title | date | comments | categories | tags | permalink | |||
---|---|---|---|---|---|---|---|---|
多路I/O转接服务器之poll
|
2020/3/19
|
true
|
|
|
6.14
|
epoll功能说明
epoll是目前比较流行的多路I/O复用技术,技术人员必备技能,其实最最重要的是掌握三个基础PI函数的使用
1、创建一个epoll句柄,参数size用来告诉内核监听的文件描述符个数,需要考虑内存的大小
#include
int epoll_create(int size) //size:监听数目
2、控制某个epoll监控的文件描述符上的事件:注册,修改,删除。
#include
epoll_ctl(int epfd, int op, int fd, struct epoll_event *event)
epfd:为epoll_creat的句柄
op: 表示动作,用3个宏来表示:
EPOLL_CTL_ADD (注册新的fd到epfd),
EPOLL_CTL_MOD (修改已经注册的fd的监听事件),
EPOLL_CTL_DEL (从epfd删除一个fd);
event:告诉内核需要监听的事件
struct epoll_event {
__uint32_t events; /* Epoll events */
epoll_data_t data; /* User data variable */
};
typedef union epoll_data {
void *ptr;
int fd;
uint32_t u32;
uint64_t u64;
} epoll_data_t;
EPOLLIN : 表示对应的文件描述符可以读(包括对端SOCKET正常关闭)
EPOLLOUT: 表示对应的文件描述符可以写
EPOLLPRI: 表示对应的文件描述符有紧急的数据可读(这里应该表示有带外数据到来)
EPOLLERR: 表示对应的文件描述符发生错误
EPOLLHUP: 表示对应的文件描述符被挂断;
EPOLLET: 将EPOLL设为边缘触发(Edge Triggered)模式,这是相对于水平触发(Level Triggered)而言的
EPOLLONESHOT:只监听一次事件,当监听完这次事件之后,如果还需要继续监听这个socket的话,需要再次把这个socket加入到EPOLL队列里
3、等待所监控文件描述符上有事件的产生,类似于调用select
#include
int epoll_wait(int epfd, struct epoll_event *events, int maxevents, int timeout)
events: 用来存内核得到事件的集合,
maxevents: 告之内核这个events有多大,这个maxevents的值不能大于创建epoll_create()时的size,
timeout: 是超时时间
-1: 阻塞
0: 立即返回,非阻塞
>0: 指定毫秒
返回值: 成功返回有多少文件描述符就绪,时间到时返回0,出错返回-1
epoll事件模型
epoll有两种时间模型
- Edge Triggered(ET) 边缘触发,只有数据到来才触发,不管缓冲区是否有数据.这种模式下必须使用非阻塞套接口,以避免由于一个文件句柄的阻塞读/阻塞写操作把处理多个文件描述符的任务饿死。
- Level Trieggered(LT) 水平触发,只要缓冲区有数据就会触发
代码示例
示例1
#include
#include
#include
#include
#include
#include //MAC系统没有这个头文件
#include
#include "wrap.h"
#define MAXLINE 80
#define SERV_PORT 6666
#define OPEN_MAX 1024
int main(int argc, char *argv[])
{
int i, j, maxi, listenfd, connfd, sockfd;
int nready, efd, res;
ssize_t n;
char buf[MAXLINE], str[INET_ADDRSTRLEN];
socklen_t clilen;
int client[OPEN_MAX];
struct sockaddr_in cliaddr, servaddr;
struct epoll_event tep, ep[OPEN_MAX];
listenfd = Socket(AF_INET, SOCK_STREAM, 0);
bzero(&servaddr, sizeof(servaddr));
servaddr.sin_family = AF_INET;
servaddr.sin_addr.s_addr = htonl(INADDR_ANY);
servaddr.sin_port = htons(SERV_PORT);
Bind(listenfd, (struct sockaddr *) &servaddr, sizeof(servaddr));
Listen(listenfd, 20);
for (i = 0; i < OPEN_MAX; i++)
client[i] = -1;
maxi = -1;
efd = epoll_create(OPEN_MAX);
if (efd == -1)
perr_exit("epoll_create");
tep.events = EPOLLIN; tep.data.fd = listenfd;
res = epoll_ctl(efd, EPOLL_CTL_ADD, listenfd, &tep);
if (res == -1)
perr_exit("epoll_ctl");
while (1) {
nready = epoll_wait(efd, ep, OPEN_MAX, -1); /* 阻塞监听 */
if (nready == -1)
perr_exit("epoll_wait");
for (i = 0; i < nready; i++) {
if (!(ep[i].events & EPOLLIN))
continue;
if (ep[i].data.fd == listenfd) {
clilen = sizeof(cliaddr);
connfd = Accept(listenfd, (struct sockaddr *)&cliaddr, &clilen);
printf("received from %s at PORT %d\n",
inet_ntop(AF_INET, &cliaddr.sin_addr, str, sizeof(str)),
ntohs(cliaddr.sin_port));
for (j = 0; j < OPEN_MAX; j++) {
if (client[j] < 0) {
