Tracking Issue for Mutex Trait RFC [#377]

[korken89]

This is the tracking issue for the Mutex trait RFC #377, the following questions need to be ironed out before the RFC is implemented:

1. • Decide on crate name (core-mutex was proposed, mutex-trait was selected)
2. • Decide on crate location (freestanding crate under the HAL team was proposed)
3. • Add and implement crate (proposed reference implementation: https://github.com/korken89/core-mutex)
4. • Release crate
5. • Create PRs for the ecosystem
     • Identify where the crate needs to be re-exported
6. • Add a book chapter on it?
7. • (optional) Provide reference implementations for different mutexes
8. • (optional) Add to the 1.0 stabilization pipeline

Lets start with 1 and 2.

CC @rust-embedded/all

[Disasm]

As for the name, I suggest mutex-trait or mutex_trait (like stable_deref_trait)

[korken89]

@Disasm Thanks!

---

I think we should have a time period for proposing names and then we go for a vote.
Maybe 2 weeks now for proposing names, and 1 week to vote after that.
What does the rest of @rust-embedded/all think?

[adamgreig]

Having already had a full RFC for the concept I don't think we really need a vote on the name; I'd be happy for you to go ahead with whatever you think is sensible.

[therealprof]

I also don't think we need a vote. It's bikeshedding really...

[almindor]

I agree with @adamgreig and @therealprof. We shouldn't watershed the name choice, it's not that important.

[nickray]

As an outside view: given that

• in the RFC discussion and chat there has been confusion of this being an implementation, and
• this being called the "Mutex Trait RFC"

calling it mutex-trait sounds very reasonable :)

[geomatsi]

Hi @korken89
Link 'this part of the RTFM book' is broken.

[korken89]

Thanks for the input, let's have a small discussion (yay/nay) on mutex-trait and use that. :)

@geomasi, I'll fix it now.

[korken89]

In today's WG meeting we decided to go with mutex-trait, this will not be set in stone until the first release but lets have it as the current name.

[korken89]

Lets continue, @rust-embedded/hal do you agree that this fits under HAL or would you propose another group?

[therealprof]

@korken89 HAL is maybe not ideal, how do we feel about @rust-embedded/resources ?

[jamesmunns]

IMO HAL team (focused on abstractions and code portability) makes more sense than Resources (focused on docs, PR, and guides).

[therealprof]

I don't quite see a Mutex trait as a Hardware abstraction but more of a implementation convention which seems closer to a resource. But hey, just throwing it out there, I'm not too attached to it. ;)