[OpenMP] Initial implementation of code generation for pragma 'distribute parallel for' on host

[carlo.bertolli]

Summary

This patch makes the following additions:

1. It abstracts away loop bound generation code from procedures associated with pragma 'for' and loops in general, in such a way that the same procedures can be used for 'distribute parallel for' without the need for a full re-implementation.

2. It implements code generation for 'distribute parallel for' and adds regression tests. It includes tests for clauses.

It is important to notice that most of the clauses are implemented as part of existing procedures. For instance, firstprivate is already implemented for 'distribute' and 'for' as separate pragmas. As the implementation of 'distribute parallel for' is based on the same procedures, then we automatically obtain implementation for such clauses without the need to add new code. However, this requires regression tests that verify correctness of produced code.

Looking forward to comments.

[ABataev]

Please provide diff with full context.
Also, the patch is too big. Could you split it into several patches?

lib/CodeGen/CGStmtOpenMP.cpp

1251

Do not use SmallVector as parameter type, use SmallVectorImpl instead.

[carlo.bertolli]

[OpenMP] Updating the patch to show full context

[carlo.bertolli]

I will split this patch into smaller patches, but I will keep this open in case I need to refer back to it: some of the changes that can go into independent patches require a justification and this patch shows it.
Thanks for your comments!

[ABataev]

In D29508#669800, @carlo.bertolli wrote:

I will split this patch into smaller patches, but I will keep this open in case I need to refer back to it: some of the changes that can go into independent patches require a justification and this patch shows it.
Thanks for your comments!

Sounds good, thanks!

[ABataev]

General note: many new lambdas are the same, it is worth it to convert them to functions, if possible

lib/CodeGen/CGOpenMPRuntime.h

678

This function has too many parameters, maybe it's worth it to put (some of)them in a struct.

lib/CodeGen/CGStmtOpenMP.cpp

1257

Do we really need the bounds emission generator in a parallel-like stuff?

1284–1288

I believe it's better to do it in Sema. You need to create variables for UB/LB and capture them, so you don't need to add non-parallel stuff to parallel code.

2012–2014

Better to use CGF.EmitScalarConversion()

2016–2018

Same here

[carlo.bertolli]

About this:

General note: many new lambdas are the same, it is worth it to convert them to functions, if possible

I added two static functions that return, respectively, two lambdas that are being used in the code gen for various compositions of pragma 'for' (e.g. 'pragma omp for',
'pragma omp for simd', etc.). As these are all the same, I could merge those in two functions.

lib/CodeGen/CGStmtOpenMP.cpp

1257

See next comment below.

1284–1288

About this one: I need help to understand what you precisely mean. I searched how to add a new stmt in the capture list, but by the time I am doing ActOnOpenMPDistributeParallelForDirective, the AssociateStmt list has already been created by parsing the content of the 'distribute parallel for' region.

Let me tell you what the generated code looks currently like and why I chose to have the code gen for the bound parameters in this file.
This is, roughly, what we need to generate:

void main() {
  ; forgive my lazyness in the following two lines..
  %lb.param = (int64_t) %lb
  %ub.param = (int64_t) %ub
  call @kmpc_fork_call (outlined, %lb.param, %ub.param, ..)

}

void outlined(%prev.lb, %prev.ub) {

  %lb = alloca i64,
  %ub = alloca i64,

  store %prev.lb, %lb
  store %prev.ub, %ub
}

A previous patch already takes care of adding the prev.lb and prev.ub to the formal parameter list of the outlined function. This is done in SemaOpenMP.cpp function ActOnOpenMPRegionStart.

The storing of the prev.lb and prev.ub parameters into the #for loop upper and lower bound variables (%lb, %ub) is also done in SemaOpenMP.cpp, CheckOpenMPLoop function. This corresponds to the code in the outlined function shown above.

For the caller, I followed the same logic as for 'taskloop' and 'taskloop simd', where the lower and upper bound actual parameters are populated in CGOpenMPRuntime.cpp, emitTaskLoopCall function.
This corresponds to the code in the caller of fork_call shown above.

Of course, I understand that taskloop is different from 'distribute parallel for', as the former is not a composite one.

[carlo.bertolli]

Address comments on initial patch after splitting. Added a dependency with sema patch.

[ABataev]

lib/CodeGen/CGOpenMPRuntime.cpp

2488

Seems to Chunk = part is erroneous

lib/CodeGen/CGOpenMPRuntime.h

675–680

Use /// style of comments here and add comments to all fields in this struct

682

Please, add comments for this function too

lib/CodeGen/CGStmtOpenMP.cpp

1249–1251

You don't need to use param names here, just types (inside function type)

1284–1288

This is what I meant. You're trying to capture variables %lb.param and %ub.param during codegen. But I think it is better to do it during sema analysis. You need to build some artificial variables for UB and LB and capture them in the captured region, if this is possible

1316

S is not required

1744–1748

Again, could you try to reduce number of parameters here?

1979–1980

Parameters are not needed here

1993

Why you're capturing S here if it is a parameter already?

1997

This can be defined outside of CGParallelFor lambda, too many enclosing levels

2035

Also, could you define it outside of CGParallelFor lambda?

2057

Do not capture S here (moreover, do not capture it by value), it is a parameter already. And move it out of CGParallelFor

2417–2426

No, you can convert lambda to a function and pass a pointer to function instead of lambda

3124–3128

Again, you can define it as a function and pass a pointer to this function in line 3130

4031

Name S is not needed

lib/CodeGen/CodeGenFunction.h

20

Do you really need this include?

2712–2713

Can it be converted into a static function local to a .cpp module? Also, this question is for other similar function, which are using new function_ref types introduced in lines 177-194. If you can do them static, these types also can be moved to .cpp file

[carlo.bertolli]

[OpenMP] Refactor after split of original patch in two patches (sema and cg). Transform lambdas to static functions, where possible, define default constructors, format, and update against trunk.

[ABataev]

LG

[carlo.bertolli]

Committed revision 301223.

[carlo.bertolli]

Re-Committed revision 301340.

include/clang/AST/StmtOpenMP.h

   /// allocated: loop counters, their updates and final values.
   /// PrevLowerBound and PrevUpperBound are used to communicate blocking
   /// information in composite constructs which require loop blocking
+  /// DistInc is used to generate the increment expression for the distribute
+  /// loop when combined with a further nested loop
+  /// PrevEnsureUpperBound is used as the EnsureUpperBound expression for the
+  /// for loop when combined with a previous distribute loop in the same pragma
+  /// (e.g. 'distribute parallel for')
   ///
   enum {
     AssociatedStmtOffset = 0,

     // specify the offset to the end (and start of the following counters/
     // updates/finals arrays).
     DefaultEnd = 9,
-    // The following 7 exprs are used by worksharing loops only.
+    // The following 12 exprs are used by worksharing and distribute loops only.
     IsLastIterVariableOffset = 9,
     LowerBoundVariableOffset = 10,
     UpperBoundVariableOffset = 11,

     NumIterationsOffset = 16,
     PrevLowerBoundVariableOffset = 17,
     PrevUpperBoundVariableOffset = 18,
+    DistIncOffset = 19,
+    PrevEnsureUpperBoundOffset = 20,
     // Offset to the end (and start of the following counters/updates/finals
     // arrays) for worksharing loop directives.
-    WorksharingEnd = 19,
+    WorksharingEnd = 21,
   };
 
   /// \brief Get the counters storage.

            "expected worksharing loop directive");
     *std::next(child_begin(), PrevUpperBoundVariableOffset) = PrevUB;
   }
+  void setDistInc(Expr *DistInc) {
+    assert((isOpenMPWorksharingDirective(getDirectiveKind()) ||
+            isOpenMPTaskLoopDirective(getDirectiveKind()) ||
+            isOpenMPDistributeDirective(getDirectiveKind())) &&
+           "expected worksharing loop directive");
+    *std::next(child_begin(), DistIncOffset) = DistInc;
+  }
+  void setPrevEnsureUpperBound(Expr *PrevEUB) {
+    assert((isOpenMPWorksharingDirective(getDirectiveKind()) ||
+            isOpenMPTaskLoopDirective(getDirectiveKind()) ||
+            isOpenMPDistributeDirective(getDirectiveKind())) &&
+           "expected worksharing loop directive");
+    *std::next(child_begin(), PrevEnsureUpperBoundOffset) = PrevEUB;
+  }
   void setCounters(ArrayRef<Expr *> A);
   void setPrivateCounters(ArrayRef<Expr *> A);
   void setInits(ArrayRef<Expr *> A);

     Expr *UB;
     /// \brief Stride - local variable passed to runtime.
     Expr *ST;
-    /// \brief EnsureUpperBound -- expression LB = min(LB, NumIterations).
+    /// \brief EnsureUpperBound -- expression UB = min(UB, NumIterations).
     Expr *EUB;
     /// \brief Update of LowerBound for statically sheduled 'omp for' loops.
     Expr *NLB;

     /// \brief PreviousUpperBound - local variable passed to runtime in the
     /// enclosing schedule or null if that does not apply.
     Expr *PrevUB;
+    /// \brief DistInc - increment expression for distribute loop when found
+    /// combined with a further loop level (e.g. in 'distribute parallel for')
+    /// expression IV = IV + ST
+    Expr *DistInc;
+    /// \brief PrevEUB - expression similar to EUB but to be used when loop
+    /// scheduling uses PrevLB and PrevUB (e.g.  in 'distribute parallel for'
+    /// when ensuring that the UB is either the calculated UB by the runtime or
+    /// the end of the assigned distribute chunk)
+    /// expression UB = min (UB, PrevUB)
+    Expr *PrevEUB;
     /// \brief Counters Loop counters.
     SmallVector<Expr *, 4> Counters;
     /// \brief PrivateCounters Loop counters.

       NumIterations = nullptr;
       PrevLB = nullptr;
       PrevUB = nullptr;
+      DistInc = nullptr;
+      PrevEUB = nullptr;
       Counters.resize(Size);
       PrivateCounters.resize(Size);
       Inits.resize(Size);

     return const_cast<Expr *>(reinterpret_cast<const Expr *>(
         *std::next(child_begin(), PrevUpperBoundVariableOffset)));
   }
+  Expr *getDistInc() const {
+    assert((isOpenMPWorksharingDirective(getDirectiveKind()) ||
+            isOpenMPTaskLoopDirective(getDirectiveKind()) ||
+            isOpenMPDistributeDirective(getDirectiveKind())) &&
+           "expected worksharing loop directive");
+    return const_cast<Expr *>(reinterpret_cast<const Expr *>(
+        *std::next(child_begin(), DistIncOffset)));
+  }
+  Expr *getPrevEnsureUpperBound() const {
+    assert((isOpenMPWorksharingDirective(getDirectiveKind()) ||
+            isOpenMPTaskLoopDirective(getDirectiveKind()) ||
+            isOpenMPDistributeDirective(getDirectiveKind())) &&
+           "expected worksharing loop directive");
+    return const_cast<Expr *>(reinterpret_cast<const Expr *>(
+        *std::next(child_begin(), PrevEnsureUpperBoundOffset)));
+  }
   const Stmt *getBody() const {
     // This relies on the loop form is already checked by Sema.
     Stmt *Body = getAssociatedStmt()->IgnoreContainers(true);

lib/AST/StmtOpenMP.cpp

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

   Dir->setNumIterations(Exprs.NumIterations);
   Dir->setPrevLowerBoundVariable(Exprs.PrevLB);
   Dir->setPrevUpperBoundVariable(Exprs.PrevUB);
+  Dir->setDistInc(Exprs.DistInc);
+  Dir->setPrevEnsureUpperBound(Exprs.PrevEUB);
   Dir->setCounters(Exprs.Counters);
   Dir->setPrivateCounters(Exprs.PrivateCounters);
   Dir->setInits(Exprs.Inits);

lib/Basic/OpenMPKinds.cpp

   case OMPD_parallel_for:
   case OMPD_parallel_for_simd:
   case OMPD_parallel_sections:
+  case OMPD_distribute_parallel_for:
     CaptureRegions.push_back(OMPD_parallel);
     break;
   case OMPD_target_teams:

   case OMPD_taskloop_simd:
   case OMPD_distribute_parallel_for_simd:
   case OMPD_distribute_simd:
-  case OMPD_distribute_parallel_for:
   case OMPD_teams_distribute:
   case OMPD_teams_distribute_simd:
   case OMPD_teams_distribute_parallel_for_simd:

lib/CodeGen/CGOpenMPRuntime.h

   virtual void emitForDispatchInit(CodeGenFunction &CGF, SourceLocation Loc,
                                    const OpenMPScheduleTy &ScheduleKind,
                                    unsigned IVSize, bool IVSigned, bool Ordered,
-                                   llvm::Value *UB,
+                                   llvm::Value *LB, llvm::Value *UB,
                                    llvm::Value *Chunk = nullptr);
 
   /// \brief Call the appropriate runtime routine to initialize it before start

lib/CodeGen/CGOpenMPRuntime.cpp

                                           SourceLocation Loc,
                                           const OpenMPScheduleTy &ScheduleKind,
                                           unsigned IVSize, bool IVSigned,
-                                          bool Ordered, llvm::Value *UB,
-                                          llvm::Value *Chunk) {
+                                          bool Ordered, llvm::Value *LB,
+                                          llvm::Value *UB, llvm::Value *Chunk) {
   if (!CGF.HaveInsertPoint())
     return;
   OpenMPSchedType Schedule =

       emitUpdateLocation(CGF, Loc), getThreadID(CGF, Loc),
       CGF.Builder.getInt32(addMonoNonMonoModifier(
           Schedule, ScheduleKind.M1, ScheduleKind.M2)), // Schedule type
-      CGF.Builder.getIntN(IVSize, 0),                   // Lower
+      LB,                                               // Lower
       UB,                                               // Upper
       CGF.Builder.getIntN(IVSize, 1),                   // Stride
       Chunk                                             // Chunk

lib/CodeGen/CGStmtOpenMP.cpp

 class OMPParallelScope final : public OMPLexicalScope {
   bool EmitPreInitStmt(const OMPExecutableDirective &S) {
     OpenMPDirectiveKind Kind = S.getDirectiveKind();
-    return !isOpenMPTargetExecutionDirective(Kind) &&
+    return !(isOpenMPTargetExecutionDirective(Kind) ||
+             isOpenMPLoopBoundSharingDirective(Kind)) &&
            isOpenMPParallelDirective(Kind);
   }
 

     CGF.EmitBlock(DoneBB, /*IsFinished=*/true);
 }
 
-static void emitCommonOMPParallelDirective(CodeGenFunction &CGF,
-                                           const OMPExecutableDirective &S,
-                                           OpenMPDirectiveKind InnermostKind,
-                                           const RegionCodeGenTy &CodeGen) {
+// Codegen lambda for appending distribute lower and upper bounds to outlined
+// parallel funciton. This is necessary for combined constructs such as
+// 'distribute parallel for'
+typedef llvm::function_ref<void(CodeGenFunction &CGF,
+                                const OMPExecutableDirective &S,
+                                llvm::SmallVector<llvm::Value *, 16> &)>
+    CodeGenBoundParametersTy;
+
+static void emitCommonOMPParallelDirective(
+    CodeGenFunction &CGF, const OMPExecutableDirective &S,
+    OpenMPDirectiveKind InnermostKind, const RegionCodeGenTy &CodeGen,
+    const CodeGenBoundParametersTy &CodeGenBoundParameters) {
   const CapturedStmt *CS = S.getCapturedStmt(OMPD_parallel);
   auto OutlinedFn = CGF.CGM.getOpenMPRuntime().emitParallelOutlinedFunction(
       S, *CS->getCapturedDecl()->param_begin(), InnermostKind, CodeGen);

 
   OMPParallelScope Scope(CGF, S);
   llvm::SmallVector<llvm::Value *, 16> CapturedVars;
+  // Combining 'distribute' with 'for' requires sharing each 'distribute' chunk
+  // lower and upper bounds with the pragma 'for' chunking mechanism.
+  // The following lambda takes care of appending the lower and upper bound
+  // parameters when necessary
+  CodeGenBoundParameters(CGF, S, CapturedVars);
   CGF.GenerateOpenMPCapturedVars(*CS, CapturedVars);
   CGF.CGM.getOpenMPRuntime().emitParallelCall(CGF, S.getLocStart(), OutlinedFn,
                                               CapturedVars, IfCond);

     CGF.EmitStmt(cast<CapturedStmt>(S.getAssociatedStmt())->getCapturedStmt());
     CGF.EmitOMPReductionClauseFinal(S);
   };
-  emitCommonOMPParallelDirective(*this, S, OMPD_parallel, CodeGen);
+  emitCommonOMPParallelDirective(*this, S, OMPD_parallel, CodeGen,
+                                 [](CodeGenFunction &,
+                                    const OMPExecutableDirective &S,
+                                    llvm::SmallVector<llvm::Value *, 16> &) {});
   emitPostUpdateForReductionClause(
       *this, S, [](CodeGenFunction &) -> llvm::Value * { return nullptr; });
 }

   CGM.getOpenMPRuntime().emitInlinedDirective(*this, OMPD_simd, CodeGen);
 }
 
-void CodeGenFunction::EmitOMPOuterLoop(bool DynamicOrOrdered, bool IsMonotonic,
-    const OMPLoopDirective &S, OMPPrivateScope &LoopScope, bool Ordered,
-    Address LB, Address UB, Address ST, Address IL, llvm::Value *Chunk) {
+void CodeGenFunction::EmitOMPOuterLoop(
+    bool DynamicOrOrdered, bool IsMonotonic, const OMPLoopDirective &S,
+    OMPPrivateScope &LoopScope, Address LB, Address UB, Address ST, Address IL,
+    llvm::Value *Chunk, Expr *EUB, Expr *IncExpr,
+    const CodeGenLoopTy &CodeGenLoop, const CodeGenOrderedTy &CodeGenOrdered) {
   auto &RT = CGM.getOpenMPRuntime();
 
   const Expr *IVExpr = S.getIterationVariable();

 
   llvm::Value *BoolCondVal = nullptr;
   if (!DynamicOrOrdered) {
-    // UB = min(UB, GlobalUB)
-    EmitIgnoredExpr(S.getEnsureUpperBound());
+    // UB = min(UB, GlobalUB) or
+    // UB = min(UB, PrevUB) for combined loop sharing constructs (e.g.
+    // 'distribute parallel for')
+    EmitIgnoredExpr(EUB);
     // IV = LB
     EmitIgnoredExpr(S.getInit());
     // IV < UB

     EmitOMPSimdInit(S, IsMonotonic);
 
   SourceLocation Loc = S.getLocStart();
-  EmitOMPInnerLoop(S, LoopScope.requiresCleanups(), S.getCond(), S.getInc(),
-                   [&S, LoopExit](CodeGenFunction &CGF) {
-                     CGF.EmitOMPLoopBody(S, LoopExit);
-                     CGF.EmitStopPoint(&S);
-                   },
-                   [Ordered, IVSize, IVSigned, Loc](CodeGenFunction &CGF) {
-                     if (Ordered) {
-                       CGF.CGM.getOpenMPRuntime().emitForOrderedIterationEnd(
-                           CGF, Loc, IVSize, IVSigned);
-                     }
-                   });
+
+  // when 'distribute' is not combined with a 'for':
+  // while (idx <= UB) { BODY; ++idx; }
+  // when 'distribute' is combined with a 'for'
+  // (e.g. 'distribute parallel for')
+  // while (idx <= UB) { <CodeGen rest of pragma>; idx += ST; }
+  EmitOMPInnerLoop(
+      S, LoopScope.requiresCleanups(), S.getCond(), IncExpr,
+      [&S, LoopExit, &CodeGenLoop](CodeGenFunction &CGF) {
+        CodeGenLoop(CGF, S, LoopExit);
+      },
+      [IVSize, IVSigned, Loc, &CodeGenOrdered](CodeGenFunction &CGF) {
+        CodeGenOrdered(CGF, Loc, IVSize, IVSigned);
+      });
 
   EmitBlock(Continue.getBlock());
   BreakContinueStack.pop_back();

 void CodeGenFunction::EmitOMPForOuterLoop(
     const OpenMPScheduleTy &ScheduleKind, bool IsMonotonic,
     const OMPLoopDirective &S, OMPPrivateScope &LoopScope, bool Ordered,
-    Address LB, Address UB, Address ST, Address IL, llvm::Value *Chunk) {
+    Address LB, Address UB, Address ST, Address IL, llvm::Value *Chunk,
+    Expr *EUB, const CodeGenDispatchBoundsTy &CGDispatchBounds) {
   auto &RT = CGM.getOpenMPRuntime();
 
   // Dynamic scheduling of the outer loop (dynamic, guided, auto, runtime).

   const bool IVSigned = IVExpr->getType()->hasSignedIntegerRepresentation();
 
   if (DynamicOrOrdered) {
-    llvm::Value *UBVal = EmitScalarExpr(S.getLastIteration());
+    auto DispatchBounds = CGDispatchBounds(*this, LB, UB);
+    llvm::Value *LBVal = DispatchBounds.first;
+    llvm::Value *UBVal = DispatchBounds.second;
     RT.emitForDispatchInit(*this, S.getLocStart(), ScheduleKind, IVSize,
-                           IVSigned, Ordered, UBVal, Chunk);
+                           IVSigned, Ordered, LBVal, UBVal, Chunk);
   } else {
     RT.emitForStaticInit(*this, S.getLocStart(), ScheduleKind, IVSize, IVSigned,
                          Ordered, IL, LB, UB, ST, Chunk);
   }
 
-  EmitOMPOuterLoop(DynamicOrOrdered, IsMonotonic, S, LoopScope, Ordered, LB, UB,
-                   ST, IL, Chunk);
+  auto &&CodeGenLoop = [](CodeGenFunction &CGF, const OMPLoopDirective &S,
+                          JumpDest LoopExit) {
+    CGF.EmitOMPLoopBody(S, LoopExit);
+    CGF.EmitStopPoint(&S);
+  };
+
+  auto &&CodeGenOrdered = [Ordered](CodeGenFunction &CGF, SourceLocation Loc,
+                                    const unsigned IVSize,
+                                    const bool IVSigned) {
+    if (Ordered) {
+      CGF.CGM.getOpenMPRuntime().emitForOrderedIterationEnd(CGF, Loc, IVSize,
+                                                            IVSigned);
+    }
+  };
+
+  EmitOMPOuterLoop(DynamicOrOrdered, IsMonotonic, S, LoopScope, LB, UB, ST, IL,
+                   Chunk, EUB, S.getInc(), CodeGenLoop, CodeGenOrdered);
 }
 
 void CodeGenFunction::EmitOMPDistributeOuterLoop(
-    OpenMPDistScheduleClauseKind ScheduleKind,
-    const OMPDistributeDirective &S, OMPPrivateScope &LoopScope,
-    Address LB, Address UB, Address ST, Address IL, llvm::Value *Chunk) {
+    OpenMPDistScheduleClauseKind ScheduleKind, const OMPLoopDirective &S,
+    OMPPrivateScope &LoopScope, Address LB, Address UB, Address ST, Address IL,
+    llvm::Value *Chunk, const CodeGenLoopTy &CodeGenLoopContent) {
 
   auto &RT = CGM.getOpenMPRuntime();
 

                               IVSize, IVSigned, /* Ordered = */ false,
                               IL, LB, UB, ST, Chunk);
 
-  EmitOMPOuterLoop(/* DynamicOrOrdered = */ false, /* IsMonotonic = */ false,
-                   S, LoopScope, /* Ordered = */ false, LB, UB, ST, IL, Chunk);
+  // for combined 'distribute' and 'for' the increment expression of distribute
+  // is store in DistInc. For 'distribute' alone, it is in Inc.
+  Expr *IncExpr = nullptr;
+  if (isOpenMPLoopBoundSharingDirective(S.getDirectiveKind()))
+    IncExpr = S.getDistInc();
+  else
+    IncExpr = S.getInc();
+
+  EmitOMPOuterLoop(/* DynamicOrOrdered = */ false, /* IsMonotonic = */ false, S,
+                   LoopScope, LB, UB, ST, IL, Chunk, S.getEnsureUpperBound(),
+                   IncExpr, CodeGenLoopContent,
+                   [](CodeGenFunction &, SourceLocation Loc,
+                      const unsigned IVSize, const bool IVSigned) {});
+}
+
+/// \brief Emit a helper variable and return corresponding lvalue.
+static LValue EmitOMPHelperVar(CodeGenFunction &CGF,
+                               const DeclRefExpr *Helper) {
+  auto VDecl = cast<VarDecl>(Helper->getDecl());
+  CGF.EmitVarDecl(*VDecl);
+  return CGF.EmitLValue(Helper);
 }
 
 void CodeGenFunction::EmitOMPDistributeParallelForDirective(
     const OMPDistributeParallelForDirective &S) {
+  auto &&CGParallelFor = [&S](CodeGenFunction &CGF, const OMPLoopDirective &S,
+                              JumpDest LoopExit) {
+    auto &&CGInlinedWorksharingLoop = [&S](CodeGenFunction &CGF,
+                                           PrePostActionTy &) {
+      auto &&CodeGenBounds = [&S](CodeGenFunction &CGF) {
+        LValue LB =
+            EmitOMPHelperVar(CGF, cast<DeclRefExpr>(S.getLowerBoundVariable()));
+        LValue UB =
+            EmitOMPHelperVar(CGF, cast<DeclRefExpr>(S.getUpperBoundVariable()));
+
+        // When composing 'distribute' with 'for' (e.g. as in 'distribute
+        // parallel for') we need to use the 'distribute'
+        // chunk lower and upper bounds rather than the whole loop iteration
+        // space. These are parameters to the outlined function for 'parallel'
+        // and we copy the bounds of the previous schedule into the
+        // the current ones.
+        const unsigned IVSize = CGF.getContext().getTypeSize(
+            cast<DeclRefExpr>(S.getIterationVariable())->getType());
+        LValue PrevLB = CGF.EmitLValue(S.getPrevLowerBoundVariable());
+        LValue PrevUB = CGF.EmitLValue(S.getPrevUpperBoundVariable());
+        auto PrevLBVal = CGF.EmitLoadOfScalar(PrevLB, SourceLocation());
+        PrevLBVal =
+            CGF.Builder.CreateIntCast(PrevLBVal, CGF.Builder.getIntNTy(IVSize),
+                                      /*isSigned=*/false);
+        auto PrevUBVal = CGF.EmitLoadOfScalar(PrevUB, SourceLocation());
+        PrevUBVal =
+            CGF.Builder.CreateIntCast(PrevUBVal, CGF.Builder.getIntNTy(IVSize),
+                                      /*isSigned=*/false);
+        CGF.EmitStoreOfScalar(PrevLBVal, LB);
+        CGF.EmitStoreOfScalar(PrevUBVal, UB);
+
+        return std::make_pair(LB, UB);
+      };
+
+      // if the 'for' loop has a dispatch schedule (e.g. dynamic, guided) then
+      // we need to use the LB and UB expressions generated by the worksharing
+      // code generation support, whereas in non combined situations we would
+      // just emit 0 and the LastIteration expression
+      // This lambda is necessary due to the difference of the LB and UB
+      // types for the RT emission routines for 'for_static_init' and
+      // 'for_dispatch_init'
+      auto &&CodeGenDispatchBounds = [&S](CodeGenFunction &CGF, Address LB,
+                                          Address UB) {
+        const Expr *IVExpr = S.getIterationVariable();
+        // when implementing a dynamic schedule for a 'for' combined with a
+        // 'distribute' (e.g. 'distribute parallel for'), the 'for' loop
+        // is not normalized as each team only executes its own assigned
+        // distribute chunk
+        QualType IteratorTy = IVExpr->getType();
+        llvm::Value *LBVal = CGF.EmitLoadOfScalar(LB, /*Volatile=*/false,
+                                                  IteratorTy, SourceLocation());
+        llvm::Value *UBVal = CGF.EmitLoadOfScalar(UB, /*Volatile=*/false,
+                                                  IteratorTy, SourceLocation());
+        return std::make_pair(LBVal, UBVal);
+      };
+
+      CGF.EmitOMPWorksharingLoop(S, S.getPrevEnsureUpperBound(), CodeGenBounds,
+                                 CodeGenDispatchBounds);
+    };
+
+    auto &&CodeGenDistributeBoundParams =
+        [S](CodeGenFunction &CGF, const OMPExecutableDirective &S,
+            llvm::SmallVector<llvm::Value *, 16> &CapturedVars) {
+          const OMPLoopDirective &Dir = cast<OMPLoopDirective>(S);
+          LValue LB =
+              CGF.EmitLValue(cast<DeclRefExpr>(Dir.getLowerBoundVariable()));
+          auto LBCast =
+              CGF.Builder.CreateIntCast(CGF.Builder.CreateLoad(LB.getAddress()),
+                                        CGF.SizeTy, /*isSigned=*/false);
+          CapturedVars.push_back(LBCast);
+          LValue UB =
+              CGF.EmitLValue(cast<DeclRefExpr>(Dir.getUpperBoundVariable()));
+          auto UBCast =
+              CGF.Builder.CreateIntCast(CGF.Builder.CreateLoad(UB.getAddress()),
+                                        CGF.SizeTy, /*isSigned=*/false);
+          CapturedVars.push_back(UBCast);
+        };
+    emitCommonOMPParallelDirective(CGF, S, OMPD_for, CGInlinedWorksharingLoop,
+                                   CodeGenDistributeBoundParams);
+  };
+
+  auto &&CodeGen = [&S, &CGParallelFor](CodeGenFunction &CGF,
+                                        PrePostActionTy &) {
+    CGF.EmitOMPDistributeLoop(S, CGParallelFor, S.getDistInc());
+  };
   OMPLexicalScope Scope(*this, S, /*AsInlined=*/true);
-  CGM.getOpenMPRuntime().emitInlinedDirective(
-      *this, OMPD_distribute_parallel_for,
-      [&S](CodeGenFunction &CGF, PrePostActionTy &) {
-        OMPLoopScope PreInitScope(CGF, S);
-        OMPCancelStackRAII CancelRegion(CGF, OMPD_distribute_parallel_for,
-                                        /*HasCancel=*/false);
-        CGF.EmitStmt(
-            cast<CapturedStmt>(S.getAssociatedStmt())->getCapturedStmt());
-      });
+  OMPCancelStackRAII CancelRegion(*this, OMPD_distribute_parallel_for,
+                                  /*HasCancel=*/false);
+  CGM.getOpenMPRuntime().emitInlinedDirective(*this, OMPD_distribute, CodeGen,
+                                              false);
 }
 
 void CodeGenFunction::EmitOMPDistributeParallelForSimdDirective(

       });
 }
 
-/// \brief Emit a helper variable and return corresponding lvalue.
-static LValue EmitOMPHelperVar(CodeGenFunction &CGF,
-                               const DeclRefExpr *Helper) {
-  auto VDecl = cast<VarDecl>(Helper->getDecl());
-  CGF.EmitVarDecl(*VDecl);
-  return CGF.EmitLValue(Helper);
-}
-
 namespace {
   struct ScheduleKindModifiersTy {
     OpenMPScheduleClauseKind Kind;

   };
 } // namespace
 
-bool CodeGenFunction::EmitOMPWorksharingLoop(const OMPLoopDirective &S) {
+bool CodeGenFunction::EmitOMPWorksharingLoop(
+    const OMPLoopDirective &S, Expr *EUB,
+    const CodeGenLoopBoundsTy &CodeGenLoopBounds,
+    const CodeGenDispatchBoundsTy &CGDispatchBounds) {
   // Emit the loop iteration variable.
   auto IVExpr = cast<DeclRefExpr>(S.getIterationVariable());
   auto IVDecl = cast<VarDecl>(IVExpr->getDecl());

     emitAlignedClause(*this, S);
     EmitOMPLinearClauseInit(S);
     // Emit helper vars inits.
-    LValue LB =
-        EmitOMPHelperVar(*this, cast<DeclRefExpr>(S.getLowerBoundVariable()));
-    LValue UB =
-        EmitOMPHelperVar(*this, cast<DeclRefExpr>(S.getUpperBoundVariable()));
+
+    std::pair<LValue, LValue> Bounds = CodeGenLoopBounds(*this);
+    LValue LB = Bounds.first;
+    LValue UB = Bounds.second;
     LValue ST =
         EmitOMPHelperVar(*this, cast<DeclRefExpr>(S.getStrideVariable()));
     LValue IL =

         // runtime and runs the inner loop to process it.
         EmitOMPForOuterLoop(ScheduleKind, IsMonotonic, S, LoopScope, Ordered,
                             LB.getAddress(), UB.getAddress(), ST.getAddress(),
-                            IL.getAddress(), Chunk);
+                            IL.getAddress(), Chunk, EUB, CGDispatchBounds);
       }
       if (isOpenMPSimdDirective(S.getDirectiveKind())) {
         EmitOMPSimdFinal(S,

   auto &&CodeGen = [&S, &HasLastprivates](CodeGenFunction &CGF,
                                           PrePostActionTy &) {
     OMPCancelStackRAII CancelRegion(CGF, OMPD_for, S.hasCancel());
-    HasLastprivates = CGF.EmitOMPWorksharingLoop(S);
+
+    auto &&CodeGenBounds = [&S](CodeGenFunction &CGF) {
+      LValue LB =
+          EmitOMPHelperVar(CGF, cast<DeclRefExpr>(S.getLowerBoundVariable()));
+      LValue UB =
+          EmitOMPHelperVar(CGF, cast<DeclRefExpr>(S.getUpperBoundVariable()));
+      return std::make_pair(LB, UB);
+    };
+
+    // when dealing with dispatch schedules (e.g. dynamic, guided) we do not
+    // consider the lower and upper bound expressions generated by the
+    // worksharing
+    // loop support, but we use 0 and the iteration space size as constants
+    auto &&CodeGenDispatchBounds = [&S](CodeGenFunction &CGF, Address LB,
+                                        Address UB) {
+      const Expr *IVExpr = S.getIterationVariable();
+      const unsigned IVSize = CGF.getContext().getTypeSize(IVExpr->getType());
+      llvm::Value *LBVal = CGF.Builder.getIntN(IVSize, 0);
+      llvm::Value *UBVal = CGF.EmitScalarExpr(S.getLastIteration());
+      return std::make_pair(LBVal, UBVal);
+    };
+
+    HasLastprivates = CGF.EmitOMPWorksharingLoop(
+        S, S.getEnsureUpperBound(), CodeGenBounds, CodeGenDispatchBounds);
   };
   {
     OMPLexicalScope Scope(*this, S, /*AsInlined=*/true);

   bool HasLastprivates = false;
   auto &&CodeGen = [&S, &HasLastprivates](CodeGenFunction &CGF,
                                           PrePostActionTy &) {
-    HasLastprivates = CGF.EmitOMPWorksharingLoop(S);
+    auto &&CodeGenBounds = [&S](CodeGenFunction &CGF) {
+      LValue LB =
+          EmitOMPHelperVar(CGF, cast<DeclRefExpr>(S.getLowerBoundVariable()));
+      LValue UB =
+          EmitOMPHelperVar(CGF, cast<DeclRefExpr>(S.getUpperBoundVariable()));
+      return std::make_pair(LB, UB);
+    };
+
+    // when dealing with dispatch schedules (e.g. dynamic, guided) we do not
+    // consider the lower and upper bound expressions generated by the
+    // worksharing
+    // loop support, but we use 0 and the iteration space size as constants
+    auto &&CodeGenDispatchBounds = [&S](CodeGenFunction &CGF, Address LB,
+                                        Address UB) {
+      const Expr *IVExpr = S.getIterationVariable();
+      const unsigned IVSize = CGF.getContext().getTypeSize(IVExpr->getType());
+      llvm::Value *LBVal = CGF.Builder.getIntN(IVSize, 0);
+      llvm::Value *UBVal = CGF.EmitScalarExpr(S.getLastIteration());
+      return std::make_pair(LBVal, UBVal);
+    };
+
+    HasLastprivates = CGF.EmitOMPWorksharingLoop(
+        S, S.getEnsureUpperBound(), CodeGenBounds, CodeGenDispatchBounds);
   };
   {
     OMPLexicalScope Scope(*this, S, /*AsInlined=*/true);

   // directives: 'parallel' with 'for' directive.
   auto &&CodeGen = [&S](CodeGenFunction &CGF, PrePostActionTy &) {
     OMPCancelStackRAII CancelRegion(CGF, OMPD_parallel_for, S.hasCancel());
-    CGF.EmitOMPWorksharingLoop(S);
+
+    auto &&CodeGenBounds = [&S](CodeGenFunction &CGF) {
+      LValue LB =
+          EmitOMPHelperVar(CGF, cast<DeclRefExpr>(S.getLowerBoundVariable()));
+      LValue UB =
+          EmitOMPHelperVar(CGF, cast<DeclRefExpr>(S.getUpperBoundVariable()));
+      return std::make_pair(LB, UB);
+    };
+
+    // when dealing with dispatch schedules (e.g. dynamic, guided) we do not
+    // consider the lower and upper bound expressions generated by the
+    // worksharing
+    // loop support, but we use 0 and the iteration space size as constants
+    auto &&CodeGenDispatchBounds = [&S](CodeGenFunction &CGF, Address LB,
+                                        Address UB) {
+      const Expr *IVExpr = S.getIterationVariable();
+      const unsigned IVSize = CGF.getContext().getTypeSize(IVExpr->getType());
+      llvm::Value *LBVal = CGF.Builder.getIntN(IVSize, 0);
+      llvm::Value *UBVal = CGF.EmitScalarExpr(S.getLastIteration());
+      return std::make_pair(LBVal, UBVal);
+    };
+    CGF.EmitOMPWorksharingLoop(S, S.getEnsureUpperBound(), CodeGenBounds,
+                               CodeGenDispatchBounds);
   };
-  emitCommonOMPParallelDirective(*this, S, OMPD_for, CodeGen);
+  emitCommonOMPParallelDirective(*this, S, OMPD_for, CodeGen,
+                                 [](CodeGenFunction &,
+                                    const OMPExecutableDirective &S,
+                                    llvm::SmallVector<llvm::Value *, 16> &) {});
 }
 
 void CodeGenFunction::EmitOMPParallelForSimdDirective(

   // Emit directive as a combined directive that consists of two implicit
   // directives: 'parallel' with 'for' directive.
   auto &&CodeGen = [&S](CodeGenFunction &CGF, PrePostActionTy &) {
-    CGF.EmitOMPWorksharingLoop(S);
+    auto &&CodeGenBounds = [&S](CodeGenFunction &CGF) {
+      LValue LB =
+          EmitOMPHelperVar(CGF, cast<DeclRefExpr>(S.getLowerBoundVariable()));
+      LValue UB =
+          EmitOMPHelperVar(CGF, cast<DeclRefExpr>(S.getUpperBoundVariable()));
+      return std::make_pair(LB, UB);
+    };
+    // when dealing with dispatch schedules (e.g. dynamic, guided) we do not
+    // consider the lower and upper bound expressions generated by the
+    // worksharing
+    // loop support, but we use 0 and the iteration space size as constants
+    auto &&CodeGenDispatchBounds = [&S](CodeGenFunction &CGF, Address LB,
+                                        Address UB) {
+      const Expr *IVExpr = S.getIterationVariable();
+      const unsigned IVSize = CGF.getContext().getTypeSize(IVExpr->getType());
+      llvm::Value *LBVal = CGF.Builder.getIntN(IVSize, 0);
+      llvm::Value *UBVal = CGF.EmitScalarExpr(S.getLastIteration());
+      return std::make_pair(LBVal, UBVal);
+    };
+
+    CGF.EmitOMPWorksharingLoop(S, S.getEnsureUpperBound(), CodeGenBounds,
+                               CodeGenDispatchBounds);
   };
-  emitCommonOMPParallelDirective(*this, S, OMPD_simd, CodeGen);
+  emitCommonOMPParallelDirective(*this, S, OMPD_simd, CodeGen,
+                                 [](CodeGenFunction &,
+                                    const OMPExecutableDirective &S,
+                                    llvm::SmallVector<llvm::Value *, 16> &) {});
 }
 
 void CodeGenFunction::EmitOMPParallelSectionsDirective(

   auto &&CodeGen = [&S](CodeGenFunction &CGF, PrePostActionTy &) {
     CGF.EmitSections(S);
   };
-  emitCommonOMPParallelDirective(*this, S, OMPD_sections, CodeGen);
+  emitCommonOMPParallelDirective(*this, S, OMPD_sections, CodeGen,
+                                 [](CodeGenFunction &,
+                                    const OMPExecutableDirective &S,
+                                    llvm::SmallVector<llvm::Value *, 16> &) {});
 }
 
 void CodeGenFunction::EmitOMPTaskBasedDirective(const OMPExecutableDirective &S,

   }(), S.getLocStart());
 }
 
-void CodeGenFunction::EmitOMPDistributeLoop(const OMPDistributeDirective &S) {
+void CodeGenFunction::EmitOMPDistributeLoop(const OMPLoopDirective &S,
+                                            const CodeGenLoopTy &CodeGenLoop,
+                                            Expr *IncExpr) {
   // Emit the loop iteration variable.
   auto IVExpr = cast<DeclRefExpr>(S.getIterationVariable());
   auto IVDecl = cast<VarDecl>(IVExpr->getDecl());

         EmitIgnoredExpr(S.getEnsureUpperBound());
         // IV = LB;
         EmitIgnoredExpr(S.getInit());
+
+        // for distribute alone,  codegen
         // while (idx <= UB) { BODY; ++idx; }
-        EmitOMPInnerLoop(S, LoopScope.requiresCleanups(), S.getCond(),
-                         S.getInc(),
-                         [&S, LoopExit](CodeGenFunction &CGF) {
-                           CGF.EmitOMPLoopBody(S, LoopExit);
-                           CGF.EmitStopPoint(&S);
+        // when combined with 'for' (e.g. as in 'distribute parallel for')
+        // while (idx <= UB) { <CodeGen rest of pragma>; idx += ST; }
+        EmitOMPInnerLoop(S, LoopScope.requiresCleanups(), S.getCond(), IncExpr,
+                         [&S, LoopExit, &CodeGenLoop](CodeGenFunction &CGF) {
+                           CodeGenLoop(CGF, S, LoopExit);
                          },
                          [](CodeGenFunction &) {});
+
         EmitBlock(LoopExit.getBlock());
         // Tell the runtime we are done.
         RT.emitForStaticFinish(*this, S.getLocStart());
       } else {
         // Emit the outer loop, which requests its work chunk [LB..UB] from
         // runtime and runs the inner loop to process it.
-        EmitOMPDistributeOuterLoop(ScheduleKind, S, LoopScope,
-                            LB.getAddress(), UB.getAddress(), ST.getAddress(),
-                            IL.getAddress(), Chunk);
+        EmitOMPDistributeOuterLoop(ScheduleKind, S, LoopScope, LB.getAddress(),
+                                   UB.getAddress(), ST.getAddress(),
+                                   IL.getAddress(), Chunk, CodeGenLoop);
       }
 
