diff options
Diffstat (limited to 'net/ipv4/tcp_output.c')
| -rw-r--r-- | net/ipv4/tcp_output.c | 2043 |
1 files changed, 1435 insertions, 608 deletions
diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 26a2948dca95..479afb714bdf 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -1,3 +1,4 @@ +// SPDX-License-Identifier: GPL-2.0-only /* * INET An implementation of the TCP/IP protocol suite for the LINUX * operating system. INET is implemented using the BSD Socket @@ -37,11 +38,17 @@ #define pr_fmt(fmt) "TCP: " fmt #include <net/tcp.h> +#include <net/tcp_ecn.h> +#include <net/mptcp.h> +#include <net/smc.h> +#include <net/proto_memory.h> +#include <net/psp.h> #include <linux/compiler.h> #include <linux/gfp.h> #include <linux/module.h> #include <linux/static_key.h> +#include <linux/skbuff_ref.h> #include <trace/events/tcp.h> @@ -52,12 +59,8 @@ void tcp_mstamp_refresh(struct tcp_sock *tp) { u64 val = tcp_clock_ns(); - if (val > tp->tcp_clock_cache) - tp->tcp_clock_cache = val; - - val = div_u64(val, NSEC_PER_USEC); - if (val > tp->tcp_mstamp) - tp->tcp_mstamp = val; + tp->tcp_clock_cache = val; + tp->tcp_mstamp = div_u64(val, NSEC_PER_USEC); } static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle, @@ -70,17 +73,21 @@ static void tcp_event_new_data_sent(struct sock *sk, struct sk_buff *skb) struct tcp_sock *tp = tcp_sk(sk); unsigned int prior_packets = tp->packets_out; - tp->snd_nxt = TCP_SKB_CB(skb)->end_seq; + WRITE_ONCE(tp->snd_nxt, TCP_SKB_CB(skb)->end_seq); __skb_unlink(skb, &sk->sk_write_queue); tcp_rbtree_insert(&sk->tcp_rtx_queue, skb); + if (tp->highest_sack == NULL) + tp->highest_sack = skb; + tp->packets_out += tcp_skb_pcount(skb); if (!prior_packets || icsk->icsk_pending == ICSK_TIME_LOSS_PROBE) tcp_rearm_rto(sk); NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPORIGDATASENT, tcp_skb_pcount(skb)); + tcp_check_space(sk); } /* SND.NXT, if window was not shrunk or the amount of shrunk was less than one @@ -141,7 +148,7 @@ void tcp_cwnd_restart(struct sock *sk, s32 delta) { struct tcp_sock *tp = tcp_sk(sk); u32 restart_cwnd = tcp_init_cwnd(tp, __sk_dst_get(sk)); - u32 cwnd = tp->snd_cwnd; + u32 cwnd = tcp_snd_cwnd(tp); tcp_ca_event(sk, CA_EVENT_CWND_RESTART); @@ -150,7 +157,7 @@ void tcp_cwnd_restart(struct sock *sk, s32 delta) while ((delta -= inet_csk(sk)->icsk_rto) > 0 && cwnd > restart_cwnd) cwnd >>= 1; - tp->snd_cwnd = max(cwnd, restart_cwnd); + tcp_snd_cwnd_set(tp, max(cwnd, restart_cwnd)); tp->snd_cwnd_stamp = tcp_jiffies32; tp->snd_cwnd_used = 0; } @@ -168,29 +175,28 @@ static void tcp_event_data_sent(struct tcp_sock *tp, tp->lsndtime = now; /* If it is a reply for ato after last received - * packet, enter pingpong mode. + * packet, increase pingpong count. */ if ((u32)(now - icsk->icsk_ack.lrcvtime) < icsk->icsk_ack.ato) - icsk->icsk_ack.pingpong = 1; + inet_csk_inc_pingpong_cnt(sk); } /* Account for an ACK we sent. */ -static inline void tcp_event_ack_sent(struct sock *sk, unsigned int pkts, - u32 rcv_nxt) +static inline void tcp_event_ack_sent(struct sock *sk, u32 rcv_nxt) { struct tcp_sock *tp = tcp_sk(sk); - if (unlikely(tp->compressed_ack > TCP_FASTRETRANS_THRESH)) { + if (unlikely(tp->compressed_ack)) { NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPACKCOMPRESSED, - tp->compressed_ack - TCP_FASTRETRANS_THRESH); - tp->compressed_ack = TCP_FASTRETRANS_THRESH; + tp->compressed_ack); + tp->compressed_ack = 0; if (hrtimer_try_to_cancel(&tp->compressed_ack_timer) == 1) __sock_put(sk); } if (unlikely(rcv_nxt != tp->rcv_nxt)) return; /* Special ACK sent by DCTCP to reflect ECN */ - tcp_dec_quickack_mode(sk, pkts); + tcp_dec_quickack_mode(sk); inet_csk_clear_xmit_timer(sk, ICSK_TIME_DACK); } @@ -202,16 +208,17 @@ static inline void tcp_event_ack_sent(struct sock *sk, unsigned int pkts, * This MUST be enforced by all callers. */ void tcp_select_initial_window(const struct sock *sk, int __space, __u32 mss, - __u32 *rcv_wnd, __u32 *window_clamp, + __u32 *rcv_wnd, __u32 *__window_clamp, int wscale_ok, __u8 *rcv_wscale, __u32 init_rcv_wnd) { unsigned int space = (__space < 0 ? 0 : __space); + u32 window_clamp = READ_ONCE(*__window_clamp); /* If no clamp set the clamp to the max possible scaled window */ - if (*window_clamp == 0) - (*window_clamp) = (U16_MAX << TCP_MAX_WSCALE); - space = min(*window_clamp, space); + if (window_clamp == 0) + window_clamp = (U16_MAX << TCP_MAX_WSCALE); + space = min(window_clamp, space); /* Quantize space offering to a multiple of mss if possible. */ if (space > mss) @@ -225,10 +232,10 @@ void tcp_select_initial_window(const struct sock *sk, int __space, __u32 mss, * which we interpret as a sign the remote TCP is not * misinterpreting the window field as a signed quantity. */ - if (sock_net(sk)->ipv4.sysctl_tcp_workaround_signed_windows) + if (READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_workaround_signed_windows)) (*rcv_wnd) = min(space, MAX_TCP_WINDOW); else - (*rcv_wnd) = min_t(u32, space, U16_MAX); + (*rcv_wnd) = space; if (init_rcv_wnd) *rcv_wnd = min(*rcv_wnd, init_rcv_wnd * mss); @@ -236,16 +243,17 @@ void tcp_select_initial_window(const struct sock *sk, int __space, __u32 mss, *rcv_wscale = 0; if (wscale_ok) { /* Set window scaling on max possible window */ - space = max_t(u32, space, sock_net(sk)->ipv4.sysctl_tcp_rmem[2]); - space = max_t(u32, space, sysctl_rmem_max); - space = min_t(u32, space, *window_clamp); + space = max_t(u32, space, READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_rmem[2])); + space = max_t(u32, space, READ_ONCE(sysctl_rmem_max)); + space = min_t(u32, space, window_clamp); *rcv_wscale = clamp_t(int, ilog2(space) - 15, 0, TCP_MAX_WSCALE); } /* Set the clamp no higher than max representable value */ - (*window_clamp) = min_t(__u32, U16_MAX << (*rcv_wscale), *window_clamp); + WRITE_ONCE(*__window_clamp, + min_t(__u32, U16_MAX << (*rcv_wscale), window_clamp)); } -EXPORT_SYMBOL(tcp_select_initial_window); +EXPORT_IPV6_MOD(tcp_select_initial_window); /* Chose a new window to advertise, update state in tcp_sock for the * socket, and return result with RFC1323 scaling applied. The return @@ -255,11 +263,22 @@ EXPORT_SYMBOL(tcp_select_initial_window); static u16 tcp_select_window(struct sock *sk) { struct tcp_sock *tp = tcp_sk(sk); + struct net *net = sock_net(sk); u32 old_win = tp->rcv_wnd; - u32 cur_win = tcp_receive_window(tp); - u32 new_win = __tcp_select_window(sk); + u32 cur_win, new_win; + + /* Make the window 0 if we failed to queue the data because we + * are out of memory. + */ + if (unlikely(inet_csk(sk)->icsk_ack.pending & ICSK_ACK_NOMEM)) { + tp->pred_flags = 0; + tp->rcv_wnd = 0; + tp->rcv_wup = tp->rcv_nxt; + return 0; + } - /* Never shrink the offered window */ + cur_win = tcp_receive_window(tp); + new_win = __tcp_select_window(sk); if (new_win < cur_win) { /* Danger Will Robinson! * Don't update rcv_wup/rcv_wnd here or else @@ -268,11 +287,14 @@ static u16 tcp_select_window(struct sock *sk) * * Relax Will Robinson. */ - if (new_win == 0) - NET_INC_STATS(sock_net(sk), - LINUX_MIB_TCPWANTZEROWINDOWADV); - new_win = ALIGN(cur_win, 1 << tp->rx_opt.rcv_wscale); + if (!READ_ONCE(net->ipv4.sysctl_tcp_shrink_window) || !tp->rx_opt.rcv_wscale) { + /* Never shrink the offered window */ + if (new_win == 0) + NET_INC_STATS(net, LINUX_MIB_TCPWANTZEROWINDOWADV); + new_win = ALIGN(cur_win, 1 << tp->rx_opt.rcv_wscale); + } } + tp->rcv_wnd = new_win; tp->rcv_wup = tp->rcv_nxt; @@ -280,7 +302,7 @@ static u16 tcp_select_window(struct sock *sk) * scaled window. */ if (!tp->rx_opt.rcv_wscale && - sock_net(sk)->ipv4.sysctl_tcp_workaround_signed_windows) + READ_ONCE(net->ipv4.sysctl_tcp_workaround_signed_windows)) new_win = min(new_win, MAX_TCP_WINDOW); else new_win = min(new_win, (65535U << tp->rx_opt.rcv_wscale)); @@ -292,69 +314,14 @@ static u16 tcp_select_window(struct sock *sk) if (new_win == 0) { tp->pred_flags = 0; if (old_win) - NET_INC_STATS(sock_net(sk), - LINUX_MIB_TCPTOZEROWINDOWADV); + NET_INC_STATS(net, LINUX_MIB_TCPTOZEROWINDOWADV); } else if (old_win == 0) { - NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPFROMZEROWINDOWADV); + NET_INC_STATS(net, LINUX_MIB_TCPFROMZEROWINDOWADV); } return new_win; } -/* Packet ECN state for a SYN-ACK */ -static void tcp_ecn_send_synack(struct sock *sk, struct sk_buff *skb) -{ - const struct tcp_sock *tp = tcp_sk(sk); - - TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_CWR; - if (!(tp->ecn_flags & TCP_ECN_OK)) - TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_ECE; - else if (tcp_ca_needs_ecn(sk) || - tcp_bpf_ca_needs_ecn(sk)) - INET_ECN_xmit(sk); -} - -/* Packet ECN state for a SYN. */ -static void tcp_ecn_send_syn(struct sock *sk, struct sk_buff *skb) -{ - struct tcp_sock *tp = tcp_sk(sk); - bool bpf_needs_ecn = tcp_bpf_ca_needs_ecn(sk); - bool use_ecn = sock_net(sk)->ipv4.sysctl_tcp_ecn == 1 || - tcp_ca_needs_ecn(sk) || bpf_needs_ecn; - - if (!use_ecn) { - const struct dst_entry *dst = __sk_dst_get(sk); - - if (dst && dst_feature(dst, RTAX_FEATURE_ECN)) - use_ecn = true; - } - - tp->ecn_flags = 0; - - if (use_ecn) { - TCP_SKB_CB(skb)->tcp_flags |= TCPHDR_ECE | TCPHDR_CWR; - tp->ecn_flags = TCP_ECN_OK; - if (tcp_ca_needs_ecn(sk) || bpf_needs_ecn) - INET_ECN_xmit(sk); - } -} - -static void tcp_ecn_clear_syn(struct sock *sk, struct sk_buff *skb) -{ - if (sock_net(sk)->ipv4.sysctl_tcp_ecn_fallback) - /* tp->ecn_flags are cleared at a later point in time when - * SYN ACK is ultimatively being received. - */ - TCP_SKB_CB(skb)->tcp_flags &= ~(TCPHDR_ECE | TCPHDR_CWR); -} - -static void -tcp_ecn_make_synack(const struct request_sock *req, struct tcphdr *th) -{ - if (inet_rsk(req)->ecn_ok) - th->ece = 1; -} - /* Set up ECN state for a packet on a ESTABLISHED socket that is about to * be sent. */ @@ -363,7 +330,15 @@ static void tcp_ecn_send(struct sock *sk, struct sk_buff *skb, { struct tcp_sock *tp = tcp_sk(sk); - if (tp->ecn_flags & TCP_ECN_OK) { + if (!tcp_ecn_mode_any(tp)) + return; + + if (tcp_ecn_mode_accecn(tp)) { + if (!tcp_accecn_ace_fail_recv(tp)) + INET_ECN_xmit(sk); + tcp_accecn_set_ace(tp, skb, th); + skb_shinfo(skb)->gso_type |= SKB_GSO_TCP_ACCECN; + } else { /* Not-retransmitted data segment: set ECT and inject CWR. */ if (skb->len != tcp_header_len && !before(TCP_SKB_CB(skb)->seq, tp->snd_nxt)) { @@ -385,14 +360,15 @@ static void tcp_ecn_send(struct sock *sk, struct sk_buff *skb, /* Constructs common control bits of non-data skb. If SYN/FIN is present, * auto increment end seqno. */ -static void tcp_init_nondata_skb(struct sk_buff *skb, u32 seq, u8 flags) +static void tcp_init_nondata_skb(struct sk_buff *skb, struct sock *sk, + u32 seq, u16 flags) { skb->ip_summed = CHECKSUM_PARTIAL; TCP_SKB_CB(skb)->tcp_flags = flags; - TCP_SKB_CB(skb)->sacked = 0; tcp_skb_pcount_set(skb, 1); + psp_enqueue_set_decrypted(sk, skb); TCP_SKB_CB(skb)->seq = seq; if (flags & (TCPHDR_SYN | TCPHDR_FIN)) @@ -405,12 +381,15 @@ static inline bool tcp_urg_mode(const struct tcp_sock *tp) return tp->snd_una != tp->snd_up; } -#define OPTION_SACK_ADVERTISE (1 << 0) -#define OPTION_TS (1 << 1) -#define OPTION_MD5 (1 << 2) -#define OPTION_WSCALE (1 << 3) -#define OPTION_FAST_OPEN_COOKIE (1 << 8) -#define OPTION_SMC (1 << 9) +#define OPTION_SACK_ADVERTISE BIT(0) +#define OPTION_TS BIT(1) +#define OPTION_MD5 BIT(2) +#define OPTION_WSCALE BIT(3) +#define OPTION_FAST_OPEN_COOKIE BIT(8) +#define OPTION_SMC BIT(9) +#define OPTION_MPTCP BIT(10) +#define OPTION_AO BIT(11) +#define OPTION_ACCECN BIT(12) static void smc_options_write(__be32 *ptr, u16 *options) { @@ -432,12 +411,210 @@ struct tcp_out_options { u16 mss; /* 0 to disable */ u8 ws; /* window scale, 0 to disable */ u8 num_sack_blocks; /* number of SACK blocks to include */ + u8 num_accecn_fields:7, /* number of AccECN fields needed */ + use_synack_ecn_bytes:1; /* Use synack_ecn_bytes or not */ u8 hash_size; /* bytes in hash_location */ + u8 bpf_opt_len; /* length of BPF hdr option */ __u8 *hash_location; /* temporary pointer, overloaded */ __u32 tsval, tsecr; /* need to include OPTION_TS */ struct tcp_fastopen_cookie *fastopen_cookie; /* Fast open cookie */ + struct mptcp_out_options mptcp; }; +static void mptcp_options_write(struct tcphdr *th, __be32 *ptr, + struct tcp_sock *tp, + struct tcp_out_options *opts) +{ +#if IS_ENABLED(CONFIG_MPTCP) + if (unlikely(OPTION_MPTCP & opts->options)) + mptcp_write_options(th, ptr, tp, &opts->mptcp); +#endif +} + +#ifdef CONFIG_CGROUP_BPF +static int bpf_skops_write_hdr_opt_arg0(struct sk_buff *skb, + enum tcp_synack_type synack_type) +{ + if (unlikely(!skb)) + return BPF_WRITE_HDR_TCP_CURRENT_MSS; + + if (unlikely(synack_type == TCP_SYNACK_COOKIE)) + return BPF_WRITE_HDR_TCP_SYNACK_COOKIE; + + return 0; +} + +/* req, syn_skb and synack_type are used when writing synack */ +static void bpf_skops_hdr_opt_len(struct sock *sk, struct sk_buff *skb, + struct request_sock *req, + struct sk_buff *syn_skb, + enum tcp_synack_type synack_type, + struct tcp_out_options *opts, + unsigned int *remaining) +{ + struct bpf_sock_ops_kern sock_ops; + int err; + + if (likely(!BPF_SOCK_OPS_TEST_FLAG(tcp_sk(sk), + BPF_SOCK_OPS_WRITE_HDR_OPT_CB_FLAG)) || + !