diff --git a/lib/Transforms/Scalar/SpeculativeExecution.cpp b/lib/Transforms/Scalar/SpeculativeExecution.cpp
index 6c210edb254..9bf2d620681 100644
--- a/lib/Transforms/Scalar/SpeculativeExecution.cpp
+++ b/lib/Transforms/Scalar/SpeculativeExecution.cpp
@@ -62,6 +62,7 @@
 //===----------------------------------------------------------------------===//
 
 #include "llvm/ADT/SmallSet.h"
+#include "llvm/Analysis/GlobalsModRef.h"
 #include "llvm/Analysis/TargetTransformInfo.h"
 #include "llvm/Analysis/ValueTracking.h"
 #include "llvm/IR/Instructions.h"
@@ -138,6 +139,7 @@ INITIALIZE_PASS_END(SpeculativeExecution, "speculative-execution",
 
 void SpeculativeExecution::getAnalysisUsage(AnalysisUsage &AU) const {
   AU.addRequired<TargetTransformInfoWrapperPass>();
+  AU.addPreserved<GlobalsAAWrapperPass>();
 }
 
 bool SpeculativeExecution::runOnFunction(Function &F) {
diff --git a/test/Transforms/PhaseOrdering/globalaa-retained.ll b/test/Transforms/PhaseOrdering/globalaa-retained.ll
new file mode 100644
index 00000000000..bce193b5e85
--- /dev/null
+++ b/test/Transforms/PhaseOrdering/globalaa-retained.ll
@@ -0,0 +1,26 @@
+; RUN: opt -O3 -S < %s | FileCheck %s
+target datalayout = "e-m:e-i64:64-i128:128-n32:64-S128"
+target triple = "aarch64"
+
+@v = internal unnamed_addr global i32 0, align 4
+@p = common global i32* null, align 8
+
+; Function Attrs: norecurse nounwind
+define void @f(i32 %n) {
+entry:
+  %0 = load i32, i32* @v, align 4
+  %inc = add nsw i32 %0, 1
+  store i32 %inc, i32* @v, align 4
+  %1 = load i32*, i32** @p, align 8
+  store i32 %n, i32* %1, align 4
+  %2 = load i32, i32* @v, align 4
+  %inc1 = add nsw i32 %2, 1
+  store i32 %inc1, i32* @v, align 4
+  ret void
+}
+
+; check variable v is loaded only once after optimization, which should be
+; prove that globalsAA survives until the optimization that can use it to
+; optimize away the duplicate load/stores on variable v.
+; CHECK:     load i32, i32* @v, align 4
+; CHECK-NOT: load i32, i32* @v, align 4