       // Emit final copy of the lastprivate variables if IsLastIter != 0.

 void CodeGenFunction::EmitOMPDistributeDirective(
     const OMPDistributeDirective &S) {
   auto &&CodeGen = [&S](CodeGenFunction &CGF, PrePostActionTy &) {
-    CGF.EmitOMPDistributeLoop(S);
+    auto &&CodeGenLoop = [](CodeGenFunction &CGF, const OMPLoopDirective &S,
+                            JumpDest LoopExit) {
+      CGF.EmitOMPLoopBody(S, LoopExit);
+      CGF.EmitStopPoint(&S);
+    };
+
+    CGF.EmitOMPDistributeLoop(S, CodeGenLoop, S.getInc());
   };
   OMPLexicalScope Scope(*this, S, /*AsInlined=*/true);
   CGM.getOpenMPRuntime().emitInlinedDirective(*this, OMPD_distribute, CodeGen,

     // TODO: Add support for clauses.
     CGF.EmitStmt(CS->getCapturedStmt());
   };
-  emitCommonOMPParallelDirective(CGF, S, OMPD_parallel, CodeGen);
+  emitCommonOMPParallelDirective(CGF, S, OMPD_parallel, CodeGen,
+                                 [](CodeGenFunction &,
+                                    const OMPExecutableDirective &S,
+                                    llvm::SmallVector<llvm::Value *, 16> &) {});
 }
 
 void CodeGenFunction::EmitOMPTargetParallelDeviceFunction(

lib/CodeGen/CodeGenFunction.h

 #include "CGBuilder.h"
 #include "CGDebugInfo.h"
 #include "CGLoopInfo.h"
+#include "CGOpenMPRuntime.h"
 #include "CGValue.h"
 #include "CodeGenModule.h"
 #include "CodeGenPGO.h"

   // because of jumps.
   VarBypassDetector Bypasses;
 
+  // CodeGen lambda for loops and support for ordered clause
+  typedef llvm::function_ref<void(CodeGenFunction &, const OMPLoopDirective &,
+                                  JumpDest)>
+      CodeGenLoopTy;
+  typedef llvm::function_ref<void(CodeGenFunction &, SourceLocation,
+                                  const unsigned, const bool)>
+      CodeGenOrderedTy;
+
+  // Codegen lambda for loop bounds in worksharing loop constructs
+  typedef llvm::function_ref<std::pair<LValue, LValue>(CodeGenFunction &)>
+      CodeGenLoopBoundsTy;
+
+  // Codegen lambda for loop bounds in dispatch-based loop implementation
+  typedef llvm::function_ref<std::pair<llvm::Value *, llvm::Value *>(
+      CodeGenFunction &, Address LB, Address UB)>
+      CodeGenDispatchBoundsTy;
+
   /// \brief CGBuilder insert helper. This function is called after an
   /// instruction is created using Builder.
   void InsertHelper(llvm::Instruction *I, const llvm::Twine &Name,

   void EmitOMPTaskLoopDirective(const OMPTaskLoopDirective &S);
   void EmitOMPTaskLoopSimdDirective(const OMPTaskLoopSimdDirective &S);
   void EmitOMPDistributeDirective(const OMPDistributeDirective &S);
-  void EmitOMPDistributeLoop(const OMPDistributeDirective &S);
+  void EmitOMPDistributeLoop(const OMPLoopDirective &S,
+                             const CodeGenLoopTy &CodeGenLoop, Expr *IncExpr);
   void EmitOMPDistributeParallelForDirective(
       const OMPDistributeParallelForDirective &S);
   void EmitOMPDistributeParallelForSimdDirective(

   /// \brief Emit code for the worksharing loop-based directive.
   /// \return true, if this construct has any lastprivate clause, false -
   /// otherwise.
-  bool EmitOMPWorksharingLoop(const OMPLoopDirective &S);
-  void EmitOMPOuterLoop(bool IsMonotonic, bool DynamicOrOrdered,
-      const OMPLoopDirective &S, OMPPrivateScope &LoopScope, bool Ordered,
-      Address LB, Address UB, Address ST, Address IL, llvm::Value *Chunk);
+  bool EmitOMPWorksharingLoop(const OMPLoopDirective &S, Expr *EUB,
+                              const CodeGenLoopBoundsTy &CodeGenLoopBounds,
+                              const CodeGenDispatchBoundsTy &CGDispatchBounds);
+  void EmitOMPOuterLoop(bool DynamicOrOrdered, bool IsMonotonic,
+                        const OMPLoopDirective &S, OMPPrivateScope &LoopScope,
+                        Address LB, Address UB, Address ST, Address IL,
+                        llvm::Value *Chunk, Expr *EUB, Expr *IncExpr,
+                        const CodeGenLoopTy &CodeGenLoop,
+                        const CodeGenOrderedTy &CodeGenOrdered);
   void EmitOMPForOuterLoop(const OpenMPScheduleTy &ScheduleKind,
                            bool IsMonotonic, const OMPLoopDirective &S,
                            OMPPrivateScope &LoopScope, bool Ordered, Address LB,
                            Address UB, Address ST, Address IL,
-                           llvm::Value *Chunk);
-  void EmitOMPDistributeOuterLoop(
-      OpenMPDistScheduleClauseKind ScheduleKind,
-      const OMPDistributeDirective &S, OMPPrivateScope &LoopScope,
-      Address LB, Address UB, Address ST, Address IL, llvm::Value *Chunk);
+                           llvm::Value *Chunk, Expr *EUB,
+                           const CodeGenDispatchBoundsTy &CGDispatchBounds);
+  void EmitOMPDistributeOuterLoop(OpenMPDistScheduleClauseKind ScheduleKind,
+                                  const OMPLoopDirective &S,
+                                  OMPPrivateScope &LoopScope, Address LB,
+                                  Address UB, Address ST, Address IL,
+                                  llvm::Value *Chunk,
+                                  const CodeGenLoopTy &CodeGenLoopContent);
   /// \brief Emit code for sections directive.
   void EmitSections(const OMPExecutableDirective &S);
 

lib/Sema/SemaOpenMP.cpp

       return 0;
   }
 
+  // Create: increment expression for distribute loop when combined in a same
+  // directive with for as IV = IV + ST; ensure upper bound expression based
+  // on PrevUB instead of NumIterations - used to implement 'for' when found
+  // in combination with 'distribute', like in 'distribute parallel for'
+  SourceLocation DistIncLoc;
+  ExprResult DistCond, DistInc, PrevEUB;
+  if (isOpenMPLoopBoundSharingDirective(DKind)) {
+    DistCond = SemaRef.BuildBinOp(CurScope, CondLoc, BO_LE, IV.get(), UB.get());
+    assert(DistCond.isUsable() && "distribute cond expr was not built");
+
+    DistInc =
+        SemaRef.BuildBinOp(CurScope, DistIncLoc, BO_Add, IV.get(), ST.get());
+    assert(DistInc.isUsable() && "distribute inc expr was not built");
+    DistInc = SemaRef.BuildBinOp(CurScope, DistIncLoc, BO_Assign, IV.get(),
+                                 DistInc.get());
+    DistInc = SemaRef.ActOnFinishFullExpr(DistInc.get());
+    assert(DistInc.isUsable() && "distribute inc expr was not built");
+
+    // Build expression: UB = min(UB, prevUB) for #for in composite or combined
+    // construct
+    SourceLocation DistEUBLoc;
+    ExprResult IsUBGreater =
+        SemaRef.BuildBinOp(CurScope, DistEUBLoc, BO_GT, UB.get(), PrevUB.get());
+    ExprResult CondOp = SemaRef.ActOnConditionalOp(
+        DistEUBLoc, DistEUBLoc, IsUBGreater.get(), PrevUB.get(), UB.get());
+    PrevEUB = SemaRef.BuildBinOp(CurScope, DistIncLoc, BO_Assign, UB.get(),
+                                 CondOp.get());
+    PrevEUB = SemaRef.ActOnFinishFullExpr(PrevEUB.get());
+  }
+
   // Build updates and final values of the loop counters.
   bool HasErrors = false;
   Built.Counters.resize(NestedLoopCount);

   Built.NUB = NextUB.get();
   Built.PrevLB = PrevLB.get();
   Built.PrevUB = PrevUB.get();
+  Built.DistInc = DistInc.get();
+  Built.PrevEUB = PrevEUB.get();
 
   Expr *CounterVal = SemaRef.DefaultLvalueConversion(IV.get()).get();
   // Fill data for doacross depend clauses.

lib/Serialization/ASTReaderStmt.cpp

   if (isOpenMPLoopBoundSharingDirective(D->getDirectiveKind())) {
     D->setPrevLowerBoundVariable(Record.readSubExpr());
     D->setPrevUpperBoundVariable(Record.readSubExpr());
+    D->setDistInc(Record.readSubExpr());
+    D->setPrevEnsureUpperBound(Record.readSubExpr());
   }
   SmallVector<Expr *, 4> Sub;
   unsigned CollapsedNum = D->getCollapsedNumber();

lib/Serialization/ASTWriterStmt.cpp

   if (isOpenMPLoopBoundSharingDirective(D->getDirectiveKind())) {
     Record.AddStmt(D->getPrevLowerBoundVariable());
     Record.AddStmt(D->getPrevUpperBoundVariable());
+    Record.AddStmt(D->getDistInc());
+    Record.AddStmt(D->getPrevEnsureUpperBound());
   }
   for (auto I : D->counters()) {
     Record.AddStmt(I);

