[flang][acc] Add support for lowering combined constructs (#86696)

PR#80319 added support to record combined construct semantics via an
attribute. Add lowering support for this.

6 files changed, 273 insertions, 216 deletions

diff --git a/flang/lib/Lower/OpenACC.cpp b/flang/lib/Lower/OpenACC.cpp
index 7b7e4a8..6e67144 100644
--- a/flang/lib/Lower/OpenACC.cpp
+++ b/flang/lib/Lower/OpenACC.cpp
@@ -1667,15 +1667,17 @@ static void privatizeIv(Fortran::lower::AbstractConverter &converter,
   ivPrivate.push_back(privateValue);
 }
 
-static mlir::acc::LoopOp
-createLoopOp(Fortran::lower::AbstractConverter &converter,
-             mlir::Location currentLocation,
-             Fortran::semantics::SemanticsContext &semanticsContext,
-             Fortran::lower::StatementContext &stmtCtx,
-             const Fortran::parser::DoConstruct &outerDoConstruct,
-             Fortran::lower::pft::Evaluation &eval,
-             const Fortran::parser::AccClauseList &accClauseList,
-             bool needEarlyReturnHandling = false) {
+static mlir::acc::LoopOp createLoopOp(
+    Fortran::lower::AbstractConverter &converter,
+    mlir::Location currentLocation,
+    Fortran::semantics::SemanticsContext &semanticsContext,
+    Fortran::lower::StatementContext &stmtCtx,
+    const Fortran::parser::DoConstruct &outerDoConstruct,
+    Fortran::lower::pft::Evaluation &eval,
+    const Fortran::parser::AccClauseList &accClauseList,
+    std::optional<mlir::acc::CombinedConstructsType> combinedConstructs =
+        std::nullopt,
+    bool needEarlyReturnHandling = false) {
   fir::FirOpBuilder &builder = converter.getFirOpBuilder();
   llvm::SmallVector<mlir::Value> tileOperands, privateOperands, ivPrivate,
       reductionOperands, cacheOperands, vectorOperands, workerNumOperands,
@@ -2015,6 +2017,10 @@ createLoopOp(Fortran::lower::AbstractConverter &converter,
   if (!collapseDeviceTypes.empty())
     loopOp.setCollapseDeviceTypeAttr(builder.getArrayAttr(collapseDeviceTypes));
 
+  if (combinedConstructs)
+    loopOp.setCombinedAttr(mlir::acc::CombinedConstructsTypeAttr::get(
+        builder.getContext(), *combinedConstructs));
+
   return loopOp;
 }
 
@@ -2060,7 +2066,7 @@ genACC(Fortran::lower::AbstractConverter &converter,
       std::get<std::optional<Fortran::parser::DoConstruct>>(loopConstruct.t);
   auto loopOp = createLoopOp(converter, currentLocation, semanticsContext,
                              stmtCtx, *outerDoConstruct, eval, accClauseList,
-                             needEarlyExitHandling);
+                             /*combinedConstructs=*/{}, needEarlyExitHandling);
   if (needEarlyExitHandling)
     return loopOp.getResult(0);
 
@@ -2092,14 +2098,14 @@ static void genDataOperandOperationsWithModifier(
 }
 
 template <typename Op>
-static Op
-createComputeOp(Fortran::lower::AbstractConverter &converter,
-                mlir::Location currentLocation,
-                Fortran::lower::pft::Evaluation &eval,
-                Fortran::semantics::SemanticsContext &semanticsContext,
-                Fortran::lower::StatementContext &stmtCtx,
-                const Fortran::parser::AccClauseList &accClauseList,
-                bool outerCombined = false) {
+static Op createComputeOp(
+    Fortran::lower::AbstractConverter &converter,
+    mlir::Location currentLocation, Fortran::lower::pft::Evaluation &eval,
+    Fortran::semantics::SemanticsContext &semanticsContext,
+    Fortran::lower::StatementContext &stmtCtx,
+    const Fortran::parser::AccClauseList &accClauseList,
+    std::optional<mlir::acc::CombinedConstructsType> combinedConstructs =
+        std::nullopt) {
 
   // Parallel operation operands
   mlir::Value ifCond;
@@ -2292,7 +2298,7 @@ createComputeOp(Fortran::lower::AbstractConverter &converter,
     } else if (const auto *privateClause =
                    std::get_if<Fortran::parser::AccClause::Private>(
                        &clause.u)) {
-      if (!outerCombined)
+      if (!combinedConstructs)
         genPrivatizations<mlir::acc::PrivateRecipeOp>(
             privateClause->v, converter, semanticsContext, stmtCtx,
             privateOperands, privatizations);
@@ -2310,7 +2316,7 @@ createComputeOp(Fortran::lower::AbstractConverter &converter,
       // combined - delay it to the loop. However, a reduction clause on a
       // combined construct implies a copy clause so issue an implicit copy
       // instead.
-      if (!outerCombined) {
+      if (!combinedConstructs) {
         genReductions(reductionClause->v, converter, semanticsContext, stmtCtx,
                       reductionOperands, reductionRecipes);
       } else {
@@ -2362,11 +2368,11 @@ createComputeOp(Fortran::lower::AbstractConverter &converter,
   if constexpr (std::is_same_v<Op, mlir::acc::KernelsOp>)
     computeOp = createRegionOp<Op, mlir::acc::TerminatorOp>(
         builder, currentLocation, currentLocation, eval, operands,
-        operandSegments, outerCombined);
+        operandSegments, /*outerCombined=*/combinedConstructs.has_value());
   else
     computeOp = createRegionOp<Op, mlir::acc::YieldOp>(
         builder, currentLocation, currentLocation, eval, operands,
-        operandSegments, outerCombined);
+        operandSegments, /*outerCombined=*/combinedConstructs.has_value());
 
   if (addSelfAttr)
     computeOp.setSelfAttrAttr(builder.getUnitAttr());
@@ -2419,6 +2425,9 @@ createComputeOp(Fortran::lower::AbstractConverter &converter,
           mlir::ArrayAttr::get(builder.getContext(), firstPrivatizations));
   }
 
+  if (combinedConstructs)
+    computeOp.setCombinedAttr(builder.getUnitAttr());
+
   auto insPt = builder.saveInsertionPoint();
   builder.setInsertionPointAfter(computeOp);
 
@@ -2734,21 +2743,24 @@ genACC(Fortran::lower::AbstractConverter &converter,
   if (combinedDirective.v == llvm::acc::ACCD_kernels_loop) {
     createComputeOp<mlir::acc::KernelsOp>(
         converter, currentLocation, eval, semanticsContext, stmtCtx,
-        accClauseList, /*outerCombined=*/true);
+        accClauseList, mlir::acc::CombinedConstructsType::KernelsLoop);
     createLoopOp(converter, currentLocation, semanticsContext, stmtCtx,
-                 *outerDoConstruct, eval, accClauseList);
+                 *outerDoConstruct, eval, accClauseList,
+                 mlir::acc::CombinedConstructsType::KernelsLoop);
   } else if (combinedDirective.v == llvm::acc::ACCD_parallel_loop) {
     createComputeOp<mlir::acc::ParallelOp>(
         converter, currentLocation, eval, semanticsContext, stmtCtx,
-        accClauseList, /*outerCombined=*/true);
+        accClauseList, mlir::acc::CombinedConstructsType::ParallelLoop);
     createLoopOp(converter, currentLocation, semanticsContext, stmtCtx,
-                 *outerDoConstruct, eval, accClauseList);
+                 *outerDoConstruct, eval, accClauseList,
+                 mlir::acc::CombinedConstructsType::ParallelLoop);
   } else if (combinedDirective.v == llvm::acc::ACCD_serial_loop) {
-    createComputeOp<mlir::acc::SerialOp>(converter, currentLocation, eval,
-                                         semanticsContext, stmtCtx,
-                                         accClauseList, /*outerCombined=*/true);
+    createComputeOp<mlir::acc::SerialOp>(
+        converter, currentLocation, eval, semanticsContext, stmtCtx,
+        accClauseList, mlir::acc::CombinedConstructsType::SerialLoop);
     createLoopOp(converter, currentLocation, semanticsContext, stmtCtx,
-                 *outerDoConstruct, eval, accClauseList);
+                 *outerDoConstruct, eval, accClauseList,
+                 mlir::acc::CombinedConstructsType::SerialLoop);
   } else {
     llvm::report_fatal_error("Unknown combined construct encountered");
   }
diff --git a/flang/test/Lower/OpenACC/acc-kernels-loop.f90 b/flang/test/Lower/OpenACC/acc-kernels-loop.f90
index e85065e..e5791f0 100644
--- a/flang/test/Lower/OpenACC/acc-kernels-loop.f90
+++ b/flang/test/Lower/OpenACC/acc-kernels-loop.f90
@@ -37,13 +37,27 @@ subroutine acc_kernels_loop
 ! CHECK: %[[IFCONDITION:.*]] = fir.address_of(@{{.*}}ifcondition) : !fir.ref<!fir.logical<4>>
 ! CHECK: %[[DECLIFCONDITION:.*]]:2 = hlfir.declare %[[IFCONDITION]]
 
-  !$acc kernels loop
+  !$acc kernels
+  !$acc loop
   DO i = 1, n
     a(i) = b(i)
   END DO
+  !$acc end kernels
 
