Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Patch release for all crossbeam crates. #474

Merged
merged 7 commits into from Feb 22, 2020

Conversation

jonhoo
Copy link
Contributor

@jonhoo jonhoo commented Feb 17, 2020

This bumps the patch version for all sub-crates and the top-level crate and updates the changelogs with references to PRs where applicable. As far as I can tell, there are no breaking changes. The top-level crate has not change at all, though I bumped rand to 0.7 in all the crates, which I guess technically requires a patch bump if we do another release.

Fixes #473.
Closes #472.
Closes #468.
Closes #409.
Not sure if this also solves #347 by virtue of #458?

This also bumps the `rand` `dev-dependency` to `0.7`.
No breaking changes as far as I can tell.
This also bumps the `rand` `dev-dependency` to `0.7`.
No breaking changes as far as I can tell.
This also bumps the `rand` `dev-dependency` to `0.7`.
No breaking changes as far as I can tell.
This also bumps the `rand` `dev-dependency` to `0.7`.
No breaking changes as far as I can tell.
This also bumps the `rand` `dev-dependency` to `0.7`.
No breaking changes as far as I can tell.
This just bumps the `rand` `dev-dependency` to `0.7`.
None of the underlying crates had breaking changes, and this crate has
no changes, so a patch release is fine.
@jeehoonkang
Copy link
Contributor

It looks good to me. Thanks!

bors r+

I'll publish the crates soon.

@jeehoonkang
Copy link
Contributor

bors r+

@bors
Copy link
Contributor

bors bot commented Feb 17, 2020

Already running a review

bors bot added a commit that referenced this pull request Feb 17, 2020
474: Patch release for all crossbeam crates. r=jeehoonkang a=jonhoo

This bumps the patch version for all sub-crates and the top-level crate and updates the changelogs with references to PRs where applicable. As far as I can tell, there are no breaking changes. The top-level crate has not change at all, though I bumped `rand` to `0.7` in all the crates, which I guess technically requires a patch bump if we do another release.

Fixes #473.
Closes #472.
Closes #468.
Closes #409.
Not sure if this also solves #347 by virtue of #458?

Co-authored-by: Jon Gjengset <jon@thesquareplanet.com>
@jeehoonkang
Copy link
Contributor

CI fails with e.g. https://travis-ci.org/crossbeam-rs/crossbeam/jobs/651554494#L233 @jonhoo any ideas how to fix?

@jonhoo
Copy link
Contributor Author

jonhoo commented Feb 17, 2020

Bah, looks like rand 0.7 requires support for the 2018 edition, which I believe is 1.32.
That said, it is a dev-dependency, and generally MSRV applies only to the library itself, not the library's tests. Would you consider changing the CI to only run cargo check --lib --examples on MSRV?

@jeehoonkang
Copy link
Contributor

Do we have an option to leave rand as is? If rand 0.6 doesn't have a critical bug, I would rather wait for a few more months before bumping it. What do you think?

rand 0.7 requires support for the 2018 edition, which only works on
1.32 (our current MSRV is 1.28).

This also means that the root crate no longer needs a version bump.
@bors
Copy link
Contributor

bors bot commented Feb 17, 2020

Canceled

@jonhoo
Copy link
Contributor Author

jonhoo commented Feb 17, 2020

Reverted the rand bump

@jeehoonkang
Copy link
Contributor

bors r+

bors bot added a commit that referenced this pull request Feb 17, 2020
474: Patch release for all crossbeam crates. r=jeehoonkang a=jonhoo

This bumps the patch version for all sub-crates and the top-level crate and updates the changelogs with references to PRs where applicable. As far as I can tell, there are no breaking changes. The top-level crate has not change at all, though I bumped `rand` to `0.7` in all the crates, which I guess technically requires a patch bump if we do another release.

Fixes #473.
Closes #472.
Closes #468.
Closes #409.
Not sure if this also solves #347 by virtue of #458?

Co-authored-by: Jon Gjengset <jon@thesquareplanet.com>
@bors
Copy link
Contributor

bors bot commented Feb 17, 2020

Build failed

@jeehoonkang
Copy link
Contributor

uh-oh, seems like a real bug... https://travis-ci.org/crossbeam-rs/crossbeam/jobs/651567571#L538

@jonhoo
Copy link
Contributor Author

jonhoo commented Feb 17, 2020

Ouch... As far as I can tell from the history, this basically has to be #458 if the bug wasn't already present in the previous release. There is, I suppose, also a slim chance that it is related to #454 in crossbeam-utils? /cc @manuthambi @cynecx.

@jonhoo
Copy link
Contributor Author

jonhoo commented Feb 17, 2020

It could also be that the test itself is just racy?

