linux-kernel - Re: [PATCH 1/3] rust: maple

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: <aIefe_MEPd_yicde@tardis-2.local>
Date: Mon, 28 Jul 2025 09:04:11 -0700
From: Boqun Feng <boqun.feng@...il.com>
To: Alice Ryhl <aliceryhl@...gle.com>
Cc: Andrew Morton <akpm@...ux-foundation.org>,
	"Liam R. Howlett" <Liam.Howlett@...cle.com>,
	Lorenzo Stoakes <lorenzo.stoakes@...cle.com>,
	Miguel Ojeda <ojeda@...nel.org>,
	Andrew Ballance <andrewjballance@...il.com>,
	Gary Guo <gary@...yguo.net>,
	Björn Roy Baron <bjorn3_gh@...tonmail.com>,
	Benno Lossin <lossin@...nel.org>,
	Andreas Hindborg <a.hindborg@...nel.org>,
	Trevor Gross <tmgross@...ch.edu>,
	Danilo Krummrich <dakr@...nel.org>, linux-kernel@...r.kernel.org,
	maple-tree@...ts.infradead.org, rust-for-linux@...r.kernel.org,
	linux-mm@...ck.org
Subject: Re: [PATCH 1/3] rust: maple_tree: add MapleTree

On Sat, Jul 26, 2025 at 01:23:22PM +0000, Alice Ryhl wrote:
> The maple tree will be used in the Tyr driver to allocate and keep track
> of GPU allocations created internally (i.e. not by userspace). It will
> likely also be used in the Nova driver eventually.
> 
> This adds the simplest methods for additional and removal that do not
> require any special care with respect to concurrency.
> 
> This implementation is based on the RFC by Andrew but with significant
> changes to simplify the implementation.
> 
> Co-developed-by: Andrew Ballance <andrewjballance@...il.com>
> Signed-off-by: Andrew Ballance <andrewjballance@...il.com>
> Signed-off-by: Alice Ryhl <aliceryhl@...gle.com>
> ---
[...]
> +    /// Free all `T` instances in this tree.
> +    ///
> +    /// # Safety
> +    ///
> +    /// This frees Rust data referenced by the maple tree without removing it from the maple tree.
> +    /// The caller must ensure that no reference that remains in the maple tree is used incorrectly
> +    /// after this call.
> +    unsafe fn free_all_entries(self: Pin<&mut Self>) {
> +        // SAFETY: The pointer references a valid maple tree.
> +        let ma_state = unsafe { Opaque::new(bindings::MA_STATE(self.tree.get(), 0, usize::MAX)) };
> +

A meta comment here for the future direction: I think it really makes a
lot of sense if we could have the Rust abstraction for struct ma_state,
that'll allow us to have flexible locking strategy and Iterator-like
interface. Maybe it's something Andrew can take a deeper look when
MapleTree binding is in-tree (no word play intented ;-))?

For example, with a ma_state binding, we can do:

    let mas = MAState::new(self, 0..);

    while let Some(v) = mas.next() {
    	drop(v)
    }

Regards,
Boqun

> +        loop {
> +            // SAFETY: The maple tree is valid. This call to `free_all_entries` has exclusive
> +            // access to the maple tree, so no further synchronization is required.
> +            let ptr = unsafe { bindings::mas_find(ma_state.get(), usize::MAX) };
> +            if ptr.is_null() {
> +                break;
> +            }
> +            // SAFETY: By the type invariants, this pointer references a valid value of type `T`.
> +            // By the safety requirements, it is okay to free it without removing it from the maple
> +            // tree.
> +            unsafe { drop(T::from_foreign(ptr)) };
> +        }
> +    }
> +}
> +
[...]