[Mlir-commits] [mlir] [mlir][amdgpu] Add conversion from arith.scaling_extf to amdgpu (PR #146372)

[Oleksandr Alex Zinenko]

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

Please find a better way to check the conversion broadcast that 400 lines of check-next, this is not very maintainable :)

https://github.com/llvm/llvm-project/pull/146372