squiid/llvm
1
0
Fork 0
Code Issues Pull requests Activity
llvm/clang/test/OpenMP/ordered_doacross_codegen.cpp
PeixinQiao a42380ce83 [OMPIRBuilder] Add ordered directive to OMPBuilder 
Add support for ordered directive in the OpenMPIRBuilder.

This patch also modidies clang to use the ordered directive when the
option -fopenmp-enable-irbuilder is enabled.

Also fix one ICE when parsing one canonical for loop with the relational
operator LE or GE in openmp region by replacing unary increment
operation of the expression of the variable "Expr A" minus the variable
"Expr B" (++(Expr A - Expr B)) with binary addition operation of the
experssion of the variable "Expr A" minus the variable "Expr B" and the
expression with constant value "1" (Expr A - Expr B + "1").

Reviewed By: Meinersbur, kiranchandramohan

Differential Revision: https://reviews.llvm.org/D107430
2021-09-03 09:37:58 +08:00

281 lines
14 KiB
C++
Raw Blame History

// RUN: %clang_cc1 -verify -fopenmp -x c++ -triple x86_64-unknown-unknown -emit-llvm %s -fexceptions -fcxx-exceptions -o - | FileCheck %s --check-prefixes=CHECK,CHECK-NORMAL
// 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 -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefixes=CHECK,CHECK-NORMAL
// RUN: %clang_cc1 -verify -fopenmp -fopenmp-enable-irbuilder -x c++ -triple x86_64-unknown-unknown -emit-llvm %s -fexceptions -fcxx-exceptions -o - | FileCheck %s --check-prefixes=CHECK,CHECK-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 -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefixes=CHECK,CHECK-IRBUILDER
// RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -triple x86_64-unknown-unknown -emit-llvm %s -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
// RUN: %clang_cc1 -fopenmp-simd -x c++ -triple x86_64-unknown-unknown -fexceptions -fcxx-exceptions -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck --check-prefix SIMD-ONLY0 %s
// SIMD-ONLY0-NOT: {{__kmpc|__tgt}}
// expected-no-diagnostics
#ifndef HEADER
#define HEADER
// CHECK: [[KMP_DIM:%.+]] = type { i64, i64, i64 }
extern int n;
int a[10], b[10], c[10], d[10];
void foo();
// CHECK-LABEL:bar
void bar() {
int i,j;
// CHECK: call void @__kmpc_doacross_init(
// CHECK: call void @__kmpc_doacross_fini(
#pragma omp parallel for ordered(2)
for (i = 0; i < n; ++i)
for (j = 0; j < n; ++j)
a[i] = b[i] + 1;
// CHECK: call void @__kmpc_doacross_init(
// CHECK: call void @__kmpc_doacross_fini(
#pragma omp for collapse(2) ordered(2)
for (int i = 0; i < n; i++)
for (int j = 0; j < n; j++)
;
}
// CHECK-LABEL: @main()
int main() {
int i;
// CHECK: [[DIMS:%.+]] = alloca [1 x [[KMP_DIM]]],
// CHECK-NORMAL: [[GTID:%.+]] = call i32 @__kmpc_global_thread_num([[IDENT:%.+]])
// CHECK: icmp
// CHECK-NEXT: br i1 %
// CHECK: [[CAST:%.+]] = bitcast [1 x [[KMP_DIM]]]* [[DIMS]] to i8*
// CHECK: call void @llvm.memset.p0i8.i64(i8* align 8 [[CAST]], i8 0, i64 24, i1 false)
// CHECK: [[DIM:%.+]] = getelementptr inbounds [1 x [[KMP_DIM]]], [1 x [[KMP_DIM]]]* [[DIMS]], i64 0, i64 0
