| 1 | //===-- unittests/Runtime/CUDA/Allocatable.cpp ------------------*- C++ -*-===// |
| 2 | // |
| 3 | // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. |
| 4 | // See https://llvm.org/LICENSE.txt for license information. |
| 5 | // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception |
| 6 | // |
| 7 | //===----------------------------------------------------------------------===// |
| 8 | |
| 9 | #include "flang/Runtime/allocatable.h" |
| 10 | #include "cuda_runtime.h" |
| 11 | #include "gtest/gtest.h" |
| 12 | #include "flang-rt/runtime/allocator-registry.h" |
| 13 | #include "flang-rt/runtime/descriptor.h" |
| 14 | #include "flang-rt/runtime/terminator.h" |
| 15 | #include "flang/Runtime/CUDA/allocator.h" |
| 16 | #include "flang/Runtime/CUDA/common.h" |
| 17 | #include "flang/Runtime/CUDA/descriptor.h" |
| 18 | #include "flang/Support/Fortran.h" |
| 19 | |
| 20 | using namespace Fortran::runtime; |
| 21 | using namespace Fortran::runtime::cuda; |
| 22 | |
| 23 | static OwningPtr<Descriptor> createAllocatable( |
| 24 | Fortran::common::TypeCategory tc, int kind, int rank = 1) { |
| 25 | return Descriptor::Create(TypeCode{tc, kind}, kind, nullptr, rank, nullptr, |
| 26 | CFI_attribute_allocatable); |
| 27 | } |
| 28 | |
| 29 | TEST(AllocatableCUFTest, SimpleDeviceAllocatable) { |
| 30 | using Fortran::common::TypeCategory; |
| 31 | RTNAME(CUFRegisterAllocator)(); |
| 32 | // REAL(4), DEVICE, ALLOCATABLE :: a(:) |
| 33 | auto a{createAllocatable(TypeCategory::Real, 4)}; |
| 34 | a->SetAllocIdx(kDeviceAllocatorPos); |
| 35 | EXPECT_EQ((int)kDeviceAllocatorPos, a->GetAllocIdx()); |
| 36 | EXPECT_FALSE(a->HasAddendum()); |
| 37 | RTNAME(AllocatableSetBounds)(*a, 0, 1, 10); |
| 38 | |
| 39 | // Emulate a device descriptor for the purpose of unit testing part of the |
| 40 | // code. |
| 41 | Descriptor *device_desc; |
| 42 | CUDA_REPORT_IF_ERROR(cudaMalloc(&device_desc, a->SizeInBytes())); |
| 43 | |
| 44 | RTNAME(AllocatableAllocate) |
| 45 | (*a, kNoAsyncObject, /*hasStat=*/false, /*errMsg=*/nullptr, __FILE__, |
| 46 | __LINE__); |
| 47 | EXPECT_TRUE(a->IsAllocated()); |
| 48 | RTNAME(CUFDescriptorSync)(device_desc, a.get(), __FILE__, __LINE__); |
| 49 | cudaDeviceSynchronize(); |
| 50 | |
| 51 | EXPECT_EQ(cudaSuccess, cudaGetLastError()); |
| 52 | |
| 53 | RTNAME(AllocatableDeallocate) |
| 54 | (*a, /*hasStat=*/false, /*errMsg=*/nullptr, __FILE__, __LINE__); |
| 55 | EXPECT_FALSE(a->IsAllocated()); |
| 56 | |
| 57 | RTNAME(CUFDescriptorSync)(device_desc, a.get(), __FILE__, __LINE__); |
| 58 | cudaDeviceSynchronize(); |
| 59 | |
| 60 | EXPECT_EQ(cudaSuccess, cudaGetLastError()); |
| 61 | } |
