *dst /* align width (8 or 16) */, \ const uint8_t *src /* align 1 */, \
uint8_t *src /* align 1 */, ptrdiff_t stride); 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_mc11_old_c)
DEF_OLD_QPEL(qpel16_mc31_old_c) DEF_OLD_QPEL(qpel16_mc31_old_c)
...@@ -60,7 +63,8 @@ DEF_OLD_QPEL(qpel8_mc13_old_c) ...@@ -60,7 +63,8 @@ DEF_OLD_QPEL(qpel8_mc13_old_c)
DEF_OLD_QPEL(qpel8_mc33_old_c) DEF_OLD_QPEL(qpel8_mc33_old_c)
typedef void (*qpel_mc_func)(uint8_t *dst /* align width (8 or 16) */, 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 * quarterpel DSP context
......
...@@ -31,7 +31,7 @@ ...@@ -31,7 +31,7 @@
#include "rv34dsp.h" #include "rv34dsp.h"
#define RV30_LOWPASS(OPNAME, OP) \ #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 int h = 8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\ const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
int i;\ int i;\
...@@ -50,7 +50,7 @@ static av_unused void OPNAME ## rv30_tpel8_h_lowpass(uint8_t *dst, uint8_t *src, ...@@ -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 int w = 8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\ const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
int i;\ int i;\
...@@ -80,7 +80,7 @@ static void OPNAME ## rv30_tpel8_v_lowpass(uint8_t *dst, uint8_t *src, int dstSt ...@@ -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 w = 8;\
const int h = 8;\ const int h = 8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\ 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 ...@@ -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 w = 8;\
const int h = 8;\ const int h = 8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\ 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 ...@@ -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 w = 8;\
const int h = 8;\ const int h = 8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\ 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 ...@@ -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 w = 8;\
const int h = 8;\ const int h = 8;\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\ 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 ...@@ -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 , src , dstStride, srcStride, C1, C2);\
OPNAME ## rv30_tpel8_v_lowpass(dst+8, src+8, dstStride, srcStride, C1, C2);\ OPNAME ## rv30_tpel8_v_lowpass(dst+8, src+8, dstStride, srcStride, C1, C2);\
src += 8*srcStride;\ src += 8*srcStride;\
...@@ -164,7 +164,7 @@ static void OPNAME ## rv30_tpel16_v_lowpass(uint8_t *dst, uint8_t *src, int dstS ...@@ -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);\ 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 , src , dstStride, srcStride, C1, C2);\
OPNAME ## rv30_tpel8_h_lowpass(dst+8, src+8, dstStride, srcStride, C1, C2);\ OPNAME ## rv30_tpel8_h_lowpass(dst+8, src+8, dstStride, srcStride, C1, C2);\
src += 8*srcStride;\ src += 8*srcStride;\
...@@ -173,7 +173,7 @@ static void OPNAME ## rv30_tpel16_h_lowpass(uint8_t *dst, uint8_t *src, int dstS ...@@ -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);\ 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 , src , dstStride, srcStride);\
OPNAME ## rv30_tpel8_hv_lowpass(dst+8, src+8, dstStride, srcStride);\ OPNAME ## rv30_tpel8_hv_lowpass(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\ src += 8*srcStride;\
...@@ -182,7 +182,7 @@ static void OPNAME ## rv30_tpel16_hv_lowpass(uint8_t *dst, uint8_t *src, int dst ...@@ -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);\ 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 , src , dstStride, srcStride);\
OPNAME ## rv30_tpel8_hhv_lowpass(dst+8, src+8, dstStride, srcStride);\ OPNAME ## rv30_tpel8_hhv_lowpass(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\ src += 8*srcStride;\
...@@ -191,7 +191,7 @@ static void OPNAME ## rv30_tpel16_hhv_lowpass(uint8_t *dst, uint8_t *src, int ds ...@@ -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);\ 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 , src , dstStride, srcStride);\
OPNAME ## rv30_tpel8_hvv_lowpass(dst+8, src+8, dstStride, srcStride);\ OPNAME ## rv30_tpel8_hvv_lowpass(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\ src += 8*srcStride;\
...@@ -200,7 +200,7 @@ static void OPNAME ## rv30_tpel16_hvv_lowpass(uint8_t *dst, uint8_t *src, int ds ...@@ -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);\ 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 , src , dstStride, srcStride);\
OPNAME ## rv30_tpel8_hhvv_lowpass(dst+8, src+8, dstStride, srcStride);\ OPNAME ## rv30_tpel8_hhvv_lowpass(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\ src += 8*srcStride;\
...@@ -211,42 +211,42 @@ static void OPNAME ## rv30_tpel16_hhvv_lowpass(uint8_t *dst, uint8_t *src, int d ...@@ -211,42 +211,42 @@ static void OPNAME ## rv30_tpel16_hhvv_lowpass(uint8_t *dst, uint8_t *src, int d
\ \
#define RV30_MC(OPNAME, SIZE) \ #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);\ 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);\ 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);\ 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);\ 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);\ 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);\ 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);\ 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);\ OPNAME ## rv30_tpel ## SIZE ## _hhvv_lowpass(dst, src, stride, stride);\
}\ }\
......
