From e6d1a700e06ba8a985c29ba212b359c9e33bcc55 Mon Sep 17 00:00:00 2001 From: Miguel Freitas Date: Mon, 4 Dec 2006 22:25:13 +0000 Subject: trying an updated ffmpeg version (51.25.0) CVS patchset: 8405 CVS date: 2006/12/04 22:25:13 --- src/libffmpeg/libavcodec/dsputil.c | 284 +++++++++++++++++++------------------ 1 file changed, 145 insertions(+), 139 deletions(-) (limited to 'src/libffmpeg/libavcodec/dsputil.c') diff --git a/src/libffmpeg/libavcodec/dsputil.c b/src/libffmpeg/libavcodec/dsputil.c index 9b79b8659..51eddbc60 100644 --- a/src/libffmpeg/libavcodec/dsputil.c +++ b/src/libffmpeg/libavcodec/dsputil.c @@ -3,18 +3,20 @@ * Copyright (c) 2000, 2001 Fabrice Bellard. * Copyright (c) 2002-2004 Michael Niedermayer * - * This library is free software; you can redistribute it and/or + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public * License as published by the Free Software Foundation; either - * version 2 of the License, or (at your option) any later version. + * version 2.1 of the License, or (at your option) any later version. * - * This library is distributed in the hope that it will be useful, + * FFmpeg is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU * Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public - * License along with this library; if not, write to the Free Software + * License along with FFmpeg; if not, write to the Free Software * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA * * gmc & q-pel & 32/64 bit based MC by Michael Niedermayer @@ -35,8 +37,11 @@ /* snow.c */ void ff_spatial_dwt(int *buffer, int width, int height, int stride, int type, int decomposition_count); -uint8_t cropTbl[256 + 2 * MAX_NEG_CROP] = {0, }; -uint32_t squareTbl[512] = {0, }; +/* vorbis.c */ +void vorbis_inverse_coupling(float *mag, float *ang, int blocksize); + +uint8_t ff_cropTbl[256 + 2 * MAX_NEG_CROP] = {0, }; +uint32_t ff_squareTbl[512] = {0, }; const uint8_t ff_zigzag_direct[64] = { 0, 1, 8, 16, 9, 2, 3, 10, @@ -88,7 +93,7 @@ const uint8_t ff_alternate_vertical_scan[64] = { }; /* a*inverse[b]>>32 == a/b for all 0<=a<=65536 && 2<=b<=255 */ -const uint32_t inverse[256]={ +const uint32_t ff_inverse[256]={ 0, 4294967295U,2147483648U,1431655766, 1073741824, 858993460, 715827883, 613566757, 536870912, 477218589, 429496730, 390451573, 357913942, 330382100, 306783379, 286331154, 268435456, 252645136, 238609295, 226050911, 214748365, 204522253, 195225787, 186737709, @@ -160,7 +165,7 @@ static int pix_sum_c(uint8_t * pix, int line_size) static int pix_norm1_c(uint8_t * pix, int line_size) { int s, i, j; - uint32_t *sq = squareTbl + 256; + uint32_t *sq = ff_squareTbl + 256; s = 0; for (i = 0; i < 16; i++) { @@ -226,7 +231,7 @@ static void bswap_buf(uint32_t *dst, uint32_t *src, int w){ static int sse4_c(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, int h) { int s, i; - uint32_t *sq = squareTbl + 256; + uint32_t *sq = ff_squareTbl + 256; s = 0; for (i = 0; i < h; i++) { @@ -243,7 +248,7 @@ static int sse4_c(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, int h) static int sse8_c(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, int h) { int s, i; - uint32_t *sq = squareTbl + 256; + uint32_t *sq = ff_squareTbl + 256; s = 0; for (i = 0; i < h; i++) { @@ -264,7 +269,7 @@ static int sse8_c(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, int h) static int sse16_c(void *v, uint8_t *pix1, uint8_t *pix2, int line_size, int h) { int s, i; - uint32_t *sq = squareTbl + 256; + uint32_t *sq = ff_squareTbl + 256; s = 0; for (i = 0; i < h; i++) { @@ -353,7 +358,7 @@ static inline int w_c(void *v, uint8_t * pix1, uint8_t * pix2, int