TCP发送数据包后，会收到对端的ACK。通过处理ACK，TCP可以进行拥塞控制和流控制，所以

ACK的处理是TCP的一个重要内容。tcp_ack()用于处理接收到的ACK。

本文主要内容：TCP接收ACK处理，tcp_ack()的实现。

内核版本：3.2.12

Author：zhangskd @ csdn

基础

在我们开始探究tcp_ack()的处理流程前，不妨先来回顾一些即将涉及到的概念和数据结构。

struct tcp_sock {.../* Packets which are "in flight"，发送且未确认的数据包个数*/u32 packets_out ; /* Retransmitted packets out ，重传的且未确认数据包个数 */u32 retrans_out ;/* 被SACKED数据段的个数 */u32 sacked_out ;/* 丢失的数据包个数 */u32 loss_out ;/* FACK'd packets*/u32 fackets_out;/* snd_nxt at onset of  congestion */u32 high_seq;u32 bytes_acked; /* Appropriate Byte Counting - RFC3465*/u32 prior_ssthresh; /* ssthresh saved at recovery start */u8 reordering; /* Packet reordering metric. *//* Options received (usually on last paket, some only on SYN packets).*/struct tcp_options_received rx_opt;u32 snd_wl1; /* Sequence for window update，记录更新发送窗口的ACK段序号*/...
}

TCP的控制信息块：

/* This is what the send packet queuing engine uses to pass * TCP per-packet control information to the transmission code.* We also store the host-order sequence numbers in here too.* This is 44 bytes if IPv6 is enabled.* If this grows please adjust sk_buff.h : skbuff->cb[xxx] size appropriately.*/
struct tcp_skb_cb {union {struct inet_skb_parm h4 ;
#if defined (CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE)struct inet6_skb_parm h6 ;
#endif} header ; /* For incoming frames ，主要是ip头信息*/__u32 seq ; /* Starting sequence number */__u32 end_seq ; /* SEQ + FIN + SYN + datalen */__u32 when ; /* used to compute rtt's */__u8 flags ; /* TCP header flags 是tcp头的标志位 */__u8 sacked ; /* State flags for SACK/FACK */#define TCPCB_SACKED_ACKED 0x01 /* SKB ACK'd by a SACK block */
#define TCPCB_SACKED_RETRANS 0x02 /* SKB retransmitted */
#define TCPCB_LOST 0x04 /* SKB is lost */
#define TCPCB_TAGBITS 0x07 /* All tag bits */
#define TCPCB_EVER_RETRANS 0x80 /* Ever retransmitted frame */
#define TCPCB_RETRANS (TCP_SACKED_RETRANS | TCP_EVER_RETRANS)__u32 ack_seq ; /* Sequence number ACK'd */
}#define TCP_SKB_CB(__skb)   ((struct tcp_skb_cb *) &((__skb)->cb[0]))

拥塞控制用到的变量：

/* 已发送且还没被ACK的数据包个数 */

packets_out ;

/* 重传且未未被ACK的数据包个数 */

retrans_out ;

/* 被SACKED的数据包个数 */

sacked_out ;

/* 丢失的数据包个数，这是个猜测值 */

loss_out ;

left_out = sacked_out + loss_out ; /* 表示离开网络且未被确认的数据包个数*/

in_flight = packets_out + retrans_out - left_out ; /* 表示处于网络中，还未被确认的数据包个数 */

/* 离开网络且未被确认的数据包个数 */

static inline unsigned int tcp_left_out(const struct tcp_sock *tp)
{return tp->sacked_out + tp->loss_out ;
}

以下是这些变量的详细介绍：

packets_out is the number of originally transmitted segments above snd.una.

sacked_out is the number of segments acknowledged by SACK blocks.

lost_out is an estimation of the number of segments lost in the network.

retrans_out is the number of retransmitted segments.

Determining the lost_out parameter depends on the selected recovery method.

For example, when FACK is in use, all unacknowledged segments between the highest SACK block

and the cumulative acknowledgement are counted in lost_out.

