profile/ivi/pulseaudio.git
12 years agopacmd: dynamically allocate ibuf and obuf
Pino Toscano [Mon, 14 Nov 2011 11:07:12 +0000 (12:07 +0100)]
pacmd: dynamically allocate ibuf and obuf

Use pa_pipe_buf to determine the minimum size for ibuf and obuf, taking into
account the two descriptors that use each of them.

See bug #42715

12 years agomutex: handle gracefully if a PTHREAD_PRIO_INHERIT protocol cannot be set
Pino Toscano [Mon, 14 Nov 2011 10:58:25 +0000 (11:58 +0100)]
mutex: handle gracefully if a PTHREAD_PRIO_INHERIT protocol cannot be set

This adds an additional check for unavailable PTHREAD_PRIO_INHERIT to the
fallback work done in ca717643ee768307475fc36ea29d920a13db0a8e

See bug #42715

12 years agomodule-rtp-recv: fail when SO_TIMESTAMP is not defined
Pino Toscano [Mon, 14 Nov 2011 10:48:47 +0000 (11:48 +0100)]
module-rtp-recv: fail when SO_TIMESTAMP is not defined

SO_TIMESTAMP is not POSIX and not available in any platform, so just fail
if the current platform does not have it.

See bug #42715

12 years agortp: use the right type when checking cmsg_type
Pino Toscano [Mon, 14 Nov 2011 10:44:43 +0000 (11:44 +0100)]
rtp: use the right type when checking cmsg_type

Use SCM_* instead of SO_* when checking the type of each cmsghdr.

See bug #42715

12 years agopipe: use pa_pipe_buf instead of the macro PIPE_BUF
Pino Toscano [Mon, 14 Nov 2011 10:33:47 +0000 (11:33 +0100)]
pipe: use pa_pipe_buf instead of the macro PIPE_BUF

This helps when porting to platforms that lack PIPE_BUF.

See bug #42715

12 years agodoc: Clarify pa_stream_get_latency() return value
Arun Raghavan [Fri, 30 Dec 2011 18:19:59 +0000 (23:49 +0530)]
doc: Clarify pa_stream_get_latency() return value

Clarifies that the latency is returned via an inout parameter and the
return value is an error code or 0 on success.

12 years agobuild-sys: Make esound bits optional
Arun Raghavan [Thu, 8 Dec 2011 10:36:50 +0000 (16:06 +0530)]
build-sys: Make esound bits optional

This is really quite irrelevant on most modern systems today, and more
so for embedded systems.

12 years agocore: comment typo
Peter Meerwald [Mon, 19 Dec 2011 11:28:44 +0000 (12:28 +0100)]
core: comment typo

12 years agocore: fix typo in logging
Peter Meerwald [Mon, 19 Dec 2011 11:28:43 +0000 (12:28 +0100)]
core: fix typo in logging

12 years agocore: sample_spec.rate is in Hz, not kHz; change logging output
Peter Meerwald [Mon, 19 Dec 2011 11:28:42 +0000 (12:28 +0100)]
core: sample_spec.rate is in Hz, not kHz; change logging output

12 years agocore: fix potential memory leak
Peter Meerwald [Mon, 19 Dec 2011 11:28:41 +0000 (12:28 +0100)]
core: fix potential memory leak

12 years agobuild-sys: Fix po/ build with --disable-nls
Arun Raghavan [Mon, 19 Dec 2011 04:50:30 +0000 (10:20 +0530)]
build-sys: Fix po/ build with --disable-nls

We can't call IT_PROG_INTLTOOL because it requires intltool even with
--disable-nls, but it provides USE_NLS which is required for the po/
build. This might take a while to get fixed upstream, so working around
this in our build system for now.

12 years agobuild-sys: Fix map-file check
Arun Raghavan [Mon, 19 Dec 2011 04:46:05 +0000 (10:16 +0530)]
build-sys: Fix map-file check

configure doesn't provide an abs_top_srcdir early enough, so we just use
srcdir for that. There was also a problem with the path we were setting
in VERSIONING_LDFLAGS that is fixed now.

