2000-07-04 16:35:15 +00:00
|
|
|
/* $FreeBSD$ */
|
2002-04-19 04:46:24 +00:00
|
|
|
/* $KAME: if_stf.c,v 1.73 2001/12/03 11:08:30 keiichi Exp $ */
|
2000-07-04 16:35:15 +00:00
|
|
|
|
2005-01-07 01:45:51 +00:00
|
|
|
/*-
|
2000-07-04 16:35:15 +00:00
|
|
|
* Copyright (C) 2000 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.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
/*
|
2001-06-11 12:39:29 +00:00
|
|
|
* 6to4 interface, based on RFC3056.
|
2000-07-04 16:35:15 +00:00
|
|
|
*
|
|
|
|
|
* 6to4 interface is NOT capable of link-layer (I mean, IPv4) multicasting.
|
|
|
|
|
* There is no address mapping defined from IPv6 multicast address to IPv4
|
|
|
|
|
* address. Therefore, we do not have IFF_MULTICAST on the interface.
|
|
|
|
|
*
|
|
|
|
|
* Due to the lack of address mapping for link-local addresses, we cannot
|
|
|
|
|
* throw packets toward link-local addresses (fe80::x). Also, we cannot throw
|
|
|
|
|
* packets to link-local multicast addresses (ff02::x).
|
|
|
|
|
*
|
|
|
|
|
* Here are interesting symptoms due to the lack of link-local address:
|
|
|
|
|
*
|
|
|
|
|
* Unicast routing exchange:
|
|
|
|
|
* - RIPng: Impossible. Uses link-local multicast packet toward ff02::9,
|
|
|
|
|
* and link-local addresses as nexthop.
|
|
|
|
|
* - OSPFv6: Impossible. OSPFv6 assumes that there's link-local address
|
|
|
|
|
* assigned to the link, and makes use of them. Also, HELLO packets use
|
|
|
|
|
* link-local multicast addresses (ff02::5 and ff02::6).
|
|
|
|
|
* - BGP4+: Maybe. You can only use global address as nexthop, and global
|
|
|
|
|
* address as TCP endpoint address.
|
|
|
|
|
*
|
|
|
|
|
* Multicast routing protocols:
|
|
|
|
|
* - PIM: Hello packet cannot be used to discover adjacent PIM routers.
|
|
|
|
|
* Adjacent PIM routers must be configured manually (is it really spec-wise
|
|
|
|
|
* correct thing to do?).
|
|
|
|
|
*
|
|
|
|
|
* ICMPv6:
|
|
|
|
|
* - Redirects cannot be used due to the lack of link-local address.
|
|
|
|
|
*
|
2001-06-11 12:39:29 +00:00
|
|
|
* stf interface does not have, and will not need, a link-local address.
|
|
|
|
|
* It seems to have no real benefit and does not help the above symptoms much.
|
|
|
|
|
* Even if we assign link-locals to interface, we cannot really
|
|
|
|
|
* use link-local unicast/multicast on top of 6to4 cloud (since there's no
|
|
|
|
|
* encapsulation defined for link-local address), and the above analysis does
|
|
|
|
|
* not change. RFC3056 does not mandate the assignment of link-local address
|
|
|
|
|
* either.
|
2000-07-04 16:35:15 +00:00
|
|
|
*
|
|
|
|
|
* 6to4 interface has security issues. Refer to
|
|
|
|
|
* http://playground.iijlab.net/i-d/draft-itojun-ipv6-transition-abuse-00.txt
|
|
|
|
|
* for details. The code tries to filter out some of malicious packets.
|
|
|
|
|
* Note that there is no way to be 100% secure.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
#include "opt_inet.h"
|
|
|
|
|
#include "opt_inet6.h"
|
2002-10-20 22:39:55 +00:00
|
|
|
#include "opt_mac.h"
|
2000-07-04 16:35:15 +00:00
|
|
|
|
|
|
|
|
#include <sys/param.h>
|
|
|
|
|
#include <sys/systm.h>
|
|
|
|
|
#include <sys/socket.h>
|
|
|
|
|
#include <sys/sockio.h>
|
|
|
|
|
#include <sys/mbuf.h>
|
|
|
|
|
#include <sys/errno.h>
|
|
|
|
|
#include <sys/kernel.h>
|
2004-05-30 20:27:19 +00:00
|
|
|
#include <sys/module.h>
|
2001-09-19 00:13:00 +00:00
|
|
|
#include <sys/protosw.h>
|
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
Reviewed by: several including rwatson, bz and mlair (parts each)
Obtained from: Ironport systems/Cisco
2008-05-09 23:03:00 +00:00
|
|
|
#include <sys/proc.h>
|
2001-09-19 00:13:00 +00:00
|
|
|
#include <sys/queue.h>
|
2008-09-25 12:35:01 +00:00
|
|
|
#include <sys/sysctl.h>
|
2000-07-04 16:35:15 +00:00
|
|
|
#include <machine/cpu.h>
|
|
|
|
|
|
|
|
|
|
#include <sys/malloc.h>
|
Commit step 1 of the vimage project, (network stack)
virtualization work done by Marko Zec (zec@).
This is the first in a series of commits over the course
of the next few weeks.
Mark all uses of global variables to be virtualized
with a V_ prefix.
Use macros to map them back to their global names for
now, so this is a NOP change only.
We hope to have caught at least 85-90% of what is needed
so we do not invalidate a lot of outstanding patches again.
Obtained from: //depot/projects/vimage-commit2/...
Reviewed by: brooks, des, ed, mav, julian,
jamie, kris, rwatson, zec, ...
(various people I forgot, different versions)
md5 (with a bit of help)
Sponsored by: NLnet Foundation, The FreeBSD Foundation
X-MFC after: never
V_Commit_Message_Reviewed_By: more people than the patch
2008-08-17 23:27:27 +00:00
|
|
|
#include <sys/vimage.h>
|
2000-07-04 16:35:15 +00:00
|
|
|
|
|
|
|
|
#include <net/if.h>
|
Major overhaul of pseudo-interface cloning. Highlights include:
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-22 20:13:25 +00:00
|
|
|
#include <net/if_clone.h>
|
2000-07-04 16:35:15 +00:00
|
|
|
#include <net/route.h>
|
|
|
|
|
#include <net/netisr.h>
|
|
|
|
|
#include <net/if_types.h>
|
|
|
|
|
#include <net/if_stf.h>
|
|
|
|
|
|
|
|
|
|
#include <netinet/in.h>
|
|
|
|
|
#include <netinet/in_systm.h>
|
|
|
|
|
#include <netinet/ip.h>
|
|
|
|
|
#include <netinet/ip_var.h>
|
|
|
|
|
#include <netinet/in_var.h>
|
2008-12-02 21:37:28 +00:00
|
|
|
#include <netinet/vinet.h>
|
2000-07-04 16:35:15 +00:00
|
|
|
|
|
|
|
|
#include <netinet/ip6.h>
|
|
|
|
|
#include <netinet6/ip6_var.h>
|
|
|
|
|
#include <netinet6/in6_var.h>
|
|
|
|
|
#include <netinet/ip_ecn.h>
|
|
|
|
|
|
|
|
|
|
#include <netinet/ip_encap.h>
|
|
|
|
|
|
|
|
|
|
#include <machine/stdarg.h>
|
|
|
|
|
|
|
|
|
|
#include <net/bpf.h>
|
|
|
|
|
|
2006-10-22 11:52:19 +00:00
|
|
|
#include <security/mac/mac_framework.h>
|
|
|
|
|
|
2008-09-25 12:35:01 +00:00
|
|
|
SYSCTL_DECL(_net_link);
|
|
|
|
|
SYSCTL_NODE(_net_link, IFT_STF, stf, CTLFLAG_RW, 0, "6to4 Interface");
|
|
|
|
|
|
|
|
|
|
static int stf_route_cache = 1;
|
|
|
|
|
SYSCTL_INT(_net_link_stf, OID_AUTO, route_cache, CTLFLAG_RW,
|
|
|
|
|
&stf_route_cache, 0, "Caching of IPv4 routes for 6to4 Output");
|
|
|
|
|
|
2001-09-19 00:13:00 +00:00
|
|
|
#define STFNAME "stf"
|
Major overhaul of pseudo-interface cloning. Highlights include:
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-22 20:13:25 +00:00
|
|
|
#define STFUNIT 0
|
2001-09-19 00:13:00 +00:00
|
|
|
|
2000-07-04 16:35:15 +00:00
|
|
|
#define IN6_IS_ADDR_6TO4(x) (ntohs((x)->s6_addr16[0]) == 0x2002)
|
2003-01-15 20:09:52 +00:00
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* XXX: Return a pointer with 16-bit aligned. Don't cast it to
|
|
|
|
|
* struct in_addr *; use bcopy() instead.
