[clang] 56c3b2b - [OPENMP]Fix capturing of global variables in OpenMP regions.
Alexey Bataev via cfe-commits
cfe-commits at lists.llvm.org
Thu Apr 9 07:40:01 PDT 2020
Author: Alexey Bataev
Date: 2020-04-09T10:36:09-04:00
New Revision: 56c3b2b091bf94ed5c3d614f3d10e8e973a72d17
URL: https://github.com/llvm/llvm-project/commit/56c3b2b091bf94ed5c3d614f3d10e8e973a72d17
DIFF: https://github.com/llvm/llvm-project/commit/56c3b2b091bf94ed5c3d614f3d10e8e973a72d17.diff
LOG: [OPENMP]Fix capturing of global variables in OpenMP regions.
Summary:
The global variable should be captured in the region only if it was
privitized in the region or in any of the outer regions. Otherwise, it
should not be captured.
Reviewers: jdoerfert
Subscribers: yaxunl, guansong, cfe-commits, caomhin
Tags: #clang
Differential Revision: https://reviews.llvm.org/D77731
Added:
Modified:
clang/lib/Sema/SemaOpenMP.cpp
clang/test/OpenMP/parallel_codegen.cpp
Removed:
################################################################################
diff --git a/clang/lib/Sema/SemaOpenMP.cpp b/clang/lib/Sema/SemaOpenMP.cpp
index 6718656f6133..c675672f4f29 100644
--- a/clang/lib/Sema/SemaOpenMP.cpp
+++ b/clang/lib/Sema/SemaOpenMP.cpp
@@ -2110,21 +2110,22 @@ VarDecl *Sema::isOpenMPCapturedDecl(ValueDecl *D, bool CheckScopeInfo,
isImplicitOrExplicitTaskingRegion(DSAStack->getCurrentDirective())) ||
(VD && DSAStack->isForceVarCapturing()))
return VD ? VD : Info.second;
- DSAStackTy::DSAVarData DVarPrivate =
+ DSAStackTy::DSAVarData DVarTop =
DSAStack->getTopDSA(D, DSAStack->isClauseParsingMode());
- if (DVarPrivate.CKind != OMPC_unknown && isOpenMPPrivate(DVarPrivate.CKind))
- return VD ? VD : cast<VarDecl>(DVarPrivate.PrivateCopy->getDecl());
+ if (DVarTop.CKind != OMPC_unknown && isOpenMPPrivate(DVarTop.CKind))
+ return VD ? VD : cast<VarDecl>(DVarTop.PrivateCopy->getDecl());
// Threadprivate variables must not be captured.
- if (isOpenMPThreadPrivate(DVarPrivate.CKind))
- return nullptr;
- // Global shared must not be captured.
- if (VD && !VD->hasLocalStorage() && DVarPrivate.CKind == OMPC_shared)
+ if (isOpenMPThreadPrivate(DVarTop.CKind))
return nullptr;
// The variable is not private or it is the variable in the directive with
// default(none) clause and not used in any clause.
- DVarPrivate = DSAStack->hasDSA(D, isOpenMPPrivate,
- [](OpenMPDirectiveKind) { return true; },
- DSAStack->isClauseParsingMode());
+ DSAStackTy::DSAVarData DVarPrivate = DSAStack->hasDSA(
+ D, isOpenMPPrivate, [](OpenMPDirectiveKind) { return true; },
+ DSAStack->isClauseParsingMode());
+ // Global shared must not be captured.
