[Mlir-commits] [mlir] [mlir][x86]Convert a `linalg.generic` with BF16/Int8 accumulation to F32/Int32 accumulation (PR #190779)

[Adam Siemieniuk]

adam-smnk wrote:

As discussed offline, we still need to iterate on this and see if it's really necessary.
@arun-thmn Maybe let's close this for now?

https://github.com/llvm/llvm-project/pull/190779