/*
- * Copyright (c) 2000-2011 Apple Inc. All rights reserved.
+ * Copyright (c) 2000-2012 Apple Inc. All rights reserved.
*
* @APPLE_OSREFERENCE_LICENSE_HEADER_START@
*
#include <sys/kdebug.h>
#include <mach/sdt.h>
+#include <netinet/lro_ext.h>
+
#define DBG_FNC_TCP_CLOSE NETDBG_CODE(DBG_NETTCP, ((5 << 8) | 2))
extern int tcp_lq_overflow;
"Default TCP Maximum Segment Size for IPv6");
#endif
+extern int tcp_do_autorcvbuf;
+
/*
* Minimum MSS we accept and use. This prevents DoS attacks where
* we are forced to a ridiculous low MSS like 20 and send hundreds
SYSCTL_INT(_net_inet_tcp, OID_AUTO, background_sockets, CTLFLAG_RD | CTLFLAG_LOCKED,
&tcp_cc_ledbat.num_sockets, 0, "Number of sockets using background transport");
+__private_extern__ int tcp_win_scale = 3;
+SYSCTL_INT(_net_inet_tcp, OID_AUTO, win_scale_factor, CTLFLAG_RW | CTLFLAG_LOCKED,
+ &tcp_win_scale, 0, "Window scaling factor");
+
static void tcp_cleartaocache(void);
static void tcp_notify(struct inpcb *, int);
static void tcp_cc_init(void);
struct zone *sack_hole_zone;
struct zone *tcp_reass_zone;
+struct zone *tcp_bwmeas_zone;
/* The array containing pointers to currently implemented TCP CC algorithms */
struct tcp_cc_algo* tcp_cc_algo_list[TCP_CC_ALGO_COUNT];
-extern unsigned int total_mb_cnt;
-extern unsigned int total_cl_cnt;
-extern int sbspace_factor;
-extern int tcp_sockthreshold;
extern int slowlink_wsize; /* window correction for slow links */
extern int path_mtu_discovery;
+extern u_int32_t tcp_autorcvbuf_max;
+extern u_int32_t tcp_autorcvbuf_inc_shift;
+static void tcp_sbrcv_grow_rwin(struct tcpcb *tp, struct sockbuf *sb);
+
+#define TCP_BWMEAS_BURST_MINSIZE 6
+#define TCP_BWMEAS_BURST_MAXSIZE 25
+
+static uint32_t bwmeas_elm_size;
/*
* Target size of TCP PCB hash tables. Must be a power of two.
static lck_attr_t *tcp_uptime_mtx_attr = NULL; /* mutex attributes */
static lck_grp_t *tcp_uptime_mtx_grp = NULL; /* mutex group definition */
static lck_grp_attr_t *tcp_uptime_mtx_grp_attr = NULL; /* mutex group attributes */
+int tcp_notsent_lowat_check(struct socket *so);
int get_inpcb_str_size(void)
zone_change(tcp_reass_zone, Z_CALLERACCT, FALSE);
zone_change(tcp_reass_zone, Z_EXPAND, TRUE);
+ bwmeas_elm_size = P2ROUNDUP(sizeof(struct bwmeas), sizeof(u_int64_t));
+ tcp_bwmeas_zone = zinit(bwmeas_elm_size, (100 * bwmeas_elm_size), 0, "tcp_bwmeas_zone");
+ if (tcp_bwmeas_zone == NULL) {
+ panic("%s: failed allocating tcp_bwmeas_zone", __func__);
+ /* NOTREACHED */
+ }
+ zone_change(tcp_bwmeas_zone, Z_CALLERACCT, FALSE);
+ zone_change(tcp_bwmeas_zone, Z_EXPAND, TRUE);
+
#if INET6
#define TCP_MINPROTOHDR (sizeof(struct ip6_hdr) + sizeof(struct tcphdr))
#else /* INET6 */
#define TCP_MINPROTOHDR (sizeof(struct tcpiphdr))
#endif /* INET6 */
- if (max_protohdr < TCP_MINPROTOHDR)
