rt_thread 中tcp server,在lwip_accept()函数前调用lwip_select()函数，然后客户端连上去以后，select()无法检测到事件的变化

jason shi

unread,

Jul 3, 2014, 9:33:10 PM7/3/14

to rt-threa...@googlegroups.com

最近在看rt_thread lwip，自己写了一个小demo，发现了这个问题：

lwip_accept()函数前调用lwip_select()函数，用lwip 网络测试工具 NetAssist.exe 作为客户端连上去以后，服务端没有任何响应。不知道是怎么回事？下面是我的代码：

------------

void tcpserv(void *parameter)

{

struct sockaddr_in server_addr, client_addr;

int listenfd = -1, confd = -1, opt = 1;

int ret = -1;

int ndfs;

struct timeval timeout;

fd_set allset;

fd_set readset;

if( (listenfd = lwip_socket(AF_INET, SOCK_STREAM, 0)) < 0)

{

rt_kprintf("create socket faild!\n\r");

return ;

}

server_addr.sin_family = AF_INET;

server_addr.sin_port = htons(PORT);

server_addr.sin_addr.s_addr = INADDR_ANY;

rt_memset(&(server_addr.sin_zero),8, sizeof(server_addr.sin_zero));

if(lwip_bind(listenfd, (struct sockaddr*)&server_addr, sizeof(server_addr)) < 0)

{

rt_kprintf("bind faild!\n\r");

lwip_close(listenfd);

return ;

}

if(listen(listenfd, BACKLOG) < 0)

{

rt_kprintf("listen faild !\n\r");

lwip_close(listenfd);

return ;

}

ndfs = listenfd + 1;

FD_ZERO(&allset);

FD_ZERO(&readset);

FD_SET(listenfd, &readset);

while(1)

{

timeout.tv_sec = 1;

timeout.tv_usec = 0;

readset = allset;

ret = lwip_select(ndfs, &readset, NULL, 0, &timeout);

if(ret < 0)

{

rt_kprintf("select error!\n\r");

lwip_close(listenfd);

return ;

}else if(ret == 0)

{

rt_kprintf("select..........\n\r");

continue ;

}else

{

rt_kprintf("client connect ----->\n\r");

}

--------------------------------

有人遇到过这个问题吗？谢谢！ THX！

Ming Bai

unread,

Jul 3, 2014, 9:44:23 PM7/3/14

to rt-thread用户组

while(1)

{

timeout.tv_sec = 1;

timeout.tv_usec = 0;

readset = allset;

ret = lwip_select(ndfs, &readset, NULL, 0, &timeout);

--
You received this message because you are subscribed to the Google Groups "rt-thread用户组" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rt-thread-cnus...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

jason shi

unread,

Jul 3, 2014, 11:14:36 PM7/3/14

to rt-threa...@googlegroups.com

谢谢 Ming,

现在我按照你的提示，将上面的代码修改为这样：

---

ndfs = listenfd + 1;

FD_ZERO(&allset);

FD_ZERO(&readset);

FD_SET(listenfd, &readset);

while(1)

{

timeout.tv_sec = 1;

timeout.tv_usec = 0;

// readset = allset;

ret = lwip_select(ndfs, &readset, NULL, 0, &timeout);

if(ret < 0)

{

rt_kprintf("select error!\n\r");

lwip_close(listenfd);

return ;

}else if(ret == 0)

{

rt_kprintf("select..........\n\r");

continue ;

}else

{

rt_kprintf("client connect ----->\n\r");

}

----------------------------------------------------------------------------

然后在 lwip_selscan(int maxfdp1, fd_set *readset, fd_set *writeset, fd_set *exceptset){} 的这里加了一行打印：

rt_kprintf("hello I/m here------------------->\n\r");

------------------------------------------------------------------------------

static int

lwip_selscan(int maxfdp1, fd_set *readset, fd_set *writeset, fd_set *exceptset)

