* DSP utils
*/
+#include "libavcore/imgutils.h"
#include "avcodec.h"
#include "dsputil.h"
#include "simple_idct.h"
#include "mathops.h"
#include "mpegvideo.h"
#include "config.h"
-#include "lpc.h"
#include "ac3dec.h"
#include "vorbis.h"
#include "png.h"
38, 46, 54, 62, 39, 47, 55, 63,
};
-/* a*inverse[b]>>32 == a/b for all 0<=a<=16909558 && 2<=b<=256
- * for a>16909558, is an overestimate by less than 1 part in 1<<24 */
-const uint32_t ff_inverse[257]={
- 0, 4294967295U,2147483648U,1431655766, 1073741824, 858993460, 715827883, 613566757,
- 536870912, 477218589, 429496730, 390451573, 357913942, 330382100, 306783379, 286331154,
- 268435456, 252645136, 238609295, 226050911, 214748365, 204522253, 195225787, 186737709,
- 178956971, 171798692, 165191050, 159072863, 153391690, 148102321, 143165577, 138547333,
- 134217728, 130150525, 126322568, 122713352, 119304648, 116080198, 113025456, 110127367,
- 107374183, 104755300, 102261127, 99882961, 97612894, 95443718, 93368855, 91382283,
- 89478486, 87652394, 85899346, 84215046, 82595525, 81037119, 79536432, 78090315,
- 76695845, 75350304, 74051161, 72796056, 71582789, 70409300, 69273667, 68174085,
- 67108864, 66076420, 65075263, 64103990, 63161284, 62245903, 61356676, 60492498,
- 59652324, 58835169, 58040099, 57266231, 56512728, 55778797, 55063684, 54366675,
- 53687092, 53024288, 52377650, 51746594, 51130564, 50529028, 49941481, 49367441,
- 48806447, 48258060, 47721859, 47197443, 46684428, 46182445, 45691142, 45210183,
- 44739243, 44278014, 43826197, 43383509, 42949673, 42524429, 42107523, 41698712,
- 41297763, 40904451, 40518560, 40139882, 39768216, 39403370, 39045158, 38693400,
- 38347923, 38008561, 37675152, 37347542, 37025581, 36709123, 36398028, 36092163,
- 35791395, 35495598, 35204650, 34918434, 34636834, 34359739, 34087043, 33818641,
- 33554432, 33294321, 33038210, 32786010, 32537632, 32292988, 32051995, 31814573,
- 31580642, 31350127, 31122952, 30899046, 30678338, 30460761, 30246249, 30034737,
- 29826162, 29620465, 29417585, 29217465, 29020050, 28825284, 28633116, 28443493,
- 28256364, 28071682, 27889399, 27709467, 27531842, 27356480, 27183338, 27012373,
- 26843546, 26676816, 26512144, 26349493, 26188825, 26030105, 25873297, 25718368,
- 25565282, 25414008, 25264514, 25116768, 24970741, 24826401, 24683721, 24542671,
- 24403224, 24265352, 24129030, 23994231, 23860930, 23729102, 23598722, 23469767,
- 23342214, 23216040, 23091223, 22967740, 22845571, 22724695, 22605092, 22486740,
- 22369622, 22253717, 22139007, 22025474, 21913099, 21801865, 21691755, 21582751,
- 21474837, 21367997, 21262215, 21157475, 21053762, 20951060, 20849356, 20748635,
- 20648882, 20550083, 20452226, 20355296, 20259280, 20164166, 20069941, 19976593,
- 19884108, 19792477, 19701685, 19611723, 19522579, 19434242, 19346700, 19259944,
