1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
|
// MIT License
//
// Copyright (c) 2022-2024 Advanced Micro Devices, Inc. All rights reserved.
//
// Permission is hereby granted, free of charge, to any person obtaining a copy
// of this software and associated documentation files (the "Software"), to deal
// in the Software without restriction, including without limitation the rights
// to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
// copies of the Software, and to permit persons to whom the Software is
// furnished to do so, subject to the following conditions:
//
// The above copyright notice and this permission notice shall be included in
// all copies or substantial portions of the Software.
//
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
// FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
// AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
// LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
// OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
// SOFTWARE.
#include "common_benchmark_header.hpp"
// HIP API
#include "hipcub/block/block_load.hpp"
#include "hipcub/block/block_radix_rank.hpp"
#include "hipcub/block/block_store.hpp"
#include "hipcub/block/radix_rank_sort_operations.hpp"
#ifndef DEFAULT_N
const size_t DEFAULT_N = 1024 * 1024 * 128;
#endif
enum class RadixRankAlgorithm
{
RADIX_RANK_BASIC,
RADIX_RANK_MEMOIZE,
RADIX_RANK_MATCH,
};
template<class T,
unsigned int RadixBits,
bool Descending,
RadixRankAlgorithm BenchmarkKind,
unsigned int BlockSize,
unsigned int ItemsPerThread,
unsigned int Trials>
__global__ __launch_bounds__(BlockSize) void rank_kernel(const T* keys_input, int* ranks_output)
{
const unsigned int lid = hipThreadIdx_x;
const unsigned int block_offset = hipBlockIdx_x * ItemsPerThread * BlockSize;
T keys[ItemsPerThread];
hipcub::LoadDirectBlocked(lid, keys_input + block_offset, keys);
using KeyTraits = hipcub::Traits<T>;
using UnsignedBits = typename KeyTraits::UnsignedBits;
using DigitExtractor = hipcub::BFEDigitExtractor<T>;
UnsignedBits(&unsigned_keys)[ItemsPerThread]
= reinterpret_cast<UnsignedBits(&)[ItemsPerThread]>(keys);
using RankType = std::conditional_t<
BenchmarkKind == RadixRankAlgorithm::RADIX_RANK_MATCH,
hipcub::BlockRadixRankMatch<BlockSize, RadixBits, Descending>,
hipcub::BlockRadixRank<BlockSize,
RadixBits,
Descending,
BenchmarkKind == RadixRankAlgorithm::RADIX_RANK_MEMOIZE>>;
#pragma unroll
for(unsigned int key = 0; key < ItemsPerThread; key++)
{
unsigned_keys[key] = KeyTraits::TwiddleIn(unsigned_keys[key]);
}
int ranks[ItemsPerThread];
#pragma nounroll
for(unsigned int trial = 0; trial < Trials; trial++)
{
__shared__ typename RankType::TempStorage storage;
RankType rank(storage);
unsigned begin_bit = 0;
const unsigned end_bit = sizeof(T) * 8;
while(begin_bit < end_bit)
{
const unsigned pass_bits = min(RadixBits, end_bit - begin_bit);
DigitExtractor digit_extractor(begin_bit, pass_bits);
rank.RankKeys(unsigned_keys, ranks, digit_extractor);
begin_bit += RadixBits;
}
}
hipcub::StoreDirectBlocked(lid, ranks_output + block_offset, ranks);
}
template<class T,
RadixRankAlgorithm BenchmarkKind,
unsigned int BlockSize,
unsigned int ItemsPerThread,
unsigned int Trials = 10>
void run_benchmark(benchmark::State& state, hipStream_t stream, size_t N)
{
constexpr unsigned int items_per_block = BlockSize * ItemsPerThread;
const unsigned int size = items_per_block * ((N + items_per_block - 1) / items_per_block);
std::vector<T> input
= benchmark_utils::get_random_data<T>(size,
