linux-kernel - Re: [PATCH v2 03/14] rust: str: introduce `NullBorrowFormatter`

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: <aG9-MAwbNbjuoR0i@google.com>
Date: Thu, 10 Jul 2025 08:47:44 +0000
From: Alice Ryhl <aliceryhl@...gle.com>
To: Andreas Hindborg <a.hindborg@...nel.org>
Cc: Boqun Feng <boqun.feng@...il.com>, Miguel Ojeda <ojeda@...nel.org>, 
	Alex Gaynor <alex.gaynor@...il.com>, Gary Guo <gary@...yguo.net>, 
	"Björn Roy Baron" <bjorn3_gh@...tonmail.com>, Benno Lossin <lossin@...nel.org>, 
	Trevor Gross <tmgross@...ch.edu>, Danilo Krummrich <dakr@...nel.org>, Jens Axboe <axboe@...nel.dk>, 
	linux-block@...r.kernel.org, rust-for-linux@...r.kernel.org, 
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 03/14] rust: str: introduce `NullBorrowFormatter`

On Wed, Jul 09, 2025 at 05:49:57PM +0200, Andreas Hindborg wrote:
> "Alice Ryhl" <aliceryhl@...gle.com> writes:
> 
> > On Tue, Jul 08, 2025 at 09:44:58PM +0200, Andreas Hindborg wrote:
> >> Add `NullBorrowFormatter`, a formatter that writes a null terminated string
> >> to an array or slice buffer. Because this type needs to manage the trailing
> >> null marker, the existing formatters cannot be used to implement this type.
> >>
> >> Signed-off-by: Andreas Hindborg <a.hindborg@...nel.org>
> >> ---
> >>  rust/kernel/str.rs | 59 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
> >>  1 file changed, 59 insertions(+)
> >>
> >> diff --git a/rust/kernel/str.rs b/rust/kernel/str.rs
> >> index 78b2f95eb3171..05d79cf40c201 100644
> >> --- a/rust/kernel/str.rs
> >> +++ b/rust/kernel/str.rs
> >> @@ -860,6 +860,65 @@ fn deref_mut(&mut self) -> &mut Self::Target {
> >>      }
> >>  }
> >>
> >> +/// A mutable reference to a byte buffer where a string can be written into.
> >> +///
> >> +/// The buffer will be automatically null terminated after the last written character.
> >> +///
> >> +/// # Invariants
> >> +///
> >> +/// `buffer` is always null terminated.
> >> +pub(crate) struct NullBorrowFormatter<'a> {
> >> +    buffer: &'a mut [u8],
> >> +    pos: usize,
> >> +}
> >
> > Do you need `pos`? Often I see this kind of code subslice `buffer`
> > instead.
> 
> How would that work? Can I move the start index of `buffer` in some way
> without an unsafe block?

Yes. I think this will work:

let buffer = mem::take(&mut self.buffer);
self.buffer = &mut buffer[pos..];

Temporarily storing an empty slice avoids lifetime issues.

> >> +    #[expect(dead_code)]
> >> +    pub(crate) fn from_array<const N: usize>(
> >> +        a: &'a mut [crate::ffi::c_char; N],
> >> +    ) -> Result<NullBorrowFormatter<'a>> {
> >> +        Self::new(
> >> +            // SAFETY: the buffer of `a` is valid for read and write as `u8` for
> >> +            // at least `N` bytes.
> >> +            unsafe { core::slice::from_raw_parts_mut(a.as_mut_ptr().cast::<u8>(), N) },
> >> +        )
> >> +    }
> >
> > Arrays automatically coerce to slices, so I don't think this is
> > necessary. You can just call `new`.
> 
> Nice!

I'm guessing it used to be necessary back when we used core::ffi::c_char
since &[i8;N] doesn't coerce to &[u8]. But now that we use the right
c_char definition, that isn't the case anymore.

> >> +impl Write for NullBorrowFormatter<'_> {
> >> +    fn write_str(&mut self, s: &str) -> fmt::Result {
> >> +        let bytes = s.as_bytes();
> >> +        let len = bytes.len();
> >> +
> >> +        // We want space for a null terminator
> >> +        if self.pos + len > self.buffer.len() - 1 {
> >
> > Integer overflow?
> 
> In the subtraction? `buffer.len()` is at least 1, because of the type invariant.
> 
> Or do you mean I should do a checked add for `self.pos + len`?

Ah, I guess self.pos and len are both <= the length of a slice, which is
at most isize::MAX, so the addition can't overflow an usize. Would be
good to comment this, though.

Alice