- max_protohdr = TCP_MINPROTOHDR;
- if (max_linkhdr + TCP_MINPROTOHDR > MHLEN)
+ if (max_protohdr < TCP_MINPROTOHDR) {
+ _max_protohdr = TCP_MINPROTOHDR;
+ _max_protohdr = max_protohdr; /* round it up */
+ }
+ if (max_linkhdr + max_protohdr > MHLEN)
panic("tcp_init");
#undef TCP_MINPROTOHDR
/* Initialize TCP congestion control algorithms list */
tcp_cc_init();
+
+ /* Initialize TCP LRO data structures */
+ tcp_lro_init();
}
/*
struct ip6_hdr *ip6;
int isipv6;
#endif /* INET6 */
- unsigned int outif;
+ struct ifnet *outif;
#if INET6
isipv6 = IP_VHL_V(((struct ip *)ipgen)->ip_vhl) == 6;
bcopy((caddr_t)ip6, mtod(m, caddr_t),
sizeof(struct ip6_hdr));
ip6 = mtod(m, struct ip6_hdr *);
- nth = (struct tcphdr *)(ip6 + 1);
+ nth = (struct tcphdr *)(void *)(ip6 + 1);
} else
#endif /* INET6 */
{
bcopy((caddr_t)ip, mtod(m, caddr_t), sizeof(struct ip));
ip = mtod(m, struct ip *);
- nth = (struct tcphdr *)(ip + 1);
+ nth = (struct tcphdr *)(void *)(ip + 1);
}
bcopy((caddr_t)th, (caddr_t)nth, sizeof(struct tcphdr));
flags = TH_ACK;
#define xchg(a,b,type) { type t; t=a; a=b; b=t; }
#if INET6
if (isipv6) {
+ /* Expect 32-bit aligned IP on strict-align platforms */
+ IP6_HDR_STRICT_ALIGNMENT_CHECK(ip6);
xchg(ip6->ip6_dst, ip6->ip6_src, struct in6_addr);
- nth = (struct tcphdr *)(ip6 + 1);
+ nth = (struct tcphdr *)(void *)(ip6 + 1);
} else
#endif /* INET6 */
{
+ /* Expect 32-bit aligned IP on strict-align platforms */
+ IP_HDR_STRICT_ALIGNMENT_CHECK(ip);
xchg(ip->ip_dst.s_addr, ip->ip_src.s_addr, n_long);
- nth = (struct tcphdr *)(ip + 1);
+ nth = (struct tcphdr *)(void *)(ip + 1);
}
if (th != nth) {
/*
}
#endif
- if (tp != NULL)
- set_packet_tclass(m, tp->t_inpcb->inp_socket, MBUF_TC_UNSPEC, isipv6);
+ if (tp != NULL) {
+ u_int32_t svc_flags = 0;
+ if (isipv6) {
+ svc_flags |= PKT_SCF_IPV6;
+ }
+ set_packet_service_class(m, tp->t_inpcb->inp_socket,
+ MBUF_SC_UNSPEC, svc_flags);
+
+ /* Embed flowhash and flow control flags */
+ m->m_pkthdr.m_flowhash = tp->t_inpcb->inp_flowhash;
+ m->m_pkthdr.m_fhflags |=
+ (PF_TAG_TCP | PF_TAG_FLOWHASH | PF_TAG_FLOWADV);
+ }
#if INET6
if (isipv6) {
- struct ip6_out_args ip6oa = { ifscope, nocell };
+ struct ip6_out_args ip6oa = { ifscope, { 0 },
+ IP6OAF_SELECT_SRCIF | IP6OAF_BOUND_SRCADDR };
+
+ if (ifscope != IFSCOPE_NONE)
+ ip6oa.ip6oa_flags |= IP6OAF_BOUND_IF;
+ if (nocell)
+ ip6oa.ip6oa_flags |= IP6OAF_NO_CELLULAR;
(void) ip6_output(m, NULL, ro6, IPV6_OUTARGS, NULL,
NULL, &ip6oa);
if (ro6 == &sro6) {
rtfree(ro6->ro_rt);
ro6->ro_rt = NULL;
- } else if ((outif = ro6->ro_rt->rt_ifp->if_index) !=
- tp->t_inpcb->in6p_last_outif) {
- tp->t_inpcb->in6p_last_outif = outif;
+ } else if ((outif = ro6->ro_rt->rt_ifp) !=
+ tp->t_inpcb->in6p_last_outifp) {
+ tp->t_inpcb->in6p_last_outifp = outif;