*remaining) + return; + + /* *remaining has already been aligned to 4 bytes, so *remaining >= 4 */ + + /* init sock_ops */ + memset(&sock_ops, 0, offsetof(struct bpf_sock_ops_kern, temp)); + + sock_ops.op = BPF_SOCK_OPS_HDR_OPT_LEN_CB; + + if (req) { + /* The listen "sk" cannot be passed here because + * it is not locked. It would not make too much + * sense to do bpf_setsockopt(listen_sk) based + * on individual connection request also. + * + * Thus, "req" is passed here and the cgroup-bpf-progs + * of the listen "sk" will be run. + * + * "req" is also used here for fastopen even the "sk" here is + * a fullsock "child" sk. It is to keep the behavior + * consistent between fastopen and non-fastopen on + * the bpf programming side. + */ + sock_ops.sk = (struct sock *)req; + sock_ops.syn_skb = syn_skb; + } else { + sock_owned_by_me(sk); + + sock_ops.is_fullsock = 1; + sock_ops.is_locked_tcp_sock = 1; + sock_ops.sk = sk; + } + + sock_ops.args[0] = bpf_skops_write_hdr_opt_arg0(skb, synack_type); + sock_ops.remaining_opt_len = *remaining; + /* tcp_current_mss() does not pass a skb */ + if (skb) + bpf_skops_init_skb(&sock_ops, skb, 0); + + err = BPF_CGROUP_RUN_PROG_SOCK_OPS_SK(&sock_ops, sk); + + if (err || sock_ops.remaining_opt_len == *remaining) + return; + + opts->bpf_opt_len = *remaining - sock_ops.remaining_opt_len; + /* round up to 4 bytes */ + opts->bpf_opt_len = (opts->bpf_opt_len + 3) & ~3; + + *remaining -= opts->bpf_opt_len; +} + +static void bpf_skops_write_hdr_opt(struct sock *sk, struct sk_buff *skb, + struct request_sock *req, + struct sk_buff *syn_skb, + enum tcp_synack_type synack_type, + struct tcp_out_options *opts) +{ + u8 first_opt_off, nr_written, max_opt_len = opts->bpf_opt_len; + struct bpf_sock_ops_kern sock_ops; + int err; + + if (likely(!max_opt_len)) + return; + + memset(&sock_ops, 0, offsetof(struct bpf_sock_ops_kern, temp)); + + sock_ops.op = BPF_SOCK_OPS_WRITE_HDR_OPT_CB; + + if (req) { + sock_ops.sk = (struct sock *)req; + sock_ops.syn_skb = syn_skb; + } else { + sock_owned_by_me(sk); + + sock_ops.is_fullsock = 1; + sock_ops.is_locked_tcp_sock = 1; + sock_ops.sk = sk; + } + + sock_ops.args[0] = bpf_skops_write_hdr_opt_arg0(skb, synack_type); + sock_ops.remaining_opt_len = max_opt_len; + first_opt_off = tcp_hdrlen(skb) - max_opt_len; + bpf_skops_init_skb(&sock_ops, skb, first_opt_off); + + err = BPF_CGROUP_RUN_PROG_SOCK_OPS_SK(&sock_ops, sk); + + if (err) + nr_written = 0; + else + nr_written = max_opt_len - sock_ops.remaining_opt_len; + + if (nr_written < max_opt_len) + memset(skb->data + first_opt_off + nr_written, TCPOPT_NOP, + max_opt_len - nr_written); +} +#else +static void bpf_skops_hdr_opt_len(struct sock *sk, struct sk_buff *skb, + struct request_sock *req, + struct sk_buff *syn_skb, + enum tcp_synack_type synack_type, + struct tcp_out_options *opts, + unsigned int *remaining) +{ +} + +static void bpf_skops_write_hdr_opt(struct sock *sk, struct sk_buff *skb, + struct request_sock *req, + struct sk_buff *syn_skb, + enum tcp_synack_type synack_type, + struct tcp_out_options *opts) +{ +} +#endif + +static __be32 *process_tcp_ao_options(struct tcp_sock *tp, + const struct tcp_request_sock *tcprsk, + struct tcp_out_options *opts, + struct tcp_key *key, __be32 *ptr) +{ +#ifdef CONFIG_TCP_AO + u8 maclen = tcp_ao_maclen(key->ao_key); + + if (tcprsk) { + u8 aolen = maclen + sizeof(struct tcp_ao_hdr); + + *ptr++ = htonl((TCPOPT_AO << 24) | (aolen << 16) | + (tcprsk->ao_keyid << 8) | + (tcprsk->ao_rcv_next)); + } else { + struct tcp_ao_key *rnext_key; + struct tcp_ao_info *ao_info; + + ao_info = rcu_dereference_check(tp->ao_info, + lockdep_sock_is_held(&tp->inet_conn.icsk_inet.sk)); + rnext_key = READ_ONCE(ao_info->rnext_key); + if (WARN_ON_ONCE(!rnext_key)) + return ptr; + *ptr++ = htonl((TCPOPT_AO << 24) | + (tcp_ao_len(key->ao_key) << 16) | + (key->ao_key->sndid << 8) | + (rnext_key->rcvid)); + } + opts->hash_location = (__u8 *)ptr; + ptr += maclen / sizeof(*ptr); + if (unlikely(maclen % sizeof(*ptr))) { + memset(ptr, TCPOPT_NOP, sizeof(*ptr)); + ptr++; + } +#endif + return ptr; +} + +/* Initial values for AccECN option, ordered is based on ECN field bits + * similar to received_ecn_bytes. Used for SYN/ACK AccECN option. + */ +static const u32 synack_ecn_bytes[3] = { 0, 0, 0 }; + /* Write previously computed TCP options to the packet. * * Beware: Something in the Internet is very sensitive to the ordering of @@ -451,19 +628,25 @@ struct tcp_out_options { * At least SACK_PERM as the first option is known to lead to a disaster * (but it may well be that other scenarios fail similarly). */ -static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp, - struct tcp_out_options *opts) +static void tcp_options_write(struct tcphdr *th, struct tcp_sock *tp, + const struct tcp_request_sock *tcprsk, + struct tcp_out_options *opts, + struct tcp_key *key) { + u8 leftover_highbyte = TCPOPT_NOP; /* replace 1st NOP if avail */ + u8 leftover_lowbyte = TCPOPT_NOP; /* replace 2nd NOP in succession */ + __be32 *ptr = (__be32 *)(th + 1); u16 options = opts->options; /* mungable copy */ - if (unlikely(OPTION_MD5 & options)) { + if (tcp_key_is_md5(key)) { *ptr++ = htonl((TCPOPT_NOP << 24) | (TCPOPT_NOP << 16) | (TCPOPT_MD5SIG << 8) | TCPOLEN_MD5SIG); /* overload cookie hash location */ opts->hash_location = (__u8 *)ptr; ptr += 4; + } else if (tcp_key_is_ao(key)) { + ptr = process_tcp_ao_options(tp, tcprsk, opts, key, ptr); } - if (unlikely(opts->mss)) { *ptr++ = htonl((TCPOPT_MSS << 24) | (TCPOLEN_MSS << 16) | @@ -487,15 +670,75 @@ static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp, *ptr++ = htonl(opts->tsecr); } + if (OPTION_ACCECN & options) { + const u32 *ecn_bytes = opts->use_synack_ecn_bytes ? + synack_ecn_bytes : + tp->received_ecn_bytes; + const u8 ect0_idx = INET_ECN_ECT_0 - 1; + const u8 ect1_idx = INET_ECN_ECT_1 - 1; + const u8 ce_idx = INET_ECN_CE - 1; + u32 e0b; + u32 e1b; + u32 ceb; + u8 len; + + e0b = ecn_bytes[ect0_idx] + TCP_ACCECN_E0B_INIT_OFFSET; + e1b = ecn_bytes[ect1_idx] + TCP_ACCECN_E1B_INIT_OFFSET; + ceb = ecn_bytes[ce_idx] + TCP_ACCECN_CEB_INIT_OFFSET; + len = TCPOLEN_ACCECN_BASE + + opts->num_accecn_fields * TCPOLEN_ACCECN_PERFIELD; + + if (opts->num_accecn_fields == 2) { + *ptr++ = htonl((TCPOPT_ACCECN1 << 24) | (len << 16) | + ((e1b >> 8) & 0xffff)); + *ptr++ = htonl(((e1b & 0xff) << 24) | + (ceb & 0xffffff)); + } else if (opts->num_accecn_fields == 1) { + *ptr++ = htonl((TCPOPT_ACCECN1 << 24) | (len << 16) | + ((e1b >> 8) & 0xffff)); + leftover_highbyte = e1b & 0xff; + leftover_lowbyte = TCPOPT_NOP; + } else if (opts->num_accecn_fields == 0) { + leftover_highbyte = TCPOPT_ACCECN1; + leftover_lowbyte = len; + } else if (opts->num_accecn_fields == 3) { + *ptr++ = htonl((TCPOPT_ACCECN1 << 24) | (len << 16) | + ((e1b >> 8) & 0xffff)); + *ptr++ = htonl(((e1b & 0xff) << 24) | + (ceb & 0xffffff)); + *ptr++ = htonl(((e0b & 0xffffff) << 8) | + TCPOPT_NOP); + } + if (tp) { + tp->accecn_minlen = 0; + tp->accecn_opt_tstamp = tp->tcp_mstamp; + if (tp->accecn_opt_demand) + tp->accecn_opt_demand--; + } + } + if (unlikely(OPTION_SACK_ADVERTISE & options)) { - *ptr++ = htonl((TCPOPT_NOP << 24) | - (TCPOPT_NOP << 16) | + *ptr++ = htonl((leftover_highbyte << 24) | + (leftover_lowbyte << 16) | (TCPOPT_SACK_PERM << 8) | TCPOLEN_SACK_PERM); + leftover_highbyte = TCPOPT_NOP; + leftover_lowbyte = TCPOPT_NOP; } if (unlikely(OPTION_WSCALE & options)) { - *ptr++ = htonl((TCPOPT_NOP << 24) | + u8 highbyte = TCPOPT_NOP; + + /* Do not split the leftover 2-byte to fit into a single + * NOP, i.e., replace this NOP only when 1 byte is leftover + * within leftover_highbyte. + */ + if (unlikely(leftover_highbyte != TCPOPT_NOP && + leftover_lowbyte == TCPOPT_NOP)) { + highbyte = leftover_highbyte; + leftover_highbyte = TCPOPT_NOP; + } + *ptr++ = htonl((highbyte << 24) | (TCPOPT_WINDOW << 16) | (TCPOLEN_WINDOW << 8) | opts->ws); @@ -506,11 +749,13 @@ static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp, tp->duplicate_sack : tp->selective_acks; int this_sack; - *ptr++ = htonl((TCPOPT_NOP << 24) | - (TCPOPT_NOP << 16) | + *ptr++ = htonl((leftover_highbyte << 24) | + (leftover_lowbyte << 16) | (TCPOPT_SACK << 8) | (TCPOLEN_SACK_BASE + (opts->num_sack_blocks * TCPOLEN_SACK_PERBLOCK))); + leftover_highbyte = TCPOPT_NOP; + leftover_lowbyte = TCPOPT_NOP; for (this_sack = 0; this_sack < opts->num_sack_blocks; ++this_sack) { @@ -519,6 +764,14 @@ static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp, } tp->rx_opt.dsack = 0; + } else if (unlikely(leftover_highbyte != TCPOPT_NOP || + leftover_lowbyte != TCPOPT_NOP)) { + *ptr++ = htonl((leftover_highbyte << 24) | + (leftover_lowbyte << 16) | + (TCPOPT_NOP << 8) | + TCPOPT_NOP); + leftover_highbyte = TCPOPT_NOP; + leftover_lowbyte = TCPOPT_NOP; } if (unlikely(OPTION_FAST_OPEN_COOKIE & options)) { @@ -546,63 +799,159 @@ static void tcp_options_write(__be32 *ptr, struct tcp_sock *tp, } smc_options_write(ptr, &options); + + mptcp_options_write(th, ptr, tp, opts); } -static void smc_set_option(const struct tcp_sock *tp, +static void smc_set_option(struct tcp_sock *tp, struct tcp_out_options *opts, unsigned int *remaining) { #if IS_ENABLED(CONFIG_SMC) - if (static_branch_unlikely(&tcp_have_smc)) { - if (tp->syn_smc) { - if (*remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) { - opts->options |= OPTION_SMC; - *remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED; - } + if (static_branch_unlikely(&tcp_have_smc) && tp->syn_smc) { + tp->syn_smc = !!smc_call_hsbpf(1, tp, syn_option); + /* re-check syn_smc */ + if (tp->syn_smc && + *remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) { + opts->options |= OPTION_SMC; + *remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED; } } #endif } static void smc_set_option_cond(const struct tcp_sock *tp, - const struct inet_request_sock *ireq, + struct inet_request_sock *ireq, struct tcp_out_options *opts, unsigned int *remaining) { #if IS_ENABLED(CONFIG_SMC) - if (static_branch_unlikely(&tcp_have_smc)) { - if (tp->syn_smc && ireq->smc_ok) { - if (*remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) { - opts->options |= OPTION_SMC; - *remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED; - } + if (static_branch_unlikely(&tcp_have_smc) && tp->syn_smc && ireq->smc_ok) { + ireq->smc_ok = !!smc_call_hsbpf(1, tp, synack_option, ireq); + /* re-check smc_ok */ + if (ireq->smc_ok && + *remaining >= TCPOLEN_EXP_SMC_BASE_ALIGNED) { + opts->options |= OPTION_SMC; + *remaining -= TCPOLEN_EXP_SMC_BASE_ALIGNED; } } #endif } +static void mptcp_set_option_cond(const struct request_sock *req, + struct tcp_out_options *opts, + unsigned int *remaining) +{ + if (rsk_is_mptcp(req)) { + unsigned int size; + + if (mptcp_synack_options(req, &size, &opts->mptcp)) { + if (*remaining >= size) { + opts->options |= OPTION_MPTCP; + *remaining -= size; + } + } + } +} + +static u32 tcp_synack_options_combine_saving(struct tcp_out_options *opts) +{ + /* How much there's room for combining with the alignment padding? */ + if ((opts->options & (OPTION_SACK_ADVERTISE | OPTION_TS)) == + OPTION_SACK_ADVERTISE) + return 2; + else if (opts->options & OPTION_WSCALE) + return 1; + return 0; +} + +/* Calculates how long AccECN option will fit to @remaining option space. + * + * AccECN option can sometimes replace NOPs used for alignment of other + * TCP options (up to @max_combine_saving available). + * + * Only solutions with at least @required AccECN fields are accepted. + * + * Returns: The size of the AccECN option excluding space repurposed from + * the alignment of the other options. + */ +static int tcp_options_fit_accecn(struct tcp_out_options *opts, int required, + int remaining) +{ + int size = TCP_ACCECN_MAXSIZE; + int sack_blocks_reduce = 0; + int max_combine_saving; + int rem = remaining; + int align_size; + + if (opts->use_synack_ecn_bytes) + max_combine_saving = tcp_synack_options_combine_saving(opts); + else + max_combine_saving = opts->num_sack_blocks > 0 ? 2 : 0; + opts->num_accecn_fields = TCP_ACCECN_NUMFIELDS; + while (opts->num_accecn_fields >= required) { + /* Pad to dword if cannot combine */ + if ((size & 0x3) > max_combine_saving) + align_size = ALIGN(size, 4); + else + align_size = ALIGN_DOWN(size, 4); + + if (rem >= align_size) { + size = align_size; + break; + } else if (opts->num_accecn_fields == required && + opts->num_sack_blocks > 2 && + required > 0) { + /* Try to fit the option by removing one SACK block */ + opts->num_sack_blocks--; + sack_blocks_reduce++; + rem = rem + TCPOLEN_SACK_PERBLOCK; + + opts->num_accecn_fields = TCP_ACCECN_NUMFIELDS; + size = TCP_ACCECN_MAXSIZE; + continue; + } + + opts->num_accecn_fields--; + size -= TCPOLEN_ACCECN_PERFIELD; + } + if (sack_blocks_reduce > 0) { + if (opts->num_accecn_fields >= required) + size -= sack_blocks_reduce * TCPOLEN_SACK_PERBLOCK; + else + opts->num_sack_blocks += sack_blocks_reduce; + } + if (opts->num_accecn_fields < required) + return 0; + + opts->options |= OPTION_ACCECN; + return size; +} + /* Compute TCP options for SYN packets. This is not the final * network wire format yet. */ static unsigned int tcp_syn_options(struct sock *sk, struct sk_buff *skb, struct tcp_out_options *opts, - struct tcp_md5sig_key **md5) + struct tcp_key *key) { struct tcp_sock *tp = tcp_sk(sk); unsigned int remaining = MAX_TCP_OPTION_SPACE; struct tcp_fastopen_request *fastopen = tp->fastopen_req; + bool timestamps; - *md5 = NULL; -#ifdef CONFIG_TCP_MD5SIG - if (static_key_false(&tcp_md5_needed) && - rcu_access_pointer(tp->md5sig_info)) { - *md5 = tp->af_specific->md5_lookup(sk, sk); - if (*md5) { - opts->options |= OPTION_MD5; - remaining -= TCPOLEN_MD5SIG_ALIGNED; + /* Better than switch (key.type) as it has static branches */ + if (tcp_key_is_md5(key)) { + timestamps = false; + opts->options |= OPTION_MD5; + remaining -= TCPOLEN_MD5SIG_ALIGNED; + } else { + timestamps = READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_timestamps); + if (tcp_key_is_ao(key)) { + opts->options |= OPTION_AO; + remaining -= tcp_ao_len_aligned(key->ao_key); } } -#endif /* We always get an MSS option. The option bytes which will be seen in * normal data packets should timestamps be used, must be in the MSS @@ -616,18 +965,18 @@ static unsigned int tcp_syn_options(struct sock *sk, struct sk_buff *skb, opts->mss = tcp_advertise_mss(sk); remaining -= TCPOLEN_MSS_ALIGNED; - if (likely(sock_net(sk)->ipv4.sysctl_tcp_timestamps && !*md5)) { + if (likely(timestamps)) { opts->options |= OPTION_TS; - opts->tsval = tcp_skb_timestamp(skb) + tp->tsoffset; + opts->tsval = tcp_skb_timestamp_ts(tp->tcp_usec_ts, skb) + tp->tsoffset; opts->tsecr = tp->rx_opt.ts_recent; remaining -= TCPOLEN_TSTAMP_ALIGNED; } - if (likely(sock_net(sk)->ipv4.sysctl_tcp_window_scaling)) { + if (likely(READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_window_scaling))) { opts->ws = tp->rx_opt.rcv_wscale; opts->options |= OPTION_WSCALE; remaining -= TCPOLEN_WSCALE_ALIGNED; } - if (likely(sock_net(sk)->ipv4.sysctl_tcp_sack)) { + if (likely(READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_sack))) { opts->options |= OPTION_SACK_ADVERTISE; if (unlikely(!