Stefan Sauer [Mon, 20 Feb 2012 19:39:59 +0000 (20:39 +0100)]
seek: fix format strings
Sebastian Dröge [Mon, 20 Feb 2012 15:44:07 +0000 (16:44 +0100)]
seek: Add support for sending navigation commands
Sebastian Dröge [Mon, 20 Feb 2012 14:29:49 +0000 (15:29 +0100)]
seek: Make the seek-bar insensitive for DVD menus
Sebastian Dröge [Mon, 20 Feb 2012 13:57:36 +0000 (14:57 +0100)]
seek: Add support for the navigation interface
Sebastian Dröge [Mon, 20 Feb 2012 13:57:11 +0000 (14:57 +0100)]
seek: Keep a reference to the xoverlay element
Sebastian Dröge [Mon, 20 Feb 2012 13:35:37 +0000 (14:35 +0100)]
seek: Add window-embedding support for OS X/Quartz
osxvideosink implements the GstXOverlay interface since some time now.
Sebastian Dröge [Mon, 20 Feb 2012 13:09:28 +0000 (14:09 +0100)]
seek: Explicitely call gst_x_overlay_expose() from the GtkDrawingArea expose/draw signal
David Schleef [Sun, 19 Feb 2012 08:03:03 +0000 (00:03 -0800)]
videoscale: fix AYUV64 scaling
David Schleef [Sun, 19 Feb 2012 01:37:29 +0000 (17:37 -0800)]
riff: Add v210, r210 formats
Tim-Philipp Müller [Fri, 17 Feb 2012 15:08:36 +0000 (15:08 +0000)]
docs: add new audio base class API to docs and .def file
Ognyan Tonchev [Mon, 30 Jan 2012 14:55:26 +0000 (15:55 +0100)]
rtspconnection: only send new data immediately if there are no queued messages
Even if watch->messages->length is 0 there may still be some
data from a message that was only written partially at the
previous attempt stored in watch->write_data, so check for
that as well. We don't want to write data into the middle
of another message, which could happen when there wasn't
enough bandwidth.
https://bugzilla.gnome.org/show_bug.cgi?id=669039
Mark Nauwelaerts [Thu, 16 Feb 2012 11:19:20 +0000 (12:19 +0100)]
audiodecoder: add some properties to tweak baseclass behaviour
... so subclass can also rely upon never being bothered with some NULL buffer
it can't do any interesting with, or with any data before it received
any format configuration (and setup properly).
Mark Nauwelaerts [Thu, 16 Feb 2012 11:18:03 +0000 (12:18 +0100)]
audioencoder: add some properties to tweak baseclass behaviour
... so subclass can also rely upon never being bothered with less data
than it desires or with some NULL buffer it can't do any interesting with.
Mark Nauwelaerts [Thu, 16 Feb 2012 11:15:47 +0000 (12:15 +0100)]
audiodecoder: assert some more that subclass parsed frame has proper len
Olivier Aubert [Tue, 17 Jan 2012 09:54:48 +0000 (10:54 +0100)]
docs: fix playbin2 documentation about DVD URIs
and playbin => playbin2 in example pipelines.
https://bugzilla.gnome.org/show_bug.cgi?id=668081
Tim-Philipp Müller [Tue, 14 Feb 2012 19:23:27 +0000 (19:23 +0000)]
discoverer: try harder to obtain a duration if we don't get one right away
If we don't get a duration right away, set the pipeline to playing
and sleep a bit, then try again. This is ugly, but the least worst
we can do right now. The alternative would be to make parsers etc.
return some bogus duration estimate even after only having pushed
a single frame, for example.
Fixes discoverer showing 0 durations for some mp3 and aac files
(e.g. soweto-adts.aac).
Tim-Philipp Müller [Wed, 8 Feb 2012 19:39:00 +0000 (19:39 +0000)]
typefindfunctions: make h264 typefinder more picky when returning "likely" probability
Only return LIKELY probability if we've seen an SPS, PPS and an
IDR slice nal, i.e. try harder to avoid false positives such
as with certain VC-1 files.
https://bugzilla.gnome.org/show_bug.cgi?id=668565
Tim-Philipp Müller [Wed, 8 Feb 2012 19:34:57 +0000 (19:34 +0000)]
typefindfunctions: minor cosmetic change
Don't write < 1 when we mean == 0.
