r261209 - [OPENMP] Fix codegen for lastprivate loop counters.
Alexey Bataev via cfe-commits
cfe-commits at lists.llvm.org
Thu Feb 18 05:48:16 PST 2016
Author: abataev
Date: Thu Feb 18 07:48:15 2016
New Revision: 261209
URL: http://llvm.org/viewvc/llvm-project?rev=261209&view=rev
Log:
[OPENMP] Fix codegen for lastprivate loop counters.
Patch fixes bug with codegen for lastprivate loop counters. Also it may
improve performance for lastprivates calculations in some cases.
Modified:
cfe/trunk/lib/CodeGen/CGStmtOpenMP.cpp
cfe/trunk/test/OpenMP/for_lastprivate_codegen.cpp
Modified: cfe/trunk/lib/CodeGen/CGStmtOpenMP.cpp
URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CGStmtOpenMP.cpp?rev=261209&r1=261208&r2=261209&view=diff
==============================================================================
--- cfe/trunk/lib/CodeGen/CGStmtOpenMP.cpp (original)
+++ cfe/trunk/lib/CodeGen/CGStmtOpenMP.cpp Thu Feb 18 07:48:15 2016
@@ -653,71 +653,48 @@ void CodeGenFunction::EmitOMPLastprivate
EmitBlock(ThenBB);
}
llvm::DenseMap<const Decl *, const Expr *> LoopCountersAndUpdates;
- const Expr *LastIterVal = nullptr;
- const Expr *IVExpr = nullptr;
- const Expr *IncExpr = nullptr;
if (auto *LoopDirective = dyn_cast<OMPLoopDirective>(&D)) {
- if (isOpenMPWorksharingDirective(D.getDirectiveKind())) {
- LastIterVal = cast<VarDecl>(cast<DeclRefExpr>(
- LoopDirective->getUpperBoundVariable())
- ->getDecl())
- ->getAnyInitializer();
- IVExpr = LoopDirective->getIterationVariable();
- IncExpr = LoopDirective->getInc();
- auto IUpdate = LoopDirective->updates().begin();
- for (auto *E : LoopDirective->counters()) {
- auto *D = cast<DeclRefExpr>(E)->getDecl()->getCanonicalDecl();
- LoopCountersAndUpdates[D] = *IUpdate;
- ++IUpdate;
- }
+ auto IC = LoopDirective->counters().begin();
+ for (auto F : LoopDirective->finals()) {
+ auto *D = cast<DeclRefExpr>(*IC)->getDecl()->getCanonicalDecl();
+ LoopCountersAndUpdates[D] = F;
+ ++IC;
}
}
- {
- llvm::DenseSet<const VarDecl *> AlreadyEmittedVars;
- bool FirstLCV = true;
- for (const auto *C : D.getClausesOfKind<OMPLastprivateClause>()) {
- auto IRef = C->varlist_begin();
- auto ISrcRef = C->source_exprs().begin();
- auto IDestRef = C->destination_exprs().begin();
- for (auto *AssignOp : C->assignment_ops()) {
- auto *PrivateVD = cast<VarDecl>(cast<DeclRefExpr>(*IRef)->getDecl());
- QualType Type = PrivateVD->getType();
- auto *CanonicalVD = PrivateVD->getCanonicalDecl();
- if (AlreadyEmittedVars.insert(CanonicalVD).second) {
- // If lastprivate variable is a loop control variable for loop-based
- // directive, update its value before copyin back to original
- // variable.
- if (auto *UpExpr = LoopCountersAndUpdates.lookup(CanonicalVD)) {
- if (FirstLCV && LastIterVal) {
- EmitAnyExprToMem(LastIterVal, EmitLValue(IVExpr).getAddress(),
- IVExpr->getType().getQualifiers(),
- /*IsInitializer=*/false);
- EmitIgnoredExpr(IncExpr);
- FirstLCV = false;
- }
- EmitIgnoredExpr(UpExpr);
- }
- auto *SrcVD = cast<VarDecl>(cast<DeclRefExpr>(*ISrcRef)->getDecl());
- auto *DestVD = cast<VarDecl>(cast<DeclRefExpr>(*IDestRef)->getDecl());
- // Get the address of the original variable.
- Address OriginalAddr = GetAddrOfLocalVar(DestVD);
- // Get the address of the private variable.