(OPTION_TS & opts->options))) remaining -= TCPOLEN_SACKPERM_ALIGNED; @@ -650,6 +999,33 @@ static unsigned int tcp_syn_options(struct sock *sk, struct sk_buff *skb, smc_set_option(tp, opts, &remaining); + if (sk_is_mptcp(sk)) { + unsigned int size; + + if (mptcp_syn_options(sk, skb, &size, &opts->mptcp)) { + if (remaining >= size) { + opts->options |= OPTION_MPTCP; + remaining -= size; + } + } + } + + /* Simultaneous open SYN/ACK needs AccECN option but not SYN. + * It is attempted to negotiate the use of AccECN also on the first + * retransmitted SYN, as mentioned in "3.1.4.1. Retransmitted SYNs" + * of AccECN draft. + */ + if (unlikely((TCP_SKB_CB(skb)->tcp_flags & TCPHDR_ACK) && + tcp_ecn_mode_accecn(tp) && + inet_csk(sk)->icsk_retransmits < 2 && + READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_ecn_option) && + remaining >= TCPOLEN_ACCECN_BASE)) { + opts->use_synack_ecn_bytes = 1; + remaining -= tcp_options_fit_accecn(opts, 0, remaining); + } + + bpf_skops_hdr_opt_len(sk, skb, NULL, NULL, 0, opts, &remaining); + return MAX_TCP_OPTION_SPACE - remaining; } @@ -658,14 +1034,16 @@ static unsigned int tcp_synack_options(const struct sock *sk, struct request_sock *req, unsigned int mss, struct sk_buff *skb, struct tcp_out_options *opts, - const struct tcp_md5sig_key *md5, - struct tcp_fastopen_cookie *foc) + const struct tcp_key *key, + struct tcp_fastopen_cookie *foc, + enum tcp_synack_type synack_type, + struct sk_buff *syn_skb) { struct inet_request_sock *ireq = inet_rsk(req); unsigned int remaining = MAX_TCP_OPTION_SPACE; + struct tcp_request_sock *treq = tcp_rsk(req); -#ifdef CONFIG_TCP_MD5SIG - if (md5) { + if (tcp_key_is_md5(key)) { opts->options |= OPTION_MD5; remaining -= TCPOLEN_MD5SIG_ALIGNED; @@ -674,9 +1052,13 @@ static unsigned int tcp_synack_options(const struct sock *sk, * rather than TS in order to fit in better with old, * buggy kernels, but that was deemed to be unnecessary. */ + if (synack_type != TCP_SYNACK_COOKIE) + ireq->tstamp_ok &= !ireq->sack_ok; + } else if (tcp_key_is_ao(key)) { + opts->options |= OPTION_AO; + remaining -= tcp_ao_len_aligned(key->ao_key); ireq->tstamp_ok &= !ireq->sack_ok; } -#endif /* We always send an MSS option. */ opts->mss = mss; @@ -689,7 +1071,14 @@ static unsigned int tcp_synack_options(const struct sock *sk, } if (likely(ireq->tstamp_ok)) { opts->options |= OPTION_TS; - opts->tsval = tcp_skb_timestamp(skb) + tcp_rsk(req)->ts_off; + opts->tsval = tcp_skb_timestamp_ts(tcp_rsk(req)->req_usec_ts, skb) + + tcp_rsk(req)->ts_off; + if (!tcp_rsk(req)->snt_tsval_first) { + if (!opts->tsval) + opts->tsval = ~0U; + tcp_rsk(req)->snt_tsval_first = opts->tsval; + } + WRITE_ONCE(tcp_rsk(req)->snt_tsval_last, opts->tsval); opts->tsecr = req->ts_recent; remaining -= TCPOLEN_TSTAMP_ALIGNED; } @@ -711,8 +1100,20 @@ static unsigned int tcp_synack_options(const struct sock *sk, } } + mptcp_set_option_cond(req, opts, &remaining); + smc_set_option_cond(tcp_sk(sk), ireq, opts, &remaining); + if (treq->accecn_ok && + READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_ecn_option) && + req->num_timeout < 1 && remaining >= TCPOLEN_ACCECN_BASE) { + opts->use_synack_ecn_bytes = 1; + remaining -= tcp_options_fit_accecn(opts, 0, remaining); + } + + bpf_skops_hdr_opt_len((struct sock *)sk, skb, req, syn_skb, + synack_type, opts, &remaining); + return MAX_TCP_OPTION_SPACE - remaining; } @@ -721,7 +1122,7 @@ static unsigned int tcp_synack_options(const struct sock *sk, */ static unsigned int tcp_established_options(struct sock *sk, struct sk_buff *skb, struct tcp_out_options *opts, - struct tcp_md5sig_key **md5) + struct tcp_key *key) { struct tcp_sock *tp = tcp_sk(sk); unsigned int size = 0; @@ -729,34 +1130,78 @@ static unsigned int tcp_established_options(struct sock *sk, struct sk_buff *skb opts->options = 0; - *md5 = NULL; -#ifdef CONFIG_TCP_MD5SIG - if (static_key_false(&tcp_md5_needed) && - rcu_access_pointer(tp->md5sig_info)) { - *md5 = tp->af_specific->md5_lookup(sk, sk); - if (*md5) { - opts->options |= OPTION_MD5; - size += TCPOLEN_MD5SIG_ALIGNED; - } + /* Better than switch (key.type) as it has static branches */ + if (tcp_key_is_md5(key)) { + opts->options |= OPTION_MD5; + size += TCPOLEN_MD5SIG_ALIGNED; + } else if (tcp_key_is_ao(key)) { + opts->options |= OPTION_AO; + size += tcp_ao_len_aligned(key->ao_key); } -#endif if (likely(tp->rx_opt.tstamp_ok)) { opts->options |= OPTION_TS; - opts->tsval = skb ? tcp_skb_timestamp(skb) + tp->tsoffset : 0; + opts->tsval = skb ? tcp_skb_timestamp_ts(tp->tcp_usec_ts, skb) + + tp->tsoffset : 0; opts->tsecr = tp->rx_opt.ts_recent; size += TCPOLEN_TSTAMP_ALIGNED; } + /* MPTCP options have precedence over SACK for the limited TCP + * option space because a MPTCP connection would be forced to + * fall back to regular TCP if a required multipath option is + * missing. SACK still gets a chance to use whatever space is + * left. + */ + if (sk_is_mptcp(sk)) { + unsigned int remaining = MAX_TCP_OPTION_SPACE - size; + unsigned int opt_size = 0; + + if (mptcp_established_options(sk, skb, &opt_size, remaining, + &opts->mptcp)) { + opts->options |= OPTION_MPTCP; + size += opt_size; + } + } + eff_sacks = tp->rx_opt.num_sacks + tp->rx_opt.dsack; if (unlikely(eff_sacks)) { const unsigned int remaining = MAX_TCP_OPTION_SPACE - size; - opts->num_sack_blocks = - min_t(unsigned int, eff_sacks, - (remaining - TCPOLEN_SACK_BASE_ALIGNED) / - TCPOLEN_SACK_PERBLOCK); - size += TCPOLEN_SACK_BASE_ALIGNED + - opts->num_sack_blocks * TCPOLEN_SACK_PERBLOCK; + if (likely(remaining >= TCPOLEN_SACK_BASE_ALIGNED + + TCPOLEN_SACK_PERBLOCK)) { + opts->num_sack_blocks = + min_t(unsigned int, eff_sacks, + (remaining - TCPOLEN_SACK_BASE_ALIGNED) / + TCPOLEN_SACK_PERBLOCK); + + size += TCPOLEN_SACK_BASE_ALIGNED + + opts->num_sack_blocks * TCPOLEN_SACK_PERBLOCK; + } else { + opts->num_sack_blocks = 0; + } + } else { + opts->num_sack_blocks = 0; + } + + if (tcp_ecn_mode_accecn(tp)) { + int ecn_opt = READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_ecn_option); + + if (ecn_opt && tp->saw_accecn_opt && !tcp_accecn_opt_fail_send(tp) && + (ecn_opt >= TCP_ACCECN_OPTION_FULL || tp->accecn_opt_demand || + tcp_accecn_option_beacon_check(sk))) { + opts->use_synack_ecn_bytes = 0; + size += tcp_options_fit_accecn(opts, tp->accecn_minlen, + MAX_TCP_OPTION_SPACE - size); + } + } + + if (unlikely(BPF_SOCK_OPS_TEST_FLAG(tp, + BPF_SOCK_OPS_WRITE_HDR_OPT_CB_FLAG))) { + unsigned int remaining = MAX_TCP_OPTION_SPACE - size; + + bpf_skops_hdr_opt_len(sk, skb, NULL, NULL, 0, opts, &remaining); + + size = MAX_TCP_OPTION_SPACE - remaining; } return size; @@ -773,15 +1218,15 @@ static unsigned int tcp_established_options(struct sock *sk, struct sk_buff *skb * needs to be reallocated in a driver. * The invariant being skb->truesize subtracted from sk->sk_wmem_alloc * - * Since transmit from skb destructor is forbidden, we use a tasklet + * Since transmit from skb destructor is forbidden, we use a BH work item * to process all sockets that eventually need to send more skbs. - * We use one tasklet per cpu, with its own queue of sockets. + * We use one work item per cpu, with its own queue of sockets. */ -struct tsq_tasklet { - struct tasklet_struct tasklet; +struct tsq_work { + struct work_struct work; struct list_head head; /* queue of tcp sockets */ }; -static DEFINE_PER_CPU(struct tsq_tasklet, tsq_tasklet); +static DEFINE_PER_CPU(struct tsq_work, tsq_work); static void tcp_tsq_write(struct sock *sk) { @@ -791,7 +1236,7 @@ static void tcp_tsq_write(struct sock *sk) struct tcp_sock *tp = tcp_sk(sk); if (tp->lost_out > tp->retrans_out && - tp->snd_cwnd > tcp_packets_in_flight(tp)) { + tcp_snd_cwnd(tp) > tcp_packets_in_flight(tp)) { tcp_mstamp_refresh(tp); tcp_xmit_retransmit_queue(sk); } @@ -811,14 +1256,14 @@ static void tcp_tsq_handler(struct sock *sk) bh_unlock_sock(sk); } /* - * One tasklet per cpu tries to send more skbs. - * We run in tasklet context but need to disable irqs when + * One work item per cpu tries to send more skbs. + * We run in BH context but need to disable irqs when * transferring tsq->head because tcp_wfree() might * interrupt us (non NAPI drivers) */ -static void tcp_tasklet_func(unsigned long data) +static void tcp_tsq_workfn(struct work_struct *work) { - struct tsq_tasklet *tsq = (struct tsq_tasklet *)data; + struct tsq_work *tsq = container_of(work, struct tsq_work, work); LIST_HEAD(list); unsigned long flags; struct list_head *q, *n; @@ -845,7 +1290,8 @@ static void tcp_tasklet_func(unsigned long data) #define TCP_DEFERRED_ALL (TCPF_TSQ_DEFERRED | \ TCPF_WRITE_TIMER_DEFERRED | \ TCPF_DELACK_TIMER_DEFERRED | \ - TCPF_MTU_REDUCED_DEFERRED) + TCPF_MTU_REDUCED_DEFERRED | \ + TCPF_ACK_DEFERRED) /** * tcp_release_cb - tcp release_sock() callback * @sk: socket @@ -855,30 +1301,20 @@ static void tcp_tasklet_func(unsigned long data) */ void tcp_release_cb(struct sock *sk) { - unsigned long flags, nflags; + unsigned long flags = smp_load_acquire(&sk->sk_tsq_flags); + unsigned long nflags; /* perform an atomic operation only if at least one flag is set */ do { - flags = sk->sk_tsq_flags; if (!(flags & TCP_DEFERRED_ALL)) return; nflags = flags & ~TCP_DEFERRED_ALL; - } while (cmpxchg(&sk->sk_tsq_flags, flags, nflags) != flags); + } while (!try_cmpxchg(&sk->sk_tsq_flags, &flags, nflags)); if (flags & TCPF_TSQ_DEFERRED) { tcp_tsq_write(sk); __sock_put(sk); } - /* Here begins the tricky part : - * We are called from release_sock() with : - * 1) BH disabled - * 2) sk_lock.slock spinlock held - * 3) socket owned by us (sk->sk_lock.owned == 1) - * - * But following code is meant to be called from BH handlers, - * so we should keep BH disabled, but early release socket ownership - */ - sock_release_ownership(sk); if (flags & TCPF_WRITE_TIMER_DEFERRED) { tcp_write_timer_handler(sk); @@ -892,20 +1328,20 @@ void tcp_release_cb(struct sock *sk) inet_csk(sk)->icsk_af_ops->mtu_reduced(sk); __sock_put(sk); } + if ((flags & TCPF_ACK_DEFERRED) && inet_csk_ack_scheduled(sk)) + tcp_send_ack(sk); } -EXPORT_SYMBOL(tcp_release_cb); +EXPORT_IPV6_MOD(tcp_release_cb); -void __init tcp_tasklet_init(void) +void __init tcp_tsq_work_init(void) { int i; for_each_possible_cpu(i) { - struct tsq_tasklet *tsq = &per_cpu(tsq_tasklet, i); + struct tsq_work *tsq = &per_cpu(tsq_work, i); INIT_LIST_HEAD(&tsq->head); - tasklet_init(&tsq->tasklet, - tcp_tasklet_func, - (unsigned long)tsq); + INIT_WORK(&tsq->work, tcp_tsq_workfn); } } @@ -919,9 +1355,11 @@ void tcp_wfree(struct sk_buff *skb) struct sock *sk = skb->sk; struct tcp_sock *tp = tcp_sk(sk); unsigned long flags, nval, oval; + struct tsq_work *tsq; + bool empty; /* Keep one reference on sk_wmem_alloc. - * Will be released by sk_free() from here or tcp_tasklet_func() + * Will be released by sk_free() from here or tcp_tsq_workfn() */ WARN_ON(refcount_sub_and_test(skb->truesize - 1, &sk->sk_wmem_alloc)); @@ -935,28 +1373,23 @@ void tcp_wfree(struct sk_buff *skb) if (refcount_read(&sk->sk_wmem_alloc) >= SKB_TRUESIZE(1) && this_cpu_ksoftirqd() == current) goto out; - for (oval = READ_ONCE(sk->sk_tsq_flags);; oval = nval) { - struct tsq_tasklet *tsq; - bool empty; - + oval = smp_load_acquire(&sk->sk_tsq_flags); + do { if (!(oval & TSQF_THROTTLED) || (oval & TSQF_QUEUED)) goto out; nval = (oval & ~TSQF_THROTTLED) | TSQF_QUEUED; - nval = cmpxchg(&sk->sk_tsq_flags, oval, nval); - if (nval != oval) - continue; + } while (!try_cmpxchg(&sk->sk_tsq_flags, &oval, nval)); - /* queue this socket to tasklet queue */ - local_irq_save(flags); - tsq = this_cpu_ptr(&tsq_tasklet); - empty = list_empty(&tsq->head); - list_add(&tp->tsq_node, &tsq->head); - if (empty) - tasklet_schedule(&tsq->tasklet); - local_irq_restore(flags); - return; - } + /* queue this socket to BH workqueue */ + local_irq_save(flags); + tsq = this_cpu_ptr(&tsq_work); + empty = list_empty(&tsq->head); + list_add(&tp->tsq_node, &tsq->head); + if (empty) + queue_work(system_bh_wq, &tsq->work); + local_irq_restore(flags); + return; out: sk_free(sk); } @@ -981,7 +1414,7 @@ static void tcp_update_skb_after_send(struct sock *sk, struct sk_buff *skb, struct tcp_sock *tp = tcp_sk(sk); if (sk->sk_pacing_status != SK_PACING_NONE) { - unsigned long rate = sk->sk_pacing_rate; + unsigned long rate = READ_ONCE(sk->sk_pacing_rate); /* Original sch_fq does not pace first 10 MSS * Note that tp->data_segs_out overflows after 2^32 packets, @@ -999,6 +1432,10 @@ static void tcp_update_skb_after_send(struct sock *sk, struct sk_buff *skb, list_move_tail(&skb->tcp_tsorted_anchor, &tp->tsorted_sent_queue); } +INDIRECT_CALLABLE_DECLARE(int ip_queue_xmit(struct sock *sk, struct sk_buff *skb, struct flowi *fl)); +INDIRECT_CALLABLE_DECLARE(int inet6_csk_xmit(struct sock *sk, struct sk_buff *skb, struct flowi *fl)); +INDIRECT_CALLABLE_DECLARE(void tcp_v4_send_check(struct sock *sk, struct sk_buff *skb)); + /* This routine actually transmits TCP packets queued in by * tcp_do_sendmsg(). This is used by both the initial * transmission and possible later retransmissions. @@ -1020,7 +1457,7 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, struct tcp_out_options opts; unsigned int tcp_options_size, tcp_header_size; struct sk_buff *oskb = NULL; - struct tcp_md5sig_key *md5; + struct tcp_key key; struct tcphdr *th; u64 prior_wstamp; int err; @@ -1029,10 +1466,8 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, tp = tcp_sk(sk); prior_wstamp = tp->tcp_wstamp_ns; tp->tcp_wstamp_ns = max(tp->tcp_wstamp_ns, tp->tcp_clock_cache); - skb->skb_mstamp_ns = tp->tcp_wstamp_ns; + skb_set_delivery_time(skb, tp->tcp_wstamp_ns, SKB_CLOCK_MONOTONIC); if (clone_it) { - TCP_SKB_CB(skb)->tx.in_flight = TCP_SKB_CB(skb)->end_seq - - tp->snd_una; oskb = skb; tcp_skb_tsorted_save(oskb) { @@ -1044,27 +1479,49 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, if (unlikely(!skb)) return -ENOBUFS; + /* retransmit skbs might have a non zero value in skb->dev + * because skb->dev is aliased with skb->rbnode.rb_left + */ + skb->dev = NULL; } inet = inet_sk(sk); tcb = TCP_SKB_CB(skb); memset(&opts, 0, sizeof(opts)); - if (unlikely(tcb->tcp_flags & TCPHDR_SYN)) - tcp_options_size = tcp_syn_options(sk, skb, &opts, &md5); - else - tcp_options_size = tcp_established_options(sk, skb, &opts, - &md5); + tcp_get_current_key(sk, &key); + if (unlikely(tcb->tcp_flags & TCPHDR_SYN)) { + tcp_options_size = tcp_syn_options(sk, skb, &opts, &key); + } else { + tcp_options_size = tcp_established_options(sk, skb, &opts, &key); + /* Force a PSH flag on all (GSO) packets to expedite GRO flush + * at receiver : This slightly improve GRO performance. + * Note that we do not force the PSH flag for non GSO packets, + * because they might be sent under high congestion events, + * and in this case it is better to delay the delivery of 1-MSS + * packets and thus the corresponding ACK packet that would + * release the following packet. + */ + if (tcp_skb_pcount(skb) > 1) + tcb->tcp_flags |= TCPHDR_PSH; + } tcp_header_size = tcp_options_size + sizeof(struct tcphdr); - /* if no packet is in qdisc/device queue, then allow XPS to select - * another queue. We can be called from tcp_tsq_handler() - * which holds one reference to sk. - * - * TODO: Ideally, in-flight pure ACK packets should not matter here. - * One way to get this would be to set skb->truesize = 2 on them. + /* We set skb->ooo_okay to one if this packet can select + * a different TX queue than prior packets of this flow, + * to avoid self inflicted reorders. + * The 'other' queue decision is based on current cpu number + * if XPS is enabled, or sk->sk_txhash otherwise. + * We can switch to another (and better) queue if: + * 1) No packet with payload is in qdisc/device queues. + * Delays in TX completion can defeat the test + * even if packets were already sent. + * 2) Or rtx queue is empty. + * This mitigates above case if ACK packets for + * all prior packets were already processed. */ - skb->ooo_okay = sk_wmem_alloc_get(sk) < SKB_TRUESIZE(1); + skb->ooo_okay = sk_wmem_alloc_get(sk) < SKB_TRUESIZE(1) || + tcp_rtx_queue_empty(sk); /* If we had to use memory reserve to allocate this skb, * this might cause drops if packet is looped back : @@ -1079,10 +1536,9 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, skb_orphan(skb); skb->sk = sk; skb->destructor = skb_is_tcp_pure_ack(skb) ? __sock_wfree : tcp_wfree; - skb_set_hash_from_sk(skb, sk); refcount_add(skb->truesize, &sk->sk_wmem_alloc); - skb_set_dst_pending_confirm(skb, sk->sk_dst_pending_confirm); + skb_set_dst_pending_confirm(skb, READ_ONCE(sk->sk_dst_pending_confirm)); /* Build TCP header and checksum it. */ th = (struct tcphdr *)skb->data; @@ -1091,7 +1547,7 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, th->seq = htonl(tcb->seq); th->ack_seq = htonl(rcv_nxt); *(((__be16 *)th) + 6) = htons(((tcp_header_size >> 2) << 12) | - tcb->tcp_flags); + (tcb->tcp_flags & TCPHDR_FLAGS_MASK)); th->check = 0; th->urg_ptr = 0; @@ -1107,7 +1563,6 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, } } - tcp_options_write((__be32 *)(th + 1), tp, &opts); skb_shinfo(skb)->gso_type = sk->sk_gso_type; if (likely(!