dolysis
/
benchmark
mirror of https://github.com/google/benchmark.git
3
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

[Tools] A new, more versatile benchmark output compare tool (#474) 

* [Tools] A new, more versatile benchmark output compare tool

Sometimes, there is more than one implementation of some functionality.
And the obvious use-case is to benchmark them, which is better?

Currently, there is no easy way to compare the benchmarking results
in that case:
    The obvious solution is to have multiple binaries, each one
containing/running one implementation. And each binary must use
exactly the same benchmark family name, which is super bad,
because now the binary name should contain all the info about
benchmark family...

What if i tell you that is not the solution?
What if we could avoid producing one binary per benchmark family,
with the same family name used in each binary,
but instead could keep all the related families in one binary,
with their proper names, AND still be able to compare them?

There are three modes of operation:
1. Just compare two benchmarks, what `compare_bench.py` did:
```
$ ../tools/compare.py benchmarks ./a.out ./a.out
RUNNING: ./a.out --benchmark_out=/tmp/tmprBT5nW
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:16:44
------------------------------------------------------
Benchmark               Time           CPU Iterations
------------------------------------------------------
BM_memcpy/8            36 ns         36 ns   19101577   211.669MB/s
BM_memcpy/64           76 ns         76 ns    9412571   800.199MB/s
BM_memcpy/512          84 ns         84 ns    8249070   5.64771GB/s
BM_memcpy/1024        116 ns        116 ns    6181763   8.19505GB/s
BM_memcpy/8192        643 ns        643 ns    1062855   11.8636GB/s
BM_copy/8             222 ns        222 ns    3137987   34.3772MB/s
BM_copy/64           1608 ns       1608 ns     432758   37.9501MB/s
BM_copy/512         12589 ns      12589 ns      54806   38.7867MB/s
BM_copy/1024        25169 ns      25169 ns      27713   38.8003MB/s
BM_copy/8192       201165 ns     201112 ns       3486   38.8466MB/s
RUNNING: ./a.out --benchmark_out=/tmp/tmpt1wwG_
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:16:53
------------------------------------------------------
Benchmark               Time           CPU Iterations
------------------------------------------------------
BM_memcpy/8            36 ns         36 ns   19397903   211.255MB/s
BM_memcpy/64           73 ns         73 ns    9691174   839.635MB/s
BM_memcpy/512          85 ns         85 ns    8312329   5.60101GB/s
BM_memcpy/1024        118 ns        118 ns    6438774   8.11608GB/s
BM_memcpy/8192        656 ns        656 ns    1068644   11.6277GB/s
BM_copy/8             223 ns        223 ns    3146977   34.2338MB/s
BM_copy/64           1611 ns       1611 ns     435340   37.8751MB/s
BM_copy/512         12622 ns      12622 ns      54818   38.6844MB/s
BM_copy/1024        25257 ns      25239 ns      27779   38.6927MB/s
BM_copy/8192       205013 ns     205010 ns       3479    38.108MB/s
Comparing ./a.out to ./a.out
Benchmark                 Time             CPU      Time Old      Time New       CPU Old       CPU New
------------------------------------------------------------------------------------------------------
BM_memcpy/8            +0.0020         +0.0020            36            36            36            36
BM_memcpy/64           -0.0468         -0.0470            76            73            76            73
BM_memcpy/512          +0.0081         +0.0083            84            85            84            85
BM_memcpy/1024         +0.0098         +0.0097           116           118           116           118
BM_memcpy/8192         +0.0200         +0.0203           643           656           643           656
BM_copy/8              +0.0046         +0.0042           222           223           222           223
BM_copy/64             +0.0020         +0.0020          1608          1611          1608          1611
BM_copy/512            +0.0027         +0.0026         12589         12622         12589         12622
BM_copy/1024           +0.0035         +0.0028         25169         25257         25169         25239
BM_copy/8192           +0.0191         +0.0194        201165        205013        201112        205010
```

2. Compare two different filters of one benchmark:
(for simplicity, the benchmark is executed twice)
```
$ ../tools/compare.py filters ./a.out BM_memcpy BM_copy
RUNNING: ./a.out --benchmark_filter=BM_memcpy --benchmark_out=/tmp/tmpBWKk0k
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:37:28
------------------------------------------------------
Benchmark               Time           CPU Iterations
------------------------------------------------------
BM_memcpy/8            36 ns         36 ns   17891491   211.215MB/s
BM_memcpy/64           74 ns         74 ns    9400999   825.646MB/s
BM_memcpy/512          87 ns         87 ns    8027453   5.46126GB/s
BM_memcpy/1024        111 ns        111 ns    6116853    8.5648GB/s
BM_memcpy/8192        657 ns        656 ns    1064679   11.6247GB/s
RUNNING: ./a.out --benchmark_filter=BM_copy --benchmark_out=/tmp/tmpAvWcOM
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:37:33
----------------------------------------------------
Benchmark             Time           CPU Iterations
----------------------------------------------------
BM_copy/8           227 ns        227 ns    3038700   33.6264MB/s
BM_copy/64         1640 ns       1640 ns     426893   37.2154MB/s
BM_copy/512       12804 ns      12801 ns      55417   38.1444MB/s
BM_copy/1024      25409 ns      25407 ns      27516   38.4365MB/s
BM_copy/8192     202986 ns     202990 ns       3454   38.4871MB/s
Comparing BM_memcpy to BM_copy (from ./a.out)
Benchmark                               Time             CPU      Time Old      Time New       CPU Old       CPU New
--------------------------------------------------------------------------------------------------------------------
[BM_memcpy vs. BM_copy]/8            +5.2829         +5.2812            36           227            36           227
[BM_memcpy vs. BM_copy]/64          +21.1719        +21.1856            74          1640            74          1640
[BM_memcpy vs. BM_copy]/512        +145.6487       +145.6097            87         12804            87         12801
[BM_memcpy vs. BM_copy]/1024       +227.1860       +227.1776           111         25409           111         25407
[BM_memcpy vs. BM_copy]/8192       +308.1664       +308.2898           657        202986           656        202990
```

