Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/disruptek/golden
a benchmark for compile-time and/or runtime Nim ๐
https://github.com/disruptek/golden
bench benchmark compile golden nim profile profiling run runtime
Last synced: about 1 month ago
JSON representation
a benchmark for compile-time and/or runtime Nim ๐
- Host: GitHub
- URL: https://github.com/disruptek/golden
- Owner: disruptek
- License: mit
- Created: 2019-09-30T22:22:23.000Z (over 4 years ago)
- Default Branch: master
- Last Pushed: 2021-01-14T19:33:19.000Z (over 3 years ago)
- Last Synced: 2024-05-02T01:13:38.111Z (about 2 months ago)
- Topics: bench, benchmark, compile, golden, nim, profile, profiling, run, runtime
- Language: Nim
- Homepage:
- Size: 149 KB
- Stars: 27
- Watchers: 3
- Forks: 0
- Open Issues: 10
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Lists
- awesome-nim - golden - A benchmark for compile-time and/or runtime Nim. (Development Tools / Benchmarking)
- awesome-nim-cn - golden - ไธไธชๅบๅๆต่ฏๅทฅๅ ท๏ผ็จไบๆต้ๅ่ฎฐๅฝไปปไฝๅฏๆง่กๆไปถ็่ฟ่กๆถ้ดใ[MIT] (ๆง่ฝๆต่ฏ)
README
# golden
A benchmarking tool that measures and records runtime of any executable and
also happens to know how to compile Nim.
- `cpp +/ nim-1.0` [](https://travis-ci.org/disruptek/golden)
- `arc +/ cpp +/ nim-1.3` [](https://travis-ci.org/disruptek/golden)
The idea here is that we're gonna make a record of everything we run,
everything we build, and be able to discover and pinpoint regressions
automatically using native git-fu performed by the tool. Fire and forget!
## Installation
### Nimph
```
$ nimph clone golden
```
### Nimble
```
$ nimble install golden
```
## Usage
If you pass it a binary, it'll run it a bunch of times and report some runtime
statistics periodically.
If you pass it some Nim source, it will compile it for you and report some
compilation and runtime statistics periodically.
By default, it will run until you interrupt it.
```
$ golden --truth=0.002 bench.nim
compilations after 0s
โโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโ
โ Builds โ Min โ Max โ Mean โ StdDev โ
โโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโค
โ 1 โ 0.396129 โ 0.396129 โ 0.396129 โ 0.000000 โ
โโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโ
benchmark after 1s
โโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโ
โ Runs โ Min โ Max โ Mean โ StdDev โ
โโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโค
โ 1 โ 1.959187 โ 1.959187 โ 1.959187 โ 0.000000 โ
โโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโ
benchmark after 3s
โโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโ
โ Runs โ Min โ Max โ Mean โ StdDev โ
โโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโค
โ 2 โ 1.958892 โ 1.959187 โ 1.959039 โ 0.000147 โ
โโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโ
completed benchmark after 5s
โโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโ
โ Runs โ Min โ Max โ Mean โ StdDev โ
โโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโค
โ 3 โ 1.958892 โ 1.961293 โ 1.959791 โ 0.001069 โ
โโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโ
```
Benchmarking the compilation of Nim itself:
```
$ cd ~/git/Nim
$ golden koch -- boot -d:danger
# ...
โโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโฌโโโโโโโโโโโ
โ # โ Min โ Max โ Mean โ StdDev โ
โโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโผโโโโโโโโโโโค
โ 12 โ 8.846606 โ 9.485832 โ 8.945023 โ 0.165638 โ
โโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโดโโโโโโโโโโโ
```
Benchmarking compilation of slow-to-compile Nim:
```
$ golden --compilation openapi.nim
โโโโโโโโโโฌโโโโโโโโโโโโฌโโโโโโโโโโโโฌโโโโโโโโโโโโฌโโโโโโโโโโโ
โ # โ Min โ Max โ Mean โ StdDev โ
โโโโโโโโโโผโโโโโโโโโโโโผโโโโโโโโโโโโผโโโโโโโโโโโโผโโโโโโโโโโโค
โ 1 โ 91.946370 โ 91.946370 โ 91.946370 โ 0.000000 โ
โโโโโโโโโโดโโโโโโโโโโโโดโโโโโโโโโโโโดโโโโโโโโโโโโดโโโโโโโโโโโ
โโโโโโโโโโฌโโโโโโโโโโโโฌโโโโโโโโโโโโฌโโโโโโโโโโโโฌโโโโโโโโโโโโ
โ # โ Min โ Max โ Mean โ StdDev โ
โโโโโโโโโโผโโโโโโโโโโโโผโโโโโโโโโโโโผโโโโโโโโโโโโผโโโโโโโโโโโโค
โ 2 โ 29.271556 โ 91.946370 โ 60.608963 โ 31.337407 โ
โโโโโโโโโโดโโโโโโโโโโโโดโโโโโโโโโโโโดโโโโโโโโโโโโดโโโโโโโโโโโโ
```
## Command Line Options
- `truth` a float percentage indicating how much jitter you'll accept
- `runtime` a float of seconds after which we should kill each invocation
- `iterations` a number of invocations after which we should stop the benchmark
- `storage` the path to a database file you wish to use; must end in `.golden-lmdb`
- `interactive-forced` assume output friendly to humans
- `json-output` assume output friendly to machines _(work in progress)_
- `color-forced` enable color output when not in `interactive` mode
- `prune-outliers` throw out this percentage of aberrant invocations with long runtime in order to clean up the histogram
- `dry-run` don't write any results to the database
- `histogram-classes` the number of points in the histogram
- `compilation-only` benchmark the Nim compiler on the given source(s)
- `brief` only output the statistics at the completion of the benchmark
- `never-output` never emit anything via stdout/stderr
- `dump-output` always print the stdout/stderr of the benchmarked program
- `--` the following arguments are passed to the compiler and runtime. Note that if you supply `-- cpp` for compilation via C++, you will need to supply your own defines such as `-d:danger`.
## License
MIT