Commit graph

136 commits

Author SHA1 Message Date
Tim-Philipp Müller
f325935314 pulse, speexenc, rtpgsmpay: don't use g_assert() for error handling
Don't use g_assert() for error handling, even if they're highly unlikely.
Either we *know* that something can't happen, in which case we
should just not handle it, or we think something can happen, but it is
very very unlikely that it will ever happen, in which case we should
handle it like any other error instead of asserting.

g_assert() is best left for conditions we have control of, like checking
internal consistency of our code, not checking return values of external
code.

Fixes a bunch of warnings when compiling with -DG_DISABLE_ASSERT:
gstrtpgsmpay.c: In function 'gst_rtp_gsm_pay_handle_buffer':
gstrtpgsmpay.c:130:17: warning: variable 'rtpgsmpay' set but not used
gstspeexenc.c: In function 'gst_speex_enc_encode':
gstspeexenc.c:904:19: warning: variable 'written' set but not used
pulsesink.c: In function 'gst_pulsesink_change_state':
pulsesink.c:2725:9: warning: variable 'res' set but not used
pulsesrc.c: In function 'gst_pulsesrc_change_state':
pulsesrc.c:1253:7: warning: variable 'e' set but not used
2011-04-16 18:15:43 +01:00
Robert Swain
098cf89e34 jpegenc: pulsesink: raw1394: Address unused but set variables
GCC 4.6.x spits warnings about such usage of variables. The variables in
raw1394 were marked with G_GNUC_UNUSED as this seemed omre appropriate.
The others were removed.
2011-04-16 13:13:04 +01:00
Sebastian Dröge
11bcac7c90 Revert "Pulsesink: Allow chunks up to bufsize instead of segsize"
This reverts commit 1e2c1467ae.

The commit causes pulsesink to ignore the latency-time baseaudiosink property.
2011-04-08 14:35:04 +02:00
Arun Raghavan
dc48eaac13 pulsesink: Always call pa_stream_new_with_proplist()
pa_stream_new_with_proplist() can take a NULL proplist, so we don't need
to concern ourselves with whether it's NULL or not.
2011-04-04 17:23:21 +05:30
Mark Nauwelaerts
176b8ffbff pulsesink: also uncork during EOS waiting (and after EOS is rendered)
Pulsesink was recently changed to defer uncorking until there is data
to write. This condition will however never occur when EOS in being
rendered (since that marks the end of data). Changing to PAUSED state
while EOS is being waited on results in a hang: pausing corks the
stream, which will never be undone since there is no more data when
going back to PLAYING. If pulsesink is the clock provider, deadlock
ensues since time doesn't continue in corked state and the clock id
for EOS wait never fires.

Fixes #645961.
2011-03-31 13:25:19 +02:00
Philip Jägenstedt
84decd731f pulsesink: Better fix for deadlock on failed connect
This reverts the previous fix that would cause a double-unlock when the
stream connect failed.

https://bugzilla.gnome.org/show_bug.cgi?id=644510
2011-03-12 00:48:03 +05:30
Arun Raghavan
387f533a07 pulsesink: Fix deadlock if connecting to PA fails
Commit dd4ec22e introduced a deadlock in the failure path while trying
to connect to PulseAudio. This makes sure we drop the lock on the
resource mutex to avoid this.

https://bugzilla.gnome.org/show_bug.cgi?id=644510
2011-03-11 23:19:34 +05:30
Tim-Philipp Müller
14e1d9de3d docs: fix pulsesink gtk-doc markup 2011-03-11 09:54:36 +00:00
Stefan Kost
89914439b7 pulsesink: add a doc example for setting stream-properties 2011-03-04 15:50:49 +02:00
Stefan Kost
edfa99ef31 pulsesink: fix the xml in the docs 2011-03-04 15:50:49 +02:00
Philip Jägenstedt
dd4ec22e7e pulsesink: release pa_shared_resource_mutex before pa_threaded_mainloop_wait
Not doing so can result in a deadlock when two threads enter
gst_pulseringbuffer_open_device at the same time, as
pa_threaded_mainloop_wait releases the mainloop lock while waiting,
allowing another thread to take it, resulting in a deadlock as two
threads waits for the lock the other is holding.

https://bugzilla.gnome.org/show_bug.cgi?id=643087
2011-02-28 16:20:22 +05:30
Philip Jägenstedt
683c96b37a pulsesink: s/ressource/resource/
https://bugzilla.gnome.org/show_bug.cgi?id=643087
2011-02-28 16:20:20 +05:30
David Henningsson
1e2c1467ae Pulsesink: Allow chunks up to bufsize instead of segsize
By allowing larger chunks to be sent, PulseAudio will have a
lower CPU usage. This is especially important on low-end machines,
where PulseAudio can crash if packets are coming in at a higher
rate than PulseAudio can process them.

