NetBSD/usr.bin/netstat/inet6.c

1538 lines
41 KiB
C
Raw Normal View History

/* $NetBSD: inet6.c,v 1.69 2016/12/17 09:12:22 mlelstv Exp $ */
1999-07-01 22:40:35 +04:00
/* BSDI inet.c,v 2.3 1995/10/24 02:19:29 prb Exp */
/*
* Copyright (C) 1995, 1996, 1997, 1998, and 1999 WIDE Project.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the project nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE PROJECT AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE PROJECT OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
1999-07-01 22:40:35 +04:00
/*
* Copyright (c) 1983, 1988, 1993
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. Neither the name of the University nor the names of its contributors
1999-07-01 22:40:35 +04:00
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
#ifndef lint
#if 0
static char sccsid[] = "@(#)inet.c 8.4 (Berkeley) 4/20/94";
#else
__RCSID("$NetBSD: inet6.c,v 1.69 2016/12/17 09:12:22 mlelstv Exp $");
1999-07-01 22:40:35 +04:00
#endif
#endif /* not lint */
#define _CALLOUT_PRIVATE
1999-07-01 22:40:35 +04:00
#include <sys/param.h>
#include <sys/socket.h>
#include <sys/socketvar.h>
#include <sys/ioctl.h>
1999-07-01 22:40:35 +04:00
#include <sys/mbuf.h>
#include <sys/protosw.h>
#include <sys/sysctl.h>
1999-07-01 22:40:35 +04:00
#include <net/route.h>
#include <net/if.h>
#include <netinet/in.h>
#include <netinet/ip6.h>
#include <netinet/icmp6.h>
#include <netinet/in_systm.h>
#ifndef TCP6
#include <netinet/ip.h>
#include <netinet/ip_var.h>
#endif
#include <netinet6/ip6_var.h>
1999-07-01 22:40:35 +04:00
#include <netinet6/in6_pcb.h>
#include <netinet6/in6_var.h>
1999-07-01 22:40:35 +04:00
#ifdef TCP6
#include <netinet/tcp6.h>
#include <netinet/tcp6_seq.h>
1999-07-01 22:40:35 +04:00
#define TCP6STATES
#include <netinet/tcp6_fsm.h>
1999-07-01 22:40:35 +04:00
#define TCP6TIMERS
#include <netinet/tcp6_timer.h>
#include <netinet/tcp6_var.h>
#include <netinet/tcp6_debug.h>
1999-07-01 22:40:35 +04:00
#else
#define TCP6T_NTIMERS TCPT_NTIMERS
#define tcp6timers tcptimers
#define tcp6states tcpstates
#define TCP6_NSTATES TCP_NSTATES
#define tcp6cb tcpcb
1999-07-01 22:40:35 +04:00
#include <netinet/tcp.h>
#include <netinet/tcpip.h>
#include <netinet/tcp_seq.h>
#include <netinet/tcp_fsm.h>
extern const char * const tcpstates[];
extern const char * const tcptimers[];
1999-07-01 22:40:35 +04:00
#include <netinet/tcp_timer.h>
#include <netinet/tcp_var.h>
#include <netinet/tcp_debug.h>
#endif /*TCP6*/
#include <netinet6/udp6.h>
#include <netinet6/udp6_var.h>
#include <netinet6/pim6_var.h>
2001-10-18 13:26:16 +04:00
#include <netinet6/raw_ip6.h>
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
#include <netinet/tcp_vtw.h>
1999-07-01 22:40:35 +04:00
#include <arpa/inet.h>
#if 0
#include "gethostbyname2.h"
#endif
#include <netdb.h>
#include <err.h>
#include <errno.h>
#include <kvm.h>
1999-07-01 22:40:35 +04:00
#include <stdio.h>
#include <stdlib.h>
1999-07-01 22:40:35 +04:00
#include <string.h>
#include <unistd.h>
2015-02-07 22:36:06 +03:00
#include <util.h>
1999-07-01 22:40:35 +04:00
#include "netstat.h"
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
#include "vtw.h"
#include "prog_ops.h"
1999-07-01 22:40:35 +04:00
#ifdef INET6
struct in6pcb in6pcb;
#ifdef TCP6
struct tcp6cb tcp6cb;
#else
struct tcpcb tcpcb;
#endif
struct socket sockb;
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
char *inet6name(const struct in6_addr *);
void inet6print(const struct in6_addr *, int, const char *);
void print_vtw_v6(const vtw_t *);
1999-07-01 22:40:35 +04:00
/*
* Print a summary of connections related to an Internet
* protocol. For TCP, also give state of connection.
* Listening processes (aflag) are suppressed unless the
* -a (all) flag is specified.
*/
static int width;
static int compact;
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
/* VTW-related variables. */
static struct timeval now;
static void
ip6protoprhdr(void)
{
printf("Active Internet6 connections");
if (aflag)
printf(" (including servers)");
putchar('\n');
if (Aflag) {
printf("%-8.8s ", "PCB");
width = 18;
}
printf(
Vflag ? "%-5.5s %-6.6s %-6.6s %*.*s %*.*s %-13.13s Expires\n"
: "%-5.5s %-6.6s %-6.6s %*.*s %*.*s %s\n",
"Proto", "Recv-Q", "Send-Q",
-width, width, "Local Address",
-width, width, "Foreign Address", "(state)");
}
static void
ip6protopr0(intptr_t ppcb, u_long rcv_sb_cc, u_long snd_sb_cc,
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
const struct in6_addr *laddr, u_int16_t lport,
const struct in6_addr *faddr, u_int16_t fport,
short t_state, const char *name, const struct timeval *expires)
{
static const char *shorttcpstates[] = {
"CLOSED", "LISTEN", "SYNSEN", "SYSRCV",
"ESTABL", "CLWAIT", "FWAIT1", "CLOSNG",
"LASTAK", "FWAIT2", "TMWAIT",
};
int istcp;
istcp = strcmp(name, "tcp6") == 0;
if (Aflag)
printf("%8" PRIxPTR " ", ppcb);
printf("%-5.5s %6ld %6ld%s", name, rcv_sb_cc, snd_sb_cc,
compact ? "" : " ");
inet6print(laddr, (int)lport, name);
inet6print(faddr, (int)fport, name);
if (istcp) {
#ifdef TCP6
if (t_state < 0 || t_state >= TCP6_NSTATES)
printf(" %d", t_state);
else
printf(" %s", tcp6states[t_state]);
#else
if (t_state < 0 || t_state >= TCP_NSTATES)
printf(" %d", t_state);
else
printf(" %s", compact ? shorttcpstates[t_state] :
tcpstates[t_state]);
#endif
}
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
if (Vflag && expires != NULL) {
if (expires->tv_sec == 0 && expires->tv_usec == -1)
printf(" reclaimed");
else {
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
struct timeval delta;
timersub(expires, &now, &delta);
printf(" %.3fms",
delta.tv_sec * 1000.0 + delta.tv_usec / 1000.0);
}
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
}
putchar('\n');
}
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
static void
dbg_printf(const char *fmt, ...)
