/*
* DSP utils
- * Copyright (c) 2000, 2001 Fabrice Bellard.
+ * Copyright (c) 2000, 2001 Fabrice Bellard
* Copyright (c) 2002-2004 Michael Niedermayer <michaelni@gmx.at>
*
* gmc & q-pel & 32/64 bit based MC by Michael Niedermayer <michaelni@gmx.at>
*/
/**
- * @file dsputil.c
+ * @file libavcodec/dsputil.c
* DSP utils
*/
#include "simple_idct.h"
#include "faandct.h"
#include "faanidct.h"
+#include "mathops.h"
#include "h263.h"
#include "snow.h"
+#include "mpegvideo.h"
+#include "config.h"
/* snow.c */
void ff_spatial_dwt(int *buffer, int width, int height, int stride, int type, int decomposition_count);
/* ac3dec.c */
void ff_ac3_downmix_c(float (*samples)[256], float (*matrix)[2], int out_ch, int in_ch, int len);
-/* flacenc.c */
-void ff_flac_compute_autocorr(const int32_t *data, int len, int lag, double *autoc);
+/* lpc.c */
+void ff_lpc_compute_autocorr(const int32_t *data, int len, int lag, double *autoc);
/* pngdec.c */
void ff_add_png_paeth_prediction(uint8_t *dst, uint8_t *src, uint8_t *top, int w, int bpp);
};
/* not permutated inverse zigzag_direct + 1 for MMX quantizer */
-DECLARE_ALIGNED_8(uint16_t, inv_zigzag_direct16[64]) = {0, };
+DECLARE_ALIGNED_16(uint16_t, inv_zigzag_direct16[64]);
const uint8_t ff_alternate_horizontal_scan[64] = {
0, 1, 2, 3, 8, 9, 16, 17,
38, 46, 54, 62, 39, 47, 55, 63,
};
-/* a*inverse[b]>>32 == a/b for all 0<=a<=65536 && 2<=b<=255 */
-const uint32_t ff_inverse[256]={
+/* 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,
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 */
int j;
j = src_scantable[i];
st->permutated[i] = permutation[j];
-#ifdef ARCH_POWERPC
+#if ARCH_PPC
st->inverse[j] = i;
#endif
}
}
-#ifdef CONFIG_SNOW_ENCODER //dwt is in snow.c
+#if CONFIG_SNOW_ENCODER //dwt is in snow.c
static inline int w_c(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, int w, int h, int type){
int s, i, j;
const int dec_count= w==8 ? 3 : 4;
#undef op_avg
#undef op_put
-static void put_no_rnd_h264_chroma_mc8_c(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){
+static void put_no_rnd_vc1_chroma_mc8_c(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){
const int A=(8-x)*(8-y);
const int B=( x)*(8-y);
const int C=(8-x)*( y);
}
}
+static void avg_no_rnd_vc1_chroma_mc8_c(uint8_t *dst/*align 8*/, uint8_t *src/*align 1*/, int stride, int h, int x, int y){
+ const int A=(8-x)*(8-y);
+ const int B=( x)*(8-y);
+ const int C=(8-x)*( y);
+ const int D=( x)*( y);
+ int i;
+
+ assert(x<8 && y<8 && x>=0 && y>=0);
+
+ for(i=0; i<h; i++)
+ {
+ dst[0] = avg2(dst[0], ((A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1] + 32 - 4) >> 6));
+ dst[1] = avg2(dst[1], ((A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2] + 32 - 4) >> 6));
+ dst[2] = avg2(dst[2], ((A*src[2] + B*src[3] + C*src[stride+2] + D*src[stride+3] + 32 - 4) >> 6));
+ dst[3] = avg2(dst[3], ((A*src[3] + B*src[4] + C*src[stride+3] + D*src[stride+4] + 32 - 4) >> 6));
+ dst[4] = avg2(dst[4], ((A*src[4] + B*src[5] + C*src[stride+4] + D*src[stride+5] + 32 - 4) >> 6));
+ dst[5] = avg2(dst[5], ((A*src[5] + B*src[6] + C*src[stride+5] + D*src[stride+6] + 32 - 4) >> 6));
+ dst[6] = avg2(dst[6], ((A*src[6] + B*src[7] + C*src[stride+6] + D*src[stride+7] + 32 - 4) >> 6));
+ dst[7] = avg2(dst[7], ((A*src[7] + B*src[8] + C*src[stride+7] + D*src[stride+8] + 32 - 4) >> 6));
