linux-kernel - Re: [PATCH v2 3/4] rust: alloc: refactor `Vec::truncate` using `dec

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: <Z9qUBe_8SM4c-9UI@google.com>
Date: Wed, 19 Mar 2025 09:53:09 +0000
From: Alice Ryhl <aliceryhl@...gle.com>
To: Tamir Duberstein <tamird@...il.com>
Cc: Danilo Krummrich <dakr@...nel.org>, Andrew Ballance <andrewjballance@...il.com>, 
	Miguel Ojeda <ojeda@...nel.org>, Alex Gaynor <alex.gaynor@...il.com>, 
	Boqun Feng <boqun.feng@...il.com>, Gary Guo <gary@...yguo.net>, 
	"Björn Roy Baron" <bjorn3_gh@...tonmail.com>, Benno Lossin <benno.lossin@...ton.me>, 
	Andreas Hindborg <a.hindborg@...nel.org>, Trevor Gross <tmgross@...ch.edu>, 
	rust-for-linux@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 3/4] rust: alloc: refactor `Vec::truncate` using `dec_len`

On Tue, Mar 18, 2025 at 04:13:55PM -0400, Tamir Duberstein wrote:
> Use `checked_sub` to satisfy the safety requirements of `dec_len` and
> replace nearly the whole body of `truncate` with a call to `dec_len`.
> 
> Signed-off-by: Tamir Duberstein <tamird@...il.com>
> ---
>  rust/kernel/alloc/kvec.rs | 29 +++++++++++------------------
>  1 file changed, 11 insertions(+), 18 deletions(-)
> 
> diff --git a/rust/kernel/alloc/kvec.rs b/rust/kernel/alloc/kvec.rs
> index 97cc5ab11e2a..6f4dc89ef7f8 100644
> --- a/rust/kernel/alloc/kvec.rs
> +++ b/rust/kernel/alloc/kvec.rs
> @@ -489,25 +489,18 @@ pub fn reserve(&mut self, additional: usize, flags: Flags) -> Result<(), AllocEr
>      /// # Ok::<(), Error>(())
>      /// ```
>      pub fn truncate(&mut self, len: usize) {
> -        if len >= self.len() {
> -            return;
> +        match self.len().checked_sub(len) {
> +            None => {}
> +            Some(count) => {

This could be simplified as:
if let Some(count) = self.len().checked_sub(len) {
    // logic here
}

or
let Some(count) = self.len().checked_sub(len) else {
    return;
}
// logic here

> +                // SAFETY: `count` is `self.len() - len` so it is guaranteed to be less than or
> +                // equal to `self.len()`.
> +                let tail = unsafe { self.dec_len(count) };
> +
> +                // SAFETY: the contract of `dec_len` guarantees that the elements in `tail` are
> +                // valid elements whose ownership has been transferred to the caller.
> +                unsafe { ptr::drop_in_place(ptr) };

We have a mutable reference to these elements until after the
`drop_in_place` call, but the elements are invalidated by that call.
This means that we have a mutable reference to invalid values, which
violates the invariants for mutable references.

Consider converting to a raw pointer when creating `tail` instead to
avoid that:

let tail: *mut [T] = unsafe { self.dec_len(count) };
unsafe { ptr::drop_in_place(ptr) };

Alice