12 years agosvolume_{mmx, sse}, sconv_sse: Fix compilation errors with X32 toolchain
Paul Menzel [Sat, 10 Dec 2011 09:03:51 +0000 (10:03 +0100)]
svolume_{mmx, sse}, sconv_sse: Fix compilation errors with X32 toolchain

From d8b81d5393df36085009bf9f69d41fa85e2ae58a Mon Sep 17 00:00:00 2001
From: Nitin A Kamble <nitin.a.kamble@intel.com>
Date: Sat, 10 Dec 2011 09:09:06 +0100

Make assembly syntax compatible to the X32 toolchain and fix the
following kind of compilations errors with X32 gcc.

| pulsecore/svolume_mmx.c: Assembler messages:
| pulsecore/svolume_mmx.c:107: Error: `(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:135: Error: `(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:161: Error: `(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:162: Error: `8(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:180: Error: `(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:210: Error: `(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:244: Error: `(%esi,%rdi,4)' is not a valid base/index expression
| pulsecore/svolume_mmx.c:245: Error: `8(%esi,%rdi,4)' is not a valid base/index expression
| make[3]: *** [libpulsecore_1.1_la-svolume_mmx.lo] Error 1

Originally these assembly lines were written for x86_64 ABI, now they
are also compatible with X32 ABI [3][4].

The patch was submitted to the OpenEmbedded-Core list [1][2].

[1] http://lists.linuxtogo.org/pipermail/openembedded-core/2011-December/014189.html
[2] http://git.yoctoproject.org/cgit.cgi/poky-contrib/commit/?h=nitin/x32&id=2d8eec54f755c51f2eff600390f5a4b3cc2a7662
[3] https://wiki.yoctoproject.org/wiki/X32_abi
[4] http://en.wikipedia.org/wiki/X32_ABI

12 years agox11: Fix build without NLS support
Arun Raghavan [Wed, 14 Dec 2011 16:27:01 +0000 (21:57 +0530)]
x11: Fix build without NLS support

12 years agotests: Fix resampler-test compilation without NLS support
Peter Meerwald [Wed, 14 Dec 2011 15:07:57 +0000 (16:07 +0100)]
tests: Fix resampler-test compilation without NLS support

12 years agoecho-cancel: Better handling of error conditions in test
Peter Meerwald [Tue, 13 Dec 2011 14:51:54 +0000 (15:51 +0100)]
echo-cancel: Better handling of error conditions in test

12 years agoecho-cancel: Begin log message with uppercase letter
Peter Meerwald [Tue, 13 Dec 2011 14:51:53 +0000 (15:51 +0100)]
echo-cancel: Begin log message with uppercase letter

12 years agoecho-cancel: Set file mode to binary in test code
Peter Meerwald [Tue, 13 Dec 2011 14:51:52 +0000 (15:51 +0100)]
echo-cancel: Set file mode to binary in test code

12 years agoprotocol-native: Fix 'auth-group-enabled' modarg
Arun Raghavan [Wed, 14 Dec 2011 15:18:12 +0000 (20:48 +0530)]
protocol-native: Fix 'auth-group-enabled' modarg

This is actually implemented in module-protocol-stub as
'auth-group-enable'. An unfortunate typo because the other argument is
spelt as 'enabled', but it's better to be slightly inconsistent than to
change the public interface.

12 years agoresampler: Move some peak resampler asserts around
Arun Raghavan [Tue, 13 Dec 2011 03:56:44 +0000 (09:26 +0530)]
resampler: Move some peak resampler asserts around

This moves a couple of asserts from peak_resample() to peaks_init()
since they're resampler parameters that shouldn't change after
initialisation.

12 years agoresampler: Remove invalid channel asserts in peak and trivial
Arun Raghavan [Tue, 13 Dec 2011 03:55:04 +0000 (09:25 +0530)]
resampler: Remove invalid channel asserts in peak and trivial

The incoming channel count will be fixed up by the remapping code before
the resampler is invoked.