+ dst+= stride;
+ src+= stride;
+ }
+}
+
#define QPEL_MC(r, OPNAME, RND, OP) \
static void OPNAME ## mpeg4_qpel8_h_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h){\
uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;\
}
}
-#ifdef CONFIG_CAVS_DECODER
+#if CONFIG_CAVS_DECODER
/* AVS specific */
void ff_cavsdsp_init(DSPContext* c, AVCodecContext *avctx);
}
#endif /* CONFIG_CAVS_DECODER */
-#if defined(CONFIG_VC1_DECODER) || defined(CONFIG_WMV3_DECODER)
+void ff_mlp_init(DSPContext* c, AVCodecContext *avctx);
+
+#if CONFIG_VC1_DECODER
/* VC-1 specific */
void ff_vc1dsp_init(DSPContext* c, AVCodecContext *avctx);
void ff_put_vc1_mspel_mc00_c(uint8_t *dst, uint8_t *src, int stride, int rnd) {
put_pixels8_c(dst, src, stride, 8);
}
-#endif /* CONFIG_VC1_DECODER||CONFIG_WMV3_DECODER */
+void ff_avg_vc1_mspel_mc00_c(uint8_t *dst, uint8_t *src, int stride, int rnd) {
+ avg_pixels8_c(dst, src, stride, 8);
+}
+#endif /* CONFIG_VC1_DECODER */
void ff_intrax8dsp_init(DSPContext* c, AVCodecContext *avctx);
/* H264 specific */
void ff_h264dspenc_init(DSPContext* c, AVCodecContext *avctx);
-#if defined(CONFIG_RV40_DECODER)
+#if CONFIG_RV30_DECODER
+void ff_rv30dsp_init(DSPContext* c, AVCodecContext *avctx);
+#endif /* CONFIG_RV30_DECODER */
+
+#if CONFIG_RV40_DECODER
static void put_rv40_qpel16_mc33_c(uint8_t *dst, uint8_t *src, int stride){
put_pixels16_xy2_c(dst, src, stride, 16);
}
}
static void h263_v_loop_filter_c(uint8_t *src, int stride, int qscale){
- if(ENABLE_ANY_H263) {
+ if(CONFIG_ANY_H263) {
int x;
const int strength= ff_h263_loop_filter_strength[qscale];
}
static void h263_h_loop_filter_c(uint8_t *src, int stride, int qscale){
- if(ENABLE_ANY_H263) {
+ if(CONFIG_ANY_H263) {
int y;
const int strength= ff_h263_loop_filter_strength[qscale];
h264_loop_filter_luma_c(pix, 1, stride, alpha, beta, tc0);
}
+static inline void h264_loop_filter_luma_intra_c(uint8_t *pix, int xstride, int ystride, int alpha, int beta)
+{
+ int d;
+ for( d = 0; d < 16; d++ ) {
+ const int p2 = pix[-3*xstride];
+ const int p1 = pix[-2*xstride];
+ const int p0 = pix[-1*xstride];
+
+ const int q0 = pix[ 0*xstride];
+ const int q1 = pix[ 1*xstride];
+ const int q2 = pix[ 2*xstride];
+
+ if( FFABS( p0 - q0 ) < alpha &&
+ FFABS( p1 - p0 ) < beta &&
+ FFABS( q1 - q0 ) < beta ) {
+
+ if(FFABS( p0 - q0 ) < (( alpha >> 2 ) + 2 )){
+ if( FFABS( p2 - p0 ) < beta)
+ {
+ const int p3 = pix[-4*xstride];
+ /* p0', p1', p2' */
+ pix[-1*xstride] = ( p2 + 2*p1 + 2*p0 + 2*q0 + q1 + 4 ) >> 3;
+ pix[-2*xstride] = ( p2 + p1 + p0 + q0 + 2 ) >> 2;
+ pix[-3*xstride] = ( 2*p3 + 3*p2 + p1 + p0 + q0 + 4 ) >> 3;
+ } else {
+ /* p0' */
+ pix[-1*xstride] = ( 2*p1 + p0 + q1 + 2 ) >> 2;
+ }
+ if( FFABS( q2 - q0 ) < beta)
+ {
+ const int q3 = pix[3*xstride];
+ /* q0', q1', q2' */
+ pix[0*xstride] = ( p1 + 2*p0 + 2*q0 + 2*q1 + q2 + 4 ) >> 3;
+ pix[1*xstride] = ( p0 + q0 + q1 + q2 + 2 ) >> 2;
+ pix[2*xstride] = ( 2*q3 + 3*q2 + q1 + q0 + p0 + 4 ) >> 3;
+ } else {
+ /* q0' */
+ pix[0*xstride] = ( 2*q1 + q0 + p1 + 2 ) >> 2;
+ }
+ }else{
+ /* p0', q0' */
+ pix[-1*xstride] = ( 2*p1 + p0 + q1 + 2 ) >> 2;
+ pix[ 0*xstride] = ( 2*q1 + q0 + p1 + 2 ) >> 2;
+ }
+ }
+ pix += ystride;
+ }
+}
+static void h264_v_loop_filter_luma_intra_c(uint8_t *pix, int stride, int alpha, int beta)
