[Mlir-commits] [mlir] [vector][mlir] Canonicalize to shape_cast where possible (PR #140583)

Mehdi Amini llvmlistbot at llvm.org
Wed Dec 17 07:12:20 PST 2025


joker-eph wrote:

> You can write down the same example with transpose / shape_cast inverted and come to the opposite conclusion.

Are you saying that there are sequences of transposes that can't be all shape_casts that would fold themselves?


https://github.com/llvm/llvm-project/pull/140583


More information about the Mlir-commits mailing list