test/OpenMP/distribute_parallel_for_codegen.cpp

+// Test host code gen
+// RUN: %clang_cc1 -DLAMBDA -verify -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-64
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-64
+// RUN: %clang_cc1 -DLAMBDA -verify -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-32
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-32
+
+// RUN: %clang_cc1  -verify -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-64
+// RUN: %clang_cc1  -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1  -fopenmp -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-64
+// RUN: %clang_cc1  -verify -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-32
+// RUN: %clang_cc1  -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1  -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-32
+// expected-no-diagnostics
+#ifndef HEADER
+#define HEADER
+
+template <typename T>
+T tmain() {
+  T *a, *b, *c;
+  int n = 10000;
+  int ch = 100;
+
+  // no schedule clauses
+  #pragma omp target
+  #pragma omp teams
+  #pragma omp distribute parallel for
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+  }
+
+  // dist_schedule: static no chunk
+  #pragma omp target
+  #pragma omp teams
+  #pragma omp distribute parallel for dist_schedule(static)
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+  }
+
+  // dist_schedule: static chunk
+  #pragma omp target
+  #pragma omp teams
+  #pragma omp distribute parallel for dist_schedule(static, ch)
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+  }
+
+  // schedule: static no chunk
+  #pragma omp target
+  #pragma omp teams
+  #pragma omp distribute parallel for schedule(static)
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+  }
+
+  // schedule: static chunk
+  #pragma omp target
+  #pragma omp teams
+  #pragma omp distribute parallel for schedule(static, ch)
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+  }
+
+  // schedule: dynamic no chunk
+  #pragma omp target
+  #pragma omp teams
+  #pragma omp distribute parallel for schedule(dynamic)
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+  }
+
+  // schedule: dynamic chunk
+  #pragma omp target
+  #pragma omp teams
+  #pragma omp distribute parallel for schedule(dynamic, ch)
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+  }
+
+  return T();
+}
+
+int main() {
+  double *a, *b, *c;
+  int n = 10000;
+  int ch = 100;
+
+  #ifdef LAMBDA
+  // LAMBDA-LABEL: @main
+  // LAMBDA: call{{.*}} void [[OUTER_LAMBDA:@.+]](
+  [&]() {
+    // LAMBDA: define{{.*}} internal{{.*}} void [[OUTER_LAMBDA]](
+
+    // LAMBDA: call i{{[0-9]+}} @__tgt_target_teams(
+    // LAMBDA: call void [[OFFLOADING_FUN_1:@.+]](
+
+    // LAMBDA: call i{{[0-9]+}} @__tgt_target_teams(
+    // LAMBDA: call void [[OFFLOADING_FUN_2:@.+]](
+
+    // LAMBDA: call i{{[0-9]+}} @__tgt_target_teams(
+    // LAMBDA: call void [[OFFLOADING_FUN_3:@.+]](
+
+    // LAMBDA: call i{{[0-9]+}} @__tgt_target_teams(
+    // LAMBDA: call void [[OFFLOADING_FUN_4:@.+]](
+
+    // LAMBDA: call i{{[0-9]+}} @__tgt_target_teams(
+    // LAMBDA: call void [[OFFLOADING_FUN_5:@.+]](
+
+    // LAMBDA: call i{{[0-9]+}} @__tgt_target_teams(
+    // LAMBDA: call void [[OFFLOADING_FUN_6:@.+]](
+
+    // LAMBDA: call i{{[0-9]+}} @__tgt_target_teams(
+    // LAMBDA: call void [[OFFLOADING_FUN_7:@.+]](
+
+    // no schedule clauses
+    #pragma omp target
+    #pragma omp teams
+    // LAMBDA: define{{.+}} void [[OFFLOADING_FUN_1]](
+    // LAMBDA: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED_1:@.+]] to {{.+}})
+
+    #pragma omp distribute parallel for
+    for (int i = 0; i < n; ++i) {
+      a[i] = b[i] + c[i];
+      // LAMBDA: define{{.+}} void [[OMP_OUTLINED_1]](
+      // LAMBDA-DAG: [[OMP_IV:%.omp.iv]] = alloca
+      // LAMBDA-DAG: [[OMP_LB:%.omp.lb]] = alloca
+      // LAMBDA-DAG: [[OMP_UB:%.omp.ub]] = alloca
+      // LAMBDA-DAG: [[OMP_ST:%.omp.stride]] = alloca
+
+      // LAMBDA: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+
+      // check EUB for distribute
+      // LAMBDA-DAG: [[OMP_UB_VAL_1:%.+]] = load{{.+}} [[OMP_UB]],
+      // LAMBDA: [[NUM_IT_1:%.+]] = load{{.+}},
+      // LAMBDA-DAG: [[CMP_UB_NUM_IT:%.+]] = icmp sgt {{.+}}  [[OMP_UB_VAL_1]], [[NUM_IT_1]]
+      // LAMBDA: br {{.+}} [[CMP_UB_NUM_IT]], label %[[EUB_TRUE:.+]], label %[[EUB_FALSE:.+]]
+      // LAMBDA-DAG: [[EUB_TRUE]]:
+      // LAMBDA: [[NUM_IT_2:%.+]] = load{{.+}},
+      // LAMBDA: br label %[[EUB_END:.+]]
+      // LAMBDA-DAG: [[EUB_FALSE]]:
+      // LAMBDA: [[OMP_UB_VAL2:%.+]] = load{{.+}} [[OMP_UB]],
+      // LAMBDA: br label %[[EUB_END]]
+      // LAMBDA-DAG: [[EUB_END]]:
+      // LAMBDA-DAG: [[EUB_RES:%.+]] = phi{{.+}} [ [[NUM_IT_2]], %[[EUB_TRUE]] ], [ [[OMP_UB_VAL2]], %[[EUB_FALSE]] ]
+      // LAMBDA: store{{.+}} [[EUB_RES]], {{.+}}* [[OMP_UB]],
+
+      // initialize omp.iv
+      // LAMBDA: [[OMP_LB_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[OMP_LB]],
+      // LAMBDA: store {{.+}} [[OMP_LB_VAL_1]], {{.+}}* [[OMP_IV]],
+      // LAMBDA: br label %[[OMP_JUMP_BACK:.+]]
+
+      // check exit condition
+      // LAMBDA: [[OMP_JUMP_BACK]]:
+      // LAMBDA-DAG: [[OMP_IV_VAL_1:%.+]] = load {{.+}} [[OMP_IV]],
+      // LAMBDA-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_UB]],
+      // LAMBDA: [[CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_IV_VAL_1]], [[OMP_UB_VAL_3]]
+      // LAMBDA: br {{.+}} [[CMP_IV_UB]], label %[[DIST_BODY:.+]], label %[[DIST_END:.+]]
+
+      // check that PrevLB and PrevUB are passed to the 'for'
+      // LAMBDA: [[DIST_BODY]]:
+      // LAMBDA-DAG: [[OMP_PREV_LB:%.+]] = load {{.+}}, {{.+}} [[OMP_LB]],
+      // LAMBDA-64-DAG: [[OMP_PREV_LB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_LB]] to {{.+}}
+      // LAMBDA-DAG: [[OMP_PREV_UB:%.+]] = load {{.+}}, {{.+}} [[OMP_UB]],
+      // LAMBDA-64-DAG: [[OMP_PREV_UB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_UB]] to {{.+}}
+      // check that distlb and distub are properly passed to fork_call
+      // LAMBDA-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_1:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_EXT]], i{{[0-9]+}} [[OMP_PREV_UB_EXT]], {{.+}})
+      // LAMBDA-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_1:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB]], i{{[0-9]+}} [[OMP_PREV_UB]], {{.+}})
+      // LAMBDA: br label %[[DIST_INC:.+]]
+
+      // increment by stride (distInc - 'parallel for' executes the whole chunk) and latch
+      // LAMBDA: [[DIST_INC]]:
+      // LAMBDA-DAG: [[OMP_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_IV]],
+      // LAMBDA-DAG: [[OMP_ST_VAL_1:%.+]] = load {{.+}}, {{.+}}* [[OMP_ST]],
+      // LAMBDA: [[OMP_IV_INC:%.+]] = add{{.+}} [[OMP_IV_VAL_2]], [[OMP_ST_VAL_1]]
+      // LAMBDA: store{{.+}} [[OMP_IV_INC]], {{.+}}* [[OMP_IV]],
+      // LAMBDA: br label %[[OMP_JUMP_BACK]]
+
+      // LAMBDA-DAG: call void @__kmpc_for_static_fini(
+      // LAMBDA: ret
+
+      // implementation of 'parallel for'
+      // LAMBDA: define{{.+}} void [[OMP_PARFOR_OUTLINED_1]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}})
+
+      // LAMBDA-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+
+      // initialize lb and ub to PrevLB and PrevUB
+      // LAMBDA-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+      // LAMBDA-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+      // LAMBDA-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+      // LAMBDA-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+      // LAMBDA-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+      // LAMBDA-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+      // LAMBDA-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 34, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]],{{.+}})
+
+      // PrevEUB is only used when 'for' has a chunked schedule, otherwise EUB is used
+      // In this case we use EUB
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL_1:%.+]] = load{{.+}} [[OMP_PF_UB]],
+      // LAMBDA: [[PF_NUM_IT_1:%.+]] = load{{.+}},
+      // LAMBDA-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_1]], [[PF_NUM_IT_1]]
+      // LAMBDA: br i1 [[PF_CMP_UB_NUM_IT]], label %[[PF_EUB_TRUE:.+]], label %[[PF_EUB_FALSE:.+]]
+      // LAMBDA: [[PF_EUB_TRUE]]:
+      // LAMBDA: [[PF_NUM_IT_2:%.+]] = load{{.+}},
+      // LAMBDA: br label %[[PF_EUB_END:.+]]
+      // LAMBDA-DAG: [[PF_EUB_FALSE]]:
+      // LAMBDA: [[OMP_PF_UB_VAL2:%.+]] = load{{.+}} [[OMP_PF_UB]],
+      // LAMBDA: br label %[[PF_EUB_END]]
+      // LAMBDA-DAG: [[PF_EUB_END]]:
+      // LAMBDA-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_NUM_IT_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL2]], %[[PF_EUB_FALSE]] ]
+      // LAMBDA: store{{.+}} [[PF_EUB_RES]],{{.+}}  [[OMP_PF_UB]],
+
+      // initialize omp.iv
+      // LAMBDA: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+      // LAMBDA: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: br label %[[OMP_PF_JUMP_BACK:.+]]
+
+      // check exit condition
+      // LAMBDA: [[OMP_PF_JUMP_BACK]]:
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_1:%.+]] = load {{.+}} [[OMP_PF_IV]],
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL_3:%.+]] = load {{.+}} [[OMP_PF_UB]],
+      // LAMBDA: [[PF_CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_PF_IV_VAL_1]], [[OMP_PF_UB_VAL_3]]
+      // LAMBDA: br {{.+}} [[PF_CMP_IV_UB]], label %[[PF_BODY:.+]], label %[[PF_END:.+]]
+
+      // check that PrevLB and PrevUB are passed to the 'for'
+      // LAMBDA: [[PF_BODY]]:
+      // LAMBDA-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: br label {{.+}}
+
+      // check stride 1 for 'for' in 'distribute parallel for'
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: [[OMP_PF_IV_INC:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_2]], 1
+      // LAMBDA: store{{.+}} [[OMP_PF_IV_INC]], {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: br label %[[OMP_PF_JUMP_BACK]]
+
+      // LAMBDA-DAG: call void @__kmpc_for_static_fini(
+      // LAMBDA: ret
+
+      [&]() {
+  a[i] = b[i] + c[i];
+      }();
+    }
+
+    // dist_schedule: static no chunk (same sa default - no dist_schedule)
+    #pragma omp target
+    #pragma omp teams
+    // LAMBDA: define{{.+}} void [[OFFLOADING_FUN_2]](
+    // LAMBDA: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED_2:@.+]] to {{.+}})
+
+    #pragma omp distribute parallel for dist_schedule(static)
+    for (int i = 0; i < n; ++i) {
+      a[i] = b[i] + c[i];
+      // LAMBDA: define{{.+}} void [[OMP_OUTLINED_2]](
+      // LAMBDA-DAG: [[OMP_IV:%.omp.iv]] = alloca
+      // LAMBDA-DAG: [[OMP_LB:%.omp.lb]] = alloca
+      // LAMBDA-DAG: [[OMP_UB:%.omp.ub]] = alloca
+      // LAMBDA-DAG: [[OMP_ST:%.omp.stride]] = alloca
+
+      // LAMBDA: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+
+      // check EUB for distribute
+      // LAMBDA-DAG: [[OMP_UB_VAL_1:%.+]] = load{{.+}} [[OMP_UB]],
+      // LAMBDA: [[NUM_IT_1:%.+]] = load{{.+}},
+      // LAMBDA-DAG: [[CMP_UB_NUM_IT:%.+]] = icmp sgt {{.+}}  [[OMP_UB_VAL_1]], [[NUM_IT_1]]
+      // LAMBDA: br {{.+}} [[CMP_UB_NUM_IT]], label %[[EUB_TRUE:.+]], label %[[EUB_FALSE:.+]]
+      // LAMBDA-DAG: [[EUB_TRUE]]:
+      // LAMBDA: [[NUM_IT_2:%.+]] = load{{.+}},
+      // LAMBDA: br label %[[EUB_END:.+]]
+      // LAMBDA-DAG: [[EUB_FALSE]]:
+      // LAMBDA: [[OMP_UB_VAL2:%.+]] = load{{.+}} [[OMP_UB]],
+      // LAMBDA: br label %[[EUB_END]]
+      // LAMBDA-DAG: [[EUB_END]]:
+      // LAMBDA-DAG: [[EUB_RES:%.+]] = phi{{.+}} [ [[NUM_IT_2]], %[[EUB_TRUE]] ], [ [[OMP_UB_VAL2]], %[[EUB_FALSE]] ]
+      // LAMBDA: store{{.+}} [[EUB_RES]], {{.+}}* [[OMP_UB]],
+
+      // initialize omp.iv
+      // LAMBDA: [[OMP_LB_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[OMP_LB]],
+      // LAMBDA: store {{.+}} [[OMP_LB_VAL_1]], {{.+}}* [[OMP_IV]],
+      // LAMBDA: br label %[[OMP_JUMP_BACK:.+]]
+
+      // check exit condition
+      // LAMBDA: [[OMP_JUMP_BACK]]:
+      // LAMBDA-DAG: [[OMP_IV_VAL_1:%.+]] = load {{.+}} [[OMP_IV]],
+      // LAMBDA-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_UB]],
+      // LAMBDA: [[CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_IV_VAL_1]], [[OMP_UB_VAL_3]]
+      // LAMBDA: br {{.+}} [[CMP_IV_UB]], label %[[DIST_BODY:.+]], label %[[DIST_END:.+]]
+
+      // check that PrevLB and PrevUB are passed to the 'for'
+      // LAMBDA: [[DIST_BODY]]:
+      // LAMBDA-DAG: [[OMP_PREV_LB:%.+]] = load {{.+}}, {{.+}} [[OMP_LB]],
+      // LAMBDA-64-DAG: [[OMP_PREV_LB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_LB]] to {{.+}}
+      // LAMBDA-DAG: [[OMP_PREV_UB:%.+]] = load {{.+}}, {{.+}} [[OMP_UB]],
+      // LAMBDA-64-DAG: [[OMP_PREV_UB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_UB]] to {{.+}}
+      // check that distlb and distub are properly passed to fork_call
+      // LAMBDA-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_2:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_EXT]], i{{[0-9]+}} [[OMP_PREV_UB_EXT]], {{.+}})
+      // LAMBDA-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_2:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB]], i{{[0-9]+}} [[OMP_PREV_UB]], {{.+}})
+      // LAMBDA: br label %[[DIST_INC:.+]]
+
+      // increment by stride (distInc - 'parallel for' executes the whole chunk) and latch
+      // LAMBDA: [[DIST_INC]]:
+      // LAMBDA-DAG: [[OMP_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_IV]],
+      // LAMBDA-DAG: [[OMP_ST_VAL_1:%.+]] = load {{.+}}, {{.+}}* [[OMP_ST]],
+      // LAMBDA: [[OMP_IV_INC:%.+]] = add{{.+}} [[OMP_IV_VAL_2]], [[OMP_ST_VAL_1]]
+      // LAMBDA: store{{.+}} [[OMP_IV_INC]], {{.+}}* [[OMP_IV]],
+      // LAMBDA: br label %[[OMP_JUMP_BACK]]
+
+      // LAMBDA-DAG: call void @__kmpc_for_static_fini(
+      // LAMBDA: ret
+
+      // implementation of 'parallel for'
+      // LAMBDA: define{{.+}} void [[OMP_PARFOR_OUTLINED_2]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}})
+
+      // LAMBDA-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+
+      // initialize lb and ub to PrevLB and PrevUB
+      // LAMBDA-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+      // LAMBDA-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+      // LAMBDA-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+      // LAMBDA-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+      // LAMBDA-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+      // LAMBDA-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+      // LAMBDA-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 34, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]],{{.+}})
+
+      // PrevEUB is only used when 'for' has a chunked schedule, otherwise EUB is used
+      // In this case we use EUB
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL_1:%.+]] = load{{.+}} [[OMP_PF_UB]],
+      // LAMBDA: [[PF_NUM_IT_1:%.+]] = load{{.+}},
+      // LAMBDA-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_1]], [[PF_NUM_IT_1]]
+      // LAMBDA: br i1 [[PF_CMP_UB_NUM_IT]], label %[[PF_EUB_TRUE:.+]], label %[[PF_EUB_FALSE:.+]]
+      // LAMBDA: [[PF_EUB_TRUE]]:
+      // LAMBDA: [[PF_NUM_IT_2:%.+]] = load{{.+}},
+      // LAMBDA: br label %[[PF_EUB_END:.+]]
+      // LAMBDA-DAG: [[PF_EUB_FALSE]]:
+      // LAMBDA: [[OMP_PF_UB_VAL2:%.+]] = load{{.+}} [[OMP_PF_UB]],
+      // LAMBDA: br label %[[PF_EUB_END]]
+      // LAMBDA-DAG: [[PF_EUB_END]]:
+      // LAMBDA-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_NUM_IT_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL2]], %[[PF_EUB_FALSE]] ]
+      // LAMBDA: store{{.+}} [[PF_EUB_RES]],{{.+}}  [[OMP_PF_UB]],
+
+      // initialize omp.iv
+      // LAMBDA: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+      // LAMBDA: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: br label %[[OMP_PF_JUMP_BACK:.+]]
+
+      // check exit condition
+      // LAMBDA: [[OMP_PF_JUMP_BACK]]:
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_1:%.+]] = load {{.+}} [[OMP_PF_IV]],
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL_3:%.+]] = load {{.+}} [[OMP_PF_UB]],
+      // LAMBDA: [[PF_CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_PF_IV_VAL_1]], [[OMP_PF_UB_VAL_3]]
+      // LAMBDA: br {{.+}} [[PF_CMP_IV_UB]], label %[[PF_BODY:.+]], label %[[PF_END:.+]]
+
+      // check that PrevLB and PrevUB are passed to the 'for'
+      // LAMBDA: [[PF_BODY]]:
+      // LAMBDA-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: br label {{.+}}
+
+      // check stride 1 for 'for' in 'distribute parallel for'
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: [[OMP_PF_IV_INC:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_2]], 1
+      // LAMBDA: store{{.+}} [[OMP_PF_IV_INC]], {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: br label %[[OMP_PF_JUMP_BACK]]
+
+      // LAMBDA-DAG: call void @__kmpc_for_static_fini(
+      // LAMBDA: ret
+      [&]() {
+  a[i] = b[i] + c[i];
+      }();
+    }
+
+    // dist_schedule: static chunk
+    #pragma omp target
+    #pragma omp teams
+    // LAMBDA: define{{.+}} void [[OFFLOADING_FUN_3]](
+    // LAMBDA: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 5, {{.+}}* [[OMP_OUTLINED_3:@.+]] to {{.+}})
+
+    #pragma omp distribute parallel for dist_schedule(static, ch)
+    for (int i = 0; i < n; ++i) {
+      a[i] = b[i] + c[i];
+      // LAMBDA: define{{.+}} void [[OMP_OUTLINED_3]](
+      // LAMBDA-DAG: [[OMP_IV:%.omp.iv]] = alloca
+      // LAMBDA-DAG: [[OMP_LB:%.omp.lb]] = alloca
+      // LAMBDA-DAG: [[OMP_UB:%.omp.ub]] = alloca
+      // LAMBDA-DAG: [[OMP_ST:%.omp.stride]] = alloca
+
+      // unlike the previous tests, in this one we have a outer and inner loop for 'distribute'
+      // LAMBDA: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 91,
+      // LAMBDA: br label %[[DIST_OUTER_LOOP_HEADER:.+]]
+
+      // LAMBDA: [[DIST_OUTER_LOOP_HEADER]]:
+      // check EUB for distribute
+      // LAMBDA-DAG: [[OMP_UB_VAL_1:%.+]] = load{{.+}} [[OMP_UB]],
+      // LAMBDA: [[NUM_IT_1:%.+]] = load{{.+}},
+      // LAMBDA-DAG: [[CMP_UB_NUM_IT:%.+]] = icmp sgt {{.+}}  [[OMP_UB_VAL_1]], [[NUM_IT_1]]
+      // LAMBDA: br {{.+}} [[CMP_UB_NUM_IT]], label %[[EUB_TRUE:.+]], label %[[EUB_FALSE:.+]]
+      // LAMBDA-DAG: [[EUB_TRUE]]:
+      // LAMBDA: [[NUM_IT_2:%.+]] = load{{.+}},
+      // LAMBDA: br label %[[EUB_END:.+]]
+      // LAMBDA-DAG: [[EUB_FALSE]]:
+      // LAMBDA: [[OMP_UB_VAL2:%.+]] = load{{.+}} [[OMP_UB]],
+      // LAMBDA: br label %[[EUB_END]]
+      // LAMBDA-DAG: [[EUB_END]]:
+      // LAMBDA-DAG: [[EUB_RES:%.+]] = phi{{.+}} [ [[NUM_IT_2]], %[[EUB_TRUE]] ], [ [[OMP_UB_VAL2]], %[[EUB_FALSE]] ]
+      // LAMBDA: store{{.+}} [[EUB_RES]], {{.+}}* [[OMP_UB]],
+
+      // initialize omp.iv
+      // LAMBDA: [[OMP_LB_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[OMP_LB]],
+      // LAMBDA: store {{.+}} [[OMP_LB_VAL_1]], {{.+}}* [[OMP_IV]],
+
+      // check exit condition
+      // LAMBDA-DAG: [[OMP_IV_VAL_1:%.+]] = load {{.+}} [[OMP_IV]],
+      // LAMBDA-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_UB]],
+      // LAMBDA: [[CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_IV_VAL_1]], [[OMP_UB_VAL_3]]
+      // LAMBDA: br {{.+}} [[CMP_IV_UB]], label %[[DIST_OUTER_LOOP_BODY:.+]], label %[[DIST_OUTER_LOOP_END:.+]]
+
+      // LAMBDA: [[DIST_OUTER_LOOP_BODY]]:
+      // LAMBDA: br label %[[DIST_INNER_LOOP_HEADER:.+]]
+
+      // LAMBDA: [[DIST_INNER_LOOP_HEADER]]:
+      // LAMBDA-DAG: [[OMP_IV_VAL_2:%.+]] = load {{.+}} [[OMP_IV]],
+      // LAMBDA-DAG: [[OMP_UB_VAL_4:%.+]] = load {{.+}} [[OMP_UB]],
+      // LAMBDA: [[CMP_IV_UB_2:%.+]] = icmp sle {{.+}} [[OMP_IV_VAL_2]], [[OMP_UB_VAL_4]]
+      // LAMBDA: br{{.+}} [[CMP_IV_UB_2]], label %[[DIST_INNER_LOOP_BODY:.+]], label %[[DIST_INNER_LOOP_END:.+]]
+
+      // check that PrevLB and PrevUB are passed to the 'for'
+      // LAMBDA: [[DIST_INNER_LOOP_BODY]]:
+      // LAMBDA-DAG: [[OMP_PREV_LB:%.+]] = load {{.+}}, {{.+}} [[OMP_LB]],
+      // LAMBDA-64-DAG: [[OMP_PREV_LB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_LB]] to {{.+}}
+      // LAMBDA-DAG: [[OMP_PREV_UB:%.+]] = load {{.+}}, {{.+}} [[OMP_UB]],
+      // LAMBDA-64-DAG: [[OMP_PREV_UB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_UB]] to {{.+}}
+      // check that distlb and distub are properly passed to fork_call
+      // LAMBDA-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_3:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_EXT]], i{{[0-9]+}} [[OMP_PREV_UB_EXT]], {{.+}})
+      // LAMBDA-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_3:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB]], i{{[0-9]+}} [[OMP_PREV_UB]], {{.+}})
+      // LAMBDA: br label %[[DIST_INNER_LOOP_INC:.+]]
+
+      // check DistInc
+      // LAMBDA: [[DIST_INNER_LOOP_INC]]:
+      // LAMBDA-DAG: [[OMP_IV_VAL_3:%.+]] = load {{.+}}, {{.+}}* [[OMP_IV]],
+      // LAMBDA-DAG: [[OMP_ST_VAL_1:%.+]] = load {{.+}}, {{.+}}* [[OMP_ST]],
+      // LAMBDA: [[OMP_IV_INC:%.+]] = add{{.+}} [[OMP_IV_VAL_3]], [[OMP_ST_VAL_1]]
+      // LAMBDA: store{{.+}} [[OMP_IV_INC]], {{.+}}* [[OMP_IV]],
+      // LAMBDA: br label %[[DIST_INNER_LOOP_HEADER]]
+
+      // LAMBDA: [[DIST_INNER_LOOP_END]]:
+      // LAMBDA: br label %[[DIST_OUTER_LOOP_INC:.+]]
+
+      // LAMBDA: [[DIST_OUTER_LOOP_INC]]:
+      // check NextLB and NextUB
+      // LAMBDA-DAG: [[OMP_LB_VAL_2:%.+]] = load{{.+}}, {{.+}} [[OMP_LB]],
+      // LAMBDA-DAG: [[OMP_ST_VAL_2:%.+]] = load{{.+}}, {{.+}} [[OMP_ST]],
+      // LAMBDA-DAG: [[OMP_LB_NEXT:%.+]] = add{{.+}} [[OMP_LB_VAL_2]], [[OMP_ST_VAL_2]]
+      // LAMBDA: store{{.+}} [[OMP_LB_NEXT]], {{.+}}* [[OMP_LB]],
+      // LAMBDA-DAG: [[OMP_UB_VAL_5:%.+]] = load{{.+}}, {{.+}} [[OMP_UB]],
+      // LAMBDA-DAG: [[OMP_ST_VAL_3:%.+]] = load{{.+}}, {{.+}} [[OMP_ST]],
+      // LAMBDA-DAG: [[OMP_UB_NEXT:%.+]] = add{{.+}} [[OMP_UB_VAL_5]], [[OMP_ST_VAL_3]]
+      // LAMBDA: store{{.+}} [[OMP_UB_NEXT]], {{.+}}* [[OMP_UB]],
+      // LAMBDA: br label %[[DIST_OUTER_LOOP_HEADER]]
+
+      // outer loop exit
+      // LAMBDA: [[DIST_OUTER_LOOP_END]]:
+      // LAMBDA-DAG: call void @__kmpc_for_static_fini(
+      // LAMBDA: ret
+
+      // skip implementation of 'parallel for': using default scheduling and was tested above
+      [&]() {
+  a[i] = b[i] + c[i];
+      }();
+    }
+
+    // schedule: static no chunk
+    #pragma omp target
+    #pragma omp teams
+    // LAMBDA: define{{.+}} void [[OFFLOADING_FUN_4]](
+    // LAMBDA: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED_4:@.+]] to {{.+}})
+
+    #pragma omp distribute parallel for schedule(static)
+    for (int i = 0; i < n; ++i) {
+      a[i] = b[i] + c[i];
+      // LAMBDA: define{{.+}} void [[OMP_OUTLINED_4]](
+      // LAMBDA-DAG: [[OMP_IV:%.omp.iv]] = alloca
+      // LAMBDA-DAG: [[OMP_LB:%.omp.lb]] = alloca
+      // LAMBDA-DAG: [[OMP_UB:%.omp.ub]] = alloca
+      // LAMBDA-DAG: [[OMP_ST:%.omp.stride]] = alloca
+
+      // LAMBDA: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+      // LAMBDA: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_4:@.+]] to {{.+}},
+      // skip rest of implementation of 'distribute' as it is tested above for default dist_schedule case
+      // LAMBDA: ret
+
+      // 'parallel for' implementation is the same as the case without schedule clase (static no chunk is the default)
+      // LAMBDA: define{{.+}} void [[OMP_PARFOR_OUTLINED_4]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}})
+
+      // LAMBDA-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+
+      // initialize lb and ub to PrevLB and PrevUB
+      // LAMBDA-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+      // LAMBDA-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+      // LAMBDA-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+      // LAMBDA-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+      // LAMBDA-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+      // LAMBDA-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+      // LAMBDA-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 34, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]],{{.+}})
+
+      // PrevEUB is only used when 'for' has a chunked schedule, otherwise EUB is used
+      // In this case we use EUB
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL_1:%.+]] = load{{.+}} [[OMP_PF_UB]],
+      // LAMBDA: [[PF_NUM_IT_1:%.+]] = load{{.+}},
+      // LAMBDA-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_1]], [[PF_NUM_IT_1]]
+      // LAMBDA: br i1 [[PF_CMP_UB_NUM_IT]], label %[[PF_EUB_TRUE:.+]], label %[[PF_EUB_FALSE:.+]]
+      // LAMBDA: [[PF_EUB_TRUE]]:
+      // LAMBDA: [[PF_NUM_IT_2:%.+]] = load{{.+}},
+      // LAMBDA: br label %[[PF_EUB_END:.+]]
+      // LAMBDA-DAG: [[PF_EUB_FALSE]]:
+      // LAMBDA: [[OMP_PF_UB_VAL2:%.+]] = load{{.+}} [[OMP_PF_UB]],
+      // LAMBDA: br label %[[PF_EUB_END]]
+      // LAMBDA-DAG: [[PF_EUB_END]]:
+      // LAMBDA-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_NUM_IT_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL2]], %[[PF_EUB_FALSE]] ]
+      // LAMBDA: store{{.+}} [[PF_EUB_RES]],{{.+}}  [[OMP_PF_UB]],
+
+      // initialize omp.iv
+      // LAMBDA: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+      // LAMBDA: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: br label %[[OMP_PF_JUMP_BACK:.+]]
+
+      // check exit condition
+      // LAMBDA: [[OMP_PF_JUMP_BACK]]:
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_1:%.+]] = load {{.+}} [[OMP_PF_IV]],
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL_3:%.+]] = load {{.+}} [[OMP_PF_UB]],
+      // LAMBDA: [[PF_CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_PF_IV_VAL_1]], [[OMP_PF_UB_VAL_3]]
+      // LAMBDA: br {{.+}} [[PF_CMP_IV_UB]], label %[[PF_BODY:.+]], label %[[PF_END:.+]]
+
+      // check that PrevLB and PrevUB are passed to the 'for'
+      // LAMBDA: [[PF_BODY]]:
+      // LAMBDA-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: br label {{.+}}
+
+      // check stride 1 for 'for' in 'distribute parallel for'
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: [[OMP_PF_IV_INC:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_2]], 1
+      // LAMBDA: store{{.+}} [[OMP_PF_IV_INC]], {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: br label %[[OMP_PF_JUMP_BACK]]
+
+      // LAMBDA-DAG: call void @__kmpc_for_static_fini(
+      // LAMBDA: ret
+
+      [&]() {
+  a[i] = b[i] + c[i];
+      }();
+    }
+
+    // schedule: static chunk
+    #pragma omp target
+    #pragma omp teams
+    // LAMBDA: define{{.+}} void [[OFFLOADING_FUN_5]](
+    // LAMBDA: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 5, {{.+}}* [[OMP_OUTLINED_5:@.+]] to {{.+}})
+
+    #pragma omp distribute parallel for schedule(static, ch)
+    for (int i = 0; i < n; ++i) {
+      a[i] = b[i] + c[i];
+      // LAMBDA: define{{.+}} void [[OMP_OUTLINED_5]](
+      // LAMBDA: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+      // LAMBDA: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_5:@.+]] to {{.+}},
+      // skip rest of implementation of 'distribute' as it is tested above for default dist_schedule case
+      // LAMBDA: ret
+
+      // 'parallel for' implementation using outer and inner loops and PrevEUB
+      // LAMBDA: define{{.+}} void [[OMP_PARFOR_OUTLINED_5]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}}, {{.+}})
+      // LAMBDA-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_ST:%.omp.stride]] = alloca{{.+}},
+
+      // initialize lb and ub to PrevLB and PrevUB
+      // LAMBDA-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+      // LAMBDA-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+      // LAMBDA-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+      // LAMBDA-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+      // LAMBDA-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+      // LAMBDA-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+      // LAMBDA-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 33, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]],{{.+}})
+      // LAMBDA: br label %[[OMP_PF_OUTER_LOOP_HEADER:.+]]
+
+      // check PrevEUB (using PrevUB instead of NumIt as upper bound)
+      // LAMBDA: [[OMP_PF_OUTER_LOOP_HEADER]]:
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL_1:%.+]] = load{{.+}} [[OMP_PF_UB]],
+      // LAMBDA-64-DAG: [[OMP_PF_UB_VAL_CONV:%.+]] = sext{{.+}} [[OMP_PF_UB_VAL_1]] to
+      // LAMBDA: [[PF_PREV_UB_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+      // LAMBDA-64-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_CONV]], [[PF_PREV_UB_VAL_1]]
+      // LAMBDA-32-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_1]], [[PF_PREV_UB_VAL_1]]
+      // LAMBDA: br i1 [[PF_CMP_UB_NUM_IT]], label %[[PF_EUB_TRUE:.+]], label %[[PF_EUB_FALSE:.+]]
+      // LAMBDA: [[PF_EUB_TRUE]]:
+      // LAMBDA: [[PF_PREV_UB_VAL_2:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+      // LAMBDA: br label %[[PF_EUB_END:.+]]
+      // LAMBDA-DAG: [[PF_EUB_FALSE]]:
+      // LAMBDA: [[OMP_PF_UB_VAL_2:%.+]] = load{{.+}} [[OMP_PF_UB]],
+      // LAMBDA-64: [[OMP_PF_UB_VAL_2_CONV:%.+]] = sext{{.+}} [[OMP_PF_UB_VAL_2]] to
+      // LAMBDA: br label %[[PF_EUB_END]]
+      // LAMBDA-DAG: [[PF_EUB_END]]:
+      // LAMBDA-64-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_PREV_UB_VAL_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL_2_CONV]], %[[PF_EUB_FALSE]] ]
+      // LAMBDA-32-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_PREV_UB_VAL_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL_2]], %[[PF_EUB_FALSE]] ]
+      // LAMBDA-64-DAG: [[PF_EUB_RES_CONV:%.+]] = trunc{{.+}} [[PF_EUB_RES]] to
+      // LAMBDA-64: store{{.+}} [[PF_EUB_RES_CONV]],{{.+}}  [[OMP_PF_UB]],
+      // LAMBDA-32: store{{.+}} [[PF_EUB_RES]], {{.+}} [[OMP_PF_UB]],
+
+      // initialize omp.iv (IV = LB)
+      // LAMBDA: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+      // LAMBDA: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+
+      // outer loop: while (IV < UB) {
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL_3:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA: [[PF_CMP_IV_UB_1:%.+]] = icmp{{.+}} [[OMP_PF_IV_VAL_1]], [[OMP_PF_UB_VAL_3]]
+      // LAMBDA: br{{.+}} [[PF_CMP_IV_UB_1]], label %[[OMP_PF_OUTER_LOOP_BODY:.+]], label %[[OMP_PF_OUTER_LOOP_END:.+]]
+
+      // LAMBDA: [[OMP_PF_OUTER_LOOP_BODY]]:
+      // LAMBDA: br label %[[OMP_PF_INNER_FOR_HEADER:.+]]
+
+      // LAMBDA: [[OMP_PF_INNER_FOR_HEADER]]:
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL_4:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA: [[PF_CMP_IV_UB_2:%.+]] = icmp{{.+}} [[OMP_PF_IV_VAL_2]], [[OMP_PF_UB_VAL_4]]
+      // LAMBDA: br{{.+}} [[PF_CMP_IV_UB_2]], label %[[OMP_PF_INNER_LOOP_BODY:.+]], label %[[OMP_PF_INNER_LOOP_END:.+]]
+
+      // LAMBDA: [[OMP_PF_INNER_LOOP_BODY]]:
+      // LAMBDA-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // skip body branch
+      // LAMBDA: br{{.+}}
+      // LAMBDA: br label %[[OMP_PF_INNER_LOOP_INC:.+]]
+
+      // IV = IV + 1 and inner loop latch
+      // LAMBDA: [[OMP_PF_INNER_LOOP_INC]]:
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_3:%.+]] = load{{.+}}, {{.+}}* [[OMP_IV]],
+      // LAMBDA-DAG: [[OMP_PF_NEXT_IV:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_3]], 1
+      // LAMBDA-DAG: store{{.+}} [[OMP_PF_NEXT_IV]], {{.+}}* [[OMP_IV]],
+      // LAMBDA: br label %[[OMP_PF_INNER_FOR_HEADER]]
+
+      // check NextLB and NextUB
+      // LAMBDA: [[OMP_PF_INNER_LOOP_END]]:
+      // LAMBDA: br label %[[OMP_PF_OUTER_LOOP_INC:.+]]
+
+      // LAMBDA: [[OMP_PF_OUTER_LOOP_INC]]:
+      // LAMBDA-DAG: [[OMP_PF_LB_VAL_2:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+      // LAMBDA-DAG: [[OMP_PF_ST_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_ST]],
+      // LAMBDA-DAG: [[OMP_PF_LB_NEXT:%.+]] = add{{.+}} [[OMP_PF_LB_VAL_2]], [[OMP_PF_ST_VAL_1]]
+      // LAMBDA: store{{.+}} [[OMP_PF_LB_NEXT]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL_5:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_UB]],
+      // LAMBDA-DAG: [[OMP_PF_ST_VAL_2:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_ST]],
+      // LAMBDA-DAG: [[OMP_PF_UB_NEXT:%.+]] = add{{.+}} [[OMP_PF_UB_VAL_5]], [[OMP_PF_ST_VAL_2]]
+      // LAMBDA: store{{.+}} [[OMP_PF_UB_NEXT]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA: br label %[[OMP_PF_OUTER_LOOP_HEADER]]
+
+      // LAMBDA: [[OMP_PF_OUTER_LOOP_END]]:
+      // LAMBDA-DAG: call void @__kmpc_for_static_fini(
+      // LAMBDA: ret
+      [&]() {
+  a[i] = b[i] + c[i];
+      }();
+    }
+
+    // schedule: dynamic no chunk
+    #pragma omp target
+    #pragma omp teams
+    // LAMBDA: define{{.+}} void [[OFFLOADING_FUN_6]](
+    // LAMBDA: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED_6:@.+]] to {{.+}})
+
+    #pragma omp distribute parallel for schedule(dynamic)
+    for (int i = 0; i < n; ++i) {
+      a[i] = b[i] + c[i];
+      // LAMBDA: define{{.+}} void [[OMP_OUTLINED_6]](
+      // LAMBDA: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+      // LAMBDA: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_6:@.+]] to {{.+}},
+      // skip rest of implementation of 'distribute' as it is tested above for default dist_schedule case
+      // LAMBDA: ret
+
+      // 'parallel for' implementation using outer and inner loops and PrevEUB
+      // LAMBDA: define{{.+}} void [[OMP_PARFOR_OUTLINED_6]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}})
+      // LAMBDA-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_ST:%.omp.stride]] = alloca{{.+}},
+
+      // initialize lb and ub to PrevLB and PrevUB
+      // LAMBDA-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+      // LAMBDA-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+      // LAMBDA-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+      // LAMBDA-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+      // LAMBDA-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+      // LAMBDA-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+      // LAMBDA-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA-DAG: [[OMP_PF_LB_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_UB]],
+      // LAMBDA: call void @__kmpc_dispatch_init_4({{.+}}, {{.+}}, {{.+}} 35, {{.+}} [[OMP_PF_LB_VAL]], {{.+}} [[OMP_PF_UB_VAL]], {{.+}}, {{.+}})
+      // LAMBDA: br label %[[OMP_PF_OUTER_LOOP_HEADER:.+]]
+
+      // LAMBDA: [[OMP_PF_OUTER_LOOP_HEADER]]:
+      // LAMBDA: [[IS_FIN:%.+]] = call{{.+}} @__kmpc_dispatch_next_4({{.+}}, {{.+}}, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]], {{.+}}* [[OMP_PF_ST]])
+      // LAMBDA: [[IS_FIN_CMP:%.+]] = icmp{{.+}} [[IS_FIN]], 0
+      // LAMBDA: br{{.+}} [[IS_FIN_CMP]], label %[[OMP_PF_OUTER_LOOP_BODY:.+]], label %[[OMP_PF_OUTER_LOOP_END:.+]]
+
+      // initialize omp.iv (IV = LB)
+      // LAMBDA: [[OMP_PF_OUTER_LOOP_BODY]]:
+      // LAMBDA-DAG: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+      // LAMBDA-DAG: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: br label %[[OMP_PF_INNER_LOOP_HEADER:.+]]
+
+      // LAMBDA: [[OMP_PF_INNER_LOOP_HEADER]]:
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL_4:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA: [[PF_CMP_IV_UB_2:%.+]] = icmp{{.+}} [[OMP_PF_IV_VAL_2]], [[OMP_PF_UB_VAL_4]]
+      // LAMBDA: br{{.+}} [[PF_CMP_IV_UB_2]], label %[[OMP_PF_INNER_LOOP_BODY:.+]], label %[[OMP_PF_INNER_LOOP_END:.+]]
+
+      // LAMBDA: [[OMP_PF_INNER_LOOP_BODY]]:
+      // LAMBDA-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // skip body branch
+      // LAMBDA: br{{.+}}
+      // LAMBDA: br label %[[OMP_PF_INNER_LOOP_INC:.+]]
+
+      // IV = IV + 1 and inner loop latch
+      // LAMBDA: [[OMP_PF_INNER_LOOP_INC]]:
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_3:%.+]] = load{{.+}}, {{.+}}* [[OMP_IV]],
+      // LAMBDA-DAG: [[OMP_PF_NEXT_IV:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_3]], 1
+      // LAMBDA-DAG: store{{.+}} [[OMP_PF_NEXT_IV]], {{.+}}* [[OMP_IV]],
+      // LAMBDA: br label %[[OMP_PF_INNER_FOR_HEADER]]
+
+      // check NextLB and NextUB
+      // LAMBDA: [[OMP_PF_INNER_LOOP_END]]:
+      // LAMBDA: br label %[[OMP_PF_OUTER_LOOP_INC:.+]]
+
+      // LAMBDA: [[OMP_PF_OUTER_LOOP_INC]]:
+      // LAMBDA: br label %[[OMP_PF_OUTER_LOOP_HEADER]]
+
+      // LAMBDA: [[OMP_PF_OUTER_LOOP_END]]:
+      // LAMBDA: ret
+      [&]() {
+  a[i] = b[i] + c[i];
+      }();
+    }
+
+    // schedule: dynamic chunk
+    #pragma omp target
+    #pragma omp teams
+    // LAMBDA: define{{.+}} void [[OFFLOADING_FUN_7]](
+    // LAMBDA: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 5, {{.+}}* [[OMP_OUTLINED_7:@.+]] to {{.+}})
+
+    #pragma omp distribute parallel for schedule(dynamic, ch)
+    for (int i = 0; i < n; ++i) {
+      a[i] = b[i] + c[i];
+      // LAMBDA: define{{.+}} void [[OMP_OUTLINED_7]](
+      // LAMBDA: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+      // LAMBDA: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_7:@.+]] to {{.+}},
+      // skip rest of implementation of 'distribute' as it is tested above for default dist_schedule case
+      // LAMBDA: ret
+
+      // 'parallel for' implementation using outer and inner loops and PrevEUB
+      // LAMBDA: define{{.+}} void [[OMP_PARFOR_OUTLINED_7]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}}, {{.+}})
+      // LAMBDA-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+      // LAMBDA-DAG: [[OMP_PF_ST:%.omp.stride]] = alloca{{.+}},
+
+      // initialize lb and ub to PrevLB and PrevUB
+      // LAMBDA-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+      // LAMBDA-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+      // LAMBDA-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+      // LAMBDA-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+      // LAMBDA-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+      // LAMBDA-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+      // LAMBDA-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+      // LAMBDA-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA-DAG: [[OMP_PF_LB_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_UB]],
+      // LAMBDA: call void @__kmpc_dispatch_init_4({{.+}}, {{.+}}, {{.+}} 35, {{.+}} [[OMP_PF_LB_VAL]], {{.+}} [[OMP_PF_UB_VAL]], {{.+}}, {{.+}})
+      // LAMBDA: br label %[[OMP_PF_OUTER_LOOP_HEADER:.+]]
+
+      // LAMBDA: [[OMP_PF_OUTER_LOOP_HEADER]]:
+      // LAMBDA: [[IS_FIN:%.+]] = call{{.+}} @__kmpc_dispatch_next_4({{.+}}, {{.+}}, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]], {{.+}}* [[OMP_PF_ST]])
+      // LAMBDA: [[IS_FIN_CMP:%.+]] = icmp{{.+}} [[IS_FIN]], 0
+      // LAMBDA: br{{.+}} [[IS_FIN_CMP]], label %[[OMP_PF_OUTER_LOOP_BODY:.+]], label %[[OMP_PF_OUTER_LOOP_END:.+]]
+
+      // initialize omp.iv (IV = LB)
+      // LAMBDA: [[OMP_PF_OUTER_LOOP_BODY]]:
+      // LAMBDA-DAG: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+      // LAMBDA-DAG: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA: br label %[[OMP_PF_INNER_LOOP_HEADER:.+]]
+
+      // LAMBDA: [[OMP_PF_INNER_LOOP_HEADER]]:
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // LAMBDA-DAG: [[OMP_PF_UB_VAL_4:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_UB]],
+      // LAMBDA: [[PF_CMP_IV_UB_2:%.+]] = icmp{{.+}} [[OMP_PF_IV_VAL_2]], [[OMP_PF_UB_VAL_4]]
+      // LAMBDA: br{{.+}} [[PF_CMP_IV_UB_2]], label %[[OMP_PF_INNER_LOOP_BODY:.+]], label %[[OMP_PF_INNER_LOOP_END:.+]]
+
+      // LAMBDA: [[OMP_PF_INNER_LOOP_BODY]]:
+      // LAMBDA-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+      // skip body branch
+      // LAMBDA: br{{.+}}
+      // LAMBDA: br label %[[OMP_PF_INNER_LOOP_INC:.+]]
+
+      // IV = IV + 1 and inner loop latch
+      // LAMBDA: [[OMP_PF_INNER_LOOP_INC]]:
+      // LAMBDA-DAG: [[OMP_PF_IV_VAL_3:%.+]] = load{{.+}}, {{.+}}* [[OMP_IV]],
+      // LAMBDA-DAG: [[OMP_PF_NEXT_IV:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_3]], 1
+      // LAMBDA-DAG: store{{.+}} [[OMP_PF_NEXT_IV]], {{.+}}* [[OMP_IV]],
+      // LAMBDA: br label %[[OMP_PF_INNER_FOR_HEADER]]
+
+      // check NextLB and NextUB
+      // LAMBDA: [[OMP_PF_INNER_LOOP_END]]:
+      // LAMBDA: br label %[[OMP_PF_OUTER_LOOP_INC:.+]]
+
+      // LAMBDA: [[OMP_PF_OUTER_LOOP_INC]]:
+      // LAMBDA: br label %[[OMP_PF_OUTER_LOOP_HEADER]]
+
+      // LAMBDA: [[OMP_PF_OUTER_LOOP_END]]:
+      // LAMBDA: ret
+      [&]() {
+  a[i] = b[i] + c[i];
+      }();
+    }
+  }();
+  return 0;
+  #else
+  // CHECK-LABEL: @main
+
+  // CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+  // CHECK: call void [[OFFLOADING_FUN_1:@.+]](
+
+  // CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+  // CHECK: call void [[OFFLOADING_FUN_2:@.+]](
+
+  // CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+  // CHECK: call void [[OFFLOADING_FUN_3:@.+]](
+
+  // CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+  // CHECK: call void [[OFFLOADING_FUN_4:@.+]](
+
+  // CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+  // CHECK: call void [[OFFLOADING_FUN_5:@.+]](
+
+  // CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+  // CHECK: call void [[OFFLOADING_FUN_6:@.+]](
+
+  // CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+  // CHECK: call void [[OFFLOADING_FUN_7:@.+]](
+
+  // CHECK: call{{.+}} [[TMAIN:@.+]]()
+
+  // no schedule clauses
+  #pragma omp target
+  #pragma omp teams
+  // CHECK: define internal void [[OFFLOADING_FUN_1]](
+  // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED_1:@.+]] to {{.+}})
+
+  #pragma omp distribute parallel for
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+    // CHECK: define{{.+}} void [[OMP_OUTLINED_1]](
+    // CHECK-DAG: [[OMP_IV:%.omp.iv]] = alloca
+    // CHECK-DAG: [[OMP_LB:%.omp.lb]] = alloca
+    // CHECK-DAG: [[OMP_UB:%.omp.ub]] = alloca
+    // CHECK-DAG: [[OMP_ST:%.omp.stride]] = alloca
+
+    // CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+
+    // check EUB for distribute
+    // CHECK-DAG: [[OMP_UB_VAL_1:%.+]] = load{{.+}} [[OMP_UB]],
+    // CHECK: [[NUM_IT_1:%.+]] = load{{.+}},
+    // CHECK-DAG: [[CMP_UB_NUM_IT:%.+]] = icmp sgt {{.+}}  [[OMP_UB_VAL_1]], [[NUM_IT_1]]
+    // CHECK: br {{.+}} [[CMP_UB_NUM_IT]], label %[[EUB_TRUE:.+]], label %[[EUB_FALSE:.+]]
+    // CHECK-DAG: [[EUB_TRUE]]:
+    // CHECK: [[NUM_IT_2:%.+]] = load{{.+}},
+    // CHECK: br label %[[EUB_END:.+]]
+    // CHECK-DAG: [[EUB_FALSE]]:
+    // CHECK: [[OMP_UB_VAL2:%.+]] = load{{.+}} [[OMP_UB]],
+    // CHECK: br label %[[EUB_END]]
+    // CHECK-DAG: [[EUB_END]]:
+    // CHECK-DAG: [[EUB_RES:%.+]] = phi{{.+}} [ [[NUM_IT_2]], %[[EUB_TRUE]] ], [ [[OMP_UB_VAL2]], %[[EUB_FALSE]] ]
+    // CHECK: store{{.+}} [[EUB_RES]], {{.+}}* [[OMP_UB]],
+
+    // initialize omp.iv
+    // CHECK: [[OMP_LB_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[OMP_LB]],
+    // CHECK: store {{.+}} [[OMP_LB_VAL_1]], {{.+}}* [[OMP_IV]],
+    // CHECK: br label %[[OMP_JUMP_BACK:.+]]
+
+    // check exit condition
+    // CHECK: [[OMP_JUMP_BACK]]:
+    // CHECK-DAG: [[OMP_IV_VAL_1:%.+]] = load {{.+}} [[OMP_IV]],
+    // CHECK-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_UB]],
+    // CHECK: [[CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_IV_VAL_1]], [[OMP_UB_VAL_3]]
+    // CHECK: br {{.+}} [[CMP_IV_UB]], label %[[DIST_BODY:.+]], label %[[DIST_END:.+]]
+
+    // check that PrevLB and PrevUB are passed to the 'for'
+    // CHECK: [[DIST_BODY]]:
+    // CHECK-DAG: [[OMP_PREV_LB:%.+]] = load {{.+}}, {{.+}} [[OMP_LB]],
+    // CHECK-64-DAG: [[OMP_PREV_LB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_LB]] to {{.+}}
+    // CHECK-DAG: [[OMP_PREV_UB:%.+]] = load {{.+}}, {{.+}} [[OMP_UB]],
+    // CHECK-64-DAG: [[OMP_PREV_UB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_UB]] to {{.+}}
+    // check that distlb and distub are properly passed to fork_call
+    // CHECK-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_1:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_EXT]], i{{[0-9]+}} [[OMP_PREV_UB_EXT]], {{.+}})
+    // CHECK-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_1:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB]], i{{[0-9]+}} [[OMP_PREV_UB]], {{.+}})
+    // CHECK: br label %[[DIST_INC:.+]]
+
+    // increment by stride (distInc - 'parallel for' executes the whole chunk) and latch
+    // CHECK: [[DIST_INC]]:
+    // CHECK-DAG: [[OMP_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_IV]],
+    // CHECK-DAG: [[OMP_ST_VAL_1:%.+]] = load {{.+}}, {{.+}}* [[OMP_ST]],
+    // CHECK: [[OMP_IV_INC:%.+]] = add{{.+}} [[OMP_IV_VAL_2]], [[OMP_ST_VAL_1]]
+    // CHECK: store{{.+}} [[OMP_IV_INC]], {{.+}}* [[OMP_IV]],
+    // CHECK: br label %[[OMP_JUMP_BACK]]
+
+    // CHECK-DAG: call void @__kmpc_for_static_fini(
+    // CHECK: ret
+
+    // implementation of 'parallel for'
+    // CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_1]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}})
+
+    // CHECK-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+
+    // initialize lb and ub to PrevLB and PrevUB
+    // CHECK-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+    // CHECK-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+    // CHECK-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+    // CHECK-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+    // CHECK-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+    // CHECK-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+    // CHECK-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 34, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]],{{.+}})
+
+    // PrevEUB is only used when 'for' has a chunked schedule, otherwise EUB is used
+    // In this case we use EUB
+    // CHECK-DAG: [[OMP_PF_UB_VAL_1:%.+]] = load{{.+}} [[OMP_PF_UB]],
+    // CHECK: [[PF_NUM_IT_1:%.+]] = load{{.+}},
+    // CHECK-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_1]], [[PF_NUM_IT_1]]
+    // CHECK: br i1 [[PF_CMP_UB_NUM_IT]], label %[[PF_EUB_TRUE:.+]], label %[[PF_EUB_FALSE:.+]]
+    // CHECK: [[PF_EUB_TRUE]]:
+    // CHECK: [[PF_NUM_IT_2:%.+]] = load{{.+}},
+    // CHECK: br label %[[PF_EUB_END:.+]]
+    // CHECK-DAG: [[PF_EUB_FALSE]]:
+    // CHECK: [[OMP_PF_UB_VAL2:%.+]] = load{{.+}} [[OMP_PF_UB]],
+    // CHECK: br label %[[PF_EUB_END]]
+    // CHECK-DAG: [[PF_EUB_END]]:
+    // CHECK-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_NUM_IT_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL2]], %[[PF_EUB_FALSE]] ]
+    // CHECK: store{{.+}} [[PF_EUB_RES]],{{.+}}  [[OMP_PF_UB]],
+
+    // initialize omp.iv
+    // CHECK: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+    // CHECK: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+    // CHECK: br label %[[OMP_PF_JUMP_BACK:.+]]
+
+    // check exit condition
+    // CHECK: [[OMP_PF_JUMP_BACK]]:
+    // CHECK-DAG: [[OMP_PF_IV_VAL_1:%.+]] = load {{.+}} [[OMP_PF_IV]],
+    // CHECK-DAG: [[OMP_PF_UB_VAL_3:%.+]] = load {{.+}} [[OMP_PF_UB]],
+    // CHECK: [[PF_CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_PF_IV_VAL_1]], [[OMP_PF_UB_VAL_3]]
+    // CHECK: br {{.+}} [[PF_CMP_IV_UB]], label %[[PF_BODY:.+]], label %[[PF_END:.+]]
+
+    // check that PrevLB and PrevUB are passed to the 'for'
+    // CHECK: [[PF_BODY]]:
+    // CHECK-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // CHECK: br label {{.+}}
+
+    // check stride 1 for 'for' in 'distribute parallel for'
+    // CHECK-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // CHECK: [[OMP_PF_IV_INC:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_2]], 1
+    // CHECK: store{{.+}} [[OMP_PF_IV_INC]], {{.+}}* [[OMP_PF_IV]],
+    // CHECK: br label %[[OMP_PF_JUMP_BACK]]
+
+    // CHECK-DAG: call void @__kmpc_for_static_fini(
+    // CHECK: ret
+  }
+
+  // dist_schedule: static no chunk
+  #pragma omp target
+  #pragma omp teams
+  // CHECK: define{{.+}} void [[OFFLOADING_FUN_2]](
+  // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED_2:@.+]] to {{.+}})
+
+  #pragma omp distribute parallel for dist_schedule(static)
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+    // CHECK: define{{.+}} void [[OMP_OUTLINED_2]](
+    // CHECK-DAG: [[OMP_IV:%.omp.iv]] = alloca
+    // CHECK-DAG: [[OMP_LB:%.omp.lb]] = alloca
+    // CHECK-DAG: [[OMP_UB:%.omp.ub]] = alloca
+    // CHECK-DAG: [[OMP_ST:%.omp.stride]] = alloca
+
+    // CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+
+    // check EUB for distribute
+    // CHECK-DAG: [[OMP_UB_VAL_1:%.+]] = load{{.+}} [[OMP_UB]],
+    // CHECK: [[NUM_IT_1:%.+]] = load{{.+}},
+    // CHECK-DAG: [[CMP_UB_NUM_IT:%.+]] = icmp sgt {{.+}}  [[OMP_UB_VAL_1]], [[NUM_IT_1]]
+    // CHECK: br {{.+}} [[CMP_UB_NUM_IT]], label %[[EUB_TRUE:.+]], label %[[EUB_FALSE:.+]]
+    // CHECK-DAG: [[EUB_TRUE]]:
+    // CHECK: [[NUM_IT_2:%.+]] = load{{.+}},
+    // CHECK: br label %[[EUB_END:.+]]
+    // CHECK-DAG: [[EUB_FALSE]]:
+    // CHECK: [[OMP_UB_VAL2:%.+]] = load{{.+}} [[OMP_UB]],
+    // CHECK: br label %[[EUB_END]]
+    // CHECK-DAG: [[EUB_END]]:
+    // CHECK-DAG: [[EUB_RES:%.+]] = phi{{.+}} [ [[NUM_IT_2]], %[[EUB_TRUE]] ], [ [[OMP_UB_VAL2]], %[[EUB_FALSE]] ]
+    // CHECK: store{{.+}} [[EUB_RES]], {{.+}}* [[OMP_UB]],
+
+    // initialize omp.iv
+    // CHECK: [[OMP_LB_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[OMP_LB]],
+    // CHECK: store {{.+}} [[OMP_LB_VAL_1]], {{.+}}* [[OMP_IV]],
+    // CHECK: br label %[[OMP_JUMP_BACK:.+]]
+
+    // check exit condition
+    // CHECK: [[OMP_JUMP_BACK]]:
+    // CHECK-DAG: [[OMP_IV_VAL_1:%.+]] = load {{.+}} [[OMP_IV]],
+    // CHECK-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_UB]],
+    // CHECK: [[CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_IV_VAL_1]], [[OMP_UB_VAL_3]]
+    // CHECK: br {{.+}} [[CMP_IV_UB]], label %[[DIST_BODY:.+]], label %[[DIST_END:.+]]
+
+    // check that PrevLB and PrevUB are passed to the 'for'
+    // CHECK: [[DIST_BODY]]:
+    // CHECK-DAG: [[OMP_PREV_LB:%.+]] = load {{.+}}, {{.+}} [[OMP_LB]],
+    // CHECK-64-DAG: [[OMP_PREV_LB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_LB]] to {{.+}}
+    // CHECK-DAG: [[OMP_PREV_UB:%.+]] = load {{.+}}, {{.+}} [[OMP_UB]],
+    // CHECK-64-DAG: [[OMP_PREV_UB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_UB]] to {{.+}}
+    // check that distlb and distub are properly passed to fork_call
+    // CHECK-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_2:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_EXT]], i{{[0-9]+}} [[OMP_PREV_UB_EXT]], {{.+}})
+    // CHECK-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_2:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB]], i{{[0-9]+}} [[OMP_PREV_UB]], {{.+}})
+    // CHECK: br label %[[DIST_INC:.+]]
+
+    // increment by stride (distInc - 'parallel for' executes the whole chunk) and latch
+    // CHECK: [[DIST_INC]]:
+    // CHECK-DAG: [[OMP_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_IV]],
+    // CHECK-DAG: [[OMP_ST_VAL_1:%.+]] = load {{.+}}, {{.+}}* [[OMP_ST]],
+    // CHECK: [[OMP_IV_INC:%.+]] = add{{.+}} [[OMP_IV_VAL_2]], [[OMP_ST_VAL_1]]
+    // CHECK: store{{.+}} [[OMP_IV_INC]], {{.+}}* [[OMP_IV]],
+    // CHECK: br label %[[OMP_JUMP_BACK]]
+
+    // CHECK-DAG: call void @__kmpc_for_static_fini(
+    // CHECK: ret
+
+    // implementation of 'parallel for'
+    // CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_2]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}})
+
+    // CHECK-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+
+    // initialize lb and ub to PrevLB and PrevUB
+    // CHECK-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+    // CHECK-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+    // CHECK-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+    // CHECK-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+    // CHECK-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+    // CHECK-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+    // CHECK-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 34, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]],{{.+}})
+
+    // PrevEUB is only used when 'for' has a chunked schedule, otherwise EUB is used
+    // In this case we use EUB
+    // CHECK-DAG: [[OMP_PF_UB_VAL_1:%.+]] = load{{.+}} [[OMP_PF_UB]],
+    // CHECK: [[PF_NUM_IT_1:%.+]] = load{{.+}},
+    // CHECK-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_1]], [[PF_NUM_IT_1]]
+    // CHECK: br i1 [[PF_CMP_UB_NUM_IT]], label %[[PF_EUB_TRUE:.+]], label %[[PF_EUB_FALSE:.+]]
+    // CHECK: [[PF_EUB_TRUE]]:
+    // CHECK: [[PF_NUM_IT_2:%.+]] = load{{.+}},
+    // CHECK: br label %[[PF_EUB_END:.+]]
+    // CHECK-DAG: [[PF_EUB_FALSE]]:
+    // CHECK: [[OMP_PF_UB_VAL2:%.+]] = load{{.+}} [[OMP_PF_UB]],
+    // CHECK: br label %[[PF_EUB_END]]
+    // CHECK-DAG: [[PF_EUB_END]]:
+    // CHECK-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_NUM_IT_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL2]], %[[PF_EUB_FALSE]] ]
+    // CHECK: store{{.+}} [[PF_EUB_RES]],{{.+}}  [[OMP_PF_UB]],
+
+    // initialize omp.iv
+    // CHECK: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+    // CHECK: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+    // CHECK: br label %[[OMP_PF_JUMP_BACK:.+]]
+
+    // check exit condition
+    // CHECK: [[OMP_PF_JUMP_BACK]]:
+    // CHECK-DAG: [[OMP_PF_IV_VAL_1:%.+]] = load {{.+}} [[OMP_PF_IV]],
+    // CHECK-DAG: [[OMP_PF_UB_VAL_3:%.+]] = load {{.+}} [[OMP_PF_UB]],
+    // CHECK: [[PF_CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_PF_IV_VAL_1]], [[OMP_PF_UB_VAL_3]]
+    // CHECK: br {{.+}} [[PF_CMP_IV_UB]], label %[[PF_BODY:.+]], label %[[PF_END:.+]]
+
+    // check that PrevLB and PrevUB are passed to the 'for'
+    // CHECK: [[PF_BODY]]:
+    // CHECK-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // CHECK: br label {{.+}}
+
+    // check stride 1 for 'for' in 'distribute parallel for'
+    // CHECK-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // CHECK: [[OMP_PF_IV_INC:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_2]], 1
+    // CHECK: store{{.+}} [[OMP_PF_IV_INC]], {{.+}}* [[OMP_PF_IV]],
+    // CHECK: br label %[[OMP_PF_JUMP_BACK]]
+
+    // CHECK-DAG: call void @__kmpc_for_static_fini(
+    // CHECK: ret
+  }
+
+  // dist_schedule: static chunk
+  #pragma omp target
+  #pragma omp teams
+  // CHECK: define{{.+}} void [[OFFLOADING_FUN_3]](
+  // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 5, {{.+}}* [[OMP_OUTLINED_3:@.+]] to {{.+}})
+
+  #pragma omp distribute parallel for dist_schedule(static, ch)
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+    // CHECK: define{{.+}} void [[OMP_OUTLINED_3]](
+    // CHECK-DAG: [[OMP_IV:%.omp.iv]] = alloca
+    // CHECK-DAG: [[OMP_LB:%.omp.lb]] = alloca
+    // CHECK-DAG: [[OMP_UB:%.omp.ub]] = alloca
+    // CHECK-DAG: [[OMP_ST:%.omp.stride]] = alloca
+
+    // unlike the previous tests, in this one we have a outer and inner loop for 'distribute'
+    // CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 91,
+    // CHECK: br label %[[DIST_OUTER_LOOP_HEADER:.+]]
+
+    // CHECK: [[DIST_OUTER_LOOP_HEADER]]:
+    // check EUB for distribute
+    // CHECK-DAG: [[OMP_UB_VAL_1:%.+]] = load{{.+}} [[OMP_UB]],
+    // CHECK: [[NUM_IT_1:%.+]] = load{{.+}},
+    // CHECK-DAG: [[CMP_UB_NUM_IT:%.+]] = icmp sgt {{.+}}  [[OMP_UB_VAL_1]], [[NUM_IT_1]]
+    // CHECK: br {{.+}} [[CMP_UB_NUM_IT]], label %[[EUB_TRUE:.+]], label %[[EUB_FALSE:.+]]
+    // CHECK-DAG: [[EUB_TRUE]]:
+    // CHECK: [[NUM_IT_2:%.+]] = load{{.+}},
+    // CHECK: br label %[[EUB_END:.+]]
+    // CHECK-DAG: [[EUB_FALSE]]:
+    // CHECK: [[OMP_UB_VAL2:%.+]] = load{{.+}} [[OMP_UB]],
+    // CHECK: br label %[[EUB_END]]
+    // CHECK-DAG: [[EUB_END]]:
+    // CHECK-DAG: [[EUB_RES:%.+]] = phi{{.+}} [ [[NUM_IT_2]], %[[EUB_TRUE]] ], [ [[OMP_UB_VAL2]], %[[EUB_FALSE]] ]
+    // CHECK: store{{.+}} [[EUB_RES]], {{.+}}* [[OMP_UB]],
+
+    // initialize omp.iv
+    // CHECK: [[OMP_LB_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[OMP_LB]],
+    // CHECK: store {{.+}} [[OMP_LB_VAL_1]], {{.+}}* [[OMP_IV]],
+
+    // check exit condition
+    // CHECK-DAG: [[OMP_IV_VAL_1:%.+]] = load {{.+}} [[OMP_IV]],
+    // CHECK-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_UB]],
+    // CHECK: [[CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_IV_VAL_1]], [[OMP_UB_VAL_3]]
+    // CHECK: br {{.+}} [[CMP_IV_UB]], label %[[DIST_OUTER_LOOP_BODY:.+]], label %[[DIST_OUTER_LOOP_END:.+]]
+
+    // CHECK: [[DIST_OUTER_LOOP_BODY]]:
+    // CHECK: br label %[[DIST_INNER_LOOP_HEADER:.+]]
+
+    // CHECK: [[DIST_INNER_LOOP_HEADER]]:
+    // CHECK-DAG: [[OMP_IV_VAL_2:%.+]] = load {{.+}} [[OMP_IV]],
+    // CHECK-DAG: [[OMP_UB_VAL_4:%.+]] = load {{.+}} [[OMP_UB]],
+    // CHECK: [[CMP_IV_UB_2:%.+]] = icmp sle {{.+}} [[OMP_IV_VAL_2]], [[OMP_UB_VAL_4]]
+    // CHECK: br{{.+}} [[CMP_IV_UB_2]], label %[[DIST_INNER_LOOP_BODY:.+]], label %[[DIST_INNER_LOOP_END:.+]]
+
+    // check that PrevLB and PrevUB are passed to the 'for'
+    // CHECK: [[DIST_INNER_LOOP_BODY]]:
+    // CHECK-DAG: [[OMP_PREV_LB:%.+]] = load {{.+}}, {{.+}} [[OMP_LB]],
+    // CHECK-64-DAG: [[OMP_PREV_LB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_LB]] to {{.+}}
+    // CHECK-DAG: [[OMP_PREV_UB:%.+]] = load {{.+}}, {{.+}} [[OMP_UB]],
+    // CHECK-64-DAG: [[OMP_PREV_UB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_UB]] to {{.+}}
+    // check that distlb and distub are properly passed to fork_call
+    // CHECK-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_3:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_EXT]], i{{[0-9]+}} [[OMP_PREV_UB_EXT]], {{.+}})
+    // CHECK-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_3:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB]], i{{[0-9]+}} [[OMP_PREV_UB]], {{.+}})
+    // CHECK: br label %[[DIST_INNER_LOOP_INC:.+]]
+
+    // check DistInc
+    // CHECK: [[DIST_INNER_LOOP_INC]]:
+    // CHECK-DAG: [[OMP_IV_VAL_3:%.+]] = load {{.+}}, {{.+}}* [[OMP_IV]],
+    // CHECK-DAG: [[OMP_ST_VAL_1:%.+]] = load {{.+}}, {{.+}}* [[OMP_ST]],
+    // CHECK: [[OMP_IV_INC:%.+]] = add{{.+}} [[OMP_IV_VAL_3]], [[OMP_ST_VAL_1]]
+    // CHECK: store{{.+}} [[OMP_IV_INC]], {{.+}}* [[OMP_IV]],
+    // CHECK: br label %[[DIST_INNER_LOOP_HEADER]]
+
+    // CHECK: [[DIST_INNER_LOOP_END]]:
+    // CHECK: br label %[[DIST_OUTER_LOOP_INC:.+]]
+
+    // CHECK: [[DIST_OUTER_LOOP_INC]]:
+    // check NextLB and NextUB
+    // CHECK-DAG: [[OMP_LB_VAL_2:%.+]] = load{{.+}}, {{.+}} [[OMP_LB]],
+    // CHECK-DAG: [[OMP_ST_VAL_2:%.+]] = load{{.+}}, {{.+}} [[OMP_ST]],
+    // CHECK-DAG: [[OMP_LB_NEXT:%.+]] = add{{.+}} [[OMP_LB_VAL_2]], [[OMP_ST_VAL_2]]
+    // CHECK: store{{.+}} [[OMP_LB_NEXT]], {{.+}}* [[OMP_LB]],
+    // CHECK-DAG: [[OMP_UB_VAL_5:%.+]] = load{{.+}}, {{.+}} [[OMP_UB]],
+    // CHECK-DAG: [[OMP_ST_VAL_3:%.+]] = load{{.+}}, {{.+}} [[OMP_ST]],
+    // CHECK-DAG: [[OMP_UB_NEXT:%.+]] = add{{.+}} [[OMP_UB_VAL_5]], [[OMP_ST_VAL_3]]
+    // CHECK: store{{.+}} [[OMP_UB_NEXT]], {{.+}}* [[OMP_UB]],
+    // CHECK: br label %[[DIST_OUTER_LOOP_HEADER]]
+
+    // outer loop exit
+    // CHECK: [[DIST_OUTER_LOOP_END]]:
+    // CHECK-DAG: call void @__kmpc_for_static_fini(
+    // CHECK: ret
+
+    // skip implementation of 'parallel for': using default scheduling and was tested above
+  }
+
+  // schedule: static no chunk
+  #pragma omp target
+  #pragma omp teams
+  // CHECK: define{{.+}} void [[OFFLOADING_FUN_4]](
+  // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED_4:@.+]] to {{.+}})
+
+  #pragma omp distribute parallel for schedule(static)
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+    // CHECK: define{{.+}} void [[OMP_OUTLINED_4]](
+    // CHECK-DAG: [[OMP_IV:%.omp.iv]] = alloca
+    // CHECK-DAG: [[OMP_LB:%.omp.lb]] = alloca
+    // CHECK-DAG: [[OMP_UB:%.omp.ub]] = alloca
+    // CHECK-DAG: [[OMP_ST:%.omp.stride]] = alloca
+
+    // CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+    // CHECK: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_4:@.+]] to {{.+}},
+    // skip rest of implementation of 'distribute' as it is tested above for default dist_schedule case
+    // CHECK: ret
+
+    // 'parallel for' implementation is the same as the case without schedule clase (static no chunk is the default)
+    // CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_4]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}})
+
+    // CHECK-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+
+    // initialize lb and ub to PrevLB and PrevUB
+    // CHECK-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+    // CHECK-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+    // CHECK-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+    // CHECK-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+    // CHECK-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+    // CHECK-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+    // CHECK-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 34, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]],{{.+}})
+
+    // PrevEUB is only used when 'for' has a chunked schedule, otherwise EUB is used
+    // In this case we use EUB
+    // CHECK-DAG: [[OMP_PF_UB_VAL_1:%.+]] = load{{.+}} [[OMP_PF_UB]],
+    // CHECK: [[PF_NUM_IT_1:%.+]] = load{{.+}},
+    // CHECK-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_1]], [[PF_NUM_IT_1]]
+    // CHECK: br i1 [[PF_CMP_UB_NUM_IT]], label %[[PF_EUB_TRUE:.+]], label %[[PF_EUB_FALSE:.+]]
+    // CHECK: [[PF_EUB_TRUE]]:
+    // CHECK: [[PF_NUM_IT_2:%.+]] = load{{.+}},
+    // CHECK: br label %[[PF_EUB_END:.+]]
+    // CHECK-DAG: [[PF_EUB_FALSE]]:
+    // CHECK: [[OMP_PF_UB_VAL2:%.+]] = load{{.+}} [[OMP_PF_UB]],
+    // CHECK: br label %[[PF_EUB_END]]
+    // CHECK-DAG: [[PF_EUB_END]]:
+    // CHECK-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_NUM_IT_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL2]], %[[PF_EUB_FALSE]] ]
+    // CHECK: store{{.+}} [[PF_EUB_RES]],{{.+}}  [[OMP_PF_UB]],
+
+    // initialize omp.iv
+    // CHECK: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+    // CHECK: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+    // CHECK: br label %[[OMP_PF_JUMP_BACK:.+]]
+
+    // check exit condition
+    // CHECK: [[OMP_PF_JUMP_BACK]]:
+    // CHECK-DAG: [[OMP_PF_IV_VAL_1:%.+]] = load {{.+}} [[OMP_PF_IV]],
+    // CHECK-DAG: [[OMP_PF_UB_VAL_3:%.+]] = load {{.+}} [[OMP_PF_UB]],
+    // CHECK: [[PF_CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_PF_IV_VAL_1]], [[OMP_PF_UB_VAL_3]]
+    // CHECK: br {{.+}} [[PF_CMP_IV_UB]], label %[[PF_BODY:.+]], label %[[PF_END:.+]]
+
+    // check that PrevLB and PrevUB are passed to the 'for'
+    // CHECK: [[PF_BODY]]:
+    // CHECK-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // CHECK: br label {{.+}}
+
+    // check stride 1 for 'for' in 'distribute parallel for'
+    // CHECK-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // CHECK: [[OMP_PF_IV_INC:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_2]], 1
+    // CHECK: store{{.+}} [[OMP_PF_IV_INC]], {{.+}}* [[OMP_PF_IV]],
+    // CHECK: br label %[[OMP_PF_JUMP_BACK]]
+
+    // CHECK-DAG: call void @__kmpc_for_static_fini(
+    // CHECK: ret
+  }
+
+  // schedule: static chunk
+  #pragma omp target
+  #pragma omp teams
+  // CHECK: define{{.+}} void [[OFFLOADING_FUN_5]](
+  // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 5, {{.+}}* [[OMP_OUTLINED_5:@.+]] to {{.+}})
+
+  #pragma omp distribute parallel for schedule(static, ch)
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+    // CHECK: define{{.+}} void [[OMP_OUTLINED_5]](
+    // CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+    // CHECK: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_5:@.+]] to {{.+}},
+    // skip rest of implementation of 'distribute' as it is tested above for default dist_schedule case
+    // CHECK: ret
+
+    // 'parallel for' implementation using outer and inner loops and PrevEUB
+    // CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_5]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}}, {{.+}})
+    // CHECK-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_ST:%.omp.stride]] = alloca{{.+}},
+
+    // initialize lb and ub to PrevLB and PrevUB
+    // CHECK-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+    // CHECK-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+    // CHECK-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+    // CHECK-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+    // CHECK-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+    // CHECK-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+    // CHECK-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 33, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]],{{.+}})
+    // CHECK: br label %[[OMP_PF_OUTER_LOOP_HEADER:.+]]
+
+    // check PrevEUB (using PrevUB instead of NumIt as upper bound)
+    // CHECK: [[OMP_PF_OUTER_LOOP_HEADER]]:
+    // CHECK-DAG: [[OMP_PF_UB_VAL_1:%.+]] = load{{.+}} [[OMP_PF_UB]],
+    // CHECK-64-DAG: [[OMP_PF_UB_VAL_CONV:%.+]] = sext{{.+}} [[OMP_PF_UB_VAL_1]] to
+    // CHECK: [[PF_PREV_UB_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+    // CHECK-64-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_CONV]], [[PF_PREV_UB_VAL_1]]
+    // CHECK-32-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_1]], [[PF_PREV_UB_VAL_1]]
+    // CHECK: br i1 [[PF_CMP_UB_NUM_IT]], label %[[PF_EUB_TRUE:.+]], label %[[PF_EUB_FALSE:.+]]
+    // CHECK: [[PF_EUB_TRUE]]:
+    // CHECK: [[PF_PREV_UB_VAL_2:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+    // CHECK: br label %[[PF_EUB_END:.+]]
+    // CHECK-DAG: [[PF_EUB_FALSE]]:
+    // CHECK: [[OMP_PF_UB_VAL_2:%.+]] = load{{.+}} [[OMP_PF_UB]],
+    // CHECK-64: [[OMP_PF_UB_VAL_2_CONV:%.+]] = sext{{.+}} [[OMP_PF_UB_VAL_2]] to
+    // CHECK: br label %[[PF_EUB_END]]
+    // CHECK-DAG: [[PF_EUB_END]]:
+    // CHECK-64-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_PREV_UB_VAL_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL_2_CONV]], %[[PF_EUB_FALSE]] ]
+    // CHECK-32-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_PREV_UB_VAL_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL_2]], %[[PF_EUB_FALSE]] ]
+    // CHECK-64-DAG: [[PF_EUB_RES_CONV:%.+]] = trunc{{.+}} [[PF_EUB_RES]] to
+    // CHECK-64: store{{.+}} [[PF_EUB_RES_CONV]],{{.+}}  [[OMP_PF_UB]],
+    // CHECK-32: store{{.+}} [[PF_EUB_RES]], {{.+}} [[OMP_PF_UB]],
+
+    // initialize omp.iv (IV = LB)
+    // CHECK: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+    // CHECK: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+
+    // outer loop: while (IV < UB) {
+    // CHECK-DAG: [[OMP_PF_IV_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // CHECK-DAG: [[OMP_PF_UB_VAL_3:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_UB]],
+    // CHECK: [[PF_CMP_IV_UB_1:%.+]] = icmp{{.+}} [[OMP_PF_IV_VAL_1]], [[OMP_PF_UB_VAL_3]]
+    // CHECK: br{{.+}} [[PF_CMP_IV_UB_1]], label %[[OMP_PF_OUTER_LOOP_BODY:.+]], label %[[OMP_PF_OUTER_LOOP_END:.+]]
+
+    // CHECK: [[OMP_PF_OUTER_LOOP_BODY]]:
+    // CHECK: br label %[[OMP_PF_INNER_FOR_HEADER:.+]]
+
+    // CHECK: [[OMP_PF_INNER_FOR_HEADER]]:
+    // CHECK-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // CHECK-DAG: [[OMP_PF_UB_VAL_4:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_UB]],
+    // CHECK: [[PF_CMP_IV_UB_2:%.+]] = icmp{{.+}} [[OMP_PF_IV_VAL_2]], [[OMP_PF_UB_VAL_4]]
+    // CHECK: br{{.+}} [[PF_CMP_IV_UB_2]], label %[[OMP_PF_INNER_LOOP_BODY:.+]], label %[[OMP_PF_INNER_LOOP_END:.+]]
+
+    // CHECK: [[OMP_PF_INNER_LOOP_BODY]]:
+    // CHECK-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // skip body branch
+    // CHECK: br{{.+}}
+    // CHECK: br label %[[OMP_PF_INNER_LOOP_INC:.+]]
+
+    // IV = IV + 1 and inner loop latch
+    // CHECK: [[OMP_PF_INNER_LOOP_INC]]:
+    // CHECK-DAG: [[OMP_PF_IV_VAL_3:%.+]] = load{{.+}}, {{.+}}* [[OMP_IV]],
+    // CHECK-DAG: [[OMP_PF_NEXT_IV:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_3]], 1
+    // CHECK-DAG: store{{.+}} [[OMP_PF_NEXT_IV]], {{.+}}* [[OMP_IV]],
+    // CHECK: br label %[[OMP_PF_INNER_FOR_HEADER]]
+
+    // check NextLB and NextUB
+    // CHECK: [[OMP_PF_INNER_LOOP_END]]:
+    // CHECK: br label %[[OMP_PF_OUTER_LOOP_INC:.+]]
+
+    // CHECK: [[OMP_PF_OUTER_LOOP_INC]]:
+    // CHECK-DAG: [[OMP_PF_LB_VAL_2:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+    // CHECK-DAG: [[OMP_PF_ST_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_ST]],
+    // CHECK-DAG: [[OMP_PF_LB_NEXT:%.+]] = add{{.+}} [[OMP_PF_LB_VAL_2]], [[OMP_PF_ST_VAL_1]]
+    // CHECK: store{{.+}} [[OMP_PF_LB_NEXT]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-DAG: [[OMP_PF_UB_VAL_5:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_UB]],
+    // CHECK-DAG: [[OMP_PF_ST_VAL_2:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_ST]],
+    // CHECK-DAG: [[OMP_PF_UB_NEXT:%.+]] = add{{.+}} [[OMP_PF_UB_VAL_5]], [[OMP_PF_ST_VAL_2]]
+    // CHECK: store{{.+}} [[OMP_PF_UB_NEXT]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK: br label %[[OMP_PF_OUTER_LOOP_HEADER]]
+
+    // CHECK: [[OMP_PF_OUTER_LOOP_END]]:
+    // CHECK-DAG: call void @__kmpc_for_static_fini(
+    // CHECK: ret
+  }
+
+  // schedule: dynamic no chunk
+  #pragma omp target
+  #pragma omp teams
+  // CHECK: define{{.+}} void [[OFFLOADING_FUN_6]](
+  // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED_6:@.+]] to {{.+}})
+
+  #pragma omp distribute parallel for schedule(dynamic)
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+    // CHECK: define{{.+}} void [[OMP_OUTLINED_6]](
+    // CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+    // CHECK: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_6:@.+]] to {{.+}},
+    // skip rest of implementation of 'distribute' as it is tested above for default dist_schedule case
+    // CHECK: ret
+
+    // 'parallel for' implementation using outer and inner loops and PrevEUB
+    // CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_6]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}})
+    // CHECK-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_ST:%.omp.stride]] = alloca{{.+}},
+
+    // initialize lb and ub to PrevLB and PrevUB
+    // CHECK-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+    // CHECK-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+    // CHECK-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+    // CHECK-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+    // CHECK-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+    // CHECK-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+    // CHECK-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK-DAG: [[OMP_PF_LB_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+    // CHECK-DAG: [[OMP_PF_UB_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_UB]],
+    // CHECK: call void @__kmpc_dispatch_init_4({{.+}}, {{.+}}, {{.+}} 35, {{.+}} [[OMP_PF_LB_VAL]], {{.+}} [[OMP_PF_UB_VAL]], {{.+}}, {{.+}})
+    // CHECK: br label %[[OMP_PF_OUTER_LOOP_HEADER:.+]]
+
+    // CHECK: [[OMP_PF_OUTER_LOOP_HEADER]]:
+    // CHECK: [[IS_FIN:%.+]] = call{{.+}} @__kmpc_dispatch_next_4({{.+}}, {{.+}}, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]], {{.+}}* [[OMP_PF_ST]])
+    // CHECK: [[IS_FIN_CMP:%.+]] = icmp{{.+}} [[IS_FIN]], 0
+    // CHECK: br{{.+}} [[IS_FIN_CMP]], label %[[OMP_PF_OUTER_LOOP_BODY:.+]], label %[[OMP_PF_OUTER_LOOP_END:.+]]
+
+    // initialize omp.iv (IV = LB)
+    // CHECK: [[OMP_PF_OUTER_LOOP_BODY]]:
+    // CHECK-DAG: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+    // CHECK-DAG: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+    // CHECK: br label %[[OMP_PF_INNER_LOOP_HEADER:.+]]
+
+    // CHECK: [[OMP_PF_INNER_LOOP_HEADER]]:
+    // CHECK-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // CHECK-DAG: [[OMP_PF_UB_VAL_4:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_UB]],
+    // CHECK: [[PF_CMP_IV_UB_2:%.+]] = icmp{{.+}} [[OMP_PF_IV_VAL_2]], [[OMP_PF_UB_VAL_4]]
+    // CHECK: br{{.+}} [[PF_CMP_IV_UB_2]], label %[[OMP_PF_INNER_LOOP_BODY:.+]], label %[[OMP_PF_INNER_LOOP_END:.+]]
+
+    // CHECK: [[OMP_PF_INNER_LOOP_BODY]]:
+    // CHECK-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // skip body branch
+    // CHECK: br{{.+}}
+    // CHECK: br label %[[OMP_PF_INNER_LOOP_INC:.+]]
+
+    // IV = IV + 1 and inner loop latch
+    // CHECK: [[OMP_PF_INNER_LOOP_INC]]:
+    // CHECK-DAG: [[OMP_PF_IV_VAL_3:%.+]] = load{{.+}}, {{.+}}* [[OMP_IV]],
+    // CHECK-DAG: [[OMP_PF_NEXT_IV:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_3]], 1
+    // CHECK-DAG: store{{.+}} [[OMP_PF_NEXT_IV]], {{.+}}* [[OMP_IV]],
+    // CHECK: br label %[[OMP_PF_INNER_FOR_HEADER]]
+
+    // check NextLB and NextUB
+    // CHECK: [[OMP_PF_INNER_LOOP_END]]:
+    // CHECK: br label %[[OMP_PF_OUTER_LOOP_INC:.+]]
+
+    // CHECK: [[OMP_PF_OUTER_LOOP_INC]]:
+    // CHECK: br label %[[OMP_PF_OUTER_LOOP_HEADER]]
+
+    // CHECK: [[OMP_PF_OUTER_LOOP_END]]:
+    // CHECK: ret
+  }
+
+  // schedule: dynamic chunk
+  #pragma omp target
+  #pragma omp teams
+  // CHECK: define{{.+}} void [[OFFLOADING_FUN_7]](
+  // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 5, {{.+}}* [[OMP_OUTLINED_7:@.+]] to {{.+}})
+
+  #pragma omp distribute parallel for schedule(dynamic, ch)
+  for (int i = 0; i < n; ++i) {
+    a[i] = b[i] + c[i];
+    // CHECK: define{{.+}} void [[OMP_OUTLINED_7]](
+    // CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+    // CHECK: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_7:@.+]] to {{.+}},
+    // skip rest of implementation of 'distribute' as it is tested above for default dist_schedule case
+    // CHECK: ret
+
+    // 'parallel for' implementation using outer and inner loops and PrevEUB
+    // CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_7]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}}, {{.+}})