+{
+ h264_loop_filter_luma_intra_c(pix, stride, 1, alpha, beta);
+}
+static void h264_h_loop_filter_luma_intra_c(uint8_t *pix, int stride, int alpha, int beta)
+{
+ h264_loop_filter_luma_intra_c(pix, 1, stride, alpha, beta);
+}
+
static inline void h264_loop_filter_chroma_c(uint8_t *pix, int xstride, int ystride, int alpha, int beta, int8_t *tc0)
{
int i, d;
void ff_set_cmp(DSPContext* c, me_cmp_func *cmp, int type){
int i;
- memset(cmp, 0, sizeof(void*)*5);
+ memset(cmp, 0, sizeof(void*)*6);
- for(i=0; i<5; i++){
+ for(i=0; i<6; i++){
switch(type&0xFF){
case FF_CMP_SAD:
cmp[i]= c->sad[i];
case FF_CMP_NSSE:
cmp[i]= c->nsse[i];
break;
-#ifdef CONFIG_SNOW_ENCODER
+#if CONFIG_SNOW_ENCODER
case FF_CMP_W53:
cmp[i]= c->w53[i];
break;
static void diff_bytes_c(uint8_t *dst, uint8_t *src1, uint8_t *src2, int w){
long i;
-#ifndef HAVE_FAST_UNALIGNED
+#if !HAVE_FAST_UNALIGNED
if((long)src2 & (sizeof(long)-1)){
for(i=0; i+7<w; i+=8){
dst[i+0] = src1[i+0]-src2[i+0];
dst[i+0] = src1[i+0]-src2[i+0];
}
-static void sub_hfyu_median_prediction_c(uint8_t *dst, uint8_t *src1, uint8_t *src2, int w, int *left, int *left_top){
+static void add_hfyu_median_prediction_c(uint8_t *dst, const uint8_t *src1, const uint8_t *diff, int w, int *left, int *left_top){
+ int i;
+ uint8_t l, lt;
+
+ l= *left;
+ lt= *left_top;
+
+ for(i=0; i<w; i++){
+ l= mid_pred(l, src1[i], (l + src1[i] - lt)&0xFF) + diff[i];
+ lt= src1[i];
+ dst[i]= l;
+ }
+
+ *left= l;
+ *left_top= lt;
+}
+
+static void sub_hfyu_median_prediction_c(uint8_t *dst, const uint8_t *src1, const uint8_t *src2, int w, int *left, int *left_top){
int i;
uint8_t l, lt;
*left_top= lt;
}
+static int add_hfyu_left_prediction_c(uint8_t *dst, const uint8_t *src, int w, int acc){
+ int i;
+
+ for(i=0; i<w-1; i++){
+ acc+= src[i];
+ dst[i]= acc;
+ i++;
+ acc+= src[i];
+ dst[i]= acc;
+ }
+
+ for(; i<w; i++){
+ acc+= src[i];
+ dst[i]= acc;
+ }
+
+ return acc;
+}
+
+#if HAVE_BIGENDIAN
+#define B 3
+#define G 2
+#define R 1
+#else
+#define B 0
+#define G 1
+#define R 2
+#endif
+static void add_hfyu_left_prediction_bgr32_c(uint8_t *dst, const uint8_t *src, int w, int *red, int *green, int *blue){
+ int i;
+ int r,g,b;
+ r= *red;
+ g= *green;
+ b= *blue;
+
+ for(i=0; i<w; i++){
+ b+= src[4*i+B];
+ g+= src[4*i+G];
+ r+= src[4*i+R];
+
+ dst[4*i+B]= b;
+ dst[4*i+G]= g;
+ dst[4*i+R]= r;
+ }
+
+ *red= r;
+ *green= g;
+ *blue= b;
+}
+#undef B
+#undef G
+#undef R
+
#define BUTTERFLY2(o1,o2,i1,i2) \
o1= (i1)+(i2);\
o2= (i1)-(i2);
return s->dsp.sum_abs_dctelem(temp);
}
-#ifdef CONFIG_GPL
+#if CONFIG_GPL
#define DCT8_1D {\
const int s07 = SRC(0) + SRC(7);\
const int s16 = SRC(1) + SRC(6);\
static int dct_max8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, int stride, int h){
MpegEncContext * const s= (MpegEncContext *)c;
- DECLARE_ALIGNED_8(uint64_t, aligned_temp[sizeof(DCTELEM)*64/8]);
+ DECLARE_ALIGNED_16(uint64_t, aligned_temp[sizeof(DCTELEM)*64/8]);
DCTELEM * const temp= (DCTELEM*)aligned_temp;
int sum=0, i;
static int quant_psnr8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, int stride, int h){
MpegEncContext * const s= (MpegEncContext *)c;
- DECLARE_ALIGNED_8 (uint64_t, aligned_temp[sizeof(DCTELEM)*64*2/8]);
+ DECLARE_ALIGNED_16(uint64_t, aligned_temp[sizeof(DCTELEM)*64*2/8]);
DCTELEM * const temp= (DCTELEM*)aligned_temp;
