This patch changes the code we generate to enter a target region on the device. This is in-line with the new definition in the runtime that was added previously. Additionally we implement this in the OpenMPIRBuilder so that this code can be shared with Flang in the future. Reviewed By: ABataev Differential Revision: https://reviews.llvm.org/D128550
85 lines
7.8 KiB
C++
85 lines
7.8 KiB
C++
// expected-no-diagnostics
|
|
#ifndef HEADER
|
|
#define HEADER
|
|
|
|
///==========================================================================///
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -verify -fopenmp -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-unknown-unknown -emit-llvm %s -o - | FileCheck -allow-deprecated-dag-overlap %s --check-prefix CK16 --check-prefix CK16-64
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -fopenmp -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -emit-pch -o %t %s
|
|
// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-unknown-unknown -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck -allow-deprecated-dag-overlap %s --check-prefix CK16 --check-prefix CK16-64
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -verify -fopenmp -fopenmp-targets=i386-pc-linux-gnu -x c++ -triple i386-unknown-unknown -emit-llvm %s -o - | FileCheck -allow-deprecated-dag-overlap %s --check-prefix CK16 --check-prefix CK16-32
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -fopenmp -fopenmp-targets=i386-pc-linux-gnu -x c++ -std=c++11 -triple i386-unknown-unknown -emit-pch -o %t %s
|
|
// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-targets=i386-pc-linux-gnu -x c++ -triple i386-unknown-unknown -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck -allow-deprecated-dag-overlap %s --check-prefix CK16 --check-prefix CK16-32
|
|
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -verify -fopenmp -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-unknown-unknown -emit-llvm %s -o - | FileCheck -allow-deprecated-dag-overlap %s --check-prefix CK16 --check-prefix CK16-64
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -fopenmp -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -emit-pch -o %t %s
|
|
// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-unknown-unknown -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck -allow-deprecated-dag-overlap %s --check-prefix CK16 --check-prefix CK16-64
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -verify -fopenmp -fopenmp-version=45 -fopenmp-targets=i386-pc-linux-gnu -x c++ -triple i386-unknown-unknown -emit-llvm %s -o - | FileCheck -allow-deprecated-dag-overlap %s --check-prefix CK16 --check-prefix CK16-32
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -fopenmp -fopenmp-version=45 -fopenmp-targets=i386-pc-linux-gnu -x c++ -std=c++11 -triple i386-unknown-unknown -emit-pch -o %t %s
|
|
// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-version=45 -fopenmp-targets=i386-pc-linux-gnu -x c++ -triple i386-unknown-unknown -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck -allow-deprecated-dag-overlap %s --check-prefix CK16 --check-prefix CK16-32
|
|
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -verify -fopenmp -fopenmp-version=50 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-unknown-unknown -emit-llvm %s -o - | FileCheck -allow-deprecated-dag-overlap %s --check-prefix CK16 --check-prefix CK16-64
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -fopenmp -fopenmp-version=50 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -emit-pch -o %t %s
|
|
// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-version=50 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-unknown-unknown -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck -allow-deprecated-dag-overlap %s --check-prefix CK16 --check-prefix CK16-64
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -verify -fopenmp -fopenmp-version=50 -fopenmp-targets=i386-pc-linux-gnu -x c++ -triple i386-unknown-unknown -emit-llvm %s -o - | FileCheck -allow-deprecated-dag-overlap %s --check-prefix CK16 --check-prefix CK16-32