David Schleef [Mon, 30 Jan 2012 16:21:54 +0000 (08:21 -0800)]
videoscale: Add nearest/linear scaling for NV12
David Schleef [Wed, 25 Jan 2012 23:49:00 +0000 (15:49 -0800)]
videoscale: Add AYUV64 path to Lanczos
David Schleef [Wed, 31 Aug 2011 02:02:51 +0000 (19:02 -0700)]
theoraenc: Use GAP flag when possible
Set TH_ENCCTL_SET_DUPLICATE_FLAG when we see a gap flag, to
indicate to the encoder that the current frame is a duplicate
of the previous frame.
Ryan Lortie [Sun, 8 Jan 2012 04:09:23 +0000 (23:09 -0500)]
autogen.sh: allow calling from out-of-tree
https://bugzilla.gnome.org/show_bug.cgi?id=667665
Vincent Penquerc'h [Wed, 1 Feb 2012 15:28:45 +0000 (15:28 +0000)]
oggdemux: fix granpos interpolation violating max keyframe distance
In case many packets fit on a page, we may not see a granpos for
a while, and granpos interpolation can wrap the 'frames since last
keyframe' part of the granpos, generating a granpos which is smaller
than what it should be.
This is fixed by detecting keyframe packets (at least for Theora),
and updating the last keyframe granpos from this.
This may still be generating potentially wrong granpos for streams
which have a Theora like granpos (keyframes, a max keyframe distance
and a count of frames since last keyframe), and which allow implicit
granules on packets. For these streams, a custom keyframe detection
routine should be plugged into their GstOggStream mapper.
https://bugzilla.gnome.org/show_bug.cgi?id=669164
Vincent Penquerc'h [Wed, 1 Feb 2012 16:46:13 +0000 (16:46 +0000)]
vorbisparse: pedantically recognize undefined headers too
Vincent Penquerc'h [Wed, 1 Feb 2012 16:32:24 +0000 (16:32 +0000)]
vorbisparse: fix header detection
It was matching non header packets.
This fixes various leaks, where buffers would be pushed onto a headers
list, but never popped.
Might also fix corruption as those buffers were dropped from the output
silently...
https://bugzilla.gnome.org/show_bug.cgi?id=669167
George Kiagiadakis [Tue, 10 Jan 2012 19:17:58 +0000 (21:17 +0200)]
tests: test 16-bit rgb formats in test_parse_caps_rgb
https://bugzilla.gnome.org/show_bug.cgi?id=667681
George Kiagiadakis [Tue, 10 Jan 2012 19:02:48 +0000 (21:02 +0200)]
video: Use host endianness when generating caps for 16-bit rgb formats
This is necessary in order to match what the caps strings in
video.h contain for 16-bit rgb formats and also to match how
gst_video_format_parse_caps expects them.
https://bugzilla.gnome.org/show_bug.cgi?id=667681
Vincent Penquerc'h [Fri, 27 Jan 2012 13:52:30 +0000 (13:52 +0000)]
v4l: include the glib compatiblity header for the deprecated mutex API
Edward Hervey [Wed, 25 Jan 2012 11:50:44 +0000 (12:50 +0100)]
Suppress deprecations in selected files
Sebastian Dröge [Wed, 25 Jan 2012 11:25:05 +0000 (12:25 +0100)]
Revert "decodebin2: Prune old groups before switching to the new one"
This reverts commit
e2a038acee2969ed0b558093fa1c8b7422073e40.
This wasn't entirely correct yet and needs some changes here
and there.
David Schleef [Mon, 23 Jan 2012 17:28:18 +0000 (09:28 -0800)]
propertyprobe: fix documentation
Vincent Penquerc'h [Wed, 18 Jan 2012 14:58:08 +0000 (14:58 +0000)]
playbin2: do not try to deactivate an inactive group
A group may have failed to activate due to an error (for instance,
having set the URI to a non existent location in about-to-finish).
https://bugzilla.gnome.org/show_bug.cgi?id=666395
Anssi Hannula [Tue, 17 Jan 2012 14:05:41 +0000 (16:05 +0200)]
subtitleoverlay: fix state change stall on PAUSED->READY->PAUSED
After a PAUSED->READY change the sink pads are currently not set to
blocking state. When the element is set back to PAUSED, the change will
be done asynchronously, but as the _pad_blocked_cb() callback is now not
called, the state change never completes.
Fix that by setting the sink pads to blocking state on a PAUSED->READY
change, which ensures that the _pad_blocked_cb() is called when needed
on any future READY->PAUSED change. The sink pads are already put to
blocking state on NULL->READY change, so this behavior is consistent.