+ if (VD && !VD->hasLocalStorage() && DVarPrivate.CKind == OMPC_unknown &&
+ (DSAStack->getDefaultDSA() != DSA_none || DVarTop.CKind == OMPC_shared))
+ return nullptr;
if (DVarPrivate.CKind != OMPC_unknown ||
(VD && DSAStack->getDefaultDSA() == DSA_none))
return VD ? VD : cast<VarDecl>(DVarPrivate.PrivateCopy->getDecl());
diff --git a/clang/test/OpenMP/parallel_codegen.cpp b/clang/test/OpenMP/parallel_codegen.cpp
index d42312193709..b38e9ae706bd 100644
--- a/clang/test/OpenMP/parallel_codegen.cpp
+++ b/clang/test/OpenMP/parallel_codegen.cpp
@@ -1,9 +1,9 @@
// RUN: %clang_cc1 -verify -fopenmp -x c++ -emit-llvm %s -triple %itanium_abi_triple -fexceptions -fcxx-exceptions -o - | FileCheck %s --check-prefixes=ALL,CHECK
// RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -emit-pch -o %t %s
// RUN: %clang_cc1 -fopenmp -x c++ -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -debug-info-kind=limited -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck --check-prefixes=ALL-DEBUG,CHECK-DEBUG %s
-// RUN: %clang_cc1 -verify -fopenmp -fopenmp-enable-irbuilder -x c++ -emit-llvm %s -triple %itanium_abi_triple -fexceptions -fcxx-exceptions -o - | FileCheck %s --check-prefixes=ALL,IRBUILDER
-// RUN: %clang_cc1 -fopenmp -fopenmp-enable-irbuilder -x c++ -std=c++11 -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -emit-pch -o %t %s
-// RUN: %clang_cc1 -fopenmp -fopenmp-enable-irbuilder -x c++ -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -debug-info-kind=limited -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck --check-prefixes=ALL-DEBUG,IRBUILDER-DEBUG %s
+// RUN: %clang_cc1 -verify -fopenmp -fopenmp-enable-irbuilder -DIRBUILDER -x c++ -emit-llvm %s -triple %itanium_abi_triple -fexceptions -fcxx-exceptions -o - | FileCheck %s --check-prefixes=ALL,IRBUILDER
+// RUN: %clang_cc1 -fopenmp -fopenmp-enable-irbuilder -DIRBUILDER -x c++ -std=c++11 -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -emit-pch -o %t %s
+// RUN: %clang_cc1 -fopenmp -fopenmp-enable-irbuilder -DIRBUILDER -x c++ -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -debug-info-kind=limited -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck --check-prefixes=ALL-DEBUG,IRBUILDER-DEBUG %s
// RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -emit-llvm %s -triple %itanium_abi_triple -fexceptions -fcxx-exceptions -o - | FileCheck --check-prefix SIMD-ONLY0 %s
// RUN: %clang_cc1 -fopenmp-simd -x c++ -std=c++11 -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -emit-pch -o %t %s
@@ -45,7 +45,18 @@ int global;
int main (int argc, char **argv) {
int a[argc];
#pragma omp parallel shared(global, a) default(none)
- (void)global, foo(a[1]);
+ foo(a[1]), a[1] = global;
+#ifndef IRBUILDER
+// TODO: Support for privates in IRBuilder.
+#pragma omp parallel private(global, a) default(none)
+#pragma omp parallel shared(global, a) default(none)
+ foo(a[1]), a[1] = global;
+// FIXME: IRBuilder crashes in void llvm::OpenMPIRBuilder::finalize()
+// Assertion `Extractor.isEligible() && "Expected OpenMP outlining to be possible!"' failed.
+#pragma omp parallel shared(global, a) default(none)
+#pragma omp parallel shared(global, a) default(none)
+ foo(a[1]), a[1] = global;
+#endif // IRBUILDER
return tmain(argv);
}
@@ -53,6 +64,8 @@ int main (int argc, char **argv) {
// ALL: store i32 %argc, i32* [[ARGC_ADDR:%.+]],
// ALL: [[VLA:%.+]] = alloca i32, i{{[0-9]+}} [[VLA_SIZE:%[^,]+]],
// CHECK: call {{.*}}void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* [[DEF_LOC_2]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i{{[0-9]+}}, i32*)* [[OMP_OUTLINED:@.+]] to void (i32*, i32*, ...)*), i{{[0-9]+}} [[VLA_SIZE]], i32* [[VLA]])
+// CHECK: call {{.*}}void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* [[DEF_LOC_2]], i32 1, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64)* [[OMP_OUTLINED1:@.+]] to void (i32*, i32*, ...)*), i{{[0-9]+}} [[VLA_SIZE]])
+// CHECK: call {{.*}}void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* [[DEF_LOC_2]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i{{[0-9]+}}, i32*)* [[OMP_OUTLINED2:@.+]] to void (i32*, i32*, ...)*), i{{[0-9]+}} [[VLA_SIZE]], i32* [[VLA]])
