diff options
Diffstat (limited to 'thirdparty/libtheora/state.c')
| -rw-r--r-- | thirdparty/libtheora/state.c | 298 |
1 files changed, 169 insertions, 129 deletions
diff --git a/thirdparty/libtheora/state.c b/thirdparty/libtheora/state.c index 42ed33a9a3..f4c6240387 100644 --- a/thirdparty/libtheora/state.c +++ b/thirdparty/libtheora/state.c @@ -11,25 +11,93 @@ ******************************************************************** function: - last mod: $Id: state.c 16503 2009-08-22 18:14:02Z giles $ + last mod: $Id$ ********************************************************************/ #include <stdlib.h> #include <string.h> -#include "internal.h" -#if defined(OC_X86_ASM) -#if defined(_MSC_VER) -# include "x86_vc/x86int.h" -#else -# include "x86/x86int.h" -#endif -#endif +#include "state.h" #if defined(OC_DUMP_IMAGES) # include <stdio.h> # include "png.h" +# include "zlib.h" #endif +/*The function used to fill in the chroma plane motion vectors for a macro + block when 4 different motion vectors are specified in the luma plane. + This version is for use with chroma decimated in the X and Y directions + (4:2:0). + _cbmvs: The chroma block-level motion vectors to fill in. + _lbmvs: The luma block-level motion vectors.*/ +static void oc_set_chroma_mvs00(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]){ + int dx; + int dy; + dx=OC_MV_X(_lbmvs[0])+OC_MV_X(_lbmvs[1]) + +OC_MV_X(_lbmvs[2])+OC_MV_X(_lbmvs[3]); + dy=OC_MV_Y(_lbmvs[0])+OC_MV_Y(_lbmvs[1]) + +OC_MV_Y(_lbmvs[2])+OC_MV_Y(_lbmvs[3]); + _cbmvs[0]=OC_MV(OC_DIV_ROUND_POW2(dx,2,2),OC_DIV_ROUND_POW2(dy,2,2)); +} + +/*The function used to fill in the chroma plane motion vectors for a macro + block when 4 different motion vectors are specified in the luma plane. + This version is for use with chroma decimated in the Y direction. + _cbmvs: The chroma block-level motion vectors to fill in. + _lbmvs: The luma block-level motion vectors.*/ +static void oc_set_chroma_mvs01(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]){ + int dx; + int dy; + dx=OC_MV_X(_lbmvs[0])+OC_MV_X(_lbmvs[2]); + dy=OC_MV_Y(_lbmvs[0])+OC_MV_Y(_lbmvs[2]); + _cbmvs[0]=OC_MV(OC_DIV_ROUND_POW2(dx,1,1),OC_DIV_ROUND_POW2(dy,1,1)); + dx=OC_MV_X(_lbmvs[1])+OC_MV_X(_lbmvs[3]); + dy=OC_MV_Y(_lbmvs[1])+OC_MV_Y(_lbmvs[3]); + _cbmvs[1]=OC_MV(OC_DIV_ROUND_POW2(dx,1,1),OC_DIV_ROUND_POW2(dy,1,1)); +} + +/*The function used to fill in the chroma plane motion vectors for a macro + block when 4 different motion vectors are specified in the luma plane. + This version is for use with chroma decimated in the X direction (4:2:2). + _cbmvs: The chroma block-level motion vectors to fill in. + _lbmvs: The luma block-level motion vectors.*/ +static void oc_set_chroma_mvs10(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]){ + int dx; + int dy; + dx=OC_MV_X(_lbmvs[0])+OC_MV_X(_lbmvs[1]); + dy=OC_MV_Y(_lbmvs[0])+OC_MV_Y(_lbmvs[1]); + _cbmvs[0]=OC_MV(OC_DIV_ROUND_POW2(dx,1,1),OC_DIV_ROUND_POW2(dy,1,1)); + dx=OC_MV_X(_lbmvs[2])+OC_MV_X(_lbmvs[3]); + dy=OC_MV_Y(_lbmvs[2])+OC_MV_Y(_lbmvs[3]); + _cbmvs[2]=OC_MV(OC_DIV_ROUND_POW2(dx,1,1),OC_DIV_ROUND_POW2(dy,1,1)); +} + +/*The function used to fill in the chroma plane motion vectors for a macro + block when 4 different motion vectors are specified in the luma plane. + This version is for use with no chroma decimation (4:4:4). + _cbmvs: The chroma block-level motion vectors to fill in. + _lmbmv: The luma macro-block level motion vector to fill in for use in + prediction. + _lbmvs: The luma block-level motion vectors.