platform/upstream/gstreamer.git
6 years agomeson: fix invalid keyword argument warnings
Tim-Philipp Müller [Fri, 20 Apr 2018 11:30:24 +0000 (12:30 +0100)]
meson: fix invalid keyword argument warnings

cc.compiles() doesn't have a 'prefix' argument (yet) and the
prefix has already been prepended to the source code snippets.

https://github.com/mesonbuild/meson/issues/2364

6 years agoRevert "docs: Minor fix in event_new_select_streams"
Thibault Saunier [Wed, 18 Apr 2018 14:35:20 +0000 (11:35 -0300)]
Revert "docs: Minor fix in event_new_select_streams"

This reverts commit f218917d02760f8f32a35e4e635e23230c47c0c6.

6 years agodocs: Minor fix in event_new_select_streams
Thibault Saunier [Tue, 17 Apr 2018 23:03:09 +0000 (20:03 -0300)]
docs: Minor fix in event_new_select_streams

6 years agomultiqueue, inputselector: show pad properties in gst-inspect-1.0
Tim-Philipp Müller [Tue, 17 Apr 2018 10:24:31 +0000 (11:24 +0100)]
multiqueue, inputselector: show pad properties in gst-inspect-1.0

6 years agoinputselector, outputselector: add guards for wrong pads being set as active pads
Tim-Philipp Müller [Tue, 17 Apr 2018 10:01:09 +0000 (11:01 +0100)]
inputselector, outputselector: add guards for wrong pads being set as active pads

Catch users wrongly setting foreign pads or wrong pads as
the selector's active pad, which leads to all kinds of
other issues. It's a programming error so handle it just
like we would if we had direct API.

https://bugzilla.gnome.org/show_bug.cgi?id=795309

6 years agocaps: Add a macro based variant of gst_caps_copy
Thibault Saunier [Tue, 17 Apr 2018 17:00:20 +0000 (14:00 -0300)]
caps: Add a macro based variant of gst_caps_copy

This way we do not hit the performance overhead of having the method
not inlined but still can use it from bindings.

6 years agopad: Handle changing sticky events in pad probes
Thibault Saunier [Mon, 16 Apr 2018 19:30:27 +0000 (16:30 -0300)]
pad: Handle changing sticky events in pad probes

In the case where the user sets a new padprobeinfo->data in a probe
where the data is a sticky event, the new sticky event should be automatically
sticked on the probed pad.

https://bugzilla.gnome.org/show_bug.cgi?id=795330

6 years agodebug: Make PADS debug background blue
Thibault Saunier [Tue, 17 Apr 2018 12:33:02 +0000 (09:33 -0300)]
debug: Make PADS debug background blue

Red on red was... suboptimal!

https://bugzilla.gnome.org/show_bug.cgi?id=795330

6 years agowin32: update defs for new exports
Tim-Philipp Müller [Tue, 17 Apr 2018 16:00:53 +0000 (17:00 +0100)]
win32: update defs for new exports

6 years agoharness: Handle harness->element not being a GstBin
Thibault Saunier [Mon, 16 Apr 2018 19:27:57 +0000 (16:27 -0300)]
harness: Handle harness->element not being a GstBin

It is totally valid but in gst_harness_find_element we were not
handling that case.

https://bugzilla.gnome.org/show_bug.cgi?id=795308

6 years agogst: Stop inlining gst_caps_copy
Thibault Saunier [Wed, 4 Apr 2018 20:36:57 +0000 (17:36 -0300)]
gst: Stop inlining gst_caps_copy

This way it gets exposed to bindings through GObject Introspection.

