/*
- * HEVC video Decoder
+ * HEVC video decoder
*
* Copyright (C) 2012 - 2013 Guillaume Martres
*
*/
#include "get_bits.h"
+#include "hevc.h"
+
#include "bit_depth_template.c"
#include "hevcdsp.h"
-#include "hevc.h"
-
-#define SET(dst, x) (dst) = (x)
-#define SCALE(dst, x) (dst) = av_clip_int16(((x) + add) >> shift)
-#define ADD_AND_SCALE(dst, x) (dst) = av_clip_pixel((dst) + av_clip_int16(((x) + add) >> shift))
-static void FUNC(put_pcm)(uint8_t *_dst, ptrdiff_t _stride, int size,
+static void FUNC(put_pcm)(uint8_t *_dst, ptrdiff_t stride, int size,
GetBitContext *gb, int pcm_bit_depth)
{
int x, y;
pixel *dst = (pixel *)_dst;
- ptrdiff_t stride = _stride / sizeof(pixel);
+ stride /= sizeof(pixel);
for (y = 0; y < size; y++) {
for (x = 0; x < size; x++)
}
static void FUNC(transquant_bypass4x4)(uint8_t *_dst, int16_t *coeffs,
- ptrdiff_t _stride)
+ ptrdiff_t stride)
{
int x, y;
pixel *dst = (pixel *)_dst;
- ptrdiff_t stride = _stride / sizeof(pixel);
+ stride /= sizeof(pixel);
for (y = 0; y < 4; y++) {
for (x = 0; x < 4; x++) {
}
static void FUNC(transquant_bypass8x8)(uint8_t *_dst, int16_t *coeffs,
- ptrdiff_t _stride)
+ ptrdiff_t stride)
{
int x, y;
pixel *dst = (pixel *)_dst;
- ptrdiff_t stride = _stride / sizeof(pixel);
+ stride /= sizeof(pixel);
for (y = 0; y < 8; y++) {
for (x = 0; x < 8; x++) {
}
static void FUNC(transquant_bypass16x16)(uint8_t *_dst, int16_t *coeffs,
- ptrdiff_t _stride)
+ ptrdiff_t stride)
{
int x, y;
pixel *dst = (pixel *)_dst;
- ptrdiff_t stride = _stride / sizeof(pixel);
+ stride /= sizeof(pixel);
for (y = 0; y < 16; y++) {
for (x = 0; x < 16; x++) {
}
static void FUNC(transquant_bypass32x32)(uint8_t *_dst, int16_t *coeffs,
- ptrdiff_t _stride)
+ ptrdiff_t stride)
{
int x, y;
pixel *dst = (pixel *)_dst;
- ptrdiff_t stride = _stride / sizeof(pixel);
+ stride /= sizeof(pixel);
for (y = 0; y < 32; y++) {
for (x = 0; x < 32; x++) {
}
static void FUNC(transform_skip)(uint8_t *_dst, int16_t *coeffs,
- ptrdiff_t _stride)
+ ptrdiff_t stride)
{
pixel *dst = (pixel *)_dst;
- ptrdiff_t stride = _stride / sizeof(pixel);
int shift = 13 - BIT_DEPTH;
#if BIT_DEPTH <= 13
int offset = 1 << (shift - 1);
#endif
int x, y;
- for (y = 0; y < 4*4; y+=4) {
- for (x = 0; x < 4; x++) {
+ stride /= sizeof(pixel);
+
+ for (y = 0; y < 4 * 4; y += 4) {
+ for (x = 0; x < 4; x++)
dst[x] = av_clip_pixel(dst[x] + ((coeffs[y + x] + offset) >> shift));
- }
dst += stride;
}
}
-static void FUNC(transform_4x4_luma_add)(uint8_t *_dst, int16_t *coeffs, ptrdiff_t _stride)
-{
-#define TR_4x4_LUMA(dst, src, step, assign) \
- do { \
- int c0 = src[0*step] + src[2*step]; \
- int c1 = src[2*step] + src[3*step]; \
