[OpenMP][NVPTX] Use single loops when generating code for distribute parallel for

[gtbercea]

Summary

This patch adds a new code generation path for bound sharing directives containing distribute parallel for. The new code generation scheme applies to chunked schedules on distribute and parallel for directives. The scheme simplifies the code that is being generated by eliminating the need for an outer for loop over chunks for both distribute and parallel for directives. In the case of distribute it applies to any sized chunk while in the parallel for case it only applies when chunk size is 1.

[gtbercea]

Rebase.

[ABataev]

lib/CodeGen/CGOpenMPRuntime.h

904

I'd rename this into isDistStaticChunked

lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp

4249

If the ChunkOne field is not required, you need to restore original code here

lib/CodeGen/CGStmtOpenMP.cpp

2360

This whole code is very similar to the unchunked case. Could you merge it?

2362

It allows you to check only the implicit case, what about if the user explicitly specifies that chunk is 1?

3421

Again, very similar to the unchunked code. Merge it.

lib/Sema/SemaOpenMP.cpp

5207

Seems to me, you need to use NumIterations instead of LastIteration

[gtbercea]

Refactor chunk one checking.

[gtbercea]

Use NumIterations.

[gtbercea]

lib/CodeGen/CGOpenMPRuntime.h

904

I've used the same naming convention as the isStaticNonchunked function for consistency.

[ABataev]

lib/CodeGen/CGOpenMPRuntime.h

904

What about this?

lib/CodeGen/CGStmtOpenMP.cpp

2360

What about this?

3421

?

lib/Sema/SemaOpenMP.cpp

5207

Add the tests for the collapsed loops.

[gtbercea]

Refactor static chunk schedules. Fix tests.

[ABataev]

What about collapsed loops?

lib/CodeGen/CGStmtOpenMP.cpp

3396

Restore the original code here, the logic can be simplified

3400

The same

3444–3464

Please, simplify this

[gtbercea]

Simplify code.

[gtbercea]

Add test for collapse.

[ABataev]

test/OpenMP/distribute_parallel_for_codegen.cpp

410

Bad check for names, you should not rely on them.

test/OpenMP/nvptx_target_teams_distribute_parallel_for_codegen.cpp

117 ↗

(On Diff #171170)

Also, bad checks.

[gtbercea]

Change tests.

[ABataev]

LG, with a nit

lib/Sema/SemaOpenMP.cpp

5308

Fix the comment here

[gtbercea]

Fix comment.

include/clang/AST/StmtOpenMP.h

enum {
     PrevEnsureUpperBoundOffset = 20,
     CombinedLowerBoundVariableOffset = 21,
     CombinedUpperBoundVariableOffset = 22,
     CombinedEnsureUpperBoundOffset = 23,
     CombinedInitOffset = 24,
     CombinedConditionOffset = 25,
     CombinedNextLowerBoundOffset = 26,
     CombinedNextUpperBoundOffset = 27,
+    CombinedDistConditionOffset = 28,
+    CombinedParForInDistConditionOffset = 29,
     // Offset to the end (and start of the following counters/updates/finals
     // arrays) for combined distribute loop directives.
-    CombinedDistributeEnd = 28,
+    CombinedDistributeEnd = 30,
   };
 
   /// Get the counters storage.
   MutableArrayRef<Expr *> getCounters() {
     Expr **Storage = reinterpret_cast<Expr **>(
         &(*(std::next(child_begin(), getArraysOffset(getDirectiveKind())))));
     return MutableArrayRef<Expr *>(Storage, CollapsedNum);
   }
assert(isOpenMPLoopBoundSharingDirective(getDirectiveKind()) &&
            "expected loop bound sharing directive");
     *std::next(child_begin(), CombinedNextLowerBoundOffset) = CombNLB;
   }
   void setCombinedNextUpperBound(Expr *CombNUB) {
     assert(isOpenMPLoopBoundSharingDirective(getDirectiveKind()) &&
            "expected loop bound sharing directive");
     *std::next(child_begin(), CombinedNextUpperBoundOffset) = CombNUB;
   }
+  void setCombinedDistCond(Expr *CombDistCond) {
+    assert(isOpenMPLoopBoundSharingDirective(getDirectiveKind()) &&
+           "expected loop bound distribute sharing directive");
+    *std::next(child_begin(), CombinedDistConditionOffset) = CombDistCond;
+  }
+  void setCombinedParForInDistCond(Expr *CombParForInDistCond) {
+    assert(isOpenMPLoopBoundSharingDirective(getDirectiveKind()) &&
+           "expected loop bound distribute sharing directive");
+    *std::next(child_begin(),
+               CombinedParForInDistConditionOffset) = CombParForInDistCond;
+  }
   void setCounters(ArrayRef<Expr *> A);
   void setPrivateCounters(ArrayRef<Expr *> A);
   void setInits(ArrayRef<Expr *> A);
   void setUpdates(ArrayRef<Expr *> A);
   void setFinals(ArrayRef<Expr *> A);
 
 public:
   /// The expressions built to support OpenMP loops in combined/composite
struct DistCombinedHelperExprs {
     ///  with 'omp for' in a same construct
     Expr *Cond;
     /// Update of LowerBound for statically scheduled omp loops for
     /// outer loop in combined constructs (e.g. 'distribute parallel for')
     Expr *NLB;
     /// Update of UpperBound for statically scheduled omp loops for
     /// outer loop in combined constructs (e.g. 'distribute parallel for')
     Expr *NUB;
+    /// Distribute Loop condition used when composing 'omp distribute'
+    ///  with 'omp for' in a same construct when schedule is chunked.
+    Expr *DistCond;
+    /// 'omp parallel for' loop condition used when composed with
+    /// 'omp distribute' in the same construct and when schedule is
+    /// chunked and the chunk size is 1.
+    Expr *ParForInDistCond;
   };
 
   /// The expressions built for the OpenMP loop CodeGen for the
   /// whole collapsed loop nest.
   struct HelperExprs {
     /// Loop iteration variable.
     Expr *IterationVarRef;
     /// Loop last iteration number.
void clear(unsigned Size) {
       PreInits = nullptr;
       DistCombinedFields.LB = nullptr;
       DistCombinedFields.UB = nullptr;
       DistCombinedFields.EUB = nullptr;
       DistCombinedFields.Init = nullptr;
       DistCombinedFields.Cond = nullptr;
       DistCombinedFields.NLB = nullptr;
       DistCombinedFields.NUB = nullptr;
+      DistCombinedFields.DistCond = nullptr;
+      DistCombinedFields.ParForInDistCond = nullptr;
     }
   };
 
   /// Get number of collapsed loops.
   unsigned getCollapsedNumber() const { return CollapsedNum; }
 
   Expr *getIterationVariable() const {
     return const_cast<Expr *>(reinterpret_cast<const Expr *>(
return const_cast<Expr *>(reinterpret_cast<const Expr *>(
         *std::next(child_begin(), CombinedNextLowerBoundOffset)));
   }
   Expr *getCombinedNextUpperBound() const {
     assert(isOpenMPLoopBoundSharingDirective(getDirectiveKind()) &&
            "expected loop bound sharing directive");
     return const_cast<Expr *>(reinterpret_cast<const Expr *>(
         *std::next(child_begin(), CombinedNextUpperBoundOffset)));
   }
+  Expr *getCombinedDistCond() const {
+    assert(isOpenMPLoopBoundSharingDirective(getDirectiveKind()) &&
+           "expected loop bound distribute sharing directive");
+    return const_cast<Expr *>(reinterpret_cast<const Expr *>(
+        *std::next(child_begin(), CombinedDistConditionOffset)));
+  }
+  Expr *getCombinedParForInDistCond() const {
+    assert(isOpenMPLoopBoundSharingDirective(getDirectiveKind()) &&
+           "expected loop bound distribute sharing directive");
+    return const_cast<Expr *>(reinterpret_cast<const Expr *>(
+        *std::next(child_begin(), CombinedParForInDistConditionOffset)));
+  }
   const Stmt *getBody() const {
     // This relies on the loop form is already checked by Sema.
     const Stmt *Body =
         getInnermostCapturedStmt()->getCapturedStmt()->IgnoreContainers();
     Body = cast<ForStmt>(Body)->getBody();
     for (unsigned Cnt = 1; Cnt < CollapsedNum; ++Cnt) {
       Body = Body->IgnoreContainers();
       Body = cast<ForStmt>(Body)->getBody();

include/clang/Basic/OpenMPKinds.h

#include "clang/Basic/OpenMPKinds.def"
   OMPC_DEFAULTMAP_MODIFIER_last
 };
 
 /// Scheduling data for loop-based OpenMP directives.
 struct OpenMPScheduleTy final {
   OpenMPScheduleClauseKind Schedule = OMPC_SCHEDULE_unknown;
   OpenMPScheduleClauseModifier M1 = OMPC_SCHEDULE_MODIFIER_unknown;
   OpenMPScheduleClauseModifier M2 = OMPC_SCHEDULE_MODIFIER_unknown;
+  bool HasChunkOne = false;
 };
 
 OpenMPDirectiveKind getOpenMPDirectiveKind(llvm::StringRef Str);
 const char *getOpenMPDirectiveName(OpenMPDirectiveKind Kind);
 
 OpenMPClauseKind getOpenMPClauseKind(llvm::StringRef Str);
 const char *getOpenMPClauseName(OpenMPClauseKind Kind);
 

lib/AST/StmtOpenMP.cpp

OMPDistributeParallelForDirective *OMPDistributeParallelForDirective::Create(
   Dir->setPreInits(Exprs.PreInits);
   Dir->setCombinedLowerBoundVariable(Exprs.DistCombinedFields.LB);
   Dir->setCombinedUpperBoundVariable(Exprs.DistCombinedFields.UB);
   Dir->setCombinedEnsureUpperBound(Exprs.DistCombinedFields.EUB);
   Dir->setCombinedInit(Exprs.DistCombinedFields.Init);
   Dir->setCombinedCond(Exprs.DistCombinedFields.Cond);
   Dir->setCombinedNextLowerBound(Exprs.DistCombinedFields.NLB);
   Dir->setCombinedNextUpperBound(Exprs.DistCombinedFields.NUB);
+  Dir->setCombinedDistCond(Exprs.DistCombinedFields.DistCond);
+  Dir->setCombinedParForInDistCond(Exprs.DistCombinedFields.ParForInDistCond);
   Dir->HasCancel = HasCancel;
   return Dir;
 }
 
 OMPDistributeParallelForDirective *
 OMPDistributeParallelForDirective::CreateEmpty(const ASTContext &C,
                                                unsigned NumClauses,
                                                unsigned CollapsedNum,
OMPDistributeParallelForSimdDirective::Create(
   Dir->setPreInits(Exprs.PreInits);
   Dir->setCombinedLowerBoundVariable(Exprs.DistCombinedFields.LB);
   Dir->setCombinedUpperBoundVariable(Exprs.DistCombinedFields.UB);
   Dir->setCombinedEnsureUpperBound(Exprs.DistCombinedFields.EUB);
   Dir->setCombinedInit(Exprs.DistCombinedFields.Init);
   Dir->setCombinedCond(Exprs.DistCombinedFields.Cond);
   Dir->setCombinedNextLowerBound(Exprs.DistCombinedFields.NLB);
   Dir->setCombinedNextUpperBound(Exprs.DistCombinedFields.NUB);
+  Dir->setCombinedDistCond(Exprs.DistCombinedFields.DistCond);
+  Dir->setCombinedParForInDistCond(Exprs.DistCombinedFields.ParForInDistCond);
   return Dir;
 }
 
 OMPDistributeParallelForSimdDirective *
 OMPDistributeParallelForSimdDirective::CreateEmpty(const ASTContext &C,
                                                    unsigned NumClauses,
                                                    unsigned CollapsedNum,
                                                    EmptyShell) {
OMPTeamsDistributeParallelForSimdDirective::Create(
   Dir->setPreInits(Exprs.PreInits);
   Dir->setCombinedLowerBoundVariable(Exprs.DistCombinedFields.LB);
   Dir->setCombinedUpperBoundVariable(Exprs.DistCombinedFields.UB);
   Dir->setCombinedEnsureUpperBound(Exprs.DistCombinedFields.EUB);
   Dir->setCombinedInit(Exprs.DistCombinedFields.Init);
   Dir->setCombinedCond(Exprs.DistCombinedFields.Cond);
   Dir->setCombinedNextLowerBound(Exprs.DistCombinedFields.NLB);
   Dir->setCombinedNextUpperBound(Exprs.DistCombinedFields.NUB);
+  Dir->setCombinedDistCond(Exprs.DistCombinedFields.DistCond);
+  Dir->setCombinedParForInDistCond(Exprs.DistCombinedFields.ParForInDistCond);
   return Dir;
 }
 
 OMPTeamsDistributeParallelForSimdDirective *
 OMPTeamsDistributeParallelForSimdDirective::CreateEmpty(const ASTContext &C,
                                                         unsigned NumClauses,
                                                         unsigned CollapsedNum,
                                                         EmptyShell) {
OMPTeamsDistributeParallelForDirective::Create(
   Dir->setPreInits(Exprs.PreInits);
   Dir->setCombinedLowerBoundVariable(Exprs.DistCombinedFields.LB);
   Dir->setCombinedUpperBoundVariable(Exprs.DistCombinedFields.UB);
   Dir->setCombinedEnsureUpperBound(Exprs.DistCombinedFields.EUB);
   Dir->setCombinedInit(Exprs.DistCombinedFields.Init);
   Dir->setCombinedCond(Exprs.DistCombinedFields.Cond);
   Dir->setCombinedNextLowerBound(Exprs.DistCombinedFields.NLB);
   Dir->setCombinedNextUpperBound(Exprs.DistCombinedFields.NUB);
+  Dir->setCombinedDistCond(Exprs.DistCombinedFields.DistCond);
+  Dir->setCombinedParForInDistCond(Exprs.DistCombinedFields.ParForInDistCond);
   Dir->HasCancel = HasCancel;
   return Dir;
 }
 
 OMPTeamsDistributeParallelForDirective *
 OMPTeamsDistributeParallelForDirective::CreateEmpty(const ASTContext &C,
                                                     unsigned NumClauses,
                                                     unsigned CollapsedNum,
OMPTargetTeamsDistributeParallelForDirective::Create(
   Dir->setPreInits(Exprs.PreInits);
   Dir->setCombinedLowerBoundVariable(Exprs.DistCombinedFields.LB);
   Dir->setCombinedUpperBoundVariable(Exprs.DistCombinedFields.UB);
   Dir->setCombinedEnsureUpperBound(Exprs.DistCombinedFields.EUB);
   Dir->setCombinedInit(Exprs.DistCombinedFields.Init);
   Dir->setCombinedCond(Exprs.DistCombinedFields.Cond);
   Dir->setCombinedNextLowerBound(Exprs.DistCombinedFields.NLB);
   Dir->setCombinedNextUpperBound(Exprs.DistCombinedFields.NUB);
+  Dir->setCombinedDistCond(Exprs.DistCombinedFields.DistCond);
+  Dir->setCombinedParForInDistCond(Exprs.DistCombinedFields.ParForInDistCond);
   Dir->HasCancel = HasCancel;
   return Dir;
 }
 
 OMPTargetTeamsDistributeParallelForDirective *
 OMPTargetTeamsDistributeParallelForDirective::CreateEmpty(const ASTContext &C,
                                                           unsigned NumClauses,
                                                           unsigned CollapsedNum,
OMPTargetTeamsDistributeParallelForSimdDirective::Create(
   Dir->setPreInits(Exprs.PreInits);
   Dir->setCombinedLowerBoundVariable(Exprs.DistCombinedFields.LB);
   Dir->setCombinedUpperBoundVariable(Exprs.DistCombinedFields.UB);
   Dir->setCombinedEnsureUpperBound(Exprs.DistCombinedFields.EUB);
   Dir->setCombinedInit(Exprs.DistCombinedFields.Init);
   Dir->setCombinedCond(Exprs.DistCombinedFields.Cond);
   Dir->setCombinedNextLowerBound(Exprs.DistCombinedFields.NLB);
   Dir->setCombinedNextUpperBound(Exprs.DistCombinedFields.NUB);
+  Dir->setCombinedDistCond(Exprs.DistCombinedFields.DistCond);
+  Dir->setCombinedParForInDistCond(Exprs.DistCombinedFields.ParForInDistCond);
   return Dir;
 }
 
 OMPTargetTeamsDistributeParallelForSimdDirective *
 OMPTargetTeamsDistributeParallelForSimdDirective::CreateEmpty(
     const ASTContext &C, unsigned NumClauses, unsigned CollapsedNum,
     EmptyShell) {
   auto Size =

lib/CodeGen/CGOpenMPRuntime.h

public:
   /// Check if the specified \a ScheduleKind is static non-chunked.
   /// This kind of distribute directive is emitted without outer loop.
   /// \param ScheduleKind Schedule kind specified in the 'dist_schedule' clause.
   /// \param Chunked True if chunk is specified in the clause.
   ///
   virtual bool isStaticNonchunked(OpenMPDistScheduleClauseKind ScheduleKind,
                                   bool Chunked) const;
 
+  /// Check if the specified \a ScheduleKind is static chunked.
+  /// \param ScheduleKind Schedule kind specified in the 'schedule' clause.
+  /// \param Chunked True if chunk is specified in the clause.
+  ///
+  virtual bool isStaticChunked(OpenMPScheduleClauseKind ScheduleKind,
+                               bool Chunked) const;
+
+  /// Check if the specified \a ScheduleKind is static non-chunked.
+  /// \param ScheduleKind Schedule kind specified in the 'dist_schedule' clause.
+  /// \param Chunked True if chunk is specified in the clause.
+  ///
+  virtual bool isStaticChunked(OpenMPDistScheduleClauseKind ScheduleKind,
+                               bool Chunked) const;
+
   /// Check if the specified \a ScheduleKind is dynamic.
   /// This kind of worksharing directive is emitted without outer loop.
   /// \param ScheduleKind Schedule Kind specified in the 'schedule' clause.
   ///
   virtual bool isDynamic(OpenMPScheduleClauseKind ScheduleKind) const;
 
   /// struct with the values to be passed to the dispatch runtime function
   struct DispatchRTInput {
public:
   /// dist_schedule clause.
   virtual void getDefaultDistScheduleAndChunk(CodeGenFunction &CGF,
       const OMPLoopDirective &S, OpenMPDistScheduleClauseKind &ScheduleKind,
       llvm::Value *&Chunk) const {}
 
   /// Choose default schedule type and chunk value for the
   /// schedule clause.
   virtual void getDefaultScheduleAndChunk(CodeGenFunction &CGF,
-      const OMPLoopDirective &S, OpenMPScheduleClauseKind &ScheduleKind,
+      const OMPLoopDirective &S, OpenMPScheduleTy &ScheduleKind,
       llvm::Value *&Chunk) const {}
 
   /// Emits call of the outlined function with the provided arguments,
   /// translating these arguments to correct target-specific arguments.
   virtual void
   emitOutlinedFunctionCall(CodeGenFunction &CGF, SourceLocation Loc,
                            llvm::Value *OutlinedFn,
                            ArrayRef<llvm::Value *> Args = llvm::None) const;

lib/CodeGen/CGOpenMPRuntime.cpp

 }
 
 bool CGOpenMPRuntime::isStaticNonchunked(
     OpenMPDistScheduleClauseKind ScheduleKind, bool Chunked) const {
   OpenMPSchedType Schedule = getRuntimeSchedule(ScheduleKind, Chunked);
   return Schedule == OMP_dist_sch_static;
 }
 
+bool CGOpenMPRuntime::isStaticChunked(OpenMPScheduleClauseKind ScheduleKind,
+                                      bool Chunked) const {
+  OpenMPSchedType Schedule =
+      getRuntimeSchedule(ScheduleKind, Chunked, /*Ordered=*/false);
+  return Schedule == OMP_sch_static_chunked;
+}
+
+bool CGOpenMPRuntime::isStaticChunked(
+    OpenMPDistScheduleClauseKind ScheduleKind, bool Chunked) const {
+  OpenMPSchedType Schedule = getRuntimeSchedule(ScheduleKind, Chunked);
+  return Schedule == OMP_dist_sch_static_chunked;
+}
 
 bool CGOpenMPRuntime::isDynamic(OpenMPScheduleClauseKind ScheduleKind) const {
   OpenMPSchedType Schedule =
       getRuntimeSchedule(ScheduleKind, /*Chunked=*/false, /*Ordered=*/false);
   assert(Schedule != OMP_sch_static_chunked && "cannot be chunked here");
   return Schedule != OMP_sch_static;
 }
 

lib/CodeGen/CGOpenMPRuntimeNVPTX.h

public:
 
   /// Choose a default value for the dist_schedule clause.
   void getDefaultDistScheduleAndChunk(CodeGenFunction &CGF,
       const OMPLoopDirective &S, OpenMPDistScheduleClauseKind &ScheduleKind,
       llvm::Value *&Chunk) const override;
 
   /// Choose a default value for the schedule clause.
   void getDefaultScheduleAndChunk(CodeGenFunction &CGF,
-      const OMPLoopDirective &S, OpenMPScheduleClauseKind &ScheduleKind,
+      const OMPLoopDirective &S, OpenMPScheduleTy &ScheduleKind,
       llvm::Value *&Chunk) const override;
 
 private:
   /// Track the execution mode when codegening directives within a target
   /// region. The appropriate mode (SPMD/NON-SPMD) is set on entry to the
   /// target region and used by containing directives such as 'parallel'
   /// to emit optimized code.
   ExecutionMode CurrentExecutionMode = EM_Unknown;

lib/CodeGen/CGOpenMPRuntimeNVPTX.cpp

if (getExecutionMode() == CGOpenMPRuntimeNVPTX::EM_SPMD) {
     return;
   }
   CGOpenMPRuntime::getDefaultDistScheduleAndChunk(
       CGF, S, ScheduleKind, Chunk);
 }
 
 void CGOpenMPRuntimeNVPTX::getDefaultScheduleAndChunk(
     CodeGenFunction &CGF, const OMPLoopDirective &S,
-    OpenMPScheduleClauseKind &ScheduleKind,
+    OpenMPScheduleTy &ScheduleKind,
     llvm::Value *&Chunk) const {
-  ScheduleKind = OMPC_SCHEDULE_static;
+  ScheduleKind.Schedule = OMPC_SCHEDULE_static;
+  ScheduleKind.HasChunkOne = true;
   Chunk = CGF.Builder.getIntN(CGF.getContext().getTypeSize(
       S.getIterationVariable()->getType()), 1);
 }

lib/CodeGen/CGStmtOpenMP.cpp

void CodeGenFunction::EmitOMPDistributeOuterLoop(
   const bool IVSigned = IVExpr->getType()->hasSignedIntegerRepresentation();
 
   CGOpenMPRuntime::StaticRTInput StaticInit(
       IVSize, IVSigned, /* Ordered = */ false, LoopArgs.IL, LoopArgs.LB,
       LoopArgs.UB, LoopArgs.ST, LoopArgs.Chunk);
   RT.emitDistributeStaticInit(*this, S.getBeginLoc(), ScheduleKind, StaticInit);
 
   // for combined 'distribute' and 'for' the increment expression of distribute
-  // is store in DistInc. For 'distribute' alone, it is in Inc.
+  // is stored in DistInc. For 'distribute' alone, it is in Inc.
   Expr *IncExpr;
   if (isOpenMPLoopBoundSharingDirective(S.getDirectiveKind()))
     IncExpr = S.getDistInc();
   else
     IncExpr = S.getInc();
 
   // this routine is shared by 'omp distribute parallel for' and
   // 'omp distribute': select the right EUB expression depending on the
// Emit 'then' code.
           Chunk = EmitScalarExpr(Ch);
           Chunk = EmitScalarConversion(Chunk, Ch->getType(),
                                        S.getIterationVariable()->getType(),
                                        S.getBeginLoc());
         }
       } else {
         // Default behaviour for schedule clause.
         CGM.getOpenMPRuntime().getDefaultScheduleAndChunk(
-            *this, S, ScheduleKind.Schedule, Chunk);
+            *this, S, ScheduleKind, Chunk);
       }
       const unsigned IVSize = getContext().getTypeSize(IVExpr->getType());
       const bool IVSigned = IVExpr->getType()->hasSignedIntegerRepresentation();
       // OpenMP 4.5, 2.7.1 Loop Construct, Description.
       // If the static schedule kind is specified or if the ordered clause is
       // specified, and if no monotonic modifier is specified, the effect will
       // be as if the monotonic modifier was specified.
       if (RT.isStaticNonchunked(ScheduleKind.Schedule,
// Emit 'then' code.
                          [](CodeGenFunction &) {});
         EmitBlock(LoopExit.getBlock());
         // Tell the runtime we are done.
         auto &&CodeGen = [&S](CodeGenFunction &CGF) {
           CGF.CGM.getOpenMPRuntime().emitForStaticFinish(CGF, S.getEndLoc(),
                                                          S.getDirectiveKind());
         };
         OMPCancelStack.emitExit(*this, S.getDirectiveKind(), CodeGen);
+      } else if (RT.isStaticChunked(ScheduleKind.Schedule,
+                                    /* Chunked */ Chunk != nullptr) &&
+                 ScheduleKind.HasChunkOne &&
+                 isOpenMPLoopBoundSharingDirective(S.getDirectiveKind())) {
+        if (isOpenMPSimdDirective(S.getDirectiveKind()))
+          EmitOMPSimdInit(S, /*IsMonotonic=*/true);
+        CGOpenMPRuntime::StaticRTInput StaticInit(
+            IVSize, IVSigned, Ordered, IL.getAddress(), LB.getAddress(),
+            UB.getAddress(), ST.getAddress(), Chunk);
+        RT.emitForStaticInit(*this, S.getBeginLoc(), S.getDirectiveKind(),
+                             ScheduleKind, StaticInit);
+        JumpDest LoopExit =
+            getJumpDestInCurrentScope(createBasicBlock("omp.loop.exit"));
+        // IV = LB;
+        EmitIgnoredExpr(S.getInit());
+
+        // Generate the following loop:
+        //
+        // while (IV <= PrevUB) {
+        //   BODY;
+        //   IV += ST;
+        // }
+        EmitOMPInnerLoop(S, LoopScope.requiresCleanups(), S.getCombinedParForInDistCond(),
+                         S.getDistInc(),
+                         [&S, LoopExit](CodeGenFunction &CGF) {
+                           CGF.EmitOMPLoopBody(S, LoopExit);
+                           CGF.EmitStopPoint(&S);
+                         },
+                         [&](CodeGenFunction &) {});
+        EmitBlock(LoopExit.getBlock());
+        // Tell the runtime we are done.
+        auto &&CodeGen = [&S](CodeGenFunction &CGF) {
+          CGF.CGM.getOpenMPRuntime().emitForStaticFinish(CGF, S.getEndLoc(),
+                                                         S.getDirectiveKind());
+        };
+        OMPCancelStack.emitExit(*this, S.getDirectiveKind(), CodeGen);
       } else {
         const bool IsMonotonic =
             Ordered || ScheduleKind.Schedule == OMPC_SCHEDULE_static ||
             ScheduleKind.Schedule == OMPC_SCHEDULE_unknown ||
             ScheduleKind.M1 == OMPC_SCHEDULE_MODIFIER_monotonic ||
             ScheduleKind.M2 == OMPC_SCHEDULE_MODIFIER_monotonic;
         // Emit the outer loop, which requests its work chunk [LB..UB] from
         // runtime and runs the inner loop to process it.
// Emit 'then' code.
         CGOpenMPRuntime::StaticRTInput StaticInit(
             IVSize, IVSigned, /* Ordered = */ false, IL.getAddress(),
             LB.getAddress(), UB.getAddress(), ST.getAddress());
         RT.emitDistributeStaticInit(*this, S.getBeginLoc(), ScheduleKind,
                                     StaticInit);
         JumpDest LoopExit =
             getJumpDestInCurrentScope(createBasicBlock("omp.loop.exit"));
         // UB = min(UB, GlobalUB);
         EmitIgnoredExpr(isOpenMPLoopBoundSharingDirective(S.getDirectiveKind())
                             ? S.getCombinedEnsureUpperBound()
                             : S.getEnsureUpperBound());
         // IV = LB;
         EmitIgnoredExpr(isOpenMPLoopBoundSharingDirective(S.getDirectiveKind())
                             ? S.getCombinedInit()
                             : S.getInit());
 
         const Expr *Cond =
             isOpenMPLoopBoundSharingDirective(S.getDirectiveKind())
                 ? S.getCombinedCond()
                 : S.getCond();
 
-        // for distribute alone,  codegen
+        // for distribute alone, codegen
         // while (idx <= UB) { BODY; ++idx; }
         // when combined with 'for' (e.g. as in 'distribute parallel for')
         // while (idx <= UB) { <CodeGen rest of pragma>; idx += ST; }
         EmitOMPInnerLoop(S, LoopScope.requiresCleanups(), Cond, IncExpr,
                          [&S, LoopExit, &CodeGenLoop](CodeGenFunction &CGF) {
                            CodeGenLoop(CGF, S, LoopExit);
                          },
                          [](CodeGenFunction &) {});
         EmitBlock(LoopExit.getBlock());
         // Tell the runtime we are done.
         RT.emitForStaticFinish(*this, S.getBeginLoc(), S.getDirectiveKind());
+      } else if (RT.isStaticChunked(ScheduleKind,
+                                    /* Chunked */ Chunk != nullptr) &&
+                 isOpenMPLoopBoundSharingDirective(S.getDirectiveKind())) {
+        // Generate a single loop when distribute is in combination with
+        // other worksharing pragmas like parallel for.
+        if (isOpenMPSimdDirective(S.getDirectiveKind()))
+          EmitOMPSimdInit(S, /*IsMonotonic=*/true);
+        CGOpenMPRuntime::StaticRTInput StaticInit(
+            IVSize, IVSigned, /* Ordered = */ false, IL.getAddress(), LB.getAddress(),
+            UB.getAddress(), ST.getAddress(), Chunk);
+        RT.emitDistributeStaticInit(*this, S.getBeginLoc(), ScheduleKind,
+                                    StaticInit);
+
+        JumpDest LoopExit =
+            getJumpDestInCurrentScope(createBasicBlock("omp.loop.exit"));
+        // UB = min(UB, GlobalUB);
+        EmitIgnoredExpr(S.getCombinedEnsureUpperBound());
+
+        // IV = LB;
+        EmitIgnoredExpr(S.getCombinedInit());
+
+        // IV < GlobalUB;
+        const Expr *Cond = S.getCombinedDistCond();
+
+        // Generate the following loop:
+        //
+        // while (IV <= GlobalUB) {
+        //   <CodeGen rest of pragma>(LB, UB);
+        //   LB += ST;
+        //   UB += ST;
+        //   UB = min(UB, GlobalUB);
+        //   IV = LB;
+        // }
+        //
+        EmitOMPInnerLoop(S, LoopScope.requiresCleanups(), Cond, IncExpr,
+                         [&S, LoopExit, &CodeGenLoop](CodeGenFunction &CGF) {
+                           CodeGenLoop(CGF, S, LoopExit);
+                         },
+                         [&S](CodeGenFunction &CGF) {
+                           CGF.EmitIgnoredExpr(S.getCombinedNextLowerBound());
+                           CGF.EmitIgnoredExpr(S.getCombinedNextUpperBound());
+                           CGF.EmitIgnoredExpr(S.getCombinedEnsureUpperBound());
+                           CGF.EmitIgnoredExpr(S.getCombinedInit());
+                         });
+        EmitBlock(LoopExit.getBlock());
+        // Tell the runtime we are done.
+        RT.emitForStaticFinish(*this, S.getBeginLoc(), S.getDirectiveKind());
       } else {
         // Emit the outer loop, which requests its work chunk [LB..UB] from
         // runtime and runs the inner loop to process it.
         const OMPLoopArguments LoopArguments = {
             LB.getAddress(), UB.getAddress(), ST.getAddress(), IL.getAddress(),
             Chunk};
         EmitOMPDistributeOuterLoop(ScheduleKind, S, LoopScope, LoopArguments,
                                    CodeGenLoop);

lib/Sema/SemaOpenMP.cpp

OpenMPDirectiveKind getDirective(unsigned Level) const {
     return Stack.back().first[Level].Directive;
   }
   /// Returns parent directive.
   OpenMPDirectiveKind getParentDirective() const {
     if (isStackEmpty() || Stack.back().first.size() == 1)
       return OMPD_unknown;
     return std::next(Stack.back().first.rbegin())->Directive;
   }
-  
   /// Add requires decl to internal vector
   void addRequiresDecl(OMPRequiresDecl *RD) {
     RequiresDecls.push_back(RD);
   }
 
   /// Checks for a duplicate clause amongst previously declared requires
   /// directives
   bool hasDuplicateRequiresClause(ArrayRef<OMPClause *> ClauseList) const {
for (OMPClause *CNew : ClauseList) {
                 << getOpenMPClauseName(CPrev->getClauseKind());
             IsDuplicate = true;
           }
         }
       }
     }
     return IsDuplicate;
   }
-  
   /// Set default data sharing attribute to none.
   void setDefaultDSANone(SourceLocation Loc) {
     assert(!isStackEmpty());
     Stack.back().first.back().DefaultAttr = DSA_none;
     Stack.back().first.back().DefaultAttrLoc = Loc;
   }
   /// Set default data sharing attribute to shared.
   void setDefaultDSAShared(SourceLocation Loc) {
checkOpenMPLoop(OpenMPDirectiveKind DKind, Expr *CollapseLoopCountExpr,
   // Loop condition (IV < NumIterations) or (IV <= UB) for worksharing loops.
   SourceLocation CondLoc = AStmt->getBeginLoc();
   ExprResult Cond =
       (isOpenMPWorksharingDirective(DKind) ||
        isOpenMPTaskLoopDirective(DKind) || isOpenMPDistributeDirective(DKind))
           ? SemaRef.BuildBinOp(CurScope, CondLoc, BO_LE, IV.get(), UB.get())
           : SemaRef.BuildBinOp(CurScope, CondLoc, BO_LT, IV.get(),
                                NumIterations.get());
+  ExprResult CombDistCond;
+  if (isOpenMPLoopBoundSharingDirective(DKind)) {
+    CombDistCond =
+        SemaRef.BuildBinOp(CurScope, CondLoc, BO_LE, IV.get(), LastIteration.get());
+  }
+
   ExprResult CombCond;
   if (isOpenMPLoopBoundSharingDirective(DKind)) {
     CombCond =
         SemaRef.BuildBinOp(CurScope, CondLoc, BO_LE, IV.get(), CombUB.get());
   }
   // Loop increment (IV = IV + 1)
   SourceLocation IncLoc = AStmt->getBeginLoc();
   ExprResult Inc =
if (isOpenMPWorksharingDirective(DKind) || isOpenMPTaskLoopDirective(DKind) ||
     }
   }
 
   // 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 = AStmt->getBeginLoc();
-  ExprResult DistCond, DistInc, PrevEUB;
+  ExprResult DistCond, DistInc, PrevEUB, ParForInDistCond;
   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 = AStmt->getBeginLoc();
     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 IV <= PrevEUB to be used in parallel for is in combination with
+    // a distribute directive with schedule(static, 1)
+    ParForInDistCond =
+        SemaRef.BuildBinOp(CurScope, CondLoc, BO_LE, IV.get(), PrevUB.get());
   }
 
   // Build updates and final values of the loop counters.
   bool HasErrors = false;
   Built.Counters.resize(NestedLoopCount);
   Built.Inits.resize(NestedLoopCount);
   Built.Updates.resize(NestedLoopCount);
   Built.Finals.resize(NestedLoopCount);
checkOpenMPLoop(OpenMPDirectiveKind DKind, Expr *CollapseLoopCountExpr,
   Built.PrevEUB = PrevEUB.get();
   Built.DistCombinedFields.LB = CombLB.get();
   Built.DistCombinedFields.UB = CombUB.get();
   Built.DistCombinedFields.EUB = CombEUB.get();
   Built.DistCombinedFields.Init = CombInit.get();
   Built.DistCombinedFields.Cond = CombCond.get();
   Built.DistCombinedFields.NLB = CombNextLB.get();
   Built.DistCombinedFields.NUB = CombNextUB.get();
+  Built.DistCombinedFields.DistCond = CombDistCond.get();
+  Built.DistCombinedFields.ParForInDistCond = ParForInDistCond.get();
 
   return NestedLoopCount;
 }
 
 static Expr *getCollapseNumberExpr(ArrayRef<OMPClause *> Clauses) {
   auto CollapseClauses =
       OMPExecutableDirective::getClausesOfKind<OMPCollapseClause>(Clauses);
   if (CollapseClauses.begin() != CollapseClauses.end())

lib/Serialization/ASTReaderStmt.cpp

if (isOpenMPLoopBoundSharingDirective(D->getDirectiveKind())) {
     D->setPrevEnsureUpperBound(Record.readSubExpr());
     D->setCombinedLowerBoundVariable(Record.readSubExpr());
     D->setCombinedUpperBoundVariable(Record.readSubExpr());
     D->setCombinedEnsureUpperBound(Record.readSubExpr());
     D->setCombinedInit(Record.readSubExpr());
     D->setCombinedCond(Record.readSubExpr());
     D->setCombinedNextLowerBound(Record.readSubExpr());
     D->setCombinedNextUpperBound(Record.readSubExpr());
+    D->setCombinedDistCond(Record.readSubExpr());
+    D->setCombinedParForInDistCond(Record.readSubExpr());
   }
   SmallVector<Expr *, 4> Sub;
   unsigned CollapsedNum = D->getCollapsedNumber();
   Sub.reserve(CollapsedNum);
   for (unsigned i = 0; i < CollapsedNum; ++i)
     Sub.push_back(Record.readSubExpr());
   D->setCounters(Sub);
   Sub.clear();

lib/Serialization/ASTWriterStmt.cpp

if (isOpenMPLoopBoundSharingDirective(D->getDirectiveKind())) {
     Record.AddStmt(D->getPrevEnsureUpperBound());
     Record.AddStmt(D->getCombinedLowerBoundVariable());
     Record.AddStmt(D->getCombinedUpperBoundVariable());
     Record.AddStmt(D->getCombinedEnsureUpperBound());
     Record.AddStmt(D->getCombinedInit());
     Record.AddStmt(D->getCombinedCond());
     Record.AddStmt(D->getCombinedNextLowerBound());
     Record.AddStmt(D->getCombinedNextUpperBound());
+    Record.AddStmt(D->getCombinedDistCond());
+    Record.AddStmt(D->getCombinedParForInDistCond());
   }
   for (auto I : D->counters()) {
     Record.AddStmt(I);
   }
   for (auto I : D->private_counters()) {
     Record.AddStmt(I);
   }
   for (auto I : D->inits()) {

test/OpenMP/distribute_parallel_for_codegen.cpp

a[i] = b[i] + c[i];
     // 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_CAPT_EXPR:%.capture_expr.1]] = alloca
       // LAMBDA-DAG: [[OMP_LB:%.omp.comb.lb]] = alloca
       // LAMBDA-DAG: [[OMP_UB:%.omp.comb.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: [[NUM_IT_1:%.+]] = load{{.+}} [[OMP_CAPT_EXPR]],
       // 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-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_CAPT_EXPR]],
       // 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:.+]]
+      // LAMBDA: br {{.+}} [[CMP_IV_UB]], 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]]:
+      // Update UB
+      // LAMBDA-DAG: [[OMP_UB_VAL_6:%.+]] = load{{.+}}, {{.+}} [[OMP_UB]],
+      // LAMBDA: [[OMP_EXPR_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_CAPT_EXPR]],
+      // LAMBDA-DAG: [[CMP_UB_NUM_IT_1:%.+]] = icmp sgt {{.+}}[[OMP_UB_VAL_6]], [[OMP_EXPR_VAL]]
+      // LAMBDA: br {{.+}} [[CMP_UB_NUM_IT_1]], label %[[EUB_TRUE_1:.+]], label %[[EUB_FALSE_1:.+]]
+      // LAMBDA-DAG: [[EUB_TRUE_1]]:
+      // LAMBDA: [[NUM_IT_3:%.+]] = load{{.+}} [[OMP_CAPT_EXPR]],
+      // LAMBDA: br label %[[EUB_END_1:.+]]
+      // LAMBDA-DAG: [[EUB_FALSE_1]]:
+      // LAMBDA: [[OMP_UB_VAL3:%.+]] = load{{.+}} [[OMP_UB]],
+      // LAMBDA: br label %[[EUB_END_1]]
+      // LAMBDA-DAG: [[EUB_END_1]]:
+      // LAMBDA-DAG: [[EUB_RES_1:%.+]] = phi{{.+}} [ [[NUM_IT_3]], %[[EUB_TRUE_1]] ], [ [[OMP_UB_VAL3]], %[[EUB_FALSE_1]] ]
+      // LAMBDA: store{{.+}} [[EUB_RES_1]], {{.+}}* [[OMP_UB]],
+
+      // Store LB in IV
+      // LAMBDA-DAG: [[OMP_LB_VAL_3:%.+]] = load{{.+}}, {{.+}} [[OMP_LB]],
+      // LAMBDA: store{{.+}} [[OMP_LB_VAL_3]], {{.+}}* [[OMP_IV]],
+
+      // LAMBDA: [[DIST_INNER_LOOP_END]]:
+      // LAMBDA: br label %[[LOOP_EXIT:.+]]
+
+      // loop exit
+      // LAMBDA: [[LOOP_EXIT]]:
       // 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];
       }();
     }
#else
   // 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_CAPT_EXPR:%.capture_expr.1]] = alloca
     // CHECK-DAG: [[OMP_LB:%.omp.comb.lb]] = alloca
     // CHECK-DAG: [[OMP_UB:%.omp.comb.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: [[NUM_IT_1:%.+]] = load{{.+}} [[OMP_CAPT_EXPR]],
     // 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-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_CAPT_EXPR]],
     // 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: br {{.+}} [[CMP_IV_UB]], 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]]:
+    // Update UB
+    // CHECK-DAG: [[OMP_UB_VAL_6:%.+]] = load{{.+}}, {{.+}} [[OMP_UB]],
+    // CHECK: [[OMP_EXPR_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_CAPT_EXPR]],
+    // CHECK-DAG: [[CMP_UB_NUM_IT_1:%.+]] = icmp sgt {{.+}}[[OMP_UB_VAL_6]], [[OMP_EXPR_VAL]]
+    // CHECK: br {{.+}} [[CMP_UB_NUM_IT_1]], label %[[EUB_TRUE_1:.+]], label %[[EUB_FALSE_1:.+]]
+    // CHECK-DAG: [[EUB_TRUE_1]]:
+    // CHECK: [[NUM_IT_3:%.+]] = load{{.+}} [[OMP_CAPT_EXPR]],
+    // CHECK: br label %[[EUB_END_1:.+]]
+    // CHECK-DAG: [[EUB_FALSE_1]]:
+    // CHECK: [[OMP_UB_VAL3:%.+]] = load{{.+}} [[OMP_UB]],
+    // CHECK: br label %[[EUB_END_1]]
+    // CHECK-DAG: [[EUB_END_1]]:
+    // CHECK-DAG: [[EUB_RES_1:%.+]] = phi{{.+}} [ [[NUM_IT_3]], %[[EUB_TRUE_1]] ], [ [[OMP_UB_VAL3]], %[[EUB_FALSE_1]] ]
+    // CHECK: store{{.+}} [[EUB_RES_1]], {{.+}}* [[OMP_UB]],
+
+    // Store LB in IV
+    // CHECK-DAG: [[OMP_LB_VAL_3:%.+]] = load{{.+}}, {{.+}} [[OMP_LB]],
+    // CHECK: store{{.+}} [[OMP_LB_VAL_3]], {{.+}}* [[OMP_IV]],
+
+    // CHECK: [[DIST_INNER_LOOP_END]]:
+    // CHECK: br label %[[LOOP_EXIT:.+]]
+
+    // loop exit
+    // CHECK: [[LOOP_EXIT]]:
     // 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

 // 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_CAPT_EXPR:%.capture_expr.1]] = alloca
 // CHECK-DAG: [[OMP_LB:%.omp.comb.lb]] = alloca
 // CHECK-DAG: [[OMP_UB:%.omp.comb.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: [[NUM_IT_1:%.+]] = load{{.+}} [[OMP_CAPT_EXPR]],
 // 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-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_CAPT_EXPR]],
 // 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: br {{.+}} [[CMP_IV_UB]], 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]]:
+// Update UB
+// CHECK-DAG: [[OMP_UB_VAL_6:%.+]] = load{{.+}}, {{.+}} [[OMP_UB]],
+// CHECK: [[OMP_EXPR_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_CAPT_EXPR]],
+// CHECK-DAG: [[CMP_UB_NUM_IT_1:%.+]] = icmp sgt {{.+}}[[OMP_UB_VAL_6]], [[OMP_EXPR_VAL]]
+// CHECK: br {{.+}} [[CMP_UB_NUM_IT_1]], label %[[EUB_TRUE_1:.+]], label %[[EUB_FALSE_1:.+]]
+// CHECK-DAG: [[EUB_TRUE_1]]:
+// CHECK: [[NUM_IT_3:%.+]] = load{{.+}} [[OMP_CAPT_EXPR]],
+// CHECK: br label %[[EUB_END_1:.+]]
+// CHECK-DAG: [[EUB_FALSE_1]]:
+// CHECK: [[OMP_UB_VAL3:%.+]] = load{{.+}} [[OMP_UB]],
+// CHECK: br label %[[EUB_END_1]]
+// CHECK-DAG: [[EUB_END_1]]:
+// CHECK-DAG: [[EUB_RES_1:%.+]] = phi{{.+}} [ [[NUM_IT_3]], %[[EUB_TRUE_1]] ], [ [[OMP_UB_VAL3]], %[[EUB_FALSE_1]] ]
+// CHECK: store{{.+}} [[EUB_RES_1]], {{.+}}* [[OMP_UB]],
+
+// Store LB in IV
+// CHECK-DAG: [[OMP_LB_VAL_3:%.+]] = load{{.+}}, {{.+}} [[OMP_LB]],
+// CHECK: store{{.+}} [[OMP_LB_VAL_3]], {{.+}}* [[OMP_IV]],
+
+// CHECK: [[DIST_INNER_LOOP_END]]:
+// CHECK: br label %[[LOOP_EXIT:.+]]
+
+// loop exit
+// CHECK: [[LOOP_EXIT]]:
 // 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 {{.+}})
 

test/OpenMP/distribute_parallel_for_simd_codegen.cpp

a[i] = b[i] + c[i];
     // LAMBDA: define{{.+}} void [[OFFLOADING_FUN_3]](
     // LAMBDA: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 5, {{.+}}* [[OMP_OUTLINED_3:@.+]] to {{.+}})
 
     #pragma omp distribute parallel for simd 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_CAPT_EXPR:%.capture_expr.1]] = alloca
       // LAMBDA-DAG: [[OMP_LB:%.omp.comb.lb]] = alloca
       // LAMBDA-DAG: [[OMP_UB:%.omp.comb.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: [[NUM_IT_1:%.+]] = load{{.+}} [[OMP_CAPT_EXPR]],
       // 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-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_CAPT_EXPR]],
       // 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:.+]]
+      // LAMBDA: br {{.+}} [[CMP_IV_UB]], 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]]:
+      // Update UB
+      // LAMBDA-DAG: [[OMP_UB_VAL_6:%.+]] = load{{.+}}, {{.+}} [[OMP_UB]],
+      // LAMBDA: [[OMP_EXPR_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_CAPT_EXPR]],
+      // LAMBDA-DAG: [[CMP_UB_NUM_IT_1:%.+]] = icmp sgt {{.+}}[[OMP_UB_VAL_6]], [[OMP_EXPR_VAL]]
+      // LAMBDA: br {{.+}} [[CMP_UB_NUM_IT_1]], label %[[EUB_TRUE_1:.+]], label %[[EUB_FALSE_1:.+]]
+      // LAMBDA-DAG: [[EUB_TRUE_1]]:
+      // LAMBDA: [[NUM_IT_3:%.+]] = load{{.+}} [[OMP_CAPT_EXPR]],
+      // LAMBDA: br label %[[EUB_END_1:.+]]
+      // LAMBDA-DAG: [[EUB_FALSE_1]]:
+      // LAMBDA: [[OMP_UB_VAL3:%.+]] = load{{.+}} [[OMP_UB]],
+      // LAMBDA: br label %[[EUB_END_1]]
+      // LAMBDA-DAG: [[EUB_END_1]]:
+      // LAMBDA-DAG: [[EUB_RES_1:%.+]] = phi{{.+}} [ [[NUM_IT_3]], %[[EUB_TRUE_1]] ], [ [[OMP_UB_VAL3]], %[[EUB_FALSE_1]] ]
+      // LAMBDA: store{{.+}} [[EUB_RES_1]], {{.+}}* [[OMP_UB]],
+
+      // Store LB in IV
+      // LAMBDA-DAG: [[OMP_LB_VAL_3:%.+]] = load{{.+}}, {{.+}} [[OMP_LB]],
+      // LAMBDA: store{{.+}} [[OMP_LB_VAL_3]], {{.+}}* [[OMP_IV]],
+
+      // LAMBDA: [[DIST_INNER_LOOP_END]]:
+      // LAMBDA: br label %[[LOOP_EXIT:.+]]
+
+      // loop exit
+      // LAMBDA: [[LOOP_EXIT]]:
       // 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];
       }();
     }
#else
   // CHECK: define{{.+}} void [[OFFLOADING_FUN_3]](
   // CHECK: call {{.*}}void {{.+}} @__kmpc_fork_teams({{.+}}, i32 5, {{.+}}* [[OMP_OUTLINED_3:@.+]] to {{.+}})
 
   #pragma omp distribute parallel for simd 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_CAPT_EXPR:%.capture_expr.1]] = alloca
     // CHECK-DAG: [[OMP_LB:%.omp.comb.lb]] = alloca
     // CHECK-DAG: [[OMP_UB:%.omp.comb.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: [[NUM_IT_1:%.+]] = load{{.+}} [[OMP_CAPT_EXPR]],
     // 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-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_CAPT_EXPR]],
     // 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: br {{.+}} [[CMP_IV_UB]], 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]]:
+    // Update UB
+    // CHECK-DAG: [[OMP_UB_VAL_6:%.+]] = load{{.+}}, {{.+}} [[OMP_UB]],
+    // CHECK: [[OMP_EXPR_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_CAPT_EXPR]],
+    // CHECK-DAG: [[CMP_UB_NUM_IT_1:%.+]] = icmp sgt {{.+}}[[OMP_UB_VAL_6]], [[OMP_EXPR_VAL]]
+    // CHECK: br {{.+}} [[CMP_UB_NUM_IT_1]], label %[[EUB_TRUE_1:.+]], label %[[EUB_FALSE_1:.+]]
+    // CHECK-DAG: [[EUB_TRUE_1]]:
+    // CHECK: [[NUM_IT_3:%.+]] = load{{.+}} [[OMP_CAPT_EXPR]],
+    // CHECK: br label %[[EUB_END_1:.+]]
+    // CHECK-DAG: [[EUB_FALSE_1]]:
+    // CHECK: [[OMP_UB_VAL3:%.+]] = load{{.+}} [[OMP_UB]],
+    // CHECK: br label %[[EUB_END_1]]
+    // CHECK-DAG: [[EUB_END_1]]:
+    // CHECK-DAG: [[EUB_RES_1:%.+]] = phi{{.+}} [ [[NUM_IT_3]], %[[EUB_TRUE_1]] ], [ [[OMP_UB_VAL3]], %[[EUB_FALSE_1]] ]
+    // CHECK: store{{.+}} [[EUB_RES_1]], {{.+}}* [[OMP_UB]],
+
+    // Store LB in IV
+    // CHECK-DAG: [[OMP_LB_VAL_3:%.+]] = load{{.+}}, {{.+}} [[OMP_LB]],
+    // CHECK: store{{.+}} [[OMP_LB_VAL_3]], {{.+}}* [[OMP_IV]],
+
+    // CHECK: [[DIST_INNER_LOOP_END]]:
+    // CHECK: br label %[[LOOP_EXIT:.+]]
+
+    // loop exit
+    // CHECK: [[LOOP_EXIT]]:
     // 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

 // 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_CAPT_EXPR:%.capture_expr.1]] = alloca
 // CHECK-DAG: [[OMP_LB:%.omp.comb.lb]] = alloca
 // CHECK-DAG: [[OMP_UB:%.omp.comb.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: [[NUM_IT_1:%.+]] = load{{.+}} [[OMP_CAPT_EXPR]],
 // 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-DAG: [[OMP_UB_VAL_3:%.+]] = load {{.+}} [[OMP_CAPT_EXPR]],
 // 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: br {{.+}} [[CMP_IV_UB]], 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]]:
+// Update UB
+// CHECK-DAG: [[OMP_UB_VAL_6:%.+]] = load{{.+}}, {{.+}} [[OMP_UB]],
+// CHECK: [[OMP_EXPR_VAL:%.+]] = load{{.+}}, {{.+}} [[OMP_CAPT_EXPR]],
+// CHECK-DAG: [[CMP_UB_NUM_IT_1:%.+]] = icmp sgt {{.+}}[[OMP_UB_VAL_6]], [[OMP_EXPR_VAL]]
+// CHECK: br {{.+}} [[CMP_UB_NUM_IT_1]], label %[[EUB_TRUE_1:.+]], label %[[EUB_FALSE_1:.+]]
+// CHECK-DAG: [[EUB_TRUE_1]]:
+// CHECK: [[NUM_IT_3:%.+]] = load{{.+}} [[OMP_CAPT_EXPR]],
+// CHECK: br label %[[EUB_END_1:.+]]
+// CHECK-DAG: [[EUB_FALSE_1]]:
+// CHECK: [[OMP_UB_VAL3:%.+]] = load{{.+}} [[OMP_UB]],
+// CHECK: br label %[[EUB_END_1]]
+// CHECK-DAG: [[EUB_END_1]]:
+// CHECK-DAG: [[EUB_RES_1:%.+]] = phi{{.+}} [ [[NUM_IT_3]], %[[EUB_TRUE_1]] ], [ [[OMP_UB_VAL3]], %[[EUB_FALSE_1]] ]
+// CHECK: store{{.+}} [[EUB_RES_1]], {{.+}}* [[OMP_UB]],
+
+// Store LB in IV
+// CHECK-DAG: [[OMP_LB_VAL_3:%.+]] = load{{.+}}, {{.+}} [[OMP_LB]],
+// CHECK: store{{.+}} [[OMP_LB_VAL_3]], {{.+}}* [[OMP_IV]],
+
+// CHECK: [[DIST_INNER_LOOP_END]]:
+// CHECK: br label %[[LOOP_EXIT:.+]]
+
+// loop exit
+// CHECK: [[LOOP_EXIT]]:
 // 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 {{.+}})