[llvm] [AMDGPU] Extend llvm.amdgcn.update.dpp intrinsic to support f64 (PR #91190)

[Matt Arsenault via llvm-commits]

arsenm wrote:

As a follow up, can you make AMDGPUInstCombineIntrinsic fold update.dpp.i64 (bitcast (x to i64)) to the different typed intrinsic 

https://github.com/llvm/llvm-project/pull/91190