[Mlir-commits] [mlir] [mlir] Add use-vector-alignment flag to ConvertVectorToLLVMPass (PR #137389)

[Krzysztof Drewniak]

================
@@ -1394,6 +1394,12 @@ def ConvertVectorToLLVMPass : Pass<"convert-vector-to-llvm"> {
            "bool", /*default=*/"true",
            "Allows compiler to assume vector indices fit in 32-bit if that "
      "yields faster code">,
+    Option<"useVectorAlignment", "use-vector-alignment",
+           "bool", /*default=*/"false",
+           "Use the preferred alignment of a vector type in load/store "
+           "operations instead of the alignment of the element type of the "
+           "memref. This flag is intended for use with hardware which requires" 
----------------
krzysz00 wrote:

"or in application contexts where it is known all vector accesses are naturally aligned"?

https://github.com/llvm/llvm-project/pull/137389