| 62 | |
| 63 | TEST(AllocatableCUFTest, StreamDeviceAllocatable) { |
| 64 | using Fortran::common::TypeCategory; |
| 65 | RTNAME(CUFRegisterAllocator)(); |
| 66 | // REAL(4), DEVICE, ALLOCATABLE :: a(:) |
| 67 | auto a{createAllocatable(TypeCategory::Real, 4)}; |
| 68 | a->SetAllocIdx(kDeviceAllocatorPos); |
| 69 | EXPECT_EQ((int)kDeviceAllocatorPos, a->GetAllocIdx()); |
| 70 | EXPECT_FALSE(a->HasAddendum()); |
| 71 | RTNAME(AllocatableSetBounds)(*a, 0, 1, 10); |
| 72 | |
| 73 | auto b{createAllocatable(TypeCategory::Real, 4)}; |
| 74 | b->SetAllocIdx(kDeviceAllocatorPos); |
| 75 | EXPECT_EQ((int)kDeviceAllocatorPos, b->GetAllocIdx()); |
| 76 | EXPECT_FALSE(b->HasAddendum()); |
| 77 | RTNAME(AllocatableSetBounds)(*b, 0, 1, 20); |
| 78 | |
| 79 | auto c{createAllocatable(TypeCategory::Real, 4)}; |
| 80 | c->SetAllocIdx(kDeviceAllocatorPos); |
| 81 | EXPECT_EQ((int)kDeviceAllocatorPos, c->GetAllocIdx()); |
| 82 | EXPECT_FALSE(b->HasAddendum()); |
| 83 | RTNAME(AllocatableSetBounds)(*c, 0, 1, 100); |
| 84 | |
| 85 | RTNAME(AllocatableAllocate) |
| 86 | (*a, /*asyncObject=*/nullptr, /*hasStat=*/false, /*errMsg=*/nullptr, __FILE__, |
| 87 | __LINE__); |
| 88 | EXPECT_TRUE(a->IsAllocated()); |
| 89 | cudaDeviceSynchronize(); |
| 90 | EXPECT_EQ(cudaSuccess, cudaGetLastError()); |
| 91 | |
| 92 | RTNAME(AllocatableAllocate) |
| 93 | (*b, /*asyncObject=*/nullptr, /*hasStat=*/false, /*errMsg=*/nullptr, __FILE__, |
| 94 | __LINE__); |
| 95 | EXPECT_TRUE(b->IsAllocated()); |
| 96 | cudaDeviceSynchronize(); |
| 97 | EXPECT_EQ(cudaSuccess, cudaGetLastError()); |
| 98 | |
| 99 | RTNAME(AllocatableAllocate) |
| 100 | (*c, /*asyncObject=*/nullptr, /*hasStat=*/false, /*errMsg=*/nullptr, __FILE__, |
| 101 | __LINE__); |
| 102 | EXPECT_TRUE(c->IsAllocated()); |
| 103 | cudaDeviceSynchronize(); |
| 104 | EXPECT_EQ(cudaSuccess, cudaGetLastError()); |
| 105 | |
| 106 | RTNAME(AllocatableDeallocate) |
| 107 | (*b, /*hasStat=*/false, /*errMsg=*/nullptr, __FILE__, __LINE__); |
| 108 | EXPECT_FALSE(b->IsAllocated()); |
| 109 | cudaDeviceSynchronize(); |
| 110 | EXPECT_EQ(cudaSuccess, cudaGetLastError()); |
| 111 | |
| 112 | RTNAME(AllocatableDeallocate) |
| 113 | (*a, /*hasStat=*/false, /*errMsg=*/nullptr, __FILE__, __LINE__); |
| 114 | EXPECT_FALSE(a->IsAllocated()); |
| 115 | cudaDeviceSynchronize(); |
| 116 | EXPECT_EQ(cudaSuccess, cudaGetLastError()); |
| 117 | |
| 118 | RTNAME(AllocatableDeallocate) |
| 119 | (*c, /*hasStat=*/false, /*errMsg=*/nullptr, __FILE__, __LINE__); |
| 120 | EXPECT_FALSE(c->IsAllocated()); |
| 121 | cudaDeviceSynchronize(); |
| 122 | EXPECT_EQ(cudaSuccess, cudaGetLastError()); |
| 123 | } |
| 124 | |