// CHECK: getelementptr inbounds [[KMP_DIM]], [[KMP_DIM]]* [[DIM]], i32 0, i32 1
// CHECK: store i64 %{{.+}}, i64* %
// CHECK: getelementptr inbounds [[KMP_DIM]], [[KMP_DIM]]* [[DIM]], i32 0, i32 2
// CHECK: store i64 1, i64* %
// CHECK: [[DIM:%.+]] = getelementptr inbounds [1 x [[KMP_DIM]]], [1 x [[KMP_DIM]]]* [[DIMS]], i64 0, i64 0
// CHECK: [[CAST:%.+]] = bitcast [[KMP_DIM]]* [[DIM]] to i8*
// CHECK-NORMAL: call void @__kmpc_doacross_init([[IDENT]], i32 [[GTID]], i32 1, i8* [[CAST]])
// CHECK-NORMAL: call void @__kmpc_for_static_init_4(%struct.ident_t* @{{.+}}, i32 [[GTID]], i32 33, i32* %{{.+}}, i32* %{{.+}}, i32* %{{.+}}, i32* %{{.+}}, i32 1, i32 1)
#pragma omp for ordered(1)
for (int i = 0; i < n; ++i) {
a[i] = b[i] + 1;
foo();
// CHECK: invoke void [[FOO:.+]](
// CHECK: load i32, i32* [[I:%.+]],
// CHECK-NEXT: sub nsw i32 %{{.+}}, 0
// CHECK-NEXT: sdiv i32 %{{.+}}, 1
// CHECK-NEXT: sext i32 %{{.+}} to i64
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [1 x i64], [1 x i64]* [[CNT:%.+]], i64 0, i64 0
// CHECK-NEXT: store i64 %{{.+}}, i64* [[TMP]],
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [1 x i64], [1 x i64]* [[CNT]], i64 0, i64 0
// CHECK-NORMAL-NEXT: call void @__kmpc_doacross_post([[IDENT]], i32 [[GTID]], i64* [[TMP]])
// CHECK-IRBUILDER-NEXT: [[GTID18:%.+]] = call i32 @__kmpc_global_thread_num([[IDENT:%.+]])
// CHECK-IRBUILDER-NEXT: call void @__kmpc_doacross_post([[IDENT]], i32 [[GTID18]], i64* [[TMP]])
#pragma omp ordered depend(source)
c[i] = c[i] + 1;
foo();
// CHECK: invoke void [[FOO]]
// CHECK: load i32, i32* [[I]],
// CHECK-NEXT: sub nsw i32 %{{.+}}, 2
// CHECK-NEXT: sub nsw i32 %{{.+}}, 0
// CHECK-NEXT: sdiv i32 %{{.+}}, 1
// CHECK-NEXT: sext i32 %{{.+}} to i64
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [1 x i64], [1 x i64]* [[CNT:%.+]], i64 0, i64 0
// CHECK-NEXT: store i64 %{{.+}}, i64* [[TMP]],
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [1 x i64], [1 x i64]* [[CNT]], i64 0, i64 0
// CHECK-NORMAL-NEXT: call void @__kmpc_doacross_wait([[IDENT]], i32 [[GTID]], i64* [[TMP]])
// CHECK-IRBUILDER-NEXT: [[GTID30:%.+]] = call i32 @__kmpc_global_thread_num([[IDENT:%.+]])
// CHECK-IRBUILDER-NEXT: call void @__kmpc_doacross_wait([[IDENT]], i32 [[GTID30]], i64* [[TMP]])
#pragma omp ordered depend(sink : i - 2)
d[i] = a[i - 2];
}
// CHECK: landingpad
// CHECK-NORMAL: call void @__kmpc_doacross_fini([[IDENT]], i32 [[GTID]])
// CHECK: br label %
// CHECK: call void @__kmpc_for_static_fini(
// CHECK-NORMAL: call void @__kmpc_doacross_fini([[IDENT]], i32 [[GTID]])
// CHECK: ret i32 0
return 0;
}
// CHECK-LABEL: main1
int main1() {
// CHECK: [[DIMS:%.+]] = alloca [1 x [[KMP_DIM]]],
// CHECK-NORMAL: [[GTID:%.+]] = call i32 @__kmpc_global_thread_num([[IDENT:%.+]])
// CHECK: icmp
// CHECK-NEXT: br i1 %
// CHECK: [[CAST:%.+]] = bitcast [1 x [[KMP_DIM]]]* [[DIMS]] to i8*
// CHECK: call void @llvm.memset.p0i8.i64(i8* align 8 [[CAST]], i8 0, i64 24, i1 false)