3. Compare filter one from benchmark one to filter two from benchmark two:
(for simplicity, the benchmark is executed twice)
```
$ ../tools/compare.py benchmarksfiltered ./a.out BM_memcpy ./a.out BM_copy
RUNNING: ./a.out --benchmark_filter=BM_memcpy --benchmark_out=/tmp/tmp_FvbYg
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:38:27
------------------------------------------------------
Benchmark               Time           CPU Iterations
------------------------------------------------------
BM_memcpy/8            37 ns         37 ns   18953482   204.118MB/s
BM_memcpy/64           74 ns         74 ns    9206578   828.245MB/s
BM_memcpy/512          91 ns         91 ns    8086195   5.25476GB/s
BM_memcpy/1024        120 ns        120 ns    5804513   7.95662GB/s
BM_memcpy/8192        664 ns        664 ns    1028363   11.4948GB/s
RUNNING: ./a.out --benchmark_filter=BM_copy --benchmark_out=/tmp/tmpDfL5iE
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:38:32
----------------------------------------------------
Benchmark             Time           CPU Iterations
----------------------------------------------------
BM_copy/8           230 ns        230 ns    2985909   33.1161MB/s
BM_copy/64         1654 ns       1653 ns     419408   36.9137MB/s
BM_copy/512       13122 ns      13120 ns      53403   37.2156MB/s
BM_copy/1024      26679 ns      26666 ns      26575   36.6218MB/s
BM_copy/8192     215068 ns     215053 ns       3221   36.3283MB/s
Comparing BM_memcpy (from ./a.out) to BM_copy (from ./a.out)
Benchmark                               Time             CPU      Time Old      Time New       CPU Old       CPU New
--------------------------------------------------------------------------------------------------------------------
[BM_memcpy vs. BM_copy]/8            +5.1649         +5.1637            37           230            37           230
[BM_memcpy vs. BM_copy]/64          +21.4352        +21.4374            74          1654            74          1653
[BM_memcpy vs. BM_copy]/512        +143.6022       +143.5865            91         13122            91         13120
[BM_memcpy vs. BM_copy]/1024       +221.5903       +221.4790           120         26679           120         26666
[BM_memcpy vs. BM_copy]/8192       +322.9059       +323.0096           664        215068           664        215053
```

* [Docs] Document tools/compare.py

* [docs] Document how the change is calculated
This commit is contained in:
Roman Lebedev 2017-11-08 00:35:25 +03:00 committed by Dominic Hamon
parent 90aa8665b5
commit 5e66248b44
4 changed files with 661 additions and 32 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