}
}
} else
#endif /* INET6 */
{
- struct ip_out_args ipoa = { ifscope, nocell };
+ struct ip_out_args ipoa = { ifscope, { 0 },
+ IPOAF_SELECT_SRCIF | IPOAF_BOUND_SRCADDR };
+
+ if (ifscope != IFSCOPE_NONE)
+ ipoa.ipoa_flags |= IPOAF_BOUND_IF;
+ if (nocell)
+ ipoa.ipoa_flags |= IPOAF_NO_CELLULAR;
if (ro != &sro) {
/* Copy the cached route and take an extra reference */
if (ro != &sro) {
if (sro.ro_rt != NULL &&
- (outif = sro.ro_rt->rt_ifp->if_index) !=
- tp->t_inpcb->inp_last_outif)
- tp->t_inpcb->inp_last_outif = outif;
+ (outif = sro.ro_rt->rt_ifp) !=
+ tp->t_inpcb->inp_last_outifp)
+ tp->t_inpcb->inp_last_outifp = outif;
/* Synchronize cached PCB route */
inp_route_copyin(tp->t_inpcb, &sro);
} else if (sro.ro_rt != NULL) {
calculate_tcp_clock();
if (so->cached_in_sock_layer == 0) {
- it = (struct inp_tp *)inp;
+ it = (struct inp_tp *)(void *)inp;
tp = &it->tcb;
}
else
- tp = (struct tcpcb *) inp->inp_saved_ppcb;
+ tp = (struct tcpcb *)(void *)inp->inp_saved_ppcb;
bzero((char *) tp, sizeof(struct tcpcb));
LIST_INIT(&tp->t_segq);
}
tp->snd_cwnd = TCP_MAXWIN << TCP_MAX_WINSHIFT;
- tp->snd_bwnd = TCP_MAXWIN << TCP_MAX_WINSHIFT;
tp->snd_ssthresh = TCP_MAXWIN << TCP_MAX_WINSHIFT;
tp->snd_ssthresh_prev = TCP_MAXWIN << TCP_MAX_WINSHIFT;
tp->t_rcvtime = tcp_now;
- tp->t_bw_rtttime = 0;
tp->tentry.timer_start = tcp_now;
tp->t_persist_timeout = tcp_max_persist_timeout;
tp->t_persist_stop = 0;
struct socket *so = tp->t_inpcb->inp_socket;
#if CONFIG_DTRACE
struct inpcb *inp = tp->t_inpcb;
-#endif /* CONFIG_DTRACE */
+#endif
if (TCPS_HAVERCVDSYN(tp->t_state)) {
DTRACE_TCP4(state__change, void, NULL, struct inpcb *, inp,
return (tcp_close(tp));
}
+void
+tcp_getrt_rtt(struct tcpcb *tp, struct rtentry *rt)
+{
+ u_int32_t rtt = rt->rt_rmx.rmx_rtt;
+ int isnetlocal = (tp->t_flags & TF_LOCAL);
+
+ if (rtt != 0) {
+ /*
+ * XXX the lock bit for RTT indicates that the value
+ * is also a minimum value; this is subject to time.
+ */
+ if (rt->rt_rmx.rmx_locks & RTV_RTT)
+ tp->t_rttmin = rtt / (RTM_RTTUNIT / TCP_RETRANSHZ);
+ else
+ tp->t_rttmin = isnetlocal ? tcp_TCPTV_MIN : TCPTV_REXMTMIN;
+ tp->t_srtt = rtt / (RTM_RTTUNIT / (TCP_RETRANSHZ * TCP_RTT_SCALE));
+ tcpstat.tcps_usedrtt++;
+ if (rt->rt_rmx.rmx_rttvar) {
+ tp->t_rttvar = rt->rt_rmx.rmx_rttvar /
+ (RTM_RTTUNIT / (TCP_RETRANSHZ * TCP_RTTVAR_SCALE));
+ tcpstat.tcps_usedrttvar++;
+ } else {
+ /* default variation is +- 1 rtt */
+ tp->t_rttvar =
+ tp->t_srtt * TCP_RTTVAR_SCALE / TCP_RTT_SCALE;
+ }
+ TCPT_RANGESET(tp->t_rxtcur,
+ ((tp->t_srtt >> 2) + tp->t_rttvar) >> 1,
+ tp->t_rttmin, TCPTV_REXMTMAX,
+ TCP_ADD_REXMTSLOP(tp));
+ }
+}
+
/*
* Close a TCP control block:
* discard all space held by the tcp
* point both flags should be cleared and we can proceed further
* with the cleanup.