+    // CHECK-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+    // CHECK-DAG: [[OMP_PF_ST:%.omp.stride]] = alloca{{.+}},
+
+    // initialize lb and ub to PrevLB and PrevUB
+    // CHECK-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+    // CHECK-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+    // CHECK-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+    // CHECK-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+    // CHECK-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+    // CHECK-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+    // CHECK-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+    // CHECK-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+    // CHECK-DAG: [[OMP_PF_LB_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+    // CHECK-DAG: [[OMP_PF_UB_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_UB]],
+    // CHECK: call void @__kmpc_dispatch_init_4({{.+}}, {{.+}}, {{.+}} 35, {{.+}} [[OMP_PF_LB_VAL]], {{.+}} [[OMP_PF_UB_VAL]], {{.+}}, {{.+}})
+    // CHECK: br label %[[OMP_PF_OUTER_LOOP_HEADER:.+]]
+
+    // CHECK: [[OMP_PF_OUTER_LOOP_HEADER]]:
+    // CHECK: [[IS_FIN:%.+]] = call{{.+}} @__kmpc_dispatch_next_4({{.+}}, {{.+}}, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]], {{.+}}* [[OMP_PF_ST]])
+    // CHECK: [[IS_FIN_CMP:%.+]] = icmp{{.+}} [[IS_FIN]], 0
+    // CHECK: br{{.+}} [[IS_FIN_CMP]], label %[[OMP_PF_OUTER_LOOP_BODY:.+]], label %[[OMP_PF_OUTER_LOOP_END:.+]]
+
+    // initialize omp.iv (IV = LB)
+    // CHECK: [[OMP_PF_OUTER_LOOP_BODY]]:
+    // CHECK-DAG: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+    // CHECK-DAG: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+    // CHECK: br label %[[OMP_PF_INNER_LOOP_HEADER:.+]]
+
+    // CHECK: [[OMP_PF_INNER_LOOP_HEADER]]:
+    // CHECK-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // CHECK-DAG: [[OMP_PF_UB_VAL_4:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_UB]],
+    // CHECK: [[PF_CMP_IV_UB_2:%.+]] = icmp{{.+}} [[OMP_PF_IV_VAL_2]], [[OMP_PF_UB_VAL_4]]
+    // CHECK: br{{.+}} [[PF_CMP_IV_UB_2]], label %[[OMP_PF_INNER_LOOP_BODY:.+]], label %[[OMP_PF_INNER_LOOP_END:.+]]
+
+    // CHECK: [[OMP_PF_INNER_LOOP_BODY]]:
+    // CHECK-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+    // skip body branch
+    // CHECK: br{{.+}}
+    // CHECK: br label %[[OMP_PF_INNER_LOOP_INC:.+]]
+
+    // IV = IV + 1 and inner loop latch
+    // CHECK: [[OMP_PF_INNER_LOOP_INC]]:
+    // CHECK-DAG: [[OMP_PF_IV_VAL_3:%.+]] = load{{.+}}, {{.+}}* [[OMP_IV]],
+    // CHECK-DAG: [[OMP_PF_NEXT_IV:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_3]], 1
+    // CHECK-DAG: store{{.+}} [[OMP_PF_NEXT_IV]], {{.+}}* [[OMP_IV]],
+    // CHECK: br label %[[OMP_PF_INNER_FOR_HEADER]]
+
+    // check NextLB and NextUB
+    // CHECK: [[OMP_PF_INNER_LOOP_END]]:
+    // CHECK: br label %[[OMP_PF_OUTER_LOOP_INC:.+]]
+
+    // CHECK: [[OMP_PF_OUTER_LOOP_INC]]:
+    // CHECK: br label %[[OMP_PF_OUTER_LOOP_HEADER]]
+
+    // CHECK: [[OMP_PF_OUTER_LOOP_END]]:
+    // CHECK: ret
+  }
+
+  return tmain<int>();
+  #endif
+}
+
+// check code
+// CHECK: define{{.+}} [[TMAIN]]()
+
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_1:@.+]](
+
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_2:@.+]](
+
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_3:@.+]](
+
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_4:@.+]](
+
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_5:@.+]](
+
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_6:@.+]](
+
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_7:@.+]](
+
+// CHECK: define{{.+}} void [[OFFLOADING_FUN_1]](
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED_1:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} void [[OMP_OUTLINED_1]](
+// CHECK-DAG: [[OMP_IV:%.omp.iv]] = alloca
+// CHECK-DAG: [[OMP_LB:%.omp.lb]] = alloca
+// CHECK-DAG: [[OMP_UB:%.omp.ub]] = alloca
+// CHECK-DAG: [[OMP_ST:%.omp.stride]] = alloca
+
+// CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+
+// check EUB for distribute
+// CHECK-DAG: [[OMP_UB_VAL_1:%.+]] = load{{.+}} [[OMP_UB]],
+// CHECK: [[NUM_IT_1:%.+]] = load{{.+}},
+// CHECK-DAG: [[CMP_UB_NUM_IT:%.+]] = icmp sgt {{.+}}  [[OMP_UB_VAL_1]], [[NUM_IT_1]]
+// CHECK: br {{.+}} [[CMP_UB_NUM_IT]], label %[[EUB_TRUE:.+]], label %[[EUB_FALSE:.+]]
+// CHECK-DAG: [[EUB_TRUE]]:
+// CHECK: [[NUM_IT_2:%.+]] = load{{.+}},
+// CHECK: br label %[[EUB_END:.+]]
+// CHECK-DAG: [[EUB_FALSE]]:
+// CHECK: [[OMP_UB_VAL2:%.+]] = load{{.+}} [[OMP_UB]],
+// CHECK: br label %[[EUB_END]]
+// CHECK-DAG: [[EUB_END]]:
+// CHECK-DAG: [[EUB_RES:%.+]] = phi{{.+}} [ [[NUM_IT_2]], %[[EUB_TRUE]] ], [ [[OMP_UB_VAL2]], %[[EUB_FALSE]] ]
+// CHECK: store{{.+}} [[EUB_RES]], {{.+}}* [[OMP_UB]],
+
+// initialize omp.iv
+// CHECK: [[OMP_LB_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[OMP_LB]],
+// CHECK: store {{.+}} [[OMP_LB_VAL_1]], {{.+}}* [[OMP_IV]],
+// CHECK: br label %[[OMP_JUMP_BACK:.+]]
+
+// check exit condition
+// CHECK: [[OMP_JUMP_BACK]]:
+// CHECK-DAG: [[OMP_IV_VAL_1:%.+]] = load {{.+}} [[OMP_IV]],
+// CHECK-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_UB]],
+// CHECK: [[CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_IV_VAL_1]], [[OMP_UB_VAL_3]]
+// CHECK: br {{.+}} [[CMP_IV_UB]], label %[[DIST_BODY:.+]], label %[[DIST_END:.+]]
+
+// check that PrevLB and PrevUB are passed to the 'for'
+// CHECK: [[DIST_BODY]]:
+// CHECK-DAG: [[OMP_PREV_LB:%.+]] = load {{.+}}, {{.+}} [[OMP_LB]],
+// CHECK-64-DAG: [[OMP_PREV_LB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_LB]] to {{.+}}
+// CHECK-DAG: [[OMP_PREV_UB:%.+]] = load {{.+}}, {{.+}} [[OMP_UB]],
+// CHECK-64-DAG: [[OMP_PREV_UB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_UB]] to {{.+}}
+// check that distlb and distub are properly passed to fork_call
+// CHECK-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_1:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_EXT]], i{{[0-9]+}} [[OMP_PREV_UB_EXT]], {{.+}})
+// CHECK-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_1:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB]], i{{[0-9]+}} [[OMP_PREV_UB]], {{.+}})
+// CHECK: br label %[[DIST_INC:.+]]
+
+// increment by stride (distInc - 'parallel for' executes the whole chunk) and latch
+// CHECK: [[DIST_INC]]:
+// CHECK-DAG: [[OMP_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_IV]],
+// CHECK-DAG: [[OMP_ST_VAL_1:%.+]] = load {{.+}}, {{.+}}* [[OMP_ST]],
+// CHECK: [[OMP_IV_INC:%.+]] = add{{.+}} [[OMP_IV_VAL_2]], [[OMP_ST_VAL_1]]
+// CHECK: store{{.+}} [[OMP_IV_INC]], {{.+}}* [[OMP_IV]],
+// CHECK: br label %[[OMP_JUMP_BACK]]
+
+// CHECK-DAG: call void @__kmpc_for_static_fini(
+// CHECK: ret
+
+// implementation of 'parallel for'
+// CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_1]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}})
+
+// CHECK-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+
+// initialize lb and ub to PrevLB and PrevUB
+// CHECK-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+// CHECK-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+// CHECK-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+// CHECK-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+// CHECK-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+// CHECK-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+// CHECK-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+// CHECK-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+// CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 34, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]],{{.+}})
+
+// PrevEUB is only used when 'for' has a chunked schedule, otherwise EUB is used
+// In this case we use EUB
+// CHECK-DAG: [[OMP_PF_UB_VAL_1:%.+]] = load{{.+}} [[OMP_PF_UB]],
+// CHECK: [[PF_NUM_IT_1:%.+]] = load{{.+}},
+// CHECK-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_1]], [[PF_NUM_IT_1]]
+// CHECK: br i1 [[PF_CMP_UB_NUM_IT]], label %[[PF_EUB_TRUE:.+]], label %[[PF_EUB_FALSE:.+]]
+// CHECK: [[PF_EUB_TRUE]]:
+// CHECK: [[PF_NUM_IT_2:%.+]] = load{{.+}},
+// CHECK: br label %[[PF_EUB_END:.+]]
+// CHECK-DAG: [[PF_EUB_FALSE]]:
+// CHECK: [[OMP_PF_UB_VAL2:%.+]] = load{{.+}} [[OMP_PF_UB]],
+// CHECK: br label %[[PF_EUB_END]]
+// CHECK-DAG: [[PF_EUB_END]]:
+// CHECK-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_NUM_IT_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL2]], %[[PF_EUB_FALSE]] ]
+// CHECK: store{{.+}} [[PF_EUB_RES]],{{.+}}  [[OMP_PF_UB]],
+
+// initialize omp.iv
+// CHECK: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+// CHECK: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+// CHECK: br label %[[OMP_PF_JUMP_BACK:.+]]
+
+// check exit condition
+// CHECK: [[OMP_PF_JUMP_BACK]]:
+// CHECK-DAG: [[OMP_PF_IV_VAL_1:%.+]] = load {{.+}} [[OMP_PF_IV]],
+// CHECK-DAG: [[OMP_PF_UB_VAL_3:%.+]] = load {{.+}} [[OMP_PF_UB]],
+// CHECK: [[PF_CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_PF_IV_VAL_1]], [[OMP_PF_UB_VAL_3]]
+// CHECK: br {{.+}} [[PF_CMP_IV_UB]], label %[[PF_BODY:.+]], label %[[PF_END:.+]]
+
+// check that PrevLB and PrevUB are passed to the 'for'
+// CHECK: [[PF_BODY]]:
+// CHECK-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+// CHECK: br label {{.+}}
+
+// check stride 1 for 'for' in 'distribute parallel for'
+// CHECK-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_PF_IV]],
+// CHECK: [[OMP_PF_IV_INC:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_2]], 1
+// CHECK: store{{.+}} [[OMP_PF_IV_INC]], {{.+}}* [[OMP_PF_IV]],
+// CHECK: br label %[[OMP_PF_JUMP_BACK]]
+
+// CHECK-DAG: call void @__kmpc_for_static_fini(
+// CHECK: ret
+
+// CHECK: define{{.+}} void [[OFFLOADING_FUN_2]](
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED_2:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} void [[OMP_OUTLINED_2]](
+// CHECK-DAG: [[OMP_IV:%.omp.iv]] = alloca
+// CHECK-DAG: [[OMP_LB:%.omp.lb]] = alloca
+// CHECK-DAG: [[OMP_UB:%.omp.ub]] = alloca
+// CHECK-DAG: [[OMP_ST:%.omp.stride]] = alloca
+
+// CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+
+// check EUB for distribute
+// CHECK-DAG: [[OMP_UB_VAL_1:%.+]] = load{{.+}} [[OMP_UB]],
+// CHECK: [[NUM_IT_1:%.+]] = load{{.+}},
+// CHECK-DAG: [[CMP_UB_NUM_IT:%.+]] = icmp sgt {{.+}}  [[OMP_UB_VAL_1]], [[NUM_IT_1]]
+// CHECK: br {{.+}} [[CMP_UB_NUM_IT]], label %[[EUB_TRUE:.+]], label %[[EUB_FALSE:.+]]
+// CHECK-DAG: [[EUB_TRUE]]:
+// CHECK: [[NUM_IT_2:%.+]] = load{{.+}},
+// CHECK: br label %[[EUB_END:.+]]
+// CHECK-DAG: [[EUB_FALSE]]:
+// CHECK: [[OMP_UB_VAL2:%.+]] = load{{.+}} [[OMP_UB]],
+// CHECK: br label %[[EUB_END]]
+// CHECK-DAG: [[EUB_END]]:
+// CHECK-DAG: [[EUB_RES:%.+]] = phi{{.+}} [ [[NUM_IT_2]], %[[EUB_TRUE]] ], [ [[OMP_UB_VAL2]], %[[EUB_FALSE]] ]
+// CHECK: store{{.+}} [[EUB_RES]], {{.+}}* [[OMP_UB]],
+
+// initialize omp.iv
+// CHECK: [[OMP_LB_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[OMP_LB]],
+// CHECK: store {{.+}} [[OMP_LB_VAL_1]], {{.+}}* [[OMP_IV]],
+// CHECK: br label %[[OMP_JUMP_BACK:.+]]
+
+// check exit condition
+// CHECK: [[OMP_JUMP_BACK]]:
+// CHECK-DAG: [[OMP_IV_VAL_1:%.+]] = load {{.+}} [[OMP_IV]],
+// CHECK-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_UB]],
+// CHECK: [[CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_IV_VAL_1]], [[OMP_UB_VAL_3]]
+// CHECK: br {{.+}} [[CMP_IV_UB]], label %[[DIST_BODY:.+]], label %[[DIST_END:.+]]
+
+// check that PrevLB and PrevUB are passed to the 'for'
+// CHECK: [[DIST_BODY]]:
+// CHECK-DAG: [[OMP_PREV_LB:%.+]] = load {{.+}}, {{.+}} [[OMP_LB]],
+// CHECK-64-DAG: [[OMP_PREV_LB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_LB]] to {{.+}}
+// CHECK-DAG: [[OMP_PREV_UB:%.+]] = load {{.+}}, {{.+}} [[OMP_UB]],
+// CHECK-64-DAG: [[OMP_PREV_UB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_UB]] to {{.+}}
+// check that distlb and distub are properly passed to fork_call
+// CHECK-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_2:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_EXT]], i{{[0-9]+}} [[OMP_PREV_UB_EXT]], {{.+}})
+// CHECK-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_2:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB]], i{{[0-9]+}} [[OMP_PREV_UB]], {{.+}})
+// CHECK: br label %[[DIST_INC:.+]]
+
+// increment by stride (distInc - 'parallel for' executes the whole chunk) and latch
+// CHECK: [[DIST_INC]]:
+// CHECK-DAG: [[OMP_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_IV]],
+// CHECK-DAG: [[OMP_ST_VAL_1:%.+]] = load {{.+}}, {{.+}}* [[OMP_ST]],
+// CHECK: [[OMP_IV_INC:%.+]] = add{{.+}} [[OMP_IV_VAL_2]], [[OMP_ST_VAL_1]]
+// CHECK: store{{.+}} [[OMP_IV_INC]], {{.+}}* [[OMP_IV]],
+// CHECK: br label %[[OMP_JUMP_BACK]]
+
+// CHECK-DAG: call void @__kmpc_for_static_fini(
+// CHECK: ret
+
+// implementation of 'parallel for'
+// CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_2]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}})
+
+// CHECK-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+
+// initialize lb and ub to PrevLB and PrevUB
+// CHECK-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+// CHECK-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+// CHECK-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+// CHECK-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+// CHECK-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+// CHECK-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+// CHECK-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+// CHECK-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+// CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 34, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]],{{.+}})
+
+// PrevEUB is only used when 'for' has a chunked schedule, otherwise EUB is used
+// In this case we use EUB
+// CHECK-DAG: [[OMP_PF_UB_VAL_1:%.+]] = load{{.+}} [[OMP_PF_UB]],
+// CHECK: [[PF_NUM_IT_1:%.+]] = load{{.+}},
+// CHECK-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_1]], [[PF_NUM_IT_1]]
+// CHECK: br i1 [[PF_CMP_UB_NUM_IT]], label %[[PF_EUB_TRUE:.+]], label %[[PF_EUB_FALSE:.+]]
+// CHECK: [[PF_EUB_TRUE]]:
+// CHECK: [[PF_NUM_IT_2:%.+]] = load{{.+}},
+// CHECK: br label %[[PF_EUB_END:.+]]
+// CHECK-DAG: [[PF_EUB_FALSE]]:
+// CHECK: [[OMP_PF_UB_VAL2:%.+]] = load{{.+}} [[OMP_PF_UB]],
+// CHECK: br label %[[PF_EUB_END]]
+// CHECK-DAG: [[PF_EUB_END]]:
+// CHECK-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_NUM_IT_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL2]], %[[PF_EUB_FALSE]] ]
+// CHECK: store{{.+}} [[PF_EUB_RES]],{{.+}}  [[OMP_PF_UB]],
+
+// initialize omp.iv
+// CHECK: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+// CHECK: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+// CHECK: br label %[[OMP_PF_JUMP_BACK:.+]]
+
+// check exit condition
+// CHECK: [[OMP_PF_JUMP_BACK]]:
+// CHECK-DAG: [[OMP_PF_IV_VAL_1:%.+]] = load {{.+}} [[OMP_PF_IV]],
+// CHECK-DAG: [[OMP_PF_UB_VAL_3:%.+]] = load {{.+}} [[OMP_PF_UB]],
+// CHECK: [[PF_CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_PF_IV_VAL_1]], [[OMP_PF_UB_VAL_3]]
+// CHECK: br {{.+}} [[PF_CMP_IV_UB]], label %[[PF_BODY:.+]], label %[[PF_END:.+]]
+
+// check that PrevLB and PrevUB are passed to the 'for'
+// CHECK: [[PF_BODY]]:
+// CHECK-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+// CHECK: br label {{.+}}
+
+// check stride 1 for 'for' in 'distribute parallel for'
+// CHECK-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_PF_IV]],
+// CHECK: [[OMP_PF_IV_INC:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_2]], 1
+// CHECK: store{{.+}} [[OMP_PF_IV_INC]], {{.+}}* [[OMP_PF_IV]],
+// CHECK: br label %[[OMP_PF_JUMP_BACK]]
+
+// CHECK-DAG: call void @__kmpc_for_static_fini(
+// CHECK: ret
+
+// CHECK: define{{.+}} void [[OFFLOADING_FUN_3]](
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 5, {{.+}}* [[OMP_OUTLINED_3:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} void [[OMP_OUTLINED_3]](
+// CHECK-DAG: [[OMP_IV:%.omp.iv]] = alloca
+// CHECK-DAG: [[OMP_LB:%.omp.lb]] = alloca
+// CHECK-DAG: [[OMP_UB:%.omp.ub]] = alloca
+// CHECK-DAG: [[OMP_ST:%.omp.stride]] = alloca
+
+// unlike the previous tests, in this one we have a outer and inner loop for 'distribute'
+// CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 91,
+// CHECK: br label %[[DIST_OUTER_LOOP_HEADER:.+]]
+
+// CHECK: [[DIST_OUTER_LOOP_HEADER]]:
+// check EUB for distribute
+// CHECK-DAG: [[OMP_UB_VAL_1:%.+]] = load{{.+}} [[OMP_UB]],
+// CHECK: [[NUM_IT_1:%.+]] = load{{.+}},
+// CHECK-DAG: [[CMP_UB_NUM_IT:%.+]] = icmp sgt {{.+}}  [[OMP_UB_VAL_1]], [[NUM_IT_1]]
+// CHECK: br {{.+}} [[CMP_UB_NUM_IT]], label %[[EUB_TRUE:.+]], label %[[EUB_FALSE:.+]]
+// CHECK-DAG: [[EUB_TRUE]]:
+// CHECK: [[NUM_IT_2:%.+]] = load{{.+}},
+// CHECK: br label %[[EUB_END:.+]]
+// CHECK-DAG: [[EUB_FALSE]]:
+// CHECK: [[OMP_UB_VAL2:%.+]] = load{{.+}} [[OMP_UB]],
+// CHECK: br label %[[EUB_END]]
+// CHECK-DAG: [[EUB_END]]:
+// CHECK-DAG: [[EUB_RES:%.+]] = phi{{.+}} [ [[NUM_IT_2]], %[[EUB_TRUE]] ], [ [[OMP_UB_VAL2]], %[[EUB_FALSE]] ]
+// CHECK: store{{.+}} [[EUB_RES]], {{.+}}* [[OMP_UB]],
+
+// initialize omp.iv
+// CHECK: [[OMP_LB_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[OMP_LB]],
+// CHECK: store {{.+}} [[OMP_LB_VAL_1]], {{.+}}* [[OMP_IV]],
+
+// check exit condition
+// CHECK-DAG: [[OMP_IV_VAL_1:%.+]] = load {{.+}} [[OMP_IV]],
+// CHECK-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_UB]],
+// CHECK: [[CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_IV_VAL_1]], [[OMP_UB_VAL_3]]
+// CHECK: br {{.+}} [[CMP_IV_UB]], label %[[DIST_OUTER_LOOP_BODY:.+]], label %[[DIST_OUTER_LOOP_END:.+]]
+
+// CHECK: [[DIST_OUTER_LOOP_BODY]]:
+// CHECK: br label %[[DIST_INNER_LOOP_HEADER:.+]]
+
+// CHECK: [[DIST_INNER_LOOP_HEADER]]:
+// CHECK-DAG: [[OMP_IV_VAL_2:%.+]] = load {{.+}} [[OMP_IV]],
+// CHECK-DAG: [[OMP_UB_VAL_4:%.+]] = load {{.+}} [[OMP_UB]],
+// CHECK: [[CMP_IV_UB_2:%.+]] = icmp sle {{.+}} [[OMP_IV_VAL_2]], [[OMP_UB_VAL_4]]
+// CHECK: br{{.+}} [[CMP_IV_UB_2]], label %[[DIST_INNER_LOOP_BODY:.+]], label %[[DIST_INNER_LOOP_END:.+]]
+
+// check that PrevLB and PrevUB are passed to the 'for'
+// CHECK: [[DIST_INNER_LOOP_BODY]]:
+// CHECK-DAG: [[OMP_PREV_LB:%.+]] = load {{.+}}, {{.+}} [[OMP_LB]],
+// CHECK-64-DAG: [[OMP_PREV_LB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_LB]] to {{.+}}
+// CHECK-DAG: [[OMP_PREV_UB:%.+]] = load {{.+}}, {{.+}} [[OMP_UB]],
+// CHECK-64-DAG: [[OMP_PREV_UB_EXT:%.+]] = zext {{.+}} [[OMP_PREV_UB]] to {{.+}}
+// check that distlb and distub are properly passed to fork_call
+// CHECK-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_3:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_EXT]], i{{[0-9]+}} [[OMP_PREV_UB_EXT]], {{.+}})
+// CHECK-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_3:@.+]] to {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB]], i{{[0-9]+}} [[OMP_PREV_UB]], {{.+}})
+// CHECK: br label %[[DIST_INNER_LOOP_INC:.+]]
+
+// check DistInc
+// CHECK: [[DIST_INNER_LOOP_INC]]:
+// CHECK-DAG: [[OMP_IV_VAL_3:%.+]] = load {{.+}}, {{.+}}* [[OMP_IV]],
+// CHECK-DAG: [[OMP_ST_VAL_1:%.+]] = load {{.+}}, {{.+}}* [[OMP_ST]],
+// CHECK: [[OMP_IV_INC:%.+]] = add{{.+}} [[OMP_IV_VAL_3]], [[OMP_ST_VAL_1]]
+// CHECK: store{{.+}} [[OMP_IV_INC]], {{.+}}* [[OMP_IV]],
+// CHECK: br label %[[DIST_INNER_LOOP_HEADER]]
+
+// CHECK: [[DIST_INNER_LOOP_END]]:
+// CHECK: br label %[[DIST_OUTER_LOOP_INC:.+]]
+
+// CHECK: [[DIST_OUTER_LOOP_INC]]:
+// check NextLB and NextUB
+// CHECK-DAG: [[OMP_LB_VAL_2:%.+]] = load{{.+}}, {{.+}} [[OMP_LB]],
+// CHECK-DAG: [[OMP_ST_VAL_2:%.+]] = load{{.+}}, {{.+}} [[OMP_ST]],
+// CHECK-DAG: [[OMP_LB_NEXT:%.+]] = add{{.+}} [[OMP_LB_VAL_2]], [[OMP_ST_VAL_2]]
+// CHECK: store{{.+}} [[OMP_LB_NEXT]], {{.+}}* [[OMP_LB]],
+// CHECK-DAG: [[OMP_UB_VAL_5:%.+]] = load{{.+}}, {{.+}} [[OMP_UB]],
+// CHECK-DAG: [[OMP_ST_VAL_3:%.+]] = load{{.+}}, {{.+}} [[OMP_ST]],
+// CHECK-DAG: [[OMP_UB_NEXT:%.+]] = add{{.+}} [[OMP_UB_VAL_5]], [[OMP_ST_VAL_3]]
+// CHECK: store{{.+}} [[OMP_UB_NEXT]], {{.+}}* [[OMP_UB]],
+// CHECK: br label %[[DIST_OUTER_LOOP_HEADER]]
+
+// outer loop exit
+// CHECK: [[DIST_OUTER_LOOP_END]]:
+// CHECK-DAG: call void @__kmpc_for_static_fini(
+// CHECK: ret
+
+// skip implementation of 'parallel for': using default scheduling and was tested above
+
+// CHECK: define{{.+}} void [[OFFLOADING_FUN_4]](
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED_4:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} void [[OMP_OUTLINED_4]](
+// CHECK-DAG: [[OMP_IV:%.omp.iv]] = alloca
+// CHECK-DAG: [[OMP_LB:%.omp.lb]] = alloca
+// CHECK-DAG: [[OMP_UB:%.omp.ub]] = alloca
+// CHECK-DAG: [[OMP_ST:%.omp.stride]] = alloca
+
+// CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+// CHECK: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_4:@.+]] to {{.+}},
+// skip rest of implementation of 'distribute' as it is tested above for default dist_schedule case
+// CHECK: ret
+
+// 'parallel for' implementation is the same as the case without schedule clase (static no chunk is the default)
+// CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_4]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}})
+
+// CHECK-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+
+// initialize lb and ub to PrevLB and PrevUB
+// CHECK-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+// CHECK-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+// CHECK-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+// CHECK-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+// CHECK-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+// CHECK-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+// CHECK-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+// CHECK-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+// CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 34, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]],{{.+}})
+
+// PrevEUB is only used when 'for' has a chunked schedule, otherwise EUB is used
+// In this case we use EUB
+// CHECK-DAG: [[OMP_PF_UB_VAL_1:%.+]] = load{{.+}} [[OMP_PF_UB]],
+// CHECK: [[PF_NUM_IT_1:%.+]] = load{{.+}},
+// CHECK-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_1]], [[PF_NUM_IT_1]]
+// CHECK: br i1 [[PF_CMP_UB_NUM_IT]], label %[[PF_EUB_TRUE:.+]], label %[[PF_EUB_FALSE:.+]]
+// CHECK: [[PF_EUB_TRUE]]:
+// CHECK: [[PF_NUM_IT_2:%.+]] = load{{.+}},
+// CHECK: br label %[[PF_EUB_END:.+]]
+// CHECK-DAG: [[PF_EUB_FALSE]]:
+// CHECK: [[OMP_PF_UB_VAL2:%.+]] = load{{.+}} [[OMP_PF_UB]],
+// CHECK: br label %[[PF_EUB_END]]
+// CHECK-DAG: [[PF_EUB_END]]:
+// CHECK-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_NUM_IT_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL2]], %[[PF_EUB_FALSE]] ]
+// CHECK: store{{.+}} [[PF_EUB_RES]],{{.+}}  [[OMP_PF_UB]],
+
+// initialize omp.iv
+// CHECK: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+// CHECK: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+// CHECK: br label %[[OMP_PF_JUMP_BACK:.+]]
+
+// check exit condition
+// CHECK: [[OMP_PF_JUMP_BACK]]:
+// CHECK-DAG: [[OMP_PF_IV_VAL_1:%.+]] = load {{.+}} [[OMP_PF_IV]],
+// CHECK-DAG: [[OMP_PF_UB_VAL_3:%.+]] = load {{.+}} [[OMP_PF_UB]],
+// CHECK: [[PF_CMP_IV_UB:%.+]] = icmp sle {{.+}} [[OMP_PF_IV_VAL_1]], [[OMP_PF_UB_VAL_3]]
+// CHECK: br {{.+}} [[PF_CMP_IV_UB]], label %[[PF_BODY:.+]], label %[[PF_END:.+]]
+
+// check that PrevLB and PrevUB are passed to the 'for'
+// CHECK: [[PF_BODY]]:
+// CHECK-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+// CHECK: br label {{.+}}
+
+// check stride 1 for 'for' in 'distribute parallel for'
+// CHECK-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load {{.+}}, {{.+}}* [[OMP_PF_IV]],
+// CHECK: [[OMP_PF_IV_INC:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_2]], 1
+// CHECK: store{{.+}} [[OMP_PF_IV_INC]], {{.+}}* [[OMP_PF_IV]],
+// CHECK: br label %[[OMP_PF_JUMP_BACK]]
+
+// CHECK-DAG: call void @__kmpc_for_static_fini(
+// CHECK: ret
+
+// CHECK: define{{.+}} void [[OFFLOADING_FUN_5]](
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 5, {{.+}}* [[OMP_OUTLINED_5:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} void [[OMP_OUTLINED_5]](
+// CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+// CHECK: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_5:@.+]] to {{.+}},
+// skip rest of implementation of 'distribute' as it is tested above for default dist_schedule case
+// CHECK: ret
+
+// 'parallel for' implementation using outer and inner loops and PrevEUB
+// CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_5]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}}, {{.+}})
+// CHECK-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_ST:%.omp.stride]] = alloca{{.+}},
+
+// initialize lb and ub to PrevLB and PrevUB
+// CHECK-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+// CHECK-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+// CHECK-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+// CHECK-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+// CHECK-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+// CHECK-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+// CHECK-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+// CHECK-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+// CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 33, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]],{{.+}})
+// CHECK: br label %[[OMP_PF_OUTER_LOOP_HEADER:.+]]
+
+// check PrevEUB (using PrevUB instead of NumIt as upper bound)
+// CHECK: [[OMP_PF_OUTER_LOOP_HEADER]]:
+// CHECK-DAG: [[OMP_PF_UB_VAL_1:%.+]] = load{{.+}} [[OMP_PF_UB]],
+// CHECK-64-DAG: [[OMP_PF_UB_VAL_CONV:%.+]] = sext{{.+}} [[OMP_PF_UB_VAL_1]] to
+// CHECK: [[PF_PREV_UB_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+// CHECK-64-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_CONV]], [[PF_PREV_UB_VAL_1]]
+// CHECK-32-DAG: [[PF_CMP_UB_NUM_IT:%.+]] = icmp{{.+}} [[OMP_PF_UB_VAL_1]], [[PF_PREV_UB_VAL_1]]
+// CHECK: br i1 [[PF_CMP_UB_NUM_IT]], label %[[PF_EUB_TRUE:.+]], label %[[PF_EUB_FALSE:.+]]
+// CHECK: [[PF_EUB_TRUE]]:
+// CHECK: [[PF_PREV_UB_VAL_2:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+// CHECK: br label %[[PF_EUB_END:.+]]
+// CHECK-DAG: [[PF_EUB_FALSE]]:
+// CHECK: [[OMP_PF_UB_VAL_2:%.+]] = load{{.+}} [[OMP_PF_UB]],
+// CHECK-64: [[OMP_PF_UB_VAL_2_CONV:%.+]] = sext{{.+}} [[OMP_PF_UB_VAL_2]] to
+// CHECK: br label %[[PF_EUB_END]]
+// CHECK-DAG: [[PF_EUB_END]]:
+// CHECK-64-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_PREV_UB_VAL_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL_2_CONV]], %[[PF_EUB_FALSE]] ]
+// CHECK-32-DAG: [[PF_EUB_RES:%.+]] = phi{{.+}} [ [[PF_PREV_UB_VAL_2]], %[[PF_EUB_TRUE]] ], [ [[OMP_PF_UB_VAL_2]], %[[PF_EUB_FALSE]] ]
+// CHECK-64-DAG: [[PF_EUB_RES_CONV:%.+]] = trunc{{.+}} [[PF_EUB_RES]] to
+// CHECK-64: store{{.+}} [[PF_EUB_RES_CONV]],{{.+}}  [[OMP_PF_UB]],
+// CHECK-32: store{{.+}} [[PF_EUB_RES]], {{.+}} [[OMP_PF_UB]],
+
+// initialize omp.iv (IV = LB)
+// CHECK: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+// CHECK: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+
+// outer loop: while (IV < UB) {
+// CHECK-DAG: [[OMP_PF_IV_VAL_1:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+// CHECK-DAG: [[OMP_PF_UB_VAL_3:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_UB]],
+// CHECK: [[PF_CMP_IV_UB_1:%.+]] = icmp{{.+}} [[OMP_PF_IV_VAL_1]], [[OMP_PF_UB_VAL_3]]
+// CHECK: br{{.+}} [[PF_CMP_IV_UB_1]], label %[[OMP_PF_OUTER_LOOP_BODY:.+]], label %[[OMP_PF_OUTER_LOOP_END:.+]]
+
+// CHECK: [[OMP_PF_OUTER_LOOP_BODY]]:
+// CHECK: br label %[[OMP_PF_INNER_FOR_HEADER:.+]]
+
+// CHECK: [[OMP_PF_INNER_FOR_HEADER]]:
+// CHECK-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+// CHECK-DAG: [[OMP_PF_UB_VAL_4:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_UB]],
+// CHECK: [[PF_CMP_IV_UB_2:%.+]] = icmp{{.+}} [[OMP_PF_IV_VAL_2]], [[OMP_PF_UB_VAL_4]]
+// CHECK: br{{.+}} [[PF_CMP_IV_UB_2]], label %[[OMP_PF_INNER_LOOP_BODY:.+]], label %[[OMP_PF_INNER_LOOP_END:.+]]
+
+// CHECK: [[OMP_PF_INNER_LOOP_BODY]]:
+// CHECK-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+// skip body branch
+// CHECK: br{{.+}}
+// CHECK: br label %[[OMP_PF_INNER_LOOP_INC:.+]]
+
+// IV = IV + 1 and inner loop latch
+// CHECK: [[OMP_PF_INNER_LOOP_INC]]:
+// CHECK-DAG: [[OMP_PF_IV_VAL_3:%.+]] = load{{.+}}, {{.+}}* [[OMP_IV]],
+// CHECK-DAG: [[OMP_PF_NEXT_IV:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_3]], 1
+// CHECK-DAG: store{{.+}} [[OMP_PF_NEXT_IV]], {{.+}}* [[OMP_IV]],
+// CHECK: br label %[[OMP_PF_INNER_FOR_HEADER]]
+
+// check NextLB and NextUB
+// CHECK: [[OMP_PF_INNER_LOOP_END]]:
+// CHECK: br label %[[OMP_PF_OUTER_LOOP_INC:.+]]
+
+// CHECK: [[OMP_PF_OUTER_LOOP_INC]]:
+// CHECK-DAG: [[OMP_PF_LB_VAL_2:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+// CHECK-DAG: [[OMP_PF_ST_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_ST]],
+// CHECK-DAG: [[OMP_PF_LB_NEXT:%.+]] = add{{.+}} [[OMP_PF_LB_VAL_2]], [[OMP_PF_ST_VAL_1]]
+// CHECK: store{{.+}} [[OMP_PF_LB_NEXT]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-DAG: [[OMP_PF_UB_VAL_5:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_UB]],
+// CHECK-DAG: [[OMP_PF_ST_VAL_2:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_ST]],
+// CHECK-DAG: [[OMP_PF_UB_NEXT:%.+]] = add{{.+}} [[OMP_PF_UB_VAL_5]], [[OMP_PF_ST_VAL_2]]
+// CHECK: store{{.+}} [[OMP_PF_UB_NEXT]], {{.+}}* [[OMP_PF_UB]],
+// CHECK: br label %[[OMP_PF_OUTER_LOOP_HEADER]]
+
+// CHECK: [[OMP_PF_OUTER_LOOP_END]]:
+// CHECK-DAG: call void @__kmpc_for_static_fini(
+// CHECK: ret
+
+// CHECK: define{{.+}} void [[OFFLOADING_FUN_6]](
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED_6:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} void [[OMP_OUTLINED_6]](
+// CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+// CHECK: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_6:@.+]] to {{.+}},
+// skip rest of implementation of 'distribute' as it is tested above for default dist_schedule case
+// CHECK: ret
+
+// 'parallel for' implementation using outer and inner loops and PrevEUB
+// CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_6]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}})
+// CHECK-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_ST:%.omp.stride]] = alloca{{.+}},
+
+// initialize lb and ub to PrevLB and PrevUB
+// CHECK-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+// CHECK-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+// CHECK-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+// CHECK-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+// CHECK-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+// CHECK-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+// CHECK-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+// CHECK-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+// CHECK-DAG: [[OMP_PF_LB_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+// CHECK-DAG: [[OMP_PF_UB_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_UB]],
+// CHECK: call void @__kmpc_dispatch_init_4({{.+}}, {{.+}}, {{.+}} 35, {{.+}} [[OMP_PF_LB_VAL]], {{.+}} [[OMP_PF_UB_VAL]], {{.+}}, {{.+}})
+// CHECK: br label %[[OMP_PF_OUTER_LOOP_HEADER:.+]]
+
+// CHECK: [[OMP_PF_OUTER_LOOP_HEADER]]:
+// CHECK: [[IS_FIN:%.+]] = call{{.+}} @__kmpc_dispatch_next_4({{.+}}, {{.+}}, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]], {{.+}}* [[OMP_PF_ST]])
+// CHECK: [[IS_FIN_CMP:%.+]] = icmp{{.+}} [[IS_FIN]], 0
+// CHECK: br{{.+}} [[IS_FIN_CMP]], label %[[OMP_PF_OUTER_LOOP_BODY:.+]], label %[[OMP_PF_OUTER_LOOP_END:.+]]
+
+// initialize omp.iv (IV = LB)
+// CHECK: [[OMP_PF_OUTER_LOOP_BODY]]:
+// CHECK-DAG: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+// CHECK-DAG: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+// CHECK: br label %[[OMP_PF_INNER_LOOP_HEADER:.+]]
+
+// CHECK: [[OMP_PF_INNER_LOOP_HEADER]]:
+// CHECK-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+// CHECK-DAG: [[OMP_PF_UB_VAL_4:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_UB]],
+// CHECK: [[PF_CMP_IV_UB_2:%.+]] = icmp{{.+}} [[OMP_PF_IV_VAL_2]], [[OMP_PF_UB_VAL_4]]
+// CHECK: br{{.+}} [[PF_CMP_IV_UB_2]], label %[[OMP_PF_INNER_LOOP_BODY:.+]], label %[[OMP_PF_INNER_LOOP_END:.+]]
+
+// CHECK: [[OMP_PF_INNER_LOOP_BODY]]:
+// CHECK-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+// skip body branch
+// CHECK: br{{.+}}
+// CHECK: br label %[[OMP_PF_INNER_LOOP_INC:.+]]
+
+// IV = IV + 1 and inner loop latch
+// CHECK: [[OMP_PF_INNER_LOOP_INC]]:
+// CHECK-DAG: [[OMP_PF_IV_VAL_3:%.+]] = load{{.+}}, {{.+}}* [[OMP_IV]],
+// CHECK-DAG: [[OMP_PF_NEXT_IV:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_3]], 1
+// CHECK-DAG: store{{.+}} [[OMP_PF_NEXT_IV]], {{.+}}* [[OMP_IV]],
+// CHECK: br label %[[OMP_PF_INNER_FOR_HEADER]]
+
+// check NextLB and NextUB
+// CHECK: [[OMP_PF_INNER_LOOP_END]]:
+// CHECK: br label %[[OMP_PF_OUTER_LOOP_INC:.+]]
+
+// CHECK: [[OMP_PF_OUTER_LOOP_INC]]:
+// CHECK: br label %[[OMP_PF_OUTER_LOOP_HEADER]]
+
+// CHECK: [[OMP_PF_OUTER_LOOP_END]]:
+// CHECK: ret
+
+// CHECK: define{{.+}} void [[OFFLOADING_FUN_7]](
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 5, {{.+}}* [[OMP_OUTLINED_7:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} void [[OMP_OUTLINED_7]](
+// CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, i32 92,
+// CHECK: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_7:@.+]] to {{.+}},
+// skip rest of implementation of 'distribute' as it is tested above for default dist_schedule case
+// CHECK: ret
+
+// 'parallel for' implementation using outer and inner loops and PrevEUB
+// CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_7]]({{.+}}, {{.+}}, i{{[0-9]+}} [[OMP_PREV_LB_IN:%.+]], i{{[0-9]+}} [[OMP_PREV_UB_IN:%.+]], {{.+}}, {{.+}}, {{.+}}, {{.+}}, {{.+}})
+// CHECK-DAG: [[OMP_PF_LB:%.omp.lb]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_UB:%.omp.ub]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_IV:%.omp.iv]] = alloca{{.+}},
+// CHECK-DAG: [[OMP_PF_ST:%.omp.stride]] = alloca{{.+}},
+
+// initialize lb and ub to PrevLB and PrevUB
+// CHECK-DAG: store{{.+}} [[OMP_PREV_LB_IN]], {{.+}}* [[PREV_LB_ADDR:%.+]],
+// CHECK-DAG: store{{.+}} [[OMP_PREV_UB_IN]], {{.+}}* [[PREV_UB_ADDR:%.+]],
+// CHECK-DAG: [[OMP_PREV_LB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_LB_ADDR]],
+// CHECK-64-DAG: [[OMP_PREV_LB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_LB_VAL]] to {{.+}}
+// CHECK-DAG: [[OMP_PREV_UB_VAL:%.+]] = load{{.+}}, {{.+}}* [[PREV_UB_ADDR]],
+// CHECK-64-DAG: [[OMP_PREV_UB_TRC:%.+]] = trunc{{.+}} [[OMP_PREV_UB_VAL]] to {{.+}}
+// CHECK-64-DAG: store{{.+}} [[OMP_PREV_LB_TRC]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-64-DAG: store{{.+}} [[OMP_PREV_UB_TRC]], {{.+}}* [[OMP_PF_UB]],
+// CHECK-32-DAG: store{{.+}} [[OMP_PREV_LB_VAL]], {{.+}}* [[OMP_PF_LB]],
+// CHECK-32-DAG: store{{.+}} [[OMP_PREV_UB_VAL]], {{.+}}* [[OMP_PF_UB]],
+// CHECK-DAG: [[OMP_PF_LB_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+// CHECK-DAG: [[OMP_PF_UB_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_UB]],
+// CHECK: call void @__kmpc_dispatch_init_4({{.+}}, {{.+}}, {{.+}} 35, {{.+}} [[OMP_PF_LB_VAL]], {{.+}} [[OMP_PF_UB_VAL]], {{.+}}, {{.+}})
+// CHECK: br label %[[OMP_PF_OUTER_LOOP_HEADER:.+]]
+
+// CHECK: [[OMP_PF_OUTER_LOOP_HEADER]]:
+// CHECK: [[IS_FIN:%.+]] = call{{.+}} @__kmpc_dispatch_next_4({{.+}}, {{.+}}, {{.+}}, {{.+}}* [[OMP_PF_LB]], {{.+}}* [[OMP_PF_UB]], {{.+}}* [[OMP_PF_ST]])
+// CHECK: [[IS_FIN_CMP:%.+]] = icmp{{.+}} [[IS_FIN]], 0
+// CHECK: br{{.+}} [[IS_FIN_CMP]], label %[[OMP_PF_OUTER_LOOP_BODY:.+]], label %[[OMP_PF_OUTER_LOOP_END:.+]]
+
+// initialize omp.iv (IV = LB)
+// CHECK: [[OMP_PF_OUTER_LOOP_BODY]]:
+// CHECK-DAG: [[OMP_PF_LB_VAL_1:%.+]] = load{{.+}}, {{.+}} [[OMP_PF_LB]],
+// CHECK-DAG: store {{.+}} [[OMP_PF_LB_VAL_1]], {{.+}}* [[OMP_PF_IV]],
+// CHECK: br label %[[OMP_PF_INNER_LOOP_HEADER:.+]]
+
+// CHECK: [[OMP_PF_INNER_LOOP_HEADER]]:
+// CHECK-DAG: [[OMP_PF_IV_VAL_2:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+// CHECK-DAG: [[OMP_PF_UB_VAL_4:%.+]] = load{{.+}}, {{.+}}* [[OMP_PF_UB]],
+// CHECK: [[PF_CMP_IV_UB_2:%.+]] = icmp{{.+}} [[OMP_PF_IV_VAL_2]], [[OMP_PF_UB_VAL_4]]
+// CHECK: br{{.+}} [[PF_CMP_IV_UB_2]], label %[[OMP_PF_INNER_LOOP_BODY:.+]], label %[[OMP_PF_INNER_LOOP_END:.+]]
+
+// CHECK: [[OMP_PF_INNER_LOOP_BODY]]:
+// CHECK-DAG: {{.+}} = load{{.+}}, {{.+}}* [[OMP_PF_IV]],
+// skip body branch
+// CHECK: br{{.+}}
+// CHECK: br label %[[OMP_PF_INNER_LOOP_INC:.+]]
+
+// IV = IV + 1 and inner loop latch
+// CHECK: [[OMP_PF_INNER_LOOP_INC]]:
+// CHECK-DAG: [[OMP_PF_IV_VAL_3:%.+]] = load{{.+}}, {{.+}}* [[OMP_IV]],
+// CHECK-DAG: [[OMP_PF_NEXT_IV:%.+]] = add{{.+}} [[OMP_PF_IV_VAL_3]], 1
+// CHECK-DAG: store{{.+}} [[OMP_PF_NEXT_IV]], {{.+}}* [[OMP_IV]],
+// CHECK: br label %[[OMP_PF_INNER_FOR_HEADER]]
+
+// check NextLB and NextUB
+// CHECK: [[OMP_PF_INNER_LOOP_END]]:
+// CHECK: br label %[[OMP_PF_OUTER_LOOP_INC:.+]]
+
+// CHECK: [[OMP_PF_OUTER_LOOP_INC]]:
+// CHECK: br label %[[OMP_PF_OUTER_LOOP_HEADER]]
+
+// CHECK: [[OMP_PF_OUTER_LOOP_END]]:
+// CHECK: ret
+#endif

