| BUILD.gn |
|
2451 |
| DEPS |
|
177 |
| mutex.h |
|
2153 |
| mutex_abseil.h |
|
1266 |
| mutex_benchmark.cc |
Results:
NB when reproducing: Remember to turn of power management features such as CPU
scaling before running!
pthreads (Linux):
----------------------------------------------------------------------
Run on (12 X 4500 MHz CPU s)
CPU Caches:
L1 Data 32 KiB (x6)
L1 Instruction 32 KiB (x6)
L2 Unified 1024 KiB (x6)
L3 Unified 8448 KiB (x1)
Load Average: 0.26, 0.28, 0.44
----------------------------------------------------------------------
Benchmark Time CPU Iterations
----------------------------------------------------------------------
BM_LockWithMutex/threads:1 13.4 ns 13.4 ns 52192906
BM_LockWithMutex/threads:2 44.2 ns 88.4 ns 8189944
BM_LockWithMutex/threads:4 52.0 ns 198 ns 3743244
BM_LockWithMutex/threads:12 84.9 ns 944 ns 733524
std::mutex performs like the pthread implementation (Linux).
Abseil (Linux):
----------------------------------------------------------------------
Run on (12 X 4500 MHz CPU s)
CPU Caches:
L1 Data 32 KiB (x6)
L1 Instruction 32 KiB (x6)
L2 Unified 1024 KiB (x6)
L3 Unified 8448 KiB (x1)
Load Average: 0.27, 0.24, 0.37
----------------------------------------------------------------------
Benchmark Time CPU Iterations
----------------------------------------------------------------------
BM_LockWithMutex/threads:1 15.0 ns 15.0 ns 46550231
BM_LockWithMutex/threads:2 91.1 ns 182 ns 4059212
BM_LockWithMutex/threads:4 40.8 ns 131 ns 5496560
BM_LockWithMutex/threads:12 37.0 ns 130 ns 5377668
|
3141 |
| mutex_critical_section.h |
|
1835 |
| mutex_gn |
|
|
| mutex_pthread.h |
|
3258 |
| mutex_unittest.cc |
|
4693 |
| sequence_checker_internal.cc |
|
3110 |
| sequence_checker_internal.h |
attach_to_current_thread |
3340 |
| sequence_checker_internal_gn |
|
|
| yield.cc |
|
975 |
| yield.h |
|
647 |
| yield_gn |
|
|
| yield_policy.cc |
|
2195 |
| yield_policy.h |
|
1286 |
| yield_policy_gn |
|
|
| yield_policy_unittest.cc |
|
2128 |