[llvm] Avoid maxnum(sNaN, x) optimizations / folds (PR #170181)

[Artem Belevich via llvm-commits]

Artem-B wrote:

> As far as I am aware, optimizations involving constant sNaN should generally be edge-cases that rarely occur, so here should hopefully be very little real-world performance impact from disabling these optimizations.

I'm curious, do we want to keep these optimizations around if the target makes no distinction between signalling and non-signalling NaNs? AFAICT, sNaN makes no difference on NVIDIA GPUs, and never causes any traps or updates any flags.

https://github.com/llvm/llvm-project/pull/170181