6 years agoAutomatic update of common submodule
Tim-Philipp Müller [Mon, 16 Apr 2018 09:52:46 +0000 (10:52 +0100)]
Automatic update of common submodule

From f0c2dc9 to ed78bee

6 years agogstconfig.h.in: initial RISC-V support
Aurelien Jarno [Sat, 14 Apr 2018 22:49:55 +0000 (00:49 +0200)]
gstconfig.h.in: initial RISC-V support

RISC-V supports unaligned accesses, but these might run extremely slowly
depending on the implementation. Therefore set GST_HAVE_UNALIGNED_ACCESS
to 0 on this architecture.

https://bugzilla.gnome.org/show_bug.cgi?id=795271

6 years agoadapter: port the buffer list from GSList to GstQueueArray
Mathieu Duponchelle [Wed, 11 Apr 2018 15:16:54 +0000 (17:16 +0200)]
adapter: port the buffer list from GSList to GstQueueArray

Significantly reduces the amount of memory allocation operations.

https://bugzilla.gnome.org/show_bug.cgi?id=795167

6 years agoAPI: gst_queue_array_peek_nth
Mathieu Duponchelle [Wed, 11 Apr 2018 13:38:36 +0000 (15:38 +0200)]
API: gst_queue_array_peek_nth

https://bugzilla.gnome.org/show_bug.cgi?id=795157

6 years agogstqueuearray: make find() return a 0-based index
Mathieu Duponchelle [Wed, 11 Apr 2018 11:44:33 +0000 (13:44 +0200)]
gstqueuearray: make find() return a 0-based index

And make the drop() functions expect a 0-based index too,
this addresses a longstanding FIXME. This will not break
backward compatibility, because the drop() functions
were previously only meant to be used with the index
returned by find().

https://bugzilla.gnome.org/show_bug.cgi?id=795156

6 years agogstsample: new API
Mathieu Duponchelle [Tue, 10 Apr 2018 22:49:02 +0000 (00:49 +0200)]
gstsample: new API

gst_sample_set_buffer
gst_sample_set_caps
gst_sample_set_segment
gst_sample_set_info
gst_sample_is_writable
gst_sample_make_writable

This commit makes it possible to reuse a sample object and avoid
unnecessary memory allocations, for example in appsink.

In addition, writability is now required to set the buffer list.

https://bugzilla.gnome.org/show_bug.cgi?id=795144

6 years agobase: fix some GIR annotations
Mark Nauwelaerts [Fri, 13 Apr 2018 18:15:46 +0000 (20:15 +0200)]
base: fix some GIR annotations

Mostly related to out parameters and their transfer

6 years agogst: add some GIR array annotations
Mark Nauwelaerts [Thu, 29 Mar 2018 16:59:43 +0000 (18:59 +0200)]
gst: add some GIR array annotations

6 years agodebugutils: Add missing parameters documentation
Víctor Manuel Jáquez Leal [Fri, 13 Apr 2018 07:58:05 +0000 (09:58 +0200)]
debugutils: Add missing parameters documentation

6 years agogstdebug: fix occasional deadlocks on windows when outputting debug logging
Tim-Philipp Müller [Wed, 11 Apr 2018 18:56:01 +0000 (19:56 +0100)]
gstdebug: fix occasional deadlocks on windows when outputting debug logging

When outputting debug logs on Windows, some sections are protected
with a non-recursive lock. Turns out though that gst_debug_message_get()
might indirectly, via our printf format extensions, call code which
in turn would try to log something when it can't handle something. If
that happens we end up in gst_debug_log_default() again recursively and
try to again take the lock that's already taken, thus deadlocking.

Format the debug message string outside of the critical section
instead to avoid this.

https://bugzilla.gnome.org/show_bug.cgi?id=784382

6 years agogsturi: include gstconfig.h earlier for GST_API define
Tim-Philipp Müller [Mon, 9 Apr 2018 13:19:19 +0000 (14:19 +0100)]
gsturi: include gstconfig.h earlier for GST_API define

6 years agogstinfo: fix debug levels being applied in the wrong order
Jan Alexander Steffens (heftig) [Tue, 27 Mar 2018 08:25:46 +0000 (10:25 +0200)]
gstinfo: fix debug levels being applied in the wrong order

Remove unneeded reapplication of patterns. Besides being
superfluous (gst_debug_reset_threshold already applies
patterns) it was also wrong and didn't stop checking patterns
after the first match (broken in 67e9d139).

