Commit 3a1df52e authored by Michael Niedermayer's avatar Michael Niedermayer

Merge remote-tracking branch 'qatar/master'

* qatar/master:
  Pretty-print RV3/4 decoder source
  Saner RV1/2 initialisation and RV2 header parsing.

Conflicts:
	libavcodec/rv10.c
Merged-by: 's avatarMichael Niedermayer <michaelni@gmx.at>
parents aae1c247 7351eb14
......@@ -34,6 +34,10 @@
//#define DEBUG
#define RV_GET_MAJOR_VER(x) ((x) >> 28)
#define RV_GET_MINOR_VER(x) (((x) >> 20) & 0xFF)
#define RV_GET_MICRO_VER(x) (((x) >> 12) & 0xFF)
#define DC_VLC_BITS 14 //FIXME find a better solution
static const uint16_t rv_lum_code[256] =
......@@ -292,6 +296,7 @@ static int rv10_decode_picture_header(MpegEncContext *s)
static int rv20_decode_picture_header(MpegEncContext *s)
{
int seq, mb_pos, i;
int rpr_bits;
#if 0
GetBitContext gb= s->gb;
......@@ -310,13 +315,6 @@ static int rv20_decode_picture_header(MpegEncContext *s)
av_log(s->avctx, AV_LOG_DEBUG, "\n");
#endif
if(s->avctx->sub_id == 0x30202002 || s->avctx->sub_id == 0x30203002){
if (get_bits(&s->gb, 3)){
av_log(s->avctx, AV_LOG_ERROR, "unknown triplet set\n");
return -1;
}
}
i= get_bits(&s->gb, 2);
switch(i){
case 0: s->pict_type= AV_PICTURE_TYPE_I; break;
......@@ -334,7 +332,7 @@ static int rv20_decode_picture_header(MpegEncContext *s)
}
if (get_bits1(&s->gb)){
av_log(s->avctx, AV_LOG_ERROR, "unknown bit set\n");
av_log(s->avctx, AV_LOG_ERROR, "reserved bit set\n");
return -1;
}
......@@ -343,23 +341,21 @@ static int rv20_decode_picture_header(MpegEncContext *s)
av_log(s->avctx, AV_LOG_ERROR, "error, qscale:0\n");
return -1;
}
if(s->avctx->sub_id == 0x30203002){
if (get_bits1(&s->gb)){
av_log(s->avctx, AV_LOG_ERROR, "unknown bit2 set\n");
return -1;
}
}
if(s->avctx->has_b_frames){
int f, new_w, new_h;
int v= s->avctx->extradata_size >= 4 ? 7&((uint8_t*)s->avctx->extradata)[1] : 0;
if(RV_GET_MINOR_VER(s->avctx->sub_id) >= 2)
s->loop_filter = get_bits1(&s->gb);
if (get_bits1(&s->gb)){
av_log(s->avctx, AV_LOG_ERROR, "unknown bit3 set\n");
}
seq= get_bits(&s->gb, 13)<<2;
if(RV_GET_MINOR_VER(s->avctx->sub_id) <= 1)
seq = get_bits(&s->gb, 8) << 7;
else
seq = get_bits(&s->gb, 13) << 2;
rpr_bits = s->avctx->extradata[1] & 7;
if(rpr_bits){
int f, new_w, new_h;
rpr_bits = FFMIN((rpr_bits >> 1) + 1, 3);
f= get_bits(&s->gb, av_log2(v)+1);
f = get_bits(&s->gb, rpr_bits);
if(f){
new_w= 4*((uint8_t*)s->avctx->extradata)[6+2*f];
......@@ -381,19 +377,12 @@ static int rv20_decode_picture_header(MpegEncContext *s)
}
if(s->avctx->debug & FF_DEBUG_PICT_INFO){
av_log(s->avctx, AV_LOG_DEBUG, "F %d/%d\n", f, v);
av_log(s->avctx, AV_LOG_DEBUG, "F %d/%d\n", f, rpr_bits);
}
}else{
seq= get_bits(&s->gb, 8)*128;