|
|
|
|
|
*/
|
|
|
|
|
#define GET_V4(x) ((caddr_t)(&(x)->s6_addr16[1]))
|
2000-07-04 16:35:15 +00:00
|
|
|
|
|
|
|
|
struct stf_softc {
|
2005-06-10 16:49:24 +00:00
|
|
|
struct ifnet *sc_ifp;
|
2000-07-04 16:35:15 +00:00
|
|
|
union {
|
|
|
|
|
struct route __sc_ro4;
|
|
|
|
|
struct route_in6 __sc_ro6; /* just for safety */
|
|
|
|
|
} __sc_ro46;
|
|
|
|
|
#define sc_ro __sc_ro46.__sc_ro4
|
2008-09-25 12:35:01 +00:00
|
|
|
struct mtx sc_ro_mtx;
|
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
Reviewed by: several including rwatson, bz and mlair (parts each)
Obtained from: Ironport systems/Cisco
2008-05-09 23:03:00 +00:00
|
|
|
u_int sc_fibnum;
|
2000-07-04 16:35:15 +00:00
|
|
|
const struct encaptab *encap_cookie;
|
|
|
|
|
};
|
2005-06-10 16:49:24 +00:00
|
|
|
#define STF2IFP(sc) ((sc)->sc_ifp)
|
2000-07-04 16:35:15 +00:00
|
|
|
|
2004-03-09 20:29:19 +00:00
|
|
|
/*
|
2008-09-25 12:35:01 +00:00
|
|
|
* Note that mutable fields in the softc are not currently locked.
|
|
|
|
|
* We do lock sc_ro in stf_output though.
|
2004-03-09 20:29:19 +00:00
|
|
|
*/
|
2001-09-19 00:13:00 +00:00
|
|
|
static MALLOC_DEFINE(M_STF, STFNAME, "6to4 Tunnel Interface");
|
2004-03-07 05:15:42 +00:00
|
|
|
static const int ip_stf_ttl = 40;
|
2000-07-04 16:35:15 +00:00
|
|
|
|
2001-07-02 21:02:09 +00:00
|
|
|
extern struct domain inetdomain;
|
2005-11-09 13:29:16 +00:00
|
|
|
struct protosw in_stf_protosw = {
|
|
|
|
|
.pr_type = SOCK_RAW,
|
|
|
|
|
.pr_domain = &inetdomain,
|
|
|
|
|
.pr_protocol = IPPROTO_IPV6,
|
|
|
|
|
.pr_flags = PR_ATOMIC|PR_ADDR,
|
|
|
|
|
.pr_input = in_stf_input,
|
|
|
|
|
.pr_output = (pr_output_t *)rip_output,
|
|
|
|
|
.pr_ctloutput = rip_ctloutput,
|
|
|
|
|
.pr_usrreqs = &rip_usrreqs
|
2001-07-02 21:02:09 +00:00
|
|
|
};
|
2000-07-04 16:35:15 +00:00
|
|
|
|
Major overhaul of pseudo-interface cloning. Highlights include:
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-22 20:13:25 +00:00
|
|
|
static char *stfnames[] = {"stf0", "stf", "6to4", NULL};
|
|
|
|
|
|
2002-03-19 21:54:18 +00:00
|
|
|
static int stfmodevent(module_t, int, void *);
|
|
|
|
|
static int stf_encapcheck(const struct mbuf *, int, int, void *);
|
|
|
|
|
static struct in6_ifaddr *stf_getsrcifa6(struct ifnet *);
|
|
|
|
|
static int stf_output(struct ifnet *, struct mbuf *, struct sockaddr *,
|
|
|
|
|
struct rtentry *);
|
2002-09-17 10:45:51 +00:00
|
|
|
static int isrfc1918addr(struct in_addr *);
|
2002-03-19 21:54:18 +00:00
|
|
|
static int stf_checkaddr4(struct stf_softc *, struct in_addr *,
|
|
|
|
|
struct ifnet *);
|
|
|
|
|
static int stf_checkaddr6(struct stf_softc *, struct in6_addr *,
|
|
|
|
|
struct ifnet *);
|
|
|
|
|
static void stf_rtrequest(int, struct rtentry *, struct rt_addrinfo *);
|
|
|
|
|
static int stf_ioctl(struct ifnet *, u_long, caddr_t);
|
|
|
|
|
|
Major overhaul of pseudo-interface cloning. Highlights include:
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-22 20:13:25 +00:00
|
|
|
static int stf_clone_match(struct if_clone *, const char *);
|
2006-07-09 06:04:01 +00:00
|
|
|
static int stf_clone_create(struct if_clone *, char *, size_t, caddr_t);
|
Major overhaul of pseudo-interface cloning. Highlights include:
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-22 20:13:25 +00:00
|
|
|
static int stf_clone_destroy(struct if_clone *, struct ifnet *);
|
|
|
|
|
struct if_clone stf_cloner = IFC_CLONE_INITIALIZER(STFNAME, NULL, 0,
|
|
|
|
|
NULL, stf_clone_match, stf_clone_create, stf_clone_destroy);
|
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
stf_clone_match(struct if_clone *ifc, const char *name)
|
|
|
|
|
{
|
|
|
|
|
int i;
|
2001-09-19 00:13:00 +00:00
|
|
|
|
Major overhaul of pseudo-interface cloning. Highlights include:
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-22 20:13:25 +00:00
|
|
|
for(i = 0; stfnames[i] != NULL; i++) {
|
|
|
|
|
if (strcmp(stfnames[i], name) == 0)
|
|
|
|
|
return (1);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return (0);
|
|
|
|
|
}
|
2001-09-19 00:13:00 +00:00
|
|
|
|
2004-04-14 00:57:49 +00:00
|
|
|
static int
|
2006-07-09 06:04:01 +00:00
|
|
|
stf_clone_create(struct if_clone *ifc, char *name, size_t len, caddr_t params)
|
2001-09-19 00:13:00 +00:00
|
|
|
{
|
Major overhaul of pseudo-interface cloning. Highlights include:
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-22 20:13:25 +00:00
|
|
|
int err, unit;
|
2001-09-19 00:13:00 +00:00
|
|
|
struct stf_softc *sc;
|
2004-04-19 05:06:27 +00:00
|
|
|
struct ifnet *ifp;
|
2001-09-19 00:13:00 +00:00
|
|
|
|
Major overhaul of pseudo-interface cloning. Highlights include:
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-22 20:13:25 +00:00
|
|
|
/*
|
|
|
|
|
* We can only have one unit, but since unit allocation is
|
|
|
|
|
* already locked, we use it to keep from allocating extra
|
|
|
|
|
* interfaces.
