[Mlir-commits] [mlir] [mlir][tosa] Check for 0-ranked-tensors during fold (PR #68512)

[Georgios Pinitas]

GeorgeARM wrote:

Looking again at the spec it says that reduce and reverse operators need to take as input tensor of rank 1 or higher. So the most appropriate solution here really is to change the TOSA operator specification to accept `AnyNon0RankedOrUnrankedTensor` as an input here.  Thoughts @eric-k256?


https://github.com/llvm/llvm-project/pull/68512