Fixes bug #668097.
Mark Nauwelaerts [Thu, 19 Jan 2012 15:40:22 +0000 (16:40 +0100)]
streamsynchronizer: avoid unlikely NULL dereference
Mark Nauwelaerts [Thu, 19 Jan 2012 15:35:54 +0000 (16:35 +0100)]
videoscale: prevent implicit upgrade to integer type and sign extension
Mark Nauwelaerts [Thu, 19 Jan 2012 15:35:04 +0000 (16:35 +0100)]
gst-discoverer: remove extraneous variable
Mark Nauwelaerts [Thu, 19 Jan 2012 15:32:37 +0000 (16:32 +0100)]
playsink: verify linking to overlay element
Mark Nauwelaerts [Thu, 19 Jan 2012 15:32:05 +0000 (16:32 +0100)]
playsink: avoid finding sink in NULL bin in corner case
Mark Nauwelaerts [Thu, 19 Jan 2012 15:29:53 +0000 (16:29 +0100)]
tag: exif: add missing break
Tim-Philipp Müller [Thu, 19 Jan 2012 14:07:34 +0000 (14:07 +0000)]
Add --disable-fatal-warnings configure option
Youness Alaoui [Tue, 13 Sep 2011 23:14:10 +0000 (23:14 +0000)]
decodebin2: Prune old groups before switching to the new one
In order to allow for proper functionality when a decoder only supports
one instance at a time (dsp), we must block the demuxer pads when they
get created if they are not part of the active group, preventing buffers
from being sent to the decoder (and initializing it through setcaps),
then after we switch to a new group, we unblock the demuxer pads for
the active groups. In the callback for the unblock, we prune the old
groups, making sure the previous decoder instance is destroyed before
we push a buffer to the new instance.
Mark Nauwelaerts [Tue, 17 Jan 2012 20:46:58 +0000 (21:46 +0100)]
baseaudiosink: commit correct number of samples when not syncing
Mark Nauwelaerts [Tue, 17 Jan 2012 17:19:30 +0000 (18:19 +0100)]
oggstream: initialize variable
... to help out challenged compiler.
Mark Nauwelaerts [Tue, 17 Jan 2012 10:52:49 +0000 (11:52 +0100)]
audiodecoder: register state change function
Vincent Penquerc'h [Mon, 16 Jan 2012 11:43:25 +0000 (11:43 +0000)]
alsasink: fix high sample rates being rejected
An ALSA sink may select a different rate (as we use the _set_rate_near
API, which is not guaranteed to set the exact target rate).
The rest of the code seems to already handle this well, as output
from a 88200 Hz file seems to have the correct pitch when selecting
a 96 kHz rate.
Vincent Penquerc'h [Mon, 16 Jan 2012 11:40:47 +0000 (11:40 +0000)]
alsasink: fix rate match message mistaking error code for sample rate
Vincent Penquerc'h [Mon, 16 Jan 2012 11:40:16 +0000 (11:40 +0000)]
alsasink: log API errors along with the error code and string
Reynaldo H. Verdejo Pinochet [Fri, 13 Jan 2012 19:57:15 +0000 (16:57 -0300)]
Android, Add explicit path for zlib
This change fixes building gst-libs/gst/tag/ code with
the Android buildsystem.
Reynaldo H. Verdejo Pinochet [Fri, 13 Jan 2012 17:50:49 +0000 (14:50 -0300)]
Fix wrong access to undefined struct member
For the USE_TREMOLO case, GstVorbisDec doesn't have
a vb member. Besides, Tremolo's vorbis_dsp_synthesis()
expects a vorbis_dsp_state to be passed as first
argument. Not a vorbis_block.
Reynaldo H. Verdejo Pinochet [Fri, 13 Jan 2012 17:47:13 +0000 (14:47 -0300)]
Fix TREMELO -> TREMOLO typo
Vincent Penquerc'h [Fri, 13 Jan 2012 16:52:23 +0000 (16:52 +0000)]
xvimagesink: fix leak when images are freed after the X context
I'm not 100% sure this is valid on any other X server than mine,
but since the XFree call does not take the context as a parameter,
it seems pretty certain it's the right thing to do, but I'll put
this caveat here in case someone checks in the future.
