Re: [PATCH] rust: init: remove impl Zeroable for Infallible
From: Laine Taffin Altman
Date: Tue Mar 19 2024 - 01:29:04 EST
On Mar 18, 2024, at 9:39 PM, Boqun Feng <boqun.feng@xxxxxxxxx> wrote:
> On Mon, Mar 18, 2024 at 08:17:07PM -0700, Laine Taffin Altman wrote:
>> On Mar 18, 2024, at 10:25 AM, Boqun Feng <boqun.feng@xxxxxxxxx> wrote:
>>> On Wed, Mar 13, 2024 at 11:09:37PM +0000, Benno Lossin wrote:
>>>> From: Laine Taffin Altman <alexanderaltman@xxxxxx>
>>>>
>>>> It is not enough for a type to be a ZST to guarantee that zeroed memory
>>>> is a valid value for it; it must also be inhabited. Creating a value of
>>>> an uninhabited type, ZST or no, is immediate UB.
>>>> Thus remove the implementation of `Zeroable` for `Infallible`, since
>>>> that type is not inhabited.
>>>>
>>>> Cc: stable@xxxxxxxxxxxxxxx
>>>> Fixes: 38cde0bd7b67 ("rust: init: add `Zeroable` trait and `init::zeroed` function")
>>>> Closes: https://github.com/Rust-for-Linux/pinned-init/pull/13
>>>> Signed-off-by: Laine Taffin Altman <alexanderaltman@xxxxxx>
>>>> Signed-off-by: Benno Lossin <benno.lossin@xxxxxxxxx>
>>>
>>> I think either in the commit log or in the code comment, there better be
>>> a link or explanation on "(un)inhabited type". The rest looks good to
>>> me.
>>
>> Would the following be okay for that purpose?
>>
>> A type is inhabited if at least one valid value of that type exists; a
>> type is uninhabited if no valid values of that type exist. The terms
>> "inhabited" and "uninhabited" in this sense originate in type theory,
>> a branch of mathematics.
>>
>> In Rust, producing an invalid value of any type is immediate undefined
>> behavior (UB); this includes via zeroing memory. Therefore, since an
>> uninhabited type has no valid values, producing any values at all for
>> it is UB.
>>
>> The Rust standard library type `core::convert::Infallible` is
>> uninhabited, by virtue of having been declared as an enum with no
>> cases, which always produces uninhabited types in Rust. Thus, remove
>> the implementation of `Zeroable` for `Infallible`, thereby avoiding
>> the UB.
>>
>
> Yeah, this works for me. Thanks!
Great! Should it be re-sent or can the new wording be incorporated upon merge?
Thank,
Laine
>
> Regards,
> Boqun
>
>> Thanks,
>> Laine
>>
>>>
>>> Reviewed-by: Boqun Feng <boqun.feng@xxxxxxxxx>
>>>
>>> Regards,
>>> Boqun
>>>
>>>> ---
>>>> rust/kernel/init.rs | 4 ++--
>>>> 1 file changed, 2 insertions(+), 2 deletions(-)
>>>>
>>>> diff --git a/rust/kernel/init.rs b/rust/kernel/init.rs
>>>> index 424257284d16..538e03cfc84a 100644
>>>> --- a/rust/kernel/init.rs
>>>> +++ b/rust/kernel/init.rs
>>>> @@ -1292,8 +1292,8 @@ macro_rules! impl_zeroable {
>>>> i8, i16, i32, i64, i128, isize,
>>>> f32, f64,
>>>>
>>>> - // SAFETY: These are ZSTs, there is nothing to zero.
>>>> - {<T: ?Sized>} PhantomData<T>, core::marker::PhantomPinned, Infallible, (),
>>>> + // SAFETY: These are inhabited ZSTs, there is nothing to zero and a valid value exists.
>>>> + {<T: ?Sized>} PhantomData<T>, core::marker::PhantomPinned, (),
>>>>
>>>> // SAFETY: Type is allowed to take any value, including all zeros.
>>>> {<T>} MaybeUninit<T>,
>>>>
>>>> base-commit: 768409cff6cc89fe1194da880537a09857b6e4db
>>>> --
>>>> 2.42.0
>>>>
>>>>
>>>>
>>