FrankReh

@Noah-Kennedy Once this passes all CI tests, even if it means the tests were commented out, I'm inclined to approve something like this as it doesn't appear to change what...

I'm going to leave this for @Noah-Kennedy to comment on. It's not the direction I was hoping for but also it's not next in line for me to figure out....

That close seemed to have been done automatically. Anyway, no harm. Definitely some file system operations left.

> Is `tokio::fs::canonicalize` also an operation we can support? The io_uring API doesn't support anything like a canonicalize (or in Unix terms, a realpath) so to avoid using the std...

Nice that you copied from `Tokio` and that you minimized it to the base essentials. I'll admit I don't know the benefits of making the macro a separate feature. Maybe...

> @FrankReh the advantage of the feature is to avoid requiring the proc macro crate to be built unless a user actually needs it. Is defining that one macro really...

> Personally I'm of the opinion that this isn't worth doing unless ... Now every time I see a `#[tokio::test]` in some other project's PRs being introduced, I wish we...

I think this is @Noah-Kennedy 's call now. And I would have preferred a PR that just was used on the no_op test, so that the git commit would show...

We can decide how to model this functionality for this crate and even whether we only support a locked down version where the crate is responsible for all use cases...

This isn't documented in the liburing man pages yet but it's been promised. There is already the function and an interesting test acting as an example in the tip of...