May 2012 - ffmpeg-cvslog - lists.ffmpeg.org

http: Add 'post_data', a new option which sets custom HTTP post data
by git＠videolan.org 31 May '12

31 May '12

ffmpeg | branch: master | Samuel Pitoiset <samuel.pitoiset(a)gmail.com> | Wed May 30 11:27:18 2012 +0200| [1876e7c0c2ec0d4a9c2daa45e0ab311b1c2ac413] | committer: Martin Storsjö http: Add 'post_data', a new option which sets custom HTTP post data This allows doing http posts with a content-length header sent in advance, avoiding chunked encoding. Signed-off-by: Martin Storsjö <martin(a)martin.st> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=1876e7c0c2ec0d4a9… --- libavformat/http.c | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-) diff --git a/libavformat/http.c b/libavformat/http.c index 65c031e..bcb1227 100644 --- a/libavformat/http.c +++ b/libavformat/http.c @@ -54,6 +54,8 @@ typedef struct { int end_chunked_post; /**< A flag which indicates if the end of chunked encoding has been sent. */ int end_header; /**< A flag which indicates we have finished to read POST reply. */ int multiple_requests; /**< A flag which indicates if we use persistent connections. */ + uint8_t *post_data; + int post_datalen; } HTTPContext; #define OFFSET(x) offsetof(HTTPContext, x) @@ -63,6 +65,7 @@ static const AVOption options[] = { {"chunked_post", "use chunked transfer-encoding for posts", OFFSET(chunked_post), AV_OPT_TYPE_INT, {.dbl = 1}, 0, 1, E }, {"headers", "custom HTTP headers, can override built in default headers", OFFSET(headers), AV_OPT_TYPE_STRING, { 0 }, 0, 0, D|E }, {"multiple_requests", "use persistent connections", OFFSET(multiple_requests), AV_OPT_TYPE_INT, {.dbl = 0}, 0, 1, D|E }, +{"post_data", "custom HTTP post data", OFFSET(post_data), AV_OPT_TYPE_BINARY, .flags = D|E }, {NULL} }; #define HTTP_CLASS(flavor)\ @@ -382,6 +385,14 @@ static int http_connect(URLContext *h, const char *path, const char *local_path, /* send http header */ post = h->flags & AVIO_FLAG_WRITE; + + if (s->post_data) { + /* force POST method and disable chunked encoding when + * custom HTTP post data is set */ + post = 1; + s->chunked_post = 0; + } + method = post ? "POST" : "GET"; authstr = ff_http_auth_create_response(&s->auth_state, auth, local_path, method); @@ -412,6 +423,9 @@ static int http_connect(URLContext *h, const char *path, const char *local_path, if (!has_header(s->headers, "\r\nHost: ")) len += av_strlcatf(headers + len, sizeof(headers) - len, "Host: %s\r\n", hoststr); + if (!has_header(s->headers, "\r\nContent-Length: ") && s->post_data) + len += av_strlcatf(headers + len, sizeof(headers) - len, + "Content-Length: %d\r\n", s->post_datalen); /* now add in custom headers */ if (s->headers) @@ -436,6 +450,10 @@ static int http_connect(URLContext *h, const char *path, const char *local_path, if (ffurl_write(s->hd, s->buffer, strlen(s->buffer)) < 0) return AVERROR(EIO); + if (s->post_data) + if ((err = ffurl_write(s->hd, s->post_data, s->post_datalen)) < 0) + return err; + /* init input buffer */ s->buf_ptr = s->buffer; s->buf_end = s->buffer; @@ -445,7 +463,7 @@ static int http_connect(URLContext *h, const char *path, const char *local_path, s->willclose = 0; s->end_chunked_post = 0; s->end_header = 0; - if (post) { + if (post && !s->post_data) { /* Pretend that it did work. We didn't read any header yet, since * we've still to send the POST data, but the code calling this * function will check http_code after we return. */

1 0

lavfi: amix: check active input count before calling request_samples
by git＠videolan.org 31 May '12

31 May '12

ffmpeg | branch: master | Justin Ruggles <justin.ruggles(a)gmail.com> | Wed May 30 08:48:26 2012 -0400| [c7bd556d482b3bb25bb6dd3bdf3542622d7cd482] | committer: Justin Ruggles lavfi: amix: check active input count before calling request_samples fixes use of the amix filter with only 1 input > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=c7bd556d482b3bb25… --- libavfilter/af_amix.c | 19 ++++++++++++------- 1 file changed, 12 insertions(+), 7 deletions(-) diff --git a/libavfilter/af_amix.c b/libavfilter/af_amix.c index 3399b7c..d4af9a4 100644 --- a/libavfilter/af_amix.c +++ b/libavfilter/af_amix.c @@ -389,6 +389,10 @@ static int request_frame(AVFilterLink *outlink) int ret; int wanted_samples, available_samples; + ret = calc_active_inputs(s); + if (ret < 0) + return ret; + if (s->input_state[0] == INPUT_OFF) { ret = request_samples(ctx, 1); if (ret < 0) @@ -419,15 +423,16 @@ static int request_frame(AVFilterLink *outlink) av_assert0(s->frame_list->nb_frames > 0); wanted_samples = frame_list_next_frame_size(s->frame_list); - ret = request_samples(ctx, wanted_samples); - if (ret < 0) - return ret; - - ret = calc_active_inputs(s); - if (ret < 0) - return ret; if (s->active_inputs > 1) { + ret = request_samples(ctx, wanted_samples); + if (ret < 0) + return ret; + + ret = calc_active_inputs(s); + if (ret < 0) + return ret; + available_samples = get_available_samples(s); if (!available_samples) return 0;