- int c2 = src[0*step] - src[3*step]; \
- int c3 = 74 * src[1*step]; \
- \
- assign(dst[2*step], 74 * (src[0*step] - src[2*step] + src[3*step])); \
- assign(dst[0*step], 29 * c0 + 55 * c1 + c3); \
- assign(dst[1*step], 55 * c2 - 29 * c1 + c3); \
- assign(dst[3*step], 55 * c0 + 29 * c2 - c3); \
+#define SET(dst, x) (dst) = (x)
+#define SCALE(dst, x) (dst) = av_clip_int16(((x) + add) >> shift)
+#define ADD_AND_SCALE(dst, x) \
+ (dst) = av_clip_pixel((dst) + av_clip_int16(((x) + add) >> shift))
+
+#define TR_4x4_LUMA(dst, src, step, assign) \
+ do { \
+ int c0 = src[0 * step] + src[2 * step]; \
+ int c1 = src[2 * step] + src[3 * step]; \
+ int c2 = src[0 * step] - src[3 * step]; \
+ int c3 = 74 * src[1 * step]; \
+ \
+ assign(dst[2 * step], 74 * (src[0 * step] - \
+ src[2 * step] + \
+ src[3 * step])); \
+ assign(dst[0 * step], 29 * c0 + 55 * c1 + c3); \
+ assign(dst[1 * step], 55 * c2 - 29 * c1 + c3); \
+ assign(dst[3 * step], 55 * c0 + 29 * c2 - c3); \
} while (0)
+static void FUNC(transform_4x4_luma_add)(uint8_t *_dst, int16_t *coeffs,
+ ptrdiff_t stride)
+{
int i;
pixel *dst = (pixel *)_dst;
- ptrdiff_t stride = _stride / sizeof(pixel);
- int shift = 7;
+ int shift = 7;
int add = 1 << (shift - 1);
int16_t *src = coeffs;
+ stride /= sizeof(pixel);
+
for (i = 0; i < 4; i++) {
TR_4x4_LUMA(src, src, 4, SCALE);
src++;
coeffs += 4;
dst += stride;
}
+}
#undef TR_4x4_LUMA
-}
-#define TR_4(dst, src, dstep, sstep, assign) \
- do { \
- const int e0 = transform[8*0][0] * src[0*sstep] + \
- transform[8*2][0] * src[2*sstep]; \
- const int e1 = transform[8*0][1] * src[0*sstep] + \
- transform[8*2][1] * src[2*sstep]; \
- const int o0 = transform[8*1][0] * src[1*sstep] + \
- transform[8*3][0] * src[3*sstep]; \
- const int o1 = transform[8*1][1] * src[1*sstep] + \
- transform[8*3][1] * src[3*sstep]; \
- \
- assign(dst[0*dstep], e0 + o0); \
- assign(dst[1*dstep], e1 + o1); \
- assign(dst[2*dstep], e1 - o1); \
- assign(dst[3*dstep], e0 - o0); \
+#define TR_4(dst, src, dstep, sstep, assign) \
+ do { \
+ const int e0 = transform[8 * 0][0] * src[0 * sstep] + \
+ transform[8 * 2][0] * src[2 * sstep]; \
+ const int e1 = transform[8 * 0][1] * src[0 * sstep] + \
+ transform[8 * 2][1] * src[2 * sstep]; \
+ const int o0 = transform[8 * 1][0] * src[1 * sstep] + \
+ transform[8 * 3][0] * src[3 * sstep]; \
+ const int o1 = transform[8 * 1][1] * src[1 * sstep] + \
+ transform[8 * 3][1] * src[3 * sstep]; \
+ \
+ assign(dst[0 * dstep], e0 + o0); \
+ assign(dst[1 * dstep], e1 + o1); \
+ assign(dst[2 * dstep], e1 - o1); \
+ assign(dst[3 * dstep], e0 - o0); \
} while (0)
#define TR_4_1(dst, src) TR_4(dst, src, 4, 4, SCALE)
#define TR_4_2(dst, src) TR_4(dst, src, 1, 1, ADD_AND_SCALE)