{
return;
}
void
print_vtw_v6(const vtw_t *vtw)
{
const vtw_v6_t *v6 = (const vtw_v6_t *)vtw;
struct timeval delta;
char buf[2][128];
static const struct timeval zero = {.tv_sec = 0, .tv_usec = 0};
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
inet_ntop(AF_INET6, &v6->laddr, buf[0], sizeof(buf[0]));
inet_ntop(AF_INET6, &v6->faddr, buf[1], sizeof(buf[1]));
timersub(&vtw->expire, &now, &delta);
if (vtw->expire.tv_sec == 0 && vtw->expire.tv_usec == -1) {
dbg_printf("%15.15s:%d %15.15s:%d reclaimed\n"
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
,buf[0], ntohs(v6->lport)
,buf[1], ntohs(v6->fport));
if (!(Vflag && vflag))
return;
} else if (vtw->expire.tv_sec == 0)
return;
else if (timercmp(&delta, &zero, <) && !(Vflag && vflag)) {
dbg_printf("%15.15s:%d %15.15s:%d expired\n"
,buf[0], ntohs(v6->lport)
,buf[1], ntohs(v6->fport));
return;
} else {
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
dbg_printf("%15.15s:%d %15.15s:%d expires in %.3fms\n"
,buf[0], ntohs(v6->lport)
,buf[1], ntohs(v6->fport)
,delta.tv_sec * 1000.0 + delta.tv_usec / 1000.0);
}
ip6protopr0(0, 0, 0,
&v6->laddr, v6->lport,
&v6->faddr, v6->fport,
TCPS_TIME_WAIT, "tcp6", &vtw->expire);
}
static struct kinfo_pcb *
getpcblist_kmem(u_long off, const char *name, size_t *len) {
2003-09-04 13:23:35 +04:00
struct inpcbtable table;
struct inpcb_hdr *next, *prev;
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
int istcp = strcmp(name, "tcp6") == 0;
struct kinfo_pcb *pcblist;
size_t size = 100, i;
struct sockaddr_in6 sin6;
struct inpcbqueue *head;
if (off == 0) {
*len = 0;
return NULL;
}
kread(off, (char *)&table, sizeof (table));
head = &table.inpt_queue;
next = TAILQ_FIRST(head);
prev = TAILQ_END(head);
if ((pcblist = malloc(size * sizeof(*pcblist))) == NULL)
err(1, "malloc");
i = 0;
while (next != TAILQ_END(head)) {
2003-09-04 13:23:35 +04:00
kread((u_long)next, (char *)&in6pcb, sizeof in6pcb);
next = TAILQ_NEXT(&in6pcb, in6p_queue);
2003-09-04 13:23:35 +04:00
prev = next;
if (in6pcb.in6p_af != AF_INET6)
continue;
kread((u_long)in6pcb.in6p_socket, (char *)&sockb,
sizeof (sockb));
1999-07-01 22:40:35 +04:00
if (istcp) {
#ifdef TCP6
kread((u_long)in6pcb.in6p_ppcb,
(char *)&tcp6cb, sizeof (tcp6cb));
#else
kread((u_long)in6pcb.in6p_ppcb,
(char *)&tcpcb, sizeof (tcpcb));
#endif
}
pcblist[i].ki_ppcbaddr =
istcp ? (uintptr_t) in6pcb.in6p_ppcb : (uintptr_t) prev;
pcblist[i].ki_rcvq = (uint64_t)sockb.so_rcv.sb_cc;
pcblist[i].ki_sndq = (uint64_t)sockb.so_snd.sb_cc;
sin6.sin6_addr = in6pcb.in6p_laddr;
sin6.sin6_port = in6pcb.in6p_lport;
memcpy(&pcblist[i].ki_s, &sin6, sizeof(sin6));
sin6.sin6_addr = in6pcb.in6p_faddr;
sin6.sin6_port = in6pcb.in6p_fport;
memcpy(&pcblist[i].ki_d, &sin6, sizeof(sin6));
pcblist[i].ki_tstate = tcpcb.t_state;
if (i++ == size) {
size += 100;
struct kinfo_pcb *n = realloc(pcblist,
size * sizeof(*pcblist));
if (n == NULL)
err(1, "realloc");
pcblist = n;
}
}
*len = i;
return pcblist;
}
void
ip6protopr(u_long off, const char *name)
{
struct kinfo_pcb *pcblist;
size_t i, len;
static int first = 1;
compact = 0;
if (Aflag) {
if (!numeric_addr)
width = 18;
else {
width = 21;
compact = 1;
}
} else
width = 22;
if (use_sysctl)
pcblist = getpcblist_sysctl(name, &len);
else
pcblist = getpcblist_kmem(off, name, &len);
for (i = 0; i < len; i++) {
struct sockaddr_in6 src, dst;
memcpy(&src, &pcblist[i].ki_s, sizeof(src));
memcpy(&dst, &pcblist[i].ki_d, sizeof(dst));
if (!aflag && IN6_IS_ADDR_UNSPECIFIED(&dst.sin6_addr))
continue;
1999-07-01 22:40:35 +04:00
if (first) {
ip6protoprhdr();
1999-07-01 22:40:35 +04:00
first = 0;
}
ip6protopr0((intptr_t) pcblist[i].ki_ppcbaddr,
pcblist[i].ki_rcvq, pcblist[i].ki_sndq,
&src.sin6_addr, src.sin6_port,
&dst.sin6_addr, dst.sin6_port,
pcblist[i].ki_tstate, name, NULL);
1999-07-01 22:40:35 +04:00
}
free(pcblist);
if (strcmp(name, "tcp6") == 0) {
struct timeval t;
timebase(&t);
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
gettimeofday(&now, NULL);
timersub(&now, &t, &now);
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
show_vtw_v6(print_vtw_v6);
}
1999-07-01 22:40:35 +04:00
}
#ifdef TCP6
/*
* Dump TCP6 statistics structure.
*/
void
tcp6_stats(u_long off, const char *name)
1999-07-01 22:40:35 +04:00
{
struct tcp6stat tcp6stat;
if (use_sysctl) {
size_t size = sizeof(tcp6stat);
if (sysctlbyname("net.inet6.tcp6.stats", &tcp6stat, &size,
NULL, 0) == -1)
return;
} else {
warnx("%s stats not available via KVM.", name);
return;
}
1999-07-01 22:40:35 +04:00
printf ("%s:\n", name);
#define p(f, m) if (tcp6stat.f || sflag <= 1) \
printf(m, tcp6stat.f, plural(tcp6stat.f))
#define p2(f1, f2, m) if (tcp6stat.f1 || tcp6stat.f2 || sflag <= 1) \
printf(m, tcp6stat.f1, plural(tcp6stat.f1), tcp6stat.f2, plural(tcp6stat.f2))
#define p3(f, m) if (tcp6stat.f || sflag <= 1) \
printf(m, tcp6stat.f, plurales(tcp6stat.f))
p(tcp6s_sndtotal, "\t%ld packet%s sent\n");
p2(tcp6s_sndpack,tcp6s_sndbyte,
"\t\t%ld data packet%s (%ld byte%s)\n");
p2(tcp6s_sndrexmitpack, tcp6s_sndrexmitbyte,
"\t\t%ld data packet%s (%ld byte%s) retransmitted\n");
p2(tcp6s_sndacks, tcp6s_delack,
"\t\t%ld ack-only packet%s (%ld packet%s delayed)\n");
p(tcp6s_sndurg, "\t\t%ld URG only packet%s\n");
p(tcp6s_sndprobe, "\t\t%ld window probe packet%s\n");
p(tcp6s_sndwinup, "\t\t%ld window update packet%s\n");
p(tcp6s_sndctrl, "\t\t%ld control packet%s\n");
p(tcp6s_rcvtotal, "\t%ld packet%s received\n");
p2(tcp6s_rcvackpack, tcp6s_rcvackbyte, "\t\t%ld ack%s (for %ld byte%s)\n");
p(tcp6s_rcvdupack, "\t\t%ld duplicate ack%s\n");
p(tcp6s_rcvacktoomuch, "\t\t%ld ack%s for unsent data\n");
p2(tcp6s_rcvpack, tcp6s_rcvbyte,
"\t\t%ld packet%s (%ld byte%s) received in-sequence\n");
p2(tcp6s_rcvduppack, tcp6s_rcvdupbyte,
"\t\t%ld completely duplicate packet%s (%ld byte%s)\n");
p(tcp6s_pawsdrop, "\t\t%ld old duplicate packet%s\n");
p2(tcp6s_rcvpartduppack, tcp6s_rcvpartdupbyte,
"\t\t%ld packet%s with some dup. data (%ld byte%s duped)\n");
p2(tcp6s_rcvoopack, tcp6s_rcvoobyte,
"\t\t%ld out-of-order packet%s (%ld byte%s)\n");
p2(tcp6s_rcvpackafterwin, tcp6s_rcvbyteafterwin,
"\t\t%ld packet%s (%ld byte%s) of data after window\n");
p(tcp6s_rcvwinprobe, "\t\t%ld window probe%s\n");
p(tcp6s_rcvwinupd, "\t\t%ld window update packet%s\n");
p(tcp6s_rcvafterclose, "\t\t%ld packet%s received after close\n");
p(tcp6s_rcvbadsum, "\t\t%ld discarded for bad checksum%s\n");
p(tcp6s_rcvbadoff, "\t\t%ld discarded for bad header offset field%s\n");
p(tcp6s_rcvshort, "\t\t%ld discarded because packet%s too short\n");
p(tcp6s_connattempt, "\t%ld connection request%s\n");
p(tcp6s_accepts, "\t%ld connection accept%s\n");
p(tcp6s_badsyn, "\t%ld bad connection attempt%s\n");
p(tcp6s_connects, "\t%ld connection%s established (including accepts)\n");
p2(tcp6s_closed, tcp6s_drops,
"\t%ld connection%s closed (including %ld drop%s)\n");
p(tcp6s_conndrops, "\t%ld embryonic connection%s dropped\n");
p2(tcp6s_rttupdated, tcp6s_segstimed,
"\t%ld segment%s updated rtt (of %ld attempt%s)\n");
p(tcp6s_rexmttimeo, "\t%ld retransmit timeout%s\n");
p(tcp6s_timeoutdrop, "\t\t%ld connection%s dropped by rexmit timeout\n");
p(tcp6s_persisttimeo, "\t%ld persist timeout%s\n");
p(tcp6s_persistdrop, "\t%ld connection%s timed out in persist\n");
p(tcp6s_keeptimeo, "\t%ld keepalive timeout%s\n");
p(tcp6s_keepprobe, "\t\t%ld keepalive probe%s sent\n");
p(tcp6s_keepdrops, "\t\t%ld connection%s dropped by keepalive\n");
p(tcp6s_predack, "\t%ld correct ACK header prediction%s\n");
p(tcp6s_preddat, "\t%ld correct data packet header prediction%s\n");
p3(tcp6s_pcbcachemiss, "\t%ld PCB cache miss%s\n");
#undef p
#undef p2
#undef p3
}
#endif
/*
* Dump UDP6 statistics structure.
