pipewire will allocate buffers aligned to the max alignment required for
the CPU. Take this into account and don't expect larger alignment.
Fixes a warning in mixer-dsp when the CPU max alignment is 16 but the
plugin requires 32 bytes alignment for the AVX2 path (that would never
be chosen on the CPU).
See #2074
Parse the quantum_limit parameters and use this to scale the buffers so
that they can contain the maximum allowed samples instead of the
hardcoded 8192 value.
See #1931
Make the alignment parameter optional when negotiating buffers.
Default to a 16 bytes alignment and adjust for the max cpu
alignment.
Remove the useless align buffer parameter in plugins, we always
set it to 16 anyway.
When we don't have any input buffers, recalculate the rate match
size field so that we can know the size of the expected buffer.
We already do this when starting but it might have been done with
a different quantum.
Scale the buffer size with the rate conversion ratio. Also make sure
that we can at least produce a maximum quantum of samples.
If we have large upconversion (8KH -> 48KHz) and small input
buffers, we would not allocate enough space for the output buffer
and cause xruns in the sink.
Fixes#1809
Follow the rate of the _io_position area and adjust the resampler
to match. This ensures that we always process at the DSP samplerate
to the target negotiated fixed rate of the device/stream.
When we add a new listener to an object, it will emit the full state
of the object. For this it temporarily sets the change_mask to all
changes. Restore the previous state after this or else we might not
emit the right change_mask for the next listener.
Consider the case where one there are two listeners on an object.
The object emits a change and the first listener wants to enumerate the
changed params. For this is adds a new listener and then triggers the
enumeration. If we set the change_mask to 0 after adding the listener,
the second listener would get a 0 change_mask and fail to update
its state.
This replaces the manual check for "true" and some (inconsistent) return value
of atoi. All those instances now require either "true" or "1" to parse as
true, any other value (including NULL) is boolean false.
SPA_MEMBER is misleading, all we're doing here is pointer+offset and a
type-casting the result. Rename to SPA_PTROFF which is more expressive (and
has the same number of characters so we don't need to re-indent).
Avoid calling memset on a large piece of memory when draining
the resampler because it might use up all the allocated time in
our realtime thread. Instead use a prealloced empty buffer.
Take the queued input samples into account when calculating the
required input size. This can be 0 when there is still enough
data queued in the input for another period.
Handle 0 read_size in alsa-source and make it push out a 0 buffer,
this will then drain the resampler and make it ask for a new buffer
size. This makes the transition from one period to another more
seamless for the resampler.
Fixes#805
Instead of requiring the upstream node to resubmit the delayed
samples, keep the samples ourselves. The benefit is probably too
small to measure but it simplifies things a lot.
This is needed for example for Clang compiler which uses different
annotations than GCC. It will make WebRTC to happily use PipeWire
since the spa library is header-only and WebRTC defaults to use
Clang with -Wimplicit-fallthrough.
Make a new DRAINED status.
Place the DRAINED status on an input IO when a stream is out of
buffers and draining.
All nodes that don't have HAVE_DATA on the input io need to copy
it to the output io and return the status. This makes sure the
DRAINED is forwarded and nodes return DRAINED from _process()
DRAINED on the resampler flushes out the last queued samples and then
forwards the DRAINED in the next iteration.
Emit a new drained signal from the context when a node returns
DRAINED. Use this to trigger the drained signal in the stream.
