If transport goes into error state too often, fail instead of trying to
acquire it again.
This avoids getting into a tight acquire->fail->reacquire loop.
We need to acquire and release all transports in the same CIG at the
same time.
Due to current kernel ISO socket limitations, this cannot be done one by
one.
Now that sinks/sources can do transport acquire asynchronously, remove
the workaround that made it synchronous. Do release still synchronously
however.
Change A2DP/BAP transport acquire and release to be async.
Since BlueZ acquiring ISO sockets blocks until all sockets in same CIG
are acquired, BAP transports must be acquired asynchronously.
Allow asynchronous changes in transport state in the sinks/sources.
Also allow transport acquire to be actually synchronous, in this case it
must set transport state during acquire call.
Separate driver start/stop from transport start/stop.
Emit any remove node events before resetting initial profile. It
indicates to the session manager that nodes if any went away before
device disconnected.
Usually the profile is removed first which removes the nodes. This
depends on ordering of events from bluez, which apparently can be
different depending on how remote device disconnects.
Add some guards against doing processing when there has been an error or
the node is not started. Set error status to IO. Continue driving on IO
errors.
In media-sink, there's no need to set RCVBUF.
In media-source, we don't need to set NONBLOCK, as reads are done with
DONTWAIT. Don't set SNDBUF as it's not needed there. Don't set RCVBUF,
but use the (big) kernel default value: decode-buffer will handle any
overruns. Small values of RCVBUF might cause problems if kernel is
sending packets in a burst faster than we wake up.
Don't just limit the max delay of samples we keep in the ALSA ringbuffer
to the buffer_size but to half of it. Make this into a max_delay
variable.
If we have a buffer size of 8192 samples and a headroom of 8192 samples,
when capturing, we would wait for the ringbuffer to contain at least
8192 samples, which would always xrun. When we limit the size to
half, we can still read the data without xruns.
Fixes#2972
On underflow in sources, pad with explicit silence. This avoids the
audioadapter from getting off sync from the cycle. That causes problems
as driver when we want to produce a buffer only a the start of the
cycle.
In some cases, it's also possible that the io already has buffer at the
start of the cycle when rate matching as driver. Currently, we don't
produce buffer in this case, but we should. Fix that by doing things in
the exact same way as ALSA sources do.
On glibc, `pthread_t` is `unsigned long int` while on musl
it has a pointer type. To avoid format string warnings,
cast it to `void *` and use the `%p` format specifier.
Delay output by one packet, so that we never need to wait for
node_process to supply more data when a packet is due out, and can write
audio packets at exactly equal intervals (up to timer/io accuracy).
In principle, this should not be necessary. However, enable it for now,
in case this improves the various stutter/etc. bug reports.
After flushing a packet, encode the next one immediately if we already
have the data. This makes the flush timing more accurate (std ~4x
smaller) as we don't need to wait for the encode.
* Add support for running the sink as a driver
* Detect which compressed formats are actually supported
* Correctly open/close/start/stop device according to the node commands
* Shift away from tinycompress and use Compress-Offload ioctls directly
to be able to access various caps information (including fragment sizes)
which are unavailable in the tinycompress API
* Implement SPA_PARAM_PropInfo and SPA_PARAM_Props support
The maximum receive buffer target of 6 packets may be too small when
there's huge jitter in reception. Increase it so that we may use all
buffer available if needed (2*quantum_limit = 370 ms @ 44100).
For SCO, explicitly set maximum buffer to 40 ms, so that latency cannot
grow too large there. For A2DP duplex, set it to 80 ms for same reason.
These are close to the old 6*packet limit.
For BAP server audio sink, set buffering target so that we try to match
the target presentation delay. Also adjust requested node latency to be
smaller than the delay.
Also fix BAP transport presentation delay value parsing, and parse also
the other BAP transport properties. Of these, transport latency value
needs to be taken into account in the total sink latency.
Use a 0 mask to handle unknown layouts. When the source or destination
is an unknown layout, pair, distribute or average. When pairing, keep
track how we paired and use that to construct the matrix later.
This fixes [ UNK UNK ] -> [ FL FR ] mapping by pairing.
Handle MONO layout as a real layout, not just like FC. This means it
does not share the FC mixing weights.
Only distribute and combine MONO channels when the target is also
MONO, enable normalization in this case.
Otherwise downmix and upmix the mono channels like any other channel,
which will make it respect the upmix and other settings.
Change some tests with this new way of doing things.
Fixes#3010