test/OpenMP/distribute_parallel_for_firstprivate_codegen.cpp

+// RxUN: %clang_cc1 -DLAMBDA -verify -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-64
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-64
+// RUN: %clang_cc1 -DLAMBDA -verify -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-32
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-32
+
+// RUN: %clang_cc1  -verify -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-64
+// RUN: %clang_cc1  -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1  -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-64
+// RUN: %clang_cc1  -verify -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-32
+// RUN: %clang_cc1  -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1  -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-32
+// expected-no-diagnostics
+#ifndef HEADER
+#define HEADER
+
+template <class T>
+struct S {
+  T f;
+  S(T a) : f(a) {}
+  S() : f() {}
+  operator T() { return T(); }
+  ~S() {}
+};
+
+// CHECK: [[S_FLOAT_TY:%.+]] = type { float }
+// CHECK: [[S_INT_TY:%.+]] = type { i{{[0-9]+}} }
+template <typename T>
+T tmain() {
+  S<T> test;
+  T t_var = T();
+  T vec[] = {1, 2};
+  S<T> s_arr[] = {1, 2};
+  S<T> &var = test;
+  #pragma omp target
+  #pragma omp teams
+  #pragma omp distribute parallel for firstprivate(t_var, vec, s_arr, s_arr, var, var)
+  for (int i = 0; i < 2; ++i) {
+    vec[i] = t_var;
+    s_arr[i] = var;
+  }
+  return T();
+}
+
+int main() {
+  static int svar;
+  volatile double g;
+  volatile double &g1 = g;
+
+  #ifdef LAMBDA
+  // LAMBDA-LABEL: @main
+  // LAMBDA: call{{.*}} void [[OUTER_LAMBDA:@.+]](
+  [&]() {
+    static float sfvar;
+    // LAMBDA: define{{.*}} internal{{.*}} void [[OUTER_LAMBDA]](
+    // LAMBDA: call i{{[0-9]+}} @__tgt_target_teams(
+    // LAMBDA: call void [[OFFLOADING_FUN:@.+]](
+
+    // LAMBDA: define{{.+}} void [[OFFLOADING_FUN]](
+    // LAMBDA: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED:@.+]] to {{.+}})
+    #pragma omp target
+    #pragma omp teams
+    #pragma omp distribute parallel for firstprivate(g, g1, svar, sfvar)
+    for (int i = 0; i < 2; ++i) {
+      // LAMBDA-64: define{{.*}} internal{{.*}} void [[OMP_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, i{{[0-9]+}} [[G_IN:%.+]], i{{[0-9]+}} [[G1_IN:%.+]], i{{[0-9]+}} [[SVAR_IN:%.+]], i{{[0-9]+}} [[SFVAR_IN:%.+]])
+      // LAMBDA-32: define{{.*}} internal{{.*}} void [[OMP_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, double* {{.+}} [[G_IN:%.+]], i{{[0-9]+}} [[G1_IN:%.+]], i{{[0-9]+}} [[SVAR_IN:%.+]], i{{[0-9]+}} [[SFVAR_IN:%.+]])
+
+      // addr alloca's
+      // LAMBDA-64: [[G_ADDR:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA-32: [[G_ADDR:%.+]] = alloca double*,
+      // LAMBDA: [[G1_ADDR:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[SVAR_ADDR:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[SFVAR_ADDR:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[G1_REF:%.+]] = alloca double*,
+      // LAMBDA: [[TMP:%.+]] = alloca double*,
+
+      // private alloca's
+      // LAMBDA: [[G_PRIV:%.+]] = alloca double,
+      // LAMBDA: [[G1_PRIV:%.+]] = alloca double,
+      // LAMBDA: [[TMP_PRIV:%.+]] = alloca double*,
+      // LAMBDA: [[SVAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[SFVAR_PRIV:%.+]] = alloca float,
+
+      // transfer input parameters into addr alloca's
+      // LAMBDA-DAG: store {{.+}} [[G_IN]], {{.+}} [[G_ADDR]],
+      // LAMBDA-DAG: store {{.+}} [[G1_IN]], {{.+}} [[G1_ADDR]],
+      // LAMBDA-DAG: store {{.+}} [[SVAR_IN]], {{.+}} [[SVAR_ADDR]],
+      // LAMBDA-DAG: store {{.+}} [[SFVAR_IN]], {{.+}} [[SFVAR_ADDR]],
+
+      // init private alloca's with addr alloca's
+      // g
+      // LAMBDA-64-DAG: [[G_CONV:%.+]] = bitcast {{.+}}* [[G_ADDR]] to
+      // LAMBDA-32-DAG: [[G_CONV:%.+]] = load {{.+}}*, {{.+}}** [[G_ADDR]]
+      // LAMBDA-DAG: [[G_ADDR_VAL:%.+]] = load {{.+}}, {{.+}}* [[G_CONV]],
+      // LAMBDA-DAG: store {{.+}} [[G_ADDR_VAL]], {{.+}}* [[G_PRIV]],
+
+      // g1
+      // LAMBDA-DAG: [[G1_CONV:%.+]] = bitcast {{.+}}* [[G1_ADDR]] to
+      // LAMBDA-DAG: store {{.+}}* [[G1_CONV]], {{.+}}** [[G1_REF]],
+      // LAMBDA-DAG: [[G1_REF_VAL:%.+]] = load {{.+}}*, {{.+}}** [[G1_REF]],
+      // LAMBDA-DAG: store {{.+}}* [[G1_REF_VAL]], {{.+}}** [[TMP]],
+      // LAMBDA-DAG: [[TMP_REF:%.+]] = load {{.+}}*, {{.+}}** [[TMP]],
+      // LAMBDA-DAG: [[TMP_VAL:%.+]] = load {{.+}}, {{.+}}* [[TMP_REF]],
+      // LAMBDA-DAG: store {{.+}} [[TMP_VAL]], {{.+}}* [[G1_PRIV]]
+      // LAMBDA-DAG: store {{.+}}* [[G1_PRIV]], {{.+}}** [[TMP_PRIV]],
+
+      // svar
+      // LAMBDA-64-DAG: [[SVAR_CONV:%.+]] = bitcast {{.+}}* [[SVAR_ADDR]] to
+      // LAMBDA-64-DAG: [[SVAR_VAL:%.+]] = load {{.+}}, {{.+}}* [[SVAR_CONV]],
+      // LAMBDA-32-DAG: [[SVAR_VAL:%.+]] = load {{.+}}, {{.+}}* [[SVAR_ADDR]],
+      // LAMBDA-DAG: store {{.+}} [[SVAR_VAL]], {{.+}}* [[SVAR_PRIV]],
+
+      // sfvar
+      // LAMBDA-DAG: [[SFVAR_CONV:%.+]] = bitcast {{.+}}* [[SFVAR_ADDR]] to
+      // LAMBDA-DAG: [[SFVAR_VAL:%.+]] = load {{.+}}, {{.+}}* [[SFVAR_CONV]],
+      // LAMBDA-DAG: store {{.+}} [[SFVAR_VAL]], {{.+}}* [[SFVAR_PRIV]],
+
+      // LAMBDA: call {{.*}}void @__kmpc_for_static_init_4(
+      // pass firstprivate parameters to parallel outlined function
+      // g
+      // LAMBDA-64-DAG: [[G_PRIV_VAL:%.+]] = load {{.+}}, {{.+}}* [[G_PRIV]],
+      // LAMBDA-64: [[G_CAST_CONV:%.+]] = bitcast {{.+}}* [[G_CAST:%.+]] to
+      // LAMBDA-64-DAG: store {{.+}} [[G_PRIV_VAL]], {{.+}}* [[G_CAST_CONV]],
+      // LAMBDA-64-DAG: [[G_PAR:%.+]] = load {{.+}}, {{.+}}* [[G_CAST]],
+
+      // g1
+      // LAMBDA-DAG: [[TMP_PRIV_VAL:%.+]] = load {{.+}}, {{.+}}* [[TMP_PRIV]],
+      // LAMBDA-DAG: [[G1_PRIV_VAL:%.+]] = load {{.+}}, {{.+}}* [[TMP_PRIV_VAL]],
+      // LAMBDA: [[G1_CAST_CONV:%.+]] = bitcast {{.+}}* [[G1_CAST:%.+]] to
+      // LAMBDA-DAG: store {{.+}} [[G1_PRIV_VAL]], {{.+}}* [[G1_CAST_CONV]],
+      // LAMBDA-DAG: [[G1_PAR:%.+]] = load {{.+}}, {{.+}}* [[G1_CAST]],
+
+      // svar
+      // LAMBDA: [[SVAR_VAL:%.+]] = load {{.+}}, {{.+}}* [[SVAR_PRIV]],
+      // LAMBDA-64-DAG: [[SVAR_CAST_CONV:%.+]] = bitcast {{.+}}* [[SVAR_CAST:%.+]] to
+      // LAMBDA-64-DAG: store {{.+}} [[SVAR_VAL]], {{.+}}* [[SVAR_CAST_CONV]],
+      // LAMBDA-32-DAG: store {{.+}} [[SVAR_VAL]], {{.+}}* [[SVAR_CAST:%.+]],
+      // LAMBDA-DAG: [[SVAR_PAR:%.+]] = load {{.+}}, {{.+}}* [[SVAR_CAST]],
+
+      // sfvar
+      // LAMBDA: [[SFVAR_VAL:%.+]] = load {{.+}}, {{.+}}* [[SFVAR_PRIV]],
+      // LAMBDA-DAG: [[SFVAR_CAST_CONV:%.+]] = bitcast {{.+}}* [[SFVAR_CAST:%.+]] to
+      // LAMBDA-DAG: store {{.+}} [[SFVAR_VAL]], {{.+}}* [[SFVAR_CAST_CONV]],
+      // LAMBDA-DAG: [[SFVAR_PAR:%.+]] = load {{.+}}, {{.+}}* [[SFVAR_CAST]],
+
+      // LAMBDA-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED:@.+]] to void ({{.+}})*), {{.+}}, {{.+}}, {{.+}} [[G_PAR]], {{.+}} [[G1_PAR]], {{.+}} [[SVAR_PAR]], {{.+}} [[SFVAR_PAR]])
+      // LAMBDA-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED:@.+]] to void ({{.+}})*), {{.+}}, {{.+}}, {{.+}} [[G_PRIV]], {{.+}} [[G1_PAR]], {{.+}} [[SVAR_PAR]], {{.+}} [[SFVAR_PAR]])
+      // LAMBDA: call {{.*}}void @__kmpc_for_static_fini(
+      // LAMBDA: ret void
+
+      
+      // LAMBDA-64: define{{.+}} void [[OMP_PARFOR_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, {{.+}}, {{.+}}, i{{[0-9]+}} [[G_IN:%.+]], i{{[0-9]+}} [[G1_IN:%.+]], i{{[0-9]+}} [[SVAR_IN:%.+]], i{{[0-9]+}} [[SFVAR_IN:%.+]])
+      // LAMBDA-32: define{{.+}} void [[OMP_PARFOR_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, {{.+}}, {{.+}}, double* {{.+}} [[G_IN:%.+]], i{{[0-9]+}} [[G1_IN:%.+]], i{{[0-9]+}} [[SVAR_IN:%.+]], i{{[0-9]+}} [[SFVAR_IN:%.+]])
+      // skip initial params
+      // LAMBDA: {{.+}} = alloca{{.+}},
+      // LAMBDA: {{.+}} = alloca{{.+}},
+      // LAMBDA: {{.+}} = alloca{{.+}},
+      // LAMBDA: {{.+}} = alloca{{.+}},
+
+      // addr alloca's
+      // LAMBDA-64: [[G_ADDR:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA-32: [[G_ADDR:%.+]] = alloca double*,
+      // LAMBDA: [[G1_ADDR:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[SVAR_ADDR:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[SFVAR_ADDR:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[G1_REF:%.+]] = alloca double*,
+
+      // private alloca's (only for 32-bit)
+      // LAMBDA-32: [[G_PRIV:%.+]] = alloca double,
+      
+      // transfer input parameters into addr alloca's
+      // LAMBDA-DAG: store {{.+}} [[G_IN]], {{.+}} [[G_ADDR]],
+      // LAMBDA-DAG: store {{.+}} [[G1_IN]], {{.+}} [[G1_ADDR]],
+      // LAMBDA-DAG: store {{.+}} [[SVAR_IN]], {{.+}} [[SVAR_ADDR]],
+      // LAMBDA-DAG: store {{.+}} [[SFVAR_IN]], {{.+}} [[SFVAR_ADDR]],
+
+      // prepare parameters for lambda
+      // g
+      // LAMBDA-64-DAG: [[G_CONV:%.+]] = bitcast {{.+}}* [[G_ADDR]] to
+      // LAMBDA-32-DAG: [[G_ADDR_REF:%.+]] = load {{.+}}*, {{.+}}** [[G_ADDR]]
+      // LAMBDA-32-DAG: [[G_ADDR_VAL:%.+]] = load {{.+}}, {{.+}}* [[G_ADDR_REF]],
+      // LAMBDA-32-DAG: store {{.+}} [[G_ADDR_VAL]], {{.+}}* [[G_PRIV]],
+      
+      // g1
+      // LAMBDA-DAG: [[G1_CONV:%.+]] = bitcast {{.+}}* [[G1_ADDR]] to
+      // LAMBDA-DAG: store {{.+}}* [[G1_CONV]], {{.+}}* [[G1_REF]],
+
+      // svar
+      // LAMBDA-64-DAG: [[SVAR_CONV:%.+]] = bitcast {{.+}}* [[SVAR_ADDR]] to
+
+      // sfvar
+      // LAMBDA-DAG: [[SFVAR_CONV:%.+]] = bitcast {{.+}}* [[SFVAR_ADDR]] to
+
+      // LAMBDA: call {{.*}}void @__kmpc_for_static_init_4(
+      g = 1;
+      g1 = 1;
+      svar = 3;
+      sfvar = 4.0;     
+      // LAMBDA-64: store double 1.0{{.+}}, double* [[G_CONV]],
+      // LAMBDA-32: store double 1.0{{.+}}, double* [[G_PRIV]],
+      // LAMBDA: [[G1_REF_REF:%.+]] = load {{.+}}*, {{.+}}** [[G1_REF]],
+      // LAMBDA: store {{.+}} 1.0{{.+}}, {{.+}}* [[G1_REF_REF]],
+      // LAMBDA-64: store {{.+}} 3, {{.+}}* [[SVAR_CONV]],
+      // LAMBDA-32: store {{.+}} 3, {{.+}}* [[SVAR_ADDR]],
+      // LAMBDA: store {{.+}} 4.0{{.+}}, {{.+}}* [[SFVAR_CONV]],
+
+      // pass params to inner lambda
+      // LAMBDA: [[G_PRIVATE_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+      // LAMBDA-64: store double* [[G_CONV]], double** [[G_PRIVATE_ADDR_REF]],
+      // LAMBDA-32: store double* [[G_PRIV]], double** [[G_PRIVATE_ADDR_REF]],
+      // LAMBDA: [[G1_PRIVATE_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
+      // LAMBDA: [[G1_REF_REF:%.+]] = load double*, double** [[G1_REF]],
+      // LAMBDA: store double* [[G1_REF_REF]], double** [[G1_PRIVATE_ADDR_REF]],
+      // LAMBDA: [[SVAR_PRIVATE_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 2
+      // LAMBDA-64: store i{{[0-9]+}}* [[SVAR_CONV]], i{{[0-9]+}}** [[SVAR_PRIVATE_ADDR_REF]]
+      // LAMBDA-32: store i{{[0-9]+}}* [[SVAR_ADDR]], i{{[0-9]+}}** [[SVAR_PRIVATE_ADDR_REF]]
+      // LAMBDA: [[SFVAR_PRIVATE_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 3
+      // LAMBDA: store float* [[SFVAR_CONV]], float** [[SFVAR_PRIVATE_ADDR_REF]]
+      // LAMBDA: call{{.*}} void [[INNER_LAMBDA:@.+]](%{{.+}}* [[ARG]])
+      // LAMBDA: call {{.*}}void @__kmpc_for_static_fini(
+      // LAMBDA: ret void
+      [&]() {
+  // LAMBDA: define {{.+}} void [[INNER_LAMBDA]](%{{.+}}* [[ARG_PTR:%.+]])
+  // LAMBDA: store %{{.+}}* [[ARG_PTR]], %{{.+}}** [[ARG_PTR_REF:%.+]],
+  g = 2;
+  g1 = 2;
+  svar = 4;
+  sfvar = 8.0;
+  // LAMBDA: [[ARG_PTR:%.+]] = load %{{.+}}*, %{{.+}}** [[ARG_PTR_REF]]
+  // LAMBDA: [[G_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+  // LAMBDA: [[G_REF:%.+]] = load double*, double** [[G_PTR_REF]]
+  // LAMBDA: store double 2.0{{.+}}, double* [[G_REF]]
+
+  // LAMBDA: [[TMP_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
+  // LAMBDA: [[G1_REF:%.+]] = load double*, double** [[TMP_PTR_REF]]
+  // LAMBDA: store double 2.0{{.+}}, double* [[G1_REF]],
+  // LAMBDA: [[SVAR_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 2
+  // LAMBDA: [[SVAR_REF:%.+]] = load i{{[0-9]+}}*, i{{[0-9]+}}** [[SVAR_PTR_REF]]
+  // LAMBDA: store i{{[0-9]+}} 4, i{{[0-9]+}}* [[SVAR_REF]]
+  // LAMBDA: [[SFVAR_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 3
+  // LAMBDA: [[SFVAR_REF:%.+]] = load float*, float** [[SFVAR_PTR_REF]]
+  // LAMBDA: store float 8.0{{.+}}, float* [[SFVAR_REF]]
+      }();
+    }
+  }();
+  return 0;
+  #else
+  S<float> test;
+  int t_var = 0;
+  int vec[] = {1, 2};
+  S<float> s_arr[] = {1, 2};
+  S<float> &var = test;
+
+  #pragma omp target
+  #pragma omp teams
+  #pragma omp distribute parallel for firstprivate(t_var, vec, s_arr, s_arr, var, var, svar)
+  for (int i = 0; i < 2; ++i) {
+    vec[i] = t_var;
+    s_arr[i] = var;
+  }
+  return tmain<int>();
+  #endif
+}
+
+// CHECK-LABEL: define{{.*}} i{{[0-9]+}} @main()
+// CHECK: [[TEST:%.+]] = alloca [[S_FLOAT_TY]],
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_CONSTR:@.+]]([[S_FLOAT_TY]]* [[TEST]])
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOAD_FUN_0:@.+]](
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_DESTR:@.+]]([[S_FLOAT_TY]]* [[TEST]])
+
+// CHECK: define{{.+}} [[OFFLOAD_FUN_0]](i{{[0-9]+}} [[T_VAR_IN:%.+]], [2 x i{{[0-9]+}}]* {{.+}} [[VEC_IN:%.+]], [2 x [[S_FLOAT_TY]]]* {{.+}} [[S_ARR_IN:%.+]], [[S_FLOAT_TY]]* {{.+}} [[VAR_IN:%.+]], i{{[0-9]+}} [[SVAR_IN:%.+]])
+// CHECK: call void (%{{.+}}*, i{{[0-9]+}}, void (i{{[0-9]+}}*, i{{[0-9]+}}*, ...)*, ...) @__kmpc_fork_teams(%{{.+}}* @{{.+}}, i{{[0-9]+}} 5, void (i{{[0-9]+}}*, i{{[0-9]+}}*, ...)* bitcast (void (i{{[0-9]+}}*, i{{[0-9]+}}*, i{{[0-9]+}}, [2 x i{{[0-9]+}}]*, [2 x [[S_FLOAT_TY]]]*, [[S_FLOAT_TY]]*, i{{[0-9]+}})* [[OMP_OUTLINED_0:@.+]] to void
+// CHECK: ret
+
+// CHECK: define internal void [[OMP_OUTLINED_0]](i{{[0-9]+}}* noalias [[GTID_ADDR:%.+]], i{{[0-9]+}}* noalias %{{.+}}, i{{[0-9]+}} [[T_VAR_IN:%.+]], [2 x i{{[0-9]+}}]* {{.+}} [[VEC_IN:%.+]], [2 x [[S_FLOAT_TY]]]* {{.+}} [[S_ARR_IN:%.+]], [[S_FLOAT_TY]]* {{.+}} [[VAR_IN:%.+]], i{{[0-9]+}} [[SVAR_IN:%.+]])
+
+// addr alloca's
+// CHECK: [[T_VAR_ADDR:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[VEC_ADDR:%.+]] = alloca [2 x i{{[0-9]+}}]*,
+// CHECK: [[S_ARR_ADDR:%.+]] = alloca [2 x [[S_FLOAT_TY]]]*,
+// CHECK: [[VAR_ADDR:%.+]] = alloca [[S_FLOAT_TY]]*,
+// CHECK: [[SVAR_ADDR:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[TMP:%.+]] = alloca [[S_FLOAT_TY]]*,
+
+// skip loop alloca's
+// CHECK: [[OMP_IV:.omp.iv+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_LB:.omp.lb+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_UB:.omp.ub+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_ST:.omp.stride+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_IS_LAST:.omp.is_last+]] = alloca i{{[0-9]+}},
+
+// private alloca's
+// CHECK: [[T_VAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[VEC_PRIV:%.+]] = alloca [2 x i{{[0-9]+}}],
+// CHECK: [[S_ARR_PRIV:%.+]] = alloca [2 x [[S_FLOAT_TY]]],
+// CHECK: [[VAR_PRIV:%.+]] = alloca [[S_FLOAT_TY]],
+// CHECK: [[TMP_PRIV:%.+]] = alloca [[S_FLOAT_TY]]*,
+// CHECK: [[SVAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+
+// CHECK: store i{{[0-9]+}}* [[GTID_ADDR]], i{{[0-9]+}}** [[GTID_ADDR_REF:%.+]]
+
+// init addr alloca's with input values
+// CHECK-DAG: store {{.+}} [[T_VAR_IN]], {{.+}}* [[T_VAR_ADDR]],
+// CHECK-DAG: store {{.+}} [[VEC_IN]], {{.+}} [[VEC_ADDR]],
+// CHECK-DAG: store {{.+}} [[S_ARR_IN]], {{.+}} [[S_ARR_ADDR]],
+// CHECK-DAG: store {{.+}} [[VAR_IN]], {{.+}} [[VAR_ADDR]],
+// CHECK-DAG: store {{.+}} [[SVAR_IN]], {{.+}} [[SVAR_ADDR]],
+
+// init private alloca's with addr alloca's
+// t-var
+// CHECK-64-DAG: [[T_VAR_CONV:%.+]] = bitcast {{.+}} [[T_VAR_ADDR]] to
+// CHECK-64-DAG: [[T_VAR_ADDR_VAL:%.+]] = load {{.+}}, {{.+}}* [[T_VAR_CONV]],
+// CHECK-32-DAG: [[T_VAR_ADDR_VAL:%.+]] = load {{.+}}, {{.+}}* [[T_VAR_ADDR]],
+// CHECK-DAG: store {{.+}} [[T_VAR_ADDR_VAL]], {{.+}} [[T_VAR_PRIV]],
+
+// vec
+// CHECK-DAG: [[VEC_ADDR_VAL:%.+]] = load {{.+}}*, {{.+}}** [[VEC_ADDR]],
+// CHECK-DAG: [[VEC_PRIV_BCAST:%.+]] = bitcast {{.+}} [[VEC_PRIV]] to
+// CHECK-DAG: [[VEC_ADDR_BCAST:%.+]] = bitcast {{.+}} [[VEC_ADDR_VAL]] to
+// CHECK-DAG: call void @llvm.memcpy{{.+}}({{.+}}* [[VEC_PRIV_BCAST]], {{.+}}* [[VEC_ADDR_BCAST]],
+
+// s_arr
+// CHECK-DAG: [[S_ARR_ADDR_VAL:%.+]] = load {{.+}}*, {{.+}}** [[S_ARR_ADDR]],
+// CHECK-DAG: [[S_ARR_BGN:%.+]] = getelementptr {{.+}}, {{.+}}* [[S_ARR_PRIV]],
+// CHECK-DAG: [[S_ARR_ADDR_BCAST:%.+]] = bitcast {{.+}}* [[S_ARR_ADDR_VAL]] to
+// CHECK-DAG: [[S_ARR_BGN_GEP:%.+]] = getelementptr {{.+}}, {{.+}}* [[S_ARR_BGN]],
+// CHECK-DAG: [[S_ARR_EMPTY:%.+]] = icmp {{.+}} [[S_ARR_BGN]], [[S_ARR_BGN_GEP]]
+// CHECK-DAG: br {{.+}} [[S_ARR_EMPTY]], label %[[CPY_DONE:.+]], label %[[CPY_BODY:.+]]
+// CHECK-DAG: [[CPY_BODY]]:
+// CHECK-DAG: call void @llvm.memcpy{{.+}}(
+// CHECK-DAG: [[CPY_DONE]]:
+
+// var
+// CHECK-DAG: [[TMP_REF:%.+]] = load {{.+}}*, {{.+}}* [[TMP]],
+// CHECK-DAG: [[VAR_PRIV_BCAST:%.+]] = bitcast {{.+}}* [[VAR_PRIV]] to
+// CHECK-DAG: [[TMP_REF_BCAST:%.+]] = bitcast {{.+}}* [[TMP_REF]] to
+// CHECK-DAG: call void @llvm.memcpy.{{.+}}({{.+}}* [[VAR_PRIV_BCAST]], {{.+}}* [[TMP_REF_BCAST]],
+// CHECK-DAG: store {{.+}}* [[VAR_PRIV]], {{.+}}** [[TMP_PRIV]],
+
+// svar
+// CHECK-64-DAG: [[SVAR_CONV:%.+]] = bitcast {{.+}}* [[SVAR_ADDR]] to
+// CHECK-64-DAG: [[SVAR_CONV_VAL:%.+]] = load {{.+}}, {{.+}}* [[SVAR_CONV]],
+// CHECK-32-DAG: [[SVAR_CONV_VAL:%.+]] = load {{.+}}, {{.+}}* [[SVAR_ADDR]],
+// CHECK-DAG: store {{.+}} [[SVAR_CONV_VAL]], {{.+}}* [[SVAR_PRIV]],
+
+// CHECK: call void @__kmpc_for_static_init_4(
+// pass private alloca's to fork
+// CHECK-DAG: [[T_VAR_PRIV_VAL:%.+]] = load {{.+}}, {{.+}}* [[T_VAR_PRIV]],
+// not dag to distinguish with S_VAR_CAST
+// CHECK-64: [[T_VAR_CAST_CONV:%.+]] = bitcast {{.+}}* [[T_VAR_CAST:%.+]] to
+// CHECK-64-DAG: store {{.+}} [[T_VAR_PRIV_VAL]], {{.+}} [[T_VAR_CAST_CONV]],
+// CHECK-32: store {{.+}} [[T_VAR_PRIV_VAL]], {{.+}} [[T_VAR_CAST:%.+]],
+// CHECK-DAG: [[T_VAR_CAST_VAL:%.+]] = load {{.+}}, {{.+}}* [[T_VAR_CAST]],
+// CHECK-DAG: [[TMP_PRIV_VAL:%.+]] = load [[S_FLOAT_TY]]*, [[S_FLOAT_TY]]** [[TMP_PRIV]],
+// CHECK-DAG: [[SVAR_PRIV_VAL:%.+]] = load {{.+}}, {{.+}}* [[SVAR_PRIV]],
+// CHECK-64-DAG: [[SVAR_CAST_CONV:%.+]] = bitcast {{.+}}* [[SVAR_CAST:%.+]] to
+// CHECK-64-DAG: store {{.+}} [[SVAR_PRIV_VAL]], {{.+}}* [[SVAR_CAST_CONV]],
+// CHECK-32-DAG: store {{.+}} [[SVAR_PRIV_VAL]], {{.+}}* [[SVAR_CAST:%.+]],
+// CHECK-DAG: [[SVAR_CAST_VAL:%.+]] = load {{.+}}, {{.+}}* [[SVAR_CAST]],
+// CHECK: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_0:@.+]] to void ({{.+}})*), {{.+}}, {{.+}}, [2 x i{{[0-9]+}}]* [[VEC_PRIV]], i{{[0-9]+}} [[T_VAR_CAST_VAL]], [2 x [[S_FLOAT_TY]]]* [[S_ARR_PRIV]], [[S_FLOAT_TY]]* [[TMP_PRIV_VAL]], i{{[0-9]+}} [[SVAR_CAST_VAL]])
+// CHECK: call void @__kmpc_for_static_fini(
+
+// call destructors: var..
+// CHECK-DAG: call {{.+}} [[S_FLOAT_TY_DEF_DESTR]]([[S_FLOAT_TY]]* [[VAR_PRIV]])
+
+// ..and s_arr
+// CHECK: {{.+}}:
+// CHECK: [[S_ARR_EL_PAST:%.+]] = phi [[S_FLOAT_TY]]*
+// CHECK: [[S_ARR_PRIV_ITEM:%.+]] = getelementptr {{.+}}, {{.+}} [[S_ARR_EL_PAST]],
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_DESTR]]([[S_FLOAT_TY]]* [[S_ARR_PRIV_ITEM]])
+
+// CHECK: ret void
+
+// By OpenMP specifications, 'firstprivate' applies to both distribute and parallel for.
+// However, the support for 'firstprivate' of 'parallel' is only used when 'parallel'
+// is found alone. Therefore we only have one 'firstprivate' support for 'parallel for'
+// in combination
+// CHECK: define internal void [[OMP_PARFOR_OUTLINED_0]]({{.+}}, {{.+}}, {{.+}}, {{.+}}, [2 x i{{[0-9]+}}]* {{.+}} [[VEC_IN:%.+]], i{{[0-9]+}} [[T_VAR_IN:%.+]], [2 x [[S_FLOAT_TY]]]* {{.+}} [[S_ARR_IN:%.+]], [[S_FLOAT_TY]]* {{.+}} [[VAR_IN:%.+]], i{{[0-9]+}} [[SVAR_IN:%.+]])
+
+// addr alloca's
+// CHECK: [[VEC_ADDR:%.+]] = alloca [2 x i{{[0-9]+}}]*,
+// CHECK: [[T_VAR_ADDR:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[S_ARR_ADDR:%.+]] = alloca [2 x [[S_FLOAT_TY]]]*,
+// CHECK: [[VAR_ADDR:%.+]] = alloca [[S_FLOAT_TY]]*,
+// CHECK: [[SVAR_ADDR:%.+]] = alloca i{{[0-9]+}},
+
+// skip loop alloca's
+// CHECK: [[OMP_IV:.omp.iv+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_LB:.omp.lb+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_UB:.omp.ub+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_ST:.omp.stride+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_IS_LAST:.omp.is_last+]] = alloca i{{[0-9]+}},
+
+// private alloca's
+// CHECK: [[VEC_PRIV:%.+]] = alloca [2 x i{{[0-9]+}}],
+// CHECK: [[S_ARR_PRIV:%.+]] = alloca [2 x [[S_FLOAT_TY]]],
+// CHECK: [[VAR_PRIV:%.+]] = alloca [[S_FLOAT_TY]],
+// CHECK: [[TMP_PRIV:%.+]] = alloca [[S_FLOAT_TY]]*,
+
+// CHECK: store i{{[0-9]+}}* [[GTID_ADDR]], i{{[0-9]+}}** [[GTID_ADDR_REF:%.+]]
+
+// init addr alloca's with input values
+// CHECK-DAG: store {{.+}} [[VEC_IN]], {{.+}} [[VEC_ADDR]],
+// CHECK-DAG: store {{.+}} [[T_VAR_IN]], {{.+}}* [[T_VAR_ADDR]],
+// CHECK-DAG: store {{.+}} [[S_ARR_IN]], {{.+}} [[S_ARR_ADDR]],
+// CHECK-DAG: store {{.+}} [[VAR_IN]], {{.+}} [[VAR_ADDR]],
+// CHECK-DAG: store {{.+}} [[SVAR_IN]], {{.+}} [[SVAR_ADDR]],
+
+// init private alloca's with addr alloca's
+// vec
+// CHECK-DAG: [[VEC_ADDR_VAL:%.+]] = load {{.+}}*, {{.+}}** [[VEC_ADDR]],
+// CHECK-DAG: [[VEC_PRIV_BCAST:%.+]] = bitcast {{.+}} [[VEC_PRIV]] to
+// CHECK-DAG: [[VEC_ADDR_BCAST:%.+]] = bitcast {{.+}} [[VEC_ADDR_VAL]] to
+// CHECK-DAG: call void @llvm.memcpy{{.+}}({{.+}}* [[VEC_PRIV_BCAST]], {{.+}}* [[VEC_ADDR_BCAST]],
+
+// s_arr
+// CHECK-DAG: [[S_ARR_ADDR_VAL:%.+]] = load {{.+}}*, {{.+}}** [[S_ARR_ADDR]],
+// CHECK-DAG: [[S_ARR_BGN:%.+]] = getelementptr {{.+}}, {{.+}}* [[S_ARR_PRIV]],
+// CHECK-DAG: [[S_ARR_ADDR_BCAST:%.+]] = bitcast {{.+}}* [[S_ARR_ADDR_VAL]] to
+// CHECK-DAG: [[S_ARR_BGN_GEP:%.+]] = getelementptr {{.+}}, {{.+}}* [[S_ARR_BGN]],
+// CHECK-DAG: [[S_ARR_EMPTY:%.+]] = icmp {{.+}} [[S_ARR_BGN]], [[S_ARR_BGN_GEP]]
+// CHECK-DAG: br {{.+}} [[S_ARR_EMPTY]], label %[[CPY_DONE:.+]], label %[[CPY_BODY:.+]]
+// CHECK-DAG: [[CPY_BODY]]:
+// CHECK-DAG: call void @llvm.memcpy{{.+}}(
+// CHECK-DAG: [[CPY_DONE]]:
+
+// var
+// CHECK-DAG: [[VAR_ADDR_REF:%.+]] = load {{.+}}*, {{.+}}* [[VAR_ADDR]],
+// CHECK-DAG: [[VAR_PRIV_BCAST:%.+]] = bitcast {{.+}}* [[VAR_PRIV]] to
+// CHECK-DAG: [[VAR_ADDR_BCAST:%.+]] = bitcast {{.+}}* [[VAR_ADDR_REF]] to
+// CHECK-DAG: call void @llvm.memcpy.{{.+}}({{.+}}* [[VAR_PRIV_BCAST]], {{.+}}* [[VAR_ADDR_BCAST]],
+// CHECK-DAG: store {{.+}}* [[VAR_PRIV]], {{.+}}** [[TMP_PRIV]],
+
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: call void @__kmpc_for_static_fini(
+
+// call destructors: var..
+// CHECK-DAG: call {{.+}} [[S_FLOAT_TY_DEF_DESTR]]([[S_FLOAT_TY]]* [[VAR_PRIV]])
+
+// ..and s_arr
+// CHECK: {{.+}}:
+// CHECK: [[S_ARR_EL_PAST:%.+]] = phi [[S_FLOAT_TY]]*
+// CHECK: [[S_ARR_PRIV_ITEM:%.+]] = getelementptr {{.+}}, {{.+}} [[S_ARR_EL_PAST]],
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_DESTR]]([[S_FLOAT_TY]]* [[S_ARR_PRIV_ITEM]])
+
+// CHECK: ret void
+
+// template tmain with S_INT_TY
+// CHECK-LABEL: define{{.*}} i{{[0-9]+}} @{{.+}}tmain{{.+}}()
+// CHECK: [[TEST:%.+]] = alloca [[S_INT_TY]],
+// CHECK: call {{.*}} [[S_INT_TY_DEF_CONSTR:@.+]]([[S_INT_TY]]* [[TEST]])
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOAD_FUN_0:@.+]](
+// CHECK: call {{.*}} [[S_INT_TY_DEF_DESTR:@.+]]([[S_INT_TY]]* [[TEST]])
+
+// CHECK: define{{.+}} [[OFFLOAD_FUN_0]](i{{[0-9]+}} [[T_VAR_IN:%.+]], [2 x i{{[0-9]+}}]* {{.+}} [[VEC_IN:%.+]], [2 x [[S_INT_TY]]]* {{.+}} [[S_ARR_IN:%.+]], [[S_INT_TY]]* {{.+}} [[VAR_IN:%.+]])
+// CHECK: call void (%{{.+}}*, i{{[0-9]+}}, void (i{{[0-9]+}}*, i{{[0-9]+}}*, ...)*, ...) @__kmpc_fork_teams(%{{.+}}* @{{.+}}, i{{[0-9]+}} 4, void (i{{[0-9]+}}*, i{{[0-9]+}}*, ...)* bitcast (void (i{{[0-9]+}}*, i{{[0-9]+}}*, i{{[0-9]+}}, [2 x i{{[0-9]+}}]*, [2 x [[S_INT_TY]]]*, [[S_INT_TY]]*)* [[OMP_OUTLINED_0:@.+]] to void
+// CHECK: ret
+
+// CHECK: define internal void [[OMP_OUTLINED_0]](i{{[0-9]+}}* noalias [[GTID_ADDR:%.+]], i{{[0-9]+}}* noalias %{{.+}}, i{{[0-9]+}} [[T_VAR_IN:%.+]], [2 x i{{[0-9]+}}]* {{.+}} [[VEC_IN:%.+]], [2 x [[S_INT_TY]]]* {{.+}} [[S_ARR_IN:%.+]], [[S_INT_TY]]* {{.+}} [[VAR_IN:%.+]])
+
+// addr alloca's
+// CHECK: [[T_VAR_ADDR:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[VEC_ADDR:%.+]] = alloca [2 x i{{[0-9]+}}]*,
+// CHECK: [[S_ARR_ADDR:%.+]] = alloca [2 x [[S_INT_TY]]]*,
+// CHECK: [[VAR_ADDR:%.+]] = alloca [[S_INT_TY]]*,
+// CHECK: [[TMP:%.+]] = alloca [[S_INT_TY]]*,
+
+// skip loop alloca's
+// CHECK: [[OMP_IV:.omp.iv+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_LB:.omp.lb+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_UB:.omp.ub+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_ST:.omp.stride+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_IS_LAST:.omp.is_last+]] = alloca i{{[0-9]+}},
+
+// private alloca's
+// CHECK: [[T_VAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[VEC_PRIV:%.+]] = alloca [2 x i{{[0-9]+}}],
+// CHECK: [[S_ARR_PRIV:%.+]] = alloca [2 x [[S_INT_TY]]],
+// CHECK: [[VAR_PRIV:%.+]] = alloca [[S_INT_TY]],
+// CHECK: [[TMP_PRIV:%.+]] = alloca [[S_INT_TY]]*,
+
+// CHECK: store i{{[0-9]+}}* [[GTID_ADDR]], i{{[0-9]+}}** [[GTID_ADDR_REF:%.+]]
+
+// init addr alloca's with input values
+// CHECK-DAG: store {{.+}} [[T_VAR_IN]], {{.+}}* [[T_VAR_ADDR]],
+// CHECK-DAG: store {{.+}} [[VEC_IN]], {{.+}} [[VEC_ADDR]],
+// CHECK-DAG: store {{.+}} [[S_ARR_IN]], {{.+}} [[S_ARR_ADDR]],
+// CHECK-DAG: store {{.+}} [[VAR_IN]], {{.+}} [[VAR_ADDR]],
+
+// init private alloca's with addr alloca's
+// t-var
+// CHECK-64-DAG: [[T_VAR_CONV:%.+]] = bitcast {{.+}} [[T_VAR_ADDR]] to
+// CHECK-64-DAG: [[T_VAR_ADDR_VAL:%.+]] = load {{.+}}, {{.+}}* [[T_VAR_CONV]],
+// CHECK-32-DAG: [[T_VAR_ADDR_VAL:%.+]] = load {{.+}}, {{.+}}* [[T_VAR_ADDR]],
+// CHECK-DAG: store {{.+}} [[T_VAR_ADDR_VAL]], {{.+}} [[T_VAR_PRIV]],
+
+// vec
+// CHECK-DAG: [[VEC_ADDR_VAL:%.+]] = load {{.+}}*, {{.+}}** [[VEC_ADDR]],
+// CHECK-DAG: [[VEC_PRIV_BCAST:%.+]] = bitcast {{.+}} [[VEC_PRIV]] to
+// CHECK-DAG: [[VEC_ADDR_BCAST:%.+]] = bitcast {{.+}} [[VEC_ADDR_VAL]] to
+// CHECK-DAG: call void @llvm.memcpy{{.+}}({{.+}}* [[VEC_PRIV_BCAST]], {{.+}}* [[VEC_ADDR_BCAST]],
+
+// s_arr
+// CHECK-DAG: [[S_ARR_ADDR_VAL:%.+]] = load {{.+}}*, {{.+}}** [[S_ARR_ADDR]],
+// CHECK-DAG: [[S_ARR_BGN:%.+]] = getelementptr {{.+}}, {{.+}}* [[S_ARR_PRIV]],
+// CHECK-DAG: [[S_ARR_ADDR_BCAST:%.+]] = bitcast {{.+}}* [[S_ARR_ADDR_VAL]] to
+// CHECK-DAG: [[S_ARR_BGN_GEP:%.+]] = getelementptr {{.+}}, {{.+}}* [[S_ARR_BGN]],
+// CHECK-DAG: [[S_ARR_EMPTY:%.+]] = icmp {{.+}} [[S_ARR_BGN]], [[S_ARR_BGN_GEP]]
+// CHECK-DAG: br {{.+}} [[S_ARR_EMPTY]], label %[[CPY_DONE:.+]], label %[[CPY_BODY:.+]]
+// CHECK-DAG: [[CPY_BODY]]:
+// CHECK-DAG: call void @llvm.memcpy{{.+}}(
+// CHECK-DAG: [[CPY_DONE]]:
+
+// var
+// CHECK-DAG: [[TMP_REF:%.+]] = load {{.+}}*, {{.+}}* [[TMP]],
+// CHECK-DAG: [[VAR_PRIV_BCAST:%.+]] = bitcast {{.+}}* [[VAR_PRIV]] to
+// CHECK-DAG: [[TMP_REF_BCAST:%.+]] = bitcast {{.+}}* [[TMP_REF]] to
+// CHECK-DAG: call void @llvm.memcpy.{{.+}}({{.+}}* [[VAR_PRIV_BCAST]], {{.+}}* [[TMP_REF_BCAST]],
+// CHECK-DAG: store {{.+}}* [[VAR_PRIV]], {{.+}}** [[TMP_PRIV]],
+
+// CHECK: call void @__kmpc_for_static_init_4(
+// pass private alloca's to fork
+// CHECK-DAG: [[T_VAR_PRIV_VAL:%.+]] = load {{.+}}, {{.+}}* [[T_VAR_PRIV]],
+// not dag to distinguish with S_VAR_CAST
+// CHECK-64: [[T_VAR_CAST_CONV:%.+]] = bitcast {{.+}}* [[T_VAR_CAST:%.+]] to
+// CHECK-64-DAG: store {{.+}} [[T_VAR_PRIV_VAL]], {{.+}} [[T_VAR_CAST_CONV]],
+// CHECK-32: store {{.+}} [[T_VAR_PRIV_VAL]], {{.+}} [[T_VAR_CAST:%.+]],
+// CHECK-DAG: [[T_VAR_CAST_VAL:%.+]] = load {{.+}}, {{.+}}* [[T_VAR_CAST]],
+// CHECK-DAG: [[TMP_PRIV_VAL:%.+]] = load [[S_INT_TY]]*, [[S_INT_TY]]** [[TMP_PRIV]],
+// CHECK: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_0:@.+]] to void ({{.+}})*), {{.+}}, {{.+}}, [2 x i{{[0-9]+}}]* [[VEC_PRIV]], i{{[0-9]+}} [[T_VAR_CAST_VAL]], [2 x [[S_INT_TY]]]* [[S_ARR_PRIV]], [[S_INT_TY]]* [[TMP_PRIV_VAL]])
+// CHECK: call void @__kmpc_for_static_fini(
+
+// call destructors: var..
+// CHECK-DAG: call {{.+}} [[S_INT_TY_DEF_DESTR]]([[S_INT_TY]]* [[VAR_PRIV]])
+
+// ..and s_arr
+// CHECK: {{.+}}:
+// CHECK: [[S_ARR_EL_PAST:%.+]] = phi [[S_INT_TY]]*
+// CHECK: [[S_ARR_PRIV_ITEM:%.+]] = getelementptr {{.+}}, {{.+}} [[S_ARR_EL_PAST]],
+// CHECK: call {{.*}} [[S_INT_TY_DEF_DESTR]]([[S_INT_TY]]* [[S_ARR_PRIV_ITEM]])
+
+// CHECK: ret void
+
+// By OpenMP specifications, 'firstprivate' applies to both distribute and parallel for.
+// However, the support for 'firstprivate' of 'parallel' is only used when 'parallel'
+// is found alone. Therefore we only have one 'firstprivate' support for 'parallel for'
+// in combination
+// CHECK: define internal void [[OMP_PARFOR_OUTLINED_0]]({{.+}}, {{.+}}, {{.+}}, {{.+}}, [2 x i{{[0-9]+}}]* {{.+}} [[VEC_IN:%.+]], i{{[0-9]+}} [[T_VAR_IN:%.+]], [2 x [[S_INT_TY]]]* {{.+}} [[S_ARR_IN:%.+]], [[S_INT_TY]]* {{.+}} [[VAR_IN:%.+]])
+
+// addr alloca's
+// CHECK: [[VEC_ADDR:%.+]] = alloca [2 x i{{[0-9]+}}]*,
+// CHECK: [[T_VAR_ADDR:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[S_ARR_ADDR:%.+]] = alloca [2 x [[S_INT_TY]]]*,
+// CHECK: [[VAR_ADDR:%.+]] = alloca [[S_INT_TY]]*,
+
+// skip loop alloca's
+// CHECK: [[OMP_IV:.omp.iv+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_LB:.omp.lb+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_UB:.omp.ub+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_ST:.omp.stride+]] = alloca i{{[0-9]+}},
+// CHECK: [[OMP_IS_LAST:.omp.is_last+]] = alloca i{{[0-9]+}},
+
+// private alloca's
+// CHECK: [[VEC_PRIV:%.+]] = alloca [2 x i{{[0-9]+}}],
+// CHECK: [[S_ARR_PRIV:%.+]] = alloca [2 x [[S_INT_TY]]],
+// CHECK: [[VAR_PRIV:%.+]] = alloca [[S_INT_TY]],
+// CHECK: [[TMP_PRIV:%.+]] = alloca [[S_INT_TY]]*,
+
+// CHECK: store i{{[0-9]+}}* [[GTID_ADDR]], i{{[0-9]+}}** [[GTID_ADDR_REF:%.+]]
+
+// init addr alloca's with input values
+// CHECK-DAG: store {{.+}} [[VEC_IN]], {{.+}} [[VEC_ADDR]],
+// CHECK-DAG: store {{.+}} [[T_VAR_IN]], {{.+}}* [[T_VAR_ADDR]],
+// CHECK-DAG: store {{.+}} [[S_ARR_IN]], {{.+}} [[S_ARR_ADDR]],
+// CHECK-DAG: store {{.+}} [[VAR_IN]], {{.+}} [[VAR_ADDR]],
+
+// init private alloca's with addr alloca's
+// vec
+// CHECK-DAG: [[VEC_ADDR_VAL:%.+]] = load {{.+}}*, {{.+}}** [[VEC_ADDR]],
+// CHECK-DAG: [[VEC_PRIV_BCAST:%.+]] = bitcast {{.+}} [[VEC_PRIV]] to
+// CHECK-DAG: [[VEC_ADDR_BCAST:%.+]] = bitcast {{.+}} [[VEC_ADDR_VAL]] to
+// CHECK-DAG: call void @llvm.memcpy{{.+}}({{.+}}* [[VEC_PRIV_BCAST]], {{.+}}* [[VEC_ADDR_BCAST]],
+
+// s_arr
+// CHECK-DAG: [[S_ARR_ADDR_VAL:%.+]] = load {{.+}}*, {{.+}}** [[S_ARR_ADDR]],
+// CHECK-DAG: [[S_ARR_BGN:%.+]] = getelementptr {{.+}}, {{.+}}* [[S_ARR_PRIV]],
+// CHECK-DAG: [[S_ARR_ADDR_BCAST:%.+]] = bitcast {{.+}}* [[S_ARR_ADDR_VAL]] to
+// CHECK-DAG: [[S_ARR_BGN_GEP:%.+]] = getelementptr {{.+}}, {{.+}}* [[S_ARR_BGN]],
+// CHECK-DAG: [[S_ARR_EMPTY:%.+]] = icmp {{.+}} [[S_ARR_BGN]], [[S_ARR_BGN_GEP]]
+// CHECK-DAG: br {{.+}} [[S_ARR_EMPTY]], label %[[CPY_DONE:.+]], label %[[CPY_BODY:.+]]
+// CHECK-DAG: [[CPY_BODY]]:
+// CHECK-DAG: call void @llvm.memcpy{{.+}}(
+// CHECK-DAG: [[CPY_DONE]]:
+
+// var
+// CHECK-DAG: [[VAR_ADDR_REF:%.+]] = load {{.+}}*, {{.+}}* [[VAR_ADDR]],
+// CHECK-DAG: [[VAR_PRIV_BCAST:%.+]] = bitcast {{.+}}* [[VAR_PRIV]] to
+// CHECK-DAG: [[VAR_ADDR_BCAST:%.+]] = bitcast {{.+}}* [[VAR_ADDR_REF]] to
+// CHECK-DAG: call void @llvm.memcpy.{{.+}}({{.+}}* [[VAR_PRIV_BCAST]], {{.+}}* [[VAR_ADDR_BCAST]],
+// CHECK-DAG: store {{.+}}* [[VAR_PRIV]], {{.+}}** [[TMP_PRIV]],
+
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: call void @__kmpc_for_static_fini(
+
+// call destructors: var..
+// CHECK-DAG: call {{.+}} [[S_INT_TY_DEF_DESTR]]([[S_INT_TY]]* [[VAR_PRIV]])
+
+// ..and s_arr
+// CHECK: {{.+}}:
+// CHECK: [[S_ARR_EL_PAST:%.+]] = phi [[S_INT_TY]]*
+// CHECK: [[S_ARR_PRIV_ITEM:%.+]] = getelementptr {{.+}}, {{.+}} [[S_ARR_EL_PAST]],
+// CHECK: call {{.*}} [[S_INT_TY_DEF_DESTR]]([[S_INT_TY]]* [[S_ARR_PRIV_ITEM]])
+
+// CHECK: ret void
+
+#endif

test/OpenMP/distribute_parallel_for_if_codegen.cpp

+// RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple %itanium_abi_triple -emit-llvm %s -o - | FileCheck %s
+// RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -std=c++11 -triple %itanium_abi_triple -emit-pch -o %t %s
+// RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple %itanium_abi_triple -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck --check-prefix=CHECK %s
+// expected-no-diagnostics
+#ifndef HEADER
+#define HEADER
+
+void fn1();
+void fn2();
+void fn3();
+void fn4();
+void fn5();
+void fn6();
+
+int Arg;
+
+// CHECK-LABEL: define {{.*}}void @{{.+}}gtid_test
+void gtid_test() {
+#pragma omp target
+#pragma omp teams
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_0:@.+]](
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_1:@.+]](
+#pragma omp distribute parallel for
+  for(int i = 0 ; i < 100; i++) {}
+  // CHECK: define internal void [[OFFLOADING_FUN_0]](
+  // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 0, {{.+}}* [[OMP_TEAMS_OUTLINED_0:@.+]] to {{.+}})
+  // CHECK: define{{.+}} void [[OMP_TEAMS_OUTLINED_0]](
+  // CHECK: call void @__kmpc_for_static_init_4(
+  // CHECK:  call void {{.+}} @__kmpc_fork_call(%{{.+}}* @{{.+}}, i{{.+}} 2, {{.+}}* [[OMP_OUTLINED_0:@.+]] to void
+  // CHECK: call void @__kmpc_for_static_fini(
+
+  // CHECK: define{{.+}} void [[OMP_OUTLINED_0]](
+  // CHECK: call void @__kmpc_for_static_init_4(
+  // CHECK: call void @__kmpc_for_static_fini(
+  // CHECK: ret
+#pragma omp target
+#pragma omp teams
+#pragma omp distribute parallel for if (parallel: false)
+  for(int i = 0 ; i < 100; i++) {
+  // CHECK: define internal void [[OFFLOADING_FUN_1]](
+  // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 0, {{.+}}* [[OMP_TEAMS_OUTLINED_1:@.+]] to {{.+}})
+  // CHECK: define{{.+}} void [[OMP_TEAMS_OUTLINED_1]](
+  // CHECK: call void @__kmpc_for_static_init_4(
+  // CHECK: call void @__kmpc_serialized_parallel(
+  // CHECK: call void [[OMP_OUTLINED_1:@.+]](
+  // CHECK: call void @__kmpc_end_serialized_parallel(
+  // CHECK: call void @__kmpc_for_static_fini(
+  // CHECK: define{{.+}} void [[OMP_OUTLINED_1]](
+  // CHECK: call void @__kmpc_for_static_init_4(
+  // CHECK: call void @{{.+}}gtid_test
+  // CHECK: call void @__kmpc_for_static_fini(
+  // CHECK: ret
+    gtid_test();
+  }
+}
+
+
+template <typename T>
+int tmain(T Arg) {
+#pragma omp target
+#pragma omp teams
+#pragma omp distribute parallel for if (true)
+  for(int i = 0 ; i < 100; i++) {
+    fn1();
+  }
+#pragma omp target
+#pragma omp teams
+#pragma omp distribute parallel for if (false)
+  for(int i = 0 ; i < 100; i++) {
+    fn2();
+  }
+#pragma omp target
+#pragma omp teams
+#pragma omp distribute parallel for if (parallel: Arg)
+  for(int i = 0 ; i < 100; i++) {   
+    fn3();
+  }
+  return 0;
+}
+
+// CHECK-LABEL: define {{.*}}i{{[0-9]+}} @main()
+int main() {
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_0:@.+]](
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_1:@.+]](
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_2:@.+]](
+// CHECK: = call {{.*}}i{{.+}} @{{.+}}tmain
+#pragma omp target
+#pragma omp teams
+#pragma omp distribute parallel for if (true)
+  for(int i = 0 ; i < 100; i++) {
+    // CHECK: define internal void [[OFFLOADING_FUN_0]](
+    // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 0, {{.+}}* [[OMP_TEAMS_OUTLINED_0:@.+]] to {{.+}})
+    // CHECK: define{{.+}} void [[OMP_TEAMS_OUTLINED_0]](
+
+    // CHECK: call void @__kmpc_for_static_init_4(
+    // CHECK: call void {{.+}} @__kmpc_fork_call(%{{.+}}* @{{.+}}, i{{.+}} 2, {{.+}}* [[OMP_OUTLINED_2:@.+]] to void
+    // CHECK: call void @__kmpc_for_static_fini(
+    // CHECK: define{{.+}} void [[OMP_OUTLINED_2]](
+    // CHECK: call void @__kmpc_for_static_init_4(
+    // CHECK: call {{.*}}void @{{.+}}fn4
+    // CHECK: call void @__kmpc_for_static_fini(
+
+    fn4();
+  }
+
+#pragma omp target
+#pragma omp teams
+#pragma omp distribute parallel for if (false)
+  for(int i = 0 ; i < 100; i++) {   
+    // CHECK: define internal void [[OFFLOADING_FUN_1]](
+    // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 0, {{.+}}* [[OMP_TEAMS_OUTLINED_1:@.+]] to {{.+}})
+    // CHECK: define{{.+}} void [[OMP_TEAMS_OUTLINED_1]](
+
+    // CHECK: call void @__kmpc_for_static_init_4(
+    // CHECK: call void @__kmpc_serialized_parallel(
+    // CHECK: call void [[OMP_OUTLINED_3:@.+]](
+    // CHECK: call void @__kmpc_end_serialized_parallel(
+    // CHECK: call void @__kmpc_for_static_fini(
+
+    // CHECK: define{{.+}} void [[OMP_OUTLINED_3]](
+    // CHECK: call void @__kmpc_for_static_init_4(
+    // CHECK: call {{.*}}void @{{.+}}fn5
+    // CHECK: call void @__kmpc_for_static_fini(
+    fn5();
+  }
+
+#pragma omp target
+#pragma omp teams
+#pragma omp distribute parallel for if (Arg)
+  for(int i = 0 ; i < 100; i++) {
+    // CHECK: define internal void [[OFFLOADING_FUN_2]](
+    // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 1, {{.+}}* [[OMP_TEAMS_OUTLINED_2:@.+]] to {{.+}})
+    // CHECK: define{{.+}} void [[OMP_TEAMS_OUTLINED_2]](
+
+    // CHECK: call void @__kmpc_for_static_init_4(
+    // CHECK: call void {{.+}} @__kmpc_fork_call(%{{.+}}* @{{.+}}, i{{.+}} 2, {{.+}}* [[OMP_OUTLINED_4:@.+]] to void
+    // CHECK: call void @__kmpc_serialized_parallel(
+    // CHECK: call void [[OMP_OUTLINED_4:@.+]](
+    // CHECK: call void @__kmpc_end_serialized_parallel(
+    // CHECK: call void @__kmpc_for_static_fini(
+
+    // CHECK: define{{.+}} void [[OMP_OUTLINED_4]](
+    // CHECK: call void @__kmpc_for_static_init_4(
+    // CHECK: call {{.*}}void @{{.+}}fn6
+    // CHECK: call void @__kmpc_for_static_fini(
+    fn6();
+  }
+
+  return tmain(Arg);
+}
+
+// CHECK-LABEL: define {{.+}} @{{.+}}tmain
+
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_call(%{{.+}}* @{{.+}}, i{{.+}} 2, void {{.+}}* [[T_OUTLINE_FUN_1:@.+]] to void
+// CHECK: call void @__kmpc_for_static_fini(
+
+// CHECK: define internal {{.*}}void [[T_OUTLINE_FUN_1]]
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: call {{.*}}void @{{.+}}fn1
+// CHECK: call void @__kmpc_for_static_fini(
+// CHECK: ret void
+
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: call {{.*}}void @__kmpc_serialized_parallel(
+// CHECK: call void [[T_OUTLINE_FUN_2:@.+]](
+// CHECK: call {{.*}}void @__kmpc_end_serialized_parallel(
+// CHECK: call void @__kmpc_for_static_fini(
+
+// CHECK: define internal {{.*}}void [[T_OUTLINE_FUN_2]]
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: call {{.*}}void @{{.+}}fn2
+// CHECK: call void @__kmpc_for_static_fini(
+// CHECK: ret void
+
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_call(%{{.+}}* @{{.+}}, i{{.+}} 2, void {{.+}}* [[T_OUTLINE_FUN_3:@.+]] to void
+// CHECK: call {{.*}}void @__kmpc_serialized_parallel(
+// call void [[T_OUTLINE_FUN_3:@.+]](
+// CHECK: call {{.*}}void @__kmpc_end_serialized_parallel(
+
+// CHECK: define internal {{.*}}void [[T_OUTLINE_FUN_3]]
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: call {{.*}}void @{{.+}}fn3
+// CHECK: call void @__kmpc_for_static_fini(
+// CHECK: ret void
+#endif