243
docs/tools.md
View File

@ -11,49 +11,232 @@ $ compare_bench.py <old-benchmark> <new-benchmark> [benchmark options]...
Where `<old-benchmark>` and `<new-benchmark>` either specify a benchmark executable file, or a JSON output file. The type of the input file is automatically detected. If a benchmark executable is specified then the benchmark is run to obtain the results. Otherwise the results are simply loaded from the output file.
`[benchmark options]` will be passed to the benchmarks invocations. They can be anything that binary accepts, be it either normal `--benchmark_*` parameters, or some custom parameters your binary takes.
The sample output using the JSON test files under `Inputs/` gives:
``` bash
$ ./compare_bench.py ./gbench/Inputs/test1_run1.json ./gbench/Inputs/test1_run2.json
Comparing ./gbench/Inputs/test1_run1.json to ./gbench/Inputs/test1_run2.json
Benchmark Time CPU
----------------------------------------------
BM_SameTimes +0.00 +0.00
BM_2xFaster -0.50 -0.50
BM_2xSlower +1.00 +1.00
BM_10PercentFaster -0.10 -0.10
BM_10PercentSlower +0.10 +0.10
Benchmark Time CPU Time Old Time New CPU Old CPU New
-------------------------------------------------------------------------------------------------------------
BM_SameTimes +0.0000 +0.0000 10 10 10 10
BM_2xFaster -0.5000 -0.5000 50 25 50 25
BM_2xSlower +1.0000 +1.0000 50 100 50 100
BM_1PercentFaster -0.0100 -0.0100 100 99 100 99
BM_1PercentSlower +0.0100 +0.0100 100 101 100 101
BM_10PercentFaster -0.1000 -0.1000 100 90 100 90
BM_10PercentSlower +0.1000 +0.1000 100 110 100 110
BM_100xSlower +99.0000 +99.0000 100 10000 100 10000
BM_100xFaster -0.9900 -0.9900 10000 100 10000 100
BM_10PercentCPUToTime +0.1000 -0.1000 100 110 100 90
BM_ThirdFaster -0.3333 -0.3334 100 67 100 67
BM_BadTimeUnit -0.9000 +0.2000 0 0 0 1
```
As you can note, the values in `Time` and `CPU` columns are calculated as `(new - old) / |old|`.
When a benchmark executable is run, the raw output from the benchmark is printed in real time to stdout. The sample output using `benchmark/basic_test` for both arguments looks like:
```
./compare_bench.py test/basic_test test/basic_test --benchmark_filter=BM_empty.*
RUNNING: test/basic_test --benchmark_filter=BM_empty.*
Run on (4 X 4228.32 MHz CPU s)
2016-08-02 19:21:33
RUNNING: test/basic_test --benchmark_filter=BM_empty.* --benchmark_out=/tmp/tmpN7LF3a
Run on (8 X 4000 MHz CPU s)
2017-11-07 23:28:36
---------------------------------------------------------------------
Benchmark Time CPU Iterations
--------------------------------------------------------------------
BM_empty 9 ns 9 ns 79545455
BM_empty/threads:4 4 ns 9 ns 75268816
BM_empty_stop_start 8 ns 8 ns 83333333
BM_empty_stop_start/threads:4 3 ns 8 ns 83333332
RUNNING: test/basic_test --benchmark_filter=BM_empty.*
Run on (4 X 4228.32 MHz CPU s)
2016-08-02 19:21:35
---------------------------------------------------------------------
BM_empty 4 ns 4 ns 170178757
BM_empty/threads:8 1 ns 7 ns 103868920
BM_empty_stop_start 0 ns 0 ns 1000000000
BM_empty_stop_start/threads:8 0 ns 0 ns 1403031720
RUNNING: /test/basic_test --benchmark_filter=BM_empty.* --benchmark_out=/tmp/tmplvrIp8
Run on (8 X 4000 MHz CPU s)
2017-11-07 23:28:38
---------------------------------------------------------------------
Benchmark Time CPU Iterations
--------------------------------------------------------------------
BM_empty 9 ns 9 ns 76086957
BM_empty/threads:4 4 ns 9 ns 76086956
BM_empty_stop_start 8 ns 8 ns 87500000
BM_empty_stop_start/threads:4 3 ns 8 ns 88607596
Comparing test/basic_test to test/basic_test
Benchmark Time CPU
---------------------------------------------------------
BM_empty +0.00 +0.00
BM_empty/threads:4 +0.00 +0.00
BM_empty_stop_start +0.00 +0.00
BM_empty_stop_start/threads:4 +0.00 +0.00
---------------------------------------------------------------------
BM_empty 4 ns 4 ns 169534855
BM_empty/threads:8 1 ns 7 ns 104188776
BM_empty_stop_start 0 ns 0 ns 1000000000
BM_empty_stop_start/threads:8 0 ns 0 ns 1404159424
Comparing ../build/test/basic_test to ../build/test/basic_test
Benchmark Time CPU Time Old Time New CPU Old CPU New
---------------------------------------------------------------------------------------------------------------------
BM_empty -0.0048 -0.0049 4 4 4 4
BM_empty/threads:8 -0.0123 -0.0054 1 1 7 7
BM_empty_stop_start -0.0000 -0.0000 0 0 0 0
BM_empty_stop_start/threads:8 -0.0029 +0.0001 0 0 0 0
```
As you can note, the values in `Time` and `CPU` columns are calculated as `(new - old) / |old|`.
