From 86fd14ad1e8c4b8f5e9a7a27b26bdade91dd4bd0 Mon Sep 17 00:00:00 2001 From: Weiping Pan Date: Fri, 18 Apr 2014 12:27:46 +0800 Subject: tcp: make tcp_cwnd_application_limited() static Make tcp_cwnd_application_limited() static and move it from tcp_input.c to tcp_output.c Signed-off-by: Weiping Pan Signed-off-by: David S. Miller --- include/net/tcp.h | 1 - 1 file changed, 1 deletion(-) (limited to 'include/net/tcp.h') diff --git a/include/net/tcp.h b/include/net/tcp.h index 87d877408188..163d2b467d78 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -558,7 +558,6 @@ void tcp_send_loss_probe(struct sock *sk); bool tcp_schedule_loss_probe(struct sock *sk); /* tcp_input.c */ -void tcp_cwnd_application_limited(struct sock *sk); void tcp_resume_early_retransmit(struct sock *sk); void tcp_rearm_rto(struct sock *sk); void tcp_reset(struct sock *sk); -- cgit v1.2.2 From e114a710aa5058c0ba4aa1dfb105132aefeb5e04 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 30 Apr 2014 11:58:13 -0700 Subject: tcp: fix cwnd limited checking to improve congestion control Yuchung discovered tcp_is_cwnd_limited() was returning false in slow start phase even if the application filled the socket write queue. All congestion modules take into account tcp_is_cwnd_limited() before increasing cwnd, so this behavior limits slow start from probing the bandwidth at full speed. The problem is that even if write queue is full (aka we are _not_ application limited), cwnd can be under utilized if TSO should auto defer or TCP Small queues decided to hold packets. So the in_flight can be kept to smaller value, and we can get to the point tcp_is_cwnd_limited() returns false. With TCP Small Queues and FQ/pacing, this issue is more visible. We fix this by having tcp_cwnd_validate(), which is supposed to track such things, take into account unsent_segs, the number of segs that we are not sending at the moment due to TSO or TSQ, but intend to send real soon. Then when we are cwnd-limited, remember this fact while we are processing the window of ACKs that comes back. For example, suppose we have a brand new connection with cwnd=10; we are in slow start, and we send a flight of 9 packets. By the time we have received ACKs for all 9 packets we want our cwnd to be 18. We implement this by setting tp->lsnd_pending to 9, and considering ourselves to be cwnd-limited while cwnd is less than twice tp->lsnd_pending (2*9 -> 18). This makes tcp_is_cwnd_limited() more understandable, by removing the GSO/TSO kludge, that tried to work around the issue. Note the in_flight parameter can be removed in a followup cleanup patch. Signed-off-by: Eric Dumazet Signed-off-by: Neal Cardwell Signed-off-by: Yuchung Cheng Signed-off-by: David S. Miller --- include/net/tcp.h | 22 +++++++++++++++++++++- 1 file changed, 21 insertions(+), 1 deletion(-) (limited to 'include/net/tcp.h') diff --git a/include/net/tcp.h b/include/net/tcp.h index 163d2b467d78..a9fe7bc4f4bb 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -974,7 +974,27 @@ static inline u32 tcp_wnd_end(const struct tcp_sock *tp) { return tp->snd_una + tp->snd_wnd; } -bool tcp_is_cwnd_limited(const struct sock *sk, u32 in_flight); + +/* We follow the spirit of RFC2861 to validate cwnd but implement a more + * flexible approach. The RFC suggests cwnd should not be raised unless + * it was fully used previously. But we allow cwnd to grow as long as the + * application has used half the cwnd. + * Example : + * cwnd is 10 (IW10), but application sends 9 frames. + * We allow cwnd to reach 18 when all frames are ACKed. + * This check is safe because it's as aggressive as slow start which already + * risks 100% overshoot. The advantage is that we discourage application to + * either send more filler packets or data to artificially blow up the cwnd + * usage, and allow application-limited process to probe bw more aggressively. + * + * TODO: remove in_flight once we can fix all callers, and their