Tim-Philipp Müller [Thu, 12 Jan 2012 17:31:44 +0000 (17:31 +0000)]
tests: discoverer test is now valgrind clean
Vincent Penquerc'h [Thu, 12 Jan 2012 16:24:01 +0000 (16:24 +0000)]
theoraparse: fix array leak
Vincent Penquerc'h [Thu, 12 Jan 2012 14:26:05 +0000 (14:26 +0000)]
discoverer: fix structure leak
I hit the 'misc' one, but let's also make sure the topology
one get freed as well, though I do not know if this can happen
twice.
Reynaldo H. Verdejo Pinochet [Wed, 11 Jan 2012 23:47:00 +0000 (20:47 -0300)]
Add missing DEFAULT_INCLUDES on androgenizer call
Fix building of the libgstvideo module on Android by adding the
missing and needed $(DEFAULT_INCLUDES) to CFLAGS for the
androgenizer call on gst-libs/gst/video/Makefile.am
Before this change, building was failing due to gst-plugins-base/
and gst-plugins-base/gst-libs/gst/video being left out of the
include path.
Vincent Penquerc'h [Wed, 11 Jan 2012 16:17:42 +0000 (16:17 +0000)]
oggdemux: fix push mode chain leak
When I first implemented push mode seeking, I removed the chain
freeing there as it could be used later. The current code does not
seem to do that though, so I'm restoring the previous freeing,
which plugs the leak while apparently not reintroducing use of
freed data with chained and normal files, both with gst-launch
playbin2 and Totem.
Vincent Penquerc'h [Wed, 11 Jan 2012 12:52:17 +0000 (12:52 +0000)]
discoverer: fix leaks caused by some base class dtors not being called
Vincent Penquerc'h [Wed, 11 Jan 2012 12:16:28 +0000 (12:16 +0000)]
discoverer: fix caps and discoverer object ref leaks
Vincent Penquerc'h [Wed, 11 Jan 2012 11:55:59 +0000 (11:55 +0000)]
discoverer: add a few consts where appropriate
Vincent Penquerc'h [Wed, 11 Jan 2012 11:55:36 +0000 (11:55 +0000)]
discoverer: fix pad leak
Vincent Penquerc'h [Tue, 10 Jan 2012 19:01:11 +0000 (19:01 +0000)]
Revert "oggmux: fix pad leak"
This reverts commit
5df30c1b905edce16f2258e414a0a4afb540d0f1.
I must have dreamt the Valgrind logs, reverting this reintroduces
no leak, and gets rid of the test failures it introduced :S
Tim-Philipp Müller [Tue, 10 Jan 2012 18:27:19 +0000 (18:27 +0000)]
discoverer: use GST_TYPE_TAG_LIST for tag lists
They may not be structures in 0.11/1.0.
Tim-Philipp Müller [Tue, 10 Jan 2012 18:07:19 +0000 (18:07 +0000)]
discoverer: fix potential tag list leaks
Not that I have ever seen these in practice, but if they
can't happen we may just as well just assign the new tag
list. Merge properly to be on the safe side, and also
avoid a useless tag list copy in the normal case where
there is no tag list yet.
Tim-Philipp Müller [Tue, 10 Jan 2012 17:48:44 +0000 (17:48 +0000)]
discoverer: fix potential caps leak
in last else chunk.
Vincent Penquerc'h [Tue, 10 Jan 2012 16:57:04 +0000 (16:57 +0000)]
oggstream: fix tag list leak
Vincent Penquerc'h [Tue, 10 Jan 2012 16:51:09 +0000 (16:51 +0000)]
oggdemux: fix pad leak
Vincent Penquerc'h [Tue, 10 Jan 2012 16:14:29 +0000 (16:14 +0000)]
oggdemux: fix hang on small truncated files
A first hang was happening when trying to locate a page backwards,
where we'd sync forever on the same page.
With that fixed, a second hang would happen after preparing an EOS
event, but with no chain created yet to send it to, the pipeline
would stay idle forever.
An element error is now emitted for this case.
Vincent Penquerc'h [Tue, 10 Jan 2012 14:35:31 +0000 (14:35 +0000)]
oggmux: fix pad leak
Mark Nauwelaerts [Mon, 9 Jan 2012 11:31:02 +0000 (12:31 +0100)]
playback: document DEINTERLACE flag
Tim-Philipp Müller [Sat, 7 Jan 2012 20:12:17 +0000 (20:12 +0000)]
rtspconnection: make hostname lookup more thread-safe
Don't write IP number string to return into a static
array which is shared amongst all threads (note: of
course a copy is returned).
https://bugzilla.gnome.org/show_bug.cgi?id=666711
Tim-Philipp Müller [Sat, 7 Jan 2012 19:39:42 +0000 (19:39 +0000)]
discoverer: make is_subtitle_caps thread-safe
Tim-Philipp Müller [Sat, 7 Jan 2012 16:43:26 +0000 (16:43 +0000)]
tests: add ogg test file and some proper unit tests for discoverer
Leaks when re-used, so blacklisted for valgrind for now.
