First we ignore request to fill the ringbuffer which are less then a segment.
The small request where causing stutter.
Then we disable flushing the stream when running against pa 0.9.12 as this
triggers an assertiong in the sound server and terminates it. It does not happen
with 0.9.10 and 0.9.14.
We can use prebuf = 0 to instruct pulse to not pause the stream on underflows.
This way we can remove the underflow callback. We however have to manually
uncork the stream now when we have no available space in the buffer or when we
are writing too far away from the current read_index.
when we switch streams, the clock will reset to 0. Make sure that the provided
clock doesn't get stuck when this happens by keeping an initial offset. We also
need to make sure that we subtract this offset in samples when writing to the
ringbuffer.
If the caps changes, the sink is reset without transitioning through
a PAUSED->PLAYING state change, resulting in a corked stream. This avoids
the problem by checking that the stream is uncorked when writing samples
to it.
When trying to write out a segment, wait until there is enough free space
for the entire segment. This helps to reduce ripple in the clock reporting,
where the app might query the playback position while only half a segment
has been written (and is therefore reported by _delay(), even though
the ring buffer has not yet been advanced)
In the event handler, gst_flac_dec_sink_event(), two functions are called on
the FLAC stream without checking if it has been initialized:
FLAC__stream_decoder_flush()
FLAC__stream_decoder_process_until_end_of_stream()
Both these FLAC__*() functions modify the internal state of the FLAC stream.
Later, when the buffers start flowing, gst_flac_dec_chain() tries to initialize
the stream. the FLAC__stream_decoder_init_stream() call will fail because the
previous calls to FLAC__*() changed the stream state so it is no longer in the
initialized state.
The flacdec API calls the write callback when performing a seek. We cannot yet
push out a buffer at that time so we must keep it and push it out later.
Flush out the upstream part of the pipeline when doing a seek.
Fixes#574275.
g_atomic_int_(get|set) only work on ints and the flags are
an enum (which on most architectures is stored as an int).
Also the way the flags were accessed atomically would still
leave a possible race condition and we don't do it in any
other mixer track implementation, let alone at any other
place where an integer could be changed from different
threads. Removing the g_atomic_int_(get|set) will only
introduce a new race condition on architectures where
integers could be half-written while reading them
which shouldn't be the case for any modern architecture
and if we really care about this we need to use
g_atomic_int_(get|set) at many other places too.
Apart from that g_atomic_int_(set|get) will result in
aliasing warnings if their argument is explicitely
casted to an int *. Fixes bug #571153.
rather than PA thread.
pa_threaded_mainloop_lock() (a.o.) and by extension get_property should
not be done from a PA thread, but the latter may occur as a result of a
property change notification. Fixes#571204 (though current situation
not ideal, e.g. post message rather than signal).
newer pulseaudio.
Fixes: #567794
* Hook pulsesink's volume property up with the stream volume -- not the
sink volume in PA.
* Read the device description directly from the sink instead of going
via the mixer.
* Properly implement _reset() methods for both sink and source to avoid
deadlocks when shutting down a pipeline.
* Replace all simple pa_threaded_mainloop_wait() by proper loops to
guarantee that we wait for the right event in case multiple events are
fired. While this is not strictly necessary in many cases it
certainly is more correct and makes me sleep better at night.
* Replace CHECK_DEAD_GOTO macros with proper functions
* Extend the number of supported channels to 32 since that is the actual
limit in PA.
* Get rid of _dispose() methods since we don't need them.
* Increase the volume property upper limit of the sink to 1000.
* Reset function pointers after we disconnect a stream/context. Better
fix for bug 556986.
* Reset the state of the element properly if open/prepare fails
* Cork the PA stream when the pipeline is paused. This allows the PA
* daemon to
close audio device on pause and thus save a bit of power.
* Set PA stream properties based on GST tags such as GST_TAG_TITLE,
GST_TAG_ARTIST, and so on.