|
|
|
|
|
*/
|
|
|
|
|
unit = STFUNIT;
|
|
|
|
|
err = ifc_alloc_unit(ifc, &unit);
|
|
|
|
|
if (err != 0)
|
|
|
|
|
return (err);
|
|
|
|
|
|
2003-02-19 05:47:46 +00:00
|
|
|
sc = malloc(sizeof(struct stf_softc), M_STF, M_WAITOK | M_ZERO);
|
2005-06-13 17:17:07 +00:00
|
|
|
ifp = STF2IFP(sc) = if_alloc(IFT_STF);
|
2005-06-10 16:49:24 +00:00
|
|
|
if (ifp == NULL) {
|
|
|
|
|
free(sc, M_STF);
|
|
|
|
|
ifc_free_unit(ifc, unit);
|
|
|
|
|
return (ENOSPC);
|
|
|
|
|
}
|
2005-06-13 17:17:07 +00:00
|
|
|
ifp->if_softc = sc;
|
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
Reviewed by: several including rwatson, bz and mlair (parts each)
Obtained from: Ironport systems/Cisco
2008-05-09 23:03:00 +00:00
|
|
|
sc->sc_fibnum = curthread->td_proc->p_fibnum;
|
2005-06-13 17:17:07 +00:00
|
|
|
|
Major overhaul of pseudo-interface cloning. Highlights include:
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-22 20:13:25 +00:00
|
|
|
/*
|
|
|
|
|
* Set the name manually rather then using if_initname because
|
|
|
|
|
* we don't conform to the default naming convention for interfaces.
|
|
|
|
|
*/
|
|
|
|
|
strlcpy(ifp->if_xname, name, IFNAMSIZ);
|
|
|
|
|
ifp->if_dname = ifc->ifc_name;
|
|
|
|
|
ifp->if_dunit = IF_DUNIT_NONE;
|
2001-09-19 00:13:00 +00:00
|
|
|
|
2008-09-25 12:35:01 +00:00
|
|
|
mtx_init(&(sc)->sc_ro_mtx, "stf ro", NULL, MTX_DEF);
|
2001-09-19 00:13:00 +00:00
|
|
|
sc->encap_cookie = encap_attach_func(AF_INET, IPPROTO_IPV6,
|
|
|
|
|
stf_encapcheck, &in_stf_protosw, sc);
|
|
|
|
|
if (sc->encap_cookie == NULL) {
|
2004-04-19 05:06:27 +00:00
|
|
|
if_printf(ifp, "attach failed\n");
|
2001-09-19 00:13:00 +00:00
|
|
|
free(sc, M_STF);
|
Major overhaul of pseudo-interface cloning. Highlights include:
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-22 20:13:25 +00:00
|
|
|
ifc_free_unit(ifc, unit);
|
2001-09-19 00:13:00 +00:00
|
|
|
return (ENOMEM);
|
|
|
|
|
}
|
|
|
|
|
|
2004-04-19 05:06:27 +00:00
|
|
|
ifp->if_mtu = IPV6_MMTU;
|
|
|
|
|
ifp->if_ioctl = stf_ioctl;
|
|
|
|
|
ifp->if_output = stf_output;
|
|
|
|
|
ifp->if_snd.ifq_maxlen = IFQ_MAXLEN;
|
|
|
|
|
if_attach(ifp);
|
2005-06-26 18:11:11 +00:00
|
|
|
bpfattach(ifp, DLT_NULL, sizeof(u_int32_t));
|
2001-09-19 00:13:00 +00:00
|
|
|
return (0);
|
|
|
|
|
}
|
|
|
|
|
|
Major overhaul of pseudo-interface cloning. Highlights include:
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-22 20:13:25 +00:00
|
|
|
static int
|
|
|
|
|
stf_clone_destroy(struct if_clone *ifc, struct ifnet *ifp)
|
2001-09-19 00:13:00 +00:00
|
|
|
{
|
2005-06-10 16:49:24 +00:00
|
|
|
struct stf_softc *sc = ifp->if_softc;
|
2005-10-12 19:52:16 +00:00
|
|
|
int err;
|
2001-09-19 00:13:00 +00:00
|
|
|
|
2005-10-12 19:52:16 +00:00
|
|
|
err = encap_detach(sc->encap_cookie);
|
|
|
|
|
KASSERT(err == 0, ("Unexpected error detaching encap_cookie"));
|
2008-09-25 12:35:01 +00:00
|
|
|
mtx_destroy(&(sc)->sc_ro_mtx);
|
2005-10-12 19:52:16 +00:00
|
|
|
bpfdetach(ifp);
|
|
|
|
|
if_detach(ifp);
|
|
|
|
|
if_free(ifp);
|
|
|
|
|
|
|
|
|
|
free(sc, M_STF);
|
Major overhaul of pseudo-interface cloning. Highlights include:
- Split the code out into if_clone.[ch].
- Locked struct if_clone. [1]
- Add a per-cloner match function rather then simply matching names of
the form <name><unit> and <name>.
- Use the match function to allow creation of <interface>.<tag>
vlan interfaces. The old way is preserved unchanged!
- Also the match function to allow creation of stf(4) interfaces named
stf0, stf, or 6to4. This is the only major user visible change in
that "ifconfig stf" creates the interface stf rather then stf0 and
does not print "stf0" to stdout.
- Allow destroy functions to fail so they can refuse to delete
interfaces. Currently, we forbid the deletion of interfaces which
were created in the init function, particularly lo0, pflog0, and
pfsync0. In the case of lo0 this was a panic implementation so it
does not count as a user visiable change. :-)
- Since most interfaces do not need the new functionality, an family of
wrapper functions, ifc_simple_*(), were created to wrap old style
cloner functions.
- The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
instead.