DCTELEM * const bak = ((DCTELEM*)aligned_temp)+64;
int sum=0, i;
static int rd8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, int stride, int h){
MpegEncContext * const s= (MpegEncContext *)c;
const uint8_t *scantable= s->intra_scantable.permutated;
- DECLARE_ALIGNED_8 (uint64_t, aligned_temp[sizeof(DCTELEM)*64/8]);
- DECLARE_ALIGNED_8 (uint64_t, aligned_bak[stride]);
+ DECLARE_ALIGNED_16(uint64_t, aligned_temp[sizeof(DCTELEM)*64/8]);
+ DECLARE_ALIGNED_16(uint64_t, aligned_src1[8]);
+ DECLARE_ALIGNED_16(uint64_t, aligned_src2[8]);
DCTELEM * const temp= (DCTELEM*)aligned_temp;
- uint8_t * const bak= (uint8_t*)aligned_bak;
+ uint8_t * const lsrc1 = (uint8_t*)aligned_src1;
+ uint8_t * const lsrc2 = (uint8_t*)aligned_src2;
int i, last, run, bits, level, distortion, start_i;
const int esc_length= s->ac_esc_length;
uint8_t * length;
assert(h==8);
- for(i=0; i<8; i++){
- ((uint32_t*)(bak + i*stride))[0]= ((uint32_t*)(src2 + i*stride))[0];
- ((uint32_t*)(bak + i*stride))[1]= ((uint32_t*)(src2 + i*stride))[1];
- }
+ copy_block8(lsrc1, src1, 8, stride, 8);
+ copy_block8(lsrc2, src2, 8, stride, 8);
- s->dsp.diff_pixels(temp, src1, src2, stride);
+ s->dsp.diff_pixels(temp, lsrc1, lsrc2, 8);
s->block_last_index[0/*FIXME*/]= last= s->fast_dct_quantize(s, temp, 0/*FIXME*/, s->qscale, &i);
s->dct_unquantize_inter(s, temp, 0, s->qscale);
}
- s->dsp.idct_add(bak, stride, temp);
+ s->dsp.idct_add(lsrc2, 8, temp);
- distortion= s->dsp.sse[1](NULL, bak, src1, stride, 8);
+ distortion= s->dsp.sse[1](NULL, lsrc2, lsrc1, 8, 8);
return distortion + ((bits*s->qscale*s->qscale*109 + 64)>>7);
}
static int bit8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2, int stride, int h){
MpegEncContext * const s= (MpegEncContext *)c;
const uint8_t *scantable= s->intra_scantable.permutated;
- DECLARE_ALIGNED_8 (uint64_t, aligned_temp[sizeof(DCTELEM)*64/8]);
+ DECLARE_ALIGNED_16(uint64_t, aligned_temp[sizeof(DCTELEM)*64/8]);
DCTELEM * const temp= (DCTELEM*)aligned_temp;
int i, last, run, bits, level, start_i;
const int esc_length= s->ac_esc_length;
return bits;
}
-static int vsad_intra16_c(/*MpegEncContext*/ void *c, uint8_t *s, uint8_t *dummy, int stride, int h){
- int score=0;
- int x,y;
-
- for(y=1; y<h; y++){
- for(x=0; x<16; x+=4){
- score+= FFABS(s[x ] - s[x +stride]) + FFABS(s[x+1] - s[x+1+stride])
- +FFABS(s[x+2] - s[x+2+stride]) + FFABS(s[x+3] - s[x+3+stride]);
- }
- s+= stride;
- }
-
- return score;
-}
+#define VSAD_INTRA(size) \
+static int vsad_intra##size##_c(/*MpegEncContext*/ void *c, uint8_t *s, uint8_t *dummy, int stride, int h){ \
+ int score=0; \
+ int x,y; \
+ \
+ for(y=1; y<h; y++){ \
+ for(x=0; x<size; x+=4){ \
+ score+= FFABS(s[x ] - s[x +stride]) + FFABS(s[x+1] - s[x+1+stride]) \
+ +FFABS(s[x+2] - s[x+2+stride]) + FFABS(s[x+3] - s[x+3+stride]); \
+ } \
+ s+= stride; \
+ } \
+ \
+ return score; \
+}
+VSAD_INTRA(8)
+VSAD_INTRA(16)
static int vsad16_c(/*MpegEncContext*/ void *c, uint8_t *s1, uint8_t *s2, int stride, int h){
int score=0;
}
#define SQ(a) ((a)*(a))
-static int vsse_intra16_c(/*MpegEncContext*/ void *c, uint8_t *s, uint8_t *dummy, int stride, int h){
- int score=0;
- int x,y;
-
- for(y=1; y<h; y++){
- for(x=0; x<16; x+=4){
- score+= SQ(s[x ] - s[x +stride]) + SQ(s[x+1] - s[x+1+stride])
