Signed-off-by: Eric Dumazet <eduma...@google.com> --- include/net/netns/ipv4.h | 1 + include/net/tcp.h | 9 ++++----- net/ipv4/sysctl_net_ipv4.c | 18 +++++++++--------- net/ipv4/tcp_input.c | 13 +++++-------- net/ipv4/tcp_ipv4.c | 1 + 5 files changed, 20 insertions(+), 22 deletions(-)
diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h index 63f91d52cbc0ad35d8e04a8da0d9f57aa960bcb0..9dbb07d4eff465428817831e55c6a4922b7208fb 100644 --- a/include/net/netns/ipv4.h +++ b/include/net/netns/ipv4.h @@ -140,6 +140,7 @@ struct netns_ipv4 { int sysctl_tcp_max_reordering; int sysctl_tcp_dsack; int sysctl_tcp_app_win; + int sysctl_tcp_adv_win_scale; struct inet_timewait_death_row tcp_death_row; int sysctl_max_syn_backlog; int sysctl_tcp_fastopen; diff --git a/include/net/tcp.h b/include/net/tcp.h index c6bee85a3dec0dea6d4402d89184ade02a637a2e..2572b57682987dd5f3700ed47d63e7238946b9a8 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -244,7 +244,6 @@ extern int sysctl_tcp_max_orphans; extern long sysctl_tcp_mem[3]; extern int sysctl_tcp_wmem[3]; extern int sysctl_tcp_rmem[3]; -extern int sysctl_tcp_adv_win_scale; extern int sysctl_tcp_frto; extern int sysctl_tcp_nometrics_save; extern int sysctl_tcp_moderate_rcvbuf; @@ -1308,9 +1307,9 @@ void tcp_select_initial_window(int __space, __u32 mss, __u32 *rcv_wnd, __u32 *window_clamp, int wscale_ok, __u8 *rcv_wscale, __u32 init_rcv_wnd); -static inline int tcp_win_from_space(int space) +static inline int tcp_win_from_space(const struct sock *sk, int space) { - int tcp_adv_win_scale = sysctl_tcp_adv_win_scale; + int tcp_adv_win_scale = sock_net(sk)->ipv4.sysctl_tcp_adv_win_scale; return tcp_adv_win_scale <= 0 ? (space>>(-tcp_adv_win_scale)) : @@ -1320,13 +1319,13 @@ static inline int tcp_win_from_space(int space) /* Note: caller must be prepared to deal with negative returns */ static inline int tcp_space(const struct sock *sk) { - return tcp_win_from_space(sk->sk_rcvbuf - + return tcp_win_from_space(sk, sk->sk_rcvbuf - atomic_read(&sk->sk_rmem_alloc)); } static inline int tcp_full_space(const struct sock *sk) { - return tcp_win_from_space(sk->sk_rcvbuf); + return tcp_win_from_space(sk, sk->sk_rcvbuf); } extern void tcp_openreq_init_rwin(struct request_sock *req, diff --git a/net/ipv4/sysctl_net_ipv4.c b/net/ipv4/sysctl_net_ipv4.c index e057788834a99cf99e141a602ddbe19b8e6fce3c..a95123e1e7da706c88bf5553b7d8ef6c2653ab50 100644 --- a/net/ipv4/sysctl_net_ipv4.c +++ b/net/ipv4/sysctl_net_ipv4.c @@ -437,15 +437,6 @@ static struct ctl_table ipv4_table[] = { .proc_handler = proc_dointvec_minmax, .extra1 = &one, }, - { - .procname = "tcp_adv_win_scale", - .data = &sysctl_tcp_adv_win_scale, - .maxlen = sizeof(int), - .mode = 0644, - .proc_handler = proc_dointvec_minmax, - .extra1 = &tcp_adv_win_scale_min, - .extra2 = &tcp_adv_win_scale_max, - }, { .procname = "tcp_frto", .data = &sysctl_tcp_frto, @@ -1145,6 +1136,15 @@ static struct ctl_table ipv4_net_table[] = { .mode = 0644, .proc_handler = proc_dointvec }, + { + .procname = "tcp_adv_win_scale", + .data = &init_net.ipv4.sysctl_tcp_adv_win_scale, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = &tcp_adv_win_scale_min, + .extra2 = &tcp_adv_win_scale_max, + }, { } }; diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 06a8c27e1a690e3b26cb6773320bafa31b06d3b3..d01f390da23dcd4100271b150bd8bc143f7328cf 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -77,9 +77,6 @@ #include <linux/errqueue.h> #include <trace/events/tcp.h> -int sysctl_tcp_adv_win_scale __read_mostly = 1; -EXPORT_SYMBOL(sysctl_tcp_adv_win_scale); - /* rfc5961 challenge ack rate limiting */ int sysctl_tcp_challenge_ack_limit = 1000; @@ -361,8 +358,8 @@ 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) >> 1; - int window = tcp_win_from_space(sysctl_tcp_rmem[2]) >> 1; + int truesize = tcp_win_from_space(sk, skb->truesize) >> 1; + int window = tcp_win_from_space(sk, sysctl_tcp_rmem[2]) >> 1; while (tp->rcv_ssthresh <= window) { if (truesize <= skb->len) @@ -387,7 +384,7 @@ static void tcp_grow_window(struct sock *sk, const struct sk_buff *skb) /* Check #2. Increase window, if skb with such overhead * will fit to rcvbuf in future. */ - if (tcp_win_from_space(skb->truesize) <= skb->len) + if (tcp_win_from_space(sk, skb->truesize) <= skb->len) incr = 2 * tp->advmss; else incr = __tcp_grow_window(sk, skb); @@ -628,7 +625,7 @@ void tcp_rcv_space_adjust(struct sock *sk) } rcvmem = SKB_TRUESIZE(tp->advmss + MAX_TCP_HEADER); - while (tcp_win_from_space(rcvmem) < tp->advmss) + while (tcp_win_from_space(sk, rcvmem) < tp->advmss) rcvmem += 128; rcvbuf = min(rcvwin / tp->advmss * rcvmem, sysctl_tcp_rmem[2]); @@ -4789,7 +4786,7 @@ tcp_collapse(struct sock *sk, struct sk_buff_head *list, struct rb_root *root, * overlaps to the next one. */ if (!(TCP_SKB_CB(skb)->tcp_flags & (TCPHDR_SYN | TCPHDR_FIN)) && - (tcp_win_from_space(skb->truesize) > skb->len || + (tcp_win_from_space(sk, skb->truesize) > skb->len || before(TCP_SKB_CB(skb)->seq, start))) { end_of_skbs = false; break; diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 189664ebd28e4cda7ef40a47591c3bd8cac3574b..1fe30fb99308b3e3fd07509b509b0e3727cc5d44 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -2491,6 +2491,7 @@ static int __net_init tcp_sk_init(struct net *net) net->ipv4.sysctl_tcp_max_reordering = 300; net->ipv4.sysctl_tcp_dsack = 1; net->ipv4.sysctl_tcp_app_win = 31; + net->ipv4.sysctl_tcp_adv_win_scale = 1; net->ipv4.sysctl_tcp_fastopen = TFO_CLIENT_ENABLE; spin_lock_init(&net->ipv4.tcp_fastopen_ctx_lock); -- 2.15.0.rc2.357.g7e34df9404-goog