sync::mutex::StaticMutex disappeared from public API, perhaps unintentionally?

[emk]

This seems to have been removed here, as part of the sync cleanup for #19274: 689ef2d

The commit message says:

Because this removes the libsync crate, it is a:

[breaking-change]

However, all uses of libsync should be able to reroute through
std::sync and std::comm instead.

My use-case for StaticMutex was protecting global state in C libraries in emk/rust-cld2 using code like:

extern crate sync;
use sync::mutex::{StaticMutex, MUTEX_INIT};

static LIBRARY_LOCK: StaticMutex = MUTEX_INIT;

fn access_global_resource() {
    let _ = LIBRARY_LOCK.lock();
    unsafe { call_thread_unsafe_c_api(); }
}

Is there a clean way to do this with guards, etc.? I suppose I could use Once to bootstrap my way up to a global mutex, and use that to guard call_thread_unsafe_c_api, but that seems a lot more complicated than using StaticMutex. Am I simply missing the obvious solution here?

Thank you very much for working to rationalize sync!

cc @bors @alexcrichton

[sfackler]

It lives in librustrt right now, though I believe that crate will be subsumed into libstd as well soon: http://doc.rust-lang.org/rustrt/mutex/

[emk]

Thank you! I've switched from StaticMutex to StaticNativeMutex, because
StaticMutex seems to have gone missing.

It's nice to have a static mutex type of some sort for protecting external
C APIs.
Le Fri Nov 28 2014 at 11:53:05, Steven Fackler notifications@github.com a
écrit :

It lives in librustrt right now, though I believe that crate will be
subsumed into libstd as well soon: http://doc.rust-lang.org/rustrt/mutex/

—
Reply to this email directly or view it on GitHub
#19379 (comment).

[kmcallister]

StaticMutex is back in std::sync. I believe there's no StaticNativeMutex anymore, because that's the only kind.