// CHECK: [[DIM:%.+]] = getelementptr inbounds [1 x [[KMP_DIM]]], [1 x [[KMP_DIM]]]* [[DIMS]], i64 0, i64 0
// CHECK: getelementptr inbounds [[KMP_DIM]], [[KMP_DIM]]* [[DIM]], i32 0, i32 1
// CHECK: store i64 %{{.+}}, i64* %
// CHECK: getelementptr inbounds [[KMP_DIM]], [[KMP_DIM]]* [[DIM]], i32 0, i32 2
// CHECK: store i64 1, i64* %
// CHECK: [[DIM:%.+]] = getelementptr inbounds [1 x [[KMP_DIM]]], [1 x [[KMP_DIM]]]* [[DIMS]], i64 0, i64 0
// CHECK: [[CAST:%.+]] = bitcast [[KMP_DIM]]* [[DIM]] to i8*
// CHECK-NORMAL: call void @__kmpc_doacross_init([[IDENT]], i32 [[GTID]], i32 1, i8* [[CAST]])
// CHECK-NORMAL: call void @__kmpc_for_static_init_4(%struct.ident_t* @{{.+}}, i32 [[GTID]], i32 33, i32* %{{.+}}, i32* %{{.+}}, i32* %{{.+}}, i32* %{{.+}}, i32 1, i32 1)
#pragma omp for ordered(1)
for (int i = n; i > 0; --i) {
a[i] = b[i] + 1;
foo();
// CHECK: invoke void [[FOO:.+]](
// CHECK: [[UB_VAL:%.+]] = load i32, i32* [[UB:%.+]],
// CHECK-NEXT: [[I_VAL:%.+]] = load i32, i32* [[I:%.+]],
// CHECK-NEXT: sub i32 [[UB_VAL]], [[I_VAL]]
// CHECK-NEXT: udiv i32 %{{.+}}, 1
// CHECK-NEXT: zext i32 %{{.+}} to i64
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [1 x i64], [1 x i64]* [[CNT:%.+]], i64 0, i64 0
// CHECK-NEXT: store i64 %{{.+}}, i64* [[TMP]],
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [1 x i64], [1 x i64]* [[CNT]], i64 0, i64 0
// CHECK-NORMAL-NEXT: call void @__kmpc_doacross_post([[IDENT]], i32 [[GTID]], i64* [[TMP]])
// CHECK-IRBUILDER-NEXT: [[GTID17:%.+]] = call i32 @__kmpc_global_thread_num([[IDENT:%.+]])
// CHECK-IRBUILDER-NEXT: call void @__kmpc_doacross_post([[IDENT]], i32 [[GTID17]], i64* [[TMP]])
#pragma omp ordered depend(source)
c[i] = c[i] + 1;
foo();
// CHECK: invoke void [[FOO]]
// CHECK: [[UB_VAL:%.+]] = load i32, i32* [[UB]],
// CHECK-NEXT: [[I_VAL:%.+]] = load i32, i32* [[I]],
// CHECK-NEXT: [[SUB:%.+]] = sub nsw i32 [[I_VAL]], 2
// CHECK-NEXT: sub i32 [[UB_VAL]], [[SUB]]
// CHECK-NEXT: udiv i32 %{{.+}}, 1
// CHECK-NEXT: zext i32 %{{.+}} to i64
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [1 x i64], [1 x i64]* [[CNT:%.+]], i64 0, i64 0
// CHECK-NEXT: store i64 %{{.+}}, i64* [[TMP]],
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [1 x i64], [1 x i64]* [[CNT]], i64 0, i64 0
// CHECK-NORMAL-NEXT: call void @__kmpc_doacross_wait([[IDENT]], i32 [[GTID]], i64* [[TMP]])
// CHECK-IRBUILDER-NEXT: [[GTID29:%.+]] = call i32 @__kmpc_global_thread_num([[IDENT:%.+]])
// CHECK-IRBUILDER-NEXT: call void @__kmpc_doacross_wait([[IDENT]], i32 [[GTID29]], i64* [[TMP]])
#pragma omp ordered depend(sink : i - 2)
d[i] = a[i - 2];
}
// CHECK: landingpad
// CHECK-NORMAL: call void @__kmpc_doacross_fini([[IDENT]], i32 [[GTID]])
// CHECK: br label %
// CHECK: call void @__kmpc_for_static_fini(
// CHECK-NORMAL: call void @__kmpc_doacross_fini([[IDENT]], i32 [[GTID]])
// CHECK: ret i32 0
return 0;
}
// CHECK: define {{.+}}TestStruct
template <typename T>
struct TestStruct {
static const int M = 10;
static const int N = 20;
T i;
T a[N][M];
T b[N][M];
T foo(T, T);
T bar(T, T, T);