-static void FUNC(transform_4x4_add)(uint8_t *_dst, int16_t *coeffs, ptrdiff_t _stride)
+static void FUNC(transform_4x4_add)(uint8_t *_dst, int16_t *coeffs,
+ ptrdiff_t stride)
{
int i;
pixel *dst = (pixel *)_dst;
- ptrdiff_t stride = _stride / sizeof(pixel);
- int shift = 7;
+ int shift = 7;
int add = 1 << (shift - 1);
int16_t *src = coeffs;
+ stride /= sizeof(pixel);
+
for (i = 0; i < 4; i++) {
TR_4_1(src, src);
src++;
#define TR_16_2(dst, src) TR_16(dst, src, 1, 1, ADD_AND_SCALE)
#define TR_32_2(dst, src) TR_32(dst, src, 1, 1, ADD_AND_SCALE)
-static void FUNC(transform_8x8_add)(uint8_t *_dst, int16_t *coeffs, ptrdiff_t _stride)
+static void FUNC(transform_8x8_add)(uint8_t *_dst, int16_t *coeffs,
+ ptrdiff_t stride)
{
int i;
pixel *dst = (pixel *)_dst;
- ptrdiff_t stride = _stride / sizeof(pixel);
- int shift = 7;
+ int shift = 7;
int add = 1 << (shift - 1);
int16_t *src = coeffs;
+ stride /= sizeof(pixel);
+
for (i = 0; i < 8; i++) {
TR_8_1(src, src);
src++;
}
static void FUNC(transform_16x16_add)(uint8_t *_dst, int16_t *coeffs,
- ptrdiff_t _stride)
+ ptrdiff_t stride)
{
int i;
pixel *dst = (pixel *)_dst;
- ptrdiff_t stride = _stride / sizeof(pixel);
- int shift = 7;
+ int shift = 7;
int add = 1 << (shift - 1);
int16_t *src = coeffs;
+ stride /= sizeof(pixel);
+
for (i = 0; i < 16; i++) {
TR_16_1(src, src);
src++;
}
static void FUNC(transform_32x32_add)(uint8_t *_dst, int16_t *coeffs,
- ptrdiff_t _stride)
+ ptrdiff_t stride)
{
#define IT32x32_even(i,w) ( src[ 0*w] * transform[ 0][i] ) + ( src[16*w] * transform[16][i] )
#define IT32x32_odd(i,w) ( src[ 8*w] * transform[ 8][i] ) + ( src[24*w] * transform[24][i] )
( src[25*w] * transform[25][i] ) + ( src[27*w] * transform[27][i] ) + ( src[29*w] * transform[29][i] ) + ( src[31*w] * transform[31][i] )
int i;
pixel *dst = (pixel *)_dst;
- ptrdiff_t stride = _stride / sizeof(pixel);
- int shift = 7;
+ int shift = 7;
int add = 1 << (shift - 1);
int16_t *src = coeffs;
+ stride /= sizeof(pixel);
+
for (i = 0; i < 32; i++) {
TR_32_1(src, src);
src++;
}
static void FUNC(sao_band_filter)(uint8_t *_dst, uint8_t *_src,
- ptrdiff_t _stride, SAOParams *sao,
+ ptrdiff_t stride, SAOParams *sao,
int *borders, int width, int height,
int c_idx, int class)
{
pixel *dst = (pixel *)_dst;
pixel *src = (pixel *)_src;
- ptrdiff_t stride = _stride / sizeof(pixel);
int offset_table[32] = { 0 };
int k, y, x;
int chroma = !!c_idx;
int sao_left_class = sao->band_position[c_idx];
int init_y = 0, init_x = 0;
+ stride /= sizeof(pixel);
+
switch (class) {
case 0:
if (!borders[2])
- width -= ((8 >> chroma) + 2);
+ width -= (8 >> chroma) + 2;
if (!borders[3])
