This is required when encoding audio in libav 0.8, otherwise
it will crash due to a division by zero.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Should fix #5919
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

This sets HAVE_FPU correctly on ARM targets if VFP instructions are
enabled. Note that __VFP_FP__ implies the use of the VFP floating point
format, not the availability of the VFP instructions. Thus software
floating emulation __SOFTFP__ (i.e. gcc -mfloat-abi=soft) is excluded
explicitly.

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

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

If we are only targeting a single architecture, then install the
architecture specific binary files instead of creating a fat binary
with a single architecture.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

When preparing the framework, guarantee that the expected
architectures are present. If they are not present, or if an
unexpected architecture is present, then reconstruct the universal
binary.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

When referring to the main_build_dir don't assume that the x86_64
subdirectory is a part of the ARCHS that we are building. Grab the
first arch in the list of ARCHS and use that as the main_build_dir.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Remove trailing backslash from the prefix variable. Other locations
in the code adds the required backslash.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

VLC_SRC_DIR is already defined in Xcode. Replace top_srcdir in favor
of VLC_SRC_DIR.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Make sure that we return the exit code when vlc/configure fails. This
forces the the xcodebuild to fail.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Follow the same style used earlier in the code and declare local
variables with the 'local' directive.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

- Remove blank lines
- Align comments
- Update comment used to partition end of vlc_install_object
function.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

- Remove white space and debug statement.
- Fix typo in comments
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

There is a dump demux for that.

The iomx module is only enabled when the user has requested it
(via settings). Currently, when the iomx module has priority 80,
it is always preferred over avcodec.

For other platforms, where one would only enable omxil if one
explicitly wants to use it, this is ok, but for android, we
ship the iomx module but allow users to enable it at runtime
if they want to.

If it isn't enabled, the priorities are currently overridden by
adding "--codec avcodec,all" to the command line, but this makes
avcodec have precedence over other decoder modules that otherwise
would have higher priority (e.g. a52).

By setting the priority to 0 for the iomx version of this
module, we can more easily enable it only when asked for,
by adding "--codec iomx,all", without affecting the relative
priorities of all other codec modules.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

This reverts commit e50cfac2.
This undid modifications made in f83d51d4

Don't close a NULL FILE* on android

This reverts commit 1a1bba5a.

Conflicts:

	include/vlc_fixups.h

don't make it point to a wrong location

Give the real host to configure

It would replace the 'ar' of 'arm-....'
Instead use "LIBRARY_LINK =.*ar" to be sure we replace the correct tool

On some devices (apparently froyo and earlier), some OMX core
(with sw-only codecs) already is loaded into the process, and
dlsym(RTLD_DEFAULT) can just as well return functions from that
one instead of the ones from the iomx wrapper that we've loaded.

This makes sure we really get the functions we want.
Signed-off-by: Rafaël Carré <funman@videolan.org>

On Linux, this simplifies the code and improves relocability.
On Windows, OS/2 and MacOS, this should fix a small race condition.

Signed-off-by: Rafaël Carré <funman@videolan.org>

This reverts commit 3bbc2f54.
Now it exists again

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

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

Using hls_Get and segment_GetSegment instead. Those check for out of
bound access.
Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>