(tcb->tcp_flags & TCPHDR_SYN))) { th->window = htons(tcp_select_window(sk)); @@ -1118,19 +1573,36 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, */ th->window = htons(min(tp->rcv_wnd, 65535U)); } + + tcp_options_write(th, tp, NULL, &opts, &key); + + if (tcp_key_is_md5(&key)) { #ifdef CONFIG_TCP_MD5SIG - /* Calculate the MD5 hash, as we have all we need now */ - if (md5) { - sk_nocaps_add(sk, NETIF_F_GSO_MASK); + /* Calculate the MD5 hash, as we have all we need now */ + sk_gso_disable(sk); tp->af_specific->calc_md5_hash(opts.hash_location, - md5, sk, skb); - } + key.md5_key, sk, skb); #endif + } else if (tcp_key_is_ao(&key)) { + int err; - icsk->icsk_af_ops->send_check(sk, skb); + err = tcp_ao_transmit_skb(sk, skb, key.ao_key, th, + opts.hash_location); + if (err) { + sk_skb_reason_drop(sk, skb, SKB_DROP_REASON_NOT_SPECIFIED); + return -ENOMEM; + } + } + + /* BPF prog is the last one writing header option */ + bpf_skops_write_hdr_opt(sk, skb, NULL, NULL, 0, &opts); + + INDIRECT_CALL_INET(icsk->icsk_af_ops->send_check, + tcp_v6_send_check, tcp_v4_send_check, + sk, skb); if (likely(tcb->tcp_flags & TCPHDR_ACK)) - tcp_event_ack_sent(sk, tcp_skb_pcount(skb), rcv_nxt); + tcp_event_ack_sent(sk, rcv_nxt); if (skb->len != tcp_header_size) { tcp_event_data_sent(tp, sk); @@ -1143,6 +1615,7 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, tcp_skb_pcount(skb)); tp->segs_out += tcp_skb_pcount(skb); + skb_set_hash_from_sk(skb, sk); /* OK, its time to fill skb_shinfo(skb)->gso_{segs|size} */ skb_shinfo(skb)->gso_segs = tcp_skb_pcount(skb); skb_shinfo(skb)->gso_size = tcp_skb_mss(skb); @@ -1153,7 +1626,11 @@ static int __tcp_transmit_skb(struct sock *sk, struct sk_buff *skb, memset(skb->cb, 0, max(sizeof(struct inet_skb_parm), sizeof(struct inet6_skb_parm))); - err = icsk->icsk_af_ops->queue_xmit(sk, skb, &inet->cork.fl); + tcp_add_tx_delay(skb, tp); + + err = INDIRECT_CALL_INET(icsk->icsk_af_ops->queue_xmit, + inet6_csk_xmit, ip_queue_xmit, + sk, skb, &inet->cork.fl); if (unlikely(err > 0)) { tcp_enter_cwr(sk); @@ -1183,26 +1660,31 @@ static void tcp_queue_skb(struct sock *sk, struct sk_buff *skb) struct tcp_sock *tp = tcp_sk(sk); /* Advance write_seq and place onto the write_queue. */ - tp->write_seq = TCP_SKB_CB(skb)->end_seq; + WRITE_ONCE(tp->write_seq, TCP_SKB_CB(skb)->end_seq); __skb_header_release(skb); + psp_enqueue_set_decrypted(sk, skb); tcp_add_write_queue_tail(sk, skb); - sk->sk_wmem_queued += skb->truesize; + sk_wmem_queued_add(sk, skb->truesize); sk_mem_charge(sk, skb->truesize); } /* Initialize TSO segments for a packet. */ -static void tcp_set_skb_tso_segs(struct sk_buff *skb, unsigned int mss_now) +static int tcp_set_skb_tso_segs(struct sk_buff *skb, unsigned int mss_now) { + int tso_segs; + if (skb->len <= mss_now) { /* Avoid the costly divide in the normal * non-TSO case. */ - tcp_skb_pcount_set(skb, 1); TCP_SKB_CB(skb)->tcp_gso_size = 0; - } else { - tcp_skb_pcount_set(skb, DIV_ROUND_UP(skb->len, mss_now)); - TCP_SKB_CB(skb)->tcp_gso_size = mss_now; + tcp_skb_pcount_set(skb, 1); + return 1; } + TCP_SKB_CB(skb)->tcp_gso_size = mss_now; + tso_segs = DIV_ROUND_UP(skb->len, mss_now); + tcp_skb_pcount_set(skb, tso_segs); + return tso_segs; } /* Pcount in the middle of the write queue got changed, we need to do various @@ -1225,11 +1707,6 @@ static void tcp_adjust_pcount(struct sock *sk, const struct sk_buff *skb, int de if (tcp_is_reno(tp) && decr > 0) tp->sacked_out -= min_t(u32, tp->sacked_out, decr); - if (tp->lost_skb_hint && - before(TCP_SKB_CB(skb)->seq, TCP_SKB_CB(tp->lost_skb_hint)->seq) && - (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED)) - tp->lost_cnt_hint -= decr; - tcp_verify_left_out(tp); } @@ -1285,28 +1762,43 @@ int tcp_fragment(struct sock *sk, enum tcp_queue tcp_queue, { struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *buff; - int nsize, old_factor; + int old_factor; + long limit; + u16 flags; int nlen; - u8 flags; if (WARN_ON(len > skb->len)) return -EINVAL; - nsize = skb_headlen(skb) - len; - if (nsize < 0) - nsize = 0; + DEBUG_NET_WARN_ON_ONCE(skb_headlen(skb)); + + /* tcp_sendmsg() can overshoot sk_wmem_queued by one full size skb. + * We need some allowance to not penalize applications setting small + * SO_SNDBUF values. + * Also allow first and last skb in retransmit queue to be split. + */ + limit = sk->sk_sndbuf + 2 * SKB_TRUESIZE(GSO_LEGACY_MAX_SIZE); + if (unlikely((sk->sk_wmem_queued >> 1) > limit && + tcp_queue != TCP_FRAG_IN_WRITE_QUEUE && + skb != tcp_rtx_queue_head(sk) && + skb != tcp_rtx_queue_tail(sk))) { + NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPWQUEUETOOBIG); + return -ENOMEM; + } - if (skb_unclone(skb, gfp)) + if (skb_unclone_keeptruesize(skb, gfp)) return -ENOMEM; /* Get a new skb... force flag on. */ - buff = sk_stream_alloc_skb(sk, nsize, gfp, true); + buff = tcp_stream_alloc_skb(sk, gfp, true); if (!buff) return -ENOMEM; /* We'll just try again later. */ + skb_copy_decrypted(buff, skb); + mptcp_skb_ext_copy(buff, skb); - sk->sk_wmem_queued += buff->truesize; + sk_wmem_queued_add(sk, buff->truesize); sk_mem_charge(sk, buff->truesize); - nlen = skb->len - len - nsize; + nlen = skb->len - len; buff->truesize += nlen; skb->truesize -= nlen; @@ -1324,9 +1816,7 @@ int tcp_fragment(struct sock *sk, enum tcp_queue tcp_queue, skb_split(skb, buff, len); - buff->ip_summed = CHECKSUM_PARTIAL; - - buff->tstamp = skb->tstamp; + skb_set_delivery_time(buff, skb->tstamp, SKB_CLOCK_MONOTONIC); tcp_fragment_tstamp(skb, buff); old_factor = tcp_skb_pcount(skb); @@ -1366,13 +1856,7 @@ static int __pskb_trim_head(struct sk_buff *skb, int len) struct skb_shared_info *shinfo; int i, k, eat; - eat = min_t(int, len, skb_headlen(skb)); - if (eat) { - __skb_pull(skb, eat); - len -= eat; - if (!len) - return 0; - } + DEBUG_NET_WARN_ON_ONCE(skb_headlen(skb)); eat = len; k = 0; shinfo = skb_shinfo(skb); @@ -1385,7 +1869,7 @@ static int __pskb_trim_head(struct sk_buff *skb, int len) } else { shinfo->frags[k] = shinfo->frags[i]; if (eat) { - shinfo->frags[k].page_offset += eat; + skb_frag_off_add(&shinfo->frags[k], eat); skb_frag_size_sub(&shinfo->frags[k], eat); eat = 0; } @@ -1404,20 +1888,17 @@ int tcp_trim_head(struct sock *sk, struct sk_buff *skb, u32 len) { u32 delta_truesize; - if (skb_unclone(skb, GFP_ATOMIC)) + if (skb_unclone_keeptruesize(skb, GFP_ATOMIC)) return -ENOMEM; delta_truesize = __pskb_trim_head(skb, len); TCP_SKB_CB(skb)->seq += len; - skb->ip_summed = CHECKSUM_PARTIAL; - if (delta_truesize) { - skb->truesize -= delta_truesize; - sk->sk_wmem_queued -= delta_truesize; + skb->truesize -= delta_truesize; + sk_wmem_queued_add(sk, -delta_truesize); + if (!skb_zcopy_pure(skb)) sk_mem_uncharge(sk, delta_truesize); - sock_set_flag(sk, SOCK_QUEUE_SHRUNK); - } /* Any change of skb->len requires recalculation of tso factor. */ if (tcp_skb_pcount(skb) > 1) @@ -1438,14 +1919,6 @@ static inline int __tcp_mtu_to_mss(struct sock *sk, int pmtu) */ mss_now = pmtu - icsk->icsk_af_ops->net_header_len - sizeof(struct tcphdr); - /* IPv6 adds a frag_hdr in case RTAX_FEATURE_ALLFRAG is set */ - if (icsk->icsk_af_ops->net_frag_header_len) { - const struct dst_entry *dst = __sk_dst_get(sk); - - if (dst && dst_allfrag(dst)) - mss_now -= icsk->icsk_af_ops->net_frag_header_len; - } - /* Clamp it (mss_clamp does not include tcp options) */ if (mss_now > tp->rx_opt.mss_clamp) mss_now = tp->rx_opt.mss_clamp; @@ -1454,8 +1927,8 @@ static inline int __tcp_mtu_to_mss(struct sock *sk, int pmtu) mss_now -= icsk->icsk_ext_hdr_len; /* Then reserve room for full set of TCP options and 8 bytes of data */ - if (mss_now < 48) - mss_now = 48; + mss_now = max(mss_now, + READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_min_snd_mss)); return mss_now; } @@ -1466,27 +1939,18 @@ int tcp_mtu_to_mss(struct sock *sk, int pmtu) return __tcp_mtu_to_mss(sk, pmtu) - (tcp_sk(sk)->tcp_header_len - sizeof(struct tcphdr)); } +EXPORT_IPV6_MOD(tcp_mtu_to_mss); /* Inverse of above */ int tcp_mss_to_mtu(struct sock *sk, int mss) { const struct tcp_sock *tp = tcp_sk(sk); const struct inet_connection_sock *icsk = inet_csk(sk); - int mtu; - mtu = mss + + return mss + tp->tcp_header_len + icsk->icsk_ext_hdr_len + icsk->icsk_af_ops->net_header_len; - - /* IPv6 adds a frag_hdr in case RTAX_FEATURE_ALLFRAG is set */ - if (icsk->icsk_af_ops->net_frag_header_len) { - const struct dst_entry *dst = __sk_dst_get(sk); - - if (dst && dst_allfrag(dst)) - mtu += icsk->icsk_af_ops->net_frag_header_len; - } - return mtu; } EXPORT_SYMBOL(tcp_mss_to_mtu); @@ -1497,15 +1961,14 @@ void tcp_mtup_init(struct sock *sk) struct inet_connection_sock *icsk = inet_csk(sk); struct net *net = sock_net(sk); - icsk->icsk_mtup.enabled = net->ipv4.sysctl_tcp_mtu_probing > 1; + icsk->icsk_mtup.enabled = READ_ONCE(net->ipv4.sysctl_tcp_mtu_probing) > 1; icsk->icsk_mtup.search_high = tp->rx_opt.mss_clamp + sizeof(struct tcphdr) + icsk->icsk_af_ops->net_header_len; - icsk->icsk_mtup.search_low = tcp_mss_to_mtu(sk, net->ipv4.sysctl_tcp_base_mss); + icsk->icsk_mtup.search_low = tcp_mss_to_mtu(sk, READ_ONCE(net->ipv4.sysctl_tcp_base_mss)); icsk->icsk_mtup.probe_size = 0; if (icsk->icsk_mtup.enabled) icsk->icsk_mtup.probe_timestamp = tcp_jiffies32; } -EXPORT_SYMBOL(tcp_mtup_init); /* This function synchronize snd mss to current pmtu/exthdr set. @@ -1549,7 +2012,7 @@ unsigned int tcp_sync_mss(struct sock *sk, u32 pmtu) return mss_now; } -EXPORT_SYMBOL(tcp_sync_mss); +EXPORT_IPV6_MOD(tcp_sync_mss); /* Compute the current effective MSS, taking SACKs and IP options, * and even PMTU discovery events into account. @@ -1561,7 +2024,7 @@ unsigned int tcp_current_mss(struct sock *sk) u32 mss_now; unsigned int header_len; struct tcp_out_options opts; - struct tcp_md5sig_key *md5; + struct tcp_key key; mss_now = tp->mss_cache; @@ -1570,8 +2033,8 @@ unsigned int tcp_current_mss(struct sock *sk) if (mtu != inet_csk(sk)->icsk_pmtu_cookie) mss_now = tcp_sync_mss(sk, mtu); } - - header_len = tcp_established_options(sk, NULL, &opts, &md5) + + tcp_get_current_key(sk, &key); + header_len = tcp_established_options(sk, NULL, &opts, &key) + sizeof(struct tcphdr); /* The mss_cache is sized based on tp->tcp_header_len, which assumes * some common options. If this is an odd packet (because we have SACK @@ -1598,9 +2061,9 @@ static void tcp_cwnd_application_limited(struct sock *sk) /* Limited by application or receiver window. */ u32 init_win = tcp_init_cwnd(tp, __sk_dst_get(sk)); u32 win_used = max(tp->snd_cwnd_used, init_win); - if (win_used < tp->snd_cwnd) { + if (win_used < tcp_snd_cwnd(tp)) { tp->snd_ssthresh = tcp_current_ssthresh(sk); - tp->snd_cwnd = (tp->snd_cwnd + win_used) >> 1; + tcp_snd_cwnd_set(tp, (tcp_snd_cwnd(tp) + win_used) >> 1); } tp->snd_cwnd_used = 0; } @@ -1612,14 +2075,20 @@ static void tcp_cwnd_validate(struct sock *sk, bool is_cwnd_limited) const struct tcp_congestion_ops *ca_ops = inet_csk(sk)->icsk_ca_ops; struct tcp_sock *tp = tcp_sk(sk); - /* Track the maximum number of outstanding packets in each - * window, and remember whether we were cwnd-limited then. + /* Track the strongest available signal of the degree to which the cwnd + * is fully utilized. If cwnd-limited then remember that fact for the + * current window. If not cwnd-limited then track the maximum number of + * outstanding packets in the current window. (If cwnd-limited then we + * chose to not update tp->max_packets_out to avoid an extra else + * clause with no functional impact.) */ - if (!before(tp->snd_una, tp->max_packets_seq) || - tp->packets_out > tp->max_packets_out) { - tp->max_packets_out = tp->packets_out; - tp->max_packets_seq = tp->snd_nxt; + if (!before(tp->snd_una, tp->cwnd_usage_seq) || + is_cwnd_limited || + (!tp->is_cwnd_limited && + tp->packets_out > tp->max_packets_out)) { tp->is_cwnd_limited = is_cwnd_limited; + tp->max_packets_out = tp->packets_out; + tp->cwnd_usage_seq = tp->snd_nxt; } if (tcp_is_cwnd_limited(sk)) { @@ -1631,7 +2100,7 @@ static void tcp_cwnd_validate(struct sock *sk, bool is_cwnd_limited) if (tp->packets_out > tp->snd_cwnd_used) tp->snd_cwnd_used = tp->packets_out; - if (sock_net(sk)->ipv4.sysctl_tcp_slow_start_after_idle && + if (READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_slow_start_after_idle) && (s32)(tcp_jiffies32 - tp->snd_cwnd_stamp) >= inet_csk(sk)->icsk_rto && !ca_ops->cong_control) tcp_cwnd_application_limited(sk); @@ -1688,25 +2157,34 @@ static bool tcp_nagle_check(bool partial, const struct tcp_sock *tp, } /* Return how many segs we'd