2005-01-07 01:45:51 +00:00
|
|
|
/*-
|
1994-05-24 10:09:53 +00:00
|
|
|
* Copyright (c) 1982, 1986, 1991, 1993
|
|
|
|
* The Regents of the University of California. All rights reserved.
|
2004-11-13 17:05:40 +00:00
|
|
|
* Copyright (C) 2001 WIDE Project. All rights reserved.
|
1994-05-24 10:09:53 +00:00
|
|
|
*
|
|
|
|
* Redistribution and use in source and binary forms, with or without
|
|
|
|
* modification, are permitted provided that the following conditions
|
|
|
|
* are met:
|
|
|
|
* 1. Redistributions of source code must retain the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer.
|
|
|
|
* 2. Redistributions in binary form must reproduce the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
|
|
* documentation and/or other materials provided with the distribution.
|
|
|
|
* 4. Neither the name of the University nor the names of its contributors
|
|
|
|
* may be used to endorse or promote products derived from this software
|
|
|
|
* without specific prior written permission.
|
|
|
|
*
|
|
|
|
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
|
|
|
|
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
|
|
|
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
|
|
|
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
|
|
|
|
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
|
|
|
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
|
|
|
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
|
|
|
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
|
|
|
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
|
|
|
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
|
|
|
* SUCH DAMAGE.
|
|
|
|
*
|
1995-09-21 17:50:45 +00:00
|
|
|
* @(#)in.c 8.4 (Berkeley) 1/9/95
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
|
|
|
|
2007-10-07 20:44:24 +00:00
|
|
|
#include <sys/cdefs.h>
|
|
|
|
__FBSDID("$FreeBSD$");
|
|
|
|
|
2010-03-09 01:11:45 +00:00
|
|
|
#include "opt_mpath.h"
|
2005-03-20 10:27:17 +00:00
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <sys/param.h>
|
1994-05-25 09:21:21 +00:00
|
|
|
#include <sys/systm.h>
|
1997-03-24 11:33:46 +00:00
|
|
|
#include <sys/sockio.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <sys/malloc.h>
|
2006-11-06 13:42:10 +00:00
|
|
|
#include <sys/priv.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <sys/socket.h>
|
2009-01-09 13:06:56 +00:00
|
|
|
#include <sys/jail.h>
|
1995-12-09 20:43:53 +00:00
|
|
|
#include <sys/kernel.h>
|
2009-01-09 13:06:56 +00:00
|
|
|
#include <sys/proc.h>
|
1995-12-09 20:43:53 +00:00
|
|
|
#include <sys/sysctl.h>
|
2009-05-12 07:41:20 +00:00
|
|
|
#include <sys/syslog.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
|
|
|
|
#include <net/if.h>
|
2009-07-27 17:08:06 +00:00
|
|
|
#include <net/if_var.h>
|
2010-11-12 22:03:02 +00:00
|
|
|
#include <net/if_arp.h>
|
2009-05-12 07:41:20 +00:00
|
|
|
#include <net/if_dl.h>
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
#include <net/if_llatbl.h>
|
1999-12-22 19:13:38 +00:00
|
|
|
#include <net/if_types.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <net/route.h>
|
2009-05-12 07:41:20 +00:00
|
|
|
#include <net/vnet.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
|
A major overhaul of the CARP implementation. The ip_carp.c was started
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
|
|
|
#include <netinet/if_ether.h>
|
1994-05-24 10:09:53 +00:00
|
|
|
#include <netinet/in.h>
|
|
|
|
#include <netinet/in_var.h>
|
2001-08-04 17:10:14 +00:00
|
|
|
#include <netinet/in_pcb.h>
|
Import rewrite of IPv4 socket multicast layer to support source-specific
and protocol-independent host mode multicast. The code is written to
accomodate IPv6, IGMPv3 and MLDv2 with only a little additional work.
This change only pertains to FreeBSD's use as a multicast end-station and
does not concern multicast routing; for an IGMPv3/MLDv2 router
implementation, consider the XORP project.
The work is based on Wilbert de Graaf's IGMPv3 code drop for FreeBSD 4.6,
which is available at: http://www.kloosterhof.com/wilbert/igmpv3.html
Summary
* IPv4 multicast socket processing is now moved out of ip_output.c
into a new module, in_mcast.c.
* The in_mcast.c module implements the IPv4 legacy any-source API in
terms of the protocol-independent source-specific API.