- 19173962, 19088744, 19004281, 18920561, 18837576, 18755316, 18673771, 18592933,
- 18512791, 18433337, 18354562, 18276457, 18199014, 18122225, 18046082, 17970575,
- 17895698, 17821442, 17747799, 17674763, 17602325, 17530479, 17459217, 17388532,
- 17318417, 17248865, 17179870, 17111424, 17043522, 16976156, 16909321, 16843010,
- 16777216
-};
-
/* Input permutation for the simple_idct_mmx */
static const uint8_t simple_mmx_permutation[64]={
0x00, 0x08, 0x04, 0x09, 0x01, 0x0C, 0x05, 0x0D,
int i;
for(i=0; i+8<=w; i+=8){
- dst[i+0]= bswap_32(src[i+0]);
- dst[i+1]= bswap_32(src[i+1]);
- dst[i+2]= bswap_32(src[i+2]);
- dst[i+3]= bswap_32(src[i+3]);
- dst[i+4]= bswap_32(src[i+4]);
- dst[i+5]= bswap_32(src[i+5]);
- dst[i+6]= bswap_32(src[i+6]);
- dst[i+7]= bswap_32(src[i+7]);
+ dst[i+0]= av_bswap32(src[i+0]);
+ dst[i+1]= av_bswap32(src[i+1]);
+ dst[i+2]= av_bswap32(src[i+2]);
+ dst[i+3]= av_bswap32(src[i+3]);
+ dst[i+4]= av_bswap32(src[i+4]);
+ dst[i+5]= av_bswap32(src[i+5]);
+ dst[i+6]= av_bswap32(src[i+6]);
+ dst[i+7]= av_bswap32(src[i+7]);
}
for(;i<w; i++){
- dst[i+0]= bswap_32(src[i+0]);
+ dst[i+0]= av_bswap32(src[i+0]);
}
}
}
/**
- * Copies a rectangular area of samples to a temporary buffer and replicates the boarder samples.
+ * Copy a rectangular area of samples to a temporary buffer and replicate the border samples.
* @param buf destination buffer
* @param src source buffer
* @param linesize number of bytes between 2 vertically adjacent samples in both the source and destination buffers
* @param w width of the source buffer
* @param h height of the source buffer
*/
-void ff_emulated_edge_mc(uint8_t *buf, uint8_t *src, int linesize, int block_w, int block_h,
+void ff_emulated_edge_mc(uint8_t *buf, const uint8_t *src, int linesize, int block_w, int block_h,
int src_x, int src_y, int w, int h){
int x, y;
int start_y, start_x, end_y, end_x;
start_x= FFMAX(0, -src_x);
end_y= FFMIN(block_h, h-src_y);
end_x= FFMIN(block_w, w-src_x);
+ assert(start_y < end_y && block_h);
+ assert(start_x < end_x && block_w);
- // copy existing part
- for(y=start_y; y<end_y; y++){
- for(x=start_x; x<end_x; x++){
- buf[x + y*linesize]= src[x + y*linesize];
- }
- }
+ w = end_x - start_x;
+ src += start_y*linesize + start_x;
+ buf += start_x;
//top
for(y=0; y<start_y; y++){
- for(x=start_x; x<end_x; x++){
- buf[x + y*linesize]= buf[x + start_y*linesize];
- }
+ memcpy(buf, src, w);
+ buf += linesize;
+ }
+
+ // copy existing part
+ for(; y<end_y; y++){
+ memcpy(buf, src, w);
+ src += linesize;
+ buf += linesize;
}
//bottom
- for(y=end_y; y<block_h; y++){
- for(x=start_x; x<end_x; x++){
- buf[x + y*linesize]= buf[x + (end_y-1)*linesize];
- }
+ src -= linesize;
+ for(; y<block_h; y++){
+ memcpy(buf, src, w);
+ buf += linesize;
}
- for(y=0; y<block_h; y++){