 ! CHECK:      acc.kernels {
-! CHECK:        acc.loop {{.*}} {
+! CHECK:        acc.loop private{{.*}} {
+! CHECK:          acc.yield
+! CHECK-NEXT:   }{{$}}
+! CHECK:        acc.terminator
+! CHECK-NEXT: }{{$}}
+
+  !$acc kernels loop
+  DO i = 1, n
+    a(i) = b(i)
+  END DO
+
+! CHECK:      acc.kernels combined(loop) {
+! CHECK:        acc.loop combined(kernels) private{{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.terminator
@@ -55,7 +69,7 @@ subroutine acc_kernels_loop
   END DO
   !$acc end kernels loop
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -68,7 +82,7 @@ subroutine acc_kernels_loop
   END DO
 
 ! CHECK:      [[ASYNC1:%.*]] = arith.constant 1 : i32
-! CHECK:      acc.kernels async([[ASYNC1]] : i32) {
+! CHECK:      acc.kernels {{.*}} async([[ASYNC1]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -81,7 +95,7 @@ subroutine acc_kernels_loop
   END DO
 
 ! CHECK:      [[ASYNC2:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:      acc.kernels async([[ASYNC2]] : i32) {
+! CHECK:      acc.kernels {{.*}} async([[ASYNC2]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -93,7 +107,7 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels wait {
+! CHECK:      acc.kernels {{.*}} wait {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -106,7 +120,7 @@ subroutine acc_kernels_loop
   END DO
 
 ! CHECK:      [[WAIT1:%.*]] = arith.constant 1 : i32
-! CHECK:      acc.kernels wait({[[WAIT1]] : i32}) {
+! CHECK:      acc.kernels {{.*}} wait({[[WAIT1]] : i32}) {
 ! CHECK:        acc.loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -120,7 +134,7 @@ subroutine acc_kernels_loop
 
 ! CHECK:      [[WAIT2:%.*]] = arith.constant 1 : i32
 ! CHECK:      [[WAIT3:%.*]] = arith.constant 2 : i32
-! CHECK:      acc.kernels wait({[[WAIT2]] : i32, [[WAIT3]] : i32}) {
+! CHECK:      acc.kernels {{.*}} wait({[[WAIT2]] : i32, [[WAIT3]] : i32}) {
 ! CHECK:        acc.loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -134,7 +148,7 @@ subroutine acc_kernels_loop
 
 ! CHECK:      [[WAIT4:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
 ! CHECK:      [[WAIT5:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:      acc.kernels wait({[[WAIT4]] : i32, [[WAIT5]] : i32}) {
+! CHECK:      acc.kernels {{.*}} wait({[[WAIT4]] : i32, [[WAIT5]] : i32}) {
 ! CHECK:        acc.loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -147,7 +161,7 @@ subroutine acc_kernels_loop
   END DO
 
 ! CHECK:      [[NUMGANGS1:%.*]] = arith.constant 1 : i32
-! CHECK:      acc.kernels num_gangs({[[NUMGANGS1]] : i32}) {
+! CHECK:      acc.kernels {{.*}} num_gangs({[[NUMGANGS1]] : i32}) {
 ! CHECK:        acc.loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -160,7 +174,7 @@ subroutine acc_kernels_loop
   END DO
 
 ! CHECK:      [[NUMGANGS2:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:      acc.kernels num_gangs({[[NUMGANGS2]] : i32}) {
+! CHECK:      acc.kernels {{.*}} num_gangs({[[NUMGANGS2]] : i32}) {
 ! CHECK:        acc.loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -173,7 +187,7 @@ subroutine acc_kernels_loop
   END DO
 
 ! CHECK:      [[NUMWORKERS1:%.*]] = arith.constant 10 : i32
-! CHECK:      acc.kernels num_workers([[NUMWORKERS1]] : i32) {
+! CHECK:      acc.kernels {{.*}} num_workers([[NUMWORKERS1]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -186,7 +200,7 @@ subroutine acc_kernels_loop
   END DO
 
 ! CHECK:      [[NUMWORKERS2:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:      acc.kernels num_workers([[NUMWORKERS2]] : i32) {
+! CHECK:      acc.kernels {{.*}} num_workers([[NUMWORKERS2]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -199,7 +213,7 @@ subroutine acc_kernels_loop
   END DO
 
 ! CHECK:      [[VECTORLENGTH1:%.*]] = arith.constant 128 : i32
-! CHECK:      acc.kernels vector_length([[VECTORLENGTH1]] : i32) {
+! CHECK:      acc.kernels {{.*}} vector_length([[VECTORLENGTH1]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -212,7 +226,7 @@ subroutine acc_kernels_loop
   END DO
 
 ! CHECK:      [[VECTORLENGTH2:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:      acc.kernels vector_length([[VECTORLENGTH2]] : i32) {
+! CHECK:      acc.kernels {{.*}} vector_length([[VECTORLENGTH2]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -225,7 +239,7 @@ subroutine acc_kernels_loop
   END DO
 
 ! CHECK:      [[IF1:%.*]] = arith.constant true
-! CHECK:      acc.kernels if([[IF1]]) {
+! CHECK:      acc.kernels {{.*}} if([[IF1]]) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -239,7 +253,7 @@ subroutine acc_kernels_loop
 
 ! CHECK:      [[IFCOND:%.*]] = fir.load %{{.*}} : !fir.ref<!fir.logical<4>>
 ! CHECK:      [[IF2:%.*]] = fir.convert [[IFCOND]] : (!fir.logical<4>) -> i1
-! CHECK:      acc.kernels if([[IF2]]) {
+! CHECK:      acc.kernels {{.*}} if([[IF2]]) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -252,7 +266,7 @@ subroutine acc_kernels_loop
   END DO
 
 ! CHECK:      [[SELF1:%.*]] = arith.constant true
-! CHECK:      acc.kernels self([[SELF1]]) {
+! CHECK:      acc.kernels {{.*}} self([[SELF1]]) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -264,7 +278,7 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}}{
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -278,7 +292,7 @@ subroutine acc_kernels_loop
 
 
 ! CHECK:      %[[SELF2:.*]] = fir.convert %[[DECLIFCONDITION]]#1 : (!fir.ref<!fir.logical<4>>) -> i1
-! CHECK:      acc.kernels self(%[[SELF2]]) {
+! CHECK:      acc.kernels {{.*}} self(%[[SELF2]]) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -293,7 +307,7 @@ subroutine acc_kernels_loop
 
 ! CHECK:      %[[COPYIN_A:.*]] = acc.copyin varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copy>, name = "a"}
 ! CHECK:      %[[COPYIN_B:.*]] = acc.copyin varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copy>, name = "b"}
-! CHECK:      acc.kernels dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.kernels {{.*}} dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -309,7 +323,7 @@ subroutine acc_kernels_loop
 
 ! CHECK:      %[[COPYIN_A:.*]] = acc.copyin varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copy>, name = "a"}
 ! CHECK:      %[[COPYIN_B:.*]] = acc.copyin varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copy>, name = "b"}
-! CHECK:      acc.kernels dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.kernels {{.*}} dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -325,7 +339,7 @@ subroutine acc_kernels_loop
 
 ! CHECK:      %[[COPYIN_A:.*]] = acc.copyin varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
 ! CHECK:      %[[COPYIN_B:.*]] = acc.copyin varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copyin_readonly>, name = "b"}
-! CHECK:      acc.kernels dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.kernels {{.*}} dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -339,7 +353,7 @@ subroutine acc_kernels_loop
 
 ! CHECK:      %[[CREATE_A:.*]] = acc.create varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copyout>, name = "a"}
 ! CHECK:      %[[CREATE_B:.*]] = acc.create varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copyout>, name = "b"}
-! CHECK:      acc.kernels dataOperands(%[[CREATE_A]], %[[CREATE_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.kernels {{.*}} dataOperands(%[[CREATE_A]], %[[CREATE_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -355,7 +369,7 @@ subroutine acc_kernels_loop
 
 ! CHECK:      %[[CREATE_B:.*]] = acc.create varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
 ! CHECK:      %[[CREATE_A:.*]] = acc.create varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_create_zero>, name = "a"}
-! CHECK:      acc.kernels dataOperands(%[[CREATE_B]], %[[CREATE_A]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.kernels {{.*}} dataOperands(%[[CREATE_B]], %[[CREATE_A]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -371,7 +385,7 @@ subroutine acc_kernels_loop
 
 ! CHECK:      %[[NOCREATE_A:.*]] = acc.nocreate varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
 ! CHECK:      %[[NOCREATE_B:.*]] = acc.nocreate varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
-! CHECK:      acc.kernels dataOperands(%[[NOCREATE_A]], %[[NOCREATE_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.kernels {{.*}} dataOperands(%[[NOCREATE_A]], %[[NOCREATE_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -385,7 +399,7 @@ subroutine acc_kernels_loop
 
 ! CHECK:      %[[PRESENT_A:.*]] = acc.present varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
 ! CHECK:      %[[PRESENT_B:.*]] = acc.present varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
-! CHECK:      acc.kernels dataOperands(%[[PRESENT_A]], %[[PRESENT_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.kernels {{.*}} dataOperands(%[[PRESENT_A]], %[[PRESENT_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -399,7 +413,7 @@ subroutine acc_kernels_loop
 
