[clang] [llvm] [CIR][X86] Implement lowering for sqrt builtins (PR #169310)

[Bruno Cardoso Lopes via llvm-commits]

bcardosolopes wrote:

> @bcardosolopes, would also appreciate your input on CIR's strategy here.

I'd wait until we get the implementation in OG, and follow along after. If there are orthogonal pieces that can be prefetched you can work on those right away though

https://github.com/llvm/llvm-project/pull/169310