Waking the other side on async consumer / producer close #30
Conversation
| let t0 = std::thread::spawn(move || { | ||
| execute!(async { | ||
| drop(prod); | ||
| assert_eq!(stage.fetch_add(1, Ordering::SeqCst), 0); | ||
| }); | ||
| }); | ||
| let t1 = std::thread::spawn(move || { | ||
| execute!(async { | ||
| cons.wait_occupied(1).await; | ||
| assert_eq!(stage_clone.fetch_add(1, Ordering::SeqCst), 1); | ||
| assert!(cons.is_closed()); | ||
| }); | ||
| }); |
There was a problem hiding this comment.
Must separate as two polling, otherwise waking the wrong party also passes this test.
There was a problem hiding this comment.
Good note, join!-ing multiple futures can actually lead to falsely passing this test because of spurious wake-ups.
Also the usage of some async runtime (like Tokio or async-std) may be considered for testing. Each async block then can be ran in separate task.
| pub trait AsyncRingBuffer: RingBuffer + AsyncProducer + AsyncConsumer { | ||
| fn wake_producer(&self); | ||
| fn wake_consumer(&self); | ||
| } |
There was a problem hiding this comment.
Add two methods for waking producer / consumer for trait AsyncRingBuffer
There was a problem hiding this comment.
I doubt that such low-level controls should be exposed to user but I can't suggest a better way to do this yet.
There was a problem hiding this comment.
Yes. Just realized it's interesting that Rust does not actually have private trait fn.
|
Okay, the request looks good. I've also began to write test on this, but yours seem to be better. I'm merging it, thank you for your contribution! |
No description provided.