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
2011-01-31 17:15:05 +01:00
Arun Raghavan
e18e25a324 pulsesink: Uncork stream while flushing the ringbuffer
After starting the ringbuffer, we wait for enough data to arrive before
uncorking the stream. This will cause the pipeline to stall if we get an
EOS (or otherwise need to flush the stream) before sufficient data
becomes available. This patch makes sure that the stream is uncorked
while flushing to avoid this problem.

Fixes issue with a webkit unit test testing reverse playback of
an MP4 H.264/AAC file.

https://bugzilla.gnome.org/show_bug.cgi?id=639740
2011-01-17 15:40:12 +00:00
Arun Raghavan
7280ea1f68 pulsesink: Make corking during pause synchronous
This makes the call to pa_stream_cork() during ringbuffer pause()
synchronous, which makes sure that the clock does not advance after we
take a snapshot for start_time.

https://bugzilla.gnome.org/show_bug.cgi?id=639240
2011-01-11 20:43:54 +00:00
Mark Nauwelaerts
53b9f87436 pulsesink: flush remaining buffered samples on EOS
... which can make a difference between all or nothing when dealing
with short streams and relatively large ringbuffer segment.
2010-12-16 15:16:54 +01:00
Wim Taymans
8ca094795a pulsesink: don't uncork in _start
Don't uncork in the _start method just yet but wait until we have written some
samples to pulseaudio. This avoid underruns on pulseaudio and less crackling
noises when starting.
2010-12-07 12:39:27 +01:00
Wim Taymans
7f48fb37e9 pulse: fix device_description in READY
Make the is_dead check more clear and add an option to check for the status of
the stream in addition to the context.
We don't need a stream to get the device_description string.

Fixes #630317
2010-09-22 19:02:35 +02:00
Stefan Kost
ae8d210fdb pulsesink: move the shared mainloop from class to static var
Just have one static var for the shared mainloop instead of one class variable
and copies in the instance.
2010-09-14 10:10:27 +03:00
Stefan Kost
f62dc6976b pulsesink: rework context sharing
We also need to share the main-loop threads as this owns the context. Thus have
a class wide main-loop thread. From this we create a context per client-name.
Instead of always looking up the context, we keep this with the instance. The
reverse mapping is only needed in pulse singal handlers. This saves a lot of
locking. Also one signal handler becomes simpler as ther eis only one mainloop
to notify.

Now valgind happy - no leaks, no bad reads/writes.

This reverts major parts of commit 69a397c32f.

Fixes #628996
2010-09-13 16:59:22 +03:00
Stefan Kost
6490c4eb3d pulsesink: remove unused code 2010-09-08 17:43:47 +03:00
Stefan Kost
9fd0adc7dc pulsesink: fixup last commit
We need to prevent the eventual leak better.
2010-09-08 14:36:48 +03:00
Stefan Kost
f6892b81ae pulsesink: code cleanups
Use g_slist_prepend as we don't care about the order. Check for list == NULL
instead of iterating the list to see if it is empty. Move ctx allocation down
to prevent leak in case of failure.
2010-09-08 14:19:34 +03:00
Wim Taymans
2148b2db0a pulsesink: simplify clock provide code
Don't leak the pulsesink element by having the clock keep a ref to the sink.
Create the clock only once in the constructor and use the baseaudiosink clock
cleanup code.
2010-09-07 19:03:33 +02:00
Stefan Kost
7fd77c07d5 pulsesink: move the context table init to _get_type phase
This seems to fix the invalid reads on context shutdown better, altough
I can't really explain.
2010-09-07 18:09:28 +03:00
Stefan Kost
f604e20499 pulse: allow setting stream properties
Add a "properties" property to the elements to allow setting extra stream
properties.

Fixes #537544
2010-09-07 14:20:21 +03:00
Stefan Kost
a313bc7524 pulsesink: don't free the context multiple times
Apparently the close function of the ring-buffer can be called multiple times.
2010-09-07 00:27:07 +03:00
Stefan Kost
fc73dd210d pulsesink: add since docs for new property. 2010-09-06 09:57:58 +03:00
Sebastian Dröge
37ee281d13 pulsesink: Fail gracefully if no threaded PA mainloop can be created
Fixes bug #628020.
2010-09-04 14:52:09 +02:00
Sebastian Dröge
998adeb974 pulsesink: Only include the server name in the context name if it's not NULL 2010-09-04 14:52:07 +02:00
Philippe Normand
7116fff3c5 pulsesink: Add "client" property to set the PA client name
Allows the application to modify the client name used to connect when
connecting to the PulseAudio daemon. Note however that updating the
property after the element reached the READY state will have no
effect until the next NULL->READY transition.

