As long as there is enough contiguous data in the avio buffer,
just return a pointer to it instead of copying it to the caller
provided buffer.
Signed-off-by: Martin Storsjö <martin@martin.st>

os_support.h contains more precise workarounds for non-POSIX mkdir().

The bitstream buffers are now private and freed by libavcodec. For
backward compatibility, the hold bitstream buffer pointer is left NULL
(applications were supposed to av_freep() it).
Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

It limits the duration of the data read from a given input.

Insert (a)trim filters on the corresponding inputs, so the extra frames
are decoded and discarded.

Using -ss 0 to drop frames with negative timestamps is a perfectly valid
use case.

This will allow using the same code when one filter can be inserted both
on input and output.

If any samples are still buffered when request_frame returns EOF, they
won't be returned currently.

Fixes an invalid free.

Bug-id: 448

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Incidentally also remove a warning.

Some code paths can call it with invalid length.

CC: libav-stable@libav.org

'filter' is too generic, since the functions in question deal with
several filters.

Also add const qualifiers to those variables.

This function should never modify the filter.

The actual usage of the pragma is within functions and older versions
of gcc only support the pragma outside of functions.

Set output_picture_number in H.264 parser.
Signed-off-by: Diego Biurrun <diego@biurrun.de>

Signed-off-by: Martin Storsjö <martin@martin.st>

Signed-off-by: Diego Biurrun <diego@biurrun.de>

Signed-off-by: Diego Biurrun <diego@biurrun.de>

Signed-off-by: Martin Storsjö <martin@martin.st>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

A separate rtcp port can already be set when opening the rtp
protocol normally, but when doing port setup as in RTSP (where
we first need to open the local ports and pass them to the peer,
and only then receive the remote peer port numbers), we didn't
check the same url parameter as in the normal open routine.
Signed-off-by: Martin Storsjö <martin@martin.st>

I doubt that anyone ever would try to send a 1 byte packet
via the RTP protocol, but check just in case - it shouldn't
crash at least.
Signed-off-by: Martin Storsjö <martin@martin.st>

Signed-off-by: Martin Storsjö <martin@martin.st>

This seems to have been missed in 636ced8e.
Signed-off-by: Martin Storsjö <martin@martin.st>