| Message ID | 20210812044354.485278-1-sehuww@mail.scut.edu.cn |
|---|---|
| State | Accepted |
| Commit | 2b77a2f723cd0af441db26dd5ee632af3040e7cb |
| Headers | show |
| Series | [FFmpeg-devel] movenc: Ensure no separate moof written for empty track | expand |
| Context | Check | Description |
|---|---|---|
| andriy/x86_make | success | Make finished |
| andriy/x86_make_fate | success | Make fate finished |
| andriy/PPC64_make | success | Make finished |
| andriy/PPC64_make_fate | success | Make fate finished |
On Thu, 12 Aug 2021, Hu Weiwen wrote: > track->mdat_buf can be not NULL while the track is still empty if the > last packet write failed. > > Signed-off-by: Hu Weiwen <sehuww@mail.scut.edu.cn> > --- > libavformat/movenc.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/libavformat/movenc.c b/libavformat/movenc.c > index bcc202300bb..a460cd9adae 100644 > --- a/libavformat/movenc.c > +++ b/libavformat/movenc.c > @@ -5422,7 +5422,7 @@ static int mov_flush_fragment(AVFormatContext *s, int force) > duration = track->start_dts + track->track_duration - > track->cluster[0].dts; > if (mov->flags & FF_MOV_FLAG_SEPARATE_MOOF) { > - if (!track->mdat_buf) > + if (!track->entry) > continue; > mdat_size = avio_tell(track->mdat_buf); > moof_tracks = i; > -- > 2.25.1 Ok I guess. If this situation happens (we allocate mdat_buf but fail to write a sample to it), I guess we can end up with some half-written packet at the start of it, which will be written as part of the next fragment. As long as byte offsets end up correct so the extra data at the start of the buffer is ignored I guess that'll work fine (I didn't check right now). // Martin
On Thu, 12 Aug 2021, Martin Storsjö wrote: > On Thu, 12 Aug 2021, Hu Weiwen wrote: > >> track->mdat_buf can be not NULL while the track is still empty if the >> last packet write failed. >> >> Signed-off-by: Hu Weiwen <sehuww@mail.scut.edu.cn> >> --- >> libavformat/movenc.c | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/libavformat/movenc.c b/libavformat/movenc.c >> index bcc202300bb..a460cd9adae 100644 >> --- a/libavformat/movenc.c >> +++ b/libavformat/movenc.c >> @@ -5422,7 +5422,7 @@ static int mov_flush_fragment(AVFormatContext *s, int >> force) >> duration = track->start_dts + track->track_duration - >> track->cluster[0].dts; >> if (mov->flags & FF_MOV_FLAG_SEPARATE_MOOF) { >> - if (!track->mdat_buf) >> + if (!track->entry) >> continue; >> mdat_size = avio_tell(track->mdat_buf); >> moof_tracks = i; >> -- >> 2.25.1 > > Ok I guess. Pushed this patch too. // Martin
diff --git a/libavformat/movenc.c b/libavformat/movenc.c index bcc202300bb..a460cd9adae 100644 --- a/libavformat/movenc.c +++ b/libavformat/movenc.c @@ -5422,7 +5422,7 @@ static int mov_flush_fragment(AVFormatContext *s, int force) duration = track->start_dts + track->track_duration - track->cluster[0].dts; if (mov->flags & FF_MOV_FLAG_SEPARATE_MOOF) { - if (!track->mdat_buf) + if (!track->entry) continue; mdat_size = avio_tell(track->mdat_buf); moof_tracks = i;
track->mdat_buf can be not NULL while the track is still empty if the last packet write failed. Signed-off-by: Hu Weiwen <sehuww@mail.scut.edu.cn> --- libavformat/movenc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)