[Mlir-commits] [mlir] [mlir][mem2reg] Support whole-buffer vector.load/store promotion (PR #224182)

[Federico Bruzzone]

================

----------------
FedericoBruzzone wrote:

 The description claims scalable slots are supported, but only the transfer ops have a scalable test. Shouldn't we add `vector.load`/`vector.store` cases (or drop the claim)?

https://github.com/llvm/llvm-project/pull/224182