From 84d3e7b9573291a1ea845bdd51b74bb484597661 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 5 Jul 2005 15:18:18 -0700 Subject: [TCP]: Move __tcp_data_snd_check into tcp_output.c It reimplements portions of tcp_snd_check(), so it we move it to tcp_output.c we can consolidate it's logic much easier in a later change. Signed-off-by: David S. Miller --- net/ipv4/tcp_input.c | 10 ---------- 1 file changed, 10 deletions(-) (limited to 'net/ipv4/tcp_input.c') diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 7bbbbc33eb4b..577424323d59 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -3346,16 +3346,6 @@ static inline void tcp_check_space(struct sock *sk) } } -static void __tcp_data_snd_check(struct sock *sk, struct sk_buff *skb) -{ - struct tcp_sock *tp = tcp_sk(sk); - - if (after(TCP_SKB_CB(skb)->end_seq, tp->snd_una + tp->snd_wnd) || - tcp_packets_in_flight(tp) >= tp->snd_cwnd || - tcp_write_xmit(sk, tp->nonagle)) - tcp_check_probe_timer(sk, tp); -} - static __inline__ void tcp_data_snd_check(struct sock *sk) { struct sk_buff *skb = sk->sk_send_head; -- cgit v1.2.2 From 55c97f3e990c1ff63957c64f6cb10711a09fd70e Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 5 Jul 2005 15:19:38 -0700 Subject: [TCP]: Fix __tcp_push_pending_frames() 'nonagle' handling. 'nonagle' should be passed to the tcp_snd_test() function as 'TCP_NAGLE_PUSH' if we are checking an SKB not at the tail of the write_queue. This is because Nagle does not apply to such frames since we cannot possibly tack more data onto them. However, while doing this __tcp_push_pending_frames() makes all of the packets in the write_queue use this modified 'nonagle' value. Fix the bug and simplify this function by just calling tcp_write_xmit() directly if sk_send_head is non-NULL. As a result, we can now make tcp_data_snd_check() just call tcp_push_pending_frames() instead of the specialized __tcp_data_snd_check(). Signed-off-by: David S. Miller --- net/ipv4/tcp_input.c | 17 +++++++---------- 1 file changed, 7 insertions(+), 10 deletions(-) (limited to 'net/ipv4/tcp_input.c') diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 577424323d59..b27be2f819ac 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -3346,12 +3346,9 @@ static inline void tcp_check_space(struct sock *sk) } } -static __inline__ void tcp_data_snd_check(struct sock *sk) +static __inline__ void tcp_data_snd_check(struct sock *sk, struct tcp_sock *tp) { - struct sk_buff *skb = sk->sk_send_head; - - if (skb != NULL) - __tcp_data_snd_check(sk, skb); + tcp_push_pending_frames(sk, tp); tcp_check_space(sk); } @@ -3645,7 +3642,7 @@ int tcp_rcv_established(struct sock *sk, struct sk_buff *skb, */ tcp_ack(sk, skb, 0); __kfree_skb(skb); - tcp_data_snd_check(sk); + tcp_data_snd_check(sk, tp); return 0; } else { /* Header too small */ TCP_INC_STATS_BH(TCP_MIB_INERRS); @@ -3711,7 +3708,7 @@ int tcp_rcv_established(struct sock *sk, struct sk_buff *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); + tcp_data_snd_check(sk, tp); if (!tcp_ack_scheduled(tp)) goto no_ack; } @@ -3789,7 +3786,7 @@ step5: /* step 7: process the segment text */ tcp_data_queue(sk, skb); - tcp_data_snd_check(sk); + tcp_data_snd_check(sk, tp); tcp_ack_snd_check(sk); return 0; @@ -4099,7 +4096,7 @@ 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); + tcp_data_snd_check(sk, tp); return 0; } @@ -4290,7 +4287,7 @@ 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); + tcp_data_snd_check(sk, tp); tcp_ack_snd_check(sk); } -- cgit v1.2.2 From a56476962e92a6c389a1a561274d4a27607b7b5f Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 5 Jul 2005 15:20:41 -0700 Subject: [TCP]: Kill bogus comment above tcp_tso_acked(). Everything stated there is out of data. tcp_trim_skb() does adjust the available socket send buffer space and skb->truesize now. Signed-off-by: David S. Miller --- net/ipv4/tcp_input.c | 9 --------- 1 file changed, 9 deletions(-) (limited to 'net/ipv4/tcp_input.c') diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index b27be2f819ac..1dba7fd438da 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -1957,15 +1957,6 @@ static inline void tcp_ack_packets_out(struct sock *sk, struct tcp_sock *tp) } } -/* There is one downside to this scheme. Although we keep the - * ACK clock ticking, adjusting packet counters and advancing - * congestion window, we do not liberate socket send buffer - * space. - * - * Mucking with skb->truesize and sk->sk_wmem_alloc et al. - * then making a write space wakeup callback is a possible - * future enhancement. WARNING: it is not trivial to make. - */ static int tcp_tso_acked(struct sock *sk, struct sk_buff *skb, __u32 now, __s32 *seq_rtt) { -- cgit v1.2.2 From cb83199a29dc0408423d6df432f28cc67fcadaf4 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 5 Jul 2005 15:20:55 -0700 Subject: [TCP]: Do not call tcp_tso_acked() if no work to do. In tcp_clean_rtx_queue(), if the TSO packet is not even partially acked, do not waste time calling tcp_tso_acked(). Signed-off-by: David S. Miller --- net/ipv4/tcp_input.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'net/ipv4/tcp_input.c') diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 1dba7fd438da..b948e4eb39b7 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -2038,7 +2038,8 @@ static int tcp_clean_rtx_queue(struct sock *sk, __s32 *seq_rtt_p, s32 *seq_usrtt * the other end. */ if (after(scb->end_seq, tp->snd_una)) { - if (tcp_skb_pcount(skb) > 1) + if (tcp_skb_pcount(skb) > 1 && + after(tp->snd_una, scb->seq)) acked |= tcp_tso_acked(sk, skb, now, &seq_rtt); break; -- cgit v1.2.2 From 0d9901df62fe4820aee86b49f1a074cdb5c6928e Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 5 Jul 2005 15:21:10 -0700 Subject: [TCP]: Break out send buffer expansion test. This makes it easier to understand, and allows easier tweaking of the heuristic later on. Signed-off-by: David S. Miller --- net/ipv4/tcp_input.c | 27 +++++++++++++++++++++++---- 1 file changed, 23 insertions(+), 4 deletions(-) (limited to 'net/ipv4/tcp_input.c') diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index b948e4eb39b7..2ef2f355b8b8 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -3300,6 +3300,28 @@ void tcp_cwnd_application_limited(struct sock *sk) tp->snd_cwnd_stamp = tcp_time_stamp; } +static inline int tcp_should_expand_sndbuf(struct sock *sk, struct tcp_sock *tp) +{ + /* If the user specified a specific send buffer setting, do + * not modify it. + */ + if (sk->sk_userlocks & SOCK_SNDBUF_LOCK) + return 0; + + /* If we are under global TCP memory pressure, do not expand. */ + if (tcp_memory_pressure) + return 0; + + /* If we are under soft global TCP memory pressure, do not expand. */ + if (atomic_read(&tcp_memory_allocated) >= sysctl_tcp_mem[0]) + return 0; + + /* If we filled the congestion window, do not expand. */ + if (tp->packets_out >= tp->snd_cwnd) + return 0; + + return 1; +} /* When incoming ACK allowed to free some skb from write_queue, * we remember this event in flag SOCK_QUEUE_SHRUNK and wake up socket @@ -3311,10 +3333,7 @@ static void tcp_new_space(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); - if (tp->packets_out < tp->snd_cwnd && - !(sk->sk_userlocks & SOCK_SNDBUF_LOCK) && - !tcp_memory_pressure && - atomic_read(&tcp_memory_allocated) < sysctl_tcp_mem[0]) { + if (tcp_should_expand_sndbuf(sk, tp)) { int sndmem = max_t(u32, tp->rx_opt.mss_clamp, tp->mss_cache_std) + MAX_TCP_HEADER + 16 + sizeof(struct sk_buff), demanded = max_t(unsigned int, tp->snd_cwnd, -- cgit v1.2.2 From c1b4a7e69576d65efc31a8cea0714173c2841244 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 5 Jul 2005 15:24:38 -0700 Subject: [TCP]: Move to new TSO segmenting scheme. Make TSO segment transmit size decisions at send time not earlier. The basic scheme is that we try to build as large a TSO frame as possible when pulling in the user data, but the size of the TSO frame output to the card is determined at transmit time. This is guided by tp->xmit_size_goal. It is always set to a multiple of MSS and tells sendmsg/sendpage how large an SKB to try and build. Later, tcp_write_xmit() and tcp_push_one() chop up the packet if necessary and conditions warrant. These routines can also decide to "defer" in order to wait for more ACKs to arrive and thus allow larger TSO frames to be emitted. A general observation is that TSO elongates the pipe, thus requiring a larger congestion window and larger buffering especially at the sender side. Therefore, it is important that applications 1) get a large enough socket send buffer (this is accomplished by our dynamic send buffer expansion code) 2) do large enough writes. Signed-off-by: David S. Miller --- net/ipv4/tcp_input.