buffer::set_final_data accepts weak_ptr<T>, not <T[]>

[fknorr]

See [4.7.2.1]:

The finalData points to where the outcome of all the buffer processing is going to be copied to at destruction time, if the buffer was involved with a write accessor.

Destination can be either an output iterator or a std::weak_ptr<T>.

Note that a raw pointer is a special case of output iterator and thus defines the host memory to which the result is to be copied.

Not sure if this was different in an earlier spec revision.

[TApplencourt]

Agree.

For example, for buffer in we have a special overload for T[]. buffer(const std::shared_ptr<T[]>& . We don't have such for final_data as you noticed.

[tomdeakin]

One more reviewer please (total 2 for this please)

[keryell]

I wonder whether this should be a specification fix to accept smart pointers on T[] also.

[keryell]

Why doing this, since you initialize it also with the pointer from the shared pointer handling the ownership? 🤔
OK, after more thoughts you do this to have the same shared pointer but returning a pointer of the array decay.
You could add a comment. 😄