*/
void
udp6_stats(u_long off, const char *name)
1999-07-01 22:40:35 +04:00
{
2008-04-15 08:43:25 +04:00
uint64_t udp6stat[UDP6_NSTATS];
u_quad_t delivered;
1999-07-01 22:40:35 +04:00
if (use_sysctl) {
size_t size = sizeof(udp6stat);
2008-04-15 08:43:25 +04:00
if (sysctlbyname("net.inet6.udp6.stats", udp6stat, &size,
NULL, 0) == -1)
return;
} else {
warnx("%s stats not available via KVM.", name);
return;
}
1999-07-01 22:40:35 +04:00
printf("%s:\n", name);
2008-04-15 08:43:25 +04:00
#define p(f, m) if (udp6stat[f] || sflag <= 1) \
printf(m, (unsigned long long)udp6stat[f], plural(udp6stat[f]))
#define p1(f, m) if (udp6stat[f] || sflag <= 1) \
printf(m, (unsigned long long)udp6stat[f])
p(UDP6_STAT_IPACKETS, "\t%llu datagram%s received\n");
p1(UDP6_STAT_HDROPS, "\t%llu with incomplete header\n");
p1(UDP6_STAT_BADLEN, "\t%llu with bad data length field\n");
p1(UDP6_STAT_BADSUM, "\t%llu with bad checksum\n");
p1(UDP6_STAT_NOSUM, "\t%llu with no checksum\n");
p1(UDP6_STAT_NOPORT, "\t%llu dropped due to no socket\n");
p(UDP6_STAT_NOPORTMCAST,
"\t%llu multicast datagram%s dropped due to no socket\n");
2008-04-15 08:43:25 +04:00
p1(UDP6_STAT_FULLSOCK, "\t%llu dropped due to full socket buffers\n");
delivered = udp6stat[UDP6_STAT_IPACKETS] -
udp6stat[UDP6_STAT_HDROPS] -
udp6stat[UDP6_STAT_BADLEN] -
udp6stat[UDP6_STAT_BADSUM] -
udp6stat[UDP6_STAT_NOPORT] -
udp6stat[UDP6_STAT_NOPORTMCAST] -
udp6stat[UDP6_STAT_FULLSOCK];
1999-07-01 22:40:35 +04:00
if (delivered || sflag <= 1)
printf("\t%llu delivered\n", (unsigned long long)delivered);
2008-04-15 08:43:25 +04:00
p(UDP6_STAT_OPACKETS, "\t%llu datagram%s output\n");
1999-07-01 22:40:35 +04:00
#undef p
#undef p1
}
static const char *ip6nh[] = {
/*0*/ "hop by hop",
1999-07-01 22:40:35 +04:00
"ICMP",
"IGMP",
NULL,
1999-07-01 22:40:35 +04:00
"IP",
/*5*/ NULL,
1999-07-01 22:40:35 +04:00
"TCP",
NULL,
NULL,
NULL,
/*10*/ NULL, NULL, NULL, NULL, NULL,
/*15*/ NULL,
NULL,
1999-07-01 22:40:35 +04:00
"UDP",
NULL,
NULL,
/*20*/ NULL,
NULL,
1999-07-01 22:40:35 +04:00
"IDP",
NULL,
NULL,
/*25*/ NULL,
NULL,
NULL,
NULL,
2013-03-01 22:25:13 +04:00
NULL,
/*30*/ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
/*40*/ NULL,
1999-07-01 22:40:35 +04:00
"IP6",
NULL,
1999-07-01 22:40:35 +04:00
"routing",
"fragment",
/*45*/ NULL, NULL, NULL, NULL, NULL,
/*50*/ "ESP",
1999-07-01 22:40:35 +04:00
"AH",
NULL,
NULL,
NULL,
/*55*/ NULL,
NULL,
NULL,
1999-07-01 22:40:35 +04:00
"ICMP6",
"no next header",
/*60*/ "destination option",
NULL,
NULL,
NULL,
NULL,
/*65*/ NULL, NULL, NULL, NULL, NULL,
/*70*/ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
2013-03-01 22:25:13 +04:00
/*80*/ NULL,
NULL,
NULL,
NULL,
NULL,
NULL,
NULL,
NULL,
NULL,
"OSPF",
/*90*/ NULL, NULL, NULL, NULL, NULL,
/*95*/ NULL,
NULL,
1999-07-01 22:40:35 +04:00
"Ethernet",
NULL,
NULL,
/*100*/ NULL,
NULL,
NULL,
1999-07-01 22:40:35 +04:00
"PIM",
NULL,
/*105*/ NULL, NULL, NULL, NULL, NULL,
/*110*/ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
/*120*/ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
/*130*/ NULL,
NULL,
"SCTP",
NULL,
NULL,
/*135*/ NULL, NULL, NULL, NULL, NULL,
/*140*/ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
/*160*/ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
/*180*/ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
/*200*/ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
/*220*/ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
/*240*/ NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL, NULL,
NULL, NULL, NULL, NULL, NULL, NULL,
1999-07-01 22:40:35 +04:00
};
/*
* Dump IP6 statistics structure.