}
// if(s->avctx->sub_id <= 0x20201002){ //0x20201002 definitely needs this
mb_pos= ff_h263_decode_mba(s);
/* }else{
mb_pos= get_bits(&s->gb, av_log2(s->mb_num-1)+1);
s->mb_x= mb_pos % s->mb_width;
s->mb_y= mb_pos / s->mb_width;
}*/
mb_pos = ff_h263_decode_mba(s);
//av_log(s->avctx, AV_LOG_DEBUG, "%d\n", seq);
seq |= s->time &~0x7FFF;
if(seq - s->time > 0x4000) seq -= 0x8000;
......@@ -420,6 +409,9 @@ static int rv20_decode_picture_header(MpegEncContext *s)
av_log(s->avctx, AV_LOG_DEBUG, "\n");*/
s->no_rounding= get_bits1(&s->gb);
if(RV_GET_MINOR_VER(s->avctx->sub_id) <= 1 && s->pict_type == AV_PICTURE_TYPE_B)
skip_bits(&s->gb, 5); // binary decoder reads 3+2 bits here but they don't seem to be used
s->f_code = 1;
s->unrestricted_mv = 1;
s->h263_aic= s->pict_type == AV_PICTURE_TYPE_I;
......@@ -444,6 +436,7 @@ static av_cold int rv10_decode_init(AVCodecContext *avctx)
{
MpegEncContext *s = avctx->priv_data;
static int done=0;
int major_ver, minor_ver, micro_ver;
if (avctx->extradata_size < 8) {
av_log(avctx, AV_LOG_ERROR, "Extradata is too small.\n");
......@@ -462,32 +455,27 @@ static av_cold int rv10_decode_init(AVCodecContext *avctx)
s->h263_long_vectors= ((uint8_t*)avctx->extradata)[3] & 1;
avctx->sub_id= AV_RB32((uint8_t*)avctx->extradata + 4);
if (avctx->sub_id == 0x10000000) {
s->rv10_version= 0;
s->low_delay=1;
} else if (avctx->sub_id == 0x10001000) {
s->rv10_version= 3;
s->low_delay=1;
} else if (avctx->sub_id == 0x10002000) {
s->rv10_version= 3;
s->low_delay=1;
s->obmc=1;
} else if (avctx->sub_id == 0x10003000) {
s->rv10_version= 3;
s->low_delay=1;
} else if (avctx->sub_id == 0x10003001) {
s->rv10_version= 3;
s->low_delay=1;
} else if ( avctx->sub_id == 0x20001000
|| (avctx->sub_id >= 0x20100000 && avctx->sub_id < 0x201a0000)) {
s->low_delay=1;
} else if ( avctx->sub_id == 0x30202002
|| avctx->sub_id == 0x30203002
|| (avctx->sub_id >= 0x20200002 && avctx->sub_id < 0x20300000)) {
s->low_delay=0;
s->avctx->has_b_frames=1;
} else
major_ver = RV_GET_MAJOR_VER(avctx->sub_id);
minor_ver = RV_GET_MINOR_VER(avctx->sub_id);
micro_ver = RV_GET_MICRO_VER(avctx->sub_id);
s->low_delay = 1;
switch (major_ver) {
case 1:
s->rv10_version = micro_ver ? 3 : 1;
s->obmc = micro_ver == 2;
break;
case 2:
if (minor_ver >= 2) {
s->low_delay = 0;
s->avctx->has_b_frames = 1;
}
break;
default:
av_log(s->avctx, AV_LOG_ERROR, "unknown header %X\n", avctx->sub_id);
av_log_missing_feature(avctx, "RV1/2 version", 1);
return AVERROR_PATCHWELCOME;
}
if(avctx->debug & FF_DEBUG_PICT_INFO){
av_log(avctx, AV_LOG_DEBUG, "ver:%X ver0:%X\n", avctx->sub_id, avctx->extradata_size >= 4 ? ((uint32_t*)avctx->extradata)[0] : -1);
......