Signed-off-by: Lennart Poettering <lennart@poettering.net>
Link to properties. Correct titles for examples. Document a few trivial cases. Keep lists in section file and docs/plugins/Makefile.am alphabetically ordered. Fix warnings that gtk-doc points out.
If libsoup-gnome is found use this as it will give us
the GNOME proxy configuration. Otherwise use normal
libsoup.
The GNOME proxy configuration will only be used if
the proxy properties are not set on souphttpsrc
and if the http_proxy environment variable is not
set.
Fixes bug #552140.
Original commit message from CVS:
Patch by: Lennart Poettering <lennart at poettering dot net>
* ext/pulse/pulseprobe.c: (gst_pulseprobe_new),
(gst_pulseprobe_free):
Fix refcount loop, resulting in a thread leak. Fixes bug #567746.
Original commit message from CVS:
* ext/pulse/pulsesink.c:
* ext/pulse/pulsesink.h:
Use a mutex to protect the current stream pointer, and ignore
callbacks for stream objects that have been destroyed already.
Fixes problems with unprepare/prepare cycles caused by the input
caps changing, without reintroducing bug #556986.
Original commit message from CVS:
* ext/pulse/pulsesink.c: (gst_pulsesink_destroy_stream):
Don't wait for the pulse mainloop when destroying the stream.
Fixes a deadlock when the pulsedaemon goes away while pulsesink
is PLAYING. Fixes bug #556986.
Original commit message from CVS:
* ext/jpeg/gstsmokeenc.c: (gst_smokeenc_init),
(gst_smokeenc_getcaps), (gst_smokeenc_setcaps),
(gst_smokeenc_chain), (gst_smokeenc_change_state):
* ext/jpeg/gstsmokeenc.h:
Implement getcaps function.
Set caps on the pad and on all outgoing buffers.
Fixes#565441.
Original commit message from CVS:
* ext/pulse/pulsemixerctrl.c:
And remove temporary comment pointing to the bug ticket.
* gst/avi/gstavimux.c:
Move reoccuring logging to LOG and log instance too.
Original commit message from CVS:
* ext/dv/gstdvdemux.c: (gst_dvdemux_handle_src_event):
Restore previous behaviour of not passing QoS and navigation
events upstream, which presumably wasn't meant to be changed.
Original commit message from CVS:
* ext/dv/gstdvdemux.c: (gst_dvdemux_add_video_pad),
(gst_dvdemux_add_audio_pad), (gst_dvdemux_remove_pads),
(gst_dvdemux_demux_audio), (gst_dvdemux_demux_video),
(gst_dvdemux_chain), (gst_dvdemux_loop),
(gst_dvdemux_change_state):
Add srcpads only when needed and remove them again when going
back to READY. This prevents stalled pipelines if there's no
audio inside the DV stream, which happens for many MXF files.
Original commit message from CVS:
* ext/dv/gstdvdemux.c: (gst_dvdemux_handle_src_event):
Forward all events upstream unless it's something we really
don't handle. This fixes latency configuration of pipelines.
Original commit message from CVS:
* ext/dv/gstdv.c: (plugin_init):
* ext/dv/gstdvdec.c: (gst_dvdec_class_init):
* ext/dv/gstdvdemux.c: (gst_dvdemux_class_init):
Really call dv_init() exactly one time, not one time for
the demuxer and one time for the decoder.
Original commit message from CVS:
Patch by: Zeeshan Ali <zeeshanak at gnome dot org>
* ext/soup/gstsouphttpsrc.c: (gst_soup_http_src_build_message):
Add transferMode.dnla.org header to HTTP requests as this is
required by the DLNA specs and doesn't hurt in other situations.
Fixes bug #561802.
Original commit message from CVS:
* ext/libpng/gstpngenc.c:
Don't flush downstream after every buffer - that's not what
this libpng callback is for at all!
Original commit message from CVS:
* ext/flac/Makefile.am:
Include $(FLAC_CFLAGS) in CFLAGS to make sure to find the FLAC headers.
