Kevin Boos
Kevin Boos
The changeset specified in this PR has been moved to the `usb_driver` branch, since this is an old PR based on the old private version of this Theseus repo.
another allocator here: https://github.com/gz/rust-slabmalloc
https://crates.io/crates/buddy_system_allocator
Instead of a lock-free allocator, which is likely impossible, I think a better solution is to use a separate allocator for allocations that occur in interrupt contexts. This is now...
We can actually take this further: by leveraging const generics, we can create MappedPages* instances parameterized with a specific size (in bytes or pages), which is known at compile time....
Note that I'm an idiot and forgot that build scripts can only run _before_ compilation of a crate. So we could do the assembly compilation in a build script, but...
I mentioned some of my thoughts here: https://github.com/theseus-os/Theseus/pull/530#issuecomment-1154547570 I'm not opposed to it, but we would indeed need to do some heavy customization of rustfmt before letting it loose on...
@hecatia-elegua ah, I completely missed your comment above, hence my confusion in #637. Disclaimer: i haven't looked at rustfmt in a while, so my understanding of its features may be...
Formatting imports would be an improvement too, as they're kind of messy and inconsistent. I like the following options: * `imports_granularity`: `Crate` * `group_imports`: `StdExternalCrate` Also it looks like the...
Yes, definitely. This has been on my mind for a while, since many things are x86-specific. The flat hierarchy is no longer a benefit.