Also fix up unit test which checked for the wrong order.

https://bugzilla.gnome.org/show_bug.cgi?id=794717

6 years agogstinfo: Simplify gst_debug_reset_threshold() implementation
Jan Alexander Steffens (heftig) [Tue, 27 Mar 2018 08:15:46 +0000 (10:15 +0200)]
gstinfo: Simplify gst_debug_reset_threshold() implementation

Replace the while+goto with a for+break and check walk to determine
whether we had a match. Move up the unlock to keep the locked section as
small as possible.

https://bugzilla.gnome.org/show_bug.cgi?id=794717

6 years agogstinfo: Reduce code duplication around level pattern matching
Jan Alexander Steffens (heftig) [Tue, 27 Mar 2018 08:14:27 +0000 (10:14 +0200)]
gstinfo: Reduce code duplication around level pattern matching

Move the match, logging and set_threshold to a new function.

The log levels are different, so choose the higher one (LOG). Having two
equivalent messages at two different levels seems like a bad idea
anyway.

https://bugzilla.gnome.org/show_bug.cgi?id=794717

6 years agostreamcollection: embed GQueue into the private struct
Tim-Philipp Müller [Tue, 27 Mar 2018 16:16:05 +0000 (17:16 +0100)]
streamcollection: embed GQueue into the private struct

6 years agodocs: Update libs documentation
Edward Hervey [Mon, 2 Apr 2018 10:44:15 +0000 (12:44 +0200)]
docs: Update libs documentation

* Make sure all libcheck headers are ignored
* Add all missing symbols

6 years agodocs: Update gst core doc
Edward Hervey [Mon, 2 Apr 2018 10:43:57 +0000 (12:43 +0200)]
docs: Update gst core doc

6 years agogst: Documentation fixes
Edward Hervey [Mon, 2 Apr 2018 10:42:30 +0000 (12:42 +0200)]
gst: Documentation fixes

* Fix copy-paste error for GstParamSpecArray documentation
* Use proper field name for tracer utils documentation

6 years agolibs: Documentation fixes
Edward Hervey [Mon, 2 Apr 2018 10:41:48 +0000 (12:41 +0200)]
libs: Documentation fixes

* Symbols not properly exposed or wrongly named

6 years agobin: fix deep-element-added signal debug log message
Matthew Waters [Thu, 29 Mar 2018 01:36:11 +0000 (12:36 +1100)]
bin: fix deep-element-added signal debug log message

Adding the bin to the child element doesn't really make sense.

6 years agomeson: bump meson req for gnome.mkenums_simple()
Tim-Philipp Müller [Thu, 22 Mar 2018 13:00:21 +0000 (13:00 +0000)]
meson: bump meson req for gnome.mkenums_simple()

6 years agomeson: remove no longer needed core enumtypes template files
Tim-Philipp Müller [Thu, 22 Mar 2018 12:18:28 +0000 (12:18 +0000)]
meson: remove no longer needed core enumtypes template files

6 years agomeson: use gnome.mkenums_simple() to generate core enumtypes
Alessandro Decina [Thu, 20 Jul 2017 08:12:43 +0000 (18:12 +1000)]
meson: use gnome.mkenums_simple() to generate core enumtypes

6 years agomeson: use gnome.mkenums_simple() to generate controller enumtypes
Alessandro Decina [Thu, 20 Jul 2017 03:03:55 +0000 (13:03 +1000)]
meson: use gnome.mkenums_simple() to generate controller enumtypes

6 years agomeson: delete unused variable
Alessandro Decina [Wed, 19 Jul 2017 09:37:02 +0000 (19:37 +1000)]
meson: delete unused variable

6 years agotest: Pool now try to reset the size
Nicolas Dufresne [Thu, 22 Mar 2018 00:02:50 +0000 (20:02 -0400)]
test: Pool now try to reset the size

As a side effect, buffers are no longer expected to be discarded on
resize.

