farzonl wrote: @cofibrant I removed the use of the flags llvm needs for the lower-matrix-intrinsics pass. I think that should have addressed your concerns so for now this is a flag that does nothing, but my follow on prs usage of the flag will have a clear distinction between 1. layout of matrix types in memory and 2. the layout used by LLVM when lowering matrix intrinsics.
We can revist merging this concept later when the `llvm.matrix.row.major.load` and `llvm.matrix.row.major.store` intrinsics are implemented if that is prefered. In anycase I'll wait for @fhahn. https://github.com/llvm/llvm-project/pull/167628 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
