P3935 R0 Rebasing <cmath> on C23

[wg21bot]

P3935R0 Rebasing <cmath> on C23 (Jan Schultke)

[mattkretz]

P3935R0: Rebasing <cmath> on C23

2026-03-25 SG6 Minutes

Chair: Matthias Kretz

Champion: Jan Schultke

Summary

SG6 reviewed all of the new functions and that they make (more or less) sense
also in the context of C++.

• Discussion whether the <fun>f32 etc. functions are going to be used or
  whether they should just not be added.

• Concern about the pointer arguments and the mismatch with classification
  functions (which are macros in C).

• Consider removing all the noexcepts.

• There should be a (new) paper to discuss the std::simd overloads, though we
  probably don't want to add all of the functions.

Polls

POLL: Don't include the fN functions for the special math functions.

SF	F	N	A	SA
0	5	3	2	0

Outcome: No consensus for change.

POLL: Don't include the fN functions for any of the math functions.

SF	F	N	A	SA
0	0	2	5	2

Outcome: No consensus for change.

POLL: Forward P3935R0 to LEWG.

SF	F	N	A	SA
8	4	0	0	0

Outcome: Strong consensus