*/
void
ip6_stats(u_long off, const char *name)
1999-07-01 22:40:35 +04:00
{
uint64_t ip6stat[IP6_NSTATS];
1999-07-01 22:40:35 +04:00
int first, i;
struct protoent *ep;
const char *n;
1999-07-01 22:40:35 +04:00
if (use_sysctl) {
size_t size = sizeof(ip6stat);
1999-07-01 22:40:35 +04:00
if (sysctlbyname("net.inet6.ip6.stats", ip6stat, &size,
NULL, 0) == -1)
return;
} else {
warnx("%s stats not available via KVM.", name);
return;
}
1999-07-01 22:40:35 +04:00
printf("%s:\n", name);
#define p(f, m) if (ip6stat[f] || sflag <= 1) \
printf(m, (unsigned long long)ip6stat[f], plural(ip6stat[f]))
#define p1(f, m) if (ip6stat[f] || sflag <= 1) \
printf(m, (unsigned long long)ip6stat[f])
p(IP6_STAT_TOTAL, "\t%llu total packet%s received\n");
p1(IP6_STAT_TOOSMALL, "\t%llu with size smaller than minimum\n");
p1(IP6_STAT_TOOSHORT, "\t%llu with data size < data length\n");
p1(IP6_STAT_BADOPTIONS, "\t%llu with bad options\n");
p1(IP6_STAT_BADVERS, "\t%llu with incorrect version number\n");
p(IP6_STAT_FRAGMENTS, "\t%llu fragment%s received\n");
p(IP6_STAT_FRAGDROPPED,
"\t%llu fragment%s dropped (dup or out of space)\n");
p(IP6_STAT_FRAGTIMEOUT, "\t%llu fragment%s dropped after timeout\n");
p(IP6_STAT_FRAGOVERFLOW, "\t%llu fragment%s that exceeded limit\n");
p(IP6_STAT_REASSEMBLED, "\t%llu packet%s reassembled ok\n");
p(IP6_STAT_DELIVERED, "\t%llu packet%s for this host\n");
p(IP6_STAT_FORWARD, "\t%llu packet%s forwarded\n");
p(IP6_STAT_FASTFORWARD, "\t%llu packet%s fast forwarded\n");
p1(IP6_STAT_FASTFORWARDFLOWS, "\t%llu fast forward flows\n");
p(IP6_STAT_CANTFORWARD, "\t%llu packet%s not forwardable\n");
p(IP6_STAT_REDIRECTSENT, "\t%llu redirect%s sent\n");
p(IP6_STAT_LOCALOUT, "\t%llu packet%s sent from this host\n");
p(IP6_STAT_RAWOUT, "\t%llu packet%s sent with fabricated ip header\n");
p(IP6_STAT_ODROPPED,
"\t%llu output packet%s dropped due to no bufs, etc.\n");
p(IP6_STAT_NOROUTE, "\t%llu output packet%s discarded due to no route\n");
p(IP6_STAT_FRAGMENTED, "\t%llu output datagram%s fragmented\n");
p(IP6_STAT_OFRAGMENTS, "\t%llu fragment%s created\n");
p(IP6_STAT_CANTFRAG, "\t%llu datagram%s that can't be fragmented\n");
p(IP6_STAT_BADSCOPE, "\t%llu packet%s that violated scope rules\n");
p(IP6_STAT_NOTMEMBER, "\t%llu multicast packet%s which we don't join\n");
1999-07-01 22:40:35 +04:00
for (first = 1, i = 0; i < 256; i++)
if (ip6stat[IP6_STAT_NXTHIST + i] != 0) {
1999-07-01 22:40:35 +04:00
if (first) {
printf("\tInput packet histogram:\n");
1999-07-01 22:40:35 +04:00
first = 0;
}
n = NULL;
if (ip6nh[i])
n = ip6nh[i];
else if ((ep = getprotobynumber(i)) != NULL)
n = ep->p_name;
if (n)
printf("\t\t%s: %llu\n", n,
(unsigned long long)ip6stat[IP6_STAT_NXTHIST + i]);
else
printf("\t\t#%d: %llu\n", i,
(unsigned long long)ip6stat[IP6_STAT_NXTHIST + i]);
1999-07-01 22:40:35 +04:00
}
printf("\tMbuf statistics:\n");
p(IP6_STAT_M1, "\t\t%llu one mbuf%s\n");
1999-07-01 22:40:35 +04:00
for (first = 1, i = 0; i < 32; i++) {
char ifbuf[IFNAMSIZ];
if (ip6stat[IP6_STAT_M2M + i] != 0) {
1999-07-01 22:40:35 +04:00
if (first) {
printf("\t\ttwo or more mbuf:\n");
first = 0;
}
printf("\t\t\t%s = %llu\n",
1999-07-01 22:40:35 +04:00
if_indextoname(i, ifbuf),
(unsigned long long)ip6stat[IP6_STAT_M2M + i]);
1999-07-01 22:40:35 +04:00
}
}
p(IP6_STAT_MEXT1, "\t\t%llu one ext mbuf%s\n");
p(IP6_STAT_MEXT2M, "\t\t%llu two or more ext mbuf%s\n");
p(IP6_STAT_EXTHDRTOOLONG,
"\t%llu packet%s whose headers are not continuous\n");
p(IP6_STAT_NOGIF, "\t%llu tunneling packet%s that can't find gif\n");
p(IP6_STAT_TOOMANYHDR,
"\t%llu packet%s discarded due to too many headers\n");
/* for debugging source address selection */
#define PRINT_SCOPESTAT(s,i) do {\
switch(i) { /* XXX hardcoding in each case */\
case 1:\
p(s, "\t\t%llu node-local%s\n");\
break;\
case 2:\
p(s, "\t\t%llu link-local%s\n");\
break;\
case 5:\
p(s, "\t\t%llu site-local%s\n");\
break;\
case 14:\
p(s, "\t\t%llu global%s\n");\
break;\
default:\
printf("\t\t%llu addresses scope=%x\n",\
(unsigned long long)ip6stat[s], i);\
}\
} while(/*CONSTCOND*/0);
p(IP6_STAT_SOURCES_NONE,
"\t%llu failure%s of source address selection\n");
for (first = 1, i = 0; i < 16; i++) {
if (ip6stat[IP6_STAT_SOURCES_SAMEIF + i]) {
if (first) {
printf("\tsource addresses on an outgoing I/F\n");
first = 0;
}
PRINT_SCOPESTAT(IP6_STAT_SOURCES_SAMEIF + i, i);
}
}
for (first = 1, i = 0; i < 16; i++) {
if (ip6stat[IP6_STAT_SOURCES_OTHERIF + i]) {
if (first) {
printf("\tsource addresses on a non-outgoing I/F\n");
first = 0;
}
PRINT_SCOPESTAT(IP6_STAT_SOURCES_OTHERIF + i, i);
}
}
for (first = 1, i = 0; i < 16; i++) {
if (ip6stat[IP6_STAT_SOURCES_SAMESCOPE + i]) {
if (first) {
printf("\tsource addresses of same scope\n");
first = 0;
}
PRINT_SCOPESTAT(IP6_STAT_SOURCES_SAMESCOPE + i, i);
}
}
for (first = 1, i = 0; i < 16; i++) {
if (ip6stat[IP6_STAT_SOURCES_OTHERSCOPE + i]) {
if (first) {
printf("\tsource addresses of a different scope\n");
first = 0;
}
PRINT_SCOPESTAT(IP6_STAT_SOURCES_OTHERSCOPE + i, i);
}
}
for (first = 1, i = 0; i < 16; i++) {
if (ip6stat[IP6_STAT_SOURCES_DEPRECATED + i]) {
if (first) {
printf("\tdeprecated source addresses\n");
first = 0;
}
PRINT_SCOPESTAT(IP6_STAT_SOURCES_DEPRECATED + i, i);
}
}
2000-07-06 16:40:19 +04:00
p1(IP6_STAT_FORWARD_CACHEHIT, "\t%llu forward cache hit\n");
p1(IP6_STAT_FORWARD_CACHEMISS, "\t%llu forward cache miss\n");
1999-07-01 22:40:35 +04:00
#undef p
#undef p1
}
/*
* Dump IPv6 per-interface statistics based on RFC 2465.