 ! CHECK:      %[[DEVICEPTR_A:.*]] = acc.deviceptr varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
 ! CHECK:      %[[DEVICEPTR_B:.*]] = acc.deviceptr varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
-! CHECK:      acc.kernels dataOperands(%[[DEVICEPTR_A]], %[[DEVICEPTR_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.kernels {{.*}} dataOperands(%[[DEVICEPTR_A]], %[[DEVICEPTR_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -417,7 +431,7 @@ subroutine acc_kernels_loop
 ! CHECK:      %[[BOX_G:.*]] = fir.load %[[DECLG]]#0 : !fir.ref<!fir.box<!fir.ptr<f32>>>
 ! CHECK:      %[[BOX_ADDR_G:.*]] = fir.box_addr %[[BOX_G]] : (!fir.box<!fir.ptr<f32>>) -> !fir.ptr<f32>
 ! CHECK:      %[[ATTACH_G:.*]] = acc.attach varPtr(%[[BOX_ADDR_G]] : !fir.ptr<f32>) -> !fir.ptr<f32> {name = "g"}
-! CHECK:      acc.kernels dataOperands(%[[ATTACH_F]], %[[ATTACH_G]] : !fir.ptr<f32>, !fir.ptr<f32>) {
+! CHECK:      acc.kernels {{.*}} dataOperands(%[[ATTACH_F]], %[[ATTACH_G]] : !fir.ptr<f32>, !fir.ptr<f32>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -429,7 +443,7 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>, seq = [#acc.device_type<none>]}
@@ -441,7 +455,7 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {auto_ = [#acc.device_type<none>], inclusiveUpperbound = array<i1: true>}
@@ -453,7 +467,7 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>, independent = [#acc.device_type<none>]}
@@ -465,8 +479,8 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
-! CHECK:        acc.loop gang {{.*}} {
+! CHECK:      acc.kernels {{.*}} {
+! CHECK:        acc.loop {{.*}} gang {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>}{{$}}
 ! CHECK:        acc.terminator
@@ -477,9 +491,9 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        [[GANGNUM1:%.*]] = arith.constant 8 : i32
-! CHECK-NEXT:   acc.loop gang({num=[[GANGNUM1]] : i32}) {{.*}} {
+! CHECK-NEXT:   acc.loop {{.*}} gang({num=[[GANGNUM1]] : i32}) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.terminator
@@ -490,9 +504,9 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        [[GANGNUM2:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK-NEXT:   acc.loop gang({num=[[GANGNUM2]] : i32}) {{.*}} {
+! CHECK-NEXT:   acc.loop {{.*}} gang({num=[[GANGNUM2]] : i32}) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.terminator
@@ -503,8 +517,8 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
-! CHECK:        acc.loop gang({num=%{{.*}} : i32, static=%{{.*}} : i32})
+! CHECK:      acc.kernels {{.*}} {
+! CHECK:        acc.loop {{.*}} gang({num=%{{.*}} : i32, static=%{{.*}} : i32})
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.terminator
@@ -514,8 +528,9 @@ subroutine acc_kernels_loop
   DO i = 1, n
     a(i) = b(i)
   END DO
-! CHECK:      acc.kernels {
-! CHECK:        acc.loop vector {{.*}} {
+
+! CHECK:      acc.kernels {{.*}} {
+! CHECK:        acc.loop {{.*}} vector {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>}{{$}}
 ! CHECK:        acc.terminator
@@ -526,9 +541,9 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        [[CONSTANT128:%.*]] = arith.constant 128 : i32
-! CHECK:        acc.loop vector([[CONSTANT128]] : i32) {{.*}} {
+! CHECK:        acc.loop {{.*}} vector([[CONSTANT128]] : i32) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.terminator
@@ -539,9 +554,9 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        [[VECTORLENGTH:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:        acc.loop vector([[VECTORLENGTH]] : i32) {{.*}} {
+! CHECK:        acc.loop {{.*}} vector([[VECTORLENGTH]] : i32) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.terminator
@@ -552,8 +567,8 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
-! CHECK:        acc.loop worker {{.*}} {
+! CHECK:      acc.kernels {{.*}} {
+! CHECK:        acc.loop {{.*}} worker {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>}{{$}}
 ! CHECK:        acc.terminator
@@ -564,9 +579,9 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        [[WORKER128:%.*]] = arith.constant 128 : i32
-! CHECK:        acc.loop worker([[WORKER128]] : i32) {{.*}} {
+! CHECK:        acc.loop {{.*}} worker([[WORKER128]] : i32) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.terminator
@@ -579,7 +594,7 @@ subroutine acc_kernels_loop
     END DO
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {collapse = [2], collapseDeviceType = [#acc.device_type<none>], inclusiveUpperbound = array<i1: true, true>}
@@ -594,7 +609,7 @@ subroutine acc_kernels_loop
     END DO
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:            acc.loop {{.*}} {
 ! CHECK:              acc.yield
@@ -609,7 +624,7 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        [[TILESIZE:%.*]] = arith.constant 2 : i32
 ! CHECK:        acc.loop {{.*}} tile({[[TILESIZE]] : i32}) {{.*}} {
 ! CHECK:          acc.yield
@@ -622,7 +637,7 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        [[TILESIZEM1:%.*]] = arith.constant -1 : i32
 ! CHECK:        acc.loop {{.*}} tile({[[TILESIZEM1]] : i32}) {{.*}} {
 ! CHECK:          acc.yield
@@ -637,7 +652,7 @@ subroutine acc_kernels_loop
     END DO
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        [[TILESIZE1:%.*]] = arith.constant 2 : i32
 ! CHECK:        [[TILESIZE2:%.*]] = arith.constant 2 : i32
 ! CHECK:        acc.loop {{.*}} tile({[[TILESIZE1]] : i32, [[TILESIZE2]] : i32}) {{.*}} {
@@ -651,7 +666,7 @@ subroutine acc_kernels_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        acc.loop {{.*}} tile({%{{.*}} : i32}) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -665,7 +680,7 @@ subroutine acc_kernels_loop
     END DO
   END DO
 
-! CHECK:      acc.kernels {
+! CHECK:      acc.kernels {{.*}} {
 ! CHECK:        acc.loop {{.*}} tile({%{{.*}} : i32, %{{.*}} : i32}) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -680,7 +695,7 @@ subroutine acc_kernels_loop
 
 ! CHECK:      %[[COPYINREDR:.*]] = acc.copyin varPtr(%{{.*}} : !fir.ref<f32>) -> !fir.ref<f32> {dataClause = #acc<data_clause acc_reduction>, implicit = true, name = "reduction_r"}
 ! CHECK:      %[[COPYINREDI:.*]] = acc.copyin varPtr(%{{.*}} : !fir.ref<i32>) -> !fir.ref<i32> {dataClause = #acc<data_clause acc_reduction>, implicit = true, name = "reduction_i"}
-! CHECK:      acc.kernels dataOperands(%[[COPYINREDR]], %[[COPYINREDI]] : !fir.ref<f32>, !fir.ref<i32>) {
+! CHECK:      acc.kernels {{.*}} dataOperands(%[[COPYINREDR]], %[[COPYINREDI]] : !fir.ref<f32>, !fir.ref<i32>) {
 ! CHECK:        acc.loop {{.*}} reduction(@reduction_add_ref_f32 -> %{{.*}} : !fir.ref<f32>, @reduction_mul_ref_i32 -> %{{.*}} : !fir.ref<i32>) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
diff --git a/flang/test/Lower/OpenACC/acc-loop.f90 b/flang/test/Lower/OpenACC/acc-loop.f90
index 1c64fab..fa910e7 100644
--- a/flang/test/Lower/OpenACC/acc-loop.f90
+++ b/flang/test/Lower/OpenACC/acc-loop.f90
@@ -324,5 +324,5 @@ end subroutine
 ! CHECK: %[[P_I:.*]] = acc.private varPtr(%[[DC_I]] : !fir.ref<i32>) -> !fir.ref<i32> {implicit = true, name = ""}
 ! CHECK: %[[P_J:.*]] = acc.private varPtr(%[[DC_J]] : !fir.ref<i32>) -> !fir.ref<i32> {implicit = true, name = ""}
 ! CHECK: %[[P_K:.*]] = acc.private varPtr(%[[DC_K]] : !fir.ref<i32>) -> !fir.ref<i32> {implicit = true, name = ""}
-! CHECK: acc.loop private(@privatization_ref_i32 -> %[[P_I]] : !fir.ref<i32>, @privatization_ref_i32 -> %[[P_J]] : !fir.ref<i32>, @privatization_ref_i32 -> %[[P_K]] : !fir.ref<i32>) control(%{{.*}} : i32, %{{.*}} : i32, %{{.*}} : i32) = (%c1{{.*}}, %c1{{.*}}, %c1{{.*}} : i32, i32, i32) to (%c10{{.*}}, %c100{{.*}}, %c200{{.*}} : i32, i32, i32)  step (%c1{{.*}}, %c1{{.*}}, %c1{{.*}} : i32, i32, i32)
+! CHECK: acc.loop combined(parallel) private(@privatization_ref_i32 -> %[[P_I]] : !fir.ref<i32>, @privatization_ref_i32 -> %[[P_J]] : !fir.ref<i32>, @privatization_ref_i32 -> %[[P_K]] : !fir.ref<i32>) control(%{{.*}} : i32, %{{.*}} : i32, %{{.*}} : i32) = (%c1{{.*}}, %c1{{.*}}, %c1{{.*}} : i32, i32, i32) to (%c10{{.*}}, %c100{{.*}}, %c200{{.*}} : i32, i32, i32)  step (%c1{{.*}}, %c1{{.*}}, %c1{{.*}} : i32, i32, i32)
 ! CHECK: } attributes {inclusiveUpperbound = array<i1: true, true, true>}
diff --git a/flang/test/Lower/OpenACC/acc-parallel-loop.f90 b/flang/test/Lower/OpenACC/acc-parallel-loop.f90
index 39f1830..48ceda0 100644
--- a/flang/test/Lower/OpenACC/acc-parallel-loop.f90
+++ b/flang/test/Lower/OpenACC/acc-parallel-loop.f90
@@ -39,13 +39,27 @@ subroutine acc_parallel_loop
 ! CHECK: %[[IFCONDITION:.*]] = fir.address_of(@{{.*}}ifcondition) : !fir.ref<!fir.logical<4>>
 ! CHECK: %[[DECLIFCONDITION:.*]]:2 = hlfir.declare %[[IFCONDITION]]
 