...@@ -35,7 +35,7 @@ ...@@ -35,7 +35,7 @@
#include "libavutil/avassert.h" #include "libavutil/avassert.h"
#define RV40_LOWPASS(OPNAME, OP) \ #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 int h, const int C1, const int C2, const int SHIFT){\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\ const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
int i;\ int i;\
...@@ -54,7 +54,7 @@ static av_unused void OPNAME ## rv40_qpel8_h_lowpass(uint8_t *dst, uint8_t *src, ...@@ -54,7 +54,7 @@ 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 int w, const int C1, const int C2, const int SHIFT){\
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\ const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP;\
int i;\ int i;\
...@@ -86,7 +86,7 @@ static void OPNAME ## rv40_qpel8_v_lowpass(uint8_t *dst, uint8_t *src, int dstSt ...@@ -86,7 +86,7 @@ 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){\ 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 , src , dstStride, srcStride, 8, C1, C2, SHIFT);\
OPNAME ## rv40_qpel8_v_lowpass(dst+8, src+8, dstStride, srcStride, 8, C1, C2, SHIFT);\ OPNAME ## rv40_qpel8_v_lowpass(dst+8, src+8, dstStride, srcStride, 8, C1, C2, SHIFT);\
...@@ -96,7 +96,7 @@ static void OPNAME ## rv40_qpel16_v_lowpass(uint8_t *dst, uint8_t *src, int dstS ...@@ -96,7 +96,7 @@ 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);\ 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){\ 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 , src , dstStride, srcStride, 8, C1, C2, SHIFT);\
OPNAME ## rv40_qpel8_h_lowpass(dst+8, src+8, dstStride, srcStride, 8, C1, C2, SHIFT);\ OPNAME ## rv40_qpel8_h_lowpass(dst+8, src+8, dstStride, srcStride, 8, C1, C2, SHIFT);\
...@@ -108,22 +108,22 @@ static void OPNAME ## rv40_qpel16_h_lowpass(uint8_t *dst, uint8_t *src, int dstS ...@@ -108,22 +108,22 @@ static void OPNAME ## rv40_qpel16_h_lowpass(uint8_t *dst, uint8_t *src, int dstS
\ \
#define RV40_MC(OPNAME, SIZE) \ #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);\ 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);\ 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);\ 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 full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\ uint8_t * const full_mid = full + SIZE*2;\
...@@ -131,7 +131,7 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc11_c(uint8_t *dst, uint8_t *src, p ...@@ -131,7 +131,7 @@ 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);\ 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 full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\ uint8_t * const full_mid = full + SIZE*2;\
...@@ -139,7 +139,7 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc21_c(uint8_t *dst, uint8_t *src, p ...@@ -139,7 +139,7 @@ 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);\ 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 full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\ uint8_t * const full_mid = full + SIZE*2;\
...@@ -147,7 +147,7 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc31_c(uint8_t *dst, uint8_t *src, p ...@@ -147,7 +147,7 @@ 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);\ 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 full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\ uint8_t * const full_mid = full + SIZE*2;\
...@@ -155,7 +155,7 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc12_c(uint8_t *dst, uint8_t *src, p ...@@ -155,7 +155,7 @@ 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);\ 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 full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\ uint8_t * const full_mid = full + SIZE*2;\
...@@ -163,7 +163,7 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc22_c(uint8_t *dst, uint8_t *src, p ...@@ -163,7 +163,7 @@ 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);\ 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 full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\ uint8_t * const full_mid = full + SIZE*2;\
...@@ -171,12 +171,12 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc32_c(uint8_t *dst, uint8_t *src, p ...@@ -171,12 +171,12 @@ 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);\ 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);\ 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 full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\ uint8_t * const full_mid = full + SIZE*2;\
...@@ -184,7 +184,7 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc13_c(uint8_t *dst, uint8_t *src, p ...@@ -184,7 +184,7 @@ 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);\ 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 full[SIZE*(SIZE+5)];\
uint8_t * const full_mid = full + SIZE*2;\ uint8_t * const full_mid = full + SIZE*2;\
...@@ -267,19 +267,19 @@ PIXOP2(put, op_put) ...@@ -267,19 +267,19 @@ PIXOP2(put, op_put)
#undef op_avg #undef op_avg
#undef op_put #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); 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); 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); 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); avg_pixels8_xy2_8_c(dst, src, stride, 8);
} }
......
