| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
| |
commit_hash:444da85d16474126d9f2373887f85e4601421460
|
| |
|
|
| |
commit_hash:c0f0e5fa6a2d9bf233fe4af617eb17b323379948
|
| |
|
|
| |
commit_hash:35957e20f0bc2213a543cff063d4b81f79621304
|
| |
|
|
| |
commit_hash:cc9c4f3b15352d534dd69064a2ba4bbecb7bc117
|
| |
|
|
|
|
| |
`TDuration::Seconds(sec)` now returns `TDuration::Max()` if `sec` is greater than the largest representable value.
Previously, a silent wrap-around could produce a significantly shorter interval (even a zero one).
commit_hash:552fde08fa3b8201ede23894bb363df2d6bbd5af
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
```
[pg@arc ~]$ ./util-ut
<----- TSaveLoadTest
[exec] TSaveLoadTest::TestSaveLoad...
[good] TSaveLoadTest::TestSaveLoad
[exec] TSaveLoadTest::TestSaveLoadEmptyStruct...
[good] TSaveLoadTest::TestSaveLoadEmptyStruct
[exec] TSaveLoadTest::TestNewStyle...
[good] TSaveLoadTest::TestNewStyle
[exec] TSaveLoadTest::TestNewNewStyle...
[good] TSaveLoadTest::TestNewNewStyle
[exec] TSaveLoadTest::TestList...
[good] TSaveLoadTest::TestList
[exec] TSaveLoadTest::TestTuple...
[good] TSaveLoadTest::TestTuple
[exec] TSaveLoadTest::TestVariant...
[good] TSaveLoadTest::TestVariant
[exec] TSaveLoadTest::TestOptional...
[good] TSaveLoadTest::TestOptional
[exec] TSaveLoadTest::TestInheritNonVirtualClass...
[good] TSaveLoadTest::TestInheritNonVirtualClass
[exec] TSaveLoadTest::TestInheritVirtualClass...
[good] TSaveLoadTest::TestInheritVirtualClass
-----> TSaveLoadTest -> ok: 10
[DONE] ok: 10
[pg@arc ~]$ uname -a
FreeBSD arc.freebsd.devtools 14.2-RELEASE-p1 FreeBSD 14.2-RELEASE-p1 GENERIC amd64
[pg@arc ~]$
```
commit_hash:229d8aa6c8ccce1376d152a833fe8bf7c5faa3bd
|
| |
|
|
|
| |
Reading /proc/uptime is about a hundred times slower than accessing the same timer via the clock_gettime call. The latter also has better precision.
commit_hash:70904860adc43c9d2d6487a415a2e6b27fe6dae7
|
| |
|
|
| |
commit_hash:b43c96b868cd36d636192fd2c6024d9f0d2fb6f8
|
| |
|
|
| |
commit_hash:9ec33d0cbe2da6aeea6e8d02fdaed5f45b6e3534
|
| |
|
|
| |
commit_hash:ec09e953e6ed7b89ab18dfdac86961de36b5b2ea
|
| |
|
|
| |
commit_hash:d57be220df393c193619ef5ed129ec4436540629
|
| |
|
|
| |
commit_hash:f3acde65f4eb25503f414d5262f52b496123bd96
|
| |
|
|
|
| |
Replacing binary search with a look-up table. The performance of which does not depend on the branch predictor, and conversion works equally fast for any dates.