-  !$acc parallel loop
+  !$acc parallel
+  !$acc loop
   DO i = 1, n
     a(i) = b(i)
   END DO
+  !$acc end parallel
 
 ! CHECK:      acc.parallel {
-! CHECK:        acc.loop {{.*}} {
+! CHECK:        acc.loop private{{.*}} {
+! CHECK:          acc.yield
+! CHECK-NEXT:   }{{$}}
+! CHECK:        acc.yield
+! CHECK-NEXT: }{{$}}
+
+  !$acc parallel loop
+  DO i = 1, n
+    a(i) = b(i)
+  END DO
+
+! CHECK:      acc.parallel combined(loop) {
+! CHECK:        acc.loop combined(parallel) private{{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.yield
@@ -57,7 +71,7 @@ subroutine acc_parallel_loop
   END DO
   !$acc end parallel loop
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -70,7 +84,7 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      [[ASYNC1:%.*]] = arith.constant 1 : i32
-! CHECK:      acc.parallel async([[ASYNC1]] : i32) {
+! CHECK:      acc.parallel {{.*}} async([[ASYNC1]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -83,7 +97,7 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      [[ASYNC2:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:      acc.parallel async([[ASYNC2]] : i32) {
+! CHECK:      acc.parallel {{.*}} async([[ASYNC2]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -95,7 +109,7 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel wait {
+! CHECK:      acc.parallel {{.*}} wait {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -108,7 +122,7 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      [[WAIT1:%.*]] = arith.constant 1 : i32
-! CHECK:      acc.parallel wait({[[WAIT1]] : i32}) {
+! CHECK:      acc.parallel {{.*}} wait({[[WAIT1]] : i32}) {
 ! CHECK:        acc.loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -122,7 +136,7 @@ subroutine acc_parallel_loop
 
 ! CHECK:      [[WAIT2:%.*]] = arith.constant 1 : i32
 ! CHECK:      [[WAIT3:%.*]] = arith.constant 2 : i32
-! CHECK:      acc.parallel wait({[[WAIT2]] : i32, [[WAIT3]] : i32}) {
+! CHECK:      acc.parallel {{.*}} wait({[[WAIT2]] : i32, [[WAIT3]] : i32}) {
 ! CHECK:        acc.loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -136,7 +150,7 @@ subroutine acc_parallel_loop
 
 ! CHECK:      [[WAIT4:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
 ! CHECK:      [[WAIT5:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:      acc.parallel wait({[[WAIT4]] : i32, [[WAIT5]] : i32}) {
+! CHECK:      acc.parallel {{.*}} wait({[[WAIT4]] : i32, [[WAIT5]] : i32}) {
 ! CHECK:        acc.loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -149,7 +163,7 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      [[NUMGANGS1:%.*]] = arith.constant 1 : i32
-! CHECK:      acc.parallel num_gangs({[[NUMGANGS1]] : i32}) {
+! CHECK:      acc.parallel {{.*}} num_gangs({[[NUMGANGS1]] : i32}) {
 ! CHECK:        acc.loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -162,7 +176,7 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      [[NUMGANGS2:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:      acc.parallel num_gangs({[[NUMGANGS2]] : i32}) {
+! CHECK:      acc.parallel {{.*}} num_gangs({[[NUMGANGS2]] : i32}) {
 ! CHECK:        acc.loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -175,7 +189,7 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      [[NUMWORKERS1:%.*]] = arith.constant 10 : i32
-! CHECK:      acc.parallel num_workers([[NUMWORKERS1]] : i32) {
+! CHECK:      acc.parallel {{.*}} num_workers([[NUMWORKERS1]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -188,7 +202,7 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      [[NUMWORKERS2:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:      acc.parallel num_workers([[NUMWORKERS2]] : i32) {
+! CHECK:      acc.parallel {{.*}} num_workers([[NUMWORKERS2]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -201,7 +215,7 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      [[VECTORLENGTH1:%.*]] = arith.constant 128 : i32
-! CHECK:      acc.parallel vector_length([[VECTORLENGTH1]] : i32) {
+! CHECK:      acc.parallel {{.*}} vector_length([[VECTORLENGTH1]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -214,7 +228,7 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      [[VECTORLENGTH2:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:      acc.parallel vector_length([[VECTORLENGTH2]] : i32) {
+! CHECK:      acc.parallel {{.*}} vector_length([[VECTORLENGTH2]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -227,7 +241,7 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      [[IF1:%.*]] = arith.constant true
-! CHECK:      acc.parallel if([[IF1]]) {
+! CHECK:      acc.parallel {{.*}} if([[IF1]]) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -241,7 +255,7 @@ subroutine acc_parallel_loop
 
 ! CHECK:      [[IFCOND:%.*]] = fir.load %{{.*}} : !fir.ref<!fir.logical<4>>
 ! CHECK:      [[IF2:%.*]] = fir.convert [[IFCOND]] : (!fir.logical<4>) -> i1
-! CHECK:      acc.parallel if([[IF2]]) {
+! CHECK:      acc.parallel {{.*}} if([[IF2]]) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -254,7 +268,7 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      [[SELF1:%.*]] = arith.constant true
-! CHECK:      acc.parallel self([[SELF1]]) {
+! CHECK:      acc.parallel {{.*}} self([[SELF1]]) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -266,7 +280,7 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -279,7 +293,7 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      %[[SELF2:.*]] = fir.convert %[[DECLIFCONDITION]]#1 : (!fir.ref<!fir.logical<4>>) -> i1
-! CHECK:      acc.parallel self(%[[SELF2]]) {
+! CHECK:      acc.parallel {{.*}} self(%[[SELF2]]) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -293,7 +307,7 @@ subroutine acc_parallel_loop
 
 ! CHECK:      %[[COPYIN_A:.*]] = acc.copyin varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copy>, name = "a"}
 ! CHECK:      %[[COPYIN_B:.*]] = acc.copyin varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copy>, name = "b"}
-! CHECK:      acc.parallel dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.parallel {{.*}} dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -309,7 +323,7 @@ subroutine acc_parallel_loop
 
 ! CHECK:      %[[COPYIN_A:.*]] = acc.copyin varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copy>, name = "a"}
 ! CHECK:      %[[COPYIN_B:.*]] = acc.copyin varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copy>, name = "b"}
-! CHECK:      acc.parallel dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.parallel {{.*}} dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -325,7 +339,7 @@ subroutine acc_parallel_loop
 
 ! CHECK:      %[[COPYIN_A:.*]] = acc.copyin varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
 ! CHECK:      %[[COPYIN_B:.*]] = acc.copyin varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copyin_readonly>, name = "b"}
-! CHECK:      acc.parallel dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.parallel {{.*}} dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -339,7 +353,7 @@ subroutine acc_parallel_loop
 
 ! CHECK:      %[[CREATE_A:.*]] = acc.create varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copyout>, name = "a"}
 ! CHECK:      %[[CREATE_B:.*]] = acc.create varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copyout>, name = "b"}
-! CHECK:      acc.parallel dataOperands(%[[CREATE_A]], %[[CREATE_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.parallel {{.*}} dataOperands(%[[CREATE_A]], %[[CREATE_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -355,7 +369,7 @@ subroutine acc_parallel_loop
 
 ! CHECK:      %[[CREATE_B:.*]] = acc.create varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
 ! CHECK:      %[[CREATE_A:.*]] = acc.create varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_create_zero>, name = "a"}
-! CHECK:      acc.parallel dataOperands(%[[CREATE_B]], %[[CREATE_A]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.parallel {{.*}} dataOperands(%[[CREATE_B]], %[[CREATE_A]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -371,7 +385,7 @@ subroutine acc_parallel_loop
 
 ! CHECK:      %[[NOCREATE_A:.*]] = acc.nocreate varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
 ! CHECK:      %[[NOCREATE_B:.*]] = acc.nocreate varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
-! CHECK:      acc.parallel dataOperands(%[[NOCREATE_A]], %[[NOCREATE_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.parallel {{.*}} dataOperands(%[[NOCREATE_A]], %[[NOCREATE_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -385,7 +399,7 @@ subroutine acc_parallel_loop
 
 ! CHECK:      %[[PRESENT_A:.*]] = acc.present varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
 ! CHECK:      %[[PRESENT_B:.*]] = acc.present varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
-! CHECK:      acc.parallel dataOperands(%[[PRESENT_A]], %[[PRESENT_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.parallel {{.*}} dataOperands(%[[PRESENT_A]], %[[PRESENT_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -399,7 +413,7 @@ subroutine acc_parallel_loop
 