void baz(T, T);
TestStruct() {
// CHECK: [[DIMS:%.+]] = alloca [2 x [[KMP_DIM]]],
// CHECK-NORMAL: [[GTID:%.+]] = call i32 @__kmpc_global_thread_num([[IDENT:%.+]])
// CHECK: [[CAST:%.+]] = bitcast [2 x [[KMP_DIM]]]* [[DIMS]] to i8*
// CHECK: call void @llvm.memset.p0i8.i64(i8* align 8 [[CAST]], i8 0, i64 48, i1 false)
// CHECK: [[DIM:%.+]] = getelementptr inbounds [2 x [[KMP_DIM]]], [2 x [[KMP_DIM]]]* [[DIMS]], i64 0, i64 0
// CHECK: getelementptr inbounds [[KMP_DIM]], [[KMP_DIM]]* [[DIM]], i32 0, i32 1
// CHECK: store i64 10, i64* %
// CHECK: getelementptr inbounds [[KMP_DIM]], [[KMP_DIM]]* [[DIM]], i32 0, i32 2
// CHECK: store i64 1, i64* %
// CHECK: [[DIM:%.+]] = getelementptr inbounds [2 x [[KMP_DIM]]], [2 x [[KMP_DIM]]]* [[DIMS]], i64 0, i64 1
// CHECK: getelementptr inbounds [[KMP_DIM]], [[KMP_DIM]]* [[DIM]], i32 0, i32 1
// CHECK: store i64 %{{.+}}, i64* %
// CHECK: getelementptr inbounds [[KMP_DIM]], [[KMP_DIM]]* [[DIM]], i32 0, i32 2
// CHECK: store i64 1, i64* %
// CHECK: [[DIM:%.+]] = getelementptr inbounds [2 x [[KMP_DIM]]], [2 x [[KMP_DIM]]]* [[DIMS]], i64 0, i64 0
// CHECK: [[CAST:%.+]] = bitcast [[KMP_DIM]]* [[DIM]] to i8*
// CHECK-NORMAL: call void @__kmpc_doacross_init([[IDENT]], i32 [[GTID]], i32 2, i8* [[CAST]])
// CHECK-NORMAL: call void @__kmpc_for_static_init_4(%struct.ident_t* @{{.+}}, i32 [[GTID]], i32 33, i32* %{{.+}}, i32* %{{.+}}, i32* %{{.+}}, i32* %{{.+}}, i32 1, i32 1)
#pragma omp for ordered(2)
for (T j = 0; j < M; j++)
for (i = 0; i < n; i += 2) {
a[i][j] = foo(i, j);
// CHECK: invoke {{.+TestStruct.+foo}}
// CHECK: load i32*, i32** %
// CHECK: load i32, i32* %
// CHECK: load i32, i32* %
// CHECK: load i32, i32* [[J:%.+]],
// CHECK-NEXT: sub nsw i32 %{{.+}}, 0
// CHECK-NEXT: sdiv i32 %{{.+}}, 1
// CHECK-NEXT: sext i32 %{{.+}} to i64
// CHECK-NORMAL-NEXT: [[TMP:%.+]] = getelementptr inbounds [2 x i64], [2 x i64]* [[CNT:%.+]], i64 0, i64 0
// CHECK-NORMAL-NEXT: store i64 %{{.+}}, i64* [[TMP:%.+]],
// CHECK-NEXT: [[I:%.+]] = load i32*, i32** [[I_REF:%.+]],
// CHECK-NEXT: load i32, i32* [[I]],
// CHECK-NEXT: sub nsw i32 %{{.+}}, 2
// CHECK-NEXT: sub nsw i32 %{{.+}}, 0
// CHECK-NEXT: sdiv i32 %{{.+}}, 2
// CHECK-NEXT: sext i32 %{{.+}} to i64
// CHECK-IRBUILDER-NEXT: [[TMP:%.+]] = getelementptr inbounds [2 x i64], [2 x i64]* [[CNT:%.+]], i64 0, i64 0
// CHECK-IRBUILDER-NEXT: store i64 %{{.+}}, i64* [[TMP:%.+]],
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [2 x i64], [2 x i64]* [[CNT]], i64 0, i64 1
// CHECK-NEXT: store i64 %{{.+}}, i64* [[TMP]],
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [2 x i64], [2 x i64]* [[CNT]], i64 0, i64 0
// CHECK-NORMAL-NEXT: call void @__kmpc_doacross_wait([[IDENT]], i32 [[GTID]], i64* [[TMP]])
// CHECK-IRBUILDER-NEXT: [[GTID18:%.+]] = call i32 @__kmpc_global_thread_num([[IDENT:%.+]])
// CHECK-IRBUILDER-NEXT: call void @__kmpc_doacross_wait([[IDENT]], i32 [[GTID18]], i64* [[TMP]])
// CHECK-NEXT: load i32, i32* [[J:%.+]],
// CHECK-NEXT: sub nsw i32 %{{.+}}, 1
