October 2019 - ffmpeg-cvslog - lists.ffmpeg.org

avfilter/vf_median: add radiusV option
by Paul B Mahol 31 Oct '19

31 Oct '19

ffmpeg | branch: master | Paul B Mahol <onemda(a)gmail.com> | Wed Oct 30 10:14:23 2019 +0100| [b414cff6304071d5ef1e2b0a1b407e4fe3f58092] | committer: Paul B Mahol avfilter/vf_median: add radiusV option > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=b414cff6304071d5e… --- doc/filters.texi | 7 ++++++- libavfilter/median.h | 1 + libavfilter/median_template.c | 15 ++++++++------- libavfilter/vf_median.c | 10 ++++++---- 4 files changed, 21 insertions(+), 12 deletions(-) diff --git a/doc/filters.texi b/doc/filters.texi index 9d387be1f5..7e9d50782a 100644 --- a/doc/filters.texi +++ b/doc/filters.texi @@ -12919,11 +12919,16 @@ This filter accepts the following options: @table @option @item radius -Set horizontal and vertical radius size. Default value is @code{1}. +Set horizontal radius size. Default value is @code{1}. Allowed range is integer from 1 to 127. @item planes Set which planes to process. Default is @code{15}, which is all available planes. + +@item radiusV +Set vertical radius size. Default value is @code{0}. +Allowed range is integer from 0 to 127. +If it is 0, value will be picked from horizontal @code{radius} option. @end table @section mergeplanes diff --git a/libavfilter/median.h b/libavfilter/median.h index bb07fddd5d..89ae0472b0 100644 --- a/libavfilter/median.h +++ b/libavfilter/median.h @@ -30,6 +30,7 @@ typedef struct MedianContext { int planes; int radius; + int radiusV; int planewidth[4]; int planeheight[4]; diff --git a/libavfilter/median_template.c b/libavfilter/median_template.c index 6953a23a5b..bb7646e0eb 100644 --- a/libavfilter/median_template.c +++ b/libavfilter/median_template.c @@ -57,6 +57,7 @@ static void fn(filter_plane)(AVFilterContext *ctx, const uint8_t *ssrc, int src_ htype *ccoarse = s->coarse[jobnr]; htype *cfine = s->fine[jobnr]; const int radius = s->radius; + const int radiusV = s->radiusV; const int t = s->t; const pixel *src = (const pixel *)ssrc; pixel *dst = (pixel *)ddst; @@ -69,16 +70,16 @@ static void fn(filter_plane)(AVFilterContext *ctx, const uint8_t *ssrc, int src_ memset(cfine, 0, s->fine_size * sizeof(*cfine)); memset(ccoarse, 0, s->coarse_size * sizeof(*ccoarse)); - srcp = src + FFMAX(0, slice_h_start - radius) * src_linesize; + srcp = src + FFMAX(0, slice_h_start - radiusV) * src_linesize; if (jobnr == 0) { for (int i = 0; i < width; i++) { - cfine[PICK_FINE_BIN(width, srcp[i], i)] += radius + 1; - ccoarse[PICK_COARSE_BIN(i, srcp[i])] += radius + 1; + cfine[PICK_FINE_BIN(width, srcp[i], i)] += radiusV + 1; + ccoarse[PICK_COARSE_BIN(i, srcp[i])] += radiusV + 1; } } - srcp = src + FFMAX(0, slice_h_start - radius - (jobnr != 0)) * src_linesize; - for (int i = 0; i < radius + (jobnr != 0) * (1 + radius); i++) { + srcp = src + FFMAX(0, slice_h_start - radiusV - (jobnr != 0)) * src_linesize; + for (int i = 0; i < radiusV + (jobnr != 0) * (1 + radiusV); i++) { for (int j = 0; j < width; j++) { cfine[PICK_FINE_BIN(width, srcp[j], j)]++; ccoarse[PICK_COARSE_BIN(j, srcp[j])]++; @@ -93,13 +94,13 @@ static void fn(filter_plane)(AVFilterContext *ctx, const uint8_t *ssrc, int src_ htype fine[BINS][BINS] = { { 0 } }; htype luc[BINS] = { 0 }; - p = srcp + src_linesize * FFMAX(0, i - radius - 1); + p = srcp + src_linesize * FFMAX(0, i - radiusV - 1); for (int j = 0; j < width; j++) { cfine[PICK_FINE_BIN(width, p[j], j)]--; ccoarse[PICK_COARSE_BIN(j, p[j])]--; } - p = srcp + src_linesize * FFMIN(height - 1, i + radius); + p = srcp + src_linesize * FFMIN(height - 1, i + radiusV); for (int j = 0; j < width; j++) { cfine[PICK_FINE_BIN(width, p[j], j)]++; ccoarse[PICK_COARSE_BIN(j, p[j])]++; diff --git a/libavfilter/vf_median.c b/libavfilter/vf_median.c index 98296bcc2b..49f2351518 100644 --- a/libavfilter/vf_median.c +++ b/libavfilter/vf_median.c @@ -59,6 +59,7 @@ static const AVOption median_options[] = { { "radius", "set median radius", OFFSET(radius), AV_OPT_TYPE_INT, {.i64=1}, 1, 127, FLAGS }, { "planes", "set planes to filter", OFFSET(planes), AV_OPT_TYPE_INT, {.i64=0xF}, 0, 0xF, FLAGS }, + { "radiusV", "set median vertical radius", OFFSET(radiusV), AV_OPT_TYPE_INT, {.i64=0},0, 127, FLAGS }, { NULL } }; @@ -119,8 +120,9 @@ static int config_input(AVFilterLink *inlink) s->planeheight[1] = s->planeheight[2] = AV_CEIL_RSHIFT(inlink->h, desc->log2_chroma_h); s->planeheight[0] = s->planeheight[3] = inlink->h; + s->radiusV = !s->radiusV ? s->radius : s->radiusV; s->nb_planes = av_pix_fmt_count_planes(inlink->format); - s->t = 2 * s->radius * s->radius + 2 * s->radius; + s->t = 2 * s->radius * s->radiusV + 2 * s->radius; for (int i = 0; i < s->nb_planes; i++) { if (!(s->planes & (1 << i))) @@ -131,13 +133,13 @@ static int config_input(AVFilterLink *inlink) return AVERROR(EINVAL); } - if (s->planeheight[i] < s->radius * 2 + 1) { - av_log(inlink->dst, AV_LOG_ERROR, "The %d plane height %d must be not less than %d\n", i, s->planeheight[i], s->radius * 2 + 1); + if (s->planeheight[i] < s->radiusV * 2 + 1) { + av_log(inlink->dst, AV_LOG_ERROR, "The %d plane height %d must be not less than %d\n", i, s->planeheight[i], s->radiusV * 2 + 1); return AVERROR(EINVAL); } } - s->nb_threads = FFMAX(1, FFMIN(s->planeheight[1] / (s->radius + 1), ff_filter_get_nb_threads(inlink->dst))); + s->nb_threads = FFMAX(1, FFMIN(s->planeheight[1] / (s->radiusV + 1), ff_filter_get_nb_threads(inlink->dst))); s->bins = 1 << ((s->depth + 1) / 2); s->fine_size = s->bins * s->bins * inlink->w; s->coarse_size = s->bins * inlink->w;

1 0

avfilter/af_afade: start crossfading only when first stream reached end
by Paul B Mahol 31 Oct '19

31 Oct '19

ffmpeg | branch: master | Paul B Mahol <onemda(a)gmail.com> | Thu Oct 31 10:22:05 2019 +0100| [86a42e954e8e75175a65ef93d05fece8e001e36b] | committer: Paul B Mahol avfilter/af_afade: start crossfading only when first stream reached end > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=86a42e954e8e75175… --- libavfilter/af_afade.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavfilter/af_afade.c b/libavfilter/af_afade.c index 23608fb59b..4edfd27a3c 100644 --- a/libavfilter/af_afade.c +++ b/libavfilter/af_afade.c @@ -485,7 +485,7 @@ static int activate(AVFilterContext *ctx) (AVRational){ 1, outlink->sample_rate }, outlink->time_base); return ff_filter_frame(outlink, in); } else if (ff_inlink_queued_samples(ctx->inputs[0]) >= s->nb_samples && - ff_inlink_queued_samples(ctx->inputs[1]) >= s->nb_samples) { + ff_inlink_queued_samples(ctx->inputs[1]) >= s->nb_samples && s->cf0_eof) { if (s->overlap) { out = ff_get_audio_buffer(outlink, s->nb_samples); if (!out)

1 0

avformat/hlsenc: fix the av_dirname path isn't include separator in the end of string
by Limin Wang 31 Oct '19

