4376fbd793
Summary: We use this `dyn_ptr` argument in Clang/OpenMP to handle the `KernelLaunchEnvironment`. This is a per-kernel argument used to share some information. Currenetly, it's prepended to the argument list and we generate storage for it in the runtime. This is bad for a few reasons: 1. It changes the ABI by shifting user arguments 2. It cannot be trivially be left uninitialized if unused 3. The runtime must allocate its own memory for it This PR changes it to be appended instead. Additionally, space for this is always emitted. This means the OMPIRBuilder itself will provide the storage, we simply need to populate it in the runtime if it is used. This means that if it's unused we don't always pay the cost and it's easier for non-OpenMP users to ignore it. Backward compatibility is maintained by auto-upgrading the kernel arguments. In `libomptarget` we completely allocate a new buffer to store this in the new format. The plugins still need to respect the old ABI of the called device object, so we simply rotate it if it's the old version.
174 lines
8.3 KiB
C++
174 lines
8.3 KiB
C++
// Test host codegen only.
|
|
// RUN: %clang_cc1 -verify -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-64
|
|
// RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s
|
|
// RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-64
|
|
// RUN: %clang_cc1 -verify -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-32
|
|
// RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s
|
|
// RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-32
|
|
|
|
// expected-no-diagnostics
|
|
#ifndef HEADER
|
|
#define HEADER
|
|
|
|
// CHECK: [[ANON_T:%.+]] = type { ptr, ptr }
|
|
// CHECK-DAG: [[SIZES_TEMPLATE:@.+]] = private {{.+}} constant [6 x i[[PTRSZ:32|64]]] [i{{32|64}} 4, i{{32|64}} 4, i{{32|64}} {{8|16}}, i{{32|64}} 0, i{{32|64}} 0, i{{32|64}} 0]
|
|
// CHECK-DAG: [[TYPES_TEMPLATE:@.+]] = private {{.+}} constant [6 x i64] [i64 800, i64 800, i64 673, i64 844424930132752, i64 844424930132752, i64 288]
|
|
// CHECK-DAG: [[SIZES:@.+]] = private {{.+}} constant [4 x i[[PTRSZ:32|64]]] [i{{32|64}} {{8|16}}, i{{32|64}} 0, i{{32|64}} 0, i{{32|64}} 0]
|
|
// CHECK-DAG: [[TYPES:@.+]] = private {{.+}} constant [4 x i64] [i64 673, i64 281474976711440, i64 281474976711440, i64 288]
|
|
// CHECK-DAG: [[TYPES3:@.+]] = private {{.+}} constant [4 x i64] [i64 545, i64 281474976711440, i64 800, i64 288]
|
|
// CHECK-DAG: [[TYPES11:@.+]] = private {{.+}} constant [6 x i64] [i64 800, i64 800, i64 549, i64 844424930132752, i64 844424930132752, i64 288]
|
|
// CHECK-DAG: [[TYPES13:@.+]] = private {{.+}} constant [3 x i64] [i64 545, i64 281474976711440, i64 288]
|
|
// CHECK-DAG: [[TYPES15:@.+]] = private {{.+}} constant [3 x i64] [i64 673, i64 281474976711440, i64 288]
|
|
|
|
template <typename F>
|
|
void omp_loop(int start, int end, F body) {
|
|
#pragma omp target teams distribute parallel for
|
|
for (int i = start; i < end; ++i) {
|
|
body(i);
|
|
}
|
|
}
|
|
|
|
template <typename F>
|
|
void omp_loop_ref(int start, int end, F& body) {
|
|
#pragma omp target teams distribute parallel for map(always, to: body)
