Commit graph

4700 commits

Author SHA1 Message Date
Edward Hervey
247b22ae1a playback: Remove subtitle-encoding property on uridecodebin3
For now, until we found a more future-proof solution
2018-03-12 13:48:18 +01:00
Mathieu Duponchelle
318eb61e23 audioaggregator: remove GstAudioAggregator->info
As we now require subclasses to use a subclass of
GstAudioAggregatorPad, we can reuse its info field

https://bugzilla.gnome.org/show_bug.cgi?id=793943
2018-03-01 15:33:25 +01:00
Mathieu Duponchelle
10835e9919 audioaggregator: refactor conversion API
For the rationale, see:

https://bugzilla.gnome.org/show_bug.cgi?id=793917

Also test audiomixer conversion of current output buffer
2018-03-01 00:40:24 +01:00
Edward Hervey
2c4dfa101a Update disted backup ORC files 2018-02-15 07:14:20 +01:00
Nicolas Dufresne
a8ee76bd30 doc: Remove obsolete Since 0.10.X marks 2018-02-14 14:39:32 -05:00
Philippe Normand
22a6287b8e playbin3: fix source-setup signal emission
Previous code would pass 0 as source element.
2018-02-14 10:23:03 +00:00
Tim-Philipp Müller
8c5725c4ea audiomixer, audiointerleave: add to build
https://bugzilla.gnome.org/show_bug.cgi?id=791218
2018-02-13 16:38:32 +00:00
Tim-Philipp Müller
ab758a9a39 audioaggregator, audiomixer, audiointerleave: move from -bad to -base
https://bugzilla.gnome.org/show_bug.cgi?id=791218
2018-02-13 15:56:49 +00:00
Edward Hervey
eacb7a77d2 playbin3: Re-enable buffering message handling
Buffering messages are only sent for the active group (in case there
is more than one).

If the inactive group posts buffering messages we keep the last one
around and will post it once it becomes the playing one.
2018-02-10 15:43:54 +01:00
François Laignel
ec7d81f67c decodebin3: high cpu usage after eos
After eos, decodebin3 enters a loop sending eos events which causes high cpu usage.

https://bugzilla.gnome.org/show_bug.cgi?id=792693
2018-02-10 13:37:27 +01:00
Edward Hervey
1f792ab89d decodebin3: Handle dual-output of STREAM_START/EOS
In order to flush out multiqueue, we send again a STREAM_START and
then a EOS event.

The problem was that was that we might end up pushing out on the
output of multiqueue (and therefore decodebin3) a series of:
* EOS / STREAM_START / EOS

Apart from the uglyness of such output, If decodebin3 is used with
elements such as concat on their output, they might potentially
block on that second STREAM_START.

In order to make sure we don't end up in that situation we send
a custom STREAM_START event when refreshing multiqueue (which we
drop on the output) and we don't special case EOS events on streams
on which we already got EOS.

At worst we now end up sending at most two EOS on the output of
multiqueue (and decodebin3).
2018-02-10 12:24:10 +01:00
Edward Hervey
c1e902a084 playbin3: Implement gapless playback
Similar in vein to the playbin2 architecture except that uridecodebin3
are prerolled much earlier and all streams of the same type are
fed through a 'concat' element.

This keeps the philosphy of having all elements connected as soon
as possible.

The 'about-to-finish' signal is emitted whenever one of the uridecodebin
is about to finish, allowing the users to set the next uri/suburi.

The notion of a group being active has changed. It now means that the
uridecodebin3 has been activated, but doesn't mean it is the one
currently being outputted by the sinks (i.e. curr_group and next_group).
This is done via detecting GST_MESSAGE_STREAM_START emission by playsink
and figuring out which group is really playing.

When the current group changes, a new thread is started to deactivate
the previous one and optionnaly fire 'about-to-finish'.
2018-02-10 12:24:10 +01:00
Edward Hervey
b9e73d0575 playbin3: Use uridecodebin3 and link/reconfigure immediately
Apologies for the big commit, but it wasn't really possible to split it
in anything smaller.

* Switch to uridecodebin3 instead of managing urisourcebin and decodebin3
ourselves. No major architectural change with this.

* Reconfigure sinks/outputs when needed. This is possible thanks to the
various streams-related API. Instead of blocking new pads and waiting
for a (fake) no-more-pads to decide what to connect, we instead reconfigure
playsink and the combiners to whatever types are currently selected. All of
this is done in reconfigure_output().
  New pads are immediately connected to (combiners and) sinks, allowing
immediate negotiation and usage.

