之前在《深入理解并行编程》的群里,有个小米的兄弟问了一个问题,服务器A发包给服务器B,Seq是1,但是在未能收到服务器B的报文回复的情况下,发送了
rst,但是rst报文中,对应的seq是1461,一堆人都在猜测,为什么seq跳变了,由于当时只看到一半的图片,所以我让他发送完整报文出来之后,我
发现其实rst的seq不是1的原因,并不是因为跳变,而是正常的,因为发送给B的报文,长度为1460,但是这个报文没有得到回复,所以在超时之后,应用程序关闭了这条连接,
导致内核协议栈发送了一个rst报文,而rst报文选取seq的时候,并不是选取的确定已经发送的seq,而是当前连接已经用掉的seq,也就是当前seq,哪怕这个报文没有收到回复,也会使用。
具体看代码:
/* We get here when a process closes a file descriptor (either due to
* an explicit close() or as a byproduct of exit()'ing) and there
* was unread data in the receive queue. This behavior is recommended
* by RFC 2525, section 2.17. -DaveM
*/
void tcp_send_active_reset(struct sock *sk, gfp_t priority)
{
struct sk_buff *skb;
/* NOTE: No TCP options attached and we never retransmit this. */
skb = alloc_skb(MAX_TCP_HEADER, priority);
if (!skb) {
NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTFAILED);
return;
}
/* Reserve space for headers and prepare control bits. */
skb_reserve(skb, MAX_TCP_HEADER);
tcp_init_nondata_skb(skb, tcp_acceptable_seq(sk),
TCPHDR_ACK | TCPHDR_RST);//注意传入的标志是rst,不是fin,可以具体参考tcp_send_fin 是怎么传参数的
/* Send it off. */ if (tcp_transmit_skb(sk, skb, 0, priority)) NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPABORTFAILED); TCP_INC_STATS(sock_net(sk), TCP_MIB_OUTRSTS); }
其中关注下报文的init过程:
/* Constructs common control bits of non-data skb. If SYN/FIN is present, * auto increment end seqno. */ static void tcp_init_nondata_skb(struct sk_buff *skb, u32 seq, u8 flags) { struct skb_shared_info *shinfo = skb_shinfo(skb); skb->ip_summed = CHECKSUM_PARTIAL; skb->csum = 0; TCP_SKB_CB(skb)->tcp_flags = flags; TCP_SKB_CB(skb)->sacked = 0; shinfo->gso_segs = 1; shinfo->gso_size = 0; shinfo->gso_type = 0; TCP_SKB_CB(skb)->seq = seq; if (flags & (TCPHDR_SYN | TCPHDR_FIN))//我们本文的标志是 TCPHDR_ACK | TCPHDR_RST ,
seq++;//此处+1 ,但进不来
TCP_SKB_CB(skb)->end_seq = seq;//所以本文应该是传入seq是多少就发送多少 }
那么传入的seq是多少呢?
/* SND.NXT, if window was not shrunk. * If window has been shrunk, what should we make? It is not clear at all. * Using SND.UNA we will fail to open window, SND.NXT is out of window. :-( * Anything in between SND.UNA...SND.UNA+SND.WND also can be already * invalid. OK, let's make this for now: */ static inline __u32 tcp_acceptable_seq(const struct sock *sk) { const struct tcp_sock *tp = tcp_sk(sk); if (!before(tcp_wnd_end(tp), tp->snd_nxt)) return tp->snd_nxt; else return tcp_wnd_end(tp); }
注释写得比较清楚,如果窗口没有shrunk,也就是tp->snd_nxt 没有out of window 的话,则取得就是tp->snd_nxt,而这个值,就是报文长度+1了,也就是1461.
如果不是rst的方式结束,而是fin的方式结束,那么这个seq则应该为多少呢?
我们来关注下 tcp_send_fin 函数,看看它怎么使用seq的,
void tcp_send_fin(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *skb = tcp_write_queue_tail(sk); int mss_now; /* Optimization, tack on the FIN if we have a queue of * unsent frames. But be careful about outgoing SACKS * and IP options. */ mss_now = tcp_current_mss(sk); if (tcp_send_head(sk) != NULL) {//说明还有空间,tcp_send_head返回值为sk->sk_send_head, TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_FIN; TCP_SKB_CB(skb)->end_seq++; tp->write_seq++; } else { /* Socket is locked, keep trying until memory is available. */ for (;;) { skb = alloc_skb_fclone(MAX_TCP_HEADER, sk->sk_allocation);//申请一个skb if (skb) break; yield(); } /* Reserve space for headers and prepare control bits. */ skb_reserve(skb, MAX_TCP_HEADER); /* FIN eats a sequence byte, write_seq advanced by tcp_queue_skb(). */ tcp_init_nondata_skb(skb, tp->write_seq, TCPHDR_ACK | TCPHDR_FIN);//fin包占用一个seq tcp_queue_skb(sk, skb); } __tcp_push_pending_frames(sk, mss_now, TCP_NAGLE_OFF); }
不管有没有空间,其实发送出去的序列号就是当前报文的下一个,因为fin也占用一个seq,所以这个seq也是上次发完的seq+报文长度+1.