1 0

vp8: move block coeff arithcoder on stack.
by git＠videolan.org 31 May '12

31 May '12

ffmpeg | branch: master | Ronald S. Bultje <rsbultje(a)gmail.com> | Tue May 29 10:56:15 2012 -0700| [6163d880c0ac8b84b45d1f7a94719c5a0a6b5cb9] | committer: Ronald S. Bultje vp8: move block coeff arithcoder on stack. This prevents gcc from assuming that contents of it may have changed between calls to vp56_range_get_prob(), thus preventing countless (and unnecessary) movs. Decoding of sintel trailer goes from (avg+SG) 9.796 +/- 0.003 to 9.635 +/- 0.010. > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=6163d880c0ac8b84b… --- libavcodec/vp8.c | 38 ++++++++++++++++++++------------------ 1 file changed, 20 insertions(+), 18 deletions(-) diff --git a/libavcodec/vp8.c b/libavcodec/vp8.c index d9fdd4d..94200f6 100644 --- a/libavcodec/vp8.c +++ b/libavcodec/vp8.c @@ -707,56 +707,58 @@ void decode_mb_mode(VP8Context *s, VP8Macroblock *mb, int mb_x, int mb_y, uint8_ * @return 0 if no coeffs were decoded * otherwise, the index of the last coeff decoded plus one */ -static int decode_block_coeffs_internal(VP56RangeCoder *c, DCTELEM block[16], +static int decode_block_coeffs_internal(VP56RangeCoder *r, DCTELEM block[16], uint8_t probs[16][3][NUM_DCT_TOKENS-1], int i, uint8_t *token_prob, int16_t qmul[2]) { + VP56RangeCoder c = *r; goto skip_eob; do { int coeff; - if (!vp56_rac_get_prob_branchy(c, token_prob[0])) // DCT_EOB - return i; + if (!vp56_rac_get_prob_branchy(&c, token_prob[0])) // DCT_EOB + break; skip_eob: - if (!vp56_rac_get_prob_branchy(c, token_prob[1])) { // DCT_0 + if (!vp56_rac_get_prob_branchy(&c, token_prob[1])) { // DCT_0 if (++i == 16) - return i; // invalid input; blocks should end with EOB + break; // invalid input; blocks should end with EOB token_prob = probs[i][0]; goto skip_eob; } - if (!vp56_rac_get_prob_branchy(c, token_prob[2])) { // DCT_1 + if (!vp56_rac_get_prob_branchy(&c, token_prob[2])) { // DCT_1 coeff = 1; token_prob = probs[i+1][1]; } else { - if (!vp56_rac_get_prob_branchy(c, token_prob[3])) { // DCT 2,3,4 - coeff = vp56_rac_get_prob_branchy(c, token_prob[4]); + if (!vp56_rac_get_prob_branchy(&c, token_prob[3])) { // DCT 2,3,4 + coeff = vp56_rac_get_prob_branchy(&c, token_prob[4]); if (coeff) - coeff += vp56_rac_get_prob(c, token_prob[5]); + coeff += vp56_rac_get_prob(&c, token_prob[5]); coeff += 2; } else { // DCT_CAT* - if (!vp56_rac_get_prob_branchy(c, token_prob[6])) { - if (!vp56_rac_get_prob_branchy(c, token_prob[7])) { // DCT_CAT1 - coeff = 5 + vp56_rac_get_prob(c, vp8_dct_cat1_prob[0]); + if (!vp56_rac_get_prob_branchy(&c, token_prob[6])) { + if (!vp56_rac_get_prob_branchy(&c, token_prob[7])) { // DCT_CAT1 + coeff = 5 + vp56_rac_get_prob(&c, vp8_dct_cat1_prob[0]); } else { // DCT_CAT2 coeff = 7; - coeff += vp56_rac_get_prob(c, vp8_dct_cat2_prob[0]) << 1; - coeff += vp56_rac_get_prob(c, vp8_dct_cat2_prob[1]); + coeff += vp56_rac_get_prob(&c, vp8_dct_cat2_prob[0]) << 1; + coeff += vp56_rac_get_prob(&c, vp8_dct_cat2_prob[1]); } } else { // DCT_CAT3 and up - int a = vp56_rac_get_prob(c, token_prob[8]); - int b = vp56_rac_get_prob(c, token_prob[9+a]); + int a = vp56_rac_get_prob(&c, token_prob[8]); + int b = vp56_rac_get_prob(&c, token_prob[9+a]); int cat = (a<<1) + b; coeff = 3 + (8<<cat); - coeff += vp8_rac_get_coeff(c, ff_vp8_dct_cat_prob[cat]); + coeff += vp8_rac_get_coeff(&c, ff_vp8_dct_cat_prob[cat]); } } token_prob = probs[i+1][2]; } - block[zigzag_scan[i]] = (vp8_rac_get(c) ? -coeff : coeff) * qmul[!!i]; + block[zigzag_scan[i]] = (vp8_rac_get(&c) ? -coeff : coeff) * qmul[!!i]; } while (++i < 16); + *r = c; return i; } #endif

1 0

mp3/ac3 probe: search for PES headers to prevent probing MPEG-PS as MP3.
by git＠videolan.org 31 May '12

