platform/upstream/gstreamer.git
11 years agoBack to development (bug-fixing)
Tim-Philipp Müller [Sun, 7 Oct 2012 15:48:25 +0000 (16:48 +0100)]
Back to development (bug-fixing)

11 years agoRelease 1.0.1 1.0.1
Tim-Philipp Müller [Sun, 7 Oct 2012 12:10:33 +0000 (13:10 +0100)]
Release 1.0.1

11 years agotests: update struct_i386.h for ABI checks
Tim-Philipp Müller [Sat, 6 Oct 2012 23:15:49 +0000 (00:15 +0100)]
tests: update struct_i386.h for ABI checks

Fixes make check on 32-bit x86.

11 years agotests: update struct_ppc32.h for ABI checks
Tim-Philipp Müller [Sat, 6 Oct 2012 16:26:21 +0000 (17:26 +0100)]
tests: update struct_ppc32.h for ABI checks

Fixes make check on 32-bit PowerPC.

11 years agoAutomatic update of common submodule
Tim-Philipp Müller [Sat, 6 Oct 2012 13:55:35 +0000 (14:55 +0100)]
Automatic update of common submodule

From 6c0b52c to 6bb6951

11 years agoexamples: .gitignore more binaries from the manual
Tim-Philipp Müller [Sat, 6 Oct 2012 11:08:34 +0000 (12:08 +0100)]
examples: .gitignore more binaries from the manual

11 years agodocs: remove obsolete part-block document
Wim Taymans [Fri, 5 Oct 2012 14:04:52 +0000 (16:04 +0200)]
docs: remove obsolete part-block document

Merge the part-block document into part-probes

11 years agopad: resend dropped events
Wim Taymans [Fri, 5 Oct 2012 07:42:59 +0000 (09:42 +0200)]
pad: resend dropped events

If we try to push sticky events but a probe dropped them, we don't mark
the event as received and mark the pad as PENDING_EVENTS. This ensures
that we resend the event the next time. For this we need to let the
custom flow return from the probe trickle up to
gst_pad_push_event_unchecked() so that we can differentiate between
OK and DROPPED probe returns.

11 years agopad: don't store sticky events on flushing/EOS pads
Wim Taymans [Fri, 5 Oct 2012 05:14:17 +0000 (07:14 +0200)]
pad: don't store sticky events on flushing/EOS pads

Don't store sticky events on flushing or EOS pads. This was done
correctly for source pads but not for sink pads.

11 years agodocs: add Since markers for new API and add it to docs and .def file
Tim-Philipp Müller [Thu, 4 Oct 2012 10:24:09 +0000 (11:24 +0100)]
docs: add Since markers for new API and add it to docs and .def file

11 years agomanual: add dynamic capsfilter example
Wim Taymans [Thu, 4 Oct 2012 09:50:58 +0000 (11:50 +0200)]
manual: add dynamic capsfilter example

11 years agocapsfilter: don't prefer passthrough
Wim Taymans [Thu, 4 Oct 2012 09:18:19 +0000 (11:18 +0200)]
capsfilter: don't prefer passthrough

Basetransform should not try to negotiate in passthrough mode but
respect the order of what we return in the transform_caps method.
A typical case is that you specify some specific new caps in the
caps property but also allow the current caps to pass.

11 years agobasetrans: add an option to prefer passthrough
Wim Taymans [Thu, 4 Oct 2012 09:15:36 +0000 (11:15 +0200)]
basetrans: add an option to prefer passthrough

Basetransform attempts to do passthrough mode regardless of the order of
the transform_caps method. Add a method to disable this.
This is needed for elements like capsfilter that want to transform caps
based on the order of the caps property.

11 years agobasetrans: improve some comments
Wim Taymans [Thu, 4 Oct 2012 08:01:19 +0000 (10:01 +0200)]
basetrans: improve some comments

11 years agomanual: talk some more about dynamic pipelines
Wim Taymans [Wed, 3 Oct 2012 15:17:02 +0000 (17:17 +0200)]
manual: talk some more about dynamic pipelines

11 years agometa: don't put essential logic in g_return_val_*
Wim Taymans [Wed, 3 Oct 2012 11:49:34 +0000 (13:49 +0200)]
meta: don't put essential logic in g_return_val_*

11 years agometa: do metadata registration threadsafe
Wim Taymans [Wed, 3 Oct 2012 11:45:22 +0000 (13:45 +0200)]
meta: do metadata registration threadsafe

We need to use g_once to register the metadata implementations
only once.