31 Oct '19

ffmpeg | branch: master | Limin Wang <lance.lmwang(a)gmail.com> | Wed Oct 23 23:55:15 2019 +0800| [7c872df3e175d1c9de4aeb294abcd7c6ab4bb0f3] | committer: Steven Liu avformat/hlsenc: fix the av_dirname path isn't include separator in the end of string Reviewed-by: Liu Steven <lq(a)chinaffmpeg.org> Signed-off-by: Limin Wang <lance.lmwang(a)gmail.com> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=7c872df3e175d1c9d… --- libavformat/hlsenc.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-) diff --git a/libavformat/hlsenc.c b/libavformat/hlsenc.c index 0876f72f13..e2c1f8a4c3 100644 --- a/libavformat/hlsenc.c +++ b/libavformat/hlsenc.c @@ -478,6 +478,12 @@ static void reflush_dynbuf(VariantStream *vs, int *range_length) avio_write(vs->out, vs->temp_buffer, *range_length);; } +#if HAVE_DOS_PATHS +#define SEPARATOR '\\' +#else +#define SEPARATOR '/' +#endif + static int hls_delete_old_segments(AVFormatContext *s, HLSContext *hls, VariantStream *vs) { @@ -544,7 +550,7 @@ static int hls_delete_old_segments(AVFormatContext *s, HLSContext *hls, while (segment) { av_log(hls, AV_LOG_DEBUG, "deleting old segment %s\n", segment->filename); - path_size = (hls->use_localtime_mkdir ? 0 : strlen(dirname)) + strlen(segment->filename) + 1; + path_size = (hls->use_localtime_mkdir ? 0 : strlen(dirname)+1) + strlen(segment->filename) + 1; path = av_malloc(path_size); if (!path) { ret = AVERROR(ENOMEM); @@ -554,8 +560,7 @@ static int hls_delete_old_segments(AVFormatContext *s, HLSContext *hls, if (hls->use_localtime_mkdir) av_strlcpy(path, segment->filename, path_size); else { // segment->filename contains basename only - av_strlcpy(path, dirname, path_size); - av_strlcat(path, segment->filename, path_size); + snprintf(path, path_size, "%s%c%s", dirname, SEPARATOR, segment->filename); } proto = avio_find_protocol_name(s->url); @@ -575,15 +580,14 @@ static int hls_delete_old_segments(AVFormatContext *s, HLSContext *hls, if ((segment->sub_filename[0] != '\0')) { char *vtt_dirname_r = av_strdup(vs->vtt_avf->url); vtt_dirname = (char*)av_dirname(vtt_dirname_r); - sub_path_size = strlen(segment->sub_filename) + 1 + strlen(vtt_dirname); + sub_path_size = strlen(segment->sub_filename) + 1 + strlen(vtt_dirname) + 1; sub_path = av_malloc(sub_path_size); if (!sub_path) { ret = AVERROR(ENOMEM); goto fail; } - av_strlcpy(sub_path, vtt_dirname, sub_path_size); - av_strlcat(sub_path, segment->sub_filename, sub_path_size); + snprintf(sub_path, sub_path_size, "%s%c%s", vtt_dirname, SEPARATOR, segment->sub_filename); av_freep(&vtt_dirname);

1 0

configure: fix standalone compilation of mxf muxer
by Lou Logan 30 Oct '19

30 Oct '19

ffmpeg | branch: master | Lou Logan <lou(a)lrcd.com> | Tue Oct 29 16:11:46 2019 -0800| [881f083a16a17509aaa53274f2c6f4f4c507dade] | committer: Lou Logan configure: fix standalone compilation of mxf muxer Signed-off-by: Lou Logan <lou(a)lrcd.com> Signed-off-by: Paul B Mahol <onemda(a)gmail.com> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=881f083a16a17509a… --- configure | 1 + 1 file changed, 1 insertion(+) diff --git a/configure b/configure index 7f63eebc9d..875b77fdf1 100755 --- a/configure +++ b/configure @@ -3273,6 +3273,7 @@ mp4_muxer_select="mov_muxer" mpegts_demuxer_select="iso_media" mpegts_muxer_select="adts_muxer latm_muxer" mpegtsraw_demuxer_select="mpegts_demuxer" +mxf_muxer_select="golomb" mxf_d10_muxer_select="mxf_muxer" mxf_opatom_muxer_select="mxf_muxer" nut_muxer_select="riffenc"

1 0

avformat/apngenc: Add deinit function
by Andreas Rheinhardt 30 Oct '19

30 Oct '19

ffmpeg | branch: master | Andreas Rheinhardt <andreas.rheinhardt(a)gmail.com> | Wed Oct 23 14:59:41 2019 +0200| [26aa04991e02f5fb83ffe177fbd832d7d7deb2fb] | committer: James Almer avformat/apngenc: Add deinit function Prevents memleaks when the trailer is never written (e.g. when there was a write error when writing the header). Fixes ticket #8347. Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt(a)gmail.com> Signed-off-by: James Almer <jamrial(a)gmail.com> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=26aa04991e02f5fb8… --- libavformat/apngenc.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/libavformat/apngenc.c b/libavformat/apngenc.c index 77c1c916c2..88cd8054d6 100644 --- a/libavformat/apngenc.c +++ b/libavformat/apngenc.c @@ -251,7 +251,6 @@ static int apng_write_trailer(AVFormatContext *format_context) if (apng->prev_packet) { ret = flush_packet(format_context, NULL); - av_freep(&apng->prev_packet); if (ret < 0) return ret; } @@ -266,12 +265,18 @@ static int apng_write_trailer(AVFormatContext *format_context) apng_write_chunk(io_context, MKBETAG('a', 'c', 'T', 'L'), buf, 8); } - av_freep(&apng->extra_data); - apng->extra_data = 0; - return 0; } +static void apng_deinit(AVFormatContext *s) +{ + APNGMuxContext *apng = s->priv_data; + + av_packet_free(&apng->prev_packet); + av_freep(&apng->extra_data); + apng->extra_data_size = 0; +} + #define OFFSET(x) offsetof(APNGMuxContext, x) #define ENC AV_OPT_FLAG_ENCODING_PARAM static const AVOption options[] = { @@ -300,6 +305,7 @@ AVOutputFormat ff_apng_muxer = { .write_header = apng_write_header, .write_packet = apng_write_packet, .write_trailer = apng_write_trailer, + .deinit = apng_deinit, .priv_class = &apng_muxer_class, .flags = AVFMT_VARIABLE_FPS, };

1 0

avfilter/af_afade: check for eof after crossfade later
by Paul B Mahol 30 Oct '19

30 Oct '19

ffmpeg | branch: master | Paul B Mahol <onemda(a)gmail.com> | Wed Oct 30 19:07:19 2019 +0100| [89389b7ed47b70f98d27b84e2aae3343d92d9682] | committer: Paul B Mahol avfilter/af_afade: check for eof after crossfade later Fixes memleaks and #8346 > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=89389b7ed47b70f98… --- libavfilter/af_afade.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/libavfilter/af_afade.c b/libavfilter/af_afade.c index 10505caf1f..23608fb59b 100644 --- a/libavfilter/af_afade.c +++ b/libavfilter/af_afade.c @@ -454,7 +454,12 @@ static int activate(AVFilterContext *ctx) if (s->crossfade_is_over) { ret = ff_inlink_consume_frame(ctx->inputs[1], &in); - if (ret < 0) { + if (ret > 0) { + in->pts = s->pts; + s->pts += av_rescale_q(in->nb_samples, + (AVRational){ 1, outlink->sample_rate }, outlink->time_base); + return ff_filter_frame(outlink, in); + } else if (ret < 0) { return ret; } else if (ff_inlink_acknowledge_status(ctx->inputs[1], &status, &pts)) { ff_outlink_set_status(ctx->outputs[0], status, pts); @@ -464,11 +469,6 @@ static int activate(AVFilterContext *ctx) ff_inlink_request_frame(ctx->inputs[1]); return 0; } - } else { - in->pts = s->pts; - s->pts += av_rescale_q(in->nb_samples, - (AVRational){ 1, outlink->sample_rate }, outlink->time_base); - return ff_filter_frame(outlink, in); } }

