Without -avoid-version, libtool creates a libwebrtc-util.so.0 and
libwebrtc-util.so.0.0.0 which are not cleaned up by make uninstall,
which in turn causes make distcheck to fail.
This moves out the webrtc bits into a small helper library to shield the
rest of module-echo-cancel from being linked with a C++ linker. This is
required because automake will _always_ link module-echo-cancel in C++
mode if any of its deps (even conditional ones) are in C++.
Without this file, pulseaudio aborts with:
E: [pulseaudio] alsa-mixer.c: Assertion 'eb = p2->elements' failed at
modules/alsa/alsa-mixer.c:3179, function path_set_condense(). Aborting.
UCM basic functions will provide another way to handle the alsa mixer
and controls. That means alsa card module will make use of alsa ucm
configurations provided by various audio systems instead of mixer and
paths configurations provided by PA. PA profiles come from UCM verb, PA
sinks/sources and ports come from UCM devices.
In case the proper UCM configurations are found, ucm branches are
activated, or we will still fall through to the original way.
Signed-off-by: Feng Wei <wei.feng@freescale.com>
This is a basic module for enabling loopback as soon as a new bluetooth A2DP
source is created. The module is given a source and a media role using command
line. This allows module-intended-roles or module-device-manager to choose a
target sink for the stream.
Several laptops have speaker ports, and/or internal mic ports, but we have
no way of detecting that. So we make the port(s) always show up for these
devices.
BugLink: https://bugs.launchpad.net/bugs/946232
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
ConsoleKit has been deprecated and replaced by systemd's logind daemon,
hence provide the same functionality of module-console-kit in
module-systemd-login. This also makes sure that the CK module becomes a
NOP if the system is booted with systemd, resp. that the systemd module
becomes a NOP if the system is booted without systemd, thus being nice
to OSes such as Debian which want to support multiple init systems.
This reverts commit 24ff719675.
If these files aren't compiled in both libpulse and
libpulsecommon, some things will try use non-public
functions from libpulse. Therefore those internal functions
have to be included directly in libpulsecommon.
This problem appears to be only visible with --as-needed,
which is why the problem wasn't noticed immediately. The
problem has existed also earlier, and it was fixed by
Maarten Bosmans in commit 2de2c735. The commit that is
now reverted basically reverted Maarten's commit (I didn't
know that when I wrote the bad patch).
A part of Xen's paravirtualized audio driver has been developed as a
pulseaudio module. This module acts as a tunnel over Xen's shared memory
mechanism and allows a domU guest to send audio data to a dom0 backend.
Reference: https://bugs.freedesktop.org/show_bug.cgi?id=43503
This module tries to switch to a port when availability changes to
"YES", and tries to switch away when availability changes to "NO".
Once there is a priority list infrastructure in place and ready,
this functionality might be redundant, but this will do as an
interim solution.
For Nvidia and Intel, support probing of up to four HDMI devices.
Also add port information to all HDMI profiles.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
It provides a virtual surround sound effect.
v2: Normalize hrir to avoid clipping, some cleanups
v3: use fabs, not abs
v4: implement changes proposed by Tanu Kaskinen
v5: likewise
v6: use channel map from hrir file
v7: remove hrir_ss and hrir_map form userdata
v8: update naming of sink
libtools causing relinking on installation, to make this succeed
libpulsecommon needs to be installed before the other libraries and the
padsp libraries needs to be installed afterwards.
Unfortunately autotools doesn't consider dependencies when running the
install target, thus we have to enforce the ordering ourselves
libpulsecommon is a private library only for use within pulseaudio, so
lets move it into a private directory
[ed: from discussion on IRC, while it looks redundant to have modules in
lib/pulse-$VER and private libraries in $pkglibdir, this is actually
desirable. For example, a multi-architecture libpulse would have a
$pkgdir per architecture (/usr/lib/<arch>/pulseaudio) whereas you'd
still want the server-specific modules in /usr/lib/pulse...) -- Arun]
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)
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)
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>
Some tests (remix-test, sig2str) only display information, so they are not
useful for automated testing. Others (interpol-test, once-test, thread-test)
do return an error on failure, so should be included in TESTS.
Instead of spilling thousands of lines of output, make check now runs the
test-suite in about 100 lines or so. If running under make check, the output of
tests is reduced. The MAKE_CHECK environment variable is used for this, so that
when running the test manually, the full output is still shown. Furthermore,
pa_log is used consistently instead of printf, so that all test output goes to
stderr by default. Colored output from make check goes to stdout.
This adds the WebRTC echo canceller as another module-echo-cancel
backend. We're exposing both the full echo canceller as well as the
mobile echo control version as modargs.
Pending items:
1. The mobile canceller doesn't seem to work at the moment.
2. We still need to add bits to hook in drift compensation (to support
sink and source from different devices).
The most controversial part of this patch would probably be the
mandatory build-time dependency on a C++ compiler. If the optional
--enable-webrtc-aec is set, then there's also a dependency on libstdc++.
This patch was already added earlier with commit ID 2f86ba4f, but the
changes got reverted by commit 3adc43b ("win32: Make once-test work").
However, this still doesn't work on OSX as here, pthread is in general
available, but the barrier APIs aren't.