See https://bugzilla.gnome.org/show_bug.cgi?id=685332

11 years agometa: handle multiple implementation registration
Wim Taymans [Wed, 3 Oct 2012 11:35:20 +0000 (13:35 +0200)]
meta: handle multiple implementation registration

First check that we can actually register the implementation before
making a GstMetaInfo. If we can't register we would otherwise end
up with an undefined type and an invalid GstMetaInfo.

It's possible that type registration fails because another metadata
with the same implementation name was already registered.

11 years agomanual: use CDATA for code blocks
Wim Taymans [Wed, 3 Oct 2012 11:12:44 +0000 (13:12 +0200)]
manual: use CDATA for code blocks

then we don't have to escape special token anymore.

11 years agomanual: add partial preroll example with probes
Wim Taymans [Wed, 3 Oct 2012 11:09:00 +0000 (13:09 +0200)]
manual: add partial preroll example with probes

11 years agomanual: add more stuff about probes
Wim Taymans [Wed, 3 Oct 2012 08:53:20 +0000 (10:53 +0200)]
manual: add more stuff about probes

11 years agomanual: start talking about dynamic pipeline changes
Wim Taymans [Tue, 2 Oct 2012 15:23:28 +0000 (17:23 +0200)]
manual: start talking about dynamic pipeline changes

11 years agomanual: move section around
Wim Taymans [Tue, 2 Oct 2012 14:47:05 +0000 (16:47 +0200)]
manual: move section around

11 years agopwg: add appsink docs
Wim Taymans [Tue, 2 Oct 2012 14:44:28 +0000 (16:44 +0200)]
pwg: add appsink docs

11 years agopwg: rewite data-access chapter
Wim Taymans [Tue, 2 Oct 2012 14:15:19 +0000 (16:15 +0200)]
pwg: rewite data-access chapter

Rewrite the data-access chapter so that we talk about appsrc instead
of the fakesrc hacks.

11 years agodocs: some 0.10 -> 1.0 changes
Wim Taymans [Tue, 2 Oct 2012 11:22:35 +0000 (13:22 +0200)]
docs: some 0.10 -> 1.0 changes

11 years agopwg: add allocation query example
Wim Taymans [Tue, 2 Oct 2012 11:12:39 +0000 (13:12 +0200)]
pwg: add allocation query example

11 years agopwg: add bufferpool docs
Wim Taymans [Tue, 2 Oct 2012 10:49:17 +0000 (12:49 +0200)]
pwg: add bufferpool docs

11 years agopwg: flesh out allocation docs
Wim Taymans [Tue, 2 Oct 2012 09:34:47 +0000 (11:34 +0200)]
pwg: flesh out allocation docs

Add more examples.
Add example for implementing new metadata.
Add programs to the docs (again?), it seems to contain useful info.

11 years agopwg: add new author
Wim Taymans [Mon, 1 Oct 2012 14:59:41 +0000 (16:59 +0200)]
pwg: add new author

11 years agopwg: add allocation docs
Wim Taymans [Mon, 1 Oct 2012 14:55:55 +0000 (16:55 +0200)]
pwg: add allocation docs

11 years agodocs: update design docs
Wim Taymans [Mon, 1 Oct 2012 14:46:03 +0000 (16:46 +0200)]
docs: update design docs

11 years agodocs: more docs fixes
Wim Taymans [Mon, 1 Oct 2012 11:28:54 +0000 (13:28 +0200)]
docs: more docs fixes

Fix allocator design doc
Add beginning of allocation chapter in the pwg

11 years agopwg: final cleanups for 1.0
Wim Taymans [Mon, 1 Oct 2012 09:47:46 +0000 (11:47 +0200)]
pwg: final cleanups for 1.0

11 years agopwg: fix events and base classes
Wim Taymans [Mon, 1 Oct 2012 09:24:52 +0000 (11:24 +0200)]
pwg: fix events and base classes

