- If the stream is already live or going live, do not reset or
transition
- If the stream is testing, transition but do not reset
- If the stream is starting a test, error out since this can take a
while
- If the attempted transition was redunant, still return a success
Fixes#5403
This single line in the entire project causes a difference between
clang-format 12.0.0 and 12.0.1. So, just rearrange the code so it works
with both instead.
Microsoft Visual Studio 2019 ships with LLVM/Clang 12 as of Visual
Studio 2019 16.11.0 (August 10, 2021). LLVM/Clang 12 is available on all
platforms. This change should make it easier to have clang-format "just
work" on dev systems and be consistent across platforms without having
to use an outdated version.
Remove lastNotifyTime throttle. Could ignore last device notification.
Will bring back if necessary, but seemed like overkill.
Make isInputDevice const for safety.
Make isDefaultDevice atomic. Allows it to be accessed simultaneously
from settings and notification callback threads. Race between setting
change in OBS and Windows is not a problem, a signal will be sent after
both to consolidate. Worst that should happen is a redundant reconnect
cycle.
Only read/write default_id from notification callback for thread safety.
Do not store IAudioClient and IAudioCaptureClient onto the source object
until initialization is almost complete, right before the capture thread
is created. We don't to retain objects from failed attempts. Also clear
them when stopping.
The previous monitor recovery logic touches the source monitoring type,
resets the monitor even if the output device hasn't been invalidated,
and was susceptible to races. Monitor invalidation should live within
the monitor abstraction.
Avoid nanosecond abstraction to reduce math operations.
Replace Sleep(0) with YieldProcessor(). We want the thread to remain
scheduled, the current CPU core to do less work, and the hyperthread
sibling to perform better.
Hacky profiling showed maybe 10-25 µs skid reduction per function call.
I think power/performance gains would be hard to measure, so I haven't
tried, but it would be shocking if they got worse.