6 years agogstbuffer: reset buffer to its original size if intact
Petr Kulhavy [Mon, 14 Nov 2016 14:35:50 +0000 (15:35 +0100)]
gstbuffer: reset buffer to its original size if intact

Enhance default_reset_buffer() to resize the buffer to its full size if the
memory hasn't changed. This allows to reuse the buffer even if the offset has
changed or the size has shrunk, rather than freeing the buffer.

Change related to: https://bugzilla.gnome.org/show_bug.cgi?id=772841

6 years agonet: Include gstnetcontrolmessagemeta.h in net.h
Sebastian Dröge [Wed, 21 Mar 2018 08:20:14 +0000 (10:20 +0200)]
net: Include gstnetcontrolmessagemeta.h in net.h

6 years agoparamspecs: Set g-i annotation values for GST_PARAM_* constants
Sebastian Dröge [Wed, 21 Mar 2018 08:13:44 +0000 (10:13 +0200)]
paramspecs: Set g-i annotation values for GST_PARAM_* constants

6 years agoelementfactory: GST_ELEMENT_FACTORY_TYPE_DECODABLE had DECRYPTOR added, update g...
Sebastian Dröge [Wed, 21 Mar 2018 08:11:30 +0000 (10:11 +0200)]
elementfactory: GST_ELEMENT_FACTORY_TYPE_DECODABLE had DECRYPTOR added, update g-i annotation value

6 years agogst: Fix compilation with latest GLib
Sebastian Dröge [Tue, 20 Mar 2018 14:11:01 +0000 (16:11 +0200)]
gst: Fix compilation with latest GLib

g_object_ref() forwards the type of its argument nowadays.

./grammar.y:409:14: error: assignment from incompatible pointer type [-Werror=incompatible-pointer-types]
gstchildproxy.c:212:7: error: assignment from incompatible pointer type [-Werror=incompatible-pointer-types]

6 years agoBack to development
Tim-Philipp Müller [Tue, 20 Mar 2018 09:02:34 +0000 (09:02 +0000)]
Back to development

6 years agoRelease 1.14.0
Tim-Philipp Müller [Mon, 19 Mar 2018 20:09:51 +0000 (20:09 +0000)]
Release 1.14.0

6 years agoUpdate docs
Tim-Philipp Müller [Mon, 19 Mar 2018 20:09:51 +0000 (20:09 +0000)]
Update docs

6 years agoRelease 1.13.91
Tim-Philipp Müller [Tue, 13 Mar 2018 19:08:54 +0000 (19:08 +0000)]
Release 1.13.91

6 years agoUpdate docs
Tim-Philipp Müller [Tue, 13 Mar 2018 19:08:54 +0000 (19:08 +0000)]
Update docs

6 years agomeson: docs: update api decorators to ignore
Tim-Philipp Müller [Tue, 13 Mar 2018 11:54:42 +0000 (11:54 +0000)]
meson: docs: update api decorators to ignore

6 years agodocs: fixup for new libs API export decorators
Tim-Philipp Müller [Mon, 12 Mar 2018 23:12:13 +0000 (23:12 +0000)]
docs: fixup for new libs API export decorators

6 years agonet: GST_EXPORT -> GST_NET_API
Tim-Philipp Müller [Mon, 12 Mar 2018 23:03:26 +0000 (23:03 +0000)]
net: GST_EXPORT -> GST_NET_API

We need different export decorators for the different libs.
For now no actual change though, just rename before the release,
and add prelude headers to define the new decorator to GST_EXPORT.

6 years agocontroller: GST_EXPORT -> GST_CONTROLLER_API
Tim-Philipp Müller [Mon, 12 Mar 2018 23:03:26 +0000 (23:03 +0000)]
controller: GST_EXPORT -> GST_CONTROLLER_API

We need different export decorators for the different libs.
For now no actual change though, just rename before the release,
and add prelude headers to define the new decorator to GST_EXPORT.