11 years agopwg: fixup tag docs
Wim Taymans [Mon, 1 Oct 2012 08:40:54 +0000 (10:40 +0200)]
pwg: fixup tag docs

11 years agopwg: patch up the section about interfaces
Wim Taymans [Mon, 1 Oct 2012 07:48:48 +0000 (09:48 +0200)]
pwg: patch up the section about interfaces

11 years agobasesrc: Fix seamless segment function
Jan Schmidt [Sat, 29 Sep 2012 18:05:36 +0000 (04:05 +1000)]
basesrc: Fix seamless segment function

The 3rd parameter of gst_base_src_new_seamless_segment in
0.10 is the time associated with the start of the new segment,
not the position in the new segment. Fix the name of the parameter,
the docs, and the implementation to match the needs of the only
extant consumer: DVD playback.

11 years agovalue: avoid duplicates when intersecting lists
Tim-Philipp Müller [Sat, 29 Sep 2012 13:35:58 +0000 (14:35 +0100)]
value: avoid duplicates when intersecting lists

Fixes negotiation taking a ridiculous amount of
time (multiple 10s of seconds on a core2) when
there are duplicate entries in lists.

Could have a negative performance impact on other
scenarios because we now have to iterate the
dest list to avoid duplicates, but we don't
have a lot of lists any more these days, and
they tend to be small anyway. The negatives
are hopefully countered by the positive effects
of reducing the list length early on in the
process. And in any case, it's the right thing
to do.

Based on patch by Andre Moreira Magalhaes.

https://bugzilla.gnome.org/show_bug.cgi?id=684981

11 years agopwg: minor update
Tim-Philipp Müller [Fri, 28 Sep 2012 23:27:03 +0000 (00:27 +0100)]
pwg: minor update

https://bugzilla.gnome.org/show_bug.cgi?id=621121

11 years agofaq: add missing </para> tag
Wim Taymans [Fri, 28 Sep 2012 21:53:53 +0000 (23:53 +0200)]
faq: add missing </para> tag

11 years agominiobject: Always reject WRITE locks on READONLY miniobjects
Olivier Crête [Fri, 28 Sep 2012 19:17:27 +0000 (15:17 -0400)]
miniobject: Always reject WRITE locks on READONLY miniobjects

Verify that mapping a read-only memory as read doesnt make it writable

11 years agodocs: purge all mention of liboil, update FAQ
Tim-Philipp Müller [Fri, 28 Sep 2012 19:38:20 +0000 (20:38 +0100)]
docs: purge all mention of liboil, update FAQ

https://bugzilla.gnome.org/show_bug.cgi?id=673285

11 years agopwg: update for 1.0
Wim Taymans [Fri, 28 Sep 2012 14:03:15 +0000 (16:03 +0200)]
pwg: update for 1.0

Rewrite clock part.
start on interfaces

11 years agopwg: rework dynamic pads docs
Wim Taymans [Fri, 28 Sep 2012 11:25:49 +0000 (13:25 +0200)]
pwg: rework dynamic pads docs

11 years agopwg: rework scheduling docs
Wim Taymans [Fri, 28 Sep 2012 11:25:30 +0000 (13:25 +0200)]
pwg: rework scheduling docs

11 years agopwg: remove some GST_BOILERPLATE
Wim Taymans [Fri, 28 Sep 2012 11:24:52 +0000 (13:24 +0200)]
pwg: remove some GST_BOILERPLATE

11 years agodocs: update activation design docs
Wim Taymans [Fri, 28 Sep 2012 09:18:11 +0000 (11:18 +0200)]
docs: update activation design docs

11 years agopad: fix activate docs
Wim Taymans [Fri, 28 Sep 2012 08:41:54 +0000 (10:41 +0200)]
pad: fix activate docs

11 years agopwg: fix more negotiation for 1.0
Wim Taymans [Fri, 28 Sep 2012 08:04:51 +0000 (10:04 +0200)]
pwg: fix more negotiation for 1.0

11 years agoinfo: do not register printf extension for %p
Olivier Blin [Thu, 27 Sep 2012 14:59:04 +0000 (16:59 +0200)]
info: do not register printf extension for %p