commit_hash:fefe9665d0d4b51c2ae09ec2b2816aed30caa57b
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
perf before
```
-------------------------------------------------------------------------------
Benchmark Time CPU Iterations
-------------------------------------------------------------------------------
BM_GmTimeR 3.61 ns 3.61 ns 194952699
BM_GmTimeRRandom/last_hour 5.04 ns 5.04 ns 136121940
BM_GmTimeRRandom/last_day 5.03 ns 5.03 ns 139883239
BM_GmTimeRRandom/last_month 8.03 ns 8.03 ns 87839823
BM_GmTimeRRandom/last_year 14.9 ns 14.9 ns 46401773
BM_GmTimeRRandom/last_decade 16.9 ns 16.9 ns 41439312
BM_GmTimeRRandom/last_half_centry 16.9 ns 16.9 ns 41225553
```
perf after
```
-------------------------------------------------------------------------------
Benchmark Time CPU Iterations
-------------------------------------------------------------------------------
BM_GmTimeR 3.27 ns 3.27 ns 215390890
BM_GmTimeRRandom/last_hour 4.52 ns 4.52 ns 154869504
BM_GmTimeRRandom/last_day 4.53 ns 4.53 ns 155409964
BM_GmTimeRRandom/last_month 7.24 ns 7.24 ns 96972146
BM_GmTimeRRandom/last_year 13.5 ns 13.5 ns 51957829
BM_GmTimeRRandom/last_decade 15.5 ns 15.5 ns 45192135
BM_GmTimeRRandom/last_half_centry 15.4 ns 15.4 ns 45526906
```
commit_hash:ac9db56d93a25c3a3018cfbb6bc5800da3d001ee
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
performance comparison againt standard gmtime\_r
```
-------------------------------------------------------------------------------
Benchmark Time CPU Iterations
-------------------------------------------------------------------------------
BM_GmTimeR 3.66 ns 3.66 ns 189855900
BM_gmtime_r 18.7 ns 18.7 ns 37340999
BM_GmTimeRRandom/last_hour 5.12 ns 5.12 ns 136126173
BM_gmtime_r_Random/last_hour 19.2 ns 19.2 ns 36556971
BM_GmTimeRRandom/last_day 5.10 ns 5.10 ns 138257644
BM_gmtime_r_Random/last_day 19.2 ns 19.2 ns 37025160
BM_GmTimeRRandom/last_month 7.10 ns 7.10 ns 98235494
BM_gmtime_r_Random/last_month 20.4 ns 20.4 ns 34637975
BM_GmTimeRRandom/last_year 15.0 ns 15.0 ns 46664304
BM_gmtime_r_Random/last_year 26.8 ns 26.8 ns 26135911
BM_GmTimeRRandom/last_decade 17.2 ns 17.2 ns 40105536
BM_gmtime_r_Random/last_decade 27.0 ns 27.0 ns 25817056
BM_GmTimeRRandom/last_half_centry 17.1 ns 17.0 ns 41066079
BM_gmtime_r_Random/last_half_century 26.8 ns 26.7 ns 26215076
```
commit_hash:33747760ad46412f50fe3cb06f4c438794a44492
|
| |
|
|
| |
commit_hash:21bc37ea2a290cb645c95fe2c94882484c7001a4
|
| |
|
|
|
| |
Часть большого ПР REVIEW:7264088
commit_hash:0f5b03fbbed0ac30f734943309e3ef5cd4d7a30e
|
| |
|
|
|
| |
FixNamespaceComments in /util
bb9152570d7c258798644ead5a59f604de05ef3b
|
| |
|
|
|
| |
Remove Y_UNIT_TEST from config.clang-format
853b6da2c4464342d4cfe7756c18ed599004fdd8
|
| |
|
|
|
| |
Set SpacesInLineCommentPrefix to 1 in /util
3853f9ec5143722c1bebd8dc0ffc9b61a6c17657
|
| |
|
|
| |
f172ae24ead4a8c8edae837d34f264a744a4d89c
|
| |
|
|
|
| |
The ProcessUptime function should now usually return a value less than 20 ms if called at the start of the program.
0961a518fdaee4fd243b01c7cf5433dd768897c4
|
| |
|
|
|
| |
Добавляем uptime для процесса
7dc898ce5f81666a4cd5f912f49a3035dc2c599c
|
| |
|
|
| |
3523ab3f5aade2bdf4c0efd5dd2defbe19f124ff
|
| | |
|
| |
|
|
| |
4475987f399172bada73db028180b5e773e58a30
|
| | |
|
| |
|
|
| |
b46c1e446c9d0f9393bf688b26db810f60add0f6
|
| | |
|
| |
|
|
| |
96265cd0cc64e1b9bb31fe97b915ed2a09caf1cb
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
в профиле перфа Маркетного репорта постоянно вижу GmTimeR
там в цикле вычитается по количеству дней в году и прибавляется по 1 году с 1970 по 2022 на каждый вызов
В Маркете проблема стреляла и раньше
в последней итерации функция становится на ~~30% быстрее системной gmtime_r
%%
----------- GmTimeR ---------------
samples: 20691
iterations: 264304536
iterations hr: 264M
run time: 5.002055225
per iteration: 39.85866484 cycles
----------- gmtime_r ---------------
samples: 17452
iterations: 188034528
iterations hr: 188M
run time: 5.001259884
per iteration: 56.96146577 cycles
%%
текущая верся почти в 2 раза медленнее системной:
%%
----------- GmTimeR ---------------
samples: 12760
iterations: 100514931
iterations hr: 101M
run time: 5.00096133
per iteration: 105.4334174 cycles
----------- gmtime_r ---------------
samples: 17667
iterations: 192697896
iterations hr: 193M
run time: 5.001356603
per iteration: 55.69031415 cycles
%%
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
ref:f8c119cf3a796a59ab5be4424e5364ba6a201d40
|
| |
|
|
| |
ref:ca7a95e8c9a9d780f96497136a152091d54e61b5
|
| | |
|
| | |
|
| | |
|