Index: include/polly/CodeGen/IslNodeBuilder.h
===================================================================
--- include/polly/CodeGen/IslNodeBuilder.h
+++ include/polly/CodeGen/IslNodeBuilder.h
@@ -262,6 +262,14 @@
   /// @param NewValues A map that maps certain llvm::Values to new llvm::Values.
   void updateValues(ValueMapT &NewValues);
 
+  /// Update the values referenced in the subtree to newer values
+  ///
+  /// When code generation continues in subfunction/kernel function, the values
+  /// must be updates with newest values if any.
+  ///
+  /// @param SubtreeValues A set of values that must updated
+  void updateSubtreeValues(SetVector<Value *> &SubtreeValues);
+
   /// Generate code for a marker now.
   ///
   /// For mark nodes with an unknown name, we just forward the code generation
Index: lib/CodeGen/IslNodeBuilder.cpp
===================================================================
--- lib/CodeGen/IslNodeBuilder.cpp
+++ lib/CodeGen/IslNodeBuilder.cpp
@@ -340,6 +340,24 @@
   }
 }
 
+void IslNodeBuilder::updateSubtreeValues(SetVector<Value *> &SubtreeValues) {
+
+  ValueMapT NewValues;
+
+  for (Value *V : SubtreeValues) {
+    Value *NewValue = ValueMap[V];
+    while (NewValue) {
+      NewValues[V] = NewValue;
+      NewValue = ValueMap[NewValue];
+    }
+  }
+
+  for (const auto &VP : NewValues) {
+    SubtreeValues.remove(VP.first);
+    SubtreeValues.insert(VP.second);
+  }
+}
+
 void IslNodeBuilder::createUserVector(__isl_take isl_ast_node *User,
                                       std::vector<Value *> &IVS,
                                       __isl_take isl_id *IteratorID,
@@ -624,6 +642,9 @@
     SubtreeValues.insert(V);
   }
 
+  // Update the values to be passed to subfunction to newer values if any.
+  updateSubtreeValues(SubtreeValues);
+
   ValueMapT NewValues;
   ParallelLoopGenerator ParallelLoopGen(Builder, LI, DT, DL);
 
Index: test/CodeGen/OpenMP/nested_parallel_loops.ll
===================================================================
--- /dev/null
+++ test/CodeGen/OpenMP/nested_parallel_loops.ll
@@ -0,0 +1,55 @@
+; RUN: opt %loadPolly -polly-opt-isl -polly-vectorizer=polly -polly-parallel -polly-tile-sizes=11,32,32 -polly-codegen -S < %s | FileCheck %s
+;
+; CHECK-DAG:     store [100 x float]* %polly.subfunc.arg.data, [100 x float]** %polly.subfn.storeaddr.polly.subfunc.arg.data
+; CHECK-DAG:     store float* %polly.subfunc.arg.mean, float** %polly.subfn.storeaddr.polly.subfunc.arg.mean
+;
+
+target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
+
+define void @func(i32 %m, i32 %n, [100 x float]* noalias %data, float* noalias %mean) {
+entry:
+  br label %entry.split
+
+entry.split:                                      ; preds = %entry
+  %cmp3 = icmp sgt i32 %m, 0
+  br i1 %cmp3, label %for.cond1.preheader.lr.ph, label %for.end10
+
+for.cond1.preheader.lr.ph:                        ; preds = %entry.split
+  br label %for.cond1.preheader
+
+for.cond1.preheader:                              ; preds = %for.cond1.preheader.lr.ph, %for.inc8
+  %indvars.iv5 = phi i64 [ 0, %for.cond1.preheader.lr.ph ], [ %indvars.iv.next6, %for.inc8 ]
+  %cmp21 = icmp sgt i32 %n, 0
+  br i1 %cmp21, label %for.body3.lr.ph, label %for.inc8
+
+for.body3.lr.ph:                                  ; preds = %for.cond1.preheader
+  br label %for.body3
+
+for.body3:                                        ; preds = %for.body3.lr.ph, %for.body3
+  %indvars.iv = phi i64 [ 0, %for.body3.lr.ph ], [ %indvars.iv.next, %for.body3 ]
+  %arrayidx5 = getelementptr inbounds [100 x float], [100 x float]* %data, i64 %indvars.iv, i64 %indvars.iv5
+  %0 = load float, float* %arrayidx5, align 4
+  %arrayidx7 = getelementptr inbounds float, float* %mean, i64 %indvars.iv5
+  %1 = load float, float* %arrayidx7, align 4
+  %add = fadd float %0, %1
+  store float %add, float* %arrayidx7, align 4
+  %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
+  %wide.trip.count = zext i32 %n to i64
+  %exitcond = icmp ne i64 %indvars.iv.next, %wide.trip.count
+  br i1 %exitcond, label %for.body3, label %for.cond1.for.inc8_crit_edge
+
+for.cond1.for.inc8_crit_edge:                     ; preds = %for.body3
+  br label %for.inc8
+
+for.inc8:                                         ; preds = %for.cond1.for.inc8_crit_edge, %for.cond1.preheader
+  %indvars.iv.next6 = add nuw nsw i64 %indvars.iv5, 1
+  %wide.trip.count7 = zext i32 %m to i64
+  %exitcond8 = icmp ne i64 %indvars.iv.next6, %wide.trip.count7
+  br i1 %exitcond8, label %for.cond1.preheader, label %for.cond.for.end10_crit_edge
+
+for.cond.for.end10_crit_edge:                     ; preds = %for.inc8
+  br label %for.end10
+
+for.end10:                                        ; preds = %for.cond.for.end10_crit_edge, %entry.split
+  ret void
+}