*/ +static void oc_set_chroma_mvs11(oc_mv _cbmvs[4],const oc_mv _lbmvs[4]){ + _cbmvs[0]=_lbmvs[0]; + _cbmvs[1]=_lbmvs[1]; + _cbmvs[2]=_lbmvs[2]; + _cbmvs[3]=_lbmvs[3]; +} + +/*A table of functions used to fill in the chroma plane motion vectors for a + macro block when 4 different motion vectors are specified in the luma + plane.*/ +const oc_set_chroma_mvs_func OC_SET_CHROMA_MVS_TABLE[TH_PF_NFORMATS]={ + (oc_set_chroma_mvs_func)oc_set_chroma_mvs00, + (oc_set_chroma_mvs_func)oc_set_chroma_mvs01, + (oc_set_chroma_mvs_func)oc_set_chroma_mvs10, + (oc_set_chroma_mvs_func)oc_set_chroma_mvs11 +}; + + + /*Returns the fragment index of the top-left block in a macro block. This can be used to test whether or not the whole macro block is valid. _sb_map: The super block map. @@ -92,7 +160,7 @@ static void oc_sb_create_plane_mapping(oc_sb_map _sb_maps[], if(jmax>4)jmax=4; else if(jmax<=0)break; /*By default, set all fragment indices to -1.*/ - memset(_sb_maps[sbi][0],0xFF,sizeof(_sb_maps[sbi])); + memset(_sb_maps[sbi],0xFF,sizeof(_sb_maps[sbi])); /*Fill in the fragment map for this super block.*/ xfrag=yfrag+x; for(i=0;i<imax;i++){ @@ -186,10 +254,14 @@ static void oc_mb_fill_cmapping10(oc_mb_map_plane _mb_map[3], This version is for use with no chroma decimation (4:4:4). This uses the already filled-in luma plane values. _mb_map: The macro block map to fill. - _fplanes: The descriptions of the fragment planes.*/ + _fplanes: The descriptions of the fragment planes. + _xfrag0: The X location of the upper-left hand fragment in the luma plane. + _yfrag0: The Y location of the upper-left hand fragment in the luma plane.*/ static void oc_mb_fill_cmapping11(oc_mb_map_plane _mb_map[3], - const oc_fragment_plane _fplanes[3]){ + const oc_fragment_plane _fplanes[3],int _xfrag0,int _yfrag0){ int k; + (void)_xfrag0; + (void)_yfrag0; for(k=0;k<4;k++){ _mb_map[1][k]=_mb_map[0][k]+_fplanes[1].froffset; _mb_map[2][k]=_mb_map[0][k]+_fplanes[2].froffset; @@ -211,7 +283,7 @@ static const oc_mb_fill_cmapping_func OC_MB_FILL_CMAPPING_TABLE[4]={ oc_mb_fill_cmapping00, oc_mb_fill_cmapping01, oc_mb_fill_cmapping10, - (oc_mb_fill_cmapping_func)oc_mb_fill_cmapping11 + oc_mb_fill_cmapping11 }; /*Fills in the mapping from macro blocks to their corresponding fragment @@ -469,7 +541,7 @@ static void oc_state_frarray_clear(oc_theora_state *_state){ unrestricted motion vectors without special casing the boundary. If chroma is decimated in either direction, the padding is reduced by a factor of 2 on the appropriate sides. - _nrefs: The number of reference buffers to init; must be 3 or 4.*/ + _nrefs: The number of reference buffers to init; must be in the range 3...6.