Submitted by: Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by: andre, mlaier
Discussed on: net
2004-06-22 20:13:25 +00:00
|
|
|
ifc_free_unit(ifc, STFUNIT);
|
|
|
|
|
|
|
|
|
|
return (0);
|
2001-09-19 00:13:00 +00:00
|
|
|
}
|
|
|
|
|
|
2001-07-02 21:02:09 +00:00
|
|
|
static int
|
|
|
|
|
stfmodevent(mod, type, data)
|
|
|
|
|
module_t mod;
|
|
|
|
|
int type;
|
|
|
|
|
void *data;
|
2000-07-04 16:35:15 +00:00
|
|
|
{
|
|
|
|
|
|
2001-07-02 21:02:09 +00:00
|
|
|
switch (type) {
|
|
|
|
|
case MOD_LOAD:
|
2001-09-19 00:13:00 +00:00
|
|
|
if_clone_attach(&stf_cloner);
|
2001-07-02 21:02:09 +00:00
|
|
|
break;
|
|
|
|
|
case MOD_UNLOAD:
|
2001-09-19 00:13:00 +00:00
|
|
|
if_clone_detach(&stf_cloner);
|
2001-07-02 21:02:09 +00:00
|
|
|
break;
|
2004-07-15 08:26:07 +00:00
|
|
|
default:
|
|
|
|
|
return (EOPNOTSUPP);
|
2000-07-04 16:35:15 +00:00
|
|
|
}
|
2001-07-02 21:02:09 +00:00
|
|
|
|
|
|
|
|
return (0);
|
2000-07-04 16:35:15 +00:00
|
|
|
}
|
|
|
|
|
|
2001-07-02 21:02:09 +00:00
|
|
|
static moduledata_t stf_mod = {
|
|
|
|
|
"if_stf",
|
|
|
|
|
stfmodevent,
|
|
|
|
|
0
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
DECLARE_MODULE(if_stf, stf_mod, SI_SUB_PSEUDO, SI_ORDER_ANY);
|
2000-07-04 16:35:15 +00:00
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
stf_encapcheck(m, off, proto, arg)
|
|
|
|
|
const struct mbuf *m;
|
|
|
|
|
int off;
|
|
|
|
|
int proto;
|
|
|
|
|
void *arg;
|
|
|
|
|
{
|
|
|
|
|
struct ip ip;
|
|
|
|
|
struct in6_ifaddr *ia6;
|
|
|
|
|
struct stf_softc *sc;
|
2003-01-05 14:03:26 +00:00
|
|
|
struct in_addr a, b, mask;
|
2000-07-04 16:35:15 +00:00
|
|
|
|
|
|
|
|
sc = (struct stf_softc *)arg;
|
|
|
|
|
if (sc == NULL)
|
|
|
|
|
return 0;
|
|
|
|
|
|
2005-06-10 16:49:24 +00:00
|
|
|
if ((STF2IFP(sc)->if_flags & IFF_UP) == 0)
|
2000-07-04 16:35:15 +00:00
|
|
|
return 0;
|
|
|
|
|
|
2001-06-11 12:39:29 +00:00
|
|
|
/* IFF_LINK0 means "no decapsulation" */
|
2005-06-10 16:49:24 +00:00
|
|
|
if ((STF2IFP(sc)->if_flags & IFF_LINK0) != 0)
|
2001-06-11 12:39:29 +00:00
|
|
|
return 0;
|
|
|
|
|
|
2000-07-04 16:35:15 +00:00
|
|
|
if (proto != IPPROTO_IPV6)
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
|
|
/* LINTED const cast */
|
2002-02-28 00:09:17 +00:00
|
|
|
m_copydata((struct mbuf *)(uintptr_t)m, 0, sizeof(ip), (caddr_t)&ip);
|
2000-07-04 16:35:15 +00:00
|
|
|
|
|
|
|
|
if (ip.ip_v != 4)
|
|
|
|
|
return 0;
|
|
|
|
|
|
2005-06-10 16:49:24 +00:00
|
|
|
ia6 = stf_getsrcifa6(STF2IFP(sc));
|
2000-07-04 16:35:15 +00:00
|
|
|
if (ia6 == NULL)
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* check if IPv4 dst matches the IPv4 address derived from the
|
|
|
|
|
* local 6to4 address.
|
|
|
|
|
* success on: dst = 10.1.1.1, ia6->ia_addr = 2002:0a01:0101:...
|
|
|
|
|
*/
|
|
|
|
|
if (bcmp(GET_V4(&ia6->ia_addr.sin6_addr), &ip.ip_dst,
|
|
|
|
|
sizeof(ip.ip_dst)) != 0)
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* check if IPv4 src matches the IPv4 address derived from the
|
|
|
|
|
* local 6to4 address masked by prefixmask.
|
|
|
|
|
* success on: src = 10.1.1.1, ia6->ia_addr = 2002:0a00:.../24
|
|
|
|
|
* fail on: src = 10.1.1.1, ia6->ia_addr = 2002:0b00:.../24
|
|
|
|
|
*/
|
|
|
|
|
bzero(&a, sizeof(a));
|
2003-01-05 14:03:26 +00:00
|
|
|
bcopy(GET_V4(&ia6->ia_addr.sin6_addr), &a, sizeof(a));
|
|
|
|
|
bcopy(GET_V4(&ia6->ia_prefixmask.sin6_addr), &mask, sizeof(mask));
|
|
|
|
|
a.s_addr &= mask.s_addr;
|
2000-07-04 16:35:15 +00:00
|
|
|
b = ip.ip_src;
|
2003-01-05 14:03:26 +00:00
|
|
|
b.s_addr &= mask.s_addr;
|
2000-07-04 16:35:15 +00:00
|
|
|
if (a.s_addr != b.s_addr)
|
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
|
|
/* stf interface makes single side match only */
|
|
|
|
|
return 32;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static struct in6_ifaddr *
|
|
|
|
|
stf_getsrcifa6(ifp)
|
|
|
|
|
struct ifnet *ifp;
|
|
|
|
|
{
|
Step 1.5 of importing the network stack virtualization infrastructure
from the vimage project, as per plan established at devsummit 08/08:
http://wiki.freebsd.org/Image/Notes200808DevSummit
Introduce INIT_VNET_*() initializer macros, VNET_FOREACH() iterator
macros, and CURVNET_SET() context setting macros, all currently
resolving to NOPs.
Prepare for virtualization of selected SYSCTL objects by introducing a
family of SYSCTL_V_*() macros, currently resolving to their global
counterparts, i.e. SYSCTL_V_INT() == SYSCTL_INT().
Move selected #defines from sys/sys/vimage.h to newly introduced header
files specific to virtualized subsystems (sys/net/vnet.h,
sys/netinet/vinet.h etc.).
All the changes are verified to have zero functional impact at this
point in time by doing MD5 comparision between pre- and post-change
object files(*).
(*) netipsec/keysock.c did not validate depending on compile time options.
Implemented by: julian, bz, brooks, zec
Reviewed by: julian, bz, brooks, kris, rwatson, ...
Approved by: julian (mentor)
Obtained from: //depot/projects/vimage-commit2/...
