[PATCH] D139095: [clang] Mark CWG405 as a duplicate of CWG218

Thu Dec 1 09:43:39 PST 2022

Endill added a comment.

> I don't think we should mark it as a dup -- we want the status in our tests to match the status on the official document, otherwise things get confusing.

We can do it the following way then: `// dr405: yes \n // NB: also dup 218`.
Do I understand correctly that superseded status should be used if and only if it's used in official document as well?

> The two issues are very closely related, but they change different words in the standard and should be tested independently as best we can

Make no mistake here: proposed wording for 405 has never made it into the standard. I double-checked this with revision 100 of CWG issues, when this issue still had open status.
P1787 <https://reviews.llvm.org/P1787> states that //CWG405 is resolved by stating that argument-dependent lookup (sometimes) occurs after an ordinary unqualified lookup (making statements like “finding a variable prevents argument-dependent lookup” formally correct).//
The only relevant wording in P1787 <https://reviews.llvm.org/P1787> I can find is for [basic.lookup.argdep] p1 and p3 (too long; not citing them here). Which, curiously enough, clearly originates from resolution of 218, which is already tested. It also has the same intent as proposed resolution for 405.

So I'd like to raise a couple of questions:

1. What test for 405 is going to be if not a copy-and-paste of a part of 218 test?
2. Is it possible to change status of 405 in the official document? Or get a technical rationale for it not being a duplicate of 218.

As a side note, I don't feel too comfortable testing name lookup via side effects like diagnostics. `#pragma clang __debug dump` is good, but not powerful enough to test ADL. Are those the only options we currently have?

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D139095/new/

https://reviews.llvm.org/D139095