函数实现

tcp_ack()用于处理接收到的带有ACK标志的段。

/* This routine deals with incoming acks, but not outgoing ones.*/
static int tcp_ack (struct sock *sk, const struct sk_buff *skb, int flag)
{struct inet_connection_sock *icsk = inet_csk(sk);struct tcp_sock *tp = tcp_sk(sk);u32 prior_snd_una = tp->snd_una; /* 此ACK之前的snd_una */u32 ack_seq = TCP_SKB_CB(skb)->seq; /* 此ACK的开始序号 */u32 ack = TCP_SKB_CB(skb)->ack_seq; /* 此ACK的确认序号 */u32 prior_in_flight;u32 prior_fackets;int prior_packets;int prior_sacked = tp->sacked_out;int newly_acked_sacked = 0;int frto_cwnd = 0;/* If the ack is older than previous acks* then we can probably ignore it.* 有负载的TCP段会顺便携带一个ACK序号，即使这个序号已经确认过。* 如果段中没有SACK选项，就不用处理了。*/if (before(ack, prior_snd_una))goto old_ack;/* If the ack includes data we haven't sent yet, discard* this segment (RFC793 Section 3.9).* 我们还没发送的段都已经说确认了，这个ACK是错误的。*/if (after(ack, tp->snd_nxt))goto invalid_ack;/* 此ACK确认了新的数据*/if (after(ack, prior_snd_una))flag |= FLAG_SND_UNA_ADVANCED;/* tcp_abc选项处理，累加这个ACK确认的字节数 */if (sysctl_tcp_abc) {if (icsk->icsk_ca_state < TCP_CA_CWR)tp->bytes_acked += ack - prior_snd_una; /* 累积确认了多少字节*/else if (icsk->icsk_ca_state == TCP_CA_Loss)/* we assume just one segment left network */tp->bytes_acked += min(ack-prior_snd_una, tp->mss_cache);}prior_fackets = tp->fackets_out;prior_in_flight = tcp_packets_in_flight(tp);/* 如果处于快速路径中*/if (! (flag & FLAG_SLOWPATH) && after(ack, prior_snd_una)) {/* Window is constant, pure forward advance.* No more checks are required.*/tcp_update_wl(tp, ack_seq); /*记录更新发送窗口的ACK段序号*/tp->snd_una = ack; /* 更新发送窗口左端*/flag |= FLAG_WIN_UPDATE; /* 设置发送窗口更新标志*/tcp_ca_event(sk, CA_EVENT_FAST_ACK); /* 快速路径拥塞事件钩子*/NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCHHPACKS);} else { /* 进入慢速路径 */if (ack_seq != TCP_SKB_CB(skb)->end_seq)flag |= FLAG_DATA; /* 此ACK携带负荷*/else /* 纯ACK*/NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPPUREACKS);flag |= tcp_ack_update_window(sk, skb, ack, ack_seq); /* 更新发送窗口*//* 根据SACK选项标志重传队列中SKB的记分牌状态*/if (TCP_SKB_CB(skb)->sacked)flag |= tcp_sacktag_write_queue(sk, skb, prior_snd_una);/* 查看ACK是否携带ECE标志 */if (TCP_ECN_rcv_ecn_echo(tp, tcp_hdr(skb)))flag |= FLAG_ECE;tcp_ca_event(sk, CA_EVENT_SLOW_ACK); /* 慢速路径拥塞事件钩子*/}/* We pass data and got it acked, remove any soft err log. Something worked...*/sk->sk_err_soft = 0;icsk->icsk_probes_out = 0;tp->rcv_tstamp = tcp_time_stamp; /*ts of last received ACK (for keepalives)*/prior_packets = tp->packets_out;if (! prior_packets) /*检查是否有发送且未确认的段*/goto no_queue; /* 持续定时器处理*//* See if we can take anything off the retransmit queue.* 删除重传队列中已经确认的数据段，并进行时延采样。*/flag |= tcp_clean_rtx_queue(sk, prior_fackets, prior_snd_una);/* 经过tcp_clean_rtx_queue之后，* 计算此次ACKED和SACKED的数据包个数* /newly_acked_sacked = (prior_packets - prior_sacked) - (tp->packets_out - tp->sacked_out) ;/* 处于FRTO处理中，frto_counter的取值为1或2*/if (tp->frto_counter)frto_cwnd = tcp_process_frto(sk, flag); /* 判断RTO是否为真的*//* Guarantee sacktag reordering detection against wrap-arounds */if (before(tp->frto_highmark, tp->snd_una))tp->frto_highmark = 0;/* 如果ACK是重复的、或者带有SACK选项、或者不是Open态*/if (tcp_ack_is_dubious(sk, flag)) { /* 如果此ACK是可疑的*//* 可疑是指已经处于拥塞状态，或刚收到拥塞信号。* 在这种条件下如果想进行拥塞避免，必须符合：* 1. 此ACK确认了新的数据* 2. 不能处于FRTO状态* 3. 处于Disorder或Loss状态*/if ((flag & FLAG_DATA_ACKED) && ! frto_cwnd && tcp_may_raise_cwnd(sk, flag))tcp_cong_avoid(sk, ack, prior_in_flight); /* 拥塞窗口的调节*//* 这里进入TCP的拥塞状态机，处理相关拥塞状态*/tcp_fastretrans_alert(sk, prior_packets - tp->packets_out,newly_acked_sacked, flag);} else { /* 至少说明ACK在Open态*//* ACK不是可疑的，如果ACK确认了新的数据，且不是frto，则进行拥塞避免*/if ( (flag & FLAG_DATA_ACKED) && ! frto_cwnd )tcp_cong_avoid(sk, ack, prior_in_flight); /* 拥塞窗口的调节*/}/* 如果ACK确认了新的段(新的数据段、SYN段、SACK段，或者接收到的ACK不是重复的，* 则确认该传输控制块的输出路由缓存项是有效的。*/if ((flag & FLAG_FORWARD_PROGRESS) || ! (flag & FLAG_NOT_DUP))dst_confirm(__sk_dst_get(sk));return 1;no_queue:/* If this ack opens up a zero window, clear backoff.* It was being used to time the probes, and is probably far higher than it* needs to be for normal retransmissions.*/if (tcp_send_head(sk)) /* 如果有数据要发送*/tcp_ack_probe(sk); /* 持续定时器处理*/return 1;invalid_ack:SOCK_DEBUG(sk, "ACK %u after %u:%u\n", ack, tp->snd_una, tp->snd_nxt);return -1;old_ack:/* 如果此ACK已经确认过，且带有SACK选项的信息*/if (TCP_SKB_CB(skb)->sacked) {/* 重新标志各个段的记分牌*/tcp_sacktag_write_queue(sk, skb, prior_snd_una);if (icsk->icsk_ca_state == TCP_CA_Open)tcp_try_keep_open(sk); /* 实际上是看看要不要进入Disorder状态*/}/* 会自动打印出调试信息*/SOCK_DEBUG(sk, "ACK %u before %u:%u\n", ack, tp->snd_una, tp->snd_nxt);
}

