[llvm] [AMDGPU][DAG] Enable ganging up of memcpy loads/stores for AMDGPU (PR #96185)

[Matt Arsenault via llvm-commits]

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

Only other nit is I would have split the memcpy and memmove cases to separate files 

https://github.com/llvm/llvm-project/pull/96185