12 years agoutils: Typo fixes in qpaeq
Arun Raghavan [Tue, 13 Dec 2011 03:46:40 +0000 (09:16 +0530)]
utils: Typo fixes in qpaeq

Pointed out by John Frankish <j-frankish@slb.com> in:

https://bugs.freedesktop.org/show_bug.cgi?id=43733

12 years agoecho-cance: Make Adrian canceller optional
Peter Meerwald [Sun, 11 Dec 2011 15:07:42 +0000 (16:07 +0100)]
echo-cance: Make Adrian canceller optional

makes the Adrian echo canceller implementation optional at compile time

this patch supersedes an earlier patch proposal and addresses the following
comments:
* separate patch from speex dependency rework (Arun)
* check that at least one EC implementation is available (Arun)
* properly align yes/no in configure summary for Adrian (Frederic)

12 years agobuild-sys: Make speex library optional
Peter Meerwald [Sun, 11 Dec 2011 15:07:41 +0000 (16:07 +0100)]
build-sys: Make speex library optional

make speex library dependency optional, this affects the resampler
and the echo canceller module

this patch supersedes an earlier patch proposal and addresses the following
comments:
* fix order of pa_echo_canceller_method_t enum and ec_table (Frederic)
* the default resampler is speex if available as before, otherwise ffmpeg (Arun)
* does not touch the Adrian EC implementation (see separate patch) (Arun)

12 years agoutils: Fixes for building without NLS
Arun Raghavan [Fri, 9 Dec 2011 15:30:11 +0000 (21:00 +0530)]
utils: Fixes for building without NLS

12 years agoiochannel: Handle missing un.h correctly
Arun Raghavan [Thu, 8 Dec 2011 08:59:06 +0000 (14:29 +0530)]
iochannel: Handle missing un.h correctly

12 years agomime: Move assert to correct position
Arun Raghavan [Thu, 8 Dec 2011 08:49:50 +0000 (14:19 +0530)]
mime: Move assert to correct position

12 years agobuild-sys: Use absolute path for map-file while linking
Arun Raghavan [Thu, 8 Dec 2011 08:47:07 +0000 (14:17 +0530)]
build-sys: Use absolute path for map-file while linking

This makes the link process more robust in some stranger configurations.

12 years agobuild-sys: Fix building without NLS
Arun Raghavan [Thu, 8 Dec 2011 08:45:06 +0000 (14:15 +0530)]
build-sys: Fix building without NLS

12 years agostream: Fix typos and formatting in stream.h doxygen documentation
Peter Meerwald [Mon, 12 Dec 2011 18:50:38 +0000 (19:50 +0100)]
stream: Fix typos and formatting in stream.h doxygen documentation

12 years agomainloop: Fix typos and rewording of thread-mainloop.h doxygen documentation
Peter Meerwald [Mon, 12 Dec 2011 18:50:37 +0000 (19:50 +0100)]
mainloop: Fix typos and rewording of thread-mainloop.h doxygen documentation

12 years agosimple: Fix typos in simple.h doxygen documentation
Peter Meerwald [Mon, 12 Dec 2011 18:50:36 +0000 (19:50 +0100)]
simple: Fix typos in simple.h doxygen documentation

12 years agostream: Fix 'e g' as 'e.g.\ '
Peter Meerwald [Mon, 12 Dec 2011 18:50:35 +0000 (19:50 +0100)]
stream: Fix 'e g' as 'e.g.\ '

12 years agoerror: Fix spelling of 'initialization' in errortab for PA_ERR_MODINITFAILED
Peter Meerwald [Mon, 12 Dec 2011 09:59:45 +0000 (10:59 +0100)]
error: Fix spelling of 'initialization' in errortab for PA_ERR_MODINITFAILED

Also update the related translations.

12 years agoproplist: Consistently use parameter name p for a pa_proplist* in prototype
Peter Meerwald [Mon, 12 Dec 2011 12:19:50 +0000 (13:19 +0100)]
proplist: Consistently use parameter name p for a pa_proplist* in prototype

12 years agoutf8: Fix doxygen file description for utf8.h
Peter Meerwald [Mon, 12 Dec 2011 12:19:49 +0000 (13:19 +0100)]
utf8: Fix doxygen file description for utf8.h

12 years agoproplist: Fix typos in doxygen documentation
Peter Meerwald [Mon, 12 Dec 2011 12:19:48 +0000 (13:19 +0100)]
proplist: Fix typos in doxygen documentation

