[llvm] LangRef: Clarify nsz on min/max operations for +0.0 vs -0.0 (PR #137567)

[Eli Friedman via llvm-commits]

================
@@ -3914,6 +3914,10 @@ following flags to enable otherwise unsafe floating-point transformations.
    No Signed Zeros - Allow optimizations to treat the sign of a zero
    argument or zero result as insignificant. This does not imply that -0.0
    is poison and/or guaranteed to not exist in the operation.
+   Don't require +0.0>-0.0 for min/max operations - Allow optimizations of the
+   min/max operation not to treat +0.0>-0.0. Note the result should be either of
+   the operands. (max|min)(-0.0, -0.0) should never return +0.0, and vice versa.
+   Note: floating compare operations always imply -0.0 == +0.0.
----------------
efriedma-quic wrote:

> When we expand the fmin(3) and fmax(3) libc calls, we need something like nsz here.

That's not what I was asking; do we have "generic" optimizations that query nsz, besides the ones that are specific to min/max?

> We already do this:

I'm not saying there's anything missing, exactly, I just think it needs to be reorganized.  Like, "For fcmp, this has no effect.  For min/max intrinsics, this allows returning either +0 or -0 if both +0 and -0 are passed.  For calls to anything other than a min/max intrinsic, arithmetic instructions, select, and phi, if the result is zero, the returned value can be a zero of either sign."

https://github.com/llvm/llvm-project/pull/137567