callers... + */ +static inline bool tcp_is_cwnd_limited(const struct sock *sk, u32 in_flight) +{ + const struct tcp_sock *tp = tcp_sk(sk); + + return tp->snd_cwnd < 2 * tp->lsnd_pending; +} static inline void tcp_check_probe_timer(struct sock *sk) { -- cgit v1.2.2 From 249015515fe3fc9818d86cb5c83bbc92505ad7dc Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 2 May 2014 21:18:05 -0700 Subject: tcp: remove in_flight parameter from cong_avoid() methods Commit e114a710aa505 ("tcp: fix cwnd limited checking to improve congestion control") obsoleted in_flight parameter from tcp_is_cwnd_limited() and its callers. This patch does the removal as promised. Signed-off-by: Eric Dumazet Acked-by: Neal Cardwell Signed-off-by: David S. Miller --- include/net/tcp.h | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) (limited to 'include/net/tcp.h') diff --git a/include/net/tcp.h b/include/net/tcp.h index a9fe7bc4f4bb..3c9418456640 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -796,7 +796,7 @@ struct tcp_congestion_ops { /* return slow start threshold (required) */ u32 (*ssthresh)(struct sock *sk); /* do new cwnd calculation (required) */ - void (*cong_avoid)(struct sock *sk, u32 ack, u32 acked, u32 in_flight); + void (*cong_avoid)(struct sock *sk, u32 ack, u32 acked); /* call before changing ca_state (optional) */ void (*set_state)(struct sock *sk, u8 new_state); /* call when cwnd event occurs (optional) */ @@ -828,7 +828,7 @@ void tcp_cong_avoid_ai(struct tcp_sock *tp, u32 w); extern struct tcp_congestion_ops tcp_init_congestion_ops; u32 tcp_reno_ssthresh(struct sock *sk); -void tcp_reno_cong_avoid(struct sock *sk, u32 ack, u32 acked, u32 in_flight); +void tcp_reno_cong_avoid(struct sock *sk, u32 ack, u32 acked); extern struct tcp_congestion_ops tcp_reno; static inline void tcp_set_ca_state(struct sock *sk, const u8 ca_state) @@ -986,10 +986,8 @@ static inline u32 tcp_wnd_end(const struct tcp_sock *tp) * risks 100% overshoot. The advantage is that we discourage application to * either send more filler packets or data to artificially blow up the cwnd * usage, and allow application-limited process to probe bw more aggressively. - * - * TODO: remove in_flight once we can fix all callers, and their callers... */ -static inline bool tcp_is_cwnd_limited(const struct sock *sk, u32 in_flight) +static inline bool tcp_is_cwnd_limited(const struct sock *sk) { const struct tcp_sock *tp = tcp_sk(sk); -- cgit v1.2.2 From 5b7ed0892f2af4e60b9a8d2c71c77774512a6cb9 Mon Sep 17 00:00:00 2001 From: Yuchung Cheng Date: Sun, 11 May 2014 20:22:09 -0700 Subject: tcp: move fastopen functions to tcp_fastopen.c Move common TFO functions that will be used by both v4 and v6 to tcp_fastopen.c. Create a helper tcp_fastopen_queue_check(). Signed-off-by: Yuchung Cheng Signed-off-by: Daniel Lee Signed-off-by: Jerry Chu Signed-off-by: Eric Dumazet Acked-by: Neal Cardwell Signed-off-by: David S. Miller --- include/net/tcp.h | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) (limited to 'include/net/tcp.h') diff --git a/include/net/tcp.h b/include/net/tcp.h index 3c9418456640..012236838583 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1329,8 +1329,14 @@ void tcp_free_fastopen_req(struct tcp_sock *tp); extern struct tcp_fastopen_context __rcu *tcp_fastopen_ctx; int tcp_fastopen_reset_cipher(void *key, unsigned int len); -void tcp_fastopen_cookie_gen(__be32 src, __be32 dst, - struct tcp_fastopen_cookie *foc); +int tcp_fastopen_create_child(struct sock *sk, + struct sk_buff *skb, + struct sk_buff *skb_synack, + struct request_sock *req); +bool tcp_fastopen_check(struct sock *sk, struct sk_buff *skb, + struct request_sock *req, + struct tcp_fastopen_cookie *foc, + struct tcp_fastopen_cookie *valid_foc); void tcp_fastopen_init_key_once(bool publish); #define TCP_FASTOPEN_KEY_LENGTH 16 -- cgit v1.2.2 From 89278c9dc922272df921042aafa18311f3398c6c Mon Sep 17 00:00:00 2001 From: Yuchung Cheng Date: Sun, 11 May 