- height -= ((4 >> chroma) + 2);
+ height -= (4 >> chroma) + 2;
break;
case 1:
init_y = -(4 >> chroma) - 2;
if (!borders[2])
- width -= ((8 >> chroma) + 2);
+ width -= (8 >> chroma) + 2;
height = (4 >> chroma) + 2;
break;
case 2:
init_x = -(8 >> chroma) - 2;
width = (8 >> chroma) + 2;
if (!borders[3])
- height -= ((4 >> chroma) + 2);
+ height -= (4 >> chroma) + 2;
break;
case 3:
init_y = -(4 >> chroma) - 2;
}
static void FUNC(sao_band_filter_3)(uint8_t *_dst, uint8_t *_src,
- ptrdiff_t _stride, SAOParams *sao,
+ ptrdiff_t stride, SAOParams *sao,
int *borders, int width, int height,
int c_idx)
{
- FUNC(sao_band_filter)(_dst, _src, _stride, sao, borders,
+ FUNC(sao_band_filter)(_dst, _src, stride, sao, borders,
width, height, c_idx, 3);
}
static void FUNC(sao_edge_filter_0)(uint8_t *_dst, uint8_t *_src,
- ptrdiff_t _stride, SAOParams *sao,
+ ptrdiff_t stride, SAOParams *sao,
int *borders, int _width, int _height,
- int c_idx,
- uint8_t vert_edge, uint8_t horiz_edge, uint8_t diag_edge)
+ int c_idx, uint8_t vert_edge,
+ uint8_t horiz_edge, uint8_t diag_edge)
{
int x, y;
pixel *dst = (pixel *)_dst;
pixel *src = (pixel *)_src;
- ptrdiff_t stride = _stride / sizeof(pixel);
int chroma = !!c_idx;
int *sao_offset_val = sao->offset_val[c_idx];
int sao_eo_class = sao->eo_class[c_idx];
+ int init_x = 0, init_y = 0, width = _width, height = _height;
static const int8_t pos[4][2][2] = {
{ { -1, 0 }, { 1, 0 } }, // horizontal
};
static const uint8_t edge_idx[] = { 1, 2, 0, 3, 4 };
- int init_x = 0, init_y = 0, width = _width, height = _height;
-
#define CMP(a, b) ((a) > (b) ? 1 : ((a) == (b) ? 0 : -1))
+ stride /= sizeof(pixel);
+
if (!borders[2])
width -= (8 >> chroma) + 2;
if (!borders[3])
}
static void FUNC(sao_edge_filter_1)(uint8_t *_dst, uint8_t *_src,
- ptrdiff_t _stride, SAOParams *sao,
+ ptrdiff_t stride, SAOParams *sao,
int *borders, int _width, int _height,
int c_idx, uint8_t vert_edge,
uint8_t horiz_edge, uint8_t diag_edge)
int x, y;
pixel *dst = (pixel *)_dst;
pixel *src = (pixel *)_src;
- ptrdiff_t stride = _stride / sizeof(pixel);
int chroma = !!c_idx;
int *sao_offset_val = sao->offset_val[c_idx];
int sao_eo_class = sao->eo_class[c_idx];
+ int init_x = 0, init_y = 0, width = _width, height = _height;
static const int8_t pos[4][2][2] = {
{ { -1, 0 }, { 1, 0 } }, // horizontal
};
static const uint8_t edge_idx[] = { 1, 2, 0, 3, 4 };
- int init_x = 0, init_y = 0, width = _width, height = _height;
-
#define CMP(a, b) ((a) > (b) ? 1 : ((a) == (b) ? 0 : -1))
+ stride /= sizeof(pixel);
+
init_y = -(4 >> chroma) - 2;
if (!borders[2])
width -= (8 >> chroma) + 2;
}
static void FUNC(sao_edge_filter_2)(uint8_t *_dst, uint8_t *_src,
