[OPENMP] Initial codegen for `target teams distribute simd` directive.

Host + generic device codegen for `target teams distribute simd`
directive.

llvm-svn: 320608

5 files changed, 69 insertions, 16 deletions

diff --git a/clang/lib/Basic/OpenMPKinds.cpp b/clang/lib/Basic/OpenMPKinds.cpp
index a5ebabd6596..09c919e2b07 100644
--- a/clang/lib/Basic/OpenMPKinds.cpp
+++ b/clang/lib/Basic/OpenMPKinds.cpp
@@ -890,6 +890,7 @@ void clang::getOpenMPCaptureRegions(
     break;
   case OMPD_target_teams:
   case OMPD_target_teams_distribute:
+  case OMPD_target_teams_distribute_simd:
     CaptureRegions.push_back(OMPD_target);
     CaptureRegions.push_back(OMPD_teams);
     break;
@@ -939,7 +940,6 @@ void clang::getOpenMPCaptureRegions(
   case OMPD_distribute_simd:
   case OMPD_target_teams_distribute_parallel_for:
   case OMPD_target_teams_distribute_parallel_for_simd:
-  case OMPD_target_teams_distribute_simd:
     CaptureRegions.push_back(OMPD_unknown);
     break;
   case OMPD_threadprivate:
diff --git a/clang/lib/CodeGen/CGOpenMPRuntime.cpp b/clang/lib/CodeGen/CGOpenMPRuntime.cpp
index 473cdfde698..2995e27e398 100644
--- a/clang/lib/CodeGen/CGOpenMPRuntime.cpp
+++ b/clang/lib/CodeGen/CGOpenMPRuntime.cpp
@@ -7144,6 +7144,10 @@ void CGOpenMPRuntime::scanForTargetRegionsFunctions(const Stmt *S,
       CodeGenFunction::EmitOMPTargetTeamsDistributeDeviceFunction(
           CGM, ParentName, cast<OMPTargetTeamsDistributeDirective>(*S));
       break;
+    case Stmt::OMPTargetTeamsDistributeSimdDirectiveClass:
+      CodeGenFunction::EmitOMPTargetTeamsDistributeSimdDeviceFunction(
+          CGM, ParentName, cast<OMPTargetTeamsDistributeSimdDirective>(*S));
+      break;
     case Stmt::OMPTargetParallelForDirectiveClass:
       CodeGenFunction::EmitOMPTargetParallelForDeviceFunction(
           CGM, ParentName, cast<OMPTargetParallelForDirective>(*S));
diff --git a/clang/lib/CodeGen/CGStmtOpenMP.cpp b/clang/lib/CodeGen/CGStmtOpenMP.cpp
index ad177b752db..a24aaa046ab 100644
--- a/clang/lib/CodeGen/CGStmtOpenMP.cpp
+++ b/clang/lib/CodeGen/CGStmtOpenMP.cpp
@@ -2118,17 +2118,6 @@ void CodeGenFunction::EmitOMPTargetTeamsDistributeParallelForSimdDirective(
       });
 }
 
-void CodeGenFunction::EmitOMPTargetTeamsDistributeSimdDirective(
-    const OMPTargetTeamsDistributeSimdDirective &S) {
-  OMPLexicalScope Scope(*this, S, /*AsInlined=*/true);
-  CGM.getOpenMPRuntime().emitInlinedDirective(
-      *this, OMPD_target_teams_distribute_simd,
-      [&S](CodeGenFunction &CGF, PrePostActionTy &) {
-        CGF.EmitStmt(
-            cast<CapturedStmt>(S.getAssociatedStmt())->getCapturedStmt());
-      });
-}
-
 namespace {
   struct ScheduleKindModifiersTy {
     OpenMPScheduleClauseKind Kind;
@@ -3928,6 +3917,51 @@ void CodeGenFunction::EmitOMPTargetTeamsDistributeDirective(
   emitCommonOMPTargetDirective(*this, S, CodeGen);
 }
 
+static void emitTargetTeamsDistributeSimdRegion(
+    CodeGenFunction &CGF, PrePostActionTy &Action,
+    const OMPTargetTeamsDistributeSimdDirective &S) {
+  Action.Enter(CGF);
+  auto &&CodeGenDistribute = [&S](CodeGenFunction &CGF, PrePostActionTy &) {
+    CGF.EmitOMPDistributeLoop(S, emitOMPLoopBodyWithStopPoint, S.getInc());
+  };
+
+  // Emit teams region as a standalone region.
+  auto &&CodeGen = [&S, &CodeGenDistribute](CodeGenFunction &CGF,
+                                            PrePostActionTy &) {
+    CodeGenFunction::OMPPrivateScope PrivateScope(CGF);
+    CGF.EmitOMPReductionClauseInit(S, PrivateScope);
+    (void)PrivateScope.Privatize();
+    CGF.CGM.getOpenMPRuntime().emitInlinedDirective(CGF, OMPD_distribute,
+                                                    CodeGenDistribute);
+    CGF.EmitOMPReductionClauseFinal(S, /*ReductionKind=*/OMPD_teams);
+  };
+  emitCommonOMPTeamsDirective(CGF, S, OMPD_distribute_simd, CodeGen);
+  emitPostUpdateForReductionClause(CGF, S,
+                                   [](CodeGenFunction &) { return nullptr; });
+}
+
+void CodeGenFunction::EmitOMPTargetTeamsDistributeSimdDeviceFunction(
+    CodeGenModule &CGM, StringRef ParentName,
+    const OMPTargetTeamsDistributeSimdDirective &S) {
+  auto &&CodeGen = [&S](CodeGenFunction &CGF, PrePostActionTy &Action) {
+    emitTargetTeamsDistributeSimdRegion(CGF, Action, S);
+  };
+  llvm::Function *Fn;
+  llvm::Constant *Addr;
+  // Emit target region as a standalone region.
+  CGM.getOpenMPRuntime().emitTargetOutlinedFunction(
+      S, ParentName, Fn, Addr, /*IsOffloadEntry=*/true, CodeGen);
+  assert(Fn && Addr && "Target device function emission failed.");
+}
+
+void CodeGenFunction::EmitOMPTargetTeamsDistributeSimdDirective(
+    const OMPTargetTeamsDistributeSimdDirective &S) {
+  auto &&CodeGen = [&S](CodeGenFunction &CGF, PrePostActionTy &Action) {
+    emitTargetTeamsDistributeSimdRegion(CGF, Action, S);
+  };
+  emitCommonOMPTargetDirective(*this, S, CodeGen);
+}
+
 void CodeGenFunction::EmitOMPTeamsDistributeDirective(
     const OMPTeamsDistributeDirective &S) {
 