line_size, in for(i=0; i=0 && y>=0); -static inline void copy_block9(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, int h) -{ - int i; for(i=0; i> 6; + dst[1] = (A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2] + 32 - 4) >> 6; + dst[2] = (A*src[2] + B*src[3] + C*src[stride+2] + D*src[stride+3] + 32 - 4) >> 6; + dst[3] = (A*src[3] + B*src[4] + C*src[stride+3] + D*src[stride+4] + 32 - 4) >> 6; + dst[4] = (A*src[4] + B*src[5] + C*src[stride+4] + D*src[stride+5] + 32 - 4) >> 6; + dst[5] = (A*src[5] + B*src[6] + C*src[stride+5] + D*src[stride+6] + 32 - 4) >> 6; + dst[6] = (A*src[6] + B*src[7] + C*src[stride+6] + D*src[stride+7] + 32 - 4) >> 6; + 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 = cropTbl + MAX_NEG_CROP;\ + uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;\ int i;\ for(i=0; i>1; + ad1= FFABS(d1)>>1; d2= clip((p0-p3)/4, -ad1, ad1); @@ -2735,7 +2687,7 @@ static void h263_h_loop_filter_c(uint8_t *src, int stride, int qscale){ src[y*stride-1] = p1; src[y*stride+0] = p2; - ad1= ABS(d1)>>1; + ad1= FFABS(d1)>>1; d2= clip((p0-p3)/4, -ad1, ad1); @@ -2787,18 +2739,18 @@ static inline void h264_loop_filter_luma_c(uint8_t *pix, int xstride, int ystrid const int q1 = pix[1*xstride]; const int q2 = pix[2*xstride]; - if( ABS( p0 - q0 ) < alpha && - ABS( p1 - p0 ) < beta && - ABS( q1 - q0 ) < beta ) { + if( FFABS( p0 - q0 ) < alpha && + FFABS( p1 - p0 ) < beta && + FFABS( q1 - q0 ) < beta ) { int tc = tc0[i]; int i_delta; - if( ABS( p2 - p0 ) < beta ) { + if( FFABS( p2 - p0 ) < beta ) { pix[-2*xstride] = p1 + clip( (( p2 + ( ( p0 + q0 + 1 ) >> 1 ) ) >> 1) - p1, -tc0[i], tc0[i] ); tc++; } - if( ABS( q2 - q0 ) < beta ) { + if( FFABS( q2 - q0 ) < beta ) { pix[ xstride] = q1 + clip( (( q2 + ( ( p0 + q0 + 1 ) >> 1 ) ) >> 1) - q1, -tc0[i], tc0[i] ); tc++; } @@ -2835,9 +2787,9 @@ static inline void h264_loop_filter_chroma_c(uint8_t *pix, int xstride, int ystr const int q0 = pix[0]; const int q1 = pix[1*xstride]; - if( ABS( p0 - q0 ) < alpha && - ABS( p1 - p0 ) < beta && - ABS( q1 - q0 ) < beta ) { + if( FFABS( p0 - q0 ) < alpha && + FFABS( p1 - p0 ) < beta && + FFABS( q1 - q0 ) < beta ) { int delta = clip( (((q0 - p0 ) << 2) + (p1 - q1) + 4) >> 3, -tc, tc ); @@ -2866,9 +2818,9 @@ static inline void h264_loop_filter_chroma_intra_c(uint8_t *pix, int xstride, in const int q0 = pix[0]; const int q1 = pix[1*xstride]; - if( ABS( p0 - q0 ) < alpha && - ABS( p1 - p0 ) < beta && - ABS( q1 - q0 ) < beta ) { + if( FFABS( p0 - q0 ) < alpha && + FFABS( p1 - p0 ) < beta && + FFABS( q1 - q0 ) < beta ) { pix[-xstride] = ( 2*p1 + p0 + q1 + 2 ) >> 2; /* p0' */ pix[0] = ( 2*q1 + q0 + p1 + 2 ) >> 2; /* q0' */ @@ -3097,9 +3049,9 @@ static int nsse16_c(void *v, uint8_t *s1, uint8_t *s2, int stride, int h){ } if(y+1avctx->nsse_weight; - else return score1 + ABS(score2)*8; + if(c) return score1 + FFABS(score2)*c->avctx->nsse_weight; + else return score1 + FFABS(score2)*8; } static int nsse8_c(void *v, uint8_t *s1, uint8_t *s2, int stride, int h){ @@ -3123,9 +3075,9 @@ static int nsse8_c(void *v, uint8_t *s1, uint8_t *s2, int stride, int h){ } if(y+1avctx->nsse_weight; - else return score1 + ABS(score2)*8; + if(c) return score1 + FFABS(score2)*c->avctx->nsse_weight; + else return score1 + FFABS(score2)*8; } static int try_8x8basis_c(int16_t rem[64], int16_t weight[64], int16_t basis[64], int scale){ @@ -3324,7 +3276,7 @@ o2= (i1)-(i2); y= a-b;\ } -#define BUTTERFLYA(x,y) (ABS((x)+(y)) + ABS((x)-(y))) +#define BUTTERFLYA(x,y) (FFABS((x)+(y)) + FFABS((x)-(y))) static int hadamard8_diff8x8_c(/*MpegEncContext*/ void *s, uint8_t *dst, uint8_t *src, int stride, int h){ int i; @@ -3421,7 +3373,7 @@ static int hadamard8_intra8x8_c(/*MpegEncContext*/ void *s, uint8_t *src, uint8_ +BUTTERFLYA(temp[8*3+i], temp[8*7+i]); } - sum -= ABS(temp[8*0] + temp[8*4]); // -mean + sum -= FFABS(temp[8*0] + temp[8*4]); // -mean return sum; } @@ -3438,7 +3390,7 @@ static int dct_sad8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2 s->dsp.fdct(temp); for(i=0; i<64; i++) - sum+= ABS(temp[i]); + sum+= FFABS(temp[i]); return sum; } @@ -3487,7 +3439,7 @@ static int dct264_sad8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *s #undef DST #define SRC(x) dct[x][i] -#define DST(x,v) sum += ABS(v) +#define DST(x,v) sum += FFABS(v) for( i = 0; i < 8; i++ ) DCT8_1D #undef SRC @@ -3508,13 +3460,11 @@ static int dct_max8x8_c(/*MpegEncContext*/ void *c, uint8_t *src1, uint8_t *src2 s->dsp.fdct(temp); for(i=0; i<64; i++) - sum= FFMAX(sum, ABS(temp[i])); + sum= FFMAX(sum, FFABS(temp[i])); return sum; } -void simple_idct(DCTELEM *block); //FIXME - 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]); @@ -3684,8 +3634,8 @@ static int vsad_intra16_c(/*MpegEncContext*/ void *c, uint8_t *s, uint8_t *dummy for(y=1; y>31; + // is this faster on some gcc/cpu combinations? +// if(tmp > 0x43c0ffff) tmp = 0xFFFF; +// else tmp = 0; + } + dst[i] = tmp - 0x8000; + } +} + /* XXX: those functions should be suppressed ASAP when all IDCTs are converted */ static void ff_jref_idct_put(uint8_t *dest, int line_size, DCTELEM *block) @@ -3787,13 +3770,13 @@ static void ff_jref_idct2_add(uint8_t *dest, int line_size, DCTELEM *block) static void ff_jref_idct1_put(uint8_t *dest, int line_size, DCTELEM *block) { - uint8_t *cm = cropTbl + MAX_NEG_CROP; + uint8_t *cm = ff_cropTbl + MAX_NEG_CROP; dest[0] = cm[(block[0] + 4)>>3]; } static void ff_jref_idct1_add(uint8_t *dest, int line_size, DCTELEM *block) { - uint8_t *cm = cropTbl + MAX_NEG_CROP; + uint8_t *cm = ff_cropTbl + MAX_NEG_CROP; dest[0] = cm[dest[0] + ((block[0] + 4)>>3)]; } @@ -3805,14 +3788,14 @@ void dsputil_static_init(void) { int i; - for(i=0;i<256;i++) cropTbl[i + MAX_NEG_CROP] = i; + for(i=0;i<256;i++) ff_cropTbl[i + MAX_NEG_CROP] = i; for(i=0;iavg_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->weight_h264_pixels_tab[0]= weight_h264_pixels16x16_c; c->weight_h264_pixels_tab[1]= weight_h264_pixels16x8_c; @@ -4075,6 +4059,7 @@ void dsputil_init(DSPContext* c, AVCodecContext *avctx) 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; c->h263_h_loop_filter= h263_h_loop_filter_c; c->h263_v_loop_filter= h263_v_loop_filter_c; @@ -4090,6 +4075,14 @@ void dsputil_init(DSPContext* c, AVCodecContext *avctx) c->inner_add_yblock = ff_snow_inner_add_yblock; #endif +#ifdef CONFIG_VORBIS_DECODER + c->vorbis_inverse_coupling = vorbis_inverse_coupling; +#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->float_to_int16 = ff_float_to_int16_c; + c->shrink[0]= ff_img_copy_plane; c->shrink[1]= ff_shrink22; c->shrink[2]= ff_shrink44; @@ -4097,6 +4090,9 @@ void dsputil_init(DSPContext* c, AVCodecContext *avctx) c->prefetch= just_return; + 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)); + #ifdef HAVE_MMX dsputil_init_mmx(c, avctx); #endif @@ -4121,6 +4117,16 @@ void dsputil_init(DSPContext* c, AVCodecContext *avctx) #ifdef ARCH_SH4 dsputil_init_sh4(c,avctx); #endif +#ifdef ARCH_BFIN + dsputil_init_bfin(c,avctx); +#endif + + for(i=0; i<64; i++){ + if(!c->put_2tap_qpel_pixels_tab[0][i]) + c->put_2tap_qpel_pixels_tab[0][i]= c->put_h264_qpel_pixels_tab[0][i]; + if(!c->avg_2tap_qpel_pixels_tab[0][i]) + c->avg_2tap_qpel_pixels_tab[0][i]= c->avg_h264_qpel_pixels_tab[0][i]; + } switch(c->idct_permutation_type){ case FF_NO_IDCT_PERM: -- cgit v1.2.3