*/ static int oc_state_ref_bufs_init(oc_theora_state *_state,int _nrefs){ th_info *info; unsigned char *ref_frame_data; @@ -481,6 +553,7 @@ static int oc_state_ref_bufs_init(oc_theora_state *_state,int _nrefs){ int yheight; int chstride; int cheight; + ptrdiff_t align; ptrdiff_t yoffset; ptrdiff_t coffset; ptrdiff_t *frag_buf_offs; @@ -489,33 +562,38 @@ static int oc_state_ref_bufs_init(oc_theora_state *_state,int _nrefs){ int vdec; int rfi; int pli; - if(_nrefs<3||_nrefs>4)return TH_EINVAL; + if(_nrefs<3||_nrefs>6)return TH_EINVAL; info=&_state->info; /*Compute the image buffer parameters for each plane.*/ hdec=!(info->pixel_fmt&1); vdec=!(info->pixel_fmt&2); yhstride=info->frame_width+2*OC_UMV_PADDING; yheight=info->frame_height+2*OC_UMV_PADDING; - chstride=yhstride>>hdec; + /*Require 16-byte aligned rows in the chroma planes.*/ + chstride=(yhstride>>hdec)+15&~15; cheight=yheight>>vdec; yplane_sz=yhstride*(size_t)yheight; cplane_sz=chstride*(size_t)cheight; yoffset=OC_UMV_PADDING+OC_UMV_PADDING*(ptrdiff_t)yhstride; coffset=(OC_UMV_PADDING>>hdec)+(OC_UMV_PADDING>>vdec)*(ptrdiff_t)chstride; - ref_frame_sz=yplane_sz+2*cplane_sz; + /*Although we guarantee the rows of the chroma planes are a multiple of 16 + bytes, the initial padding on the first row may only be 8 bytes. + Compute the offset needed to the actual image data to a multiple of 16.*/ + align=-coffset&15; + ref_frame_sz=yplane_sz+2*cplane_sz+16; ref_frame_data_sz=_nrefs*ref_frame_sz; /*Check for overflow. The same caveats apply as for oc_state_frarray_init().*/ - if(yplane_sz/yhstride!=yheight||2*cplane_sz<cplane_sz|| + if(yplane_sz/yhstride!=(size_t)yheight||2*cplane_sz+16<cplane_sz|| ref_frame_sz<yplane_sz||ref_frame_data_sz/_nrefs!=ref_frame_sz){ return TH_EIMPL; } - ref_frame_data=_ogg_malloc(ref_frame_data_sz); + ref_frame_data=oc_aligned_malloc(ref_frame_data_sz,16); frag_buf_offs=_state->frag_buf_offs= _ogg_malloc(_state->nfrags*sizeof(*frag_buf_offs)); if(ref_frame_data==NULL||frag_buf_offs==NULL){ _ogg_free(frag_buf_offs); - _ogg_free(ref_frame_data); + oc_aligned_free(ref_frame_data); return TH_EFAULT; } /*Set up the width, height and stride for the image buffers.*/ @@ -532,15 +610,15 @@ static int oc_state_ref_bufs_init(oc_theora_state *_state,int _nrefs){ memcpy(_state->ref_frame_bufs[rfi],_state->ref_frame_bufs[0], sizeof(_state->ref_frame_bufs[0])); } + _state->ref_frame_handle=ref_frame_data; /*Set up the data pointers for the image buffers.*/ for(rfi=0;rfi<_nrefs;rfi++){ - _state->ref_frame_data[rfi]=ref_frame_data; _state->ref_frame_bufs[rfi][0].data=ref_frame_data+yoffset; - ref_frame_data+=yplane_sz; + ref_frame_data+=yplane_sz+align; _state->ref_frame_bufs[rfi][1].data=ref_frame_data+coffset; ref_frame_data+=cplane_sz; _state->ref_frame_bufs[rfi][2].data=ref_frame_data+coffset; - ref_frame_data+=cplane_sz; + ref_frame_data+=cplane_sz+(16-align); /*Flip the buffer upside down. This allows us to decode Theora's bottom-up frames in their natural order, yet return a top-down buffer with a positive stride to the user.*/ @@ -550,7 +628,7 @@ static int oc_state_ref_bufs_init(oc_theora_state *_state,int _nrefs){ _state->ref_ystride[0]=-yhstride; _state->ref_ystride[1]=_state->ref_ystride[2]=-chstride; /*Initialize the fragment buffer offsets.