CWG3029 [basic.align] Note about weakest alignment for narrow char types is confusing

[limdor]

Full name of submitter: Xavier Bonaventura

Reference: basic.aling/6

Link to reflector thread: https://chat.isocpp.org/general/pl/yjx3qdhrajfoxbifprpf8p16we (no reflector thread, just mattermost discussion)

Issue description:
https://eel.is/c++draft/basic.memobj#intro.object-3 says

If a complete object is created ([expr.new]) in storage associated with another object e of type “array of N unsigned char” or of type “array of N std​::​byte” ([cstddef.syn]), that array provides storage for the created object if

However, in https://eel.is/c++draft/basic.memobj#basic.align-6 the note says

[Note 2: This enables the ordinary character types to be used as the underlying type for an aligned memory area ([dcl.align]). — end note]

Implying that any narrow character type could be use to provide storage when only "unsigned char" or "std::byte" is allowed.

Suggested resolution:

Remove Note 2 from https://eel.is/c++draft/basic.memobj#basic.align-6

[frederick-vs-ja]

However, in https://eel.is/c++draft/basic.memobj#basic.align-6 the note says

[Note 2: This enables the ordinary character types to be used as the underlying type for an aligned memory area ([dcl.align]). — end note]

Implying that any narrow character type could be use to provide storage when only "unsigned char" or "std::byte" is allowed.

No, "ordinary character types" are not "narrow character types" (since the adoption of char8_t for C++20).

Remove Note 2 from https://eel.is/c++draft/basic.memobj#basic.align-6

This change seems editorial to me, and I believe we we should either just do this or replace "the ordinary character types" with unsigned char.

---

However, per WG14 N3467 (containing changes introduced by WG14 N3197), arrays of ordinary character types (char, signed char, and unsigned char) can effectively provide storage for objects since C2y.

As C++ doesn't think char or signed char arrays can do so, perhaps we should introduce an entry in [diff.basic] once C++ is rebased onto C2y (not now).

[jensmaurer]

We definitely don't want "signed char" to provide storage for anything.

Anyway, the issue seems to talk about "narrow" vs. "ordinary", but "ordinary" is wrong in a different way (because of "signed char" not really helping for the suggested purpose).

[jensmaurer]

CWG3029