like on a TSO packet, - * to send one TSO packet per ms + * depending on current pacing rate, and how close the peer is. + * + * Rationale is: + * - For close peers, we rather send bigger packets to reduce + * cpu costs, because occasional losses will be repaired fast. + * - For long distance/rtt flows, we would like to get ACK clocking + * with 1 ACK per ms. + * + * Use min_rtt to help adapt TSO burst size, with smaller min_rtt resulting + * in bigger TSO bursts. We we cut the RTT-based allowance in half + * for every 2^9 usec (aka 512 us) of RTT, so that the RTT-based allowance + * is below 1500 bytes after 6 * ~500 usec = 3ms. */ static u32 tcp_tso_autosize(const struct sock *sk, unsigned int mss_now, int min_tso_segs) { - u32 bytes, segs; + unsigned long bytes; + u32 r; - bytes = min_t(unsigned long, - sk->sk_pacing_rate >> sk->sk_pacing_shift, - sk->sk_gso_max_size - 1 - MAX_TCP_HEADER); + bytes = READ_ONCE(sk->sk_pacing_rate) >> READ_ONCE(sk->sk_pacing_shift); - /* Goal is to send at least one packet per ms, - * not one big TSO packet every 100 ms. - * This preserves ACK clocking and is consistent - * with tcp_tso_should_defer() heuristic. - */ - segs = max_t(u32, bytes / mss_now, min_tso_segs); + r = tcp_min_rtt(tcp_sk(sk)) >> READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_tso_rtt_log); + if (r < BITS_PER_TYPE(sk->sk_gso_max_size)) + bytes += sk->sk_gso_max_size >> r; + + bytes = min_t(unsigned long, bytes, sk->sk_gso_max_size); - return segs; + return max_t(u32, bytes / mss_now, min_tso_segs); } /* Return the number of segments we want in the skb we are transmitting. @@ -1719,7 +2197,7 @@ static u32 tcp_tso_segs(struct sock *sk, unsigned int mss_now) min_tso = ca_ops->min_tso_segs ? ca_ops->min_tso_segs(sk) : - sock_net(sk)->ipv4.sysctl_tcp_min_tso_segs; + READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_min_tso_segs); tso_segs = tcp_tso_autosize(sk, mss_now, min_tso); return min_t(u32, tso_segs, sk->sk_gso_max_segs); @@ -1760,18 +2238,12 @@ static unsigned int tcp_mss_split_point(const struct sock *sk, /* Can at least one segment of SKB be sent right now, according to the * congestion window rules? If so, return how many segments are allowed. */ -static inline unsigned int tcp_cwnd_test(const struct tcp_sock *tp, - const struct sk_buff *skb) +static u32 tcp_cwnd_test(const struct tcp_sock *tp) { u32 in_flight, cwnd, halfcwnd; - /* Don't be strict about the congestion window for the final FIN. */ - if ((TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN) && - tcp_skb_pcount(skb) == 1) - return 1; - in_flight = tcp_packets_in_flight(tp); - cwnd = tp->snd_cwnd; + cwnd = tcp_snd_cwnd(tp); if (in_flight >= cwnd) return 0; @@ -1790,10 +2262,9 @@ static int tcp_init_tso_segs(struct sk_buff *skb, unsigned int mss_now) { int tso_segs = tcp_skb_pcount(skb); - if (!tso_segs || (tso_segs > 1 && tcp_skb_mss(skb) != mss_now)) { - tcp_set_skb_tso_segs(skb, mss_now); - tso_segs = tcp_skb_pcount(skb); - } + if (!tso_segs || (tso_segs > 1 && tcp_skb_mss(skb) != mss_now)) + return tcp_set_skb_tso_segs(skb, mss_now); + return tso_segs; } @@ -1843,23 +2314,23 @@ static bool tcp_snd_wnd_test(const struct tcp_sock *tp, * know that all the data is in scatter-gather pages, and that the * packet has never been sent out before (and thus is not cloned). */ -static int tso_fragment(struct sock *sk, enum tcp_queue tcp_queue, - struct sk_buff *skb, unsigned int len, +static int tso_fragment(struct sock *sk, struct sk_buff *skb, unsigned int len, unsigned int mss_now, gfp_t gfp) { - struct sk_buff *buff; int nlen = skb->len - len; - u8 flags; + struct sk_buff *buff; + u16 flags; /* All of a TSO frame must be composed of paged data. */ - if (skb->len != skb->data_len) - return tcp_fragment(sk, tcp_queue, skb, len, mss_now, gfp); + DEBUG_NET_WARN_ON_ONCE(skb->len != skb->data_len); - buff = sk_stream_alloc_skb(sk, 0, gfp, true); + buff = tcp_stream_alloc_skb(sk, gfp, true); if (unlikely(!buff)) return -ENOMEM; + skb_copy_decrypted(buff, skb); + mptcp_skb_ext_copy(buff, skb); - sk->sk_wmem_queued += buff->truesize; + sk_wmem_queued_add(sk, buff->truesize); sk_mem_charge(sk, buff->truesize); buff->truesize += nlen; skb->truesize -= nlen; @@ -1874,12 +2345,8 @@ static int tso_fragment(struct sock *sk, enum tcp_queue tcp_queue, TCP_SKB_CB(skb)->tcp_flags = flags & ~(TCPHDR_FIN | TCPHDR_PSH); TCP_SKB_CB(buff)->tcp_flags = flags; - /* This packet was never sent out yet, so no SACK bits. */ - TCP_SKB_CB(buff)->sacked = 0; - tcp_skb_fragment_eor(skb, buff); - buff->ip_summed = CHECKSUM_PARTIAL; skb_split(skb, buff, len); tcp_fragment_tstamp(skb, buff); @@ -1889,7 +2356,7 @@ static int tso_fragment(struct sock *sk, enum tcp_queue tcp_queue, /* Link BUFF into the send queue. */ __skb_header_release(buff); - tcp_insert_write_queue_after(skb, buff, sk, tcp_queue); + tcp_insert_write_queue_after(skb, buff, sk, TCP_FRAG_IN_WRITE_QUEUE); return 0; } @@ -1905,7 +2372,8 @@ static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb, u32 max_segs) { const struct inet_connection_sock *icsk = inet_csk(sk); - u32 send_win, cong_win, limit, in_flight; + u32 send_win, cong_win, limit, in_flight, threshold; + u64 srtt_in_ns, expected_ack, how_far_is_the_ack; struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *head; int win_divisor; @@ -1926,12 +2394,12 @@ static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb, in_flight = tcp_packets_in_flight(tp); BUG_ON(tcp_skb_pcount(skb) <= 1); - BUG_ON(tp->snd_cwnd <= in_flight); + BUG_ON(tcp_snd_cwnd(tp) <= in_flight); send_win = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq; /* From in_flight test above, we know that cwnd > in_flight. */ - cong_win = (tp->snd_cwnd - in_flight) * tp->mss_cache; + cong_win = (tcp_snd_cwnd(tp) - in_flight) * tp->mss_cache; limit = min(send_win, cong_win); @@ -1945,7 +2413,7 @@ static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb, win_divisor = READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_tso_win_divisor); if (win_divisor) { - u32 chunk = min(tp->snd_wnd, tp->snd_cwnd * tp->mss_cache); + u32 chunk = min(tp->snd_wnd, tcp_snd_cwnd(tp) * tp->mss_cache); /* If at least some fraction of a window is available, * just use it. @@ -1967,9 +2435,19 @@ static bool tcp_tso_should_defer(struct sock *sk, struct sk_buff *skb, head = tcp_rtx_queue_head(sk); if (!head) goto send_now; - delta = tp->tcp_clock_cache - head->tstamp; - /* If next ACK is likely to come too late (half srtt), do not defer */ - if ((s64)(delta - (u64)NSEC_PER_USEC * (tp->srtt_us >> 4)) < 0) + + srtt_in_ns = (u64)(NSEC_PER_USEC >> 3) * tp->srtt_us; + /* When is the ACK expected ? */ + expected_ack = head->tstamp + srtt_in_ns; + /* How far from now is the ACK expected ? */ + how_far_is_the_ack = expected_ack - tp->tcp_clock_cache; + + /* If next ACK is likely to come too late, + * ie in more than min(1ms, half srtt), do not defer. + */ + threshold = min(srtt_in_ns >> 1, NSEC_PER_MSEC); + + if ((s64)(how_far_is_the_ack - threshold) > 0) goto send_now; /* Ok, it looks like it is advisable to defer. @@ -2009,7 +2487,7 @@ static inline void tcp_mtu_check_reprobe(struct sock *sk) u32 interval; s32 delta; - interval = net->ipv4.sysctl_tcp_probe_interval; + interval = READ_ONCE(net->ipv4.sysctl_tcp_probe_interval); delta = tcp_jiffies32 - icsk->icsk_mtup.probe_timestamp; if (unlikely(delta >= interval * HZ)) { int mss = tcp_current_mss(sk); @@ -2035,7 +2513,7 @@ static bool tcp_can_coalesce_send_queue_head(struct sock *sk, int len) if (len <= skb->len) break; - if (unlikely(TCP_SKB_CB(skb)->eor)) + if (tcp_has_tx_tstamp(skb) || !tcp_skb_can_collapse(skb, next)) return false; len -= skb->len; @@ -2044,6 +2522,72 @@ static bool tcp_can_coalesce_send_queue_head(struct sock *sk, int len) return true; } +static int tcp_clone_payload(struct sock *sk, struct sk_buff *to, + int probe_size) +{ + skb_frag_t *lastfrag = NULL, *fragto = skb_shinfo(to)->frags; + int i, todo, len = 0, nr_frags = 0; + const struct sk_buff *skb; + + if (!sk_wmem_schedule(sk, to->truesize + probe_size)) + return -ENOMEM; + + skb_queue_walk(&sk->sk_write_queue, skb) { + const skb_frag_t *fragfrom = skb_shinfo(skb)->frags; + + if (skb_headlen(skb)) + return -EINVAL; + + for (i = 0; i < skb_shinfo(skb)->nr_frags; i++, fragfrom++) { + if (len >= probe_size) + goto commit; + todo = min_t(int, skb_frag_size(fragfrom), + probe_size - len); + len += todo; + if (lastfrag && + skb_frag_page(fragfrom) == skb_frag_page(lastfrag) && + skb_frag_off(fragfrom) == skb_frag_off(lastfrag) + + skb_frag_size(lastfrag)) { + skb_frag_size_add(lastfrag, todo); + continue; + } + if (unlikely(nr_frags == MAX_SKB_FRAGS)) + return -E2BIG; + skb_frag_page_copy(fragto, fragfrom); + skb_frag_off_copy(fragto, fragfrom); + skb_frag_size_set(fragto, todo); + nr_frags++; + lastfrag = fragto++; + } + } +commit: + WARN_ON_ONCE(len != probe_size); + for (i = 0; i < nr_frags; i++) + skb_frag_ref(to, i); + + skb_shinfo(to)->nr_frags = nr_frags; + to->truesize += probe_size; + to->len += probe_size; + to->data_len += probe_size; + __skb_header_release(to); + return 0; +} + +/* tcp_mtu_probe() and tcp_grow_skb() can both eat an skb (src) if + * all its payload was moved to another one (dst). + * Make sure to transfer tcp_flags, eor, and tstamp. + */ +static void tcp_eat_one_skb(struct sock *sk, + struct sk_buff *dst, + struct sk_buff *src) +{ + TCP_SKB_CB(dst)->tcp_flags |= TCP_SKB_CB(src)->tcp_flags; + TCP_SKB_CB(dst)->eor = TCP_SKB_CB(src)->eor; + tcp_skb_collapse_tstamp(dst, src); + tcp_unlink_write_queue(src, sk); + tcp_wmem_free_skb(sk, src); +} + /* Create a new MTU probe if we are ready. * MTU probe is regularly attempting to increase the path MTU by * deliberately sending larger packets. This discovers routing @@ -2073,7 +2617,7 @@ static int tcp_mtu_probe(struct sock *sk) if (likely(!icsk->icsk_mtup.enabled || icsk->icsk_mtup.probe_size || inet_csk(sk)->icsk_ca_state != TCP_CA_Open || - tp->snd_cwnd < 11 || + tcp_snd_cwnd(tp) < 11 || tp->rx_opt.num_sacks || tp->rx_opt.dsack)) return -1; @@ -2091,7 +2635,7 @@ static int tcp_mtu_probe(struct sock *sk) * probing process by not resetting search range to its orignal. */ if (probe_size > tcp_mtu_to_mss(sk, icsk->icsk_mtup.search_high) || - interval < net->ipv4.sysctl_tcp_probe_threshold) { + interval < READ_ONCE(net->ipv4.sysctl_tcp_probe_threshold)) { /* Check whether enough time has elaplased for * another round of probing. */ @@ -2109,7 +2653,7 @@ static int tcp_mtu_probe(struct sock *sk) return 0; /* Do we need to wait to drain cwnd? With none in flight, don't stall */ - if (tcp_packets_in_flight(tp) + 2 > tp->snd_cwnd) { + if (tcp_packets_in_flight(tp) + 2 > tcp_snd_cwnd(tp)) { if (!tcp_packets_in_flight(tp)) return -1; else @@ -2120,20 +2664,26 @@ static int tcp_mtu_probe(struct sock *sk) return -1; /* We're allowed to probe. Build it now. */ - nskb = sk_stream_alloc_skb(sk, probe_size, GFP_ATOMIC, false); + nskb = tcp_stream_alloc_skb(sk, GFP_ATOMIC, false); if (!nskb) return -1; - sk->sk_wmem_queued += nskb->truesize; + + /* build the payload, and be prepared to abort if this fails. */ + if (tcp_clone_payload(sk, nskb, probe_size)) { + tcp_skb_tsorted_anchor_cleanup(nskb); + consume_skb(nskb); + return -1; + } + sk_wmem_queued_add(sk, nskb->truesize); sk_mem_charge(sk, nskb->truesize); skb = tcp_send_head(sk); + skb_copy_decrypted(nskb, skb); + mptcp_skb_ext_copy(nskb, skb); TCP_SKB_CB(nskb)->seq = TCP_SKB_CB(skb)->seq; TCP_SKB_CB(nskb)->end_seq = TCP_SKB_CB(skb)->seq + probe_size; TCP_SKB_CB(nskb)->tcp_flags = TCPHDR_ACK; - TCP_SKB_CB(nskb)->sacked = 0; - nskb->csum = 0; - nskb->ip_summed = CHECKSUM_PARTIAL; tcp_insert_write_queue_before(nskb, skb, sk); tcp_highest_sack_replace(sk, skb, nskb); @@ -2141,27 +2691,14 @@ static int tcp_mtu_probe(struct sock *sk) len = 0; tcp_for_write_queue_from_safe(skb, next, sk) { copy = min_t(int, skb->len, probe_size - len); - skb_copy_bits(skb, 0, skb_put(nskb, copy), copy); if (skb->len <= copy) { - /* We've eaten all the data from this skb. - * Throw it away. */ - TCP_SKB_CB(nskb)->tcp_flags |= TCP_SKB_CB(skb)->tcp_flags; - /* If this is the last SKB we copy and eor is set - * we need to propagate it to the new skb. - */ - TCP_SKB_CB(nskb)->eor = TCP_SKB_CB(skb)->eor; - tcp_unlink_write_queue(skb, sk); - sk_wmem_free_skb(sk, skb); + tcp_eat_one_skb(sk, nskb, skb); } else { TCP_SKB_CB(nskb)->tcp_flags |= TCP_SKB_CB(skb)->tcp_flags & ~(TCPHDR_FIN|TCPHDR_PSH); - if (!skb_shinfo(skb)->nr_frags) { - skb_pull(skb, copy); - } else { - __pskb_trim_head(skb, copy); - tcp_set_skb_tso_segs(skb, mss_now); - } + __pskb_trim_head(skb, copy); + tcp_set_skb_tso_segs(skb, mss_now); TCP_SKB_CB(skb)->seq += copy; } @@ -2178,7 +2715,7 @@ static int tcp_mtu_probe(struct sock *sk) if (!tcp_transmit_skb(sk, nskb, 1, GFP_ATOMIC)) { /* Decrement cwnd here because we are sending * effectively two packets. */ - tp->snd_cwnd--; + tcp_snd_cwnd_set(tp, tcp_snd_cwnd(tp) - 1); tcp_event_new_data_sent(sk, nskb); icsk->icsk_mtup.probe_size = tcp_mss_to_mtu(sk, nskb->len); @@ -2210,6 +2747,18 @@ static bool tcp_pacing_check(struct sock *sk) return true; } +static bool tcp_rtx_queue_empty_or_single_skb(const struct sock *sk) +{ + const struct rb_node *node = sk->tcp_rtx_queue.rb_node; + + /* No skb in the rtx queue. */ + if (!node) + return true; + + /* Only one skb in rtx queue. */ + return !node->rb_left && !node->rb_right; +} + /* TCP Small Queues : * Control number of packets in qdisc/devices to two packets / or ~1 ms. * (These limits are doubled for retransmits) @@ -2228,19 +2777,31 @@ static bool tcp_small_queue_check(struct sock *sk, const struct sk_buff *skb, limit = max_t(unsigned long, 2 * skb->truesize, - sk->sk_pacing_rate >> sk->sk_pacing_shift); - if (sk->sk_pacing_status == SK_PACING_NONE) - limit = min_t(unsigned long, limit, - sock_net(sk)->ipv4.sysctl_tcp_limit_output_bytes); + READ_ONCE(sk->sk_pacing_rate) >> READ_ONCE(sk->sk_pacing_shift)); + limit = min_t(unsigned long, limit, + READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_limit_output_bytes)); limit <<= factor; + if (static_branch_unlikely(&tcp_tx_delay_enabled) && + tcp_sk(sk)->tcp_tx_delay) { + u64 extra_bytes = (u64)READ_ONCE(sk->sk_pacing_rate) * + tcp_sk(sk)->tcp_tx_delay; + + /* TSQ is based on skb truesize sum (sk_wmem_alloc), so we + * approximate our needs assuming an ~100% skb->truesize overhead. + * USEC_PER_SEC is approximated by 2^20. + * do_div(extra_bytes, USEC_PER_SEC/2) is