P3255 R1 Expose whether atomic notifying operations are lock-free

[wg21bot]

P3255R0 Expose whether atomic notifying operations are lock-free (Brian Bi)

[ogiroux]

SG1 in St Louis

The IS has a defect w.r.t. wait and notify lock-freedom specification
Any objection to unanimous consent?
None

We like the direction of P3255R0 as a resolution for this defect
Any objection to unanimous consent?
None

We want wait functions to be marked as signal-safe if they are
and we want to take a similar direction as P3255R0 for this (a
wait_is_signal_safe trait)
SF F N A SA
1 7 3 0 0
Unanimous consent

Forward P3255R0 to LEWG for the IS with the following changes:

• Apply the mechanical changes that add wait_is_signal_safe
  • wait_is_signal_safe member
  • atomic_wait_is_signal_safe free function
  • wait_is_always_signal_safe constexpr
  • No macro (this isn't in C)
    SF F N A SA
    4 6 1 0 0
    Unanimous consent

[wg21bot]

P3255R1 Expose whether atomic notifying operations are lock-free (Brian Bi)