[Mlir-commits] [mlir] [MLIR][NVVM] Update support for conversions to f8x2 and f6x2 types (PR #137781)

[Guray Ozen]

================
@@ -1120,6 +1121,153 @@ def NVVM_CvtToF6x2Op : NVVM_Op<"cvt.to.f6x2"> {
   }];
 }
 
+def CVTFP8E4M3 : I32EnumAttrCase<"E4M3", 0, "e4m3">;
+def CVTFP8E5M2 : I32EnumAttrCase<"E5M2", 1, "e5m2">;
+def CVTFP8UE8M0 : I32EnumAttrCase<"UE8M0", 2, "ue8m0">;
----------------
grypp wrote:

Creating type for each op won't really scale. I think should use existing types. 

https://github.com/llvm/llvm-project/pull/137781