12 years agoproplist: Add doxygen file description for proplist.h
Peter Meerwald [Mon, 12 Dec 2011 12:19:47 +0000 (13:19 +0100)]
proplist: Add doxygen file description for proplist.h

12 years agointrospect: Fix unresolved doxygen link pa_port_available_t
Peter Meerwald [Mon, 12 Dec 2011 12:19:46 +0000 (13:19 +0100)]
introspect: Fix unresolved doxygen link pa_port_available_t

12 years agoformat: Add description for format.h
Peter Meerwald [Mon, 12 Dec 2011 12:19:45 +0000 (13:19 +0100)]
format: Add description for format.h

12 years agoformat: Add periods to fix brief documentation for pa_format_info_is_pcm()
Peter Meerwald [Mon, 12 Dec 2011 12:19:44 +0000 (13:19 +0100)]
format: Add periods to fix brief documentation for pa_format_info_is_pcm()

12 years agoformat: Fix unknown doxygen command \second
Peter Meerwald [Mon, 12 Dec 2011 12:19:43 +0000 (13:19 +0100)]
format: Fix unknown doxygen command \second

12 years agodef: Document that pa_buffer_attr.maxlength is in bytes.
Peter Meerwald [Sun, 11 Dec 2011 15:07:39 +0000 (16:07 +0100)]
def: Document that pa_buffer_attr.maxlength is in bytes.

Fixes https://bugs.freedesktop.org/show_bug.cgi?id=42919

12 years agostream: Fix comments
Peter Meerwald [Sun, 11 Dec 2011 15:07:38 +0000 (16:07 +0100)]
stream: Fix comments

12 years agopstream: Fix spelling of 'receive'.
Peter Meerwald [Mon, 12 Dec 2011 22:36:39 +0000 (22:36 +0000)]
pstream: Fix spelling of 'receive'.

These functions are not exported in the map file and thus do
not form part of the API and can be corrected without any problem.

12 years agoalsa-mixer: Remove unused pa_alsa_path_set_probe() declaration.
Tanu Kaskinen [Mon, 5 Dec 2011 18:42:44 +0000 (20:42 +0200)]
alsa-mixer: Remove unused pa_alsa_path_set_probe() declaration.

12 years agodevice-port: Add a property list to ports.
David Henningsson [Fri, 2 Dec 2011 12:34:54 +0000 (13:34 +0100)]
device-port: Add a property list to ports.

They're not used for anything yet, but hopefully soon they'll be.

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
12 years agopactl: Show port availability
David Henningsson [Fri, 2 Dec 2011 13:15:28 +0000 (14:15 +0100)]
pactl: Show port availability

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
12 years agoFixup a few things in the new path probing
David Henningsson [Fri, 2 Dec 2011 09:20:23 +0000 (10:20 +0100)]
Fixup a few things in the new path probing

 * If mapping_probe_paths() fails to open the mixer, all paths are now
removed from the mapping's path sets.
 * pa_path_set.probed isn't really used for anything (removed).
 * If profile probing is configured to be skipped, mapping_paths_probe()
should still be called.

Thanks to Tanu for spotting.

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
12 years agoalsa: add card ports and path probe cache
David Henningsson [Fri, 25 Nov 2011 14:17:16 +0000 (15:17 +0100)]
alsa: add card ports and path probe cache

To be able to add ports to all profiles, we need to probe all
profiles at startup. To speed this up, we now have a cache of
probes paths which is owned by the profile set. Since paths
are now owned by the profile set, the path set must now have
a hashmap of paths instead of a linked list.

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
12 years agoalsa: Ignore the virtual "thinkpad-acpi" card
David Henningsson [Wed, 30 Nov 2011 11:51:06 +0000 (12:51 +0100)]
alsa: Ignore the virtual "thinkpad-acpi" card

The thinkpad ACPI driver sometimes creates a virtual sound card,
which at best exposes a volume control. Save some startup time, and
unnecessary error messages in the log, by ignoring it.

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
12 years agoecho-cancel: Fix memory leak in test program
Peter Meerwald [Mon, 28 Nov 2011 15:14:16 +0000 (16:14 +0100)]
echo-cancel: Fix memory leak in test program

memory for pa_echo_canceller is allocated in init_common() a second time,
hence the allocated memory in main() is leaked

12 years agoresampler: Some minor tweaks
Maarten Bosmans [Mon, 28 Nov 2011 09:36:34 +0000 (10:36 +0100)]
resampler: Some minor tweaks

that were forgotten in the previous peaks resampler commits.