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) (limited to 'net/ipv4/tcp_input.c') diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 2ef2f355b8b8..8de2f1071c2b 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -740,10 +740,10 @@ __u32 tcp_init_cwnd(struct tcp_sock *tp, struct dst_entry *dst) __u32 cwnd = (dst ? dst_metric(dst, RTAX_INITCWND) : 0); if (!cwnd) { - if (tp->mss_cache_std > 1460) + if (tp->mss_cache > 1460) cwnd = 2; else - cwnd = (tp->mss_cache_std > 1095) ? 3 : 4; + cwnd = (tp->mss_cache > 1095) ? 3 : 4; } return min_t(__u32, cwnd, tp->snd_cwnd_clamp); } @@ -914,7 +914,7 @@ tcp_sacktag_write_queue(struct sock *sk, struct sk_buff *ack_skb, u32 prior_snd_ if (sk->sk_route_caps & NETIF_F_TSO) { sk->sk_route_caps &= ~NETIF_F_TSO; sock_set_flag(sk, SOCK_NO_LARGESEND); - tp->mss_cache = tp->mss_cache_std; + tp->mss_cache = tp->mss_cache; } if (!tp->sacked_out) @@ -1077,7 +1077,7 @@ tcp_sacktag_write_queue(struct sock *sk, struct sk_buff *ack_skb, u32 prior_snd_ (IsFack(tp) || !before(lost_retrans, TCP_SKB_CB(skb)->ack_seq + tp->reordering * - tp->mss_cache_std))) { + tp->mss_cache))) { TCP_SKB_CB(skb)->sacked &= ~TCPCB_SACKED_RETRANS; tp->retrans_out -= tcp_skb_pcount(skb); @@ -3334,7 +3334,7 @@ static void tcp_new_space(struct sock *sk) struct tcp_sock *tp = tcp_sk(sk); if (tcp_should_expand_sndbuf(sk, tp)) { - int sndmem = max_t(u32, tp->rx_opt.mss_clamp, tp->mss_cache_std) + + 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, tp->reordering + 1); -- cgit v1.2.2 From b03efcfb2180289718991bb984044ce6c5b7d1b0 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Fri, 8 Jul 2005 14:57:23 -0700 Subject: [NET]: Transform skb_queue_len() binary tests into skb_queue_empty() This is part of the grand scheme to eliminate the qlen member of skb_queue_head, and subsequently remove the 'list' member of sk_buff. Most users of skb_queue_len() want to know if the queue is empty or not, and that's trivially done with skb_queue_empty() which doesn't use the skb_queue_head->qlen member and instead uses the queue list emptyness as the test. Signed-off-by: David S. Miller --- net/ipv4/tcp_input.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) (limited to 'net/ipv4/tcp_input.c') diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 8de2f1071c2b..53a8a5399f1e 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -2802,7 +2802,7 @@ static void tcp_sack_remove(struct tcp_sock *tp) int this_sack; /* Empty ofo queue, hence, all the SACKs are eaten. Clear. */ - if (skb_queue_len(&tp->out_of_order_queue) == 0) { + if (skb_queue_empty(&tp->out_of_order_queue)) { tp->rx_opt.num_sacks = 0; tp->rx_opt.eff_sacks = tp->rx_opt.dsack; return; @@ -2935,13 +2935,13 @@ queue_and_out: if(th->fin) tcp_fin(skb, sk, th); - if (skb_queue_len(&tp->out_of_order_queue)) { + if (!skb_queue_empty(&tp->out_of_order_queue)) { tcp_ofo_queue(sk); /* RFC2581. 4.2. SHOULD send immediate ACK, when * gap in queue is filled. */ - if (!skb_queue_len(&tp->out_of_order_queue)) + if (skb_queue_empty(&tp->out_of_order_queue)) tp->ack.pingpong = 0; } @@ -3249,9 +3249,8 @@ static int tcp_prune_queue(struct sock *sk) * This must not ever occur. */ /* First, purge the out_of_order queue. */ - if (skb_queue_len(&tp->out_of_order_queue)) { - NET_ADD_STATS_BH(LINUX_MIB_OFOPRUNED, - skb_queue_len(&tp->out_of_order_queue)); + if (!skb_queue_empty(&tp->out_of_order_queue)) { + NET_INC_STATS_BH(LINUX_MIB_OFOPRUNED); __skb_queue_purge(&tp->out_of_order_queue); /* Reset SACK state. A conforming SACK implementation will -- cgit v1.2.2