* Since elements are always connected, the "cached-duration" feature is gone
and queries can reach the target elements.

* The auto-plugging related code is currently disabled entirely until
we get the new proper API.

* Store collections at the GstSourceGroup level and not globally

* And more comments a bit everywhere

NOTE: gapless is still not functional, but this opens the way to be able
to handle it in a streams-aware fashion (where several uridecodebin3 can
be active at the same time).
2018-02-10 12:24:09 +01:00
Edward Hervey
ebf138e29e urisourcebin: Add 'about-to-finish' signal
With push-based sources, urisourcebin will emit this signal when
the stream has been fully consumed.

This signal can be used to know when the source is done providing
data.
2018-02-10 12:24:09 +01:00
Edward Hervey
08044ab7e6 playback: New uridecodebin3 element
In the same vein as old uridecodebin except that it also
accepts a suburi and uses urisourcebin and decodebin3 internally
2018-02-10 12:24:09 +01:00
Edward Hervey
2858400608 playbin3: Remove wrong 'notify'
Those properties doesn't exist on playbin3, don't emit a notify for that
2018-02-10 12:17:39 +01:00
Edward Hervey
8386ea706f playbin3: Remove setting 'subtitle-encoding' on decodebin
That property doesn't exist
2018-02-10 12:17:39 +01:00
Edward Hervey
7b8cfb9e53 playbin3: Clarify documentation of combiner properties 2018-02-10 12:17:39 +01:00
Edward Hervey
4ed0708f78 playbin3: Remove unused define 2018-02-10 12:17:39 +01:00
Edward Hervey
9e7401bedf decodebin3: Use GST_GROUP_ID_INVALID 2018-02-10 12:17:39 +01:00
Edward Hervey
dd7a1407a4 decodebin3: Don't forward already-handling SELECT_STREAMS
Upstream might respond negatively to the event, whereas we actually
handled it.
2018-02-10 12:17:38 +01:00
Edward Hervey
401cadfd74 decodebin3: Add new about-to-finish signal 2018-02-10 12:17:38 +01:00
Edward Hervey
c136acf27a decodebin3: Remove unused definition 2018-02-10 12:17:38 +01:00
Edward Hervey
fa710a59aa decodebin3: Don't take the lock when creating a new input
We only need to take the input lock when adding/removing
inputs from the list.
2018-02-10 12:17:38 +01:00
Edward Hervey
fa9adbe986 playbin3: Remove unused variable
The lock is never used
2018-02-10 12:17:38 +01:00
Edward Hervey
1717e33da5 urisourcebin: Remove auto-plugging signals
They were never used and we need a better system
2018-02-10 12:17:36 +01:00
Edward Hervey
e5157ca400 urisourcebin: Remove ASYNC behaviour
It is not needed in the new streams-aware world
2018-02-10 12:16:09 +01:00
Edward Hervey
f4035e430a urisourcebin: Remove 'unknown-type' signal
It was never used and makes no sense in the new streams-based world
2018-02-10 12:16:09 +01:00
Edward Hervey
0f51dde728 playbin3: Disable autoplug-related code
The signals were never emitted from decodebin3. This needs
switching to a new signalling system
2018-02-10 12:16:09 +01:00
Edward Hervey
29f57e23c7 playbin3: fix "no-more-pads" handling
That signal is never emitted by decodebin3 and is handled differently
2018-02-10 12:14:47 +01:00
Edward Hervey
9f1063e4a3 playbin3: Remove duration caching
This is now handled directly via sinks and queries through pads
2018-02-10 12:14:47 +01:00
Edward Hervey
fc93302087 parsebin: Remove async behaviour
There's no reason to do async changing
2018-02-10 12:14:17 +01:00
Jan Schmidt
577c14495c tcp/multihandlesink: Handle the case of no caps
Pass data with no caps and no streamheaders without
throwing a bunch of criticals
2018-02-09 18:07:59 +11:00
Tim-Philipp Müller
3219b704fd subparse: fix pushing out of last chunk if last line has no newline
With playbin the last subtitle chunk would not get displayed
if the last chunk was missing a newline at the end. This is
because streamsynchronizer will hold back the EOS event until
the audio and video streams are finished too, so subparse
would never forcefully push out the last chunk until the very
end when it is too late.