12 years agoecho-cancel: Fix warning for undefined HAVE_WEBRTC
Arun Raghavan [Mon, 28 Nov 2011 09:46:55 +0000 (15:16 +0530)]
echo-cancel: Fix warning for undefined HAVE_WEBRTC

12 years agotests: Don't link tests that only use public api to libpulsecommon
Maarten Bosmans [Wed, 16 Nov 2011 12:00:28 +0000 (13:00 +0100)]
tests: Don't link tests that only use public api to libpulsecommon

12 years agoresamplers: Improve performance of peaks resampler
Maarten Bosmans [Wed, 2 Nov 2011 20:54:17 +0000 (21:54 +0100)]
resamplers: Improve performance of peaks resampler

This is mainly achieved by special-casing the common 1ch float case,
which is used by applications such as pavucontrol.

12 years agoresamplers: Rework the peaks resampler
Maarten Bosmans [Wed, 2 Nov 2011 20:54:16 +0000 (21:54 +0100)]
resamplers: Rework the peaks resampler

The algorithm had been implemented the same way as the trivial resampler.  But
an important difference between the two is that the trivial resampler can write
an output as soon as the first corresponding input sample is seen, whereas the
peaks resampler must have read all input samples before writing an output
sample.

With this rework, the peaks resampler now outputs samples correctly when the
input data is spanning multiple memblocks.

12 years agoshm: Use a goto rather than early return for consistency.
Maarten Bosmans [Sun, 27 Nov 2011 16:12:10 +0000 (16:12 +0000)]
shm: Use a goto rather than early return for consistency.

This is functionally the same but is easier to understand.

12 years agowin32: Avoid some compiler warnings when cross-compiling for mingw32
Maarten Bosmans [Wed, 2 Nov 2011 20:54:13 +0000 (21:54 +0100)]
win32: Avoid some compiler warnings when cross-compiling for mingw32

Autoconf documentation says that AC_FUNC_SELECT_ARGTYPES shouldn't be used anyway.

12 years agodevice-port: Remove redundant include after the better circular dep fix.
Colin Guthrie [Sun, 27 Nov 2011 15:44:23 +0000 (15:44 +0000)]
device-port: Remove redundant include after the better circular dep fix.

12 years agox11: Drop unneeded 'struct' and use the typedef directly.
Colin Guthrie [Sun, 27 Nov 2011 15:23:55 +0000 (15:23 +0000)]
x11: Drop unneeded 'struct' and use the typedef directly.

12 years agoUse simple free function in pa_queue_free
Maarten Bosmans [Fri, 18 Nov 2011 08:58:09 +0000 (09:58 +0100)]
Use simple free function in pa_queue_free

12 years agoUse simple free function in pa_dynarray_free
Maarten Bosmans [Fri, 18 Nov 2011 08:58:08 +0000 (09:58 +0100)]
Use simple free function in pa_dynarray_free

12 years agodevice-port: Fix the circular dependency problem more cleanly.
Tanu Kaskinen [Sun, 27 Nov 2011 15:16:53 +0000 (17:16 +0200)]
device-port: Fix the circular dependency problem more cleanly.

Declaring structs before including anything ensures that the
structs are available for everybody in case of circular includes.

12 years agobluetooth: Fix property reply handling for hfgw
Mikel Astiz [Wed, 23 Nov 2011 16:08:11 +0000 (17:08 +0100)]
bluetooth: Fix property reply handling for hfgw

12 years agoresamplers: Optimize trivial resampler
Maarten Bosmans [Wed, 23 Nov 2011 10:40:08 +0000 (11:40 +0100)]
resamplers: Optimize trivial resampler

This improves the performance of a typical s16 2ch resampling by 88%.

