diff options
-rw-r--r-- | include/net/tcp.h | 25 | ||||
-rw-r--r-- | include/net/tcp_ecn.h | 11 | ||||
-rw-r--r-- | net/ipv4/tcp.c | 39 | ||||
-rw-r--r-- | net/ipv4/tcp_input.c | 145 | ||||
-rw-r--r-- | net/ipv4/tcp_output.c | 54 |
5 files changed, 158 insertions, 116 deletions
diff --git a/include/net/tcp.h b/include/net/tcp.h index e798033..43910fe 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -420,9 +420,9 @@ extern __u32 cookie_v4_init_sequence(struct sock *sk, struct sk_buff *skb, /* tcp_output.c */ -extern void __tcp_push_pending_frames(struct sock *sk, struct tcp_sock *tp, - unsigned int cur_mss, int nonagle); -extern int tcp_may_send_now(struct sock *sk, struct tcp_sock *tp); +extern void __tcp_push_pending_frames(struct sock *sk, unsigned int cur_mss, + int nonagle); +extern int tcp_may_send_now(struct sock *sk); extern int tcp_retransmit_skb(struct sock *, struct sk_buff *); extern void tcp_xmit_retransmit_queue(struct sock *); extern void tcp_simple_retransmit(struct sock *); @@ -479,8 +479,10 @@ static inline void tcp_fast_path_on(struct tcp_sock *tp) __tcp_fast_path_on(tp, tp->snd_wnd >> tp->rx_opt.snd_wscale); } -static inline void tcp_fast_path_check(struct sock *sk, struct tcp_sock *tp) +static inline void tcp_fast_path_check(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); + if (skb_queue_empty(&tp->out_of_order_queue) && tp->rcv_wnd && atomic_read(&sk->sk_rmem_alloc) < sk->sk_rcvbuf && @@ -591,10 +593,10 @@ static inline void tcp_dec_pcount_approx(__u32 *count, } } -static inline void tcp_packets_out_inc(struct sock *sk, - struct tcp_sock *tp, +static inline void tcp_packets_out_inc(struct sock *sk, const struct sk_buff *skb) { + struct tcp_sock *tp = tcp_sk(sk); int orig = tp->packets_out; tp->packets_out += tcp_skb_pcount(skb); @@ -778,18 +780,21 @@ static inline void tcp_minshall_update(struct tcp_sock *tp, int mss, tp->snd_sml = TCP_SKB_CB(skb)->end_seq; } -static inline void tcp_check_probe_timer(struct sock *sk, struct tcp_sock *tp) +static inline void tcp_check_probe_timer(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); const struct inet_connection_sock *icsk = inet_csk(sk); + if (!tp->packets_out && !icsk->icsk_pending) inet_csk_reset_xmit_timer(sk, ICSK_TIME_PROBE0, icsk->icsk_rto, TCP_RTO_MAX); } -static inline void tcp_push_pending_frames(struct sock *sk, - struct tcp_sock *tp) +static inline void tcp_push_pending_frames(struct sock *sk) { - __tcp_push_pending_frames(sk, tp, tcp_current_mss(sk, 1), tp->nonagle); + struct tcp_sock *tp = tcp_sk(sk); + + __tcp_push_pending_frames(sk, tcp_current_mss(sk, 1), tp->nonagle); } static inline void tcp_init_wl(struct tcp_sock *tp, u32 ack, u32 seq) diff --git a/include/net/tcp_ecn.h b/include/net/tcp_ecn.h index b5f7c6a..89eb3e0 100644 --- a/include/net/tcp_ecn.h +++ b/include/net/tcp_ecn.h @@ -27,9 +27,10 @@ static inline void TCP_ECN_send_synack(struct tcp_sock *tp, TCP_SKB_CB(skb)->flags &= ~TCPCB_FLAG_ECE; } -static inline void TCP_ECN_send_syn(struct sock *sk, struct tcp_sock *tp, - struct sk_buff *skb) +static inline void TCP_ECN_send_syn(struct sock *sk, struct sk_buff *skb) { + struct tcp_sock *tp = tcp_sk(sk); + tp->ecn_flags = 0; if (sysctl_tcp_ecn) { TCP_SKB_CB(skb)->flags |= TCPCB_FLAG_ECE|TCPCB_FLAG_CWR; @@ -44,9 +45,11 @@ TCP_ECN_make_synack(struct request_sock *req, struct tcphdr *th) th->ece = 1; } -static inline void TCP_ECN_send(struct sock *sk, struct tcp_sock *tp, - struct sk_buff *skb, int tcp_header_len) +static inline void TCP_ECN_send(struct sock *sk, struct sk_buff *skb, + int tcp_header_len) { + struct tcp_sock *tp = tcp_sk(sk); + if (tp->ecn_flags & TCP_ECN_OK) { /* Not-retransmitted data segment: set ECT and inject CWR. */ if (skb->len != tcp_header_len && diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 99ad52c..2cf9a89 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -460,9 +460,9 @@ static inline int forced_push(struct tcp_sock *tp) return after(tp->write_seq, tp->pushed_seq + (tp->max_window >> 1)); } -static inline void skb_entail(struct sock *sk, struct tcp_sock *tp, - struct sk_buff *skb) +static inline void skb_entail(struct sock *sk, struct sk_buff *skb) { + struct tcp_sock *tp = tcp_sk(sk); struct tcp_skb_cb *tcb = TCP_SKB_CB(skb); skb->csum = 0; @@ -486,15 +486,17 @@ static inline void tcp_mark_urg(struct tcp_sock *tp, int flags, } } -static inline void tcp_push(struct sock *sk, struct tcp_sock *tp, int flags, - int mss_now, int nonagle) +static inline void tcp_push(struct sock *sk, int flags, int mss_now, + int nonagle) { + struct tcp_sock *tp = tcp_sk(sk); + if (tcp_send_head(sk)) { struct sk_buff *skb = tcp_write_queue_tail(sk); if (!(flags & MSG_MORE) || forced_push(tp)) tcp_mark_push(tp, skb); tcp_mark_urg(tp, flags, skb); - __tcp_push_pending_frames(sk, tp, mss_now, + __tcp_push_pending_frames(sk, mss_now, (flags & MSG_MORE) ? TCP_NAGLE_CORK : nonagle); } } @@ -540,7 +542,7 @@ new_segment: if (!skb) goto wait_for_memory; - skb_entail(sk, tp, skb); + skb_entail(sk, skb); copy = size_goal; } @@ -586,7 +588,7 @@ new_segment: if (forced_push(tp)) { tcp_mark_push(tp, skb); - __tcp_push_pending_frames(sk, tp, mss_now, TCP_NAGLE_PUSH); + __tcp_push_pending_frames(sk, mss_now, TCP_NAGLE_PUSH); } else if (skb == tcp_send_head(sk)) tcp_push_one(sk, mss_now); continue; @@ -595,7 +597,7 @@ wait_for_sndbuf: set_bit(SOCK_NOSPACE, &sk->sk_socket->flags); wait_for_memory: if (copied) - tcp_push(sk, tp, flags & ~MSG_MORE, mss_now, TCP_NAGLE_PUSH); + tcp_push(sk, flags & ~MSG_MORE, mss_now, TCP_NAGLE_PUSH); if ((err = sk_stream_wait_memory(sk, &timeo)) != 0) goto do_error; @@ -606,7 +608,7 @@ wait_for_memory: out: if (copied) - tcp_push(sk, tp, flags, mss_now, tp->nonagle); + tcp_push(sk, flags, mss_now, tp->nonagle); return copied; do_error: @@ -637,8 +639,9 @@ ssize_t tcp_sendpage(struct socket *sock, struct page *page, int offset, #define TCP_PAGE(sk) (sk->sk_sndmsg_page) #define TCP_OFF(sk) (sk->sk_sndmsg_off) -static inline int select_size(struct sock *sk, struct tcp_sock *tp) +static inline int select_size(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); int tmp = tp->mss_cache; if (sk->sk_route_caps & NETIF_F_SG) { @@ -714,7 +717,7 @@ new_segment: if (!sk_stream_memory_free(sk)) goto wait_for_sndbuf; - skb = sk_stream_alloc_pskb(sk, select_size(sk, tp), + skb = sk_stream_alloc_pskb(sk, select_size(sk), 0, sk->sk_allocation); if (!skb) goto wait_for_memory; @@ -725,7 +728,7 @@ new_segment: if (sk->sk_route_caps & NETIF_F_ALL_CSUM) skb->ip_summed = CHECKSUM_PARTIAL; - skb_entail(sk, tp, skb); + skb_entail(sk, skb); copy = size_goal; } @@ -830,7 +833,7 @@ new_segment: if (forced_push(tp)) { tcp_mark_push(tp, skb); - __tcp_push_pending_frames(sk, tp, mss_now, TCP_NAGLE_PUSH); + __tcp_push_pending_frames(sk, mss_now, TCP_NAGLE_PUSH); } else if (skb == tcp_send_head(sk)) tcp_push_one(sk, mss_now); continue; @@ -839,7 +842,7 @@ wait_for_sndbuf: set_bit(SOCK_NOSPACE, &sk->sk_socket->flags); wait_for_memory: if (copied) - tcp_push(sk, tp, flags & ~MSG_MORE, mss_now, TCP_NAGLE_PUSH); + tcp_push(sk, flags & ~MSG_MORE, mss_now, TCP_NAGLE_PUSH); if ((err = sk_stream_wait_memory(sk, &timeo)) != 0) goto do_error; @@ -851,7 +854,7 @@ wait_for_memory: out: if (copied) - tcp_push(sk, tp, flags, mss_now, tp->nonagle); + tcp_push(sk, flags, mss_now, tp->nonagle); TCP_CHECK_TIMER(sk); release_sock(sk); return copied; @@ -1389,7 +1392,7 @@ do_prequeue: skip_copy: if (tp->urg_data && after(tp->copied_seq, tp->urg_seq)) { tp->urg_data = 0; - tcp_fast_path_check(sk, tp); + tcp_fast_path_check(sk); } if (used + offset < skb->len) continue; @@ -1830,7 +1833,7 @@ static int do_tcp_setsockopt(struct sock *sk, int level, * for currently