We get a STREAM_GROUP_DONE event from streamsynchronizer however,
so handle that like EOS and force out any remaining text then.

https://bugzilla.gnome.org/show_bug.cgi?id=771853
2018-01-25 18:42:08 +00:00
Ashish Kumar
b927d60371 playback-utils: Fix caps leak on failure
https://bugzilla.gnome.org/show_bug.cgi?id=789358
2018-01-17 12:18:41 +00:00
Sam Thursfield
f1324f9cb0 videotestsrc: Document the num-buffers property
It's not obvious from the existing docs that this option exists, nor
that it can be used to give a pipeline an exact length in video frames.

https://bugzilla.gnome.org/show_bug.cgi?id=777647
2018-01-12 00:49:20 +00:00
Edward Hervey
384d27002c typefind: Fix mp3 typefinding with multiple different headers
(yes, this has never worked since it was introduced, don't worry)

If we want to actually detect layer/channels/samplerate changes,
it would be better to:
* not reset the various prev_* variables at every iteration.
* and actually store the values when they change

CID #206079
CID #206080
CID #206081
2018-01-03 15:51:34 +01:00
Mathieu Duponchelle
164b5a7f94 audioaggregator: implement input conversion
https://bugzilla.gnome.org/show_bug.cgi?id=786344
2017-12-19 23:39:37 +01:00
Nicolas Dufresne
5affe09c48 videoconvert: Filter-out crop meta
To passthrough crop-meta, the converter would need to allocate and
convert buffers of the size of the originating buffer. This is currently
made difficult by GstBaseTransform since we cannot alter the caps passed
though the allocation query. We would also need to wait for the first
input buffer to be received in order to make the decision around that
size.

So the short and safe solution is just to stop pretending we can
passthrought that meta.

https://bugzilla.gnome.org/show_bug.cgi?id=791412
2017-12-17 15:50:00 -05:00
Dongil Park
a8a5905ec6 playbin3: Fix accessing invalid index in GstStream when received select-stream event
If select-stream event was send to playbin3 as missing any GstStream of ES type
(V or A or TEX) of collection then, playbin will access to invalid address of
GstStream due to invalid index limit. This caused SIGSEGV.

https://bugzilla.gnome.org/show_bug.cgi?id=791638
2017-12-15 09:03:12 +01:00
Edward Hervey
aa10e82257 typefind: Avoid overflow calculation (image/quicktime)
The qt typefinder uses guint64 values for offset and size calculation
but the typefinder system only supports gint64 values.

Make sure we don't end up using potentially overflowing values.
2017-12-08 09:12:33 +01:00
Edward Hervey
1df9b05f18 typefind: Avoid overflow calculation
The qt typefinder uses guint64 values for offset and size calculation
but the typefinder system only supports gint64 values.

Make sure we don't end up using potentially overflowing values.
2017-12-08 08:05:10 +01:00
Edward Hervey
32eae2b166 typefind: Fix previous commit
We need to make sure we have *enough* data to read (including the
next 4 bytes)
2017-12-07 16:10:22 +01:00
Edward Hervey
b1c135aa72 typefind: Optimize qt/3gp typefinding
Request the whole 'ftyp' atom data and scan within it. Avoids doing
a lot of small 4 byte pull from upstream if it's invalid.
2017-12-07 15:05:07 +01:00
Edward Hervey
0428b4afda typefind: wavpack: limit search in blocksize
The maximum blocksize is 131072 bytes, anything bigger is a corrupted
file.
2017-12-07 15:05:07 +01:00
Edward Hervey
a3bf3094d1 typefind: Fix ico out-of-bound read
The furthest we go to verify the data is reading a guint32 at offset
18, therefore make sure we can read as much.
2017-12-07 11:04:50 +01:00
Tim-Philipp Müller
6c0744a5fc Remove GstAggregator from -bad, moved to core
https://bugzilla.gnome.org/show_bug.cgi?id=739010
2017-12-02 16:14:36 +00:00
Jun Xie
f88b20397f decodebin: Fix typo in docs
https://bugzilla.gnome.org/show_bug.cgi?id=791058
2017-12-01 10:03:40 +02:00
Edward Hervey
f2f87d220f urisourcebin: Don't leak mutexes 2017-11-25 12:45:03 +01:00
Mathieu Duponchelle
9e4b3fbd82 aggregator: Remove klass->sinkpads_type
This posed problems for the python bindings (and possibly others).

Instead, subclasses now use add_pad_template_with_gtype.

https://bugzilla.gnome.org/show_bug.cgi?id=789986
2017-11-22 16:52:29 +01:00