- +SQ(s[x+2] - s[x+2+stride]) + SQ(s[x+3] - s[x+3+stride]);
- }
- s+= stride;
- }
-
- return score;
-}
+#define VSSE_INTRA(size) \
+static int vsse_intra##size##_c(/*MpegEncContext*/ void *c, uint8_t *s, uint8_t *dummy, int stride, int h){ \
+ int score=0; \
+ int x,y; \
+ \
+ for(y=1; y<h; y++){ \
+ for(x=0; x<size; x+=4){ \
+ score+= SQ(s[x ] - s[x +stride]) + SQ(s[x+1] - s[x+1+stride]) \
+ +SQ(s[x+2] - s[x+2+stride]) + SQ(s[x+3] - s[x+3+stride]); \
+ } \
+ s+= stride; \
+ } \
+ \
+ return score; \
+}
+VSSE_INTRA(8)
+VSSE_INTRA(16)
static int vsse16_c(/*MpegEncContext*/ void *c, uint8_t *s1, uint8_t *s2, int stride, int h){
int score=0;
WRAPPER8_16_SQ(hadamard8_diff8x8_c, hadamard8_diff16_c)
WRAPPER8_16_SQ(hadamard8_intra8x8_c, hadamard8_intra16_c)
WRAPPER8_16_SQ(dct_sad8x8_c, dct_sad16_c)
-#ifdef CONFIG_GPL
+#if CONFIG_GPL
WRAPPER8_16_SQ(dct264_sad8x8_c, dct264_sad16_c)
#endif
WRAPPER8_16_SQ(dct_max8x8_c, dct_max16_c)
dst[i] = src0[i] * src1[-i];
}
-void ff_vector_fmul_add_add_c(float *dst, const float *src0, const float *src1, const float *src2, int src3, int len, int step){
+static void vector_fmul_add_c(float *dst, const float *src0, const float *src1, const float *src2, int len){
int i;
for(i=0; i<len; i++)
- dst[i*step] = src0[i] * src1[i] + src2[i] + src3;
+ 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_scalar_c(float *dst, const float *src, float mul,
+ int len)
+{
+ int i;
+ for (i = 0; i < len; i++)
+ dst[i] = src[i] * mul;
+}
+
+static void vector_fmul_sv_scalar_2_c(float *dst, const float *src,
+ const float **sv, float mul, int len)
+{
+ int i;
+ for (i = 0; i < len; i += 2, sv++) {
+ dst[i ] = src[i ] * sv[0][0] * mul;
+ dst[i+1] = src[i+1] * sv[0][1] * mul;
+ }
+}
+
+static void vector_fmul_sv_scalar_4_c(float *dst, const float *src,
+ const float **sv, float mul, int len)
+{
+ int i;
+ for (i = 0; i < len; i += 4, sv++) {
+ dst[i ] = src[i ] * sv[0][0] * mul;
+ dst[i+1] = src[i+1] * sv[0][1] * mul;
+ dst[i+2] = src[i+2] * sv[0][2] * mul;
+ dst[i+3] = src[i+3] * sv[0][3] * mul;
+ }
+}
+
+static void sv_fmul_scalar_2_c(float *dst, const float **sv, float mul,
+ int len)
+{
+ int i;
+ for (i = 0; i < len; i += 2, sv++) {
+ dst[i ] = sv[0][0] * mul;
+ dst[i+1] = sv[0][1] * mul;
+ }
+}
+
+static void sv_fmul_scalar_4_c(float *dst, const float **sv, float mul,
+ int len)
+{
+ int i;
+ for (i = 0; i < len; i += 4, sv++) {
+ dst[i ] = sv[0][0] * mul;
+ dst[i+1] = sv[0][1] * mul;
+ dst[i+2] = sv[0][2] * mul;
+ dst[i+3] = sv[0][3] * mul;
+ }
+}
+
+static void butterflies_float_c(float *restrict v1, float *restrict v2,
+ int len)
+{
+ int i;
+ for (i = 0; i < len; i++) {
+ float t = v1[i] - v2[i];
+ v1[i] += v2[i];
+ v2[i] = t;
+ }
+}
+
+static float scalarproduct_float_c(const float *v1, const float *v2, int len)
+{
+ float p = 0.0;
+ int i;
+
+ for (i = 0; i < len; i++)
+ p += v1[i] * v2[i];
+
+ return p;
+}
+
static void int32_to_float_fmul_scalar_c(float *dst, const int *src, float mul, int len){
int i;
for(i=0; i<len; i++)
dst[i] = src[i] * mul;
}
+static inline uint32_t clipf_c_one(uint32_t a, uint32_t mini,
+ uint32_t maxi, uint32_t maxisign)
+{
+
+ if(a > mini) return mini;
+ else if((a^(1<<31)) > maxisign) return maxi;
+ else return a;
+}
+
+static void vector_clipf_c_opposite_sign(float *dst, const float *src, float *min, float *max, int len){
