Received-SPF: pass (google.com: domain of linux-kernel+bounces-118201-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Precedence: bulk
MIME-Version: 1.0
References: <20240325195418.166013-1-wedsonaf@gmail.com> <20240325195418.166013-8-wedsonaf@gmail.com>
 <ZgHiJ23TdOmnSGe9@boqun-archlinux>
In-Reply-To: <ZgHiJ23TdOmnSGe9@boqun-archlinux>
From: Wedson Almeida Filho <wedsonaf@gmail.com>
Date: Mon, 25 Mar 2024 21:03:58 -0300
Message-ID: <CANeycqpRfgTof1QeX4NnTV5NidEHxYCan7rvEw+T5nRyMDD9oQ@mail.gmail.com>
Subject: Re: [PATCH 07/10] rust: alloc: update `VecExt` to take allocation flags
To: Boqun Feng <boqun.feng@gmail.com>
Cc: rust-for-linux@vger.kernel.org, Miguel Ojeda <ojeda@kernel.org>, 
	Alex Gaynor <alex.gaynor@gmail.com>, Gary Guo <gary@garyguo.net>, 
	=?UTF-8?Q?Bj=C3=B6rn_Roy_Baron?= <bjorn3_gh@protonmail.com>, 
	Benno Lossin <benno.lossin@proton.me>, Andreas Hindborg <a.hindborg@samsung.com>, 
	Alice Ryhl <aliceryhl@google.com>, linux-kernel@vger.kernel.org, 
	Wedson Almeida Filho <walmeida@microsoft.com>
Content-Type: text/plain; charset="UTF-8"

On Mon, 25 Mar 2024 at 17:44, Boqun Feng <boqun.feng@gmail.com> wrote:
>
> On Mon, Mar 25, 2024 at 04:54:15PM -0300, Wedson Almeida Filho wrote:
> [...]
> > +    fn push(&mut self, v: T, flags: Flags) -> Result<(), AllocError> {
> > +        <Self as VecExt<_>>::reserve(self, 1, flags)?;
> > +        let (ptr, len, cap) = destructure(self);
> > +        // SAFETY: ptr is valid for `cap` elements. And `cap` is greater (by at least 1) than
> > +        // `len` because of the call to `reserve` above. So the pointer after offsetting by `len`
> > +        // elements is valid for write.
> > +        unsafe { ptr.wrapping_add(len).write(v) };
> > +
> > +        // SAFETY: The only difference from the values returned by `destructure` is that `length`
> > +        // is incremented by 1, which is fine because we have just initialised the element at
> > +        // offset `length`.
> > +        unsafe { rebuild(self, ptr, len + 1, cap) };
>
> probably use spare_capacity_mut() here to avoid `destructure` and
> `rebuild`?

Ah, yes, this sounds like a better approach. I will use this in v2.

>
>         https://doc.rust-lang.org/std/vec/struct.Vec.html#method.spare_capacity_mut
>
>         // .. after reserve succeed.
>         // there must be room for adding one more.
>         self.spare_capacity_mut()[0].write(v);
>         // or unsafe { self.spare_capacity_mut().as_mut_ptr().cast().write(v); }
>
>         unsafe {
>                 self.set_len(self.len() + 1);
>         }
>
> Thoughts?
>
> Regards,
> Boqun
>
> >          Ok(())
> >      }
> >
> [...]