Wim Taymans
a4ff7296b2
remote: don't try to update the source when it's gone
2018-10-24 15:45:03 +02:00
Wim Taymans
f5c78cc2a2
remote: use right io_position enum
2018-10-24 15:44:39 +02:00
Wim Taymans
06fe3af7fb
update submodules
2018-10-24 10:35:41 +02:00
Wim Taymans
1ffee101fb
jack: use spa_io_position
2018-10-24 10:34:49 +02:00
Wim Taymans
723d4d9db7
client-node: use spa_io_position
...
Remove old internal structure
2018-10-24 10:34:28 +02:00
Wim Taymans
2206deb3e8
alsa: implement set_io and handle position io
...
Use position io for deciding the size of the output buffers in
alsasrc
2018-10-24 10:27:57 +02:00
Wim Taymans
4aab226047
io: add size of current cycle to position
2018-10-24 10:27:06 +02:00
Wim Taymans
0395897856
node: only run the graph when we are the selected driver
...
Update submodules
2018-10-23 16:52:28 +02:00
Wim Taymans
98992d88d3
jack: add jack_port_rename
2018-10-23 16:50:59 +02:00
Wim Taymans
2a1328533f
stream: use pw_stream_update_properties
2018-10-23 16:50:17 +02:00
Wim Taymans
34db9be04b
io: add io areas for position information
2018-10-23 16:14:17 +02:00
Wim Taymans
66de372e9a
add node set_io function
...
Also allow setting io areas on nodes. We need this to set the
graph clock and period size on the nodes.
2018-10-23 12:31:41 +02:00
Wim Taymans
a121164df9
stream: handle update_properties before connect
2018-10-22 16:11:57 +02:00
Wim Taymans
d2570142fa
resample: don't output 0 buffers
2018-10-22 11:32:53 +02:00
Wim Taymans
8b2781b494
submodule update
2018-10-19 17:15:56 +02:00
Wim Taymans
a14e16806f
mainloop-glib: destroy the source
2018-10-19 17:15:19 +02:00
Wim Taymans
926392a701
update pulseaudio
2018-10-19 17:06:33 +02:00
Wim Taymans
e4c94bf12d
proplist: fix compilation
2018-10-19 17:00:25 +02:00
Wim Taymans
01aa9a5960
introspect: implement enough to make gnome-control-center work
...
We can see sources and sink and change the volume/mute of streams
2018-10-19 16:57:03 +02:00
Wim Taymans
faccc8d506
implement more methods
...
Implement extensions with dummies
Implement some more introspection, enough to make pavucontrol start
Implement volume changes on sink_input
2018-10-19 13:30:20 +02:00
Wim Taymans
5498d9c726
glib-mainloop: implement glib mainloop support
2018-10-19 13:27:58 +02:00
Wim Taymans
693fd3a366
meson: only add dependencies when requested
2018-10-19 13:21:52 +02:00
Wim Taymans
e59a7c0848
media-session: set node volume
2018-10-18 15:16:59 +02:00
Wim Taymans
f2e03077a8
audioconvert: add volume and mute property
...
Expose volume and mute property in channelmix and use this
in audioconvert.
2018-10-18 15:04:40 +02:00
Wim Taymans
842abad9e1
client-stream: implement enum_param and set_param better
...
Make it possible to enumerate the parameters and get/set properties
when available on the adapter.
2018-10-18 12:59:31 +02:00
Wim Taymans
c0a6c83227
client-stream: add media.class on node
...
Add the media.class property on the node as well.
2018-10-18 12:56:10 +02:00
Wim Taymans
fd5e315a10
media-session: handle default media roles better
...
Use the media.class to derive a better role.
2018-10-18 12:53:30 +02:00
Wim Taymans
79e976e5d7
audioconvert: small cleanups
2018-10-18 12:47:46 +02:00
Wim Taymans
c6551cead4
protocol-native: return errno when server failed
...
this gives a better error code
2018-10-18 12:36:53 +02:00
Wim Taymans
c28d0f1282
protocol-native: -1 is for invalid fd, not 1
2018-10-18 12:36:44 +02:00
Wim Taymans
88d6e52cc7
add systemd socket activation
2018-10-18 12:36:10 +02:00
Wim Taymans
e29229b6e0
protocol-native: return errno when server failed
...
this gives a better error code
2018-10-18 11:15:27 +02:00
Wim Taymans
6266cd0f48
protocol-native: -1 is for invalid fd, not 1
2018-10-18 11:14:52 +02:00
Wim Taymans
e5dcacdefb
add systemd socket activation
2018-10-18 10:47:21 +02:00
Wim Taymans
a086fc5c0c
port: just number unknown channels
...
Ports with unknown channel layout are just numbered.
2018-10-17 12:45:50 +02:00
Wim Taymans
6590d3a258
alsa: sanitize channel map
...
Remove unknown channels
Remove duplicate channels
Fill unknown channels with default channels
2018-10-17 12:44:05 +02:00
Wim Taymans
e22762d502
link: also handle success case
2018-10-17 10:56:47 +02:00
Wim Taymans
4b4f85afe5
link: renegotiate when no format set
...
When the port has lost the format, renegotiate.
2018-10-17 10:40:03 +02:00
Wim Taymans
d27674722a
improve debug and error handling
2018-10-17 10:02:44 +02:00
Wim Taymans
6c7d2d0e1e
alsa-utils: enumerate channels using channel map
...
When there is a channelmap, return format descriptions for each layout
otherwise return a range.
2018-10-16 18:10:43 +02:00
Wim Taymans
7186bd2583
gst: remove old format id conversion code
2018-10-16 17:13:27 +02:00
Wim Taymans
e079552dc8
link: fix negotiation
...
Don't use the current best format as a filter to get the current format.
It might be different and then we fail.
2018-10-16 17:12:12 +02:00
Wim Taymans
d91e545e1d
stream: set state to unconnected when unconnecting
2018-10-16 17:08:19 +02:00
Wim Taymans
4615c5781a
test-v4l2: call process before looking at io
2018-10-16 17:07:49 +02:00
Wim Taymans
ad35e8ef3b
pod: fix pod compare
2018-10-16 16:50:06 +02:00
Wim Taymans
aa5bc5846f
v4l2: fix preoperty get
2018-10-16 16:49:41 +02:00
Wim Taymans
6267e69638
add uninstalled target
...
Also update submodules
2018-10-16 11:30:38 +02:00
Wim Taymans
7e513ea243
alsa: clamp the period size to something sensible
...
some clients (sweep) want to negotiate the smallest possible period
size we can support, which is expressed in bytes and independent
of samplerate, format and number of channels.
For files with high samplerate and channelcount, this results in periods
of 8 frames, which is too small to handle.
Instead round up the period size to something we can likely handle,
taking into account the samplerate and number of channels.
2018-10-16 10:57:53 +02:00
Wim Taymans
b742d7533e
stream: only queue buffer when filled
...
We need to completely fill the buffer before we queue the buffer. If we
don't do this, we might have reported a larger writable size than
what we actually can do and that makes clients fail.
This is not entirely completely correct because a client can expect to
have its buffer sent as soon as it completes the write. We should really
just queue the data in a separate queue when we run out of buffers.
2018-10-16 09:58:14 +02:00
Wim Taymans
f74a0fb9bd
stream: implement _set_name
2018-10-16 09:56:12 +02:00