......@@ -275,7 +275,7 @@ AVCodec ff_rv30_decoder = {
.close = ff_rv34_decode_end,
.decode = ff_rv34_decode_frame,
.capabilities = CODEC_CAP_DR1 | CODEC_CAP_DELAY,
.flush = ff_mpeg_flush,
.long_name = NULL_IF_CONFIG_SMALL("RealVideo 3.0"),
.pix_fmts= ff_pixfmt_list_420,
.flush = ff_mpeg_flush,
.long_name = NULL_IF_CONFIG_SMALL("RealVideo 3.0"),
.pix_fmts = ff_pixfmt_list_420,
};
......@@ -29,10 +29,10 @@
#define RV30_LOWPASS(OPNAME, OP) \
static av_unused void OPNAME ## rv30_tpel8_h_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, const int C1, const int C2){\
const int h=8;\
const int h = 8;\
uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;\
int i;\
for(i=0; i<h; i++)\
for(i = 0; i < h; i++)\
{\
OP(dst[0], (-(src[-1]+src[2]) + src[0]*C1 + src[1]*C2 + 8)>>4);\
OP(dst[1], (-(src[ 0]+src[3]) + src[1]*C1 + src[2]*C2 + 8)>>4);\
......@@ -42,28 +42,28 @@ static av_unused void OPNAME ## rv30_tpel8_h_lowpass(uint8_t *dst, uint8_t *src,
OP(dst[5], (-(src[ 4]+src[7]) + src[5]*C1 + src[6]*C2 + 8)>>4);\
OP(dst[6], (-(src[ 5]+src[8]) + src[6]*C1 + src[7]*C2 + 8)>>4);\
OP(dst[7], (-(src[ 6]+src[9]) + src[7]*C1 + src[8]*C2 + 8)>>4);\
dst+=dstStride;\
src+=srcStride;\
dst += dstStride;\
src += srcStride;\
}\
}\
\
static void OPNAME ## rv30_tpel8_v_lowpass(uint8_t *dst, uint8_t *src, int dstStride, int srcStride, const int C1, const int C2){\
const int w=8;\
const int w = 8;\
uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;\
int i;\
for(i=0; i<w; i++)\
for(i = 0; i < w; i++)\
{\
const int srcA= src[-1*srcStride];\
const int src0= src[0 *srcStride];\
const int src1= src[1 *srcStride];\
const int src2= src[2 *srcStride];\
const int src3= src[3 *srcStride];\
const int src4= src[4 *srcStride];\
const int src5= src[5 *srcStride];\
const int src6= src[6 *srcStride];\
const int src7= src[7 *srcStride];\
const int src8= src[8 *srcStride];\
const int src9= src[9 *srcStride];\
const int srcA = src[-1*srcStride];\
const int src0 = src[0 *srcStride];\
const int src1 = src[1 *srcStride];\
const int src2 = src[2 *srcStride];\
const int src3 = src[3 *srcStride];\
const int src4 = src[4 *srcStride];\
const int src5 = src[5 *srcStride];\
const int src6 = src[6 *srcStride];\
const int src7 = src[7 *srcStride];\
const int src8 = src[8 *srcStride];\
const int src9 = src[9 *srcStride];\
OP(dst[0*dstStride], (-(srcA+src2) + src0*C1 + src1*C2 + 8)>>4);\
OP(dst[1*dstStride], (-(src0+src3) + src1*C1 + src2*C2 + 8)>>4);\
OP(dst[2*dstStride], (-(src1+src4) + src2*C1 + src3*C2 + 8)>>4);\
......
