[clang] [llvm] [mlir] [AMDGPU] Add the support for 45-bit buffer resource (PR #159702)

[Krzysztof Drewniak via llvm-commits]

https://github.com/krzysz00 requested changes to this pull request.

I'm not strongly tied to variadic num_records aside from it can lead to less-great codegen where you compute a 64-bit value and have a really late truncation.

More fundamentially, we're not plumbing `flags` through on gfx1250 (for example, graphics needs those for swizzles).

The hold is specifically on flags support

https://github.com/llvm/llvm-project/pull/159702