[FFmpeg-devel] [PATCH] avformat/mov: fix hang while seek on a kind of fragmented mp4.

C.H.Liu liuchh83 at gmail.com
Mon Feb 11 13:51:23 EET 2019


It is lucky that my patch works to ticket 7572. However, #7572 has a deeper
reason.

Mov demuxer consider that fragmented index is completed if a ‘sidx’ point
to the end of the file. But there may be other ‘sidx’ for other tracks. If
we skip the tail from there, we will missing the last ‘sidx’ and ‘moof’.
Then AV_NOPTS_VALUE occurs.

I tried to avoid skipping the tail simply. However, fate-test run failed on
mov-frag-encrypted.mp4.


BTW, I just got back from vacation. I am willing to follow an optimized
solution.


Thanks and Regards,

Chenhui



Carl Eugen Hoyos <ceffmpeg at gmail.com> 于2019年2月11日周一 上午6:24写道:

> 2019-02-10 23:04 GMT+01:00, Marton Balint <cus at passwd.hu>:
> >
> >
> > On Sun, 3 Feb 2019, Charles Liu wrote:
> >
> >> Binary searching would hang if the fragment items do NOT have timestamp
> >> for the specified stream.
> >>
> >> For example, a fmp4 consists of separated 'moof' boxes for each track,
> and
> >> separated 'sidx' for each segment, but no 'mfra' box.
> >> Then every fragment item only have the timestamp for one of its tracks.
> >>
> >> Signed-off-by: Charles Liu <liuchh83 at gmail.com>
> >> ---
> >> libavformat/mov.c | 21 ++++++++++++---------
> >> 1 file changed, 12 insertions(+), 9 deletions(-)
> >>
> >> diff --git a/libavformat/mov.c b/libavformat/mov.c
> >> index 9b9739f788..35cb619e9f 100644
> >> --- a/libavformat/mov.c
> >> +++ b/libavformat/mov.c
> >> @@ -1266,7 +1266,7 @@ static int64_t get_frag_time(MOVFragmentIndex
> >> *frag_index,
> >> static int search_frag_timestamp(MOVFragmentIndex *frag_index,
> >>                                  AVStream *st, int64_t timestamp)
> >> {
> >> -    int a, b, m;
> >> +    int a, b, m, m0;
> >>     int64_t frag_time;
> >>     int id = -1;
> >>
> >> @@ -1282,15 +1282,18 @@ static int
> search_frag_timestamp(MOVFragmentIndex
> >> *frag_index,
> >>     b = frag_index->nb_items;
> >>
> >>     while (b - a > 1) {
> >> -        m = (a + b) >> 1;
> >> -        frag_time = get_frag_time(frag_index, m, id);
> >> -        if (frag_time != AV_NOPTS_VALUE) {
> >> -            if (frag_time >= timestamp)
> >> -                b = m;
> >> -            if (frag_time <= timestamp)
> >> -                a = m;
> >> -        }
> >> +        m0 = m = (a + b) >> 1;
> >> +
> >> +        while (m < b &&
> >> +               (frag_time = get_frag_time(frag_index, m, id)) ==
> >> AV_NOPTS_VALUE)
> >> +            m++;
> >> +
> >> +        if (m < b && frag_time <= timestamp)
> >> +            a = m;
> >> +        else
> >> +            b = m0;
> >>     }
> >> +
> >>     return a;
> >> }
> >>
> >
> > As this fixes a hang, I will push this version soon.
>
> Please mention ticket #7572 in the commit message.
>
> Thank you, Carl Eugen
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>


More information about the ffmpeg-devel mailing list