* Source filters are lazy allocated as the common case does not use them.
They are part of per inpcb state and are covered by the inpcb lock.
* struct ip_mreqn is now supported to allow applications to specify
multicast joins by interface index in the legacy IPv4 any-source API.
* In UDP, an incoming multicast datagram only requires that the source
port matches the 4-tuple if the socket was already bound by source port.
An unbound socket SHOULD be able to receive multicasts sent from an
ephemeral source port.
* The UDP socket multicast filter mode defaults to exclusive, that is,
sources present in the per-socket list will be blocked from delivery.
* The RFC 3678 userland functions have been added to libc: setsourcefilter,
getsourcefilter, setipv4sourcefilter, getipv4sourcefilter.
* Definitions for IGMPv3 are merged but not yet used.
* struct sockaddr_storage is now referenced from <netinet/in.h>. It
is therefore defined there if not already declared in the same way
as for the C99 types.
* The RFC 1724 hack (specify 0.0.0.0/8 addresses to IP_MULTICAST_IF
which are then interpreted as interface indexes) is now deprecated.
* A patch for the Rhyolite.com routed in the FreeBSD base system
is available in the -net archives. This only affects individuals
running RIPv1 or RIPv2 via point-to-point and/or unnumbered interfaces.
* Make IPv6 detach path similar to IPv4's in code flow; functionally same.
* Bump __FreeBSD_version to 700048; see UPDATING.
This work was financially supported by another FreeBSD committer.
Obtained from: p4://bms_netdev
Submitted by: Wilbert de Graaf (original work)
Reviewed by: rwatson (locking), silence from fenner,
net@ (but with encouragement)
2007-06-12 16:24:56 +00:00
|
|
|
#include <netinet/ip_var.h>
|
A major overhaul of the CARP implementation. The ip_carp.c was started
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
|
|
|
#include <netinet/ip_carp.h>
|
2009-03-09 17:53:05 +00:00
|
|
|
#include <netinet/igmp_var.h>
|
Build on Jeff Roberson's linker-set based dynamic per-CPU allocator
(DPCPU), as suggested by Peter Wemm, and implement a new per-virtual
network stack memory allocator. Modify vnet to use the allocator
instead of monolithic global container structures (vinet, ...). This
change solves many binary compatibility problems associated with
VIMAGE, and restores ELF symbols for virtualized global variables.
Each virtualized global variable exists as a "reference copy", and also
once per virtual network stack. Virtualized global variables are
tagged at compile-time, placing the in a special linker set, which is
loaded into a contiguous region of kernel memory. Virtualized global
variables in the base kernel are linked as normal, but those in modules
are copied and relocated to a reserved portion of the kernel's vnet
region with the help of a the kernel linker.
Virtualized global variables exist in per-vnet memory set up when the
network stack instance is created, and are initialized statically from
the reference copy. Run-time access occurs via an accessor macro, which
converts from the current vnet and requested symbol to a per-vnet
address. When "options VIMAGE" is not compiled into the kernel, normal
global ELF symbols will be used instead and indirection is avoided.
This change restores static initialization for network stack global
variables, restores support for non-global symbols and types, eliminates
the need for many subsystem constructors, eliminates large per-subsystem
structures that caused many binary compatibility issues both for
monitoring applications (netstat) and kernel modules, removes the
per-function INIT_VNET_*() macros throughout the stack, eliminates the
need for vnet_symmap ksym(2) munging, and eliminates duplicate
definitions of virtualized globals under VIMAGE_GLOBALS.
Bump __FreeBSD_version and update UPDATING.