This fixes compilation if FLAC is installed in an uncommon location
that is not already handled by other CFLAGS. Fixes bug #558711.
Original commit message from CVS:
Patch by: Sjoerd Simons <sjoerd at luon dot net>
* ext/soup/gstsouphttpsrc.c: (gst_soup_http_src_class_init),
(gst_soup_http_src_set_property), (gst_soup_http_src_get_property):
Add support for souphttpsrc to act as a live source. This makes it
possible to get timestamped buffers in combination with the
"do-timestamp" property. Fixes bug #556019.
Original commit message from CVS:
* ext/flac/gstflacdec.c (gst_flac_dec_read_stream):
* ext/flac/gstflacenc.c (gst_flac_enc_write_callback):
Cast some size_t arguments to guint to avoid compiler
warnings on 64-bit systems.
Original commit message from CVS:
* ext/pulse/pulsesink.c: (gst_pulsesink_write):
* ext/pulse/pulsesrc.c: (gst_pulsesrc_read):
Return -1 instead of 0 in error cases. Fixes#554771.
Original commit message from CVS:
* ext/pulse/pulsesink.c:
Fix problems with pulsesink randomly erroring with code 'OK' after a
format change on the stream by waiting when disconnecting the stream.
Original commit message from CVS:
* ext/raw1394/gstdv1394src.c: (SEND_COMMAND):
* ext/raw1394/gsthdv1394src.c: (SEND_COMMAND):
Pretend to care about the result of write() which works around
compiler warnings.
Original commit message from CVS:
* ext/flac/gstflacenc.c: (gst_flac_enc_class_init):
Make sure the desired default values are actually set, not only
registered as defaults (actual problem is that the stereo-specific
values are only updated if channels==2, which is not the case yet
when the object is created, so the default values for the
mid-side-stereo and loose-mid-side-stereo settings are never
set in _update_quality()). Makes flacenc create smaller files by
default (for stereo input), and fixes#550791.
Original commit message from CVS:
* ext/speex/gstspeexdec.c: (speex_dec_chain_parse_data):
* ext/speex/gstspeexdec.h:
* ext/speex/gstspeexenc.c: (gst_speex_enc_encode):
* ext/speex/gstspeexenc.h:
Use integer encoding and decoding functions instead of converting
the integer input to float in the element. The libspeex integer
functions are doing this for us already or, if libspeex was compiled
in integer mode, they're doing everything using integer arithmetics.
Also saves some copying around.
Original commit message from CVS:
* ext/wavpack/gstwavpackenc.c: (gst_wavpack_enc_reset),
(gst_wavpack_enc_push_block), (gst_wavpack_enc_chain):
* ext/wavpack/gstwavpackenc.h:
Handle non-zero start timestamps and stream discontinuities
correctly. This only has an effect if we're muxing into
a container format as the raw WavPack stream must contain
continous sample numbers.
Original commit message from CVS:
* ext/speex/gstspeexenc.c: (gst_speex_enc_encode):
Correct the timestamp and granulepos calculation by one Speex
frame.
Original commit message from CVS:
* ext/speex/gstspeexdec.c: (speex_dec_chain_parse_data):
Correctly take the granulepos from upstream if possible and
correctly handle the granulepos in various calculations: the
granulepos is the sample number of the _last_ sample in a frame, not
the first.
* ext/speex/gstspeexenc.c: (gst_speex_enc_sinkevent),
(gst_speex_enc_encode), (gst_speex_enc_chain),
(gst_speex_enc_change_state):
* ext/speex/gstspeexenc.h:
Handle non-zero start timestamps in the encoder and detect/handle
stream discontinuities. Fixes bug #547075.
Original commit message from CVS:
Patch by: Craig Keogh <cskeogh at adam dot com dot au>
* ext/annodex/gstcmmlparser.c: (gst_cmml_parser_parse_chunk):
Fix compiler warnings caused by passing a string as format string
instead of "%s" and then the string. This is only exposed by -Wformat=2
as used by default on Ubuntu. Fixes bug #550015.