......@@ -181,16 +181,16 @@ static av_always_inline void rv34_row_transform(int temp[16], DCTELEM *block)
{
int i;
for(i=0; i<4; i++){
const int z0= 13*(block[i+8*0] + block[i+8*2]);
const int z1= 13*(block[i+8*0] - block[i+8*2]);
const int z2= 7* block[i+8*1] - 17*block[i+8*3];
const int z3= 17* block[i+8*1] + 7*block[i+8*3];
temp[4*i+0]= z0+z3;
temp[4*i+1]= z1+z2;
temp[4*i+2]= z1-z2;
temp[4*i+3]= z0-z3;
for(i = 0; i < 4; i++){
const int z0 = 13*(block[i+8*0] + block[i+8*2]);
const int z1 = 13*(block[i+8*0] - block[i+8*2]);
const int z2 = 7* block[i+8*1] - 17*block[i+8*3];
const int z3 = 17* block[i+8*1] + 7*block[i+8*3];
temp[4*i+0] = z0 + z3;
temp[4*i+1] = z1 + z2;
temp[4*i+2] = z1 - z2;
temp[4*i+3] = z0 - z3;
}
}
......@@ -204,16 +204,16 @@ static void rv34_inv_transform(DCTELEM *block){
rv34_row_transform(temp, block);
for(i=0; i<4; i++){
const int z0= 13*(temp[4*0+i] + temp[4*2+i]) + 0x200;
const int z1= 13*(temp[4*0+i] - temp[4*2+i]) + 0x200;
const int z2= 7* temp[4*1+i] - 17*temp[4*3+i];
const int z3= 17* temp[4*1+i] + 7*temp[4*3+i];
block[i*8+0]= (z0 + z3)>>10;
block[i*8+1]= (z1 + z2)>>10;
block[i*8+2]= (z1 - z2)>>10;
block[i*8+3]= (z0 - z3)>>10;
for(i = 0; i < 4; i++){
const int z0 = 13*(temp[4*0+i] + temp[4*2+i]) + 0x200;
const int z1 = 13*(temp[4*0+i] - temp[4*2+i]) + 0x200;
const int z2 = 7* temp[4*1+i] - 17*temp[4*3+i];
const int z3 = 17* temp[4*1+i] + 7*temp[4*3+i];
block[i*8+0] = (z0 + z3) >> 10;
block[i*8+1] = (z1 + z2) >> 10;
block[i*8+2] = (z1 - z2) >> 10;
block[i*8+3] = (z0 - z3) >> 10;
}
}
......@@ -230,16 +230,16 @@ static void rv34_inv_transform_noround(DCTELEM *block){
rv34_row_transform(temp, block);
for(i=0; i<4; i++){
const int z0= 13*(temp[4*0+i] + temp[4*2+i]);
const int z1= 13*(temp[4*0+i] - temp[4*2+i]);
const int z2= 7* temp[4*1+i] - 17*temp[4*3+i];
const int z3= 17* temp[4*1+i] + 7*temp[4*3+i];
block[i*8+0]= ((z0 + z3)*3)>>11;
block[i*8+1]= ((z1 + z2)*3)>>11;
block[i*8+2]= ((z1 - z2)*3)>>11;
block[i*8+3]= ((z0 - z3)*3)>>11;
for(i = 0; i < 4; i++){
const int z0 = 13*(temp[4*0+i] + temp[4*2+i]);
const int z1 = 13*(temp[4*0+i] - temp[4*2+i]);
const int z2 = 7* temp[4*1+i] - 17*temp[4*3+i];
const int z3 = 17* temp[4*1+i] + 7*temp[4*3+i];
block[i*8+0] = ((z0 + z3) * 3) >> 11;
block[i*8+1] = ((z1 + z2) * 3) >> 11;
block[i*8+2] = ((z1 - z2) * 3) >> 11;
block[i*8+3] = ((z0 - z3) * 3) >> 11;
}
}
......@@ -772,7 +772,7 @@ static inline void rv34_mc(RV34DecContext *r, const int block_type,
srcV += uvsrc_y * s->uvlinesize + uvsrc_x;
if( (unsigned)(src_x - !!lx*2) > s->h_edge_pos - !!lx*2 - (width <<3) - 4
|| (unsigned)(src_y - !!ly*2) > s->v_edge_pos - !!ly*2 - (height<<3) - 4){