1 0

avcodec/mips: msa optimizations for vc1dsp
by gxw 30 Oct '19

30 Oct '19

ffmpeg | branch: master | gxw <guxiwei-hf(a)loongson.cn> | Mon Oct 21 15:56:47 2019 +0800| [648b422e171d5eab18f6c6fd346e4050d717b936] | committer: Michael Niedermayer avcodec/mips: msa optimizations for vc1dsp Performance of WMV3 decoding has speed up from 3.66x to 5.23x tested on 3A4000. Reviewed-by: Shiyou Yin <yinshiyou-hf(a)loongson.cn> Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=648b422e171d5eab1… --- libavcodec/mips/Makefile | 1 + libavcodec/mips/vc1dsp_init_mips.c | 30 ++- libavcodec/mips/vc1dsp_mips.h | 23 ++ libavcodec/mips/vc1dsp_msa.c | 461 ++++++++++++++++++++++++++++++++++++ libavutil/mips/generic_macros_msa.h | 3 + 5 files changed, 514 insertions(+), 4 deletions(-) diff --git a/libavcodec/mips/Makefile b/libavcodec/mips/Makefile index c5b54d55c0..b4993f6e76 100644 --- a/libavcodec/mips/Makefile +++ b/libavcodec/mips/Makefile @@ -89,3 +89,4 @@ MMI-OBJS-$(CONFIG_WMV2DSP) += mips/wmv2dsp_mmi.o MMI-OBJS-$(CONFIG_HEVC_DECODER) += mips/hevcdsp_mmi.o MMI-OBJS-$(CONFIG_VP3DSP) += mips/vp3dsp_idct_mmi.o MMI-OBJS-$(CONFIG_VP9_DECODER) += mips/vp9_mc_mmi.o +MSA-OBJS-$(CONFIG_VC1_DECODER) += mips/vc1dsp_msa.o diff --git a/libavcodec/mips/vc1dsp_init_mips.c b/libavcodec/mips/vc1dsp_init_mips.c index 4adc9e1d4e..c0007ff650 100644 --- a/libavcodec/mips/vc1dsp_init_mips.c +++ b/libavcodec/mips/vc1dsp_init_mips.c @@ -23,6 +23,10 @@ #include "vc1dsp_mips.h" #include "config.h" +#define FN_ASSIGN(OP, X, Y, INSN) \ + dsp->OP##vc1_mspel_pixels_tab[1][X+4*Y] = ff_##OP##vc1_mspel_mc##X##Y##INSN; \ + dsp->OP##vc1_mspel_pixels_tab[0][X+4*Y] = ff_##OP##vc1_mspel_mc##X##Y##_16##INSN + #if HAVE_MMI static av_cold void vc1dsp_init_mmi(VC1DSPContext *dsp) { @@ -49,10 +53,6 @@ static av_cold void vc1dsp_init_mmi(VC1DSPContext *dsp) dsp->vc1_v_loop_filter16 = ff_vc1_v_loop_filter16_mmi; dsp->vc1_h_loop_filter16 = ff_vc1_h_loop_filter16_mmi; -#define FN_ASSIGN(OP, X, Y, INSN) \ - dsp->OP##vc1_mspel_pixels_tab[1][X+4*Y] = ff_##OP##vc1_mspel_mc##X##Y##INSN; \ - dsp->OP##vc1_mspel_pixels_tab[0][X+4*Y] = ff_##OP##vc1_mspel_mc##X##Y##_16##INSN - FN_ASSIGN(put_, 0, 0, _mmi); FN_ASSIGN(put_, 0, 1, _mmi); FN_ASSIGN(put_, 0, 2, _mmi); @@ -100,9 +100,31 @@ static av_cold void vc1dsp_init_mmi(VC1DSPContext *dsp) } #endif /* HAVE_MMI */ +#if HAVE_MSA +static av_cold void vc1dsp_init_msa(VC1DSPContext *dsp) +{ + dsp->vc1_inv_trans_8x8 = ff_vc1_inv_trans_8x8_msa; + dsp->vc1_inv_trans_4x8 = ff_vc1_inv_trans_4x8_msa; + dsp->vc1_inv_trans_8x4 = ff_vc1_inv_trans_8x4_msa; + + FN_ASSIGN(put_, 1, 1, _msa); + FN_ASSIGN(put_, 1, 2, _msa); + FN_ASSIGN(put_, 1, 3, _msa); + FN_ASSIGN(put_, 2, 1, _msa); + FN_ASSIGN(put_, 2, 2, _msa); + FN_ASSIGN(put_, 2, 3, _msa); + FN_ASSIGN(put_, 3, 1, _msa); + FN_ASSIGN(put_, 3, 2, _msa); + FN_ASSIGN(put_, 3, 3, _msa); +} +#endif /* HAVE_MSA */ + av_cold void ff_vc1dsp_init_mips(VC1DSPContext *dsp) { #if HAVE_MMI vc1dsp_init_mmi(dsp); #endif /* HAVE_MMI */ +#if HAVE_MSA + vc1dsp_init_msa(dsp); +#endif /* HAVE_MSA */ } diff --git a/libavcodec/mips/vc1dsp_mips.h b/libavcodec/mips/vc1dsp_mips.h index ad7d46e539..5897daea8c 100644 --- a/libavcodec/mips/vc1dsp_mips.h +++ b/libavcodec/mips/vc1dsp_mips.h @@ -191,4 +191,27 @@ void ff_avg_no_rnd_vc1_chroma_mc4_mmi(uint8_t *dst /* align 8 */, uint8_t *src /* align 1 */, ptrdiff_t stride, int h, int x, int y); +void ff_vc1_inv_trans_8x8_msa(int16_t block[64]); +void ff_vc1_inv_trans_8x4_msa(uint8_t *dest, ptrdiff_t linesize, int16_t *block); +void ff_vc1_inv_trans_4x8_msa(uint8_t *dest, ptrdiff_t linesize, int16_t *block); + +#define FF_PUT_VC1_MSPEL_MC_MSA(hmode, vmode) \ +void ff_put_vc1_mspel_mc ## hmode ## vmode ## _msa(uint8_t *dst, \ + const uint8_t *src, \ + ptrdiff_t stride, int rnd); \ +void ff_put_vc1_mspel_mc ## hmode ## vmode ## _16_msa(uint8_t *dst, \ + const uint8_t *src, \ + ptrdiff_t stride, int rnd); + +FF_PUT_VC1_MSPEL_MC_MSA(1, 1); +FF_PUT_VC1_MSPEL_MC_MSA(1, 2); +FF_PUT_VC1_MSPEL_MC_MSA(1, 3); + +FF_PUT_VC1_MSPEL_MC_MSA(2, 1); +FF_PUT_VC1_MSPEL_MC_MSA(2, 2); +FF_PUT_VC1_MSPEL_MC_MSA(2, 3); + +FF_PUT_VC1_MSPEL_MC_MSA(3, 1); +FF_PUT_VC1_MSPEL_MC_MSA(3, 2); +FF_PUT_VC1_MSPEL_MC_MSA(3, 3); #endif /* AVCODEC_MIPS_VC1DSP_MIPS_H */ diff --git a/libavcodec/mips/vc1dsp_msa.c b/libavcodec/mips/vc1dsp_msa.c new file mode 100644 index 0000000000..6e588e825a --- /dev/null +++ b/libavcodec/mips/vc1dsp_msa.c @@ -0,0 +1,461 @@ +/* + * Loongson SIMD optimized vc1dsp + * + * Copyright (c) 2019 Loongson Technology Corporation Limited + * gxw <guxiwei-hf(a)loongson.cn> + * + * 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 2.1 of the License, or (at your option) any later version. + * + * 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 FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include "vc1dsp_mips.h" +#include "constants.h" +#include "libavutil/mips/generic_macros_msa.h" + +void ff_vc1_inv_trans_8x8_msa(int16_t block[64]) +{ + v8i16 in0, in1, in2, in3, in4, in5, in6, in7; + v4i32 in_r0, in_r1, in_r2, in_r3, in_r4, in_r5, in_r6, in_r7; + v4i32 in_l0, in_l1, in_l2, in_l3, in_l4, in_l5, in_l6, in_l7; + v4i32 t_r1, t_r2, t_r3, t_r4, t_r5, t_r6, t_r7, t_r8; + v4i32 t_l1, t_l2, t_l3, t_l4, t_l5, t_l6, t_l7, t_l8; + v4i32 cnst_12 = {12, 12, 12, 12}; + v4i32 cnst_4 = {4, 4, 4, 4}; + v4i32 cnst_16 = {16, 16, 16, 16}; + v4i32 cnst_6 = {6, 6, 6, 6}; + v4i32 cnst_15 = {15, 15, 15, 15}; + v4i32 cnst_9 = {9, 9, 9, 9}; + v4i32 cnst_1 = {1, 1, 1, 1}; + v4i32 cnst_64 = {64, 64, 64, 64}; + + LD_SH8(block, 8, in0, in1, in2, in3, in4, in5, in6, in7); + UNPCK_SH_SW(in0, in_r0, in_l0); + UNPCK_SH_SW(in1, in_r1, in_l1); + UNPCK_SH_SW(in2, in_r2, in_l2); + UNPCK_SH_SW(in3, in_r3, in_l3); + UNPCK_SH_SW(in4, in_r4, in_l4); + UNPCK_SH_SW(in5, in_r5, in_l5); + UNPCK_SH_SW(in6, in_r6, in_l6); + UNPCK_SH_SW(in7, in_r7, in_l7); + // First loop + t_r1 = cnst_12 * (in_r0 + in_r4) + cnst_4; + t_l1 = cnst_12 * (in_l0 + in_l4) + cnst_4; + t_r2 = cnst_12 * (in_r0 - in_r4) + cnst_4; + t_l2 = cnst_12 * (in_l0 - in_l4) + cnst_4; + t_r3 = cnst_16 * in_r2 + cnst_6 * in_r6; + t_l3 = cnst_16 * in_l2 + cnst_6 * in_l6; + t_r4 = cnst_6 * in_r2 - cnst_16 * in_r6; + t_l4 = cnst_6 * in_l2 - cnst_16 * in_l6; + + ADD4(t_r1, t_r3, t_l1, t_l3, t_r2, t_r4, t_l2, t_l4, t_r5, t_l5, t_r6, t_l6); + SUB4(t_r2, t_r4, t_l2, t_l4, t_r1, t_r3, t_l1, t_l3, t_r7, t_l7, t_r8, t_l8); + t_r1 = cnst_16 * in_r1 + cnst_15 * in_r3 + cnst_9 * in_r5 + cnst_4 * in_r7; + t_l1 = cnst_16 * in_l1 + cnst_15 * in_l3 + cnst_9 * in_l5 + cnst_4 * in_l7; + t_r2 = cnst_15 * in_r1 - cnst_4 * in_r3 - cnst_16 * in_r5 - cnst_9 * in_r7; + t_l2 = cnst_15 * in_l1 - cnst_4 * in_l3 - cnst_16 * in_l5 - cnst_9 * in_l7; + t_r3 = cnst_9 * in_r1 - cnst_16 * in_r3 + cnst_4 * in_r5 + cnst_15 * in_r7; + t_l3 = cnst_9 * in_l1 - cnst_16 * in_l3 + cnst_4 * in_l5 + cnst_15 * in_l7; + t_r4 = cnst_4 * in_r1 - cnst_9 * in_r3 + cnst_15 * in_r5 - cnst_16 * in_r7; + t_l4 = cnst_4 * in_l1 - cnst_9 * in_l3 + cnst_15 * in_l5 - cnst_16 * in_l7; + + in_r0 = (t_r5 + t_r1) >> 3; + in_l0 = (t_l5 + t_l1) >> 3; + in_r1 = (t_r6 + t_r2) >> 3; + in_l1 = (t_l6 + t_l2) >> 3; + in_r2 = (t_r7 + t_r3) >> 3; + in_l2 = (t_l7 + t_l3) >> 3; + in_r3 = (t_r8 + t_r4) >> 3; + in_l3 = (t_l8 + t_l4) >> 3; + + in_r4 = (t_r8 - t_r4) >> 3; + in_l4 = (t_l8 - t_l4) >> 3; + in_r5 = (t_r7 - t_r3) >> 3; + in_l5 = (t_l7 - t_l3) >> 3; + in_r6 = (t_r6 - t_r2) >> 3; + in_l6 = (t_l6 - t_l2) >> 3; + in_r7 = (t_r5 - t_r1) >> 3; + in_l7 = (t_l5 - t_l1) >> 3; + TRANSPOSE4x4_SW_SW(in_r0, in_r1, in_r2, in_r3, in_r0, in_r1, in_r2, in_r3); + TRANSPOSE4x4_SW_SW(in_l0, in_l1, in_l2, in_l3, in_l0, in_l1, in_l2, in_l3); + TRANSPOSE4x4_SW_SW(in_r4, in_r5, in_r6, in_r7, in_r4, in_r5, in_r6, in_r7); + TRANSPOSE4x4_SW_SW(in_l4, in_l5, in_l6, in_l7, in_l4, in_l5, in_l6, in_l7); + // Second loop + t_r1 = cnst_12 * (in_r0 + in_l0) + cnst_64; + t_l1 = cnst_12 * (in_r4 + in_l4) + cnst_64; + t_r2 = cnst_12 * (in_r0 - in_l0) + cnst_64; + t_l2 = cnst_12 * (in_r4 - in_l4) + cnst_64; + t_r3 = cnst_16 * in_r2 + cnst_6 * in_l2; + t_l3 = cnst_16 * in_r6 + cnst_6 * in_l6; + t_r4 = cnst_6 * in_r2 - cnst_16 * in_l2; + t_l4 = cnst_6 * in_r6 - cnst_16 * in_l6; + + ADD4(t_r1, t_r3, t_l1, t_l3, t_r2, t_r4, t_l2, t_l4, t_r5, t_l5, t_r6, t_l6); + SUB4(t_r2, t_r4, t_l2, t_l4, t_r1, t_r3, t_l1, t_l3, t_r7, t_l7, t_r8, t_l8); + t_r1 = cnst_16 * in_r1 + cnst_15 * in_r3 + cnst_9 * in_l1 + cnst_4 * in_l3; + t_l1 = cnst_16 * in_r5 + cnst_15 * in_r7 + cnst_9 * in_l5 + cnst_4 * in_l7; + t_r2 = cnst_15 * in_r1 - cnst_4 * in_r3 - cnst_16 * in_l1 - cnst_9 * in_l3; + t_l2 = cnst_15 * in_r5 - cnst_4 * in_r7 - cnst_16 * in_l5 - cnst_9 * in_l7; + t_r3 = cnst_9 * in_r1 - cnst_16 * in_r3 + cnst_4 * in_l1 + cnst_15 * in_l3; + t_l3 = cnst_9 * in_r5 - cnst_16 * in_r7 + cnst_4 * in_l5 + cnst_15 * in_l7; + t_r4 = cnst_4 * in_r1 - cnst_9 * in_r3 + cnst_15 * in_l1 - cnst_16 * in_l3; + t_l4 = cnst_4 * in_r5 - cnst_9 * in_r7 + cnst_15 * in_l5 - cnst_16 * in_l7; + + in_r0 = (t_r5 + t_r1) >> 7; + in_l0 = (t_l5 + t_l1) >> 7; + in_r1 = (t_r6 + t_r2) >> 7; + in_l1 = (t_l6 + t_l2) >> 7; + in_r2 = (t_r7 + t_r3) >> 7; + in_l2 = (t_l7 + t_l3) >> 7; + in_r3 = (t_r8 + t_r4) >> 7; + in_l3 = (t_l8 + t_l4) >> 7; + + in_r4 = (t_r8 - t_r4 + cnst_1) >> 7; + in_l4 = (t_l8 - t_l4 + cnst_1) >> 7; + in_r5 = (t_r7 - t_r3 + cnst_1) >> 7; + in_l5 = (t_l7 - t_l3 + cnst_1) >> 7; + in_r6 = (t_r6 - t_r2 + cnst_1) >> 7; + in_l6 = (t_l6 - t_l2 + cnst_1) >> 7; + in_r7 = (t_r5 - t_r1 + cnst_1) >> 7; + in_l7 = (t_l5 - t_l1 + cnst_1) >> 7; + PCKEV_H4_SH(in_l0, in_r0, in_l1, in_r1, in_l2, in_r2, in_l3, in_r3, + in0, in1, in2, in3); + PCKEV_H4_SH(in_l4, in_r4, in_l5, in_r5, in_l6, in_r6, in_l7, in_r7, + in4, in5, in6, in7); + ST_SH8(in0, in1, in2, in3, in4, in5, in6, in7, block, 8); +} + +void ff_vc1_inv_trans_4x8_msa(uint8_t *dest, ptrdiff_t linesize, int16_t *block) +{ + v8i16 in0, in1, in2, in3, in4, in5, in6, in7; + v4i32 in_r0, in_r1, in_r2, in_r3, in_r4, in_r5, in_r6, in_r7; + v4i32 t1, t2, t3, t4, t5, t6, t7, t8; + v4i32 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; + v16i8 zero_m = { 0 }; + v4i32 cnst_17 = {17, 17, 17, 17}; + v4i32 cnst_22 = {22, 22, 22, 22}; + v4i32 cnst_10 = {10, 10, 10, 10}; + v4i32 cnst_12 = {12, 12, 12, 12}; + v4i32 cnst_64 = {64, 64, 64, 64}; + v4i32 cnst_16 = {16, 16, 16, 16}; + v4i32 cnst_15 = {15, 15, 15, 15}; + v4i32 cnst_4 = {4, 4, 4, 4}; + v4i32 cnst_6 = {6, 6, 6, 6}; + v4i32 cnst_9 = {9, 9, 9, 9}; + v4i32 cnst_1 = {1, 1, 1, 1}; + + LD_SH8(block, 8, in0, in1, in2, in3, in4, in5, in6, in7); + UNPCK_R_SH_SW(in0, in_r0); + UNPCK_R_SH_SW(in1, in_r1); + UNPCK_R_SH_SW(in2, in_r2); + UNPCK_R_SH_SW(in3, in_r3); + UNPCK_R_SH_SW(in4, in_r4); + UNPCK_R_SH_SW(in5, in_r5); + UNPCK_R_SH_SW(in6, in_r6); + UNPCK_R_SH_SW(in7, in_r7); + // First loop + TRANSPOSE4x4_SW_SW(in_r0, in_r1, in_r2, in_r3, in_r0, in_r1, in_r2, in_r3); + TRANSPOSE4x4_SW_SW(in_r4, in_r5, in_r6, in_r7, in_r4, in_r5, in_r6, in_r7); + t1 = cnst_17 * (in_r0 + in_r2) + cnst_4; + t5 = cnst_17 * (in_r4 + in_r6) + cnst_4; + t2 = cnst_17 * (in_r0 - in_r2) + cnst_4; + t6 = cnst_17 * (in_r4 - in_r6) + cnst_4; + t3 = cnst_22 * in_r1 + cnst_10 * in_r3; + t7 = cnst_22 * in_r5 + cnst_10 * in_r7; + t4 = cnst_22 * in_r3 - cnst_10 * in_r1; + t8 = cnst_22 * in_r7 - cnst_10 * in_r5; + + in_r0 = (t1 + t3) >> 3; + in_r4 = (t5 + t7) >> 3; + in_r1 = (t2 - t4) >> 3; + in_r5 = (t6 - t8) >> 3; + in_r2 = (t2 + t4) >> 3; + in_r6 = (t6 + t8) >> 3; + in_r3 = (t1 - t3) >> 3; + in_r7 = (t5 - t7) >> 3; + TRANSPOSE4x4_SW_SW(in_r0, in_r1, in_r2, in_r3, in_r0, in_r1, in_r2, in_r3); + TRANSPOSE4x4_SW_SW(in_r4, in_r5, in_r6, in_r7, in_r4, in_r5, in_r6, in_r7); + PCKEV_H4_SH(in_r1, in_r0, in_r3, in_r2, in_r5, in_r4, in_r7, in_r6, + in0, in1, in2, in3); + ST_D8(in0, in1, in2, in3, 0, 1, 0, 1, 0, 1, 0, 1, block, 8); + // Second loop + t1 = cnst_12 * (in_r0 + in_r4) + cnst_64; + t2 = cnst_12 * (in_r0 - in_r4) + cnst_64; + t3 = cnst_16 * in_r2 + cnst_6 * in_r6; + t4 = cnst_6 * in_r2 - cnst_16 * in_r6; + t5 = t1 + t3, t6 = t2 + t4; + t7 = t2 - t4, t8 = t1 - t3; + t1 = cnst_16 * in_r1 + cnst_15 * in_r3 + cnst_9 * in_r5 + cnst_4 * in_r7; + t2 = cnst_15 * in_r1 - cnst_4 * in_r3 - cnst_16 * in_r5 - cnst_9 * in_r7; + t3 = cnst_9 * in_r1 - cnst_16 * in_r3 + cnst_4 * in_r5 + cnst_15 * in_r7; + t4 = cnst_4 * in_r1 - cnst_9 * in_r3 + cnst_15 * in_r5 - cnst_16 * in_r7; + LD_SW8(dest, linesize, dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7); + ILVR_B8_SW(zero_m, dst0, zero_m, dst1, zero_m, dst2, zero_m, dst3, + zero_m, dst4, zero_m, dst5, zero_m, dst6, zero_m, dst7, + dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7); + ILVR_H4_SW(zero_m, dst0, zero_m, dst1, zero_m, dst2, zero_m, dst3, + dst0, dst1, dst2, dst3); + ILVR_H4_SW(zero_m, dst4, zero_m, dst5, zero_m, dst6, zero_m, dst7, + dst4, dst5, dst6, dst7); + in_r0 = (t5 + t1) >> 7; + in_r1 = (t6 + t2) >> 7; + in_r2 = (t7 + t3) >> 7; + in_r3 = (t8 + t4) >> 7; + in_r4 = (t8 - t4 + cnst_1) >> 7; + in_r5 = (t7 - t3 + cnst_1) >> 7; + in_r6 = (t6 - t2 + cnst_1) >> 7; + in_r7 = (t5 - t1 + cnst_1) >> 7; + ADD4(in_r0, dst0, in_r1, dst1, in_r2, dst2, in_r3, dst3, + in_r0, in_r1, in_r2, in_r3); + ADD4(in_r4, dst4, in_r5, dst5, in_r6, dst6, in_r7, dst7, + in_r4, in_r5, in_r6, in_r7); + CLIP_SW8_0_255(in_r0, in_r1, in_r2, in_r3, in_r4, in_r5, in_r6, in_r7); + PCKEV_H4_SH(in_r1, in_r0, in_r3, in_r2, in_r5, in_r4, in_r7, in_r6, + in0, in1, in2, in3); + PCKEV_B2_SH(in1, in0, in3, in2, in0, in1); + ST_W8(in0, in1, 0, 1, 2, 3, 0, 1, 2, 3, dest, linesize); +} + +void ff_vc1_inv_trans_8x4_msa(uint8_t *dest, ptrdiff_t linesize, int16_t *block) +{ + v4i32 in0, in1, in2, in3, in4, in5, in6, in7; + v4i32 t1, t2, t3, t4, t5, t6, t7, t8; + v4i32 dst0, dst1, dst2, dst3, dst4, dst5, dst6, dst7; + v16i8 zero_m = { 0 }; + v4i32 cnst_17 = {17, 17, 17, 17}; + v4i32 cnst_22 = {22, 22, 22, 22}; + v4i32 cnst_10 = {10, 10, 10, 10}; + v4i32 cnst_12 = {12, 12, 12, 12}; + v4i32 cnst_64 = {64, 64, 64, 64}; + v4i32 cnst_16 = {16, 16, 16, 16}; + v4i32 cnst_15 = {15, 15, 15, 15}; + v4i32 cnst_4 = {4, 4, 4, 4}; + v4i32 cnst_6 = {6, 6, 6, 6}; + v4i32 cnst_9 = {9, 9, 9, 9}; + + LD_SW4(block, 8, t1, t2, t3, t4); + UNPCK_SH_SW(t1, in0, in4); + UNPCK_SH_SW(t2, in1, in5); + UNPCK_SH_SW(t3, in2, in6); + UNPCK_SH_SW(t4, in3, in7); + TRANSPOSE4x4_SW_SW(in0, in1, in2, in3, in0, in1, in2, in3); + TRANSPOSE4x4_SW_SW(in4, in5, in6, in7, in4, in5, in6, in7); + // First loop + t1 = cnst_12 * (in0 + in4) + cnst_4; + t2 = cnst_12 * (in0 - in4) + cnst_4; + t3 = cnst_16 * in2 + cnst_6 * in6; + t4 = cnst_6 * in2 - cnst_16 * in6; + t5 = t1 + t3, t6 = t2 + t4; + t7 = t2 - t4, t8 = t1 - t3; + t1 = cnst_16 * in1 + cnst_15 * in3 + cnst_9 * in5 + cnst_4 * in7; + t2 = cnst_15 * in1 - cnst_4 * in3 - cnst_16 * in5 - cnst_9 * in7; + t3 = cnst_9 * in1 - cnst_16 * in3 + cnst_4 * in5 + cnst_15 * in7; + t4 = cnst_4 * in1 - cnst_9 * in3 + cnst_15 * in5 - cnst_16 * in7; + in0 = (t5 + t1) >> 3; + in1 = (t6 + t2) >> 3; + in2 = (t7 + t3) >> 3; + in3 = (t8 + t4) >> 3; + in4 = (t8 - t4) >> 3; + in5 = (t7 - t3) >> 3; + in6 = (t6 - t2) >> 3; + in7 = (t5 - t1) >> 3; + TRANSPOSE4x4_SW_SW(in0, in1, in2, in3, in0, in1, in2, in3); + TRANSPOSE4x4_SW_SW(in4, in5, in6, in7, in4, in5, in6, in7); + PCKEV_H4_SW(in4, in0, in5, in1, in6, in2, in7, in3, t1, t2, t3, t4); + ST_SW4(t1, t2, t3, t4, block, 8); + // Second loop + LD_SW4(dest, linesize, dst0, dst1, dst2, dst3); + ILVR_B4_SW(zero_m, dst0, zero_m, dst1, zero_m, dst2, zero_m, dst3, + dst0, dst1, dst2, dst3); + ILVL_H4_SW(zero_m, dst0, zero_m, dst1, zero_m, dst2, zero_m, dst3, + dst4, dst5, dst6, dst7); + ILVR_H4_SW(zero_m, dst0, zero_m, dst1, zero_m, dst2, zero_m, dst3, + dst0, dst1, dst2, dst3); + // Right part + t1 = cnst_17 * (in0 + in2) + cnst_64; + t2 = cnst_17 * (in0 - in2) + cnst_64; + t3 = cnst_22 * in1 + cnst_10 * in3; + t4 = cnst_22 * in3 - cnst_10 * in1; + in0 = (t1 + t3) >> 7; + in1 = (t2 - t4) >> 7; + in2 = (t2 + t4) >> 7; + in3 = (t1 - t3) >> 7; + ADD4(in0, dst0, in1, dst1, in2, dst2, in3, dst3, in0, in1, in2, in3); + CLIP_SW4_0_255(in0, in1, in2, in3); + // Left part + t5 = cnst_17 * (in4 + in6) + cnst_64; + t6 = cnst_17 * (in4 - in6) + cnst_64; + t7 = cnst_22 * in5 + cnst_10 * in7; + t8 = cnst_22 * in7 - cnst_10 * in5; + in4 = (t5 + t7) >> 7; + in5 = (t6 - t8) >> 7; + in6 = (t6 + t8) >> 7; + in7 = (t5 - t7) >> 7; + ADD4(in4, dst4, in5, dst5, in6, dst6, in7, dst7, in4, in5, in6, in7); + CLIP_SW4_0_255(in4, in5, in6, in7); + PCKEV_H4_SW(in4, in0, in5, in1, in6, in2, in7, in3, in0, in1, in2, in3); + PCKEV_B2_SW(in1, in0, in3, in2, in0, in1); + ST_D4(in0, in1, 0, 1, 0, 1, dest, linesize); +} + +static void put_vc1_mspel_mc_h_v_msa(uint8_t *dst, const uint8_t *src, + ptrdiff_t stride, int hmode, int vmode, + int rnd) +{ + v8i16 in_r0, in_r1, in_r2, in_r3, in_l0, in_l1, in_l2, in_l3; + v8i16 t0, t1, t2, t3, t4, t5, t6, t7; + v8i16 t8, t9, t10, t11, t12, t13, t14, t15; + v8i16 cnst_para0, cnst_para1, cnst_para2, cnst_para3, cnst_r; + static const int para_value[][4] = {{4, 53, 18, 3}, + {1, 9, 9, 1}, + {3, 18, 53, 4}}; + static const int shift_value[] = {0, 5, 1, 5}; + int shift = (shift_value[hmode] + shift_value[vmode]) >> 1; + int r = (1 << (shift - 1)) + rnd - 1; + cnst_r = __msa_fill_h(r); + src -= 1, src -= stride; + cnst_para0 = __msa_fill_h(para_value[vmode - 1][0]); + cnst_para1 = __msa_fill_h(para_value[vmode - 1][1]); + cnst_para2 = __msa_fill_h(para_value[vmode - 1][2]); + cnst_para3 = __msa_fill_h(para_value[vmode - 1][3]); + LD_SH4(src, stride, in_l0, in_l1, in_l2, in_l3); + UNPCK_UB_SH(in_l0, in_r0, in_l0); + UNPCK_UB_SH(in_l1, in_r1, in_l1); + UNPCK_UB_SH(in_l2, in_r2, in_l2); + UNPCK_UB_SH(in_l3, in_r3, in_l3); + // row 0 + t0 = cnst_para1 * in_r1 + cnst_para2 * in_r2 + - cnst_para0 * in_r0 - cnst_para3 * in_r3; + t8 = cnst_para1 * in_l1 + cnst_para2 * in_l2 + - cnst_para0 * in_l0 - cnst_para3 * in_l3; + in_l0 = LD_SH(src + 4 * stride); + UNPCK_UB_SH(in_l0, in_r0, in_l0); + // row 