David

> > may yield in nr * timeout time spent waiting > > I'm quite sure **io_uring_wait_cqe_timeout** only waits **timeout** time, not multilpied Forget I said this In your code,...

It already does that. Also, **io_uring_cq_ready** is not a syscall, it is simply two reads and a subtraction :)

Try upgrading your liburing library if that's really the case. But also, when zero cqes are returned, it's not success. It's an error and should be -ETIME

Here are my results OS: Windows 10 21H2 CPU: AMD Ryzen 7 5800X WSL Linux 5.10 epoll 140,000 /sec uring 133,000 /sec WSL Linux 6.0.0 epoll 149,000 /sec uring 170,000...

Another thing is, performance is not much worse by using IORING_OP_POLL_ADD and recv/send on the user side. If I were to guess, I would assume most of the gains are...

In analysis of `perf top`, most (80% or higher) of the CPU time is dedicated to the `send` call on a socket, which explains why the reduced system call numbers...

You can always use `Buffer.from(uint8Array.buffer)` docs at [nodejs.org](https://nodejs.org/api/buffer.html#static-method-bufferfromarraybuffer-byteoffset-length)

See https://github.com/rust-lang/rust/issues/122519 and https://github.com/rust-lang/rust/pull/122435

Hello?

Haha, no worries. I'm not sure why the tests are failing but if you could point me towards the right direction I wouldn't mind fixing them