+ buf -= block_h * linesize + start_x;
+ while (block_h--){
//left
for(x=0; x<start_x; x++){
- buf[x + y*linesize]= buf[start_x + y*linesize];
+ buf[x] = buf[start_x];
}
//right
for(x=end_x; x<block_w; x++){
- buf[x + y*linesize]= buf[end_x - 1 + y*linesize];
+ buf[x] = buf[end_x - 1];
}
+ buf += linesize;
}
}
OPNAME ## _pixels8_l2(block, pixels, pixels+line_size, line_size, line_size, line_size, h);\
}\
\
-static inline void OPNAME ## _pixels8_l4(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,\
+static inline void OPNAME ## _pixels8_l4(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, const uint8_t *src3, const uint8_t *src4,\
int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\
int i;\
for(i=0; i<h; i++){\
OPNAME ## _pixels2_l2(block, pixels, pixels+line_size, line_size, line_size, line_size, h);\
}\
\
-static inline void OPNAME ## _no_rnd_pixels8_l4(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,\
+static inline void OPNAME ## _no_rnd_pixels8_l4(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, const uint8_t *src3, const uint8_t *src4,\
int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\
int i;\
for(i=0; i<h; i++){\
OP(*((uint32_t*)&dst[i*dst_stride+4]), h0+h1+(((l0+l1)>>2)&0x0F0F0F0FUL));\
}\
}\
-static inline void OPNAME ## _pixels16_l4(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,\
+static inline void OPNAME ## _pixels16_l4(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, const uint8_t *src3, const uint8_t *src4,\
int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\
OPNAME ## _pixels8_l4(dst , src1 , src2 , src3 , src4 , dst_stride, src_stride1, src_stride2, src_stride3, src_stride4, h);\
OPNAME ## _pixels8_l4(dst+8, src1+8, src2+8, src3+8, src4+8, dst_stride, src_stride1, src_stride2, src_stride3, src_stride4, h);\
}\
-static inline void OPNAME ## _no_rnd_pixels16_l4(uint8_t *dst, const uint8_t *src1, uint8_t *src2, uint8_t *src3, uint8_t *src4,\
+static inline void OPNAME ## _no_rnd_pixels16_l4(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, const uint8_t *src3, const uint8_t *src4,\
int dst_stride, int src_stride1, int src_stride2,int src_stride3,int src_stride4, int h){\
OPNAME ## _no_rnd_pixels8_l4(dst , src1 , src2 , src3 , src4 , dst_stride, src_stride1, src_stride2, src_stride3, src_stride4, h);\
OPNAME ## _no_rnd_pixels8_l4(dst+8, src1+8, src2+8, src3+8, src4+8, dst_stride, src_stride1, src_stride2, src_stride3, src_stride4, h);\
CALL_2X_PIXELS(OPNAME ## _pixels16_x2_c , OPNAME ## _pixels8_x2_c , 8)\
CALL_2X_PIXELS(OPNAME ## _pixels16_y2_c , OPNAME ## _pixels8_y2_c , 8)\
CALL_2X_PIXELS(OPNAME ## _pixels16_xy2_c, OPNAME ## _pixels8_xy2_c, 8)\
-CALL_2X_PIXELS(OPNAME ## _no_rnd_pixels16_c , OPNAME ## _pixels8_c , 8)\
