[PATCH] D81619: [NFC] Bail out for scalable vectors before calling getNumElements

Christopher Tetreault via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Wed Jun 10 16:08:27 PDT 2020


ctetreau created this revision.
Herald added subscribers: llvm-commits, hiraditya.
Herald added a project: LLVM.

Move the bail out logic to before constructing the Result and Lane
vectors. This is both potentially faster, and avoids calling
getNumElements on a potentially scalable vector


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D81619

Files:
  llvm/lib/Analysis/ConstantFolding.cpp


Index: llvm/lib/Analysis/ConstantFolding.cpp
===================================================================
--- llvm/lib/Analysis/ConstantFolding.cpp
+++ llvm/lib/Analysis/ConstantFolding.cpp
@@ -2689,15 +2689,16 @@
                                         const DataLayout &DL,
                                         const TargetLibraryInfo *TLI,
                                         const CallBase *Call) {
-  SmallVector<Constant *, 4> Result(VTy->getNumElements());
-  SmallVector<Constant *, 4> Lane(Operands.size());
-  Type *Ty = VTy->getElementType();
-
   // Do not iterate on scalable vector. The number of elements is unknown at
   // compile-time.
   if (isa<ScalableVectorType>(VTy))
     return nullptr;
 
+  SmallVector<Constant *, 4> Result(
+      cast<FixedVectorType>(VTy)->getNumElements());
+  SmallVector<Constant *, 4> Lane(Operands.size());
+  Type *Ty = VTy->getElementType();
+
   if (IntrinsicID == Intrinsic::masked_load) {
     auto *SrcPtr = Operands[0];
     auto *Mask = Operands[2];


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D81619.269988.patch
Type: text/x-patch
Size: 1027 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20200610/a595973f/attachment.bin>


More information about the llvm-commits mailing list