X-MFC after: never
Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-10-02 15:37:58 +00:00
|
|
|
INIT_VNET_INET(ifp->if_vnet);
|
2000-07-04 16:35:15 +00:00
|
|
|
struct ifaddr *ia;
|
|
|
|
|
struct in_ifaddr *ia4;
|
|
|
|
|
struct sockaddr_in6 *sin6;
|
|
|
|
|
struct in_addr in;
|
|
|
|
|
|
2006-06-29 17:31:43 +00:00
|
|
|
TAILQ_FOREACH(ia, &ifp->if_addrlist, ifa_list) {
|
2000-07-04 16:35:15 +00:00
|
|
|
if (ia->ifa_addr->sa_family != AF_INET6)
|
|
|
|
|
continue;
|
|
|
|
|
sin6 = (struct sockaddr_in6 *)ia->ifa_addr;
|
|
|
|
|
if (!IN6_IS_ADDR_6TO4(&sin6->sin6_addr))
|
|
|
|
|
continue;
|
|
|
|
|
|
|
|
|
|
bcopy(GET_V4(&sin6->sin6_addr), &in, sizeof(in));
|
2001-09-29 05:02:36 +00:00
|
|
|
LIST_FOREACH(ia4, INADDR_HASH(in.s_addr), ia_hash)
|
2000-07-04 16:35:15 +00:00
|
|
|
if (ia4->ia_addr.sin_addr.s_addr == in.s_addr)
|
|
|
|
|
break;
|
|
|
|
|
if (ia4 == NULL)
|
|
|
|
|
continue;
|
|
|
|
|
|
|
|
|
|
return (struct in6_ifaddr *)ia;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
stf_output(ifp, m, dst, rt)
|
|
|
|
|
struct ifnet *ifp;
|
|
|
|
|
struct mbuf *m;
|
|
|
|
|
struct sockaddr *dst;
|
|
|
|
|
struct rtentry *rt;
|
|
|
|
|
{
|
|
|
|
|
struct stf_softc *sc;
|
|
|
|
|
struct sockaddr_in6 *dst6;
|
2008-09-25 12:35:01 +00:00
|
|
|
struct route *cached_route;
|
2003-01-15 20:09:52 +00:00
|
|
|
struct in_addr in4;
|
|
|
|
|
caddr_t ptr;
|
2000-07-04 16:35:15 +00:00
|
|
|
struct sockaddr_in *dst4;
|
|
|
|
|
u_int8_t tos;
|
|
|
|
|
struct ip *ip;
|
|
|
|
|
struct ip6_hdr *ip6;
|
|
|
|
|
struct in6_ifaddr *ia6;
|
2005-06-26 18:11:11 +00:00
|
|
|
u_int32_t af;
|
2002-10-20 22:39:55 +00:00
|
|
|
int error;
|
|
|
|
|
|
2008-09-25 12:35:01 +00:00
|
|
|
#ifdef MAC
|
2007-10-24 19:04:04 +00:00
|
|
|
error = mac_ifnet_check_transmit(ifp, m);
|
2002-10-20 22:39:55 +00:00
|
|
|
if (error) {
|
|
|
|
|
m_freem(m);
|
|
|
|
|
return (error);
|
|
|
|
|
}
|
|
|
|
|
#endif
|
2000-07-04 16:35:15 +00:00
|
|
|
|
2005-06-10 16:49:24 +00:00
|
|
|
sc = ifp->if_softc;
|
2000-07-04 16:35:15 +00:00
|
|
|
dst6 = (struct sockaddr_in6 *)dst;
|
|
|
|
|
|
|
|
|
|
/* just in case */
|
|
|
|
|
if ((ifp->if_flags & IFF_UP) == 0) {
|
|
|
|
|
m_freem(m);
|
2002-09-17 14:25:19 +00:00
|
|
|
ifp->if_oerrors++;
|
2000-07-04 16:35:15 +00:00
|
|
|
return ENETDOWN;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* If we don't have an ip4 address that match my inner ip6 address,
|
|
|
|
|
* we shouldn't generate output. Without this check, we'll end up
|
|
|
|
|
* using wrong IPv4 source.
|
|
|
|
|
*/
|
|
|
|
|
ia6 = stf_getsrcifa6(ifp);
|
|
|
|
|
if (ia6 == NULL) {
|
|
|
|
|
m_freem(m);
|
2002-09-17 14:25:19 +00:00
|
|
|
ifp->if_oerrors++;
|
2000-07-04 16:35:15 +00:00
|
|
|
return ENETDOWN;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (m->m_len < sizeof(*ip6)) {
|
|
|
|
|
m = m_pullup(m, sizeof(*ip6));
|
2002-09-17 14:25:19 +00:00
|
|
|
if (!m) {
|
|
|
|
|
ifp->if_oerrors++;
|
2000-07-04 16:35:15 +00:00
|
|
|
return ENOBUFS;
|
2002-09-17 14:25:19 +00:00
|
|
|
}
|
2000-07-04 16:35:15 +00:00
|
|
|
}
|
|
|
|
|
ip6 = mtod(m, struct ip6_hdr *);
|
|
|
|
|
tos = (ntohl(ip6->ip6_flow) >> 20) & 0xff;
|
|
|
|
|
|
2005-06-26 18:11:11 +00:00
|
|
|
/*
|
|
|
|
|
* BPF writes need to be handled specially.
|
|
|
|
|
* This is a null operation, nothing here checks dst->sa_family.
|
|
|
|
|
*/
|
|
|
|
|
if (dst->sa_family == AF_UNSPEC) {
|
|
|
|
|
bcopy(dst->sa_data, &af, sizeof(af));
|
|
|
|
|
dst->sa_family = af;
|
|
|
|
|
}
|
|
|
|
|
|
2001-06-11 12:39:29 +00:00
|
|
|
/*
|
|
|
|
|
* Pickup the right outer dst addr from the list of candidates.
|
|
|
|
|
* ip6_dst has priority as it may be able to give us shorter IPv4 hops.
|
|
|
|
|
*/
|
2003-01-15 20:09:52 +00:00
|
|
|
ptr = NULL;
|
2001-06-11 12:39:29 +00:00
|
|
|
if (IN6_IS_ADDR_6TO4(&ip6->ip6_dst))
|
2003-01-15 20:09:52 +00:00
|
|
|
ptr = GET_V4(&ip6->ip6_dst);
|
2001-06-11 12:39:29 +00:00
|
|
|
else if (IN6_IS_ADDR_6TO4(&dst6->sin6_addr))
|
2003-01-15 20:09:52 +00:00
|
|
|
ptr = GET_V4(&dst6->sin6_addr);
|
2001-06-11 12:39:29 +00:00
|
|
|
else {
|
|
|
|
|
m_freem(m);
|
2002-09-17 14:25:19 +00:00
|
|
|
ifp->if_oerrors++;
|
2001-06-11 12:39:29 +00:00
|
|
|
return ENETUNREACH;
|
|
|
|
|
}
|
2003-01-15 20:09:52 +00:00
|
|
|
bcopy(ptr, &in4, sizeof(in4));
|
2001-06-11 12:39:29 +00:00
|
|
|
|
2006-06-02 19:59:33 +00:00
|
|
|
if (bpf_peers_present(ifp->if_bpf)) {
|
2001-06-24 14:52:55 +00:00
|
|
|
/*
|
|
|
|
|
* We need to prepend the address family as
|
|
|
|
|
* a four byte field. Cons up a dummy header
|
|
|
|
|
* to pacify bpf. This is safe because bpf
|
|
|
|
|
* will only read from the mbuf (i.e., it won't
|
|
|
|
|
* try to free it or keep a pointer a to it).