*/
void
ip6_ifstats(const char *ifname)
{
struct in6_ifreq ifr;
int s;
#define p(f, m) if (ifr.ifr_ifru.ifru_stat.f || sflag <= 1) \
1999-12-16 03:58:17 +03:00
printf(m, (unsigned long long)ifr.ifr_ifru.ifru_stat.f, \
plural(ifr.ifr_ifru.ifru_stat.f))
#define p_5(f, m) if (ifr.ifr_ifru.ifru_stat.f || sflag <= 1) \
2000-07-06 16:40:19 +04:00
printf(m, (unsigned long long)ip6stat.f)
if ((s = socket(AF_INET6, SOCK_DGRAM, 0)) < 0) {
perror("Warning: socket(AF_INET6)");
return;
}
2001-04-06 09:10:28 +04:00
strncpy(ifr.ifr_name, ifname, sizeof(ifr.ifr_name));
printf("ip6 on %s:\n", ifname);
if (ioctl(s, SIOCGIFSTAT_IN6, (char *)&ifr) < 0) {
perror("Warning: ioctl(SIOCGIFSTAT_IN6)");
goto end;
}
p(ifs6_in_receive, "\t%llu total input datagram%s\n");
p(ifs6_in_hdrerr, "\t%llu datagram%s with invalid header received\n");
p(ifs6_in_toobig, "\t%llu datagram%s exceeded MTU received\n");
p(ifs6_in_noroute, "\t%llu datagram%s with no route received\n");
p(ifs6_in_addrerr, "\t%llu datagram%s with invalid dst received\n");
p(ifs6_in_truncated, "\t%llu truncated datagram%s received\n");
p(ifs6_in_protounknown, "\t%llu datagram%s with unknown proto received\n");
p(ifs6_in_discard, "\t%llu input datagram%s discarded\n");
p(ifs6_in_deliver,
"\t%llu datagram%s delivered to an upper layer protocol\n");
p(ifs6_out_forward, "\t%llu datagram%s forwarded to this interface\n");
p(ifs6_out_request,
"\t%llu datagram%s sent from an upper layer protocol\n");
p(ifs6_out_discard, "\t%llu total discarded output datagram%s\n");
p(ifs6_out_fragok, "\t%llu output datagram%s fragmented\n");
p(ifs6_out_fragfail, "\t%llu output datagram%s failed on fragment\n");
p(ifs6_out_fragcreat, "\t%llu output datagram%s succeeded on fragment\n");
p(ifs6_reass_reqd, "\t%llu incoming datagram%s fragmented\n");
p(ifs6_reass_ok, "\t%llu datagram%s reassembled\n");
p(ifs6_reass_fail, "\t%llu datagram%s failed on reassembling\n");
p(ifs6_in_mcast, "\t%llu multicast datagram%s received\n");
p(ifs6_out_mcast, "\t%llu multicast datagram%s sent\n");
end:
close(s);
#undef p
#undef p_5
}
static const char *icmp6names[] = {
1999-07-01 22:40:35 +04:00
"#0",
"unreach",
"packet too big",
"time exceed",
"parameter problem",
"#5",
"#6",
"#7",
"#8",
"#9",
"#10",
"#11",
"#12",
"#13",
"#14",
"#15",
"#16",
"#17",
"#18",
"#19",
"#20",
"#21",
"#22",
"#23",
"#24",
"#25",
"#26",
"#27",
"#28",
"#29",
"#30",
"#31",
"#32",
"#33",
"#34",
"#35",
"#36",
"#37",
"#38",
"#39",
"#40",
"#41",
"#42",
"#43",
"#44",
"#45",
"#46",
"#47",
"#48",
"#49",
"#50",
"#51",
"#52",
"#53",
"#54",
"#55",
"#56",
"#57",
"#58",
"#59",
"#60",
"#61",
"#62",
"#63",
"#64",
"#65",
"#66",
"#67",
"#68",
"#69",
"#70",
"#71",
"#72",
"#73",
"#74",
"#75",
"#76",
"#77",
"#78",
"#79",
"#80",
"#81",
"#82",
"#83",
"#84",
"#85",
"#86",
"#87",
"#88",
"#89",
"#80",
"#91",
"#92",
"#93",
"#94",
"#95",
"#96",
"#97",
"#98",
"#99",
"#100",
"#101",
"#102",
"#103",
"#104",
"#105",
"#106",
"#107",
"#108",
"#109",
"#110",
"#111",
"#112",
"#113",
"#114",
"#115",
"#116",
"#117",
"#118",
"#119",
"#120",
"#121",
"#122",
"#123",
"#124",
"#125",
"#126",
"#127",
"echo",
"echo reply",
"multicast listener query",
"multicast listener report",
"multicast listener done",
1999-07-01 22:40:35 +04:00
"router solicitation",
2001-06-13 06:50:25 +04:00
"router advertisement",
1999-07-01 22:40:35 +04:00
"neighbor solicitation",
2001-06-13 06:50:25 +04:00
"neighbor advertisement",
1999-07-01 22:40:35 +04:00
"redirect",
"router renumbering",
"node information request",
"node information reply",
"#141",
"#142",
"#143",
"#144",
"#145",
"#146",
"#147",
"#148",
"#149",
"#150",
"#151",
"#152",
"#153",
"#154",
"#155",
"#156",
"#157",
"#158",
"#159",
"#160",
"#161",
"#162",
"#163",
"#164",
"#165",
"#166",
"#167",
"#168",
"#169",
"#170",
"#171",
"#172",
"#173",
"#174",
"#175",
"#176",
"#177",
"#178",
"#179",
"#180",
"#181",
"#182",
"#183",
"#184",
"#185",
"#186",
"#187",
"#188",
"#189",
"#180",
"#191",
"#192",
"#193",
"#194",
"#195",
"#196",
"#197",
"#198",
"#199",
"#200",
"#201",
"#202",
"#203",
"#204",
"#205",
"#206",
"#207",
"#208",
"#209",
"#210",
"#211",
"#212",
"#213",
"#214",
"#215",
"#216",
"#217",
"#218",
"#219",
"#220",
"#221",
"#222",
"#223",
"#224",
"#225",
"#226",
"#227",
"#228",
"#229",
"#230",
"#231",
"#232",
"#233",
"#234",
"#235",
"#236",
"#237",
"#238",
"#239",
"#240",
"#241",
"#242",
"#243",
"#244",
"#245",
"#246",
"#247",
"#248",
"#249",
"#250",
"#251",
"#252",
"#253",
"#254",
"#255",
};
/*
* Dump ICMPv6 statistics.