replaced by a right shift. + */ + extra_bytes >>= (20 - 1); + limit += extra_bytes; + } if (refcount_read(&sk->sk_wmem_alloc) > limit) { - /* Always send skb if rtx queue is empty. + /* Always send skb if rtx queue is empty or has one skb. * No need to wait for TX completion to call us back, - * after softirq/tasklet schedule. + * after softirq schedule. * This helps when TX completions are delayed too much. */ - if (tcp_rtx_queue_empty(sk)) + if (tcp_rtx_queue_empty_or_single_skb(sk)) return false; set_bit(TSQ_THROTTLED, &sk->sk_tsq_flags); @@ -2297,6 +2858,35 @@ void tcp_chrono_stop(struct sock *sk, const enum tcp_chrono type) tcp_chrono_set(tp, TCP_CHRONO_BUSY); } +/* First skb in the write queue is smaller than ideal packet size. + * Check if we can move payload from the second skb in the queue. + */ +static void tcp_grow_skb(struct sock *sk, struct sk_buff *skb, int amount) +{ + struct sk_buff *next_skb = skb->next; + unsigned int nlen; + + if (tcp_skb_is_last(sk, skb)) + return; + + if (!tcp_skb_can_collapse(skb, next_skb)) + return; + + nlen = min_t(u32, amount, next_skb->len); + if (!nlen || !skb_shift(skb, next_skb, nlen)) + return; + + TCP_SKB_CB(skb)->end_seq += nlen; + TCP_SKB_CB(next_skb)->seq += nlen; + + if (!next_skb->len) { + /* In case FIN is set, we need to update end_seq */ + TCP_SKB_CB(skb)->end_seq = TCP_SKB_CB(next_skb)->end_seq; + + tcp_eat_one_skb(sk, skb, next_skb); + } +} + /* This routine writes packets to the network. It advances the * send_head. This happens as incoming acks open up the remote * window for us. @@ -2317,14 +2907,18 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle, struct tcp_sock *tp = tcp_sk(sk); struct sk_buff *skb; unsigned int tso_segs, sent_pkts; - int cwnd_quota; + u32 cwnd_quota, max_segs; int result; bool is_cwnd_limited = false, is_rwnd_limited = false; - u32 max_segs; sent_pkts = 0; tcp_mstamp_refresh(tp); + + /* AccECN option beacon depends on mstamp, it may change mss */ + if (tcp_ecn_mode_accecn(tp) && tcp_accecn_option_beacon_check(sk)) + mss_now = tcp_current_mss(sk); + if (!push_one) { /* Do MTU probing. */ result = tcp_mtu_probe(sk); @@ -2338,21 +2932,21 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle, max_segs = tcp_tso_segs(sk, mss_now); while ((skb = tcp_send_head(sk))) { unsigned int limit; + int missing_bytes; if (unlikely(tp->repair) && tp->repair_queue == TCP_SEND_QUEUE) { /* "skb_mstamp_ns" is used as a start point for the retransmit timer */ - skb->skb_mstamp_ns = tp->tcp_wstamp_ns = tp->tcp_clock_cache; + tp->tcp_wstamp_ns = tp->tcp_clock_cache; + skb_set_delivery_time(skb, tp->tcp_wstamp_ns, SKB_CLOCK_MONOTONIC); list_move_tail(&skb->tcp_tsorted_anchor, &tp->tsorted_sent_queue); + tcp_init_tso_segs(skb, mss_now); goto repair; /* Skip network transmission */ } if (tcp_pacing_check(sk)) break; - tso_segs = tcp_init_tso_segs(skb, mss_now); - BUG_ON(!tso_segs); - - cwnd_quota = tcp_cwnd_test(tp, skb); + cwnd_quota = tcp_cwnd_test(tp); if (!cwnd_quota) { if (push_one == 2) /* Force out a loss probe pkt. */ @@ -2360,6 +2954,12 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle, else break; } + cwnd_quota = min(cwnd_quota, max_segs); + missing_bytes = cwnd_quota * mss_now - skb->len; + if (missing_bytes > 0) + tcp_grow_skb(sk, skb, missing_bytes); + + tso_segs = tcp_set_skb_tso_segs(skb, mss_now); if (unlikely(!tcp_snd_wnd_test(tp, skb, mss_now))) { is_rwnd_limited = true; @@ -2381,19 +2981,24 @@ static bool tcp_write_xmit(struct sock *sk, unsigned int mss_now, int nonagle, limit = mss_now; if (tso_segs > 1 && !tcp_urg_mode(tp)) limit = tcp_mss_split_point(sk, skb, mss_now, - min_t(unsigned int, - cwnd_quota, - max_segs), + cwnd_quota, nonagle); if (skb->len > limit && - unlikely(tso_fragment(sk, TCP_FRAG_IN_WRITE_QUEUE, - skb, limit, mss_now, gfp))) + unlikely(tso_fragment(sk, skb, limit, mss_now, gfp))) break; if (tcp_small_queue_check(sk, skb, 0)) break; + /* Argh, we hit an empty skb(), presumably a thread + * is sleeping in sendmsg()/sk_stream_wait_memory(). + * We do not want to send a pure-ack packet and have + * a strange looking rtx queue with empty packet(s). + */ + if (TCP_SKB_CB(skb)->end_seq == TCP_SKB_CB(skb)->seq) + break; + if (unlikely(tcp_transmit_skb(sk, skb, 1, gfp))) break; @@ -2415,6 +3020,10 @@ repair: else tcp_chrono_stop(sk, TCP_CHRONO_RWND_LIMITED); + is_cwnd_limited |= (tcp_packets_in_flight(tp) >= tcp_snd_cwnd(tp)); + if (likely(sent_pkts || is_cwnd_limited)) + tcp_cwnd_validate(sk, is_cwnd_limited); + if (likely(sent_pkts)) { if (tcp_in_cwnd_reduction(sk)) tp->prr_out += sent_pkts; @@ -2422,8 +3031,6 @@ repair: /* Send one loss probe per tail loss episode. */ if (push_one != 2) tcp_schedule_loss_probe(sk, false); - is_cwnd_limited |= (tcp_packets_in_flight(tp) >= tp->snd_cwnd); - tcp_cwnd_validate(sk, is_cwnd_limited); return false; } return !tp->packets_out && !tcp_write_queue_empty(sk); @@ -2433,16 +3040,16 @@ bool tcp_schedule_loss_probe(struct sock *sk, bool advancing_rto) { struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); - u32 timeout, rto_delta_us; + u32 timeout, timeout_us, rto_delta_us; int early_retrans; /* Don't do any loss probe on a Fast Open connection before 3WHS * finishes. */ - if (tp->fastopen_rsk) + if (rcu_access_pointer(tp->fastopen_rsk)) return false; - early_retrans = sock_net(sk)->ipv4.sysctl_tcp_early_retrans; + early_retrans = READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_early_retrans); /* Schedule a loss probe in 2*RTT for SACK capable connections * not in loss recovery, that are either limited by cwnd or application. */ @@ -2457,11 +3064,12 @@ bool tcp_schedule_loss_probe(struct sock *sk, bool advancing_rto) * sample is available then probe after TCP_TIMEOUT_INIT. */ if (tp->srtt_us) { - timeout = usecs_to_jiffies(tp->srtt_us >> 2); + timeout_us = tp->srtt_us >> 2; if (tp->packets_out == 1) - timeout += TCP_RTO_MIN; + timeout_us += tcp_rto_min_us(sk); else - timeout += TCP_TIMEOUT_MIN; + timeout_us += TCP_TIMEOUT_MIN_US; + timeout = usecs_to_jiffies(timeout_us); } else { timeout = TCP_TIMEOUT_INIT; } @@ -2473,8 +3081,7 @@ bool tcp_schedule_loss_probe(struct sock *sk, bool advancing_rto) if (rto_delta_us > 0) timeout = min_t(u32, timeout, usecs_to_jiffies(rto_delta_us)); - tcp_reset_xmit_timer(sk, ICSK_TIME_LOSS_PROBE, timeout, - TCP_RTO_MAX, NULL); + tcp_reset_xmit_timer(sk, ICSK_TIME_LOSS_PROBE, timeout, true); return true; } @@ -2482,13 +3089,17 @@ bool tcp_schedule_loss_probe(struct sock *sk, bool advancing_rto) * a packet is still in a qdisc or driver queue. * In this case, there is very little point doing a retransmit ! */ -static bool skb_still_in_host_queue(const struct sock *sk, +static bool skb_still_in_host_queue(struct sock *sk, const struct sk_buff *skb) { if (unlikely(skb_fclone_busy(sk, skb))) { - NET_INC_STATS(sock_net(sk), - LINUX_MIB_TCPSPURIOUS_RTX_HOSTQUEUES); - return true; + set_bit(TSQ_THROTTLED, &sk->sk_tsq_flags); + smp_mb__after_atomic(); + if (skb_fclone_busy(sk, skb)) { + NET_INC_STATS(sock_net(sk), + LINUX_MIB_TCPSPURIOUS_RTX_HOSTQUEUES); + return true; + } } return false; } @@ -2503,6 +3114,11 @@ void tcp_send_loss_probe(struct sock *sk) int pcount; int mss = tcp_current_mss(sk); + /* At most one outstanding TLP */ + if (tp->tlp_high_seq) + goto rearm_timer; + + tp->tlp_retrans = 0; skb = tcp_send_head(sk); if (skb && tcp_snd_wnd_test(tp, skb, mss)) { pcount = tp->packets_out; @@ -2513,17 +3129,11 @@ void tcp_send_loss_probe(struct sock *sk) } skb = skb_rb_last(&sk->tcp_rtx_queue); if (unlikely(!skb)) { - WARN_ONCE(tp->packets_out, - "invalid inflight: %u state %u cwnd %u mss %d\n", - tp->packets_out, sk->sk_state, tp->snd_cwnd, mss); - inet_csk(sk)->icsk_pending = 0; + tcp_warn_once(sk, tp->packets_out, "invalid inflight: "); + smp_store_release(&inet_csk(sk)->icsk_pending, 0); return; } - /* At most one outstanding TLP retransmission. */ - if (tp->tlp_high_seq) - goto rearm_timer; - if (skb_still_in_host_queue(sk, skb)) goto rearm_timer; @@ -2545,13 +3155,15 @@ void tcp_send_loss_probe(struct sock *sk) if (__tcp_retransmit_skb(sk, skb, 1)) goto rearm_timer; + tp->tlp_retrans = 1; + +probe_sent: /* Record snd_nxt for loss detection. */ tp->tlp_high_seq = tp->snd_nxt; -probe_sent: NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPLOSSPROBES); /* Reset s.t. tcp_rearm_rto will restart timer from now */ - inet_csk(sk)->icsk_pending = 0; + smp_store_release(&inet_csk(sk)->icsk_pending, 0); rearm_timer: tcp_rearm_rto(sk); } @@ -2643,6 +3255,7 @@ u32 __tcp_select_window(struct sock *sk) { struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); + struct net *net = sock_net(sk); /* MSS for the peer's data. Previous versions used mss_clamp * here. I don't know if the value based on our guesses * of peer's MSS is better for the performance. It's more correct @@ -2652,20 +3265,32 @@ u32 __tcp_select_window(struct sock *sk) int mss = icsk->icsk_ack.rcv_mss; int free_space = tcp_space(sk); int allowed_space = tcp_full_space(sk); - int full_space = min_t(int, tp->window_clamp, allowed_space); - int window; + int full_space, window; + + if (sk_is_mptcp(sk)) + mptcp_space(sk, &free_space, &allowed_space); + + full_space = min_t(int, tp->window_clamp, allowed_space); if (unlikely(mss > full_space)) { mss = full_space; if (mss <= 0) return 0; } + + /* Only allow window shrink if the sysctl is enabled and we have + * a non-zero scaling factor in effect. + */ + if (READ_ONCE(net->ipv4.sysctl_tcp_shrink_window) && tp->rx_opt.rcv_wscale) + goto shrink_window_allowed; + + /* do not allow window to shrink */ + if (free_space < (full_space >> 1)) { icsk->icsk_ack.quick = 0; if (tcp_under_memory_pressure(sk)) - tp->rcv_ssthresh = min(tp->rcv_ssthresh, - 4U * tp->advmss); + tcp_adjust_rcv_ssthresh(sk); /* free_space might become our new window, make sure we don't * increase it due to wscale. @@ -2715,6 +3340,36 @@ u32 __tcp_select_window(struct sock *sk) } return window; + +shrink_window_allowed: + /* new window should always be an exact multiple of scaling factor */ + free_space = round_down(free_space, 1 << tp->rx_opt.rcv_wscale); + + if (free_space < (full_space >> 1)) { + icsk->icsk_ack.quick = 0; + + if (tcp_under_memory_pressure(sk)) + tcp_adjust_rcv_ssthresh(sk); + + /* if free space is too low, return a zero window */ + if (free_space < (allowed_space >> 4) || free_space < mss || + free_space < (1 << tp->rx_opt.rcv_wscale)) + return 0; + } + + if (free_space > tp->rcv_ssthresh) { + free_space = tp->rcv_ssthresh; + /* new window should always be an exact multiple of scaling factor + * + * For this case, we ALIGN "up" (increase free_space) because + * we know free_space is not zero here, it has been reduced from + * the memory-based limit, and rcv_ssthresh is not a hard limit + * (unlike sk_rcvbuf). + */ + free_space = ALIGN(free_space, (1 << tp->rx_opt.rcv_wscale)); + } + + return free_space; } void tcp_skb_collapse_tstamp(struct sk_buff *skb, @@ -2743,13 +3398,9 @@ static bool tcp_collapse_retrans(struct sock *sk, struct sk_buff *skb) BUG_ON(tcp_skb_pcount(skb) != 1 || tcp_skb_pcount(next_skb) != 1); - if (next_skb_size) { - if (next_skb_size <= skb_availroom(skb)) - skb_copy_bits(next_skb, 0, skb_put(skb, next_skb_size), - next_skb_size); - else if (!skb_shift(skb, next_skb, next_skb_size)) - return false; - } + if (next_skb_size && !tcp_skb_shift(skb, next_skb, 1, next_skb_size)) + return false; + tcp_highest_sack_replace(sk, next_skb, skb); /* Update sequence range on original skb. */ @@ -2765,7 +3416,6 @@ static bool tcp_collapse_retrans(struct sock *sk, struct sk_buff *skb) TCP_SKB_CB(skb)->eor = TCP_SKB_CB(next_skb)->eor; /* changed transmit queue under us so clear hints */ - tcp_clear_retrans_hints_partial(tp); if (next_skb == tp->retransmit_skb_hint) tp->retransmit_skb_hint = skb; @@ -2784,6 +3434,8 @@ static bool tcp_can_collapse(const struct sock *sk, const struct sk_buff *skb) return false; if (skb_cloned(skb)) return false; + if (!skb_frags_readable(skb)) + return false; /* Some heuristics for collapsing over SACK'd could be invented */ if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED) return false; @@ -2801,7 +3453,7 @@ static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *to, struct sk_buff *skb = to, *tmp; bool first = true; - if (!sock_net(sk)->ipv4.sysctl_tcp_retrans_collapse) + if (!READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_retrans_collapse)) return; if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN) return; @@ -2810,7 +3462,7 @@ static void tcp_retrans_try_collapse(struct sock *sk, struct sk_buff *to, if (!tcp_can_collapse(sk, skb)) break; - if (!tcp_skb_can_collapse_to(to)) + if (!tcp_skb_can_collapse(to, skb)) break; space -= skb->len; @@ -2841,65 +3493,88 @@ int __tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb, int segs) struct tcp_sock *tp = tcp_sk(sk); unsigned int cur_mss; int diff, len, err; - + int avail_wnd; /* Inconclusive MTU probe */ if (icsk->icsk_mtup.probe_size) icsk->icsk_mtup.probe_size = 0; - /* Do not sent more than we queued. 