[Mlir-commits] [mlir] [mlir][xegpu] Lower dynamic high-D nd load/store via base-pointer fold (PR #215711)

Charitha Saumya llvmlistbot at llvm.org
Wed Aug 12 11:47:46 PDT 2026


https://github.com/charithaintc commented:

LGTM. some initial comments.

what is the motivation for keeping strides and sized in create_nd now? because with this I don't see a use for them. 


https://github.com/llvm/llvm-project/pull/215711


More information about the Mlir-commits mailing list