Damien L-G c1720822b6
[libc++][atomic_ref] Use __atomic_fetch_{add,sub} builtins on floating-points whenever possible (#135685)
Fix #135109

Clang is able to emit an `atomicrmw` instruction from the
`__atomic_fetch_add` and `__atomic_fetch_sub` builtins on floating-point
types.
2025-09-23 10:33:33 -05:00
..