 ! CHECK:      %[[DEVICEPTR_A:.*]] = acc.deviceptr varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
 ! CHECK:      %[[DEVICEPTR_B:.*]] = acc.deviceptr varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
-! CHECK:      acc.parallel dataOperands(%[[DEVICEPTR_A]], %[[DEVICEPTR_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.parallel {{.*}} dataOperands(%[[DEVICEPTR_A]], %[[DEVICEPTR_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -417,7 +431,7 @@ subroutine acc_parallel_loop
 ! CHECK:      %[[BOX_G:.*]] = fir.load %[[DECLG]]#0 : !fir.ref<!fir.box<!fir.ptr<f32>>>
 ! CHECK:      %[[BOX_ADDR_G:.*]] = fir.box_addr %[[BOX_G]] : (!fir.box<!fir.ptr<f32>>) -> !fir.ptr<f32>
 ! CHECK:      %[[ATTACH_G:.*]] = acc.attach varPtr(%[[BOX_ADDR_G]] : !fir.ptr<f32>) -> !fir.ptr<f32> {name = "g"}
-! CHECK:      acc.parallel dataOperands(%[[ATTACH_F]], %[[ATTACH_G]] : !fir.ptr<f32>, !fir.ptr<f32>) {
+! CHECK:      acc.parallel {{.*}} dataOperands(%[[ATTACH_F]], %[[ATTACH_G]] : !fir.ptr<f32>, !fir.ptr<f32>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -430,9 +444,9 @@ subroutine acc_parallel_loop
   END DO
 
 ! CHECK:      %[[ACC_PRIVATE_B:.*]] = acc.firstprivate varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
-! CHECK:      acc.parallel firstprivate(@firstprivatization_section_ext10_ref_10xf32 -> %[[ACC_PRIVATE_B]] : !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.parallel {{.*}} firstprivate(@firstprivatization_section_ext10_ref_10xf32 -> %[[ACC_PRIVATE_B]] : !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:      %[[ACC_PRIVATE_A:.*]] = acc.private varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
-! CHECK:        acc.loop private({{.*}}@privatization_ref_10xf32 -> %[[ACC_PRIVATE_A]] : !fir.ref<!fir.array<10xf32>>)
+! CHECK:        acc.loop {{.*}} private({{.*}}@privatization_ref_10xf32 -> %[[ACC_PRIVATE_A]] : !fir.ref<!fir.array<10xf32>>)
 ! CHECK-NOT:      fir.do_loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -444,7 +458,7 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>, seq = [#acc.device_type<none>]}
@@ -456,7 +470,7 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {auto_ = [#acc.device_type<none>], inclusiveUpperbound = array<i1: true>}
@@ -468,7 +482,7 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>, independent = [#acc.device_type<none>]}
@@ -480,8 +494,8 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
-! CHECK:        acc.loop gang
+! CHECK:      acc.parallel {{.*}} {
+! CHECK:        acc.loop {{.*}} gang
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>}{{$}}
 ! CHECK:        acc.yield
@@ -492,9 +506,9 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        [[GANGNUM1:%.*]] = arith.constant 8 : i32
-! CHECK-NEXT:   acc.loop gang({num=[[GANGNUM1]] : i32})
+! CHECK-NEXT:   acc.loop {{.*}} gang({num=[[GANGNUM1]] : i32})
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.yield
@@ -505,9 +519,9 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        [[GANGNUM2:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK-NEXT:   acc.loop gang({num=[[GANGNUM2]] : i32})
+! CHECK-NEXT:   acc.loop {{.*}} gang({num=[[GANGNUM2]] : i32})
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.yield
@@ -518,8 +532,8 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
-! CHECK:        acc.loop gang({num=%{{.*}} : i32, static=%{{.*}} : i32})
+! CHECK:      acc.parallel {{.*}} {
+! CHECK:        acc.loop {{.*}} gang({num=%{{.*}} : i32, static=%{{.*}} : i32})
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.yield
@@ -529,8 +543,9 @@ subroutine acc_parallel_loop
   DO i = 1, n
     a(i) = b(i)
   END DO
-! CHECK:      acc.parallel {
-! CHECK:        acc.loop vector
+
+! CHECK:      acc.parallel {{.*}} {
+! CHECK:        acc.loop {{.*}} vector
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>}{{$}}
 ! CHECK:        acc.yield
@@ -541,9 +556,9 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        [[CONSTANT128:%.*]] = arith.constant 128 : i32
-! CHECK:        acc.loop vector([[CONSTANT128]] : i32) {{.*}} {
+! CHECK:        acc.loop {{.*}} vector([[CONSTANT128]] : i32) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.yield
@@ -554,9 +569,9 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        [[VECTORLENGTH:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:        acc.loop vector([[VECTORLENGTH]] : i32) {{.*}} {
+! CHECK:        acc.loop {{.*}} vector([[VECTORLENGTH]] : i32) {{.*}} {
 ! CHECK-NOT:      fir.do_loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -568,8 +583,8 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
-! CHECK:        acc.loop worker {{.*}} {
+! CHECK:      acc.parallel {{.*}} {
+! CHECK:        acc.loop {{.*}} worker {{.*}} {
 ! CHECK-NOT:      fir.do_loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>}{{$}}
@@ -581,9 +596,9 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}}{
 ! CHECK:        [[WORKER128:%.*]] = arith.constant 128 : i32
-! CHECK:        acc.loop worker([[WORKER128]] : i32) {{.*}} {
+! CHECK:        acc.loop {{.*}} worker([[WORKER128]] : i32) {{.*}} {
 ! CHECK-NOT:      fir.do_loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -597,7 +612,7 @@ subroutine acc_parallel_loop
     END DO
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {collapse = [2], collapseDeviceType = [#acc.device_type<none>], inclusiveUpperbound = array<i1: true, true>}
@@ -612,7 +627,7 @@ subroutine acc_parallel_loop
     END DO
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:            acc.loop {{.*}} {
 ! CHECK:              acc.yield
@@ -627,7 +642,7 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        [[TILESIZE:%.*]] = arith.constant 2 : i32
 ! CHECK:        acc.loop {{.*}} tile({[[TILESIZE]] : i32}) {{.*}} {
 ! CHECK:          acc.yield
@@ -640,7 +655,7 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        [[TILESIZEM1:%.*]] = arith.constant -1 : i32
 ! CHECK:        acc.loop {{.*}} tile({[[TILESIZEM1]] : i32}) {{.*}} {
 ! CHECK:          acc.yield
@@ -655,7 +670,7 @@ subroutine acc_parallel_loop
     END DO
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        [[TILESIZE1:%.*]] = arith.constant 2 : i32
 ! CHECK:        [[TILESIZE2:%.*]] = arith.constant 2 : i32
 ! CHECK:        acc.loop {{.*}} tile({[[TILESIZE1]] : i32, [[TILESIZE2]] : i32}) {{.*}} {
@@ -669,7 +684,7 @@ subroutine acc_parallel_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        acc.loop {{.*}} tile({%{{.*}} : i32}) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -683,7 +698,7 @@ subroutine acc_parallel_loop
     END DO
   END DO
 
-! CHECK:      acc.parallel {
+! CHECK:      acc.parallel {{.*}} {
 ! CHECK:        acc.loop {{.*}} tile({%{{.*}} : i32, %{{.*}} : i32}) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -698,7 +713,7 @@ subroutine acc_parallel_loop
 
 ! CHECK:      %[[COPYINREDR:.*]] = acc.copyin varPtr(%{{.*}} : !fir.ref<f32>) -> !fir.ref<f32> {dataClause = #acc<data_clause acc_reduction>, implicit = true, name = "reduction_r"}
 ! CHECK:      %[[COPYINREDI:.*]] = acc.copyin varPtr(%{{.*}} : !fir.ref<i32>) -> !fir.ref<i32> {dataClause = #acc<data_clause acc_reduction>, implicit = true, name = "reduction_i"}
-! CHECK:      acc.parallel dataOperands(%[[COPYINREDR]], %[[COPYINREDI]] : !fir.ref<f32>, !fir.ref<i32>) {
+! CHECK:      acc.parallel {{.*}} dataOperands(%[[COPYINREDR]], %[[COPYINREDI]] : !fir.ref<f32>, !fir.ref<i32>) {
 ! CHECK:        acc.loop {{.*}} reduction(@reduction_add_ref_f32 -> %{{.*}} : !fir.ref<f32>, @reduction_mul_ref_i32 -> %{{.*}} : !fir.ref<i32>) {{.*}}
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
diff --git a/flang/test/Lower/OpenACC/acc-private.f90 b/flang/test/Lower/OpenACC/acc-private.f90
index ba582d4..4d9f84b 100644
--- a/flang/test/Lower/OpenACC/acc-private.f90
+++ b/flang/test/Lower/OpenACC/acc-private.f90
@@ -228,7 +228,7 @@ program acc_private
   END DO
 
 ! CHECK: %[[FP_C:.*]] = acc.firstprivate varPtr(%[[DECLC]]#0 : !fir.ref<i32>)   -> !fir.ref<i32> {name = "c"}
-! CHECK: acc.parallel firstprivate(@firstprivatization_ref_i32 -> %[[FP_C]] : !fir.ref<i32>)
+! CHECK: acc.parallel {{.*}} firstprivate(@firstprivatization_ref_i32 -> %[[FP_C]] : !fir.ref<i32>)
 ! CHECK: acc.yield
 