Portions submitted by: bz
Reviewed by: bz, zec
Discussed with: gnn, jamie, jeff, jhb, julian, sam
Suggested by: peter
Approved by: re (kensmith)
2009-07-14 22:48:30 +00:00
|
|
|
#include <netinet/udp.h>
|
|
|
|
#include <netinet/udp_var.h>
|
1997-10-11 18:31:40 +00:00
|
|
|
|
2002-03-19 21:25:46 +00:00
|
|
|
static int in_mask2len(struct in_addr *);
|
|
|
|
static void in_len2mask(struct in_addr *, int);
|
|
|
|
static int in_lifaddr_ioctl(struct socket *, u_long, caddr_t,
|
|
|
|
struct ifnet *, struct thread *);
|
1999-12-22 19:13:38 +00:00
|
|
|
|
2002-03-19 21:25:46 +00:00
|
|
|
static void in_socktrim(struct sockaddr_in *);
|
A major overhaul of the CARP implementation. The ip_carp.c was started
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
|
|
|
static int in_ifinit(struct ifnet *, struct in_ifaddr *,
|
2012-01-16 09:53:24 +00:00
|
|
|
struct sockaddr_in *, int, int);
|
2007-03-20 00:36:10 +00:00
|
|
|
static void in_purgemaddrs(struct ifnet *);
|
1994-05-24 10:09:53 +00:00
|
|
|
|
A major overhaul of the CARP implementation. The ip_carp.c was started
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
|
|
|
static VNET_DEFINE(int, nosameprefix);
|
|
|
|
#define V_nosameprefix VNET(nosameprefix)
|
|
|
|
SYSCTL_VNET_INT(_net_inet_ip, OID_AUTO, no_same_prefix, CTLFLAG_RW,
|
|
|
|
&VNET_NAME(nosameprefix), 0,
|
2005-08-18 10:34:30 +00:00
|
|
|
"Refuse to create same prefixes on different interfaces");
|
1997-01-13 21:26:53 +00:00
|
|
|
|
2010-04-29 11:52:42 +00:00
|
|
|
VNET_DECLARE(struct inpcbinfo, ripcbinfo);
|
|
|
|
#define V_ripcbinfo VNET(ripcbinfo)
|
|
|
|
|
2010-11-12 22:03:02 +00:00
|
|
|
VNET_DECLARE(struct arpstat, arpstat); /* ARP statistics, see if_arp.h */
|
|
|
|
#define V_arpstat VNET(arpstat)
|
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
|
|
|
* Return 1 if an internet address is for a ``local'' host
|
2011-10-15 16:28:06 +00:00
|
|
|
* (one to which we have a connection).
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
1994-05-25 09:21:21 +00:00
|
|
|
int
|
2007-05-10 15:58:48 +00:00
|
|
|
in_localaddr(struct in_addr in)
|
1994-05-24 10:09:53 +00:00
|
|
|
{
|
|
|
|
register u_long i = ntohl(in.s_addr);
|
|
|
|
register struct in_ifaddr *ia;
|
|
|
|
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RLOCK();
|
2011-10-15 16:28:06 +00:00
|
|
|
TAILQ_FOREACH(ia, &V_in_ifaddrhead, ia_link) {
|
|
|
|
if ((i & ia->ia_subnetmask) == ia->ia_subnet) {
|
|
|
|
IN_IFADDR_RUNLOCK();
|
|
|
|
return (1);
|
2009-06-25 11:52:33 +00:00
|
|
|
}
|
1994-05-24 10:09:53 +00:00
|
|
|
}
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RUNLOCK();
|
1994-05-24 10:09:53 +00:00
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
2004-08-11 11:49:48 +00:00
|
|
|
/*
|
|
|
|
* Return 1 if an internet address is for the local host and configured
|
|
|
|
* on one of its interfaces.
|
|
|
|
*/
|
|
|
|
int
|
2007-05-10 15:58:48 +00:00
|
|
|
in_localip(struct in_addr in)
|
2004-08-11 11:49:48 +00:00
|
|
|
{
|
|
|
|
struct in_ifaddr *ia;
|
|
|
|
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RLOCK();
|
2004-08-11 11:49:48 +00:00
|
|
|
LIST_FOREACH(ia, INADDR_HASH(in.s_addr), ia_hash) {
|
2009-06-25 11:52:33 +00:00
|
|
|
if (IA_SIN(ia)->sin_addr.s_addr == in.s_addr) {
|
|
|
|
IN_IFADDR_RUNLOCK();
|
2008-10-26 19:17:25 +00:00
|
|
|
return (1);
|
2009-06-25 11:52:33 +00:00
|
|
|
}
|
2004-08-11 11:49:48 +00:00
|
|
|
}
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RUNLOCK();
|
2008-10-26 19:17:25 +00:00
|
|
|
return (0);
|
2004-08-11 11:49:48 +00:00
|
|
|
}
|
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
|
|
|
* Determine whether an IP address is in a reserved set of addresses
|
|
|
|
* that may not be forwarded, or whether datagrams to that destination
|
|
|
|
* may be forwarded.