31 May '12

ffmpeg | branch: master | Ronald S. Bultje <rsbultje(a)gmail.com> | Wed May 2 15:45:18 2012 -0700| [64bde8056337bb656a11f3c9e2857c10b94e2871] | committer: Ronald S. Bultje mp3/ac3 probe: search for PES headers to prevent probing MPEG-PS as MP3. > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=64bde8056337bb656… --- libavformat/ac3dec.c | 30 +++++++++++++++++++++++++----- libavformat/mp3dec.c | 30 +++++++++++++++++++++++++----- 2 files changed, 50 insertions(+), 10 deletions(-) diff --git a/libavformat/ac3dec.c b/libavformat/ac3dec.c index 60c3cf9..c2e3301 100644 --- a/libavformat/ac3dec.c +++ b/libavformat/ac3dec.c @@ -57,11 +57,31 @@ static int ac3_eac3_probe(AVProbeData *p, enum CodecID expected_codec_id) if(codec_id != expected_codec_id) return 0; // keep this in sync with mp3 probe, both need to avoid // issues with MPEG-files! - if (first_frames>=4) return AVPROBE_SCORE_MAX/2+1; - else if(max_frames>500)return AVPROBE_SCORE_MAX/2; - else if(max_frames>=4) return AVPROBE_SCORE_MAX/4; - else if(max_frames>=1) return 1; - else return 0; + if (first_frames >= 4) return AVPROBE_SCORE_MAX / 2 + 1; + + if (max_frames) { + int pes = 0, i; + unsigned int code = -1; + +#define VIDEO_ID 0x000001e0 +#define AUDIO_ID 0x000001c0 + /* do a search for mpegps headers to be able to properly bias + * towards mpegps if we detect this stream as both. */ + for (i = 0; i<p->buf_size; i++) { + code = (code << 8) + p->buf[i]; + if ((code & 0xffffff00) == 0x100) { + if ((code & 0x1f0) == VIDEO_ID) pes++; + else if((code & 0x1e0) == AUDIO_ID) pes++; + } + } + + if (pes) + max_frames = (max_frames + pes - 1) / pes; + } + if (max_frames > 500) return AVPROBE_SCORE_MAX / 2; + else if (max_frames >= 4) return AVPROBE_SCORE_MAX / 4; + else if (max_frames >= 1) return 1; + else return 0; } #if CONFIG_AC3_DEMUXER diff --git a/libavformat/mp3dec.c b/libavformat/mp3dec.c index d233209..3987520 100644 --- a/libavformat/mp3dec.c +++ b/libavformat/mp3dec.c @@ -63,11 +63,31 @@ static int mp3_read_probe(AVProbeData *p) } // keep this in sync with ac3 probe, both need to avoid // issues with MPEG-files! - if (first_frames>=4) return AVPROBE_SCORE_MAX/2+1; - else if(max_frames>500)return AVPROBE_SCORE_MAX/2; - else if(max_frames>=4) return AVPROBE_SCORE_MAX/4; - else if(max_frames>=1) return 1; - else return 0; + if (first_frames >= 4) return AVPROBE_SCORE_MAX / 2 + 1; + + if (max_frames) { + int pes = 0, i; + unsigned int code = -1; + +#define VIDEO_ID 0x000001e0 +#define AUDIO_ID 0x000001c0 + /* do a search for mpegps headers to be able to properly bias + * towards mpegps if we detect this stream as both. */ + for (i = 0; i<p->buf_size; i++) { + code = (code << 8) + p->buf[i]; + if ((code & 0xffffff00) == 0x100) { + if ((code & 0x1f0) == VIDEO_ID) pes++; + else if((code & 0x1e0) == AUDIO_ID) pes++; + } + } + + if (pes) + max_frames = (max_frames + pes - 1) / pes; + } + if (max_frames > 500) return AVPROBE_SCORE_MAX / 2; + else if (max_frames >= 4) return AVPROBE_SCORE_MAX / 4; + else if (max_frames >= 1) return 1; + else return 0; //mpegps_mp3_unrecognized_format.mpg has max_frames=3 }