   !$acc parallel loop firstprivate(b)
@@ -242,7 +242,7 @@ program acc_private
 ! CHECK: %[[UB:.*]] = arith.subi %{{.*}}, %[[C1]] : index
 ! CHECK: %[[BOUND:.*]] = acc.bounds lowerbound(%[[LB]] : index) upperbound(%[[UB]] : index) extent(%{{.*}} : index) stride(%[[C1]] : index) startIdx(%[[C1]] : index)
 ! CHECK: %[[FP_B:.*]] = acc.firstprivate varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<100xf32>>) bounds(%[[BOUND]]) -> !fir.ref<!fir.array<100xf32>> {name = "b"}
-! CHECK: acc.parallel firstprivate(@firstprivatization_section_ext100_ref_100xf32 -> %[[FP_B]] : !fir.ref<!fir.array<100xf32>>)
+! CHECK: acc.parallel {{.*}} firstprivate(@firstprivatization_section_ext100_ref_100xf32 -> %[[FP_B]] : !fir.ref<!fir.array<100xf32>>)
 ! CHECK: acc.yield
 
   !$acc parallel loop firstprivate(b(51:100))
@@ -256,7 +256,7 @@ program acc_private
 ! CHECK: %[[UB:.*]] = arith.constant 99 : index
 ! CHECK: %[[BOUND:.*]] = acc.bounds lowerbound(%[[LB]] : index) upperbound(%[[UB]] : index) extent(%{{.*}} : index) stride(%[[C1]] : index) startIdx(%[[C1]] : index)
 ! CHECK: %[[FP_B:.*]] = acc.firstprivate varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<100xf32>>) bounds(%[[BOUND]]) -> !fir.ref<!fir.array<50xf32>> {name = "b(51:100)"}
-! CHECK: acc.parallel firstprivate(@firstprivatization_section_lb50.ub99_ref_50xf32 -> %[[FP_B]] : !fir.ref<!fir.array<50xf32>>)
+! CHECK: acc.parallel {{.*}} firstprivate(@firstprivatization_section_lb50.ub99_ref_50xf32 -> %[[FP_B]] : !fir.ref<!fir.array<50xf32>>)
 
 end program
 
@@ -272,10 +272,10 @@ end subroutine
 ! CHECK-LABEL: func.func @_QPacc_private_assumed_shape(
 ! CHECK-SAME:    %[[ARG0:.*]]: !fir.box<!fir.array<?xi32>> {fir.bindc_name = "a"}
 ! CHECK: %[[DECL_A:.*]]:2 = hlfir.declare %[[ARG0]] {uniq_name = "_QFacc_private_assumed_shapeEa"} : (!fir.box<!fir.array<?xi32>>) -> (!fir.box<!fir.array<?xi32>>, !fir.box<!fir.array<?xi32>>)
-! CHECK: acc.parallel {
+! CHECK: acc.parallel {{.*}} {
 ! CHECK: %[[ADDR:.*]] = fir.box_addr %[[DECL_A]]#0 : (!fir.box<!fir.array<?xi32>>) -> !fir.ref<!fir.array<?xi32>>
 ! CHECK: %[[PRIVATE:.*]] = acc.private varPtr(%[[ADDR]] : !fir.ref<!fir.array<?xi32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<?xi32>> {name = "a"}
-! CHECK: acc.loop private({{.*}}@privatization_box_Uxi32 -> %[[PRIVATE]] : !fir.ref<!fir.array<?xi32>>)
+! CHECK: acc.loop {{.*}} private({{.*}}@privatization_box_Uxi32 -> %[[PRIVATE]] : !fir.ref<!fir.array<?xi32>>)
 
 subroutine acc_private_allocatable_array(a, n)
   integer, allocatable :: a(:)
@@ -294,11 +294,11 @@ end subroutine
 ! CHECK-LABEL: func.func @_QPacc_private_allocatable_array(
 ! CHECK-SAME: %[[ARG0:.*]]: !fir.ref<!fir.box<!fir.heap<!fir.array<?xi32>>>> {fir.bindc_name = "a"}
 ! CHECK: %[[DECLA_A:.*]]:2 = hlfir.declare %[[ARG0]] {fortran_attrs = #fir.var_attrs<allocatable>, uniq_name = "_QFacc_private_allocatable_arrayEa"} : (!fir.ref<!fir.box<!fir.heap<!fir.array<?xi32>>>>) -> (!fir.ref<!fir.box<!fir.heap<!fir.array<?xi32>>>>, !fir.ref<!fir.box<!fir.heap<!fir.array<?xi32>>>>)
-! CHECK: acc.parallel {
+! CHECK: acc.parallel {{.*}} {
 ! CHECK: %[[BOX:.*]] = fir.load %[[DECLA_A]]#0 : !fir.ref<!fir.box<!fir.heap<!fir.array<?xi32>>>>
 ! CHECK: %[[BOX_ADDR:.*]] = fir.box_addr %[[BOX]] : (!fir.box<!fir.heap<!fir.array<?xi32>>>) -> !fir.heap<!fir.array<?xi32>>
 ! CHECK: %[[PRIVATE:.*]] = acc.private varPtr(%[[BOX_ADDR]] : !fir.heap<!fir.array<?xi32>>) bounds(%{{.*}}) -> !fir.heap<!fir.array<?xi32>> {name = "a"}
-! CHECK: acc.loop private({{.*}}@privatization_box_heap_Uxi32 -> %[[PRIVATE]] : !fir.heap<!fir.array<?xi32>>)
+! CHECK: acc.loop {{.*}} private({{.*}}@privatization_box_heap_Uxi32 -> %[[PRIVATE]] : !fir.heap<!fir.array<?xi32>>)
 ! CHECK: acc.serial private(@privatization_box_heap_Uxi32 -> %{{.*}} : !fir.heap<!fir.array<?xi32>>)
 
 subroutine acc_private_pointer_array(a, n)
@@ -314,11 +314,11 @@ end subroutine
 ! CHECK-LABEL: func.func @_QPacc_private_pointer_array(
 ! CHECK-SAME: %[[ARG0:.*]]: !fir.ref<!fir.box<!fir.ptr<!fir.array<?xi32>>>> {fir.bindc_name = "a"}, %arg1: !fir.ref<i32> {fir.bindc_name = "n"}) {
 ! CHECK: %[[DECL_A:.*]]:2 = hlfir.declare %arg0 {fortran_attrs = #fir.var_attrs<pointer>, uniq_name = "_QFacc_private_pointer_arrayEa"} : (!fir.ref<!fir.box<!fir.ptr<!fir.array<?xi32>>>>) -> (!fir.ref<!fir.box<!fir.ptr<!fir.array<?xi32>>>>, !fir.ref<!fir.box<!fir.ptr<!fir.array<?xi32>>>>)
-! CHECK: acc.parallel {
+! CHECK: acc.parallel {{.*}} {
 ! CHECK: %[[BOX:.*]] = fir.load %[[DECLA_A]]#0 : !fir.ref<!fir.box<!fir.ptr<!fir.array<?xi32>>>>
 ! CHECK: %[[BOX_ADDR:.*]] = fir.box_addr %[[BOX]] : (!fir.box<!fir.ptr<!fir.array<?xi32>>>) -> !fir.ptr<!fir.array<?xi32>>
 ! CHECK: %[[PRIVATE:.*]] = acc.private varPtr(%[[BOX_ADDR]] : !fir.ptr<!fir.array<?xi32>>) bounds(%{{.*}}) -> !fir.ptr<!fir.array<?xi32>> {name = "a"}
-! CHECK: acc.loop private({{.*}}@privatization_box_ptr_Uxi32 -> %[[PRIVATE]] : !fir.ptr<!fir.array<?xi32>>)
+! CHECK: acc.loop {{.*}} private({{.*}}@privatization_box_ptr_Uxi32 -> %[[PRIVATE]] : !fir.ptr<!fir.array<?xi32>>)
 
 subroutine acc_private_dynamic_extent(a, n)
   integer :: n, i
@@ -334,10 +334,10 @@ end subroutine
 ! CHECK-SAME: %[[ARG0:.*]]: !fir.ref<!fir.array<?x?x2xi32>> {fir.bindc_name = "a"}, %[[ARG1:.*]]: !fir.ref<i32> {fir.bindc_name = "n"}) {
 ! CHECK: %[[DECL_N:.*]]:2 = hlfir.declare %[[ARG1]] {uniq_name = "_QFacc_private_dynamic_extentEn"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
 ! CHECK: %[[DECL_A:.*]]:2 = hlfir.declare %[[ARG0]](%{{.*}}) {uniq_name = "_QFacc_private_dynamic_extentEa"} : (!fir.ref<!fir.array<?x?x2xi32>>, !fir.shape<3>) -> (!fir.box<!fir.array<?x?x2xi32>>, !fir.ref<!fir.array<?x?x2xi32>>)
-! CHECK: acc.parallel {
+! CHECK: acc.parallel {{.*}} {
 ! CHECK: %[[BOX_ADDR:.*]] = fir.box_addr %[[DECL_A]]#0 : (!fir.box<!fir.array<?x?x2xi32>>) -> !fir.ref<!fir.array<?x?x2xi32>>
 ! CHECK: %[[PRIV:.*]] = acc.private varPtr(%[[BOX_ADDR]] : !fir.ref<!fir.array<?x?x2xi32>>) bounds(%{{.*}}, %{{.*}}, %{{.*}}) -> !fir.ref<!fir.array<?x?x2xi32>> {name = "a"}
-! CHECK: acc.loop private({{.*}}@privatization_box_UxUx2xi32 -> %[[PRIV]] : !fir.ref<!fir.array<?x?x2xi32>>)
+! CHECK: acc.loop {{.*}} private({{.*}}@privatization_box_UxUx2xi32 -> %[[PRIV]] : !fir.ref<!fir.array<?x?x2xi32>>)
 
 subroutine acc_firstprivate_assumed_shape(a, n)
   integer :: a(:), i, n
@@ -367,7 +367,7 @@ subroutine acc_firstprivate_dynamic_extent(a, n)
   end do
 end subroutine
 