queued segments. */ tp->nonagle |= TCP_NAGLE_OFF|TCP_NAGLE_PUSH; - tcp_push_pending_frames(sk, tp); + tcp_push_pending_frames(sk); } else { tp->nonagle &= ~TCP_NAGLE_OFF; } @@ -1854,7 +1857,7 @@ static int do_tcp_setsockopt(struct sock *sk, int level, tp->nonagle &= ~TCP_NAGLE_CORK; if (tp->nonagle&TCP_NAGLE_OFF) tp->nonagle |= TCP_NAGLE_PUSH; - tcp_push_pending_frames(sk, tp); + tcp_push_pending_frames(sk); } break; diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 2fbfc2e..6333893 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -235,9 +235,9 @@ static void tcp_fixup_sndbuf(struct sock *sk) */ /* Slow part of check#2. */ -static int __tcp_grow_window(const struct sock *sk, struct tcp_sock *tp, - const struct sk_buff *skb) +static int __tcp_grow_window(const struct sock *sk, const struct sk_buff *skb) { + struct tcp_sock *tp = tcp_sk(sk); /* Optimize this! */ int truesize = tcp_win_from_space(skb->truesize)/2; int window = tcp_win_from_space(sysctl_tcp_rmem[2])/2; @@ -252,9 +252,11 @@ static int __tcp_grow_window(const struct sock *sk, struct tcp_sock *tp, return 0; } -static void tcp_grow_window(struct sock *sk, struct tcp_sock *tp, +static void tcp_grow_window(struct sock *sk, struct sk_buff *skb) { + struct tcp_sock *tp = tcp_sk(sk); + /* Check #1 */ if (tp->rcv_ssthresh < tp->window_clamp && (int)tp->rcv_ssthresh < tcp_space(sk) && @@ -267,7 +269,7 @@ static void tcp_grow_window(struct sock *sk, struct tcp_sock *tp, if (tcp_win_from_space(skb->truesize) <= skb->len) incr = 2*tp->advmss; else - incr = __tcp_grow_window(sk, tp, skb); + incr = __tcp_grow_window(sk, skb); if (incr) { tp->rcv_ssthresh = min(tp->rcv_ssthresh + incr, tp->window_clamp); @@ -330,8 +332,9 @@ static void tcp_init_buffer_space(struct sock *sk) } /* 5. Recalculate window clamp after socket hit its memory bounds. */ -static void tcp_clamp_window(struct sock *sk, struct tcp_sock *tp) +static void tcp_clamp_window(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); struct inet_connection_sock *icsk = inet_csk(sk); icsk->icsk_ack.quick = 0; @@ -503,8 +506,9 @@ new_measure: * each ACK we send, he increments snd_cwnd and transmits more of his * queue. -DaveM */ -static void tcp_event_data_recv(struct sock *sk, struct tcp_sock *tp, struct sk_buff *skb) +static void tcp_event_data_recv(struct sock *sk, struct sk_buff *skb) { + struct tcp_sock *tp = tcp_sk(sk); struct inet_connection_sock *icsk = inet_csk(sk); u32 now; @@ -545,7 +549,7 @@ static void tcp_event_data_recv(struct sock *sk, struct tcp_sock *tp, struct sk_ TCP_ECN_check_ce(tp, skb); if (skb->len >= 128) - tcp_grow_window(sk, tp, skb); + tcp_grow_window(sk, skb); } /* Called to compute a smoothed rtt estimate. The data fed to this @@ -1541,8 +1545,10 @@ static inline int tcp_skb_timedout(struct sock *sk, struct sk_buff *skb) return (tcp_time_stamp - TCP_SKB_CB(skb)->when > inet_csk(sk)->icsk_rto); } -static inline int tcp_head_timedout(struct sock *sk, struct tcp_sock *tp) +static inline int tcp_head_timedout(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); + return tp->packets_out && tcp_skb_timedout(sk, tcp_write_queue_head(sk)); } @@ -1640,8 +1646,9 @@ static inline int tcp_head_timedout(struct sock *sk, struct tcp_sock *tp) * Main question: may we further continue forward transmission * with the same cwnd? */ -static int tcp_time_to_recover(struct sock *sk, struct tcp_sock *tp) +static int tcp_time_to_recover(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); __u32 packets_out; /* Do not perform any recovery during FRTO algorithm */ @@ -1659,7 +1666,7 @@ static int tcp_time_to_recover(struct sock *sk, struct tcp_sock *tp) /* Trick#3 : when we use RFC2988 timer restart, fast * retransmit can be triggered by timeout of queue head. */ - if (tcp_head_timedout(sk, tp)) + if (tcp_head_timedout(sk)) return 1; /* Trick#4: It is