+av_unused CALL_2X_PIXELS(OPNAME ## _no_rnd_pixels16_c , OPNAME ## _pixels8_c , 8)\
CALL_2X_PIXELS(OPNAME ## _no_rnd_pixels16_x2_c , OPNAME ## _no_rnd_pixels8_x2_c , 8)\
CALL_2X_PIXELS(OPNAME ## _no_rnd_pixels16_y2_c , OPNAME ## _no_rnd_pixels8_y2_c , 8)\
CALL_2X_PIXELS(OPNAME ## _no_rnd_pixels16_xy2_c, OPNAME ## _no_rnd_pixels8_xy2_c, 8)\
#undef op_avg
#undef op_put
+#define put_no_rnd_pixels8_c put_pixels8_c
+#define put_no_rnd_pixels16_c put_pixels16_c
+
#define avg2(a,b) ((a+b+1)>>1)
#define avg4(a,b,c,d) ((a+b+c+d+2)>>2)
}\
}\
\
-static void OPNAME ## qpel8_mc00_c (uint8_t *dst, uint8_t *src, int stride){\
- OPNAME ## pixels8_c(dst, src, stride, 8);\
-}\
-\
static void OPNAME ## qpel8_mc10_c(uint8_t *dst, uint8_t *src, int stride){\
uint8_t half[64];\
put ## RND ## mpeg4_qpel8_h_lowpass(half, src, 8, stride, 8);\
put ## RND ## mpeg4_qpel8_h_lowpass(halfH, src, 8, stride, 9);\
OPNAME ## mpeg4_qpel8_v_lowpass(dst, halfH, stride, 8);\
}\
-static void OPNAME ## qpel16_mc00_c (uint8_t *dst, uint8_t *src, int stride){\
- OPNAME ## pixels16_c(dst, src, stride, 16);\
-}\
\
static void OPNAME ## qpel16_mc10_c(uint8_t *dst, uint8_t *src, int stride){\
uint8_t half[256];\
#undef op_put
#undef op_put_no_rnd
+#define put_qpel8_mc00_c ff_put_pixels8x8_c
+#define avg_qpel8_mc00_c ff_avg_pixels8x8_c
+#define put_qpel16_mc00_c ff_put_pixels16x16_c
+#define avg_qpel16_mc00_c ff_avg_pixels16x16_c
+#define put_no_rnd_qpel8_mc00_c ff_put_pixels8x8_c
+#define put_no_rnd_qpel16_mc00_c ff_put_pixels16x16_c
+
#if 1
#define H264_LOWPASS(OPNAME, OP, OP2) \
static av_unused void OPNAME ## h264_qpel2_h_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride){\
}\
#define H264_MC(OPNAME, SIZE) \
-static void OPNAME ## h264_qpel ## SIZE ## _mc00_c (uint8_t *dst, uint8_t *src, int stride){\
+static av_unused void OPNAME ## h264_qpel ## SIZE ## _mc00_c (uint8_t *dst, uint8_t *src, int stride){\
OPNAME ## pixels ## SIZE ## _c(dst, src, stride, SIZE);\
}\
\
#undef op2_put
#endif
+#define put_h264_qpel8_mc00_c ff_put_pixels8x8_c
+#define avg_h264_qpel8_mc00_c ff_avg_pixels8x8_c
+#define put_h264_qpel16_mc00_c ff_put_pixels16x16_c
+#define avg_h264_qpel16_mc00_c ff_avg_pixels16x16_c
+
static void wmv2_mspel8_h_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h){
uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;
int i;
}
}
-#if CONFIG_CAVS_DECODER
-/* AVS specific */
-void ff_put_cavs_qpel8_mc00_c(uint8_t *dst, uint8_t *src, int stride) {
+void ff_put_pixels8x8_c(uint8_t *dst, uint8_t *src, int stride) {
put_pixels8_c(dst, src, stride, 8);
}
-void ff_avg_cavs_qpel8_mc00_c(uint8_t *dst, uint8_t *src, int stride) {
+void ff_avg_pixels8x8_c(uint8_t *dst, uint8_t *src, int stride) {
avg_pixels8_c(dst, src, stride, 8);
}
-void ff_put_cavs_qpel16_mc00_c(uint8_t *dst, uint8_t *src, int stride) {