*/
- if (tp->t_flags & (TF_CLOSING|TF_SENDINPROG)) {
+ if ((tp->t_flags & TF_CLOSING) ||
+ inp->inp_sndinprog_cnt > 0) {
tp->t_flags |= TF_CLOSING;
return (NULL);
}
if (rt == NULL)
goto no_valid_rt;
- sin6 = (struct sockaddr_in6 *)rt_key(rt);
+ sin6 = (struct sockaddr_in6 *)(void *)rt_key(rt);
if (IN6_IS_ADDR_UNSPECIFIED(&sin6->sin6_addr))
goto no_valid_rt;
}
else
#endif /* INET6 */
if (rt == NULL || !(rt->rt_flags & RTF_UP) ||
- ((struct sockaddr_in *)rt_key(rt))->sin_addr.s_addr ==
+ ((struct sockaddr_in *)(void *)rt_key(rt))->sin_addr.s_addr ==
INADDR_ANY || rt->generation_id != route_generation) {
if (tp->t_state >= TCPS_CLOSE_WAIT) {
DTRACE_TCP4(state__change, void, NULL, struct inpcb *, inp,
(void) tcp_freeq(tp);
tcp_free_sackholes(tp);
+ if (tp->t_bwmeas != NULL) {
+ tcp_bwmeas_free(tp);
+ }
/* Free the packet list */
if (tp->t_pktlist_head != NULL)
*/
sodisconnectwakeup(so);
+ /*
+ * Clean up any LRO state
+ */
+ if (tp->t_flagsext & TF_LRO_OFFLOADED) {
+ tcp_lro_remove_state(inp->inp_laddr, inp->inp_faddr,
+ inp->inp_lport,
+ inp->inp_fport);
+ tp->t_flagsext &= ~TF_LRO_OFFLOADED;
+ }
+
#if INET6
if (INP_CHECK_SOCKAF(so, AF_INET6))
in6_pcbdetach(inp);
#endif
}
+struct bwmeas*
+tcp_bwmeas_alloc(struct tcpcb *tp)
+{
+ struct bwmeas *elm;
+ elm = zalloc(tcp_bwmeas_zone);
+ if (elm == NULL)
+ return(elm);
+
+ bzero(elm, bwmeas_elm_size);
+ elm->bw_minsizepkts = TCP_BWMEAS_BURST_MINSIZE;
+ elm->bw_maxsizepkts = TCP_BWMEAS_BURST_MAXSIZE;
+ elm->bw_minsize = elm->bw_minsizepkts * tp->t_maxseg;
+ elm->bw_maxsize = elm->bw_maxsizepkts * tp->t_maxseg;
+ return(elm);
+}
+
+void
+tcp_bwmeas_free(struct tcpcb* tp)
+{
+ zfree(tcp_bwmeas_zone, tp->t_bwmeas);
+ tp->t_bwmeas = NULL;
+ tp->t_flagsext &= ~(TF_MEASURESNDBW);
+}
+
/*
* tcpcb_to_otcpcb copies specific bits of a tcpcb to a otcpcb format.
* The otcpcb data structure is passed to user space and must not change.