// IRBUILDER: call {{.*}}void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* [[DEF_LOC_2]], i32 1, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i32*)* [[OMP_OUTLINED:@.+]] to void (i32*, i32*, ...)*), i32* [[VLA]])
// ALL: [[ARGV:%.+]] = load i8**, i8*** {{%[a-z0-9.]+}}
// ALL-NEXT: [[RET:%.+]] = call {{[a-z\_\b]*[ ]?i32}} [[TMAIN:@.+tmain.+]](i8** [[ARGV]])
@@ -73,7 +86,7 @@ int main (int argc, char **argv) {
// ALL-DEBUG: ret i32
// ALL-DEBUG-NEXT: }
-// CHECK: define internal {{.*}}void [[OMP_OUTLINED]](i32* noalias %.global_tid., i32* noalias %.bound_tid., i{{[0-9]+}}{{.*}} [[VLA_SIZE:%.+]], i32* {{.+}} [[VLA_ADDR:%[^)]+]])
+// CHECK: define internal {{.*}}void [[OMP_OUTLINED]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, i{{[0-9]+}}{{.*}} [[VLA_SIZE:%.+]], i32* {{.+}} [[VLA_ADDR:%[^)]+]])
// CHECK-SAME: #[[FN_ATTRS:[0-9]+]]
// IRBUILDER: define internal {{.*}}void [[OMP_OUTLINED]](i32* noalias %{{.*}}, i32* noalias %{{.*}}, i32* [[VLA_REF:%[^)]+]])
// IRBUILDER-SAME: #[[FN_ATTRS:[0-9]+]]
@@ -83,11 +96,12 @@ int main (int argc, char **argv) {
// ALL-NEXT: [[VLA_ELEM:%.+]] = load i32, i32* [[VLA_ELEM_REF]]
// CHECK-NEXT: invoke {{.*}}void [[FOO:@.+foo.+]](i32{{[ ]?[a-z]*}} [[VLA_ELEM]])
// IRBUILDER: call {{.*}}void [[FOO:@.+foo.+]](i32{{[ ]?[a-z]*}} [[VLA_ELEM]])
+// ALL: load i32, i32* @
// CHECK: ret void
// CHECK: call {{.*}}void @{{.+terminate.*|abort}}(
// CHECK-NEXT: unreachable
// CHECK-NEXT: }
-// CHECK-DEBUG: define internal void [[OMP_OUTLINED_DEBUG:@.+]](i32* noalias %.global_tid., i32* noalias %.bound_tid., i64 [[VLA_SIZE:%.+]], i32* {{.+}} [[VLA_ADDR:%[^)]+]])
+// CHECK-DEBUG: define internal void [[OMP_OUTLINED_DEBUG:@.+]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, i64 [[VLA_SIZE:%.+]], i32* {{.+}} [[VLA_ADDR:%[^)]+]])
// CHECK-DEBUG-SAME: #[[FN_ATTRS:[0-9]+]]
// IRBUILDER-DEBUG: define internal void [[OMP_OUTLINED_DEBUG:@.+]](i32* noalias %{{.*}}, i32* noalias %{{.*}}, i32* [[VLA_REF:%[^)]+]])
// IRBUILDER-DEBUG-SAME: #[[FN_ATTRS:[0-9]+]]
@@ -105,6 +119,20 @@ int main (int argc, char **argv) {
// ALL-DAG: define linkonce_odr {{.*}}void [[FOO]]({{i32[ ]?[a-z]*}} %argc)
// ALL-DAG: declare !callback ![[cbid:[0-9]+]] {{.*}}void @__kmpc_fork_call(%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...)
// ALL-DEBUG-DAG: define linkonce_odr void [[FOO]](i32 %argc)
+
+// CHECK: define internal {{.*}}void [[OMP_OUTLINED1]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, i{{[0-9]+}}{{.*}} [[VLA_SIZE:%.+]])
+// CHECK: call {{.*}}void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* [[DEF_LOC_2]], i32 3, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i32*, i32*)* [[OMP_OUTLINED11:@.+]] to void (i32*, i32*, ...)*), i64 %{{.+}}, i32* %{{.+}}, i32* %{{.+}})
+
+// CHECK: define internal {{.*}}void [[OMP_OUTLINED11]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, i{{[0-9]+}}{{.*}} [[VLA_SIZE:%.+]], i32* {{.+}} [[VLA_ADDR:%[^)]+]], i32* {{.+}} %{{.+}})
+// CHECK-NOT: load i32, i32* @
+
+// CHECK: define internal {{.*}}void [[OMP_OUTLINED2]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, i{{[0-9]+}}{{.*}} [[VLA_SIZE:%.+]], i32* {{.+}} [[VLA_ADDR:%[^)]+]])
+// CHECK: call {{.*}}void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* [[DEF_LOC_2]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i32*)* [[OMP_OUTLINED21:@.+]] to void (i32*, i32*, ...)*), i64 %{{.+}}, i32* %{{.+}})
+
+
+// CHECK: define internal {{.*}}void [[OMP_OUTLINED21]](i32* noalias %{{.+}}, i32* noalias %{{.+}}, i{{[0-9]+}}{{.*}} [[VLA_SIZE:%.+]], i32* {{.+}} [[VLA_ADDR:%[^)]+]])
+// CHECK: load i32, i32* @
+
// ALL-DEBUG-DAG: declare !callback ![[cbid:[0-9]+]] void @__kmpc_fork_call(%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...)
// CHECK-DEBUG-DAG: define internal void [[OMP_OUTLINED]](i32* noalias %.global_tid., i32* noalias %.bound_tid., i64 [[VLA_SIZE:%.+]], i32* {{.+}} [[VLA_ADDR:%[^)]+]])
// CHECK-DEBUG-DAG: call void [[OMP_OUTLINED_DEBUG]]
More information about the cfe-commits
mailing list