As an extra, I broke try_to_switch_profile() into smaller
functions, because the two levels of loops with continue
statements inside both were a bit hard to follow.
get_latency_us() used an uninitialized variable, and an incorrect
scope for some of the AudioObjectGetPropertyData() calls. As a result,
audio would randomly not work at all.
BugLink: https://bugs.freedesktop.org/show_bug.cgi?id=65122
e->description is a pointer, not a fixed char array. Hence it
makes no sense to use strncmp.
This fixes a compiler warning when compiling under Ubuntu.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
This fixes a later assertion failure in module-stream-restore.
Buglink: https://bugs.launchpad.net/bugs/896602
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
A stationary computer usually has headphone jack(s) and line out jacks.
In some cases analog-output.conf will be a subset of
analog-output-headphones.conf, causing line outs to be unusable (because
headphones are unplugged).
This late in the cycle, this was the safest way I could think of to try
to fix this for a particular computer. In later versions of PulseAudio
we could consider making a dedicated line out path instead, and have
proper jack detection there.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
send_counter/recv_counter relate to the bytes (play stream) passed
through the queue, hence the same sample spec must be used
Signed-off-by: Peter Meerwald <p.meerwald@bct-electronic.com>
Acked-by: Stefan Huber <shuber@sthu.org>
As far as I can see, having a mono path in a stereo mapping doesn't
make any sense. It also causes breakage: if the Master Mono mixer
element has two volume channels, the analog-output path gets removed
due to being a subset of analog-output-mono, and that in turn causes
the Master element getting muted. Users generally don't like that.
BugLink: https://bugs.freedesktop.org/show_bug.cgi?id=54673
"any word starting with the letters" parts; this does not hold
any longer with commit 0e29e7365907ffbe90df768a4dea277dba40d495
core-util: Don't accept random words in pa_parse_boolean()
Signed-off-by: Peter Meerwald <p.meerwald@bct-electronic.com>
When a "Headphone Mic" jack becomes available, we do not know if
a headphone or a mic has been plugged in. Therefore, setting both
paths to "unknown" is, in theory, the correct thing to do.
However, in practice, people are more likely to plug in a headphone
rather than a mic. Therefore, allow autoswitch to the headphone port
when the jack is plugged in.
A more advanced implementation would consider what was plugged in last
time depending on what port was selected on the input side at that
time, and set availability accordingly. However, such an implementation
will have to wait (probably at least until we have our fancy routing
system implementation).
Buglink: https://bugs.launchpad.net/bugs/1169143
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
I recently came across a device without any ALSA-level mixer controls,
everything was physical knobs on the hardware.
This patch enables that device to get a port too ("Analog Input").
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
The hdmi_eld_changed callback is called by alsa-lib at shutdown.
In that case, just exit instead of trying to access something with
already closed handles.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
We need to pick the right port as early as possible, before the
first volume is picked up. Hence this module needs to be loaded
before the sound card modules are loaded.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
If a card is hot-plugged (which all cards will be when we load
this module before module-*-detect), make sure we don't start up
a sink with an unavailable port selected.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
This way port availability has been filled in when we create the
sink, which will later enable us to pick the right port directly.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
It appears that, libltdl will find the .la file in the builddir and
figure out where the real .so is.
This also requires .ifexists to be fixed up to correspondingly search in
<dlsearchpath>/.libs.
We know we always serve up LPCM, and exposing this via D-Bus lets Rygel
set the appropriate metadata while presenting the raw (i.e.
non-transcoded) stream to clients.
same for e.g. versus e.g.\ and e.g. versus E.g.
this is ueber-nitpicking: will anybody ever notice?
Signed-off-by: Peter Meerwald <p.meerwald@bct-electronic.com>
A recent patch changed the path files from PA_BUILDDIR to PA_SRCDIR.
Do the same to the profile-set files for consistency (and to fix
out of tree builds).
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
The mixer paths are not available in ${builddir} - we need to look in
${srcdir}. This should fix running an in-tree build without make install
as well as alsa-mixer-path-test in make distcheck.
Since the most straightforward way to define PA_SRCDIR was in
Makefile.am, I'm moving PA_BUILDDIR there as well for consistency.
On some machines which has a headset jack, the headset mic does not have its own
jack detection. Then we can look at the headphone jack to get some indication:
We know that if the headphone is unplugged, so is the headset mic. The opposite
is not guaranteed since the user might have plugged in a headphone, not a headset.
Also, there exist multi-function jacks which support both Headphone, Mic in headphone jack
and Headset Mic. In this case the jack name will be "Headphone Mic", not "Headphone", so
we need to include this name too.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
The old assumption seemed to be that if a sink or source has the
DYNAMIC_LATENCY flag set, it can never change, so the fixed latency
will always be zero. This assumption doesn't hold with filter sinks
and sources that are moved around.
This fixes a crash with two module-virtual-sink instances on top of
each other, when the bottom one is moved from a sink without dynamic
latency to a sink with dynamic latency. What happened was that first
the bottom virtual sink "updated" (due to this bug nothing was
actually updated) its fixed latency to match the master sink (zero
fixed latency), and then the top virtual sink updated its fixed
latency to match the master sink. The master sink was the bottom
virtual sink, whose fixed latency should have been set to zero, but it
was not, so the pa_sink_set_fixed_latency_within_thread() failed in
the assertion "latency == 0".
An example: let's say that there's an alsa sink and two filter sinks
on top of each other:
alsa-sink <- filter1 <- filter2
With the old code, if filter1 gets moved to another sink, and the
new sink doesn't have the LATENCY and DYNAMIC_LATENCY flags set
(unlike alsa-sink), filter1's flags are updated fine in the moving()
callback, but filter2 is not notified at all about the flag changes.
With this patch, the flag changes are propagated to filter2 too.
Forcing the shm file to be read-only makes shm_unlink() fail on OS X.
Thanks to Albert Zeyer for reporting the bug and investigating the
root cause.
BugLink: https://bugs.freedesktop.org/show_bug.cgi?id=62988
During a stream, most packets sent are either memblocks (with SHM info),
or requests for more data. These are only slightly bigger than the
header.
This patch makes it possible to write these packages in one write
instead of two: a memcpy of just a few bytes is worth saving extra
syscalls for write and poll.
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
