Skip to content

revise Hermit's mutex interface to support the behaviour of StaticMutex #77610

New issue

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

#

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

Already on GitHub? # to your account

Jump to bottom
Merged
merged 13 commits into from
Oct 25, 2020
Merged

revise Hermit's mutex interface to support the behaviour of StaticMutex #77610

merged 13 commits into from
Oct 25, 2020

Conversation

stlankes
Copy link
Contributor

@stlankes stlankes commented Oct 6, 2020 •
edited
Loading

#77147 simplifies things by splitting this Mutex type into two types matching the two use cases: StaticMutex and MovableMutex. To support the new behavior of StaticMutex, we move part of the mutex implementation into libstd.

The interface to the OS changed. Consequently, I removed a few functions, which aren't longer needed.

@rust-highfive
Copy link
Contributor

r? @LukasKalbertodt

(rust_highfive has picked a reviewer for you, use r? to override)

@rust-highfive rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Oct 6, 2020
@stlankes stlankes mentioned this pull request Oct 6, 2020
Closed
@m-ou-se
Copy link
Member

m-ou-se commented Oct 7, 2020 •
edited
Loading

new behavior of StaticMutex

The behaviour shouldn't have changed. I only gave it a name. Using sys::(hermit::)Mutex like this (without .init() and .destroy()), was already done by all of std's static Mutexes before #77147.

Did #77147 specifically break something for you? Or was this already a problem before?

@stlankes
Copy link
Contributor Author

stlankes commented Oct 7, 2020 •
edited
Loading

I recognize the problem just after your commit. Maybe it was already there before. In my old approach, StaticMutex' are sometimes not correctly initialized.

@stlankes
Copy link
Contributor Author

stlankes commented Oct 9, 2020

My fix to pass the format check was wrong. I fixed it and it works on my systems.

@m-ou-se
Copy link
Member

m-ou-se commented Oct 12, 2020

Does std compile for hermit now? Because it looks like type MovableMutex = Box<Mutex>; (or type MovableMutex = Mutex;) is missing for this platform in sys/hermit/mutex.rs.

@stlankes
Copy link
Contributor Author

My local version seems to work. After an update, std was broken. I fixed it with my latest commits

@stlankes
Copy link
Contributor Author

The latest commit isn't related to the Mutex interface. The alloc crate provides now a new symbol __rg_oom. Our kernel and also Hermit's "user space" based on this crate. To avoid duplicated symbols, I moved a version to our libos.

@stlankes
minor changes to pass the format check
77d9831
@stlankes
box mutex to get a movable mutex
bf268fe 
the commit avoid an alignement issue in Mutex implementation
@stlankes
Copy link
Contributor Author

stlankes commented Oct 14, 2020 •
edited
Loading

@m-ou-se When should I use type MovableMutex = Box<Mutex>;? If I don't used a boxed type, the kernel triggers some exceptions because a few data types aren't aligned.

By the way, the current version works and I created a docker container, which provide the nightly compiler for HermitCore.

@m-ou-se
Copy link
Member

m-ou-se commented Oct 14, 2020

@m-ou-se When should I use type MovableMutex = Box<Mutex>;?

A Box should only be used there if a Mutex object may not be moved (while not borrowed/used), e.g. because it might have internal pointers. Posix pthread_mutex_t needs it. Most Mutex implementations don't need it and should use type MovableMutex = Mutex;.

If I don't used a boxed type, the kernel triggers some exceptions because a few data types aren't aligned.

That sounds like a type (Mutex, SpinLock, or PriorityQueue?) is missing a #[repr(align(..))].

@stlankes
Copy link
Contributor Author

That sounds like a type (Mutex, SpinLock, or PriorityQueue?) is missing a #[repr(align(..))].

Good point, I will check it.

@stlankes
Copy link
Contributor Author

stlankes commented Oct 15, 2020 •
edited
Loading

That sounds like a type (Mutex, SpinLock, or PriorityQueue?) is missing a #[repr(align(..))].

Good point, I will check it.

No, that isn't the problem. I would say the current version is fine. @LukasKalbertodt If you like, you are able to review the PR.

@stlankes
Copy link
Contributor Author

@dtolnay Do you have time to review the code?

@m-ou-se
Copy link
Member

m-ou-se commented Oct 23, 2020

@bors r+

I did not review this in detail, as Hermit is a Tier 3 platform.

@bors
Copy link
Collaborator

bors commented Oct 23, 2020

📌 Commit bf268fe has been approved by m-ou-se

@bors bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Oct 23, 2020
@jonas-schievink jonas-schievink mentioned this pull request Oct 23, 2020
Closed
This was referenced Oct 23, 2020
Closed
Closed
@m-ou-se m-ou-se mentioned this pull request Oct 23, 2020
Closed
@jonas-schievink jonas-schievink mentioned this pull request Oct 24, 2020
Closed
@jonas-schievink jonas-schievink mentioned this pull request Oct 24, 2020
Merged
bors added a commit to rust-lang-ci/rust that referenced this pull request Oct 24, 2020
@bors
Auto merge of rust-lang#78334 - jonas-schievink:rollup-z0gzbmm, r=jon…
f58ffc9 
…as-schievink

Rollup of 12 pull requests

Successful merges:

 - rust-lang#75115 (`#[deny(unsafe_op_in_unsafe_fn)]` in sys/cloudabi)
 - rust-lang#76614 (change the order of type arguments on ControlFlow)
 - rust-lang#77610 (revise Hermit's mutex interface to support the behaviour of StaticMutex)
 - rust-lang#77830 (Simplify query proc-macros)
 - rust-lang#77930 (Do not ICE with TraitPredicates containing [type error])
 - rust-lang#78069 (Fix const core::panic!(non_literal_str).)
 - rust-lang#78072 (Cleanup constant matching in exhaustiveness checking)
 - rust-lang#78119 (Throw core::panic!("message") as &str instead of String.)
 - rust-lang#78191 (Introduce a temporary for discriminant value in MatchBranchSimplification)
 - rust-lang#78272 (const_evaluatable_checked: deal with unused nodes + div)
 - rust-lang#78318 (TyCtxt: generate single impl block with `slice_interners` macro)
 - rust-lang#78327 (resolve: Relax macro resolution consistency check to account for any errors)

Failed merges:

r? `@ghost`
@bors bors merged commit e34263d into rust-lang:master Oct 25, 2020
@rustbot rustbot added this to the 1.49.0 milestone Oct 25, 2020
This was referenced Oct 27, 2020
Merged
Closed
@mkroening mkroening deleted the dtors branch December 11, 2022 19:09
# for free to join this conversation on GitHub. Already have an account? # to comment
Reviewers
No reviews
Assignees

@LukasKalbertodt LukasKalbertodt

Labels
S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion.
Projects
None yet
Milestone
1.49.0
Development

Successfully merging this pull request may close these issues.
6 participants
@stlankes @rust-highfive @m-ou-se @bors @LukasKalbertodt @rustbot