1 + t1 = cnst_para1 * in_r2 + cnst_para2 * in_r3 + - cnst_para0 * in_r1 - cnst_para3 * in_r0; + t9 = cnst_para1 * in_l2 + cnst_para2 * in_l3 + - cnst_para0 * in_l1 - cnst_para3 * in_l0; + in_l1 = LD_SH(src + 5 * stride); + UNPCK_UB_SH(in_l1, in_r1, in_l1); + // row 2 + t2 = cnst_para1 * in_r3 + cnst_para2 * in_r0 + - cnst_para0 * in_r2 - cnst_para3 * in_r1; + t10 = cnst_para1 * in_l3 + cnst_para2 * in_l0 + - cnst_para0 * in_l2 - cnst_para3 * in_l1; + in_l2 = LD_SH(src + 6 * stride); + UNPCK_UB_SH(in_l2, in_r2, in_l2); + // row 3 + t3 = cnst_para1 * in_r0 + cnst_para2 * in_r1 + - cnst_para0 * in_r3 - cnst_para3 * in_r2; + t11 = cnst_para1 * in_l0 + cnst_para2 * in_l1 + - cnst_para0 * in_l3 - cnst_para3 * in_l2; + in_l3 = LD_SH(src + 7 * stride); + UNPCK_UB_SH(in_l3, in_r3, in_l3); + // row 4 + t4 = cnst_para1 * in_r1 + cnst_para2 * in_r2 + - cnst_para0 * in_r0 - cnst_para3 * in_r3; + t12 = cnst_para1 * in_l1 + cnst_para2 * in_l2 + - cnst_para0 * in_l0 - cnst_para3 * in_l3; + in_l0 = LD_SH(src + 8 * stride); + UNPCK_UB_SH(in_l0, in_r0, in_l0); + // row 5 + t5 = cnst_para1 * in_r2 + cnst_para2 * in_r3 + - cnst_para0 * in_r1 - cnst_para3 * in_r0; + t13 = cnst_para1 * in_l2 + cnst_para2 * in_l3 + - cnst_para0 * in_l1 - cnst_para3 * in_l0; + in_l1 = LD_SH(src + 9 * stride); + UNPCK_UB_SH(in_l1, in_r1, in_l1); + // row 6 + t6 = cnst_para1 * in_r3 + cnst_para2 * in_r0 + - cnst_para0 * in_r2 - cnst_para3 * in_r1; + t14 = cnst_para1 * in_l3 + cnst_para2 * in_l0 + - cnst_para0 * in_l2 - cnst_para3 * in_l1; + in_l2 = LD_SH(src + 10 * stride); + UNPCK_UB_SH(in_l2, in_r2, in_l2); + // row 7 + t7 = cnst_para1 * in_r0 + cnst_para2 * in_r1 + - cnst_para0 * in_r3 - cnst_para3 * in_r2; + t15 = cnst_para1 * in_l0 + cnst_para2 * in_l1 + - cnst_para0 * in_l3 - cnst_para3 * in_l2; + + ADD4(t0, cnst_r, t1, cnst_r, t2, cnst_r, t3, cnst_r, t0, t1, t2, t3); + ADD4(t4, cnst_r, t5, cnst_r, t6, cnst_r, t7, cnst_r, t4, t5, t6, t7); + ADD4(t8, cnst_r, t9, cnst_r, t10, cnst_r, t11, cnst_r, + t8, t9, t10, t11); + ADD4(t12, cnst_r, t13, cnst_r, t14, cnst_r, t15, cnst_r, + t12, t13, t14, t15); + t0 >>= shift, t1 >>= shift, t2 >>= shift, t3 >>= shift; + t4 >>= shift, t5 >>= shift, t6 >>= shift, t7 >>= shift; + t8 >>= shift, t9 >>= shift, t10 >>= shift, t11 >>= shift; + t12 >>= shift, t13 >>= shift, t14 >>= shift, t15 >>= shift; + TRANSPOSE8x8_SH_SH(t0, t1, t2, t3, t4, t5, t6, t7, + t0, t1, t2, t3, t4, t5, t6, t7); + TRANSPOSE8x8_SH_SH(t8, t9, t10, t11, t12, t13, t14, t15, + t8, t9, t10, t11, t12, t13, t14, t15); + cnst_para0 = __msa_fill_h(para_value[hmode - 1][0]); + cnst_para1 = __msa_fill_h(para_value[hmode - 1][1]); + cnst_para2 = __msa_fill_h(para_value[hmode - 1][2]); + cnst_para3 = __msa_fill_h(para_value[hmode - 1][3]); + r = 64 - rnd; + cnst_r = __msa_fill_h(r); + // col 0 ~ 7 + t0 = cnst_para1 * t1 + cnst_para2 * t2 - cnst_para0 * t0 - cnst_para3 * t3; + t1 = cnst_para1 * t2 + cnst_para2 * t3 - cnst_para0 * t1 - cnst_para3 * t4; + t2 = cnst_para1 * t3 + cnst_para2 * t4 - cnst_para0 * t2 - cnst_para3 * t5; + t3 = cnst_para1 * t4 + cnst_para2 * t5 - cnst_para0 * t3 - cnst_para3 * t6; + t4 = cnst_para1 * t5 + cnst_para2 * t6 - cnst_para0 * t4 - cnst_para3 * t7; + t5 = cnst_para1 * t6 + cnst_para2 * t7 - cnst_para0 * t5 - cnst_para3 * t8; + t6 = cnst_para1 * t7 + cnst_para2 * t8 - cnst_para0 * t6 - cnst_para3 * t9; + t7 = cnst_para1 * t8 + cnst_para2 * t9 - cnst_para0 * t7 - cnst_para3 * t10; + ADD4(t0, cnst_r, t1, cnst_r, t2, cnst_r, t3, cnst_r, t0, t1, t2, t3); + ADD4(t4, cnst_r, t5, cnst_r, t6, cnst_r, t7, cnst_r, t4, t5, t6, t7); + t0 >>= 7, t1 >>= 7, t2 >>= 7, t3 >>= 7; + t4 >>= 7, t5 >>= 7, t6 >>= 7, t7 >>= 7; + TRANSPOSE8x8_SH_SH(t0, t1, t2, t3, t4, t5, t6, t7, + t0, t1, t2, t3, t4, t5, t6, t7); + CLIP_SH8_0_255(t0, t1, t2, t3, t4, t5, t6, t7); + PCKEV_B4_SH(t1, t0, t3, t2, t5, t4, t7, t6, t0, t1, t2, t3); + ST_D8(t0, t1, t2, t3, 0, 1, 0, 1, 0, 1, 0, 1, dst, stride); +} + +#define PUT_VC1_MSPEL_MC_MSA(hmode, vmode) \ +void ff_put_vc1_mspel_mc ## hmode ## vmode ## _msa(uint8_t *dst, \ + const uint8_t *src, \ + ptrdiff_t stride, int rnd) \ +{ \ + put_vc1_mspel_mc_h_v_msa(dst, src, stride, hmode, vmode, rnd); \ +} \ +void ff_put_vc1_mspel_mc ## hmode ## vmode ## _16_msa(uint8_t *dst, \ + const uint8_t *src, \ + ptrdiff_t stride, int rnd) \ +{ \ + put_vc1_mspel_mc_h_v_msa(dst, src, stride, hmode, vmode, rnd); \ + put_vc1_mspel_mc_h_v_msa(dst + 8, src + 8, stride, hmode, vmode, rnd); \ + dst += 8 * stride, src += 8 * stride; \ + put_vc1_mspel_mc_h_v_msa(dst, src, stride, hmode, vmode, rnd); \ + put_vc1_mspel_mc_h_v_msa(dst + 8, src + 8, stride, hmode, vmode, rnd); \ +} + +PUT_VC1_MSPEL_MC_MSA(1, 1); +PUT_VC1_MSPEL_MC_MSA(1, 2); +PUT_VC1_MSPEL_MC_MSA(1, 3); + +PUT_VC1_MSPEL_MC_MSA(2, 1); +PUT_VC1_MSPEL_MC_MSA(2, 2); +PUT_VC1_MSPEL_MC_MSA(2, 3); + +PUT_VC1_MSPEL_MC_MSA(3, 1); +PUT_VC1_MSPEL_MC_MSA(3, 2); +PUT_VC1_MSPEL_MC_MSA(3, 3); diff --git a/libavutil/mips/generic_macros_msa.h b/libavutil/mips/generic_macros_msa.h index c25509e483..267d4e6ca5 100644 --- a/libavutil/mips/generic_macros_msa.h +++ b/libavutil/mips/generic_macros_msa.h @@ -299,6 +299,7 @@ #define LD_SB4(...) LD_V4(v16i8, __VA_ARGS__) #define LD_UH4(...) LD_V4(v8u16, __VA_ARGS__) #define LD_SH4(...) LD_V4(v8i16, __VA_ARGS__) +#define LD_SW4(...) LD_V4(v4i32, __VA_ARGS__) #define LD_V5(RTYPE, psrc, stride, out0, out1, out2, out3, out4) \ { \ @@ -337,6 +338,7 @@ #define LD_SB8(...) LD_V8(v16i8, __VA_ARGS__) #define LD_UH8(...) LD_V8(v8u16, __VA_ARGS__) #define LD_SH8(...) LD_V8(v8i16, __VA_ARGS__) +#define LD_SW8(...) LD_V8(v4i32, __VA_ARGS__) #define LD_V16(RTYPE, psrc, stride, \ out0, out1, out2, out3, out4, out5, out6, out7, \ @@ -1382,6 +1384,7 @@ out4, out5, out6, out7); \ } #define ILVR_B8_UH(...) ILVR_B8(v8u16, __VA_ARGS__) +#define ILVR_B8_SW(...) ILVR_B8(v4i32, __VA_ARGS__) /* Description : Interleave right half of halfword elements from vectors Arguments : Inputs - in0, in1, in2, in3, in4, in5, in6, in7