*/ - ref_frame_data=_state->ref_frame_data[0]; + ref_frame_data=_state->ref_frame_bufs[0][0].data; fragi=0; for(pli=0;pli<3;pli++){ th_img_plane *iplane; @@ -576,41 +654,44 @@ static int oc_state_ref_bufs_init(oc_theora_state *_state,int _nrefs){ vpix+=stride<<3; } } - /*Initialize the reference frame indices.*/ + /*Initialize the reference frame pointers and indices.*/ _state->ref_frame_idx[OC_FRAME_GOLD]= _state->ref_frame_idx[OC_FRAME_PREV]= - _state->ref_frame_idx[OC_FRAME_SELF]=-1; - _state->ref_frame_idx[OC_FRAME_IO]=_nrefs>3?3:-1; + _state->ref_frame_idx[OC_FRAME_GOLD_ORIG]= + _state->ref_frame_idx[OC_FRAME_PREV_ORIG]= + _state->ref_frame_idx[OC_FRAME_SELF]= + _state->ref_frame_idx[OC_FRAME_IO]=-1; + _state->ref_frame_data[OC_FRAME_GOLD]= + _state->ref_frame_data[OC_FRAME_PREV]= + _state->ref_frame_data[OC_FRAME_GOLD_ORIG]= + _state->ref_frame_data[OC_FRAME_PREV_ORIG]= + _state->ref_frame_data[OC_FRAME_SELF]= + _state->ref_frame_data[OC_FRAME_IO]=NULL; return 0; } static void oc_state_ref_bufs_clear(oc_theora_state *_state){ _ogg_free(_state->frag_buf_offs); - _ogg_free(_state->ref_frame_data[0]); + oc_aligned_free(_state->ref_frame_handle); } -void oc_state_vtable_init_c(oc_theora_state *_state){ +void oc_state_accel_init_c(oc_theora_state *_state){ + _state->cpu_flags=0; +#if defined(OC_STATE_USE_VTABLE) _state->opt_vtable.frag_copy=oc_frag_copy_c; + _state->opt_vtable.frag_copy_list=oc_frag_copy_list_c; _state->opt_vtable.frag_recon_intra=oc_frag_recon_intra_c; _state->opt_vtable.frag_recon_inter=oc_frag_recon_inter_c; _state->opt_vtable.frag_recon_inter2=oc_frag_recon_inter2_c; _state->opt_vtable.idct8x8=oc_idct8x8_c; _state->opt_vtable.state_frag_recon=oc_state_frag_recon_c; - _state->opt_vtable.state_frag_copy_list=oc_state_frag_copy_list_c; + _state->opt_vtable.loop_filter_init=oc_loop_filter_init_c; _state->opt_vtable.state_loop_filter_frag_rows= oc_state_loop_filter_frag_rows_c; _state->opt_vtable.restore_fpu=oc_restore_fpu_c; - _state->opt_data.dct_fzig_zag=OC_FZIG_ZAG; -} - -/*Initialize the accelerated function pointers.*/ -void oc_state_vtable_init(oc_theora_state *_state){ -#if defined(OC_X86_ASM) - oc_state_vtable_init_x86(_state); -#else - oc_state_vtable_init_c(_state); #endif + _state->opt_data.dct_fzig_zag=OC_FZIG_ZAG; } @@ -626,7 +707,8 @@ int oc_state_init(oc_theora_state *_state,const th_info *_info,int _nrefs){ how it is specified in the bitstream, because the Y axis is flipped in the bitstream. The displayable frame must fit inside the encoded frame. - The color space must be one known by the encoder.*/ + The color space must be one known by the encoder. + The framerate ratio must not contain a zero value.*/ if((_info->frame_width&0xF)||(_info->frame_height&0xF)|| _info->frame_width<=0||_info->frame_width>=0x100000|| _info->frame_height<=0||_info->frame_height>=0x100000|| @@ -639,7 +721,8 @@ int oc_state_init(oc_theora_state *_state,const th_info *_info,int _nrefs){ but there are a number of compilers which will mis-optimize this. It's better to live with the spurious warnings.