- ptrdiff_t _stride, SAOParams *sao,
+ ptrdiff_t stride, SAOParams *sao,
int *borders, int _width, int _height,
int c_idx, uint8_t vert_edge,
uint8_t horiz_edge, uint8_t diag_edge)
int x, y;
pixel *dst = (pixel *)_dst;
pixel *src = (pixel *)_src;
- ptrdiff_t stride = _stride / sizeof(pixel);
int chroma = !!c_idx;
int *sao_offset_val = sao->offset_val[c_idx];
int sao_eo_class = sao->eo_class[c_idx];
int init_x = 0, init_y = 0, width = _width, height = _height;
#define CMP(a, b) ((a) > (b) ? 1 : ((a) == (b) ? 0 : -1))
+ stride /= sizeof(pixel);
init_x = -(8 >> chroma) - 2;
width = (8 >> chroma) + 2;
}
static void FUNC(sao_edge_filter_3)(uint8_t *_dst, uint8_t *_src,
- ptrdiff_t _stride, SAOParams *sao,
+ ptrdiff_t stride, SAOParams *sao,
int *borders, int _width, int _height,
int c_idx, uint8_t vert_edge,
uint8_t horiz_edge, uint8_t diag_edge)
int x, y;
pixel *dst = (pixel *)_dst;
pixel *src = (pixel *)_src;
- ptrdiff_t stride = _stride / sizeof(pixel);
int chroma = !!c_idx;
int *sao_offset_val = sao->offset_val[c_idx];
int sao_eo_class = sao->eo_class[c_idx];
+ int init_x = 0, init_y = 0, width = _width, height = _height;
static const int8_t pos[4][2][2] = {
{ { -1, 0 }, { 1, 0 } }, // horizontal
};
static const uint8_t edge_idx[] = { 1, 2, 0, 3, 4 };
- int init_x = 0, init_y = 0, width = _width, height = _height;
-
#define CMP(a, b) ((a) > (b) ? 1 : ((a) == (b) ? 0 : -1))
+ stride /= sizeof(pixel);
+
init_y = -(4 >> chroma) - 2;
init_x = -(8 >> chroma) - 2;
width = (8 >> chroma) + 2;
}
#define QPEL_FILTER_1(src, stride) \
- (-src[x - 3 * stride] + 4 * src[x - 2 * stride] - 10 * src[x - stride] + 58 * src[x] + \
- 17 * src[x + stride] - 5 * src[x + 2 * stride] + 1 * src[x + 3 * stride])
+ (1 * -src[x - 3 * stride] + \
+ 4 * src[x - 2 * stride] - \
+ 10 * src[x - stride] + \
+ 58 * src[x] + \
+ 17 * src[x + stride] - \
+ 5 * src[x + 2 * stride] + \
+ 1 * src[x + 3 * stride])
#define QPEL_FILTER_2(src, stride) \
- (-src[x - 3 * stride] + 4 * src[x - 2 * stride] - 11 * src[x - stride] + 40 * src[x] + \
- 40 * src[x + stride] - 11 * src[x + 2 * stride] + 4 * src[x + 3 * stride] - src[x + 4 * stride])
+ (1 * -src[x - 3 * stride] + \
+ 4 * src[x - 2 * stride] - \
+ 11 * src[x - stride] + \
+ 40 * src[x] + \
+ 40 * src[x + stride] - \
+ 11 * src[x + 2 * stride] + \
+ 4 * src[x + 3 * stride] - \
+ 1 * src[x + 4 * stride])
#define QPEL_FILTER_3(src, stride) \
- (src[x - 2 * stride] - 5 * src[x - stride] + 17 * src[x] + 58 * src[x + stride] \
- - 10 * src[x + 2 * stride] + 4 * src[x + 3 * stride] - src[x + 4 * stride])
+ (1 * src[x - 2 * stride] - \
+ 5 * src[x - stride] + \
+ 17 * src[x] + \
+ 58 * src[x + stride] - \
