多次运行同一个 go 程序会引发恐慌：在关闭的通道上发送

在 Go 语言中，使用通道进行并发编程时，如果多次运行同一个程序，可能会遇到“在关闭通道上发送”的错误。这是因为程序中的工作协程没有在关闭通道之前完成工作。解决方法是使用 sync.WaitGroup 来等待工作协程完成后再关闭通道，确保所有协程都已完成任务，避免在已关闭的通道上发送数据。

我是 golang 新手，试图了解工作池的工作原理。如果我运行一次，以下示例程序工作正常，如果尝试运行多次，我会收到恐慌：在封闭通道上发送错误。 go版本是go1.14.2

package main
import (
    "fmt"
    "time"
)
func main() {
    jobs := make(chan int, 10)
    results := make(chan int, 10)

    for x := 1; x <= 3; x++ {
        go worker(x, jobs, results)
    }

    for j := 1; j <= 6; j++ {
        jobs <- j
    }
    close(jobs)
    for r:=range results{
        fmt.println("result received from worker: ", r)
    }
}
func worker(id int, jobs <-chan int, results chan<- int) {
    for job := range jobs {
        fmt.println("worker ", id, " is working on job ", job)
        time.sleep(1000*time.millisecond)
        fmt.println("worker ", id, " completed work on job ", job)
        results <- job
    }
    close(results)
}

第一次运行没问题

go run main.go                                                                            
worker  3  is working on job  1
worker  1  is working on job  2
worker  2  is working on job  3
worker  2  completed work on job  3
worker  2  is working on job  4
result received from worker:  3
worker  1  completed work on job  2
worker  1  is working on job  5
result received from worker:  2
worker  3  completed work on job  1
worker  3  is working on job  6
result received from worker:  1
worker  3  completed work on job  6
result received from worker:  6

第二次运行给出了这个。

Worker  3  is working on job  2
Worker  2  is working on job  3
Worker  1  is working on job  1
Worker  3  completed work on job  2
Worker  3  is working on job  4
Worker  2  completed work on job  3
Worker  2  is working on job  5
Worker  1  completed work on job  1
Worker  1  is working on job  6
Result received from worker:  2
Result received from worker:  3
Result received from worker:  1
Worker  1  completed work on job  6
Worker  3  completed work on job  4
Result received from worker:  6
panic: send on closed channel

goroutine 35 [running]:
main.worker(0x3, 0xc0000b0000, 0xc0000b00b0)
    /home/jibi_makkar/code/go/src/main.go:30 +0x252
created by main.main
    /home/jibi_makkar/code/go/src/main.go:13 +0xad
exit status 2

谁能帮我理解发生了什么？

解决方案

eli bendersky 的回答描述了这个问题。这个答案描述了修复方法。我知道您没有要求修复，但我认为您有兴趣。所以就在这里。

修复方法是等待工作协程完成后再关闭通道。使用sync.WaitGroup来实现等待。

var wg sync.waitgroup
for x := 1; x <= 3; x++ {
    wg.add(1)  // increment worker counter
    go func(x int) {
        defer wg.done() // decrement on return from goroutine
        worker(x, jobs, results)
    }(x)
}

// close results channel when workers are done.
go func() {
    wg.wait()
    close(results)
}()

在演示文稿：https://play.golang.org/p/GM-0Gqx0Gbg 上运行此 golang 程序

您有多个关闭 results 通道的 goroutine。

func worker(ID int, jobs <-chan int, results chan<- int) {
    for job := range jobs {
        fmt.Println("Worker ", ID, " is working on job ", job)
        time.Sleep(1000*time.Millisecond)
        fmt.Println("Worker ", ID, " completed work on job ", job)
        results <- job
    }
    close(results)            <<<<-------- Here
}

您在三个不同的并发 goroutine 中运行此 worker 函数。第一个到达标记线的通道将关闭通道，其他通道则尝试在循环中的 results <- job 中的关闭通道上发送。

今天带大家了解了的相关知识，希望对你有所帮助；关于Golang的技术知识我们会一点点深入介绍，欢迎大家关注golang学习网公众号，一起学习编程~