- Address PrivateAddr = GetAddrOfLocalVar(PrivateVD);
- if (auto RefTy = PrivateVD->getType()->getAs<ReferenceType>())
- PrivateAddr =
+ llvm::DenseSet<const VarDecl *> AlreadyEmittedVars;
+ for (const auto *C : D.getClausesOfKind<OMPLastprivateClause>()) {
+ auto IRef = C->varlist_begin();
+ auto ISrcRef = C->source_exprs().begin();
+ auto IDestRef = C->destination_exprs().begin();
+ for (auto *AssignOp : C->assignment_ops()) {
+ auto *PrivateVD = cast<VarDecl>(cast<DeclRefExpr>(*IRef)->getDecl());
+ QualType Type = PrivateVD->getType();
+ auto *CanonicalVD = PrivateVD->getCanonicalDecl();
+ if (AlreadyEmittedVars.insert(CanonicalVD).second) {
+ // If lastprivate variable is a loop control variable for loop-based
+ // directive, update its value before copyin back to original
+ // variable.
+ if (auto *UpExpr = LoopCountersAndUpdates.lookup(CanonicalVD))
+ EmitIgnoredExpr(UpExpr);
+ auto *SrcVD = cast<VarDecl>(cast<DeclRefExpr>(*ISrcRef)->getDecl());
+ auto *DestVD = cast<VarDecl>(cast<DeclRefExpr>(*IDestRef)->getDecl());
+ // Get the address of the original variable.
+ Address OriginalAddr = GetAddrOfLocalVar(DestVD);
+ // Get the address of the private variable.
+ Address PrivateAddr = GetAddrOfLocalVar(PrivateVD);
+ if (auto RefTy = PrivateVD->getType()->getAs<ReferenceType>())
+ PrivateAddr =
Address(Builder.CreateLoad(PrivateAddr),
getNaturalTypeAlignment(RefTy->getPointeeType()));
- EmitOMPCopy(Type, OriginalAddr, PrivateAddr, DestVD, SrcVD, AssignOp);
- }
- ++IRef;
- ++ISrcRef;
- ++IDestRef;
+ EmitOMPCopy(Type, OriginalAddr, PrivateAddr, DestVD, SrcVD, AssignOp);
}
+ ++IRef;
+ ++ISrcRef;
+ ++IDestRef;
}
}
- if (IsLastIterCond) {
+ if (IsLastIterCond)
EmitBlock(DoneBB, /*IsFinished=*/true);
- }
}
static Address castToBase(CodeGenFunction &CGF, QualType BaseTy, QualType ElTy,
@@ -1052,10 +1029,6 @@ void CodeGenFunction::EmitOMPLoopBody(co
// The end (updates/cleanups).
EmitBlock(Continue.getBlock());
BreakContinueStack.pop_back();
- // TODO: Update lastprivates if the SeparateIter flag is true.
- // This will be implemented in a follow-up OMPLastprivateClause patch, but
- // result should be still correct without it, as we do not make these
- // variables private yet.
}
void CodeGenFunction::EmitOMPInnerLoop(
Modified: cfe/trunk/test/OpenMP/for_lastprivate_codegen.cpp
URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/test/OpenMP/for_lastprivate_codegen.cpp?rev=261209&r1=261208&r2=261209&view=diff
==============================================================================
--- cfe/trunk/test/OpenMP/for_lastprivate_codegen.cpp (original)
+++ cfe/trunk/test/OpenMP/for_lastprivate_codegen.cpp Thu Feb 18 07:48:15 2016
@@ -396,20 +396,9 @@ int main() {
// CHECK: br i1 [[IS_LAST_ITER:%.+]], label %[[LAST_THEN:.+]], label %[[LAST_DONE:.+]]
// CHECK: [[LAST_THEN]]
-// Calculate last iter count
-// CHECK: store i32 1, i32* [[OMP_IV]]
-// CHECK: [[IV1_1:%.+]] = load i32, i32* [[OMP_IV]]
-// CHECK-NEXT: [[CALC_I_2:%.+]] = add nsw i32 [[IV1_1]], 1
-// CHECK-NEXT: store i32 [[CALC_I_2]], i32* [[OMP_IV]]
-// Actual copying.
-
-// original cnt=private_cnt;
// Calculate private cnt value.
-// CHECK: [[IV1_1:%.+]] = load i32, i32* [[OMP_IV]]
-// CHECK: [[MUL:%.+]] = mul nsw i32 [[IV1_1]], 1
-// CHECK: [[ADD:%.+]] = add nsw i32 0, [[MUL]]
-// CHECK: [[CONV:%.+]] = trunc i32 [[ADD]] to i8
-// CHECK: store i8 [[CONV]], i8* [[CNT_PRIV]]
+// CHECK: store i8 2, i8* [[CNT_PRIV]]
+// original cnt=private_cnt;
// CHECK: [[CNT_VAL:%.+]] = load i8, i8* [[CNT_PRIV]],
// CHECK: store i8 [[CNT_VAL]], i8* [[CNT]],
More information about the cfe-commits
mailing list