test/OpenMP/distribute_parallel_for_lastprivate_codegen.cpp

+// RUN: %clang_cc1 -DLAMBDA -verify -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-64
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-64
+// RUN: %clang_cc1 -DLAMBDA -verify -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-32
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-32
+
+// RUN: %clang_cc1  -verify -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-64
+// RUN: %clang_cc1  -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1  -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-64
+// RUN: %clang_cc1  -verify -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-32
+// RUN: %clang_cc1  -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1  -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-32
+// expected-no-diagnostics
+#ifndef HEADER
+#define HEADER
+
+template <class T>
+struct S {
+  T f;
+  S(T a) : f(a) {}
+  S() : f() {}
+  operator T() { return T(); }
+  ~S() {}
+};
+
+// CHECK: [[S_FLOAT_TY:%.+]] = type { float }
+// CHECK: [[S_INT_TY:%.+]] = type { i{{[0-9]+}} }
+template <typename T>
+T tmain() {
+  S<T> test;
+  T t_var = T();
+  T vec[] = {1, 2};
+  S<T> s_arr[] = {1, 2};
+  S<T> &var = test;
+  #pragma omp target
+  #pragma omp teams
+#pragma omp distribute parallel for lastprivate(t_var, vec, s_arr, s_arr, var, var)
+  for (int i = 0; i < 2; ++i) {
+    vec[i] = t_var;
+    s_arr[i] = var;
+  }
+  return T();
+}
+
+int main() {
+  static int svar;
+  volatile double g;
+  volatile double &g1 = g;
+
+  #ifdef LAMBDA
+  // LAMBDA-LABEL: @main
+  // LAMBDA: call{{.*}} void [[OUTER_LAMBDA:@.+]](
+  [&]() {
+    static float sfvar;
+    // LAMBDA: define{{.*}} internal{{.*}} void [[OUTER_LAMBDA]](
+    // LAMBDA: call i{{[0-9]+}} @__tgt_target_teams(
+    // LAMBDA: call void [[OFFLOADING_FUN:@.+]](
+
+    // LAMBDA: define{{.+}} void [[OFFLOADING_FUN]](
+    // LAMBDA: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 4, {{.+}}* [[OMP_OUTLINED:@.+]] to {{.+}})
+    #pragma omp target
+    #pragma omp teams
+#pragma omp distribute parallel for lastprivate(g, g1, svar, sfvar)
+    for (int i = 0; i < 2; ++i) {
+      // LAMBDA: define{{.*}} internal{{.*}} void [[OMP_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, double*{{.+}} [[G_IN:%.+]], double*{{.+}} [[G1_IN:%.+]], i{{[0-9]+}}*{{.+}} [[SVAR_IN:%.+]], float*{{.+}} [[SFVAR_IN:%.+]])
+      // LAMBDA: [[G_PRIVATE_ADDR:%.+]] = alloca double*,
+      // LAMBDA: [[G1_PRIVATE_ADDR:%.+]] = alloca double*,
+      // LAMBDA: [[SVAR_PRIVATE_ADDR:%.+]] = alloca i{{[0-9]+}}*,
+      // LAMBDA: [[SFVAR_PRIVATE_ADDR:%.+]] = alloca float*,
+      // LAMBDA: [[TMP_G1:%.+]] = alloca double*,
+      // loop variables
+      // LAMBDA: {{.+}} = alloca i{{[0-9]+}},
+      // LAMBDA: {{.+}} = alloca i{{[0-9]+}},
+      // LAMBDA: {{.+}} = alloca i{{[0-9]+}},
+      // LAMBDA: {{.+}} = alloca i{{[0-9]+}},
+      // LAMBDA: [[OMP_IS_LAST:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[G_PRIVATE:%.+]] = alloca double,
+      // LAMBDA: [[G1_PRIVATE:%.+]] = alloca double,
+      // LAMBDA: [[TMP_G1_PRIVATE:%.+]] = alloca double*,
+      // LAMBDA: [[SVAR_PRIVATE:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[SFVAR_PRIVATE:%.+]] = alloca float,
+
+      // init addr alloca's
+      // LAMBDA: store double* [[G_IN]], double** [[G_PRIVATE_ADDR]],
+      // LAMBDA: store double* [[G1_IN]], double** [[G1_PRIVATE_ADDR]],
+      // LAMBDA: store i{{[0-9]+}}* [[SVAR_IN]], i{{[0-9]+}}** [[SVAR_PRIVATE_ADDR]],
+      // LAMBDA: store float* [[SFVAR_IN]], float** [[SFVAR_PRIVATE_ADDR]],
+
+      // init private variables
+      // LAMBDA: [[G_IN_REF:%.+]] = load double*, double** [[G_PRIVATE_ADDR]],
+      // LAMBDA: [[SVAR_IN_REF:%.+]] = load i{{[0-9]+}}*, i{{[0-9]+}}** [[SVAR_PRIVATE_ADDR]],
+      // LAMBDA: [[SFVAR_IN_REF:%.+]] = load float*, float** [[SFVAR_PRIVATE_ADDR]],
+      // LAMBDA: [[G1_IN_REF:%.+]] = load double*, double** [[G1_PRIVATE_ADDR]],
+      // LAMBDA: store double* [[G1_IN_REF]], double** [[TMP_G1]],
+      // LAMBDA: [[TMP_G1_VAL:%.+]] = load double*, double** [[TMP_G1]],
+      // LAMBDA: store double* [[G1_PRIVATE]], double** [[TMP_G1_PRIVATE]],
+
+      // LAMBDA: call {{.*}}void @__kmpc_for_static_init_4(
+      // LAMBDA: [[G1_PAR:%.+]] = load{{.+}}, {{.+}} [[TMP_G1_PRIVATE]],
+      // LAMBDA-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED:@.+]] to void ({{.+}})*), {{.+}}, {{.+}}, {{.+}} [[G_PRIVATE]], {{.+}} [[G1_PAR]], {{.+}} [[SVAR_PRIVATE]], {{.+}} [[SFVAR_PRIVATE]])
+      // LAMBDA-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED:@.+]] to void ({{.+}})*), {{.+}}, {{.+}}, {{.+}} [[G_PRIVATE]], {{.+}} [[G1_PAR]], {{.+}} [[SVAR_PRIVATE]], {{.+}} [[SFVAR_PRIVATE]])
+      // LAMBDA: call {{.*}}void @__kmpc_for_static_fini(
+
+      // lastprivate
+      // LAMBDA: [[OMP_IS_LAST_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[OMP_IS_LAST]],
+      // LAMBDA: [[IS_LAST_IT:%.+]] = icmp ne i{{[0-9]+}} [[OMP_IS_LAST_VAL]], 0
+      // LAMBDA: br i1 [[IS_LAST_IT]], label %[[OMP_LASTPRIV_BLOCK:.+]], label %[[OMP_LASTPRIV_DONE:.+]]
+
+      // LAMBDA: [[OMP_LASTPRIV_BLOCK]]:
+      // LAMBDA: [[G_PRIV_VAL:%.+]] = load double, double* [[G_PRIVATE]],
+      // LAMBDA: store{{.*}} double [[G_PRIV_VAL]], double* [[G_IN_REF]],
+      // LAMBDA: [[TMP_G1_PRIV_REF:%.+]] = load double*, double** [[TMP_G1_PRIVATE]],
+      // LAMBDA: [[TMP_G1_PRIV_VAL:%.+]] = load double, double* [[TMP_G1_PRIV_REF]],
+      // LAMBDA: store{{.*}} double [[TMP_G1_PRIV_VAL]], double* [[TMP_G1_VAL]],
+
+      // LAMBDA: [[SVAR_PRIV_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[SVAR_PRIVATE]],
+      // LAMBDA: store i{{[0-9]+}} [[SVAR_PRIV_VAL]], i{{[0-9]+}}* [[SVAR_IN_REF]],
+      // LAMBDA: [[SFVAR_PRIV_VAL:%.+]] = load float, float* [[SFVAR_PRIVATE]],
+      // LAMBDA: store float [[SFVAR_PRIV_VAL]], float* [[SFVAR_IN_REF]],
+      // LAMBDA: br label %[[OMP_LASTPRIV_DONE]]
+      // LAMBDA: [[OMP_LASTPRIV_DONE]]:
+      // LAMBDA: ret
+
+      g = 1;
+      g1 = 1;
+      svar = 3;
+      sfvar = 4.0;
+      // outlined function for 'parallel for'
+      // LAMBDA-64: define{{.+}} void [[OMP_PARFOR_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, {{.+}}, {{.+}}, {{.+}} [[G_IN:%.+]], {{.+}} [[G1_IN:%.+]], {{.+}} [[SVAR_IN:%.+]], {{.+}} [[SFVAR_IN:%.+]])
+      // LAMBDA-32: define{{.+}} void [[OMP_PARFOR_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, {{.+}}, {{.+}}, {{.+}} [[G_IN:%.+]], {{.+}} [[G1_IN:%.+]], {{.+}} [[SVAR_IN:%.+]], {{.+}} [[SFVAR_IN:%.+]])
+
+      // addr alloca's
+      // LAMBDA: [[G_PRIVATE_ADDR:%.+]] = alloca double*,
+      // LAMBDA: [[G1_PRIVATE_ADDR:%.+]] = alloca double*,
+      // LAMBDA: [[SVAR_PRIVATE_ADDR:%.+]] = alloca i{{[0-9]+}}*,
+      // LAMBDA: [[SFVAR_PRIVATE_ADDR:%.+]] = alloca float*,
+
+      // loop variables
+      // LAMBDA: {{.+}} = alloca i{{[0-9]+}},
+      // LAMBDA: {{.+}} = alloca i{{[0-9]+}},
+      // LAMBDA: {{.+}} = alloca i{{[0-9]+}},
+      // LAMBDA: {{.+}} = alloca i{{[0-9]+}},
+
+      // private alloca's
+      // LAMBDA: [[OMP_IS_LAST:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[G_PRIVATE:%.+]] = alloca double,
+      // LAMBDA: [[G1_PRIVATE:%.+]] = alloca double,
+      // LAMBDA: [[TMP_G1_PRIVATE:%.+]] = alloca double*,
+      // LAMBDA: [[SVAR_PRIVATE:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[SFVAR_PRIVATE:%.+]] = alloca float,
+
+            // init addr alloca's
+      // LAMBDA: store double* [[G_IN]], double** [[G_PRIVATE_ADDR]],
+      // LAMBDA: store double* [[G1_IN]], double** [[G1_PRIVATE_ADDR]],
+      // LAMBDA: store i{{[0-9]+}}* [[SVAR_IN]], i{{[0-9]+}}** [[SVAR_PRIVATE_ADDR]],
+      // LAMBDA: store float* [[SFVAR_IN]], float** [[SFVAR_PRIVATE_ADDR]],
+
+      // init private variables
+      // LAMBDA: [[G_IN_REF:%.+]] = load double*, double** [[G_PRIVATE_ADDR]],
+      // LAMBDA: [[SVAR_IN_REF:%.+]] = load i{{[0-9]+}}*, i{{[0-9]+}}** [[SVAR_PRIVATE_ADDR]],
+      // LAMBDA: [[SFVAR_IN_REF:%.+]] = load float*, float** [[SFVAR_PRIVATE_ADDR]],
+
+      // LAMBDA: [[G1_IN_REF:%.+]] = load double*, double** [[G1_PRIVATE_ADDR]],
+      // LAMBDA: store double* [[G1_PRIVATE]], double** [[TMP_G1]],
+      
+      // LAMBDA: call {{.*}}void @__kmpc_for_static_init_4(      
+
+      // loop body
+      // LAMBDA: store double 1.0{{.+}}, double* [[G_PRIVATE]],
+      // LAMBDA: [[TMP_G1_REF:%.+]] = load double*, double** [[TMP_G1_PRIVATE]],
+      // LAMBDA: store{{.+}} double 1.0{{.+}}, double* [[TMP_G1_REF]],
+      // LAMBDA: store i{{[0-9]+}} 3, i{{[0-9]+}}* [[SVAR_PRIVATE]],
+      // LAMBDA: store float 4.0{{.+}}, float* [[SFVAR_PRIVATE]],
+      // LAMBDA: [[G_PRIVATE_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+      // LAMBDA: store double* [[G_PRIVATE]], double** [[G_PRIVATE_ADDR_REF]],
+      // LAMBDA: [[TMP_PRIVATE_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
+      // LAMBDA: [[G1_PRIVATE_ADDR_FROM_TMP:%.+]] = load double*, double** [[TMP_G1_PRIVATE]],
+      // LAMBDA: store double* [[G1_PRIVATE_ADDR_FROM_TMP]], double** [[TMP_PRIVATE_ADDR_REF]],
+      // LAMBDA: [[SVAR_PRIVATE_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 2
+      // LAMBDA: store i{{[0-9]+}}* [[SVAR_PRIVATE]], i{{[0-9]+}}** [[SVAR_PRIVATE_ADDR_REF]]
+      // LAMBDA: [[SFVAR_PRIVATE_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 3
+      // LAMBDA: store float* [[SFVAR_PRIVATE]], float** [[SFVAR_PRIVATE_ADDR_REF]]
+      // LAMBDA: call{{.*}} void [[INNER_LAMBDA:@.+]](%{{.+}}* [[ARG]])
+      
+      // LAMBDA: call {{.*}}void @__kmpc_for_static_fini(
+
+      // lastprivate
+      // LAMBDA: [[OMP_IS_LAST_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[OMP_IS_LAST]],
+      // LAMBDA: [[IS_LAST_IT:%.+]] = icmp ne i{{[0-9]+}} [[OMP_IS_LAST_VAL]], 0
+      // LAMBDA: br i1 [[IS_LAST_IT]], label %[[OMP_LASTPRIV_BLOCK:.+]], label %[[OMP_LASTPRIV_DONE:.+]]
+      // LAMBDA: [[OMP_LASTPRIV_BLOCK]]:
+      // LAMBDA: [[G_PRIV_VAL:%.+]] = load double, double* [[G_PRIVATE]],
+      // LAMBDA: store{{.*}} double [[G_PRIV_VAL]], double* [[G_IN_REF]],
+      // LAMBDA: [[TMP_G1_PRIV_REF:%.+]] = load double*, double** [[TMP_G1_PRIVATE]],
+      // LAMBDA: [[TMP_G1_PRIV_VAL:%.+]] = load double, double* [[TMP_G1_PRIV_REF]],
+      // LAMBDA: store{{.*}} double [[TMP_G1_PRIV_VAL]], double* [[G1_IN_REF]],
+      // LAMBDA: [[SVAR_PRIV_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[SVAR_PRIVATE]],
+      // LAMBDA: store i{{[0-9]+}} [[SVAR_PRIV_VAL]], i{{[0-9]+}}* [[SVAR_IN_REF]],
+      // LAMBDA: [[SFVAR_PRIV_VAL:%.+]] = load float, float* [[SFVAR_PRIVATE]],
+      // LAMBDA: store float [[SFVAR_PRIV_VAL]], float* [[SFVAR_IN_REF]],
+      // LAMBDA: br label %[[OMP_LASTPRIV_DONE]]
+      // LAMBDA: [[OMP_LASTPRIV_DONE]]:
+      // LAMBDA: ret
+      
+      [&]() {
+  // LAMBDA: define {{.+}} void [[INNER_LAMBDA]](%{{.+}}* [[ARG_PTR:%.+]])
+  // LAMBDA: store %{{.+}}* [[ARG_PTR]], %{{.+}}** [[ARG_PTR_REF:%.+]],
+  g = 2;
+  g1 = 2;
+  svar = 4;
+  sfvar = 8.0;
+  // LAMBDA: [[ARG_PTR:%.+]] = load %{{.+}}*, %{{.+}}** [[ARG_PTR_REF]]
+  // LAMBDA: [[G_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+  // LAMBDA: [[G_REF:%.+]] = load double*, double** [[G_PTR_REF]]
+  // LAMBDA: store double 2.0{{.+}}, double* [[G_REF]]
+
+  // LAMBDA: [[TMP_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
+  // LAMBDA: [[G1_REF:%.+]] = load double*, double** [[TMP_PTR_REF]]
+  // LAMBDA: store double 2.0{{.+}}, double* [[G1_REF]],
+  // LAMBDA: [[SVAR_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 2
+  // LAMBDA: [[SVAR_REF:%.+]] = load i{{[0-9]+}}*, i{{[0-9]+}}** [[SVAR_PTR_REF]]
+  // LAMBDA: store i{{[0-9]+}} 4, i{{[0-9]+}}* [[SVAR_REF]]
+  // LAMBDA: [[SFVAR_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 3
+  // LAMBDA: [[SFVAR_REF:%.+]] = load float*, float** [[SFVAR_PTR_REF]]
+  // LAMBDA: store float 8.0{{.+}}, float* [[SFVAR_REF]]
+      }();
+    }
+  }();
+  return 0;
+  #else
+  S<float> test;
+  int t_var = 0;
+  int vec[] = {1, 2};
+  S<float> s_arr[] = {1, 2};
+  S<float> &var = test;
+
+  #pragma omp target
+  #pragma omp teams
+#pragma omp distribute parallel for lastprivate(t_var, vec, s_arr, s_arr, var, var, svar)
+  for (int i = 0; i < 2; ++i) {
+    vec[i] = t_var;
+    s_arr[i] = var;
+  }
+  int i;
+
+  return tmain<int>();
+  #endif
+}
+
+// CHECK: define{{.*}} i{{[0-9]+}} @main()
+// CHECK: [[TEST:%.+]] = alloca [[S_FLOAT_TY]],
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_CONSTR:@.+]]([[S_FLOAT_TY]]* [[TEST]])
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOAD_FUN:@.+]](i{{[0-9]+}} {{.+}}, [2 x i{{[0-9]+}}]* {{.+}}, [2 x [[S_FLOAT_TY]]]* {{.+}}, [[S_FLOAT_TY]]* {{.+}}, i{{[0-9]+}} {{.+}})
+// CHECK: ret
+
+// CHECK: define{{.+}} [[OFFLOAD_FUN]](i{{[0-9]+}} {{.+}}, [2 x i{{[0-9]+}}]*{{.+}} {{.+}}, [2 x [[S_FLOAT_TY]]]*{{.+}} {{.+}}, [[S_FLOAT_TY]]*{{.+}} {{.+}}, i{{[0-9]+}} {{.+}})
+// CHECK: call void (%{{.+}}*, i{{[0-9]+}}, void (i{{[0-9]+}}*, i{{[0-9]+}}*, ...)*, ...) @__kmpc_fork_teams(
+// CHECK: ret
+//
+// CHECK: define internal void [[OMP_OUTLINED:@.+]](i{{[0-9]+}}* noalias [[GTID_ADDR:%.+]], i{{[0-9]+}}* noalias %{{.+}}, i{{[0-9]+}}*{{.+}} [[T_VAR_IN:%.+]], [2 x i{{[0-9]+}}]*{{.+}} [[VEC_IN:%.+]], [2 x [[S_FLOAT_TY]]]*{{.+}} [[S_ARR_IN:%.+]], [[S_FLOAT_TY]]*{{.+}} [[VAR_IN:%.+]], i{{[0-9]+}}*{{.*}} [[S_VAR_IN:%.+]])
+// CHECK: {{.+}} = alloca i{{[0-9]+}}*,
+// CHECK: {{.+}} = alloca i{{[0-9]+}}*,
+// CHECK: [[T_VAR_ADDR:%.+]] = alloca i{{[0-9]+}}*,
+// CHECK: [[VEC_ADDR:%.+]] = alloca [2 x i{{[0-9]+}}]*,
+// CHECK: [[S_ARR_ADDR:%.+]] = alloca [2 x [[S_FLOAT_TY]]]*,
+// CHECK: [[VAR_ADDR:%.+]] = alloca [[S_FLOAT_TY]]*,
+// CHECK: [[SVAR_ADDR:%.+]] = alloca i{{[0-9]+}}*,
+// CHECK: [[TMP:%.*]] = alloca [[S_FLOAT_TY]]*,
+// skip loop variables
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: [[OMP_IS_LAST:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[T_VAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[VEC_PRIV:%.+]] = alloca [2 x i{{[0-9]+}}],
+// CHECK: [[S_ARR_PRIV:%.+]] = alloca [2 x [[S_FLOAT_TY]]],
+// CHECK: [[VAR_PRIV:%.+]] = alloca [[S_FLOAT_TY]],
+// CHECK: [[TMP_PRIV:%.+]] = alloca [[S_FLOAT_TY]]*,
+// CHECK: [[S_VAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+
+// copy from parameters to local address variables
+// CHECK: store i{{[0-9]+}}* [[T_VAR_IN]], i{{[0-9]+}}** [[T_VAR_ADDR]],
+// CHECK: store [2 x i{{[0-9]+}}]* [[VEC_IN]], [2 x i{{[0-9]+}}]** [[VEC_ADDR]],
+// CHECK: store [2 x [[S_FLOAT_TY]]]* [[S_ARR_IN]], [2 x [[S_FLOAT_TY]]]** [[S_ARR_ADDR]],
+// CHECK: store [[S_FLOAT_TY]]* [[VAR_IN]], [[S_FLOAT_TY]]** [[VAR_ADDR]],
+// CHECK: store i{{[0-9]+}}* [[S_VAR_IN]], i{{[0-9]+}}** [[SVAR_ADDR]],
+
+// load content of local address variables
+// CHECK: [[T_VAR_ADDR_REF:%.+]] = load i{{[0-9]+}}*, i{{[0-9]+}}** [[T_VAR_ADDR]],
+// CHECK: [[VEC_ADDR_REF:%.+]] = load [2 x i{{[0-9]+}}]*, [2 x i{{[0-9]+}}]** [[VEC_ADDR]],
+// CHECK: [[S_ARR_ADDR_REF:%.+]] = load [2 x [[S_FLOAT_TY]]]*, [2 x [[S_FLOAT_TY]]]** [[S_ARR_ADDR]],
+// CHECK: [[SVAR_ADDR_REF:%.+]] = load i{{[0-9]+}}*, i{{[0-9]+}}** [[SVAR_ADDR]],
+// CHECK: [[VAR_ADDR_REF:%.+]] = load [[S_FLOAT_TY]]*, [[S_FLOAT_TY]]** [[VAR_ADDR]],
+// CHECK: store [[S_FLOAT_TY]]* [[VAR_ADDR_REF]], [[S_FLOAT_TY]]** [[TMP]],
+// CHECK: store i{{[0-9]+}} 0, i{{[0-9]+}}* [[OMP_IS_LAST]],
+
+// call constructor for s_arr
+// CHECK: [[S_ARR_BGN:%.+]] = getelementptr{{.+}} [2 x [[S_FLOAT_TY]]], [2 x [[S_FLOAT_TY]]]* [[S_ARR_PRIV]],
+// CHECK: [[S_ARR_END:%.+]] = getelementptr {{.+}} [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[S_ARR_BGN]],
+// CHECK: br label %[[S_ARR_CST_LOOP:.+]]
+// CHECK: [[S_ARR_CST_LOOP]]:
+// CHECK: [[S_ARR_CTOR:%.+]] = phi {{.+}}
+// CHECK: call void [[S_FLOAT_TY_DEF_CONSTR]]([[S_FLOAT_TY]]* [[S_ARR_CTOR]])
+// CHECK: [[S_ARR_NEXT:%.+]] = getelementptr {{.+}} [[S_ARR_CTOR]],
+// CHECK: [[S_ARR_DONE:%.+]] = icmp {{.+}} [[S_ARR_NEXT]], [[S_ARR_END]]
+// CHECK: br i1 [[S_ARR_DONE]], label %[[S_ARR_CST_END:.+]], label %[[S_ARR_CST_LOOP]]
+// CHECK: [[S_ARR_CST_END]]:
+// CHECK: [[TMP_REF:%.+]] = load [[S_FLOAT_TY]]*, [[S_FLOAT_TY]]** [[TMP]],
+// CHECK: call void [[S_FLOAT_TY_DEF_CONSTR]]([[S_FLOAT_TY]]* [[VAR_PRIV]])
+// CHECK: store [[S_FLOAT_TY]]* [[VAR_PRIV]], [[S_FLOAT_TY]]** [[TMP_PRIV]],
+
+// the distribute loop
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: [[TMP_PRIV_VAL:%.+]] = load {{.+}}, {{.+}} [[TMP_PRIV]],
+// CHECK-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED:@.+]] to void ({{.+}})*), {{.+}}, {{.+}}, {{.+}} [[VEC_PRIV]], {{.+}} [[T_VAR_PRIV]], {{.+}} [[S_ARR_PRIV]], {{.+}} [[TMP_PRIV_VAL]], {{.+}} [[S_VAR_PRIV]])
+// CHECK-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED:@.+]] to void ({{.+}})*), {{.+}}, {{.+}}, {{.+}} [[VEC_PRIV]], {{.+}} [[T_VAR_PRIV]], {{.+}} [[S_ARR_PRIV]], {{.+}} [[TMP_PRIV_VAL]], {{.+}} [[S_VAR_PRIV]])
+
+// CHECK: call void @__kmpc_for_static_fini(
+
+// lastprivates
+// CHECK: [[OMP_IS_LAST_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[OMP_IS_LAST]],
+// CHECK: [[IS_LAST_IT:%.+]] = icmp ne i{{[0-9]+}} [[OMP_IS_LAST_VAL]], 0
+// CHECK: br i1 [[IS_LAST_IT]], label %[[OMP_LASTPRIV_BLOCK:.+]], label %[[OMP_LASTPRIV_DONE:.+]]
+
+// CHECK: [[OMP_LASTPRIV_BLOCK]]:
+// CHECK: [[T_VAR_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[T_VAR_PRIV]],
+// CHECK: store i{{[0-9]+}} [[T_VAR_VAL]], i{{[0-9]+}}* [[T_VAR_ADDR_REF]],
+// CHECK: [[BCAST_VEC_ADDR_REF:%.+]] = bitcast [2 x i{{[0-9]+}}]* [[VEC_ADDR_REF]] to i8*
+// CHECK: [[BCAST_VEC_PRIV:%.+]] = bitcast [2 x i{{[0-9]+}}]* [[VEC_PRIV]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[BCAST_VEC_ADDR_REF]], i8* [[BCAST_VEC_PRIV]],
+// CHECK: [[S_ARR_BEGIN:%.+]] = getelementptr inbounds [2 x [[S_FLOAT_TY]]], [2 x [[S_FLOAT_TY]]]* [[S_ARR_ADDR_REF]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+// CHECK: [[S_ARR_PRIV_BCAST:%.+]] = bitcast [2 x [[S_FLOAT_TY]]]* [[S_ARR_PRIV]] to [[S_FLOAT_TY]]*
+// CHECK: [[S_ARR_BEGIN_GEP:%.+]] = getelementptr [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[S_ARR_BEGIN]], i{{[0-9]+}} 2
+// CHECK: [[S_ARR_IS_EMPTY:%.+]] = icmp eq [[S_FLOAT_TY]]* [[S_ARR_BEGIN]], [[S_ARR_BEGIN_GEP]]
+// CHECK: br i1 [[S_ARR_IS_EMPTY]], label %[[S_ARR_COPY_DONE:.+]], label %[[S_ARR_COPY_BLOCK:.+]]
+// CHECK: [[S_ARR_COPY_BLOCK]]:
+// CHECK: [[S_ARR_SRC_EL:%.+]] = phi [[S_FLOAT_TY]]*{{.+}}
+// CHECK: [[S_ARR_DST_EL:%.+]] = phi [[S_FLOAT_TY]]*{{.+}}
+// CHECK: [[S_ARR_DST_BCAST:%.+]] = bitcast [[S_FLOAT_TY]]* [[S_ARR_DST_EL]] to i8*
+// CHECK: [[S_ARR_SRC_BCAST:%.+]] = bitcast [[S_FLOAT_TY]]* [[S_ARR_SRC_EL]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[S_ARR_DST_BCAST]], i8* [[S_ARR_SRC_BCAST]]{{.+}})
+// CHECK: [[S_ARR_DST_NEXT:%.+]] = getelementptr [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[S_ARR_DST_EL]], i{{[0-9]+}} 1
+// CHECK: [[S_ARR_SRC_NEXT:%.+]] = getelementptr{{.+}}
+// CHECK: [[CPY_IS_FINISHED:%.+]] = icmp eq [[S_FLOAT_TY]]* [[S_ARR_DST_NEXT]], [[S_ARR_BEGIN_GEP]]
+// CHECK: br i1 [[CPY_IS_FINISHED]], label %[[S_ARR_COPY_DONE]], label %[[S_ARR_COPY_BLOCK]]
+// CHECK: [[S_ARR_COPY_DONE]]:
+// CHECK: [[TMP_VAL1:%.+]] = load [[S_FLOAT_TY]]*, [[S_FLOAT_TY]]** [[TMP_PRIV]],
+// CHECK: [[VAR_ADDR_REF_BCAST:%.+]] = bitcast [[S_FLOAT_TY]]* [[TMP_REF]] to i8*
+// CHECK: [[TMP_VAL1_BCAST:%.+]] = bitcast [[S_FLOAT_TY]]* [[TMP_VAL1]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[VAR_ADDR_REF_BCAST]], i8* [[TMP_VAL1_BCAST]],{{.+}})
+// CHECK: [[SVAR_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[S_VAR_PRIV]],
+// CHECK: store i{{[0-9]+}} [[SVAR_VAL]], i{{[0-9]+}}* [[SVAR_ADDR_REF]],
+// CHECK: ret void
+
+// outlined function for 'parallel for'
+// CHECK-64: define{{.+}} void [[OMP_PARFOR_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, {{.+}}, {{.+}}, {{.+}} [[VEC_IN:%.+]], {{.+}} [[T_VAR_IN:%.+]], {{.+}} [[S_ARR_IN:%.+]], {{.+}} [[VAR_IN:%.+]], {{.+}} [[SVAR_IN:%.+]])
+// CHECK-32: define{{.+}} void [[OMP_PARFOR_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, {{.+}}, {{.+}}, {{.+}} [[VEC_IN:%.+]], {{.+}} [[T_VAR_IN:%.+]], {{.+}} [[S_ARR_IN:%.+]], {{.+}} [[VAR_IN:%.+]], {{.+}} [[SVAR_IN:%.+]])
+
+// CHECK: {{.+}} = alloca i{{[0-9]+}}*,
+// CHECK: {{.+}} = alloca i{{[0-9]+}}*,
+// CHECK: [[VEC_ADDR:%.+]] = alloca [2 x i{{[0-9]+}}]*,
+// CHECK: [[T_VAR_ADDR:%.+]] = alloca i{{[0-9]+}}*,
+// CHECK: [[S_ARR_ADDR:%.+]] = alloca [2 x [[S_FLOAT_TY]]]*,
+// CHECK: [[VAR_ADDR:%.+]] = alloca [[S_FLOAT_TY]]*,
+// CHECK: [[SVAR_ADDR:%.+]] = alloca i{{[0-9]+}}*,
+// skip loop variables
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: [[OMP_IS_LAST:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[T_VAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[VEC_PRIV:%.+]] = alloca [2 x i{{[0-9]+}}],
+// CHECK: [[S_ARR_PRIV:%.+]] = alloca [2 x [[S_FLOAT_TY]]],
+// CHECK: [[VAR_PRIV:%.+]] = alloca [[S_FLOAT_TY]],
+// CHECK: [[TMP_PRIV:%.+]] = alloca [[S_FLOAT_TY]]*,
+// CHECK: [[S_VAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+
+// copy from parameters to local address variables
+// CHECK: store [2 x i{{[0-9]+}}]* [[VEC_IN]], [2 x i{{[0-9]+}}]** [[VEC_ADDR]],
+// CHECK: store i{{[0-9]+}}* [[T_VAR_IN]], i{{[0-9]+}}** [[T_VAR_ADDR]],
+// CHECK: store [2 x [[S_FLOAT_TY]]]* [[S_ARR_IN]], [2 x [[S_FLOAT_TY]]]** [[S_ARR_ADDR]],
+// CHECK: store [[S_FLOAT_TY]]* [[VAR_IN]], [[S_FLOAT_TY]]** [[VAR_ADDR]],
+// CHECK: store i{{[0-9]+}}* [[S_VAR_IN]], i{{[0-9]+}}** [[SVAR_ADDR]],
+
+// load content of local address variables
+// CHECK: [[VEC_ADDR_REF:%.+]] = load [2 x i{{[0-9]+}}]*, [2 x i{{[0-9]+}}]** [[VEC_ADDR]],
+// CHECK: [[T_VAR_ADDR_REF:%.+]] = load i{{[0-9]+}}*, i{{[0-9]+}}** [[T_VAR_ADDR]],
+// CHECK: [[S_ARR_ADDR_REF:%.+]] = load [2 x [[S_FLOAT_TY]]]*, [2 x [[S_FLOAT_TY]]]** [[S_ARR_ADDR]],
+// CHECK: [[SVAR_ADDR_REF:%.+]] = load i{{[0-9]+}}*, i{{[0-9]+}}** [[SVAR_ADDR]],
+// CHECK: store i{{[0-9]+}} 0, i{{[0-9]+}}* [[OMP_IS_LAST]],
+
+// call constructor for s_arr
+// CHECK: [[S_ARR_BGN:%.+]] = getelementptr{{.+}} [2 x [[S_FLOAT_TY]]], [2 x [[S_FLOAT_TY]]]* [[S_ARR_PRIV]],
+// CHECK: [[S_ARR_END:%.+]] = getelementptr {{.+}} [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[S_ARR_BGN]],
+// CHECK: br label %[[S_ARR_CST_LOOP:.+]]
+// CHECK: [[S_ARR_CST_LOOP]]:
+// CHECK: [[S_ARR_CTOR:%.+]] = phi {{.+}}
+// CHECK: call void [[S_FLOAT_TY_DEF_CONSTR]]([[S_FLOAT_TY]]* [[S_ARR_CTOR]])
+// CHECK: [[S_ARR_NEXT:%.+]] = getelementptr {{.+}} [[S_ARR_CTOR]],
+// CHECK: [[S_ARR_DONE:%.+]] = icmp {{.+}} [[S_ARR_NEXT]], [[S_ARR_END]]
+// CHECK: br i1 [[S_ARR_DONE]], label %[[S_ARR_CST_END:.+]], label %[[S_ARR_CST_LOOP]]
+// CHECK: [[S_ARR_CST_END]]:
+// CHECK: [[VAR_ADDR_REF:%.+]] = load [[S_FLOAT_TY]]*, [[S_FLOAT_TY]]** [[VAR_ADDR]],
+// CHECK: call void [[S_FLOAT_TY_DEF_CONSTR]]([[S_FLOAT_TY]]* [[VAR_PRIV]])
+// CHECK: store [[S_FLOAT_TY]]* [[VAR_PRIV]], [[S_FLOAT_TY]]** [[TMP_PRIV]],
+
+// CHECK: call void @__kmpc_for_static_init_4(
+
+// loop body
+// assignment: vec[i] = t_var;
+// CHECK: [[T_VAR_PRIV_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[T_VAR_PRIV]],
+// CHECK: [[VEC_PTR:%.+]] = getelementptr inbounds [2 x i{{[0-9]+}}], [2 x i{{[0-9]+}}]* [[VEC_PRIV]], i{{[0-9]+}} 0, i{{[0-9]+}} {{.+}}
+// CHECK:  store i{{[0-9]+}} [[T_VAR_PRIV_VAL]], i{{[0-9]+}}* [[VEC_PTR]],
+
+// assignment: s_arr[i] = var;
+// CHECK-DAG: [[S_ARR_PTR:%.+]] = getelementptr inbounds [2 x [[S_FLOAT_TY]]], [2 x [[S_FLOAT_TY]]]* [[S_ARR_PRIV]],
+// CHECK-DAG: [[TMP_VAL:%.+]] = load [[S_FLOAT_TY]]*, [[S_FLOAT_TY]]** [[TMP_PRIV]],
+// CHECK-DAG: [[S_ARR_PTR_BCAST:%.+]] = bitcast [[S_FLOAT_TY]]* [[S_ARR_PTR]] to i8*
+// CHECK-DAG: [[TMP_VAL_BCAST:%.+]] = bitcast [[S_FLOAT_TY]]* [[TMP_VAL]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[S_ARR_PTR_BCAST]], i8* [[TMP_VAL_BCAST]],
+
+// CHECK: call void @__kmpc_for_static_fini(
+
+// lastprivates
+// CHECK: [[OMP_IS_LAST_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[OMP_IS_LAST]],
+// CHECK: [[IS_LAST_IT:%.+]] = icmp ne i{{[0-9]+}} [[OMP_IS_LAST_VAL]], 0
+// CHECK: br i1 [[IS_LAST_IT]], label %[[OMP_LASTPRIV_BLOCK:.+]], label %[[OMP_LASTPRIV_DONE:.+]]
+
+// CHECK: [[OMP_LASTPRIV_BLOCK]]:
+// CHECK: [[T_VAR_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[T_VAR_PRIV]],
+// CHECK: store i{{[0-9]+}} [[T_VAR_VAL]], i{{[0-9]+}}* [[T_VAR_ADDR_REF]],
+// CHECK: [[BCAST_VEC_ADDR_REF:%.+]] = bitcast [2 x i{{[0-9]+}}]* [[VEC_ADDR_REF]] to i8*
+// CHECK: [[BCAST_VEC_PRIV:%.+]] = bitcast [2 x i{{[0-9]+}}]* [[VEC_PRIV]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[BCAST_VEC_ADDR_REF]], i8* [[BCAST_VEC_PRIV]],
+// CHECK: [[S_ARR_BEGIN:%.+]] = getelementptr inbounds [2 x [[S_FLOAT_TY]]], [2 x [[S_FLOAT_TY]]]* [[S_ARR_ADDR_REF]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+// CHECK: [[S_ARR_PRIV_BCAST:%.+]] = bitcast [2 x [[S_FLOAT_TY]]]* [[S_ARR_PRIV]] to [[S_FLOAT_TY]]*
+// CHECK: [[S_ARR_BEGIN_GEP:%.+]] = getelementptr [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[S_ARR_BEGIN]], i{{[0-9]+}} 2
+// CHECK: [[S_ARR_IS_EMPTY:%.+]] = icmp eq [[S_FLOAT_TY]]* [[S_ARR_BEGIN]], [[S_ARR_BEGIN_GEP]]
+// CHECK: br i1 [[S_ARR_IS_EMPTY]], label %[[S_ARR_COPY_DONE:.+]], label %[[S_ARR_COPY_BLOCK:.+]]
+// CHECK: [[S_ARR_COPY_BLOCK]]:
+// CHECK: [[S_ARR_SRC_EL:%.+]] = phi [[S_FLOAT_TY]]*{{.+}}
+// CHECK: [[S_ARR_DST_EL:%.+]] = phi [[S_FLOAT_TY]]*{{.+}}
+// CHECK: [[S_ARR_DST_BCAST:%.+]] = bitcast [[S_FLOAT_TY]]* [[S_ARR_DST_EL]] to i8*
+// CHECK: [[S_ARR_SRC_BCAST:%.+]] = bitcast [[S_FLOAT_TY]]* [[S_ARR_SRC_EL]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[S_ARR_DST_BCAST]], i8* [[S_ARR_SRC_BCAST]]{{.+}})
+// CHECK: [[S_ARR_DST_NEXT:%.+]] = getelementptr [[S_FLOAT_TY]], [[S_FLOAT_TY]]* [[S_ARR_DST_EL]], i{{[0-9]+}} 1
+// CHECK: [[S_ARR_SRC_NEXT:%.+]] = getelementptr{{.+}}
+// CHECK: [[CPY_IS_FINISHED:%.+]] = icmp eq [[S_FLOAT_TY]]* [[S_ARR_DST_NEXT]], [[S_ARR_BEGIN_GEP]]
+// CHECK: br i1 [[CPY_IS_FINISHED]], label %[[S_ARR_COPY_DONE]], label %[[S_ARR_COPY_BLOCK]]
+// CHECK: [[S_ARR_COPY_DONE]]:
+// CHECK: [[TMP_VAL1:%.+]] = load [[S_FLOAT_TY]]*, [[S_FLOAT_TY]]** [[TMP_PRIV]],
+// CHECK: [[VAR_ADDR_REF_BCAST:%.+]] = bitcast [[S_FLOAT_TY]]* [[VAR_ADDR_REF]] to i8*
+// CHECK: [[TMP_VAL1_BCAST:%.+]] = bitcast [[S_FLOAT_TY]]* [[TMP_VAL1]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[VAR_ADDR_REF_BCAST]], i8* [[TMP_VAL1_BCAST]],{{.+}})
+// CHECK: [[SVAR_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[S_VAR_PRIV]],
+// CHECK: store i{{[0-9]+}} [[SVAR_VAL]], i{{[0-9]+}}* [[SVAR_ADDR_REF]],
+// CHECK: ret void
+
+// template tmain
+// CHECK: define{{.*}} i{{[0-9]+}} [[TMAIN_INT:@.+]]()
+// CHECK: [[TEST:%.+]] = alloca [[S_INT_TY]],
+// CHECK: call {{.*}} [[S_INT_TY_DEF_CONSTR:@.+]]([[S_INT_TY]]* [[TEST]])
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOAD_FUN_1:@.+]](i{{[0-9]+}} {{.+}}, [2 x i{{[0-9]+}}]* {{.+}}, [2 x [[S_INT_TY]]]* {{.+}}, [[S_INT_TY]]* {{.+}})
+// CHECK: ret
+
+// CHECK: define internal void [[OFFLOAD_FUN_1]](
+// CHECK: call void (%{{.+}}*, i{{[0-9]+}}, void (i{{[0-9]+}}*, i{{[0-9]+}}*, ...)*, ...) @__kmpc_fork_teams(%{{.+}}* @{{.+}}, i{{[0-9]+}} 4,
+// CHECK: ret
+
+// CHECK: define internal void [[OMP_OUTLINED_1:@.+]](i{{[0-9]+}}* noalias [[GTID_ADDR:%.+]], i{{[0-9]+}}* noalias %{{.+}}, i{{[0-9]+}}*{{.+}} [[T_VAR_IN:%.+]], [2 x i{{[0-9]+}}]*{{.+}} [[VEC_IN:%.+]],  [2 x [[S_INT_TY]]]*{{.+}} [[S_ARR_IN:%.+]], [[S_INT_TY]]*{{.+}} [[VAR_IN:%.+]])
+// skip alloca of global_tid and bound_tid
+// CHECK: {{.+}} = alloca i{{[0-9]+}}*,
+// CHECK: {{.+}} = alloca i{{[0-9]+}}*,
+// CHECK: [[T_VAR_ADDR:%.+]] = alloca i{{[0-9]+}}*,
+// CHECK: [[VEC_ADDR:%.+]] = alloca [2 x i{{[0-9]+}}]*,
+// CHECK: [[S_ARR_ADDR:%.+]] = alloca [2 x [[S_INT_TY]]]*,
+// CHECK: [[VAR_ADDR:%.+]] = alloca [[S_INT_TY]]*,
+// CHECK: [[TMP:%.+]] = alloca [[S_INT_TY]]*,
+// skip loop variables
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: [[OMP_IS_LAST:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[T_VAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[VEC_PRIV:%.+]] = alloca [2 x i{{[0-9]+}}],
+// CHECK: [[S_ARR_PRIV:%.+]] = alloca [2 x [[S_INT_TY]]],
+// CHECK: [[VAR_PRIV:%.+]] = alloca [[S_INT_TY]],
+// CHECK: [[TMP_PRIV:%.+]] = alloca [[S_INT_TY]]*,
+
+// skip init of bound and global tid
+// CHECK: store i{{[0-9]+}}* {{.*}},
+// CHECK: store i{{[0-9]+}}* {{.*}},
+// copy from parameters to local address variables
+// CHECK: store i{{[0-9]+}}* [[T_VAR_IN]], i{{[0-9]+}}** [[T_VAR_ADDR]],
+// CHECK: store [2 x i{{[0-9]+}}]* [[VEC_IN]], [2 x i{{[0-9]+}}]** [[VEC_ADDR]],
+// CHECK: store [2 x [[S_INT_TY]]]* [[S_ARR_IN]], [2 x [[S_INT_TY]]]** [[S_ARR_ADDR]],
+// CHECK: store [[S_INT_TY]]* [[VAR_IN]], [[S_INT_TY]]** [[VAR_ADDR]],
+
+// load content of local address variables
+// CHECK: [[T_VAR_ADDR_REF:%.+]] = load i{{[0-9]+}}*, i{{[0-9]+}}** [[T_VAR_ADDR]],
+// CHECK: [[VEC_ADDR_REF:%.+]] = load [2 x i{{[0-9]+}}]*, [2 x i{{[0-9]+}}]** [[VEC_ADDR]],
+// CHECK: [[S_ARR_ADDR_REF:%.+]] = load [2 x [[S_INT_TY]]]*, [2 x [[S_INT_TY]]]** [[S_ARR_ADDR]],
+// CHECK: [[VAR_ADDR_REF:%.+]] = load [[S_INT_TY]]*, [[S_INT_TY]]** [[VAR_ADDR]],
+// CHECK-DAG: store [[S_INT_TY]]* [[VAR_ADDR_REF]], [[S_INT_TY]]** [[TMP]],
+// CHECK-DAG: store i{{[0-9]+}} 0, i{{[0-9]+}}* [[OMP_IS_LAST]],
+// CHECK-DAG: [[TMP_REF:%.+]] = load [[S_INT_TY]]*, [[S_INT_TY]]** [[TMP]],
+
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: [[TMP_PRIV_VAL:%.+]] = load {{.+}}, {{.+}} [[TMP_PRIV]],
+// CHECK-64: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED:@.+]] to void ({{.+}})*), {{.+}}, {{.+}}, {{.+}} [[VEC_PRIV]], {{.+}} [[T_VAR_PRIV]], {{.+}} [[S_ARR_PRIV]], {{.+}} [[TMP_PRIV_VAL]])
+// CHECK-32: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED:@.+]] to void ({{.+}})*), {{.+}}, {{.+}}, {{.+}} [[VEC_PRIV]], {{.+}} [[T_VAR_PRIV]], {{.+}} [[S_ARR_PRIV]], {{.+}} [[TMP_PRIV_VAL]])
+
+// CHECK: call void @__kmpc_for_static_fini(
+
+// lastprivates
+// CHECK: [[OMP_IS_LAST_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[OMP_IS_LAST]],
+// CHECK: [[IS_LAST_IT:%.+]] = icmp ne i{{[0-9]+}} [[OMP_IS_LAST_VAL]], 0
+// CHECK: br i1 [[IS_LAST_IT]], label %[[OMP_LASTPRIV_BLOCK:.+]], label %[[OMP_LASTPRIV_DONE:.+]]
+
+// CHECK: [[OMP_LASTPRIV_BLOCK]]:
+// CHECK: [[T_VAR_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[T_VAR_PRIV]],
+// CHECK: store i{{[0-9]+}} [[T_VAR_VAL]], i{{[0-9]+}}* [[T_VAR_ADDR_REF]],
+// CHECK: [[BCAST_VEC_ADDR_REF:%.+]] = bitcast [2 x i{{[0-9]+}}]* [[VEC_ADDR_REF]] to i8*
+// CHECK: [[BCAST_VEC_PRIV:%.+]] = bitcast [2 x i{{[0-9]+}}]* [[VEC_PRIV]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[BCAST_VEC_ADDR_REF]], i8* [[BCAST_VEC_PRIV]],
+// CHECK: [[S_ARR_BEGIN:%.+]] = getelementptr inbounds [2 x [[S_INT_TY]]], [2 x [[S_INT_TY]]]* [[S_ARR_ADDR_REF]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+// CHECK: [[S_ARR_PRIV_BCAST:%.+]] = bitcast [2 x [[S_INT_TY]]]* [[S_ARR_PRIV]] to [[S_INT_TY]]*
+// CHECK: [[S_ARR_BEGIN_GEP:%.+]] = getelementptr [[S_INT_TY]], [[S_INT_TY]]* [[S_ARR_BEGIN]], i{{[0-9]+}} 2
+// CHECK: [[S_ARR_IS_EMPTY:%.+]] = icmp eq [[S_INT_TY]]* [[S_ARR_BEGIN]], [[S_ARR_BEGIN_GEP]]
+// CHECK: br i1 [[S_ARR_IS_EMPTY]], label %[[S_ARR_COPY_DONE:.+]], label %[[S_ARR_COPY_BLOCK:.+]]
+// CHECK: [[S_ARR_COPY_BLOCK]]:
+// CHECK: [[S_ARR_SRC_EL:%.+]] = phi [[S_INT_TY]]*{{.+}}
+// CHECK: [[S_ARR_DST_EL:%.+]] = phi [[S_INT_TY]]*{{.+}}
+// CHECK: [[S_ARR_DST_BCAST:%.+]] = bitcast [[S_INT_TY]]* [[S_ARR_DST_EL]] to i8*
+// CHECK: [[S_ARR_SRC_BCAST:%.+]] = bitcast [[S_INT_TY]]* [[S_ARR_SRC_EL]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[S_ARR_DST_BCAST]], i8* [[S_ARR_SRC_BCAST]]{{.+}})
+// CHECK: [[S_ARR_DST_NEXT:%.+]] = getelementptr [[S_INT_TY]], [[S_INT_TY]]* [[S_ARR_DST_EL]], i{{[0-9]+}} 1
+// CHECK: [[S_ARR_SRC_NEXT:%.+]] = getelementptr{{.+}}
+// CHECK: [[CPY_IS_FINISHED:%.+]] = icmp eq [[S_INT_TY]]* [[S_ARR_DST_NEXT]], [[S_ARR_BEGIN_GEP]]
+// CHECK: br i1 [[CPY_IS_FINISHED]], label %[[S_ARR_COPY_DONE]], label %[[S_ARR_COPY_BLOCK]]
+// CHECK: [[S_ARR_COPY_DONE]]:
+// CHECK: [[TMP_VAL:%.+]] = load [[S_INT_TY]]*, [[S_INT_TY]]** [[TMP_PRIV]],
+// CHECK: [[VAR_ADDR_REF_BCAST:%.+]] = bitcast [[S_INT_TY]]* [[TMP_REF]] to i8*
+// CHECK: [[TMP_VAL_BCAST:%.+]] = bitcast [[S_INT_TY]]* [[TMP_VAL]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[VAR_ADDR_REF_BCAST]], i8* [[TMP_VAL_BCAST]],{{.+}})
+// CHECK: ret void
+
+// outlined function for 'parallel for'
+// CHECK-64: define{{.+}} void [[OMP_PARFOR_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, {{.+}}, {{.+}}, {{.+}} [[VEC_IN:%.+]], {{.+}} [[T_VAR_IN:%.+]], {{.+}} [[S_ARR_IN:%.+]], {{.+}} [[VAR_IN:%.+]])
+// CHECK-32: define{{.+}} void [[OMP_PARFOR_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, {{.+}}, {{.+}}, {{.+}} [[VEC_IN:%.+]], {{.+}} [[T_VAR_IN:%.+]], {{.+}} [[S_ARR_IN:%.+]], {{.+}} [[VAR_IN:%.+]])
+
+// CHECK: {{.+}} = alloca i{{[0-9]+}}*,
+// CHECK: {{.+}} = alloca i{{[0-9]+}}*,
+// CHECK: [[VEC_ADDR:%.+]] = alloca [2 x i{{[0-9]+}}]*,
+// CHECK: [[T_VAR_ADDR:%.+]] = alloca i{{[0-9]+}}*,
+// CHECK: [[S_ARR_ADDR:%.+]] = alloca [2 x [[S_INT_TY]]]*,
+// CHECK: [[VAR_ADDR:%.+]] = alloca [[S_INT_TY]]*,
+// skip loop variables
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: {{.+}} = alloca i{{[0-9]+}},
+// CHECK: [[OMP_IS_LAST:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[T_VAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[VEC_PRIV:%.+]] = alloca [2 x i{{[0-9]+}}],
+// CHECK: [[S_ARR_PRIV:%.+]] = alloca [2 x [[S_INT_TY]]],
+// CHECK: [[VAR_PRIV:%.+]] = alloca [[S_INT_TY]],
+// CHECK: [[TMP_PRIV:%.+]] = alloca [[S_INT_TY]]*,
+
+// copy from parameters to local address variables
+// CHECK: store [2 x i{{[0-9]+}}]* [[VEC_IN]], [2 x i{{[0-9]+}}]** [[VEC_ADDR]],
+// CHECK: store i{{[0-9]+}}* [[T_VAR_IN]], i{{[0-9]+}}** [[T_VAR_ADDR]],
+// CHECK: store [2 x [[S_INT_TY]]]* [[S_ARR_IN]], [2 x [[S_INT_TY]]]** [[S_ARR_ADDR]],
+// CHECK: store [[S_INT_TY]]* [[VAR_IN]], [[S_INT_TY]]** [[VAR_ADDR]],
+
+// load content of local address variables
+// CHECK: [[VEC_ADDR_REF:%.+]] = load [2 x i{{[0-9]+}}]*, [2 x i{{[0-9]+}}]** [[VEC_ADDR]],
+// CHECK: [[T_VAR_ADDR_REF:%.+]] = load i{{[0-9]+}}*, i{{[0-9]+}}** [[T_VAR_ADDR]],
+// CHECK: [[S_ARR_ADDR_REF:%.+]] = load [2 x [[S_INT_TY]]]*, [2 x [[S_INT_TY]]]** [[S_ARR_ADDR]],
+// CHECK: store i{{[0-9]+}} 0, i{{[0-9]+}}* [[OMP_IS_LAST]],
+
+// call constructor for s_arr
+// CHECK: [[S_ARR_BGN:%.+]] = getelementptr{{.+}} [2 x [[S_INT_TY]]], [2 x [[S_INT_TY]]]* [[S_ARR_PRIV]],
+// CHECK: [[S_ARR_END:%.+]] = getelementptr {{.+}} [[S_INT_TY]], [[S_INT_TY]]* [[S_ARR_BGN]],
+// CHECK: br label %[[S_ARR_CST_LOOP:.+]]
+// CHECK: [[S_ARR_CST_LOOP]]:
+// CHECK: [[S_ARR_CTOR:%.+]] = phi {{.+}}
+// CHECK: call void [[S_INT_TY_DEF_CONSTR]]([[S_INT_TY]]* [[S_ARR_CTOR]])
+// CHECK: [[S_ARR_NEXT:%.+]] = getelementptr {{.+}} [[S_ARR_CTOR]],
+// CHECK: [[S_ARR_DONE:%.+]] = icmp {{.+}} [[S_ARR_NEXT]], [[S_ARR_END]]
+// CHECK: br i1 [[S_ARR_DONE]], label %[[S_ARR_CST_END:.+]], label %[[S_ARR_CST_LOOP]]
+// CHECK: [[S_ARR_CST_END]]:
+// CHECK: [[VAR_ADDR_REF:%.+]] = load [[S_INT_TY]]*, [[S_INT_TY]]** [[VAR_ADDR]],
+// CHECK: call void [[S_INT_TY_DEF_CONSTR]]([[S_INT_TY]]* [[VAR_PRIV]])
+// CHECK: store [[S_INT_TY]]* [[VAR_PRIV]], [[S_INT_TY]]** [[TMP_PRIV]],
+
+// CHECK: call void @__kmpc_for_static_init_4(
+
+// assignment: vec[i] = t_var;
+// CHECK: [[IV_VAL:%.+]] =
+// CHECK: [[T_VAR_PRIV_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[T_VAR_PRIV]],
+// CHECK: [[VEC_PTR:%.+]] = getelementptr inbounds [2 x i{{[0-9]+}}], [2 x i{{[0-9]+}}]* [[VEC_PRIV]], i{{[0-9]+}} 0, i{{[0-9]+}} {{.+}}
+// CHECK:  store i{{[0-9]+}} [[T_VAR_PRIV_VAL]], i{{[0-9]+}}* [[VEC_PTR]],
+
+// assignment: s_arr[i] = var;
+// CHECK-DAG: [[S_ARR_PTR:%.+]] = getelementptr inbounds [2 x [[S_INT_TY]]], [2 x [[S_INT_TY]]]* [[S_ARR_PRIV]],
+// CHECK-DAG: [[TMP_VAL:%.+]] = load [[S_INT_TY]]*, [[S_INT_TY]]** [[TMP_PRIV]],
+// CHECK-DAG: [[S_ARR_PTR_BCAST:%.+]] = bitcast [[S_INT_TY]]* [[S_ARR_PTR]] to i8*
+// CHECK-DAG: [[TMP_VAL_BCAST:%.+]] = bitcast [[S_INT_TY]]* [[TMP_VAL]] to i8*
+// CHECK-DAG: call void @llvm.memcpy.{{.+}}(i8* [[S_ARR_PTR_BCAST]], i8* [[TMP_VAL_BCAST]],
+
+// CHECK: call void @__kmpc_for_static_fini(
+
+// lastprivates
+// CHECK: [[OMP_IS_LAST_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[OMP_IS_LAST]],
+// CHECK: [[IS_LAST_IT:%.+]] = icmp ne i{{[0-9]+}} [[OMP_IS_LAST_VAL]], 0
+// CHECK: br i1 [[IS_LAST_IT]], label %[[OMP_LASTPRIV_BLOCK:.+]], label %[[OMP_LASTPRIV_DONE:.+]]
+
+// CHECK: [[OMP_LASTPRIV_BLOCK]]:
+// CHECK: [[T_VAR_VAL:%.+]] = load i{{[0-9]+}}, i{{[0-9]+}}* [[T_VAR_PRIV]],
+// CHECK: store i{{[0-9]+}} [[T_VAR_VAL]], i{{[0-9]+}}* [[T_VAR_ADDR_REF]],
+// CHECK: [[BCAST_VEC_ADDR_REF:%.+]] = bitcast [2 x i{{[0-9]+}}]* [[VEC_ADDR_REF]] to i8*
+// CHECK: [[BCAST_VEC_PRIV:%.+]] = bitcast [2 x i{{[0-9]+}}]* [[VEC_PRIV]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[BCAST_VEC_ADDR_REF]], i8* [[BCAST_VEC_PRIV]],
+// CHECK: [[S_ARR_BEGIN:%.+]] = getelementptr inbounds [2 x [[S_INT_TY]]], [2 x [[S_INT_TY]]]* [[S_ARR_ADDR_REF]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+// CHECK: [[S_ARR_PRIV_BCAST:%.+]] = bitcast [2 x [[S_INT_TY]]]* [[S_ARR_PRIV]] to [[S_INT_TY]]*
+// CHECK: [[S_ARR_BEGIN_GEP:%.+]] = getelementptr [[S_INT_TY]], [[S_INT_TY]]* [[S_ARR_BEGIN]], i{{[0-9]+}} 2
+// CHECK: [[S_ARR_IS_EMPTY:%.+]] = icmp eq [[S_INT_TY]]* [[S_ARR_BEGIN]], [[S_ARR_BEGIN_GEP]]
+// CHECK: br i1 [[S_ARR_IS_EMPTY]], label %[[S_ARR_COPY_DONE:.+]], label %[[S_ARR_COPY_BLOCK:.+]]
+// CHECK: [[S_ARR_COPY_BLOCK]]:
+// CHECK: [[S_ARR_SRC_EL:%.+]] = phi [[S_INT_TY]]*{{.+}}
+// CHECK: [[S_ARR_DST_EL:%.+]] = phi [[S_INT_TY]]*{{.+}}
+// CHECK: [[S_ARR_DST_BCAST:%.+]] = bitcast [[S_INT_TY]]* [[S_ARR_DST_EL]] to i8*
+// CHECK: [[S_ARR_SRC_BCAST:%.+]] = bitcast [[S_INT_TY]]* [[S_ARR_SRC_EL]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[S_ARR_DST_BCAST]], i8* [[S_ARR_SRC_BCAST]]{{.+}})
+// CHECK: [[S_ARR_DST_NEXT:%.+]] = getelementptr [[S_INT_TY]], [[S_INT_TY]]* [[S_ARR_DST_EL]], i{{[0-9]+}} 1
+// CHECK: [[S_ARR_SRC_NEXT:%.+]] = getelementptr{{.+}}
+// CHECK: [[CPY_IS_FINISHED:%.+]] = icmp eq [[S_INT_TY]]* [[S_ARR_DST_NEXT]], [[S_ARR_BEGIN_GEP]]
+// CHECK: br i1 [[CPY_IS_FINISHED]], label %[[S_ARR_COPY_DONE]], label %[[S_ARR_COPY_BLOCK]]
+// CHECK: [[S_ARR_COPY_DONE]]:
+// CHECK: [[TMP_VAL1:%.+]] = load [[S_INT_TY]]*, [[S_INT_TY]]** [[TMP_PRIV]],
+// CHECK: [[VAR_ADDR_REF_BCAST:%.+]] = bitcast [[S_INT_TY]]* [[VAR_ADDR_REF]] to i8*
+// CHECK: [[TMP_VAL1_BCAST:%.+]] = bitcast [[S_INT_TY]]* [[TMP_VAL1]] to i8*
+// CHECK: call void @llvm.memcpy.{{.+}}(i8* [[VAR_ADDR_REF_BCAST]], i8* [[TMP_VAL1_BCAST]],{{.+}})
+// CHECK: ret void
+
+#endif

