[Mlir-commits] [mlir] [MLIR][ROCDL] Remove patterns for ops supported as intrinsics in the AMDGPU backend (PR #102971)

[Jan Leyonberg]

jsjodin wrote:

> Approved
> 
> (If it's possible, could you get the half-precision functions hooked up in a later PR?)

Sure, that should be pretty easy to do.

https://github.com/llvm/llvm-project/pull/102971