diff --git a/clang/lib/CodeGen/CGStmtOpenMP.cpp b/clang/lib/CodeGen/CGStmtOpenMP.cpp
@@ -49,3 +49,30 @@ void CodeGenFunction::EmitOMPParallelDirective(const OMPParallelDirective &S) {
       CGOpenMPRuntime::OMPRTL__kmpc_fork_call);
   EmitRuntimeCall(RTLFn, Args);
 }
+
+void CodeGenFunction::EmitOMPSimdDirective(const OMPSimdDirective &S) {
+  const CapturedStmt *CS = cast<CapturedStmt>(S.getAssociatedStmt());
+  const Stmt *Body = CS->getCapturedStmt();
+  LoopStack.setParallel();
+  LoopStack.setVectorizerEnable(true);
+  for (auto C : S.clauses()) {
+    switch (C->getClauseKind()) {
+    case OMPC_safelen: {
+      RValue Len = EmitAnyExpr(cast<OMPSafelenClause>(C)->getSafelen(),
+                               AggValueSlot::ignored(), true);
+      llvm::ConstantInt *Val = cast<llvm::ConstantInt>(Len.getScalarVal());
+      LoopStack.setVectorizerWidth(Val->getZExtValue());
+      // In presence of finite 'safelen', it may be unsafe to mark all
+      // the memory instructions parallel, because loop-carried
+      // dependences of 'safelen' iterations are possible.
+      LoopStack.setParallel(false);
+      break;
+    }
+    default:
+      // Not handled yet
+      ;
+    }
+  }
+  EmitStmt(Body);
+}
+
diff --git a/clang/lib/CodeGen/CMakeLists.txt b/clang/lib/CodeGen/CMakeLists.txt
@@ -38,6 +38,7 @@ add_clang_library(clangCodeGen
   CGExprComplex.cpp
   CGExprConstant.cpp
   CGExprScalar.cpp
+  CGLoopInfo.cpp
   CGObjC.cpp
   CGObjCGNU.cpp
   CGObjCMac.cpp

diff --git a/clang/lib/CodeGen/CodeGenFunction.cpp b/clang/lib/CodeGen/CodeGenFunction.cpp
@@ -35,7 +35,8 @@ using namespace CodeGen;
 
 CodeGenFunction::CodeGenFunction(CodeGenModule &cgm, bool suppressNewContext)
     : CodeGenTypeCache(cgm), CGM(cgm), Target(cgm.getTarget()),
-      Builder(cgm.getModule().getContext()), CapturedStmtInfo(nullptr),
+      Builder(cgm.getModule().getContext(), llvm::ConstantFolder(),
+              CGBuilderInserterTy(this)), CapturedStmtInfo(nullptr),
       SanitizePerformTypeCheck(CGM.getSanOpts().Null |
                                CGM.getSanOpts().Alignment |
                                CGM.getSanOpts().ObjectSize |
@@ -1644,3 +1645,30 @@ llvm::Value *CodeGenFunction::EmitFieldAnnotations(const FieldDecl *D,
 }
 
 CodeGenFunction::CGCapturedStmtInfo::~CGCapturedStmtInfo() { }
+
+void CodeGenFunction::InsertHelper(llvm::Instruction *I,
+                                   const llvm::Twine &Name,
+                                   llvm::BasicBlock *BB,
+                                   llvm::BasicBlock::iterator InsertPt) const {
+  LoopStack.InsertHelper(I);
+}
+
+template <bool PreserveNames>
+void CGBuilderInserter<PreserveNames>::InsertHelper(
+    llvm::Instruction *I, const llvm::Twine &Name, llvm::BasicBlock *BB,
+    llvm::BasicBlock::iterator InsertPt) const {
+  llvm::IRBuilderDefaultInserter<PreserveNames>::InsertHelper(I, Name, BB,
+                                                              InsertPt);
+  if (CGF)
+    CGF->InsertHelper(I, Name, BB, InsertPt);
+}
+
+#ifdef NDEBUG
+#define PreserveNames false
+#else
+#define PreserveNames true
+#endif
+template void CGBuilderInserter<PreserveNames>::InsertHelper(
+    llvm::Instruction *I, const llvm::Twine &Name, llvm::BasicBlock *BB,
+    llvm::BasicBlock::iterator InsertPt) const;
+#undef PreserveNames
diff --git a/clang/lib/CodeGen/CodeGenFunction.h b/clang/lib/CodeGen/CodeGenFunction.h
@@ -16,6 +16,7 @@
 
