[FFmpeg-devel] [PATCH 2/4] hevcdsp: remove compilation-time-fixed parameter
Christophe Gisquet
christophe.gisquet at gmail.com
Mon Jul 28 19:17:26 CEST 2014
The dststride parameter is always MAX_PB_SIZE.
---
libavcodec/hevc.c | 4 ++--
libavcodec/hevcdsp.h | 4 ++--
libavcodec/hevcdsp_template.c | 27 +++++++++++++--------------
libavcodec/x86/hevc_mc.asm | 16 ++++++++--------
libavcodec/x86/hevcdsp.h | 2 +-
libavcodec/x86/hevcdsp_init.c | 16 ++++++++--------
6 files changed, 34 insertions(+), 35 deletions(-)
diff --git a/libavcodec/hevc.c b/libavcodec/hevc.c
index 69319e2..3901c0c 100644
--- a/libavcodec/hevc.c
+++ b/libavcodec/hevc.c
@@ -1381,7 +1381,7 @@ static void luma_mc_uni(HEVCContext *s, uint8_t *dst, ptrdiff_t dststride,
src1stride = edge_emu_stride;
}
- s->hevcdsp.put_hevc_qpel[idx][!!my0][!!mx0](tmp, MAX_PB_SIZE, src0, src0stride,
+ s->hevcdsp.put_hevc_qpel[idx][!!my0][!!mx0](tmp, src0, src0stride,
block_h, mx0, my0, block_w);
if (!weight_flag)
s->hevcdsp.put_hevc_qpel_bi[idx][!!my1][!!mx1](dst, dststride, src1, src1stride, tmp, MAX_PB_SIZE,
@@ -1553,7 +1553,7 @@ static void chroma_mc_bi(HEVCContext *s, uint8_t *dst0, ptrdiff_t dststride, AVF
src2stride = edge_emu_stride;
}
- s->hevcdsp.put_hevc_epel[idx][!!my0][!!mx0](tmp, tmpstride, src1, src1stride,
+ s->hevcdsp.put_hevc_epel[idx][!!my0][!!mx0](tmp, src1, src1stride,
block_h, _mx0, _my0, block_w);
if (!weight_flag)
s->hevcdsp.put_hevc_epel_bi[idx][!!my1][!!mx1](dst0, s->frame->linesize[cidx+1],
diff --git a/libavcodec/hevcdsp.h b/libavcodec/hevcdsp.h
index be0b387..122501f 100644
--- a/libavcodec/hevcdsp.h
+++ b/libavcodec/hevcdsp.h
@@ -67,7 +67,7 @@ typedef struct HEVCDSPContext {
int _height, int c_idx, uint8_t *vert_edge,
uint8_t *horiz_edge, uint8_t *diag_edge);
- void (*put_hevc_qpel[10][2][2])(int16_t *dst, ptrdiff_t dststride, uint8_t *src, ptrdiff_t srcstride,
+ void (*put_hevc_qpel[10][2][2])(int16_t *dst, uint8_t *src, ptrdiff_t srcstride,
int height, intptr_t mx, intptr_t my, int width);
void (*put_hevc_qpel_uni[10][2][2])(uint8_t *dst, ptrdiff_t dststride, uint8_t *src, ptrdiff_t srcstride,
int height, intptr_t mx, intptr_t my, int width);
@@ -81,7 +81,7 @@ typedef struct HEVCDSPContext {
int16_t *src2, ptrdiff_t src2stride,
int height, int denom, int wx0, int wx1,
int ox0, int ox1, intptr_t mx, intptr_t my, int width);
- void (*put_hevc_epel[10][2][2])(int16_t *dst, ptrdiff_t dststride, uint8_t *src, ptrdiff_t srcstride,
+ void (*put_hevc_epel[10][2][2])(int16_t *dst, uint8_t *src, ptrdiff_t srcstride,
int height, intptr_t mx, intptr_t my, int width);
void (*put_hevc_epel_uni[10][2][2])(uint8_t *dst, ptrdiff_t dststride, uint8_t *_src, ptrdiff_t _srcstride,