1 0

h264: move q0 scan tables into context
by git＠videolan.org 31 May '12

31 May '12

ffmpeg | branch: master | Michael Niedermayer <michaelni(a)gmx.at> | Thu May 31 20:19:56 2012 +0200| [b7d14883939e756cbda376c66552be9d843910a0] | committer: Michael Niedermayer h264: move q0 scan tables into context This fixes out of global array reads. The alternative solutions of checking the index or modifying the VLC tables to prevent the index going outside are each about 1-2 cpu cyclces slower per coded 4x4 block. The alternative of padding the global tables directly is more ugly and moving them to the context should benefit cache locality. Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=b7d14883939e756cb… --- libavcodec/h264.c | 24 ++++++++++++------------ libavcodec/h264.h | 12 ++++++------ libavcodec/h264data.h | 5 ----- 3 files changed, 18 insertions(+), 23 deletions(-) diff --git a/libavcodec/h264.c b/libavcodec/h264.c index 85787f4..7ebe352 100644 --- a/libavcodec/h264.c +++ b/libavcodec/h264.c @@ -2719,19 +2719,19 @@ static void init_scan_tables(H264Context *h) #undef T } if (h->sps.transform_bypass) { // FIXME same ugly - h->zigzag_scan_q0 = zigzag_scan; - h->zigzag_scan8x8_q0 = ff_zigzag_direct; - h->zigzag_scan8x8_cavlc_q0 = zigzag_scan8x8_cavlc; - h->field_scan_q0 = field_scan; - h->field_scan8x8_q0 = field_scan8x8; - h->field_scan8x8_cavlc_q0 = field_scan8x8_cavlc; + memcpy(h->zigzag_scan_q0 , zigzag_scan , sizeof(h->zigzag_scan_q0 )); + memcpy(h->zigzag_scan8x8_q0 , ff_zigzag_direct , sizeof(h->zigzag_scan8x8_q0 )); + memcpy(h->zigzag_scan8x8_cavlc_q0 , zigzag_scan8x8_cavlc , sizeof(h->zigzag_scan8x8_cavlc_q0)); + memcpy(h->field_scan_q0 , field_scan , sizeof(h->field_scan_q0 )); + memcpy(h->field_scan8x8_q0 , field_scan8x8 , sizeof(h->field_scan8x8_q0 )); + memcpy(h->field_scan8x8_cavlc_q0 , field_scan8x8_cavlc , sizeof(h->field_scan8x8_cavlc_q0 )); } else { - h->zigzag_scan_q0 = h->zigzag_scan; - h->zigzag_scan8x8_q0 = h->zigzag_scan8x8; - h->zigzag_scan8x8_cavlc_q0 = h->zigzag_scan8x8_cavlc; - h->field_scan_q0 = h->field_scan; - h->field_scan8x8_q0 = h->field_scan8x8; - h->field_scan8x8_cavlc_q0 = h->field_scan8x8_cavlc; + memcpy(h->zigzag_scan_q0 , h->zigzag_scan , sizeof(h->zigzag_scan_q0 )); + memcpy(h->zigzag_scan8x8_q0 , h->zigzag_scan8x8 , sizeof(h->zigzag_scan8x8_q0 )); + memcpy(h->zigzag_scan8x8_cavlc_q0 , h->zigzag_scan8x8_cavlc , sizeof(h->zigzag_scan8x8_cavlc_q0)); + memcpy(h->field_scan_q0 , h->field_scan , sizeof(h->field_scan_q0 )); + memcpy(h->field_scan8x8_q0 , h->field_scan8x8 , sizeof(h->field_scan8x8_q0 )); + memcpy(h->field_scan8x8_cavlc_q0 , h->field_scan8x8_cavlc , sizeof(h->field_scan8x8_cavlc_q0 )); } } diff --git a/libavcodec/h264.h b/libavcodec/h264.h index 0e11e30..a27f82f 100644 --- a/libavcodec/h264.h +++ b/libavcodec/h264.h @@ -421,12 +421,12 @@ typedef struct H264Context { uint8_t field_scan[16]; uint8_t field_scan8x8[64]; uint8_t field_scan8x8_cavlc[64]; - const uint8_t *zigzag_scan_q0; - const uint8_t *zigzag_scan8x8_q0; - const uint8_t *zigzag_scan8x8_cavlc_q0; - const uint8_t *field_scan_q0; - const uint8_t *field_scan8x8_q0; - const uint8_t *field_scan8x8_cavlc_q0; + uint8_t zigzag_scan_q0[16]; + uint8_t zigzag_scan8x8_q0[64]; + uint8_t zigzag_scan8x8_cavlc_q0[64]; + uint8_t field_scan_q0[16]; + uint8_t field_scan8x8_q0[64]; + uint8_t field_scan8x8_cavlc_q0[64]; int x264_build; diff --git a/libavcodec/h264data.h b/libavcodec/h264data.h index b665bca..f3fc7f9 100644 --- a/libavcodec/h264data.h +++ b/libavcodec/h264data.h @@ -57,7 +57,6 @@ static const uint8_t zigzag_scan[16+1] = { 1 + 1 * 4, 2 + 0 * 4, 3 + 0 * 4, 2 + 1 * 4, 1 + 2 * 4, 0 + 3 * 4, 1 + 3 * 4, 2 + 2 * 4, 3 + 1 * 4, 3 + 2 * 4, 2 + 3 * 4, 3 + 3 * 4, - 0, }; static const uint8_t field_scan[16+1] = { @@ -65,7 +64,6 @@ static const uint8_t field_scan[16+1] = { 0 + 3 * 4, 1 + 1 * 4, 1 + 2 * 4, 1 + 3 * 4, 2 + 0 * 4, 2 + 1 * 4, 2 + 2 * 4, 2 + 3 * 4, 3 + 0 * 4, 3 + 1 * 4, 3 + 2 * 4, 3 + 3 * 4, - 0, }; static const uint8_t luma_dc_zigzag_scan[16] = { @@ -112,7 +110,6 @@ static const uint8_t zigzag_scan8x8_cavlc[64+1] = { 1 + 4 * 8, 2 + 4 * 8, 6 + 0 * 8, 4 + 3 * 8, 0 + 7 * 8, 4 + 4 * 8, 7 + 2 * 8, 3 + 6 * 8, 5 + 5 * 8, 6 + 5 * 8, 6 + 6 * 8, 7 + 7 * 8, - 0, }; static const uint8_t field_scan8x8[64+1] = { @@ -132,7 +129,6 @@ static const uint8_t field_scan8x8[64+1] = { 7 + 0 * 8, 7 + 1 * 8, 6 + 4 * 8, 6 + 5 * 8, 6 + 6 * 8, 6 + 7 * 8, 7 + 2 * 8, 7 + 3 * 8, 7 + 4 * 8, 7 + 5 * 8, 7 + 6 * 8, 7 + 7 * 8, - 0, }; static const uint8_t field_scan8x8_cavlc[64+1] = { @@ -152,7 +148,6 @@ static const uint8_t field_scan8x8_cavlc[64+1] = { 1 + 7 * 8, 3 + 2 * 8, 2 + 7 * 8, 4 + 2 * 8, 3 + 7 * 8, 5 + 2 * 8, 4 + 7 * 8, 5 + 4 * 8, 6 + 3 * 8, 6 + 5 * 8, 7 + 3 * 8, 7 + 7 * 8, - 0, }; typedef struct IMbInfo {