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -fopenmp -fopenmp-version=50 -fopenmp-targets=i386-pc-linux-gnu -x c++ -std=c++11 -triple i386-unknown-unknown -emit-pch -o %t %s
|
|
// RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-version=50 -fopenmp-targets=i386-pc-linux-gnu -x c++ -triple i386-unknown-unknown -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck -allow-deprecated-dag-overlap %s --check-prefix CK16 --check-prefix CK16-32
|
|
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -verify -fopenmp-simd -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-unknown-unknown -emit-llvm %s -o - | FileCheck -allow-deprecated-dag-overlap --check-prefix SIMD-ONLY15 %s
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -fopenmp-simd -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -emit-pch -o %t %s
|
|
// RUN: %clang_cc1 -no-opaque-pointers -fopenmp-simd -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-unknown-unknown -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck -allow-deprecated-dag-overlap --check-prefix SIMD-ONLY15 %s
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -verify -fopenmp-simd -fopenmp-targets=i386-pc-linux-gnu -x c++ -triple i386-unknown-unknown -emit-llvm %s -o - | FileCheck -allow-deprecated-dag-overlap --check-prefix SIMD-ONLY15 %s
|
|
// RUN: %clang_cc1 -no-opaque-pointers -DCK16 -fopenmp-simd -fopenmp-targets=i386-pc-linux-gnu -x c++ -std=c++11 -triple i386-unknown-unknown -emit-pch -o %t %s
|
|
// RUN: %clang_cc1 -no-opaque-pointers -fopenmp-simd -fopenmp-targets=i386-pc-linux-gnu -x c++ -triple i386-unknown-unknown -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck -allow-deprecated-dag-overlap --check-prefix SIMD-ONLY15 %s
|
|
// SIMD-ONLY15-NOT: {{__kmpc|__tgt}}
|
|
#ifdef CK16
|
|
|
|
// CK16-DAG: [[SIZES:@.+]] = {{.+}}constant [1 x i64] [i64 4]
|
|
// Map types:
|
|
// - OMP_MAP_PRIVATE_VAL + OMP_MAP_TARGET_PARAM | OMP_MAP_IMPLICIT = 800
|
|
// CK16-DAG: [[TYPES:@.+]] = {{.+}}constant [1 x i64] [i64 800]
|
|
|
|
template<int y>
|
|
int foo(int d) {
|
|
int res = d;
|
|
#pragma omp target
|
|
{
|
|
res += y;
|
|
}
|
|
return res;
|
|
}
|
|
// CK16-LABEL: implicit_maps_templated_function{{.*}}(
|
|
void implicit_maps_templated_function (int a){
|
|
int i = a;
|
|
|
|
// CK16: define {{.*}}i32 @{{.+}}foo{{.+}}(i32 {{[^,]+}})
|
|
// CK16-DAG: call i32 @__tgt_target_kernel(%struct.ident_t* @{{.+}}, i64 -1, i32 -1, i32 0, i8* @.{{.+}}.region_id, %struct.__tgt_kernel_arguments* [[ARGS:%.+]])
|
|
// CK16-DAG: [[BPARG:%.+]] = getelementptr inbounds {{.+}}[[ARGS]], i32 0, i32 2
|
|
// CK16-DAG: store i8** [[BPGEP:%.+]], i8*** [[BPARG]]
|
|
// CK16-DAG: [[PARG:%.+]] = getelementptr inbounds {{.+}}[[ARGS]], i32 0, i32 3
|
|
// CK16-DAG: store i8** [[PGEP:%.+]], i8*** [[PARG]]
|
|
// CK16-DAG: [[BPGEP]] = getelementptr inbounds {{.+}}[[BPS:%[^,]+]], i32 0, i32 0
|
|
// CK16-DAG: [[PGEP]] = getelementptr inbounds {{.+}}[[PS:%[^,]+]], i32 0, i32 0
|
|
|
|
// CK16-DAG: [[BP1:%.+]] = getelementptr inbounds {{.+}}[[BPS]], i32 0, i32 0
|
|
// CK16-DAG: [[P1:%.+]] = getelementptr inbounds {{.+}}[[PS]], i32 0, i32 0
|
|
// CK16-DAG: [[CBP1:%.+]] = bitcast i8** [[BP1]] to i[[sz:64|32]]*
|
|
// CK16-DAG: [[CP1:%.+]] = bitcast i8** [[P1]] to i[[sz]]*
|
|
// CK16-DAG: store i[[sz]] [[VAL:%.+]], i[[sz]]* [[CBP1]]
|
|
// CK16-DAG: store i[[sz]] [[VAL]], i[[sz]]* [[CP1]]
|
|
// CK16-DAG: [[VAL]] = load i[[sz]], i[[sz]]* [[ADDR:%.+]],
|
|
// CK16-64-DAG: [[CADDR:%.+]] = bitcast i[[sz]]* [[ADDR]] to i32*
|
|
// CK16-64-DAG: store i32 {{.+}}, i32* [[CADDR]],
|
|
|
|
// CK16: call void [[KERNEL:@.+]](i[[sz]] [[VAL]])
|
|
i = foo<543>(i);
|
|
}
|
|
// CK16: define internal void [[KERNEL]](i[[sz]] noundef [[ARG:%.+]])
|
|
// CK16: [[ADDR:%.+]] = alloca i[[sz]],
|
|
// CK16: store i[[sz]] [[ARG]], i[[sz]]* [[ADDR]],
|
|
// CK16-64: [[CADDR:%.+]] = bitcast i64* [[ADDR]] to i32*
|
|
// CK16-64: {{.+}} = load i32, i32* [[CADDR]],
|
|
// CK16-32: {{.+}} = load i32, i32* [[ADDR]],
|
|
|
|
#endif // CK16
|
|
#endif
|