+ int i;
+ uint32_t mini = *(uint32_t*)min;
+ uint32_t maxi = *(uint32_t*)max;
+ uint32_t maxisign = maxi ^ (1<<31);
+ uint32_t *dsti = (uint32_t*)dst;
+ const uint32_t *srci = (const uint32_t*)src;
+ for(i=0; i<len; i+=8) {
+ dsti[i + 0] = clipf_c_one(srci[i + 0], mini, maxi, maxisign);
+ dsti[i + 1] = clipf_c_one(srci[i + 1], mini, maxi, maxisign);
+ dsti[i + 2] = clipf_c_one(srci[i + 2], mini, maxi, maxisign);
+ dsti[i + 3] = clipf_c_one(srci[i + 3], mini, maxi, maxisign);
+ dsti[i + 4] = clipf_c_one(srci[i + 4], mini, maxi, maxisign);
+ dsti[i + 5] = clipf_c_one(srci[i + 5], mini, maxi, maxisign);
+ dsti[i + 6] = clipf_c_one(srci[i + 6], mini, maxi, maxisign);
+ dsti[i + 7] = clipf_c_one(srci[i + 7], mini, maxi, maxisign);
+ }
+}
+static void vector_clipf_c(float *dst, const float *src, float min, float max, int len){
+ int i;
+ if(min < 0 && max > 0) {
+ vector_clipf_c_opposite_sign(dst, src, &min, &max, len);
+ } else {
+ for(i=0; i < len; i+=8) {
+ dst[i ] = av_clipf(src[i ], min, max);
+ dst[i + 1] = av_clipf(src[i + 1], min, max);
+ dst[i + 2] = av_clipf(src[i + 2], min, max);
+ dst[i + 3] = av_clipf(src[i + 3], min, max);
+ dst[i + 4] = av_clipf(src[i + 4], min, max);
+ dst[i + 5] = av_clipf(src[i + 5], min, max);
+ dst[i + 6] = av_clipf(src[i + 6], min, max);
+ dst[i + 7] = av_clipf(src[i + 7], min, max);
+ }
+ }
+}
+
static av_always_inline int float_to_int16_one(const float *src){
int_fast32_t tmp = *(const int32_t*)src;
if(tmp & 0xf0000){
}
}
-static void add_int16_c(int16_t * v1, int16_t * v2, int order)
-{
- while (order--)
- *v1++ += *v2++;
-}
-
-static void sub_int16_c(int16_t * v1, int16_t * v2, int order)
-{
- while (order--)
- *v1++ -= *v2++;
-}
-
static int32_t scalarproduct_int16_c(int16_t * v1, 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)
+{
+ int res = 0;
+ while (order--) {
+ res += *v1 * *v2++;
+ *v1++ += mul * *v3++;
+ }
+ return res;
+}
+
#define W0 2048
#define W1 2841 /* 2048*sqrt (2)*cos (1*pi/16) */
#define W2 2676 /* 2048*sqrt (2)*cos (2*pi/16) */
static int did_fail=0;
DECLARE_ALIGNED_16(int, aligned);
- if((long)&aligned & 15){
+ if((intptr_t)&aligned & 15){
if(!did_fail){
-#if defined(HAVE_MMX) || defined(HAVE_ALTIVEC)
+#if HAVE_MMX || HAVE_ALTIVEC
av_log(NULL, AV_LOG_ERROR,
"Compiler did not align stack variables. Libavcodec has been miscompiled\n"
"and may be very slow or crash. This is not a bug in libavcodec,\n"
ff_check_alignment();
-#ifdef CONFIG_ENCODERS
+#if CONFIG_ENCODERS
if(avctx->dct_algo==FF_DCT_FASTINT) {
c->fdct = fdct_ifast;
c->fdct248 = fdct_ifast248;
#endif //CONFIG_ENCODERS
if(avctx->lowres==1){
- if(avctx->idct_algo==FF_IDCT_INT || avctx->idct_algo==FF_IDCT_AUTO || !ENABLE_H264_DECODER){
+ if(avctx->idct_algo==FF_IDCT_INT || avctx->idct_algo==FF_IDCT_AUTO || !CONFIG_H264_DECODER){
c->idct_put= ff_jref_idct4_put;
c->idct_add= ff_jref_idct4_add;
}else{
c->idct_add= ff_jref_idct_add;
c->idct = j_rev_dct;
c->idct_permutation_type= FF_LIBMPEG2_IDCT_PERM;
- }else if((ENABLE_VP3_DECODER || ENABLE_VP5_DECODER || ENABLE_VP6_DECODER || ENABLE_THEORA_DECODER ) &&
+ }else if((CONFIG_VP3_DECODER || CONFIG_VP5_DECODER || CONFIG_VP6_DECODER ) &&
avctx->idct_algo==FF_IDCT_VP3){
c->idct_put= ff_vp3_idct_put_c;
c->idct_add= ff_vp3_idct_add_c;
c->idct_add= ff_faanidct_add;