12 years agoresamplers: Use better variable name in trivial resampler
Maarten Bosmans [Wed, 23 Nov 2011 10:40:07 +0000 (11:40 +0100)]
resamplers: Use better variable name in trivial resampler

12 years agobuild-sys: Disable fastpath asserts by default
Maarten Bosmans [Wed, 23 Nov 2011 10:40:06 +0000 (11:40 +0100)]
build-sys: Disable fastpath asserts by default

12 years agoresamplers: Use fastpath assert in trivial resampler
Maarten Bosmans [Wed, 23 Nov 2011 10:40:05 +0000 (11:40 +0100)]
resamplers: Use fastpath assert in trivial resampler

When the assert is disabled, the trivial resampler gets a 35% performance boost.

12 years agopulsecore: Hardcode FRAME_SIZE_MAX_ALLOW
Maarten Bosmans [Tue, 22 Nov 2011 15:09:18 +0000 (16:09 +0100)]
pulsecore: Hardcode FRAME_SIZE_MAX_ALLOW

Instead of using PA_SCACHE_ENTRY_SIZE_MAX, the size for FRAME_SIZE_MAX_ALLOW is
set directly to the same value.  This removes the need for the core-scache.h
include, which caused an unwanted dependency of libpulsecommon on libpulsecore.

Fixes https://bugs.freedesktop.org/show_bug.cgi?id=41539

12 years agoUpdate LICENSE
Maarten Bosmans [Tue, 22 Nov 2011 15:08:48 +0000 (16:08 +0100)]
Update LICENSE

Mention gdbm, lirc and fftw there.

12 years agopulsecore: Fix issue with circuilar definitions.
Colin Guthrie [Sun, 27 Nov 2011 14:19:57 +0000 (14:19 +0000)]
pulsecore: Fix issue with circuilar definitions.

Without this fix, errors about previous definitions are generated in
numerous locations.

Example of error:

  CC     libpulsecore_1.98_la-auth-cookie.lo
In file included from ../../src/pulsecore/source.h:46:0,
                 from ../../src/pulsecore/sink.h:40,
                 from ../../src/pulsecore/core.h:50,
                 from ../../src/pulsecore/shared.h:25,
                 from ../../src/pulsecore/auth-cookie.c:33:
../../src/pulsecore/device-port.h:40:24: error: redefinition of typedef
'pa_core'
../../src/pulsecore/core.h:29:24: note: previous declaration of
'pa_core' was here
make[3]: *** [libpulsecore_1.98_la-auth-cookie.lo] Error 1

Overall it would be nicer if we could avoid this kind of fix, but it
would require further reorganisation that I'm not prepared to undertake
right now.

12 years agocli: Show card ports and jack detection status
David Henningsson [Fri, 25 Nov 2011 14:17:15 +0000 (15:17 +0100)]
cli: Show card ports and jack detection status

Expose the new stuff through pacmd.

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
12 years agoNotify port available status changes, and update protocol version
David Henningsson [Fri, 25 Nov 2011 14:17:14 +0000 (15:17 +0100)]
Notify port available status changes, and update protocol version

The recommended way of setting available status is to call
pa_device_port_set_available, which will send a subscription event
to the relevant card. It will also fire a hook.

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
12 years agoCards now has ports directly, and device port has list of profiles
David Henningsson [Fri, 25 Nov 2011 14:17:13 +0000 (15:17 +0100)]
Cards now has ports directly, and device port has list of profiles

This forms the base for being able to expose all ports of all
profiles (even inactive ones) to clients.

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
12 years agoTurn device ports into reference counted objects
David Henningsson [Fri, 25 Nov 2011 14:17:12 +0000 (15:17 +0100)]
Turn device ports into reference counted objects

Since both cards and sinks can hold references to a port, it makes
sense to reference count them. Although no current implementation
actually has sinks with ports but without a card, it felt wrong
to make it harder to make such an implementation in the future.

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
12 years agoIntroduce "available" concept for ports, and communicate that to clients. Bump protoc...
David Henningsson [Fri, 25 Nov 2011 14:17:11 +0000 (15:17 +0100)]
Introduce "available" concept for ports, and communicate that to clients. Bump protocol version to 24.

Note: There is still no notification when status availability changes.

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
12 years agofilters: Handle echo-cancel streams better
Arun Raghavan [Thu, 17 Nov 2011 14:31:59 +0000 (20:01 +0530)]
filters: Handle echo-cancel streams better

This makes handling of echo-cancel streams (and potentially other
filters that need to work with multiple streams) cleaner. We do this by
adding the ability to apply filter on a group of streams rather than
always treating streams individually.