1 0

oggparsevorbis: fix null ptr dereference
by git＠videolan.org 31 May '12

31 May '12

ffmpeg | branch: master | Michael Niedermayer <michaelni(a)gmx.at> | Thu May 31 19:06:14 2012 +0200| [18b46a494ef3592d69f4638dac1ebb613c0f548d] | committer: Michael Niedermayer oggparsevorbis: fix null ptr dereference Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=18b46a494ef3592d6… --- libavformat/oggparsevorbis.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/libavformat/oggparsevorbis.c b/libavformat/oggparsevorbis.c index 78defd7..74f4499 100644 --- a/libavformat/oggparsevorbis.c +++ b/libavformat/oggparsevorbis.c @@ -203,12 +203,12 @@ vorbis_header (AVFormatContext * s, int idx) int pkt_type = os->buf[os->pstart]; if (!(pkt_type & 1)) - return 0; + return os->private ? 0 : -1; if (!os->private) { os->private = av_mallocz(sizeof(struct oggvorbis_private)); if (!os->private) - return 0; + return -1; } if (os->psize < 1 || pkt_type > 5)

1 0

mpeg4videodec: Check that cplx_estimation_* fits in the available space
by git＠videolan.org 31 May '12

31 May '12

ffmpeg | branch: master | Michael Niedermayer <michaelni(a)gmx.at> | Thu May 31 18:54:00 2012 +0200| [b02cc2ddc610cd84bbee5923a642a8324988b28c] | committer: Michael Niedermayer mpeg4videodec: Check that cplx_estimation_* fits in the available space Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind Signed-off-by: Michael Niedermayer <michaelni(a)gmx.at> > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=b02cc2ddc610cd84b… --- libavcodec/mpeg4videodec.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/libavcodec/mpeg4videodec.c b/libavcodec/mpeg4videodec.c index bd1910b..263ac99 100644 --- a/libavcodec/mpeg4videodec.c +++ b/libavcodec/mpeg4videodec.c @@ -2050,6 +2050,10 @@ static int decode_vop_header(MpegEncContext *s, GetBitContext *gb){ if(s->pict_type == AV_PICTURE_TYPE_B) skip_bits_long(gb, s->cplx_estimation_trash_b); + if(get_bits_left(gb) < 3) { + av_log(s->avctx, AV_LOG_ERROR, "Header truncated\n"); + return -1; + } s->intra_dc_threshold= ff_mpeg4_dc_threshold[ get_bits(gb, 3) ]; if(!s->progressive_sequence){ s->top_field_first= get_bits1(gb);

