Message-ID: <CANeycqq+uVp1O5k56LPtzvJQObrjT9JY3zBiX8T0nu6Gdsbu6g@mail.gmail.com>
Date: Mon, 25 Mar 2024 21:02:33 -0300
From: Wedson Almeida Filho <wedsonaf@...il.com>
To: Benno Lossin <benno.lossin@...ton.me>
Cc: rust-for-linux@...r.kernel.org, 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>, 
	Andreas Hindborg <a.hindborg@...sung.com>, Alice Ryhl <aliceryhl@...gle.com>, 
	linux-kernel@...r.kernel.org, Wedson Almeida Filho <walmeida@...rosoft.com>
Subject: Re: [PATCH 02/10] rust: alloc: introduce the `VecExt` trait

On Mon, 25 Mar 2024 at 19:05, Benno Lossin <benno.lossin@...ton.me> wrote:
>
> On 25.03.24 20:54, Wedson Almeida Filho wrote:> From: Wedson Almeida Filho <walmeida@...rosoft.com>
> >
> > Make `try_with_capacity`, `try_push`, and `try_extend_from_slice`
> > methods available in `Vec` even though it doesn't implement them. It is
> > implemented with `try_reserve` and `push_within_capacity`.
> >
> > This is in preparation for switching to the upstream `alloc` crate.
> >
> > Suggested-by: Gary Guo <gary@...yguo.net>
> > Signed-off-by: Wedson Almeida Filho <walmeida@...rosoft.com>
> > ---
> >   rust/kernel/alloc.rs        |  1 +
> >   rust/kernel/alloc/vecext.rs | 52 +++++++++++++++++++++++++++++++++++++
> >   rust/kernel/lib.rs          |  1 +
> >   rust/kernel/prelude.rs      |  2 ++
> >   4 files changed, 56 insertions(+)
> >   create mode 100644 rust/kernel/alloc/vecext.rs
> >
> > diff --git a/rust/kernel/alloc.rs b/rust/kernel/alloc.rs
> > index ccd4149932c3..8ad57a2e693e 100644
> > --- a/rust/kernel/alloc.rs
> > +++ b/rust/kernel/alloc.rs
> > @@ -5,3 +5,4 @@
> >   #[cfg(not(test))]
> >   #[cfg(not(testlib))]
> >   mod allocator;
> > +pub mod vecext;
> > diff --git a/rust/kernel/alloc/vecext.rs b/rust/kernel/alloc/vecext.rs
> > new file mode 100644
> > index 000000000000..59e92bab534e
> > --- /dev/null
> > +++ b/rust/kernel/alloc/vecext.rs
> > @@ -0,0 +1,52 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +
> > +//! Extensions to [`Vec`] for fallible allocations.
> > +
> > +use alloc::{collections::TryReserveError, vec::Vec};
> > +use core::result::Result;
> > +
> > +/// Extensions to [`Vec`].
> > +pub trait VecExt<T>: Sized {
> > +    /// Creates a new [`Vec`] instance with at least the given capacity.
> > +    fn try_with_capacity(capacity: usize) -> Result<Self, TryReserveError>;
> > +
> > +    /// Appends an element to the back of the [`Vec`] instance.
> > +    fn try_push(&mut self, v: T) -> Result<(), TryReserveError>;
> > +
> > +    /// Pushes clones of the elements of slice into the [`Vec`] instance.
> > +    fn try_extend_from_slice(&mut self, other: &[T]) -> Result<(), TryReserveError>
> > +    where
> > +        T: Clone;
>
> All of the non `try_` prefix functions require
> `not(no_global_oom_handling)`, so we could also drop the `try_` prefix
> here.
> What do you think?

Yes, I do drop the `try_` prefixes in patch 7.

Patches 1 through 4 remove the fork with minimal changes. Patches 5 to
10 introduce new functions with flags and drop `allocator_api`.

> > +}
> > +
> > +impl<T> VecExt<T> for Vec<T> {
> > +    fn try_with_capacity(capacity: usize) -> Result<Self, TryReserveError> {
> > +        let mut v = Vec::new();
> > +        v.try_reserve(capacity)?;
> > +        Ok(v)
> > +    }
> > +
> > +    fn try_push(&mut self, v: T) -> Result<(), TryReserveError> {
> > +        if let Err(retry) = self.push_within_capacity(v) {
> > +            self.try_reserve(1)?;
> > +            let _ = self.push_within_capacity(retry);
> > +        }
> > +        Ok(())
> > +    }
> > +
> > +    fn try_extend_from_slice(&mut self, other: &[T]) -> Result<(), TryReserveError>
> > +    where
> > +        T: Clone,
> > +    {
> > +        let extra_cap = self.capacity() - self.len();
> > +        if extra_cap > 0 {
> > +            self.try_reserve(extra_cap)?;
> > +        }
>
> I am confused, why are you doing this? Did you mean to do this?:
>     let extra_cap = self.capacity() - self.len() - other.len();

I originally meant `extracap = other.len() - self.capacity() +
self.len()` because originally I thought `additional` meant additional
capacity (so the final capacity would be original capacity plus
`additional`). But after I looked at the code I learned that
`additional` is really the minimum capacity.

Then I decided that we don't need this check at all in
`try_extend_from_slice`, which I removed in patch 7, I just call
`reserve` directly with `other.len()`.

The code happens to work now because the `try_push` calls below ensure
the capacity gets expanded. In any case, I'll fix this in v2.

> --
> Cheers,
> Benno
>
> > +
> > +        for item in other {
> > +            self.try_push(item.clone())?;
> > +        }
> > +
> > +        Ok(())
> > +    }
> > +}
> > diff --git a/rust/kernel/lib.rs b/rust/kernel/lib.rs
> > index 51f30e55bd00..7f2841a18d05 100644
> > --- a/rust/kernel/lib.rs
> > +++ b/rust/kernel/lib.rs
> > @@ -19,6 +19,7 @@
> >   #![feature(offset_of)]
> >   #![feature(receiver_trait)]
> >   #![feature(unsize)]
> > +#![feature(vec_push_within_capacity)]
> >
> >   // Ensure conditional compilation based on the kernel configuration works;
> >   // otherwise we may silently break things like initcall handling.
> > diff --git a/rust/kernel/prelude.rs b/rust/kernel/prelude.rs
> > index ae21600970b3..a0177f195dec 100644
> > --- a/rust/kernel/prelude.rs
> > +++ b/rust/kernel/prelude.rs
> > @@ -14,6 +14,8 @@
> >   #[doc(no_inline)]
> >   pub use core::pin::Pin;
> >
> > +pub use crate::alloc::vecext::VecExt;
> > +
> >   #[doc(no_inline)]
> >   pub use alloc::{boxed::Box, vec::Vec};
> >
> > --
> > 2.34.1
> >

Powered by blists - more mailing lists