old_ack

Q: 什么条件下进入Disorder状态？

A: 如果检测到有被sacked的数据包，或者有重传的数据包，则进入Disorder状态。

当然，之前的状态必须为Open态。

判断条件：sacked_out、lost_out、retrans_out、undo_marker有不为0的。

static void tcp_try_keep_open(struct sock *sk)
{struct tcp_sock *tp = tcp_sk(sk);int state = TCP_CA_Open;/* 如果以下条件成立，那么就从Open进入Disorder状态*/if (tcp_left_out(tp) || tcp_any_retrans_done(sk) || tp->undo_marker)state = TCP_CA_Disorder;if (inet_csk(sk)->icsk_ca_state != state) {tcp_set_ca_state(sk, state);tp->high_seq = tp->snd_nxt;}
}#define SOCK_DEBUG(sk, msg...) do { if ((sk) && sock_flag((sk), SOCK_DBG)) \printk(KERN_DEBUG msg);} while(0)

fast path

记录更新发送窗口的ACK段序号：

static inline void tcp_update_wl(struct tcp_sock *tp, u32 seq)
{tp->snd_wl1 = seq;
}

ECN

看看是否收到了显示拥塞通知：

static inline int TCP_ECN_rcv_ecn_echo(struct tcp_sock *tp, const struct tcphdr *th)
{if (th->ece && ! th->syn && (tp->ecn_flags & TCP_ECN_OK))return 1;return 0;
}