diff --git a/clang/lib/CodeGen/CodeGenFunction.h b/clang/lib/CodeGen/CodeGenFunction.h
index b3f8888befd..e30695a99a5 100644
--- a/clang/lib/CodeGen/CodeGenFunction.h
+++ b/clang/lib/CodeGen/CodeGenFunction.h
@@ -2905,6 +2905,10 @@ public:
   static void EmitOMPTargetTeamsDistributeDeviceFunction(
       CodeGenModule &CGM, StringRef ParentName,
       const OMPTargetTeamsDistributeDirective &S);
+  /// Emit device code for the target teams distribute simd directive.
+  static void EmitOMPTargetTeamsDistributeSimdDeviceFunction(
+      CodeGenModule &CGM, StringRef ParentName,
+      const OMPTargetTeamsDistributeSimdDirective &S);
   /// Emit device code for the target simd directive.
   static void EmitOMPTargetSimdDeviceFunction(CodeGenModule &CGM,
                                               StringRef ParentName,
diff --git a/clang/lib/Sema/SemaOpenMP.cpp b/clang/lib/Sema/SemaOpenMP.cpp
index 4aa47e08792..b9323388579 100644
--- a/clang/lib/Sema/SemaOpenMP.cpp
+++ b/clang/lib/Sema/SemaOpenMP.cpp
@@ -2113,7 +2113,8 @@ void Sema::ActOnOpenMPRegionStart(OpenMPDirectiveKind DKind, Scope *CurScope) {
   case OMPD_target_parallel:
   case OMPD_target_parallel_for:
   case OMPD_target_parallel_for_simd:
-  case OMPD_target_teams_distribute: {
+  case OMPD_target_teams_distribute:
+  case OMPD_target_teams_distribute_simd: {
     Sema::CapturedParamNameType ParamsTarget[] = {
         std::make_pair(StringRef(), QualType()) // __context with shared vars
     };
@@ -2221,8 +2222,7 @@ void Sema::ActOnOpenMPRegionStart(OpenMPDirectiveKind DKind, Scope *CurScope) {
   case OMPD_distribute_parallel_for_simd:
   case OMPD_distribute_parallel_for:
   case OMPD_target_teams_distribute_parallel_for:
-  case OMPD_target_teams_distribute_parallel_for_simd:
-  case OMPD_target_teams_distribute_simd: {
+  case OMPD_target_teams_distribute_parallel_for_simd: {
     QualType KmpInt32Ty = Context.getIntTypeForBitwidth(32, 1);
     QualType KmpInt32PtrTy =
         Context.getPointerType(KmpInt32Ty).withConst().withRestrict();
@@ -7456,13 +7456,24 @@ StmtResult Sema::ActOnOpenMPTargetTeamsDistributeSimdDirective(
   // The point of exit cannot be a branch out of the structured block.
   // longjmp() and throw() must not violate the entry/exit criteria.
   CS->getCapturedDecl()->setNothrow();
+  for (int ThisCaptureLevel =
+           getOpenMPCaptureLevels(OMPD_target_teams_distribute_simd);
+       ThisCaptureLevel > 1; --ThisCaptureLevel) {
+    CS = cast<CapturedStmt>(CS->getCapturedStmt());
+    // 1.2.2 OpenMP Language Terminology
+    // Structured block - An executable statement with a single entry at the
+    // top and a single exit at the bottom.
+    // The point of exit cannot be a branch out of the structured block.
+    // longjmp() and throw() must not violate the entry/exit criteria.
+    CS->getCapturedDecl()->setNothrow();
+  }
 
   OMPLoopDirective::HelperExprs B;
   // In presence of clause 'collapse' with number of loops, it will
   // define the nested loops number.
   auto NestedLoopCount = CheckOpenMPLoop(
       OMPD_target_teams_distribute_simd, getCollapseNumberExpr(Clauses),
-      nullptr /*ordered not a clause on distribute*/, AStmt, *this, *DSAStack,
+      nullptr /*ordered not a clause on distribute*/, CS, *this, *DSAStack,
       VarsWithImplicitDSA, B);
   if (NestedLoopCount == 0)
     return StmtError();