uint8_t *uvbuf= s->edge_emu_buffer + 22 * s->linesize;
uint8_t *uvbuf = s->edge_emu_buffer + 22 * s->linesize;
srcY -= 2 + 2*s->linesize;
s->dsp.emulated_edge_mc(s->edge_emu_buffer, srcY, s->linesize, (width<<3)+6, (height<<3)+6,
......@@ -1301,8 +1301,8 @@ static int rv34_decode_slice(RV34DecContext *r, int end, const uint8_t* buf, int
}
memset(r->intra_types_hist, -1, r->intra_types_stride * 4 * 2 * sizeof(*r->intra_types_hist));
s->first_slice_line = 1;
s->resync_mb_x= s->mb_x;
s->resync_mb_y= s->mb_y;
s->resync_mb_x = s->mb_x;
s->resync_mb_y = s->mb_y;
ff_init_block_index(s);
while(!check_slice_end(r, s)) {
......@@ -1344,11 +1344,11 @@ av_cold int ff_rv34_decode_init(AVCodecContext *avctx)
MpegEncContext *s = &r->s;
MPV_decode_defaults(s);
s->avctx= avctx;
s->avctx = avctx;
s->out_format = FMT_H263;
s->codec_id= avctx->codec_id;
s->codec_id = avctx->codec_id;
s->width = avctx->width;
s->width = avctx->width;
s->height = avctx->height;
r->s.avctx = avctx;
......@@ -1404,8 +1404,8 @@ int ff_rv34_decode_frame(AVCodecContext *avctx,
if (buf_size == 0) {
/* special case for last picture */
if (s->low_delay==0 && s->next_picture_ptr) {
*pict= *(AVFrame*)s->next_picture_ptr;
s->next_picture_ptr= NULL;
*pict = *(AVFrame*)s->next_picture_ptr;
s->next_picture_ptr = NULL;
*data_size = sizeof(AVFrame);
}
......@@ -1436,13 +1436,13 @@ int ff_rv34_decode_frame(AVCodecContext *avctx,
|| avctx->skip_frame >= AVDISCARD_ALL)
return buf_size;
for(i=0; i<slice_count; i++){
int offset= get_slice_offset(avctx, slices_hdr, i);
for(i = 0; i < slice_count; i++){
int offset = get_slice_offset(avctx, slices_hdr, i);
int size;
if(i+1 == slice_count)
size= buf_size - offset;
size = buf_size - offset;
else
size= get_slice_offset(avctx, slices_hdr, i+1) - offset;
size = get_slice_offset(avctx, slices_hdr, i+1) - offset;
if(offset > buf_size){
av_log(avctx, AV_LOG_ERROR, "Slice offset is greater than frame size\n");
......@@ -1472,16 +1472,16 @@ int ff_rv34_decode_frame(AVCodecContext *avctx,
ff_er_frame_end(s);
MPV_frame_end(s);
if (s->pict_type == AV_PICTURE_TYPE_B || s->low_delay) {
*pict= *(AVFrame*)s->current_picture_ptr;
*pict = *(AVFrame*)s->current_picture_ptr;
} else if (s->last_picture_ptr != NULL) {
*pict= *(AVFrame*)s->last_picture_ptr;
*pict = *(AVFrame*)s->last_picture_ptr;
}
if(s->last_picture_ptr || s->low_delay){
*data_size = sizeof(AVFrame);
ff_print_debug_info(s, pict);
}
s->current_picture_ptr= NULL; //so we can detect if frame_end wasnt called (find some nicer solution...)
s->current_picture_ptr = NULL; //so we can detect if frame_end wasnt called (find some nicer solution...)
}
return buf_size;
}
......
