- Fail to run `test_ops_context_register_device_collection_changed_twice_*` on my MacBook Air.
- A panic in `capi_register_device_collection_changed` causes `EXC_BAD_INSTRUCTION`.
- Works fine if replacing `register_device_collection_changed: Option<unsafe extern "C" fn(..,) -> c_int>` to `register_device_collection_changed: unsafe extern "C" fn(..,) -> c_int`
- Test them in `AudioUnitContext` directly instead of calling them via `OPS` for now.
- Fail to run `test_configure_input_with_zero_latency_frames` and `test_configure_input` at the same time.
-`audiounit_set_buffer_size` cannot be called in parallel
- We should not set `kAudioDevicePropertyBufferFrameSize` in parallel when another stream using the same device with smaller buffer size is active. See [here][chg-buf-sz] for reference.
- *Buffer frame size* within same device may be overwritten (no matter the *AudioUnit*s are different or not) ?
- A dummy mutex like `Mutex<()>` should work (see [`test_dummy_mutex_multithread`][ocs-rust]) as what `owned_critical_section` does in [_C_ version][ocs], but it doens't has equivalent API for `assert_current_thread_owns`.
- It's hard to debug with the variables using `OwnedCriticalSection`. Within a test with a variable using `OwnedCriticalSection` that will get a panic, if the `OwnedCriticalSection` used in the test isn't be dropped **before** where the code get a panic, then the test might get a crash in `OwnedCriticalSection` rather than the line having a panic. One example is [`test_stream_get_panic_before_releasing_mutex`](src/backend/test.rs). The tests must be created very carefully.
- Unworkable API: [`dispatch_async`][dis-async] and [`dispatch_sync`][dis-sync]
- The second parameter of [`dispatch_async`][dis-async] and [`dispatch_sync`][dis-sync] is [`dispatch_block_t`][dis-block], which is defined by `typedef void (^dispatch_block_t)(void)`.
- For now, we implement an API [`async_dispatch`][async-dis] and [`sync_dispatch`][sync-dis] to replace [`dispatch_async`][dis-async] and [`dispatch_sync`][dis-sync] (prototype on [gist][osx-dis-gist].)
- [`sync_dispatch`][sync-dis] is based on [`dispatch_sync_f`][dis-sync-f].
- [`async_dispatch`][async-dis] and [`sync_dispatch`][sync-dis] take [_Rust closures_][rs-closure], instead of [Apple's _block_][apple-block], as one of their parameters.
- The [_Rust closure_][rs-closure] (it's actually a struct) will be `box`ed, which means the _closure_ will be moved into heap, so the _closure_ cannot be optimized as _inline_ code. (Need to find a way to optimize it?)
- Since the _closure_ will be run on an asynchronous thread, we need to move the _closure_ to heap to make sure it's alive and then it will be destroyed after the task of the _closure_ is done.
- Borrowing Issues
1. Pass `AudioUnitContext` across threads. In _C_ version, we [pass the pointer to `cubeb` context across threads][cubeb-au-ptr-across-threads], but it's forbidden in _Rust_. A workaround here is to
2. We have a [`mutex`][ocs-rust] in `AudioUnitContext`, and we have a _reference_ to `AudioUnitContext` in `AudioUnitStream`. To sync what we do in [_C version_][cubeb-au-init-stream], we need to _lock_ the `mutex` in `AudioUnitContext` then pass a _reference_ to `AudioUnitContext` to `AudioUnitStream::new(...)`. To _lock_ the `mutex` in `AudioUnitContext`, we call `AutoLock::new(&mut AudioUnitContext.mutex)`. That is, we will borrow a reference to `AudioUnitContext` as a mutable first then borrow it again. It's forbidden in _Rust_. Some workarounds are
1. Replace `AutoLock` by calling `mutex.lock()` and `mutex.unlock()` explicitly.
2. Save the pointer to `mutex` first, then call `AutoLock::new(unsafe { &mut (*mutex_ptr) })`.
3. Cast immutable reference to a `*const` then to a `*mut`: `pthread_mutex_lock(&self.mutex as *const pthread_mutex_t as *mut pthread_mutex_t)`
- The APIs depending on `audiounit_set_buffer_size` cannot be called in parallel
-`kAudioDevicePropertyBufferFrameSize` cannot be set when another stream using the same device with smaller buffer size is active. See [here][chg-buf-sz] for reference.
- The *buffer frame size* within same device may be overwritten (no matter the *AudioUnit*s are different or not) ?
[integrate-with-cubeb]: https://github.com/ChunMinChang/cubeb-coreaudio-rs/commit/e84c554f18ef054376134c79a112a84cb8f923b4 "patch for integrating within cubeb"
[c-ext-block]: https://en.wikipedia.org/wiki/Blocks_(C_language_extension) "Blocks: C language extension"
[apple-block]: https://developer.apple.com/library/archive/documentation/Cocoa/Conceptual/ProgrammingWithObjectiveC/WorkingwithBlocks/WorkingwithBlocks.html "Working with Blocks"
[cubeb-backend-stm-reg-dev-chg-cb]: cubeb-backend-stream_register_device_changed_callback.diff "Implementation of stream_register_device_changed_callback"
[cubeb-pulse-rs-reg-dev-chg-cb]: cubeb-pulse-rs-register_device_changed_callback.diff "Impelement of register_device_changed_callback"
