[Mlir-commits] [mlir] Support float8_e3m4 and float8_e4m3 in np_to_memref (PR #186453)

[llvmlistbot at llvm.org]

================
@@ -42,6 +42,15 @@ class F8E5M2(ctypes.Structure):
 
     _fields_ = [("f8E5M2", ctypes.c_int8)]
 
+class F8E3M4(ctypes.Structure):
+    """A ctype representation for MLIR's Float8E3M4."""
+
+    _fields_ = [("f8E3M4", ctypes.c_int8)]
----------------
srcarroll wrote:

to be honest, i'm not 100% this is the right represenation.

https://github.com/llvm/llvm-project/pull/186453