1
0
mirror of https://github.com/golang/go synced 2024-11-22 00:24:41 -07:00

testing: allow benchmarks to print and fail

Refactors the benchmarks and test code.
Now benchmarks can call Errorf, Fail, etc.,
and the runner will act accordingly.

Because functionality has been folded into an
embedded type, a number of methods' docs
no longer appear in godoc output. A fix is
underway; if it doesn't happen fast enough,
I'll add wrapper methods to restore the
documentation.

R=bradfitz, adg, rsc
CC=golang-dev
https://golang.org/cl/5492060
This commit is contained in:
Rob Pike 2011-12-20 09:51:39 -08:00
parent 7be6229f9f
commit c50e4f5e2f
3 changed files with 126 additions and 59 deletions

View File

@ -25,12 +25,12 @@ type InternalBenchmark struct {
// B is a type passed to Benchmark functions to manage benchmark
// timing and to specify the number of iterations to run.
type B struct {
common
N int
benchmark InternalBenchmark
ns time.Duration
bytes int64
start time.Time
timerOn bool
result BenchmarkResult
}
// StartTimer starts timing a test. This function is called automatically
@ -48,7 +48,7 @@ func (b *B) StartTimer() {
// want to measure.
func (b *B) StopTimer() {
if b.timerOn {
b.ns += time.Now().Sub(b.start)
b.duration += time.Now().Sub(b.start)
b.timerOn = false
}
}
@ -59,7 +59,7 @@ func (b *B) ResetTimer() {
if b.timerOn {
b.start = time.Now()
}
b.ns = 0
b.duration = 0
}
// SetBytes records the number of bytes processed in a single operation.
@ -70,7 +70,7 @@ func (b *B) nsPerOp() int64 {
if b.N <= 0 {
return 0
}
return b.ns.Nanoseconds() / int64(b.N)
return b.duration.Nanoseconds() / int64(b.N)
}
// runN runs a single benchmark for the specified number of iterations.
@ -127,17 +127,25 @@ func roundUp(n int) int {
return 10 * base
}
// run times the benchmark function. It gradually increases the number
// run times the benchmark function in a separate goroutine.
func (b *B) run() BenchmarkResult {
go b.launch()
<-b.signal
return b.result
}
// launch launches the benchmark function. It gradually increases the number
// of benchmark iterations until the benchmark runs for a second in order
// to get a reasonable measurement. It prints timing information in this form
// testing.BenchmarkHello 100000 19 ns/op
func (b *B) run() BenchmarkResult {
// launch is run by the fun function as a separate goroutine.
func (b *B) launch() {
// Run the benchmark for a single iteration in case it's expensive.
n := 1
b.runN(n)
// Run the benchmark for at least the specified amount of time.
d := time.Duration(*benchTime * float64(time.Second))
for b.ns < d && n < 1e9 {
for !b.failed && b.duration < d && n < 1e9 {
last := n
// Predict iterations/sec.
if b.nsPerOp() == 0 {
@ -153,7 +161,8 @@ func (b *B) run() BenchmarkResult {
n = roundUp(n)
b.runN(n)
}
return BenchmarkResult{b.N, b.ns, b.bytes}
b.result = BenchmarkResult{b.N, b.duration, b.bytes}
b.signal <- b
}
// The results of a benchmark run.
@ -215,14 +224,32 @@ func RunBenchmarks(matchString func(pat, str string) (bool, error), benchmarks [
}
for _, procs := range cpuList {
runtime.GOMAXPROCS(procs)
b := &B{benchmark: Benchmark}
b := &B{
common: common{
signal: make(chan interface{}),
},
benchmark: Benchmark,
}
benchName := Benchmark.Name
if procs != 1 {
benchName = fmt.Sprintf("%s-%d", Benchmark.Name, procs)
}
fmt.Printf("%s\t", benchName)
r := b.run()
if b.failed {
// The output could be very long here, but probably isn't.
// We print it all, regardless, because we don't want to trim the reason
// the benchmark failed.
fmt.Printf("--- FAIL: %s\n%s", benchName, b.output)
continue
}
fmt.Printf("%v\n", r)
// Unlike with tests, we ignore the -chatty flag and always print output for
// benchmarks since the output generation time will skew the results.
if len(b.output) > 0 {
b.trimOutput()
fmt.Printf("--- BENCH: %s\n%s", benchName, b.output)
}
if p := runtime.GOMAXPROCS(-1); p != procs {
fmt.Fprintf(os.Stderr, "testing: %s left GOMAXPROCS set to %d\n", benchName, p)
}
@ -230,9 +257,31 @@ func RunBenchmarks(matchString func(pat, str string) (bool, error), benchmarks [
}
}
// trimOutput shortens the output from a benchmark, which can be very long.
func (b *B) trimOutput() {
// The output is likely to appear multiple times because the benchmark
// is run multiple times, but at least it will be seen. This is not a big deal
// because benchmarks rarely print, but just in case, we trim it if it's too long.
const maxNewlines = 10
for nlCount, j := 0, 0; j < len(b.output); j++ {
if b.output[j] == '\n' {
nlCount++
if nlCount >= maxNewlines {
b.output = append(b.output[:j], "\n\t... [output truncated]\n"...)
break
}
}
}
}
// Benchmark benchmarks a single function. Useful for creating
// custom benchmarks that do not use gotest.
func Benchmark(f func(b *B)) BenchmarkResult {
b := &B{benchmark: InternalBenchmark{"", f}}
b := &B{
common: common{
signal: make(chan interface{}),
},
benchmark: InternalBenchmark{"", f},
}
return b.run()
}

