Make an eventfd for each node and listen for events when the node
is activated.
Reorganize some graphs links to make it possible to activiate nodes
by signaling the eventfd
Pass the peer node to each remote node and let the remote node
directly activate the peer when needed.
Let each node signal the driver node when finished.
With this we don't need to go through the daemon to schedule the
graph, nodes will simply activate eachother. We only go to the
server when there is a server node to schedule.
Keep stats about the state of each node and the time it was
triggered, running and finished.
Allocate a per node a piece of shared memory where we place the
activation structure with the graph state and io_position.
We can then give this info to nodes so that they can get the position
in the graph directly but also later, activate the next node in
the graph.
Automatically parse and build key/value when in objects without having
to prefix the key with ":"
Automatically build control/value when in sequence without the "."
prefix.
Remove the builder with key/pod, taking a reference to the stack built
temporary pods is not allowed in c++. We can use the varargs version
with the same convenient syntax.
Remove the parser "*" option, it is unused.
Improve spa_pod_builder_add_* and spa_pod_parser_get_* and make them
look similar.
The driver root is the graph that should be run when a remote driver
starts the graph and can be different from the node itself when it's
part of a subgraph like in client-stream.
Make a profile param that can be used to configure a node with a
certain profile.
Use the profile to configure the ports on splitter and merger and
get rid of the dynamic ports.
Use the profile to configure the client-node and audio-dsp.
Don't try to link more ports than available between client-node and
dsp.
Remove the split/merge functionality from fmtconvert.
Make audioconvert use the splitter/merger.
Let client-stream decide what mode to configure audioconvert in.
Make special format types for planar and interleaved instead of
having a field.
Add enum for audio channel positions
Add some default audio channel layouts
Place the channel layout in the audio format when possible
alsa: place audio channel positions in format
Add sse optimized channel mixing for some common cases
Remove name from port info, it's not mandatory and in the properties
Add direction to port info
Put a pw_control object on all ports that support control messages.
Allow linking of control objects (and thus ports).
Make export-sink generate control messages to update the contrast
property.
spa_io_control_range -> spa_io_range
Add helpers for sequence and control pod objects
Implement control properties in audiotestsrc, update test example.
SPA_ID_* -> SPA_TYPE_*. We use 'type' for all registered types.
ID refers to either a registered type or an enum.
Improve introspection of object ids. Make 0 an invalid enum and
use it to look up the object id type.
Move some type-info files together.
Swap type and id of the object, we first specify the type and then
the id because the meaning of the id depends on the object type.
Param ids and IO ids are now simple enums.
Move some type info in one place, delete some type-info files
Fix type debug
Make audio layout an enum
Mark more enums as enums in types so they show us with their names in
the debug.
Do not use dynamic types anymore. The reason is that it's difficult:
- to maintain a shared type database over a network.
- the extra overhead when translating between processes and for
maintaining the translation tables.
- race conditions in translating in RT-threads, this is a problem
because we want to make event streams.
We now have simple enums with types and extension points for all
types. This is also nicer to use in general.
We don't need the mapper anymore or pass strings around as types.
There is a parallel type info system to get more info about ids and
enums and their hierarchy. It can also be used for debugging.
Make it possible to assign an arbitary node as the port mixer.
Also remove dynamically added ports.
Improve negotiation and allocation on the mixer ports
Add some more SSE optimisations
Move float mixer from the audio dsp to the port
Remove pw_node_get_free_port() and do things more explicitly.
Handle mixer ports in client-node
Keep track of queued data in the stream
Pass delay field around to make it possible to know about the raw
software read/write pointer and the hardware one.
Start stream position at 0
Make sure we set the right parent id and owner for the client node
and client-stream
Move the remote id in the proxy object and use it in the stream to
get the node id.
