The convolver will parse integers as a sample delay (as before) but will
now also fall back to parsing it as a float in the time domain.
Specifying the delay in time is better because it does not depend on the
samplerate of the graph or the IR.
Store the biquad type in the biquad, remove some unused fields.
Init all the biquads with the identify before loading the channel
specific biquads. We will process the max of all the channel biquads.
Make sure we skip processing for the indentity ones.
When we allocate data on an output port, always set it immediately on
the plugin. We let output nodes allocate data and input ports consume
it.
When we have no data on an output port, use discard data or NULL.
This ensure that the ports only have data when they are used by an input
port and otherwise use discards or NULL data.
Add sse optimized biquads.
Make a new dsp function to run multiple biquads on multiple channels.
This makes it possible to unroll some operations and run the channels
in parallel later.
Simply load a filter-chain with the new param_eq node and the
given filename in the config.
This fixes a number of issues such as not copying global properties to the
streams, setting a unique node.name etc...
Don't use %f to serialize floats to JSON but use the json formatter
because in some locales, the decimal point becomes a , which does not
parse as a float anymore.
Also reformat some lines.
Add a new followerClock block in the profiler info. This is only set
when the follower could be a driver and it contains the clock info used
for following the driver, mostly the rate difference and delay.
Dump this info in pw-profiler -J
Make sure we always set the info in the clock, especially also when we
are following.
Use node.param.Props for the patternType on videotestsr.
Make the audiotestsrc non-live. Otherwise it will be live but not a
driver and fail to schedule things.
See #4341
It's broken because it only iterates one Props property and most nodes
have an adapter, and so two sets of Props.
The node.param.Props = { } property is a better way of doing things
these days.
Flush with drain calls the drained callback for each cycle until paused
or resumed. Setting the stream to active again, clears the drained state
and makes things resume.
Flush without drain does not set the state to PAUSED but simply clears
the queued data. This is mostly useful when pausing or stopping.
At no point should the flush operation result in a PAUSED state change.
Put the newline and comma right after printing the line so that it gets
flushed immediately. Otherwise we only see the last line in the next
cycle. We then need to end the JSON output with an empty object.
Pretty print the node status instead of the value.
Convert the RequestProcess event to a command and send this to the
driver node. This ensures that any future properties on the Event will be
passed to the Command as well, such as timestamps etc.
Save the complete RequestProcess command when we need to send it later
when the node is RUNNING so that we preserver the properties.
Make a function to emit an event from a stream. Use this function
internally to emit the RequestProcess.
Make the RequestProcess event construction a little more efficient.
When combine-stream initiation code was moved around in b46673b4, a
bitwise or of flags was accidentally dropped, and thus flags were
overwritten instead of added to.
Don't queue an async state change completion for exported nodes. The
server sends a ping to check for completion and we want this ping reply
to happen after the state completion.
Consider the case where we have a follower and a driver, the follower is
sent the Start/Ping commands and replies to the ping but is still
processing the state change async. The server can then Start the driver,
which will then try to schedule the (still starting) follower and fail.
We could add the ping to the work queue as well but that creates
complications because modules (clients) and server share the same work
queues right now and block each other completions.
We could also make a method to process the work queue immediately but
that would be dangerous as well because it could contain a BUSY item
from some module that would block things.
Set the corked state on streams so that we can use this in sink-input
and source-output info without guessing.
The problem is that when a stream starts un-corked, the state is less
than RUNNING and so before this patch, pulse-server reports a corked
stream, which is not what pulseaudio reports.
When we receive a message with fds and we are at the end of the
buffer, we will call clear_buffer, which moves the next fds over the
fds of this message before we copy the fds into the message. This
results in the fd being leaked and the message using the fd of the next
message instead.
Avoid this by first copying the fds into the message and then move the
new ones over the old ones.
This fixes some wrong fds being used by clients.
In line also with da8e207de9,
make sure that the rtp-sap receive socket listens only to the configured
sap.ip in unicast mode, instead of accepting all packets on the port.
This additionally fixes breakage that was inadvertently introduced in
f2f204d604. Because the
`struct sockaddr_storage *sa` argument in `make_recv_socket` points
to `impl->sap_addr`, changing the address to INADDR_ANY in the unicast
code path would also silently cause the sender socket to try to
connect to INADDR_ANY and fail.