+void ff_put_pixels16x16_c(uint8_t *dst, uint8_t *src, int stride) {
put_pixels16_c(dst, src, stride, 16);
}
-void ff_avg_cavs_qpel16_mc00_c(uint8_t *dst, uint8_t *src, int stride) {
+void ff_avg_pixels16x16_c(uint8_t *dst, uint8_t *src, int stride) {
avg_pixels16_c(dst, src, stride, 16);
}
-#endif /* CONFIG_CAVS_DECODER */
-
-#if CONFIG_VC1_DECODER
-/* VC-1 specific */
-void ff_put_vc1_mspel_mc00_c(uint8_t *dst, const uint8_t *src, int stride, int rnd) {
- put_pixels8_c(dst, src, stride, 8);
-}
-void ff_avg_vc1_mspel_mc00_c(uint8_t *dst, const uint8_t *src, int stride, int rnd) {
- avg_pixels8_c(dst, src, stride, 8);
-}
-#endif /* CONFIG_VC1_DECODER */
#if CONFIG_RV40_DECODER
static void put_rv40_qpel16_mc33_c(uint8_t *dst, uint8_t *src, int stride){
}
}
-static void put_mspel8_mc00_c (uint8_t *dst, uint8_t *src, int stride){
- put_pixels8_c(dst, src, stride, 8);
-}
-
static void put_mspel8_mc10_c(uint8_t *dst, uint8_t *src, int stride){
uint8_t half[64];
wmv2_mspel8_h_lowpass(half, src, 8, stride, 8);
WRAPPER8_16_SQ(rd8x8_c, rd16_c)
WRAPPER8_16_SQ(bit8x8_c, bit16_c)
-static void vector_fmul_c(float *dst, const float *src, int len){
+static void vector_fmul_c(float *dst, const float *src0, const float *src1, int len){
int i;
for(i=0; i<len; i++)
- dst[i] *= src[i];
+ dst[i] = src0[i] * src1[i];
}
static void vector_fmul_reverse_c(float *dst, const float *src0, const float *src1, int len){
dst[i] = src0[i] * src1[i] + src2[i];
}
-void ff_vector_fmul_window_c(float *dst, const float *src0, const float *src1, const float *win, float add_bias, int len){
+static void vector_fmul_window_c(float *dst, const float *src0,
+ const float *src1, const float *win, int len)
+{
int i,j;
dst += len;
win += len;
float s1 = src1[j];
float wi = win[i];
float wj = win[j];
- dst[i] = s0*wj - s1*wi + add_bias;
- dst[j] = s0*wi + s1*wj + add_bias;
+ dst[i] = s0*wj - s1*wi;
+ dst[j] = s0*wi + s1*wj;
}
}
}
static av_always_inline int float_to_int16_one(const float *src){
- int_fast32_t tmp = *(const int32_t*)src;
- if(tmp & 0xf0000){
- tmp = (0x43c0ffff - tmp)>>31;
- // is this faster on some gcc/cpu combinations?
-// if(tmp > 0x43c0ffff) tmp = 0xFFFF;
-// else tmp = 0;
- }
- return tmp - 0x8000;
+ return av_clip_int16(lrintf(*src));
}
-void ff_float_to_int16_c(int16_t *dst, const float *src, long len){
+static void ff_float_to_int16_c(int16_t *dst, const float *src, long len){
int i;
for(i=0; i<len; i++)
dst[i] = float_to_int16_one(src+i);
}
-void ff_float_to_int16_interleave_c(int16_t *dst, const float **src, long len, int channels){
+static void ff_float_to_int16_interleave_c(int16_t *dst, const float **src, long len, int channels){
int i,j,c;
if(channels==2){
for(i=0; i<len; i++){
}
}
-static int32_t scalarproduct_int16_c(int16_t * v1, int16_t * v2, int order, int shift)
