================ ---------------- purnima-nlp wrote: @FedericoBruzzone Thanks! I shortened both regression tests by using smaller vector shapes while keeping the same rank-mismatch and unit-dimension cases covered. https://github.com/llvm/llvm-project/pull/227351