6 years agocheck: GST_EXPORT -> GST_CHECK_API
Tim-Philipp Müller [Mon, 12 Mar 2018 23:03:26 +0000 (23:03 +0000)]
check: GST_EXPORT -> GST_CHECK_API

We need different export decorators for the different libs.
For now no actual change though, just rename before the release,
and add prelude headers to define the new decorator to GST_EXPORT.

6 years agobase: GST_EXPORT -> GST_BASE_API
Tim-Philipp Müller [Mon, 12 Mar 2018 23:03:26 +0000 (23:03 +0000)]
base: GST_EXPORT -> GST_BASE_API

We need different export decorators for the different libs.
For now no actual change though, just rename before the release,
and add prelude headers to define the new decorator to GST_EXPORT.

6 years agogst: GST_EXPORT -> GST_API
Tim-Philipp Müller [Mon, 12 Mar 2018 23:03:26 +0000 (23:03 +0000)]
gst: GST_EXPORT -> GST_API

We need different export decorators for the different libs.
For now no actual change though, just rename before the release,
and add prelude headers to define the new decorator to GST_EXPORT.

6 years agopromise: be more explicit in docs about who/when to use reply/interrupt/expire
Matthew Waters [Thu, 8 Mar 2018 02:30:30 +0000 (13:30 +1100)]
promise: be more explicit in docs about who/when to use reply/interrupt/expire

https://bugzilla.gnome.org/show_bug.cgi?id=794153

6 years agobasesrc: Balance unlock/unlock_stop in _src_stop()
Nicolas Dufresne [Wed, 7 Mar 2018 16:19:25 +0000 (11:19 -0500)]
basesrc: Balance unlock/unlock_stop in _src_stop()

Otherwise it's possible that we won't be able to start again
depending the implementation. We do start/stop in normal use cases
whenever GST_QUERY_SCHEDULING happens before we are started.

https://bugzilla.gnome.org/show_bug.cgi?id=794149

6 years agobasesrc: No need to stop flushing in start_complete
Nicolas Dufresne [Wed, 7 Mar 2018 16:16:00 +0000 (11:16 -0500)]
basesrc: No need to stop flushing in start_complete

The flushing state is handled a bit differently, there is no need
to stop flushing in start_complete. This would other result in
unlock_stop being called without unlock_start.

Unlike what the old comment says, there is no need to take the live
lock here, we are still single threaded at this point (app thread
or the state change thread). Also, we will wait for playing state
in create/getrange, no need to do that twice.

https://bugzilla.gnome.org/show_bug.cgi?id=794149

6 years agodebugutils: Change dot-file functions documentation to proper gtk-doc
Sebastian Dröge [Mon, 5 Mar 2018 09:52:24 +0000 (11:52 +0200)]
debugutils: Change dot-file functions documentation to proper gtk-doc

This way gobject-introspection also picks it up and handles our
annotations.

See https://gitlab.gnome.org/GNOME/gobject-introspection/issues/194

6 years agoqueuearray: Implement pop_tail_struct() for completeness
Sebastian Dröge [Sun, 4 Mar 2018 08:53:10 +0000 (10:53 +0200)]
queuearray: Implement pop_tail_struct() for completeness

All other variants of {peek,pop}_{head,tail}_{,struct} were already
implemented.

https://bugzilla.gnome.org/show_bug.cgi?id=794035

6 years agogst: Add some more (type filename) annotations
Sebastian Dröge [Sun, 4 Mar 2018 08:24:49 +0000 (10:24 +0200)]
gst: Add some more (type filename) annotations

6 years agoRelease 1.13.90
Tim-Philipp Müller [Sat, 3 Mar 2018 21:51:49 +0000 (21:51 +0000)]
Release 1.13.90

6 years agoUpdate docs
Tim-Philipp Müller [Sat, 3 Mar 2018 21:51:49 +0000 (21:51 +0000)]
Update docs

6 years agoplugins: Don't force 64-bit file/seek functions variants on android
Matthew Waters [Tue, 13 Feb 2018 11:20:18 +0000 (22:20 +1100)]
plugins: Don't force 64-bit file/seek functions variants on android