still not OK... But will it be useful to delay @@ -1668,7 +1675,7 @@ static int tcp_time_to_recover(struct sock *sk, struct tcp_sock *tp) packets_out = tp->packets_out; if (packets_out <= tp->reordering && tp->sacked_out >= max_t(__u32, packets_out/2, sysctl_tcp_reordering) && - !tcp_may_send_now(sk, tp)) { + !tcp_may_send_now(sk)) { /* We have nothing to send. This connection is limited * either by receiver window or by application. */ @@ -1708,8 +1715,10 @@ static void tcp_add_reno_sack(struct sock *sk) /* Account for ACK, ACKing some data in Reno Recovery phase. */ -static void tcp_remove_reno_sacks(struct sock *sk, struct tcp_sock *tp, int acked) +static void tcp_remove_reno_sacks(struct sock *sk, int acked) { + struct tcp_sock *tp = tcp_sk(sk); + if (acked > 0) { /* One ACK acked hole. The rest eat duplicate ACKs. */ if (acked-1 >= tp->sacked_out) @@ -1728,9 +1737,10 @@ static inline void tcp_reset_reno_sack(struct tcp_sock *tp) } /* Mark head of queue up as lost. */ -static void tcp_mark_head_lost(struct sock *sk, struct tcp_sock *tp, +static void tcp_mark_head_lost(struct sock *sk, int packets, u32 high_seq) { + struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *skb; int cnt; @@ -1771,15 +1781,17 @@ static void tcp_mark_head_lost(struct sock *sk, struct tcp_sock *tp, /* Account newly detected lost packet(s) */ -static void tcp_update_scoreboard(struct sock *sk, struct tcp_sock *tp) +static void tcp_update_scoreboard(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); + if (IsFack(tp)) { int lost = tp->fackets_out - tp->reordering; if (lost <= 0) lost = 1; - tcp_mark_head_lost(sk, tp, lost, tp->high_seq); + tcp_mark_head_lost(sk, lost, tp->high_seq); } else { - tcp_mark_head_lost(sk, tp, 1, tp->high_seq); + tcp_mark_head_lost(sk, 1, tp->high_seq); } /* New heuristics: it is possible only after we switched @@ -1787,7 +1799,7 @@ static void tcp_update_scoreboard(struct sock *sk, struct tcp_sock *tp) * Hence, we can detect timed out packets during fast * retransmit without falling to slow start. */ - if (!IsReno(tp) && tcp_head_timedout(sk, tp)) { + if (!IsReno(tp) && tcp_head_timedout(sk)) { struct sk_buff *skb; skb = tp->scoreboard_skb_hint ? tp->scoreboard_skb_hint @@ -1867,9 +1879,11 @@ static inline int tcp_packet_delayed(struct tcp_sock *tp) /* Undo procedures. */ #if FASTRETRANS_DEBUG > 1 -static void DBGUNDO(struct sock *sk, struct tcp_sock *tp, const char *msg) +static void DBGUNDO(struct sock *sk, const char *msg) { + struct tcp_sock *tp = tcp_sk(sk); struct inet_sock *inet = inet_sk(sk); + printk(KERN_DEBUG "Undo %s %u.%u.%u.%u/%u c%u l%u ss%u/%u p%u\n", msg, NIPQUAD(inet->daddr), ntohs(inet->dport), @@ -1915,13 +1929,15 @@ static inline int tcp_may_undo(struct tcp_sock *tp) } /* People celebrate: "We love our President!" */ -static int tcp_try_undo_recovery(struct sock *sk, struct tcp_sock *tp) +static int tcp_try_undo_recovery(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); + if (tcp_may_undo(tp)) { /* Happy end! We did not retransmit anything * or our original transmission succeeded. */ - DBGUNDO(sk, tp, inet_csk(sk)->icsk_ca_state == TCP_CA_Loss ? "loss" : "retrans"); + DBGUNDO(sk, inet_csk(sk)->icsk_ca_state == TCP_CA_Loss ? "loss" : "retrans"); tcp_undo_cwr(sk, 1); if (inet_csk(sk)->icsk_ca_state == TCP_CA_Loss) NET_INC_STATS_BH(LINUX_MIB_TCPLOSSUNDO); @@ -1941,10 +1957,12 @@ static int tcp_try_undo_recovery(struct sock *sk, struct tcp_sock *tp) } /* Try to undo cwnd reduction, because D-SACKs acked all retransmitted data */ -static void tcp_try_undo_dsack(struct sock *sk, struct tcp_sock *tp) +static void tcp_try_undo_dsack(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); + if (tp->undo_marker && !tp->undo_retrans) { - DBGUNDO(sk, tp, "D-SACK"); + DBGUNDO(sk, "D-SACK"); tcp_undo_cwr(sk, 1); tp->undo_marker = 0; NET_INC_STATS_BH(LINUX_MIB_TCPDSACKUNDO); @@ -1953,9 +1971,9 @@ static void tcp_try_undo_dsack(struct sock *sk, struct tcp_sock *tp) /* Undo during fast recovery after partial ACK. */ -static int tcp_try_undo_partial(struct sock *sk, struct tcp_sock *tp, - int acked) +static int tcp_try_undo_partial(struct sock *sk, int acked) { + struct tcp_sock *tp = tcp_sk(sk); /* Partial ACK arrived. Force Hoe's retransmit. */ int failed = IsReno(tp) || tp->fackets_out>tp->reordering; @@ -1968,7 +1986,7 @@ static int tcp_try_undo_partial(struct sock *sk, struct tcp_sock *tp, tcp_update_reordering(sk, tcp_fackets_out(tp) + acked, 1); - DBGUNDO(sk, tp, "Hoe"); + DBGUNDO(sk, "Hoe"); tcp_undo_cwr(sk, 0); NET_INC_STATS_BH(LINUX_MIB_TCPPARTIALUNDO); @@ -1982,8 +2000,10 @@ static int tcp_try_undo_partial(struct sock *sk, struct tcp_sock *tp, } /* Undo during loss recovery after partial ACK. */ -static int tcp_try_undo_loss(struct sock *sk, struct tcp_sock *tp) +static int tcp_try_undo_loss(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); + if (tcp_may_undo(tp)) { struct sk_buff *skb; tcp_for_write_queue(skb, sk) { @@ -1994,7 +2014,7 @@ static int tcp_try_undo_loss(struct sock *sk, struct tcp_sock *tp) clear_all_retrans_hints(tp); - DBGUNDO(sk, tp, "partial loss"); + DBGUNDO(sk, "partial loss"); tp->lost_out = 0; tp->left_out = tp->sacked_out; tcp_undo_cwr(sk, 1); @@ -2016,8 +2036,10 @@ static inline void tcp_complete_cwr(struct sock *sk) tcp_ca_event(sk, CA_EVENT_COMPLETE_CWR); } -static void tcp_try_to_open(struct sock *sk, struct tcp_sock *tp, int flag) +static void tcp_try_to_open(struct sock *sk, int flag) { + struct tcp_sock *tp = tcp_sk(sk); + tp->left_out = tp->sacked_out; if (tp->retrans_out == 0) @@ -2111,7 +2133,7 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, before(tp->snd_una, tp->high_seq) && icsk->icsk_ca_state != TCP_CA_Open && tp->fackets_out > tp->reordering) { - tcp_mark_head_lost(sk, tp, tp->fackets_out-tp->reordering, tp->high_seq); + tcp_mark_head_lost(sk, tp->fackets_out-tp->reordering, tp->high_seq); NET_INC_STATS_BH(LINUX_MIB_TCPLOSS); } @@ -2127,7 +2149,7 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, switch (icsk->icsk_ca_state) { case TCP_CA_Loss: icsk->icsk_retransmits = 0; - if (tcp_try_undo_recovery(sk, tp)) + if (tcp_try_undo_recovery(sk)) return; break; @@ -2141,7 +2163,7 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, break; case TCP_CA_Disorder: - tcp_try_undo_dsack(sk, tp); + tcp_try_undo_dsack(sk); if (!tp->undo_marker || /* For SACK case do not Open to allow to undo * catching for all duplicate ACKs. */ @@ -2154,7 +2176,7 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, case TCP_CA_Recovery: if (IsReno(tp)) tcp_reset_reno_sack(tp); - if (tcp_try_undo_recovery(sk, tp)) + if (tcp_try_undo_recovery(sk)) return; tcp_complete_cwr(sk); break; @@ -2170,14 +2192,14 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, } else { int acked = prior_packets - tp->packets_out; if (IsReno(tp)) - tcp_remove_reno_sacks(sk, tp, acked); - is_dupack = tcp_try_undo_partial(sk, tp, acked); + tcp_remove_reno_sacks(sk, acked); + is_dupack = tcp_try_undo_partial(sk, acked); } break; case TCP_CA_Loss: if (flag&FLAG_DATA_ACKED) icsk->icsk_retransmits = 0; - if (!tcp_try_undo_loss(sk, tp)) { + if (!tcp_try_undo_loss(sk)) { tcp_moderate_cwnd(tp); tcp_xmit_retransmit_queue(sk); return; @@ -2194,10 +2216,10 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, } if (icsk->icsk_ca_state == TCP_CA_Disorder) - tcp_try_undo_dsack(sk, tp); + tcp_try_undo_dsack(sk); - if (!tcp_time_to_recover(sk, tp)) { - tcp_try_to_open(sk, tp, flag); + if (!tcp_time_to_recover(sk)) { + tcp_try_to_open(sk, flag); return; } @@ -2236,8 +2258,8 @@ tcp_fastretrans_alert(struct sock *sk, u32 prior_snd_una, tcp_set_ca_state(sk, TCP_CA_Recovery); } - if (is_dupack || tcp_head_timedout(sk, tp)) - tcp_update_scoreboard(sk, tp); + if (is_dupack || tcp_head_timedout(sk)) + tcp_update_scoreboard(sk); tcp_cwnd_down(sk); tcp_xmit_retransmit_queue(sk); } @@ -2313,8 +2335,10 @@ static void tcp_cong_avoid(struct sock *sk, u32 ack, u32 rtt, * RFC2988 recommends to restart timer to now+rto. */ -static void tcp_ack_packets_out(struct sock *sk, struct tcp_sock *tp) +static void tcp_ack_packets_out(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); + if (!tp->packets_out) { inet_csk_clear_xmit_timer(sk, ICSK_TIME_RETRANS); } else { @@ -2471,7 +2495,7 @@ static int tcp_clean_rtx_queue(struct sock *sk, __s32 *seq_rtt_p) if (acked&FLAG_ACKED) { tcp_ack_update_rtt(sk, acked, seq_rtt); - tcp_ack_packets_out(sk, tp); + tcp_ack_packets_out(sk); if (rtt_sample && !(acked & FLAG_RETRANS_DATA_ACKED)) (*rtt_sample)(sk, tcp_usrtt(&tv)); @@ -2556,9 +2580,10 @@ static inline int tcp_may_update_window(const struct tcp_sock *tp, const u32 ack * Window update algorithm, described in RFC793/RFC1122 (used in linux-2.2 * and in FreeBSD. NetBSD's one is even worse.) is wrong. */ -static int tcp_ack_update_window(struct sock *sk, struct tcp_sock *tp, - struct sk_buff *skb, u32 ack, u32 ack_seq) +static int tcp_ack_update_window(struct sock *sk, struct sk_buff *skb, u32 ack, + u32 ack_seq) { + struct tcp_sock *tp = tcp_sk(sk); int flag = 0; u32 nwin = ntohs(tcp_hdr(skb)->window); @@ -2576,7 +2601,7 @@ static int tcp_ack_update_window(struct sock *sk, struct tcp_sock *tp, * fast path is recovered for sending TCP. */ tp->pred_flags = 0; - tcp_fast_path_check(sk, tp); + tcp_fast_path_check(sk); if (nwin > tp->max_window) { tp->max_window = nwin; @@ -2762,7 +2787,7 @@ static int tcp_ack(struct sock *sk, struct sk_buff *skb, int flag) else NET_INC_STATS_BH(LINUX_MIB_TCPPUREACKS); - flag |= tcp_ack_update_window(sk, tp, skb, ack, ack_seq); + flag |= tcp_ack_update_window(sk, skb, ack, ack_seq); if (TCP_SKB_CB(skb)->sacked) flag |= tcp_sacktag_write_queue(sk, skb, prior_snd_una); @@ -3426,7 +3451,7 @@ queue_and_out: } tp->rcv_nxt = TCP_SKB_CB(skb)->end_seq; if (skb->len) - tcp_event_data_recv(sk, tp, skb); + tcp_event_data_recv(sk, skb); if (th->fin) tcp_fin(skb, sk, th); @@ -3443,7 +3468,7 @@ queue_and_out: if (tp->rx_opt.num_sacks) tcp_sack_remove(tp); - tcp_fast_path_check(sk, tp); + tcp_fast_path_check(sk); if (eaten > 0) __kfree_skb(skb); @@ -3734,7 +3759,7 @@ static int tcp_prune_queue(struct sock *sk) NET_INC_STATS_BH(LINUX_MIB_PRUNECALLED); if (atomic_read(&sk->sk_rmem_alloc) >= sk->sk_rcvbuf) - tcp_clamp_window(sk, tp); + tcp_clamp_window(sk); else if (tcp_memory_pressure) tp->rcv_ssthresh = min(tp->rcv_ssthresh, 4U * tp->advmss); @@ -3803,8 +3828,10 @@ void tcp_cwnd_application_limited(struct sock *sk) tp->snd_cwnd_stamp = tcp_time_stamp; } -static int tcp_should_expand_sndbuf(struct sock *sk, struct tcp_sock *tp) +static int tcp_should_expand_sndbuf(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); + /* If the user specified a specific send buffer setting, do * not modify it. */ @@ -3836,7 +3863,7 @@ static void tcp_new_space(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); - if (tcp_should_expand_sndbuf(sk, tp)) { + if (tcp_should_expand_sndbuf(sk)) { int sndmem = max_t(u32, tp->rx_opt.mss_clamp, tp->mss_cache) + MAX_TCP_HEADER + 16 + sizeof(struct sk_buff), demanded = max_t(unsigned int, tp->snd_cwnd, @@ -3860,9 +3887,9 @@ static void tcp_check_space(struct sock *sk) } } -static inline void tcp_data_snd_check(struct sock *sk, struct tcp_sock *tp) +static inline void tcp_data_snd_check(struct sock *sk) { - tcp_push_pending_frames(sk, tp); + tcp_push_pending_frames(sk); tcp_check_space(sk); } @@ -4196,7 +4223,7 @@ int tcp_rcv_established(struct sock *sk, struct sk_buff *skb, */ tcp_ack(sk, skb, 0); __kfree_skb(skb); - tcp_data_snd_check(sk, tp); + tcp_data_snd_check(sk); return 0; } else { /* Header too small */ TCP_INC_STATS_BH(TCP_MIB_INERRS); @@ -4267,12 +4294,12 @@ int tcp_rcv_established(struct sock *sk, struct sk_buff *skb, tp->rcv_nxt = TCP_SKB_CB(skb)->end_seq; } - tcp_event_data_recv(sk, tp, skb); + tcp_event_data_recv(sk, skb); if (TCP_SKB_CB(skb)->ack_seq != tp->snd_una) { /* Well, only one small jumplet in fast path... */ tcp_ack(sk, skb, FLAG_DATA); - tcp_data_snd_check(sk, tp); + tcp_data_snd_check(sk); if (!inet_csk_ack_scheduled(sk)) goto no_ack; } @@ -4355,7 +4382,7 @@ step5: /* step 7: process the segment text */ tcp_data_queue(sk, skb); - tcp_data_snd_check(sk, tp); + tcp_data_snd_check(sk); tcp_ack_snd_check(sk); return 0; @@ -4672,7 +4699,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, /* Do step6 onward by hand. */ tcp_urg(sk, skb, th); __kfree_skb(skb); - tcp_data_snd_check(sk, tp); + tcp_data_snd_check(sk); return 0; } @@ -4864,7 +4891,7 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb, /* tcp_data could move socket to TIME-WAIT */ if (sk->sk_state != TCP_CLOSE) { - tcp_data_snd_check(sk, tp); + tcp_data_snd_check(sk); tcp_ack_snd_check(sk); } diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 94d9f0c..3a60aea 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -62,12 +62,13 @@ int sysctl_tcp_base_mss __read_mostly = 512; /* By default, RFC2861 behavior. */ int sysctl_tcp_slow_start_after_idle __read_mostly = 1; -static void update_send_head(struct sock *sk, struct tcp_sock *tp, - struct sk_buff *skb) +static void update_send_head(struct sock *sk, struct sk_buff *skb) { + struct tcp_sock *tp = tcp_sk(sk); + tcp_advance_send_head(sk, skb); tp->snd_nxt = TCP_SKB_CB(skb)->end_seq; - tcp_packets_out_inc(sk, tp, skb); + tcp_packets_out_inc(sk, skb); } /* SND.NXT, if window was not shrunk. @@ -76,8 +77,10 @@ static void update_send_head(struct sock *sk, struct tcp_sock *tp, * 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(struct sock *sk, struct tcp_sock *tp) +static inline __u32 tcp_acceptable_seq(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); + if (!before(tp->snd_una+tp->snd_wnd, tp->snd_nxt)) return tp->snd_nxt; else @@ -516,7 +519,7 @@ static int tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, int clone_it, md5 ? &md5_hash_location : #endif NULL); - TCP_ECN_send(sk, tp, skb, tcp_header_size); + TCP_ECN_send(sk, skb, tcp_header_size); } #ifdef CONFIG_TCP_MD5SIG @@ -927,8 +930,9 @@ unsigned int tcp_current_mss(struct sock *sk, int large_allowed) /* Congestion window validation. (RFC2861) */ -static void tcp_cwnd_validate(struct sock *sk, struct tcp_sock *tp) +static void tcp_cwnd_validate(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); __u32 packets_out = tp->packets_out; if (packets_out >= tp->snd_cwnd) { @@ -1076,8 +1080,9 @@ static unsigned int tcp_snd_test(struct sock *sk, struct sk_buff *skb, return cwnd_quota; } -int tcp_may_send_now(struct sock *sk, struct tcp_sock *tp) +int tcp_may_send_now(struct sock *sk) { + struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *skb = tcp_send_head(sk); return (skb && @@ -1144,8 +1149,9 @@ static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len, * * This algorithm is from John Heffner. */ -static int tcp_tso_should_defer(struct sock *sk, struct tcp_sock *tp, struct sk_buff *skb) +static int tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb) { + struct tcp_sock *tp = tcp_sk(sk); const struct inet_connection_sock *icsk = inet_csk(sk); u32 send_win, cong_win, limit, in_flight; @@ -1324,7 +1330,7 @@ static int tcp_mtu_probe(struct sock *sk) /* Decrement cwnd here because we are sending * effectively two packets. */ tp->snd_cwnd--; - update_send_head(sk, tp, nskb); + update_send_head(sk, nskb); icsk->icsk_mtup.probe_size = tcp_mss_to_mtu(sk, nskb->len); tp->mtu_probe.probe_seq_start = TCP_SKB_CB(nskb)->seq; @@ -1387,7 +1393,7 @@ static int tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle) nonagle : TCP_NAGLE_PUSH)))) break; } else { - if (tcp_tso_should_defer(sk, tp, skb)) + if (tcp_tso_should_defer(sk, skb)) break; } @@ -1416,14 +1422,14 @@ static int tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle) /* Advance the send_head. This one is sent out. * This call will increment packets_out. */ - update_send_head(sk, tp, skb); + update_send_head(sk, skb); tcp_minshall_update(tp, mss_now, skb); sent_pkts++; } if (likely(sent_pkts)) { - tcp_cwnd_validate(sk, tp); + tcp_cwnd_validate(sk); return 0; } return !tp->packets_out && tcp_send_head(sk); @@ -1433,14 +1439,14 @@ static int tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle) * TCP_CORK or attempt at coalescing tiny packets. * The socket must be locked by the caller. */ -void __tcp_push_pending_frames(struct sock *sk, struct tcp_sock *tp, - unsigned int cur_mss, int nonagle) +void __tcp_push_pending_frames(struct sock *sk, unsigned int cur_mss, + int nonagle) { struct sk_buff *skb = tcp_send_head(sk); if (skb) { if (tcp_write_xmit(sk, cur_mss, nonagle)) - tcp_check_probe_timer(sk, tp); + tcp_check_probe_timer(sk); } } @@ -1484,8 +1490,8 @@ void tcp_push_one(struct sock *sk, unsigned int mss_now) TCP_SKB_CB(skb)->when = tcp_time_stamp; if (likely(!tcp_transmit_skb(sk, skb, 1, sk->sk_allocation))) { - update_send_head(sk, tp, skb); - tcp_cwnd_validate(sk, tp); + update_send_head(sk, skb); + tcp_cwnd_validate(sk); return; } } @@ -1933,7 +1939,7 @@ void tcp_xmit_retransmit_queue(struct sock *sk) * segments to send. */ - if (tcp_may_send_now(sk, tp)) + if (tcp_may_send_now(sk)) return; if (tp->forward_skb_hint) { @@ -2023,7 +2029,7 @@ void tcp_send_fin(struct sock *sk) TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(skb)->seq + 1; tcp_queue_skb(sk, skb); } - __tcp_push_pending_frames(sk, tp, mss_now, TCP_NAGLE_OFF); + __tcp_push_pending_frames(sk, mss_now, TCP_NAGLE_OFF); } /* We get here when a process closes a file descriptor (either due to @@ -2033,7 +2039,6 @@ void tcp_send_fin(struct sock *sk) */ void tcp_send_active_reset(struct sock *sk, gfp_t priority) { - struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *skb; /* NOTE: No TCP options attached and we never retransmit this. */ @@ -2053,7 +2058,7 @@ void tcp_send_active_reset(struct sock *sk, gfp_t priority) skb_shinfo(skb)->gso_type = 0; /* Send it off. */ - TCP_SKB_CB(skb)->seq = tcp_acceptable_seq(sk, tp); + TCP_SKB_CB(skb)->seq = tcp_acceptable_seq(sk); TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(skb)->seq; TCP_SKB_CB(skb)->when = tcp_time_stamp; if (tcp_transmit_skb(sk, skb, 0, priority)) @@ -2271,7 +2276,7 @@ int tcp_connect(struct sock *sk) skb_reserve(buff, MAX_TCP_HEADER); TCP_SKB_CB(buff)->flags = TCPCB_FLAG_SYN; - TCP_ECN_send_syn(sk, tp, buff); + TCP_ECN_send_syn(sk, buff); TCP_SKB_CB(buff)->sacked = 0; skb_shinfo(buff)->gso_segs = 1; skb_shinfo(buff)->gso_size = 0; @@ -2363,7 +2368,6 @@ void tcp_send_ack(struct sock *sk) { /* If we have been reset, we may not send again. */ if (sk->sk_state != TCP_CLOSE) { - struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *buff; /* We are not putting this on the write queue, so @@ -2389,7 +2393,7 @@ void tcp_send_ack(struct sock *sk) skb_shinfo(buff)->gso_type = 0; /* Send it off, this clears delayed acks for us. */ - TCP_SKB_CB(buff)->seq = TCP_SKB_CB(buff)->end_seq = tcp_acceptable_seq(sk, tp); + TCP_SKB_CB(buff)->seq = TCP_SKB_CB(buff)->end_seq = tcp_acceptable_seq(sk); TCP_SKB_CB(buff)->when = tcp_time_stamp; tcp_transmit_skb(sk, buff, 0, GFP_ATOMIC); } @@ -2467,7 +2471,7 @@ int tcp_write_wakeup(struct sock *sk) TCP_SKB_CB(skb)->when = tcp_time_stamp; err = tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC); if (!err) { - update_send_head(sk, tp, skb); + update_send_head(sk, skb); } return err; } else { |