1999-07-01 22:40:35 +04:00
*/
void
icmp6_stats(u_long off, const char *name)
1999-07-01 22:40:35 +04:00
{
uint64_t icmp6stat[ICMP6_NSTATS];
2011-05-04 04:55:19 +04:00
int i, first;
1999-07-01 22:40:35 +04:00
if (use_sysctl) {
size_t size = sizeof(icmp6stat);
if (sysctlbyname("net.inet6.icmp6.stats", icmp6stat, &size,
NULL, 0) == -1)
return;
} else {
warnx("%s stats not available via KVM.", name);
return;
}
1999-07-01 22:40:35 +04:00
printf("%s:\n", name);
#define p(f, m) if (icmp6stat[f] || sflag <= 1) \
printf(m, (unsigned long long)icmp6stat[f], plural(icmp6stat[f]))
#define p_oerr(f, m) if (icmp6stat[ICMP6_STAT_OUTERRHIST + f] || sflag <= 1) \
printf(m, (unsigned long long)icmp6stat[ICMP6_STAT_OUTERRHIST + f])
1999-07-01 22:40:35 +04:00
p(ICMP6_STAT_ERROR, "\t%llu call%s to icmp6_error\n");
p(ICMP6_STAT_CANTERROR,
2000-05-17 15:54:48 +04:00
"\t%llu error%s not generated because old message was icmp6 or so\n");
p(ICMP6_STAT_TOOFREQ,
2000-12-11 20:52:43 +03:00
"\t%llu error%s not generated because of rate limitation\n");
1999-07-01 22:40:35 +04:00
for (first = 1, i = 0; i < 256; i++)
if (icmp6stat[ICMP6_STAT_OUTHIST + i] != 0) {
1999-07-01 22:40:35 +04:00
if (first) {
printf("\tOutput packet histogram:\n");
1999-07-01 22:40:35 +04:00
first = 0;
}
printf("\t\t%s: %llu\n", icmp6names[i],
(unsigned long long)icmp6stat[ICMP6_STAT_OUTHIST + i]);
1999-07-01 22:40:35 +04:00
}
p(ICMP6_STAT_BADCODE, "\t%llu message%s with bad code fields\n");
p(ICMP6_STAT_TOOSHORT, "\t%llu message%s < minimum length\n");
p(ICMP6_STAT_CHECKSUM, "\t%llu bad checksum%s\n");
p(ICMP6_STAT_BADLEN, "\t%llu message%s with bad length\n");
1999-07-01 22:40:35 +04:00
for (first = 1, i = 0; i < ICMP6_MAXTYPE; i++)
if (icmp6stat[ICMP6_STAT_INHIST + i] != 0) {
1999-07-01 22:40:35 +04:00
if (first) {
printf("\tInput packet histogram:\n");
1999-07-01 22:40:35 +04:00
first = 0;
}
printf("\t\t%s: %llu\n", icmp6names[i],
(unsigned long long)icmp6stat[ICMP6_STAT_INHIST + i]);
1999-07-01 22:40:35 +04:00
}
2000-07-06 16:40:19 +04:00
printf("\tHistogram of error messages to be generated:\n");
p_oerr(ICMP6_ERRSTAT_DST_UNREACH_NOROUTE, "\t\t%llu no route\n");
p_oerr(ICMP6_ERRSTAT_DST_UNREACH_ADMIN, "\t\t%llu administratively prohibited\n");
p_oerr(ICMP6_ERRSTAT_DST_UNREACH_BEYONDSCOPE, "\t\t%llu beyond scope\n");
p_oerr(ICMP6_ERRSTAT_DST_UNREACH_ADDR, "\t\t%llu address unreachable\n");
p_oerr(ICMP6_ERRSTAT_DST_UNREACH_NOPORT, "\t\t%llu port unreachable\n");
p_oerr(ICMP6_ERRSTAT_PACKET_TOO_BIG, "\t\t%llu packet too big\n");
p_oerr(ICMP6_ERRSTAT_TIME_EXCEED_TRANSIT, "\t\t%llu time exceed transit\n");
p_oerr(ICMP6_ERRSTAT_TIME_EXCEED_REASSEMBLY, "\t\t%llu time exceed reassembly\n");
p_oerr(ICMP6_ERRSTAT_PARAMPROB_HEADER, "\t\t%llu erroneous header field\n");
p_oerr(ICMP6_ERRSTAT_PARAMPROB_NEXTHEADER, "\t\t%llu unrecognized next header\n");
p_oerr(ICMP6_ERRSTAT_PARAMPROB_OPTION, "\t\t%llu unrecognized option\n");
p_oerr(ICMP6_ERRSTAT_REDIRECT, "\t\t%llu redirect\n");
p_oerr(ICMP6_ERRSTAT_UNKNOWN, "\t\t%llu unknown\n");
p(ICMP6_STAT_REFLECT, "\t%llu message response%s generated\n");
p(ICMP6_STAT_ND_TOOMANYOPT, "\t%llu message%s with too many ND options\n");
p(ICMP6_STAT_ND_BADOPT, "\t%llu message%s with bad ND options\n");
p(ICMP6_STAT_BADNS, "\t%llu bad neighbor solicitation message%s\n");
p(ICMP6_STAT_BADNA, "\t%llu bad neighbor advertisement message%s\n");
p(ICMP6_STAT_BADRS, "\t%llu bad router solicitation message%s\n");
p(ICMP6_STAT_BADRA, "\t%llu bad router advertisement message%s\n");
p(ICMP6_STAT_DROPPED_RAROUTE, "\t%llu router advertisement route%s dropped\n");
p(ICMP6_STAT_BADREDIRECT, "\t%llu bad redirect message%s\n");
p(ICMP6_STAT_PMTUCHG, "\t%llu path MTU change%s\n");
#undef p
#undef p_oerr
}
/*
* Dump ICMPv6 per-interface statistics based on RFC 2466.
*/
void
icmp6_ifstats(const char *ifname)
{
struct in6_ifreq ifr;
int s;
#define p(f, m) if (ifr.ifr_ifru.ifru_icmp6stat.f || sflag <= 1) \
1999-12-16 03:58:17 +03:00
printf(m, (unsigned long long)ifr.ifr_ifru.ifru_icmp6stat.f, \
plural(ifr.ifr_ifru.ifru_icmp6stat.f))
if ((s = socket(AF_INET6, SOCK_DGRAM, 0)) < 0) {
perror("Warning: socket(AF_INET6)");
return;
}
2001-04-06 09:10:28 +04:00
strncpy(ifr.ifr_name, ifname, sizeof(ifr.ifr_name));
printf("icmp6 on %s:\n", ifname);
if (ioctl(s, SIOCGIFSTAT_ICMP6, (char *)&ifr) < 0) {
perror("Warning: ioctl(SIOCGIFSTAT_ICMP6)");
goto end;
}
p(ifs6_in_msg, "\t%llu total input message%s\n");
p(ifs6_in_error, "\t%llu total input error message%s\n");
p(ifs6_in_dstunreach, "\t%llu input destination unreachable error%s\n");
p(ifs6_in_adminprohib, "\t%llu input administratively prohibited error%s\n");
p(ifs6_in_timeexceed, "\t%llu input time exceeded error%s\n");
p(ifs6_in_paramprob, "\t%llu input parameter problem error%s\n");
p(ifs6_in_pkttoobig, "\t%llu input packet too big error%s\n");
p(ifs6_in_echo, "\t%llu input echo request%s\n");
p(ifs6_in_echoreply, "\t%llu input echo reply%s\n");
p(ifs6_in_routersolicit, "\t%llu input router solicitation%s\n");
p(ifs6_in_routeradvert, "\t%llu input router advertisement%s\n");
p(ifs6_in_neighborsolicit, "\t%llu input neighbor solicitation%s\n");
p(ifs6_in_neighboradvert, "\t%llu input neighbor advertisement%s\n");
p(ifs6_in_redirect, "\t%llu input redirect%s\n");
p(ifs6_in_mldquery, "\t%llu input MLD query%s\n");
p(ifs6_in_mldreport, "\t%llu input MLD report%s\n");
p(ifs6_in_mlddone, "\t%llu input MLD done%s\n");
p(ifs6_out_msg, "\t%llu total output message%s\n");
p(ifs6_out_error, "\t%llu total output error message%s\n");
p(ifs6_out_dstunreach, "\t%llu output destination unreachable error%s\n");
p(ifs6_out_adminprohib, "\t%llu output administratively prohibited error%s\n");
p(ifs6_out_timeexceed, "\t%llu output time exceeded error%s\n");
p(ifs6_out_paramprob, "\t%llu output parameter problem error%s\n");
p(ifs6_out_pkttoobig, "\t%llu output packet too big error%s\n");
p(ifs6_out_echo, "\t%llu output echo request%s\n");
p(ifs6_out_echoreply, "\t%llu output echo reply%s\n");
p(ifs6_out_routersolicit, "\t%llu output router solicitation%s\n");
p(ifs6_out_routeradvert, "\t%llu output router advertisement%s\n");
p(ifs6_out_neighborsolicit, "\t%llu output neighbor solicitation%s\n");
p(ifs6_out_neighboradvert, "\t%llu output neighbor advertisement%s\n");
p(ifs6_out_redirect, "\t%llu output redirect%s\n");
p(ifs6_out_mldquery, "\t%llu output MLD query%s\n");
p(ifs6_out_mldreport, "\t%llu output MLD report%s\n");
p(ifs6_out_mlddone, "\t%llu output MLD done%s\n");
end:
close(s);
1999-07-01 22:40:35 +04:00
#undef p
}
/*
* Dump PIM statistics structure.