Original commit message from CVS:
* ext/pulse/pulsesrc.c: (gst_pulsesrc_class_init),
(gst_pulsesrc_create_stream), (gst_pulsesrc_negotiate),
(gst_pulsesrc_prepare):
* ext/pulse/pulseutil.c: (gst_pulse_gst_to_channel_map),
(gst_pulse_channel_map_to_gst):
* ext/pulse/pulseutil.h:
If downstream provides no channel layout and >2 channels should be
used use the default layout that pulseaudio chooses and also
add this layout to the caps. Fixes bug #547258.
Original commit message from CVS:
* ext/pulse/pulsesink.c: (gst_pulsesink_prepare):
* ext/pulse/pulsesrc.c: (gst_pulsesrc_prepare):
The bytes_per_sample and silence_sample fields of the GstRingBufferSpec
are already filled with the correct values by
gst_ring_buffer_parse_caps() so there's no need to set them again
with wrong values.
Original commit message from CVS:
* ext/pulse/pulsesink.c: (gst_pulsesink_class_init),
(gst_pulsesink_init), (gst_pulsesink_finalize),
(gst_pulsesink_set_volume), (gst_pulsesink_get_volume),
(gst_pulsesink_set_property), (gst_pulsesink_get_property),
(gst_pulsesink_prepare), (gst_pulsesink_change_state):
* ext/pulse/pulsesink.h:
Add "device-name" property to pulsesink too and currently commented
out and not working support for a "volume" property.
Original commit message from CVS:
Patch by: Laszlo Pandy <laszlok2 at gmail dot com>
* ext/pulse/pulsesrc.c: (gst_pulsesrc_class_init),
(gst_pulsesrc_get_property):
Add "device-name" property, which provides a human readable string
for the audio device, to make it more consisten with other audio
sources. Fixes bug #547519.
Original commit message from CVS:
* ext/pulse/pulsemixer.c: (gst_pulsemixer_change_state):
* ext/pulse/pulsemixerctrl.c: (gst_pulsemixer_ctrl_subscribe_cb),
(gst_pulsemixer_ctrl_open), (gst_pulsemixer_ctrl_new),
(gst_pulsemixer_ctrl_free), (gst_pulsemixer_ctrl_timeout_event):
* ext/pulse/pulsemixerctrl.h:
* ext/pulse/pulseprobe.c: (gst_pulseprobe_open),
(gst_pulseprobe_enumerate), (gst_pulseprobe_new),
(gst_pulseprobe_free), (gst_pulseprobe_needs_probe),
(gst_pulseprobe_probe_property), (gst_pulseprobe_get_values):
* ext/pulse/pulseprobe.h:
* ext/pulse/pulsesink.c: (gst_pulsesink_init):
* ext/pulse/pulsesrc.c: (gst_pulsesrc_init), (gst_pulsesrc_delay),
(gst_pulsesrc_change_state):
Improve debugging a bit by including the parent object in pulsemixerctrl
and pulseprobe objects and using GST_WARNING_OBJECT instead of
GST_WARNING.
Use the parent GObject subclass instead of a random struct as GObject
parameter for G_OBJECT_WARN_INVALID_PROPERTY_ID. This fixes a crash
when probing for another property than "device".
Original commit message from CVS:
Patch by: Laszlo Pandy <laszlok2 at gmail dot com>
* ext/pulse/pulsemixer.c: (gst_pulsemixer_set_property):
Fix property probing after the device property is set by calling
set_server when the server property changes. Fixes bug #547518.
Original commit message from CVS:
Patch by: Laszlo Pandy <laszlok2 at gmail dot com>
* ext/pulse/pulsemixer.c: (gst_pulsemixer_set_property):
Fix property probing after the device property is set by calling
set_server when the server property changes. Fixes bug #547518.