// CHECK-NEXT: sub nsw i32 %{{.+}}, 0
// CHECK-NEXT: sdiv i32 %{{.+}}, 1
// CHECK-NEXT: sext i32 %{{.+}} to i64
// CHECK-NORMAL-NEXT: [[TMP:%.+]] = getelementptr inbounds [2 x i64], [2 x i64]* [[CNT:%.+]], i64 0, i64 0
// CHECK-NORMAL-NEXT: store i64 %{{.+}}, i64* [[TMP]],
// CHECK-NEXT: [[I:%.+]] = load i32*, i32** [[I_REF]],
// CHECK-NEXT: load i32, i32* [[I]],
// CHECK-NEXT: sub nsw i32 %{{.+}}, 0
// CHECK-NEXT: sdiv i32 %{{.+}}, 2
// CHECK-NEXT: sext i32 %{{.+}} to i64
// CHECK-IRBUILDER-NEXT: [[TMP:%.+]] = getelementptr inbounds [2 x i64], [2 x i64]* [[CNT:%.+]], i64 0, i64 0
// CHECK-IRBUILDER-NEXT: store i64 %{{.+}}, i64* [[TMP]],
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [2 x i64], [2 x i64]* [[CNT]], i64 0, i64 1
// CHECK-NEXT: store i64 %{{.+}}, i64* [[TMP]],
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [2 x i64], [2 x i64]* [[CNT]], i64 0, i64 0
// CHECK-NORMAL-NEXT: call void @__kmpc_doacross_wait([[IDENT]], i32 [[GTID]], i64* [[TMP]])
// CHECK-IRBUILDER-NEXT: [[GTID27:%.+]] = call i32 @__kmpc_global_thread_num([[IDENT:%.+]])
// CHECK-IRBUILDER-NEXT: call void @__kmpc_doacross_wait([[IDENT]], i32 [[GTID27]], i64* [[TMP]])
#pragma omp ordered depend(sink : j, i - 2) depend(sink : j - 1, i)
b[i][j] = bar(a[i][j], b[i - 1][j], b[i][j - 1]);
// CHECK: invoke {{.+TestStruct.+bar}}
// CHECK: load i32*, i32** %
// CHECK: load i32, i32* %
// CHECK: load i32, i32* %
// CHECK: load i32, i32* [[J]],
// CHECK-NEXT: sub nsw i32 %{{.+}}, 0
// CHECK-NEXT: sdiv i32 %{{.+}}, 1
// CHECK-NEXT: sext i32 %{{.+}} to i64
// CHECK-NORMAL-NEXT: [[TMP:%.+]] = getelementptr inbounds [2 x i64], [2 x i64]* [[CNT:%.+]], i64 0, i64 0
// CHECK-NORMAL-NEXT: store i64 %{{.+}}, i64* [[TMP]],
// CHECK-NEXT: [[I:%.+]] = load i32*, i32** [[I_REF]],
// CHECK-NEXT: load i32, i32* [[I]],
// CHECK-NEXT: sub nsw i32 %{{.+}}, 0
// CHECK-NEXT: sdiv i32 %{{.+}}, 2
// CHECK-NEXT: sext i32 %{{.+}} to i64
// CHECK-IRBUILDER-NEXT: [[TMP:%.+]] = getelementptr inbounds [2 x i64], [2 x i64]* [[CNT:%.+]], i64 0, i64 0
// CHECK-IRBUILDER-NEXT: store i64 %{{.+}}, i64* [[TMP]],
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [2 x i64], [2 x i64]* [[CNT]], i64 0, i64 1
// CHECK-NEXT: store i64 %{{.+}}, i64* [[TMP]],
// CHECK-NEXT: [[TMP:%.+]] = getelementptr inbounds [2 x i64], [2 x i64]* [[CNT]], i64 0, i64 0
// CHECK-NORMAL-NEXT: call void @__kmpc_doacross_post([[IDENT]], i32 [[GTID]], i64* [[TMP]])
// CHECK-IRBUILDER-NEXT: [[GTID58:%.+]] = call i32 @__kmpc_global_thread_num([[IDENT:%.+]])
// CHECK-IRBUILDER-NEXT: call void @__kmpc_doacross_post([[IDENT]], i32 [[GTID58]], i64* [[TMP]])
#pragma omp ordered depend(source)
baz(a[i][j], b[i][j]);
}
}
// CHECK: landingpad
// CHECK-NORMAL: call void @__kmpc_doacross_fini([[IDENT]], i32 [[GTID]])
// CHECK: br label %
// CHECK: call void @__kmpc_for_static_fini(
// CHECK-NORMAL: call void @__kmpc_doacross_fini([[IDENT]], i32 [[GTID]])
// CHECK: ret
};
TestStruct<int> s;
#endif // HEADER
Reference in a new issue View git blame Copy permalink