[dcl.array] The first element has the same address as the array

[languagelawyer]

This provides normative wording for the Note from [basic.compound]/4:

[ Note: An array object and its first element are not pointer-interconvertible, even though they have the same address.— end note ]

[jensmaurer]

Editorial teleconference: Send to CWG with given proposed resolution.

[languagelawyer]

Sent to CWG: http://lists.isocpp.org/core/2020/01/8174.php

[frederick-vs-ja]

Do we still need this given #4061 is merged? I think it's probably deducible from the current wording that the first element has the same address as the array.

[jensmaurer]

Let's close this, then.

[languagelawyer]

How did #4061 affect this? The concern about first element address is/was not due to padding

[frederick-vs-ja]

Hmm, as it's clarified that there can't be paddings, I think it's already clear that the element at the lowest address has the same address as the array. Is there still possibility that element can be so arranged that the first element doesn't has the same address as the array?

[languagelawyer]

it's already clear that the element at the lowest address has the same address as the array

If by "the lowest address" you mean the address represented by a pointer to the array object, then this is trivially true.

However, is this "element at the lowest address" the same object as "the first element of the array" (from [conv.array], which, I suppose, means array element 0 for the purposes of http://eel.is/c++draft/expr.add#4.2)?

[jensmaurer]

[expr.rel] p4 might help.