experiment: try to only use tokio::mpsc in iroh-net #2614
Conversation
|
Pushed a fix. But this isn't really send_blocking, more like send_blocking without runtime and send_detached or whatever in case a runtime is found. It might be that this is OK in most cases. But then the fn should be called send_sync or something and return just unit to signal that it is fire and forget / best effort. |
|
Documentation for this PR has been generated and is available at: https://n0-computer.github.io/iroh/pr/2614/docs/iroh/ Last updated: 2024-08-13T10:00:22Z |
|
As you mention, the calls to send_blocking can be replaced by a normal send call just by making some fns async. So should we just do that? send_blocking works now, except that it is not really blocking in the case where a runtime is found. But we don't need it at all in iroh-net. |
yes, we should try and do that |
|
closing in favor of #2620 |
## Description Removes async-channel from iroh-net and replaces it with the tokio mpsc channel. This is a mergeable version of #2614 ## Breaking Changes LocalSwarmDiscovery is no longer UnwindSafe ## Notes & open questions Open question: I am using blocking_send from inside the windows RouteMonitor. This depends on that the callback thread is not a tokio thread. I have no idea if this is a reasonable assumption to make. Otherwise we would have to bring back the runtime check condition or just capture a runtime handle and spawn a task. Note: what even is this? The only message there is is NetworkMessage::Change, and there is no timestamp or anything. So could I just use try_send? If the queue is full, there is already a Change in it, so it is as good as the new one... ## Change checklist - [x] Self-review. - [ ] Documentation updates following the [style guide](https://rust-lang.github.io/rfcs/1574-more-api-documentation-conventions.html#appendix-a-full-conventions-text), if relevant. - [ ] Tests if relevant. - [ ] All breaking changes documented. --------- Co-authored-by: dignifiedquire <me@dignifiedquire.com>
I am failing to implement the
send_blocking, if anyone has ideas, I would love to hear them