[PATCH] D31767: [LLVM][X86][SSE] Update MOVNTDQA non-temporal loads to generic implementation

Craig Topper via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Mon Apr 10 11:14:11 PDT 2017


craig.topper added inline comments.


================
Comment at: lib/IR/AutoUpgrade.cpp:1884
+      Value *Ptr = CI->getArgOperand(0);
+      VectorType *VTy = cast<VectorType>(CI->getType());
+
----------------
RKSimon wrote:
> craig.topper wrote:
> > Do we need a VectorType for the two calls that use this or could it just be generic Type*?
> IIRC we need the VectorType for the getBitWidth() call.
Right, forgot getBitWidth has to multiply the elements with the scalar width.


Repository:
  rL LLVM

https://reviews.llvm.org/D31767





More information about the llvm-commits mailing list