1 0

vf_overlay: rewrite request/push logic.
by git＠videolan.org 31 May '12

31 May '12

ffmpeg | branch: master | Nicolas George <nicolas.george(a)normalesup.org> | Tue Apr 24 15:45:08 2012 +0200| [06bf6d3bc04979bd39ecdc7311d0daf8aee7e10f] | committer: Nicolas George vf_overlay: rewrite request/push logic. The old implementation, upon receiving a frame on the main input, would request an overlay frame if necessary. This could generate an infinite recursion, off-by-one frame mismatch and other harmful effects, especially it the split filter is present upgraph. The new implementation uses the linear approach: it uses two buffer queues for frames received out of turn and forwards request_frame calls to the input where a frame is necessary. > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=06bf6d3bc04979bd3… --- libavfilter/vf_overlay.c | 225 +++++++++++++++++++++++++++++++--------------- 1 file changed, 154 insertions(+), 71 deletions(-) diff --git a/libavfilter/vf_overlay.c b/libavfilter/vf_overlay.c index 8227d35..65712a8 100644 --- a/libavfilter/vf_overlay.c +++ b/libavfilter/vf_overlay.c @@ -36,6 +36,7 @@ #include "libavutil/mathematics.h" #include "libavutil/timestamp.h" #include "internal.h" +#include "bufferqueue.h" #include "drawutils.h" static const char *const var_names[] = { @@ -71,6 +72,8 @@ typedef struct { int x, y; ///< position of overlayed picture int allow_packed_rgb; + uint8_t frame_requested; + uint8_t overlay_eof; uint8_t main_is_packed_rgb; uint8_t main_rgba_map[4]; uint8_t main_has_alpha; @@ -78,7 +81,9 @@ typedef struct { uint8_t overlay_rgba_map[4]; uint8_t overlay_has_alpha; - AVFilterBufferRef *overpicref, *overpicref_next; + AVFilterBufferRef *overpicref; + struct FFBufQueue queue_main; + struct FFBufQueue queue_over; int main_pix_step[4]; ///< steps per pixel for each plane of the main output int overlay_pix_step[4]; ///< steps per pixel for each plane of the overlay @@ -149,8 +154,8 @@ static av_cold void uninit(AVFilterContext *ctx) if (over->overpicref) avfilter_unref_buffer(over->overpicref); - if (over->overpicref_next) - avfilter_unref_buffer(over->overpicref_next); + ff_bufqueue_discard_all(&over->queue_main); + ff_bufqueue_discard_all(&over->queue_over); } static int query_formats(AVFilterContext *ctx) @@ -304,51 +309,6 @@ static AVFilterBufferRef *get_video_buffer(AVFilterLink *link, int perms, int w, return avfilter_get_video_buffer(link->dst->outputs[0], perms, w, h); } -static void start_frame(AVFilterLink *inlink, AVFilterBufferRef *inpicref) -{ - AVFilterBufferRef *outpicref = avfilter_ref_buffer(inpicref, ~0); - AVFilterContext *ctx = inlink->dst; - OverlayContext *over = ctx->priv; - av_unused AVFilterLink *outlink = ctx->outputs[0]; - - inlink->dst->outputs[0]->out_buf = outpicref; - outpicref->pts = av_rescale_q(outpicref->pts, ctx->inputs[MAIN]->time_base, - ctx->outputs[0]->time_base); - - if (!over->overpicref || over->overpicref->pts < outpicref->pts) { - if (!over->overpicref_next) - avfilter_request_frame(ctx->inputs[OVERLAY]); - - if (over->overpicref && over->overpicref_next && - over->overpicref_next->pts <= outpicref->pts) { - avfilter_unref_buffer(over->overpicref); - over->overpicref = over->overpicref_next; - over->overpicref_next = NULL; - } - } - - av_dlog(ctx, "main_pts:%s main_pts_time:%s", - av_ts2str(outpicref->pts), av_ts2timestr(outpicref->pts, &outlink->time_base)); - if (over->overpicref) - av_dlog(ctx, " over_pts:%s over_pts_time:%s", - av_ts2str(over->overpicref->pts), av_ts2timestr(over->overpicref->pts, &outlink->time_base)); - av_dlog(ctx, "\n"); - - avfilter_start_frame(inlink->dst->outputs[0], outpicref); -} - -static void start_frame_overlay(AVFilterLink *inlink, AVFilterBufferRef *inpicref) -{ - AVFilterContext *ctx = inlink->dst; - OverlayContext *over = ctx->priv; - - inpicref->pts = av_rescale_q(inpicref->pts, ctx->inputs[OVERLAY]->time_base, - ctx->outputs[0]->time_base); - - if (!over->overpicref) over->overpicref = inpicref; - else over->overpicref_next = inpicref; -} - // divide by 255 and round to nearest // apply a fast variant: (X+127)/255 = ((X+127)*257+257)>>16 = ((X+128)*257)>>16 #define FAST_DIV255(x) ((((x) + 128) * 257) >> 16) @@ -483,16 +443,100 @@ static void blend_slice(AVFilterContext *ctx, } } -static void draw_slice(AVFilterLink *inlink, int y, int h, int slice_dir) +static int try_start_frame(AVFilterContext *ctx, AVFilterBufferRef *mainpic) { - AVFilterContext *ctx = inlink->dst; + OverlayContext *over = ctx->priv; + AVFilterLink *outlink = ctx->outputs[0]; + AVFilterBufferRef *next_overpic, *outpicref; + + /* Discard obsolete overlay frames: if there is a next frame with pts is + * before the main frame, we can drop the current overlay. */ + while (1) { + next_overpic = ff_bufqueue_peek(&over->queue_over, 0); + if (!next_overpic || next_overpic->pts > mainpic->pts) + break; + ff_bufqueue_get(&over->queue_over); + avfilter_unref_buffer(over->overpicref); + over->overpicref = next_overpic; + } + /* If there is no next frame and no EOF and the overlay frame is before + * the main frame, we can not know yet if it will be superseded. */ + if (!over->queue_over.available && !over->overlay_eof && + (!over->overpicref || over->overpicref->pts < mainpic->pts)) + return AVERROR(EAGAIN); + /* At this point, we know that the current overlay frame extends to the + * time of the main frame. */ + outlink->out_buf = outpicref = avfilter_ref_buffer(mainpic, ~0); + + av_dlog(ctx, "main_pts:%s main_pts_time:%s", + av_ts2str(outpicref->pts), av_ts2timestr(outpicref->pts, &outlink->time_base)); + if (over->overpicref) + av_dlog(ctx, " over_pts:%s