s = 0;
for (i = 0; i < 16; i++) {
for (j = 0; j < 16; j += 8) {
-#if LONG_MAX > 2147483647
+#if 0
+ s += sq[pix[0]];
+ s += sq[pix[1]];
+ s += sq[pix[2]];
+ s += sq[pix[3]];
+ s += sq[pix[4]];
+ s += sq[pix[5]];
+ s += sq[pix[6]];
+ s += sq[pix[7]];
+#else
+#if HAVE_FAST_64BIT
register uint64_t x=*(uint64_t*)pix;
s += sq[x&0xff];
s += sq[(x>>8)&0xff];
s += sq[(x>>8)&0xff];
s += sq[(x>>16)&0xff];
s += sq[(x>>24)&0xff];
+#endif
#endif
pix += 8;
}
return s;
}
-static void get_pixels_c(DCTELEM *restrict block, const uint8_t *pixels, int line_size)
-{
- int i;
-
- /* read the pixels */
- for(i=0;i<8;i++) {
- block[0] = pixels[0];
- block[1] = pixels[1];
- block[2] = pixels[2];
- block[3] = pixels[3];
- block[4] = pixels[4];
- block[5] = pixels[5];
- block[6] = pixels[6];
- block[7] = pixels[7];
- pixels += line_size;
- block += 8;
- }
-}
-
static void diff_pixels_c(DCTELEM *restrict block, const uint8_t *s1,
const uint8_t *s2, int stride){
int i;
}
}
-static void put_pixels_nonclamped_c(const DCTELEM *block, uint8_t *restrict pixels,
- int line_size)
-{
- int i;
-
- /* read the pixels */
- for(i=0;i<8;i++) {
- pixels[0] = block[0];
- pixels[1] = block[1];
- pixels[2] = block[2];
- pixels[3] = block[3];
- pixels[4] = block[4];
- pixels[5] = block[5];
- pixels[6] = block[6];
- pixels[7] = block[7];
-
- pixels += line_size;
- block += 8;
- }
-}
-
void ff_add_pixels_clamped_c(const DCTELEM *block, uint8_t *restrict pixels,
int line_size)
{
}
}
-static void scale_block_c(const uint8_t src[64]/*align 8*/, uint8_t *dst/*align 8*/, int linesize)
-{
- int i, j;
- uint16_t *dst1 = (uint16_t *) dst;
- uint16_t *dst2 = (uint16_t *)(dst + linesize);
-
- for (j = 0; j < 8; j++) {
- for (i = 0; i < 8; i++) {
- dst1[i] = dst2[i] = src[i] * 0x0101;
- }
- src += 8;
- dst1 += linesize;
- dst2 += linesize;
- }
-}
-
#define avg2(a,b) ((a+b+1)>>1)
#define avg4(a,b,c,d) ((a+b+c+d+2)>>2)
}
#if CONFIG_RV40_DECODER
-static void put_rv40_qpel16_mc33_c(uint8_t *dst, uint8_t *src, int stride){
+void ff_put_rv40_qpel16_mc33_c(uint8_t *dst, uint8_t *src, int stride){
put_pixels16_xy2_8_c(dst, src, stride, 16);
}
-static void avg_rv40_qpel16_mc33_c(uint8_t *dst, uint8_t *src, int stride){
+void ff_avg_rv40_qpel16_mc33_c(uint8_t *dst, uint8_t *src, int stride){
avg_pixels16_xy2_8_c(dst, src, stride, 16);
}
-static void put_rv40_qpel8_mc33_c(uint8_t *dst, uint8_t *src, int stride){
+void ff_put_rv40_qpel8_mc33_c(uint8_t *dst, uint8_t *src, int stride){
put_pixels8_xy2_8_c(dst, src, stride, 8);
}
-static void avg_rv40_qpel8_mc33_c(uint8_t *dst, uint8_t *src, int stride){
+void ff_avg_rv40_qpel8_mc33_c(uint8_t *dst, uint8_t *src, int stride){
avg_pixels8_xy2_8_c(dst, src, stride, 8);
}
#endif /* CONFIG_RV40_DECODER */
s->block_last_index[0/*FIXME*/]= s->fast_dct_quantize(s, temp, 0/*FIXME*/, s->qscale, &i);
s->dct_unquantize_inter(s, temp, 0, s->qscale);
- ff_simple_idct(temp); //FIXME
