github-repos/benchmark
1
0
Fork 0
mirror of https://github.com/google/benchmark.git synced 2024-12-28 13:30:15 +08:00
Code Issues Packages Projects Releases Wiki Activity
9d47552020
benchmark/test/basic_test.cc

105 lines
2.7 KiB
C++
Raw Normal View History

Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
#include <cstddef>
fix minimal_benchmark.h include
2015-03-18 02:02:02 +08:00
#include "benchmark/benchmark_api.h"
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
#define BASIC_BENCHMARK_TEST(x) \
BENCHMARK(x)->Arg(8)->Arg(512)->Arg(8192)
void BM_empty(benchmark::State& state) {
while (state.KeepRunning()) {
Add 'benchmark::DoNotOptimize(...)' to help users prevent optimizations
2015-03-28 04:35:46 +08:00
benchmark::DoNotOptimize(state.iterations());
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
}
}
BENCHMARK(BM_empty);
BENCHMARK(BM_empty)->ThreadPerCpu();
void BM_spin_empty(benchmark::State& state) {
while (state.KeepRunning()) {
for (int x = 0; x < state.range_x(); ++x) {
Add 'benchmark::DoNotOptimize(...)' to help users prevent optimizations
2015-03-28 04:35:46 +08:00
benchmark::DoNotOptimize(x);
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
}
}
}
BASIC_BENCHMARK_TEST(BM_spin_empty);
BASIC_BENCHMARK_TEST(BM_spin_empty)->ThreadPerCpu();
void BM_spin_pause_before(benchmark::State& state) {
for (int i = 0; i < state.range_x(); ++i) {
Add 'benchmark::DoNotOptimize(...)' to help users prevent optimizations
2015-03-28 04:35:46 +08:00
benchmark::DoNotOptimize(i);
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
}
while(state.KeepRunning()) {
for (int i = 0; i < state.range_x(); ++i) {
Add 'benchmark::DoNotOptimize(...)' to help users prevent optimizations
2015-03-28 04:35:46 +08:00
benchmark::DoNotOptimize(i);
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
}
}
}
BASIC_BENCHMARK_TEST(BM_spin_pause_before);
BASIC_BENCHMARK_TEST(BM_spin_pause_before)->ThreadPerCpu();
void BM_spin_pause_during(benchmark::State& state) {
while(state.KeepRunning()) {
state.PauseTiming();
for (int i = 0; i < state.range_x(); ++i) {
Add 'benchmark::DoNotOptimize(...)' to help users prevent optimizations
2015-03-28 04:35:46 +08:00
benchmark::DoNotOptimize(i);
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
}
state.ResumeTiming();
for (int i = 0; i < state.range_x(); ++i) {
Add 'benchmark::DoNotOptimize(...)' to help users prevent optimizations
2015-03-28 04:35:46 +08:00
benchmark::DoNotOptimize(i);
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
}
}
}
BASIC_BENCHMARK_TEST(BM_spin_pause_during);
BASIC_BENCHMARK_TEST(BM_spin_pause_during)->ThreadPerCpu();
Introduce use of std::chrono and remove Walltime printing
2015-03-27 02:26:07 +08:00
void BM_pause_during(benchmark::State& state) {
while(state.KeepRunning()) {
state.PauseTiming();
state.ResumeTiming();
}
}
add walltime benchmark and fix unused variable.
2015-03-27 02:56:52 +08:00
BENCHMARK(BM_pause_during);
BENCHMARK(BM_pause_during)->ThreadPerCpu();
merge new changes in basic_test.cc
2015-03-28 04:37:53 +08:00
BENCHMARK(BM_pause_during)->UseRealTime();
BENCHMARK(BM_pause_during)->UseRealTime()->ThreadPerCpu();
Introduce use of std::chrono and remove Walltime printing
2015-03-27 02:26:07 +08:00
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
void BM_spin_pause_after(benchmark::State& state) {
while(state.KeepRunning()) {
for (int i = 0; i < state.range_x(); ++i) {
Add 'benchmark::DoNotOptimize(...)' to help users prevent optimizations
2015-03-28 04:35:46 +08:00
benchmark::DoNotOptimize(i);
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
}
}
for (int i = 0; i < state.range_x(); ++i) {
Add 'benchmark::DoNotOptimize(...)' to help users prevent optimizations
2015-03-28 04:35:46 +08:00
benchmark::DoNotOptimize(i);
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
}
}
BASIC_BENCHMARK_TEST(BM_spin_pause_after);
BASIC_BENCHMARK_TEST(BM_spin_pause_after)->ThreadPerCpu();
void BM_spin_pause_before_and_after(benchmark::State& state) {
for (int i = 0; i < state.range_x(); ++i) {
Add 'benchmark::DoNotOptimize(...)' to help users prevent optimizations
2015-03-28 04:35:46 +08:00
benchmark::DoNotOptimize(i);
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
}
while(state.KeepRunning()) {
for (int i = 0; i < state.range_x(); ++i) {
Add 'benchmark::DoNotOptimize(...)' to help users prevent optimizations
2015-03-28 04:35:46 +08:00
benchmark::DoNotOptimize(i);
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
}
}
for (int i = 0; i < state.range_x(); ++i) {
Add 'benchmark::DoNotOptimize(...)' to help users prevent optimizations
2015-03-28 04:35:46 +08:00
benchmark::DoNotOptimize(i);
Adopt new benchmark timing internals. This patch adopts a new internal structure for how timings are performed. Currently every iteration of a benchmark checks to see if it has been running for an appropriate amount of time. Checking the clock introduces noise into the timings and this can cause inconsistent output from each benchmark. Now every iteration of a benchmark only checks an iteration count to see if it should stop running. The iteration count is determined before hand by testing the benchmark on a series of increasing iteration counts until a suitable count is found. This increases the amount of time it takes to run the actual benchmarks but it also greatly increases the accuracy of the results. This patch introduces some breaking changes. The notable breaking changes are: 1. Benchmarks run on multiple threads no generate a report per thread. Instead only a single report is generated. 2. ::benchmark::UseRealTime() was removed and replaced with State::UseRealTime().
2015-03-13 06:03:33 +08:00
}
}
BASIC_BENCHMARK_TEST(BM_spin_pause_before_and_after);
BASIC_BENCHMARK_TEST(BM_spin_pause_before_and_after)->ThreadPerCpu();
void BM_empty_stop_start(benchmark::State& state) {
while (state.KeepRunning()) { }
}
BENCHMARK(BM_empty_stop_start);
BENCHMARK(BM_empty_stop_start)->ThreadPerCpu();
BENCHMARK_MAIN()
Reference in New Issue Copy Permalink