{

int i, nready = 0;

fd_set lreadset, lwriteset, lexceptset;

struct lwip_socket *p_sock;

FD_ZERO(&lreadset);

FD_ZERO(&lwriteset);

FD_ZERO(&lexceptset);

/* Go through each socket in each list to count number of sockets which

currently match */

for(i = 0; i < maxfdp1; i++) {

if (FD_ISSET(i, readset)) {

/* See if netconn of this socket is ready for read */

//这里是我自己添加的 add by myself

rt_kprintf("hello I/m here------------------->\n\r");

p_sock = get_socket(i);

if (p_sock && (p_sock->lastdata || (p_sock->rcvevent > 0))) {

FD_SET(i, &lreadset);

LWIP_DEBUGF(SOCKETS_DEBUG, ("lwip_selscan: fd=%d ready for reading\n", i));

nready++;

//这里是我自己添加的 add by myself

rt_kprintf(""lwip_selscan: ready for reading\n\r");

}

if (FD_ISSET(i, writeset)) {

/* See if netconn of this socket is ready for write */

p_sock = get_socket(i);

if (p_sock && p_sock->sendevent) {

FD_SET(i, &lwriteset);

LWIP_DEBUGF(SOCKETS_DEBUG, ("lwip_selscan: fd=%d ready for writing\n", i));

nready++;

}

*readset = lreadset;

*writeset = lwriteset;

FD_ZERO(exceptset);

return nready;

}

--------------------------------------------------------------------

但是最后发现，客户端连上以后，通过串口打印，根本就不执行 rt_kprintf("hello I/m here------------------->\n\r");也就是FD_ISSET()测试失败，然后就执行下面去了。最后就导致了 select()返回为0.

===========

如果，我去掉 if (FD_ISSET(i, readset)) 这个约束条件后，程序能够执行 rt_kprintf("hello I/m here------------------->\n\r");和 rt_kprintf(""lwip_selscan: ready for reading\n\r"); 通过串口打印，我已经观察到了这两个打印。

------

像这样：

------

for(i = 0; i < maxfdp1; i++) {

// if (FD_ISSET(i, readset)) {

/* See if netconn of this socket is ready for read */

rt_kprintf("hello I/m here------------------->\n\r");

p_sock = get_socket(i);

if (p_sock && (p_sock->lastdata || (p_sock->rcvevent > 0))) {

FD_SET(i, &lreadset);

LWIP_DEBUGF(SOCKETS_DEBUG, ("lwip_selscan: fd=%d ready for reading\n", i));

nready++;

rt_kprintf(""lwip_selscan: ready for reading\n\r");

}

// }

if (FD_ISSET(i, writeset)) {

/* See if netconn of this socket is ready for write */

p_sock = get_socket(i);

if (p_sock && p_sock->sendevent) {

FD_SET(i, &lwriteset);

LWIP_DEBUGF(SOCKETS_DEBUG, ("lwip_selscan: fd=%d ready for writing\n", i));

nready++;

}

----------

由于MDK没找到 FD_ISSET()的实现，不知道为什么无法检测到呢

THX ！

--
You received this message because you are subscribed to a topic in the Google Groups "rt-thread用户组" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/rt-thread-cnusers/-Jla_auV0Dk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to rt-thread-cnus...@googlegroups.com.

Ming Bai

unread,

Jul 3, 2014, 11:27:53 PM7/3/14

to rt-thread用户组

这些都是socket编程的问题，不一定要在lwip上做这种试验，先在PC上练练呗

jason shi

unread,

Jul 4, 2014, 12:34:28 AM7/4/14

to rt-threa...@googlegroups.com

谢谢！

jason shi

unread,

Aug 25, 2014, 5:18:42 AM8/25/14

to rt-threa...@googlegroups.com

谢谢各位大侠，确实是自己没搞清楚 select()的性质：程序每执行一次select()，就会把里面的某些参数重置一次，我已经写了笔记了，供有需要的新人查看吧，链接地址：