Obviously this example doesn't give any useful output, but it's intended to show the output format when 'compare_bench.py' needs to run benchmarks.
## compare.py
The `compare.py` can be used to compare the result of benchmarks.
There are three modes of operation:
1. Just compare two benchmarks, what `compare_bench.py` did.
The program is invoked like:
``` bash
$ compare.py benchmarks <benchmark_baseline> <benchmark_contender> [benchmark options]...
```
Where `<benchmark_baseline>` and `<benchmark_contender>` either specify a benchmark executable file, or a JSON output file. The type of the input file is automatically detected. If a benchmark executable is specified then the benchmark is run to obtain the results. Otherwise the results are simply loaded from the output file.
`[benchmark options]` will be passed to the benchmarks invocations. They can be anything that binary accepts, be it either normal `--benchmark_*` parameters, or some custom parameters your binary takes.
Example output:
```
$ ./compare.py benchmarks ./a.out ./a.out
RUNNING: ./a.out --benchmark_out=/tmp/tmprBT5nW
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:16:44
------------------------------------------------------
Benchmark Time CPU Iterations
------------------------------------------------------
BM_memcpy/8 36 ns 36 ns 19101577 211.669MB/s
BM_memcpy/64 76 ns 76 ns 9412571 800.199MB/s
BM_memcpy/512 84 ns 84 ns 8249070 5.64771GB/s
BM_memcpy/1024 116 ns 116 ns 6181763 8.19505GB/s
BM_memcpy/8192 643 ns 643 ns 1062855 11.8636GB/s
BM_copy/8 222 ns 222 ns 3137987 34.3772MB/s
BM_copy/64 1608 ns 1608 ns 432758 37.9501MB/s
BM_copy/512 12589 ns 12589 ns 54806 38.7867MB/s
BM_copy/1024 25169 ns 25169 ns 27713 38.8003MB/s
BM_copy/8192 201165 ns 201112 ns 3486 38.8466MB/s
RUNNING: ./a.out --benchmark_out=/tmp/tmpt1wwG_
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:16:53
------------------------------------------------------
Benchmark Time CPU Iterations
------------------------------------------------------
BM_memcpy/8 36 ns 36 ns 19397903 211.255MB/s
BM_memcpy/64 73 ns 73 ns 9691174 839.635MB/s
BM_memcpy/512 85 ns 85 ns 8312329 5.60101GB/s
BM_memcpy/1024 118 ns 118 ns 6438774 8.11608GB/s
BM_memcpy/8192 656 ns 656 ns 1068644 11.6277GB/s
BM_copy/8 223 ns 223 ns 3146977 34.2338MB/s
BM_copy/64 1611 ns 1611 ns 435340 37.8751MB/s
BM_copy/512 12622 ns 12622 ns 54818 38.6844MB/s
BM_copy/1024 25257 ns 25239 ns 27779 38.6927MB/s
BM_copy/8192 205013 ns 205010 ns 3479 38.108MB/s
Comparing ./a.out to ./a.out
Benchmark Time CPU Time Old Time New CPU Old CPU New
------------------------------------------------------------------------------------------------------
BM_memcpy/8 +0.0020 +0.0020 36 36 36 36
BM_memcpy/64 -0.0468 -0.0470 76 73 76 73
BM_memcpy/512 +0.0081 +0.0083 84 85 84 85
BM_memcpy/1024 +0.0098 +0.0097 116 118 116 118
BM_memcpy/8192 +0.0200 +0.0203 643 656 643 656
BM_copy/8 +0.0046 +0.0042 222 223 222 223
BM_copy/64 +0.0020 +0.0020 1608 1611 1608 1611
BM_copy/512 +0.0027 +0.0026 12589 12622 12589 12622
BM_copy/1024 +0.0035 +0.0028 25169 25257 25169 25239
BM_copy/8192 +0.0191 +0.0194 201165 205013 201112 205010
```
What it does is for the every benchmark from the first run it looks for the benchmark with exactly the same name in the second run, and then compares the results. If the names differ, the benchmark is omitted from the diff.
As you can note, the values in `Time` and `CPU` columns are calculated as `(new - old) / |old|`.
2. Compare two different filters of one benchmark
The program is invoked like:
``` bash
$ compare.py filters <benchmark> <filter_baseline> <filter_contender> [benchmark options]...
```
Where `<benchmark>` either specify a benchmark executable file, or a JSON output file. The type of the input file is automatically detected. If a benchmark executable is specified then the benchmark is run to obtain the results. Otherwise the results are simply loaded from the output file.
Where `<filter_baseline>` and `<filter_contender>` are the same regex filters that you would pass to the `[--benchmark_filter=<regex>]` parameter of the benchmark binary.
`[benchmark options]` will be passed to the benchmarks invocations. They can be anything that binary accepts, be it either normal `--benchmark_*` parameters, or some custom parameters your binary takes.
Example output:
```
$ ./compare.py filters ./a.out BM_memcpy BM_copy
RUNNING: ./a.out --benchmark_filter=BM_memcpy --benchmark_out=/tmp/tmpBWKk0k
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:37:28
------------------------------------------------------
Benchmark Time CPU Iterations