1 0

avfilter/f_sidedata: fix Wtautological-constant-out-of-range-compare
by Zhao Zhili 30 Oct '19

30 Oct '19

ffmpeg | branch: master | Zhao Zhili <quinkblack(a)foxmail.com> | Wed Oct 16 01:17:54 2019 +0800| [af70c94c63cc0ccf6a8078e87c81d061b8765889] | committer: Michael Niedermayer avfilter/f_sidedata: fix Wtautological-constant-out-of-range-compare Reviewed-by: Paul B Mahol <onemda(a)gmail.com> Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=af70c94c63cc0ccf6… --- libavfilter/f_sidedata.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavfilter/f_sidedata.c b/libavfilter/f_sidedata.c index 381da5a052..4210dcac4c 100644 --- a/libavfilter/f_sidedata.c +++ b/libavfilter/f_sidedata.c @@ -39,7 +39,7 @@ typedef struct SideDataContext { const AVClass *class; int mode; - enum AVFrameSideDataType type; + int type; // enum AVFrameSideDataType or -1 for delete side data mode } SideDataContext; #define OFFSET(x) offsetof(SideDataContext, x)

1 0

FATE: add a test for freeezedetect
by Limin Wang 30 Oct '19

30 Oct '19

ffmpeg | branch: master | Limin Wang <lance.lmwang(a)gmail.com> | Mon Oct 21 18:12:13 2019 +0800| [78c0030f344c408d71eb35db7f880ec96a1be03a] | committer: Michael Niedermayer FATE: add a test for freeezedetect Signed-off-by: Limin Wang <lance.lmwang(a)gmail.com> Signed-off-by: Michael Niedermayer <michael(a)niedermayer.cc> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=78c0030f344c408d7… --- tests/fate/filter-video.mak | 4 + tests/ref/fate/filter-metadata-freezedetect | 251 ++++++++++++++++++++++++++++ 2 files changed, 255 insertions(+) diff --git a/tests/fate/filter-video.mak b/tests/fate/filter-video.mak index 9e6351be3e..7f55c7380b 100644 --- a/tests/fate/filter-video.mak +++ b/tests/fate/filter-video.mak @@ -744,6 +744,10 @@ FATE_METADATA_FILTER-$(call ALLYES, $(CROPDETECT_DEPS)) += fate-filter-metadata- fate-filter-metadata-cropdetect: SRC = $(TARGET_SAMPLES)/filter/cropdetect.mp4 fate-filter-metadata-cropdetect: CMD = run $(FILTER_METADATA_COMMAND) "sws_flags=+accurate_rnd+bitexact;movie='$(SRC)',cropdetect=max_outliers=3" +FREEZEDETECT_DEPS = FFPROBE AVDEVICE LAVFI_INDEV MPTESTSRC_FILTER SCALE_FILTER FREEZEDETECT_FILTER +FATE_METADATA_FILTER-$(call ALLYES, $(FREEZEDETECT_DEPS)) += fate-filter-metadata-freezedetect +fate-filter-metadata-freezedetect: CMD = run $(FILTER_METADATA_COMMAND) "sws_flags=+accurate_rnd+bitexact;mptestsrc=r=25:d=10:m=51,freezedetect" + SILENCEDETECT_DEPS = FFPROBE AVDEVICE LAVFI_INDEV AMOVIE_FILTER TTA_DEMUXER TTA_DECODER SILENCEDETECT_FILTER FATE_METADATA_FILTER-$(call ALLYES, $(SILENCEDETECT_DEPS)) += fate-filter-metadata-silencedetect fate-filter-metadata-silencedetect: SRC = $(TARGET_SAMPLES)/lossless-audio/inside.tta diff --git a/tests/ref/fate/filter-metadata-freezedetect b/tests/ref/fate/filter-metadata-freezedetect new file mode 100644 index 0000000000..a0ee38e458 --- /dev/null +++ b/tests/ref/fate/filter-metadata-freezedetect @@ -0,0 +1,251 @@ +pkt_pts=0 +pkt_pts=1 +pkt_pts=2 +pkt_pts=3 +pkt_pts=4 +pkt_pts=5 +pkt_pts=6 +pkt_pts=7 +pkt_pts=8 +pkt_pts=9 +pkt_pts=10 +pkt_pts=11 +pkt_pts=12 +pkt_pts=13 +pkt_pts=14 +pkt_pts=15 +pkt_pts=16 +pkt_pts=17 +pkt_pts=18 +pkt_pts=19 +pkt_pts=20 +pkt_pts=21 +pkt_pts=22 +pkt_pts=23 +pkt_pts=24 +pkt_pts=25 +pkt_pts=26 +pkt_pts=27 +pkt_pts=28 +pkt_pts=29 +pkt_pts=30 +pkt_pts=31 +pkt_pts=32 +pkt_pts=33 +pkt_pts=34 +pkt_pts=35 +pkt_pts=36 +pkt_pts=37 +pkt_pts=38 +pkt_pts=39 +pkt_pts=40 +pkt_pts=41 +pkt_pts=42 +pkt_pts=43 +pkt_pts=44 +pkt_pts=45 +pkt_pts=46 +pkt_pts=47 +pkt_pts=48 +pkt_pts=49 +pkt_pts=50 +pkt_pts=51 +pkt_pts=52 +pkt_pts=53 +pkt_pts=54 +pkt_pts=55 +pkt_pts=56 +pkt_pts=57 +pkt_pts=58 +pkt_pts=59 +pkt_pts=60 +pkt_pts=61 +pkt_pts=62 +pkt_pts=63 +pkt_pts=64 +pkt_pts=65 +pkt_pts=66 +pkt_pts=67 +pkt_pts=68 +pkt_pts=69 +pkt_pts=70 +pkt_pts=71 +pkt_pts=72 +pkt_pts=73 +pkt_pts=74 +pkt_pts=75 +pkt_pts=76 +pkt_pts=77 +pkt_pts=78 +pkt_pts=79 +pkt_pts=80 +pkt_pts=81 +pkt_pts=82 +pkt_pts=83 +pkt_pts=84 +pkt_pts=85 +pkt_pts=86 +pkt_pts=87 +pkt_pts=88 +pkt_pts=89 +pkt_pts=90 +pkt_pts=91 +pkt_pts=92 +pkt_pts=93 +pkt_pts=94 +pkt_pts=95 +pkt_pts=96 +pkt_pts=97 +pkt_pts=98 +pkt_pts=99 +pkt_pts=100 +pkt_pts=101 +pkt_pts=102 +pkt_pts=103 +pkt_pts=104 +pkt_pts=105 +pkt_pts=106 +pkt_pts=107 +pkt_pts=108 +pkt_pts=109 +pkt_pts=110 +pkt_pts=111 +pkt_pts=112 +pkt_pts=113 +pkt_pts=114 +pkt_pts=115 +pkt_pts=116 +pkt_pts=117 +pkt_pts=118 +pkt_pts=119 +pkt_pts=120 +pkt_pts=121 +pkt_pts=122 +pkt_pts=123 +pkt_pts=124 +pkt_pts=125 +pkt_pts=126 +pkt_pts=127 +pkt_pts=128 +pkt_pts=129 +pkt_pts=130 +pkt_pts=131 +pkt_pts=132 +pkt_pts=133 +pkt_pts=134 +pkt_pts=135 +pkt_pts=136 +pkt_pts=137 +pkt_pts=138 +pkt_pts=139 +pkt_pts=140 +pkt_pts=141 +pkt_pts=142 +pkt_pts=143 +pkt_pts=144 +pkt_pts=145 +pkt_pts=146 +pkt_pts=147 +pkt_pts=148 +pkt_pts=149 +pkt_pts=150 +pkt_pts=151 +pkt_pts=152 +pkt_pts=153|tag:lavfi.freezedetect.freeze_start=4.12|tag:lavfi.freezedetect.freeze_duration=2|tag:lavfi.freezedetect.freeze_end=6.12 +pkt_pts=154 +pkt_pts=155 +pkt_pts=156 +pkt_pts=157 +pkt_pts=158 +pkt_pts=159 +pkt_pts=160 +pkt_pts=161 +pkt_pts=162 +pkt_pts=163 +pkt_pts=164 +pkt_pts=165 +pkt_pts=166 +pkt_pts=167 +pkt_pts=168 +pkt_pts=169 +pkt_pts=170 +pkt_pts=171 +pkt_pts=172 +pkt_pts=173 +pkt_pts=174 +pkt_pts=175 +pkt_pts=176 +pkt_pts=177 +pkt_pts=178 +pkt_pts=179 +pkt_pts=180 +pkt_pts=181 +pkt_pts=182 +pkt_pts=183 +pkt_pts=184 +pkt_pts=185 +pkt_pts=186 +pkt_pts=187 +pkt_pts=188 +pkt_pts=189 +pkt_pts=190 +pkt_pts=191 +pkt_pts=192 +pkt_pts=193 +pkt_pts=194 +pkt_pts=195 +pkt_pts=196 +pkt_pts=197 +pkt_pts=198 +pkt_pts=199 +pkt_pts=200 +pkt_pts=201 +pkt_pts=202 +pkt_pts=203 +pkt_pts=204|tag:lavfi.freezedetect.freeze_start=6.16|tag:lavfi.freezedetect.freeze_duration=2|tag:lavfi.freezedetect.freeze_end=8.16 +pkt_pts=205 +pkt_pts=206 +pkt_pts=207 +pkt_pts=208 +pkt_pts=209 +pkt_pts=210 +pkt_pts=211 +pkt_pts=212 +pkt_pts=213 +pkt_pts=214 +pkt_pts=215 +pkt_pts=216 +pkt_pts=217 +pkt_pts=218 +pkt_pts=219 +pkt_pts=220 +pkt_pts=221 +pkt_pts=222 +pkt_pts=223 +pkt_pts=224 +pkt_pts=225 +pkt_pts=226 +pkt_pts=227 +pkt_pts=228 +pkt_pts=229 +pkt_pts=230 +pkt_pts=231 +pkt_pts=232 +pkt_pts=233 +pkt_pts=234 +pkt_pts=235 +pkt_pts=236 +pkt_pts=237 +pkt_pts=238 +pkt_pts=239 +pkt_pts=240 +pkt_pts=241 +pkt_pts=242 +pkt_pts=243 +pkt_pts=244 +pkt_pts=245 +pkt_pts=246 +pkt_pts=247 +pkt_pts=248 +pkt_pts=249 +pkt_pts=250