1/4 is reserved for possible - * copying overhead: fragmentation, tunneling, mangling etc. - */ - if (refcount_read(&sk->sk_wmem_alloc) > - min_t(u32, sk->sk_wmem_queued + (sk->sk_wmem_queued >> 2), - sk->sk_sndbuf)) - return -EAGAIN; - - if (skb_still_in_host_queue(sk, skb)) - return -EBUSY; + if (skb_still_in_host_queue(sk, skb)) { + err = -EBUSY; + goto out; + } +start: if (before(TCP_SKB_CB(skb)->seq, tp->snd_una)) { + if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) { + TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_SYN; + TCP_SKB_CB(skb)->seq++; + goto start; + } if (unlikely(before(TCP_SKB_CB(skb)->end_seq, tp->snd_una))) { WARN_ON_ONCE(1); - return -EINVAL; + err = -EINVAL; + goto out; + } + if (tcp_trim_head(sk, skb, tp->snd_una - TCP_SKB_CB(skb)->seq)) { + err = -ENOMEM; + goto out; } - if (tcp_trim_head(sk, skb, tp->snd_una - TCP_SKB_CB(skb)->seq)) - return -ENOMEM; } - if (inet_csk(sk)->icsk_af_ops->rebuild_header(sk)) - return -EHOSTUNREACH; /* Routing failure or similar. */ + if (inet_csk(sk)->icsk_af_ops->rebuild_header(sk)) { + err = -EHOSTUNREACH; /* Routing failure or similar. */ + goto out; + } cur_mss = tcp_current_mss(sk); + avail_wnd = tcp_wnd_end(tp) - TCP_SKB_CB(skb)->seq; /* If receiver has shrunk his window, and skb is out of * new window, do not retransmit it. The exception is the * case, when window is shrunk to zero. In this case - * our retransmit serves as a zero window probe. + * our retransmit of one segment serves as a zero window probe. */ - if (!before(TCP_SKB_CB(skb)->seq, tcp_wnd_end(tp)) && - TCP_SKB_CB(skb)->seq != tp->snd_una) - return -EAGAIN; + if (avail_wnd <= 0) { + if (TCP_SKB_CB(skb)->seq != tp->snd_una) { + err = -EAGAIN; + goto out; + } + avail_wnd = cur_mss; + } len = cur_mss * segs; + if (len > avail_wnd) { + len = rounddown(avail_wnd, cur_mss); + if (!len) + len = avail_wnd; + } if (skb->len > len) { if (tcp_fragment(sk, TCP_FRAG_IN_RTX_QUEUE, skb, len, - cur_mss, GFP_ATOMIC)) - return -ENOMEM; /* We'll try again later. */ + cur_mss, GFP_ATOMIC)) { + err = -ENOMEM; /* We'll try again later. */ + goto out; + } } else { - if (skb_unclone(skb, GFP_ATOMIC)) - return -ENOMEM; + if (skb_unclone_keeptruesize(skb, GFP_ATOMIC)) { + err = -ENOMEM; + goto out; + } diff = tcp_skb_pcount(skb); tcp_set_skb_tso_segs(skb, cur_mss); diff -= tcp_skb_pcount(skb); if (diff) tcp_adjust_pcount(sk, skb, diff); - if (skb->len < cur_mss) - tcp_retrans_try_collapse(sk, skb, cur_mss); + avail_wnd = min_t(int, avail_wnd, cur_mss); + if (skb->len < avail_wnd) + tcp_retrans_try_collapse(sk, skb, avail_wnd); } - /* RFC3168, section 6.1.1.1. ECN fallback */ + /* RFC3168, section 6.1.1.1. ECN fallback + * As AccECN uses the same SYN flags (+ AE), this check covers both + * cases. + */ if ((TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN_ECN) == TCPHDR_SYN_ECN) tcp_ecn_clear_syn(sk, skb); @@ -2921,8 +3596,12 @@ int __tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb, int segs) tcp_skb_tsorted_save(skb) { nskb = __pskb_copy(skb, MAX_TCP_HEADER, GFP_ATOMIC); - err = nskb ? tcp_transmit_skb(sk, nskb, 0, GFP_ATOMIC) : - -ENOBUFS; + if (nskb) { + nskb->dev = NULL; + err = tcp_transmit_skb(sk, nskb, 0, GFP_ATOMIC); + } else { + err = -ENOBUFS; + } } tcp_skb_tsorted_restore(skb); if (!err) { @@ -2933,20 +3612,20 @@ int __tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb, int segs) err = tcp_transmit_skb(sk, skb, 1, GFP_ATOMIC); } - /* To avoid taking spuriously low RTT samples based on a timestamp - * for a transmit that never happened, always mark EVER_RETRANS - */ - TCP_SKB_CB(skb)->sacked |= TCPCB_EVER_RETRANS; - if (BPF_SOCK_OPS_TEST_FLAG(tp, BPF_SOCK_OPS_RETRANS_CB_FLAG)) tcp_call_bpf_3arg(sk, BPF_SOCK_OPS_RETRANS_CB, TCP_SKB_CB(skb)->seq, segs, err); - if (likely(!err)) { - trace_tcp_retransmit_skb(sk, skb); - } else if (err != -EBUSY) { + if (unlikely(err) && err != -EBUSY) NET_ADD_STATS(sock_net(sk), LINUX_MIB_TCPRETRANSFAIL, segs); - } + + /* To avoid taking spuriously low RTT samples based on a timestamp + * for a transmit that never happened, always mark EVER_RETRANS + */ + TCP_SKB_CB(skb)->sacked |= TCPCB_EVER_RETRANS; + +out: + trace_tcp_retransmit_skb(sk, skb, err); return err; } @@ -2967,7 +3646,7 @@ int tcp_retransmit_skb(struct sock *sk, struct sk_buff *skb, int segs) /* Save stamp of the first (attempted) retransmit. */ if (!tp->retrans_stamp) - tp->retrans_stamp = tcp_skb_timestamp(skb); + tp->retrans_stamp = tcp_skb_timestamp_ts(tp->tcp_usec_ts, skb); if (tp->undo_retrans < 0) tp->undo_retrans = 0; @@ -2985,6 +3664,7 @@ void tcp_xmit_retransmit_queue(struct sock *sk) const struct inet_connection_sock *icsk = inet_csk(sk); struct sk_buff *skb, *rtx_head, *hole = NULL; struct tcp_sock *tp = tcp_sk(sk); + bool rearm_timer = false; u32 max_segs; int mib_idx; @@ -3005,9 +3685,9 @@ void tcp_xmit_retransmit_queue(struct sock *sk) if (!hole) tp->retransmit_skb_hint = skb; - segs = tp->snd_cwnd - tcp_packets_in_flight(tp); + segs = tcp_snd_cwnd(tp) - tcp_packets_in_flight(tp); if (segs <= 0) - return; + break; sacked = TCP_SKB_CB(skb)->sacked; /* In case tcp_shift_skb_data() have aggregated large skbs, * we need to make sure not sending too bigs TSO packets @@ -3032,10 +3712,10 @@ void tcp_xmit_retransmit_queue(struct sock *sk) continue; if (tcp_small_queue_check(sk, skb, 1)) - return; + break; if (tcp_retransmit_skb(sk, skb, segs)) - return; + break; NET_ADD_STATS(sock_net(sk), mib_idx, tcp_skb_pcount(skb)); @@ -3044,11 +3724,12 @@ void tcp_xmit_retransmit_queue(struct sock *sk) if (skb == rtx_head && icsk->icsk_pending != ICSK_TIME_REO_TIMEOUT) - tcp_reset_xmit_timer(sk, ICSK_TIME_RETRANS, - inet_csk(sk)->icsk_rto, - TCP_RTO_MAX, - skb); + rearm_timer = true; + } + if (rearm_timer) + tcp_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + inet_csk(sk)->icsk_rto, true); } /* We allow to exceed memory limits for FIN packets to expedite @@ -3060,16 +3741,22 @@ void tcp_xmit_retransmit_queue(struct sock *sk) */ void sk_forced_mem_schedule(struct sock *sk, int size) { - int amt; + int delta, amt; - if (size <= sk->sk_forward_alloc) + delta = size - sk->sk_forward_alloc; + if (delta <= 0) return; - amt = sk_mem_pages(size); - sk->sk_forward_alloc += amt * SK_MEM_QUANTUM; - sk_memory_allocated_add(sk, amt); - if (mem_cgroup_sockets_enabled && sk->sk_memcg) - mem_cgroup_charge_skmem(sk->sk_memcg, amt); + amt = sk_mem_pages(delta); + sk_forward_alloc_add(sk, amt << PAGE_SHIFT); + + if (mem_cgroup_sk_enabled(sk)) + mem_cgroup_sk_charge(sk, amt, gfp_memcg_charge() | __GFP_NOFAIL); + + if (sk->sk_bypass_prot_mem) + return; + + sk_memory_allocated_add(sk, amt); } /* Send a FIN. The caller locks the socket for us. @@ -3077,7 +3764,7 @@ void sk_forced_mem_schedule(struct sock *sk, int size) */ void tcp_send_fin(struct sock *sk) { - struct sk_buff *skb, *tskb = tcp_write_queue_tail(sk); + struct sk_buff *skb, *tskb, *tail = tcp_write_queue_tail(sk); struct tcp_sock *tp = tcp_sk(sk); /* Optimization, tack on the FIN if we have one skb in write queue and @@ -3085,36 +3772,36 @@ void tcp_send_fin(struct sock *sk) * Note: in the latter case, FIN packet will be sent after a timeout, * as TCP stack thinks it has already been transmitted. */ + tskb = tail; if (!tskb && tcp_under_memory_pressure(sk)) tskb = skb_rb_last(&sk->tcp_rtx_queue); if (tskb) { -coalesce: TCP_SKB_CB(tskb)->tcp_flags |= TCPHDR_FIN; TCP_SKB_CB(tskb)->end_seq++; tp->write_seq++; - if (tcp_write_queue_empty(sk)) { + if (!tail) { /* This means tskb was already sent. * Pretend we included the FIN on previous transmit. * We need to set tp->snd_nxt to the value it would have * if FIN had been sent. This is because retransmit path * does not change tp->snd_nxt. */ - tp->snd_nxt++; + WRITE_ONCE(tp->snd_nxt, tp->snd_nxt + 1); return; } } else { - skb = alloc_skb_fclone(MAX_TCP_HEADER, sk->sk_allocation); - if (unlikely(!skb)) { - if (tskb) - goto coalesce; + skb = alloc_skb_fclone(MAX_TCP_HEADER, + sk_gfp_mask(sk, GFP_ATOMIC | + __GFP_NOWARN)); + if (unlikely(!skb)) return; - } + INIT_LIST_HEAD(&skb->tcp_tsorted_anchor); skb_reserve(skb, MAX_TCP_HEADER); sk_forced_mem_schedule(sk, skb->truesize); /* FIN eats a sequence byte, write_seq advanced by tcp_queue_skb(). */ - tcp_init_nondata_skb(skb, tp->write_seq, + tcp_init_nondata_skb(skb, sk, tp->write_seq, TCPHDR_ACK | TCPHDR_FIN); tcp_queue_skb(sk, skb); } @@ -3126,7 +3813,8 @@ coalesce: * was unread data in the receive queue. This behavior is recommended * by RFC 2525, section 2.17. -DaveM */ -void tcp_send_active_reset(struct sock *sk, gfp_t priority) +void tcp_send_active_reset(struct sock *sk, gfp_t priority, + enum sk_rst_reason reason) { struct sk_buff *skb; @@ -3141,7 +3829,7 @@ void tcp_send_active_reset(struct sock *sk, gfp_t priority) /* Reserve space for headers and prepare control bits. */ skb_reserve(skb, MAX_TCP_HEADER); - tcp_init_nondata_skb(skb, tcp_acceptable_seq(sk), + tcp_init_nondata_skb(skb, sk, tcp_acceptable_seq(sk), TCPHDR_ACK | TCPHDR_RST); tcp_mstamp_refresh(tcp_sk(sk)); /* Send it off. */ @@ -3151,7 +3839,7 @@ void tcp_send_active_reset(struct sock *sk, gfp_t priority) /* skb of trace_tcp_send_reset() keeps the skb that caused RST, * skb here is different to the troublesome skb, so use NULL */ - trace_tcp_send_reset(sk, NULL); + trace_tcp_send_reset(sk, NULL, reason); } /* Send a crossed SYN-ACK during socket establishment. @@ -3179,10 +3867,11 @@ int tcp_send_synack(struct sock *sk) if (!nskb) return -ENOMEM; INIT_LIST_HEAD(&nskb->tcp_tsorted_anchor); + tcp_highest_sack_replace(sk, skb, nskb); tcp_rtx_queue_unlink_and_free(skb, sk); __skb_header_release(nskb); tcp_rbtree_insert(&sk->tcp_rtx_queue, nskb); - sk->sk_wmem_queued += nskb->truesize; + sk_wmem_queued_add(sk, nskb->truesize); sk_mem_charge(sk, nskb->truesize); skb = nskb; } @@ -3194,27 +3883,30 @@ int tcp_send_synack(struct sock *sk) } /** - * tcp_make_synack - Prepare a SYN-ACK. - * sk: listener socket - * dst: dst entry attached to the SYNACK - * req: request_sock pointer - * - * Allocate one skb and build a SYNACK packet. - * @dst is consumed : Caller should not use it again. + * tcp_make_synack - Allocate one skb and build a SYNACK packet. + * @sk: listener socket + * @dst: dst entry attached to the SYNACK. It is consumed and caller + * should not use it again. + * @req: request_sock pointer + * @foc: cookie for tcp fast open + * @synack_type: Type of synack to prepare + * @syn_skb: SYN packet just received. It could be NULL for rtx case. */ struct sk_buff *tcp_make_synack(const struct sock *sk, struct dst_entry *dst, struct request_sock *req, struct tcp_fastopen_cookie *foc, - enum tcp_synack_type synack_type) + enum tcp_synack_type synack_type, + struct sk_buff *syn_skb) { struct inet_request_sock *ireq = inet_rsk(req); const struct tcp_sock *tp = tcp_sk(sk); - struct tcp_md5sig_key *md5 = NULL; struct tcp_out_options opts; + struct tcp_key key = {}; struct sk_buff *skb; int tcp_header_size; struct tcphdr *th; int mss; + u64 now; skb = alloc_skb(MAX_TCP_HEADER, GFP_ATOMIC); if (unlikely(!skb)) { @@ -3226,7 +3918,7 @@ struct sk_buff *tcp_make_synack(const struct sock *sk, struct dst_entry *dst, switch (synack_type) { case TCP_SYNACK_NORMAL: - skb_set_owner_w(skb, req_to_sk(req)); + skb_set_owner_edemux(skb, req_to_sk(req)); break; case TCP_SYNACK_COOKIE: /* Under synflood, we do not attach skb to a socket, @@ -3246,20 +3938,60 @@ struct sk_buff *tcp_make_synack(const struct sock *sk, struct dst_entry *dst, mss = tcp_mss_clamp(tp, dst_metric_advmss(dst)); memset(&opts, 0, sizeof(opts)); + now = tcp_clock_ns(); #ifdef CONFIG_SYN_COOKIES - if (unlikely(req->cookie_ts)) - skb->skb_mstamp_ns = cookie_init_timestamp(req); + if (unlikely(synack_type == TCP_SYNACK_COOKIE && ireq->tstamp_ok)) + skb_set_delivery_time(skb, cookie_init_timestamp(req, now), + SKB_CLOCK_MONOTONIC); else #endif - skb->skb_mstamp_ns = tcp_clock_ns(); + { + skb_set_delivery_time(skb, now, SKB_CLOCK_MONOTONIC); + if (!tcp_rsk(req)->snt_synack) /* Timestamp first SYNACK */ + tcp_rsk(req)->snt_synack = tcp_skb_timestamp_us(skb); + } -#ifdef CONFIG_TCP_MD5SIG +#if defined(CONFIG_TCP_MD5SIG) || defined(CONFIG_TCP_AO) rcu_read_lock(); - md5 = tcp_rsk(req)->af_specific->req_md5_lookup(sk, req_to_sk(req)); #endif - skb_set_hash(skb, tcp_rsk(req)->txhash, PKT_HASH_TYPE_L4); - tcp_header_size = tcp_synack_options(sk, req, mss, skb, &opts, md5, - foc) + sizeof(*th); + if (tcp_rsk_used_ao(req)) { +#ifdef CONFIG_TCP_AO + struct tcp_ao_key *ao_key = NULL; + u8 keyid = tcp_rsk(req)->ao_keyid; + u8 rnext = tcp_rsk(req)->ao_rcv_next; + + ao_key = tcp_sk(sk)->af_specific->ao_lookup(sk, req_to_sk(req), + keyid, -1); + /* If there is no matching key - avoid sending anything, + * especially usigned segments. It could try harder and lookup + * for another peer-matching key, but the peer has requested + * ao_keyid (RFC5925 RNextKeyID), so let's keep it simple here. + */ + if (unlikely(!ao_key)) { + trace_tcp_ao_synack_no_key(sk, keyid, rnext); + rcu_read_unlock(); + kfree_skb(skb); + net_warn_ratelimited("TCP-AO: the keyid %u from SYN packet is not present - not sending SYNACK\n", + keyid); + return NULL; + } + key.ao_key = ao_key; + key.type = TCP_KEY_AO; +#endif + } else { +#ifdef CONFIG_TCP_MD5SIG + key.md5_key = tcp_rsk(req)->af_specific->req_md5_lookup(sk, + req_to_sk(req)); + if (key.md5_key) + key.type = TCP_KEY_MD5; +#endif + } + skb_set_hash(skb, READ_ONCE(tcp_rsk(req)->txhash), PKT_HASH_TYPE_L4); + /* bpf program will be interested in the tcp_flags */ + TCP_SKB_CB(skb)->tcp_flags = TCPHDR_SYN | TCPHDR_ACK; + tcp_header_size = tcp_synack_options(sk, req, mss, skb, &opts, + &key, foc, synack_type, syn_skb) + + sizeof(*th); skb_push(skb, tcp_header_size); skb_reset_transport_header(skb); @@ -3279,23 +4011,36 @@ struct sk_buff *tcp_make_synack(const struct sock *sk, struct dst_entry *dst, /* RFC1323: The window in SYN & SYN/ACK segments is never scaled. */ th->window = htons(min(req->rsk_rcv_wnd, 65535U)); - tcp_options_write((__be32 *)(th + 1), NULL, &opts); + tcp_options_write(th, NULL, tcp_rsk(req), &opts, &key); th->doff = (tcp_header_size >> 2); - __TCP_INC_STATS(sock_net(sk), TCP_MIB_OUTSEGS); + TCP_INC_STATS(sock_net(sk), TCP_MIB_OUTSEGS); -#ifdef CONFIG_TCP_MD5SIG /* Okay, we have all we need - do the md5 hash if needed */ - if (md5) + if (tcp_key_is_md5(&key)) { +#ifdef CONFIG_TCP_MD5SIG tcp_rsk(req)->af_specific->calc_md5_hash(opts.hash_location, - md5, req_to_sk(req), skb); + key.md5_key, req_to_sk(req), skb); +#endif + } else if (tcp_key_is_ao(&key)) { +#ifdef CONFIG_TCP_AO + tcp_rsk(req)->af_specific->ao_synack_hash(opts.hash_location, + key.ao_key, req, skb, + opts.hash_location - (u8 *)th, 0); +#endif + } +#if defined(CONFIG_TCP_MD5SIG) || defined(CONFIG_TCP_AO) rcu_read_unlock(); #endif - /* Do not fool tcpdump (if any), clean our debris */ - skb->tstamp = 0; + bpf_skops_write_hdr_opt((struct sock *)sk, skb, req, syn_skb, + synack_type, &opts); + + skb_set_delivery_time(skb, now, SKB_CLOCK_MONOTONIC); + tcp_add_tx_delay(skb, tp); + return skb; } -EXPORT_SYMBOL(tcp_make_synack); +EXPORT_IPV6_MOD(tcp_make_synack); static void tcp_ca_dst_init(struct sock *sk, const struct dst_entry *dst) { @@ -3308,8 +4053,8 @@ static void tcp_ca_dst_init(struct sock *sk, const struct dst_entry *dst) rcu_read_lock(); ca = tcp_ca_find_key(ca_key); - if (likely(ca && try_module_get(ca->owner))) { - module_put(icsk->icsk_ca_ops->owner); + if (likely(ca && bpf_try_module_get(ca, ca->owner))) { + bpf_module_put(icsk->icsk_ca_ops, icsk->icsk_ca_ops->owner); icsk->icsk_ca_dst_locked = tcp_ca_dst_locked(dst); icsk->icsk_ca_ops = ca; } @@ -3322,23 +4067,22 @@ static void tcp_connect_init(struct sock *sk) const struct dst_entry *dst = __sk_dst_get(sk); struct tcp_sock *tp = tcp_sk(sk); __u8 rcv_wscale; + u16 user_mss; u32 rcv_wnd; /* We'll fix this up when we get a