*/
void
pim6_stats(u_long off, const char *name)
1999-07-01 22:40:35 +04:00
{
2008-04-15 09:40:15 +04:00
uint64_t pim6stat[PIM6_NSTATS];
1999-07-01 22:40:35 +04:00
if (use_sysctl) {
size_t size = sizeof(pim6stat);
2008-04-15 09:40:15 +04:00
if (sysctlbyname("net.inet6.pim6.stats", pim6stat, &size,
NULL, 0) == -1)
return;
} else {
warnx("%s stats not available via KVM.", name);
return;
}
1999-07-01 22:40:35 +04:00
printf("%s:\n", name);
2008-04-15 09:40:15 +04:00
#define p(f, m) if (pim6stat[f] || sflag <= 1) \
printf(m, (unsigned long long)pim6stat[f], plural(pim6stat[f]))
p(PIM6_STAT_RCV_TOTAL, "\t%llu message%s received\n");
p(PIM6_STAT_RCV_TOOSHORT, "\t%llu message%s received with too few bytes\n");
p(PIM6_STAT_RCV_BADSUM, "\t%llu message%s received with bad checksum\n");
p(PIM6_STAT_RCV_BADVERSION, "\t%llu message%s received with bad version\n");
p(PIM6_STAT_RCV_REGISTERS, "\t%llu register%s received\n");
p(PIM6_STAT_RCV_BADREGISTERS, "\t%llu bad register%s received\n");
p(PIM6_STAT_SND_REGISTERS, "\t%llu register%s sent\n");
1999-07-01 22:40:35 +04:00
#undef p
}
2001-10-18 13:26:16 +04:00
/*
* Dump raw ip6 statistics structure.
*/
void
rip6_stats(u_long off, const char *name)
2001-10-18 13:26:16 +04:00
{
2008-04-15 09:13:37 +04:00
uint64_t rip6stat[RIP6_NSTATS];
2001-10-18 13:26:16 +04:00
u_quad_t delivered;
if (use_sysctl) {
size_t size = sizeof(rip6stat);
2008-04-15 09:13:37 +04:00
if (sysctlbyname("net.inet6.raw6.stats", rip6stat, &size,
NULL, 0) == -1)
return;
} else {
warnx("%s stats not available via KVM.", name);
return;
}
2001-10-18 13:26:16 +04:00
printf("%s:\n", name);
2008-04-15 09:13:37 +04:00
#define p(f, m) if (rip6stat[f] || sflag <= 1) \
printf(m, (unsigned long long)rip6stat[f], plural(rip6stat[f]))
p(RIP6_STAT_IPACKETS, "\t%llu message%s received\n");
p(RIP6_STAT_ISUM, "\t%llu checksum calculation%s on inbound\n");
p(RIP6_STAT_BADSUM, "\t%llu message%s with bad checksum\n");
p(RIP6_STAT_NOSOCK, "\t%llu message%s dropped due to no socket\n");
p(RIP6_STAT_NOSOCKMCAST,
2001-10-18 13:26:16 +04:00
"\t%llu multicast message%s dropped due to no socket\n");
2008-04-15 09:13:37 +04:00
p(RIP6_STAT_FULLSOCK,
2001-10-18 13:26:16 +04:00
"\t%llu message%s dropped due to full socket buffers\n");
2008-04-15 09:13:37 +04:00
delivered = rip6stat[RIP6_STAT_IPACKETS] -
rip6stat[RIP6_STAT_BADSUM] -
rip6stat[RIP6_STAT_NOSOCK] -
rip6stat[RIP6_STAT_NOSOCKMCAST] -
rip6stat[RIP6_STAT_FULLSOCK];
2001-10-18 13:26:16 +04:00
if (delivered || sflag <= 1)
printf("\t%llu delivered\n", (unsigned long long)delivered);
2008-04-15 09:13:37 +04:00
p(RIP6_STAT_OPACKETS, "\t%llu datagram%s output\n");
2001-10-18 13:26:16 +04:00
#undef p
}
1999-07-01 22:40:35 +04:00
/*
* Pretty print an Internet address (net address + port).
* Take numeric_addr and numeric_port into consideration.
1999-07-01 22:40:35 +04:00
*/
void
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
inet6print(const struct in6_addr *in6, int port, const char *proto)
1999-07-01 22:40:35 +04:00
{
#define GETSERVBYPORT6(port, proto, ret)\
do {\
1999-07-01 22:40:35 +04:00
if (strcmp((proto), "tcp6") == 0)\
(ret) = getservbyport((int)(port), "tcp");\
else if (strcmp((proto), "udp6") == 0)\
(ret) = getservbyport((int)(port), "udp");\
else\
(ret) = getservbyport((int)(port), (proto));\
} while (0)
1999-07-01 22:40:35 +04:00
struct servent *sp = 0;
char line[80], *cp;
int lwidth;
1999-07-01 22:40:35 +04:00
lwidth = Aflag ? 12 : 16;
if (vflag && lwidth < (int)strlen(inet6name(in6)))
lwidth = strlen(inet6name(in6));
snprintf(line, sizeof(line), "%.*s.", lwidth, inet6name(in6));
2002-06-09 06:44:55 +04:00
cp = strchr(line, '\0');
if (!numeric_port && port)
1999-07-01 22:40:35 +04:00
GETSERVBYPORT6(port, proto, sp);
if (sp || port == 0)
2001-04-06 09:10:28 +04:00
snprintf(cp, sizeof(line) - (cp - line),
"%s", sp ? sp->s_name : "*");
1999-07-01 22:40:35 +04:00
else
2001-04-06 09:10:28 +04:00
snprintf(cp, sizeof(line) - (cp - line),
"%d", ntohs((u_short)port));
lwidth = Aflag ? 18 : 22;
if (vflag && lwidth < (int)strlen(line))
lwidth = strlen(line);
printf(" %-*.*s", lwidth, lwidth, line);
1999-07-01 22:40:35 +04:00
}
/*
* Construct an Internet address representation.
* If the numeric_addr has been supplied, give
1999-07-01 22:40:35 +04:00
* numeric value, otherwise try for symbolic name.
*/
char *
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
inet6name(const struct in6_addr *in6p)
1999-07-01 22:40:35 +04:00
{
2011-05-04 04:55:19 +04:00
char *cp;
static char line[NI_MAXHOST];
1999-07-01 22:40:35 +04:00
struct hostent *hp;
static char domain[MAXHOSTNAMELEN + 1];
static int first = 1;
char hbuf[NI_MAXHOST];
struct sockaddr_in6 sin6;
const int niflag = NI_NUMERICHOST;
1999-07-01 22:40:35 +04:00
if (first && !numeric_addr) {
1999-07-01 22:40:35 +04:00
first = 0;
if (gethostname(domain, MAXHOSTNAMELEN) == 0 &&
2002-06-09 06:44:55 +04:00
(cp = strchr(domain, '.')))
