[clang] [AMDGPU][SPIRV] Correctly lower huge device function arguments (PR #176921)

[Manuel Carrasco via cfe-commits]

mgcarrasco wrote:

LGTM. Just asking to make sure that I got it right, the frontend now emits IR matching the allocation sizes. Before this PR, it wasn't due to the overflow. Some backends might not support the IR but that is a different issue, right?

https://github.com/llvm/llvm-project/pull/176921