client[j] = connfd; /* save descriptor */
break;
}
}
if (j == OPEN_MAX)
perr_exit("too many clients");
if (j > maxi)
maxi = j; /* max index in client[] array */
tep.events = EPOLLIN;
tep.data.fd = connfd;
res = epoll_ctl(efd, EPOLL_CTL_ADD, connfd, &tep);
if (res == -1)
perr_exit("epoll_ctl");
} else {
sockfd = ep[i].data.fd;
n = Read(sockfd, buf, MAXLINE);
if (n == 0) {
for (j = 0; j <= maxi; j++) {
if (client[j] == sockfd) {
client[j] = -1;
break;
}
}
res = epoll_ctl(efd, EPOLL_CTL_DEL, sockfd, NULL);
if (res == -1)
perr_exit("epoll_ctl");
Close(sockfd);
printf("client[%d] closed connection\n", j);
} else {
for (j = 0; j < n; j++)
buf[j] = toupper(buf[j]);
Writen(sockfd, buf, n);
}
}
}
}
close(listenfd);
close(efd);
return 0;
}
示例2
//基于管道的epoll ET模式
#include
#include
#include
#include
#include
#define MAXLINE 10
int main(int argc, char *argv[])
{
int efd, i;
int pfd[2];
pid_t pid;
char buf[MAXLINE], ch = 'a';
pipe(pfd);
pid = fork();
if (pid == 0) {
close(pfd[0]);
while (1) {
for (i = 0; i < MAXLINE/2; i++)
buf[i] = ch;
buf[i-1] = '\n';
ch++;
for (; i < MAXLINE; i++)
buf[i] = ch;
buf[i-1] = '\n';
ch++;
write(pfd[1], buf, sizeof(buf));
sleep(2);
}
close(pfd[1]);
} else if (pid > 0) {
struct epoll_event event;
struct epoll_event resevent[10];
int res, len;
close(pfd[1]);
efd = epoll_create(10);
/* event.events = EPOLLIN; */
event.events = EPOLLIN | EPOLLET; /* ET 边沿触发 ,默认是水平触发 */
event.data.fd = pfd[0];
epoll_ctl(efd, EPOLL_CTL_ADD, pfd[0], &event);
while (1) {
res = epoll_wait(efd, resevent, 10, -1);
printf("res %d\n", res);
if (resevent[0].data.fd == pfd[0]) {
len = read(pfd[0], buf, MAXLINE/2);
write(STDOUT_FILENO, buf, len);
}
}
close(pfd[0]);
close(efd);
} else {
perror("fork");
exit(-1);
}
return 0;
}
示例3
//基于C/S架构的epoll ET触发模式
#include
#include
#include
#include
#include
#include
#include
#include
#include
#define MAXLINE 10
#define SERV_PORT 8080
int main(void)
{
struct sockaddr_in servaddr, cliaddr;
socklen_t cliaddr_len;
int listenfd, connfd;
char buf[MAXLINE];
char str[INET_ADDRSTRLEN];
int i, efd;
listenfd = socket(AF_INET, SOCK_STREAM, 0);
bzero(&servaddr, sizeof(servaddr));
servaddr.sin_family = AF_INET;
servaddr.sin_addr.s_addr = htonl(INADDR_ANY);
servaddr.sin_port = htons(SERV_PORT);
bind(listenfd, (struct sockaddr *)&servaddr, sizeof(servaddr));
listen(listenfd, 20);
struct epoll_event event;
struct epoll_event resevent[10];
int res, len;
efd = epoll_create(10);
event.events = EPOLLIN | EPOLLET; /* ET 边沿触发 ,默认是水平触发 */
printf("Accepting connections ...\n");
cliaddr_len = sizeof(cliaddr);
connfd = accept(listenfd, (struct sockaddr *)&cliaddr, &cliaddr_len);
printf("received from %s at PORT %d\n",
inet_ntop(AF_INET, &cliaddr.sin_addr, str, sizeof(str)),
ntohs(cliaddr.sin_port));
event.data.fd = connfd;
epoll_ctl(efd, EPOLL_CTL_ADD, connfd, &event);
while (1) {
res = epoll_wait(efd, resevent, 10, -1);
printf("res %d\n", res);
if (resevent[0].data.fd == connfd) {
len = read(connfd, buf, MAXLINE/2);
write(STDOUT_FILENO, buf, len);
}
}
return 0;
}