netdev - Re: [PATCH 0/5] Rust abstractions for network device drivers

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Fri, 16 Jun 2023 15:23:01 +0200
From: Miguel Ojeda <miguel.ojeda.sandonis@...il.com>
To: FUJITA Tomonori <fujita.tomonori@...il.com>
Cc: kuba@...nel.org, netdev@...r.kernel.org, rust-for-linux@...r.kernel.org, 
	aliceryhl@...gle.com, andrew@...n.ch
Subject: Re: [PATCH 0/5] Rust abstractions for network device drivers

On Fri, Jun 16, 2023 at 2:18 PM FUJITA Tomonori
<fujita.tomonori@...il.com> wrote:
>
> As far as I know, no subsystem has accepted Rust bindings yet.

For abstractions in general (see my previous reply for "real HW"
etc.), the KUnit subsystem [1] which is onboard and taking some
patches through their tree / ownership of the code.

[1] https://lore.kernel.org/rust-for-linux/CABVgOSnprvxzi-z42KFjOZsiRUv7u7E2poVGJNmTfS2OU4x4AA@mail.gmail.com/

> Replacing the existing C driver for real HW with Rust new one doesn't
> make sense, right? So a necessary condition of getting Rust bindings
> for a subsystem accepted is that a HW verndor implements both a driver
> and bindings for their new HW?

Not necessarily. It is true that, in general, the kernel does not
want/accept duplicate implementations.

However, this is a bit of a special situation, and there may be some
reasons to allow for it in a given subsystem. For instance:

  - The need to experiment with Rust.

  - To have an actual in-tree user that allows to develop the
abstractions for a subsystem, so that later they are ready to be used
for future, actual new drivers.

  - Pending redesigns: sometimes subsystems may have a
redesign/refactor/experiment that they have wanted to do for a while,
so they may take the chance to also try to write it in Rust anyway. Of
course, that could conflate two experiments, but... :)

  - Security: there may be some modules that have been problematic in
the past (especially if due to memory safety issues / data races), and
the subsystem may be willing to accept a parallel implementation to
see if it would be an improvement thanks to Rust's properties.

Cheers,
Miguel