------------------------------------------------------
BM_memcpy/8 36 ns 36 ns 17891491 211.215MB/s
BM_memcpy/64 74 ns 74 ns 9400999 825.646MB/s
BM_memcpy/512 87 ns 87 ns 8027453 5.46126GB/s
BM_memcpy/1024 111 ns 111 ns 6116853 8.5648GB/s
BM_memcpy/8192 657 ns 656 ns 1064679 11.6247GB/s
RUNNING: ./a.out --benchmark_filter=BM_copy --benchmark_out=/tmp/tmpAvWcOM
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:37:33
----------------------------------------------------
Benchmark Time CPU Iterations
----------------------------------------------------
BM_copy/8 227 ns 227 ns 3038700 33.6264MB/s
BM_copy/64 1640 ns 1640 ns 426893 37.2154MB/s
BM_copy/512 12804 ns 12801 ns 55417 38.1444MB/s
BM_copy/1024 25409 ns 25407 ns 27516 38.4365MB/s
BM_copy/8192 202986 ns 202990 ns 3454 38.4871MB/s
Comparing BM_memcpy to BM_copy (from ./a.out)
Benchmark Time CPU Time Old Time New CPU Old CPU New
--------------------------------------------------------------------------------------------------------------------
[BM_memcpy vs. BM_copy]/8 +5.2829 +5.2812 36 227 36 227
[BM_memcpy vs. BM_copy]/64 +21.1719 +21.1856 74 1640 74 1640
[BM_memcpy vs. BM_copy]/512 +145.6487 +145.6097 87 12804 87 12801
[BM_memcpy vs. BM_copy]/1024 +227.1860 +227.1776 111 25409 111 25407
[BM_memcpy vs. BM_copy]/8192 +308.1664 +308.2898 657 202986 656 202990
```
As you can see, it applies filter to the benchmarks, both when running the benchmark, and before doing the diff. And to make the diff work, the matches are replaced with some common string. Thus, you can compare two different benchmark families within one benchmark binary.
As you can note, the values in `Time` and `CPU` columns are calculated as `(new - old) / |old|`.
3. Compare filter one from benchmark one to filter two from benchmark two:
The program is invoked like:
``` bash
$ compare.py filters <benchmark_baseline> <filter_baseline> <benchmark_contender> <filter_contender> [benchmark options]...
```
Where `<benchmark_baseline>` and `<benchmark_contender>` either specify a benchmark executable file, or a JSON output file. The type of the input file is automatically detected. If a benchmark executable is specified then the benchmark is run to obtain the results. Otherwise the results are simply loaded from the output file.
Where `<filter_baseline>` and `<filter_contender>` are the same regex filters that you would pass to the `[--benchmark_filter=<regex>]` parameter of the benchmark binary.
`[benchmark options]` will be passed to the benchmarks invocations. They can be anything that binary accepts, be it either normal `--benchmark_*` parameters, or some custom parameters your binary takes.
Example output:
```
$ ./compare.py benchmarksfiltered ./a.out BM_memcpy ./a.out BM_copy
RUNNING: ./a.out --benchmark_filter=BM_memcpy --benchmark_out=/tmp/tmp_FvbYg
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:38:27
------------------------------------------------------
Benchmark Time CPU Iterations
------------------------------------------------------
BM_memcpy/8 37 ns 37 ns 18953482 204.118MB/s
BM_memcpy/64 74 ns 74 ns 9206578 828.245MB/s
BM_memcpy/512 91 ns 91 ns 8086195 5.25476GB/s
BM_memcpy/1024 120 ns 120 ns 5804513 7.95662GB/s
BM_memcpy/8192 664 ns 664 ns 1028363 11.4948GB/s
RUNNING: ./a.out --benchmark_filter=BM_copy --benchmark_out=/tmp/tmpDfL5iE
Run on (8 X 4000 MHz CPU s)
2017-11-07 21:38:32
----------------------------------------------------
Benchmark Time CPU Iterations
----------------------------------------------------
BM_copy/8 230 ns 230 ns 2985909 33.1161MB/s
BM_copy/64 1654 ns 1653 ns 419408 36.9137MB/s
BM_copy/512 13122 ns 13120 ns 53403 37.2156MB/s
BM_copy/1024 26679 ns 26666 ns 26575 36.6218MB/s
BM_copy/8192 215068 ns 215053 ns 3221 36.3283MB/s
Comparing BM_memcpy (from ./a.out) to BM_copy (from ./a.out)
Benchmark Time CPU Time Old Time New CPU Old CPU New
--------------------------------------------------------------------------------------------------------------------
[BM_memcpy vs. BM_copy]/8 +5.1649 +5.1637 37 230 37 230
[BM_memcpy vs. BM_copy]/64 +21.4352 +21.4374 74 1654 74 1653
[BM_memcpy vs. BM_copy]/512 +143.6022 +143.5865 91 13122 91 13120
[BM_memcpy vs. BM_copy]/1024 +221.5903 +221.4790 120 26679 120 26666
[BM_memcpy vs. BM_copy]/8192 +322.9059 +323.0096 664 215068 664 215053
```
This is a mix of the previous two modes, two (potentially different) benchmark binaries are run, and a different filter is applied to each one.
As you can note, the values in `Time` and `CPU` columns are calculated as `(new - old) / |old|`.