c->idct = ff_faanidct;
c->idct_permutation_type= FF_NO_IDCT_PERM;
- }else if(ENABLE_EATGQ_DECODER && avctx->idct_algo==FF_IDCT_EA) {
+ }else if(CONFIG_EATGQ_DECODER && avctx->idct_algo==FF_IDCT_EA) {
c->idct_put= ff_ea_idct_put_c;
c->idct_permutation_type= FF_NO_IDCT_PERM;
}else{ //accurate/default
}
}
- if (ENABLE_H264_DECODER) {
+ if (CONFIG_H264_DECODER) {
c->h264_idct_add= ff_h264_idct_add_c;
c->h264_idct8_add= ff_h264_idct8_add_c;
c->h264_idct_dc_add= ff_h264_idct_dc_add_c;
c->h264_idct8_dc_add= ff_h264_idct8_dc_add_c;
+ c->h264_idct_add16 = ff_h264_idct_add16_c;
+ c->h264_idct8_add4 = ff_h264_idct8_add4_c;
+ c->h264_idct_add8 = ff_h264_idct_add8_c;
+ c->h264_idct_add16intra= ff_h264_idct_add16intra_c;
}
c->get_pixels = get_pixels_c;
c->avg_h264_chroma_pixels_tab[0]= avg_h264_chroma_mc8_c;
c->avg_h264_chroma_pixels_tab[1]= avg_h264_chroma_mc4_c;
c->avg_h264_chroma_pixels_tab[2]= avg_h264_chroma_mc2_c;
- c->put_no_rnd_h264_chroma_pixels_tab[0]= put_no_rnd_h264_chroma_mc8_c;
+ c->put_no_rnd_vc1_chroma_pixels_tab[0]= put_no_rnd_vc1_chroma_mc8_c;
+ c->avg_no_rnd_vc1_chroma_pixels_tab[0]= avg_no_rnd_vc1_chroma_mc8_c;
c->weight_h264_pixels_tab[0]= weight_h264_pixels16x16_c;
c->weight_h264_pixels_tab[1]= weight_h264_pixels16x8_c;
c->draw_edges = draw_edges_c;
-#ifdef CONFIG_CAVS_DECODER
+#if CONFIG_CAVS_DECODER
ff_cavsdsp_init(c,avctx);
#endif
-#if defined(CONFIG_VC1_DECODER) || defined(CONFIG_WMV3_DECODER)
+
+#if CONFIG_MLP_DECODER || CONFIG_TRUEHD_DECODER
+ ff_mlp_init(c, avctx);
+#endif
+#if CONFIG_VC1_DECODER
ff_vc1dsp_init(c,avctx);
#endif
-#if defined(CONFIG_WMV2_DECODER) || defined(CONFIG_VC1_DECODER) || defined(CONFIG_WMV3_DECODER)
+#if CONFIG_WMV2_DECODER || CONFIG_VC1_DECODER
ff_intrax8dsp_init(c,avctx);
#endif
-#if defined(CONFIG_H264_ENCODER)
- ff_h264dspenc_init(c,avctx);
+#if CONFIG_RV30_DECODER
+ ff_rv30dsp_init(c,avctx);
#endif
-#if defined(CONFIG_RV40_DECODER)
+#if CONFIG_RV40_DECODER
ff_rv40dsp_init(c,avctx);
c->put_rv40_qpel_pixels_tab[0][15] = put_rv40_qpel16_mc33_c;
c->avg_rv40_qpel_pixels_tab[0][15] = avg_rv40_qpel16_mc33_c;
SET_CMP_FUNC(hadamard8_diff)
c->hadamard8_diff[4]= hadamard8_intra16_c;
+ c->hadamard8_diff[5]= hadamard8_intra8x8_c;
SET_CMP_FUNC(dct_sad)
SET_CMP_FUNC(dct_max)
-#ifdef CONFIG_GPL
+#if CONFIG_GPL
SET_CMP_FUNC(dct264_sad)
#endif
c->sad[0]= pix_abs16_c;
SET_CMP_FUNC(bit)
c->vsad[0]= vsad16_c;
c->vsad[4]= vsad_intra16_c;
+ c->vsad[5]= vsad_intra8_c;
c->vsse[0]= vsse16_c;
c->vsse[4]= vsse_intra16_c;
+ c->vsse[5]= vsse_intra8_c;
c->nsse[0]= nsse16_c;
c->nsse[1]= nsse8_c;
-#ifdef CONFIG_SNOW_ENCODER
+#if CONFIG_SNOW_ENCODER
c->w53[0]= w53_16_c;
c->w53[1]= w53_8_c;
c->w97[0]= w97_16_c;
c->add_bytes= add_bytes_c;
c->add_bytes_l2= add_bytes_l2_c;
c->diff_bytes= diff_bytes_c;
+ c->add_hfyu_median_prediction= add_hfyu_median_prediction_c;
c->sub_hfyu_median_prediction= sub_hfyu_median_prediction_c;
+ c->add_hfyu_left_prediction = add_hfyu_left_prediction_c;
+ c->add_hfyu_left_prediction_bgr32 = add_hfyu_left_prediction_bgr32_c;
c->bswap_buf= bswap_buf;
-#ifdef CONFIG_PNG_DECODER
+#if CONFIG_PNG_DECODER
c->add_png_paeth_prediction= ff_add_png_paeth_prediction;
#endif
c->h264_v_loop_filter_luma= h264_v_loop_filter_luma_c;
c->h264_h_loop_filter_luma= h264_h_loop_filter_luma_c;
