d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

TCP: Misc cleanups of tcp_subr.c

Browse Source 
In going through all the TCP stacks I have found we have a few little bugs and niggles that need to
be cleaned up in tcp_subr.c including the following:

a) Set tcp_restoral_thresh to 450 (45%) not 550. This is a better proven value in my testing.
b) Lets track when we try to do pacing but fail via a counter for connections that do pace.
c) If a switch away from the default stack occurs and it fails we need to make sure the time
   scale is in the right mode (just in case the other stack changed it but then failed).
d) Use the TP_RXTCUR() macro when starting the TT_REXMT timer.
e) When we end a default flow lets log that in BBlogs as well as cleanup any t_acktime (disable).
f) When we respond with a RST lets make sure to update the log_end_status properly.
g) When starting a new pcb lets assure that all LRO features are off.
h) When discarding a connection lets make sure that any t_in_pkt's that might be there are freed properly.

Reviewed by: tuexen
Sponsored by: Netflix Inc
Differential Revision:https://reviews.freebsd.org/D39501
This commit is contained in:
Randall Stewart 2023-04-13 09:29:05 -04:00
parent d8e36cd2b1
commit 25685b7537
1 changed files with 54 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

57
sys/netinet/tcp_subr.c
View File

@ -168,10 +168,10 @@ SYSCTL_INT(_net_inet_tcp_sack_attack, OID_AUTO, move_thresh,
CTLFLAG_RW,
&tcp_sack_to_move_thresh, 600,
"Percentage of sack moves we must see above (10.1 percent is 101)");
int32_t tcp_restoral_thresh = 650; /* 65 % (sack:2:ack -5%) */
int32_t tcp_restoral_thresh = 450; /* 45 % (sack:2:ack -25%) (mv:ratio -15%) **/
SYSCTL_INT(_net_inet_tcp_sack_attack, OID_AUTO, restore_thresh,
CTLFLAG_RW,
&tcp_restoral_thresh, 550,
&tcp_restoral_thresh, 450,
"Percentage of sack to ack percentage we must see below to restore(10.1 percent is 101)");
int32_t tcp_sad_decay_val = 800;
SYSCTL_INT(_net_inet_tcp_sack_attack, OID_AUTO, decay_per,
@ -290,6 +290,7 @@ SYSCTL_INT(_net_inet_tcp, OID_AUTO, ts_offset_per_conn, CTLFLAG_VNET | CTLFLAG_R
/* How many connections are pacing */
static volatile uint32_t number_of_tcp_connections_pacing = 0;
static uint32_t shadow_num_connections = 0;
static counter_u64_t tcp_pacing_failures;
static int tcp_pacing_limit = 10000;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, pacing_limit, CTLFLAG_RW,
@ -299,6 +300,9 @@ SYSCTL_INT(_net_inet_tcp, OID_AUTO, pacing_limit, CTLFLAG_RW,
SYSCTL_UINT(_net_inet_tcp, OID_AUTO, pacing_count, CTLFLAG_RD,
&shadow_num_connections, 0, "Number of TCP connections being paced");
SYSCTL_COUNTER_U64(_net_inet_tcp, OID_AUTO, pacing_failures, CTLFLAG_RD,
&tcp_pacing_failures, "Number of times we failed to enable pacing to avoid exceeding the limit");
static int tcp_log_debug = 0;
SYSCTL_INT(_net_inet_tcp, OID_AUTO, log_debug, CTLFLAG_RW,
&tcp_log_debug, 0, "Log errors caused by incoming TCP segments");
@ -378,6 +382,7 @@ static struct inpcb *tcp_mtudisc(struct inpcb *, int);
static struct inpcb *tcp_drop_syn_sent(struct inpcb *, int);
static char * tcp_log_addr(struct in_conninfo *inc, struct tcphdr *th,
const void *ip4hdr, const void *ip6hdr);
static void tcp_default_switch_failed(struct tcpcb *tp);
