[llvm] [NVPTX] Don't propagate `ninf` and `nnan` in `lowerFREM` (PR #147125)

[Matt Arsenault via llvm-commits]

https://github.com/arsenm commented:

FYI this lowering is also just wrong and should be an approximation requiring afn. We have essentially the same expansion in AMDGPU to be replaced by https://github.com/llvm/llvm-project/pull/130988

https://github.com/llvm/llvm-project/pull/147125