View File

@ -23,8 +23,8 @@
// }
// The benchmark package will vary b.N until the benchmark function lasts
// long enough to be timed reliably. The output
// testing.BenchmarkHello 500000 4076 ns/op
// means that the loop ran 500000 times at a speed of 4076 ns per loop.
// testing.BenchmarkHello 10000000 282 ns/op
// means that the loop ran 10000000 times at a speed of 282 ns per loop.
//
// If a benchmark needs some expensive setup before running, the timer
// may be stopped:
@ -70,6 +70,17 @@ var (
cpuList []int
)
// common holds the elements common between T and B and
// captures common methods such as Errorf.
type common struct {
output []byte // Output generated by test or benchmark.
failed bool // Test or benchmark has failed.
start time.Time // Time test or benchmark started
duration time.Duration
self interface{} // To be sent on signal channel when done.
signal chan interface{} // Output for serial tests.
}
// Short reports whether the -test.short flag is set.
func Short() bool {
return *short
@ -111,69 +122,67 @@ func decorate(s string, addFileLine bool) string {
// T is a type passed to Test functions to manage test state and support formatted test logs.
// Logs are accumulated during execution and dumped to standard error when done.
type T struct {
common
name string // Name of test.
errors string // Error string from test.
failed bool // Test has failed.
ch chan *T // Output for serial tests.
startParallel chan bool // Parallel tests will wait on this.
start time.Time // Time test started
dt time.Duration // Length of test
}
// Fail marks the Test function as having failed but continues execution.
func (t *T) Fail() { t.failed = true }
// Fail marks the function as having failed but continues execution.
func (c *common) Fail() { c.failed = true }
// Failed returns whether the Test function has failed.
func (t *T) Failed() bool { return t.failed }
// Failed returns whether the function has failed.
func (c *common) Failed() bool { return c.failed }
// FailNow marks the Test function as having failed and stops its execution.
// FailNow marks the function as having failed and stops its execution.
// Execution will continue at the next Test.
func (t *T) FailNow() {
t.dt = time.Now().Sub(t.start)
t.Fail()
t.ch <- t
func (c *common) FailNow() {
c.duration = time.Now().Sub(c.start)
c.Fail()
c.signal <- c.self
runtime.Goexit()
}
// log generates the output. It's always at the same stack depth.
func (t *T) log(s string) { t.errors += decorate(s, true) }
func (c *common) log(s string) {
c.output = append(c.output, decorate(s, true)...)
}
// Log formats its arguments using default formatting, analogous to Print(),
// and records the text in the error log.
func (t *T) Log(args ...interface{}) { t.log(fmt.Sprintln(args...)) }
func (c *common) Log(args ...interface{}) { c.log(fmt.Sprintln(args...)) }
// Logf formats its arguments according to the format, analogous to Printf(),
// and records the text in the error log.
func (t *T) Logf(format string, args ...interface{}) { t.log(fmt.Sprintf(format, args...)) }
func (c *common) Logf(format string, args ...interface{}) { c.log(fmt.Sprintf(format, args...)) }
// Error is equivalent to Log() followed by Fail().
func (t *T) Error(args ...interface{}) {
t.log(fmt.Sprintln(args...))
t.Fail()
func (c *common) Error(args ...interface{}) {
c.log(fmt.Sprintln(args...))
c.Fail()
}
// Errorf is equivalent to Logf() followed by Fail().