response from the other end. * See tcp_input.c:tcp_rcv_state_process case TCP_SYN_SENT. */ tp->tcp_header_len = sizeof(struct tcphdr); - if (sock_net(sk)->ipv4.sysctl_tcp_timestamps) + if (READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_timestamps)) tp->tcp_header_len += TCPOLEN_TSTAMP_ALIGNED; -#ifdef CONFIG_TCP_MD5SIG - if (tp->af_specific->md5_lookup(sk, sk)) - tp->tcp_header_len += TCPOLEN_MD5SIG_ALIGNED; -#endif + tcp_ao_connect_init(sk); /* If user gave his TCP_MAXSEG, record it to clamp */ - if (tp->rx_opt.user_mss) - tp->rx_opt.mss_clamp = tp->rx_opt.user_mss; + user_mss = READ_ONCE(tp->rx_opt.user_mss); + if (user_mss) + tp->rx_opt.mss_clamp = user_mss; tp->max_window = 0; tcp_mtup_init(sk); tcp_sync_mss(sk, dst_mtu(dst)); @@ -3346,7 +4090,7 @@ static void tcp_connect_init(struct sock *sk) tcp_ca_dst_init(sk, dst); if (!tp->window_clamp) - tp->window_clamp = dst_metric(dst, RTAX_WINDOW); + WRITE_ONCE(tp->window_clamp, dst_metric(dst, RTAX_WINDOW)); tp->advmss = tcp_mss_clamp(tp, dst_metric_advmss(dst)); tcp_initialize_rcv_mss(sk); @@ -3354,7 +4098,7 @@ static void tcp_connect_init(struct sock *sk) /* limit the window selection if the user enforce a smaller rx buffer */ if (sk->sk_userlocks & SOCK_RCVBUF_LOCK && (tp->window_clamp > tcp_full_space(sk) || tp->window_clamp == 0)) - tp->window_clamp = tcp_full_space(sk); + WRITE_ONCE(tp->window_clamp, tcp_full_space(sk)); rcv_wnd = tcp_rwnd_init_bpf(sk); if (rcv_wnd == 0) @@ -3364,14 +4108,14 @@ static void tcp_connect_init(struct sock *sk) tp->advmss - (tp->rx_opt.ts_recent_stamp ? tp->tcp_header_len - sizeof(struct tcphdr) : 0), &tp->rcv_wnd, &tp->window_clamp, - sock_net(sk)->ipv4.sysctl_tcp_window_scaling, + READ_ONCE(sock_net(sk)->ipv4.sysctl_tcp_window_scaling), &rcv_wscale, rcv_wnd); tp->rx_opt.rcv_wscale = rcv_wscale; tp->rcv_ssthresh = tp->rcv_wnd; - sk->sk_err = 0; + WRITE_ONCE(sk->sk_err, 0); sock_reset_flag(sk, SOCK_DONE); tp->snd_wnd = 0; tcp_init_wl(tp, 0); @@ -3379,17 +4123,17 @@ static void tcp_connect_init(struct sock *sk) tp->snd_una = tp->write_seq; tp->snd_sml = tp->write_seq; tp->snd_up = tp->write_seq; - tp->snd_nxt = tp->write_seq; + WRITE_ONCE(tp->snd_nxt, tp->write_seq); if (likely(!tp->repair)) tp->rcv_nxt = 0; else tp->rcv_tstamp = tcp_jiffies32; tp->rcv_wup = tp->rcv_nxt; - tp->copied_seq = tp->rcv_nxt; + WRITE_ONCE(tp->copied_seq, tp->rcv_nxt); inet_csk(sk)->icsk_rto = tcp_timeout_init(sk); - inet_csk(sk)->icsk_retransmits = 0; + WRITE_ONCE(inet_csk(sk)->icsk_retransmits, 0); tcp_clear_retrans(tp); } @@ -3400,9 +4144,9 @@ static void tcp_connect_queue_skb(struct sock *sk, struct sk_buff *skb) tcb->end_seq += skb->len; __skb_header_release(skb); - sk->sk_wmem_queued += skb->truesize; + sk_wmem_queued_add(sk, skb->truesize); sk_mem_charge(sk, skb->truesize); - tp->write_seq = tcb->end_seq; + WRITE_ONCE(tp->write_seq, tcb->end_seq); tp->packets_out += tcp_skb_pcount(skb); } @@ -3415,10 +4159,12 @@ static void tcp_connect_queue_skb(struct sock *sk, struct sk_buff *skb) */ static int tcp_send_syn_data(struct sock *sk, struct sk_buff *syn) { + struct inet_connection_sock *icsk = inet_csk(sk); struct tcp_sock *tp = tcp_sk(sk); struct tcp_fastopen_request *fo = tp->fastopen_req; - int space, err = 0; + struct page_frag *pfrag = sk_page_frag(sk); struct sk_buff *syn_data; + int space, err = 0; tp->rx_opt.mss_clamp = tp->advmss; /* If MSS is not cached */ if (!tcp_fastopen_cookie_check(sk, &tp->rx_opt.mss_clamp, &fo->cookie)) @@ -3429,32 +4175,39 @@ static int tcp_send_syn_data(struct sock *sk, struct sk_buff *syn) * private TCP options. The cost is reduced data space in SYN :( */ tp->rx_opt.mss_clamp = tcp_mss_clamp(tp, tp->rx_opt.mss_clamp); + /* Sync mss_cache after updating the mss_clamp */ + tcp_sync_mss(sk, icsk->icsk_pmtu_cookie); - space = __tcp_mtu_to_mss(sk, inet_csk(sk)->icsk_pmtu_cookie) - + space = __tcp_mtu_to_mss(sk, icsk->icsk_pmtu_cookie) - MAX_TCP_OPTION_SPACE; space = min_t(size_t, space, fo->size); - /* limit to order-0 allocations */ - space = min_t(size_t, space, SKB_MAX_HEAD(MAX_TCP_HEADER)); - - syn_data = sk_stream_alloc_skb(sk, space, sk->sk_allocation, false); + if (space && + !skb_page_frag_refill(min_t(size_t, space, PAGE_SIZE), + pfrag, sk->sk_allocation)) + goto fallback; + syn_data = tcp_stream_alloc_skb(sk, sk->sk_allocation, false); if (!syn_data) goto fallback; - syn_data->ip_summed = CHECKSUM_PARTIAL; memcpy(syn_data->cb, syn->cb, sizeof(syn->cb)); if (space) { - int copied = copy_from_iter(skb_put(syn_data, space), space, - &fo->data->msg_iter); - if (unlikely(!copied)) { + space = min_t(size_t, space, pfrag->size - pfrag->offset); + space = tcp_wmem_schedule(sk, space); + } + if (space) { + space = copy_page_from_iter(pfrag->page, pfrag->offset, + space, &fo->data->msg_iter); + if (unlikely(!space)) { tcp_skb_tsorted_anchor_cleanup(syn_data); kfree_skb(syn_data); goto fallback; } - if (copied != space) { - skb_trim(syn_data, copied); - space = copied; - } + skb_fill_page_desc(syn_data, 0, pfrag->page, + pfrag->offset, space); + page_ref_inc(pfrag->page); + pfrag->offset += space; + skb_len_add(syn_data, space); skb_zcopy_set(syn_data, fo->uarg, NULL); } /* No more data pending in inet_wait_for_connect() */ @@ -3468,7 +4221,7 @@ static int tcp_send_syn_data(struct sock *sk, struct sk_buff *syn) err = tcp_transmit_skb(sk, syn_data, 1, sk->sk_allocation); - syn->skb_mstamp_ns = syn_data->skb_mstamp_ns; + skb_set_delivery_time(syn, syn_data->skb_mstamp_ns, SKB_CLOCK_MONOTONIC); /* Now full SYN+DATA was cloned and sent (or not), * remove the SYN from the original skb (syn_data) @@ -3509,6 +4262,53 @@ int tcp_connect(struct sock *sk) tcp_call_bpf(sk, BPF_SOCK_OPS_TCP_CONNECT_CB, 0, NULL); +#if defined(CONFIG_TCP_MD5SIG) && defined(CONFIG_TCP_AO) + /* Has to be checked late, after setting daddr/saddr/ops. + * Return error if the peer has both a md5 and a tcp-ao key + * configured as this is ambiguous. + */ + if (unlikely(rcu_dereference_protected(tp->md5sig_info, + lockdep_sock_is_held(sk)))) { + bool needs_ao = !!tp->af_specific->ao_lookup(sk, sk, -1, -1); + bool needs_md5 = !!tp->af_specific->md5_lookup(sk, sk); + struct tcp_ao_info *ao_info; + + ao_info = rcu_dereference_check(tp->ao_info, + lockdep_sock_is_held(sk)); + if (ao_info) { + /* This is an extra check: tcp_ao_required() in + * tcp_v{4,6}_parse_md5_keys() should prevent adding + * md5 keys on ao_required socket. + */ + needs_ao |= ao_info->ao_required; + WARN_ON_ONCE(ao_info->ao_required && needs_md5); + } + if (needs_md5 && needs_ao) + return -EKEYREJECTED; + + /* If we have a matching md5 key and no matching tcp-ao key + * then free up ao_info if allocated. + */ + if (needs_md5) { + tcp_ao_destroy_sock(sk, false); + } else if (needs_ao) { + tcp_clear_md5_list(sk); + kfree(rcu_replace_pointer(tp->md5sig_info, NULL, + lockdep_sock_is_held(sk))); + } + } +#endif +#ifdef CONFIG_TCP_AO + if (unlikely(rcu_dereference_protected(tp->ao_info, + lockdep_sock_is_held(sk)))) { + /* Don't allow connecting if ao is configured but no + * matching key is found. + */ + if (!tp->af_specific->ao_lookup(sk, sk, -1, -1)) + return -EKEYREJECTED; + } +#endif + if (inet_csk(sk)->icsk_af_ops->rebuild_header(sk)) return -EHOSTUNREACH; /* Routing failure or similar. */ @@ -3519,13 +4319,16 @@ int tcp_connect(struct sock *sk) return 0; } - buff = sk_stream_alloc_skb(sk, 0, sk->sk_allocation, true); + buff = tcp_stream_alloc_skb(sk, sk->sk_allocation, true); if (unlikely(!buff)) return -ENOBUFS; - tcp_init_nondata_skb(buff, tp->write_seq++, TCPHDR_SYN); + /* SYN eats a sequence byte, write_seq updated by + * tcp_connect_queue_skb(). + */ + tcp_init_nondata_skb(buff, sk, tp->write_seq, TCPHDR_SYN); tcp_mstamp_refresh(tp); - tp->retrans_stamp = tcp_time_stamp(tp); + tp->retrans_stamp = tcp_time_stamp_ts(tp); tcp_connect_queue_skb(sk, buff); tcp_ecn_send_syn(sk, buff); tcp_rbtree_insert(&sk->tcp_rtx_queue, buff); @@ -3539,22 +4342,29 @@ int tcp_connect(struct sock *sk) /* We change tp->snd_nxt after the tcp_transmit_skb() call * in order to make this packet get counted in tcpOutSegs. */ - tp->snd_nxt = tp->write_seq; + WRITE_ONCE(tp->snd_nxt, tp->write_seq); tp->pushed_seq = tp->write_seq; buff = tcp_send_head(sk); if (unlikely(buff)) { - tp->snd_nxt = TCP_SKB_CB(buff)->seq; + WRITE_ONCE(tp->snd_nxt, TCP_SKB_CB(buff)->seq); tp->pushed_seq = TCP_SKB_CB(buff)->seq; } TCP_INC_STATS(sock_net(sk), TCP_MIB_ACTIVEOPENS); /* Timer for repeating the SYN until an answer. */ - inet_csk_reset_xmit_timer(sk, ICSK_TIME_RETRANS, - inet_csk(sk)->icsk_rto, TCP_RTO_MAX); + tcp_reset_xmit_timer(sk, ICSK_TIME_RETRANS, + inet_csk(sk)->icsk_rto, false); return 0; } EXPORT_SYMBOL(tcp_connect); +u32 tcp_delack_max(const struct sock *sk) +{ + u32 delack_from_rto_min = max(tcp_rto_min(sk), 2) - 1; + + return min(READ_ONCE(inet_csk(sk)->icsk_delack_max), delack_from_rto_min); +} + /* Send out a delayed ack, the caller does the policy checking * to see if we should even be here. See tcp_input.c:tcp_ack_snd_check() * for details. @@ -3569,7 +4379,7 @@ void tcp_send_delayed_ack(struct sock *sk) const struct tcp_sock *tp = tcp_sk(sk); int max_ato = HZ / 2; - if (icsk->icsk_ack.pingpong || + if (inet_csk_in_pingpong_mode(sk) || (icsk->icsk_ack.pending & ICSK_ACK_PUSHED)) max_ato = TCP_DELACK_MAX; @@ -3590,30 +4400,29 @@ void tcp_send_delayed_ack(struct sock *sk) ato = min(ato, max_ato); } + ato = min_t(u32, ato, tcp_delack_max(sk)); + /* Stay within the limit we were given */ timeout = jiffies + ato; /* Use new timeout only if there wasn't a older one earlier. */ if (icsk->icsk_ack.pending & ICSK_ACK_TIMER) { - /* If delack timer was blocked or is about to expire, - * send ACK now. - */ - if (icsk->icsk_ack.blocked || - time_before_eq(icsk->icsk_ack.timeout, jiffies + (ato >> 2))) { + /* If delack timer is about to expire, send ACK now. */ + if (time_before_eq(icsk_delack_timeout(icsk), jiffies + (ato >> 2))) { tcp_send_ack(sk); return; } - if (!time_before(timeout, icsk->icsk_ack.timeout)) - timeout = icsk->icsk_ack.timeout; + if (!time_before(timeout, icsk_delack_timeout(icsk))) + timeout = icsk_delack_timeout(icsk); } - icsk->icsk_ack.pending |= ICSK_ACK_SCHED | ICSK_ACK_TIMER; - icsk->icsk_ack.timeout = timeout; + smp_store_release(&icsk->icsk_ack.pending, + icsk->icsk_ack.pending | ICSK_ACK_SCHED | ICSK_ACK_TIMER); sk_reset_timer(sk, &icsk->icsk_delack_timer, timeout); } /* This routine sends an ack and also updates the window. */ -void __tcp_send_ack(struct sock *sk, u32 rcv_nxt) +void __tcp_send_ack(struct sock *sk, u32 rcv_nxt, u16 flags) { struct sk_buff *buff; @@ -3628,16 +4437,22 @@ void __tcp_send_ack(struct sock *sk, u32 rcv_nxt) buff = alloc_skb(MAX_TCP_HEADER, sk_gfp_mask(sk, GFP_ATOMIC | __GFP_NOWARN)); if (unlikely(!buff)) { + struct inet_connection_sock *icsk = inet_csk(sk); + unsigned long delay; + + delay = TCP_DELACK_MAX << icsk->icsk_ack.retry; + if (delay < tcp_rto_max(sk)) + icsk->icsk_ack.retry++; inet_csk_schedule_ack(sk); - inet_csk(sk)->icsk_ack.ato = TCP_ATO_MIN; - inet_csk_reset_xmit_timer(sk, ICSK_TIME_DACK, - TCP_DELACK_MAX, TCP_RTO_MAX); + icsk->icsk_ack.ato = TCP_ATO_MIN; + tcp_reset_xmit_timer(sk, ICSK_TIME_DACK, delay, false); return; } /* Reserve space for headers and prepare control bits. */ skb_reserve(buff, MAX_TCP_HEADER); - tcp_init_nondata_skb(buff, tcp_acceptable_seq(sk), TCPHDR_ACK); + tcp_init_nondata_skb(buff, sk, + tcp_acceptable_seq(sk), TCPHDR_ACK | flags); /* We do not want pure acks influencing TCP Small Queues or fq/pacing * too much. @@ -3652,7 +4467,7 @@ EXPORT_SYMBOL_GPL(__tcp_send_ack); void tcp_send_ack(struct sock *sk) { - __tcp_send_ack(sk, tcp_sk(sk)->rcv_nxt); + __tcp_send_ack(sk, tcp_sk(sk)->rcv_nxt, 0); } /* This routine sends a packet with an out of date sequence @@ -3683,7 +4498,7 @@ static int tcp_xmit_probe_skb(struct sock *sk, int urgent, int mib) * end to send an ack. Don't queue or clone SKB, just * send it. */ - tcp_init_nondata_skb(skb, tp->snd_una - !urgent, TCPHDR_ACK); + tcp_init_nondata_skb(skb, sk, tp->snd_una - !urgent, TCPHDR_ACK); NET_INC_STATS(sock_net(sk), mib); return tcp_transmit_skb(sk, skb, 0, (__force gfp_t)0); } @@ -3757,23 +4572,26 @@ void tcp_send_probe0(struct sock *sk) if (tp->packets_out || tcp_write_queue_empty(sk)) { /* Cancel probe timer, if it is not required. */ - icsk->icsk_probes_out = 0; + WRITE_ONCE(icsk->icsk_probes_out, 0); icsk->icsk_backoff = 0; + icsk->icsk_probes_tstamp = 0; return; } - icsk->icsk_probes_out++; + WRITE_ONCE(icsk->icsk_probes_out, icsk->icsk_probes_out + 1); if (err <= 0) { - if (icsk->icsk_backoff < net->ipv4.sysctl_tcp_retries2) + if (icsk->icsk_backoff < READ_ONCE(net->ipv4.sysctl_tcp_retries2)) icsk->icsk_backoff++; - timeout = tcp_probe0_when(sk, TCP_RTO_MAX); + timeout = tcp_probe0_when(sk, tcp_rto_max(sk)); } else { /* If packet was not sent due to local congestion, * Let senders fight for local resources conservatively. */ timeout = TCP_RESOURCE_PROBE_INTERVAL; } - tcp_reset_xmit_timer(sk, ICSK_TIME_PROBE0, timeout, TCP_RTO_MAX, NULL); + + timeout = tcp_clamp_probe0_to_user_timeout(sk, timeout); + tcp_reset_xmit_timer(sk, ICSK_TIME_PROBE0, timeout, true); } int tcp_rtx_synack(const struct sock *sk, struct request_sock *req) @@ -3782,15 +4600,24 @@ int tcp_rtx_synack(const struct sock *sk, struct request_sock *req) struct flowi fl; int res; - tcp_rsk(req)->txhash = net_tx_rndhash(); - res = af_ops->send_synack(sk, NULL, &fl, req, NULL, TCP_SYNACK_NORMAL); + /* Paired with WRITE_ONCE() in sock_setsockopt() */ + if (READ_ONCE(sk->sk_txrehash) == SOCK_TXREHASH_ENABLED) + WRITE_ONCE(tcp_rsk(req)->txhash, net_tx_rndhash()); + res = af_ops->send_synack(sk, NULL, &fl, req, NULL, TCP_SYNACK_NORMAL, + NULL); if (!res) { - __TCP_INC_STATS(sock_net(sk), TCP_MIB_RETRANSSEGS); - __NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSYNRETRANS); - if (unlikely(tcp_passive_fastopen(sk))) - tcp_sk(sk)->total_retrans++; + TCP_INC_STATS(sock_net(sk), TCP_MIB_RETRANSSEGS); + NET_INC_STATS(sock_net(sk), LINUX_MIB_TCPSYNRETRANS); + if (unlikely(tcp_passive_fastopen(sk))) { + /* sk has const attribute because listeners are lockless. + * However in this case, we are dealing with a passive fastopen + * socket thus we can change total_retrans value. + */ + tcp_sk_rw(sk)->total_retrans++; + } trace_tcp_retransmit_synack(sk, req); + WRITE_ONCE(req->num_retrans, req->num_retrans + 1); } return res; } -EXPORT_SYMBOL(tcp_rtx_synack); +EXPORT_IPV6_MOD(tcp_rtx_synack); |