312
tools/compare.py Executable file
View File

@ -0,0 +1,312 @@
#!/usr/bin/env python
"""
compare.py - versatile benchmark output compare tool
"""
import argparse
from argparse import ArgumentParser
import sys
import gbench
from gbench import util, report
from gbench.util import *
def check_inputs(in1, in2, flags):
"""
Perform checking on the user provided inputs and diagnose any abnormalities
"""
in1_kind, in1_err = classify_input_file(in1)
in2_kind, in2_err = classify_input_file(in2)
output_file = find_benchmark_flag('--benchmark_out=', flags)
output_type = find_benchmark_flag('--benchmark_out_format=', flags)
if in1_kind == IT_Executable and in2_kind == IT_Executable and output_file:
print(("WARNING: '--benchmark_out=%s' will be passed to both "
"benchmarks causing it to be overwritten") % output_file)
if in1_kind == IT_JSON and in2_kind == IT_JSON and len(flags) > 0:
print("WARNING: passing optional flags has no effect since both "
"inputs are JSON")
if output_type is not None and output_type != 'json':
print(("ERROR: passing '--benchmark_out_format=%s' to 'compare.py`"
" is not supported.") % output_type)
sys.exit(1)
def create_parser():
parser = ArgumentParser(
description='versatile benchmark output compare tool')
subparsers = parser.add_subparsers(
help='This tool has multiple modes of operation:',
dest='mode')
parser_a = subparsers.add_parser(
'benchmarks',
help='The most simple use-case, compare all the output of these two benchmarks')
baseline = parser_a.add_argument_group(
'baseline', 'The benchmark baseline')
baseline.add_argument(
'test_baseline',
metavar='test_baseline',
type=argparse.FileType('r'),
nargs=1,
help='A benchmark executable or JSON output file')
contender = parser_a.add_argument_group(
'contender', 'The benchmark that will be compared against the baseline')
contender.add_argument(
'test_contender',
metavar='test_contender',
type=argparse.FileType('r'),
nargs=1,
help='A benchmark executable or JSON output file')
parser_a.add_argument(
'benchmark_options',
metavar='benchmark_options',
nargs=argparse.REMAINDER,
help='Arguments to pass when running benchmark executables')
parser_b = subparsers.add_parser(
'filters', help='Compare filter one with the filter two of benchmark')
baseline = parser_b.add_argument_group(
'baseline', 'The benchmark baseline')
baseline.add_argument(
'test',
metavar='test',
type=argparse.FileType('r'),
nargs=1,
help='A benchmark executable or JSON output file')
baseline.add_argument(
'filter_baseline',
metavar='filter_baseline',
type=str,
nargs=1,
help='The first filter, that will be used as baseline')
contender = parser_b.add_argument_group(
'contender', 'The benchmark that will be compared against the baseline')
contender.add_argument(
'filter_contender',
metavar='filter_contender',
type=str,
nargs=1,
help='The second filter, that will be compared against the baseline')
parser_b.add_argument(
'benchmark_options',
metavar='benchmark_options',
nargs=argparse.REMAINDER,
help='Arguments to pass when running benchmark executables')
parser_c = subparsers.add_parser(
'benchmarksfiltered',
help='Compare filter one of first benchmark with filter two of the second benchmark')
baseline = parser_c.add_argument_group(
'baseline', 'The benchmark baseline')
baseline.add_argument(
'test_baseline',
metavar='test_baseline',
type=argparse.FileType('r'),
nargs=1,
help='A benchmark executable or JSON output file')
baseline.add_argument(
'filter_baseline',
metavar='filter_baseline',
type=str,
nargs=1,
help='The first filter, that will be used as baseline')
contender = parser_c.add_argument_group(
'contender', 'The benchmark that will be compared against the baseline')
contender.add_argument(
'test_contender',
metavar='test_contender',
type=argparse.FileType('r'),
nargs=1,
help='The second benchmark executable or JSON output file, that will be compared against the baseline')
contender.add_argument(
'filter_contender',
metavar='filter_contender',
type=str,
nargs=1,
help='The second filter, that will be compared against the baseline')
parser_c.add_argument(
'benchmark_options',
metavar='benchmark_options',
nargs=argparse.REMAINDER,
help='Arguments to pass when running benchmark executables')
return parser
def main():
# Parse the command line flags
parser = create_parser()
args, unknown_args = parser.parse_known_args()
assert not unknown_args
benchmark_options = args.benchmark_options
if args.mode == 'benchmarks':
test_baseline = args.test_baseline[0].name
test_contender = args.test_contender[0].name
filter_baseline = ''
filter_contender = ''
# NOTE: if test_baseline == test_contender, you are analyzing the stdev
description = 'Comparing %s to %s' % (test_baseline, test_contender)
elif args.mode == 'filters':
test_baseline = args.test[0].name
test_contender = args.test[0].name
filter_baseline = args.filter_baseline[0]
filter_contender = args.filter_contender[0]
# NOTE: if filter_baseline == filter_contender, you are analyzing the
# stdev
description = 'Comparing %s to %s (from %s)' % (
filter_baseline, filter_contender, args.test[0].name)
elif args.mode == 'benchmarksfiltered':
test_baseline = args.test_baseline[0].name
test_contender = args.test_contender[0].name
filter_baseline = args.filter_baseline[0]
filter_contender = args.filter_contender[0]
# NOTE: if test_baseline == test_contender and
# filter_baseline == filter_contender, you are analyzing the stdev
description = 'Comparing %s (from %s) to %s (from %s)' % (
