e78a86d30b
Cleanup of the CMake script.
2026-01-02 15:50:15 +10:00
46b2cdd0cf
Improve SDL2 integration in the CMake script.
2026-01-02 14:44:26 +10:00
ce41f6cfc9
SDL2: Fix a compilation error due.
...
This is due to calling an internal miniaudio function.
2026-01-02 14:14:10 +10:00
da764a5a28
PipeWire: Fix an error when runtime linking is disabled.
2026-01-02 12:34:42 +10:00
e9fad62f44
Improve default device enumeration with the SDL2 backend.
2026-01-01 19:36:37 +10:00
d0392288c4
Remove an old device job thread object.
...
This was used by AAudio for dealing with rerouting, but with the new
backend architecture it is no longer needed.
2026-01-01 19:04:49 +10:00
a5ef023cee
Remove some global variables.
...
These are replaced by functions. The problem is that declaring a
variable as MA_API will result in errors when it's declared as static.
2026-01-01 18:39:43 +10:00
45c7a64299
Add support for device selection the deviceio test.
2026-01-01 15:58:37 +10:00
1d308f69bc
Add a basic single-threaded test to deviceio test.
2026-01-01 14:52:01 +10:00
db6bc371eb
WebAudio: Remove dependency on -sASYNCIFY for the Audio Worklets path.
...
Public issue https://github.com/mackron/miniaudio/issues/863
2026-01-01 13:10:15 +10:00
1f2e59548b
Web Audio: Minor change for multi-threaded mode.
2026-01-01 10:47:39 +10:00
edc44cbcaa
PipeWire: Optimization to multi-threaded mode.
2026-01-01 09:38:21 +10:00
1d0a598485
Rename MA_THREADING_MODE_MULTITHREADED.
...
This not having an underscore while SINGLE_THREADED did was driving me
crazy.
2026-01-01 09:19:43 +10:00
ff9ff16341
CMake: Remove is_backend_enabled()
...
This is no longer necessary since we can just use MINIAUDIO_NO_[BACKEND]
as the standard way to check if a backend is enabled.
2025-12-31 20:25:11 +10:00
7cbb204abb
CMake: Experiment to make it easier to check if a backend is enabled.
2025-12-31 20:17:46 +10:00
0e6eaae06b
CMake: Don't compile the the PipeWire backend if device IO is disabled.
2025-12-31 19:34:58 +10:00
c6429bb866
Try fixing a CMake error.
2025-12-31 19:24:16 +10:00
552cf613b3
PipeWire: Add support for disabling runtime linking.
2025-12-31 18:19:52 +10:00
114b9594c1
PipeWire: Finish work on removing the SPA dependency.
2025-12-31 18:18:17 +10:00
e1dfc9d77f
PipeWire: Remove dependency on spa_format_audio_raw_parse().
2025-12-31 13:30:49 +10:00
d2a93567df
PipeWire: Remove dependency on spa_pod_builder.
2025-12-31 09:04:00 +10:00
c67b29de38
PipeWire: More work on removing the SPA dependency.
2025-12-30 18:00:07 +10:00
9ebc3efa95
PipeWire: First tentative step towards removing the SPA dependency.
2025-12-30 16:40:16 +10:00
e086de4eb3
Web Audio: Get single-threaded mode working to spec.
2025-12-30 14:54:36 +10:00
07af5ea6f7
Remove an unnecessary function.
...
The naming of this function is now a bit misleading with the new changes
to the backend system.
2025-12-30 10:32:09 +10:00
f073a40a1c
Fix some warnings for the PipeWire backend.
2025-12-30 10:18:56 +10:00
d04c1985d6
Fix an error with the Web Audio backend.
2025-12-30 10:13:44 +10:00
cc4db76c26
Minor update to documentation for the single-threaded example.
2025-12-30 09:32:11 +10:00
a16756ec3b
Add an example for single-threaded mode.
2025-12-30 09:26:38 +10:00
350cc0be1f
Fix a bug with starting and stopping a device in single-threaded mode.
2025-12-30 09:25:37 +10:00
7483e55aee
Add ma_device_get_threading_mode().
2025-12-29 19:39:29 +10:00
0f4bc3ca9a
Add support for configuring the threading mode for a device.
2025-12-29 19:12:34 +10:00
b0894c34b1
Add ma_device_step() in preparation for single-threaded mode.
2025-12-29 19:02:18 +10:00
fa2c1b49b3
Remove unused callbacks from the device backend system.
2025-12-29 18:02:51 +10:00
3d79043d7d
WASAPI: Move thread characteristic stuff from start/stop to init/uninit.
...
All audio functions now run from the same thread, so this feels like a
more appropriate location for these functions to me.
2025-12-29 15:43:27 +10:00
9df9b19a59
WASAPI: Refactoring of the device rerouting system.
...
This now performs rerouting from the step callback which is always done
from the main thread. This should address issues relating to COM
initialization and some rerouting race conditions.
There is a very slight change in behaviour here. Previously when a
device is detached (such as headphones being unplugged), the device
would actually be stopped proper, but now it is no longer stopped and
the stop notification is not fired when the device is deactivated.
2025-12-29 15:32:23 +10:00
d72811c127
WASAPI: Remove some unnecessary variables.
2025-12-29 10:35:00 +10:00
aaf868f92a
WASAPI: Refactoring for the new backend architecture.
2025-12-29 10:31:25 +10:00
d0b8b07c49
DirectSound: Refactoring for the new backend architecture.
2025-12-28 19:55:59 +10:00
a4406088f4
WinMM: Refactoring for the new backend architecture.
2025-12-28 11:07:55 +10:00
c6d44b67e1
PulseAudio: Stop firing the stop callback in response to a suspension.
2025-12-26 19:26:43 +10:00
521c224b02
JACK: Stop double-posting the stop callback.
2025-12-26 19:16:55 +10:00
5d1c089d15
OpenSL: Stop double-posting the stop callback.
2025-12-26 19:16:43 +10:00
f270f515e4
AAudio: Stop double-posting the stop callback.
2025-12-26 19:15:03 +10:00
d10526fb3a
Core Audio: Stop double-posting stop callbacks.
2025-12-26 19:04:10 +10:00
ef59b93a4d
Fix a bug with asynchronous style backends.
...
This was resulting in glitching in duplex mode when internal sample
rates differ.
2025-12-26 18:36:11 +10:00
b6799e92e0
Update gitignore.
2025-12-26 12:03:50 +10:00
f932077982
Core Audio: Refactoring for the new backend architecture.
2025-12-26 12:02:57 +10:00
5296384289
ALSA: Fix an error with non-blocking mode.
2025-12-25 11:29:10 +10:00
4cb0851ccc
sndio: Refactoring for the new backend architecture.
2025-12-25 11:28:14 +10:00
David Reid