 #include "CGBuilder.h"
 #include "CGDebugInfo.h"
+#include "CGLoopInfo.h"
 #include "CGValue.h"
 #include "CodeGenModule.h"
 #include "CodeGenPGO.h"
@@ -129,8 +130,15 @@ class CodeGenFunction : public CodeGenTypeCache {
   const TargetInfo &Target;
 
   typedef std::pair<llvm::Value *, llvm::Value *> ComplexPairTy;
+  LoopInfoStack LoopStack;
   CGBuilderTy Builder;
 
+  /// \brief CGBuilder insert helper. This function is called after an
+  /// instruction is created using Builder.
+  void InsertHelper(llvm::Instruction *I, const llvm::Twine &Name,
+                    llvm::BasicBlock *BB,
+                    llvm::BasicBlock::iterator InsertPt) const;
+
   /// CurFuncDecl - Holds the Decl for the current outermost
   /// non-closure context.
   const Decl *CurFuncDecl;
@@ -1883,6 +1891,7 @@ class CodeGenFunction : public CodeGenTypeCache {
   llvm::Value *GenerateCapturedStmtArgument(const CapturedStmt &S);
 
   void EmitOMPParallelDirective(const OMPParallelDirective &S);
+  void EmitOMPSimdDirective(const OMPSimdDirective &S);
 
   //===--------------------------------------------------------------------===//
   //                         LValue Expression Emission

diff --git a/clang/test/OpenMP/simd_metadata.c b/clang/test/OpenMP/simd_metadata.c
@@ -0,0 +1,52 @@
+// RUN: %clang_cc1 -fopenmp=libiomp5 -emit-llvm %s -o - | FileCheck %s
+
+void h1(float *c, float *a, float *b, int size)
+{
+// CHECK-LABEL: define void @h1
+  int t = 0;
+#pragma omp simd safelen(16) linear(t)
+  for (int i = 0; i < size; ++i) {
+    c[i] = a[i] * a[i] + b[i] * b[t];
+    ++t;
+// do not emit parallel_loop_access metadata due to usage of safelen clause.
+// CHECK-NOT: store float {{.+}}, float* {{.+}}, align {{.+}}, !llvm.mem.parallel_loop_access {{![0-9]+}}
+  }
+}
+
+void h2(float *c, float *a, float *b, int size)
+{
+// CHECK-LABEL: define void @h2
+  int t = 0;
+#pragma omp simd linear(t)
+  for (int i = 0; i < size; ++i) {
+    c[i] = a[i] * a[i] + b[i] * b[t];
+    ++t;
+// CHECK: store float {{.+}}, float* {{.+}}, align {{.+}}, !llvm.mem.parallel_loop_access [[LOOP_H2_HEADER:![0-9]+]]
+  }
+}
+
+void h3(float *c, float *a, float *b, int size)
+{
+// CHECK-LABEL: define void @h3
+#pragma omp simd
+  for (int i = 0; i < size; ++i) {
+    for (int j = 0; j < size; ++j) {
+      c[j*i] = a[i] * b[j];
+    }
+  }
+// do not emit parallel_loop_access for nested loop.
+// CHECK-NOT: store float {{.+}}, float* {{.+}}, align {{.+}}, !llvm.mem.parallel_loop_access {{![0-9]+}}
+}
+
+// Metadata for h1:
+// CHECK: [[LOOP_H1_HEADER:![0-9]+]] = metadata !{metadata [[LOOP_H1_HEADER]], metadata [[LOOP_WIDTH_16:![0-9]+]], metadata [[LOOP_VEC_ENABLE:![0-9]+]]}
+// CHECK: [[LOOP_WIDTH_16]] = metadata !{metadata !"llvm.vectorizer.width", i32 16}
+// CHECK: [[LOOP_VEC_ENABLE]] = metadata !{metadata !"llvm.vectorizer.enable", i1 true}
+//
+// Metadata for h2:
+// CHECK: [[LOOP_H2_HEADER]] = metadata !{metadata [[LOOP_H2_HEADER]], metadata [[LOOP_VEC_ENABLE]]}
+//
+// Metadata for h3:
+// CHECK: [[LOOP_H3_HEADER:![0-9]+]] = metadata !{metadata [[LOOP_H3_HEADER]], metadata [[LOOP_VEC_ENABLE]]}
+//
+