filter_baseline, test_baseline, filter_contender, test_contender)
else:
# should never happen
print("Unrecognized mode of operation: '%s'" % args.mode)
exit(1)
check_inputs(test_baseline, test_contender, benchmark_options)
options_baseline = []
options_contender = []
if filter_baseline and filter_contender:
options_baseline = ['--benchmark_filter=%s' % filter_baseline]
options_contender = ['--benchmark_filter=%s' % filter_contender]
# Run the benchmarks and report the results
json1 = json1_orig = gbench.util.run_or_load_benchmark(
test_baseline, benchmark_options + options_baseline)
json2 = json2_orig = gbench.util.run_or_load_benchmark(
test_contender, benchmark_options + options_contender)
# Now, filter the benchmarks so that the difference report can work
if filter_baseline and filter_contender:
replacement = '[%s vs. %s]' % (filter_baseline, filter_contender)
json1 = gbench.report.filter_benchmark(
json1_orig, filter_baseline, replacement)
json2 = gbench.report.filter_benchmark(
json2_orig, filter_contender, replacement)
# Diff and output
output_lines = gbench.report.generate_difference_report(json1, json2)
print(description)
for ln in output_lines:
print(ln)
import unittest
class TestParser(unittest.TestCase):
def setUp(self):
self.parser = create_parser()
testInputs = os.path.join(
os.path.dirname(
os.path.realpath(__file__)),
'gbench',
'Inputs')
self.testInput0 = os.path.join(testInputs, 'test_baseline_run1.json')
self.testInput1 = os.path.join(testInputs, 'test_baseline_run2.json')
def test_benchmarks_basic(self):
parsed = self.parser.parse_args(
['benchmarks', self.testInput0, self.testInput1])
self.assertEqual(parsed.mode, 'benchmarks')
self.assertEqual(parsed.test_baseline[0].name, self.testInput0)
self.assertEqual(parsed.test_contender[0].name, self.testInput1)
self.assertFalse(parsed.benchmark_options)
def test_benchmarks_with_remainder(self):
parsed = self.parser.parse_args(
['benchmarks', self.testInput0, self.testInput1, 'd'])
self.assertEqual(parsed.mode, 'benchmarks')
self.assertEqual(parsed.test_baseline[0].name, self.testInput0)
self.assertEqual(parsed.test_contender[0].name, self.testInput1)
self.assertEqual(parsed.benchmark_options, ['d'])
def test_benchmarks_with_remainder_after_doubleminus(self):
parsed = self.parser.parse_args(
['benchmarks', self.testInput0, self.testInput1, '--', 'e'])
self.assertEqual(parsed.mode, 'benchmarks')
self.assertEqual(parsed.test_baseline[0].name, self.testInput0)
self.assertEqual(parsed.test_contender[0].name, self.testInput1)
self.assertEqual(parsed.benchmark_options, ['e'])
def test_filters_basic(self):
parsed = self.parser.parse_args(
['filters', self.testInput0, 'c', 'd'])
self.assertEqual(parsed.mode, 'filters')
self.assertEqual(parsed.test[0].name, self.testInput0)
self.assertEqual(parsed.filter_baseline[0], 'c')
self.assertEqual(parsed.filter_contender[0], 'd')
self.assertFalse(parsed.benchmark_options)
def test_filters_with_remainder(self):
parsed = self.parser.parse_args(
['filters', self.testInput0, 'c', 'd', 'e'])
self.assertEqual(parsed.mode, 'filters')
self.assertEqual(parsed.test[0].name, self.testInput0)
self.assertEqual(parsed.filter_baseline[0], 'c')
self.assertEqual(parsed.filter_contender[0], 'd')
self.assertEqual(parsed.benchmark_options, ['e'])
def test_filters_with_remainder_after_doubleminus(self):
parsed = self.parser.parse_args(
['filters', self.testInput0, 'c', 'd', '--', 'f'])
self.assertEqual(parsed.mode, 'filters')
self.assertEqual(parsed.test[0].name, self.testInput0)
self.assertEqual(parsed.filter_baseline[0], 'c')
self.assertEqual(parsed.filter_contender[0], 'd')
self.assertEqual(parsed.benchmark_options, ['f'])
def test_benchmarksfiltered_basic(self):
parsed = self.parser.parse_args(
['benchmarksfiltered', self.testInput0, 'c', self.testInput1, 'e'])
self.assertEqual(parsed.mode, 'benchmarksfiltered')
self.assertEqual(parsed.test_baseline[0].name, self.testInput0)
self.assertEqual(parsed.filter_baseline[0], 'c')
self.assertEqual(parsed.test_contender[0].name, self.testInput1)
self.assertEqual(parsed.filter_contender[0], 'e')
self.assertFalse(parsed.benchmark_options)
def test_benchmarksfiltered_with_remainder(self):
parsed = self.parser.parse_args(
['benchmarksfiltered', self.testInput0, 'c', self.testInput1, 'e', 'f'])
self.assertEqual(parsed.mode, 'benchmarksfiltered')
self.assertEqual(parsed.test_baseline[0].name, self.testInput0)
self.assertEqual(parsed.filter_baseline[0], 'c')
self.assertEqual(parsed.test_contender[0].name, self.testInput1)
self.assertEqual(parsed.filter_contender[0], 'e')
self.assertEqual(parsed.benchmark_options[0], 'f')
def test_benchmarksfiltered_with_remainder_after_doubleminus(self):
parsed = self.parser.parse_args(
['benchmarksfiltered', self.testInput0, 'c', self.testInput1, 'e', '--', 'g'])
self.assertEqual(parsed.mode, 'benchmarksfiltered')
self.assertEqual(parsed.test_baseline[0].name, self.testInput0)
self.assertEqual(parsed.filter_baseline[0], 'c')
self.assertEqual(parsed.test_contender[0].name, self.testInput1)
self.assertEqual(parsed.filter_contender[0], 'e')
self.assertEqual(parsed.benchmark_options[0], 'g')
if __name__ == '__main__':
# unittest.main()
main()
# vim: tabstop=4 expandtab shiftwidth=4 softtabstop=4
# kate: tab-width: 4; replace-tabs on; indent-width 4; tab-indents: off;
# kate: indent-mode python; remove-trailing-spaces modified;