*/ _info->colorspace<0||_info->colorspace>=TH_CS_NSPACES|| - _info->pixel_fmt<0||_info->pixel_fmt>=TH_PF_NFORMATS){ + _info->pixel_fmt<0||_info->pixel_fmt>=TH_PF_NFORMATS|| + _info->fps_numerator<1||_info->fps_denominator<1){ return TH_EINVAL; } memset(_state,0,sizeof(*_state)); @@ -648,7 +731,7 @@ int oc_state_init(oc_theora_state *_state,const th_info *_info,int _nrefs){ system.*/ _state->info.pic_y=_info->frame_height-_info->pic_height-_info->pic_y; _state->frame_type=OC_UNKWN_FRAME; - oc_state_vtable_init(_state); + oc_state_accel_init(_state); ret=oc_state_frarray_init(_state); if(ret>=0)ret=oc_state_ref_bufs_init(_state,_nrefs); if(ret<0){ @@ -758,11 +841,10 @@ void oc_state_borders_fill(oc_theora_state *_state,int _refi){ _offsets[1] is set if the motion vector has non-zero fractional components. _pli: The color plane index. - _dx: The X component of the motion vector. - _dy: The Y component of the motion vector. + _mv: The motion vector. Return: The number of offsets returned: 1 or 2.*/ int oc_state_get_mv_offsets(const oc_theora_state *_state,int _offsets[2], - int _pli,int _dx,int _dy){ + int _pli,oc_mv _mv){ /*Here is a brief description of how Theora handles motion vectors: Motion vector components are specified to half-pixel accuracy in undecimated directions of each plane, and quarter-pixel accuracy in @@ -785,21 +867,25 @@ int oc_state_get_mv_offsets(const oc_theora_state *_state,int _offsets[2], int xfrac; int yfrac; int offs; + int dx; + int dy; ystride=_state->ref_ystride[_pli]; /*These two variables decide whether we are in half- or quarter-pixel precision in each component.*/ xprec=1+(_pli!=0&&!(_state->info.pixel_fmt&1)); yprec=1+(_pli!=0&&!(_state->info.pixel_fmt&2)); + dx=OC_MV_X(_mv); + dy=OC_MV_Y(_mv); /*These two variables are either 0 if all the fractional bits are zero or -1 if any of them are non-zero.*/ - xfrac=OC_SIGNMASK(-(_dx&(xprec|1))); - yfrac=OC_SIGNMASK(-(_dy&(yprec|1))); - offs=(_dx>>xprec)+(_dy>>yprec)*ystride; + xfrac=OC_SIGNMASK(-(dx&(xprec|1))); + yfrac=OC_SIGNMASK(-(dy&(yprec|1))); + offs=(dx>>xprec)+(dy>>yprec)*ystride; if(xfrac||yfrac){ int xmask; int ymask; - xmask=OC_SIGNMASK(_dx); - ymask=OC_SIGNMASK(_dy); + xmask=OC_SIGNMASK(dx); + ymask=OC_SIGNMASK(dy); yfrac&=ystride; _offsets[0]=offs-(xfrac&xmask)+(yfrac&ymask); _offsets[1]=offs-(xfrac&~xmask)+(yfrac&~ymask); @@ -848,13 +934,17 @@ int oc_state_get_mv_offsets(const oc_theora_state *_state,int _offsets[2], int mx2; int my2; int offs; + int dx; + int dy; ystride=_state->ref_ystride[_pli]; qpy=_pli!=0&&!(_state->info.pixel_fmt&2); - my=OC_MVMAP[qpy][_dy+31]; - my2=OC_MVMAP2[qpy][_dy+31]; + dx=OC_MV_X(_mv); + dy=OC_MV_Y(_mv); + my=OC_MVMAP[qpy][dy+31]; + my2=OC_MVMAP2[qpy][dy+31]; qpx=_pli!=0&&!(_state->info.pixel_fmt&1); - mx=OC_MVMAP[qpx][_dx+31]; - mx2=OC_MVMAP2[qpx][_dx+31]; + mx=OC_MVMAP[qpx][dx+31]; + mx2=OC_MVMAP2[qpx][dx+31]; offs=my*ystride+mx; if(mx2||my2){ _offsets[1]=offs+my2*ystride+mx2; @@ -866,18 +956,12 @@ int oc_state_get_mv_offsets(const oc_theora_state *_state,int _offsets[2], #endif } -void oc_state_frag_recon(const oc_theora_state *_state,ptrdiff_t _fragi, - int _pli,ogg_int16_t _dct_coeffs[64],int _last_zzi,ogg_uint16_t _dc_quant){ - _state->opt_vtable.state_frag_recon(_state,_fragi,_pli,_dct_coeffs, - _last_zzi,_dc_quant); -} - void oc_state_frag_recon_c(const oc_theora_state *_state,ptrdiff_t _fragi, - int _pli,ogg_int16_t _dct_coeffs[64],int _last_zzi,ogg_uint16_t _dc_quant){ + int _pli,ogg_int16_t _dct_coeffs[128],int _last_zzi,ogg_uint16_t _dc_quant){ unsigned char *dst; ptrdiff_t frag_buf_off; int ystride; - int mb_mode; + int refi; /*Apply the inverse transform.