This uses the GstWlClock to schedule frames. This is a first
implementation, some improvement are required.

*Implement preroll handling*

We need to wait for the preroll to complete before rendering another
frame. See gst_base_sink_wait_preroll()

*Handle play/pause transition*

We need to immediatly render any pending frame when going to pause. And
we need to reset the clock calibration when doing to pause in case the
master clock was not progressing like an audio clock.

This fixes the GstWlCLock implementation which was wrongly assuming
one could have both access to the control pipe and be able to use the
flushing mechanism.

Add a adjust function that does not change the internal state of
the clock. Otherwise the time will freeze and cause hangs.

If the compositor supports presentation feedback we created a clock
derived from system clock which would mean 2 threads per sink. Instead,
implement our own clock over the GstWlDisplay thread.

This is an experiment where we do an immediate estimate of the
render delay and update the property.

In slow path, we may need to copy the frame to different type of memory.
We can reduce the effect on latency by doing so before waiting on the
clock.

This is useful start to analyse the data and figure-out how to use it.

This is not strictly related to wayland, it's just a POSIX system
clock that supports more clock types. This is needed as the clk_id
passed to clock_gettime() can be anything supported on the target
OS.

Joan Pau Beltran authored Oct 11, 2016 and Nicolas Dufresne committed Oct 14, 2016

https://bugzilla.gnome.org/show_bug.cgi?id=758717

Arun Raghavan authored Sep 29, 2016 and Nicolas Dufresne committed Oct 13, 2016

There was a small omission in the code. It still worked, but the pool
configuration may not have been optimimal.

Fixes accidental common change when commit patch.

Edward Hervey authored Oct 11, 2016 and Edward Hervey committed Oct 12, 2016

If the parent bin can handle it, only add/remove the new/gone stream
instead of re-adding/re-moving everything

https://bugzilla.gnome.org/show_bug.cgi?id=772742

Edward Hervey authored Oct 12, 2016 and Edward Hervey committed Oct 12, 2016

This was a regression.

We only have a upstream-id via STREAM_START if we were in push-mode.
In pull-mode we need to create one.

Note: It would be good to eventually have that method (copied from
gst_pad_get_stream_id_internal()) public in the future

Edward Hervey authored Oct 11, 2016 and Edward Hervey committed Oct 11, 2016

Some variables were hidden or unused

Edward Hervey authored Jul 29, 2015 and Edward Hervey committed Oct 11, 2016

For each MpegTSBaseStream, we have a GstStream object which
subclasses can extend with information.

For each program a GstStreamCollection is created with all
GstStream from each stream.

Edward Hervey authored Oct 11, 2016 and Edward Hervey committed Oct 11, 2016

Just have one function to check for private section streams

Arnaud Vrac authored Oct 10, 2016 and Tim-Philipp Müller committed Oct 11, 2016

https://bugzilla.gnome.org/show_bug.cgi?id=772706

Arnaud Vrac authored Oct 10, 2016 and Tim-Philipp Müller committed Oct 11, 2016

GstHLSDemuxStream inherits GstAdaptiveDemuxStream, not
GstAdaptiveDemux. Fixes memory corruption issues.

https://bugzilla.gnome.org/show_bug.cgi?id=772704

Fixes wrong default failover when parsing libdvbv5 formatted
channel configuration files.

Avoids relying on the H/W default when the information
can be extracted from the configuration file.

Edward Hervey authored Oct 08, 2016 and Edward Hervey committed Oct 08, 2016

When dealing with TIME-based input, the incoming stream could have
potentially changed completely.

In order to check whether it did or not, we need to re-check all sections
(PAT, PMT...). If it didn't, we will keep using the existing streams/pad,
and if it did we will act as if there was a program switch.

Fixes HLS streaming with decodebin3/playbin3

Edward Hervey authored Oct 02, 2016 and Edward Hervey committed Oct 08, 2016

In order to calculate the *actual* bitrate for downloading a fragment
we need to take into account the time since we requested the fragment.

Without this, the bitrate calculations (previously reported by queue2)
would be biased since they wouldn't take into account the request latency
(that is the time between the moment we request a specific URI and the
moment we receive the first byte of that request).

Such examples were it would be biased would be high-bandwith but high-latency
networks. If you download 5MB in 500ms, but it takes 200ms to get the first
byte, queue2 would report 80Mbit/s (5Mb in 500ms) , but taking the request
into account it is only 57Mbit/s (5Mb in 700ms).

While this would not cause too much issues if the above fragment represented
a much longer duration (5s of content), it would cause issues with short
ones (say 1s, or when doing keyframe-only requests which are even shorter)
where the code would expect to be able to download up to 80Mbit/s ... whereas
if we take the request time into account it's much lower (and we would
therefore end up doing late requests).

Also calculate the request latency for debugging purposes and further
usage (it could allow us to figure out the maximum request rate for
example).

https://bugzilla.gnome.org/show_bug.cgi?id=733959

https://bugzilla.gnome.org/show_bug.cgi?id=772330

Jimmy Ohn authored Oct 06, 2016 and Sebastian Dröge committed Oct 06, 2016

gst_caps_to_string function returned allocated memory.
So, It should be freed using g_free function.

https://bugzilla.gnome.org/show_bug.cgi?id=772499

Using g_thread_join() in _finalize() handlers may result in a deadlock
joining the current thread when the last reference is held by a signal
handler.

e.g.:

error 'Resource deadlock avoided' during 'pthread_join (pt->system_thread, NULL)'

The backtrace looks like this:
[...]
g_thread_join ()
gst_gl_window_finalize ()
gst_gl_window_x11_finalize ()
g_object_unref ()
g_value_unset ()
g_signal_emit_valist ()
g_signal_emit ()
gst_gl_window_send_mouse_event ()
gst_gl_window_mouse_event_cb ()
g_main_dispatch ()
[..]
g_main_loop_run ()
gst_gl_window_navigation_thread ()
g_thread_proxy ()
start_thread ()
clone ()

When only linking the element, the upload object will be created from
_transform_caps() but will never be unreffed as the only case is in _stop().

Add an unref if non-NULL to a new finalize handler for this case.

It's possible that the window may have been destroyed when a winsys
event comes in for it.

Fixes an assertion in make -C tests/check generic/states.check

Fixes an assertion in make -C tests/check generic/states.check

e.g. the final glGetError() must also be completed on the GL thread

It hasn't been used in ages and is dead code.

Wonchul Lee authored Sep 13, 2016 and Olivier Crête committed Oct 03, 2016

https://bugzilla.gnome.org/show_bug.cgi?id=770528