Most functions are automatically chosen from the _FILE_OFFSET_BITS
define, the remaining one (fstat) is only available on API >= 21 so
check for that

6 years agoAdd new symbol to docs and .def file
Tim-Philipp Müller [Thu, 1 Mar 2018 22:21:17 +0000 (22:21 +0000)]
Add new symbol to docs and .def file

Fixes make check

6 years agoqueue: Ignore thresholds if a query is queued
Olivier Crête [Thu, 1 Mar 2018 21:19:09 +0000 (16:19 -0500)]
queue: Ignore thresholds if a query is queued

The queue gets filled by the tail, so a query will always be the tail
object, not the head object. Also add a _peek_tail_struct() method to the
GstQueueArray to enable looking at the tail.

With unit test to prevent future regression.

https://bugzilla.gnome.org/show_bug.cgi?id=762875

6 years agomeson: -Wformat-* require -Wformat
Mathieu Duponchelle [Thu, 1 Mar 2018 17:38:01 +0000 (18:38 +0100)]
meson: -Wformat-* require -Wformat

6 years agomeson: enable more warnings
Mathieu Duponchelle [Thu, 1 Mar 2018 16:20:06 +0000 (17:20 +0100)]
meson: enable more warnings

Modeled on the autotools build, -W flags are only
added if the compiler supports them.

https://bugzilla.gnome.org/show_bug.cgi?id=793958

6 years agogstaggregator: pads must inherit from #GstAggregatorPad
Mathieu Duponchelle [Wed, 28 Feb 2018 23:31:11 +0000 (00:31 +0100)]
gstaggregator: pads must inherit from #GstAggregatorPad

Document this, and take advantage of that fact to use
GstAggregator.srcpad.segment instead of GstAggregator.segment

https://bugzilla.gnome.org/show_bug.cgi?id=793942

6 years agoRevert "gstaggregator: pads must inherit from #GstAggregatorPad"
Mathieu Duponchelle [Thu, 1 Mar 2018 00:15:34 +0000 (01:15 +0100)]
Revert "gstaggregator: pads must inherit from #GstAggregatorPad"

This reverts commit 9774b3775d8483e5697f9196a26c1e5831113bd6.

Pushed by mistake

6 years agoghostpad: ensure we build a ghost pad ..
Mathieu Duponchelle [Thu, 1 Mar 2018 00:12:07 +0000 (01:12 +0100)]
ghostpad: ensure we build a ghost pad ..

When we construct from a custom GType

6 years agopad: fix mixed declarations
Mathieu Duponchelle [Thu, 1 Mar 2018 00:09:48 +0000 (01:09 +0100)]
pad: fix mixed declarations

6 years agogstaggregator: pads must inherit from #GstAggregatorPad
Mathieu Duponchelle [Wed, 28 Feb 2018 23:31:11 +0000 (00:31 +0100)]
gstaggregator: pads must inherit from #GstAggregatorPad

Document this, and take advantage of that fact to use
GstAggregator.srcpad.segment instead of GstAggregator.segment

https://bugzilla.gnome.org/show_bug.cgi?id=793942

6 years agoaggregator: allow src GstAggregatorPads
Mathieu Duponchelle [Wed, 28 Feb 2018 18:53:42 +0000 (19:53 +0100)]
aggregator: allow src GstAggregatorPads

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

https://bugzilla.gnome.org/show_bug.cgi?id=793934

6 years agopad, ghostpad: use the template gtype if specified
Mathieu Duponchelle [Wed, 28 Feb 2018 18:51:44 +0000 (19:51 +0100)]
pad, ghostpad: use the template gtype if specified