This is required, for example, when changing the input source for an
echo-cancel source output. When this happens, we want to change the
module-echo-cancel instance to actually cancel between the new source
and old sink.

To do this when one stream in a group moves, we create a new filter
instance between the current master sink/source pair and move the
relevant streams to the virtual sink/source from this new instance.

12 years agoproplist: Add internal API to get stream group
Arun Raghavan [Wed, 23 Nov 2011 10:56:40 +0000 (16:26 +0530)]
proplist: Add internal API to get stream group

This moves out code from module-stream-restore and makes an internal API
out of it to get a "stream group" for a given sink input or source output.
This is factored out for reuse in module-filter-*.

The stream group basically provides some means of attaching a logical
identification to the stream (by role, application id, etc.).

12 years agofilters: Allow a filter to have both sink and source
Arun Raghavan [Fri, 11 Nov 2011 09:38:24 +0000 (15:08 +0530)]
filters: Allow a filter to have both sink and source

This sets the base for better representing module-echo-cancel. This
patch itself should change no behaviour -- it only makes way for other
changes.

12 years agofilters: Fix the master source/sink when autoloaded
Arun Raghavan [Wed, 23 Nov 2011 11:16:12 +0000 (16:46 +0530)]
filters: Fix the master source/sink when autoloaded

When autoloaded, it is expected that module-filter-apply (or whatever is
loading us) will take care of applying the filter on the correct
sink/source master. Instead of adding complexity by tracking what is
currently being filtered, we just disallow filtering anything except the
original master sink/source and let module-filter-apply or whatever is
loading us deal with dynamic sink/source changes.

12 years agoecho-cancel: Use more human-friendly descriptions
Arun Raghavan [Tue, 22 Nov 2011 07:54:22 +0000 (13:24 +0530)]
echo-cancel: Use more human-friendly descriptions

This makes what devices are being cancelled clearer in the UI (at the
cost of being somewhat less clear when multiple devices of the same name
are plugged, but at least that's a much smaller set than everyone).

12 years agoecho-cancel: Use speex by default if webrtc isn't available
Arun Raghavan [Tue, 15 Nov 2011 07:20:10 +0000 (12:50 +0530)]
echo-cancel: Use speex by default if webrtc isn't available

Thanks to Peter Meerwald <pmeerw@pmeerw.net> for pointing this out.

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=42925

12 years agocore: Look up /etc/machine-id if D-Bus machine-id is not found
Arun Raghavan [Mon, 14 Nov 2011 08:12:21 +0000 (13:42 +0530)]
core: Look up /etc/machine-id if D-Bus machine-id is not found

It appears that this is currently the fallback for early boot and other
such cases where /var might not be available. Relevant upstream commit:

http://cgit.freedesktop.org/dbus/dbus/commit/?id=66e52541d5bdd4927a5c702963749760643313f4

Thanks to Samuli Suominen <ssuominen@gentoo.org> for pointing this out
in https://bugs.gentoo.org/show_bug.cgi?id=390287

12 years agooss: Use log2 function from core-util instead of own implementation
Maarten Bosmans [Wed, 2 Nov 2011 20:54:15 +0000 (21:54 +0100)]
oss: Use log2 function from core-util instead of own implementation

12 years agoSquash unused variable compiler warning
Maarten Bosmans [Wed, 2 Nov 2011 20:54:14 +0000 (21:54 +0100)]
Squash unused variable compiler warning

12 years agoRemove vector.h and vector-test
Maarten Bosmans [Wed, 2 Nov 2011 20:54:12 +0000 (21:54 +0100)]
Remove vector.h and vector-test

This code was never used and vectorization is handled by Orc anyway.

12 years agobuild-sys: Remove time-smoother and shm from libpulsecore
Maarten Bosmans [Wed, 19 Oct 2011 12:10:13 +0000 (14:10 +0200)]
build-sys: Remove time-smoother and shm from libpulsecore

As they are already included in libpulsecommon.

