GitLab

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Technically it could be set to 0
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

We require libavformat >= 57 now.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

The entries within p_sys->attachments are of course allocated with
vlc_input_attachment_New, meaning that they shall be released using
the corresponding vlc_input_attachment_Delete.

fixes: #18288
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

we might want to override timestamps, not just
deal with video pts

Some parameters are not correctly updated
(ex: AAC rate/channels) when reading from
deprecated AVContext

av_read_frame expects 0 to be returned on end-of-file, and negative
values are reserved for fatal stream-errors. The previous
implementation would return -1 upon EOF (vlc_stream_Read returning 0),
causing premature EOF from modules/demux/avformat/demux.c:Demux.

These changes make sure that we honor the contract associated with the
read-callback, and that we only return -1 if there is a fatal error.
Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

In FFmpeg a call to avformat_write_header() then avio_flush() doesn't
necessarily write the data to the IO so our b_write_header detection never sees
anything.
We can rely on the new write system that tells what kind of data that is being
written.

This fixes multiple connections on the HTTP server when using the FFmpeg muxer.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

refs #17174
Same issue as raw HEVC, no timestamps output
so no PCR, and PCR can't be set by packetizer.

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

no pkt dts/pts, so no pcr.
Just demuxes at max cpu/speed.

&p_input->fmt == p_input->p_fmt
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

As mentioned by documentation
"It may be freed and replaced with a new buffer by libavformat."
" AVIOContext.buffer holds the buffer currently in use"

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

avformat needs nul padding after the probe data.