2014 20:22:10 -0700 Subject: tcp: simplify fast open cookie processing Consolidate various cookie checking and generation code to simplify the fast open processing. The main goal is to reduce code duplication in tcp_v4_conn_request() for IPv6 support. Removes two experimental sysctl flags TFO_SERVER_ALWAYS and TFO_SERVER_COOKIE_NOT_CHKD used primarily for developmental debugging purposes. Signed-off-by: Yuchung Cheng Signed-off-by: Daniel Lee Signed-off-by: Jerry Chu Signed-off-by: Eric Dumazet Acked-by: Neal Cardwell Signed-off-by: David S. Miller --- include/net/tcp.h | 9 +-------- 1 file changed, 1 insertion(+), 8 deletions(-) (limited to 'include/net/tcp.h') diff --git a/include/net/tcp.h b/include/net/tcp.h index 012236838583..17d7c6a3d037 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -220,8 +220,6 @@ void tcp_time_wait(struct sock *sk, int state, int timeo); #define TFO_SERVER_ENABLE 2 #define TFO_CLIENT_NO_COOKIE 4 /* Data in SYN w/o cookie option */ -/* Process SYN data but skip cookie validation */ -#define TFO_SERVER_COOKIE_NOT_CHKED 0x100 /* Accept SYN data w/o any cookie option */ #define TFO_SERVER_COOKIE_NOT_REQD 0x200 @@ -230,10 +228,6 @@ void tcp_time_wait(struct sock *sk, int state, int timeo); */ #define TFO_SERVER_WO_SOCKOPT1 0x400 #define TFO_SERVER_WO_SOCKOPT2 0x800 -/* Always create TFO child sockets on a TFO listener even when - * cookie/data not present. (For testing purpose!) - */ -#define TFO_SERVER_ALWAYS 0x1000 extern struct inet_timewait_death_row tcp_death_row; @@ -1335,8 +1329,7 @@ int tcp_fastopen_create_child(struct sock *sk, struct request_sock *req); bool tcp_fastopen_check(struct sock *sk, struct sk_buff *skb, struct request_sock *req, - struct tcp_fastopen_cookie *foc, - struct tcp_fastopen_cookie *valid_foc); + struct tcp_fastopen_cookie *foc); void tcp_fastopen_init_key_once(bool publish); #define TCP_FASTOPEN_KEY_LENGTH 16 -- cgit v1.2.2 From 843f4a55e336e6d0c7bb92e7f9621535bc8d5fcd Mon Sep 17 00:00:00 2001 From: Yuchung Cheng Date: Sun, 11 May 2014 20:22:11 -0700 Subject: tcp: use tcp_v4_send_synack on first SYN-ACK To avoid large code duplication in IPv6, we need to first simplify the complicate SYN-ACK sending code in tcp_v4_conn_request(). To use tcp_v4(6)_send_synack() to send all SYN-ACKs, we need to initialize the mini socket's receive window before trying to create the child socket and/or building the SYN-ACK packet. So we move that initialization from tcp_make_synack() to tcp_v4_conn_request() as a new function tcp_openreq_init_req_rwin(). After this refactoring the SYN-ACK sending code is simpler and easier to implement Fast Open for IPv6. Signed-off-by: Yuchung Cheng Signed-off-by: Daniel Lee Signed-off-by: Jerry Chu Acked-by: Neal Cardwell Signed-off-by: David S. Miller --- include/net/tcp.h | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) (limited to 'include/net/tcp.h') diff --git a/include/net/tcp.h b/include/net/tcp.h index 17d7c6a3d037..f5d6ca4a9d28 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -1114,6 +1114,9 @@ static inline void tcp_openreq_init(struct request_sock *req, ireq->ir_num = ntohs(tcp_hdr(skb)->dest); } +extern void tcp_openreq_init_rwin(struct request_sock *req, + struct sock *sk, struct dst_entry *dst); + void tcp_enter_memory_pressure(struct sock *sk); static inline int keepalive_intvl_when(const struct tcp_sock *tp) @@ -1323,13 +1326,10 @@ void tcp_free_fastopen_req(struct tcp_sock *tp); extern struct tcp_fastopen_context __rcu *tcp_fastopen_ctx; int tcp_fastopen_reset_cipher(void *key, unsigned int len); -int tcp_fastopen_create_child(struct sock *sk, - struct sk_buff *skb, - struct sk_buff *skb_synack, - struct request_sock *req); -bool tcp_fastopen_check(struct sock *sk, struct sk_buff *skb, - struct request_sock *req, - struct tcp_fastopen_cookie *foc); +bool tcp_try_fastopen(struct sock *sk, struct sk_buff *skb, + struct request_sock *req, + struct tcp_fastopen_cookie *foc, + struct