+ 10 * src[x + 2 * stride] + \
+ 4 * src[x + 3 * stride] - \
+ 1 * src[x + 4 * stride])
#define PUT_HEVC_QPEL_H(H) \
}
#define PUT_HEVC_QPEL_HV(H, V) \
-static void FUNC(put_hevc_qpel_h ## H ## v ## V)(int16_t *dst, ptrdiff_t dststride, \
- uint8_t *_src, ptrdiff_t _srcstride,\
+static void FUNC(put_hevc_qpel_h ## H ## v ## V)(int16_t *dst, \
+ ptrdiff_t dststride, \
+ uint8_t *_src, \
+ ptrdiff_t _srcstride, \
int width, int height, \
int16_t* mcbuffer) \
{ \
}
#define EPEL_FILTER(src, stride) \
- (filter_0 * src[x - stride] + filter_1 * src[x] + filter_2 * src[x + stride] + filter_3 * src[x + 2 * stride])
+ (filter_0 * src[x - stride] + \
+ filter_1 * src[x] + \
+ filter_2 * src[x + stride] + \
+ filter_3 * src[x + 2 * stride])
static void FUNC(put_hevc_epel_h)(int16_t *dst, ptrdiff_t dststride,
uint8_t *_src, ptrdiff_t _srcstride,
int16_t *src, ptrdiff_t srcstride,
int width, int height)
{
- int shift;
- int log2Wd;
- int wx;
- int ox;
- int x, y;
- int offset;
+ int shift, log2Wd, wx, ox, x, y, offset;
pixel *dst = (pixel *)_dst;
ptrdiff_t dststride = _dststride / sizeof(pixel);
ptrdiff_t srcstride,
int width, int height)
{
- int shift;
- int log2Wd;
- int w0;
- int w1;
- int o0;
- int o1;
- int x, y;
+ int shift, log2Wd, w0, w1, o0, o1, x, y;
pixel *dst = (pixel *)_dst;
ptrdiff_t dststride = _dststride / sizeof(pixel);
#define P3 pix[-4 * xstride]
#define P2 pix[-3 * xstride]
#define P1 pix[-2 * xstride]
-#define P0 pix[-xstride]
-#define Q0 pix[0]
-#define Q1 pix[xstride]
+#define P0 pix[-1 * xstride]
+#define Q0 pix[0 * xstride]
+#define Q1 pix[1 * xstride]
#define Q2 pix[2 * xstride]
#define Q3 pix[3 * xstride]
#define TP3 pix[-4 * xstride + 3 * ystride]
#define TP2 pix[-3 * xstride + 3 * ystride]
#define TP1 pix[-2 * xstride + 3 * ystride]
-#define TP0 pix[-xstride + 3 * ystride]
-#define TQ0 pix[3 * ystride]
-#define TQ1 pix[xstride + 3 * ystride]
+#define TP0 pix[-1 * xstride + 3 * ystride]
+#define TQ0 pix[0 * xstride + 3 * ystride]
+#define TQ1 pix[1 * xstride + 3 * ystride]
#define TQ2 pix[2 * xstride + 3 * ystride]
#define TQ3 pix[3 * xstride + 3 * ystride]
const int dq3 = abs(TQ2 - 2 * TQ1 + TQ0);
const int d0 = dp0 + dq0;
const int d3 = dp3 + dq3;
- int beta = _beta[j] << (BIT_DEPTH - 8);
+ const int beta = _beta[j] << (BIT_DEPTH - 8);
const int tc = _tc[j] << (BIT_DEPTH - 8);
const int no_p = _no_p[j];
const int no_q = _no_q[j];
ptrdiff_t _ystride, int *_tc,
uint8_t *_no_p, uint8_t *_no_q)
{
- int d, j;
- int no_p, no_q;
+ int d, j, no_p, no_q;
pixel *pix = (pixel *)_pix;
ptrdiff_t xstride = _xstride / sizeof(pixel);
ptrdiff_t ystride = _ystride / sizeof(pixel);
projects / ffmpeg / blobdiff