inpcb_to_compat(inp, &xt.xt_inp);
inp_ppcb = inp->inp_ppcb;
if (inp_ppcb != NULL) {
- tcpcb_to_otcpcb((struct tcpcb *)inp_ppcb,
+ tcpcb_to_otcpcb((struct tcpcb *)(void *)inp_ppcb,
&xt.xt_tp);
} else {
bzero((char *) &xt.xt_tp, sizeof xt.xt_tp);
tcp_pcblist_n, "S,xtcpcb_n", "List of active TCP connections");
+__private_extern__ void
+tcp_get_ports_used(unsigned int ifindex, uint8_t *bitfield)
+{
+ inpcb_get_ports_used(ifindex, bitfield, &tcbinfo);
+}
+
+__private_extern__ uint32_t
+tcp_count_opportunistic(unsigned int ifindex, u_int32_t flags)
+{
+ return inpcb_count_opportunistic(ifindex, &tcbinfo, flags);
+}
+
void
tcp_ctlinput(cmd, sa, vip)
int cmd;
{
tcp_seq icmp_tcp_seq;
struct ip *ip = vip;
- struct tcphdr *th;
struct in_addr faddr;
struct inpcb *inp;
struct tcpcb *tp;
-
- void (*notify)(struct inpcb *, int) = tcp_notify;
- struct icmp *icp;
+ void (*notify)(struct inpcb *, int) = tcp_notify;
- faddr = ((struct sockaddr_in *)sa)->sin_addr;
+ faddr = ((struct sockaddr_in *)(void *)sa)->sin_addr;
if (sa->sa_family != AF_INET || faddr.s_addr == INADDR_ANY)
return;
else if ((unsigned)cmd > PRC_NCMDS || inetctlerrmap[cmd] == 0)
return;
if (ip) {
- icp = (struct icmp *)((caddr_t)ip
- - offsetof(struct icmp, icmp_ip));
- th = (struct tcphdr *)((caddr_t)ip
- + (IP_VHL_HL(ip->ip_vhl) << 2));
- inp = in_pcblookup_hash(&tcbinfo, faddr, th->th_dport,
- ip->ip_src, th->th_sport, 0, NULL);
+ struct tcphdr th;
+ struct icmp *icp;
+
+ icp = (struct icmp *)(void *)
+ ((caddr_t)ip - offsetof(struct icmp, icmp_ip));
+ bcopy(((caddr_t)ip + (IP_VHL_HL(ip->ip_vhl) << 2)),
+ &th, sizeof (th));
+ inp = in_pcblookup_hash(&tcbinfo, faddr, th.th_dport,
+ ip->ip_src, th.th_sport, 0, NULL);
if (inp != NULL && inp->inp_socket != NULL) {
tcp_lock(inp->inp_socket, 1, 0);
if (in_pcb_checkstate(inp, WNT_RELEASE, 1) == WNT_STOPUSING) {
tcp_unlock(inp->inp_socket, 1, 0);
return;
}
- icmp_tcp_seq = htonl(th->th_seq);
+ icmp_tcp_seq = htonl(th.th_seq);
tp = intotcpcb(inp);
if (SEQ_GEQ(icmp_tcp_seq, tp->snd_una) &&
SEQ_LT(icmp_tcp_seq, tp->snd_max)) {
ro->ro_dst.sa_family = AF_INET;
ro->ro_dst.sa_len = sizeof(struct sockaddr_in);
- ((struct sockaddr_in *) &ro->ro_dst)->sin_addr =
+ ((struct sockaddr_in *)(void *)&ro->ro_dst)->sin_addr =
inp->inp_faddr;
/*
* input_ifscope is IFSCOPE_NONE).
*/
ifscope = (inp->inp_flags & INP_BOUND_IF) ?
- inp->inp_boundif : input_ifscope;
+ inp->inp_boundifp->if_index : input_ifscope;
if (rt != NULL)
RT_UNLOCK(rt);
* input_ifscope is IFSCOPE_NONE).
*/
ifscope = (inp->inp_flags & INP_BOUND_IF) ?
- inp->inp_boundif : input_ifscope;
+ inp->inp_boundifp->if_index : input_ifscope;
if (rt != NULL)
RT_UNLOCK(rt);
#if INET6
if ((inp->inp_vflag & INP_IPV6) != 0) {
ip6 = mtod(m, struct ip6_hdr *);
- th = (struct tcphdr *)(ip6 + 1);
+ th = (struct tcphdr *)(void *)(ip6 + 1);
m->m_pkthdr.len = m->m_len =
sizeof(struct ip6_hdr) + sizeof(struct tcphdr);
tcp_fillheaders(tp, ip6, th);
}
}
+/* Determine if we can grow the recieve socket buffer to avoid sending
+ * a zero window update to the peer. We allow even socket buffers that
+ * have fixed size (set by the application) to grow if the resource
+ * constraints are met. They will also be trimmed after the application
+ * reads data.
