[all-commits] [llvm/llvm-project] 557f7e: [libc++] P2167R3: Improved Proposed Wording for LW...

[A. Jiang via All-commits]

  Branch: refs/heads/main
  Home:   https://github.com/llvm/llvm-project
  Commit: 557f7e1398e13c0957c7a0cbc013c1468f47a237
      https://github.com/llvm/llvm-project/commit/557f7e1398e13c0957c7a0cbc013c1468f47a237
  Author: A. Jiang <de34 at live.cn>
  Date:   2024-10-03 (Thu, 03 Oct 2024)

  Changed paths:
    M libcxx/docs/Status/Cxx23Papers.csv
    M libcxx/include/__compare/compare_partial_order_fallback.h
    M libcxx/include/__compare/compare_strong_order_fallback.h
    M libcxx/include/__compare/compare_weak_order_fallback.h
    M libcxx/test/std/language.support/cmp/cmp.alg/compare_partial_order_fallback.pass.cpp
    M libcxx/test/std/language.support/cmp/cmp.alg/compare_strong_order_fallback.pass.cpp
    M libcxx/test/std/language.support/cmp/cmp.alg/compare_weak_order_fallback.pass.cpp

  Log Message:
  -----------
  [libc++] P2167R3: Improved Proposed Wording for LWG 2114 (#109102)

Only the [cmp.alg] part (for `comparison_meow_fallback` CPOs) in the
paper required changes. Other parts merely fixed preconditions of some
standard library functions.

I strongly feel that P2167R3 should be a DR despite that it is not a DR
officially: CPOs -> C++20; remain parts -> C++98/11 (except that
_`boolean-testable`_ should be transformed into the original
_BooleanTestable_ requirements in the old resolution of LWG2114).

Note that P2167R3 damaged the resolution of LWG3465: the type of `F < E`
was left underconstrained. I've tried to submit an LWG issue for this,
which is now LWG4157.

Drive-by change:
- enable some test coverages in `compare_strong_order_fallback.pass.cpp`
when `TEST_LONG_DOUBLE_IS_DOUBLE`, following up #106742

Closes #105241.



To unsubscribe from these emails, change your notification settings at https://github.com/llvm/llvm-project/settings/notifications