[Mlir-commits] [mlir] [mlir][ArithToAMDGPU] Add option for saturating truncation to fp8 (PR #74153)

[Jakub Kuderski]

https://github.com/kuhar commented:

Left some suggestions, but overall the clamping logic looks good to me. Some of the nits for naming come from me being used to different conventions, so feel free to disregard in this PR -- just wanted to surface things that surprised me.

https://github.com/llvm/llvm-project/pull/74153