test/OpenMP/distribute_parallel_for_num_threads_codegen.cpp

+// RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple %itanium_abi_triple -emit-llvm %s -fexceptions -fcxx-exceptions -o - | FileCheck %s
+// RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -std=c++11 -triple %itanium_abi_triple -fexceptions -fcxx-exceptions -emit-pch -o %t %s
+// RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple %itanium_abi_triple -fexceptions -fcxx-exceptions -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s
+// expected-no-diagnostics
+#ifndef HEADER
+#define HEADER
+
+typedef __INTPTR_TYPE__ intptr_t;
+
+// CHECK-DAG: [[IDENT_T_TY:%.+]] = type { i32, i32, i32, i32, i8* }
+// CHECK-DAG: [[S_TY:%.+]] = type { [[INTPTR_T_TY:i[0-9]+]], [[INTPTR_T_TY]], [[INTPTR_T_TY]] }
+// CHECK-DAG: [[STR:@.+]] = private unnamed_addr constant [23 x i8] c";unknown;unknown;0;0;;\00"
+// CHECK-DAG: [[DEF_LOC_2:@.+]] = private unnamed_addr constant [[IDENT_T_TY]] { i32 0, i32 2, i32 0, i32 0, i8* getelementptr inbounds ([23 x i8], [23 x i8]* [[STR]], i32 0, i32 0) }
+
+void foo();
+
+struct S {
+  intptr_t a, b, c;
+  S(intptr_t a) : a(a) {}
+  operator char() { return a; }
+  ~S() {}
+};
+
+template <typename T, int C>
+int tmain() {
+#pragma omp target
+#pragma omp teams
+#pragma omp distribute parallel for num_threads(C)
+  for (int i = 0; i < 100; i++)
+    foo();
+#pragma omp target
+#pragma omp teams
+#pragma omp distribute parallel for num_threads(T(23))
+  for (int i = 0; i < 100; i++)
+    foo();
+  return 0;
+}
+
+int main() {
+  S s(0);
+  char a = s;
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_0:@.+]](
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOADING_FUN_1:@.+]](
+// CHECK: invoke{{.+}} [[TMAIN_5:@.+]]()
+// CHECK: invoke{{.+}} [[TMAIN_1:@.+]]()
+#pragma omp target
+#pragma omp teams
+  // CHECK: define internal void [[OFFLOADING_FUN_0]](
+  // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 0, {{.+}}* [[OMP_TEAMS_OUTLINED_0:@.+]] to {{.+}})
+#pragma omp distribute parallel for num_threads(2)
+  for (int i = 0; i < 100; i++) {
+    // CHECK: define{{.+}} void [[OMP_TEAMS_OUTLINED_0]](
+    // CHECK:       call {{.*}}void @__kmpc_push_num_threads([[IDENT_T_TY]]* [[DEF_LOC_2]], i32 {{.+}}, i32 2)
+    // CHECK:       call {{.*}}void {{.*}} @__kmpc_fork_call(    
+    foo();
+  }
+#pragma omp target
+#pragma omp teams
+  // CHECK: define internal void [[OFFLOADING_FUN_1]](
+  
+  // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 1, {{.+}}* [[OMP_TEAMS_OUTLINED_1:@.+]] to {{.+}})
+#pragma omp distribute parallel for num_threads(a) 
+  for (int i = 0; i < 100; i++) {
+    // CHECK: define{{.+}} void [[OMP_TEAMS_OUTLINED_1]](
+    // CHECK-DAG: [[A_ADDR:%.+]] = alloca i8*,
+    // CHECK-DAG: [[A_REF:%.+]] = load i8*, i8** [[A_ADDR]],
+    // CHECK-DAG: [[A_VAL:%.+]] = load i8, i8* [[A_REF]],
+    // CHECK-DAG: [[A_EXT:%.+]] = sext i8 [[A_VAL]] to {{.+}}
+    // CHECK: call {{.*}}void @__kmpc_push_num_threads([[IDENT_T_TY]]* [[DEF_LOC_2]], i32 {{.+}}, i32 [[A_EXT]])
+    // CHECK: call {{.*}}void {{.*}} @__kmpc_fork_call(
+    foo();
+  }
+  return a + tmain<char, 5>() + tmain<S, 1>();
+}
+
+// tmain 5
+// CHECK-DAG: define {{.*}}i{{[0-9]+}} [[TMAIN_5]]()
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[T_OFFLOADING_FUN_0:@.+]](
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[T_OFFLOADING_FUN_1:@.+]](
+
+// tmain 1
+// CHECK-DAG: define {{.*}}i{{[0-9]+}} [[TMAIN_1]]()
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[T_OFFLOADING_FUN_2:@.+]](
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[T_OFFLOADING_FUN_3:@.+]](
+
+// CHECK: define internal void [[T_OFFLOADING_FUN_0]](
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 0, {{.+}}* [[T_OMP_TEAMS_OUTLINED_0:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} void [[T_OMP_TEAMS_OUTLINED_0]](
+// CHECK:       call {{.*}}void @__kmpc_push_num_threads([[IDENT_T_TY]]* [[DEF_LOC_2]], i32 {{.+}}, i32 5)
+// CHECK:       call {{.*}}void {{.*}} @__kmpc_fork_call(    
+
+// CHECK: define internal void [[T_OFFLOADING_FUN_1]](
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 0, {{.+}}* [[T_OMP_TEAMS_OUTLINED_1:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} void [[T_OMP_TEAMS_OUTLINED_1]](
+// CHECK:       call {{.*}}void @__kmpc_push_num_threads([[IDENT_T_TY]]* [[DEF_LOC_2]], i32 {{.+}}, i32 23)
+// CHECK:       call {{.*}}void {{.*}} @__kmpc_fork_call(    
+
+// CHECK: define internal void [[T_OFFLOADING_FUN_2]](
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 0, {{.+}}* [[T_OMP_TEAMS_OUTLINED_2:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} void [[T_OMP_TEAMS_OUTLINED_2]](
+// CHECK:       call {{.*}}void @__kmpc_push_num_threads([[IDENT_T_TY]]* [[DEF_LOC_2]], i32 {{.+}}, i32 1)
+// CHECK:       call {{.*}}void {{.*}} @__kmpc_fork_call(    
+
+// CHECK: define internal void [[T_OFFLOADING_FUN_3]](
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 0, {{.+}}* [[T_OMP_TEAMS_OUTLINED_3:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} void [[T_OMP_TEAMS_OUTLINED_3]](
+// CHECK-DAG:   [[CALL_RES:%.+]] = invoke{{.+}} i8 [[S_TY_CHAR_OP:@.+]]([[S_TY]]* {{.+}})
+// CHECK-DAG:   [[CALL_RES_SEXT:%.+]] = sext i8 [[CALL_RES]] to {{.+}}
+// CHECK:       call {{.*}}void @__kmpc_push_num_threads([[IDENT_T_TY]]* [[DEF_LOC_2]], i32 {{.+}}, i32 [[CALL_RES_SEXT]])
+// CHECK:       call {{.*}}void {{.*}} @__kmpc_fork_call(
+#endif

test/OpenMP/distribute_parallel_for_private_codegen.cpp

+// RUN: %clang_cc1 -DLAMBDA -verify -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-64
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-64
+// RUN: %clang_cc1 -DLAMBDA -verify -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-32
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-32
+
+// RUN: %clang_cc1  -verify -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-64
+// RUN: %clang_cc1  -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1  -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-64
+// RUN: %clang_cc1  -verify -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-32
+// RUN: %clang_cc1  -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1  -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-32
+// expected-no-diagnostics
+#ifndef HEADER
+#define HEADER
+
+template <class T>
+struct S {
+  T f;
+  S(T a) : f(a) {}
+  S() : f() {}
+  operator T() { return T(); }
+  ~S() {}
+};
+
+// CHECK: [[S_FLOAT_TY:%.+]] = type { float }
+// CHECK: [[S_INT_TY:%.+]] = type { i{{[0-9]+}} }
+template <typename T>
+T tmain() {
+  S<T> test;
+  T t_var = T();
+  T vec[] = {1, 2};
+  S<T> s_arr[] = {1, 2};
+  S<T> &var = test;
+  #pragma omp target
+  #pragma omp teams
+  #pragma omp distribute parallel for private(t_var, vec, s_arr, s_arr, var, var)
+  for (int i = 0; i < 2; ++i) {
+    vec[i] = t_var;
+    s_arr[i] = var;
+  }
+  return T();
+}
+
+int main() {
+  static int svar;
+  volatile double g;
+  volatile double &g1 = g;
+
+  #ifdef LAMBDA
+  // LAMBDA-LABEL: @main
+  // LAMBDA: call{{.*}} void [[OUTER_LAMBDA:@.+]](
+  [&]() {
+    static float sfvar;
+    // LAMBDA: define{{.*}} internal{{.*}} void [[OUTER_LAMBDA]](
+    // LAMBDA: call i{{[0-9]+}} @__tgt_target_teams(
+    // LAMBDA: call void [[OFFLOADING_FUN:@.+]](
+
+    // LAMBDA: define{{.+}} void [[OFFLOADING_FUN]]()
+    // LAMBDA: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 0, {{.+}}* [[OMP_OUTLINED:@.+]] to {{.+}})
+    #pragma omp target
+    #pragma omp teams
+    #pragma omp distribute parallel for private(g, g1, svar, sfvar)
+    for (int i = 0; i < 2; ++i) {
+      // LAMBDA: define{{.*}} internal{{.*}} void [[OMP_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}})
+      // LAMBDA: [[G_PRIVATE_ADDR:%.+]] = alloca double,
+      // LAMBDA: [[G1_PRIVATE_ADDR:%.+]] = alloca double,
+      // LAMBDA: [[TMP_PRIVATE_ADDR:%.+]] = alloca double*,
+      // LAMBDA: [[SVAR_PRIVATE_ADDR:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[SFVAR_PRIVATE_ADDR:%.+]] = alloca float,
+      // LAMBDA: store double* [[G1_PRIVATE_ADDR]], double** [[TMP_PRIVATE_ADDR]],
+      // LAMBDA: call {{.*}}void @__kmpc_for_static_init_4(
+      // LAMBDA: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED:@.+]] to {{.+}},
+      // LAMBDA: call {{.*}}void @__kmpc_for_static_fini(
+      // LAMBDA: ret void
+
+      // LAMBDA: define{{.+}} void [[OMP_PARFOR_OUTLINED]](
+      // LAMBDA: [[G_PRIVATE_ADDR:%.+]] = alloca double,
+      // LAMBDA: [[G1_PRIVATE_ADDR:%.+]] = alloca double,
+      // LAMBDA: [[TMP_PRIVATE_ADDR:%.+]] = alloca double*,
+      // LAMBDA: [[SVAR_PRIVATE_ADDR:%.+]] = alloca i{{[0-9]+}},
+      // LAMBDA: [[SFVAR_PRIVATE_ADDR:%.+]] = alloca float,
+
+      g = 1;
+      g1 = 1;
+      svar = 3;
+      sfvar = 4.0;     
+      // LAMBDA: store double* [[G1_PRIVATE_ADDR]], double** [[TMP_PRIVATE_ADDR]],      
+      // LAMBDA: store double 1.0{{.+}}, double* [[G_PRIVATE_ADDR]],
+      // LAMBDA: store i{{[0-9]+}} 3, i{{[0-9]+}}* [[SVAR_PRIVATE_ADDR]],
+      // LAMBDA: store float 4.0{{.+}}, float* [[SFVAR_PRIVATE_ADDR]],
+      // LAMBDA: [[G_PRIVATE_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+      // LAMBDA: store double* [[G_PRIVATE_ADDR]], double** [[G_PRIVATE_ADDR_REF]],
+      // LAMBDA: [[TMP_PRIVATE_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
+      // LAMBDA: [[G1_PRIVATE_ADDR_FROM_TMP:%.+]] = load double*, double** [[TMP_PRIVATE_ADDR]],
+      // LAMBDA: store double* [[G1_PRIVATE_ADDR_FROM_TMP]], double** [[TMP_PRIVATE_ADDR_REF]],
+      // LAMBDA: [[SVAR_PRIVATE_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 2
+      // LAMBDA: store i{{[0-9]+}}* [[SVAR_PRIVATE_ADDR]], i{{[0-9]+}}** [[SVAR_PRIVATE_ADDR_REF]]
+      // LAMBDA: [[SFVAR_PRIVATE_ADDR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG:%.+]], i{{[0-9]+}} 0, i{{[0-9]+}} 3
+      // LAMBDA: store float* [[SFVAR_PRIVATE_ADDR]], float** [[SFVAR_PRIVATE_ADDR_REF]]
+      // LAMBDA: call{{.*}} void [[INNER_LAMBDA:@.+]](%{{.+}}* [[ARG]])
+      // LAMBDA: call {{.*}}void @__kmpc_for_static_fini(
+      // LAMBDA: ret void
+      [&]() {
+  // LAMBDA: define {{.+}} void [[INNER_LAMBDA]](%{{.+}}* [[ARG_PTR:%.+]])
+  // LAMBDA: store %{{.+}}* [[ARG_PTR]], %{{.+}}** [[ARG_PTR_REF:%.+]],
+  g = 2;
+  g1 = 2;
+  svar = 4;
+  sfvar = 8.0;
+  // LAMBDA: [[ARG_PTR:%.+]] = load %{{.+}}*, %{{.+}}** [[ARG_PTR_REF]]
+  // LAMBDA: [[G_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 0
+  // LAMBDA: [[G_REF:%.+]] = load double*, double** [[G_PTR_REF]]
+  // LAMBDA: store double 2.0{{.+}}, double* [[G_REF]]
+
+  // LAMBDA: [[TMP_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 1
+  // LAMBDA: [[G1_REF:%.+]] = load double*, double** [[TMP_PTR_REF]]
+  // LAMBDA: store double 2.0{{.+}}, double* [[G1_REF]],
+  // LAMBDA: [[SVAR_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 2
+  // LAMBDA: [[SVAR_REF:%.+]] = load i{{[0-9]+}}*, i{{[0-9]+}}** [[SVAR_PTR_REF]]
+  // LAMBDA: store i{{[0-9]+}} 4, i{{[0-9]+}}* [[SVAR_REF]]
+  // LAMBDA: [[SFVAR_PTR_REF:%.+]] = getelementptr inbounds %{{.+}}, %{{.+}}* [[ARG_PTR]], i{{[0-9]+}} 0, i{{[0-9]+}} 3
+  // LAMBDA: [[SFVAR_REF:%.+]] = load float*, float** [[SFVAR_PTR_REF]]
+  // LAMBDA: store float 8.0{{.+}}, float* [[SFVAR_REF]]
+      }();
+    }
+  }();
+  return 0;
+  #else
+  S<float> test;
+  int t_var = 0;
+  int vec[] = {1, 2};
+  S<float> s_arr[] = {1, 2};
+  S<float> &var = test;
+
+  #pragma omp target
+  #pragma omp teams
+  #pragma omp distribute parallel for private(t_var, vec, s_arr, s_arr, var, var, svar)
+  for (int i = 0; i < 2; ++i) {
+    vec[i] = t_var;
+    s_arr[i] = var;
+  }
+  return tmain<int>();
+  #endif
+}
+
+// CHECK: define{{.*}} i{{[0-9]+}} @main()
+// CHECK: [[TEST:%.+]] = alloca [[S_FLOAT_TY]],
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_CONSTR:@.+]]([[S_FLOAT_TY]]* [[TEST]])
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOAD_FUN_0:@.+]](
+
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_DESTR:@.+]]([[S_FLOAT_TY]]* [[TEST]])
+// CHECK: ret
+
+// CHECK: define{{.+}} [[OFFLOAD_FUN_0]]()
+// CHECK: call void (%{{.+}}*, i{{[0-9]+}}, void (i{{[0-9]+}}*, i{{[0-9]+}}*, ...)*, ...) @__kmpc_fork_teams(%{{.+}}* @{{.+}}, i{{[0-9]+}} 0, void (i{{[0-9]+}}*, i{{[0-9]+}}*, ...)* bitcast (void (i{{[0-9]+}}*, i{{[0-9]+}}*)* [[OMP_OUTLINED_0:@.+]] to void
+// CHECK: ret
+//
+// CHECK: define internal void [[OMP_OUTLINED_0]](i{{[0-9]+}}* noalias [[GTID_ADDR:%.+]], i{{[0-9]+}}* noalias %{{.+}})
+// CHECK: [[T_VAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[VEC_PRIV:%.+]] = alloca [2 x i{{[0-9]+}}],
+// CHECK: [[S_ARR_PRIV:%.+]] = alloca [2 x [[S_FLOAT_TY]]],
+// CHECK-NOT: alloca [2 x [[S_FLOAT_TY]]],
+// CHECK: [[VAR_PRIV:%.+]] = alloca [[S_FLOAT_TY]],
+// CHECK-NOT: alloca [[S_FLOAT_TY]],
+// CHECK: [[S_VAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+// CHECK: store i{{[0-9]+}}* [[GTID_ADDR]], i{{[0-9]+}}** [[GTID_ADDR_REF:%.+]]
+// CHECK-NOT: [[T_VAR_PRIV]]
+// CHECK-NOT: [[VEC_PRIV]]
+// this is the ctor loop
+// CHECK: {{.+}}:
+// CHECK: [[S_ARR_PRIV_ITEM:%.+]] = phi [[S_FLOAT_TY]]*
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_CONSTR]]([[S_FLOAT_TY]]* [[S_ARR_PRIV_ITEM]])
+// CHECK-NOT: [[T_VAR_PRIV]]
+// CHECK-NOT: [[VEC_PRIV]]
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_CONSTR]]([[S_FLOAT_TY]]* [[VAR_PRIV]])
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_0:@.+]] to {{.+}},
+// CHECK: call void @__kmpc_for_static_fini(
+
+// call destructors: var..
+// CHECK-DAG: call {{.+}} [[S_FLOAT_TY_DEF_DESTR]]([[S_FLOAT_TY]]* [[VAR_PRIV]])
+
+// ..and s_arr
+// CHECK: {{.+}}:
+// CHECK: [[S_ARR_EL_PAST:%.+]] = phi [[S_FLOAT_TY]]*
+// CHECK: [[S_ARR_PRIV_ITEM:%.+]] = getelementptr {{.+}}, {{.+}} [[S_ARR_EL_PAST]],
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_DESTR]]([[S_FLOAT_TY]]* [[S_ARR_PRIV_ITEM]])
+
+// CHECK: ret void
+
+// By OpenMP specifications, private applies to both distribute and parallel for.
+// However, the support for 'private' of 'parallel' is only used when 'parallel'
+// is found alone. Therefore we only have one 'private' support for 'parallel for'
+// in combination
+// CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_0]](
+// CHECK: [[T_VAR_PRIV:%t_var+]] = alloca i{{[0-9]+}},
+// CHECK: [[VEC_PRIV:%vec+]] = alloca [2 x i{{[0-9]+}}],
+// CHECK: [[S_ARR_PRIV:%s_arr+]] = alloca [2 x [[S_FLOAT_TY]]],
+// CHECK-NOT: alloca [2 x [[S_FLOAT_TY]]],
+// CHECK: [[VAR_PRIV:%var+]] = alloca [[S_FLOAT_TY]],
+// CHECK-NOT: alloca [[S_FLOAT_TY]],
+// CHECK: [[S_VAR_PRIV:%svar+]] = alloca i{{[0-9]+}},
+// CHECK: store i{{[0-9]+}}* [[GTID_ADDR]], i{{[0-9]+}}** [[GTID_ADDR_REF:%.+]]
+// CHECK-NOT: [[T_VAR_PRIV]]
+// CHECK-NOT: [[VEC_PRIV]]
+// this is the ctor loop
+// CHECK: {{.+}}:
+// CHECK: [[S_ARR_PRIV_ITEM:%.+]] = phi [[S_FLOAT_TY]]*
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_CONSTR]]([[S_FLOAT_TY]]* [[S_ARR_PRIV_ITEM]])
+// CHECK-NOT: [[T_VAR_PRIV]]
+// CHECK-NOT: [[VEC_PRIV]]
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_CONSTR]]([[S_FLOAT_TY]]* [[VAR_PRIV]])
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: call void @__kmpc_for_static_fini(
+
+// call destructors: var..
+// CHECK-DAG: call {{.+}} [[S_FLOAT_TY_DEF_DESTR]]([[S_FLOAT_TY]]* [[VAR_PRIV]])
+
+// ..and s_arr
+// CHECK: {{.+}}:
+// CHECK: [[S_ARR_EL_PAST:%.+]] = phi [[S_FLOAT_TY]]*
+// CHECK: [[S_ARR_PRIV_ITEM:%.+]] = getelementptr {{.+}}, {{.+}} [[S_ARR_EL_PAST]],
+// CHECK: call {{.*}} [[S_FLOAT_TY_DEF_DESTR]]([[S_FLOAT_TY]]* [[S_ARR_PRIV_ITEM]])
+
+// CHECK: ret void
+
+// template tmain with S_INT_TY
+// CHECK: define{{.*}} i{{[0-9]+}} [[TMAIN_INT:@.+]]()
+// CHECK: [[TEST:%.+]] = alloca [[S_INT_TY]],
+// CHECK: call {{.*}} [[S_INT_TY_DEF_CONSTR:@.+]]([[S_INT_TY]]* [[TEST]])
+// CHECK: call i{{[0-9]+}} @__tgt_target_teams(
+// CHECK: call void [[OFFLOAD_FUN_1:@.+]](
+// CHECK: call {{.*}} [[S_INT_TY_DEF_DESTR:@.+]]([[S_INT_TY]]* [[TEST]])
+// CHECK: ret
+
+// CHECK: ret
+
+// CHECK: define internal void [[OFFLOAD_FUN_1]]()
+// CHECK: call void (%{{.+}}*, i{{[0-9]+}}, void (i{{[0-9]+}}*, i{{[0-9]+}}*, ...)*, ...) @__kmpc_fork_teams(%{{.+}}* @{{.+}}, i{{[0-9]+}} 0, void (i{{[0-9]+}}*, i{{[0-9]+}}*, ...)* bitcast (void (i{{[0-9]+}}*, i{{[0-9]+}}*)* [[OMP_OUTLINED_1:@.+]] to void
+// CHECK: ret
+//
+// CHECK: define internal void [[OMP_OUTLINED_1]](i{{[0-9]+}}* noalias [[GTID_ADDR:%.+]], i{{[0-9]+}}* noalias %{{.+}})
+// CHECK: [[T_VAR_PRIV:%.+]] = alloca i{{[0-9]+}},
+// CHECK: [[VEC_PRIV:%.+]] = alloca [2 x i{{[0-9]+}}],
+// CHECK: [[S_ARR_PRIV:%.+]] = alloca [2 x [[S_INT_TY]]],
+// CHECK-NOT: alloca [2 x [[S_INT_TY]]],
+// CHECK: [[VAR_PRIV:%.+]] = alloca [[S_INT_TY]],
+// CHECK-NOT: alloca [[S_INT_TY]],
+// CHECK: store i{{[0-9]+}}* [[GTID_ADDR]], i{{[0-9]+}}** [[GTID_ADDR_REF:%.+]]
+// CHECK-NOT: [[T_VAR_PRIV]]
+// CHECK-NOT: [[VEC_PRIV]]
+// CHECK: {{.+}}:
+// CHECK: [[S_ARR_PRIV_ITEM:%.+]] = phi [[S_INT_TY]]*
+// CHECK: call {{.*}} [[S_INT_TY_DEF_CONSTR]]([[S_INT_TY]]* [[S_ARR_PRIV_ITEM]])
+// CHECK-NOT: [[T_VAR_PRIV]]
+// CHECK-NOT: [[VEC_PRIV]]
+// CHECK: call {{.*}} [[S_INT_TY_DEF_CONSTR]]([[S_INT_TY]]* [[VAR_PRIV]])
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: call{{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}}[[OMP_PARFOR_OUTLINED_1:@.+]] to {{.+}},
+// CHECK: call void @__kmpc_for_static_fini(
+// CHECK: ret void
+
+// CHECK: define{{.+}} void [[OMP_PARFOR_OUTLINED_1]](
+// CHECK: [[T_VAR_PRIV:%t_var+]] = alloca i{{[0-9]+}},
+// CHECK: [[VEC_PRIV:%vec+]] = alloca [2 x i{{[0-9]+}}],
+// CHECK: [[S_ARR_PRIV:%s_arr+]] = alloca [2 x [[S_INT_TY]]],
+// CHECK-NOT: alloca [2 x [[S_INT_TY]]],
+// CHECK: [[VAR_PRIV:%var+]] = alloca [[S_INT_TY]],
+// CHECK-NOT: alloca [[S_INT_TY]],
+// CHECK: store i{{[0-9]+}}* [[GTID_ADDR]], i{{[0-9]+}}** [[GTID_ADDR_REF:%.+]]
+// CHECK-NOT: [[T_VAR_PRIV]]
+// CHECK-NOT: [[VEC_PRIV]]
+// this is the ctor loop
+// CHECK: {{.+}}:
+// CHECK: [[S_ARR_PRIV_ITEM:%.+]] = phi [[S_INT_TY]]*
+// CHECK: call {{.*}} [[S_INT_TY_DEF_CONSTR]]([[S_INT_TY]]* [[S_ARR_PRIV_ITEM]])
+// CHECK-NOT: [[T_VAR_PRIV]]
+// CHECK-NOT: [[VEC_PRIV]]
+// CHECK: call {{.*}} [[S_INT_TY_DEF_CONSTR]]([[S_INT_TY]]* [[VAR_PRIV]])
+// CHECK: call void @__kmpc_for_static_init_4(
+// CHECK: call void @__kmpc_for_static_fini(
+
+// call destructors: var..
+// CHECK-DAG: call {{.+}} [[S_INT_TY_DEF_DESTR]]([[S_INT_TY]]* [[VAR_PRIV]])
+
+// ..and s_arr
+// CHECK: {{.+}}:
+// CHECK: [[S_ARR_EL_PAST:%.+]] = phi [[S_INT_TY]]*
+// CHECK: [[S_ARR_PRIV_ITEM:%.+]] = getelementptr {{.+}}, {{.+}} [[S_ARR_EL_PAST]],
+// CHECK: call {{.*}} [[S_INT_TY_DEF_DESTR]]([[S_INT_TY]]* [[S_ARR_PRIV_ITEM]])
+
+// CHECK: ret void
+
+#endif

test/OpenMP/distribute_parallel_for_proc_bind_codegen.cpp

+// add -fopenmp-targets
+
+// RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s
+// RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
+// RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s
+// expected-no-diagnostics
+#ifndef HEADER
+#define HEADER
+
+typedef __INTPTR_TYPE__ intptr_t;
+
+// CHECK-DAG: [[IDENT_T_TY:%.+]] = type { i32, i32, i32, i32, i8* }
+// CHECK-DAG: [[STR:@.+]] = private unnamed_addr constant [23 x i8] c";unknown;unknown;0;0;;\00"
+// CHECK-DAG: [[DEF_LOC_2:@.+]] = private unnamed_addr constant [[IDENT_T_TY]] { i32 0, i32 2, i32 0, i32 0, i8* getelementptr inbounds ([23 x i8], [23 x i8]* [[STR]], i32 0, i32 0) }
+
+void foo();
+
+struct S {
+  intptr_t a, b, c;
+  S(intptr_t a) : a(a) {}
+  operator char() { return a; }
+  ~S() {}
+};
+
+template <typename T>
+T tmain() {
+#pragma omp target
+#pragma omp teams
+#pragma omp distribute parallel for proc_bind(master)
+  for(int i = 0; i < 1000; i++) {}
+  return T();
+}
+
+int main() {
+  // CHECK-LABEL: @main
+#pragma omp target
+#pragma omp teams
+#pragma omp distribute parallel for proc_bind(spread)
+  for(int i = 0; i < 1000; i++) {}  
+#pragma omp target
+#pragma omp teams
+#pragma omp distribute parallel for proc_bind(close)
+  for(int i = 0; i < 1000; i++) {}
+  return tmain<int>();
+}
+
+// CHECK: call {{.*}}@__tgt_target_teams({{.+}})
+// CHECK: call void [[OFFL1:@.+]]()
+// CHECK: call {{.*}}@__tgt_target_teams({{.+}})
+// CHECK: call void [[OFFL2:@.+]]()
+// CHECK: [[CALL_RET:%.+]] = call{{.+}} i32 [[TMAIN:@.+]]()
+// CHECK: ret i32 [[CALL_RET]]
+
+// CHECK: define{{.+}} void [[OFFL1]](
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, {{.+}}, {{.+}}* [[OMP_OUTLINED_1:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} [[OMP_OUTLINED_1]](i32* {{.+}} [[GTID_IN:%.+]],
+// CHECK: [[GTID_ADDR:%.+]] = alloca i32*,
+// CHECK: store i32* [[GTID_IN]], i32** [[GTID_ADDR]],
+// CHECK: [[GTID_REF:%.+]] = load i32*, i32** [[GTID_ADDR]],
+// CHECK: [[GTID_VAL:%.+]] = load i32, i32* [[GTID_REF]],
+// CHECK: call {{.*}}void @__kmpc_push_proc_bind([[IDENT_T_TY]]* [[DEF_LOC_2]], i32 [[GTID_VAL]], i32 4)
+// CHECK: call {{.*}}void (%ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(
+// CHECK: ret void
+
+// CHECK: define{{.+}} [[OFFL2]]()
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, {{.+}}, {{.+}}* [[OMP_OUTLINED_1:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} [[OMP_OUTLINED_1]](i32* {{.+}} [[GTID_IN:%.+]],
+// CHECK: [[GTID_ADDR:%.+]] = alloca i32*,
+// CHECK: store i32* [[GTID_IN]], i32** [[GTID_ADDR]],
+// CHECK: [[GTID_REF:%.+]] = load i32*, i32** [[GTID_ADDR]],
+// CHECK: [[GTID_VAL:%.+]] = load i32, i32* [[GTID_REF]],
+// CHECK: call {{.*}}void @__kmpc_push_proc_bind([[IDENT_T_TY]]* [[DEF_LOC_2]], i32 [[GTID_VAL]], i32 3)
+// CHECK: call {{.*}}void (%ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(
+// CHECK: ret void
+
+// CHECK: define{{.+}} [[TMAIN]]()
+// CHECK: call {{.*}}@__tgt_target_teams({{.+}})
+// CHECK: call void [[OFFL3:@.+]]()
+
+// CHECK: define{{.+}} [[OFFL3]]()
+// CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, {{.+}}, {{.+}}* [[OMP_OUTLINED_3:@.+]] to {{.+}})
+
+// CHECK: define{{.+}} [[OMP_OUTLINED_3]](i32* {{.+}} [[GTID_IN:%.+]],
+// CHECK: [[GTID_ADDR:%.+]] = alloca i32*,
+// CHECK: store i32* [[GTID_IN]], i32** [[GTID_ADDR]],
+// CHECK: [[GTID_REF:%.+]] = load i32*, i32** [[GTID_ADDR]],
+// CHECK: [[GTID_VAL:%.+]] = load i32, i32* [[GTID_REF]],
+// CHECK: call {{.*}}void @__kmpc_push_proc_bind([[IDENT_T_TY]]* [[DEF_LOC_2]], i32 [[GTID_VAL]], i32 2)
+// CHECK: call {{.*}}void (%ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(
+// CHECK: ret void
+#endif