|
|
for (int i = start; i < end; ++i) {
|
|
body(i);
|
|
}
|
|
int *p;
|
|
const auto &body_ref = [=](int i) {p[i]=0;};
|
|
#pragma omp target map(to: body_ref)
|
|
body_ref(10);
|
|
#pragma omp target
|
|
body_ref(10);
|
|
}
|
|
|
|
template <class FTy>
|
|
struct C {
|
|
static void xoo(const FTy& f) {
|
|
int x = 10;
|
|
#pragma omp target map(to:f)
|
|
f(x);
|
|
}
|
|
};
|
|
|
|
template <class FTy>
|
|
void zoo(const FTy &functor) {
|
|
C<FTy>::xoo(functor);
|
|
}
|
|
|
|
// CHECK: define {{.*}}[[MAIN:@.+]](
|
|
int main()
|
|
{
|
|
int* p = new int[100];
|
|
int* q = new int[100];
|
|
auto body = [=](int i){
|
|
p[i] = q[i];
|
|
};
|
|
zoo([=](int i){p[i] = 0;});
|
|
|
|
#pragma omp target teams distribute parallel for
|
|
for (int i = 0; i < 100; ++i) {
|
|
body(i);
|
|
}
|
|
|
|
// CHECK: [[BASE_PTRS:%.+]] = alloca [4 x ptr]{{.+}}
|
|
// CHECK: [[PTRS:%.+]] = alloca [4 x ptr]{{.+}}
|
|
|
|
// First gep of pointers inside lambdas to store the values across function call need to be ignored
|
|
// CHECK: {{%.+}} = getelementptr inbounds nuw [[ANON_T]], ptr %{{.+}}, i{{.+}} 0, i{{.+}} 0
|
|
// CHECK: {{%.+}} = getelementptr inbounds nuw [[ANON_T]], ptr %{{.+}}, i{{.+}} 0, i{{.+}} 1
|
|
|
|
// access of pointers inside lambdas
|
|
// CHECK: [[BASE_PTR1:%.+]] = getelementptr inbounds nuw [[ANON_T]], ptr %{{.+}}, i{{.+}} 0, i{{.+}} 0
|
|
// CHECK: [[PTR1:%.+]] = load ptr, ptr [[BASE_PTR1]]
|
|
// CHECK: [[BASE_PTR2:%.+]] = getelementptr inbounds nuw [[ANON_T]], ptr %{{.+}}, i{{.+}} 0, i{{.+}} 1
|
|
// CHECK: [[PTR2:%.+]] = load ptr, ptr [[BASE_PTR2]]
|
|
|
|
// storage of pointers in baseptrs and ptrs arrays
|
|
// CHECK: [[LOC_LAMBDA:%.+]] = getelementptr inbounds [4 x ptr], ptr [[BASE_PTRS]], i{{.+}} 0, i{{.+}} 0
|
|
// CHECK: store ptr %{{.+}}, ptr [[LOC_LAMBDA]]{{.+}}
|
|
// CHECK: [[LOC_LAMBDA:%.+]] = getelementptr inbounds [4 x ptr], ptr [[PTRS]], i{{.+}} 0, i{{.+}} 0
|
|
// CHECK: store ptr %{{.+}}, ptr [[LOC_LAMBDA]]{{.+}}
|
|
|
|
// CHECK: [[LOC_PTR1:%.+]] = getelementptr inbounds [4 x ptr], ptr [[BASE_PTRS]], i{{.+}} 0, i{{.+}} 1
|
|
// CHECK: store ptr [[BASE_PTR1]], ptr [[LOC_PTR1]]{{.+}}
|
|
// CHECK: [[LOC_PTR1:%.+]] = getelementptr inbounds [4 x ptr], ptr [[PTRS]], i{{.+}} 0, i{{.+}} 1
|
|
// CHECK: store ptr [[PTR1]], ptr [[LOC_PTR1]]{{.+}}
|
|
|
|
|
|
// CHECK: [[LOC_PTR2:%.+]] = getelementptr inbounds [4 x ptr], ptr [[BASE_PTRS]], i{{.+}} 0, i{{.+}} 2
|
|
// CHECK: store ptr [[BASE_PTR2]], ptr [[LOC_PTR2]]{{.+}}
|
|
// CHECK: [[LOC_PTR2:%.+]] = getelementptr inbounds [4 x ptr], ptr [[PTRS]], i{{.+}} 0, i{{.+}} 2
|
|
// CHECK: store ptr [[PTR2]], ptr [[LOC_PTR2]]{{.+}}
|
|
|
|
// actual target invocation
|
|
// CHECK: [[BASES_GEP:%.+]] = getelementptr {{.+}} [4 x ptr], ptr [[BASE_PTRS]], {{.+}} 0, {{.+}} 0
|
|
// CHECK: [[PTRS_GEP:%.+]] = getelementptr {{.+}} [4 x ptr], ptr [[PTRS]], {{.+}} 0, {{.+}} 0
|
|
// CHECK: {{%.+}} = call i32 @__tgt_target_kernel(ptr @{{.+}}, i64 -1, i32 0, i32 0, ptr @.{{.+}}.region_id, ptr %{{.+}})
|
|
|
|
omp_loop(0,100,body);
|
|
omp_loop_ref(0,100,body);
|
|
}
|
|
|
|
// CHECK: [[BASE_PTRS:%.+]] = alloca [6 x ptr]{{.+}}
