[cpp.cond] Missing `__has_cpp_attribute(assume)` CWG2615

[cpplearner]

P1774R8 adds the assumption attribute assume, but does not modify Table 22 "__has_cpp_attribute values" to handle __has_cpp_attribute(assume).

This appears to be an obvious mistake. Can it be fixed editorially?

[tkoeppe]

@jensmaurer Has the feature test macro been discussed?

[jensmaurer]

The change history of the paper says

R5, 2021-12-09: Updated wording (removed feature-test macro, allowed duplicate attributes, added clarifications). Updated and restructured text, expanding semantics section to reflect discussion in EWG and on the WG21 reflectors. Added Peter Dimov’s refcounted smart pointer code example.

I don't remember any discussion about a feature-test macro in CWG, and the CWG minutes don't show any discussion either. I presume everybody assumed __cpp_has_attribute would just magically work.

[tkoeppe]

Thanks -- I don't have a problem adding it, but there has been pushback on other missing macros being "normative changes that should go through the regular issue process". Maybe we can make this a CWG issue then?

[tkoeppe]

(Please relabel/reassign as you see fit!)

[jensmaurer]

CWG2615

[frederick-vs-ja]

Should have been fixed by #5988 (82c4158).