[llvm] [GlobalISel][AArch64][AMDGPU] Expand FPOWI into series of multiplication (PR #95217)

[David Green via llvm-commits]

https://github.com/davemgreen commented:

Thanks. I had actually misread isBeneficialToExpandPowI and hadn't realized the heuristic was OptSize only - that the tree of multiplies was efficient enough for any possible constant. That's cool it works like that.

Other than calling isBeneficialToExpandPowI directly this looks OK to me.

https://github.com/llvm/llvm-project/pull/95217