+ c->h264_v_loop_filter_luma_intra= h264_v_loop_filter_luma_intra_c;
+ c->h264_h_loop_filter_luma_intra= h264_h_loop_filter_luma_intra_c;
c->h264_v_loop_filter_chroma= h264_v_loop_filter_chroma_c;
c->h264_h_loop_filter_chroma= h264_h_loop_filter_chroma_c;
c->h264_v_loop_filter_chroma_intra= h264_v_loop_filter_chroma_intra_c;
c->h264_h_loop_filter_chroma_intra= h264_h_loop_filter_chroma_intra_c;
c->h264_loop_filter_strength= NULL;
- if (ENABLE_ANY_H263) {
+ if (CONFIG_ANY_H263) {
c->h263_h_loop_filter= h263_h_loop_filter_c;
c->h263_v_loop_filter= h263_v_loop_filter_c;
}
- if (ENABLE_VP3_DECODER || ENABLE_THEORA_DECODER) {
+ if (CONFIG_VP3_DECODER) {
c->vp3_h_loop_filter= ff_vp3_h_loop_filter_c;
c->vp3_v_loop_filter= ff_vp3_v_loop_filter_c;
}
+ if (CONFIG_VP6_DECODER) {
+ c->vp6_filter_diag4= ff_vp6_filter_diag4_c;
+ }
c->h261_loop_filter= h261_loop_filter_c;
c->try_8x8basis= try_8x8basis_c;
c->add_8x8basis= add_8x8basis_c;
-#ifdef CONFIG_SNOW_DECODER
+#if CONFIG_SNOW_DECODER
c->vertical_compose97i = ff_snow_vertical_compose97i;
c->horizontal_compose97i = ff_snow_horizontal_compose97i;
c->inner_add_yblock = ff_snow_inner_add_yblock;
#endif
-#ifdef CONFIG_VORBIS_DECODER
+#if CONFIG_VORBIS_DECODER
c->vorbis_inverse_coupling = vorbis_inverse_coupling;
#endif
-#ifdef CONFIG_AC3_DECODER
+#if CONFIG_AC3_DECODER
c->ac3_downmix = ff_ac3_downmix_c;
#endif
-#ifdef CONFIG_FLAC_ENCODER
- c->flac_compute_autocorr = ff_flac_compute_autocorr;
+#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_add = ff_vector_fmul_add_add_c;
+ c->vector_fmul_add = vector_fmul_add_c;
c->vector_fmul_window = ff_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->float_to_int16_interleave = ff_float_to_int16_interleave_c;
- c->add_int16 = add_int16_c;
- c->sub_int16 = sub_int16_c;
c->scalarproduct_int16 = scalarproduct_int16_c;
+ c->scalarproduct_and_madd_int16 = scalarproduct_and_madd_int16_c;
+ c->scalarproduct_float = scalarproduct_float_c;
+ c->butterflies_float = butterflies_float_c;
+ c->vector_fmul_scalar = vector_fmul_scalar_c;
+
+ c->vector_fmul_sv_scalar[0] = vector_fmul_sv_scalar_2_c;
+ c->vector_fmul_sv_scalar[1] = vector_fmul_sv_scalar_4_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[1]= ff_shrink22;
memset(c->put_2tap_qpel_pixels_tab, 0, sizeof(c->put_2tap_qpel_pixels_tab));
memset(c->avg_2tap_qpel_pixels_tab, 0, sizeof(c->avg_2tap_qpel_pixels_tab));
- if (ENABLE_MMX) dsputil_init_mmx (c, avctx);
- if (ENABLE_ARM) dsputil_init_arm (c, avctx);
- if (ENABLE_MLIB) dsputil_init_mlib (c, avctx);
- if (ENABLE_VIS) dsputil_init_vis (c, avctx);
- if (ENABLE_ALPHA) dsputil_init_alpha (c, avctx);
- if (ENABLE_POWERPC) dsputil_init_ppc (c, avctx);
- if (ENABLE_MMI) dsputil_init_mmi (c, avctx);
- if (ENABLE_SH4) dsputil_init_sh4 (c, avctx);
- if (ENABLE_BFIN) dsputil_init_bfin (c, avctx);
+ if (HAVE_MMX) dsputil_init_mmx (c, avctx);
+ if (ARCH_ARM) dsputil_init_arm (c, avctx);
+ if (CONFIG_MLIB) dsputil_init_mlib (c, avctx);
+ if (HAVE_VIS) dsputil_init_vis (c, avctx);
+ if (ARCH_ALPHA) dsputil_init_alpha (c, avctx);
+ if (ARCH_PPC) dsputil_init_ppc (c, avctx);
+ if (HAVE_MMI) dsputil_init_mmi (c, avctx);
+ if (ARCH_SH4) dsputil_init_sh4 (c, avctx);
+ if (ARCH_BFIN) dsputil_init_bfin (c, avctx);
for(i=0; i<64; i++){
if(!c->put_2tap_qpel_pixels_tab[0][i])
projects / ffmpeg / blobdiff