......@@ -677,7 +677,7 @@ AVCodec ff_rv40_decoder = {
.close = ff_rv34_decode_end,
.decode = ff_rv34_decode_frame,
.capabilities = CODEC_CAP_DR1 | CODEC_CAP_DELAY,
.flush = ff_mpeg_flush,
.long_name = NULL_IF_CONFIG_SMALL("RealVideo 4.0"),
.pix_fmts= ff_pixfmt_list_420,
.flush = ff_mpeg_flush,
.long_name = NULL_IF_CONFIG_SMALL("RealVideo 4.0"),
.pix_fmts = ff_pixfmt_list_420,
};
......@@ -32,7 +32,7 @@ static av_unused void OPNAME ## rv40_qpel8_h_lowpass(uint8_t *dst, uint8_t *src,
const int h, const int C1, const int C2, const int SHIFT){\
uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;\
int i;\
for(i=0; i<h; i++)\
for(i = 0; i < h; i++)\
{\
OP(dst[0], (src[-2] + src[ 3] - 5*(src[-1]+src[2]) + src[0]*C1 + src[1]*C2 + (1<<(SHIFT-1))) >> SHIFT);\
OP(dst[1], (src[-1] + src[ 4] - 5*(src[ 0]+src[3]) + src[1]*C1 + src[2]*C2 + (1<<(SHIFT-1))) >> SHIFT);\
......@@ -42,8 +42,8 @@ static av_unused void OPNAME ## rv40_qpel8_h_lowpass(uint8_t *dst, uint8_t *src,
OP(dst[5], (src[ 3] + src[ 8] - 5*(src[ 4]+src[7]) + src[5]*C1 + src[6]*C2 + (1<<(SHIFT-1))) >> SHIFT);\
OP(dst[6], (src[ 4] + src[ 9] - 5*(src[ 5]+src[8]) + src[6]*C1 + src[7]*C2 + (1<<(SHIFT-1))) >> SHIFT);\
OP(dst[7], (src[ 5] + src[10] - 5*(src[ 6]+src[9]) + src[7]*C1 + src[8]*C2 + (1<<(SHIFT-1))) >> SHIFT);\
dst+=dstStride;\
src+=srcStride;\
dst += dstStride;\
src += srcStride;\
}\
}\
\
......@@ -51,21 +51,21 @@ static void OPNAME ## rv40_qpel8_v_lowpass(uint8_t *dst, uint8_t *src, int dstSt
const int w, const int C1, const int C2, const int SHIFT){\
uint8_t *cm = ff_cropTbl + MAX_NEG_CROP;\
int i;\
for(i=0; i<w; i++)\
for(i = 0; i < w; i++)\
{\
const int srcB = src[-2*srcStride];\
const int srcA = src[-1*srcStride];\
const int src0 = src[0 *srcStride];\
const int src1 = src[1 *srcStride];\
const int src2 = src[2 *srcStride];\
const int src3 = src[3 *srcStride];\
const int src4 = src[4 *srcStride];\
const int src5 = src[5 *srcStride];\
const int src6 = src[6 *srcStride];\
const int src7 = src[7 *srcStride];\
const int src8 = src[8 *srcStride];\
const int src9 = src[9 *srcStride];\
const int src10= src[10*srcStride];\
const int srcB = src[-2*srcStride];\
const int srcA = src[-1*srcStride];\
const int src0 = src[0 *srcStride];\
const int src1 = src[1 *srcStride];\
const int src2 = src[2 *srcStride];\
const int src3 = src[3 *srcStride];\
const int src4 = src[4 *srcStride];\
const int src5 = src[5 *srcStride];\
const int src6 = src[6 *srcStride];\
const int src7 = src[7 *srcStride];\
const int src8 = src[8 *srcStride];\
const int src9 = src[9 *srcStride];\
const int src10 = src[10*srcStride];\
OP(dst[0*dstStride], (srcB + src3 - 5*(srcA+src2) + src0*C1 + src1*C2 + (1<<(SHIFT-1))) >> SHIFT);\
OP(dst[1*dstStride], (srcA + src4 - 5*(src0+src3) + src1*C1 + src2*C2 + (1<<(SHIFT-1))) >> SHIFT);\
OP(dst[2*dstStride], (src0 + src5 - 5*(src1+src4) + src2*C1 + src3*C2 + (1<<(SHIFT-1))) >> SHIFT);\
......@@ -119,21 +119,21 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc01_c(uint8_t *dst, uint8_t *src, i
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc11_c(uint8_t *dst, uint8_t *src, int stride){\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid= full + SIZE*2;\
uint8_t * const full_mid = full + SIZE*2;\
put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 52, 20, 6);\
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 52, 20, 6);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc21_c(uint8_t *dst, uint8_t *src, int stride){\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid= full + SIZE*2;\
uint8_t * const full_mid = full + SIZE*2;\
put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 20, 20, 5);\
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 52, 20, 6);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc31_c(uint8_t *dst, uint8_t *src, int stride){\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid= full + SIZE*2;\
uint8_t * const full_mid = full + SIZE*2;\
put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 20, 52, 6);\
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 52, 20, 6);\
}\
......@@ -144,21 +144,21 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc02_c(uint8_t *dst, uint8_t *src, i
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc12_c(uint8_t *dst, uint8_t *src, int stride){\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid= full + SIZE*2;\
uint8_t * const full_mid = full + SIZE*2;\
put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 52, 20, 6);\
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 20, 5);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc22_c(uint8_t *dst, uint8_t *src, int stride){\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid= full + SIZE*2;\
uint8_t * const full_mid = full + SIZE*2;\
put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 20, 20, 5);\
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 20, 5);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc32_c(uint8_t *dst, uint8_t *src, int stride){\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid= full + SIZE*2;\
uint8_t * const full_mid = full + SIZE*2;\
put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 20, 52, 6);\
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 20, 5);\
}\
......@@ -169,14 +169,14 @@ static void OPNAME ## rv40_qpel ## SIZE ## _mc03_c(uint8_t *dst, uint8_t *src, i
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc13_c(uint8_t *dst, uint8_t *src, int stride){\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid= full + SIZE*2;\
uint8_t * const full_mid = full + SIZE*2;\
put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 52, 20, 6);\
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 52, 6);\
}\
\
static void OPNAME ## rv40_qpel ## SIZE ## _mc23_c(uint8_t *dst, uint8_t *src, int stride){\
uint8_t full[SIZE*(SIZE+5)];\
uint8_t * const full_mid= full + SIZE*2;\
uint8_t * const full_mid = full + SIZE*2;\
put_rv40_qpel ## SIZE ## _h_lowpass(full, src - 2*stride, SIZE, stride, SIZE+5, 20, 20, 5);\
OPNAME ## rv40_qpel ## SIZE ## _v_lowpass(dst, full_mid, stride, SIZE, SIZE, 20, 52, 6);\
}\
......@@ -205,50 +205,50 @@ static const int rv40_bias[4][4] = {
#define RV40_CHROMA_MC(OPNAME, OP)\
static void OPNAME ## rv40_chroma_mc4_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);\
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;\
int bias = rv40_bias[y>>1][x>>1];\
