Implement AsRawFd for EventLoop<'l, Data>

[kchibisov]

This should allow inserting one EventLoop into the other.

[codecov]

Codecov Report

Attention: 6 lines in your changes are missing coverage. Please review.

Comparison is base (342ef70) 86.65% compared to head (86d4026) 86.39%.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #159      +/-   ##
==========================================
- Coverage   86.65%   86.39%   -0.26%     
==========================================
  Files          16       16              
  Lines        1933     1941       +8     
==========================================
+ Hits         1675     1677       +2     
- Misses        258      264       +6     

Flag	Coverage Δ
macos-latest	85.47% <25.00%> (-0.28%)	⬇️
ubuntu-latest	85.95% <25.00%> (-0.27%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files	Coverage Δ
src/sys.rs	92.18% <ø> (ø)
src/loop_logic.rs	85.37% <25.00%> (-1.29%)	⬇️

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[elinorbgr]

I'm thinking that if we're going that route, we could even add an AsFd impl, as well as an EventSource that does the composition (and tests for that).

[kchibisov]

The issue with AsFd is that the inner Poller is inside the Rc<RefCell<..>> chain, so returning something with the lifetime doesn't work because of that.

The only way would be to clone fd, but that's about it.

[notgull]

The issue with AsFd is that the inner Poller is inside the Rc<RefCell<..>> chain, so returning something with the lifetime doesn't work because of that.

The only way would be to clone fd, but that's about it.

The Poller is inside of an Arc, so it can be cheaply cloned to the main structure and used from there.

[kchibisov]

The Poller is inside of an Arc, so it can be cheaply cloned to the main structure and used from there

That's fair. Fixed.

[elinorbgr]

The issue with AsFd is that the inner Poller is inside the Rc<RefCell<..>> chain, so returning something with the lifetime doesn't work because of that.

Actually, it looks like this RefCell is an artifact of the past that is no longer needed, all the methods of Poll take a &self, so I think this RefCell can just be removed.

[kchibisov]

Actually, it looks like this RefCell is an artifact of the past that is no longer needed, all the methods of Poll take a &self, so I think this RefCell can just be removed.

The problem is not methods on the Poll but methods taking &mut Poll. Refactoring them is a breaking change, which you can do separately to what I'm doing here.