Tim-Philipp Müller [Sat, 7 Jan 2012 14:44:51 +0000 (14:44 +0000)]
win32: .def file should be sorted for make check-exports
Vincent Penquerc'h [Wed, 14 Dec 2011 14:14:47 +0000 (14:14 +0000)]
video: overlays may now have premultiplied alpha
https://bugzilla.gnome.org/show_bug.cgi?id=666177
Havard Graff [Tue, 1 Nov 2011 16:57:59 +0000 (17:57 +0100)]
Fix various unlikely, but still potential memoryleaks in error code paths
https://bugzilla.gnome.org/show_bug.cgi?id=667311
Havard Graff [Sat, 22 Oct 2011 14:41:23 +0000 (16:41 +0200)]
appsrc: implement get_caps vfunc
This allows downstream elements to query what caps are available.
https://bugzilla.gnome.org/show_bug.cgi?id=667312
Tim-Philipp Müller [Thu, 5 Jan 2012 12:23:08 +0000 (12:23 +0000)]
tools: avoid unportable vararg macro construct in gst-discoverer
https://bugzilla.gnome.org/show_bug.cgi?id=667306
Idar Tollefsen [Sun, 1 Jan 2012 19:44:08 +0000 (20:44 +0100)]
build: Run platform check for platform specific configuration.
Pascal Buhler [Wed, 12 Oct 2011 09:28:10 +0000 (11:28 +0200)]
rtcpbuffer: prevent overflow of 16bit header length.
RTCP header can be (2^16 + 1) * 4 bytes long, so when validating a bogus
packet it was possible to get a 16bit overflow resulting in a length of 0.
This would put the gst_rtcp_buffer_validate_data function in a endless loop.
https://bugzilla.gnome.org/show_bug.cgi?id=667313
Havard Graff [Sat, 24 Sep 2011 12:05:42 +0000 (14:05 +0200)]
videotestsrc: keep the calculation fixed-point
https://bugzilla.gnome.org/show_bug.cgi?id=667315
Idar Tollefsen [Thu, 4 Aug 2011 09:30:05 +0000 (11:30 +0200)]
pango: changes includes from brackets to quotes for local files
https://bugzilla.gnome.org/show_bug.cgi?id=667316
Philip Flarsheim [Thu, 5 Jan 2012 01:51:35 +0000 (01:51 +0000)]
docs: add win32 code snippets to GstXOverlay Gtk+ example
Tim-Philipp Müller [Tue, 3 Jan 2012 21:20:04 +0000 (21:20 +0000)]
videotestsrc: don't build generate_sine_table utility by default
Mark Nauwelaerts [Tue, 3 Jan 2012 10:04:23 +0000 (11:04 +0100)]
streamsynchronizer: force fallback buffer_alloc when other pad not available
... to avoid unnecessary spurious errors (upon e.g. shutdown).
If a real error is applicable in this unusual circumstance (missing other pad),
other (STREAM_LOCK protected) call paths can take care of that.
Mark Nauwelaerts [Tue, 3 Jan 2012 10:02:17 +0000 (11:02 +0100)]
streamsynchronizer: avoid crashing when operating on released pad
Tim-Philipp Müller [Mon, 2 Jan 2012 00:59:39 +0000 (00:59 +0000)]
tag: add function to check whether a string is a valid language code
API: gst_tag_check_language_code()
Tim-Philipp Müller [Tue, 20 Dec 2011 21:48:29 +0000 (21:48 +0000)]
audio, rtsp: remove private/protected gtk-doc markup for enums
This confuses glib-mkenums, and is not really useful anyway.
https://bugzilla.gnome.org/show_bug.cgi?id=666618
Thiago Santos [Tue, 27 Dec 2011 17:37:26 +0000 (14:37 -0300)]
oggmux: fix leak when initializing pads
Pads are initialized twice: when requesting pads and when
initializing collectpads. Avoid double initialization by
checking if collectpads are still going to be initialized when
creating request pads.
