The number of threads impacts the number of surfaces that should be
allocated by the HW acceleration backend. If not enough surfaces are
allocated, the VAAPI and DXVA modules returned the oldest surface
currently in used. This technique can cause many visual glitches with
some samples. The number of allocated surfaces now scales with the
number of threads used by avcodec.

Fix #9887
Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>

Signed-off-by: Felix Paul Kühne <fkuehne@videolan.org>

(see also fccfbd2f)

This should perhaps be const... but libavcodec seems to require
non-const even when const ought to be enough :-(

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

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

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

Close #10832

We don't do any actual call to gettext here, so the warning
is not really needed.

dequeueOutputBuffer should have an higher priority since releasing an
output buffer often allows MediaCodec to make progress

Most of them are memleaks, uninitialized values, unused code
and type mismatches (float / int vs. CGFloat, uint vs. int)

close #10761

Copy LibVLC headers from the installation directory: $(prefix)/include
rather than from the source directory, where the version file is still
unprocessed.
Signed-off-by: Felix Paul Kühne <fkuehne@videolan.org>

Ref #6649

Close #9833

Ref #9833

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

Increase Volume / Decrease Volume / Mute / Speed->Faster(fine) / Speed->Slower(fine)

Used Existing resources, nothing more added.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

This patch removes a calculation which seems not to be required.
It also "fixes" potential issues when i_visible_width/height != i_width/height.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Remove extra slash when computing relative uri with an absolute path.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

This reverts commit 4ec3dd7b.

This broke windows packaging.

Signed-off-by: Adrien Maglo <magsoft@videolan.org>

See CVE-2014-0092

Signed-off-by: Felix Paul Kühne <fkuehne@videolan.org>

src/missing: remove return after assert case since this is only needed for non-debug builds which are not used by developers anyway

So far only network protocols (well, only RTMP) are aliased.

This works more or less like SNDVOL (and like PulseAudio with flat
volumes not disabled): the VLC volume represents the product of the
endpoint volume by the simple volume.