This happened when glib was not using system printf, and caused the
internal gstreamer printf extensions to be used for all %p printfs,
causing crashes.

https://bugzilla.gnome.org/show_bug.cgi?id=684970

11 years agopwg: fix some negotiation to 1.0
Wim Taymans [Thu, 27 Sep 2012 15:21:53 +0000 (17:21 +0200)]
pwg: fix some negotiation to 1.0

11 years agopwg: more updates for 1.0
Wim Taymans [Thu, 27 Sep 2012 12:42:07 +0000 (14:42 +0200)]
pwg: more updates for 1.0

11 years agopwg: more updates for 1.0
Wim Taymans [Thu, 27 Sep 2012 11:57:46 +0000 (13:57 +0200)]
pwg: more updates for 1.0

11 years agopwg: update boiler to 1.0
Wim Taymans [Thu, 27 Sep 2012 09:53:36 +0000 (11:53 +0200)]
pwg: update boiler to 1.0

11 years agoghostpad: also ref the internal pad for activate functions
Wim Taymans [Thu, 27 Sep 2012 09:06:06 +0000 (11:06 +0200)]
ghostpad: also ref the internal pad for activate functions

Also take a ref to the internal pad in the activate functions

11 years agoproxypad: Hold a reference to the internal pad while pushing through it
Olivier Crête [Mon, 24 Sep 2012 22:26:16 +0000 (18:26 -0400)]
proxypad: Hold a reference to the internal pad while pushing through it

https://bugzilla.gnome.org/show_bug.cgi?id=684809

11 years agotests: Test the case where ghost pads are removed while streaming
Olivier Crête [Tue, 25 Sep 2012 18:44:54 +0000 (14:44 -0400)]
tests: Test the case where ghost pads are removed while streaming

https://bugzilla.gnome.org/show_bug.cgi?id=684809

11 years agotests: enable library abi checks
Wim Taymans [Thu, 27 Sep 2012 07:44:07 +0000 (09:44 +0200)]
tests: enable library abi checks

11 years agodocs: fix up basesrc/basesink docs formatting
Tim-Philipp Müller [Wed, 26 Sep 2012 22:32:35 +0000 (23:32 +0100)]
docs: fix up basesrc/basesink docs formatting

11 years agotests: add abi checks
Wim Taymans [Wed, 26 Sep 2012 15:08:30 +0000 (17:08 +0200)]
tests: add abi checks

Enable abi checks again.
Fix abi sizes for x86_64, copy the file to other archs.

11 years agoupdate docs for 1.0 API
Wim Taymans [Wed, 26 Sep 2012 14:26:48 +0000 (16:26 +0200)]
update docs for 1.0 API

11 years agouri: use proper 'transfer floating' annotation
Mark Nauwelaerts [Wed, 26 Sep 2012 12:15:01 +0000 (14:15 +0200)]
uri: use proper 'transfer floating' annotation

https://bugzilla.gnome.org/show_bug.cgi?id=664099

11 years agotypefind: send STREAM-START event
Wim Taymans [Wed, 26 Sep 2012 11:19:13 +0000 (13:19 +0200)]
typefind: send STREAM-START event

Send a STREAM_START event when we are operating in pull mode.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=684424

11 years agosegment: mark GstSegmentFlags as flags rather than enum
Mark Nauwelaerts [Wed, 26 Sep 2012 08:55:28 +0000 (10:55 +0200)]
segment: mark GstSegmentFlags as flags rather than enum

... which really makes a difference when trying to serialize
a flags value which is a combination of flags, which is hard
to do as an enum type.

11 years agoidentity: retimestamp both pts and dts when doing so
Mark Nauwelaerts [Wed, 26 Sep 2012 08:54:06 +0000 (10:54 +0200)]
identity: retimestamp both pts and dts when doing so

11 years agobaseparse: Move some run of the mill debug statements to LOG level
Jan Schmidt [Wed, 26 Sep 2012 05:01:42 +0000 (15:01 +1000)]
baseparse: Move some run of the mill debug statements to LOG level

11 years agobaseparse: Output timestamps after a seek.
Jan Schmidt [Wed, 26 Sep 2012 04:23:52 +0000 (14:23 +1000)]
baseparse: Output timestamps after a seek.