+ ff_simple_idct_8(temp); //FIXME
for(i=0; i<64; i++)
sum+= (temp[i]-bak[i])*(temp[i]-bak[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 ff_check_alignment(void){
static int did_fail=0;
- LOCAL_ALIGNED_16(int, aligned);
+ LOCAL_ALIGNED_16(int, aligned, [4]);
- if((intptr_t)&aligned & 15){
+ if((intptr_t)aligned & 15){
if(!did_fail){
#if HAVE_MMX || HAVE_ALTIVEC
av_log(NULL, AV_LOG_ERROR,
ff_check_alignment();
#if CONFIG_ENCODERS
- if(avctx->dct_algo==FF_DCT_FASTINT) {
- c->fdct = fdct_ifast;
- c->fdct248 = fdct_ifast248;
- }
- else if(avctx->dct_algo==FF_DCT_FAAN) {
- c->fdct = ff_faandct;
- c->fdct248 = ff_faandct248;
- }
- else {
- c->fdct = ff_jpeg_fdct_islow; //slow/accurate/default
- c->fdct248 = ff_fdct248_islow;
+ if (avctx->bits_per_raw_sample == 10) {
+ c->fdct = ff_jpeg_fdct_islow_10;
+ c->fdct248 = ff_fdct248_islow_10;
+ } else {
+ if(avctx->dct_algo==FF_DCT_FASTINT) {
+ c->fdct = fdct_ifast;
+ c->fdct248 = fdct_ifast248;
+ }
+ else if(avctx->dct_algo==FF_DCT_FAAN) {
+ c->fdct = ff_faandct;
+ c->fdct248 = ff_faandct248;
+ }
+ else {
+ c->fdct = ff_jpeg_fdct_islow_8; //slow/accurate/default
+ c->fdct248 = ff_fdct248_islow_8;
+ }
}
#endif //CONFIG_ENCODERS
if(avctx->lowres==1){
- 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{
- if (avctx->codec_id != CODEC_ID_H264) {
- c->idct_put= ff_h264_lowres_idct_put_8_c;
- c->idct_add= ff_h264_lowres_idct_add_8_c;
- } else {
- switch (avctx->bits_per_raw_sample) {
- case 9:
- c->idct_put= ff_h264_lowres_idct_put_9_c;
- c->idct_add= ff_h264_lowres_idct_add_9_c;
- break;
- case 10:
- c->idct_put= ff_h264_lowres_idct_put_10_c;
- c->idct_add= ff_h264_lowres_idct_add_10_c;
- break;
- default:
- c->idct_put= ff_h264_lowres_idct_put_8_c;
- c->idct_add= ff_h264_lowres_idct_add_8_c;
- }
- }
- }
+ c->idct_put= ff_jref_idct4_put;
+ c->idct_add= ff_jref_idct4_add;
c->idct = j_rev_dct4;
c->idct_permutation_type= FF_NO_IDCT_PERM;
}else if(avctx->lowres==2){
c->idct = j_rev_dct1;
c->idct_permutation_type= FF_NO_IDCT_PERM;
}else{
+ if (avctx->bits_per_raw_sample == 10) {
+ c->idct_put = ff_simple_idct_put_10;
+ c->idct_add = ff_simple_idct_add_10;
+ c->idct = ff_simple_idct_10;
+ c->idct_permutation_type = FF_NO_IDCT_PERM;
+ } else {
if(avctx->idct_algo==FF_IDCT_INT){
c->idct_put= ff_jref_idct_put;
c->idct_add= ff_jref_idct_add;
}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 if(CONFIG_BINK_DECODER && avctx->idct_algo==FF_IDCT_BINK) {
- c->idct = ff_bink_idct_c;
- c->idct_add = ff_bink_idct_add_c;
- c->idct_put = ff_bink_idct_put_c;
- c->idct_permutation_type = FF_NO_IDCT_PERM;
}else{ //accurate/default
- c->idct_put= ff_simple_idct_put;
- c->idct_add= ff_simple_idct_add;
- c->idct = ff_simple_idct;
+ c->idct_put = ff_simple_idct_put_8;
+ c->idct_add = ff_simple_idct_add_8;
+ c->idct = ff_simple_idct_8;
c->idct_permutation_type= FF_NO_IDCT_PERM;
}
+ }
}
- c->get_pixels = get_pixels_c;
