Update tokio-core uses to tokio-0.1#304
Conversation
|
I'm probably missing some bit of information, but what's the motivation behind these changes? |
|
Just a follow up of #286. |
| let server = listener.incoming().for_each(move |(socket, _addr)| { | ||
| let server = listener.incoming().for_each(move |socket| { | ||
| trace!("incoming connection"); | ||
| SccacheProto.bind_server(&handle, socket, service.clone()); |
There was a problem hiding this comment.
Forgot I need to cleanup this part. tokio-proto is unmaintained so I had to string together a minimal version of it instead but it still uses some of tokio-proto's types which makes it a bit of a mess.
b02a522 to
f10d3c5
Compare
|
Marwes@0faa567 might be a bit more palatable, it is on this branch https://github.com/Marwes/sccache/commits/fmt_plus_tokio which formats the entire project first so that any accidental formatting changes are in a separate commit, and all the behavioural changes are in the second commit. |
519718e to
6d748f6
Compare
|
Seeing a deprecation warning about |
|
Thanks for the PR! I think this will need to wait until we land the last few sccache-dist PRs. #323 should be the second-to-last one. |
|
Hi! I think I've been running into #248 (which might be fixed by this) - any updates on getting this landed? |
|
We're very close to finishing landing all the distributed compilation work--I'll get this landed as soon as that's all done. |
1990a52 to
3b8c700
Compare
|
OK, I apologize for how long this has taken, but if you want to rebase this one last time I will merge it immediately. |
3b8c700 to
89d7b71
Compare
|
Rebased but seem to have broken something in the distributed compilation. Will try and debug it. |
89d7b71 to
cd05ab2
Compare
Also updates request to 0.9 (as that uses hyper 0.12)
cd05ab2 to
3f831ff
Compare
d7811b6 to
1373063
Compare
|
Thanks! Sorry that this took so long to merge that you had to rebase it several times. |
reqwesthas received an update to drop itstokio-coredependency so it is now possible to move to usingtokiohere as well.hyperxwere pulled it to paper overhyper's removal of typed headers but were otherwise avoided.Since there are quite a few non-Send futures I had to use
current_thread::Runtimeinstead of the standard multi-threaded runtime.