Re: [PATCH v2 11/12] rust: init: add `{pin_}chain` functions to `{Pin}Init<T, E>`

[Benno Lossin]

On 21.07.23 02:23, Martin Rodriguez Reboredo wrote:
> On 7/19/23 11:21, Benno Lossin wrote:
>> +/// An initializer returned by [`PinInit::pin_chain`].
>> +pub struct ChainPinInit<I, F, T: ?Sized, E>(I, F, __internal::Invariant<(E, Box<T>)>);
>> +
>> +// SAFETY: the `__pinned_init` function is implemented such that it
>> +// - returns `Ok(())` on successful initialization,
>> +// - returns `Err(err)` on error and in this case `slot` will be dropped.
>> +// - considers `slot` pinned.
>> +unsafe impl<T: ?Sized, E, I, F> PinInit<T, E> for ChainPinInit<I, F, T, E>
>> +where
>> +    I: PinInit<T, E>,
>> +    F: FnOnce(Pin<&mut T>) -> Result<(), E>,
>> +{
>> +    unsafe fn __pinned_init(self, slot: *mut T) -> Result<(), E> {
>> +        // SAFETY: all requirements fulfilled since this function is `__pinned_init`.
>> +        unsafe { self.0.__pinned_init(slot)? };
>> +        // SAFETY: The above call initialized `slot` and we still have unique access.
>> +        let val = unsafe { &mut *slot };
>> +        // SAFETY: `slot` is considered pinned
>> +        let val = unsafe { Pin::new_unchecked(val) };
>> +        (self.1)(val).map_err(|e| {
>> +            // SAFETY: `slot` was initialized above.
>> +            unsafe { core::ptr::drop_in_place(slot) };
>> +            e
> 
> I might stumble upon an error like EAGAIN if I call `pin_chain` but that
> means `slot` will be dropped. So my recommendation is to either not drop
> the value or detail in `pin_chain`'s doc comment that the closure will
> drop on error.

This is a bit confusing to me, because dropping the value on returning `Err`
is a safety requirement of `PinInit`. Could you elaborate why this is
surprising? I can of course add it to the documentation, but I do not see
how it could be implemented differently. Since if you do not drop the value
here, nobody would know that it is still initialized.

-- 
Cheers,
Benno