|
|
|
|
|
*/
|
2005-06-26 18:11:11 +00:00
|
|
|
af = AF_INET6;
|
2003-12-28 03:56:00 +00:00
|
|
|
bpf_mtap2(ifp->if_bpf, &af, sizeof(af), m);
|
2001-06-24 14:52:55 +00:00
|
|
|
}
|
|
|
|
|
|
2003-02-19 05:47:46 +00:00
|
|
|
M_PREPEND(m, sizeof(struct ip), M_DONTWAIT);
|
2000-07-04 16:35:15 +00:00
|
|
|
if (m && m->m_len < sizeof(struct ip))
|
|
|
|
|
m = m_pullup(m, sizeof(struct ip));
|
2002-09-17 14:25:19 +00:00
|
|
|
if (m == NULL) {
|
|
|
|
|
ifp->if_oerrors++;
|
2000-07-04 16:35:15 +00:00
|
|
|
return ENOBUFS;
|
2002-09-17 14:25:19 +00:00
|
|
|
}
|
2000-07-04 16:35:15 +00:00
|
|
|
ip = mtod(m, struct ip *);
|
|
|
|
|
|
|
|
|
|
bzero(ip, sizeof(*ip));
|
|
|
|
|
|
|
|
|
|
bcopy(GET_V4(&((struct sockaddr_in6 *)&ia6->ia_addr)->sin6_addr),
|
|
|
|
|
&ip->ip_src, sizeof(ip->ip_src));
|
2003-01-15 20:09:52 +00:00
|
|
|
bcopy(&in4, &ip->ip_dst, sizeof(ip->ip_dst));
|
2000-07-04 16:35:15 +00:00
|
|
|
ip->ip_p = IPPROTO_IPV6;
|
2001-07-02 21:02:09 +00:00
|
|
|
ip->ip_ttl = ip_stf_ttl;
|
2000-07-04 16:35:15 +00:00
|
|
|
ip->ip_len = m->m_pkthdr.len; /*host order*/
|
|
|
|
|
if (ifp->if_flags & IFF_LINK1)
|
|
|
|
|
ip_ecn_ingress(ECN_ALLOWED, &ip->ip_tos, &tos);
|
2001-06-11 12:39:29 +00:00
|
|
|
else
|
|
|
|
|
ip_ecn_ingress(ECN_NOCARE, &ip->ip_tos, &tos);
|
2000-07-04 16:35:15 +00:00
|
|
|
|
2008-09-25 12:35:01 +00:00
|
|
|
if (!stf_route_cache) {
|
|
|
|
|
cached_route = NULL;
|
|
|
|
|
goto sendit;
|
|
|
|
|
}
|
|
|
|
|
|
2004-03-09 20:29:19 +00:00
|
|
|
/*
|
2008-09-25 12:35:01 +00:00
|
|
|
* Do we have a cached route?
|
2004-03-09 20:29:19 +00:00
|
|
|
*/
|
2008-09-25 12:35:01 +00:00
|
|
|
mtx_lock(&(sc)->sc_ro_mtx);
|
2000-07-04 16:35:15 +00:00
|
|
|
dst4 = (struct sockaddr_in *)&sc->sc_ro.ro_dst;
|
|
|
|
|
if (dst4->sin_family != AF_INET ||
|
|
|
|
|
bcmp(&dst4->sin_addr, &ip->ip_dst, sizeof(ip->ip_dst)) != 0) {
|
|
|
|
|
/* cache route doesn't match */
|
|
|
|
|
dst4->sin_family = AF_INET;
|
|
|
|
|
dst4->sin_len = sizeof(struct sockaddr_in);
|
|
|
|
|
bcopy(&ip->ip_dst, &dst4->sin_addr, sizeof(dst4->sin_addr));
|
|
|
|
|
if (sc->sc_ro.ro_rt) {
|
|
|
|
|
RTFREE(sc->sc_ro.ro_rt);
|
|
|
|
|
sc->sc_ro.ro_rt = NULL;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
if (sc->sc_ro.ro_rt == NULL) {
|
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
Reviewed by: several including rwatson, bz and mlair (parts each)
Obtained from: Ironport systems/Cisco
2008-05-09 23:03:00 +00:00
|
|
|
rtalloc_fib(&sc->sc_ro, sc->sc_fibnum);
|
2000-07-04 16:35:15 +00:00
|
|
|
if (sc->sc_ro.ro_rt == NULL) {
|
|
|
|
|
m_freem(m);
|
2008-09-25 12:35:01 +00:00
|
|
|
mtx_unlock(&(sc)->sc_ro_mtx);
|
2002-09-17 14:25:19 +00:00
|
|
|
ifp->if_oerrors++;
|
2000-07-04 16:35:15 +00:00
|
|
|
return ENETUNREACH;
|
|
|
|
|
}
|
|
|
|
|
}
|
2008-09-25 12:35:01 +00:00
|
|
|
cached_route = &sc->sc_ro;
|
2000-07-04 16:35:15 +00:00
|
|
|
|
2008-09-25 12:35:01 +00:00
|
|
|
sendit:
|
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
Reviewed by: several including rwatson, bz and mlair (parts each)
Obtained from: Ironport systems/Cisco
2008-05-09 23:03:00 +00:00
|
|
|
M_SETFIB(m, sc->sc_fibnum);
|
2002-09-17 14:25:19 +00:00
|
|
|
ifp->if_opackets++;
|
2008-09-25 12:35:01 +00:00
|
|
|
error = ip_output(m, NULL, cached_route, 0, NULL, NULL);
|
|
|
|
|
|
|
|
|
|
if (cached_route != NULL)
|
|
|
|
|
mtx_unlock(&(sc)->sc_ro_mtx);
|
|
|
|
|
return error;
|
2000-07-04 16:35:15 +00:00
|
|
|
}
|
|
|
|
|
|
2002-09-17 10:45:51 +00:00
|
|
|
static int
|
|
|
|
|
isrfc1918addr(in)
|
2003-01-15 20:09:52 +00:00
|
|
|
struct in_addr *in;
|
2002-09-17 10:45:51 +00:00
|
|
|
{
|
|
|
|
|
/*
|
|
|
|
|
* returns 1 if private address range:
|
|
|
|
|
* 10.0.0.0/8 172.16.0.0/12 192.168.0.0/16
|
|
|
|
|
*/
|
2003-01-15 20:09:52 +00:00
|
|
|
if ((ntohl(in->s_addr) & 0xff000000) >> 24 == 10 ||
|
|
|
|
|
(ntohl(in->s_addr) & 0xfff00000) >> 16 == 172 * 256 + 16 ||
|
|
|
|
|
(ntohl(in->s_addr) & 0xffff0000) >> 16 == 192 * 256 + 168)
|
2002-09-17 10:45:51 +00:00
|
|
|
return 1;
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
2000-07-04 16:35:15 +00:00
|
|
|
static int
|
2001-06-11 12:39:29 +00:00
|
|
|
stf_checkaddr4(sc, in, inifp)
|
|
|
|
|
struct stf_softc *sc;
|
2003-01-15 20:09:52 +00:00
|
|
|
struct in_addr *in;
|
2001-06-11 12:39:29 +00:00
|
|
|
struct ifnet *inifp; /* incoming interface */
|
2000-07-04 16:35:15 +00:00
|
|
|
{
|
Step 1.5 of importing the network stack virtualization infrastructure
from the vimage project, as per plan established at devsummit 08/08:
http://wiki.freebsd.org/Image/Notes200808DevSummit
Introduce INIT_VNET_*() initializer macros, VNET_FOREACH() iterator
macros, and CURVNET_SET() context setting macros, all currently
resolving to NOPs.
Prepare for virtualization of selected SYSCTL objects by introducing a
family of SYSCTL_V_*() macros, currently resolving to their global
counterparts, i.e. SYSCTL_V_INT() == SYSCTL_INT().
Move selected #defines from sys/sys/vimage.h to newly introduced header
files specific to virtualized subsystems (sys/net/vnet.h,
sys/netinet/vinet.h etc.).
All the changes are verified to have zero functional impact at this
point in time by doing MD5 comparision between pre- and post-change
object files(*).
(*) netipsec/keysock.c did not validate depending on compile time options.
Implemented by: julian, bz, brooks, zec
Reviewed by: julian, bz, brooks, kris, rwatson, ...
Approved by: julian (mentor)
Obtained from: //depot/projects/vimage-commit2/...
X-MFC after: never
Sponsored by: NLnet Foundation, The FreeBSD Foundation
2008-10-02 15:37:58 +00:00
|
|
|
INIT_VNET_INET(curvnet);
|
2000-07-04 16:35:15 +00:00
|
|
|
struct in_ifaddr *ia4;
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* reject packets with the following address:
|
|
|
|
|
* 224.0.0.0/4 0.0.0.0/8 127.0.0.0/8 255.0.0.0/8
|
|
|
|
|
*/
|
2003-01-15 20:09:52 +00:00
|
|
|
if (IN_MULTICAST(ntohl(in->s_addr)))
|
2000-07-04 16:35:15 +00:00
|
|
|
return -1;
|
2003-01-15 20:09:52 +00:00
|
|
|
switch ((ntohl(in->s_addr) & 0xff000000) >> 24) {
|
|
|
|
|
case 0: case 127: case 255:
|
2000-07-04 16:35:15 +00:00
|
|
|
return -1;
|
2003-01-15 20:09:52 +00:00
|
|
|
}
|
2000-07-04 16:35:15 +00:00
|
|
|
|
2002-09-17 10:45:51 +00:00
|
|
|
/*
|
|
|
|
|
* reject packets with private address range.
|
|
|
|
|
* (requirement from RFC3056 section 2 1st paragraph)
|
|
|
|
|
*/
|
|
|
|
|
if (isrfc1918addr(in))
|
|
|
|
|
return -1;
|
|
|
|
|
|
2000-07-04 16:35:15 +00:00
|
|
|
/*
|
|
|
|
|
* reject packets with broadcast
|
|
|
|
|
*/
|
Commit step 1 of the vimage project, (network stack)
virtualization work done by Marko Zec (zec@).
This is the first in a series of commits over the course
of the next few weeks.
Mark all uses of global variables to be virtualized
with a V_ prefix.
Use macros to map them back to their global names for
now, so this is a NOP change only.
We hope to have caught at least 85-90% of what is needed
so we do not invalidate a lot of outstanding patches again.
Obtained from: //depot/projects/vimage-commit2/...
Reviewed by: brooks, des, ed, mav, julian,
jamie, kris, rwatson, zec, ...
(various people I forgot, different versions)
md5 (with a bit of help)
Sponsored by: NLnet Foundation, The FreeBSD Foundation
X-MFC after: never
V_Commit_Message_Reviewed_By: more people than the patch
2008-08-17 23:27:27 +00:00
|
|
|
for (ia4 = TAILQ_FIRST(&V_in_ifaddrhead);
|
2000-07-04 16:35:15 +00:00
|
|
|
ia4;
|
|
|
|
|
ia4 = TAILQ_NEXT(ia4, ia_link))
|
|
|
|
|
{
|
|
|
|
|
if ((ia4->ia_ifa.ifa_ifp->if_flags & IFF_BROADCAST) == 0)
|
|
|
|
|
continue;
|
2003-01-15 20:09:52 +00:00
|
|
|
if (in->s_addr == ia4->ia_broadaddr.sin_addr.s_addr)
|
2000-07-04 16:35:15 +00:00
|
|
|
return -1;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* perform ingress filter
|
|
|
|
|
*/
|
2005-06-10 16:49:24 +00:00
|
|
|
if (sc && (STF2IFP(sc)->if_flags & IFF_LINK2) == 0 && inifp) {
|
2000-07-04 16:35:15 +00:00
|
|
|
struct sockaddr_in sin;
|
|
|
|
|
struct rtentry *rt;
|
|
|
|
|
|
|
|
|
|
bzero(&sin, sizeof(sin));
|
|
|
|
|
sin.sin_family = AF_INET;
|
|
|
|
|
sin.sin_len = sizeof(struct sockaddr_in);
|
2003-01-15 20:09:52 +00:00
|
|
|
sin.sin_addr = *in;
|
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
Reviewed by: several including rwatson, bz and mlair (parts each)
Obtained from: Ironport systems/Cisco
2008-05-09 23:03:00 +00:00
|
|
|
rt = rtalloc1_fib((struct sockaddr *)&sin, 0,
|
|
|
|
|
0UL, sc->sc_fibnum);
|
2001-06-11 12:39:29 +00:00
|
|
|
if (!rt || rt->rt_ifp != inifp) {
|
|
|
|
|
#if 0
|
|
|
|
|
log(LOG_WARNING, "%s: packet from 0x%x dropped "
|
2005-06-10 16:49:24 +00:00
|
|
|
"due to ingress filter\n", if_name(STF2IFP(sc)),
|
2001-06-11 12:39:29 +00:00
|
|
|
(u_int32_t)ntohl(sin.sin_addr.s_addr));
|
|
|
|
|
#endif
|
|
|
|
|
if (rt)
|
2007-09-23 17:50:17 +00:00
|
|
|
RTFREE_LOCKED(rt);
|
2000-07-04 16:35:15 +00:00
|
|
|
return -1;
|
|
|
|
|
}
|
2007-09-23 17:50:17 +00:00
|
|
|
RTFREE_LOCKED(rt);
|
2000-07-04 16:35:15 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int
|
2001-06-11 12:39:29 +00:00
|
|
|
stf_checkaddr6(sc, in6, inifp)
|
|
|
|
|
struct stf_softc *sc;
|
2000-07-04 16:35:15 +00:00
|
|
|
struct in6_addr *in6;
|
2001-06-11 12:39:29 +00:00
|
|
|
struct ifnet *inifp; /* incoming interface */
|
2000-07-04 16:35:15 +00:00
|
|
|
{
|
|
|
|
|
/*
|
|
|
|
|
* check 6to4 addresses
|
|
|
|
|
*/
|
2003-01-15 20:09:52 +00:00
|
|
|
if (IN6_IS_ADDR_6TO4(in6)) {
|
|
|
|
|
struct in_addr in4;
|
|
|
|
|
bcopy(GET_V4(in6), &in4, sizeof(in4));
|
|
|
|
|
return stf_checkaddr4(sc, &in4, inifp);
|
|
|
|
|
}
|
2000-07-04 16:35:15 +00:00
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* reject anything that look suspicious. the test is implemented
|
|
|
|
|
* in ip6_input too, but we check here as well to
|
|
|
|
|
* (1) reject bad packets earlier, and
|
|
|
|
|
* (2) to be safe against future ip6_input change.
|
|
|
|
|
*/
|
|
|
|
|
if (IN6_IS_ADDR_V4COMPAT(in6) || IN6_IS_ADDR_V4MAPPED(in6))
|
|
|
|
|
return -1;
|
|
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
void
|
2001-09-07 07:19:12 +00:00
|
|
|
in_stf_input(m, off)
|
2001-06-11 12:39:29 +00:00
|
|
|
struct mbuf *m;
|
2001-09-07 07:19:12 +00:00
|
|
|
int off;
|
2000-07-04 16:35:15 +00:00
|
|
|
{
|
2001-09-07 07:19:12 +00:00
|
|
|
int proto;
|
2000-07-04 16:35:15 +00:00
|
|
|
struct stf_softc *sc;
|
|
|
|
|
struct ip *ip;
|
|
|
|
|
struct ip6_hdr *ip6;
|
|
|
|
|
u_int8_t otos, itos;
|
|
|
|
|
struct ifnet *ifp;
|
|
|
|
|
|
2001-09-03 20:03:55 +00:00
|
|
|
proto = mtod(m, struct ip *)->ip_p;
|
2000-07-04 16:35:15 +00:00
|
|
|
|
|
|
|
|
if (proto != IPPROTO_IPV6) {
|
|
|
|
|
m_freem(m);
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
ip = mtod(m, struct ip *);
|
|
|
|
|
|
|
|
|
|
sc = (struct stf_softc *)encap_getarg(m);
|
|
|
|
|
|
2005-06-10 16:49:24 +00:00
|
|
|
if (sc == NULL || (STF2IFP(sc)->if_flags & IFF_UP) == 0) {
|
2000-07-04 16:35:15 +00:00
|
|
|
m_freem(m);
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
2005-06-10 16:49:24 +00:00
|
|
|
ifp = STF2IFP(sc);
|
2000-07-04 16:35:15 +00:00
|
|
|
|
2002-10-20 22:39:55 +00:00
|
|
|
#ifdef MAC
|
2007-10-24 19:04:04 +00:00
|
|
|
mac_ifnet_create_mbuf(ifp, m);
|
2002-10-20 22:39:55 +00:00
|
|
|
#endif
|
|
|
|
|
|
2000-07-04 16:35:15 +00:00
|
|
|
/*
|
|
|
|
|
* perform sanity check against outer src/dst.
|
|
|
|
|
* for source, perform ingress filter as well.
|
|
|
|
|
*/
|
2001-06-11 12:39:29 +00:00
|
|
|
if (stf_checkaddr4(sc, &ip->ip_dst, NULL) < 0 ||
|
|
|
|
|
stf_checkaddr4(sc, &ip->ip_src, m->m_pkthdr.rcvif) < 0) {
|
2000-07-04 16:35:15 +00:00
|
|
|
m_freem(m);
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
otos = ip->ip_tos;
|
|
|
|
|
m_adj(m, off);
|
|
|
|
|
|
|
|
|
|
if (m->m_len < sizeof(*ip6)) {
|
|
|
|
|
m = m_pullup(m, sizeof(*ip6));
|
|
|
|
|
if (!m)
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
ip6 = mtod(m, struct ip6_hdr *);
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* perform sanity check against inner src/dst.
|
|
|
|
|
* for source, perform ingress filter as well.
|
|
|
|
|
*/
|
2001-06-11 12:39:29 +00:00
|
|
|
if (stf_checkaddr6(sc, &ip6->ip6_dst, NULL) < 0 ||
|
|
|
|
|
stf_checkaddr6(sc, &ip6->ip6_src, m->m_pkthdr.rcvif) < 0) {
|
2000-07-04 16:35:15 +00:00
|
|
|
m_freem(m);
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
itos = (ntohl(ip6->ip6_flow) >> 20) & 0xff;
|
|
|
|
|
if ((ifp->if_flags & IFF_LINK1) != 0)
|
|
|
|
|
ip_ecn_egress(ECN_ALLOWED, &otos, &itos);
|
2001-06-11 12:39:29 +00:00
|
|
|
else
|
|
|
|
|
ip_ecn_egress(ECN_NOCARE, &otos, &itos);
|
2000-07-04 16:35:15 +00:00
|
|
|
ip6->ip6_flow &= ~htonl(0xff << 20);
|
|
|
|
|
ip6->ip6_flow |= htonl((u_int32_t)itos << 20);
|
|
|
|
|
|
|
|
|
|
m->m_pkthdr.rcvif = ifp;
|
|
|
|
|
|
2006-06-02 19:59:33 +00:00
|
|
|
if (bpf_peers_present(ifp->if_bpf)) {
|
2000-07-04 16:35:15 +00:00
|
|
|
/*
|
|
|
|
|
* We need to prepend the address family as
|
|
|
|
|
* a four byte field. Cons up a dummy header
|
|
|
|
|
* to pacify bpf. This is safe because bpf
|
|
|
|
|
* will only read from the mbuf (i.e., it won't
|
|
|
|
|
* try to free it or keep a pointer a to it).
|
|
|
|
|
*/
|
2001-06-11 12:39:29 +00:00
|
|
|
u_int32_t af = AF_INET6;
|
2005-01-11 07:05:56 +00:00
|
|
|
bpf_mtap2(ifp->if_bpf, &af, sizeof(af), m);
|
2000-07-04 16:35:15 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
|
* Put the packet to the network layer input queue according to the
|
|
|
|
|
* specified address family.
|
|
|
|
|
* See net/if_gif.c for possible issues with packet processing
|
|
|
|
|
* reorder due to extra queueing.
|
|
|
|
|
*/
|
|
|
|
|
ifp->if_ipackets++;
|
2003-03-04 23:19:55 +00:00
|
|
|
ifp->if_ibytes += m->m_pkthdr.len;
|
|
|
|
|
netisr_dispatch(NETISR_IPV6, m);
|
2000-07-04 16:35:15 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/* ARGSUSED */
|
|
|
|
|
static void
|
2001-10-17 18:07:05 +00:00
|
|
|
stf_rtrequest(cmd, rt, info)
|
2000-07-04 16:35:15 +00:00
|
|
|
int cmd;
|
|
|
|
|
struct rtentry *rt;
|
2001-10-17 18:07:05 +00:00
|
|
|
struct rt_addrinfo *info;
|
2000-07-04 16:35:15 +00:00
|
|
|
{
|
2003-10-04 03:44:50 +00:00
|
|
|
RT_LOCK_ASSERT(rt);
|
2005-02-24 01:34:01 +00:00
|
|
|
rt->rt_rmx.rmx_mtu = IPV6_MMTU;
|
2000-07-04 16:35:15 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static int
|
|
|
|
|
stf_ioctl(ifp, cmd, data)
|
|
|
|
|
struct ifnet *ifp;
|
|
|
|
|
u_long cmd;
|
|
|
|
|
caddr_t data;
|
|
|
|
|
{
|
|
|
|
|
struct ifaddr *ifa;
|
|
|
|
|
struct ifreq *ifr;
|
|
|
|
|
struct sockaddr_in6 *sin6;
|
2003-01-15 20:09:52 +00:00
|
|
|
struct in_addr addr;
|
2000-07-04 16:35:15 +00:00
|
|
|
int error;
|
|
|
|
|
|
|
|
|
|
error = 0;
|
|
|
|
|
switch (cmd) {
|
|
|
|
|
case SIOCSIFADDR:
|
|
|
|
|
ifa = (struct ifaddr *)data;
|
|
|
|
|
if (ifa == NULL || ifa->ifa_addr->sa_family != AF_INET6) {
|
|
|
|
|
error = EAFNOSUPPORT;
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
sin6 = (struct sockaddr_in6 *)ifa->ifa_addr;
|
2003-01-15 20:09:52 +00:00
|
|
|
if (!IN6_IS_ADDR_6TO4(&sin6->sin6_addr)) {
|
2000-07-04 16:35:15 +00:00
|
|
|
error = EINVAL;
|
2003-01-15 20:09:52 +00:00
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
bcopy(GET_V4(&sin6->sin6_addr), &addr, sizeof(addr));
|
|
|
|
|
if (isrfc1918addr(&addr)) {
|
|
|
|
|
error = EINVAL;
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
ifa->ifa_rtrequest = stf_rtrequest;
|
|
|
|
|
ifp->if_flags |= IFF_UP;
|
2000-07-04 16:35:15 +00:00
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
case SIOCADDMULTI:
|
|
|
|
|
case SIOCDELMULTI:
|
|
|
|
|
ifr = (struct ifreq *)data;
|
|
|
|
|
if (ifr && ifr->ifr_addr.sa_family == AF_INET6)
|
|
|
|
|
;
|
|
|
|
|
else
|
|
|
|
|
error = EAFNOSUPPORT;
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
|
|
default:
|
|
|
|
|
error = EINVAL;
|
|
|
|
|
break;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
return error;
|
|
|
|
|
}
|