2001-04-06 09:10:28 +04:00
(void) strlcpy(domain, cp + 1, sizeof(domain));
1999-07-01 22:40:35 +04:00
else
domain[0] = 0;
}
cp = 0;
if (!numeric_addr && !IN6_IS_ADDR_UNSPECIFIED(in6p)) {
Reduces the resources demanded by TCP sessions in TIME_WAIT-state using methods called Vestigial Time-Wait (VTW) and Maximum Segment Lifetime Truncation (MSLT). MSLT and VTW were contributed by Coyote Point Systems, Inc. Even after a TCP session enters the TIME_WAIT state, its corresponding socket and protocol control blocks (PCBs) stick around until the TCP Maximum Segment Lifetime (MSL) expires. On a host whose workload necessarily creates and closes down many TCP sockets, the sockets & PCBs for TCP sessions in TIME_WAIT state amount to many megabytes of dead weight in RAM. Maximum Segment Lifetimes Truncation (MSLT) assigns each TCP session to a class based on the nearness of the peer. Corresponding to each class is an MSL, and a session uses the MSL of its class. The classes are loopback (local host equals remote host), local (local host and remote host are on the same link/subnet), and remote (local host and remote host communicate via one or more gateways). Classes corresponding to nearer peers have lower MSLs by default: 2 seconds for loopback, 10 seconds for local, 60 seconds for remote. Loopback and local sessions expire more quickly when MSLT is used. Vestigial Time-Wait (VTW) replaces a TIME_WAIT session's PCB/socket dead weight with a compact representation of the session, called a "vestigial PCB". VTW data structures are designed to be very fast and memory-efficient: for fast insertion and lookup of vestigial PCBs, the PCBs are stored in a hash table that is designed to minimize the number of cacheline visits per lookup/insertion. The memory both for vestigial PCBs and for elements of the PCB hashtable come from fixed-size pools, and linked data structures exploit this to conserve memory by representing references with a narrow index/offset from the start of a pool instead of a pointer. When space for new vestigial PCBs runs out, VTW makes room by discarding old vestigial PCBs, oldest first. VTW cooperates with MSLT. It may help to think of VTW as a "FIN cache" by analogy to the SYN cache. A 2.8-GHz Pentium 4 running a test workload that creates TIME_WAIT sessions as fast as it can is approximately 17% idle when VTW is active versus 0% idle when VTW is inactive. It has 103 megabytes more free RAM when VTW is active (approximately 64k vestigial PCBs are created) than when it is inactive.
2011-05-03 22:28:44 +04:00
hp = gethostbyaddr((const char *)in6p, sizeof(*in6p), AF_INET6);
1999-07-01 22:40:35 +04:00
if (hp) {
2002-06-09 06:44:55 +04:00
if ((cp = strchr(hp->h_name, '.')) &&
1999-07-01 22:40:35 +04:00
!strcmp(cp + 1, domain))
*cp = 0;
cp = hp->h_name;
}
}
if (IN6_IS_ADDR_UNSPECIFIED(in6p))
strlcpy(line, "*", sizeof(line));
1999-07-01 22:40:35 +04:00
else if (cp)
strlcpy(line, cp, sizeof(line));
else {
memset(&sin6, 0, sizeof(sin6));
sin6.sin6_len = sizeof(sin6);
sin6.sin6_family = AF_INET6;
sin6.sin6_addr = *in6p;
inet6_getscopeid(&sin6, INET6_IS_ADDR_LINKLOCAL|
INET6_IS_ADDR_MC_LINKLOCAL);
if (getnameinfo((struct sockaddr *)&sin6, sin6.sin6_len,
hbuf, sizeof(hbuf), NULL, 0, niflag) != 0)
2001-04-06 09:10:28 +04:00
strlcpy(hbuf, "?", sizeof(hbuf));
strlcpy(line, hbuf, sizeof(line));
}
1999-07-01 22:40:35 +04:00
return (line);
}
/*
* Dump the contents of a TCP6 PCB.
*/
void
tcp6_dump(u_long off, const char *name, u_long pcbaddr)
1999-07-01 22:40:35 +04:00
{
callout_impl_t *ci;
int i, hardticks;
struct kinfo_pcb *pcblist;
#ifdef TCP6
#define mypcb tcp6cb
#else
#define mypcb tcpcb
#endif
size_t j, len;
if (use_sysctl)
pcblist = getpcblist_sysctl(name, &len);
else
pcblist = getpcblist_kmem(off, name, &len);
for (j = 0; j < len; j++)
if (pcblist[j].ki_ppcbaddr == pcbaddr)
break;
free(pcblist);
if (j == len)
errx(1, "0x%lx is not a valid pcb address", pcbaddr);
1999-07-01 22:40:35 +04:00
kread(pcbaddr, (char *)&mypcb, sizeof(mypcb));
hardticks = get_hardticks();
1999-07-01 22:40:35 +04:00
printf("TCP Protocol Control Block at 0x%08lx:\n\n", pcbaddr);
printf("Timers:\n");
for (i = 0; i < TCP6T_NTIMERS; i++) {
2015-02-07 22:36:06 +03:00
char buf[128];
ci = (callout_impl_t *)&tcpcb.t_timer[i];
2015-02-07 22:36:06 +03:00
snprintb(buf, sizeof(buf), CALLOUT_FMT, ci->c_flags);
printf("\t%s\t%s", tcptimers[i], buf);
if (ci->c_flags & CALLOUT_PENDING)
printf("\t%d\n", ci->c_time - hardticks);
else
printf("\n");
}
1999-07-01 22:40:35 +04:00
printf("\n\n");
if (mypcb.t_state < 0 || mypcb.t_state >= TCP6_NSTATES)
printf("State: %d", mypcb.t_state);
1999-07-01 22:40:35 +04:00
else
printf("State: %s", tcp6states[mypcb.t_state]);
printf(", flags 0x%x, in6pcb 0x%lx\n\n", mypcb.t_flags,
(u_long)mypcb.t_in6pcb);
1999-07-01 22:40:35 +04:00
printf("rxtshift %d, rxtcur %d, dupacks %d\n", mypcb.t_rxtshift,
mypcb.t_rxtcur, mypcb.t_dupacks);
#ifdef TCP6
printf("peermaxseg %u, maxseg %u, force %d\n\n", mypcb.t_peermaxseg,
mypcb.t_maxseg, mypcb.t_force);
#else
printf("peermss %u, ourmss %u, segsz %u, segqlen %u\n\n",
tcpcb.t_peermss, tcpcb.t_ourmss, tcpcb.t_segsz, tcpcb.t_segqlen);
#endif
1999-07-01 22:40:35 +04:00
printf("snd_una %u, snd_nxt %u, snd_up %u\n",
mypcb.snd_una, mypcb.snd_nxt, mypcb.snd_up);
printf("snd_wl1 %u, snd_wl2 %u, iss %u, snd_wnd %llu\n\n",
mypcb.snd_wl1, mypcb.snd_wl2, mypcb.iss,
(unsigned long long)mypcb.snd_wnd);
1999-07-01 22:40:35 +04:00
printf("rcv_wnd %llu, rcv_nxt %u, rcv_up %u, irs %u\n\n",
(unsigned long long)mypcb.rcv_wnd, mypcb.rcv_nxt,
mypcb.rcv_up, mypcb.irs);
1999-07-01 22:40:35 +04:00
printf("rcv_adv %u, snd_max %u, snd_cwnd %llu, snd_ssthresh %llu\n",
mypcb.rcv_adv, mypcb.snd_max, (unsigned long long)mypcb.snd_cwnd,
(unsigned long long)mypcb.snd_ssthresh);
1999-07-01 22:40:35 +04:00
#ifdef TCP6
printf("idle %d, rtt %d, " mypcb.t_idle, mypcb.t_rtt);
#else
printf("rcvtime %u, rtttime %u, ", tcpcb.t_rcvtime, tcpcb.t_rtttime);
#endif
printf("rtseq %u, srtt %d, rttvar %d, rttmin %d, "
"max_sndwnd %llu\n\n", mypcb.t_rtseq,
mypcb.t_srtt, mypcb.t_rttvar, mypcb.t_rttmin,
(unsigned long long)mypcb.max_sndwnd);
1999-07-01 22:40:35 +04:00
printf("oobflags %d, iobc %d, softerror %d\n\n", mypcb.t_oobflags,
mypcb.t_iobc, mypcb.t_softerror);
1999-07-01 22:40:35 +04:00
printf("snd_scale %d, rcv_scale %d, req_r_scale %d, req_s_scale %d\n",
mypcb.snd_scale, mypcb.rcv_scale, mypcb.request_r_scale,
mypcb.requested_s_scale);
1999-07-01 22:40:35 +04:00
printf("ts_recent %u, ts_regent_age %d, last_ack_sent %u\n",
mypcb.ts_recent, mypcb.ts_recent_age, mypcb.last_ack_sent);
1999-07-01 22:40:35 +04:00
}
#endif /*INET6*/