over_pts_time:%s", + av_ts2str(over->overpicref->pts), av_ts2timestr(over->overpicref->pts, &outlink->time_base)); + av_dlog(ctx, "\n"); + + avfilter_start_frame(ctx->outputs[0], avfilter_ref_buffer(outpicref, ~0)); + over->frame_requested = 0; + return 0; +} + +static int try_start_next_frame(AVFilterContext *ctx) +{ + OverlayContext *over = ctx->priv; + AVFilterBufferRef *next_mainpic = ff_bufqueue_peek(&over->queue_main, 0); + if (!next_mainpic || try_start_frame(ctx, next_mainpic) < 0) + return AVERROR(EAGAIN); + avfilter_unref_buffer(ff_bufqueue_get(&over->queue_main)); + return 0; +} + +static int try_push_frame(AVFilterContext *ctx) +{ + OverlayContext *over = ctx->priv; AVFilterLink *outlink = ctx->outputs[0]; AVFilterBufferRef *outpicref = outlink->out_buf; + + if (try_start_next_frame(ctx) < 0) + return AVERROR(EAGAIN); + outpicref = outlink->out_buf; + if (over->overpicref) + blend_slice(ctx, outpicref, over->overpicref, over->x, over->y, + over->overpicref->video->w, over->overpicref->video->h, + 0, outpicref->video->w, outpicref->video->h); + avfilter_draw_slice(outlink, 0, outpicref->video->h, +1); + avfilter_unref_bufferp(&outlink->out_buf); + avfilter_end_frame(outlink); + return 0; +} + +static void flush_frames(AVFilterContext *ctx) +{ + while (!try_push_frame(ctx)); +} + +static void start_frame_main(AVFilterLink *inlink, AVFilterBufferRef *inpicref) +{ + AVFilterContext *ctx = inlink->dst; + OverlayContext *over = ctx->priv; + + flush_frames(ctx); + inpicref->pts = av_rescale_q(inpicref->pts, ctx->inputs[MAIN]->time_base, + ctx->outputs[0]->time_base); + if (try_start_frame(ctx, inpicref) < 0) + ff_bufqueue_add(ctx, &over->queue_main, inpicref); +} + +static void draw_slice_main(AVFilterLink *inlink, int y, int h, int slice_dir) +{ + AVFilterContext *ctx = inlink->dst; OverlayContext *over = ctx->priv; + AVFilterLink *outlink = ctx->outputs[0]; + AVFilterBufferRef *outpicref = outlink->out_buf; + if (!outpicref) + return; if (over->overpicref && - !(over->x >= outpicref->video->w || over->y >= outpicref->video->h || - y+h < over->y || y >= over->y + over->overpicref->video->h)) { + y + h > over->y && y < over->y + over->overpicref->video->h) { blend_slice(ctx, outpicref, over->overpicref, over->x, over->y, over->overpicref->video->w, over->overpicref->video->h, y, outpicref->video->w, h); @@ -500,28 +544,67 @@ static void draw_slice(AVFilterLink *inlink, int y, int h, int slice_dir) avfilter_draw_slice(outlink, y, h, slice_dir); } -static void end_frame(AVFilterLink *inlink) +static void end_frame_main(AVFilterLink *inlink) { - avfilter_end_frame(inlink->dst->outputs[0]); - avfilter_unref_buffer(inlink->cur_buf); -} + AVFilterContext *ctx = inlink->dst; + AVFilterLink *outlink = ctx->outputs[0]; + AVFilterBufferRef *outpicref = outlink->out_buf; + flush_frames(ctx); -static void null_draw_slice(AVFilterLink *inlink, int y, int h, int slice_dir) { } + if (!outpicref) + return; + avfilter_unref_bufferp(&inlink->cur_buf); + avfilter_unref_bufferp(&outlink->out_buf); + avfilter_end_frame(ctx->outputs[0]); +} -static void null_end_frame(AVFilterLink *inlink) { } +static void start_frame_over(AVFilterLink *inlink, AVFilterBufferRef *inpicref) +{ +} -static int poll_frame(AVFilterLink *link) +static void end_frame_over(AVFilterLink *inlink) { - AVFilterContext *s = link->src; - OverlayContext *over = s->priv; - int ret = avfilter_poll_frame(s->inputs[OVERLAY]); + AVFilterContext *ctx = inlink->dst; + OverlayContext *over = ctx->priv; + AVFilterBufferRef *inpicref = inlink->cur_buf; - if (ret == AVERROR_EOF) - ret = !!over->overpicref; + flush_frames(ctx); + inpicref->pts = av_rescale_q(inpicref->pts, ctx->inputs[OVERLAY]->time_base, + ctx->outputs[0]->time_base); + ff_bufqueue_add(ctx, &over->queue_over, inpicref); + try_push_frame(ctx); +} - return ret && avfilter_poll_frame(s->inputs[MAIN]); +static int request_frame(AVFilterLink *outlink) +{ + AVFilterContext *ctx = outlink->src; + OverlayContext *over = ctx->priv; + int input, ret; + + if (!try_push_frame(ctx)) + return 0; + over->frame_requested = 1; + while (over->frame_requested) { + /* TODO if we had a frame duration, we could guess more accurately */ + input = !over->overlay_eof && (over->queue_main.available || + over->queue_over.available < 2) ? + OVERLAY : MAIN; + ret = avfilter_request_frame(ctx->inputs[input]); + /* EOF on main is reported immediately */ + if (ret == AVERROR_EOF && input == OVERLAY) { + over->overlay_eof = 1; + if (!try_start_next_frame(ctx)) + return 0; + ret = 0; /* continue requesting frames on main */ + } + if (ret < 0) + return ret; + } + return 0; } +static void null_draw_slice(AVFilterLink *inlink, int y, int h, int slice_dir) { } + AVFilter avfilter_vf_overlay = { .name = "overlay", .description = NULL_IF_CONFIG_SMALL("Overlay a video source on top of the input."), @@ -535,25 +618,25 @@ AVFilter avfilter_vf_overlay = { .inputs = (const AVFilterPad[]) {{ .name = "main", .type = AVMEDIA_TYPE_VIDEO, - .start_frame = start_frame, .get_video_buffer= get_video_buffer, .config_props = config_input_main, - .draw_slice = draw_slice, - .end_frame = end_frame, + .start_frame = start_frame_main, + .draw_slice = draw_slice_main, + .end_frame = end_frame_main, .min_perms = AV_PERM_READ, .rej_perms = AV_PERM_REUSE2|AV_PERM_PRESERVE, }, { .name = "overlay", .type = AVMEDIA_TYPE_VIDEO, - .start_frame = start_frame_overlay, .config_props = config_input_overlay, + .start_frame = start_frame_over, .draw_slice = null_draw_slice, - .end_frame = null_end_frame, + .end_frame = end_frame_over, .min_perms = AV_PERM_READ, .rej_perms = AV_PERM_REUSE2, }, { .name = NULL}}, .outputs = (const AVFilterPad[]) {{ .name = "default", .type = AVMEDIA_TYPE_VIDEO, .config_props = config_output, - .poll_frame = poll_frame }, + .request_frame = request_frame, }, { .name = NULL}}, };

