1 files changed, 21 insertions, 5 deletions
diff --git a/llvm/lib/Analysis/InstructionSimplify.cpp b/llvm/lib/Analysis/InstructionSimplify.cpp
index 609cd26..0cb2c78 100644
--- a/llvm/lib/Analysis/InstructionSimplify.cpp
+++ b/llvm/lib/Analysis/InstructionSimplify.cpp
@@ -3944,6 +3944,22 @@ static Value *SimplifyRelativeLoad(Constant *Ptr, Constant *Offset,
   return ConstantExpr::getBitCast(LoadedLHSPtr, Int8PtrTy);
 }
 
+static bool maskIsAllZeroOrUndef(Value *Mask) {
+  auto *ConstMask = dyn_cast<Constant>(Mask);
+  if (!ConstMask)
+    return false;
+  if (ConstMask->isNullValue() || isa<UndefValue>(ConstMask))
+    return true;
+  for (unsigned I = 0, E = ConstMask->getType()->getVectorNumElements(); I != E;
+       ++I) {
+    if (auto *MaskElt = ConstMask->getAggregateElement(I))
+      if (MaskElt->isNullValue() || isa<UndefValue>(MaskElt))
+        continue;
+    return false;
+  }
+  return true;
+}
+
 template <typename IterTy>
 static Value *SimplifyIntrinsic(Function *F, IterTy ArgBegin, IterTy ArgEnd,
                                 const Query &Q, unsigned MaxRecurse) {
@@ -3993,11 +4009,11 @@ static Value *SimplifyIntrinsic(Function *F, IterTy ArgBegin, IterTy ArgEnd,
 
   // Simplify calls to llvm.masked.load.*
   if (IID == Intrinsic::masked_load) {
-    IterTy MaskArg = ArgBegin + 2;
-    // If the mask is all zeros, the "passthru" argument is the result.
-    if (auto *ConstMask = dyn_cast<Constant>(*MaskArg))
-      if (ConstMask->isNullValue())
-        return ArgBegin[3];
+    Value *MaskArg = ArgBegin[2];
+    Value *PassthruArg = ArgBegin[3];
+    // If the mask is all zeros or undef, the "passthru" argument is the result.
+    if (maskIsAllZeroOrUndef(MaskArg))
+      return PassthruArg;
   }
 
   // Perform idempotent optimizations