News for hackers. Stuff that matters.

A comment on Linux kernel in-tree Rust support. Parent comment

Re:Linux kernel in-tree Rust support by steveklabnik on Saturday July 11, 2020 @12:55PM (#23803557)
It will force your code to be memory safe, but it can't enforce that you have the correct semantics.

https://doc.rust-lang.org/std/sync/atomic/fn.fence.html exists, but Rust can't force you to say, use the correct ordering.

[ Parent | on HN ]
Re:Linux kernel in-tree Rust support by fullstop on Saturday July 11, 2020 @01:10PM (#23803670)
Right, so no different than C in this regard.
[ Parent | on HN ]
Re:Linux kernel in-tree Rust support by comex on Sunday July 12, 2020 @05:29AM (#23810115)
Really, you should rarely be using raw memory barriers even in C, IMO. The C11/C++11 atomics model (which Rust also uses) is substantially nicer. Though, unfortunately, still extremely tricky to get right.
[ Parent | on HN ]
Re:Linux kernel in-tree Rust support by fullstop on Monday July 13, 2020 @10:23PM (#23827861)
The only time that I ever had to worry about that was when writing drivers for the Linux kernel.
[ Parent | on HN ]