*/ /*Special case only having a DC component.*/ if(_last_zzi<2){ @@ -887,69 +971,35 @@ void oc_state_frag_recon_c(const oc_theora_state *_state,ptrdiff_t _fragi, no iDCT rounding.*/ p=(ogg_int16_t)(_dct_coeffs[0]*(ogg_int32_t)_dc_quant+15>>5); /*LOOP VECTORIZES.*/ - for(ci=0;ci<64;ci++)_dct_coeffs[ci]=p; + for(ci=0;ci<64;ci++)_dct_coeffs[64+ci]=p; } else{ /*First, dequantize the DC coefficient.*/ _dct_coeffs[0]=(ogg_int16_t)(_dct_coeffs[0]*(int)_dc_quant); - oc_idct8x8(_state,_dct_coeffs,_last_zzi); + oc_idct8x8(_state,_dct_coeffs+64,_dct_coeffs,_last_zzi); } /*Fill in the target buffer.*/ frag_buf_off=_state->frag_buf_offs[_fragi]; - mb_mode=_state->frags[_fragi].mb_mode; + refi=_state->frags[_fragi].refi; ystride=_state->ref_ystride[_pli]; - dst=_state->ref_frame_data[_state->ref_frame_idx[OC_FRAME_SELF]]+frag_buf_off; - if(mb_mode==OC_MODE_INTRA)oc_frag_recon_intra(_state,dst,ystride,_dct_coeffs); + dst=_state->ref_frame_data[OC_FRAME_SELF]+frag_buf_off; + if(refi==OC_FRAME_SELF)oc_frag_recon_intra(_state,dst,ystride,_dct_coeffs+64); else{ const unsigned char *ref; int mvoffsets[2]; - ref= - _state->ref_frame_data[_state->ref_frame_idx[OC_FRAME_FOR_MODE(mb_mode)]] - +frag_buf_off; + ref=_state->ref_frame_data[refi]+frag_buf_off; if(oc_state_get_mv_offsets(_state,mvoffsets,_pli, - _state->frag_mvs[_fragi][0],_state->frag_mvs[_fragi][1])>1){ + _state->frag_mvs[_fragi])>1){ oc_frag_recon_inter2(_state, - dst,ref+mvoffsets[0],ref+mvoffsets[1],ystride,_dct_coeffs); + dst,ref+mvoffsets[0],ref+mvoffsets[1],ystride,_dct_coeffs+64); + } + else{ + oc_frag_recon_inter(_state,dst,ref+mvoffsets[0],ystride,_dct_coeffs+64); } - else oc_frag_recon_inter(_state,dst,ref+mvoffsets[0],ystride,_dct_coeffs); - } -} - -/*Copies the fragments specified by the lists of fragment indices from one - frame to another. - _fragis: A pointer to a list of fragment indices. - _nfragis: The number of fragment indices to copy. - _dst_frame: The reference frame to copy to. - _src_frame: The reference frame to copy from. - _pli: The color plane the fragments lie in.*/ -void oc_state_frag_copy_list(const oc_theora_state *_state, - const ptrdiff_t *_fragis,ptrdiff_t _nfragis, - int _dst_frame,int _src_frame,int _pli){ - _state->opt_vtable.state_frag_copy_list(_state,_fragis,_nfragis,_dst_frame, - _src_frame,_pli); -} - -void oc_state_frag_copy_list_c(const oc_theora_state *_state, - const ptrdiff_t *_fragis,ptrdiff_t _nfragis, - int _dst_frame,int _src_frame,int _pli){ - const ptrdiff_t *frag_buf_offs; - const unsigned char *src_frame_data; - unsigned char *dst_frame_data; - ptrdiff_t fragii; - int