1 0

lavfi: add a generic API for buffer queues.
by git＠videolan.org 31 May '12

31 May '12

ffmpeg | branch: master | Nicolas George <nicolas.george(a)normalesup.org> | Wed Apr 25 00:31:22 2012 +0200| [6d679c435990281e69a74075eeaba499a35f4a80] | committer: Nicolas George lavfi: add a generic API for buffer queues. > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=6d679c435990281e6… --- libavfilter/bufferqueue.h | 111 +++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 111 insertions(+) diff --git a/libavfilter/bufferqueue.h b/libavfilter/bufferqueue.h new file mode 100644 index 0000000..a27fb86 --- /dev/null +++ b/libavfilter/bufferqueue.h @@ -0,0 +1,111 @@ +/* + * Generic buffer queue + * Copyright (c) 2012 Nicolas George + * + * 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 + */ + +#ifndef AVFILTER_BUFFERQUEUE_H +#define AVFILTER_BUFFERQUEUE_H + +/** + * FFBufQueue: simple AVFilterBufferRef queue API + * + * Note: this API is not thread-safe. Concurrent access to the same queue + * must be protected by a mutex or any synchronization mechanism. + */ + +/** + * Maximum size of the queue. + * + * This value can be overridden by definying it before including this + * header. + * Powers of 2 are recommended. + */ +#ifndef FF_BUFQUEUE_SIZE +#define FF_BUFQUEUE_SIZE 32 +#endif + +#include "avfilter.h" +#include "libavutil/avassert.h" + +/** + * Structure holding the queue + */ +struct FFBufQueue { + AVFilterBufferRef *queue[FF_BUFQUEUE_SIZE]; + unsigned short head; + unsigned short available; /**< number of available buffers */ +}; + +#define BUCKET(i) queue->queue[(queue->head + (i)) % FF_BUFQUEUE_SIZE] + +/** + * Add a buffer to the queue. + * + * If the queue is already full, then the current last buffer is dropped + * (and unrefed) with a warning before adding the new buffer. + */ +static inline void ff_bufqueue_add(void *log, struct FFBufQueue *queue, + AVFilterBufferRef *buf) +{ + if (queue->available == FF_BUFQUEUE_SIZE) { + av_log(log, AV_LOG_WARNING, "Buffer queue overflow, dropping.\n"); + avfilter_unref_buffer(BUCKET(--queue->available)); + } + BUCKET(queue->available++) = buf; +} + +/** + * Get a buffer from the queue without altering it. + * + * Buffer with index 0 is the first buffer in the queue. + * Return NULL if the queue has not enough buffers. + */ +static inline AVFilterBufferRef *ff_bufqueue_peek(struct FFBufQueue *queue, + unsigned index) +{ + return index < queue->available ? BUCKET(index) : NULL; +} + +/** + * Get the first buffer from the queue and remove it. + * + * Do not use on an empty queue. + */ +static inline AVFilterBufferRef *ff_bufqueue_get(struct FFBufQueue *queue) +{ + AVFilterBufferRef *ret = queue->queue[queue->head]; + av_assert0(queue->available); + queue->available--; + queue->queue[queue->head] = NULL; + queue->head = (queue->head + 1) % FF_BUFQUEUE_SIZE; + return ret; +} + +/** + * Unref and remove all buffers from the queue. + */ +static inline void ff_bufqueue_discard_all(struct FFBufQueue *queue) +{ + while (queue->available) + avfilter_unref_buffer(ff_bufqueue_get(queue)); +} + +#undef BUCKET + +#endif /* AVFILTER_BUFFERQUEUE_H */

1 0

sink_buffer: guard against NULL frames.
by git＠videolan.org 31 May '12

31 May '12

ffmpeg | branch: master | Nicolas George <nicolas.george(a)normalesup.org> | Tue Apr 24 12:51:37 2012 +0200| [72f3786d0f0897ed0fc3356f4fec167219de54a4] | committer: Nicolas George sink_buffer: guard against NULL frames. This only happens if some filter somewhere is bogus, but it helps debugging. > http://git.videolan.org/gitweb.cgi/ffmpeg.git/?a=commit;h=72f3786d0f0897ed0… --- libavfilter/sink_buffer.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/libavfilter/sink_buffer.c b/libavfilter/sink_buffer.c index 3eb66ca..d83f21b 100644 --- a/libavfilter/sink_buffer.c +++ b/libavfilter/sink_buffer.c @@ -23,6 +23,7 @@ * buffer video sink */ +#include "libavutil/avassert.h" #include "libavutil/fifo.h" #include "avfilter.h" #include "buffersink.h" @@ -98,6 +99,7 @@ static void end_frame(AVFilterLink *inlink) AVFilterContext *ctx = inlink->dst; BufferSinkContext *buf = inlink->dst->priv; + av_assert1(inlink->cur_buf); if (av_fifo_space(buf->fifo) < sizeof(AVFilterBufferRef *)) { /* realloc fifo size */ if (av_fifo_realloc2(buf->fifo, av_fifo_size(buf->fifo) * 2) < 0) {

1 0