无法在辅助接口上接收 UDP/IPv6 数据包

本文介绍了无法在辅助接口上接收 UDP/IPv6 数据包的处理方法，对大家解决问题具有一定的参考价值，需要的朋友们下面随着小编来一起学习吧！ 问题描述

我有以下用于 UDP 服务器的 C 代码，它将绑定到辅助接口 tap0 的 ipv6 地址.

I have the following C code for UDP server which would bind to the ipv6 address of the secondary interface tap0.

context_t * new_context(const struct sockaddr *listen_addr, size_t addr_size) { context_t *c = malloc( sizeof( context_t ) ); time_t now; int reuse = 1; if (!listen_addr) { fprintf(stderr, "no listen address specified "); return NULL; } srand( getpid() ^ time(&now) ); if ( !c ) { perror("init: malloc:"); return NULL; } memset(c, 0, sizeof( context_t ) ); c->sockfd = socket(listen_addr->sa_family, SOCK_DGRAM, 0); if ( c->sockfd < 0 ) { perror("new_context: socket"); goto onerror; } if ( setsockopt( c->sockfd, SOL_SOCKET, SO_REUSEADDR, &reuse, sizeof(reuse) ) < 0 ) perror("setsockopt SO_REUSEADDR"); if ( bind (c->sockfd, listen_addr, addr_size) < 0 ) { perror("new_context: bind"); goto onerror; } return c; onerror: if ( c->sockfd >= 0 ) close ( c->sockfd ); free( c ); return NULL; } context_t * get_context(const char *ipaddress, const char *port, unsigned int scopeId) { int s; context_t* ctx; struct addrinfo hints; struct addrinfo *result, *rp; memset(&hints, 0, sizeof(struct addrinfo)); hints.ai_family = AF_UNSPEC; /* Allow IPv4 or IPv6 */ hints.ai_socktype = SOCK_DGRAM; /* Coap uses UDP */ hints.ai_flags = AI_PASSIVE | AI_NUMERICHOST | AI_NUMERICSERV | AI_ALL; s = getaddrinfo(ipaddress, port, &hints, &result); if ( s != 0 ) { fprintf(stderr, "getaddrinfo: %s ", gai_strerror(s)); return NULL; } /* iterate through results until success */ for (rp = result; rp != NULL; rp = rp->ai_next) { ctx = new_context(rp->ai_addr, rp->ai_addrlen); if (ctx) { if ( rp->ai_family == PF_INET6 ) { struct sockaddr_in6* pSadrIn6 = (struct sockaddr_in6*) rp->ai_addr; if ( pSadrIn6->sin6_scope_id == 0 ) { pSadrIn6->sin6_scope_id = scopeId; } /* End IF the scope ID wasn't set. */ } goto finish; } } fprintf(stderr, "no context available for interface '%s' ", node); finish: freeaddrinfo(result); return ctx; }

设备 tap0 有以下详细信息:

Device tap0 has following details:

tap0 Link encap:Ethernet HWaddr ce:23:fc:81:7f:65 inet6 addr: fe80::cc23:fcff:fe81:7f65/64 Scope:Link inet6 addr: aaaa::1/64 Scope:Global

我使用以下命令运行上述代码:

I run the above code with the following command:

./server -A aaaa::1

但它无法在 tap0 上监听到地址 aaaa::1 的消息.我可以在 tap0 接口上的 wireshark 转储上看到数据包.有趣的是，从上述命令(和代码)运行的服务器可以通过 localhost 接收消息.因此，如果我执行以下命令，服务器会收到消息:

But it's not able to listen to the messages coming to the address aaaa::1 on tap0. I can see the packets on the wireshark dump on tap0 interface. Interestingly, the server ran from the above command(and code) can receive the messages through the localhost. So, if I execute the following command, the server receives the messages:

nc -6 -vvv -u aaaa::1 61616 < /tmp/send_to_node_raw

上面的命令发送和接收是通过localhost完成的.有什么方法可以在辅助接口上以编程方式接收 UDP/IPv6 消息?

The above command sending and reception is done through localhost. Is there any way I can programatically receive UDP/IPv6 messages on secondary interfaces?

推荐答案

问题一定出在其他地方.使用上面的代码，我可以分别使用 aaaa::1/64 和 aaaa::2/64 作为本地和远程地址成功地接收来自另一个节点的数据包.

The problem must be somewhere else. Using the code above I can successfully receive packets from another node using aaaa::1/64 and aaaa::2/64 as the local and remote addresses, respectively.

尝试绑定到 :: 并查看它是否以这种方式工作.也尝试使用 netcat6 绑定到 aaaa::1 并从另一个 netcat6 实例接收.

Try to bind to :: and see if it works that way. Also try to use netcat6 to bind to aaaa::1 and receive from another netcat6 instance.