-! CHECK: acc.parallel firstprivate(@firstprivatization_box_UxUx2xi32 -> %{{.*}} : !fir.ref<!fir.array<?x?x2xi32>>)
+! CHECK: acc.parallel {{.*}} firstprivate(@firstprivatization_box_UxUx2xi32 -> %{{.*}} : !fir.ref<!fir.array<?x?x2xi32>>)
 
 module acc_declare_equivalent
   integer, parameter :: n = 10
@@ -398,6 +398,6 @@ end
 ! CHECK: acc.parallel
 ! CHECK: %[[PRIV_I:.*]] = acc.private varPtr(%[[DECL_I]]#1 : !fir.ref<i32>) -> !fir.ref<i32> {implicit = true, name = ""}
 ! CHECK: %[[DECL_PRIV_I:.*]]:2 = hlfir.declare %[[PRIV_I]] {uniq_name = "_QFacc_private_useEi"} : (!fir.ref<i32>) -> (!fir.ref<i32>, !fir.ref<i32>)
-! CHECK: acc.loop private(@privatization_ref_i32 -> %[[PRIV_I]] : !fir.ref<i32>) control(%[[IV0:.*]] : i32) = (%c1{{.*}} : i32) to (%c10{{.*}} : i32) step (%c1{{.*}} : i32)
+! CHECK: acc.loop {{.*}} private(@privatization_ref_i32 -> %[[PRIV_I]] : !fir.ref<i32>) control(%[[IV0:.*]] : i32) = (%c1{{.*}} : i32) to (%c10{{.*}} : i32) step (%c1{{.*}} : i32)
 ! CHECK:   fir.store %[[IV0]] to %[[DECL_PRIV_I]]#0 : !fir.ref<i32>
 ! CHECK:   %{{.*}} = fir.load %[[DECL_PRIV_I]]#0 : !fir.ref<i32>
diff --git a/flang/test/Lower/OpenACC/acc-serial-loop.f90 b/flang/test/Lower/OpenACC/acc-serial-loop.f90
index 77f5e1e..fa3b3f7 100644
--- a/flang/test/Lower/OpenACC/acc-serial-loop.f90
+++ b/flang/test/Lower/OpenACC/acc-serial-loop.f90
@@ -58,13 +58,27 @@ subroutine acc_serial_loop
 ! CHECK: %[[IFCONDITION:.*]] = fir.address_of(@{{.*}}ifcondition) : !fir.ref<!fir.logical<4>>
 ! CHECK: %[[DECLIFCONDITION:.*]]:2 = hlfir.declare %[[IFCONDITION]]
 
-  !$acc serial loop
+  !$acc serial
+  !$acc loop
   DO i = 1, n
     a(i) = b(i)
   END DO
+  !$acc end serial
 
 ! CHECK:      acc.serial {
-! CHECK:        acc.loop {{.*}} {
+! CHECK:        acc.loop private{{.*}} {
+! CHECK:          acc.yield
+! CHECK-NEXT:   }{{$}}
+! CHECK:        acc.yield
+! CHECK-NEXT: }{{$}}
+
+  !$acc serial loop
+  DO i = 1, n
+    a(i) = b(i)
+  END DO
+
+! CHECK:      acc.serial combined(loop) {
+! CHECK:        acc.loop combined(serial) private{{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.yield
@@ -76,7 +90,7 @@ subroutine acc_serial_loop
   END DO
   !$acc end serial loop
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -89,7 +103,7 @@ subroutine acc_serial_loop
   END DO
 
 ! CHECK:      [[ASYNC1:%.*]] = arith.constant 1 : i32
-! CHECK:      acc.serial async([[ASYNC1]] : i32) {
+! CHECK:      acc.serial {{.*}} async([[ASYNC1]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -102,7 +116,7 @@ subroutine acc_serial_loop
   END DO
 
 ! CHECK:      [[ASYNC2:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:      acc.serial async([[ASYNC2]] : i32) {
+! CHECK:      acc.serial {{.*}} async([[ASYNC2]] : i32) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -114,7 +128,7 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial wait {
+! CHECK:      acc.serial {{.*}} wait {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -127,7 +141,7 @@ subroutine acc_serial_loop
   END DO
 
 ! CHECK:      [[WAIT1:%.*]] = arith.constant 1 : i32
-! CHECK:      acc.serial wait({[[WAIT1]] : i32}) {
+! CHECK:      acc.serial {{.*}} wait({[[WAIT1]] : i32}) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -141,7 +155,7 @@ subroutine acc_serial_loop
 
 ! CHECK:      [[WAIT2:%.*]] = arith.constant 1 : i32
 ! CHECK:      [[WAIT3:%.*]] = arith.constant 2 : i32
-! CHECK:      acc.serial wait({[[WAIT2]] : i32, [[WAIT3]] : i32}) {
+! CHECK:      acc.serial {{.*}} wait({[[WAIT2]] : i32, [[WAIT3]] : i32}) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -155,7 +169,7 @@ subroutine acc_serial_loop
 
 ! CHECK:      [[WAIT4:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
 ! CHECK:      [[WAIT5:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:      acc.serial wait({[[WAIT4]] : i32, [[WAIT5]] : i32}) {
+! CHECK:      acc.serial {{.*}} wait({[[WAIT4]] : i32, [[WAIT5]] : i32}) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -168,7 +182,7 @@ subroutine acc_serial_loop
   END DO
 
 ! CHECK:      [[IF1:%.*]] = arith.constant true
-! CHECK:      acc.serial if([[IF1]]) {
+! CHECK:      acc.serial {{.*}} if([[IF1]]) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -182,7 +196,7 @@ subroutine acc_serial_loop
 
 ! CHECK:      [[IFCOND:%.*]] = fir.load %{{.*}} : !fir.ref<!fir.logical<4>>
 ! CHECK:      [[IF2:%.*]] = fir.convert [[IFCOND]] : (!fir.logical<4>) -> i1
-! CHECK:      acc.serial if([[IF2]]) {
+! CHECK:      acc.serial {{.*}} if([[IF2]]) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -195,7 +209,7 @@ subroutine acc_serial_loop
   END DO
 
 ! CHECK:      [[SELF1:%.*]] = arith.constant true
-! CHECK:      acc.serial self([[SELF1]]) {
+! CHECK:      acc.serial {{.*}} self([[SELF1]]) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -207,7 +221,7 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -220,7 +234,7 @@ subroutine acc_serial_loop
   END DO
 
 ! CHECK:      %[[SELF2:.*]] = fir.convert %[[DECLIFCONDITION]]#1 : (!fir.ref<!fir.logical<4>>) -> i1
-! CHECK:      acc.serial self(%[[SELF2]]) {
+! CHECK:      acc.serial {{.*}} self(%[[SELF2]]) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -234,7 +248,7 @@ subroutine acc_serial_loop
 
 ! CHECK:      %[[COPYIN_A:.*]] = acc.copyin varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copy>, name = "a"}
 ! CHECK:      %[[COPYIN_B:.*]] = acc.copyin varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copy>, name = "b"}
-! CHECK:      acc.serial dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.serial {{.*}} dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -250,7 +264,7 @@ subroutine acc_serial_loop
 
 ! CHECK:      %[[COPYIN_A:.*]] = acc.copyin varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copy>, name = "a"}
 ! CHECK:      %[[COPYIN_B:.*]] = acc.copyin varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copy>, name = "b"}
-! CHECK:      acc.serial dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.serial {{.*}} dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -266,7 +280,7 @@ subroutine acc_serial_loop
 
 ! CHECK:      %[[COPYIN_A:.*]] = acc.copyin varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
 ! CHECK:      %[[COPYIN_B:.*]] = acc.copyin varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copyin_readonly>, name = "b"}
-! CHECK:      acc.serial dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.serial {{.*}} dataOperands(%[[COPYIN_A]], %[[COPYIN_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -280,7 +294,7 @@ subroutine acc_serial_loop
 
 ! CHECK:      %[[CREATE_A:.*]] = acc.create varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copyout>, name = "a"}
 ! CHECK:      %[[CREATE_B:.*]] = acc.create varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_copyout>, name = "b"}
-! CHECK:      acc.serial dataOperands(%[[CREATE_A]], %[[CREATE_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.serial {{.*}} dataOperands(%[[CREATE_A]], %[[CREATE_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -296,7 +310,7 @@ subroutine acc_serial_loop
 
 ! CHECK:      %[[CREATE_B:.*]] = acc.create varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
 ! CHECK:      %[[CREATE_A:.*]] = acc.create varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {dataClause = #acc<data_clause acc_create_zero>, name = "a"}
-! CHECK:      acc.serial dataOperands(%[[CREATE_B]], %[[CREATE_A]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.serial {{.*}} dataOperands(%[[CREATE_B]], %[[CREATE_A]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -312,7 +326,7 @@ subroutine acc_serial_loop
 
 ! CHECK:      %[[NOCREATE_A:.*]] = acc.nocreate varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
 ! CHECK:      %[[NOCREATE_B:.*]] = acc.nocreate varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
-! CHECK:      acc.serial dataOperands(%[[NOCREATE_A]], %[[NOCREATE_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.serial {{.*}} dataOperands(%[[NOCREATE_A]], %[[NOCREATE_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -326,7 +340,7 @@ subroutine acc_serial_loop
 
