When g_option_context_parse fails, context and error variables are not getting free'd
which results in memory leaks. Free'ing the same.

And replacing g_error_free with g_clear_error, which checks if the error being passed
is not NULL and sets the variable to NULL on free'ing.

https://bugzilla.gnome.org/show_bug.cgi?id=753851

When looking for exact matches in the sequence, this results
in much simpler code than when using g_sequence_search.

https://bugzilla.gnome.org/show_bug.cgi?id=755498

The '++' got incidentially added during the refactoring in
2fe3939c.

The default padding I introduced in d4f81fb4 is
actually only 4 pointers and on 32bit platforms already smaller than the union.
Replace it with a fixed 64byte padding. Don't add the normal padding for now.

Fixes #755822

Some OpenCV plugins use haar cascade files that are included in the
GStreamer sources. To be able to use these from uninstalled, they need
to be found through an environment variable.

Adding this environment variable pointing to haar cascade files to
gst-uninstalled.

Broke this when I removed the G_GNUC_PRINTF in a previous
commit to fix indentation, since it was not really needed.
Turns out unlike gcc clang warns though if a non-literal
format string is passed then. Fix indentation differently.

http://clang.llvm.org/docs/AttributeReference.html#format-gnu-format

While this technically is an abi break, we decided to do this:
1) the struct is documented to be internal
2) the struct is alloced and freed inside the library
3) there are no public methods that receive or return instances
4) the only code known to use this struct are classes containd here

When adding an element to a bin we need to propagate the GstContext's
to/from the element.

This moves the GstContext list from GstBin to GstElement and adds
convenience functions to get the currently set list of GstContext's.

This does not deal with the collection of GstContext's propagated
using GST_CONTEXT_QUERY.  Element subclasses are advised to call
gst_element_set_context if they need to propagate GstContext's
received from the context query.

https://bugzilla.gnome.org/show_bug.cgi?id=705579

This new mode won't overshoot the min/max y values set by the control-points.
Fixes #754678
API: GST_INTERPOLATION_MODE_CUBIC_MONO

Extract common code that looks up the control-points around the timestamp. Add
some comments for future investigation.

This is in preparation for new modes to be added. In particullar it demonstrates
how the cubic splines overshoot the range.

The design is to only do a local check

A proxy-pad should always proxy the caps related queries
and events to its down or upstream peers on the other side
of the element. Falling back to a caps query seems wrong.

https://bugzilla.gnome.org/show_bug.cgi?id=754112

gst_segment_to_position might cause confusion, especially with the addition of
gst_segment_position_from_stream_time . Deprecated gst_segment_to_position
now, and replaced it with gst_segment_position_from_running_time.

Also added unit tests.

Renamed the "result" variable to "stream_time" for better readability.

gst_segment_position_from_stream_time() will convert stream time into a
position in the segment so that gst_segment_to_stream_time() with that
position returns the same stream time. It will return -1 if the stream time
given is not inside the segment.

The new wording makes it easier to understand exactly what each field of the
GstSegment struct represents.

When a running-time-offset is stored in the event, it could become smaller
than 0 although the event is otherwise correct. This can happen when pad
offsets are used.

To prevent this, we set the timestamp to -diff, so that in the end the sum of
both is exactly 0.

https://bugzilla.gnome.org/show_bug.cgi?id=754356

If the second parameter is an absolute URI, the base should have no effect and
the second parameter should be returned again.

https://bugzilla.gnome.org/show_bug.cgi?id=755134

https://bugzilla.gnome.org/show_bug.cgi?id=755607

Iterator may need to be resynced, for instance if pads are released
during state change.

got_eos should be protected by the object lock of the element, not of
the pad, as is the case throughout the rest of the funnel code.

https://bugzilla.gnome.org/show_bug.cgi?id=755343

This fixes a race where a state change may return failure if it has
request pads that are deactivated and removed (and thus have no
parent) at the same time as the element changes state and (de)activates
its pads.

https://bugzilla.gnome.org/show_bug.cgi?id=755342

https://bugzilla.gnome.org/show_bug.cgi?id=755511

After doing gst_pad_push() in case of sync_streams and cache_buffers,
if the buffer can not be kept in cache, it should be unreffed to avoid
memory leackage.

https://bugzilla.gnome.org/show_bug.cgi?id=755141

gst_caps_can_intersect() & GST_PAD_IS_ACCEPT_INTERSECT()
documentation typo fix.

https://bugzilla.gnome.org/show_bug.cgi?id=755257

https://bugzilla.gnome.org/show_bug.cgi?id=755019

https://bugzilla.gnome.org/show_bug.cgi?id=755019