ystride; - dst_frame_data=_state->ref_frame_data[_state->ref_frame_idx[_dst_frame]]; - src_frame_data=_state->ref_frame_data[_state->ref_frame_idx[_src_frame]]; - ystride=_state->ref_ystride[_pli]; - frag_buf_offs=_state->frag_buf_offs; - for(fragii=0;fragii<_nfragis;fragii++){ - ptrdiff_t frag_buf_off; - frag_buf_off=frag_buf_offs[_fragis[fragii]]; - oc_frag_copy(_state,dst_frame_data+frag_buf_off, - src_frame_data+frag_buf_off,ystride); } } -static void loop_filter_h(unsigned char *_pix,int _ystride,int *_bv){ +static void loop_filter_h(unsigned char *_pix,int _ystride,signed char *_bv){ int y; _pix-=2; for(y=0;y<8;y++){ @@ -965,7 +1015,7 @@ static void loop_filter_h(unsigned char *_pix,int _ystride,int *_bv){ } } -static void loop_filter_v(unsigned char *_pix,int _ystride,int *_bv){ +static void loop_filter_v(unsigned char *_pix,int _ystride,signed char *_bv){ int x; _pix-=_ystride*2; for(x=0;x<8;x++){ @@ -982,20 +1032,16 @@ static void loop_filter_v(unsigned char *_pix,int _ystride,int *_bv){ /*Initialize the bounding values array used by the loop filter. _bv: Storage for the array. - Return: 0 on success, or a non-zero value if no filtering need be applied.*/ -int oc_state_loop_filter_init(oc_theora_state *_state,int _bv[256]){ - int flimit; + _flimit: The filter limit as defined in Section 7.10 of the spec.*/ +void oc_loop_filter_init_c(signed char _bv[256],int _flimit){ int i; - flimit=_state->loop_filter_limits[_state->qis[0]]; - if(flimit==0)return 1; memset(_bv,0,sizeof(_bv[0])*256); - for(i=0;i<flimit;i++){ - if(127-i-flimit>=0)_bv[127-i-flimit]=i-flimit; - _bv[127-i]=-i; - _bv[127+i]=i; - if(127+i+flimit<256)_bv[127+i+flimit]=flimit-i; + for(i=0;i<_flimit;i++){ + if(127-i-_flimit>=0)_bv[127-i-_flimit]=(signed char)(i-_flimit); + _bv[127-i]=(signed char)(-i); + _bv[127+i]=(signed char)(i); + if(127+i+_flimit<256)_bv[127+i+_flimit]=(signed char)(_flimit-i); } - return 0; } /*Apply the loop filter to a given set of fragment rows in the given plane. @@ -1006,14 +1052,8 @@ int oc_state_loop_filter_init(oc_theora_state *_state,int _bv[256]){ _pli: The color plane to filter. _fragy0: The Y coordinate of the first fragment row to filter. _fragy_end: The Y coordinate of the fragment row to stop filtering at.*/ -void oc_state_loop_filter_frag_rows(const oc_theora_state *_state,int _bv[256], - int _refi,int _pli,int _fragy0,int _fragy_end){ - _state->opt_vtable.state_loop_filter_frag_rows(_state,_bv,_refi,_pli, - _fragy0,_fragy_end); -} - -void oc_state_loop_filter_frag_rows_c(const oc_theora_state *_state,int *_bv, - int _refi,int _pli,int _fragy0,int _fragy_end){ +void oc_state_loop_filter_frag_rows_c(const oc_theora_state *_state, + signed char *_bv,int _refi,int _pli,int _fragy0,int _fragy_end){ const oc_fragment_plane *fplane; const oc_fragment *frags; const ptrdiff_t *frag_buf_offs; @@ -1030,7 +1070,7 @@ void oc_state_loop_filter_frag_rows_c(const oc_theora_state *_state,int *_bv, fragi_top=fplane->froffset; fragi_bot=fragi_top+fplane->nfrags; fragi0=fragi_top+_fragy0*(ptrdiff_t)nhfrags; - fragi0_end=fragi0+(_fragy_end-_fragy0)*(ptrdiff_t)nhfrags; + fragi0_end=fragi_top+_fragy_end*(ptrdiff_t)nhfrags; ystride=_state->ref_ystride[_pli]; frags=_state->frags; frag_buf_offs=_state->frag_buf_offs; |