@jonhoo
Copy link
Contributor Author

jonhoo commented Feb 17, 2020

It certainly looks like that might be the case, because it passed just fine here: https://travis-ci.org/crossbeam-rs/crossbeam/builds/651566762

jonhoo added a commit to mit-pdos/noria that referenced this pull request Feb 17, 2020
Remove `ctrlc` since `3.1.4` is now out.
Remove `crossbeam` since it's about to get a release:
 crossbeam-rs/crossbeam#474
@jeehoonkang
Copy link
Contributor

Hmm... I feel we should ping @stjepang for this matter. I'm not familiar with crossbeam-channel. Stjepan, do you think we can publish new crates even in the presence of the bug that popped up in the CI?

@cuviper
Copy link
Contributor

cuviper commented Feb 20, 2020

uh-oh, seems like a real bug... https://travis-ci.org/crossbeam-rs/crossbeam/jobs/651567571#L538

I was just looking at Fedora failed builds in that same test, on the current crossbeam-channel 0.4.0. In other words, I think this isn't new, at least.

$ for (( i=1 ;; i+=1 )); do echo "*** test $i ***"; cargo test --test ready fairness || break; done

This loops until failure -- usually I make it 20-30 times successfully. It seems to fail even more frequently with --release, so maybe the test just needs more allowance?

Note: testing "fairness" matches fairness1 as well, but it's always fairness2 that fails.

@jonhoo
Copy link
Contributor Author

jonhoo commented Feb 20, 2020

Yeah, that'd be my guess. My inclination here would be to merge this and then do a release. The test is probably racy, and isn't caused by this release, so it's not worth holding up the release we wait for Stjepan to have a chance to look at the test to see if the tolerance should be increased.

@jonhoo
Copy link
Contributor Author

jonhoo commented Feb 21, 2020

@jeehoonkang What do you think?

@jeehoonkang
Copy link
Contributor

Okay, I'm trying again to merge

bors retry

@bors
Copy link
Contributor

bors bot commented Feb 22, 2020

Build succeeded

@bors bors bot merged commit 6d6591a into crossbeam-rs:master Feb 22, 2020
@jeehoonkang
Copy link
Contributor

@jonhoo I'll release crates now. Would you please create an issue on the CI failure?

@jeehoonkang
Copy link
Contributor

  • crossbeam-deque is not published due to the following error:
error[E0599]: no method named `into_handle` found for type `epoch::Owned<Buffer<T>>` in the current scope
   --> src/lib.rs:230:33
    |
230 |             buffer.into_owned().into_handle().dealloc();
    |                                 ^^^^^^^^^^^ method not found in `epoch::Owned<Buffer<T>>`
  • crossbeam-queue is not published because I don't have a permission to do so. @stjepang would you please give me permission?

@jonhoo jonhoo deleted the version-bump branch February 22, 2020 15:47
@jonhoo
Copy link
Contributor Author

jonhoo commented Feb 22, 2020

@jeehoonkang
Racy test filed as #477.
As for the compile error, cargo publish --dry-run runs just fine for me on crossbeam-deque:

$ cargo publish --dry-run
    Updating crates.io index
   Packaging crossbeam-deque v0.7.3 (/home/jon/dev/others/crossbeam/crossbeam-deque)
   Verifying crossbeam-deque v0.7.3 (/home/jon/dev/others/crossbeam/crossbeam-deque)
   Compiling crossbeam-utils v0.7.2
   Compiling crossbeam-epoch v0.8.2
   Compiling crossbeam-deque v0.7.3 (/home/jon/.cargo-target/package/crossbeam-deque-0.7.3)
    Finished dev [unoptimized + debuginfo] target(s) in 1.35s
   Uploading crossbeam-deque v0.7.3 (/home/jon/dev/others/crossbeam/crossbeam-deque)

I think maybe what happened is that crossbeam-deque relies on the new version of crossbeam-epoch or crossbeam-epoch, and that new version wasn't published yet when you tried to run cargo publish. cargo publish will never use path dependencies, and will instead just use whatever the latest version on crates.io is. This makes me think that we need to bump the minimum version requirement for one of the internal dependencies of crossbeam-deque before uploading it to crates.io. Specifically, we need to set it to the version that introduced into_handle for epoch::Owned<Buffer<T>>.

@jeehoonkang
Copy link
Contributor

@jonhoo sorry, it was my bad. It was resolved in #476. I just published a new version of deque: https://crates.io/crates/crossbeam-deque

@jeehoonkang
Copy link
Contributor

crossbeam-queue is not published because I don't have a permission to do so. @stjepang would you please give me permission?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
3 participants