cortex-m-semihosting: Ship a 1.0 release

[thejpster]

Part of rust-embedded/wg#383

Blockers:

• interrupt::free is not enough to access static mut safely cortex-m-semihosting#49 – interrupt::free is not enough to access static mut safely

[jonas-schievink]

rust-embedded/cortex-m-semihosting#41 and rust-embedded/cortex-m-semihosting#48 are both breaking changes. Would be good to get them in before 1.0.

[japaric]

I would say rust-embedded/cortex-m-semihosting#49 is a blocker: can we even get hprintln! to work on multi-core hardware? if not, how should we indicate that's unsound in multi-core context? make it unsafe to call? remove it? Cargo feature it?

make it unsafe to call?

if we go this route then device-specific crates (PAC, HAL and the like) can expose a safe wrapper to it if they know the crate will only ever be used on a single-core device

[jonas-schievink]

Yeah, added rust-embedded/cortex-m-semihosting#49.

if we go this route then device-specific crates (PAC, HAL and the like) can expose a safe wrapper to it if they know the crate will only ever be used on a single-core device

That doesn't sound horrible. I was thinking of doing something like this for bare_metal::Mutex itself too.