dst_entry *dst); void tcp_fastopen_init_key_once(bool publish); #define TCP_FASTOPEN_KEY_LENGTH 16 -- cgit v1.2.2 From ca8a22634381537c92b5a10308652e1c38fd9edf Mon Sep 17 00:00:00 2001 From: Neal Cardwell Date: Thu, 22 May 2014 10:41:08 -0400 Subject: tcp: make cwnd-limited checks measurement-based, and gentler Experience with the recent e114a710aa50 ("tcp: fix cwnd limited checking to improve congestion control") has shown that there are common cases where that commit can cause cwnd to be much larger than necessary. This leads to TSO autosizing cooking skbs that are too large, among other things. The main problems seemed to be: (1) That commit attempted to predict the future behavior of the connection by looking at the write queue (if TSO or TSQ limit sending). That prediction sometimes overestimated future outstanding packets. (2) That commit always allowed cwnd to grow to twice the number of outstanding packets (even in congestion avoidance, where this is not needed). This commit improves both of these, by: (1) Switching to a measurement-based approach where we explicitly track the largest number of packets in flight during the past window ("max_packets_out"), and remember whether we were cwnd-limited at the moment we finished sending that flight. (2) Only allowing cwnd to grow to twice the number of outstanding packets ("max_packets_out") in slow start. In congestion avoidance mode we now only allow cwnd to grow if it was fully utilized. Signed-off-by: Neal Cardwell Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/tcp.h | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) (limited to 'include/net/tcp.h') diff --git a/include/net/tcp.h b/include/net/tcp.h index f5d6ca4a9d28..e80abe4486cb 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -971,8 +971,9 @@ static inline u32 tcp_wnd_end(const struct tcp_sock *tp) /* We follow the spirit of RFC2861 to validate cwnd but implement a more * flexible approach. The RFC suggests cwnd should not be raised unless - * it was fully used previously. But we allow cwnd to grow as long as the - * application has used half the cwnd. + * it was fully used previously. And that's exactly what we do in + * congestion avoidance mode. But in slow start we allow cwnd to grow + * as long as the application has used half the cwnd. * Example : * cwnd is 10 (IW10), but application sends 9 frames. * We allow cwnd to reach 18 when all frames are ACKed. @@ -985,7 +986,11 @@ static inline bool tcp_is_cwnd_limited(const struct sock *sk) { const struct tcp_sock *tp = tcp_sk(sk); - return tp->snd_cwnd < 2 * tp->lsnd_pending; + /* If in slow start, ensure cwnd grows to twice what was ACKed. */ + if (tp->snd_cwnd <= tp->snd_ssthresh) + return tp->snd_cwnd < 2 * tp->max_packets_out; + + return tp->is_cwnd_limited; } static inline void tcp_check_probe_timer(struct sock *sk) -- cgit v1.2.2 From 6cc55e096f1f2a8585cf8dc9049862f2376f66d4 Mon Sep 17 00:00:00 2001 From: Octavian Purdila Date: Fri, 6 Jun 2014 17:32:37 +0300 Subject: tcp: add gfp parameter to tcp_fragment tcp_fragment can be called from process context (from tso_fragment). Add a new gfp parameter to allow it to preserve atomic memory if possible. Signed-off-by: Octavian Purdila Reviewed-by: Christoph Paasch Signed-off-by: David S. Miller --- include/net/tcp.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/net/tcp.h') diff --git a/include/net/tcp.h b/include/net/tcp.h index e80abe4486cb..7286db80e8b8 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -535,7 +535,7 @@ void tcp_retransmit_timer(struct sock *sk); void tcp_xmit_retransmit_queue(struct sock *); void tcp_simple_retransmit(struct sock *); int tcp_trim_head(struct sock *, struct sk_buff *, u32); -int tcp_fragment(struct sock *, struct sk_buff *, u32, unsigned int); +int tcp_fragment(struct sock *, struct sk_buff *, u32, unsigned int, gfp_t); void tcp_send_probe0(struct sock *); void tcp_send_partial(struct sock *); -- cgit v1.2.2