81
tools/gbench/Inputs/test2_run.json Normal file
View File

@ -0,0 +1,81 @@
{
"context": {
"date": "2016-08-02 17:44:46",
"num_cpus": 4,
"mhz_per_cpu": 4228,
"cpu_scaling_enabled": false,
"library_build_type": "release"
},
"benchmarks": [
{
"name": "BM_Hi",
"iterations": 1234,
"real_time": 42,
"cpu_time": 24,
"time_unit": "ms"
},
{
"name": "BM_Zero",
"iterations": 1000,
"real_time": 10,
"cpu_time": 10,
"time_unit": "ns"
},
{
"name": "BM_Zero/4",
"iterations": 4000,
"real_time": 40,
"cpu_time": 40,
"time_unit": "ns"
},
{
"name": "Prefix/BM_Zero",
"iterations": 2000,
"real_time": 20,
"cpu_time": 20,
"time_unit": "ns"
},
{
"name": "Prefix/BM_Zero/3",
"iterations": 3000,
"real_time": 30,
"cpu_time": 30,
"time_unit": "ns"
},
{
"name": "BM_One",
"iterations": 5000,
"real_time": 5,
"cpu_time": 5,
"time_unit": "ns"
},
{
"name": "BM_One/4",
"iterations": 2000,
"real_time": 20,
"cpu_time": 20,
"time_unit": "ns"
},
{
"name": "Prefix/BM_One",
"iterations": 1000,
"real_time": 10,
"cpu_time": 10,
"time_unit": "ns"
},
{
"name": "Prefix/BM_One/3",
"iterations": 1500,
"real_time": 15,
"cpu_time": 15,
"time_unit": "ns"
},
{
"name": "BM_Bye",
"iterations": 5321,
"real_time": 11,
"cpu_time": 63,
"time_unit": "ns"
}
]
}

57
tools/gbench/report.py
View File

@ -1,6 +1,8 @@
"""report.py - Utilities for reporting statistics about benchmark results
"""
import os
import re
import copy
class BenchmarkColor(object):
def __init__(self, name, code):
@ -66,6 +68,22 @@ def calculate_change(old_val, new_val):
return float(new_val - old_val) / abs(old_val)
def filter_benchmark(json_orig, family, replacement=""):
"""
Apply a filter to the json, and only leave the 'family' of benchmarks.
"""
regex = re.compile(family)
filtered = {}
filtered['benchmarks'] = []
for be in json_orig['benchmarks']:
if not regex.search(be['name']):
continue
filteredbench = copy.deepcopy(be) # Do NOT modify the old name!
filteredbench['name'] = regex.sub(replacement, filteredbench['name'])
filtered['benchmarks'].append(filteredbench)
return filtered
def generate_difference_report(json1, json2, use_color=True):
"""
Calculate and report the difference between each test of two benchmarks
@ -77,8 +95,9 @@ def generate_difference_report(json1, json2, use_color=True):
if b['name'] == name:
return b
return None
first_line = "{:<{}s} Time CPU Time Old Time New CPU Old CPU New".format(
'Benchmark', first_col_width)
first_col_width = max(first_col_width, len('Benchmark'))
first_line = "{:<{}s}Time CPU Time Old Time New CPU Old CPU New".format(
'Benchmark', 12 + first_col_width)
output_strs = [first_line, '-' * len(first_line)]
gen = (bn for bn in json1['benchmarks'] if 'real_time' in bn and 'cpu_time' in bn)
@ -151,5 +170,39 @@ class TestReportDifference(unittest.TestCase):
self.assertEqual(parts, expect_lines[i])
class TestReportDifferenceBetweenFamilies(unittest.TestCase):
def load_result(self):
import json
testInputs = os.path.join(os.path.dirname(os.path.realpath(__file__)), 'Inputs')
testOutput = os.path.join(testInputs, 'test2_run.json')
with open(testOutput, 'r') as f:
json = json.load(f)
return json
def test_basic(self):
expect_lines = [
['.', '-0.5000', '-0.5000', '10', '5', '10', '5'],
['./4', '-0.5000', '-0.5000', '40', '20', '40', '20'],
['Prefix/.', '-0.5000', '-0.5000', '20', '10', '20', '10'],
['Prefix/./3', '-0.5000', '-0.5000', '30', '15', '30', '15'],
]
json = self.load_result()
json1 = filter_benchmark(json, "BM_Z.ro", ".")
json2 = filter_benchmark(json, "BM_O.e", ".")
output_lines_with_header = generate_difference_report(json1, json2, use_color=False)
output_lines = output_lines_with_header[2:]
print "\n"
print("\n".join(output_lines_with_header))
self.assertEqual(len(output_lines), len(expect_lines))
for i in range(0, len(output_lines)):
parts = [x for x in output_lines[i].split(' ') if x]
self.assertEqual(len(parts), 7)
self.assertEqual(parts, expect_lines[i])
if __name__ == '__main__':
unittest.main()
# vim: tabstop=4 expandtab shiftwidth=4 softtabstop=4
# kate: tab-width: 4; replace-tabs on; indent-width 4; tab-indents: off;
# kate: indent-mode python; remove-trailing-spaces modified;