Also make sure the GType passed to the with_gtype versions
of the template constructors is_a GstPad

https://bugzilla.gnome.org/show_bug.cgi?id=793933

6 years agobaseparse: Fix integer overflow in bitrate calculation
Nicolas Dufresne [Thu, 22 Feb 2018 03:25:25 +0000 (22:25 -0500)]
baseparse: Fix integer overflow in bitrate calculation

https://bugzilla.gnome.org/show_bug.cgi?id=793284

6 years agobaseparse: Avoid overflow in update_interval calculation
Nicolas Dufresne [Thu, 22 Feb 2018 03:01:36 +0000 (22:01 -0500)]
baseparse: Avoid overflow in update_interval calculation

https://bugzilla.gnome.org/show_bug.cgi?id=793284

6 years agobaseparse: Fix check for update_interval
Nicolas Dufresne [Thu, 22 Feb 2018 02:43:59 +0000 (21:43 -0500)]
baseparse: Fix check for update_interval

update_interval may be -1

https://bugzilla.gnome.org/show_bug.cgi?id=793284

6 years agomeson: Use .dylib suffix if darwin
Justin Kim [Mon, 19 Feb 2018 06:39:46 +0000 (15:39 +0900)]
meson: Use .dylib suffix if darwin

For Mac OS, GST_EXTRA_MODULE_SUFFIX should be set as '.dylib'.
Otherwise, GStreamer fails to load its plugins.

https://bugzilla.gnome.org/show_bug.cgi?id=793584

6 years agoqueuearray: add _peek_tail() and _pop_tail()
Tim-Philipp Müller [Thu, 1 Feb 2018 18:29:27 +0000 (18:29 +0000)]
queuearray: add _peek_tail() and _pop_tail()

API: gst_queue_array_pop_tail()
API: gst_queue_array_peek_tail()

These will be needed later for appsrc.

6 years agogst: fix some GIR annotations
Mark Nauwelaerts [Tue, 13 Feb 2018 11:38:33 +0000 (12:38 +0100)]
gst: fix some GIR annotations

Mostly related to out parameters and their transfer

6 years agogstbasesink: Include segment.offset in the computation of position
Alicia Boya García [Wed, 10 Jan 2018 03:08:57 +0000 (04:08 +0100)]
gstbasesink: Include segment.offset in the computation of position

Position queries with GST_FORMAT_TIME are supposed to return stream
time.

gst_base_sink_get_position() estimates the current stream time on its
own instead of using gst_segment_to_stream_time(), but the algorithm
used was not taking segment.offset into account, resulting in invalid
values when this field was set to a non-zero value.

https://bugzilla.gnome.org/show_bug.cgi?id=792434

6 years agotracers: latency: allow for non parented pads to send latency probes
Matthew Waters [Thu, 15 Feb 2018 01:58:43 +0000 (12:58 +1100)]
tracers: latency: allow for non parented pads to send latency probes

Such a setup is used in rtspsrc for its TCP connection

https://bugzilla.gnome.org/show_bug.cgi?id=793478

6 years agoBack to development
Tim-Philipp Müller [Thu, 15 Feb 2018 19:44:14 +0000 (19:44 +0000)]
Back to development

6 years agoRelease 1.13.1
Tim-Philipp Müller [Thu, 15 Feb 2018 16:31:16 +0000 (16:31 +0000)]
Release 1.13.1

6 years agotests: pipeline: try to make test_pipeline_reset_start_time more reliable
Tim-Philipp Müller [Thu, 15 Feb 2018 13:36:26 +0000 (13:36 +0000)]
tests: pipeline: try to make test_pipeline_reset_start_time more reliable

Occasionally this test would fail, especially if the system is under load,
because the position query would pick up the last position from the
last buffer timestamp which has a lower timestamp than what we're
looking for. The sleep is long enough, however. It's unclear to me why
exactly this happens but there seems to be some kind of scheduling
issue going on as the streaming thread floods the sink with buffers.

Let's throttle the fakesrc to 100 buffers per second and make the sink
sync to the clock to restore some sanity. It should be totally sufficient
to test what we want to test, and seems to make things reliable here.

6 years agotests: uri: fix build without -DGST_DISABLE_DEPRECATED
Tim-Philipp Müller [Thu, 15 Feb 2018 12:03:20 +0000 (12:03 +0000)]
tests: uri: fix build without -DGST_DISABLE_DEPRECATED

Must undefine it before including gst headers, since the test
tests deprecated API.

