[PATCH] allow-passthrough module

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Fri, 2015-12-11 at 17:08 +0100, Guillaume Desmottes wrote:
> Hi there,
> 
> A while ago we (Collabora) implemented a module for Valve changing
> the
> default policy regarding passthrough streams.
> You can find some context on the original thread: http://lists.freede
> sk
> top.org/archives/pulseaudio-discuss/2014-May/020644.html
> 
> This module never reached upstream but is still shipped and used by
> default on SteamOS.
> 
> I'd be interested resurrecting this work and bring it to a state it
> could be merged upstream.
> What would be needed to reach this? Would fixing Arun's comments from
> h
> ttp://lists.freedesktop.org/archives/pulseaudio-discuss/2014
> -May/020745.html be enough?

Re-enumerating those, plus additional comments:

1. You need to deal with the passthrough sink moving to another sink as
well (pretty much the same as the passthrough sink being unlinked)

2. If the sink goes away the corresponding null-sink should too. You
may find a situation where a passthrough stream on this sink moves to
another sink successfully, in which case you might need to set up a
null sink for the new sink and apply the same logic as a new
passthrough stream

3. There's a "fake-passthrough" hack that has a FIXME against it -- not
sure what this is for

4. The (stream != i) condition in passthrough_stream_removed() seems
like a noop? i-> sink will never be equal to null_sink

5. Might be nice to name the null sinks to have the "parent" sink's
name

This isn't a blocker, but I'd also like a more descriptive name than
allow-passthrough, but I can't think of one. :) Maybe someone else has
a suggestion.

-- Arun


[Index of Archives]     [Linux Audio Users]     [AMD Graphics]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux