[OpenMP] Refactor OMPScheduleType enum.

[Meinersbur]

Summary

The OMPScheduleType enum stores the constants from libomp's internal sched_type in kmp.h and are used by several kmp API functions. The enum values have an internal structure, namely each scheduling algorithm (e.g.) exists in four variants: unordered, orderend, normerge unordered, and nomerge ordered.

This patch (basically a followup to D114940) splits the "ordered" and "nomerge" bits into separate flags, as was already done for the "monotonic" and "nonmonotonic", so we can apply bit flags operations on them. It also now contains all possible combinations according to kmp's sched_type. Deriving of the OMPScheduleType enum from clause parameters has been moved form MLIR's OpenMPToLLVMIRTranslation.cpp to OpenMPIRBuilder to make available for clang as well. Since the primary purpose of the flag is the binary interface to libomp, it has been made more private to LLVMFrontend. The primary interface for generating worksharing-loop using OpenMPIRBuilder code becomes applyWorkshareLoop which derives the OMPScheduleType automatically and calls the appropriate emitter function.

While this is mostly a NFC refactor, it still applies the following functional changes:

• The logic from OpenMPToLLVMIRTranslation to derive the OMPScheduleType also applies to clang. Most notably, it now applies the nonmonotonic flag for non-static schedules by default.
• In OpenMPToLLVMIRTranslation, the nonmonotonic default flag was previously not applied if the simd modifier was used. I assume this was a bug, since the effect was due to loop.schedule_modifier() returning mlir::omp::ScheduleModifier::none instead of llvm::Optional::None.
• In OpenMPToLLVMIRTranslation, the nonmonotonic default flag was set even if ordered was specified, in breach to what the comment before citing the OpenMP specification says. I assume this was an oversight.

The ordered flag with parameter was not considered in this patch. Changes will need to be made (e.g. adding/modifying function parameters) when support for it is added. The lengthy names of the enum values can be discussed, for the moment this is avoiding reusing previously existing enum value names such as StaticChunked to avoid confusion.

[peixin]

A few comments. Mostly nits.

clang/lib/CodeGen/CGStmtOpenMP.cpp

3767

llvm/include/llvm/Frontend/OpenMP/OMPConstants.h

87

BaseRuntime = 5,
BaseAuto = 6,

135

Why not using the following to be consistent with the name in kmp.h?
StaticBalancedChunked
GuidedSimd
RuntimeSimd

llvm/unittests/Frontend/OpenMPIRBuilderTest.cpp

2114

Nit

2254

2255

2256

mlir/test/Target/LLVMIR/openmp-llvm.mlir

809

Is this one tab?

[Meinersbur]

llvm/include/llvm/Frontend/OpenMP/OMPConstants.h

135

As mentioned in the summary. to avoid confusion by not using the original name. StaticBalancedChunked could mean either the algorithm to use (now BaseStaticBalancedChunked, as in omp_sched_t/enum kmp_sched), or that algorithm with the unordered flag set (now UnorderedStaticBalancedChunked ). I would the former because that's how the enum is structured.

The name in kmp.h for it is actually kmp_sch_static_balanced_chunked. sch for "unordered"?

[Meinersbur]

• address review

[peixin]

llvm/include/llvm/Frontend/OpenMP/OMPConstants.h

135

OK. Agree.

The name in kmp.h for it is actually kmp_sch_static_balanced_chunked. sch for "unordered"?

Yes, I think so. With simd modifier, it cannot be ordered in semantics.

clang/lib/CodeGen/CGStmtOpenMP.cpp

if (UseOMPIRBuilder) {
       const Stmt *Inner = S.getRawStmt();
       llvm::CanonicalLoopInfo *CLI =
           EmitOMPCollapsedCanonicalLoopNest(Inner, 1);
 
       llvm::OpenMPIRBuilder &OMPBuilder =
           CGM.getOpenMPRuntime().getOMPBuilder();
       llvm::OpenMPIRBuilder::InsertPointTy AllocaIP(
           AllocaInsertPt->getParent(), AllocaInsertPt->getIterator());
-      OMPBuilder.applyWorkshareLoop(Builder.getCurrentDebugLocation(), CLI,
-                                    AllocaIP, NeedsBarrier, SchedKind,
-                                    ChunkSize);
+      OMPBuilder.applyWorkshareLoop(
+          Builder.getCurrentDebugLocation(), CLI, AllocaIP, NeedsBarrier,
+          SchedKind, ChunkSize, /*HasSimdModifier=*/false,
+          /*HasMonotonicModifier=*/false, /*HasNonmonotonicModifier=*/false,
+          /*HasOrderedClause=*/false);
       return;
     }
 
     HasLastprivates = emitWorksharingDirective(CGF, S, S.hasCancel());
   };
   {
     auto LPCRegion =
         CGOpenMPRuntime::LastprivateConditionalRAII::disable(*this, S);

clang/test/OpenMP/irbuilder_for_unsigned_auto.c

 // CHECK-NEXT:    %[[DOTCOUNT:.+]] = load i32, i32* %[[DOTCOUNT_ADDR]], align 4
 // CHECK-NEXT:    br label %[[OMP_LOOP_PREHEADER:.+]]
 // CHECK-EMPTY:
 // CHECK-NEXT:  [[OMP_LOOP_PREHEADER]]:
 // CHECK-NEXT:    store i32 1, i32* %[[P_LOWERBOUND]], align 4
 // CHECK-NEXT:    store i32 %[[DOTCOUNT]], i32* %[[P_UPPERBOUND]], align 4
 // CHECK-NEXT:    store i32 1, i32* %[[P_STRIDE]], align 4
 // CHECK-NEXT:    %[[OMP_GLOBAL_THREAD_NUM:.+]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @1)
-// CHECK-NEXT:    call void @__kmpc_dispatch_init_4u(%struct.ident_t* @1, i32 %[[OMP_GLOBAL_THREAD_NUM]], i32 38, i32 1, i32 %[[DOTCOUNT]], i32 1, i32 1)
+// CHECK-NEXT:    call void @__kmpc_dispatch_init_4u(%struct.ident_t* @1, i32 %[[OMP_GLOBAL_THREAD_NUM]], i32 1073741862, i32 1, i32 %[[DOTCOUNT]], i32 1, i32 1)
 // CHECK-NEXT:    br label %[[OMP_LOOP_PREHEADER_OUTER_COND:.+]]
 // CHECK-EMPTY:
 // CHECK-NEXT:  [[OMP_LOOP_HEADER:.*]]:
 // CHECK-NEXT:    %[[OMP_LOOP_IV:.+]] = phi i32 [ %[[LB:.+]], %[[OMP_LOOP_PREHEADER_OUTER_COND]] ], [ %[[OMP_LOOP_NEXT:.+]], %[[OMP_LOOP_INC:.+]] ]
 // CHECK-NEXT:    br label %[[OMP_LOOP_COND:.+]]
 // CHECK-EMPTY:
 // CHECK-NEXT:  [[OMP_LOOP_COND]]:
 // CHECK-NEXT:    %[[UB:.+]] = load i32, i32* %[[P_UPPERBOUND]], align 4

clang/test/OpenMP/irbuilder_for_unsigned_dynamic.c

 // CHECK-NEXT:    %[[DOTCOUNT:.+]] = load i32, i32* %[[DOTCOUNT_ADDR]], align 4
 // CHECK-NEXT:    br label %[[OMP_LOOP_PREHEADER:.+]]
 // CHECK-EMPTY:
 // CHECK-NEXT:  [[OMP_LOOP_PREHEADER]]:
 // CHECK-NEXT:    store i32 1, i32* %[[P_LOWERBOUND]], align 4
 // CHECK-NEXT:    store i32 %[[DOTCOUNT]], i32* %[[P_UPPERBOUND]], align 4
 // CHECK-NEXT:    store i32 1, i32* %[[P_STRIDE]], align 4
 // CHECK-NEXT:    %[[OMP_GLOBAL_THREAD_NUM:.+]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @1)
-// CHECK-NEXT:    call void @__kmpc_dispatch_init_4u(%struct.ident_t* @1, i32 %[[OMP_GLOBAL_THREAD_NUM]], i32 35, i32 1, i32 %[[DOTCOUNT]], i32 1, i32 1)
+// CHECK-NEXT:    call void @__kmpc_dispatch_init_4u(%struct.ident_t* @1, i32 %[[OMP_GLOBAL_THREAD_NUM]], i32 1073741859, i32 1, i32 %[[DOTCOUNT]], i32 1, i32 1)
 // CHECK-NEXT:    br label %[[OMP_LOOP_PREHEADER_OUTER_COND:.+]]
 // CHECK-EMPTY:
 // CHECK-NEXT:  [[OMP_LOOP_HEADER:.*]]:
 // CHECK-NEXT:    %[[OMP_LOOP_IV:.+]] = phi i32 [ %[[LB:.+]], %[[OMP_LOOP_PREHEADER_OUTER_COND]] ], [ %[[OMP_LOOP_NEXT:.+]], %[[OMP_LOOP_INC:.+]] ]
 // CHECK-NEXT:    br label %[[OMP_LOOP_COND:.+]]
 // CHECK-EMPTY:
 // CHECK-NEXT:  [[OMP_LOOP_COND]]:
 // CHECK-NEXT:    %[[UB:.+]] = load i32, i32* %[[P_UPPERBOUND]], align 4

clang/test/OpenMP/irbuilder_for_unsigned_dynamic_chunked.c

 // CHECK-NEXT:    %[[DOTCOUNT:.+]] = load i32, i32* %[[DOTCOUNT_ADDR]], align 4
 // CHECK-NEXT:    br label %[[OMP_LOOP_PREHEADER:.+]]
 // CHECK-EMPTY:
 // CHECK-NEXT:  [[OMP_LOOP_PREHEADER]]:
 // CHECK-NEXT:    store i32 1, i32* %[[P_LOWERBOUND]], align 4
 // CHECK-NEXT:    store i32 %[[DOTCOUNT]], i32* %[[P_UPPERBOUND]], align 4
 // CHECK-NEXT:    store i32 1, i32* %[[P_STRIDE]], align 4
 // CHECK-NEXT:    %[[OMP_GLOBAL_THREAD_NUM:.+]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @1)
-// CHECK-NEXT:    call void @__kmpc_dispatch_init_4u(%struct.ident_t* @1, i32 %[[OMP_GLOBAL_THREAD_NUM]], i32 35, i32 1, i32 %[[DOTCOUNT]], i32 1, i32 5)
+// CHECK-NEXT:    call void @__kmpc_dispatch_init_4u(%struct.ident_t* @1, i32 %[[OMP_GLOBAL_THREAD_NUM]], i32 1073741859, i32 1, i32 %[[DOTCOUNT]], i32 1, i32 5)
 // CHECK-NEXT:    br label %[[OMP_LOOP_PREHEADER_OUTER_COND:.+]]
 // CHECK-EMPTY:
 // CHECK-NEXT:  [[OMP_LOOP_HEADER:.*]]:
 // CHECK-NEXT:    %[[OMP_LOOP_IV:.+]] = phi i32 [ %[[LB:.+]], %[[OMP_LOOP_PREHEADER_OUTER_COND]] ], [ %[[OMP_LOOP_NEXT:.+]], %[[OMP_LOOP_INC:.+]] ]
 // CHECK-NEXT:    br label %[[OMP_LOOP_COND:.+]]
 // CHECK-EMPTY:
 // CHECK-NEXT:  [[OMP_LOOP_COND]]:
 // CHECK-NEXT:    %[[UB:.+]] = load i32, i32* %[[P_UPPERBOUND]], align 4

clang/test/OpenMP/irbuilder_for_unsigned_runtime.c

 // CHECK-NEXT:    %[[DOTCOUNT:.+]] = load i32, i32* %[[DOTCOUNT_ADDR]], align 4
 // CHECK-NEXT:    br label %[[OMP_LOOP_PREHEADER:.+]]
 // CHECK-EMPTY:
 // CHECK-NEXT:  [[OMP_LOOP_PREHEADER]]:
 // CHECK-NEXT:    store i32 1, i32* %[[P_LOWERBOUND]], align 4
 // CHECK-NEXT:    store i32 %[[DOTCOUNT]], i32* %[[P_UPPERBOUND]], align 4
 // CHECK-NEXT:    store i32 1, i32* %[[P_STRIDE]], align 4
 // CHECK-NEXT:    %[[OMP_GLOBAL_THREAD_NUM:.+]] = call i32 @__kmpc_global_thread_num(%struct.ident_t* @1)
-// CHECK-NEXT:    call void @__kmpc_dispatch_init_4u(%struct.ident_t* @1, i32 %[[OMP_GLOBAL_THREAD_NUM]], i32 37, i32 1, i32 %[[DOTCOUNT]], i32 1, i32 1)
+// CHECK-NEXT:    call void @__kmpc_dispatch_init_4u(%struct.ident_t* @1, i32 %[[OMP_GLOBAL_THREAD_NUM]], i32 1073741861, i32 1, i32 %[[DOTCOUNT]], i32 1, i32 1)
 // CHECK-NEXT:    br label %[[OMP_LOOP_PREHEADER_OUTER_COND:.+]]
 // CHECK-EMPTY:
 // CHECK-NEXT:  [[OMP_LOOP_HEADER:.*]]:
 // CHECK-NEXT:    %[[OMP_LOOP_IV:.+]] = phi i32 [ %[[LB:.+]], %[[OMP_LOOP_PREHEADER_OUTER_COND]] ], [ %[[OMP_LOOP_NEXT:.+]], %[[OMP_LOOP_INC:.+]] ]
 // CHECK-NEXT:    br label %[[OMP_LOOP_COND:.+]]
 // CHECK-EMPTY:
 // CHECK-NEXT:  [[OMP_LOOP_COND]]:
 // CHECK-NEXT:    %[[UB:.+]] = load i32, i32* %[[P_UPPERBOUND]], align 4

llvm/include/llvm/Frontend/OpenMP/OMPConstants.h

#include "llvm/Frontend/OpenMP/OMPKinds.def"
   LLVM_MARK_AS_BITMASK_ENUM(0x7FFFFFFF)
 };
 
 #define OMP_IDENT_FLAG(Enum, ...) constexpr auto Enum = omp::IdentFlag::Enum;
 #include "llvm/Frontend/OpenMP/OMPKinds.def"
 
 /// \note This needs to be kept in sync with kmp.h enum sched_type.
 /// Todo: Update kmp.h to include this file, and remove the enums in kmp.h
-///       To complete this, more enum values will need to be moved here.
 enum class OMPScheduleType {
-  StaticChunked = 33,
-  Static = 34, // static unspecialized
-  DynamicChunked = 35,
-  GuidedChunked = 36, // guided unspecialized
-  Runtime = 37,
-  Auto = 38, // auto
-
-  StaticBalancedChunked = 45, // static with chunk adjustment (e.g., simd)
-  GuidedSimd = 46,            // guided with chunk adjustment
-  RuntimeSimd = 47,           // runtime with chunk adjustment
-
-  OrderedStaticChunked = 65,
-  OrderedStatic = 66, // ordered static unspecialized
-  OrderedDynamicChunked = 67,
-  OrderedGuidedChunked = 68,
-  OrderedRuntime = 69,
-  OrderedAuto = 70, // ordered auto
-
-  DistributeChunked = 91, // distribute static chunked
-  Distribute = 92,        // distribute static unspecialized
-
-  ModifierMonotonic =
-      (1 << 29), // Set if the monotonic schedule modifier was present
-  ModifierNonmonotonic =
-      (1 << 30), // Set if the nonmonotonic schedule modifier was present
-  ModifierMask = ModifierMonotonic | ModifierNonmonotonic,
+  // For typed comparisons, not a valid schedule
+  None = 0,
+
+  // Schedule algorithms
+  BaseStaticChunked = 1,
+  BaseStatic = 2,
+  BaseDynamicChunked = 3,
+  BaseGuidedChunked = 4,
+  BaseRuntime = 5,
+  BaseAuto = 6,
+  BaseTrapezoidal = 7,
+  BaseGreedy = 8,
+  BaseBalanced = 9,
+  BaseGuidedIterativeChunked = 10,
+  BaseGuidedAnalyticalChunked = 11,
+  BaseSteal = 12,
+
+  // with chunk adjustment (e.g., simd)
+  BaseStaticBalancedChunked = 13,
+  BaseGuidedSimd = 14,
+  BaseRuntimeSimd = 15,
+
+  // static schedules algorithims for distribute
+  BaseDistributeChunked = 27,
+  BaseDistribute = 28,
+
+  // Modifier flags to be combined with schedule algorithms
+  ModifierUnordered = (1 << 5),
+  ModifierOrdered = (1 << 6),
+  ModifierNomerge = (1 << 7),
+  ModifierMonotonic = (1 << 29),
+  ModifierNonmonotonic = (1 << 30),
+
+  // Masks combining multiple flags
+  OrderingMask = ModifierUnordered | ModifierOrdered | ModifierNomerge,
+  MonotonicityMask = ModifierMonotonic | ModifierNonmonotonic,
+  ModifierMask = OrderingMask | MonotonicityMask,
+
+  // valid schedule type values, without monotonicity flags
+  UnorderedStaticChunked = BaseStaticChunked | ModifierUnordered,   // 33
+  UnorderedStatic = BaseStatic | ModifierUnordered,                 // 34
+  UnorderedDynamicChunked = BaseDynamicChunked | ModifierUnordered, // 35
+  UnorderedGuidedChunked = BaseGuidedChunked | ModifierUnordered,   // 36
+  UnorderedRuntime = BaseRuntime | ModifierUnordered,               // 37
+  UnorderedAuto = BaseAuto | ModifierUnordered,                     // 38
+  UnorderedTrapezoidal = BaseTrapezoidal | ModifierUnordered,       // 39
+  UnorderedGreedy = BaseGreedy | ModifierUnordered,                 // 40
+  UnorderedBalanced = BaseBalanced | ModifierUnordered,             // 41
+  UnorderedGuidedIterativeChunked =
+      BaseGuidedIterativeChunked | ModifierUnordered, // 42
+  UnorderedGuidedAnalyticalChunked =
+      BaseGuidedAnalyticalChunked | ModifierUnordered, // 43
+  UnorderedSteal = BaseSteal | ModifierUnordered,      // 44
+
+  UnorderedStaticBalancedChunked =
+      BaseStaticBalancedChunked | ModifierUnordered,          // 45
+  UnorderedGuidedSimd = BaseGuidedSimd | ModifierUnordered,   // 46
+  UnorderedRuntimeSimd = BaseRuntimeSimd | ModifierUnordered, // 47
+
+  OrderedStaticChunked = BaseStaticChunked | ModifierOrdered,   // 65
+  OrderedStatic = BaseStatic | ModifierOrdered,                 // 66
+  OrderedDynamicChunked = BaseDynamicChunked | ModifierOrdered, // 67
+  OrderedGuidedChunked = BaseGuidedChunked | ModifierOrdered,   // 68
+  OrderedRuntime = BaseRuntime | ModifierOrdered,               // 69
+  OrderedAuto = BaseAuto | ModifierOrdered,                     // 70
+  OrderdTrapezoidal = BaseTrapezoidal | ModifierOrdered,        // 71
+
+  OrderedDistributeChunked = BaseDistributeChunked | ModifierOrdered, // 91
+  OrderedDistribute = BaseDistribute | ModifierOrdered,               // 92
+
+  NomergeUnorderedStaticChunked =
+      BaseStaticChunked | ModifierUnordered | ModifierNomerge, // 161
+  NomergeUnorderedStatic =
+      BaseStatic | ModifierUnordered | ModifierNomerge, // 162
+  NomergeUnorderedDynamicChunked =
+      BaseDynamicChunked | ModifierUnordered | ModifierNomerge, // 163
+  NomergeUnorderedGuidedChunked =
+      BaseGuidedChunked | ModifierUnordered | ModifierNomerge, // 164
+  NomergeUnorderedRuntime =
+      BaseRuntime | ModifierUnordered | ModifierNomerge,                 // 165
+  NomergeUnorderedAuto = BaseAuto | ModifierUnordered | ModifierNomerge, // 166
+  NomergeUnorderedTrapezoidal =
+      BaseTrapezoidal | ModifierUnordered | ModifierNomerge, // 167
+  NomergeUnorderedGreedy =
+      BaseGreedy | ModifierUnordered | ModifierNomerge, // 168
+  NomergeUnorderedBalanced =
+      BaseBalanced | ModifierUnordered | ModifierNomerge, // 169
+  NomergeUnorderedGuidedIterativeChunked =
+      BaseGuidedIterativeChunked | ModifierUnordered | ModifierNomerge, // 170
+  NomergeUnorderedGuidedAnalyticalChunked =
+      BaseGuidedAnalyticalChunked | ModifierUnordered | ModifierNomerge, // 171
+  NomergeUnorderedSteal =
+      BaseSteal | ModifierUnordered | ModifierNomerge, // 172
+
+  NomergeOrderedStaticChunked =
+      BaseStaticChunked | ModifierOrdered | ModifierNomerge,             // 193
+  NomergeOrderedStatic = BaseStatic | ModifierOrdered | ModifierNomerge, // 194
+  NomergeOrderedDynamicChunked =
+      BaseDynamicChunked | ModifierOrdered | ModifierNomerge, // 195
+  NomergeOrderedGuidedChunked =
+      BaseGuidedChunked | ModifierOrdered | ModifierNomerge, // 196
+  NomergeOrderedRuntime =
+      BaseRuntime | ModifierOrdered | ModifierNomerge,               // 197
+  NomergeOrderedAuto = BaseAuto | ModifierOrdered | ModifierNomerge, // 198
+  NomergeOrderedTrapezoidal =
+      BaseTrapezoidal | ModifierOrdered | ModifierNomerge, // 199
+
   LLVM_MARK_AS_BITMASK_ENUM(/* LargestValue */ ModifierMask)
 };
 
 enum OMPTgtExecModeFlags : int8_t {
   OMP_TGT_EXEC_MODE_GENERIC = 1 << 0,
   OMP_TGT_EXEC_MODE_SPMD = 1 << 1,
   OMP_TGT_EXEC_MODE_GENERIC_SPMD =
       OMP_TGT_EXEC_MODE_GENERIC | OMP_TGT_EXEC_MODE_SPMD,

llvm/include/llvm/Frontend/OpenMP/OMPIRBuilder.h

public:
   ///                  trip count. If not set, defaults to before the generated
   ///                  loop.
   ///
   /// \returns The CanonicalLoopInfo object representing the collapsed loop.
   CanonicalLoopInfo *collapseLoops(DebugLoc DL,
                                    ArrayRef<CanonicalLoopInfo *> Loops,
                                    InsertPointTy ComputeIP);
 
+private:
   /// Modifies the canonical loop to be a statically-scheduled workshare loop.
   ///
   /// This takes a \p LoopInfo representing a canonical loop, such as the one
   /// created by \p createCanonicalLoop and emits additional instructions to
   /// turn it into a workshare loop. In particular, it calls to an OpenMP
   /// runtime function in the preheader to obtain the loop bounds to be used in
   /// the current thread, updates the relevant instructions in the canonical
   /// loop and calls to an OpenMP runtime finalization function after the loop.
private:
   /// \param CLI      A descriptor of the canonical loop to workshare.
   /// \param AllocaIP An insertion point for Alloca instructions usable in the
   ///                 preheader of the loop.
   /// \param SchedType Type of scheduling to be passed to the init function.
   /// \param NeedsBarrier Indicates whether a barrier must be insterted after
   ///                     the loop.
   /// \param Chunk    The size of loop chunk considered as a unit when
   ///                 scheduling. If \p nullptr, defaults to 1.
-  /// \param Ordered  Indicates whether the ordered clause is specified without
-  ///                 parameter.
   ///
   /// \returns Point where to insert code after the workshare construct.
   InsertPointTy applyDynamicWorkshareLoop(DebugLoc DL, CanonicalLoopInfo *CLI,
                                           InsertPointTy AllocaIP,
                                           omp::OMPScheduleType SchedType,
                                           bool NeedsBarrier,
-                                          Value *Chunk = nullptr,
-                                          bool Ordered = false);
+                                          Value *Chunk = nullptr);
 
+public:
   /// Modifies the canonical loop to be a workshare loop.
   ///
   /// This takes a \p LoopInfo representing a canonical loop, such as the one
   /// created by \p createCanonicalLoop and emits additional instructions to
   /// turn it into a workshare loop. In particular, it calls to an OpenMP
   /// runtime function in the preheader to obtain the loop bounds to be used in
   /// the current thread, updates the relevant instructions in the canonical
   /// loop and calls to an OpenMP runtime finalization function after the loop.
   ///
   /// The concrete transformation is done by applyStaticWorkshareLoop,
   /// applyStaticChunkedWorkshareLoop, or applyDynamicWorkshareLoop, depending
   /// on the value of \p SchedKind and \p ChunkSize.
   ///
   /// \param DL       Debug location for instructions added for the
   ///                 workshare-loop construct itself.
   /// \param CLI      A descriptor of the canonical loop to workshare.
   /// \param AllocaIP An insertion point for Alloca instructions usable in the
   ///                 preheader of the loop.
   /// \param NeedsBarrier Indicates whether a barrier must be insterted after
   ///                     the loop.
   /// \param SchedKind Scheduling algorithm to use.
   /// \param ChunkSize The chunk size for the inner loop.
+  /// \param HasSimdModifier Whether the simd modifier is present in the
+  ///                        schedule clause.
+  /// \param HasMonotonicModifier Whether the monotonic modifier is present in
+  ///                             the schedule clause.
+  /// \param HasNonmonotonicModifier Whether the nonmonotonic modifier is
+  ///                                present in the schedule clause.
+  /// \param HasOrderedClause Whether the (parameterless) ordered clause is
+  ///                         present.
   ///
   /// \returns Point where to insert code after the workshare construct.
   InsertPointTy applyWorkshareLoop(
       DebugLoc DL, CanonicalLoopInfo *CLI, InsertPointTy AllocaIP,
       bool NeedsBarrier,
       llvm::omp::ScheduleKind SchedKind = llvm::omp::OMP_SCHEDULE_Default,
-      Value *ChunkSize = nullptr);
+      Value *ChunkSize = nullptr, bool HasSimdModifier = false,
+      bool HasMonotonicModifier = false, bool HasNonmonotonicModifier = false,
+      bool HasOrderedClause = false);
 
   /// Tile a loop nest.
   ///
   /// Tiles the loops of \p Loops by the tile sizes in \p TileSizes. Loops in
   /// \p/ Loops must be perfectly nested, from outermost to innermost loop
   /// (i.e. Loops.front() is the outermost loop). The trip count llvm::Value
   /// of every loop and every tile sizes must be usable in the outermost
   /// loop's preheader. This implies that the loop nest is rectangular.

llvm/lib/Frontend/OpenMP/OMPIRBuilder.cpp

 /// instance, if both point to the same instruction, two IRBuilders alternating
 /// creating instruction will cause the instructions to be interleaved.
 static bool isConflictIP(IRBuilder<>::InsertPoint IP1,
                          IRBuilder<>::InsertPoint IP2) {
   if (!IP1.isSet() || !IP2.isSet())
     return false;
   return IP1.getBlock() == IP2.getBlock() && IP1.getPoint() == IP2.getPoint();
 }
+
+static bool isValidWorkshareLoopScheduleType(OMPScheduleType SchedType) {
+  // Valid ordered/unordered and base algorithm combinations.
+  switch (SchedType & ~OMPScheduleType::MonotonicityMask) {
+  case OMPScheduleType::UnorderedStaticChunked:
+  case OMPScheduleType::UnorderedStatic:
+  case OMPScheduleType::UnorderedDynamicChunked:
+  case OMPScheduleType::UnorderedGuidedChunked:
+  case OMPScheduleType::UnorderedRuntime:
+  case OMPScheduleType::UnorderedAuto:
+  case OMPScheduleType::UnorderedTrapezoidal:
+  case OMPScheduleType::UnorderedGreedy:
+  case OMPScheduleType::UnorderedBalanced:
+  case OMPScheduleType::UnorderedGuidedIterativeChunked:
+  case OMPScheduleType::UnorderedGuidedAnalyticalChunked:
+  case OMPScheduleType::UnorderedSteal:
+  case OMPScheduleType::UnorderedStaticBalancedChunked:
+  case OMPScheduleType::UnorderedGuidedSimd:
+  case OMPScheduleType::UnorderedRuntimeSimd:
+  case OMPScheduleType::OrderedStaticChunked:
+  case OMPScheduleType::OrderedStatic:
+  case OMPScheduleType::OrderedDynamicChunked:
+  case OMPScheduleType::OrderedGuidedChunked:
+  case OMPScheduleType::OrderedRuntime:
+  case OMPScheduleType::OrderedAuto:
+  case OMPScheduleType::OrderdTrapezoidal:
+  case OMPScheduleType::NomergeUnorderedStaticChunked:
+  case OMPScheduleType::NomergeUnorderedStatic:
+  case OMPScheduleType::NomergeUnorderedDynamicChunked:
+  case OMPScheduleType::NomergeUnorderedGuidedChunked:
+  case OMPScheduleType::NomergeUnorderedRuntime:
+  case OMPScheduleType::NomergeUnorderedAuto:
+  case OMPScheduleType::NomergeUnorderedTrapezoidal:
+  case OMPScheduleType::NomergeUnorderedGreedy:
+  case OMPScheduleType::NomergeUnorderedBalanced:
+  case OMPScheduleType::NomergeUnorderedGuidedIterativeChunked:
+  case OMPScheduleType::NomergeUnorderedGuidedAnalyticalChunked:
+  case OMPScheduleType::NomergeUnorderedSteal:
+  case OMPScheduleType::NomergeOrderedStaticChunked:
+  case OMPScheduleType::NomergeOrderedStatic:
+  case OMPScheduleType::NomergeOrderedDynamicChunked:
+  case OMPScheduleType::NomergeOrderedGuidedChunked:
+  case OMPScheduleType::NomergeOrderedRuntime:
+  case OMPScheduleType::NomergeOrderedAuto:
+  case OMPScheduleType::NomergeOrderedTrapezoidal:
+    break;
+  default:
+    return false;
+  }
+
+  // Must not set both monotonicity modifiers at the same time.
+  OMPScheduleType MonotonicityFlags =
+      SchedType & OMPScheduleType::MonotonicityMask;
+  if (MonotonicityFlags == OMPScheduleType::MonotonicityMask)
+    return false;
+
+  return true;
+}
 #endif
 
+/// Determine which scheduling algorithm to use, determined from schedule clause
+/// arguments.
+static OMPScheduleType
+getOpenMPBaseScheduleType(llvm::omp::ScheduleKind ClauseKind, bool HasChunks,
+                          bool HasSimdModifier) {
+  // Currently, the default schedule it static.
+  switch (ClauseKind) {
+  case OMP_SCHEDULE_Default:
+  case OMP_SCHEDULE_Static:
+    return HasChunks ? OMPScheduleType::BaseStaticChunked
+                     : OMPScheduleType::BaseStatic;
+  case OMP_SCHEDULE_Dynamic:
+    return OMPScheduleType::BaseDynamicChunked;
+  case OMP_SCHEDULE_Guided:
+    return HasSimdModifier ? OMPScheduleType::BaseGuidedSimd
+                           : OMPScheduleType::BaseGuidedChunked;
+  case OMP_SCHEDULE_Auto:
+    return llvm::omp::OMPScheduleType::BaseAuto;
+  case OMP_SCHEDULE_Runtime:
+    return HasSimdModifier ? OMPScheduleType::BaseRuntimeSimd
+                           : OMPScheduleType::BaseRuntime;
+  }
+  llvm_unreachable("unhandled schedule clause argument");
+}
+
+/// Adds ordering modifier flags to schedule type.
+static OMPScheduleType
+getOpenMPOrderingScheduleType(OMPScheduleType BaseScheduleType,
+                              bool HasOrderedClause) {
+  assert((BaseScheduleType & OMPScheduleType::ModifierMask) ==
+             OMPScheduleType::None &&
+         "Must not have ordering nor monotonicity flags already set");
+
+  OMPScheduleType OrderingModifier = HasOrderedClause
+                                         ? OMPScheduleType::ModifierOrdered
+                                         : OMPScheduleType::ModifierUnordered;
+  OMPScheduleType OrderingScheduleType = BaseScheduleType | OrderingModifier;
+
+  // Unsupported combinations
+  if (OrderingScheduleType ==
+      (OMPScheduleType::BaseGuidedSimd | OMPScheduleType::ModifierOrdered))
+    return OMPScheduleType::OrderedGuidedChunked;
+  else if (OrderingScheduleType == (OMPScheduleType::BaseRuntimeSimd |
+                                    OMPScheduleType::ModifierOrdered))
+    return OMPScheduleType::OrderedRuntime;
+
+  return OrderingScheduleType;
+}
+
+/// Adds monotonicity modifier flags to schedule type.
+static OMPScheduleType
+getOpenMPMonotonicityScheduleType(OMPScheduleType ScheduleType,
+                                  bool HasSimdModifier, bool HasMonotonic,
+                                  bool HasNonmonotonic, bool HasOrderedClause) {
+  assert((ScheduleType & OMPScheduleType::MonotonicityMask) ==
+             OMPScheduleType::None &&
+         "Must not have monotonicity flags already set");
+  assert((!HasMonotonic || !HasNonmonotonic) &&
+         "Monotonic and Nonmonotonic are contradicting each other");
+
+  if (HasMonotonic) {
+    return ScheduleType | OMPScheduleType::ModifierMonotonic;
+  } else if (HasNonmonotonic) {
+    return ScheduleType | OMPScheduleType::ModifierNonmonotonic;
+  } else {
+    // OpenMP 5.1, 2.11.4 Worksharing-Loop Construct, Description.
+    // If the static schedule kind is specified or if the ordered clause is
+    // specified, and if the nonmonotonic modifier is not specified, the
+    // effect is as if the monotonic modifier is specified. Otherwise, unless
+    // the monotonic modifier is specified, the effect is as if the
+    // nonmonotonic modifier is specified.
+    OMPScheduleType BaseScheduleType =
+        ScheduleType & ~OMPScheduleType::ModifierMask;
+    if ((BaseScheduleType == OMPScheduleType::BaseStatic) ||
+        (BaseScheduleType == OMPScheduleType::BaseStaticChunked) ||
+        HasOrderedClause) {
+      // The monotonic is used by default in openmp runtime library, so no need
+      // to set it.
+      return ScheduleType;
+    } else {
+      return ScheduleType | OMPScheduleType::ModifierNonmonotonic;
+    }
+  }
+}
+
+/// Determine the schedule type using schedule and ordering clause arguments.
+static OMPScheduleType
+computeOpenMPScheduleType(ScheduleKind ClauseKind, bool HasChunks,
+                          bool HasSimdModifier, bool HasMonotonicModifier,
+                          bool HasNonmonotonicModifier, bool HasOrderedClause) {
+  OMPScheduleType BaseSchedule =
+      getOpenMPBaseScheduleType(ClauseKind, HasChunks, HasSimdModifier);
+  OMPScheduleType OrderedSchedule =
+      getOpenMPOrderingScheduleType(BaseSchedule, HasOrderedClause);
+  OMPScheduleType Result = getOpenMPMonotonicityScheduleType(
+      OrderedSchedule, HasSimdModifier, HasMonotonicModifier,
+      HasNonmonotonicModifier, HasOrderedClause);
+
+  assert(isValidWorkshareLoopScheduleType(Result));
+  return Result;
+}
+
 /// Make \p Source branch to \p Target.
 ///
 /// Handles two situations:
 /// * \p Source already has an unconditional branch.
 /// * \p Source is a degenerate block (no terminator because the BB is
 ///             the current head of the IR construction).
 static void redirectTo(BasicBlock *Source, BasicBlock *Target, DebugLoc DL) {
   if (Instruction *Term = Source->getTerminator()) {
OpenMPIRBuilder::applyStaticWorkshareLoop(DebugLoc DL, CanonicalLoopInfo *CLI,
   Constant *One = ConstantInt::get(IVTy, 1);
   Builder.CreateStore(Zero, PLowerBound);
   Value *UpperBound = Builder.CreateSub(CLI->getTripCount(), One);
   Builder.CreateStore(UpperBound, PUpperBound);
   Builder.CreateStore(One, PStride);
 
   Value *ThreadNum = getOrCreateThreadID(SrcLoc);
 
-  Constant *SchedulingType =
-      ConstantInt::get(I32Type, static_cast<int>(OMPScheduleType::Static));
+  Constant *SchedulingType = ConstantInt::get(
+      I32Type, static_cast<int>(OMPScheduleType::UnorderedStatic));
 
   // Call the "init" function and update the trip count of the loop with the
   // value it produced.
   Builder.CreateCall(StaticInit,
                      {SrcLoc, ThreadNum, SchedulingType, PLastIter, PLowerBound,
                       PUpperBound, PStride, One, Zero});
   Value *LowerBound = Builder.CreateLoad(IVTy, PLowerBound);
   Value *InclusiveUpperBound = Builder.CreateLoad(IVTy, PUpperBound);
OpenMPIRBuilder::InsertPointTy OpenMPIRBuilder::applyStaticChunkedWorkshareLoop(
 
   // TODO: Detect overflow in ubsan or max-out with current tripcount.
   Value *CastedChunkSize =
       Builder.CreateZExtOrTrunc(ChunkSize, InternalIVTy, "chunksize");
   Value *CastedTripCount =
       Builder.CreateZExt(OrigTripCount, InternalIVTy, "tripcount");
 
   Constant *SchedulingType = ConstantInt::get(
-      I32Type, static_cast<int>(OMPScheduleType::StaticChunked));
+      I32Type, static_cast<int>(OMPScheduleType::UnorderedStaticChunked));
   Builder.CreateStore(Zero, PLowerBound);
   Value *OrigUpperBound = Builder.CreateSub(CastedTripCount, One);
   Builder.CreateStore(OrigUpperBound, PUpperBound);
   Builder.CreateStore(One, PStride);
 
   // Call the "init" function and update the trip count of the loop with the
   // value it produced.
   uint32_t SrcLocStrSize;
#ifndef NDEBUG
   // Even though we currently do not support applying additional methods to it,
   // the chunk loop should remain a canonical loop.
   CLI->assertOK();
 #endif
 
   return {DispatchAfter, DispatchAfter->getFirstInsertionPt()};
 }
 
-OpenMPIRBuilder::InsertPointTy
-OpenMPIRBuilder::applyWorkshareLoop(DebugLoc DL, CanonicalLoopInfo *CLI,
-                                    InsertPointTy AllocaIP, bool NeedsBarrier,
-                                    llvm::omp::ScheduleKind SchedKind,
-                                    llvm::Value *ChunkSize) {
-  switch (SchedKind) {
-  case llvm::omp::ScheduleKind::OMP_SCHEDULE_Default:
-    assert(!ChunkSize && "No chunk size with default schedule (which for clang "
-                         "is static non-chunked)");
-    LLVM_FALLTHROUGH;
-  case llvm::omp::ScheduleKind::OMP_SCHEDULE_Static:
-    if (ChunkSize)
-      return applyStaticChunkedWorkshareLoop(DL, CLI, AllocaIP, NeedsBarrier,
-                                             ChunkSize);
-    return applyStaticWorkshareLoop(DL, CLI, AllocaIP, NeedsBarrier);
-  case llvm::omp::ScheduleKind::OMP_SCHEDULE_Auto:
-    assert(!ChunkSize && "Chunk size with auto scheduling not user-defined");
-    return applyDynamicWorkshareLoop(DL, CLI, AllocaIP, OMPScheduleType::Auto,
-                                     NeedsBarrier, nullptr);
-  case llvm::omp::ScheduleKind::OMP_SCHEDULE_Dynamic:
-    return applyDynamicWorkshareLoop(DL, CLI, AllocaIP,
-                                     OMPScheduleType::DynamicChunked,
+OpenMPIRBuilder::InsertPointTy OpenMPIRBuilder::applyWorkshareLoop(
+    DebugLoc DL, CanonicalLoopInfo *CLI, InsertPointTy AllocaIP,
+    bool NeedsBarrier, llvm::omp::ScheduleKind SchedKind,
+    llvm::Value *ChunkSize, bool HasSimdModifier, bool HasMonotonicModifier,
+    bool HasNonmonotonicModifier, bool HasOrderedClause) {
+  OMPScheduleType EffectiveScheduleType = computeOpenMPScheduleType(
+      SchedKind, ChunkSize, HasSimdModifier, HasMonotonicModifier,
+      HasNonmonotonicModifier, HasOrderedClause);
+
+  bool IsOrdered = (EffectiveScheduleType & OMPScheduleType::ModifierOrdered) ==
+                   OMPScheduleType::ModifierOrdered;
+  switch (EffectiveScheduleType & ~OMPScheduleType::ModifierMask) {
+  case OMPScheduleType::BaseStatic:
+    assert(!ChunkSize && "No chunk size with static-chunked schedule");
+    if (IsOrdered)
+      return applyDynamicWorkshareLoop(DL, CLI, AllocaIP, EffectiveScheduleType,
                                        NeedsBarrier, ChunkSize);
-  case llvm::omp::ScheduleKind::OMP_SCHEDULE_Guided:
-    return applyDynamicWorkshareLoop(DL, CLI, AllocaIP,
-                                     OMPScheduleType::GuidedChunked,
+    // FIXME: Monotonicity ignored?
+    return applyStaticWorkshareLoop(DL, CLI, AllocaIP, NeedsBarrier);
+
+  case OMPScheduleType::BaseStaticChunked:
+    if (IsOrdered)
+      return applyDynamicWorkshareLoop(DL, CLI, AllocaIP, EffectiveScheduleType,
                                        NeedsBarrier, ChunkSize);
-  case llvm::omp::ScheduleKind::OMP_SCHEDULE_Runtime:
+    // FIXME: Monotonicity ignored?
+    return applyStaticChunkedWorkshareLoop(DL, CLI, AllocaIP, NeedsBarrier,
+                                           ChunkSize);
+
+  case OMPScheduleType::BaseRuntime:
+  case OMPScheduleType::BaseAuto:
+  case OMPScheduleType::BaseGreedy:
+  case OMPScheduleType::BaseBalanced:
+  case OMPScheduleType::BaseSteal:
+  case OMPScheduleType::BaseGuidedSimd:
+  case OMPScheduleType::BaseRuntimeSimd:
     assert(!ChunkSize &&
-           "Chunk size with runtime scheduling implied to be one");
-    return applyDynamicWorkshareLoop(
-        DL, CLI, AllocaIP, OMPScheduleType::Runtime, NeedsBarrier, nullptr);
-  }
+           "schedule type does not support user-defined chunk sizes");
+    LLVM_FALLTHROUGH;
+  case OMPScheduleType::BaseDynamicChunked:
+  case OMPScheduleType::BaseGuidedChunked:
+  case OMPScheduleType::BaseGuidedIterativeChunked:
+  case OMPScheduleType::BaseGuidedAnalyticalChunked:
+  case OMPScheduleType::BaseStaticBalancedChunked:
+    return applyDynamicWorkshareLoop(DL, CLI, AllocaIP, EffectiveScheduleType,
+                                     NeedsBarrier, ChunkSize);
 
+  default:
     llvm_unreachable("Unknown/unimplemented schedule kind");
   }
+}
 
 /// Returns an LLVM function to call for initializing loop bounds using OpenMP
 /// dynamic scheduling depending on `type`. Only i32 and i64 are supported by
 /// the runtime. Always interpret integers as unsigned similarly to
 /// CanonicalLoopInfo.
 static FunctionCallee
 getKmpcForDynamicInitForType(Type *Ty, Module &M, OpenMPIRBuilder &OMPBuilder) {
   unsigned Bitwidth = Ty->getIntegerBitWidth();
getKmpcForDynamicFiniForType(Type *Ty, Module &M, OpenMPIRBuilder &OMPBuilder) {
   if (Bitwidth == 64)
     return OMPBuilder.getOrCreateRuntimeFunction(
         M, omp::RuntimeFunction::OMPRTL___kmpc_dispatch_fini_8u);
   llvm_unreachable("unknown OpenMP loop iterator bitwidth");
 }
 
 OpenMPIRBuilder::InsertPointTy OpenMPIRBuilder::applyDynamicWorkshareLoop(
     DebugLoc DL, CanonicalLoopInfo *CLI, InsertPointTy AllocaIP,
-    OMPScheduleType SchedType, bool NeedsBarrier, Value *Chunk, bool Ordered) {
+    OMPScheduleType SchedType, bool NeedsBarrier, Value *Chunk) {
   assert(CLI->isValid() && "Requires a valid canonical loop");
   assert(!isConflictIP(AllocaIP, CLI->getPreheaderIP()) &&
          "Require dedicated allocate IP");
+  assert(isValidWorkshareLoopScheduleType(SchedType) &&
+         "Require valid schedule type");
+
+  bool Ordered = (SchedType & OMPScheduleType::ModifierOrdered) ==
+                 OMPScheduleType::ModifierOrdered;
 
   // Set up the source location value for OpenMP runtime.
   Builder.SetCurrentDebugLocation(DL);
 
   uint32_t SrcLocStrSize;
   Constant *SrcLocStr = getOrCreateSrcLocStr(DL, SrcLocStrSize);
   Value *SrcLoc = getOrCreateIdent(SrcLocStr, SrcLocStrSize);
 

llvm/lib/Transforms/IPO/OpenMPOpt.cpp

void initialize(Attributor &A) override {
     case OMPRTL___kmpc_for_static_init_8u: {
       // Check the schedule and allow static schedule in SPMD mode.
       unsigned ScheduleArgOpNo = 2;
       auto *ScheduleTypeCI =
           dyn_cast<ConstantInt>(CB.getArgOperand(ScheduleArgOpNo));
       unsigned ScheduleTypeVal =
           ScheduleTypeCI ? ScheduleTypeCI->getZExtValue() : 0;
       switch (OMPScheduleType(ScheduleTypeVal)) {
-      case OMPScheduleType::Static:
-      case OMPScheduleType::StaticChunked:
-      case OMPScheduleType::Distribute:
-      case OMPScheduleType::DistributeChunked:
+      case OMPScheduleType::UnorderedStatic:
+      case OMPScheduleType::UnorderedStaticChunked:
+      case OMPScheduleType::OrderedDistribute:
+      case OMPScheduleType::OrderedDistributeChunked:
         break;
       default:
         SPMDCompatibilityTracker.indicatePessimisticFixpoint();
         SPMDCompatibilityTracker.insert(&CB);
         break;
       };
     } break;
     case OMPRTL___kmpc_target_init:

llvm/unittests/Frontend/OpenMPIRBuilderTest.cpp

static CallInst *findSingleCall(Function *F, omp::RuntimeFunction FnID,
   SmallVector<CallInst *, 1> Calls;
   findCalls(F, FnID, OMPBuilder, Calls);
   EXPECT_EQ(1u, Calls.size());
   if (Calls.size() != 1)
     return nullptr;
   return Calls.front();
 }
 
+static omp::ScheduleKind getSchedKind(omp::OMPScheduleType SchedType) {
+  switch (SchedType & ~omp::OMPScheduleType::ModifierMask) {
+  case omp::OMPScheduleType::BaseDynamicChunked:
+    return omp::OMP_SCHEDULE_Dynamic;
+  case omp::OMPScheduleType::BaseGuidedChunked:
+    return omp::OMP_SCHEDULE_Guided;
+  case omp::OMPScheduleType::BaseAuto:
+    return omp::OMP_SCHEDULE_Auto;
+  case omp::OMPScheduleType::BaseRuntime:
+    return omp::OMP_SCHEDULE_Runtime;
+  default:
+    llvm_unreachable("unknown type for this test");
+  }
+}
+
 class OpenMPIRBuilderTest : public testing::Test {
 protected:
   void SetUp() override {
     M.reset(new Module("MyModule", Ctx));
     FunctionType *FTy =
         FunctionType::get(Type::getVoidTy(Ctx), {Type::getInt32Ty(Ctx)},
                           /*isVarArg=*/false);
     F = Function::Create(FTy, Function::ExternalLinkage, "", M.get());
TEST_F(OpenMPIRBuilderTest, StaticWorkShareLoop) {
   BasicBlock *Preheader = CLI->getPreheader();
   BasicBlock *Body = CLI->getBody();
   Value *IV = CLI->getIndVar();
   BasicBlock *ExitBlock = CLI->getExit();
 
   Builder.SetInsertPoint(BB, BB->getFirstInsertionPt());
   InsertPointTy AllocaIP = Builder.saveIP();
 
-  OMPBuilder.applyStaticWorkshareLoop(DL, CLI, AllocaIP, /*NeedsBarrier=*/true);
+  OMPBuilder.applyWorkshareLoop(DL, CLI, AllocaIP, /*NeedsBarrier=*/true,
+                                OMP_SCHEDULE_Static);
 
   BasicBlock *Cond = Body->getSinglePredecessor();
   Instruction *Cmp = &*Cond->begin();
   Value *TripCount = Cmp->getOperand(1);
 
   auto AllocaIter = BB->begin();
   ASSERT_GE(std::distance(BB->begin(), BB->end()), 4);
   AllocaInst *PLastIter = dyn_cast<AllocaInst>(&*(AllocaIter++));
TEST_P(OpenMPIRBuilderTestWithIVBits, StaticChunkedWorkshareLoop) {
 
   Instruction *OrigIndVar = CLI->getIndVar();
   EXPECT_EQ(Call->getOperand(1), OrigIndVar);
 
   Type *LCTy = Type::getInt32Ty(Ctx);
   Value *ChunkSize = ConstantInt::get(LCTy, 5);
   InsertPointTy AllocaIP{&F->getEntryBlock(),
                          F->getEntryBlock().getFirstInsertionPt()};
-  OMPBuilder.applyStaticChunkedWorkshareLoop(DL, CLI, AllocaIP,
-                                             /*NeedsBarrier=*/true, ChunkSize);
+  OMPBuilder.applyWorkshareLoop(DL, CLI, AllocaIP, /*NeedsBarrier=*/true,
+                                OMP_SCHEDULE_Static, ChunkSize);
 
   OMPBuilder.finalize();
   EXPECT_FALSE(verifyModule(*M, &errs()));
 
   BasicBlock *Entry = &F->getEntryBlock();
   BasicBlock *Preheader = Entry->getSingleSuccessor();
 
   BasicBlock *DispatchPreheader = Preheader->getSingleSuccessor();
TEST_P(OpenMPIRBuilderTestWithParams, DynamicWorkShareLoop) {
   using InsertPointTy = OpenMPIRBuilder::InsertPointTy;
   OpenMPIRBuilder OMPBuilder(*M);
   OMPBuilder.initialize();
   IRBuilder<> Builder(BB);
   OpenMPIRBuilder::LocationDescription Loc({Builder.saveIP(), DL});
 
   omp::OMPScheduleType SchedType = GetParam();
   uint32_t ChunkSize = 1;
-  switch (SchedType & ~omp::OMPScheduleType::ModifierMask) {
-  case omp::OMPScheduleType::DynamicChunked:
-  case omp::OMPScheduleType::GuidedChunked:
+  switch (SchedType & ~OMPScheduleType::ModifierMask) {
+  case omp::OMPScheduleType::BaseDynamicChunked:
+  case omp::OMPScheduleType::BaseGuidedChunked:
     ChunkSize = 7;
     break;
-  case omp::OMPScheduleType::Auto:
-  case omp::OMPScheduleType::Runtime:
+  case omp::OMPScheduleType::BaseAuto:
+  case omp::OMPScheduleType::BaseRuntime:
     ChunkSize = 1;
     break;
   default:
     assert(0 && "unknown type for this test");
     break;
   }
 
   Type *LCTy = Type::getInt32Ty(Ctx);
   Value *StartVal = ConstantInt::get(LCTy, 10);
   Value *StopVal = ConstantInt::get(LCTy, 52);
   Value *StepVal = ConstantInt::get(LCTy, 2);
-  Value *ChunkVal = ConstantInt::get(LCTy, ChunkSize);
+  Value *ChunkVal =
+      (ChunkSize == 1) ? nullptr : ConstantInt::get(LCTy, ChunkSize);
   auto LoopBodyGen = [&](InsertPointTy, llvm::Value *) {};
 
   CanonicalLoopInfo *CLI = OMPBuilder.createCanonicalLoop(
       Loc, LoopBodyGen, StartVal, StopVal, StepVal,
       /*IsSigned=*/false, /*InclusiveStop=*/false);
 
   Builder.SetInsertPoint(BB, BB->getFirstInsertionPt());
   InsertPointTy AllocaIP = Builder.saveIP();
 
   // Collect all the info from CLI, as it isn't usable after the call to
   // createDynamicWorkshareLoop.
   InsertPointTy AfterIP = CLI->getAfterIP();
   BasicBlock *Preheader = CLI->getPreheader();
   BasicBlock *ExitBlock = CLI->getExit();
   BasicBlock *LatchBlock = CLI->getLatch();
   Value *IV = CLI->getIndVar();
 
-  InsertPointTy EndIP =
-      OMPBuilder.applyDynamicWorkshareLoop(DL, CLI, AllocaIP, SchedType,
-                                           /*NeedsBarrier=*/true, ChunkVal,
-                                           /*Ordered=*/false);
+  InsertPointTy EndIP = OMPBuilder.applyWorkshareLoop(
+      DL, CLI, AllocaIP, /*NeedsBarrier=*/true, getSchedKind(SchedType),
+      ChunkVal, /*Simd=*/false,
+      (SchedType & omp::OMPScheduleType::ModifierMonotonic) ==
+          omp::OMPScheduleType::ModifierMonotonic,
+      (SchedType & omp::OMPScheduleType::ModifierNonmonotonic) ==
+          omp::OMPScheduleType::ModifierNonmonotonic,
+
   // The returned value should be the "after" point.
   ASSERT_EQ(EndIP.getBlock(), AfterIP.getBlock());
   ASSERT_EQ(EndIP.getPoint(), AfterIP.getPoint());
 
   auto AllocaIter = BB->begin();
   ASSERT_GE(std::distance(BB->begin(), BB->end()), 4);
   AllocaInst *PLastIter = dyn_cast<AllocaInst>(&*(AllocaIter++));
   AllocaInst *PLowerBound = dyn_cast<AllocaInst>(&*(AllocaIter++));
TEST_P(OpenMPIRBuilderTestWithParams, DynamicWorkShareLoop) {
   CallInst *InitCall = dyn_cast<CallInst>(&*PreheaderIter);
 
   ASSERT_NE(InitCall, nullptr);
   EXPECT_EQ(InitCall->getCalledFunction()->getName(),
             "__kmpc_dispatch_init_4u");
   EXPECT_EQ(InitCall->arg_size(), 7U);
   EXPECT_EQ(InitCall->getArgOperand(6), ConstantInt::get(LCTy, ChunkSize));
   ConstantInt *SchedVal = cast<ConstantInt>(InitCall->getArgOperand(2));
-  EXPECT_EQ(SchedVal->getValue(), static_cast<uint64_t>(SchedType));
+  if ((SchedType & OMPScheduleType::MonotonicityMask) ==
+      OMPScheduleType::None) {
+    // Implementation is allowed to add default nonmonotonicity flag
+    EXPECT_EQ(
+        static_cast<OMPScheduleType>(SchedVal->getValue().getZExtValue()) |
+            OMPScheduleType::ModifierNonmonotonic,
+        SchedType | OMPScheduleType::ModifierNonmonotonic);
+  } else {
+    EXPECT_EQ(static_cast<OMPScheduleType>(SchedVal->getValue().getZExtValue()),
+              SchedType);
+  }
 
   ConstantInt *OrigLowerBound =
       dyn_cast<ConstantInt>(LowerBoundStore->getValueOperand());
   ConstantInt *OrigUpperBound =
       dyn_cast<ConstantInt>(UpperBoundStore->getValueOperand());
   ConstantInt *OrigStride =
       dyn_cast<ConstantInt>(StrideStore->getValueOperand());
   ASSERT_NE(OrigLowerBound, nullptr);
TEST_P(OpenMPIRBuilderTestWithParams, DynamicWorkShareLoop) {
   Builder.restoreIP(EndIP);
   Builder.CreateRetVoid();
   OMPBuilder.finalize();
   EXPECT_FALSE(verifyModule(*M, &errs()));
 }
 
