[llvm] [LangRef] Clarify specification for float min/max operations (PR #172012)

[Ralf Jung via llvm-commits]

================
@@ -20379,9 +20301,17 @@ The '``llvm.vector.reduce.fmax.*``' intrinsics do a floating-point
 ``MAX`` reduction of a vector, returning the result as a scalar. The return type
 matches the element-type of the vector input.
 
-This instruction has the same comparison semantics as the '``llvm.maxnum.*``'
-intrinsic.  If the intrinsic call has the ``nnan`` fast-math flag, then the
-operation can assume that NaNs are not present in the input vector.
+This instruction has the same comparison and ``nsz`` semantics as the
+'``llvm.maxnum.*``' intrinsic.
+
+If any of the vector elements is a signaling NaN, the intrinsic will
+non-deterministically either:
+
+ * Return a :ref:`NaN <floatnan>`.
+ * Treat the signaling NaN as a quiet NaN.
+
+If the intrinsic call has the ``nnan`` fast-math flag, then the operation can
+assume that NaNs are not present in the input vector.
----------------
RalfJung wrote:

If the assumption is violated, then ... UB? poison? undef?

https://github.com/llvm/llvm-project/pull/172012