Fixes bug #627174.
2010-09-04 14:52:07 +02:00
Sebastian Dröge
03f2e81f69 pulsesink: Free the clock on state change failures too 2010-09-04 14:52:07 +02:00
Sebastian Dröge
bc5345bbd3 pulsesink: Free the GstPulseContext after usage 2010-09-04 14:52:06 +02:00
Philippe Normand
69a397c32f pulsesink: share the PA context between all clients with the same name
Avoid to create a new PA context for each new client by using a hash
table containing the list of ring-buffers and the shared PA context
for each client. Doing this will improve application memory usage in
the cases where multiple pipelines involving multiple pulsesink
elements are used.

Fixes bug #624338.
2010-09-04 14:52:06 +02:00
Philippe Normand
2c5976d355 pulsesink: clear the PA mainloop if baseaudiosink failed to open the ring_buffer
If the application requests a state-change and pulsesink fails to open
the ring_buffer device the mainloop attribute of the sink should be
cleaned up to avoid future state-change (NULL->READY) failures.
2010-09-04 14:52:06 +02:00
Tim-Philipp Müller
74b2af63fd pulse: fix printf format in some debugging messages 2010-08-09 00:36:36 +01:00
Philippe Normand
864a52d8aa pulsesink: use G_TYPE_DEFINE to define ring buffer type
The existing get_type() implementation is racy, and the
g_type_class_ref() workaround didn't actually work because
it was in the wrong function. Since class creation in GObject
is thread-safe these days (since 2.16), the class_ref workaround
is no longer needed and it is sufficient to ensure the _get_type()
function is thread-safe, which G_TYPE_DEFINE does.

https://bugzilla.gnome.org/show_bug.cgi?id=624338
2010-08-04 16:19:09 +01:00
Sebastian Dröge
f6240a4773 pulsesink: Post CLOCK-LOST/CLOCK-PROVIDE when going to/from READY
Otherwise the clocks are redistributed every time the pipeline
goes to PAUSED, which is quite expensive.
2010-08-04 15:21:26 +02:00
Tim-Philipp Müller
7055264e2e pulsesink: correctly check what version of gst-plugins-base we're compiling against
We need to check the gst-plugins-base version, not the core version
(even if both should be the same in any sane setup).
2010-07-26 18:18:02 +02:00
Sebastian Dröge
10424a4064 pulsesink: Only use gst_audio_clock_new() when compiling against newer base 2010-07-16 17:53:55 +02:00
Sebastian Dröge
9e92bd7b1a pulsesink: Post clock-provide/clock-lost when going to/from PAUSED
Also use gst_audio_clock_new_full() to prevent crashes when the
clock is used after the element was destroyed.
2010-07-16 17:45:38 +02:00
Sebastian Dröge
4113a1dad1 pulsesink: Don't request more shared memory than needed 2010-06-09 20:53:06 +02:00
Sebastian Dröge
bd5987a3d2 pulsesink: Flush shm buffer immediately if it's full 2010-06-09 16:22:27 +02:00
Sebastian Dröge
57813d469f pulsesink: Fix writing of buffers larger than segsize
Fixes bug #620540.
2010-06-09 16:21:55 +02:00
Sebastian Dröge
b78a27beb2 pulsesink: Fix playback if PA doesn't give us a large enough shared memory buffer 2010-06-09 15:46:59 +02:00
Tim-Philipp Müller
3486f49808 pulse: log message printf format fixes 2010-06-06 17:25:36 +01:00
Sebastian Dröge
dcae55852e ext: Don't use GST_DEBUG_FUNCPTR for GObject vfuncs 2010-06-06 18:00:22 +02:00
Sebastian Dröge
5072c47624 Revert "pulsesink: Add comments to remove the provide-clock message posting once we depend on base 0.10.30"
This reverts commit 8f3708f38a.

The baseaudiosink commit was reverted
2010-06-03 13:45:32 +02:00
Sebastian Dröge
8f3708f38a pulsesink: Add comments to remove the provide-clock message posting once we depend on base 0.10.30
baseaudiosink does all this for us now.
2010-06-03 10:28:08 +02:00
Pierre-Louis Bossart
66a76d1f65 pulsesink: optimize communication with PulseAudio using pa_stream_begin_write 2010-06-02 13:47:06 +02:00