benchmark_utils::generate_limits<T>::min(),
benchmark_utils::generate_limits<T>::max());
T* d_input;
int* d_output;
HIP_CHECK(hipMalloc(&d_input, size * sizeof(T)));
HIP_CHECK(hipMalloc(&d_output, size * sizeof(int)));
HIP_CHECK(hipMemcpy(d_input, input.data(), size * sizeof(T), hipMemcpyHostToDevice));
HIP_CHECK(hipDeviceSynchronize());
for(auto _ : state)
{
auto start = std::chrono::high_resolution_clock::now();
hipLaunchKernelGGL(
HIP_KERNEL_NAME(
rank_kernel<T, 4, false, BenchmarkKind, BlockSize, ItemsPerThread, Trials>),
dim3(size / items_per_block),
dim3(BlockSize),
0,
stream,
d_input,
d_output);
HIP_CHECK(hipPeekAtLastError());
HIP_CHECK(hipDeviceSynchronize());
auto end = std::chrono::high_resolution_clock::now();
auto elapsed_seconds
= std::chrono::duration_cast<std::chrono::duration<double>>(end - start);
state.SetIterationTime(elapsed_seconds.count());
}
state.SetBytesProcessed(state.iterations() * Trials * size * sizeof(T));
state.SetItemsProcessed(state.iterations() * Trials * size);
HIP_CHECK(hipFree(d_input));
HIP_CHECK(hipFree(d_output));
}
#define CREATE_BENCHMARK(T, KIND, BS, IPT) \
benchmark::RegisterBenchmark(std::string("block_radix_rank<data_type:" #T ",kind:" #KIND \
",block_size:" #BS ",items_per_thread:" #IPT ">." \
+ name) \
.c_str(), \
&run_benchmark<T, KIND, BS, IPT>, \
stream, \
size)
// clang-format off
#define CREATE_BENCHMARK_KINDS(type, block, ipt) \
CREATE_BENCHMARK(type, RadixRankAlgorithm::RADIX_RANK_BASIC, block, ipt), \
CREATE_BENCHMARK(type, RadixRankAlgorithm::RADIX_RANK_MEMOIZE, block, ipt), \
CREATE_BENCHMARK(type, RadixRankAlgorithm::RADIX_RANK_MATCH, block, ipt)
#define BENCHMARK_TYPE(type, block) \
CREATE_BENCHMARK_KINDS(type, block, 1), \
CREATE_BENCHMARK_KINDS(type, block, 4), \
CREATE_BENCHMARK_KINDS(type, block, 8), \
CREATE_BENCHMARK_KINDS(type, block, 16), \
CREATE_BENCHMARK_KINDS(type, block, 32)
// clang-format on
void add_benchmarks(const std::string& name,
std::vector<benchmark::internal::Benchmark*>& benchmarks,
hipStream_t stream,
size_t size)
{
std::vector<benchmark::internal::Benchmark*> bs = {
BENCHMARK_TYPE(int, 128),
BENCHMARK_TYPE(int, 256),
BENCHMARK_TYPE(int, 512),
BENCHMARK_TYPE(uint8_t, 128),
BENCHMARK_TYPE(uint8_t, 256),
BENCHMARK_TYPE(uint8_t, 512),
BENCHMARK_TYPE(long long, 128),
BENCHMARK_TYPE(long long, 256),
BENCHMARK_TYPE(long long, 512),
};
benchmarks.insert(benchmarks.end(), bs.begin(), bs.end());
}
int main(int argc, char* argv[])
{
cli::Parser parser(argc, argv);
parser.set_optional<size_t>("size", "size", DEFAULT_N, "number of values");
parser.set_optional<int>("trials", "trials", -1, "number of iterations");
parser.run_and_exit_if_error();
// Parse argv
benchmark::Initialize(&argc, argv);
const size_t size = parser.get<size_t>("size");
const int trials = parser.get<int>("trials");
// HIP
hipStream_t stream = 0; // default
hipDeviceProp_t devProp;
int device_id = 0;
HIP_CHECK(hipGetDevice(&device_id));
HIP_CHECK(hipGetDeviceProperties(&devProp, device_id));
std::cout << "benchmark_block_radix_rank" << std::endl;
std::cout << "[HIP] Device name: " << devProp.name << std::endl;
// Add benchmarks
std::vector<benchmark::internal::Benchmark*> benchmarks;
add_benchmarks("rank", benchmarks, stream, size);
// Use manual timing
for(auto& b : benchmarks)
{
b->UseManualTime();
b->Unit(benchmark::kMillisecond);
}
// Force number of iterations
if(trials > 0)
{
for(auto& b : benchmarks)
{
b->Iterations(trials);
}
}
// Run benchmarks
benchmark::RunSpecifiedBenchmarks();
return 0;
}
|