c->diff_pixels = diff_pixels_c;
c->put_pixels_clamped = ff_put_pixels_clamped_c;
c->put_signed_pixels_clamped = ff_put_signed_pixels_clamped_c;
- c->put_pixels_nonclamped = put_pixels_nonclamped_c;
c->add_pixels_clamped = ff_add_pixels_clamped_c;
c->sum_abs_dctelem = sum_abs_dctelem_c;
c->gmc1 = gmc1_c;
c->fill_block_tab[0] = fill_block16_c;
c->fill_block_tab[1] = fill_block8_c;
- c->scale_block = scale_block_c;
/* TODO [0] 16 [1] 8 */
c->pix_abs[0][0] = pix_abs16_c;
#if CONFIG_WMV2_DECODER || CONFIG_VC1_DECODER
ff_intrax8dsp_init(c,avctx);
#endif
-#if CONFIG_RV30_DECODER
- ff_rv30dsp_init(c,avctx);
-#endif
-#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;
- c->put_rv40_qpel_pixels_tab[1][15] = put_rv40_qpel8_mc33_c;
- c->avg_rv40_qpel_pixels_tab[1][15] = avg_rv40_qpel8_mc33_c;
-#endif
c->put_mspel_pixels_tab[0]= ff_put_pixels8x8_c;
c->put_mspel_pixels_tab[1]= put_mspel8_mc10_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]= av_image_copy_plane;
c->shrink[1]= ff_shrink22;
c->shrink[2]= ff_shrink44;
c->PFX ## _pixels_tab[IDX][15] = FUNCC(PFX ## NUM ## _mc33, depth)
-#define BIT_DEPTH_FUNCS(depth)\
+#define BIT_DEPTH_FUNCS(depth, dct)\
+ c->get_pixels = FUNCC(get_pixels ## dct , depth);\
c->draw_edges = FUNCC(draw_edges , depth);\
c->emulated_edge_mc = FUNC (ff_emulated_edge_mc , depth);\
- c->clear_block = FUNCC(clear_block , depth);\
- c->clear_blocks = FUNCC(clear_blocks , depth);\
- c->add_pixels8 = FUNCC(add_pixels8 , depth);\
- c->add_pixels4 = FUNCC(add_pixels4 , depth);\
+ c->clear_block = FUNCC(clear_block ## dct , depth);\
+ c->clear_blocks = FUNCC(clear_blocks ## dct , depth);\
+ c->add_pixels8 = FUNCC(add_pixels8 ## dct , depth);\
+ c->add_pixels4 = FUNCC(add_pixels4 ## dct , depth);\
c->put_no_rnd_pixels_l2[0] = FUNCC(put_no_rnd_pixels16_l2, depth);\
c->put_no_rnd_pixels_l2[1] = FUNCC(put_no_rnd_pixels8_l2 , depth);\
\
dspfunc2(avg_h264_qpel, 1, 8, depth);\
dspfunc2(avg_h264_qpel, 2, 4, depth);
- if (avctx->codec_id != CODEC_ID_H264 || avctx->bits_per_raw_sample == 8) {
- BIT_DEPTH_FUNCS(8)
- } else {
- switch (avctx->bits_per_raw_sample) {
- case 9:
- BIT_DEPTH_FUNCS(9)
- break;
- case 10:
- BIT_DEPTH_FUNCS(10)
- break;
- default:
- av_log(avctx, AV_LOG_DEBUG, "Unsupported bit depth: %d\n", avctx->bits_per_raw_sample);
- BIT_DEPTH_FUNCS(8)
- break;
+ switch (avctx->bits_per_raw_sample) {
+ case 9:
+ if (c->dct_bits == 32) {
+ BIT_DEPTH_FUNCS(9, _32);
+ } else {
+ BIT_DEPTH_FUNCS(9, _16);
+ }
+ break;
+ case 10:
+ if (c->dct_bits == 32) {
+ BIT_DEPTH_FUNCS(10, _32);
+ } else {
+ BIT_DEPTH_FUNCS(10, _16);
}
+ break;
+ default:
+ av_log(avctx, AV_LOG_DEBUG, "Unsupported bit depth: %d\n", avctx->bits_per_raw_sample);
+ case 8:
+ BIT_DEPTH_FUNCS(8, _16);
+ break;
}
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++)
av_log(avctx, AV_LOG_ERROR, "Internal error, IDCT permutation not set\n");
}
}
-
projects / ffmpeg / blobdiff