...@@ -139,7 +139,7 @@ static void wmv2_idct_put_c(uint8_t *dest, int line_size, int16_t *block) ...@@ -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) int dstStride, int srcStride, int h)
{ {
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP; 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, ...@@ -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) int dstStride, int srcStride, int w)
{ {
const uint8_t *cm = ff_crop_tab + MAX_NEG_CROP; 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, ...@@ -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]; uint8_t half[64];
...@@ -198,12 +198,12 @@ static void put_mspel8_mc10_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride) ...@@ -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); 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); 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]; uint8_t half[64];
...@@ -211,12 +211,12 @@ static void put_mspel8_mc30_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride) ...@@ -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); 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); 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 halfH[88];
uint8_t halfV[64]; uint8_t halfV[64];
...@@ -228,7 +228,7 @@ static void put_mspel8_mc12_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride) ...@@ -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); 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 halfH[88];
uint8_t halfV[64]; uint8_t halfV[64];
...@@ -240,7 +240,7 @@ static void put_mspel8_mc32_c(uint8_t *dst, uint8_t *src, ptrdiff_t stride) ...@@ -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); 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]; uint8_t halfH[88];
......
...@@ -336,7 +336,7 @@ static void cavs_idct8_add_mmx(uint8_t *dst, int16_t *block, int stride) ...@@ -336,7 +336,7 @@ static void cavs_idct8_add_mmx(uint8_t *dst, int16_t *block, int stride)
} }
#define QPEL_CAVS(OPNAME, OP, MMX)\ #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;\ int h=8;\
__asm__ volatile(\ __asm__ volatile(\
"pxor %%mm7, %%mm7 \n\t"\ "pxor %%mm7, %%mm7 \n\t"\
...@@ -384,43 +384,43 @@ static void OPNAME ## cavs_qpel8_h_ ## MMX(uint8_t *dst, uint8_t *src, int dstSt ...@@ -384,43 +384,43 @@ 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) \ 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_42) \ QPEL_CAVSVNUM(QPEL_CAVSV2,OP,ff_pw_4,ff_pw_5,ff_pw_42) \
}\ }\
\ \
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) \ 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);\ 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 , src , dstStride, srcStride, 16);\
OPNAME ## cavs_qpel8or16_v1_ ## MMX(dst+8, src+8, 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);\ 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 , src , dstStride, srcStride, 16);\
OPNAME ## cavs_qpel8or16_v2_ ## MMX(dst+8, src+8, 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);\ 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 , src , dstStride, srcStride, 16);\
OPNAME ## cavs_qpel8or16_v3_ ## MMX(dst+8, src+8, 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 , src , dstStride, srcStride);\
OPNAME ## cavs_qpel8_h_ ## MMX(dst+8, src+8, dstStride, srcStride);\ OPNAME ## cavs_qpel8_h_ ## MMX(dst+8, src+8, dstStride, srcStride);\
src += 8*srcStride;\ src += 8*srcStride;\
...@@ -430,22 +430,22 @@ static void OPNAME ## cavs_qpel16_h_ ## MMX(uint8_t *dst, uint8_t *src, int dstS ...@@ -430,22 +430,22 @@ static void OPNAME ## cavs_qpel16_h_ ## MMX(uint8_t *dst, uint8_t *src, int dstS
}\ }\
#define CAVS_MC(OPNAME, SIZE, MMX) \ #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);\ 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);\ 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);\ 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);\ OPNAME ## cavs_qpel ## SIZE ## _v3_ ## MMX(dst, src, stride, stride);\
}\ }\
...@@ -463,49 +463,49 @@ static void OPNAME ## cavs_qpel ## SIZE ## _mc03_ ## MMX(uint8_t *dst, uint8_t * ...@@ -463,49 +463,49 @@ static void OPNAME ## cavs_qpel ## SIZE ## _mc03_ ## MMX(uint8_t *dst, uint8_t *
#endif /* (HAVE_MMXEXT_INLINE || HAVE_AMD3DNOW_INLINE) */ #endif /* (HAVE_MMXEXT_INLINE || HAVE_AMD3DNOW_INLINE) */
#if HAVE_MMX_EXTERNAL #if HAVE_MMX_EXTERNAL
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) ptrdiff_t stride)
{ {
ff_put_pixels8_mmx(dst, src, stride, 8); 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) ptrdiff_t stride)
{ {
ff_avg_pixels8_mmx(dst, src, stride, 8); ff_avg_pixels8_mmx(dst, src, stride, 8);
} }
static void avg_cavs_qpel8_mc00_mmxext(uint8_t *dst, uint8_t *src, static void avg_cavs_qpel8_mc00_mmxext(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride) ptrdiff_t stride)
{ {
ff_avg_pixels8_mmxext(dst, src, stride, 8); ff_avg_pixels8_mmxext(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) ptrdiff_t stride)
{ {
ff_put_pixels16_mmx(dst, src, stride, 16); 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) ptrdiff_t stride)
{ {
ff_avg_pixels16_mmx(dst, src, stride, 16); ff_avg_pixels16_mmx(dst, src, stride, 16);
} }
static void avg_cavs_qpel16_mc00_mmxext(uint8_t *dst, uint8_t *src, static void avg_cavs_qpel16_mc00_mmxext(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride) ptrdiff_t stride)
{ {
ff_avg_pixels16_mmxext(dst, src, stride, 16); ff_avg_pixels16_mmxext(dst, src, stride, 16);
} }
static void put_cavs_qpel16_mc00_sse2(uint8_t *dst, uint8_t *src, static void put_cavs_qpel16_mc00_sse2(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride) ptrdiff_t stride)
{ {
ff_put_pixels16_sse2(dst, src, stride, 16); ff_put_pixels16_sse2(dst, src, stride, 16);
} }
static void avg_cavs_qpel16_mc00_sse2(uint8_t *dst, uint8_t *src, static void avg_cavs_qpel16_mc00_sse2(uint8_t *dst, const uint8_t *src,
ptrdiff_t stride) ptrdiff_t stride)
{ {
ff_avg_pixels16_sse2(dst, src, stride, 16); ff_avg_pixels16_sse2(dst, src, stride, 16);
......
This diff is collapsed.
This diff is collapsed.
...@@ -33,7 +33,7 @@ ...@@ -33,7 +33,7 @@
#include "hpeldsp.h" #include "hpeldsp.h"
#define DEFINE_FN(op, size, insn) \ #define DEFINE_FN(op, size, insn) \
static void op##_rv40_qpel##size##_mc33_##insn(uint8_t *dst, uint8_t *src, \ static void op##_rv40_qpel##size##_mc33_##insn(uint8_t *dst, const uint8_t *src, \
ptrdiff_t stride) \ ptrdiff_t stride) \
{ \ { \
ff_##op##_pixels##size##_xy2_##insn(dst, src, stride, size); \ ff_##op##_pixels##size##_xy2_##insn(dst, src, stride, size); \
...@@ -77,7 +77,7 @@ DECLARE_WEIGHT(ssse3) ...@@ -77,7 +77,7 @@ DECLARE_WEIGHT(ssse3)
*/ */
#define QPEL_FUNC_DECL(OP, SIZE, PH, PV, OPT) \ #define QPEL_FUNC_DECL(OP, SIZE, PH, PV, OPT) \
static void OP ## rv40_qpel ##SIZE ##_mc ##PH ##PV ##OPT(uint8_t *dst, \ static void OP ## rv40_qpel ##SIZE ##_mc ##PH ##PV ##OPT(uint8_t *dst, \
uint8_t *src, \ const uint8_t *src, \
ptrdiff_t stride) \ ptrdiff_t stride) \
{ \ { \
int i; \ int i; \
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment