Andrew Kelley
45d5691254
Merge pull request #71 from michaelmaltese/feature-more-coreaudio-formats
...
coreaudio: support more formats: S16LE, S32LE, and Float64LE
2016-04-26 16:46:31 -07:00
Michael Maltese
8742fe215d
coreaudio: support more formats: S16LE, S32LE, and Float64LE
2016-04-26 16:24:37 -07:00
Michael Maltese
99f5b4e4b6
coreaudio: obey MAX_CHANNELS when creating devices
...
fixes #57 , which was caused by a buffer overflow with devices with more
than MAX_CHANNELS channels
2016-04-25 15:16:12 -07:00
Andrew Kelley
658c4f60c5
fix compile error with use of ATOMIC_FLAG_INIT
...
closes #60
2016-04-22 10:33:44 -07:00
Andrew Kelley
e8b908243d
add missing include directive
...
fixes compilation when no backends are available. closes #67
2016-04-22 10:24:32 -07:00
Andrew Kelley
ca0ebfda5c
os: fix not cleaning up fd resource when mmap fails
2016-02-17 18:05:08 -07:00
Andrew Kelley
c2836e494b
alsa: refactor for readability
2016-01-26 20:18:39 -07:00
Charles Lehner
cdadcee7ed
Handle ALSA devices with null description
2016-01-26 21:42:44 -05:00
Raphaël Londeix
a205359a08
Fix missing none backend initializer
2016-01-06 10:23:33 +01:00
Raphaël Londeix
ce60176e99
Go back to statically initialized arrays
2016-01-06 09:27:22 +01:00
Raphaël Londeix
45cd6a7af2
Fix WASAPI implementation when building in C++ mode.
2015-12-07 11:28:06 +01:00
Raphaël Londeix
699b00c427
Prevent WASAPI GUID from being re-defined in all compilation units.
2015-12-07 11:28:06 +01:00
Raphaël Londeix
8587a22b21
Fix compiler attributes for MSVC
2015-12-07 11:28:06 +01:00
Raphaël Londeix
d1eb1d8042
Do not use C99 array initialization syntax.
2015-12-07 11:28:06 +01:00
Raphaël Londeix
406ffa668c
Wrap atomic_flag usage behind SoundIo macros and types
...
As done for other atomic types, we hide atomic_flag real type in
order to use both the C and the C++ standard libraries.
2015-12-05 13:33:47 +00:00
Raphaël Londeix
af98891b4f
Use a different set of macros and types for atomics when compiled in C++
...
Previously, the `atomics.h` header relied solely on the presence of
the C11 specific header `stdatomic.h`. Now, when compiled in C++,
we use the `std::atomic` type family provided by the STL in C++11.
2015-12-05 13:33:47 +00:00
Raphaël Londeix
ff1af0e4cb
ALLOCATE_* family macros cast to the destination type
...
They were previously evaluated as an expression of `void *` type.
This allows a C++ compiler to assign it a variable of the
destination type.
2015-12-05 13:33:47 +00:00
Andrew Kelley
ce42b89da3
no longer ship config.h header
...
Instead provide these functions:
* soundio_version_string
* soundio_version_major
* soundio_version_minor
* soundio_version_patch
2015-11-23 13:51:25 -07:00
Andrew Kelley
72fde44073
PulseAudio: fix incorrect outstream software_latency
...
Previously, when using PulseAudio, unintended integer division
caused SoundIoOutStream::software_latency to be incorrect.
2015-11-20 23:51:16 -07:00
Andrew Kelley
f61acfd953
ALSA: fix potential cleanup deadlock
...
The main ALSA poll now includes another file descriptor which we
write to when cleaning up, to ensure that the poll terminates
when soundio_outstream_destroy is invoked.
closes #36
2015-11-14 22:16:29 -07:00
Andrew Kelley
8f1e1b8752
ALSA: let alsa lib choose period settings
...
Closes #46
2015-11-14 21:48:58 -07:00
Andrew Kelley
d168a7d192
ALSA: better device detection
...
* Don't suppress sysdefault.
* If default and sysdefault are missing, use the first device as
the default device.
* In sio_record example, don't segfault when device not found or
probe error.
* Workaround for Raspberry Pi driver that incorrectly reports
itself as Output when it is actually Input.
See #44 .
2015-11-10 04:23:27 -07:00
Andrew Kelley
ee7c0d3e11
convert source code to pure C
...
List<T> is now a really ugly macro.
Added a workaround for jack.h not putting `void` in function
prototypes for functions that take no arguments. I made upstream
pull requests to jack1 and jack2 but I don't have high hopes
about them getting merged.
I removed the lock-free atomic asserts. clang reports
non-lock-free atomics when in fact it does have lock-free
atomics. I inspected the generated code for gcc and clang
for fetch_add, load, and store, on x86_64 and armhf, and
it's all lock free.
Closes #45 .
2015-11-10 02:24:18 -07:00
Andrew Kelley
2633b1bd60
PulseAudio: pass the latency test
2015-11-01 12:39:21 -07:00
Andrew Kelley
8e30053962
JACK: delete broken pause implementation
...
Previously, calling soundio_outstream_pause or soundio_instream_pause
during the write_callback or read_callback would cause a deadlock. Now,
attempting to pause always results in SoundIoErrorBackendIncompatible.
Closes #39
2015-10-28 16:01:08 -07:00
Andrew Kelley
2f14aa569e
ALSA: use period size to calculate buffer size
...
determining the buffer size before the period size sometimes results
in EINVAL.
closes #33
2015-10-13 00:47:18 -07:00
Andrew Kelley
db0e3d136c
fix double clean up corruption when opening stream fails
...
affects ALSA, CoreAudio, JACK, WASAPI, and Dummy.
2015-10-08 17:10:21 -07:00
Andrew Kelley
277bb3106b
dummy: fix deadlock when pause called from write_callback
2015-10-06 19:36:57 -07:00
Andrew Kelley
a3c4f85489
ALSA: fix deadlock under some circumstances
...
If the write_callback came upon an error during the prepare
phase, and never wrote any frames to the hardware buffer,
and another thread tried to soundio_outstream_destroy, this
would cause a deadlock.
This is now fixed.
2015-10-06 18:50:56 -07:00
Andrew Kelley
0a0715eb72
architecture independent header files
...
Do endian detection in the header file instead of cmake.
2015-09-24 09:26:17 -07:00
Andrew Kelley
be6e654098
fix and add test for soundio_device_nearest_sample_rate
2015-09-18 22:08:36 -07:00
Andrew Kelley
71e31b88fe
os: fix crash when creating non high priority thread fails
2015-09-17 15:04:34 -07:00
Andrew Kelley
dbed33abd2
os: remove superfluous condition on init
...
remove superfluous #include math.h
2015-09-17 13:55:47 -07:00
Andrew Kelley
f52080a1fd
fix mirrored memory not working on linux
...
add mirrored memory unit test
2015-09-14 14:56:13 -07:00
Andrew Kelley
7e95d9f9c0
PulseAudio: fix soundio_outstream_pause triggering assertion
...
when called from within write_callback. closes #24
2015-09-14 04:15:02 -07:00
Andrew Kelley
2a68f63079
ALSA: pause gives error when state is invalid
2015-09-11 20:33:40 -07:00
Andrew Kelley
4a0b3579d4
soundio_disconnect: allow NULL parameter
2015-09-11 19:10:09 -07:00
Andrew Kelley
0887b1ee3d
soundio_disconnect: null out more function pointers
2015-09-11 18:25:06 -07:00
Andrew Kelley
c3e3af2061
dummy backend: fix sometimes calling write_callback with 0 avail
2015-09-11 10:24:30 -07:00
Andrew Kelley
c62c2e4661
macos: avoid allocation when getting time
2015-09-11 10:20:46 -07:00
Andrew Kelley
b6c0c84294
macos: fix not cleaning up condition variables
2015-09-10 17:14:16 -07:00
Andrew Kelley
f0356597e2
ALSA: remove stray print statement
2015-09-08 15:05:44 -07:00
Andrew Kelley
369ea0f608
os: fix some variables not static
2015-09-08 10:03:23 -07:00
Andrew Kelley
749654b922
PulseAudio: better clear buffer implementation
...
seek instead of pa_stream_flush
2015-09-07 16:58:46 -07:00
Andrew Kelley
e2e483a4d5
no longer depend on or link against libm
2015-09-05 14:56:36 -07:00
Andrew Kelley
f8cb816139
ALSA: remove duplicate assert
...
Thanks to Devin Anderson for pointing it out.
2015-09-04 11:11:14 -07:00
Andrew Kelley
9cc6495306
JACK: fix infinite loop when refreshing devices
...
closes #9
2015-09-03 20:20:59 -07:00
Andrew Kelley
918183adc6
ALSA: be ready for /dev/snd to not be available
...
If this occurs, the system has no ALSA support in the kernel,
and it is appropriate for the backend to fail.
closes #6
2015-09-03 15:40:36 -07:00
Andrew Kelley
85c2190675
build: fix build on linux with clang
...
closes #8
2015-09-03 14:05:19 -07:00
Andrew Kelley
b55a225b9c
CoreAudio: pass the latency test
...
See #2
2015-09-03 10:47:38 -07:00