\
assert(x<8 && y<8 && x>=0 && y>=0);\
\
if(D){\
for(i=0; i<h; i++){\
for(i = 0; i < h; i++){\
OP(dst[0], (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1] + bias));\
OP(dst[1], (A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2] + bias));\
OP(dst[2], (A*src[2] + B*src[3] + C*src[stride+2] + D*src[stride+3] + bias));\
OP(dst[3], (A*src[3] + B*src[4] + C*src[stride+3] + D*src[stride+4] + bias));\
dst+= stride;\
src+= stride;\
dst += stride;\
src += stride;\
}\
}else{\
const int E= B+C;\
const int step= C ? stride : 1;\
for(i=0; i<h; i++){\
const int E = B + C;\
const int step = C ? stride : 1;\
for(i = 0; i < h; i++){\
OP(dst[0], (A*src[0] + E*src[step+0] + bias));\
OP(dst[1], (A*src[1] + E*src[step+1] + bias));\
OP(dst[2], (A*src[2] + E*src[step+2] + bias));\
OP(dst[3], (A*src[3] + E*src[step+3] + bias));\
dst+= stride;\
src+= stride;\
dst += stride;\
src += stride;\
}\
}\
}\
\
static void OPNAME ## rv40_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);\
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;\
int bias = rv40_bias[y>>1][x>>1];\
\
assert(x<8 && y<8 && x>=0 && y>=0);\
\
if(D){\
for(i=0; i<h; i++){\
for(i = 0; i < h; i++){\
OP(dst[0], (A*src[0] + B*src[1] + C*src[stride+0] + D*src[stride+1] + bias));\
OP(dst[1], (A*src[1] + B*src[2] + C*src[stride+1] + D*src[stride+2] + bias));\
OP(dst[2], (A*src[2] + B*src[3] + C*src[stride+2] + D*src[stride+3] + bias));\
......@@ -257,13 +257,13 @@ static void OPNAME ## rv40_chroma_mc8_c(uint8_t *dst/*align 8*/, uint8_t *src/*a
OP(dst[5], (A*src[5] + B*src[6] + C*src[stride+5] + D*src[stride+6] + bias));\
OP(dst[6], (A*src[6] + B*src[7] + C*src[stride+6] + D*src[stride+7] + bias));\
OP(dst[7], (A*src[7] + B*src[8] + C*src[stride+7] + D*src[stride+8] + bias));\
dst+= stride;\
src+= stride;\
dst += stride;\
src += stride;\
}\
}else{\
const int E= B+C;\
const int step= C ? stride : 1;\
for(i=0; i<h; i++){\
const int E = B + C;\
const int step = C ? stride : 1;\
for(i = 0; i < h; i++){\
OP(dst[0], (A*src[0] + E*src[step+0] + bias));\
OP(dst[1], (A*src[1] + E*src[step+1] + bias));\
OP(dst[2], (A*src[2] + E*src[step+2] + bias));\
......@@ -272,8 +272,8 @@ static void OPNAME ## rv40_chroma_mc8_c(uint8_t *dst/*align 8*/, uint8_t *src/*a
OP(dst[5], (A*src[5] + E*src[step+5] + bias));\
OP(dst[6], (A*src[6] + E*src[step+6] + bias));\
OP(dst[7], (A*src[7] + E*src[step+7] + bias));\
dst+= stride;\
src+= stride;\
dst += stride;\
src += stride;\
}\
}\
}
......@@ -346,8 +346,8 @@ void ff_rv40dsp_init(DSPContext* c, AVCodecContext *avctx) {
c->avg_rv40_qpel_pixels_tab[1][13] = avg_rv40_qpel8_mc13_c;
c->avg_rv40_qpel_pixels_tab[1][14] = avg_rv40_qpel8_mc23_c;
c->put_rv40_chroma_pixels_tab[0]= put_rv40_chroma_mc8_c;
c->put_rv40_chroma_pixels_tab[1]= put_rv40_chroma_mc4_c;
c->avg_rv40_chroma_pixels_tab[0]= avg_rv40_chroma_mc8_c;
c->avg_rv40_chroma_pixels_tab[1]= avg_rv40_chroma_mc4_c;
c->put_rv40_chroma_pixels_tab[0] = put_rv40_chroma_mc8_c;
c->put_rv40_chroma_pixels_tab[1] = put_rv40_chroma_mc4_c;
c->avg_rv40_chroma_pixels_tab[0] = avg_rv40_chroma_mc8_c;
c->avg_rv40_chroma_pixels_tab[1] = avg_rv40_chroma_mc4_c;
}
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment