1
0
mirror of https://git.FreeBSD.org/src.git synced 2024-12-22 11:17:19 +00:00
freebsd/usr.bin/netstat/route.c

1138 lines
25 KiB
C
Raw Normal View History

/*-
1994-05-27 12:33:43 +00:00
* 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.
* 4. Neither the name of the University 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 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.
*/
#if 0
#ifndef lint
1996-01-14 23:33:13 +00:00
static char sccsid[] = "From: @(#)route.c 8.6 (Berkeley) 4/28/95";
1994-05-27 12:33:43 +00:00
#endif /* not lint */
#endif
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
1994-05-27 12:33:43 +00:00
#include <sys/param.h>
#include <sys/protosw.h>
#include <sys/socket.h>
Restore netstat -M functionality for most statistics on core dumps. In general, when support was added to netstat for fetching data using sysctl, no provision was left for fetching equivalent data from a core dump, and in fact, netstat would _always_ fetch data from the live kernel using sysctl even when -M was specified resulting in the user believing they were getting data from coredumps when they actually weren't. Some specific changes: - Add a global 'live' variable that is true if netstat is running against the live kernel and false if -M has been specified. - Stop abusing the sysctl flag in the protocol tables to hold the protocol number. Instead, the protocol is now its own field in the tables, and it is passed as a separate parameter to the PCB and stat routines rather than overloading the KVM offset parameter. - Don't run PCB or stats functions who don't have a namelist offset if we are being run against a crash dump (!live). - For the inet and unix PCB routines, we generate the same buffer from KVM that the sysctl usually generates complete with the header and trailer. - Don't run bpf stats for !live (before it would just silently always run live). - kread() no longer trashes memory when opening the buffer if there is an error on open and the passed in buffer is smaller than _POSIX2_LINE_MAX. - The multicast routing code doesn't fallback to kvm on live kernels if the sysctl fails. Keeping this made the code rather hairy, and netstat is already tied to the kernel ABI anyway (even when using sysctl's since things like xinpcb contain an inpcb) so any kernels this is run against that have the multicast routing stuff should have the sysctls. - Don't try to dig around in the kernel linker in the netgraph PCB routine for core dumps. Other notes: - sctp's PCB routine only works on live kernels, it looked rather complicated to generate all the same stuff via KVM. Someone can always add it later if desired though. - Fix the ipsec removal bug where N_xxx for IPSEC stats weren't renumbered. - Use sysctlbyname() everywhere rather than hardcoded mib values. MFC after: 1 week Approved by: re (rwatson)
2007-07-16 17:15:55 +00:00
#include <sys/socketvar.h>
#include <sys/time.h>
1994-05-27 12:33:43 +00:00
#include <net/ethernet.h>
1994-05-27 12:33:43 +00:00
#include <net/if.h>
#include <net/if_var.h>
1994-05-27 12:33:43 +00:00
#include <net/if_dl.h>
#include <net/if_types.h>
#include <net/radix.h>
1994-05-27 12:33:43 +00:00
#include <net/route.h>
#include <netinet/in.h>
#include <netipx/ipx.h>
#include <netatalk/at.h>
#include <netgraph/ng_socket.h>
1994-05-27 12:33:43 +00:00
#include <sys/sysctl.h>
#include <arpa/inet.h>
#include <libutil.h>
1994-05-27 12:33:43 +00:00
#include <netdb.h>
#include <stdint.h>
1994-05-27 12:33:43 +00:00
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sysexits.h>
#include <unistd.h>
#include <err.h>
1994-05-27 12:33:43 +00:00
#include "netstat.h"
#define kget(p, d) (kread((u_long)(p), (char *)&(d), sizeof (d)))
1994-05-27 12:33:43 +00:00
/*
* Definitions for showing gateway flags.
*/
struct bits {
u_long b_mask;
1994-05-27 12:33:43 +00:00
char b_val;
} bits[] = {
{ RTF_UP, 'U' },
{ RTF_GATEWAY, 'G' },
{ RTF_HOST, 'H' },
{ RTF_REJECT, 'R' },
{ RTF_DYNAMIC, 'D' },
{ RTF_MODIFIED, 'M' },
{ RTF_DONE, 'd' }, /* Completed -- for routing messages only */
{ RTF_XRESOLVE, 'X' },
{ RTF_STATIC, 'S' },
{ RTF_PROTO1, '1' },
{ RTF_PROTO2, '2' },
{ RTF_PRCLONING,'c' },
{ RTF_PROTO3, '3' },
{ RTF_BLACKHOLE,'B' },
{ RTF_BROADCAST,'b' },
#ifdef RTF_LLINFO
{ RTF_LLINFO, 'L' },
#endif
#ifdef RTF_WASCLONED
{ RTF_WASCLONED,'W' },
#endif
#ifdef RTF_CLONING
{ RTF_CLONING, 'C' },
#endif
{ 0 , 0 }
1994-05-27 12:33:43 +00:00
};
typedef union {
long dummy; /* Helps align structure. */
1994-05-27 12:33:43 +00:00
struct sockaddr u_sa;
u_short u_data[128];
} sa_u;
static sa_u pt_u;
1994-05-27 12:33:43 +00:00
int do_rtent = 0;
struct rtentry rtentry;
struct radix_node rnode;
struct radix_mask rmask;
struct radix_node_head **rt_tables;
1994-05-27 12:33:43 +00:00
int NewTree = 0;
struct timespec uptime;
static struct sockaddr *kgetsa(struct sockaddr *);
static void size_cols(int ef, struct radix_node *rn);
static void size_cols_tree(struct radix_node *rn);
static void size_cols_rtentry(struct rtentry *rt);
static void p_tree(struct radix_node *);
static void p_rtnode(void);
static void ntreestuff(void);
static void np_rtentry(struct rt_msghdr *);
static void p_sockaddr(struct sockaddr *, struct sockaddr *, int, int);
static const char *fmt_sockaddr(struct sockaddr *sa, struct sockaddr *mask,
2008-09-01 15:04:38 +00:00
int flags);
static void p_flags(int, const char *);
static const char *fmt_flags(int f);
static void p_rtentry(struct rtentry *);
static void domask(char *, in_addr_t, u_long);
1994-05-27 12:33:43 +00:00
/*
* Print routing tables.
*/
void
2001-06-15 23:35:13 +00:00
routepr(u_long rtree)
1994-05-27 12:33:43 +00:00
{
struct radix_node_head **rnhp, *rnh, head;
2008-09-01 15:04:38 +00:00
size_t intsize;
int fam, fibnum, numfibs;
Add code to allow the system to handle multiple routing tables. This particular implementation is designed to be fully backwards compatible and to be MFC-able to 7.x (and 6.x) Currently the only protocol that can make use of the multiple tables is IPv4 Similar functionality exists in OpenBSD and Linux. From my notes: ----- One thing where FreeBSD has been falling behind, and which by chance I have some time to work on is "policy based routing", which allows different packet streams to be routed by more than just the destination address. Constraints: ------------ I want to make some form of this available in the 6.x tree (and by extension 7.x) , but FreeBSD in general needs it so I might as well do it in -current and back port the portions I need. One of the ways that this can be done is to have the ability to instantiate multiple kernel routing tables (which I will now refer to as "Forwarding Information Bases" or "FIBs" for political correctness reasons). Which FIB a particular packet uses to make the next hop decision can be decided by a number of mechanisms. The policies these mechanisms implement are the "Policies" referred to in "Policy based routing". One of the constraints I have if I try to back port this work to 6.x is that it must be implemented as a EXTENSION to the existing ABIs in 6.x so that third party applications do not need to be recompiled in timespan of the branch. This first version will not have some of the bells and whistles that will come with later versions. It will, for example, be limited to 16 tables in the first commit. Implementation method, Compatible version. (part 1) ------------------------------- For this reason I have implemented a "sufficient subset" of a multiple routing table solution in Perforce, and back-ported it to 6.x. (also in Perforce though not always caught up with what I have done in -current/P4). The subset allows a number of FIBs to be defined at compile time (8 is sufficient for my purposes in 6.x) and implements the changes needed to allow IPV4 to use them. I have not done the changes for ipv6 simply because I do not need it, and I do not have enough knowledge of ipv6 (e.g. neighbor discovery) needed to do it. Other protocol families are left untouched and should there be users with proprietary protocol families, they should continue to work and be oblivious to the existence of the extra FIBs. To understand how this is done, one must know that the current FIB code starts everything off with a single dimensional array of pointers to FIB head structures (One per protocol family), each of which in turn points to the trie of routes available to that family. The basic change in the ABI compatible version of the change is to extent that array to be a 2 dimensional array, so that instead of protocol family X looking at rt_tables[X] for the table it needs, it looks at rt_tables[Y][X] when for all protocol families except ipv4 Y is always 0. Code that is unaware of the change always just sees the first row of the table, which of course looks just like the one dimensional array that existed before. The entry points rtrequest(), rtalloc(), rtalloc1(), rtalloc_ign() are all maintained, but refer only to the first row of the array, so that existing callers in proprietary protocols can continue to do the "right thing". Some new entry points are added, for the exclusive use of ipv4 code called in_rtrequest(), in_rtalloc(), in_rtalloc1() and in_rtalloc_ign(), which have an extra argument which refers the code to the correct row. In addition, there are some new entry points (currently called rtalloc_fib() and friends) that check the Address family being looked up and call either rtalloc() (and friends) if the protocol is not IPv4 forcing the action to row 0 or to the appropriate row if it IS IPv4 (and that info is available). These are for calling from code that is not specific to any particular protocol. The way these are implemented would change in the non ABI preserving code to be added later. One feature of the first version of the code is that for ipv4, the interface routes show up automatically on all the FIBs, so that no matter what FIB you select you always have the basic direct attached hosts available to you. (rtinit() does this automatically). You CAN delete an interface route from one FIB should you want to but by default it's there. ARP information is also available in each FIB. It's assumed that the same machine would have the same MAC address, regardless of which FIB you are using to get to it. This brings us as to how the correct FIB is selected for an outgoing IPV4 packet. Firstly, all packets have a FIB associated with them. if nothing has been done to change it, it will be FIB 0. The FIB is changed in the following ways. Packets fall into one of a number of classes. 1/ locally generated packets, coming from a socket/PCB. Such packets select a FIB from a number associated with the socket/PCB. This in turn is inherited from the process, but can be changed by a socket option. The process in turn inherits it on fork. I have written a utility call setfib that acts a bit like nice.. setfib -3 ping target.example.com # will use fib 3 for ping. It is an obvious extension to make it a property of a jail but I have not done so. It can be achieved by combining the setfib and jail commands. 2/ packets received on an interface for forwarding. By default these packets would use table 0, (or possibly a number settable in a sysctl(not yet)). but prior to routing the firewall can inspect them (see below). (possibly in the future you may be able to associate a FIB with packets received on an interface.. An ifconfig arg, but not yet.) 3/ packets inspected by a packet classifier, which can arbitrarily associate a fib with it on a packet by packet basis. A fib assigned to a packet by a packet classifier (such as ipfw) would over-ride a fib associated by a more default source. (such as cases 1 or 2). 4/ a tcp listen socket associated with a fib will generate accept sockets that are associated with that same fib. 5/ Packets generated in response to some other packet (e.g. reset or icmp packets). These should use the FIB associated with the packet being reponded to. 6/ Packets generated during encapsulation. gif, tun and other tunnel interfaces will encapsulate using the FIB that was in effect withthe proces that set up the tunnel. thus setfib 1 ifconfig gif0 [tunnel instructions] will set the fib for the tunnel to use to be fib 1. Routing messages would be associated with their process, and thus select one FIB or another. messages from the kernel would be associated with the fib they refer to and would only be received by a routing socket associated with that fib. (not yet implemented) In addition Netstat has been edited to be able to cope with the fact that the array is now 2 dimensional. (It looks in system memory using libkvm (!)). Old versions of netstat see only the first FIB. In addition two sysctls are added to give: a) the number of FIBs compiled in (active) b) the default FIB of the calling process. Early testing experience: ------------------------- Basically our (IronPort's) appliance does this functionality already using ipfw fwd but that method has some drawbacks. For example, It can't fully simulate a routing table because it can't influence the socket's choice of local address when a connect() is done. Testing during the generating of these changes has been remarkably smooth so far. Multiple tables have co-existed with no notable side effects, and packets have been routes accordingly. ipfw has grown 2 new keywords: setfib N ip from anay to any count ip from any to any fib N In pf there seems to be a requirement to be able to give symbolic names to the fibs but I do not have that capacity. I am not sure if it is required. SCTP has interestingly enough built in support for this, called VRFs in Cisco parlance. it will be interesting to see how that handles it when it suddenly actually does something. Where to next: -------------------- After committing the ABI compatible version and MFCing it, I'd like to proceed in a forward direction in -current. this will result in some roto-tilling in the routing code. Firstly: the current code's idea of having a separate tree per protocol family, all of the same format, and pointed to by the 1 dimensional array is a bit silly. Especially when one considers that there is code that makes assumptions about every protocol having the same internal structures there. Some protocols don't WANT that sort of structure. (for example the whole idea of a netmask is foreign to appletalk). This needs to be made opaque to the external code. My suggested first change is to add routing method pointers to the 'domain' structure, along with information pointing the data. instead of having an array of pointers to uniform structures, there would be an array pointing to the 'domain' structures for each protocol address domain (protocol family), and the methods this reached would be called. The methods would have an argument that gives FIB number, but the protocol would be free to ignore it. When the ABI can be changed it raises the possibilty of the addition of a fib entry into the "struct route". Currently, the structure contains the sockaddr of the desination, and the resulting fib entry. To make this work fully, one could add a fib number so that given an address and a fib, one can find the third element, the fib entry. Interaction with the ARP layer/ LL layer would need to be revisited as well. Qing Li has been working on this already. This work was sponsored by Ironport Systems/Cisco PR: Reviewed by: several including rwatson, bz and mlair (parts each) Approved by: Obtained from: Ironport systems/Cisco MFC after: Security:
2008-05-09 23:00:22 +00:00
2008-05-10 09:22:17 +00:00
intsize = sizeof(int);
if (sysctlbyname("net.my_fibnum", &fibnum, &intsize, NULL, 0) == -1)
Add code to allow the system to handle multiple routing tables. This particular implementation is designed to be fully backwards compatible and to be MFC-able to 7.x (and 6.x) Currently the only protocol that can make use of the multiple tables is IPv4 Similar functionality exists in OpenBSD and Linux. From my notes: ----- One thing where FreeBSD has been falling behind, and which by chance I have some time to work on is "policy based routing", which allows different packet streams to be routed by more than just the destination address. Constraints: ------------ I want to make some form of this available in the 6.x tree (and by extension 7.x) , but FreeBSD in general needs it so I might as well do it in -current and back port the portions I need. One of the ways that this can be done is to have the ability to instantiate multiple kernel routing tables (which I will now refer to as "Forwarding Information Bases" or "FIBs" for political correctness reasons). Which FIB a particular packet uses to make the next hop decision can be decided by a number of mechanisms. The policies these mechanisms implement are the "Policies" referred to in "Policy based routing". One of the constraints I have if I try to back port this work to 6.x is that it must be implemented as a EXTENSION to the existing ABIs in 6.x so that third party applications do not need to be recompiled in timespan of the branch. This first version will not have some of the bells and whistles that will come with later versions. It will, for example, be limited to 16 tables in the first commit. Implementation method, Compatible version. (part 1) ------------------------------- For this reason I have implemented a "sufficient subset" of a multiple routing table solution in Perforce, and back-ported it to 6.x. (also in Perforce though not always caught up with what I have done in -current/P4). The subset allows a number of FIBs to be defined at compile time (8 is sufficient for my purposes in 6.x) and implements the changes needed to allow IPV4 to use them. I have not done the changes for ipv6 simply because I do not need it, and I do not have enough knowledge of ipv6 (e.g. neighbor discovery) needed to do it. Other protocol families are left untouched and should there be users with proprietary protocol families, they should continue to work and be oblivious to the existence of the extra FIBs. To understand how this is done, one must know that the current FIB code starts everything off with a single dimensional array of pointers to FIB head structures (One per protocol family), each of which in turn points to the trie of routes available to that family. The basic change in the ABI compatible version of the change is to extent that array to be a 2 dimensional array, so that instead of protocol family X looking at rt_tables[X] for the table it needs, it looks at rt_tables[Y][X] when for all protocol families except ipv4 Y is always 0. Code that is unaware of the change always just sees the first row of the table, which of course looks just like the one dimensional array that existed before. The entry points rtrequest(), rtalloc(), rtalloc1(), rtalloc_ign() are all maintained, but refer only to the first row of the array, so that existing callers in proprietary protocols can continue to do the "right thing". Some new entry points are added, for the exclusive use of ipv4 code called in_rtrequest(), in_rtalloc(), in_rtalloc1() and in_rtalloc_ign(), which have an extra argument which refers the code to the correct row. In addition, there are some new entry points (currently called rtalloc_fib() and friends) that check the Address family being looked up and call either rtalloc() (and friends) if the protocol is not IPv4 forcing the action to row 0 or to the appropriate row if it IS IPv4 (and that info is available). These are for calling from code that is not specific to any particular protocol. The way these are implemented would change in the non ABI preserving code to be added later. One feature of the first version of the code is that for ipv4, the interface routes show up automatically on all the FIBs, so that no matter what FIB you select you always have the basic direct attached hosts available to you. (rtinit() does this automatically). You CAN delete an interface route from one FIB should you want to but by default it's there. ARP information is also available in each FIB. It's assumed that the same machine would have the same MAC address, regardless of which FIB you are using to get to it. This brings us as to how the correct FIB is selected for an outgoing IPV4 packet. Firstly, all packets have a FIB associated with them. if nothing has been done to change it, it will be FIB 0. The FIB is changed in the following ways. Packets fall into one of a number of classes. 1/ locally generated packets, coming from a socket/PCB. Such packets select a FIB from a number associated with the socket/PCB. This in turn is inherited from the process, but can be changed by a socket option. The process in turn inherits it on fork. I have written a utility call setfib that acts a bit like nice.. setfib -3 ping target.example.com # will use fib 3 for ping. It is an obvious extension to make it a property of a jail but I have not done so. It can be achieved by combining the setfib and jail commands. 2/ packets received on an interface for forwarding. By default these packets would use table 0, (or possibly a number settable in a sysctl(not yet)). but prior to routing the firewall can inspect them (see below). (possibly in the future you may be able to associate a FIB with packets received on an interface.. An ifconfig arg, but not yet.) 3/ packets inspected by a packet classifier, which can arbitrarily associate a fib with it on a packet by packet basis. A fib assigned to a packet by a packet classifier (such as ipfw) would over-ride a fib associated by a more default source. (such as cases 1 or 2). 4/ a tcp listen socket associated with a fib will generate accept sockets that are associated with that same fib. 5/ Packets generated in response to some other packet (e.g. reset or icmp packets). These should use the FIB associated with the packet being reponded to. 6/ Packets generated during encapsulation. gif, tun and other tunnel interfaces will encapsulate using the FIB that was in effect withthe proces that set up the tunnel. thus setfib 1 ifconfig gif0 [tunnel instructions] will set the fib for the tunnel to use to be fib 1. Routing messages would be associated with their process, and thus select one FIB or another. messages from the kernel would be associated with the fib they refer to and would only be received by a routing socket associated with that fib. (not yet implemented) In addition Netstat has been edited to be able to cope with the fact that the array is now 2 dimensional. (It looks in system memory using libkvm (!)). Old versions of netstat see only the first FIB. In addition two sysctls are added to give: a) the number of FIBs compiled in (active) b) the default FIB of the calling process. Early testing experience: ------------------------- Basically our (IronPort's) appliance does this functionality already using ipfw fwd but that method has some drawbacks. For example, It can't fully simulate a routing table because it can't influence the socket's choice of local address when a connect() is done. Testing during the generating of these changes has been remarkably smooth so far. Multiple tables have co-existed with no notable side effects, and packets have been routes accordingly. ipfw has grown 2 new keywords: setfib N ip from anay to any count ip from any to any fib N In pf there seems to be a requirement to be able to give symbolic names to the fibs but I do not have that capacity. I am not sure if it is required. SCTP has interestingly enough built in support for this, called VRFs in Cisco parlance. it will be interesting to see how that handles it when it suddenly actually does something. Where to next: -------------------- After committing the ABI compatible version and MFCing it, I'd like to proceed in a forward direction in -current. this will result in some roto-tilling in the routing code. Firstly: the current code's idea of having a separate tree per protocol family, all of the same format, and pointed to by the 1 dimensional array is a bit silly. Especially when one considers that there is code that makes assumptions about every protocol having the same internal structures there. Some protocols don't WANT that sort of structure. (for example the whole idea of a netmask is foreign to appletalk). This needs to be made opaque to the external code. My suggested first change is to add routing method pointers to the 'domain' structure, along with information pointing the data. instead of having an array of pointers to uniform structures, there would be an array pointing to the 'domain' structures for each protocol address domain (protocol family), and the methods this reached would be called. The methods would have an argument that gives FIB number, but the protocol would be free to ignore it. When the ABI can be changed it raises the possibilty of the addition of a fib entry into the "struct route". Currently, the structure contains the sockaddr of the desination, and the resulting fib entry. To make this work fully, one could add a fib number so that given an address and a fib, one can find the third element, the fib entry. Interaction with the ARP layer/ LL layer would need to be revisited as well. Qing Li has been working on this already. This work was sponsored by Ironport Systems/Cisco PR: Reviewed by: several including rwatson, bz and mlair (parts each) Approved by: Obtained from: Ironport systems/Cisco MFC after: Security:
2008-05-09 23:00:22 +00:00
fibnum = 0;
2008-05-10 09:22:17 +00:00
if (sysctlbyname("net.fibs", &numfibs, &intsize, NULL, 0) == -1)
Add code to allow the system to handle multiple routing tables. This particular implementation is designed to be fully backwards compatible and to be MFC-able to 7.x (and 6.x) Currently the only protocol that can make use of the multiple tables is IPv4 Similar functionality exists in OpenBSD and Linux. From my notes: ----- One thing where FreeBSD has been falling behind, and which by chance I have some time to work on is "policy based routing", which allows different packet streams to be routed by more than just the destination address. Constraints: ------------ I want to make some form of this available in the 6.x tree (and by extension 7.x) , but FreeBSD in general needs it so I might as well do it in -current and back port the portions I need. One of the ways that this can be done is to have the ability to instantiate multiple kernel routing tables (which I will now refer to as "Forwarding Information Bases" or "FIBs" for political correctness reasons). Which FIB a particular packet uses to make the next hop decision can be decided by a number of mechanisms. The policies these mechanisms implement are the "Policies" referred to in "Policy based routing". One of the constraints I have if I try to back port this work to 6.x is that it must be implemented as a EXTENSION to the existing ABIs in 6.x so that third party applications do not need to be recompiled in timespan of the branch. This first version will not have some of the bells and whistles that will come with later versions. It will, for example, be limited to 16 tables in the first commit. Implementation method, Compatible version. (part 1) ------------------------------- For this reason I have implemented a "sufficient subset" of a multiple routing table solution in Perforce, and back-ported it to 6.x. (also in Perforce though not always caught up with what I have done in -current/P4). The subset allows a number of FIBs to be defined at compile time (8 is sufficient for my purposes in 6.x) and implements the changes needed to allow IPV4 to use them. I have not done the changes for ipv6 simply because I do not need it, and I do not have enough knowledge of ipv6 (e.g. neighbor discovery) needed to do it. Other protocol families are left untouched and should there be users with proprietary protocol families, they should continue to work and be oblivious to the existence of the extra FIBs. To understand how this is done, one must know that the current FIB code starts everything off with a single dimensional array of pointers to FIB head structures (One per protocol family), each of which in turn points to the trie of routes available to that family. The basic change in the ABI compatible version of the change is to extent that array to be a 2 dimensional array, so that instead of protocol family X looking at rt_tables[X] for the table it needs, it looks at rt_tables[Y][X] when for all protocol families except ipv4 Y is always 0. Code that is unaware of the change always just sees the first row of the table, which of course looks just like the one dimensional array that existed before. The entry points rtrequest(), rtalloc(), rtalloc1(), rtalloc_ign() are all maintained, but refer only to the first row of the array, so that existing callers in proprietary protocols can continue to do the "right thing". Some new entry points are added, for the exclusive use of ipv4 code called in_rtrequest(), in_rtalloc(), in_rtalloc1() and in_rtalloc_ign(), which have an extra argument which refers the code to the correct row. In addition, there are some new entry points (currently called rtalloc_fib() and friends) that check the Address family being looked up and call either rtalloc() (and friends) if the protocol is not IPv4 forcing the action to row 0 or to the appropriate row if it IS IPv4 (and that info is available). These are for calling from code that is not specific to any particular protocol. The way these are implemented would change in the non ABI preserving code to be added later. One feature of the first version of the code is that for ipv4, the interface routes show up automatically on all the FIBs, so that no matter what FIB you select you always have the basic direct attached hosts available to you. (rtinit() does this automatically). You CAN delete an interface route from one FIB should you want to but by default it's there. ARP information is also available in each FIB. It's assumed that the same machine would have the same MAC address, regardless of which FIB you are using to get to it. This brings us as to how the correct FIB is selected for an outgoing IPV4 packet. Firstly, all packets have a FIB associated with them. if nothing has been done to change it, it will be FIB 0. The FIB is changed in the following ways. Packets fall into one of a number of classes. 1/ locally generated packets, coming from a socket/PCB. Such packets select a FIB from a number associated with the socket/PCB. This in turn is inherited from the process, but can be changed by a socket option. The process in turn inherits it on fork. I have written a utility call setfib that acts a bit like nice.. setfib -3 ping target.example.com # will use fib 3 for ping. It is an obvious extension to make it a property of a jail but I have not done so. It can be achieved by combining the setfib and jail commands. 2/ packets received on an interface for forwarding. By default these packets would use table 0, (or possibly a number settable in a sysctl(not yet)). but prior to routing the firewall can inspect them (see below). (possibly in the future you may be able to associate a FIB with packets received on an interface.. An ifconfig arg, but not yet.) 3/ packets inspected by a packet classifier, which can arbitrarily associate a fib with it on a packet by packet basis. A fib assigned to a packet by a packet classifier (such as ipfw) would over-ride a fib associated by a more default source. (such as cases 1 or 2). 4/ a tcp listen socket associated with a fib will generate accept sockets that are associated with that same fib. 5/ Packets generated in response to some other packet (e.g. reset or icmp packets). These should use the FIB associated with the packet being reponded to. 6/ Packets generated during encapsulation. gif, tun and other tunnel interfaces will encapsulate using the FIB that was in effect withthe proces that set up the tunnel. thus setfib 1 ifconfig gif0 [tunnel instructions] will set the fib for the tunnel to use to be fib 1. Routing messages would be associated with their process, and thus select one FIB or another. messages from the kernel would be associated with the fib they refer to and would only be received by a routing socket associated with that fib. (not yet implemented) In addition Netstat has been edited to be able to cope with the fact that the array is now 2 dimensional. (It looks in system memory using libkvm (!)). Old versions of netstat see only the first FIB. In addition two sysctls are added to give: a) the number of FIBs compiled in (active) b) the default FIB of the calling process. Early testing experience: ------------------------- Basically our (IronPort's) appliance does this functionality already using ipfw fwd but that method has some drawbacks. For example, It can't fully simulate a routing table because it can't influence the socket's choice of local address when a connect() is done. Testing during the generating of these changes has been remarkably smooth so far. Multiple tables have co-existed with no notable side effects, and packets have been routes accordingly. ipfw has grown 2 new keywords: setfib N ip from anay to any count ip from any to any fib N In pf there seems to be a requirement to be able to give symbolic names to the fibs but I do not have that capacity. I am not sure if it is required. SCTP has interestingly enough built in support for this, called VRFs in Cisco parlance. it will be interesting to see how that handles it when it suddenly actually does something. Where to next: -------------------- After committing the ABI compatible version and MFCing it, I'd like to proceed in a forward direction in -current. this will result in some roto-tilling in the routing code. Firstly: the current code's idea of having a separate tree per protocol family, all of the same format, and pointed to by the 1 dimensional array is a bit silly. Especially when one considers that there is code that makes assumptions about every protocol having the same internal structures there. Some protocols don't WANT that sort of structure. (for example the whole idea of a netmask is foreign to appletalk). This needs to be made opaque to the external code. My suggested first change is to add routing method pointers to the 'domain' structure, along with information pointing the data. instead of having an array of pointers to uniform structures, there would be an array pointing to the 'domain' structures for each protocol address domain (protocol family), and the methods this reached would be called. The methods would have an argument that gives FIB number, but the protocol would be free to ignore it. When the ABI can be changed it raises the possibilty of the addition of a fib entry into the "struct route". Currently, the structure contains the sockaddr of the desination, and the resulting fib entry. To make this work fully, one could add a fib number so that given an address and a fib, one can find the third element, the fib entry. Interaction with the ARP layer/ LL layer would need to be revisited as well. Qing Li has been working on this already. This work was sponsored by Ironport Systems/Cisco PR: Reviewed by: several including rwatson, bz and mlair (parts each) Approved by: Obtained from: Ironport systems/Cisco MFC after: Security:
2008-05-09 23:00:22 +00:00
numfibs = 1;
rt_tables = calloc(numfibs * (AF_MAX+1),
sizeof(struct radix_node_head *));
Add code to allow the system to handle multiple routing tables. This particular implementation is designed to be fully backwards compatible and to be MFC-able to 7.x (and 6.x) Currently the only protocol that can make use of the multiple tables is IPv4 Similar functionality exists in OpenBSD and Linux. From my notes: ----- One thing where FreeBSD has been falling behind, and which by chance I have some time to work on is "policy based routing", which allows different packet streams to be routed by more than just the destination address. Constraints: ------------ I want to make some form of this available in the 6.x tree (and by extension 7.x) , but FreeBSD in general needs it so I might as well do it in -current and back port the portions I need. One of the ways that this can be done is to have the ability to instantiate multiple kernel routing tables (which I will now refer to as "Forwarding Information Bases" or "FIBs" for political correctness reasons). Which FIB a particular packet uses to make the next hop decision can be decided by a number of mechanisms. The policies these mechanisms implement are the "Policies" referred to in "Policy based routing". One of the constraints I have if I try to back port this work to 6.x is that it must be implemented as a EXTENSION to the existing ABIs in 6.x so that third party applications do not need to be recompiled in timespan of the branch. This first version will not have some of the bells and whistles that will come with later versions. It will, for example, be limited to 16 tables in the first commit. Implementation method, Compatible version. (part 1) ------------------------------- For this reason I have implemented a "sufficient subset" of a multiple routing table solution in Perforce, and back-ported it to 6.x. (also in Perforce though not always caught up with what I have done in -current/P4). The subset allows a number of FIBs to be defined at compile time (8 is sufficient for my purposes in 6.x) and implements the changes needed to allow IPV4 to use them. I have not done the changes for ipv6 simply because I do not need it, and I do not have enough knowledge of ipv6 (e.g. neighbor discovery) needed to do it. Other protocol families are left untouched and should there be users with proprietary protocol families, they should continue to work and be oblivious to the existence of the extra FIBs. To understand how this is done, one must know that the current FIB code starts everything off with a single dimensional array of pointers to FIB head structures (One per protocol family), each of which in turn points to the trie of routes available to that family. The basic change in the ABI compatible version of the change is to extent that array to be a 2 dimensional array, so that instead of protocol family X looking at rt_tables[X] for the table it needs, it looks at rt_tables[Y][X] when for all protocol families except ipv4 Y is always 0. Code that is unaware of the change always just sees the first row of the table, which of course looks just like the one dimensional array that existed before. The entry points rtrequest(), rtalloc(), rtalloc1(), rtalloc_ign() are all maintained, but refer only to the first row of the array, so that existing callers in proprietary protocols can continue to do the "right thing". Some new entry points are added, for the exclusive use of ipv4 code called in_rtrequest(), in_rtalloc(), in_rtalloc1() and in_rtalloc_ign(), which have an extra argument which refers the code to the correct row. In addition, there are some new entry points (currently called rtalloc_fib() and friends) that check the Address family being looked up and call either rtalloc() (and friends) if the protocol is not IPv4 forcing the action to row 0 or to the appropriate row if it IS IPv4 (and that info is available). These are for calling from code that is not specific to any particular protocol. The way these are implemented would change in the non ABI preserving code to be added later. One feature of the first version of the code is that for ipv4, the interface routes show up automatically on all the FIBs, so that no matter what FIB you select you always have the basic direct attached hosts available to you. (rtinit() does this automatically). You CAN delete an interface route from one FIB should you want to but by default it's there. ARP information is also available in each FIB. It's assumed that the same machine would have the same MAC address, regardless of which FIB you are using to get to it. This brings us as to how the correct FIB is selected for an outgoing IPV4 packet. Firstly, all packets have a FIB associated with them. if nothing has been done to change it, it will be FIB 0. The FIB is changed in the following ways. Packets fall into one of a number of classes. 1/ locally generated packets, coming from a socket/PCB. Such packets select a FIB from a number associated with the socket/PCB. This in turn is inherited from the process, but can be changed by a socket option. The process in turn inherits it on fork. I have written a utility call setfib that acts a bit like nice.. setfib -3 ping target.example.com # will use fib 3 for ping. It is an obvious extension to make it a property of a jail but I have not done so. It can be achieved by combining the setfib and jail commands. 2/ packets received on an interface for forwarding. By default these packets would use table 0, (or possibly a number settable in a sysctl(not yet)). but prior to routing the firewall can inspect them (see below). (possibly in the future you may be able to associate a FIB with packets received on an interface.. An ifconfig arg, but not yet.) 3/ packets inspected by a packet classifier, which can arbitrarily associate a fib with it on a packet by packet basis. A fib assigned to a packet by a packet classifier (such as ipfw) would over-ride a fib associated by a more default source. (such as cases 1 or 2). 4/ a tcp listen socket associated with a fib will generate accept sockets that are associated with that same fib. 5/ Packets generated in response to some other packet (e.g. reset or icmp packets). These should use the FIB associated with the packet being reponded to. 6/ Packets generated during encapsulation. gif, tun and other tunnel interfaces will encapsulate using the FIB that was in effect withthe proces that set up the tunnel. thus setfib 1 ifconfig gif0 [tunnel instructions] will set the fib for the tunnel to use to be fib 1. Routing messages would be associated with their process, and thus select one FIB or another. messages from the kernel would be associated with the fib they refer to and would only be received by a routing socket associated with that fib. (not yet implemented) In addition Netstat has been edited to be able to cope with the fact that the array is now 2 dimensional. (It looks in system memory using libkvm (!)). Old versions of netstat see only the first FIB. In addition two sysctls are added to give: a) the number of FIBs compiled in (active) b) the default FIB of the calling process. Early testing experience: ------------------------- Basically our (IronPort's) appliance does this functionality already using ipfw fwd but that method has some drawbacks. For example, It can't fully simulate a routing table because it can't influence the socket's choice of local address when a connect() is done. Testing during the generating of these changes has been remarkably smooth so far. Multiple tables have co-existed with no notable side effects, and packets have been routes accordingly. ipfw has grown 2 new keywords: setfib N ip from anay to any count ip from any to any fib N In pf there seems to be a requirement to be able to give symbolic names to the fibs but I do not have that capacity. I am not sure if it is required. SCTP has interestingly enough built in support for this, called VRFs in Cisco parlance. it will be interesting to see how that handles it when it suddenly actually does something. Where to next: -------------------- After committing the ABI compatible version and MFCing it, I'd like to proceed in a forward direction in -current. this will result in some roto-tilling in the routing code. Firstly: the current code's idea of having a separate tree per protocol family, all of the same format, and pointed to by the 1 dimensional array is a bit silly. Especially when one considers that there is code that makes assumptions about every protocol having the same internal structures there. Some protocols don't WANT that sort of structure. (for example the whole idea of a netmask is foreign to appletalk). This needs to be made opaque to the external code. My suggested first change is to add routing method pointers to the 'domain' structure, along with information pointing the data. instead of having an array of pointers to uniform structures, there would be an array pointing to the 'domain' structures for each protocol address domain (protocol family), and the methods this reached would be called. The methods would have an argument that gives FIB number, but the protocol would be free to ignore it. When the ABI can be changed it raises the possibilty of the addition of a fib entry into the "struct route". Currently, the structure contains the sockaddr of the desination, and the resulting fib entry. To make this work fully, one could add a fib number so that given an address and a fib, one can find the third element, the fib entry. Interaction with the ARP layer/ LL layer would need to be revisited as well. Qing Li has been working on this already. This work was sponsored by Ironport Systems/Cisco PR: Reviewed by: several including rwatson, bz and mlair (parts each) Approved by: Obtained from: Ironport systems/Cisco MFC after: Security:
2008-05-09 23:00:22 +00:00
if (rt_tables == NULL)
err(EX_OSERR, "memory allocation failed");
/*
* Since kernel & userland use different timebase
* (time_uptime vs time_second) and we are reading kernel memory
* directly we should do rt_rmx.rmx_expire --> expire_time conversion.
*/
if (clock_gettime(CLOCK_UPTIME, &uptime) < 0)
err(EX_OSERR, "clock_gettime() failed");
1994-05-27 12:33:43 +00:00
printf("Routing tables\n");
if (Aflag == 0 && NewTree)
ntreestuff();
else {
if (rtree == 0) {
printf("rt_tables: symbol not in namelist\n");
return;
}
if (kread((u_long)(rtree), (char *)(rt_tables), (numfibs *
(AF_MAX+1) * sizeof(struct radix_node_head *))) != 0)
return;
for (fam = 0; fam <= AF_MAX; fam++) {
Add code to allow the system to handle multiple routing tables. This particular implementation is designed to be fully backwards compatible and to be MFC-able to 7.x (and 6.x) Currently the only protocol that can make use of the multiple tables is IPv4 Similar functionality exists in OpenBSD and Linux. From my notes: ----- One thing where FreeBSD has been falling behind, and which by chance I have some time to work on is "policy based routing", which allows different packet streams to be routed by more than just the destination address. Constraints: ------------ I want to make some form of this available in the 6.x tree (and by extension 7.x) , but FreeBSD in general needs it so I might as well do it in -current and back port the portions I need. One of the ways that this can be done is to have the ability to instantiate multiple kernel routing tables (which I will now refer to as "Forwarding Information Bases" or "FIBs" for political correctness reasons). Which FIB a particular packet uses to make the next hop decision can be decided by a number of mechanisms. The policies these mechanisms implement are the "Policies" referred to in "Policy based routing". One of the constraints I have if I try to back port this work to 6.x is that it must be implemented as a EXTENSION to the existing ABIs in 6.x so that third party applications do not need to be recompiled in timespan of the branch. This first version will not have some of the bells and whistles that will come with later versions. It will, for example, be limited to 16 tables in the first commit. Implementation method, Compatible version. (part 1) ------------------------------- For this reason I have implemented a "sufficient subset" of a multiple routing table solution in Perforce, and back-ported it to 6.x. (also in Perforce though not always caught up with what I have done in -current/P4). The subset allows a number of FIBs to be defined at compile time (8 is sufficient for my purposes in 6.x) and implements the changes needed to allow IPV4 to use them. I have not done the changes for ipv6 simply because I do not need it, and I do not have enough knowledge of ipv6 (e.g. neighbor discovery) needed to do it. Other protocol families are left untouched and should there be users with proprietary protocol families, they should continue to work and be oblivious to the existence of the extra FIBs. To understand how this is done, one must know that the current FIB code starts everything off with a single dimensional array of pointers to FIB head structures (One per protocol family), each of which in turn points to the trie of routes available to that family. The basic change in the ABI compatible version of the change is to extent that array to be a 2 dimensional array, so that instead of protocol family X looking at rt_tables[X] for the table it needs, it looks at rt_tables[Y][X] when for all protocol families except ipv4 Y is always 0. Code that is unaware of the change always just sees the first row of the table, which of course looks just like the one dimensional array that existed before. The entry points rtrequest(), rtalloc(), rtalloc1(), rtalloc_ign() are all maintained, but refer only to the first row of the array, so that existing callers in proprietary protocols can continue to do the "right thing". Some new entry points are added, for the exclusive use of ipv4 code called in_rtrequest(), in_rtalloc(), in_rtalloc1() and in_rtalloc_ign(), which have an extra argument which refers the code to the correct row. In addition, there are some new entry points (currently called rtalloc_fib() and friends) that check the Address family being looked up and call either rtalloc() (and friends) if the protocol is not IPv4 forcing the action to row 0 or to the appropriate row if it IS IPv4 (and that info is available). These are for calling from code that is not specific to any particular protocol. The way these are implemented would change in the non ABI preserving code to be added later. One feature of the first version of the code is that for ipv4, the interface routes show up automatically on all the FIBs, so that no matter what FIB you select you always have the basic direct attached hosts available to you. (rtinit() does this automatically). You CAN delete an interface route from one FIB should you want to but by default it's there. ARP information is also available in each FIB. It's assumed that the same machine would have the same MAC address, regardless of which FIB you are using to get to it. This brings us as to how the correct FIB is selected for an outgoing IPV4 packet. Firstly, all packets have a FIB associated with them. if nothing has been done to change it, it will be FIB 0. The FIB is changed in the following ways. Packets fall into one of a number of classes. 1/ locally generated packets, coming from a socket/PCB. Such packets select a FIB from a number associated with the socket/PCB. This in turn is inherited from the process, but can be changed by a socket option. The process in turn inherits it on fork. I have written a utility call setfib that acts a bit like nice.. setfib -3 ping target.example.com # will use fib 3 for ping. It is an obvious extension to make it a property of a jail but I have not done so. It can be achieved by combining the setfib and jail commands. 2/ packets received on an interface for forwarding. By default these packets would use table 0, (or possibly a number settable in a sysctl(not yet)). but prior to routing the firewall can inspect them (see below). (possibly in the future you may be able to associate a FIB with packets received on an interface.. An ifconfig arg, but not yet.) 3/ packets inspected by a packet classifier, which can arbitrarily associate a fib with it on a packet by packet basis. A fib assigned to a packet by a packet classifier (such as ipfw) would over-ride a fib associated by a more default source. (such as cases 1 or 2). 4/ a tcp listen socket associated with a fib will generate accept sockets that are associated with that same fib. 5/ Packets generated in response to some other packet (e.g. reset or icmp packets). These should use the FIB associated with the packet being reponded to. 6/ Packets generated during encapsulation. gif, tun and other tunnel interfaces will encapsulate using the FIB that was in effect withthe proces that set up the tunnel. thus setfib 1 ifconfig gif0 [tunnel instructions] will set the fib for the tunnel to use to be fib 1. Routing messages would be associated with their process, and thus select one FIB or another. messages from the kernel would be associated with the fib they refer to and would only be received by a routing socket associated with that fib. (not yet implemented) In addition Netstat has been edited to be able to cope with the fact that the array is now 2 dimensional. (It looks in system memory using libkvm (!)). Old versions of netstat see only the first FIB. In addition two sysctls are added to give: a) the number of FIBs compiled in (active) b) the default FIB of the calling process. Early testing experience: ------------------------- Basically our (IronPort's) appliance does this functionality already using ipfw fwd but that method has some drawbacks. For example, It can't fully simulate a routing table because it can't influence the socket's choice of local address when a connect() is done. Testing during the generating of these changes has been remarkably smooth so far. Multiple tables have co-existed with no notable side effects, and packets have been routes accordingly. ipfw has grown 2 new keywords: setfib N ip from anay to any count ip from any to any fib N In pf there seems to be a requirement to be able to give symbolic names to the fibs but I do not have that capacity. I am not sure if it is required. SCTP has interestingly enough built in support for this, called VRFs in Cisco parlance. it will be interesting to see how that handles it when it suddenly actually does something. Where to next: -------------------- After committing the ABI compatible version and MFCing it, I'd like to proceed in a forward direction in -current. this will result in some roto-tilling in the routing code. Firstly: the current code's idea of having a separate tree per protocol family, all of the same format, and pointed to by the 1 dimensional array is a bit silly. Especially when one considers that there is code that makes assumptions about every protocol having the same internal structures there. Some protocols don't WANT that sort of structure. (for example the whole idea of a netmask is foreign to appletalk). This needs to be made opaque to the external code. My suggested first change is to add routing method pointers to the 'domain' structure, along with information pointing the data. instead of having an array of pointers to uniform structures, there would be an array pointing to the 'domain' structures for each protocol address domain (protocol family), and the methods this reached would be called. The methods would have an argument that gives FIB number, but the protocol would be free to ignore it. When the ABI can be changed it raises the possibilty of the addition of a fib entry into the "struct route". Currently, the structure contains the sockaddr of the desination, and the resulting fib entry. To make this work fully, one could add a fib number so that given an address and a fib, one can find the third element, the fib entry. Interaction with the ARP layer/ LL layer would need to be revisited as well. Qing Li has been working on this already. This work was sponsored by Ironport Systems/Cisco PR: Reviewed by: several including rwatson, bz and mlair (parts each) Approved by: Obtained from: Ironport systems/Cisco MFC after: Security:
2008-05-09 23:00:22 +00:00
int tmpfib;
switch (fam) {
case AF_INET6:
case AF_INET:
Add code to allow the system to handle multiple routing tables. This particular implementation is designed to be fully backwards compatible and to be MFC-able to 7.x (and 6.x) Currently the only protocol that can make use of the multiple tables is IPv4 Similar functionality exists in OpenBSD and Linux. From my notes: ----- One thing where FreeBSD has been falling behind, and which by chance I have some time to work on is "policy based routing", which allows different packet streams to be routed by more than just the destination address. Constraints: ------------ I want to make some form of this available in the 6.x tree (and by extension 7.x) , but FreeBSD in general needs it so I might as well do it in -current and back port the portions I need. One of the ways that this can be done is to have the ability to instantiate multiple kernel routing tables (which I will now refer to as "Forwarding Information Bases" or "FIBs" for political correctness reasons). Which FIB a particular packet uses to make the next hop decision can be decided by a number of mechanisms. The policies these mechanisms implement are the "Policies" referred to in "Policy based routing". One of the constraints I have if I try to back port this work to 6.x is that it must be implemented as a EXTENSION to the existing ABIs in 6.x so that third party applications do not need to be recompiled in timespan of the branch. This first version will not have some of the bells and whistles that will come with later versions. It will, for example, be limited to 16 tables in the first commit. Implementation method, Compatible version. (part 1) ------------------------------- For this reason I have implemented a "sufficient subset" of a multiple routing table solution in Perforce, and back-ported it to 6.x. (also in Perforce though not always caught up with what I have done in -current/P4). The subset allows a number of FIBs to be defined at compile time (8 is sufficient for my purposes in 6.x) and implements the changes needed to allow IPV4 to use them. I have not done the changes for ipv6 simply because I do not need it, and I do not have enough knowledge of ipv6 (e.g. neighbor discovery) needed to do it. Other protocol families are left untouched and should there be users with proprietary protocol families, they should continue to work and be oblivious to the existence of the extra FIBs. To understand how this is done, one must know that the current FIB code starts everything off with a single dimensional array of pointers to FIB head structures (One per protocol family), each of which in turn points to the trie of routes available to that family. The basic change in the ABI compatible version of the change is to extent that array to be a 2 dimensional array, so that instead of protocol family X looking at rt_tables[X] for the table it needs, it looks at rt_tables[Y][X] when for all protocol families except ipv4 Y is always 0. Code that is unaware of the change always just sees the first row of the table, which of course looks just like the one dimensional array that existed before. The entry points rtrequest(), rtalloc(), rtalloc1(), rtalloc_ign() are all maintained, but refer only to the first row of the array, so that existing callers in proprietary protocols can continue to do the "right thing". Some new entry points are added, for the exclusive use of ipv4 code called in_rtrequest(), in_rtalloc(), in_rtalloc1() and in_rtalloc_ign(), which have an extra argument which refers the code to the correct row. In addition, there are some new entry points (currently called rtalloc_fib() and friends) that check the Address family being looked up and call either rtalloc() (and friends) if the protocol is not IPv4 forcing the action to row 0 or to the appropriate row if it IS IPv4 (and that info is available). These are for calling from code that is not specific to any particular protocol. The way these are implemented would change in the non ABI preserving code to be added later. One feature of the first version of the code is that for ipv4, the interface routes show up automatically on all the FIBs, so that no matter what FIB you select you always have the basic direct attached hosts available to you. (rtinit() does this automatically). You CAN delete an interface route from one FIB should you want to but by default it's there. ARP information is also available in each FIB. It's assumed that the same machine would have the same MAC address, regardless of which FIB you are using to get to it. This brings us as to how the correct FIB is selected for an outgoing IPV4 packet. Firstly, all packets have a FIB associated with them. if nothing has been done to change it, it will be FIB 0. The FIB is changed in the following ways. Packets fall into one of a number of classes. 1/ locally generated packets, coming from a socket/PCB. Such packets select a FIB from a number associated with the socket/PCB. This in turn is inherited from the process, but can be changed by a socket option. The process in turn inherits it on fork. I have written a utility call setfib that acts a bit like nice.. setfib -3 ping target.example.com # will use fib 3 for ping. It is an obvious extension to make it a property of a jail but I have not done so. It can be achieved by combining the setfib and jail commands. 2/ packets received on an interface for forwarding. By default these packets would use table 0, (or possibly a number settable in a sysctl(not yet)). but prior to routing the firewall can inspect them (see below). (possibly in the future you may be able to associate a FIB with packets received on an interface.. An ifconfig arg, but not yet.) 3/ packets inspected by a packet classifier, which can arbitrarily associate a fib with it on a packet by packet basis. A fib assigned to a packet by a packet classifier (such as ipfw) would over-ride a fib associated by a more default source. (such as cases 1 or 2). 4/ a tcp listen socket associated with a fib will generate accept sockets that are associated with that same fib. 5/ Packets generated in response to some other packet (e.g. reset or icmp packets). These should use the FIB associated with the packet being reponded to. 6/ Packets generated during encapsulation. gif, tun and other tunnel interfaces will encapsulate using the FIB that was in effect withthe proces that set up the tunnel. thus setfib 1 ifconfig gif0 [tunnel instructions] will set the fib for the tunnel to use to be fib 1. Routing messages would be associated with their process, and thus select one FIB or another. messages from the kernel would be associated with the fib they refer to and would only be received by a routing socket associated with that fib. (not yet implemented) In addition Netstat has been edited to be able to cope with the fact that the array is now 2 dimensional. (It looks in system memory using libkvm (!)). Old versions of netstat see only the first FIB. In addition two sysctls are added to give: a) the number of FIBs compiled in (active) b) the default FIB of the calling process. Early testing experience: ------------------------- Basically our (IronPort's) appliance does this functionality already using ipfw fwd but that method has some drawbacks. For example, It can't fully simulate a routing table because it can't influence the socket's choice of local address when a connect() is done. Testing during the generating of these changes has been remarkably smooth so far. Multiple tables have co-existed with no notable side effects, and packets have been routes accordingly. ipfw has grown 2 new keywords: setfib N ip from anay to any count ip from any to any fib N In pf there seems to be a requirement to be able to give symbolic names to the fibs but I do not have that capacity. I am not sure if it is required. SCTP has interestingly enough built in support for this, called VRFs in Cisco parlance. it will be interesting to see how that handles it when it suddenly actually does something. Where to next: -------------------- After committing the ABI compatible version and MFCing it, I'd like to proceed in a forward direction in -current. this will result in some roto-tilling in the routing code. Firstly: the current code's idea of having a separate tree per protocol family, all of the same format, and pointed to by the 1 dimensional array is a bit silly. Especially when one considers that there is code that makes assumptions about every protocol having the same internal structures there. Some protocols don't WANT that sort of structure. (for example the whole idea of a netmask is foreign to appletalk). This needs to be made opaque to the external code. My suggested first change is to add routing method pointers to the 'domain' structure, along with information pointing the data. instead of having an array of pointers to uniform structures, there would be an array pointing to the 'domain' structures for each protocol address domain (protocol family), and the methods this reached would be called. The methods would have an argument that gives FIB number, but the protocol would be free to ignore it. When the ABI can be changed it raises the possibilty of the addition of a fib entry into the "struct route". Currently, the structure contains the sockaddr of the desination, and the resulting fib entry. To make this work fully, one could add a fib number so that given an address and a fib, one can find the third element, the fib entry. Interaction with the ARP layer/ LL layer would need to be revisited as well. Qing Li has been working on this already. This work was sponsored by Ironport Systems/Cisco PR: Reviewed by: several including rwatson, bz and mlair (parts each) Approved by: Obtained from: Ironport systems/Cisco MFC after: Security:
2008-05-09 23:00:22 +00:00
tmpfib = fibnum;
break;
default:
tmpfib = 0;
}
rnhp = (struct radix_node_head **)*rt_tables;
/* Calculate the in-kernel address. */
rnhp += tmpfib * (AF_MAX+1) + fam;
/* Read the in kernel rhn pointer. */
if (kget(rnhp, rnh) != 0)
continue;
if (rnh == NULL)
1994-05-27 12:33:43 +00:00
continue;
/* Read the rnh data. */
if (kget(rnh, head) != 0)
continue;
if (fam == AF_UNSPEC) {
1994-05-27 12:33:43 +00:00
if (Aflag && af == 0) {
printf("Netmasks:\n");
p_tree(head.rnh_treetop);
}
} else if (af == AF_UNSPEC || af == fam) {
size_cols(fam, head.rnh_treetop);
pr_family(fam);
1994-05-27 12:33:43 +00:00
do_rtent = 1;
pr_rthdr(fam);
1994-05-27 12:33:43 +00:00
p_tree(head.rnh_treetop);
}
}
}
}
/*
* Print address family header before a section of the routing table.
*/
void
pr_family(int af1)
1994-05-27 12:33:43 +00:00
{
const char *afname;
1994-05-27 12:33:43 +00:00
switch (af1) {
1994-05-27 12:33:43 +00:00
case AF_INET:
afname = "Internet";
break;
#ifdef INET6
case AF_INET6:
afname = "Internet6";
break;
#endif /*INET6*/
case AF_IPX:
afname = "IPX";
break;
1994-05-27 12:33:43 +00:00
case AF_ISO:
afname = "ISO";
break;
case AF_APPLETALK:
afname = "AppleTalk";
break;
1994-05-27 12:33:43 +00:00
case AF_CCITT:
afname = "X.25";
break;
case AF_NETGRAPH:
afname = "Netgraph";
break;
1994-05-27 12:33:43 +00:00
default:
afname = NULL;
break;
}
if (afname)
printf("\n%s:\n", afname);
else
printf("\nProtocol Family %d:\n", af1);
1994-05-27 12:33:43 +00:00
}
/* column widths; each followed by one space */
#ifndef INET6
#define WID_DST_DEFAULT(af) 18 /* width of destination column */
#define WID_GW_DEFAULT(af) 18 /* width of gateway column */
#define WID_IF_DEFAULT(af) (Wflag ? 8 : 6) /* width of netif column */
#else
#define WID_DST_DEFAULT(af) \
((af) == AF_INET6 ? (numeric_addr ? 33: 18) : 18)
#define WID_GW_DEFAULT(af) \
((af) == AF_INET6 ? (numeric_addr ? 29 : 18) : 18)
#define WID_IF_DEFAULT(af) ((af) == AF_INET6 ? 8 : (Wflag ? 8 : 6))
#endif /*INET6*/
1994-05-27 12:33:43 +00:00
static int wid_dst;
static int wid_gw;
static int wid_flags;
static int wid_refs;
static int wid_use;
static int wid_mtu;
static int wid_if;
static int wid_expire;
static void
size_cols(int ef __unused, struct radix_node *rn)
{
wid_dst = WID_DST_DEFAULT(ef);
wid_gw = WID_GW_DEFAULT(ef);
wid_flags = 6;
wid_refs = 6;
wid_use = 8;
wid_mtu = 6;
wid_if = WID_IF_DEFAULT(ef);
wid_expire = 6;
if (Wflag)
size_cols_tree(rn);
}
static void
size_cols_tree(struct radix_node *rn)
{
again:
if (kget(rn, rnode) != 0)
return;
if (!(rnode.rn_flags & RNF_ACTIVE))
return;
if (rnode.rn_bit < 0) {
if ((rnode.rn_flags & RNF_ROOT) == 0) {
if (kget(rn, rtentry) != 0)
return;
size_cols_rtentry(&rtentry);
}
if ((rn = rnode.rn_dupedkey))
goto again;
} else {
rn = rnode.rn_right;
size_cols_tree(rnode.rn_left);
size_cols_tree(rn);
}
}
static void
size_cols_rtentry(struct rtentry *rt)
{
static struct ifnet ifnet, *lastif;
static char buffer[100];
const char *bp;
struct sockaddr *sa;
sa_u addr, mask;
int len;
bzero(&addr, sizeof(addr));
if ((sa = kgetsa(rt_key(rt))))
bcopy(sa, &addr, sa->sa_len);
bzero(&mask, sizeof(mask));
if (rt_mask(rt) && (sa = kgetsa(rt_mask(rt))))
bcopy(sa, &mask, sa->sa_len);
bp = fmt_sockaddr(&addr.u_sa, &mask.u_sa, rt->rt_flags);
len = strlen(bp);
wid_dst = MAX(len, wid_dst);
bp = fmt_sockaddr(kgetsa(rt->rt_gateway), NULL, RTF_HOST);
len = strlen(bp);
wid_gw = MAX(len, wid_gw);
bp = fmt_flags(rt->rt_flags);
len = strlen(bp);
wid_flags = MAX(len, wid_flags);
if (addr.u_sa.sa_family == AF_INET || Wflag) {
len = snprintf(buffer, sizeof(buffer), "%d", rt->rt_refcnt);
wid_refs = MAX(len, wid_refs);
len = snprintf(buffer, sizeof(buffer), "%lu", rt->rt_use);
wid_use = MAX(len, wid_use);
if (Wflag && rt->rt_rmx.rmx_mtu != 0) {
len = snprintf(buffer, sizeof(buffer),
"%lu", rt->rt_rmx.rmx_mtu);
wid_mtu = MAX(len, wid_mtu);
}
}
if (rt->rt_ifp) {
if (rt->rt_ifp != lastif) {
if (kget(rt->rt_ifp, ifnet) == 0)
len = strlen(ifnet.if_xname);
else
len = strlen("---");
lastif = rt->rt_ifp;
wid_if = MAX(len, wid_if);
}
if (rt->rt_rmx.rmx_expire) {
time_t expire_time;
if ((expire_time =
rt->rt_rmx.rmx_expire - uptime.tv_sec) > 0) {
len = snprintf(buffer, sizeof(buffer), "%d",
(int)expire_time);
wid_expire = MAX(len, wid_expire);
}
}
}
}
1994-05-27 12:33:43 +00:00
/*
* Print header for routing table columns.
*/
void
pr_rthdr(int af1)
1994-05-27 12:33:43 +00:00
{
1994-05-27 12:33:43 +00:00
if (Aflag)
printf("%-8.8s ","Address");
if (af1 == AF_INET || Wflag) {
if (Wflag) {
printf("%-*.*s %-*.*s %-*.*s %*.*s %*.*s %*.*s %*.*s %*s\n",
wid_dst, wid_dst, "Destination",
wid_gw, wid_gw, "Gateway",
wid_flags, wid_flags, "Flags",
wid_refs, wid_refs, "Refs",
wid_use, wid_use, "Use",
wid_mtu, wid_mtu, "Mtu",
wid_if, wid_if, "Netif",
wid_expire, "Expire");
} else {
printf("%-*.*s %-*.*s %-*.*s %*.*s %*.*s %*.*s %*s\n",
wid_dst, wid_dst, "Destination",
wid_gw, wid_gw, "Gateway",
wid_flags, wid_flags, "Flags",
wid_refs, wid_refs, "Refs",
wid_use, wid_use, "Use",
wid_if, wid_if, "Netif",
wid_expire, "Expire");
}
} else {
printf("%-*.*s %-*.*s %-*.*s %*.*s %*s\n",
wid_dst, wid_dst, "Destination",
wid_gw, wid_gw, "Gateway",
wid_flags, wid_flags, "Flags",
wid_if, wid_if, "Netif",
wid_expire, "Expire");
}
1994-05-27 12:33:43 +00:00
}
static struct sockaddr *
2001-06-15 23:35:13 +00:00
kgetsa(struct sockaddr *dst)
1994-05-27 12:33:43 +00:00
{
if (kget(dst, pt_u.u_sa) != 0)
return (NULL);
1994-05-27 12:33:43 +00:00
if (pt_u.u_sa.sa_len > sizeof (pt_u.u_sa))
kread((u_long)dst, (char *)pt_u.u_data, pt_u.u_sa.sa_len);
return (&pt_u.u_sa);
}
static void
2001-06-15 23:35:13 +00:00
p_tree(struct radix_node *rn)
1994-05-27 12:33:43 +00:00
{
again:
if (kget(rn, rnode) != 0)
return;
if (!(rnode.rn_flags & RNF_ACTIVE))
return;
2000-04-23 14:18:18 +00:00
if (rnode.rn_bit < 0) {
1994-05-27 12:33:43 +00:00
if (Aflag)
printf("%-8.8lx ", (u_long)rn);
1994-05-27 12:33:43 +00:00
if (rnode.rn_flags & RNF_ROOT) {
if (Aflag)
printf("(root node)%s",
rnode.rn_dupedkey ? " =>\n" : "\n");
} else if (do_rtent) {
if (kget(rn, rtentry) == 0) {
p_rtentry(&rtentry);
if (Aflag)
p_rtnode();
}
1994-05-27 12:33:43 +00:00
} else {
p_sockaddr(kgetsa((struct sockaddr *)rnode.rn_key),
1996-01-14 23:33:13 +00:00
NULL, 0, 44);
1994-05-27 12:33:43 +00:00
putchar('\n');
}
if ((rn = rnode.rn_dupedkey))
1994-05-27 12:33:43 +00:00
goto again;
} else {
if (Aflag && do_rtent) {
printf("%-8.8lx ", (u_long)rn);
1994-05-27 12:33:43 +00:00
p_rtnode();
}
2000-04-23 14:18:18 +00:00
rn = rnode.rn_right;
p_tree(rnode.rn_left);
1994-05-27 12:33:43 +00:00
p_tree(rn);
}
}
char nbuf[20];
static void
2001-06-15 23:35:13 +00:00
p_rtnode(void)
1994-05-27 12:33:43 +00:00
{
struct radix_mask *rm = rnode.rn_mklist;
2000-04-23 14:18:18 +00:00
if (rnode.rn_bit < 0) {
1994-05-27 12:33:43 +00:00
if (rnode.rn_mask) {
printf("\t mask ");
p_sockaddr(kgetsa((struct sockaddr *)rnode.rn_mask),
1996-01-14 23:33:13 +00:00
NULL, 0, -1);
1994-05-27 12:33:43 +00:00
} else if (rm == 0)
return;
} else {
2000-04-23 14:18:18 +00:00
sprintf(nbuf, "(%d)", rnode.rn_bit);
printf("%6.6s %8.8lx : %8.8lx", nbuf, (u_long)rnode.rn_left, (u_long)rnode.rn_right);
1994-05-27 12:33:43 +00:00
}
while (rm) {
if (kget(rm, rmask) != 0)
break;
1994-05-27 12:33:43 +00:00
sprintf(nbuf, " %d refs, ", rmask.rm_refs);
printf(" mk = %8.8lx {(%d),%s",
2000-04-23 14:18:18 +00:00
(u_long)rm, -1 - rmask.rm_bit, rmask.rm_refs ? nbuf : " ");
1996-01-14 23:33:13 +00:00
if (rmask.rm_flags & RNF_NORMAL) {
struct radix_node rnode_aux;
printf(" <normal>, ");
if (kget(rmask.rm_leaf, rnode_aux) == 0)
p_sockaddr(kgetsa((struct sockaddr *)rnode_aux.rn_mask),
1996-01-14 23:33:13 +00:00
NULL, 0, -1);
else
p_sockaddr(NULL, NULL, 0, -1);
1996-01-14 23:33:13 +00:00
} else
p_sockaddr(kgetsa((struct sockaddr *)rmask.rm_mask),
NULL, 0, -1);
1994-05-27 12:33:43 +00:00
putchar('}');
if ((rm = rmask.rm_mklist))
1994-05-27 12:33:43 +00:00
printf(" ->");
}
putchar('\n');
}
static void
2001-06-15 23:35:13 +00:00
ntreestuff(void)
1994-05-27 12:33:43 +00:00
{
size_t needed;
int mib[6];
char *buf, *next, *lim;
struct rt_msghdr *rtm;
1994-05-27 12:33:43 +00:00
mib[0] = CTL_NET;
mib[1] = PF_ROUTE;
mib[2] = 0;
mib[3] = 0;
mib[4] = NET_RT_DUMP;
mib[5] = 0;
if (sysctl(mib, 6, NULL, &needed, NULL, 0) < 0) {
err(1, "sysctl: net.route.0.0.dump estimate");
}
if ((buf = malloc(needed)) == 0) {
errx(2, "malloc(%lu)", (unsigned long)needed);
}
if (sysctl(mib, 6, buf, &needed, NULL, 0) < 0) {
err(1, "sysctl: net.route.0.0.dump");
}
1994-05-27 12:33:43 +00:00
lim = buf + needed;
for (next = buf; next < lim; next += rtm->rtm_msglen) {
rtm = (struct rt_msghdr *)next;
np_rtentry(rtm);
}
}
static void
2001-06-15 23:35:13 +00:00
np_rtentry(struct rt_msghdr *rtm)
1994-05-27 12:33:43 +00:00
{
struct sockaddr *sa = (struct sockaddr *)(rtm + 1);
1994-05-27 12:33:43 +00:00
#ifdef notdef
static int masks_done, banner_printed;
#endif
static int old_af;
int af1 = 0, interesting = RTF_UP | RTF_GATEWAY | RTF_HOST;
1994-05-27 12:33:43 +00:00
#ifdef notdef
/* for the moment, netmasks are skipped over */
if (!banner_printed) {
printf("Netmasks:\n");
banner_printed = 1;
}
if (masks_done == 0) {
if (rtm->rtm_addrs != RTA_DST ) {
masks_done = 1;
af1 = sa->sa_family;
1994-05-27 12:33:43 +00:00
}
} else
#endif
af1 = sa->sa_family;
if (af1 != old_af) {
pr_family(af1);
old_af = af1;
1994-05-27 12:33:43 +00:00
}
if (rtm->rtm_addrs == RTA_DST)
1996-01-14 23:33:13 +00:00
p_sockaddr(sa, NULL, 0, 36);
1994-05-27 12:33:43 +00:00
else {
1996-01-14 23:33:13 +00:00
p_sockaddr(sa, NULL, rtm->rtm_flags, 16);
2004-04-13 11:24:43 +00:00
sa = (struct sockaddr *)(SA_SIZE(sa) + (char *)sa);
1996-01-14 23:33:13 +00:00
p_sockaddr(sa, NULL, 0, 18);
1994-05-27 12:33:43 +00:00
}
p_flags(rtm->rtm_flags & interesting, "%-6.6s ");
putchar('\n');
}
static void
2001-06-15 23:35:13 +00:00
p_sockaddr(struct sockaddr *sa, struct sockaddr *mask, int flags, int width)
1994-05-27 12:33:43 +00:00
{
const char *cp;
cp = fmt_sockaddr(sa, mask, flags);
if (width < 0 )
printf("%s ", cp);
else {
if (numeric_addr)
printf("%-*s ", width, cp);
else
printf("%-*.*s ", width, width, cp);
}
}
static const char *
fmt_sockaddr(struct sockaddr *sa, struct sockaddr *mask, int flags)
{
static char workbuf[128];
const char *cp;
1994-05-27 12:33:43 +00:00
if (sa == NULL)
return ("null");
1994-05-27 12:33:43 +00:00
switch(sa->sa_family) {
case AF_INET:
{
struct sockaddr_in *sockin = (struct sockaddr_in *)sa;
1994-05-27 12:33:43 +00:00
if ((sockin->sin_addr.s_addr == INADDR_ANY) &&
mask &&
ntohl(((struct sockaddr_in *)mask)->sin_addr.s_addr)
==0L)
cp = "default" ;
1996-01-14 23:33:13 +00:00
else if (flags & RTF_HOST)
cp = routename(sockin->sin_addr.s_addr);
1996-01-14 23:33:13 +00:00
else if (mask)
cp = netname(sockin->sin_addr.s_addr,
1996-01-14 23:33:13 +00:00
ntohl(((struct sockaddr_in *)mask)
->sin_addr.s_addr));
else
cp = netname(sockin->sin_addr.s_addr, 0L);
1994-05-27 12:33:43 +00:00
break;
}
#ifdef INET6
case AF_INET6:
{
struct sockaddr_in6 *sa6 = (struct sockaddr_in6 *)sa;
in6_fillscopeid(sa6);
if (flags & RTF_HOST)
cp = routename6(sa6);
else if (mask)
cp = netname6(sa6,
&((struct sockaddr_in6 *)mask)->sin6_addr);
else {
cp = netname6(sa6, NULL);
}
break;
}
#endif /*INET6*/
case AF_IPX:
{
struct ipx_addr work = ((struct sockaddr_ipx *)sa)->sipx_addr;
if (ipx_nullnet(satoipx_addr(work)))
cp = "default";
else
cp = ipx_print(sa);
break;
}
case AF_APPLETALK:
{
if (!(flags & RTF_HOST) && mask)
cp = atalk_print2(sa,mask,9);
else
cp = atalk_print(sa,11);
break;
}
case AF_NETGRAPH:
{
strlcpy(workbuf, ((struct sockaddr_ng *)sa)->sg_data,
sizeof(workbuf));
cp = workbuf;
break;
}
1994-05-27 12:33:43 +00:00
case AF_LINK:
{
struct sockaddr_dl *sdl = (struct sockaddr_dl *)sa;
1994-05-27 12:33:43 +00:00
if (sdl->sdl_nlen == 0 && sdl->sdl_alen == 0 &&
sdl->sdl_slen == 0) {
1994-05-27 12:33:43 +00:00
(void) sprintf(workbuf, "link#%d", sdl->sdl_index);
cp = workbuf;
} else
switch (sdl->sdl_type) {
case IFT_ETHER:
case IFT_L2VLAN:
case IFT_BRIDGE:
if (sdl->sdl_alen == ETHER_ADDR_LEN) {
cp = ether_ntoa((struct ether_addr *)
2002-07-16 05:57:21 +00:00
(sdl->sdl_data + sdl->sdl_nlen));
break;
}
/* FALLTHROUGH */
default:
cp = link_ntoa(sdl);
break;
1994-05-27 12:33:43 +00:00
}
break;
}
default:
{
u_char *s = (u_char *)sa->sa_data, *slim;
char *cq, *cqlim;
1994-05-27 12:33:43 +00:00
cq = workbuf;
1994-05-27 12:33:43 +00:00
slim = sa->sa_len + (u_char *) sa;
cqlim = cq + sizeof(workbuf) - 6;
cq += sprintf(cq, "(%d)", sa->sa_family);
while (s < slim && cq < cqlim) {
cq += sprintf(cq, " %02x", *s++);
1994-05-27 12:33:43 +00:00
if (s < slim)
cq += sprintf(cq, "%02x", *s++);
1994-05-27 12:33:43 +00:00
}
cp = workbuf;
}
}
return (cp);
1994-05-27 12:33:43 +00:00
}
static void
p_flags(int f, const char *format)
1994-05-27 12:33:43 +00:00
{
printf(format, fmt_flags(f));
}
static const char *
fmt_flags(int f)
{
static char name[33];
char *flags;
struct bits *p = bits;
1994-05-27 12:33:43 +00:00
for (flags = name; p->b_mask; p++)
if (p->b_mask & f)
*flags++ = p->b_val;
*flags = '\0';
return (name);
1994-05-27 12:33:43 +00:00
}
static void
2001-06-15 23:35:13 +00:00
p_rtentry(struct rtentry *rt)
1994-05-27 12:33:43 +00:00
{
static struct ifnet ifnet, *lastif;
static char buffer[128];
static char prettyname[128];
1996-01-14 23:33:13 +00:00
struct sockaddr *sa;
sa_u addr, mask;
bzero(&addr, sizeof(addr));
if ((sa = kgetsa(rt_key(rt))))
bcopy(sa, &addr, sa->sa_len);
bzero(&mask, sizeof(mask));
if (rt_mask(rt) && (sa = kgetsa(rt_mask(rt))))
bcopy(sa, &mask, sa->sa_len);
p_sockaddr(&addr.u_sa, &mask.u_sa, rt->rt_flags, wid_dst);
p_sockaddr(kgetsa(rt->rt_gateway), NULL, RTF_HOST, wid_gw);
snprintf(buffer, sizeof(buffer), "%%-%d.%ds ", wid_flags, wid_flags);
p_flags(rt->rt_flags, buffer);
if (addr.u_sa.sa_family == AF_INET || Wflag) {
printf("%*d %*lu ", wid_refs, rt->rt_refcnt,
wid_use, rt->rt_use);
if (Wflag) {
if (rt->rt_rmx.rmx_mtu != 0)
printf("%*lu ", wid_mtu, rt->rt_rmx.rmx_mtu);
else
printf("%*s ", wid_mtu, "");
}
}
1994-05-27 12:33:43 +00:00
if (rt->rt_ifp) {
if (rt->rt_ifp != lastif) {
if (kget(rt->rt_ifp, ifnet) == 0)
strlcpy(prettyname, ifnet.if_xname,
sizeof(prettyname));
else
strlcpy(prettyname, "---", sizeof(prettyname));
1994-05-27 12:33:43 +00:00
lastif = rt->rt_ifp;
}
printf("%*.*s", wid_if, wid_if, prettyname);
if (rt->rt_rmx.rmx_expire) {
time_t expire_time;
if ((expire_time =
rt->rt_rmx.rmx_expire - uptime.tv_sec) > 0)
printf(" %*d", wid_expire, (int)expire_time);
}
if (rt->rt_nodes[0].rn_dupedkey)
printf(" =>");
1994-05-27 12:33:43 +00:00
}
putchar('\n');
}
char *
routename(in_addr_t in)
1994-05-27 12:33:43 +00:00
{
char *cp;
static char line[MAXHOSTNAMELEN];
1994-05-27 12:33:43 +00:00
struct hostent *hp;
1994-05-27 12:33:43 +00:00
cp = 0;
if (!numeric_addr) {
hp = gethostbyaddr(&in, sizeof (struct in_addr), AF_INET);
1994-05-27 12:33:43 +00:00
if (hp) {
cp = hp->h_name;
trimdomain(cp, strlen(cp));
1994-05-27 12:33:43 +00:00
}
}
if (cp) {
2008-10-17 21:14:50 +00:00
strlcpy(line, cp, sizeof(line));
} else {
#define C(x) ((x) & 0xff)
1994-05-27 12:33:43 +00:00
in = ntohl(in);
sprintf(line, "%u.%u.%u.%u",
1994-05-27 12:33:43 +00:00
C(in >> 24), C(in >> 16), C(in >> 8), C(in));
}
return (line);
}
#define NSHIFT(m) ( \
(m) == IN_CLASSA_NET ? IN_CLASSA_NSHIFT : \
(m) == IN_CLASSB_NET ? IN_CLASSB_NSHIFT : \
(m) == IN_CLASSC_NET ? IN_CLASSC_NSHIFT : \
0)
1996-01-14 23:33:13 +00:00
static void
domask(char *dst, in_addr_t addr __unused, u_long mask)
1996-01-14 23:33:13 +00:00
{
int b, i;
1996-01-14 23:33:13 +00:00
if (mask == 0 || (!numeric_addr && NSHIFT(mask) != 0)) {
1996-01-14 23:33:13 +00:00
*dst = '\0';
return;
}
i = 0;
for (b = 0; b < 32; b++)
if (mask & (1 << b)) {
int bb;
1996-01-14 23:33:13 +00:00
i = b;
for (bb = b+1; bb < 32; bb++)
if (!(mask & (1 << bb))) {
i = -1; /* noncontig */
break;
}
break;
}
if (i == -1)
sprintf(dst, "&0x%lx", mask);
else
sprintf(dst, "/%d", 32-i);
}
1994-05-27 12:33:43 +00:00
/*
* Return the name of the network whose address is given.
* The address is assumed to be that of a net or subnet, not a host.
*/
char *
netname(in_addr_t in, u_long mask)
1994-05-27 12:33:43 +00:00
{
char *cp = 0;
static char line[MAXHOSTNAMELEN];
1994-05-27 12:33:43 +00:00
struct netent *np = 0;
in_addr_t i;
1994-05-27 12:33:43 +00:00
i = ntohl(in);
if (!numeric_addr && i) {
np = getnetbyaddr(i >> NSHIFT(mask), AF_INET);
if (np != NULL) {
1994-05-27 12:33:43 +00:00
cp = np->n_name;
trimdomain(cp, strlen(cp));
}
1994-05-27 12:33:43 +00:00
}
if (cp != NULL) {
2008-10-17 21:14:50 +00:00
strlcpy(line, cp, sizeof(line));
} else {
inet_ntop(AF_INET, &in, line, sizeof(line) - 1);
}
domask(line + strlen(line), i, mask);
1994-05-27 12:33:43 +00:00
return (line);
}
#undef NSHIFT
#ifdef INET6
void
in6_fillscopeid(struct sockaddr_in6 *sa6)
{
#if defined(__KAME__)
/*
* XXX: This is a special workaround for KAME kernels.
* sin6_scope_id field of SA should be set in the future.
*/
if (IN6_IS_ADDR_LINKLOCAL(&sa6->sin6_addr) ||
IN6_IS_ADDR_MC_NODELOCAL(&sa6->sin6_addr) ||
IN6_IS_ADDR_MC_LINKLOCAL(&sa6->sin6_addr)) {
/* XXX: override is ok? */
sa6->sin6_scope_id =
ntohs(*(u_int16_t *)&sa6->sin6_addr.s6_addr[2]);
sa6->sin6_addr.s6_addr[2] = sa6->sin6_addr.s6_addr[3] = 0;
}
#endif
}
const char *
2001-06-15 23:35:13 +00:00
netname6(struct sockaddr_in6 *sa6, struct in6_addr *mask)
{
static char line[MAXHOSTNAMELEN];
u_char *p = (u_char *)mask;
u_char *lim;
int masklen, illegal = 0, flag = 0;
if (mask) {
for (masklen = 0, lim = p + 16; p < lim; p++) {
switch (*p) {
case 0xff:
masklen += 8;
break;
case 0xfe:
masklen += 7;
break;
case 0xfc:
masklen += 6;
break;
case 0xf8:
masklen += 5;
break;
case 0xf0:
masklen += 4;
break;
case 0xe0:
masklen += 3;
break;
case 0xc0:
masklen += 2;
break;
case 0x80:
masklen += 1;
break;
case 0x00:
break;
default:
illegal ++;
break;
}
}
if (illegal)
fprintf(stderr, "illegal prefixlen\n");
}
else
masklen = 128;
if (masklen == 0 && IN6_IS_ADDR_UNSPECIFIED(&sa6->sin6_addr))
return("default");
if (numeric_addr)
flag |= NI_NUMERICHOST;
getnameinfo((struct sockaddr *)sa6, sa6->sin6_len, line, sizeof(line),
NULL, 0, flag);
if (numeric_addr)
sprintf(&line[strlen(line)], "/%d", masklen);
return line;
}
char *
2001-06-15 23:35:13 +00:00
routename6(struct sockaddr_in6 *sa6)
{
static char line[MAXHOSTNAMELEN];
int flag = 0;
/* use local variable for safety */
struct sockaddr_in6 sa6_local;
sa6_local.sin6_family = AF_INET6;
sa6_local.sin6_len = sizeof(sa6_local);
sa6_local.sin6_addr = sa6->sin6_addr;
sa6_local.sin6_scope_id = sa6->sin6_scope_id;
if (numeric_addr)
flag |= NI_NUMERICHOST;
getnameinfo((struct sockaddr *)&sa6_local, sa6_local.sin6_len,
line, sizeof(line), NULL, 0, flag);
return line;
}
#endif /*INET6*/
1994-05-27 12:33:43 +00:00
/*
* Print routing statistics
*/
void
rt_stats(u_long rtsaddr, u_long rttaddr)
1994-05-27 12:33:43 +00:00
{
struct rtstat rtstat;
int rttrash;
1994-05-27 12:33:43 +00:00
if (rtsaddr == 0) {
1994-05-27 12:33:43 +00:00
printf("rtstat: symbol not in namelist\n");
return;
}
if (rttaddr == 0) {
printf("rttrash: symbol not in namelist\n");
return;
}
kread(rtsaddr, (char *)&rtstat, sizeof (rtstat));
kread(rttaddr, (char *)&rttrash, sizeof (rttrash));
1994-05-27 12:33:43 +00:00
printf("routing:\n");
#define p(f, m) if (rtstat.f || sflag <= 1) \
printf(m, rtstat.f, plural(rtstat.f))
p(rts_badredirect, "\t%hu bad routing redirect%s\n");
p(rts_dynamic, "\t%hu dynamically created route%s\n");
p(rts_newgateway, "\t%hu new gateway%s due to redirects\n");
p(rts_unreach, "\t%hu destination%s found unreachable\n");
p(rts_wildcard, "\t%hu use%s of a wildcard route\n");
#undef p
if (rttrash || sflag <= 1)
printf("\t%u route%s not in table but not freed\n",
rttrash, plural(rttrash));
1994-05-27 12:33:43 +00:00
}
char *
2001-06-15 23:35:13 +00:00
ipx_print(struct sockaddr *sa)
{
u_short port;
struct servent *sp = 0;
const char *net = "", *host = "";
char *p;
u_char *q;
struct ipx_addr work = ((struct sockaddr_ipx *)sa)->sipx_addr;
static char mybuf[50];
char cport[10], chost[15], cnet[15];
port = ntohs(work.x_port);
if (ipx_nullnet(work) && ipx_nullhost(work)) {
if (port) {
if (sp)
sprintf(mybuf, "*.%s", sp->s_name);
else
sprintf(mybuf, "*.%x", port);
} else
sprintf(mybuf, "*.*");
return (mybuf);
}
if (ipx_wildnet(work))
net = "any";
else if (ipx_nullnet(work))
net = "*";
else {
q = work.x_net.c_net;
sprintf(cnet, "%02x%02x%02x%02x",
q[0], q[1], q[2], q[3]);
for (p = cnet; *p == '0' && p < cnet + 8; p++)
continue;
net = p;
}
if (ipx_wildhost(work))
host = "any";
else if (ipx_nullhost(work))
host = "*";
else {
q = work.x_host.c_host;
sprintf(chost, "%02x%02x%02x%02x%02x%02x",
q[0], q[1], q[2], q[3], q[4], q[5]);
for (p = chost; *p == '0' && p < chost + 12; p++)
continue;
host = p;
}
if (port) {
if (strcmp(host, "*") == 0)
host = "";
if (sp)
snprintf(cport, sizeof(cport),
"%s%s", *host ? "." : "", sp->s_name);
else
snprintf(cport, sizeof(cport),
"%s%x", *host ? "." : "", port);
} else
*cport = 0;
snprintf(mybuf, sizeof(mybuf), "%s.%s%s", net, host, cport);
return(mybuf);
}
char *
2001-06-15 23:35:13 +00:00
ipx_phost(struct sockaddr *sa)
{
struct sockaddr_ipx *sipx = (struct sockaddr_ipx *)sa;
struct sockaddr_ipx work;
static union ipx_net ipx_zeronet;
char *p;
struct ipx_addr in;
work = *sipx;
in = work.sipx_addr;
work.sipx_addr.x_port = 0;
work.sipx_addr.x_net = ipx_zeronet;
p = ipx_print((struct sockaddr *)&work);
if (strncmp("*.", p, 2) == 0) p += 2;
return(p);
}
1994-05-27 12:33:43 +00:00
void
2001-06-15 23:35:13 +00:00
upHex(char *p0)
1994-05-27 12:33:43 +00:00
{
char *p = p0;
1994-05-27 12:33:43 +00:00
for (; *p; p++)
switch (*p) {
case 'a':
case 'b':
case 'c':
case 'd':
case 'e':
case 'f':
*p += ('A' - 'a');
break;
}
1994-05-27 12:33:43 +00:00
}