The following benchmark is modified from
https://github.com/golang/go/issues/30802
package main
import (
"testing"
"time"
)
func SleepTest(milliseconds time.Duration) {
time.Sleep(time.Millisecond * milliseconds)
}
func BenchmarkTestSleep_2000(b *testing.B) {
for i := 0; i < b.N; i++ {
SleepTest(2000)
}
}
func BenchmarkTestSleep_1000(b *testing.B) {
for i := 0; i < b.N; i++ {
SleepTest(1000)
}
}
func BenchmarkTestSleep_500(b *testing.B) {
for i := 0; i < b.N; i++ {
SleepTest(500)
}
}
func BenchmarkTestSleep_100(b *testing.B) {
for i := 0; i < b.N; i++ {
SleepTest(100)
}
}
func BenchmarkTestSleep_10(b *testing.B) {
for i := 0; i < b.N; i++ {
SleepTest(10)
}
}
The result:
$ go test -bench=.
goos: linux
goarch: amd64
pkg: app/t
BenchmarkTestSleep_2000-4 1 2000142946 ns/op
BenchmarkTestSleep_1000-4 1 1000192805 ns/op
BenchmarkTestSleep_500-4 2 500191727 ns/op
BenchmarkTestSleep_100-4 10 100179544 ns/op
BenchmarkTestSleep_10-4 100 10164319 ns/op
It is some strange that when the argument is less than 1s, the smaller it
is, the more allocations.
Bu, or my understanding is wrong?
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to golang-nuts+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
