Drop client connections when pipewire goes away. pipewire-pulse daemon can remain running and pulseaudio clients will be able to connect again once pipewire is up and running.
Even though all event sources are kept in a list in the current
implementation of the event loop in the SPA support plugin, relying
on it freeing all sources could lead to unbounded memory use
if the profiler module is loaded and unloaded over and over again.
Handle if the global is destroyed (e.g. `pw-cli destroy X`) to
avoid a use-after-free in `pw_global_destroy()` when it is called
with a dangling pointer from `module_destroy()`.
When a global is removed, also remove the proxies. We can have multiple
proxies to a global otherwise when a client has a sequence of events
queued from the server like this:
Global 1 added
Global 1 removed
Global 1 added
The proxy we added in the firt event would not be destroyed because the
server did not know about it when global 1 was removed and then a
second one would be made.
Sort ports per node and then per node-specific port_id. This will
keep related ports together and will sort them based on when they
were created.
Fixes#1780
Simplify the key-value pair parsing for AT+IPHONEACCEV commands,
and move the handling of {key,indicator}-value pairs into
dedicated functions. Furthermore, fix a comment.
Reorganize the latency setup in one place, return a desired device
latency for use as quantum.
PulseAudio assigns half of the (tlength - minreq) latency to the sink
but we can't do that because our sinks have a max-quantum of latency.
Fix this by clamping our calculated sink latency to the quantum
PulseAudio subtracts the sink latency from the tlength in adjust latency
mode, so we need to do the same.
This makes PULSE_LATENCY_MSEC values bahave more like pulseaudio.
See #1769
Keep track of the negotiated parameters and use those to construct
the Buffers param.
Use the FrameAllocator to allocate buffer memory. Keep a request for
each buffer and queue them.
In requestComplete, put the request metadata info into the buffer and
header and write the finished buffer id to a ringbuffer. Then wake up
the data thread to pop the finished buffer_id and push it in the
graph.
Remove some old files.
Keep track of the proxies and remove them when the global is removed.
Otherwise we might end up in a case where the have 2 proxies when
an add/remove/add sequence of messages was queued.
When an object is created, it is marked creating until all roundtrips
complete. If the object is removed in between, we don't remove it
because find_object does not return creating objects.
Make find_object also return the creating objects to fix this.
The current compile-time-check only tests for the getrandom() syscall
wrapper of libc. The presence of this wrapper however does not relate to
the presence of the actual syscall at runtime.
gstreamer alsasink can change sample while stream is already created.
In that case, gstreamer calls snd_pcm_ioplug_callback_t::stop,
snd_pcm_ioplug_callback_t::hw_params and snd_pcm_ioplug_callback_t::prepare
Add new flag to the snd_pcm_pipewire_t that is set every time
snd_pcm_pipewire_hw_params is called to prevent using stream with old
sample rate
Signed-off-by: Martin Geier <martin.geier@streamunlimited.com>
pw->time.delay is delay in number of frames in pw->time.rate domain,
however snd_pcm_pipewire_delay function is suppose to return number of
frames in io->rate domain. Convert pw->time.delay to io->rate domain to
increase precision when the io->rate is not equal to the pw->time.rate
snd_pcm_pipewire_delay should return how many frames are queued in
pipewire, pw_stream_get_time returns numbers of the queued frames before
snd_pcm_pipewire_process is called, however this function inserts (or
removes) some frames from pipewire. Therefore newly inserted (removed)
frames should be added to pw->time.delay to increase precision.
Signed-off-by: Martin Geier <martin.geier@streamunlimited.com>
Instead of doing pointer arithmetic through actual pointers, cast them
to the proper integer types.
This is functionally equivalent to the previous code but is clearer for
the compiler.
Fixes#1018
