/*
* 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"
};
/* 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,
#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;\
}
#endif /* CONFIG_CAVS_DECODER */
-#if CONFIG_VC1_DECODER || 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);
}
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];
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];
dst[i+0] = src1[i+0]-src2[i+0];
}
+static void add_hfyu_median_prediction_c(uint8_t *dst, uint8_t *src1, 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, uint8_t *src1, uint8_t *src2, int w, int *left, int *left_top){
int i;
uint8_t l, lt;
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;
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 int did_fail=0;
DECLARE_ALIGNED_16(int, aligned);
- if((long)&aligned & 15){
+ if((intptr_t)&aligned & 15){
if(!did_fail){
#if HAVE_MMX || HAVE_ALTIVEC
av_log(NULL, AV_LOG_ERROR,
#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->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;
#if CONFIG_CAVS_DECODER
ff_cavsdsp_init(c,avctx);
#endif
-#if CONFIG_VC1_DECODER || 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 CONFIG_WMV2_DECODER || CONFIG_VC1_DECODER || CONFIG_WMV3_DECODER
+#if CONFIG_WMV2_DECODER || CONFIG_VC1_DECODER
ff_intrax8dsp_init(c,avctx);
#endif
-#if CONFIG_H264_ENCODER
- ff_h264dspenc_init(c,avctx);
-#endif
#if CONFIG_RV30_DECODER
ff_rv30dsp_init(c,avctx);
#endif
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)
#if CONFIG_GPL
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;
#if CONFIG_SNOW_ENCODER
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->bswap_buf= bswap_buf;
#if CONFIG_PNG_DECODER
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->vector_fmul_add_add = ff_vector_fmul_add_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;
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_PPC) 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