func (t *T) Errorf(format string, args ...interface{}) {
t.log(fmt.Sprintf(format, args...))
t.Fail()
func (c *common) Errorf(format string, args ...interface{}) {
c.log(fmt.Sprintf(format, args...))
c.Fail()
}
// Fatal is equivalent to Log() followed by FailNow().
func (t *T) Fatal(args ...interface{}) {
t.log(fmt.Sprintln(args...))
t.FailNow()
func (c *common) Fatal(args ...interface{}) {
c.log(fmt.Sprintln(args...))
c.FailNow()
}
// Fatalf is equivalent to Logf() followed by FailNow().
func (t *T) Fatalf(format string, args ...interface{}) {
t.log(fmt.Sprintf(format, args...))
t.FailNow()
func (c *common) Fatalf(format string, args ...interface{}) {
c.log(fmt.Sprintf(format, args...))
c.FailNow()
}
// Parallel signals that this test is to be run in parallel with (and only with)
// other parallel tests in this CPU group.
func (t *T) Parallel() {
t.ch <- nil // Release main testing loop
t.signal <- nil // Release main testing loop
<-t.startParallel // Wait for serial tests to finish
}
@ -187,8 +196,8 @@ type InternalTest struct {
func tRunner(t *T, test *InternalTest) {
t.start = time.Now()
test.F(t)
t.dt = time.Now().Sub(t.start)
t.ch <- t
t.duration = time.Now().Sub(t.start)
t.signal <- t
}
// An internal function but exported because it is cross-package; part of the implementation
@ -211,13 +220,13 @@ func Main(matchString func(pat, str string) (bool, error), tests []InternalTest,
after()
}
func report(t *T) {
tstr := fmt.Sprintf("(%.2f seconds)", t.dt.Seconds())
func (t *T) report() {
tstr := fmt.Sprintf("(%.2f seconds)", t.duration.Seconds())
format := "--- %s: %s %s\n%s"
if t.failed {
fmt.Printf(format, "FAIL", t.name, tstr, t.errors)
fmt.Printf(format, "FAIL", t.name, tstr, t.output)
} else if *chatty {
fmt.Printf(format, "PASS", t.name, tstr, t.errors)
fmt.Printf(format, "PASS", t.name, tstr, t.output)
}
}
@ -227,7 +236,9 @@ func RunTests(matchString func(pat, str string) (bool, error), tests []InternalT
fmt.Fprintln(os.Stderr, "testing: warning: no tests to run")
return
}
ch := make(chan *T)
// TODO: each test should have its own channel, although that means
// keeping track of the channels when we're running parallel tests.
signal := make(chan interface{})
for _, procs := range cpuList {
runtime.GOMAXPROCS(procs)
@ -247,17 +258,24 @@ func RunTests(matchString func(pat, str string) (bool, error), tests []InternalT
if procs != 1 {
testName = fmt.Sprintf("%s-%d", tests[i].Name, procs)
}
t := &T{ch: ch, name: testName, startParallel: startParallel}
t := &T{
common: common{
signal: signal,
},
name: testName,
startParallel: startParallel,
}
t.self = t
if *chatty {
fmt.Printf("=== RUN %s\n", t.name)
}
go tRunner(t, &tests[i])
out := <-t.ch
out := (<-t.signal).(*T)
if out == nil { // Parallel run.
numParallel++
continue
}
report(t)
t.report()
ok = ok && !out.failed
}
@ -269,8 +287,8 @@ func RunTests(matchString func(pat, str string) (bool, error), tests []InternalT
numParallel--
continue
}
t := <-ch
report(t)
t := (<-signal).(*T)
t.report()
ok = ok && !t.failed
running--
}

View File

@ -12,9 +12,9 @@ func main() {
var t testing.T
// make sure error mentions that
// ch is unexported, not just "ch not found".
// name is unexported, not just "name not found".
t.ch = nil // ERROR "unexported"
t.name = nil // ERROR "unexported"
println(testing.anyLowercaseName("asdf")) // ERROR "unexported" "undefined: testing.anyLowercaseName"
}