 INSTANTIATE_TEST_SUITE_P(
     OpenMPWSLoopSchedulingTypes, OpenMPIRBuilderTestWithParams,
-    ::testing::Values(omp::OMPScheduleType::DynamicChunked,
-                      omp::OMPScheduleType::GuidedChunked,
-                      omp::OMPScheduleType::Auto, omp::OMPScheduleType::Runtime,
-                      omp::OMPScheduleType::DynamicChunked |
+    ::testing::Values(omp::OMPScheduleType::UnorderedDynamicChunked,
+                      omp::OMPScheduleType::UnorderedGuidedChunked,
+                      omp::OMPScheduleType::UnorderedAuto,
+                      omp::OMPScheduleType::UnorderedRuntime,
+                      omp::OMPScheduleType::UnorderedDynamicChunked |
                           omp::OMPScheduleType::ModifierMonotonic,
-                      omp::OMPScheduleType::DynamicChunked |
+                      omp::OMPScheduleType::UnorderedDynamicChunked |
                           omp::OMPScheduleType::ModifierNonmonotonic,
-                      omp::OMPScheduleType::GuidedChunked |
+                      omp::OMPScheduleType::UnorderedGuidedChunked |
                           omp::OMPScheduleType::ModifierMonotonic,
-                      omp::OMPScheduleType::GuidedChunked |
+                      omp::OMPScheduleType::UnorderedGuidedChunked |
                           omp::OMPScheduleType::ModifierNonmonotonic,
-                      omp::OMPScheduleType::Auto |
+                      omp::OMPScheduleType::UnorderedAuto |
                           omp::OMPScheduleType::ModifierMonotonic,
-                      omp::OMPScheduleType::Runtime |
+                      omp::OMPScheduleType::UnorderedRuntime |
                           omp::OMPScheduleType::ModifierMonotonic));
 
 TEST_F(OpenMPIRBuilderTest, DynamicWorkShareLoopOrdered) {
   using InsertPointTy = OpenMPIRBuilder::InsertPointTy;
   OpenMPIRBuilder OMPBuilder(*M);
   OMPBuilder.initialize();
   IRBuilder<> Builder(BB);
   OpenMPIRBuilder::LocationDescription Loc({Builder.saveIP(), DL});
 
-  omp::OMPScheduleType SchedType = omp::OMPScheduleType::OrderedStaticChunked;
   uint32_t ChunkSize = 1;
   Type *LCTy = Type::getInt32Ty(Ctx);
   Value *StartVal = ConstantInt::get(LCTy, 10);
   Value *StopVal = ConstantInt::get(LCTy, 52);
   Value *StepVal = ConstantInt::get(LCTy, 2);
   Value *ChunkVal = ConstantInt::get(LCTy, ChunkSize);
   auto LoopBodyGen = [&](InsertPointTy, llvm::Value *) {};
 
   CanonicalLoopInfo *CLI = OMPBuilder.createCanonicalLoop(
       Loc, LoopBodyGen, StartVal, StopVal, StepVal,
       /*IsSigned=*/false, /*InclusiveStop=*/false);
 
   Builder.SetInsertPoint(BB, BB->getFirstInsertionPt());
   InsertPointTy AllocaIP = Builder.saveIP();
 
   // Collect all the info from CLI, as it isn't usable after the call to
   // createDynamicWorkshareLoop.
   BasicBlock *Preheader = CLI->getPreheader();
   BasicBlock *ExitBlock = CLI->getExit();
   BasicBlock *LatchBlock = CLI->getLatch();
   Value *IV = CLI->getIndVar();
 
-  InsertPointTy EndIP =
-      OMPBuilder.applyDynamicWorkshareLoop(DL, CLI, AllocaIP, SchedType,
-                                           /*NeedsBarrier=*/true, ChunkVal,
-                                           /*Ordered=*/true);
+  InsertPointTy EndIP = OMPBuilder.applyWorkshareLoop(
+      DL, CLI, AllocaIP, /*NeedsBarrier=*/true, OMP_SCHEDULE_Static, ChunkVal,
+      /*HasSimdModifier=*/false, /*HasMonotonicModifier=*/false,
+      /*HasNonmonotonicModifier=*/false,
+      /*HasOrderedClause=*/true);
 
   // Add a termination to our block and check that it is internally consistent.
   Builder.restoreIP(EndIP);
   Builder.CreateRetVoid();
   OMPBuilder.finalize();
   EXPECT_FALSE(verifyModule(*M, &errs()));
 
   CallInst *InitCall = nullptr;
   for (Instruction &EI : *Preheader) {
     Instruction *Cur = &EI;
     if (isa<CallInst>(Cur)) {
       InitCall = cast<CallInst>(Cur);
       if (InitCall->getCalledFunction()->getName() == "__kmpc_dispatch_init_4u")
         break;
       InitCall = nullptr;
     }
   }
   EXPECT_NE(InitCall, nullptr);
   EXPECT_EQ(InitCall->arg_size(), 7U);
   ConstantInt *SchedVal = cast<ConstantInt>(InitCall->getArgOperand(2));
-  EXPECT_EQ(SchedVal->getValue(), static_cast<uint64_t>(SchedType));
+  EXPECT_EQ(SchedVal->getValue(),
+            static_cast<uint64_t>(OMPScheduleType::OrderedStaticChunked));
 
   CallInst *FiniCall = dyn_cast<CallInst>(
       &*(LatchBlock->getTerminator()->getPrevNonDebugInstruction(true)));
   ASSERT_NE(FiniCall, nullptr);
   EXPECT_EQ(FiniCall->getCalledFunction()->getName(),
             "__kmpc_dispatch_fini_4u");
   EXPECT_EQ(FiniCall->arg_size(), 2U);
   EXPECT_EQ(InitCall->getArgOperand(0), FiniCall->getArgOperand(0));

mlir/lib/Target/LLVMIR/Dialect/OpenMP/OpenMPToLLVMIRTranslation.cpp

 #include "llvm/ADT/TypeSwitch.h"
 #include "llvm/Frontend/OpenMP/OMPIRBuilder.h"
 #include "llvm/IR/DebugInfoMetadata.h"
 #include "llvm/IR/IRBuilder.h"
 
 using namespace mlir;
 
 namespace {
+static llvm::omp::ScheduleKind
+convertToScheduleKind(Optional<omp::ClauseScheduleKind> schedKind) {
+  if (!schedKind.hasValue())
+    return llvm::omp::OMP_SCHEDULE_Default;
+  switch (schedKind.getValue()) {
+  case omp::ClauseScheduleKind::Static:
+    return llvm::omp::OMP_SCHEDULE_Static;
+  case omp::ClauseScheduleKind::Dynamic:
+    return llvm::omp::OMP_SCHEDULE_Dynamic;
+  case omp::ClauseScheduleKind::Guided:
+    return llvm::omp::OMP_SCHEDULE_Guided;
+  case omp::ClauseScheduleKind::Auto:
+    return llvm::omp::OMP_SCHEDULE_Auto;
+  case omp::ClauseScheduleKind::Runtime:
+    return llvm::omp::OMP_SCHEDULE_Runtime;
+  }
+  llvm_unreachable("unhandled schedule clause argument");
+}
+
 /// ModuleTranslation stack frame for OpenMP operations. This keeps track of the
 /// insertion points for allocas.
 class OpenMPAllocaStackFrame
     : public LLVM::ModuleTranslation::StackFrameBase<OpenMPAllocaStackFrame> {
 public:
   MLIR_DEFINE_EXPLICIT_INTERNAL_INLINE_TYPE_ID(OpenMPAllocaStackFrame)
 
   explicit OpenMPAllocaStackFrame(llvm::OpenMPIRBuilder::InsertPointTy allocaIP)
convertOmpWsLoop(Operation &opInst, llvm::IRBuilderBase &builder,
   // Collapse loops. Store the insertion point because LoopInfos may get
   // invalidated.
   llvm::IRBuilderBase::InsertPoint afterIP = loopInfos.front()->getAfterIP();
   llvm::CanonicalLoopInfo *loopInfo =
       ompBuilder->collapseLoops(ompLoc.DL, loopInfos, {});
 
   allocaIP = findAllocaInsertPoint(builder, moduleTranslation);
 
+  // TODO: Handle doacross loops when the ordered clause has a parameter.
+  bool isOrdered = loop.ordered_val().hasValue();
+  Optional<omp::ScheduleModifier> scheduleModifier = loop.schedule_modifier();
   bool isSimd = loop.simd_modifier();
 