+static int32_t scalarproduct_int16_c(const int16_t * v1, const int16_t * v2, int order, int shift)
{
int res = 0;
return res;
}
-static int32_t scalarproduct_and_madd_int16_c(int16_t *v1, int16_t *v2, int16_t *v3, int order, int mul)
+static int32_t scalarproduct_and_madd_int16_c(int16_t *v1, const int16_t *v2, const int16_t *v3, int order, int mul)
{
int res = 0;
while (order--) {
c->add_pixels8 = add_pixels8_c;
c->add_pixels4 = add_pixels4_c;
c->sum_abs_dctelem = sum_abs_dctelem_c;
+ c->emulated_edge_mc = ff_emulated_edge_mc;
c->gmc1 = gmc1_c;
c->gmc = ff_gmc_c;
c->clear_block = clear_block_c;
c->draw_edges = draw_edges_c;
-#if CONFIG_CAVS_DECODER
- ff_cavsdsp_init(c,avctx);
-#endif
-
#if CONFIG_MLP_DECODER || CONFIG_TRUEHD_DECODER
ff_mlp_init(c, avctx);
#endif
c->avg_rv40_qpel_pixels_tab[1][15] = avg_rv40_qpel8_mc33_c;
#endif
- c->put_mspel_pixels_tab[0]= put_mspel8_mc00_c;
+ c->put_mspel_pixels_tab[0]= ff_put_pixels8x8_c;
c->put_mspel_pixels_tab[1]= put_mspel8_mc10_c;
c->put_mspel_pixels_tab[2]= put_mspel8_mc20_c;
c->put_mspel_pixels_tab[3]= put_mspel8_mc30_c;
c->vp3_v_loop_filter= ff_vp3_v_loop_filter_c;
c->vp3_idct_dc_add= ff_vp3_idct_dc_add_c;
}
- if (CONFIG_VP6_DECODER) {
- c->vp6_filter_diag4= ff_vp6_filter_diag4_c;
- }
c->h261_loop_filter= h261_loop_filter_c;
#endif
#if CONFIG_AC3_DECODER
c->ac3_downmix = ff_ac3_downmix_c;
-#endif
-#if CONFIG_LPC
- c->lpc_compute_autocorr = ff_lpc_compute_autocorr;
#endif
c->vector_fmul = vector_fmul_c;
c->vector_fmul_reverse = vector_fmul_reverse_c;
c->vector_fmul_add = vector_fmul_add_c;
- c->vector_fmul_window = ff_vector_fmul_window_c;
+ c->vector_fmul_window = vector_fmul_window_c;
c->int32_to_float_fmul_scalar = int32_to_float_fmul_scalar_c;
c->vector_clipf = vector_clipf_c;
c->float_to_int16 = ff_float_to_int16_c;
c->sv_fmul_scalar[0] = sv_fmul_scalar_2_c;
c->sv_fmul_scalar[1] = sv_fmul_scalar_4_c;
- c->shrink[0]= ff_img_copy_plane;
+ c->shrink[0]= av_image_copy_plane;
c->shrink[1]= ff_shrink22;
c->shrink[2]= ff_shrink44;
c->shrink[3]= ff_shrink88;
c->avg_2tap_qpel_pixels_tab[0][i]= c->avg_h264_qpel_pixels_tab[0][i];
}
+ c->put_rv30_tpel_pixels_tab[0][0] = c->put_h264_qpel_pixels_tab[0][0];
+ c->put_rv30_tpel_pixels_tab[1][0] = c->put_h264_qpel_pixels_tab[1][0];
+ c->avg_rv30_tpel_pixels_tab[0][0] = c->avg_h264_qpel_pixels_tab[0][0];
+ c->avg_rv30_tpel_pixels_tab[1][0] = c->avg_h264_qpel_pixels_tab[1][0];
+
+ c->put_rv40_qpel_pixels_tab[0][0] = c->put_h264_qpel_pixels_tab[0][0];
+ c->put_rv40_qpel_pixels_tab[1][0] = c->put_h264_qpel_pixels_tab[1][0];
+ c->avg_rv40_qpel_pixels_tab[0][0] = c->avg_h264_qpel_pixels_tab[0][0];
+ c->avg_rv40_qpel_pixels_tab[1][0] = c->avg_h264_qpel_pixels_tab[1][0];
+
switch(c->idct_permutation_type){
case FF_NO_IDCT_PERM:
for(i=0; i<64; i++)