Upgrade to hashbrown 0.15

[djc]

RUSTSEC-2024-0402 briefly stated that it affected hashbrown 0.14, but that was later corrected. Probably still good to upgrade!

[djc]

Current CI failure:

     Running tests/linked_hash_set.rs (target/x86_64-unknown-linux-gnu/debug/deps/linked_hash_set-4316cf24f2740083)
MemorySanitizer: CHECK failed: msan_linux.cpp:193 "((personality(old_personality | ADDR_NO_RANDOMIZE))) != ((-1))" (0xffffffffffffffff, 0xffffffffffffffff) (tid=4338)
    <empty stack>

error: test failed, to rerun pass `--test linked_hash_set`

Caused by:
  process didn't exit successfully: `/home/circleci/project/target/x86_64-unknown-linux-gnu/debug/deps/linked_hash_set-4316cf24f2740083` (exit status: 1)

Not sure if that's caused by my clippy fixes here...

[kyren]

Current CI failure:

     Running tests/linked_hash_set.rs (target/x86_64-unknown-linux-gnu/debug/deps/linked_hash_set-4316cf24f2740083)
MemorySanitizer: CHECK failed: msan_linux.cpp:193 "((personality(old_personality | ADDR_NO_RANDOMIZE))) != ((-1))" (0xffffffffffffffff, 0xffffffffffffffff) (tid=4338)
    <empty stack>

error: test failed, to rerun pass `--test linked_hash_set`

Caused by:
  process didn't exit successfully: `/home/circleci/project/target/x86_64-unknown-linux-gnu/debug/deps/linked_hash_set-4316cf24f2740083` (exit status: 1)

Not sure if that's caused by my clippy fixes here...

I don't think it's you, I just tried the build of the master branch and it has started failing in exactly the same way.

Interestingly, gc-arena has also started failing with msan in the same way recently (also some weird linking issue that I never fully tracked down). The only thing in common with the two projects I can think of would be hashbrown itself? I'm not going to pretend to be knowledgeable enough to track that down right now, but I am a little concerned (but also it wouldn't surprise me if it was a false positive either, I just can't be sure).

Do you happen to know what kind of things (if there even are any) that msan would catch that wouldn't be caught by miri?

[djc]

Nope. @Amanieu would you be able to take a look?

[kyren]

The only thing in common with the two projects I can think of would be hashbrown itself?

Also this is meaningless because it's not even used in tests right now, it's just an optional dependency to implement a trait on it. gc-arena does use it via std in tests right now ofc but... so does basically everything haha. Anyway I highly doubt it's anything to do with hashbrown, I only mentioned that to say if it's not a false positive it's two entirely separate projects that potentially made the same error uncovered at the exact same time?

[djc]

I don't think it's you, I just tried the build of the master branch and it has started failing in exactly the same way.

Interestingly, gc-arena has also started failing with msan in the same way recently (also some weird linking issue that I never fully tracked down). The only thing in common with the two projects I can think of would be hashbrown itself? I'm not going to pretend to be knowledgeable enough to track that down right now, but I am a little concerned (but also it wouldn't surprise me if it was a false positive either, I just can't be sure).

Might be interesting to look at what else changed. Maybe Rust 1.82 vs 1.83?

[kyren]

Thank you for all of the cleanup, everything looks great to me. I'll merge this now, but if I can I want to figure out the CI failure before doing another release? I guess... that's not actually a good reason to wait since whatever the cause is is also in the previous release (if it is even real), I just want to see if anyone has any insight really. If I have time I'll try and look into it but I'm out of my normal depth so it would probably take me a long time.

[djc]

Thank you for all of the cleanup, everything looks great to me. I'll merge this now, but if I can I want to figure out the CI failure before doing another release? I guess... that's not actually a good reason to wait since whatever the cause is is also in the previous release (if it is even real), I just want to see if anyone has any insight really. If I have time I'll try and look into it but I'm out of my normal depth so it would probably take me a long time.

Yeah, I think it would still make sense to release since it's not a regression from the previous release. But, ultimately up to you.

[kyren]

Yeah, I think it would still make sense to release since it's not a regression from the previous release. But, ultimately up to you.

I won't let it wait very long, it's not a regression so if nobody has time to look into it within the next few days I'll just do a new release.

[kyren]

Yeah, I think it would still make sense to release since it's not a regression from the previous release. But, ultimately up to you.

I won't let it wait very long, it's not a regression so if nobody has time to look into it within the next few days I'll just do a new release.

released!