6 years agogstconfig.h: want deprecation warnings if GST_DISABLE_DEPRECATED is *set*
Tim-Philipp Müller [Thu, 15 Feb 2018 12:09:31 +0000 (12:09 +0000)]
gstconfig.h: want deprecation warnings if GST_DISABLE_DEPRECATED is *set*

Fix inverted logic. If GST_DISABLE_DEPRECATED is undefined,
we don't want warnings about deprecated API, and if it's
defined we do want warnings.

6 years agopo: update translations
Tim-Philipp Müller [Thu, 15 Feb 2018 11:28:23 +0000 (11:28 +0000)]
po: update translations

6 years agodocs: add flow combiner ref/unref to docs
Tim-Philipp Müller [Wed, 14 Feb 2018 19:37:35 +0000 (19:37 +0000)]
docs: add flow combiner ref/unref to docs

So new-in-1.12 index actually has some entries.

6 years agodocs: add index for new symbols in 1.14
Tim-Philipp Müller [Wed, 14 Feb 2018 19:13:28 +0000 (19:13 +0000)]
docs: add index for new symbols in 1.14

6 years agodocs: add index for new symbols in 1.12
Tim-Philipp Müller [Wed, 14 Feb 2018 19:12:06 +0000 (19:12 +0000)]
docs: add index for new symbols in 1.12

6 years agomeson: make version numbers ints and fix int/string comparison
Tim-Philipp Müller [Thu, 8 Feb 2018 17:22:14 +0000 (17:22 +0000)]
meson: make version numbers ints and fix int/string comparison

WARNING: Trying to compare values of different types (str, int).
The result of this is undefined and will become a hard error
in a future Meson release.

6 years agoautotools: use -fno-strict-aliasing where supported
Tim-Philipp Müller [Sat, 3 Feb 2018 16:56:04 +0000 (17:56 +0100)]
autotools: use -fno-strict-aliasing where supported

https://bugzilla.gnome.org/show_bug.cgi?id=769183

6 years agobuffer: fix gtk-doc warning regarding _get_n_meta() declaration
Tim-Philipp Müller [Sat, 3 Feb 2018 16:55:29 +0000 (17:55 +0100)]
buffer: fix gtk-doc warning regarding _get_n_meta() declaration

6 years agogstpromise: add since 1.14 markers
Matthew Waters [Thu, 1 Feb 2018 13:24:20 +0000 (00:24 +1100)]
gstpromise: add since 1.14 markers

6 years agogst-uninstalled: add webrtc to libraries
Matthew Waters [Wed, 18 Oct 2017 10:24:19 +0000 (21:24 +1100)]
gst-uninstalled: add webrtc to libraries

6 years agogstbuffer.h: move FLAG_LAST documentation back to the bottom
Mathieu Duponchelle [Wed, 31 Jan 2018 13:01:36 +0000 (14:01 +0100)]
gstbuffer.h: move FLAG_LAST documentation back to the bottom

6 years agobasetransform: annotate virtual methods
Mathieu Duponchelle [Wed, 31 Jan 2018 12:36:15 +0000 (13:36 +0100)]
basetransform: annotate virtual methods

6 years agogstbuffer: add GST_BUFFER_FLAG_NON_DROPPABLE
Mathieu Duponchelle [Tue, 30 Jan 2018 15:41:39 +0000 (16:41 +0100)]
gstbuffer: add GST_BUFFER_FLAG_NON_DROPPABLE

This can be used to identify buffers for which a higher percentage
of redundancy should be allocated when performing forward error
correction, or to prevent still video frames from being dropped by
elements due to QoS.

https://bugzilla.gnome.org/show_bug.cgi?id=793008

6 years agomeson: use -fno-strict-aliasing if supported
Tim-Philipp Müller [Tue, 30 Jan 2018 20:30:47 +0000 (20:30 +0000)]
meson: use -fno-strict-aliasing if supported

https://bugzilla.gnome.org/show_bug.cgi?id=769183