linux-kernel - Re: [PATCH] rust: fix outdated safety note in `Revocable::revoke

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]

Message-Id: <DB2ONAGQF92H.1ARBZW47KMVV8@kernel.org>
Date: Thu, 03 Jul 2025 21:55:44 +0200
From: "Benno Lossin" <lossin@...nel.org>
To: Onur Özkan <work@...rozkan.dev>,
 <rust-for-linux@...r.kernel.org>, <linux-kernel@...r.kernel.org>
Cc: <ojeda@...nel.org>, <alex.gaynor@...il.com>, <boqun.feng@...il.com>,
 <gary@...yguo.net>, <bjorn3_gh@...tonmail.com>, <a.hindborg@...nel.org>,
 <aliceryhl@...gle.com>, <tmgross@...ch.edu>, <dakr@...nel.org>,
 <acourbot@...dia.com>, <joelagnelf@...dia.com>, <wedsonaf@...il.com>
Subject: Re: [PATCH] rust: fix outdated safety note in
 `Revocable::revoke_internal`

Something went wrong with your TO addresses, merging Alex's and
Boqun's...

On Thu Jul 3, 2025 at 7:26 PM CEST, Onur Özkan wrote:
> The code used to use `compare_exchange` in the initial version
> but it was changed to `swap` after a reviewer suggestion (see [1]),
> and then the safety comment was not updated and became incorrect.
>
> Link: https://lore.kernel.org/all/20241211104742.533392-1-benoit@dugarreau.fr [1]
>
> Signed-off-by: Onur Özkan <work@...rozkan.dev>
> ---
>  rust/kernel/revocable.rs | 6 ++++--
>  1 file changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/rust/kernel/revocable.rs b/rust/kernel/revocable.rs
> index 06a3cdfce344..5c0b7afa76fb 100644
> --- a/rust/kernel/revocable.rs
> +++ b/rust/kernel/revocable.rs
> @@ -163,8 +163,10 @@ unsafe fn revoke_internal<const SYNC: bool>(&self) -> bool {
>                  unsafe { bindings::synchronize_rcu() };
>              }
>  
> -            // SAFETY: We know `self.data` is valid because only one CPU can succeed the
> -            // `compare_exchange` above that takes `is_available` from `true` to `false`.
> +            // SAFETY: We just used an atomic `swap` to check if the data was still marked
> +            // as available. If it returns `true`, that means we are the first (and only)
> +            // thread to see it as available and mark it as unavailable. So no other thread
> +            // can access or drop the data after this. That makes it safe to drop the data here.

I think this is already addressed by this series:

    https://lore.kernel.org/all/20250626165927.66498-1-marcelomoreira1905@gmail.com

---
Cheers,
Benno

>              unsafe { drop_in_place(self.data.get()) };
>          }
>