1 0

doc/filters: move drawgraph to video filters chapter
by Paul B Mahol 30 Oct '19

30 Oct '19

ffmpeg | branch: master | Paul B Mahol <onemda(a)gmail.com> | Wed Oct 30 17:55:44 2019 +0100| [d4ba2506ce9e05dda598f702fa63c183580927b4] | committer: Paul B Mahol doc/filters: move drawgraph to video filters chapter > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=d4ba2506ce9e05dda… --- doc/filters.texi | 215 ++++++++++++++++++++++++++++--------------------------- 1 file changed, 110 insertions(+), 105 deletions(-) diff --git a/doc/filters.texi b/doc/filters.texi index 8fdcb196b5..9d387be1f5 100644 --- a/doc/filters.texi +++ b/doc/filters.texi @@ -9037,6 +9037,111 @@ The command accepts the same syntax of the corresponding option. If the specified expression is not valid, it is kept at its current value. +@anchor{drawgraph} +@section drawgraph +Draw a graph using input video metadata. + +It accepts the following parameters: + +@table @option +@item m1 +Set 1st frame metadata key from which metadata values will be used to draw a graph. + +@item fg1 +Set 1st foreground color expression. + +@item m2 +Set 2nd frame metadata key from which metadata values will be used to draw a graph. + +@item fg2 +Set 2nd foreground color expression. + +@item m3 +Set 3rd frame metadata key from which metadata values will be used to draw a graph. + +@item fg3 +Set 3rd foreground color expression. + +@item m4 +Set 4th frame metadata key from which metadata values will be used to draw a graph. + +@item fg4 +Set 4th foreground color expression. + +@item min +Set minimal value of metadata value. + +@item max +Set maximal value of metadata value. + +@item bg +Set graph background color. Default is white. + +@item mode +Set graph mode. + +Available values for mode is: +@table @samp +@item bar +@item dot +@item line +@end table + +Default is @code{line}. + +@item slide +Set slide mode. + +Available values for slide is: +@table @samp +@item frame +Draw new frame when right border is reached. + +@item replace +Replace old columns with new ones. + +@item scroll +Scroll from right to left. + +@item rscroll +Scroll from left to right. + +@item picture +Draw single picture. +@end table + +Default is @code{frame}. + +@item size +Set size of graph video. For the syntax of this option, check the +@ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}. +The default value is @code{900x256}. + +The foreground color expressions can use the following variables: +@table @option +@item MIN +Minimal value of metadata value. + +@item MAX +Maximal value of metadata value. + +@item VAL +Current metadata key value. +@end table + +The color is defined as 0xAABBGGRR. +@end table + +Example using metadata from @ref{signalstats} filter: +@example +signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255 +@end example + +Example using metadata from @ref{ebur128} filter: +@example +ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5 +@end example + @section drawgrid Draw a grid on the input image. @@ -21498,6 +21603,11 @@ draw channels. Unrecognized or missing colors will be replaced by white color. @end table +@section adrawgraph +Draw a graph using input audio metadata. + +See @ref{drawgraph} + @section agraphmonitor See @ref{graphmonitor}. @@ -21863,111 +21973,6 @@ This filter supports the following commands: Close the current segment and step to the next one @end table -@section drawgraph, adrawgraph - -Draw a graph using input video or audio metadata. - -It accepts the following parameters: - -@table @option -@item m1 -Set 1st frame metadata key from which metadata values will be used to draw a graph. - -@item fg1 -Set 1st foreground color expression. - -@item m2 -Set 2nd frame metadata key from which metadata values will be used to draw a graph. - -@item fg2 -Set 2nd foreground color expression. - -@item m3 -Set 3rd frame metadata key from which metadata values will be used to draw a graph. - -@item fg3 -Set 3rd foreground color expression. - -@item m4 -Set 4th frame metadata key from which metadata values will be used to draw a graph. - -@item fg4 -Set 4th foreground color expression. - -@item min -Set minimal value of metadata value. - -@item max -Set maximal value of metadata value. - -@item bg -Set graph background color. Default is white. - -@item mode -Set graph mode. - -Available values for mode is: -@table @samp -@item bar -@item dot -@item line -@end table - -Default is @code{line}. - -@item slide -Set slide mode. - -Available values for slide is: -@table @samp -@item frame -Draw new frame when right border is reached. - -@item replace -Replace old columns with new ones. - -@item scroll -Scroll from right to left. - -@item rscroll -Scroll from left to right. - -@item picture -Draw single picture. -@end table - -Default is @code{frame}. - -@item size -Set size of graph video. For the syntax of this option, check the -@ref{video size syntax,,"Video size" section in the ffmpeg-utils manual,ffmpeg-utils}. -The default value is @code{900x256}. - -The foreground color expressions can use the following variables: -@table @option -@item MIN -Minimal value of metadata value. - -@item MAX -Maximal value of metadata value. - -@item VAL -Current metadata key value. -@end table - -The color is defined as 0xAABBGGRR. -@end table - -Example using metadata from @ref{signalstats} filter: -@example -signalstats,drawgraph=lavfi.signalstats.YAVG:min=0:max=255 -@end example - -Example using metadata from @ref{ebur128} filter: -@example -ebur128=metadata=1,adrawgraph=lavfi.r128.M:min=-120:max=5 -@end example - @anchor{ebur128} @section ebur128

1 0