diff --git a/libavcodec/hevcdsp_template.c b/libavcodec/hevcdsp_template.c
index 03a1c68..df49fc6 100644
--- a/libavcodec/hevcdsp_template.c
+++ b/libavcodec/hevcdsp_template.c
@@ -543,7 +543,7 @@ static void FUNC(sao_edge_filter_1)(uint8_t *_dst, uint8_t *_src,
////////////////////////////////////////////////////////////////////////////////
//
////////////////////////////////////////////////////////////////////////////////
-static void FUNC(put_hevc_pel_pixels)(int16_t *dst, ptrdiff_t dststride,
+static void FUNC(put_hevc_pel_pixels)(int16_t *dst,
uint8_t *_src, ptrdiff_t _srcstride,
int height, intptr_t mx, intptr_t my, int width)
{
@@ -555,7 +555,7 @@ static void FUNC(put_hevc_pel_pixels)(int16_t *dst, ptrdiff_t dststride,
for (x = 0; x < width; x++)
dst[x] = src[x] << (14 - BIT_DEPTH);
src += srcstride;
- dst += dststride;
+ dst += MAX_PB_SIZE;
}
}
@@ -664,7 +664,7 @@ static void FUNC(put_hevc_pel_bi_w_pixels)(uint8_t *_dst, ptrdiff_t _dststride,
filter[6] * src[x + 3 * stride] + \
filter[7] * src[x + 4 * stride])
-static void FUNC(put_hevc_qpel_h)(int16_t *dst, ptrdiff_t dststride,
+static void FUNC(put_hevc_qpel_h)(int16_t *dst,
uint8_t *_src, ptrdiff_t _srcstride,
int height, intptr_t mx, intptr_t my, int width)
{
@@ -676,11 +676,11 @@ static void FUNC(put_hevc_qpel_h)(int16_t *dst, ptrdiff_t dststride,
for (x = 0; x < width; x++)
dst[x] = QPEL_FILTER(src, 1) >> (BIT_DEPTH - 8);
src += srcstride;
- dst += dststride;
+ dst += MAX_PB_SIZE;
}
}
-static void FUNC(put_hevc_qpel_v)(int16_t *dst, ptrdiff_t dststride,
+static void FUNC(put_hevc_qpel_v)(int16_t *dst,
uint8_t *_src, ptrdiff_t _srcstride,
int height, intptr_t mx, intptr_t my, int width)
{
@@ -692,12 +692,11 @@ static void FUNC(put_hevc_qpel_v)(int16_t *dst, ptrdiff_t dststride,
for (x = 0; x < width; x++)
dst[x] = QPEL_FILTER(src, srcstride) >> (BIT_DEPTH - 8);
src += srcstride;
- dst += dststride;
+ dst += MAX_PB_SIZE;
}
}
static void FUNC(put_hevc_qpel_hv)(int16_t *dst,
- ptrdiff_t dststride,
uint8_t *_src,
ptrdiff_t _srcstride,
int height, intptr_t mx,
@@ -725,7 +724,7 @@ static void FUNC(put_hevc_qpel_hv)(int16_t *dst,
for (x = 0; x < width; x++)
dst[x] = QPEL_FILTER(tmp, MAX_PB_SIZE) >> 6;
tmp += MAX_PB_SIZE;
- dst += dststride;
+ dst += MAX_PB_SIZE;
}
}
@@ -1118,7 +1117,7 @@ static void FUNC(put_hevc_qpel_bi_w_hv)(uint8_t *_dst, ptrdiff_t _dststride, uin
filter[2] * src[x + stride] + \
filter[3] * src[x + 2 * stride])
-static void FUNC(put_hevc_epel_h)(int16_t *dst, ptrdiff_t dststride,
+static void FUNC(put_hevc_epel_h)(int16_t *dst,
uint8_t *_src, ptrdiff_t _srcstride,
int height, intptr_t mx, intptr_t my, int width)
{
@@ -1130,11 +1129,11 @@ static void FUNC(put_hevc_epel_h)(int16_t *dst, ptrdiff_t dststride,
for (x = 0; x < width; x++)
dst[x] = EPEL_FILTER(src, 1) >> (BIT_DEPTH - 8);
src += srcstride;
- dst += dststride;
+ dst += MAX_PB_SIZE;
}
}
-static void FUNC(put_hevc_epel_v)(int16_t *dst, ptrdiff_t dststride,
+static void FUNC(put_hevc_epel_v)(int16_t *dst,
uint8_t *_src, ptrdiff_t _srcstride,
int height, intptr_t mx, intptr_t my, int width)
{
@@ -1147,11 +1146,11 @@ static void FUNC(put_hevc_epel_v)(int16_t *dst, ptrdiff_t dststride,
for (x = 0; x < width; x++)
dst[x] = EPEL_FILTER(src, srcstride) >> (BIT_DEPTH - 8);
src += srcstride;
- dst += dststride;
+ dst += MAX_PB_SIZE;
}
}
-static void FUNC(put_hevc_epel_hv)(int16_t *dst, ptrdiff_t dststride,
+static void FUNC(put_hevc_epel_hv)(int16_t *dst,
uint8_t *_src, ptrdiff_t _srcstride,
int height, intptr_t mx, intptr_t my, int width)
{
@@ -1178,7 +1177,7 @@ static void FUNC(put_hevc_epel_hv)(int16_t *dst, ptrdiff_t dststride,
for (x = 0; x < width; x++)
dst[x] = EPEL_FILTER(tmp, MAX_PB_SIZE) >> 6;
tmp += MAX_PB_SIZE;
- dst += dststride;
+ dst += MAX_PB_SIZE;
}
}
diff --git a/libavcodec/x86/hevc_mc.asm b/libavcodec/x86/hevc_mc.asm
index 56cb6b7..0f7e3d5 100644
--- a/libavcodec/x86/hevc_mc.asm
+++ b/libavcodec/x86/hevc_mc.asm
@@ -548,7 +548,7 @@ INIT_XMM sse4 ; adds ff_ and _sse4 to functio
; ******************************
%macro HEVC_PUT_HEVC_PEL_PIXELS 2
-cglobal hevc_put_hevc_pel_pixels%1_%2, 5, 5, 3, dst, dststride, src, srcstride,height
+cglobal hevc_put_hevc_pel_pixels%1_%2, 4, 4, 3, dst, src, srcstride,height
pxor m2, m2
.loop
SIMPLE_LOAD %1, %2, srcq, m0
@@ -596,7 +596,7 @@ cglobal hevc_put_hevc_bi_pel_pixels%1_%2, 7, 7, 6, dst, dststride, src, srcstrid
%macro HEVC_PUT_HEVC_EPEL 2
-cglobal hevc_put_hevc_epel_h%1_%2, 6, 7, 6, dst, dststride, src, srcstride, height, mx, rfilter
+cglobal hevc_put_hevc_epel_h%1_%2, 5, 6, 6, dst, src, srcstride, height, mx, rfilter
%assign %%stride ((%2 + 7)/8)
EPEL_FILTER %2, mx, m4, m5
.loop
@@ -644,7 +644,7 @@ cglobal hevc_put_hevc_bi_epel_h%1_%2, 8, 9, 7, dst, dststride, src, srcstride, s
; int16_t* mcbuffer)
; ******************************
-cglobal hevc_put_hevc_epel_v%1_%2, 7, 8, 6, dst, dststride, src, srcstride, height, r3src, my, rfilter
+cglobal hevc_put_hevc_epel_v%1_%2, 6, 7, 6, dst, src, srcstride, height, r3src, my, rfilter
lea r3srcq, [srcstrideq*3]
sub srcq, srcstrideq
EPEL_FILTER %2, my, m4, m5
@@ -699,7 +699,7 @@ cglobal hevc_put_hevc_bi_epel_v%1_%2, 9, 10, 7, dst, dststride, src, srcstride,
; ******************************
%macro HEVC_PUT_HEVC_EPEL_HV 2
-cglobal hevc_put_hevc_epel_hv%1_%2, 7, 9, 12 , dst, dststride, src, srcstride, height, mx, my, r3src, rfilter
+cglobal hevc_put_hevc_epel_hv%1_%2, 6, 8, 12 , dst, src, srcstride, height, mx, my, r3src, rfilter
%assign %%stride ((%2 + 7)/8)
sub srcq, srcstrideq
EPEL_HV_FILTER %2
@@ -820,7 +820,7 @@ cglobal hevc_put_hevc_bi_epel_hv%1_%2, 9, 11, 16, dst, dststride, src, srcstride
; ******************************
%macro HEVC_PUT_HEVC_QPEL 2
-cglobal hevc_put_hevc_qpel_h%1_%2, 6, 7, 15 , dst, dststride, src, srcstride, height, mx, rfilter
+cglobal hevc_put_hevc_qpel_h%1_%2, 5, 6, 15, dst, src, srcstride, height, mx, rfilter
QPEL_FILTER %2, mx
.loop
QPEL_H_LOAD %2, srcq, %1, 10
@@ -875,11 +875,11 @@ cglobal hevc_put_hevc_bi_qpel_h%1_%2, 8, 9, 16 , dst, dststride, src, srcstride,
; int width, int height, int mx, int my)
; ******************************
-cglobal hevc_put_hevc_qpel_v%1_%2, 7, 9, 15, dst, dststride, src, srcstride, height, r3src, my, rfilter
+cglobal hevc_put_hevc_qpel_v%1_%2, 6, 8, 15, dst, src, srcstride, height, r3src, my, rfilter
lea r3srcq, [srcstrideq*3]
QPEL_FILTER %2, my
.loop
- QPEL_V_LOAD %2, srcq, srcstride, %1, r8
+ QPEL_V_LOAD %2, srcq, srcstride, %1, r7
QPEL_COMPUTE %1, %2
%if %2 > 8
packssdw m0, m1
@@ -934,7 +934,7 @@ cglobal hevc_put_hevc_bi_qpel_v%1_%2, 9, 11, 16, dst, dststride, src, srcstride,
; int height, int mx, int my)
; ******************************
%macro HEVC_PUT_HEVC_QPEL_HV 2
-cglobal hevc_put_hevc_qpel_hv%1_%2, 7, 9, 12 , dst, dststride, src, srcstride, height, mx, my, r3src, rfilter
+cglobal hevc_put_hevc_qpel_hv%1_%2, 6, 8, 12, dst, src, srcstride, height, mx, my, r3src, rfilter
lea mxq, [mxq*8-8]
lea myq, [myq*8-8]
lea r3srcq, [srcstrideq*3]
diff --git a/libavcodec/x86/hevcdsp.h b/libavcodec/x86/hevcdsp.h
index 2f57352..9f5fa4b 100644
--- a/libavcodec/x86/hevcdsp.h
+++ b/libavcodec/x86/hevcdsp.h
@@ -74,7 +74,7 @@ dst ## _bi_w[idx1][idx2][idx3] = ff_hevc_put_hevc_bi_w_ ## name ## _ ## D ## _##
#define PEL_PROTOTYPE(name, D, opt) \
-void ff_hevc_put_hevc_ ## name ## _ ## D ## _##opt(int16_t *dst, ptrdiff_t dststride,uint8_t *_src, ptrdiff_t _srcstride, int height, intptr_t mx, intptr_t my,int width); \
+void ff_hevc_put_hevc_ ## name ## _ ## D ## _##opt(int16_t *dst, uint8_t *_src, ptrdiff_t _srcstride, int height, intptr_t mx, intptr_t my,int width); \
void ff_hevc_put_hevc_bi_ ## name ## _ ## D ## _##opt(uint8_t *_dst, ptrdiff_t _dststride, uint8_t *_src, ptrdiff_t _srcstride, int16_t *src2, ptrdiff_t src2stride, int height, intptr_t mx, intptr_t my, int width); \
void ff_hevc_put_hevc_uni_ ## name ## _ ## D ## _##opt(uint8_t *_dst, ptrdiff_t _dststride, uint8_t *_src, ptrdiff_t _srcstride, int height, intptr_t mx, intptr_t my, int width); \
void ff_hevc_put_hevc_uni_w_ ## name ## _ ## D ## _##opt(uint8_t *_dst, ptrdiff_t _dststride, uint8_t *_src, ptrdiff_t _srcstride, int height, int denom, int wx, int ox, intptr_t mx, intptr_t my, int width); \
diff --git a/libavcodec/x86/hevcdsp_init.c b/libavcodec/x86/hevcdsp_init.c
index f7bdc62..b5a50b6 100644
--- a/libavcodec/x86/hevcdsp_init.c
+++ b/libavcodec/x86/hevcdsp_init.c
@@ -66,7 +66,7 @@ IDCT_FUNCS(16x16, avx2);
IDCT_FUNCS(32x32, avx2);
#define mc_rep_func(name, bitd, step, W, opt) \
-void ff_hevc_put_hevc_##name##W##_##bitd##_##opt(int16_t *_dst, ptrdiff_t dststride, \
+void ff_hevc_put_hevc_##name##W##_##bitd##_##opt(int16_t *_dst, \
uint8_t *_src, ptrdiff_t _srcstride, int height, \
intptr_t mx, intptr_t my, int width) \
{ \
@@ -76,7 +76,7 @@ void ff_hevc_put_hevc_##name##W##_##bitd##_##opt(int16_t *_dst, ptrdiff_t dststr
for (i = 0; i < W; i += step) { \
src = _src + (i * ((bitd + 7) / 8)); \
dst = _dst + i; \
- ff_hevc_put_hevc_##name##step##_##bitd##_##opt(dst, dststride, src, _srcstride, height, mx, my, width); \
+ ff_hevc_put_hevc_##name##step##_##bitd##_##opt(dst, src, _srcstride, height, mx, my, width); \
} \
}
#define mc_rep_uni_func(name, bitd, step, W, opt) \
@@ -321,9 +321,9 @@ void ff_hevc_put_hevc_uni_w_##name##W##_##bitd##_##opt(uint8_t *_dst, ptrdiff_t
int _wx, int _ox, \
intptr_t mx, intptr_t my, int width) \
{ \
- LOCAL_ALIGNED_16(int16_t, temp, [71 * 64]); \
- ff_hevc_put_hevc_##name##W##_##bitd##_##opt(temp, 64, _src, _srcstride, height, mx, my, width); \
- ff_hevc_put_hevc_uni_w##W##_##bitd##_##opt(_dst, _dststride, temp, 64, height, denom, _wx, _ox);\
+ LOCAL_ALIGNED_16(int16_t, temp, [71 * MAX_PB_SIZE]); \
+ ff_hevc_put_hevc_##name##W##_##bitd##_##opt(temp, _src, _srcstride, height, mx, my, width); \
+ ff_hevc_put_hevc_uni_w##W##_##bitd##_##opt(_dst, _dststride, temp, MAX_PB_SIZE, height, denom, _wx, _ox);\
}
#define mc_uni_w_funcs(name, bitd, opt) \
@@ -380,9 +380,9 @@ void ff_hevc_put_hevc_bi_w_##name##W##_##bitd##_##opt(uint8_t *_dst, ptrdiff_t _
int _wx0, int _wx1, int _ox0, int _ox1, \
intptr_t mx, intptr_t my, int width) \
{ \
- LOCAL_ALIGNED_16(int16_t, temp, [71 * 64]); \
- ff_hevc_put_hevc_##name##W##_##bitd##_##opt(temp, 64, _src, _srcstride, height, mx, my, width); \
- ff_hevc_put_hevc_bi_w##W##_##bitd##_##opt(_dst, _dststride, temp, 64, _src2, _src2stride, \
+ LOCAL_ALIGNED_16(int16_t, temp, [71 * MAX_PB_SIZE]); \
+ ff_hevc_put_hevc_##name##W##_##bitd##_##opt(temp, _src, _srcstride, height, mx, my, width); \
+ ff_hevc_put_hevc_bi_w##W##_##bitd##_##opt(_dst, _dststride, temp, MAX_PB_SIZE, _src2, _src2stride, \
height, denom, _wx0, _wx1, _ox0, _ox1); \
}
--
1.9.2.msysgit.0
More information about the ffmpeg-devel
mailing list