[Mlir-commits] [mlir] [mlir][math] Fix the description of the semantics of math.clampf (PR #175012)

[Fabian Mora]

fabianmcg wrote:

> > Maybe worth mentioning that both implementations are equivalent. I don't think either is wrong. I don't know which one is the way to go tbh.
> 
> Can we check with something like alive2?

I would just go with a published standard. AFAIK there's no IEEE guarantees for them to be equivalent.
Downstreams can introduce a different lowering if it's beneficial or equivalent in their targets.


https://github.com/llvm/llvm-project/pull/175012