-  // The orderedVal refers to the value obtained from the ordered[(n)] clause.
-  //   orderedVal == -1: No ordered[(n)] clause specified.
-  //   orderedVal == 0: The ordered clause specified without a parameter.
-  //   orderedVal > 0: The ordered clause specified with a parameter (n).
-  // TODO: Handle doacross loop init when orderedVal is greater than 0.
-  int64_t orderedVal =
-      loop.ordered_val().hasValue() ? loop.ordered_val().getValue() : -1;
-  if (schedule == omp::ClauseScheduleKind::Static && orderedVal != 0) {
-    ompBuilder->applyWorkshareLoop(ompLoc.DL, loopInfo, allocaIP,
-                                   !loop.nowait(),
-                                   llvm::omp::OMP_SCHEDULE_Static, chunk);
-  } else {
-    llvm::omp::OMPScheduleType schedType;
-    switch (schedule) {
-    case omp::ClauseScheduleKind::Static:
-      if (loop.schedule_chunk_var())
-        schedType = llvm::omp::OMPScheduleType::OrderedStaticChunked;
-      else
-        schedType = llvm::omp::OMPScheduleType::OrderedStatic;
-      break;
-    case omp::ClauseScheduleKind::Dynamic:
-      if (orderedVal == 0)
-        schedType = llvm::omp::OMPScheduleType::OrderedDynamicChunked;
-      else
-        schedType = llvm::omp::OMPScheduleType::DynamicChunked;
-      break;
-    case omp::ClauseScheduleKind::Guided:
-      if (orderedVal == 0) {
-        schedType = llvm::omp::OMPScheduleType::OrderedGuidedChunked;
-      } else {
-        if (isSimd)
-          schedType = llvm::omp::OMPScheduleType::GuidedSimd;
-        else
-          schedType = llvm::omp::OMPScheduleType::GuidedChunked;
-      }
-      break;
-    case omp::ClauseScheduleKind::Auto:
-      if (orderedVal == 0)
-        schedType = llvm::omp::OMPScheduleType::OrderedAuto;
-      else
-        schedType = llvm::omp::OMPScheduleType::Auto;
-      break;
-    case omp::ClauseScheduleKind::Runtime:
-      if (orderedVal == 0) {
-        schedType = llvm::omp::OMPScheduleType::OrderedRuntime;
-      } else {
-        if (isSimd)
-          schedType = llvm::omp::OMPScheduleType::RuntimeSimd;
-        else
-          schedType = llvm::omp::OMPScheduleType::Runtime;
-      }
-      break;
-    }
-
-    if (Optional<omp::ScheduleModifier> modifier = loop.schedule_modifier()) {
-      switch (*modifier) {
-      case omp::ScheduleModifier::monotonic:
-        schedType |= llvm::omp::OMPScheduleType::ModifierMonotonic;
-        break;
-      case omp::ScheduleModifier::nonmonotonic:
-        schedType |= llvm::omp::OMPScheduleType::ModifierNonmonotonic;
-        break;
-      default:
-        // Nothing to do here.
-        break;
-      }
-    } else {
-      // OpenMP 5.1, 2.11.4 Worksharing-Loop Construct, Description.
-      // If the static schedule kind is specified or if the ordered clause is
-      // specified, and if the nonmonotonic modifier is not specified, the
-      // effect is as if the monotonic modifier is specified. Otherwise, unless
-      // the monotonic modifier is specified, the effect is as if the
-      // nonmonotonic modifier is specified.
-      // The monotonic is used by default in openmp runtime library, so no need
-      // to set it.
-      if (!(schedType == llvm::omp::OMPScheduleType::OrderedStatic ||
-            schedType == llvm::omp::OMPScheduleType::OrderedStaticChunked))
-        schedType |= llvm::omp::OMPScheduleType::ModifierNonmonotonic;
-    }
-
-    ompBuilder->applyDynamicWorkshareLoop(ompLoc.DL, loopInfo, allocaIP,
-                                          schedType, !loop.nowait(), chunk,
-                                          /*ordered*/ orderedVal == 0);
-  }
+  ompBuilder->applyWorkshareLoop(
+      ompLoc.DL, loopInfo, allocaIP, !loop.nowait(),
+      convertToScheduleKind(schedule), chunk, isSimd,
+      scheduleModifier == omp::ScheduleModifier::monotonic,
+      scheduleModifier == omp::ScheduleModifier::nonmonotonic, isOrdered);
 
   // Continue building IR after the loop. Note that the LoopInfo returned by
   // `collapseLoops` points inside the outermost loop and is intended for
   // potential further loop transformations. Use the insertion point stored
   // before collapsing loops instead.
   builder.restoreIP(afterIP);
 
   // Process the reductions if required.

mlir/test/Target/LLVMIR/openmp-llvm.mlir

 
 // -----
 
 llvm.func @body(i64)
 
 llvm.func @test_omp_wsloop_runtime_simd(%lb : i64, %ub : i64, %step : i64) -> () {
   omp.wsloop schedule(runtime, simd)
   for (%iv) : i64 = (%lb) to (%ub) step (%step) {
-    // CHECK: call void @__kmpc_dispatch_init_8u(%struct.ident_t* @{{.*}}, i32 %{{.*}}, i32 47
+    // CHECK: call void @__kmpc_dispatch_init_8u(%struct.ident_t* @{{.*}}, i32 %{{.*}}, i32 1073741871
     // CHECK: %[[continue:.*]] = call i32 @__kmpc_dispatch_next_8u
     // CHECK: %[[cond:.*]] = icmp ne i32 %[[continue]], 0
     // CHECK  br i1 %[[cond]], label %omp_loop.header{{.*}}, label %omp_loop.exit{{.*}}
     llvm.call @body(%iv) : (i64) -> ()
     omp.yield
   }
   llvm.return
 }
 
 // -----
 
 llvm.func @body(i64)
 
 llvm.func @test_omp_wsloop_guided_simd(%lb : i64, %ub : i64, %step : i64) -> () {
   omp.wsloop schedule(guided, simd)
   for (%iv) : i64 = (%lb) to (%ub) step (%step) {
-    // CHECK: call void @__kmpc_dispatch_init_8u(%struct.ident_t* @{{.*}}, i32 %{{.*}}, i32 46
+    // CHECK: call void @__kmpc_dispatch_init_8u(%struct.ident_t* @{{.*}}, i32 %{{.*}}, i32 1073741870
     // CHECK: %[[continue:.*]] = call i32 @__kmpc_dispatch_next_8u
     // CHECK: %[[cond:.*]] = icmp ne i32 %[[continue]], 0
     // CHECK  br i1 %[[cond]], label %omp_loop.header{{.*}}, label %omp_loop.exit{{.*}}
     llvm.call @body(%iv) : (i64) -> ()
     omp.yield
   }
   llvm.return
 }

 
 // -----
 
 llvm.func @body(i64)
 
 llvm.func @test_omp_wsloop_dynamic_ordered(%lb : i64, %ub : i64, %step : i64) -> () {
  omp.wsloop schedule(dynamic) ordered(0)
  for (%iv) : i64 = (%lb) to (%ub) step (%step) {
-  // CHECK: call void @__kmpc_dispatch_init_8u(%struct.ident_t* @{{.*}}, i32 %{{.*}}, i32 1073741891, i64 1, i64 %{{.*}}, i64 1, i64 1)
+  // CHECK: call void @__kmpc_dispatch_init_8u(%struct.ident_t* @{{.*}}, i32 %{{.*}}, i32 67, i64 1, i64 %{{.*}}, i64 1, i64 1)
   // CHECK: call void @__kmpc_dispatch_fini_8u
   // CHECK: %[[continue:.*]] = call i32 @__kmpc_dispatch_next_8u
   // CHECK: %[[cond:.*]] = icmp ne i32 %[[continue]], 0
   // CHECK  br i1 %[[cond]], label %omp_loop.header{{.*}}, label %omp_loop.exit{{.*}}
    llvm.call @body(%iv) : (i64) -> ()
    omp.yield
  }
  llvm.return
 }
 
 // -----
 
 llvm.func @body(i64)
 
 llvm.func @test_omp_wsloop_auto_ordered(%lb : i64, %ub : i64, %step : i64) -> () {
  omp.wsloop schedule(auto) ordered(0)
  for (%iv) : i64 = (%lb) to (%ub) step (%step) {
-  // CHECK: call void @__kmpc_dispatch_init_8u(%struct.ident_t* @{{.*}}, i32 %{{.*}}, i32 1073741894, i64 1, i64 %{{.*}}, i64 1, i64 1)
+  // CHECK: call void @__kmpc_dispatch_init_8u(%struct.ident_t* @{{.*}}, i32 %{{.*}}, i32 70, i64 1, i64 %{{.*}}, i64 1, i64 1)
   // CHECK: call void @__kmpc_dispatch_fini_8u
   // CHECK: %[[continue:.*]] = call i32 @__kmpc_dispatch_next_8u
   // CHECK: %[[cond:.*]] = icmp ne i32 %[[continue]], 0
   // CHECK  br i1 %[[cond]], label %omp_loop.header{{.*}}, label %omp_loop.exit{{.*}}
    llvm.call @body(%iv) : (i64) -> ()
    omp.yield
  }
  llvm.return
 }
 
 // -----
 
 llvm.func @body(i64)
 
 llvm.func @test_omp_wsloop_runtime_ordered(%lb : i64, %ub : i64, %step : i64) -> () {
  omp.wsloop schedule(runtime) ordered(0)
  for (%iv) : i64 = (%lb) to (%ub) step (%step) {
-  // CHECK: call void @__kmpc_dispatch_init_8u(%struct.ident_t* @{{.*}}, i32 %{{.*}}, i32 1073741893, i64 1, i64 %{{.*}}, i64 1, i64 1)
+  // CHECK: call void @__kmpc_dispatch_init_8u(%struct.ident_t* @{{.*}}, i32 %{{.*}}, i32 69, i64 1, i64 %{{.*}}, i64 1, i64 1)
   // CHECK: call void @__kmpc_dispatch_fini_8u
   // CHECK: %[[continue:.*]] = call i32 @__kmpc_dispatch_next_8u
   // CHECK: %[[cond:.*]] = icmp ne i32 %[[continue]], 0
   // CHECK  br i1 %[[cond]], label %omp_loop.header{{.*}}, label %omp_loop.exit{{.*}}
    llvm.call @body(%iv) : (i64) -> ()
    omp.yield
  }
  llvm.return
 }
 
 // -----
 
 llvm.func @body(i64)
 
 llvm.func @test_omp_wsloop_guided_ordered(%lb : i64, %ub : i64, %step : i64) -> () {
  omp.wsloop schedule(guided) ordered(0)
  for (%iv) : i64 = (%lb) to (%ub) step (%step) {
-  // CHECK: call void @__kmpc_dispatch_init_8u(%struct.ident_t* @{{.*}}, i32 %{{.*}}, i32 1073741892, i64 1, i64 %{{.*}}, i64 1, i64 1)
+  // CHECK: call void @__kmpc_dispatch_init_8u(%struct.ident_t* @{{.*}}, i32 %{{.*}}, i32 68, i64 1, i64 %{{.*}}, i64 1, i64 1)
   // CHECK: call void @__kmpc_dispatch_fini_8u
   // CHECK: %[[continue:.*]] = call i32 @__kmpc_dispatch_next_8u
   // CHECK: %[[cond:.*]] = icmp ne i32 %[[continue]], 0
   // CHECK  br i1 %[[cond]], label %omp_loop.header{{.*}}, label %omp_loop.exit{{.*}}
    llvm.call @body(%iv) : (i64) -> ()
    omp.yield
  }
  llvm.return