Tim-Philipp Müller [Fri, 23 Dec 2011 22:51:59 +0000 (22:51 +0000)]
theoraenc: fix template caps creation on big endian systems
Tim-Philipp Müller [Fri, 23 Dec 2011 22:24:44 +0000 (22:24 +0000)]
tag: fix writing of Exif tag payloads <= 4 bytes
When the payload for an Exif tag is less than or equal to 4 bytes,
the data is simply put into the offset field. Fix writing these
kinds of payloads on big endian systems (and possibly also on
little endian systems). The caller will have already formatted
the bytes in memory according to the writer's endianness, so just
write out the bytes as they are in this case. Fixes tags unit test
on big endian systems.
Mark Nauwelaerts [Thu, 22 Dec 2011 15:54:18 +0000 (16:54 +0100)]
audioencoder: add a few more debug statements
Mark Nauwelaerts [Thu, 22 Dec 2011 15:53:49 +0000 (16:53 +0100)]
audiodecoder: tweak documentation
Thiago Santos [Thu, 22 Dec 2011 10:53:39 +0000 (07:53 -0300)]
tag: xmp: Keep compatibility with our old generated xmp
We used to add a trailing \n to the end of generated xmp packets.
Windows viewer was unhappy with it and we fixed it in
96d2120c2bb0b29e1849098198f5fbef81939cdd
The problem is that this caused xmp generated before this fix
to not be recognized and parsed anymore. This patch makes it
recognize xmp with the trailing \n and without, fixing the
regression. Also adds tests for it.
Vincent Penquerc'h [Wed, 14 Dec 2011 16:34:39 +0000 (16:34 +0000)]
gstvideo: fix a RGB ordering mixup in colorspace conversion code
Oleksij Rempel (Alexey Fisher) [Mon, 19 Dec 2011 16:41:23 +0000 (17:41 +0100)]
theoraenc: add "dup-on-gap" option
This option will produce duplicate frames if we get
a frame with GAP flag. This will reduce CPU load and file size.
This option should be disabled for real time applications, because it
collects GAP frames and waits until it gets a non GAP frame to start
encoding.
v30.06.2011: make some spell changes.
v03.07.2011: add handling of EOS and discontinuous for dup-on-gap.
v19.12.2011: fix pointer dangling in theora_timefifo_free
v20.12.2010: fix timestamp bug for dup-on-gap=0
Bugzilla: https://bugzilla.gnome.org/show_bug.cgi?id=627459
Signed-off-by: Oleksij Rempel (Alexey Fisher) <bug-track@fisher-privat.net>
Mark Nauwelaerts [Tue, 20 Dec 2011 11:42:18 +0000 (12:42 +0100)]
audiodecoder: set a non-zero default maximum tolerated errors
Whereas the previous default 0 was backwards compatible in that it lead
to erroring out immediately upon any error, elements that are really
ported and using the base class error macro can be assumed to intend to
improve behaviour rather than maintaining the old one. So, make it easy
on those and any future one and tolerate some errors by default, as intended.
Fixes #666579.
Vincent Penquerc'h [Fri, 16 Dec 2011 17:32:41 +0000 (17:32 +0000)]
adder: do not send too many flush-stop events
GstCollectPads2 now allows us to override the event function,
so we can withhold flush stop events if none are to be sent.
https://bugzilla.gnome.org/show_bug.cgi?id=666379
Vincent Penquerc'h [Fri, 16 Dec 2011 17:31:06 +0000 (17:31 +0000)]
adder: use the stream lock where appropriate
GstCollectPads2 locking was changed from GstCollectPads to use
the stream lock instead of the object lock for those cases, so
change it so here as well to match.
https://bugzilla.gnome.org/show_bug.cgi?id=666379
Vincent Penquerc'h [Fri, 16 Dec 2011 17:25:10 +0000 (17:25 +0000)]
adder: send a flush event before trying to get the stream lock
This avoids hanging when the streaming thread is busy in _chain
waiting for preroll.
https://bugzilla.gnome.org/show_bug.cgi?id=666379
Vincent Penquerc'h [Fri, 16 Dec 2011 15:27:24 +0000 (15:27 +0000)]
oggdemux: assume live stream if byte size cannot be determined
This prevents trying to seek and failing, then ending up unable
to stream because we can't get back at the headers.
A more robust way would be to find a good place to reinject the
headers when a seek fails, but I can't seem to get this to work.