|
|
// CHECK: [[PTRS:%.+]] = alloca [6 x ptr]{{.+}}
|
|
|
|
// access of pointers inside lambdas
|
|
// CHECK: [[BASE_PTR1:%.+]] = getelementptr inbounds nuw [[ANON_T]], ptr %{{.+}}, i{{.+}} 0, i{{.+}} 0
|
|
// CHECK: [[PTR1:%.+]] = load ptr, ptr [[BASE_PTR1]]
|
|
// CHECK: [[BASE_PTR2:%.+]] = getelementptr inbounds nuw [[ANON_T]], ptr %{{.+}}, i{{.+}} 0, i{{.+}} 1
|
|
// CHECK: [[PTR2:%.+]] = load ptr, ptr [[BASE_PTR2]]
|
|
|
|
// storage of pointers in baseptrs and ptrs arrays
|
|
// CHECK: [[LOC_LAMBDA:%.+]] = getelementptr inbounds [6 x ptr], ptr [[BASE_PTRS]], i{{.+}} 0, i{{.+}} 2
|
|
// CHECK: store ptr %{{.+}}, ptr [[LOC_LAMBDA]]{{.+}}
|
|
// CHECK: [[LOC_LAMBDA:%.+]] = getelementptr inbounds [6 x ptr], ptr [[PTRS]], i{{.+}} 0, i{{.+}} 2
|
|
// CHECK: store ptr %{{.+}}, ptr [[LOC_LAMBDA]]{{.+}}
|
|
|
|
// CHECK: [[LOC_PTR1:%.+]] = getelementptr inbounds [6 x ptr], ptr [[BASE_PTRS]], i{{.+}} 0, i{{.+}} 3
|
|
// CHECK: store ptr [[BASE_PTR1]], ptr [[LOC_PTR1]]{{.+}}
|
|
// CHECK: [[LOC_PTR1:%.+]] = getelementptr inbounds [6 x ptr], ptr [[PTRS]], i{{.+}} 0, i{{.+}} 3
|
|
// CHECK: store ptr [[PTR1]], ptr [[LOC_PTR1]]{{.+}}
|
|
|
|
|
|
// CHECK: [[LOC_PTR2:%.+]] = getelementptr inbounds [6 x ptr], ptr [[BASE_PTRS]], i{{.+}} 0, i{{.+}} 4
|
|
// CHECK: store ptr [[BASE_PTR2]], ptr [[LOC_PTR2]]{{.+}}
|
|
// CHECK: [[LOC_PTR2:%.+]] = getelementptr inbounds [6 x ptr], ptr [[PTRS]], i{{.+}} 0, i{{.+}} 4
|
|
// CHECK: store ptr [[PTR2]], ptr [[LOC_PTR2]]{{.+}}
|
|
|
|
// actual target invocation
|
|
// CHECK: [[BASES_GEP:%.+]] = getelementptr {{.+}} [6 x ptr], ptr [[BASE_PTRS]], {{.+}} 0, {{.+}} 0
|
|
// CHECK: [[PTRS_GEP:%.+]] = getelementptr {{.+}} [6 x ptr], ptr [[PTRS]], {{.+}} 0, {{.+}} 0
|
|
|
|
// CHECK: define internal void @{{.+}}omp_loop_ref{{.+}}(
|
|
// CHECK: [[BODY:%body.addr]] = alloca ptr
|
|
// CHECK: [[TMP:%tmp]] = alloca ptr
|
|
// CHECK: [[BODY_REF:%body_ref]] = alloca ptr
|
|
// CHECK: [[REF_TMP:%ref.tmp]] = alloca %class.anon.1
|
|
// CHECK: [[TMP8:%tmp.+]] = alloca ptr
|
|
// CHECK: [[L0:%.+]] = load ptr, ptr [[BODY]]
|
|
// CHECK: store ptr [[L0]], ptr [[TMP]]
|
|
// CHECK: [[L5:%.+]] = load ptr, ptr [[TMP]]
|
|
// CHECK-NOT [[L6:%.+]] = load ptr, ptr [[TMP]]
|
|
// CHECK-NOT [[L7:%.+]] = load ptr, ptr [[TMP]]
|
|
// CHECK: store ptr [[REF_TMP]], ptr [[BODY_REF]]
|
|
// CHECK:[[L47:%.+]] = load ptr, ptr [[BODY_REF]]
|
|
// CHECK: store ptr [[L47]], ptr [[TMP8]]
|
|
// CHECK: [[L48:%.+]] = load ptr, ptr [[TMP8]]
|
|
// CHECK-NOT: [[L49:%.+]] = load ptr, ptr [[TMP8]]
|
|
// CHECK-NOT: [[L50:%.+]] = load ptr, ptr [[TMP8]]
|
|
// CHECK: ret void
|
|
|
|
// CHECK: define internal void @{{.+}}xoo{{.+}}(
|
|
// CHECK: [[FADDR:%f.addr]] = alloca ptr
|
|
// CHECK: [[L0:%.+]] = load ptr, ptr [[FADDR]]
|
|
// CHECK: store ptr [[L0]], ptr [[TMP:%tmp]]
|
|
// CHECK: [[L1:%.+]] = load ptr, ptr [[TMP]]
|
|
// CHECK-NOT: %4 = load ptr, ptr [[TMP]]
|
|
// CHECK-NOT: %5 = load ptr, ptr [[TMP]]
|
|
// CHECK: [[L4:%.+]] = getelementptr inbounds nuw %class.anon.0, ptr [[L1]], i32 0, i32 0
|
|
// CHECK: [[L5:%.+]] = load ptr, ptr [[L4]]
|
|
// CHECK: ret void
|
|
|
|
#endif
|