[llvm] [AMDGPU] When allocating VGPRs, VGPR spills are not part of the prologue (PR #109439)

[Matt Arsenault via llvm-commits]

https://github.com/arsenm approved this pull request.

I still think this prolog concept is a bit broken. This is also really tough to get a test out of, but I'm still trying (I'm hoping #110229 helps reduce it) 

https://github.com/llvm/llvm-project/pull/109439