 ! CHECK:      %[[PRESENT_A:.*]] = acc.present varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
 ! CHECK:      %[[PRESENT_B:.*]] = acc.present varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
-! CHECK:      acc.serial dataOperands(%[[PRESENT_A]], %[[PRESENT_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.serial {{.*}} dataOperands(%[[PRESENT_A]], %[[PRESENT_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -340,7 +354,7 @@ subroutine acc_serial_loop
 
 ! CHECK:      %[[DEVICEPTR_A:.*]] = acc.deviceptr varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
 ! CHECK:      %[[DEVICEPTR_B:.*]] = acc.deviceptr varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
-! CHECK:      acc.serial dataOperands(%[[DEVICEPTR_A]], %[[DEVICEPTR_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.serial {{.*}} dataOperands(%[[DEVICEPTR_A]], %[[DEVICEPTR_B]] : !fir.ref<!fir.array<10xf32>>, !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -358,7 +372,7 @@ subroutine acc_serial_loop
 ! CHECK:      %[[BOX_G:.*]] = fir.load %[[DECLG]]#0 : !fir.ref<!fir.box<!fir.ptr<f32>>>
 ! CHECK:      %[[BOX_ADDR_G:.*]] = fir.box_addr %[[BOX_G]] : (!fir.box<!fir.ptr<f32>>) -> !fir.ptr<f32>
 ! CHECK:      %[[ATTACH_G:.*]] = acc.attach varPtr(%[[BOX_ADDR_G]] : !fir.ptr<f32>) -> !fir.ptr<f32> {name = "g"}
-! CHECK:      acc.serial dataOperands(%[[ATTACH_F]], %[[ATTACH_G]] : !fir.ptr<f32>, !fir.ptr<f32>) {
+! CHECK:      acc.serial {{.*}} dataOperands(%[[ATTACH_F]], %[[ATTACH_G]] : !fir.ptr<f32>, !fir.ptr<f32>) {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -371,9 +385,9 @@ subroutine acc_serial_loop
   END DO
 
 ! CHECK:      %[[ACC_FPRIVATE_B:.*]] = acc.firstprivate varPtr(%[[DECLB]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "b"}
-! CHECK:      acc.serial firstprivate(@firstprivatization_section_ext10_ref_10xf32 -> %[[ACC_FPRIVATE_B]] : !fir.ref<!fir.array<10xf32>>) {
+! CHECK:      acc.serial {{.*}} firstprivate(@firstprivatization_section_ext10_ref_10xf32 -> %[[ACC_FPRIVATE_B]] : !fir.ref<!fir.array<10xf32>>) {
 ! CHECK:      %[[ACC_PRIVATE_A:.*]] = acc.private varPtr(%[[DECLA]]#0 : !fir.ref<!fir.array<10xf32>>) bounds(%{{.*}}) -> !fir.ref<!fir.array<10xf32>> {name = "a"}
-! CHECK:        acc.loop private({{.*}}@privatization_ref_10xf32 -> %[[ACC_PRIVATE_A]] : !fir.ref<!fir.array<10xf32>>)
+! CHECK:        acc.loop {{.*}} private({{.*}}@privatization_ref_10xf32 -> %[[ACC_PRIVATE_A]] : !fir.ref<!fir.array<10xf32>>)
 ! CHECK-NOT:          fir.do_loop
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -385,7 +399,7 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>, seq = [#acc.device_type<none>]}
@@ -397,7 +411,7 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {auto_ = [#acc.device_type<none>], inclusiveUpperbound = array<i1: true>}
@@ -409,7 +423,7 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>, independent = [#acc.device_type<none>]}
@@ -421,8 +435,8 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
-! CHECK:        acc.loop gang {{.*}} {
+! CHECK:      acc.serial {{.*}} {
+! CHECK:        acc.loop {{.*}} gang {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>}{{$}}
 ! CHECK:        acc.yield
@@ -433,9 +447,9 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        [[GANGNUM1:%.*]] = arith.constant 8 : i32
-! CHECK-NEXT:   acc.loop gang({num=[[GANGNUM1]] : i32}) {{.*}} {
+! CHECK-NEXT:   acc.loop {{.*}} gang({num=[[GANGNUM1]] : i32}) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.yield
@@ -446,9 +460,9 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        [[GANGNUM2:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK-NEXT:   acc.loop gang({num=[[GANGNUM2]] : i32}) {{.*}} {
+! CHECK-NEXT:   acc.loop {{.*}} gang({num=[[GANGNUM2]] : i32}) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.yield
@@ -459,8 +473,8 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
-! CHECK:        acc.loop gang({num=%{{.*}} : i32, static=%{{.*}} : i32}) {{.*}} {
+! CHECK:      acc.serial {{.*}} {
+! CHECK:        acc.loop {{.*}} gang({num=%{{.*}} : i32, static=%{{.*}} : i32}) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.yield
@@ -470,8 +484,9 @@ subroutine acc_serial_loop
   DO i = 1, n
     a(i) = b(i)
   END DO
-! CHECK:      acc.serial {
-! CHECK:        acc.loop vector {{.*}} {
+
+! CHECK:      acc.serial {{.*}} {
+! CHECK:        acc.loop {{.*}} vector {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>}{{$}}
 ! CHECK:        acc.yield
@@ -482,9 +497,9 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        [[CONSTANT128:%.*]] = arith.constant 128 : i32
-! CHECK:        acc.loop vector([[CONSTANT128]] : i32) {{.*}} {
+! CHECK:        acc.loop {{.*}} vector([[CONSTANT128]] : i32) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.yield
@@ -495,9 +510,9 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        [[VECTORLENGTH:%.*]] = fir.load %{{.*}} : !fir.ref<i32>
-! CHECK:        acc.loop vector([[VECTORLENGTH]] : i32) {{.*}} {
+! CHECK:        acc.loop {{.*}} vector([[VECTORLENGTH]] : i32) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.yield
@@ -508,8 +523,8 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
-! CHECK:        acc.loop worker {{.*}} {
+! CHECK:      acc.serial {{.*}} {
+! CHECK:        acc.loop {{.*}} worker {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   } attributes {inclusiveUpperbound = array<i1: true>}{{$}}
 ! CHECK:        acc.yield
@@ -520,9 +535,9 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        [[WORKER128:%.*]] = arith.constant 128 : i32
-! CHECK:        acc.loop worker([[WORKER128]] : i32) {{.*}} {
+! CHECK:        acc.loop {{.*}} worker([[WORKER128]] : i32) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
 ! CHECK:        acc.yield
@@ -535,7 +550,7 @@ subroutine acc_serial_loop
     END DO
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK-NOT:            fir.do_loop
 ! CHECK:          acc.yield
@@ -551,7 +566,7 @@ subroutine acc_serial_loop
     END DO
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        acc.loop {{.*}} {
 ! CHECK:            acc.loop {{.*}} {
 ! CHECK:              acc.yield
@@ -566,7 +581,7 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        [[TILESIZE:%.*]] = arith.constant 2 : i32
 ! CHECK:        acc.loop {{.*}} tile({[[TILESIZE]] : i32}) {{.*}} {
 ! CHECK:          acc.yield
@@ -579,7 +594,7 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        [[TILESIZEM1:%.*]] = arith.constant -1 : i32
 ! CHECK:        acc.loop {{.*}} tile({[[TILESIZEM1]] : i32}) {{.*}} {
 ! CHECK:          acc.yield
@@ -594,7 +609,7 @@ subroutine acc_serial_loop
     END DO
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        [[TILESIZE1:%.*]] = arith.constant 2 : i32
 ! CHECK:        [[TILESIZE2:%.*]] = arith.constant 2 : i32
 ! CHECK:        acc.loop {{.*}} tile({[[TILESIZE1]] : i32, [[TILESIZE2]] : i32}) {{.*}} {
@@ -608,7 +623,7 @@ subroutine acc_serial_loop
     a(i) = b(i)
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        acc.loop {{.*}} tile({%{{.*}} : i32}) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -622,7 +637,7 @@ subroutine acc_serial_loop
     END DO
   END DO
 
-! CHECK:      acc.serial {
+! CHECK:      acc.serial {{.*}} {
 ! CHECK:        acc.loop {{.*}} tile({%{{.*}} : i32, %{{.*}} : i32}) {{.*}} {
 ! CHECK:          acc.yield
 ! CHECK-NEXT:   }{{$}}
@@ -637,7 +652,7 @@ subroutine acc_serial_loop
 
 ! CHECK:      %[[COPYINREDR:.*]] = acc.copyin varPtr(%{{.*}} : !fir.ref<f32>) -> !fir.ref<f32> {dataClause = #acc<data_clause acc_reduction>, implicit = true, name = "reduction_r"}
 ! CHECK:      %[[COPYINREDI:.*]] = acc.copyin varPtr(%{{.*}} : !fir.ref<i32>) -> !fir.ref<i32> {dataClause = #acc<data_clause acc_reduction>, implicit = true, name = "reduction_i"}
-! CHECK:      acc.serial dataOperands(%[[COPYINREDR]], %[[COPYINREDI]] : !fir.ref<f32>, !fir.ref<i32>) {
+! CHECK:      acc.serial {{.*}} dataOperands(%[[COPYINREDR]], %[[COPYINREDI]] : !fir.ref<f32>, !fir.ref<i32>) {
 ! CHECK:        acc.loop {{.*}} reduction(@reduction_add_ref_f32 -> %{{.*}} : !fir.ref<f32>, @reduction_mul_ref_i32 -> %{{.*}} : !fir.ref<i32>)
 ! CHECK-NOT:      fir.do_loop
 ! CHECK:          acc.yield