[flang-commits] [flang] [flang][AMDGPU] Convert math ops to AMD GPU library calls instead of libm calls (PR #99517)

[Matt Arsenault via flang-commits]

arsenm wrote:

> @arsenm, if I change the conversion pass to map to the supported llvm intinsics, which you pointed out, would that be acceptable for now?

Yes. Hopefully in the future we will have a sensible intrinsic lowering solution 

https://github.com/llvm/llvm-project/pull/99517