Move some of the tracking code for the DLL to where it is used.
Add resync.ms (default 10) option at which we give up rate adjusting
and instead do a hard resync. This results in a jump in the position
of the graph clock.
Devices may advertise other values, but not certain they will work well
in duplex configuration.
E.g. my Samsung Galaxy Buds2 Pro emits buzzing sound with 48kHz duplex
input.
Don't believe QoS values recommended by the device, which may be
suboptimal. Instead, pick the values from the BAP v1.0.1 Table 5.2.
Link: https://github.com/bluez/bluez/issues/713
The PAC profile UUIDs do not appear in the UUID list, but are still
useful to know before SelectProperties.
Set them ahead of time based on the visible remote endpoints.
The "default" codec is the one with fill_caps != NULL, and should be
picked if we don't know which one we are using.
Fixes showing AAC-ELD as supported when it's not, which happened because
it's ordered before the default AAC in the codec list unlike the other
"shared endpoint" codecs.
We need a database of pixel format properties to remove hardcoded magic
numbers. This commit creates a simple function to query the bytes per
pixel for common RGBA formats.
This should be promoted to or superseeded by a common spa namespaced
version.
This splots the format enumeration into a helper for dsp and raw pod
creation and a common function for the compute filters.
Counting formats with modifiers will be usefull to announce dsp and
raw formats from the same node.
No functional changes.
Validation of properties is moved from the import and allocation
functions into a separate function. This allows later separation of
modifier informations from the base struct.
This extension is required for the aquire and release barriers
to transfer the buffer ownership from and two an external queue.
This removes a warning from the validation layers.
We can increase the MAX_LATENCY again if we increase the amount of
buffers when we are using a small buffer.
Normally we ask for 4 * quantum-limit as the buffer. This should be good
to use 1 buffer and quantum-limit as the quantum with enough headroom
to not run out of buffers.
If we are however using less buffer-frames we need to be careful and
allocate an extra buffer. Imagine using a buffer of 4096 frames, we can
support a quantum of up to 2048 frames if we use 2 buffers.
See #3744
Half of the buffersize is not enough to support as a max-quantum, we
need to divide by (4 * frame_scale) to allow some headroom and account
for the DSD scaling. We do the same calculation to suggest a buffer size
using the quantum-limit.
See #3744
When the filter has no format property, just enumerate all possible
framerates. Handle error case where the filter has the wrong type.
Makes gst-launch gstpipewiresrc ! video/x-raw ! fakesink work.
See #1793
Actually count the number of frame fractions we add. If we added 0, we
don't have any supported framerate that intersects with the filter and we
try the next frame size.
See #1793
Not waiting for HFP when no HFP backend should be checked via
adapter_connectable_profiles in spa_bt_device_check_profiles where the
relevant logic is.
Cleanup by moving the checks there.
When starting the converter, calculate the initial size needed by
the resampler to fill one quantum.
This makes it possible to get the requested amount of samples before
the first process call is made.
Unknown transports visible in DBus usually belong to a different
sound server instance that is talking to BlueZ.
Explain this in the warning message that we log, so that people can more
easily understand why things are not working.
When freewheeling we will immediately schedule a new graph cycle when we
get a process call because the graph completed.
When the process call is not done, because of some xrun or
because some node was removed that causes the graph to fail completion,
The next cycle will happen after a timeout.
This timeout was calculated as the ideal wakeup time (after a quantum of
time) and would accumulate for each timeout. The result is that the
timeout ended up far in the future and would stall the freewheel driver
for a long time.
Fix this by always setting the next timeout to wakeup time + freewheel.timeout
seconds. Also add a config property for the timeout (10 seconds, like
jack2 by default).
Fix the following compiler warning:
| In file included from /usr/include/spa-0.2/spa/utils/dict.h:14,
| from ../src/util_pipewire_objects.c:15:
| /usr/include/spa-0.2/spa/utils/defs.h: In function 'spa_ptr_inside_and_aligned':
| /usr/include/spa-0.2/spa/utils/defs.h:275:56: error: conversion to 'long unsigned int' from 'long int' may change the sign of the result [-Werror=sign-conversion]
| 275 | #define SPA_PTR_ALIGNMENT(p,align) ((intptr_t)(p) & ((align)-1))
| | ^
| /usr/include/spa-0.2/spa/utils/defs.h:276:42: note: in expansion of macro 'SPA_PTR_ALIGNMENT'
| 276 | #define SPA_IS_ALIGNED(p,align) (SPA_PTR_ALIGNMENT(p,align) == 0)
| | ^~~~~~~~~~~~~~~~~
| /usr/include/spa-0.2/spa/utils/defs.h:308:13: note: in expansion of macro 'SPA_IS_ALIGNED'
| 308 | if (SPA_IS_ALIGNED(p2, align)) {
| | ^~~~~~~~~~~~~~
In multi-ASE configurations there can be multiple transports per device,
each corresponding to different channels.
Emit sink/source nodes for each BAP transport present.
Combine them into a single sink/source in the same way as we do for
device sets.
For multi-ASE configurations, BlueZ does the channel allocation itself,
and passes us the result in the ChannelAllocation parameter.
If it is present, don't do the allocation ourselves but use that value
instead.