Reinitialise the DTS after a seek so as to continue
generating timestamps when baseparse is not downstream
of a demuxer.

Fixes: #684538

11 years agodocs: updates
Wim Taymans [Tue, 25 Sep 2012 15:06:45 +0000 (17:06 +0200)]
docs: updates

MIME-type -> Media type
Fix some old gst-inspect output

11 years agopwg: update for 1.0 API
Wim Taymans [Tue, 25 Sep 2012 14:53:08 +0000 (16:53 +0200)]
pwg: update for 1.0 API

11 years agodocs: add section for metadata
Wim Taymans [Tue, 25 Sep 2012 13:11:33 +0000 (15:11 +0200)]
docs: add section for metadata

11 years agoelementfactory: Fail if no valid element factory metadata is set
Sebastian Dröge [Tue, 25 Sep 2012 11:09:38 +0000 (13:09 +0200)]
elementfactory: Fail if no valid element factory metadata is set

11 years agoplugin: Fail if no valid plugin metadata is set
Sebastian Dröge [Tue, 25 Sep 2012 11:09:28 +0000 (13:09 +0200)]
plugin: Fail if no valid plugin metadata is set

11 years agoidentity: also track and store segment info in single segment mode
Mark Nauwelaerts [Tue, 25 Sep 2012 13:06:43 +0000 (15:06 +0200)]
identity: also track and store segment info in single segment mode

11 years agomanual: fix up the manual
Wim Taymans [Tue, 25 Sep 2012 12:40:20 +0000 (14:40 +0200)]
manual: fix up the manual

MIME-type -> media types
Fix up the manual in various places with the 1.0 way of doing things
such as probes, static elements, scheduling, ...
Add porting from 0.10 to 1.0 chapter.
Add probe example to build.
Remove some docs for remove components such as GstMixer and
GstPropertyProbe, XML...

11 years agodocs: gst-python is no more
Wim Taymans [Mon, 24 Sep 2012 14:50:50 +0000 (16:50 +0200)]
docs: gst-python is no more

gst-python is no more and gst-libav is one of the main modules that
we release.

11 years agodocs: fix basesink docs
Wim Taymans [Mon, 24 Sep 2012 14:31:34 +0000 (16:31 +0200)]
docs: fix basesink docs

11 years agodocs: update FAQ
Wim Taymans [Mon, 24 Sep 2012 14:25:36 +0000 (16:25 +0200)]
docs: update FAQ

Change versions.
Use tools with version prefix.

11 years agopo: update translations for typo fix
Tim-Philipp Müller [Tue, 25 Sep 2012 12:15:14 +0000 (13:15 +0100)]
po: update translations for typo fix

11 years agotaglist: fix typo in translated string
Tim-Philipp Müller [Tue, 25 Sep 2012 12:14:53 +0000 (13:14 +0100)]
taglist: fix typo in translated string

Spotted by Chris Leonard.

https://bugzilla.gnome.org/show_bug.cgi?id=684755

11 years agopluginfeature: Remove 0.11.9X->1.0.0 version mangling
Sebastian Dröge [Tue, 25 Sep 2012 07:27:59 +0000 (09:27 +0200)]
pluginfeature: Remove 0.11.9X->1.0.0 version mangling

11 years agobenchmarks: printf format fixes to make intel compiler happy
Josep Torra Valles [Tue, 25 Sep 2012 00:02:03 +0000 (01:02 +0100)]
benchmarks: printf format fixes to make intel compiler happy

https://bugzilla.gnome.org/show_bug.cgi?id=552657

11 years agoMake intel compiler happier
Josep Torra Valles [Mon, 24 Sep 2012 23:55:59 +0000 (00:55 +0100)]
Make intel compiler happier

https://bugzilla.gnome.org/show_bug.cgi?id=552657

11 years agoBack to development (bug fixing)
Tim-Philipp Müller [Mon, 24 Sep 2012 15:31:54 +0000 (16:31 +0100)]
Back to development (bug fixing)

11 years agoRelease 1.0.0 1.0.0
Tim-Philipp Müller [Mon, 24 Sep 2012 11:19:16 +0000 (12:19 +0100)]
Release 1.0.0

11 years agodocs: update 0.11 references in porting guide
Tim-Philipp Müller [Sun, 23 Sep 2012 23:39:26 +0000 (00:39 +0100)]
docs: update 0.11 references in porting guide

11 years agodocs: rename porting-to-0.11.txt to porting-to-1.0.txt
Tim-Philipp Müller [Sun, 23 Sep 2012 23:37:27 +0000 (00:37 +0100)]
docs: rename porting-to-0.11.txt to porting-to-1.0.txt

11 years agocheck: fix FIXME printing for tcase_skip_broken_test()
Tim-Philipp Müller [Sun, 23 Sep 2012 18:56:43 +0000 (19:56 +0100)]
check: fix FIXME printing for tcase_skip_broken_test()

11 years agodocs: update release doc
Tim-Philipp Müller [Sun, 23 Sep 2012 16:30:50 +0000 (17:30 +0100)]
docs: update release doc

Create tags for releases without the ugly RELEASE- prefix.

11 years agocollectpads: don't forward random stream-start event
Tim-Philipp Müller [Sun, 23 Sep 2012 11:42:01 +0000 (12:42 +0100)]
collectpads: don't forward random stream-start event

It's not right, and we don't know what extra properties
that event might have set in future (e.g. sparseness).
This change means collectpad users need to create their
own stream-start event now. We could add a utility
function that creates a stream-start event based on
the input stream-start events.

11 years agoAutomatic update of common submodule
Tim-Philipp Müller [Sat, 22 Sep 2012 15:07:15 +0000 (16:07 +0100)]
Automatic update of common submodule

From 4f962f7 to 6c0b52c

11 years agomanual: update controller documentation
Mark Nauwelaerts [Fri, 21 Sep 2012 19:13:27 +0000 (21:13 +0200)]
manual: update controller documentation

11 years agoobject: update controller documentation
Mark Nauwelaerts [Fri, 21 Sep 2012 19:13:13 +0000 (21:13 +0200)]
object: update controller documentation

11 years agoman: Fix syntax for value lists in caps strings
Bastian Winkler [Tue, 18 Sep 2012 13:22:03 +0000 (15:22 +0200)]
man: Fix syntax for value lists in caps strings

Value lists use curly brackets instead of parentheses

https://bugzilla.gnome.org/show_bug.cgi?id=684293

11 years agopad: Remove pad probes only once
Olivier Crête [Thu, 20 Sep 2012 18:48:17 +0000 (14:48 -0400)]
pad: Remove pad probes only once

Also add test to make sure that if a pad probe is removed while it's
callback is running, the cleanup_hook isn't called again if it
returns GST_PAD_PROBE_REMOVE

11 years agopad: Add functions to safely access GstProbeInfo data pointer
Olivier Crête [Wed, 19 Sep 2012 19:01:46 +0000 (15:01 -0400)]
pad: Add functions to safely access GstProbeInfo data pointer

This is so that introspection based bindings can access it.

https://bugzilla.gnome.org/show_bug.cgi?id=684402

11 years agodocs: remove reference to 0.8 GstBin API from manual
Tim-Philipp Müller [Wed, 19 Sep 2012 22:25:54 +0000 (23:25 +0100)]
docs: remove reference to 0.8 GstBin API from manual

https://bugzilla.gnome.org/show_bug.cgi?id=684048

11 years agoidentity: transform GAP event in single segment mode
Mark Nauwelaerts [Wed, 19 Sep 2012 13:14:35 +0000 (15:14 +0200)]
identity: transform GAP event in single segment mode

11 years agodocs: collectpads doc fixes
Tim-Philipp Müller [Wed, 19 Sep 2012 08:44:08 +0000 (09:44 +0100)]
docs: collectpads doc fixes

11 years agobasetransform: check acquire result value
Wim Taymans [Tue, 18 Sep 2012 19:49:41 +0000 (21:49 +0200)]
basetransform: check acquire result value

Check the result value from _buffer_pool_acquire() and return the
value when allocation failed.

Fixes https://bugzilla.gnome.org/show_bug.cgi?id=684285