Re: [PATCH 2/2] rust: alloc: add `Vec::dec_len`

[Danilo Krummrich <dakr@kernel.org>]

On Sun, Mar 16, 2025 at 06:47:42PM -0400, Tamir Duberstein wrote:
> On Sun, Mar 16, 2025 at 6:42 PM Danilo Krummrich <dakr@kernel.org> wrote:
> >
> > On Sun, Mar 16, 2025 at 06:32:01PM -0400, Tamir Duberstein wrote:
> > > Add `Vec::dec_len` that reduces the length of the receiver. This method
> > > is intended to be used from methods that remove elements from `Vec` such
> > > as `truncate`, `pop`, `remove`, and others. This method is intentionally
> > > not `pub`.
> > >
> > > Signed-off-by: Tamir Duberstein <tamird@gmail.com>
> > > ---
> > >  rust/kernel/alloc/kvec.rs | 15 +++++++++++++++
> > >  1 file changed, 15 insertions(+)
> > >
> > > diff --git a/rust/kernel/alloc/kvec.rs b/rust/kernel/alloc/kvec.rs
> > > index d43a1d609434..5d604e04b9a5 100644
> > > --- a/rust/kernel/alloc/kvec.rs
> > > +++ b/rust/kernel/alloc/kvec.rs
> > > @@ -195,6 +195,21 @@ pub unsafe fn inc_len(&mut self, additional: usize) {
> > >          self.len += additional;
> > >      }
> > >
> > > +    /// Decreases `self.len` by `count`.
> > > +    ///
> > > +    /// Returns a mutable reference to the removed elements.
> > > +    ///
> > > +    /// # Safety
> > > +    ///
> > > +    /// - `count` must be less than or equal to `self.len`.
> >
> > Why? We can catch this, no?
> >
> > We can keep the debug_assert!(), but use self.len.saturating_sub(count) instead.
> 
> This is why I didn't want to write this until we had an actual caller :)

That just defers this question, the methods you mention in your commit message
will be added, hence I think it's better to do it right away.

> We can, but it's not clear why that's better. What does it mean if the
> caller asked to decrement by more than self.len?

It tells us that the caller is buggy, but that's what the debug_assert!() is
for.

But to me both is fine, it's also good when the caller has to justify.

Forgot to mention, for dec_len(), please add the corresponding invariant comment
when adjusting self.len.