12 years agortpoll: Update comment
Maarten Bosmans [Thu, 13 Oct 2011 18:54:17 +0000 (20:54 +0200)]
rtpoll: Update comment

to reflect changes made in
32e2cd6d3216f780c4cffed0f8eb3c30f2c8d732
core: get rid of rt sig/timer handling since modern Linux' ppoll() is finally fixed for granularity

12 years agostream: Fix upload samples' cleanup
Antti-Ville Jansson [Fri, 11 Nov 2011 14:22:24 +0000 (16:22 +0200)]
stream: Fix upload samples' cleanup

In pa_create_stream_callback, a stream is inserted into
s->context->record_streams only if it's a record stream. Otherwise it's
inserted into s->context->playback_streams. However, in stream_unlink the
stream is removed from s->context->playback_streams only if it's a playback
stream and otherwise it's removed from s->context->record_streams.

Thus, if the stream is an upload stream, we first insert it into
s->context->playback_streams in pa_create_stream_callback and then try to
remove it unsuccessfully from s->context->record_streams in stream_unlink. This
means that we are leaking hashmap entries until the context is freed,
constantly consuming more memory with applications that upload and unload a
large number of samples through one context.

Of course, this begs the question whether upload streams even belong in either
of those hashmaps. I don't want to mess around with the code too much at this
point though, so this patch should be a sufficient improvement.

12 years agoalsa: Minor debug code cleanup
Arun Raghavan [Sat, 12 Nov 2011 05:48:25 +0000 (11:18 +0530)]
alsa: Minor debug code cleanup

Remove a trap on underrun and uncomment a couple of useful debug prints
(still enabled only when DEBUG_TIMING is defined).

12 years agotests: Add playback tests to check-daemon
Arun Raghavan [Thu, 10 Nov 2011 07:54:07 +0000 (13:24 +0530)]
tests: Add playback tests to check-daemon

Adds the standard/extended API playback tests to the check-daemon set.

12 years agocore: Make debugging a bit simpler
Arun Raghavan [Thu, 10 Nov 2011 07:13:09 +0000 (12:43 +0530)]
core: Make debugging a bit simpler

This takes out a bunch of commented debug prints and puts them in
defines. Makes it easier to turn them all on/off at a single point.

13 years agotests: Add a target for running manual tests
Maarten Bosmans [Mon, 7 Nov 2011 18:39:34 +0000 (00:09 +0530)]
tests: Add a target for running manual tests

This adds a check-daemon make target which requires a running PulseAudio
server to test against.

13 years agoecho-cancel: Make WebRTC the default canceller
Arun Raghavan [Mon, 7 Nov 2011 14:05:28 +0000 (19:35 +0530)]
echo-cancel: Make WebRTC the default canceller

Works pretty well, and isn't a CPU hog, so let's use it!

13 years agoecho-cancel: Turn WebRTC analog gain control on by default
Arun Raghavan [Mon, 7 Nov 2011 14:05:05 +0000 (19:35 +0530)]
echo-cancel: Turn WebRTC analog gain control on by default

Seems to be pretty well-behaved.

13 years agoecho-cancel: Hook up WebRTC analog gain control
Arun Raghavan [Fri, 4 Nov 2011 11:08:30 +0000 (16:38 +0530)]
echo-cancel: Hook up WebRTC analog gain control

13 years agoecho-cancel: Add infrastructure for cancellers to do AGC
Arun Raghavan [Fri, 4 Nov 2011 10:47:26 +0000 (16:17 +0530)]
echo-cancel: Add infrastructure for cancellers to do AGC

This adds some infrastructure for canceller implementations to also
perform acoustic gain control. Cancellers now have a couple of new API
calls that allow them to get/set capture volume.

This is made slightly complex by the fact that cancellation happens in
thread context while most volume mangling needs to be done in main
context. To deal with this, while getting the volume we save source
volume updates as they are propagated to thread context and use this
cached value for queries. To set the volume, we send an async message to
main context and let that set the source volume.

13 years agosink,source: Update sample rate if possible on stream uncork
Arun Raghavan [Sun, 6 Nov 2011 13:42:19 +0000 (19:12 +0530)]
sink,source: Update sample rate if possible on stream uncork

If a stream is being uncorked and no other stream is running, we try to
update the sink/source sample rate to match it and avoid resampling.