static ipproto_ctlinput_t tcp_ctlinput;
static udp_tun_icmp_t tcp_ctlinput_viaudp;
@ -389,6 +394,7 @@ static struct tcp_function_block tcp_def_funcblk = {
.tfb_tcp_handoff_ok = tcp_default_handoff_ok,
.tfb_tcp_fb_init = tcp_default_fb_init,
.tfb_tcp_fb_fini = tcp_default_fb_fini,
.tfb_switch_failed = tcp_default_switch_failed,
};
static int tcp_fb_cnt = 0;
@ -1107,7 +1113,7 @@ tcp_default_fb_init(struct tcpcb *tp, void **ptr)
(int32_t)sbavail(&so->so_snd))
tcp_setpersist(tp);
else
tcp_timer_activate(tp, TT_REXMT, tp->t_rxtcur);
tcp_timer_activate(tp, TT_REXMT, TP_RXTCUR(tp));
}
/* All non-embryonic sessions get a keepalive timer. */
@ -1140,6 +1146,12 @@ tcp_default_fb_fini(struct tcpcb *tp, int tcb_is_purged)
{
INP_WLOCK_ASSERT(tptoinpcb(tp));
#ifdef TCP_BLACKBOX
tcp_log_flowend(tp);
#endif
tp->t_acktime = 0;
return;
}
/*
@ -1561,6 +1573,7 @@ tcp_init(void *arg __unused)
tcp_comp_total = counter_u64_alloc(M_WAITOK);
tcp_uncomp_total = counter_u64_alloc(M_WAITOK);
tcp_bad_csums = counter_u64_alloc(M_WAITOK);
tcp_pacing_failures = counter_u64_alloc(M_WAITOK);
#ifdef TCPPCAP
tcp_pcap_init();
#endif
@ -1932,6 +1945,8 @@ tcp_respond(struct tcpcb *tp, void *ipgen, struct tcphdr *th, struct mbuf *m,
m_freem(m->m_next);
m->m_next = NULL;
m->m_data = (caddr_t)ipgen;
/* clear any receive flags for proper bpf timestamping */
m->m_flags &= ~(M_TSTMP | M_TSTMP_LRO);
/* m_len is set later */
#ifdef INET6
if (isipv6) {
@ -2068,6 +2083,10 @@ tcp_respond(struct tcpcb *tp, void *ipgen, struct tcphdr *th, struct mbuf *m,
nth->th_ack = htonl(ack);
nth->th_off = (sizeof (struct tcphdr) + optlen) >> 2;
tcp_set_flags(nth, flags);
if (tp && (flags & TH_RST)) {
/* Log the reset */
tcp_log_end_status(tp, TCP_EI_STATUS_SERVER_RST);
}
if (tp != NULL)
nth->th_win = htons((u_short) (win >> tp->rcv_scale));
else
@ -2249,6 +2268,9 @@ tcp_newtcpcb(struct inpcb *inp)
#endif /* INET6 */
V_tcp_mssdflt;
/* All mbuf queue/ack compress flags should be off */
tcp_lro_features_off(tptoinpcb(tp));
callout_init_rw(&tp->t_callout, &inp->inp_lock, CALLOUT_RETURNUNLOCKED);
for (int i = 0; i < TT_N; i++)
tp->t_timers[i] = SBT_MAX;
@ -2403,6 +2425,17 @@ tcp_discardcb(struct tcpcb *tp)
#ifdef TCP_BLACKBOX
tcp_log_tcpcbfini(tp);
#endif
if (tp->t_in_pkt) {
struct mbuf *m, *n;
m = tp->t_in_pkt;
tp->t_in_pkt = tp->t_tail_pkt = NULL;
while (m) {
n = m->m_nextpkt;
m_freem(m);
m = n;
}
}
TCPSTATES_DEC(tp->t_state);
if (tp->t_fb->tfb_tcp_fb_fini)
(*tp->t_fb->tfb_tcp_fb_fini)(tp, 1);
@ -3984,6 +4017,7 @@ tcp_can_enable_pacing(void)
shadow_num_connections = number_of_tcp_connections_pacing;
return (1);
} else {
counter_u64_add(tcp_pacing_failures, 1);
return (0);
}
}
@ -4009,6 +4043,23 @@ tcp_decrement_paced_conn(void)
}
}
static void
tcp_default_switch_failed(struct tcpcb *tp)
{
/*
* If a switch fails we only need to
* care about two things:
* a) The inp_flags2
* and
* b) The timer granularity.
* Timeouts, at least for now, don't use the
* old callout system in the other stacks so
* those are hopefully safe.
*/
tcp_lro_features_off(tptoinpcb(tp));
tcp_change_time_units(tp, TCP_TMR_GRANULARITY_TICKS);
}
#ifdef TCP_ACCOUNTING
int
tcp_do_ack_accounting(struct tcpcb *tp, struct tcphdr *th, struct tcpopt *to, uint32_t tiwin, int mss)