Hi,
I just joined to the mailing list, so my answer might not go to the
original thread.
> Hi,
>
> I try to understand the configuration of the PulseAudio Murphy IVI
> module. Right now I try to wrap my head around routing_groups and
> nodes. This is my understanding so far, please correct me if I am
> wrong:
>
> (1) As I understand, an application and a device can be a node, right?
That's correct. All possible routing start and endpoints are a nodes.
>
> (2) OK, a routing_group is set of nodes, right?
A routing group is a prioritized list of possible routing endpoints.
Source nodes are assigned to a single routing group based on their node
type. Routing groups are used for implicit, ie. default, routing. When
implicit routing is used first the routing group is picked based on the
type of the node that represents the stream. Next, the list of routing
endpoints in that routing group walked through in priority order and the
highest priority available node is picked as the routing target for the
input stream. In the last step the input stream is routed to the sink of
the picked routing target.
However, explicit routes requested directly by the applications do not
use the routing groups.
>
> (3) Application_class.roles define the resources an application wants
> to bind to itself, right?
That's correct. The applicable resource definitions are picked either by
a matching media.role property or binary name, ie, either the media.role
property of the stream or the binary name of the application must match
one of the entries in the relevant list in the application_class
definition.
>
>
> What I don't understand:
>
> (A) What is the purpose of the routing_group.compare method?
As we said earlier a routing group is basically a prioritized list of
possible routing targets. The list is defined by two functions: 'accept'
and 'compare'.
The 'accept' function is used to decide whether a node belong to the
routing group, eg. when an output node is created the 'accept' function
is called for each routing group and the node is added to the routing
group if 'accept' would return 'true'.
'compare' is used to find for the newly created node the insertion point
in the prioritized list. 'compare' takes two nodes as input and return
-1, 0 or +1 just like 'strcmp' does.
>
> (B) How do I map nodes, like phone input, to real sources or sink
> (devices)?
For streams the media.role property or the binary name can be used.
For dynamically loaded device nodes there are hardwired heuristics to
determine the node type. For manually loaded alsa devices a node.type
sink property can be specified. For instance in your default.pa you
might load a sink like:
load-module module-alsa-sink device=hw:0,1 name=speakers
sink_properties="node.type='speakers'"
Currently the following node.types are supported: 'speakers',
'front-speakers', 'rear-speakers', 'microphone', 'jack', 'hdmi' and
'spdif'
Phone inputs/outputs, if Bluez is used, are handled by the
hardwired heuristics and for the phone inputs/outputs loop-back
streams are automatically created. In turn those are routed implicitly.
In case you wanted to change the loop-back stuff, pls. consult the
pa_classify_loopback_stream() function in classify.c. If needed, the
table there can be made configurable.
>
> (C) How is the data flow coordinated between nodes?
Could you be pls. a bit more specific on this?
>
>
> I hope someone can help me with my questions.
>
>
> Cheers,
> Robert
br
-janos-
---------------------------------------------------------------------
Intel Finland Oy
Registered Address: PL 281, 00181 Helsinki
Business Identity Code: 0357606 - 4
Domiciled in Helsinki
This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
_______________________________________________
IVI mailing list
[email protected]
https://lists.tizen.org/listinfo/ivi