+ */
+static void
+tcp_sbrcv_grow_rwin(struct tcpcb *tp, struct sockbuf *sb) {
+ u_int32_t rcvbufinc = tp->t_maxseg << tcp_autorcvbuf_inc_shift;
+ if (tcp_do_autorcvbuf == 1 &&
+ tcp_cansbgrow(sb) &&
+ (tp->t_flags & TF_SLOWLINK) == 0 &&
+ (sb->sb_hiwat - sb->sb_cc) < rcvbufinc &&
+ (sb->sb_hiwat < tcp_autorcvbuf_max)) {
+ sbreserve(sb, (sb->sb_hiwat + rcvbufinc));
+ }
+}
+
int32_t
tcp_sbspace(struct tcpcb *tp)
{
struct sockbuf *sb = &tp->t_inpcb->inp_socket->so_rcv;
- int32_t space, newspace;
+ int32_t space;
+
+ tcp_sbrcv_grow_rwin(tp, sb);
space = ((int32_t) imin((sb->sb_hiwat - sb->sb_cc),
(sb->sb_mbmax - sb->sb_mbcnt)));
if (((tp->t_flags & TF_SLOWLINK) != 0) && slowlink_wsize > 0 )
return imin(space, slowlink_wsize);
- /*
- * Check for ressources constraints before over-ajusting the amount of space we can
- * advertise in the TCP window size updates.
- */
-
- if (sbspace_factor && (tp->t_inpcb->inp_pcbinfo->ipi_count < tcp_sockthreshold) &&
- (total_mb_cnt / 8) < (mbstat.m_clusters / sbspace_factor)) {
- if (space < (int32_t)(sb->sb_maxused - sb->sb_cc)) {/* make sure we don't constrain the window if we have enough ressources */
- space = (int32_t) imax((sb->sb_maxused - sb->sb_cc), tp->rcv_maxbyps);
- }
- newspace = (int32_t) imax(((int32_t)sb->sb_maxused - sb->sb_cc), (int32_t)tp->rcv_maxbyps);
-
- if (newspace > space)
- space = newspace;
- }
return space;
}
/*
return;
}
+/* Compute receive window scaling that we are going to request
+ * for this connection based on sb_hiwat. Try to leave some
+ * room to potentially increase the window size upto a maximum
+ * defined by the constant tcp_autorcvbuf_max.
+ */
+void
+tcp_set_max_rwinscale(struct tcpcb *tp, struct socket *so) {
+ u_int32_t maxsockbufsize;
+
+ tp->request_r_scale = max(tcp_win_scale, tp->request_r_scale);
+ maxsockbufsize = ((so->so_rcv.sb_flags & SB_USRSIZE) != 0) ?
+ so->so_rcv.sb_hiwat : tcp_autorcvbuf_max;
+
+ while (tp->request_r_scale < TCP_MAX_WINSHIFT &&
+ (TCP_MAXWIN << tp->request_r_scale) < maxsockbufsize)
+ tp->request_r_scale++;
+ tp->request_r_scale = min(tp->request_r_scale, TCP_MAX_WINSHIFT);
+
+}
+
+int
+tcp_notsent_lowat_check(struct socket *so) {
+ struct inpcb *inp = sotoinpcb(so);
+ struct tcpcb *tp = NULL;
+ int notsent = 0;
+ if (inp != NULL) {
+ tp = intotcpcb(inp);
+ }
+
+ notsent = so->so_snd.sb_cc -
+ (tp->snd_nxt - tp->snd_una);
+
+ /* When we send a FIN or SYN, not_sent can be negative.
+ * In that case also we need to send a write event to the
+ * process if it is waiting. In the FIN case, it will
+ * get an error from send because cantsendmore will be set.
+ */
+ if (notsent <= tp->t_notsent_lowat) {
+ return(1);
+ }
+
+ /* When Nagle's algorithm is not disabled, it is better
+ * to wakeup the client until there is atleast one
+ * maxseg of data to write.
+ */
+ if ((tp->t_flags & TF_NODELAY) == 0 &&
+ notsent > 0 && notsent < tp->t_maxseg) {
+ return(1);
+ }
+ return(0);
+}
+
+
/* DSEP Review Done pl-20051213-v02 @3253,@3391,@3400 */
projects / apple / xnu.git / blobdiff