|
|
|
|
*/
|
1994-05-25 09:21:21 +00:00
|
|
|
int
|
2007-05-10 15:58:48 +00:00
|
|
|
in_canforward(struct in_addr in)
|
1994-05-24 10:09:53 +00:00
|
|
|
{
|
|
|
|
register u_long i = ntohl(in.s_addr);
|
|
|
|
register u_long net;
|
|
|
|
|
2007-02-03 06:45:51 +00:00
|
|
|
if (IN_EXPERIMENTAL(i) || IN_MULTICAST(i) || IN_LINKLOCAL(i))
|
1994-05-24 10:09:53 +00:00
|
|
|
return (0);
|
|
|
|
if (IN_CLASSA(i)) {
|
|
|
|
net = i & IN_CLASSA_NET;
|
|
|
|
if (net == 0 || net == (IN_LOOPBACKNET << IN_CLASSA_NSHIFT))
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
return (1);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Trim a mask in a sockaddr
|
|
|
|
*/
|
1995-11-14 20:34:56 +00:00
|
|
|
static void
|
2007-05-10 15:58:48 +00:00
|
|
|
in_socktrim(struct sockaddr_in *ap)
|
1994-05-24 10:09:53 +00:00
|
|
|
{
|
|
|
|
register char *cplim = (char *) &ap->sin_addr;
|
|
|
|
register char *cp = (char *) (&ap->sin_addr + 1);
|
|
|
|
|
|
|
|
ap->sin_len = 0;
|
1994-11-03 21:04:21 +00:00
|
|
|
while (--cp >= cplim)
|
2004-08-16 18:32:07 +00:00
|
|
|
if (*cp) {
|
1994-05-24 10:09:53 +00:00
|
|
|
(ap)->sin_len = cp - (char *) (ap) + 1;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
1999-12-22 19:13:38 +00:00
|
|
|
static int
|
|
|
|
in_mask2len(mask)
|
|
|
|
struct in_addr *mask;
|
|
|
|
{
|
|
|
|
int x, y;
|
|
|
|
u_char *p;
|
|
|
|
|
|
|
|
p = (u_char *)mask;
|
|
|
|
for (x = 0; x < sizeof(*mask); x++) {
|
|
|
|
if (p[x] != 0xff)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
y = 0;
|
|
|
|
if (x < sizeof(*mask)) {
|
|
|
|
for (y = 0; y < 8; y++) {
|
|
|
|
if ((p[x] & (0x80 >> y)) == 0)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2008-10-26 19:17:25 +00:00
|
|
|
return (x * 8 + y);
|
1999-12-22 19:13:38 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static void
|
2007-05-10 15:58:48 +00:00
|
|
|
in_len2mask(struct in_addr *mask, int len)
|
1999-12-22 19:13:38 +00:00
|
|
|
{
|
|
|
|
int i;
|
|
|
|
u_char *p;
|
|
|
|
|
|
|
|
p = (u_char *)mask;
|
|
|
|
bzero(mask, sizeof(*mask));
|
|
|
|
for (i = 0; i < len / 8; i++)
|
|
|
|
p[i] = 0xff;
|
|
|
|
if (len % 8)
|
|
|
|
p[i] = (0xff00 >> (len % 8)) & 0xff;
|
|
|
|
}
|
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
|
|
|
* Generic internet control operations (ioctl's).
|
2009-04-23 21:41:37 +00:00
|
|
|
*
|
|
|
|
* ifp is NULL if not an interface-specific ioctl.
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
|
|
|
/* ARGSUSED */
|
1994-05-25 09:21:21 +00:00
|
|
|
int
|
2007-05-10 15:58:48 +00:00
|
|
|
in_control(struct socket *so, u_long cmd, caddr_t data, struct ifnet *ifp,
|
|
|
|
struct thread *td)
|
1994-05-24 10:09:53 +00:00
|
|
|
{
|
|
|
|
register struct ifreq *ifr = (struct ifreq *)data;
|
2008-10-26 19:17:25 +00:00
|
|
|
register struct in_ifaddr *ia, *iap;
|
1994-05-24 10:09:53 +00:00
|
|
|
register struct ifaddr *ifa;
|
2007-03-29 21:39:22 +00:00
|
|
|
struct in_addr allhosts_addr;
|
2001-09-29 04:34:11 +00:00
|
|
|
struct in_addr dst;
|
2009-03-09 17:53:05 +00:00
|
|
|
struct in_ifinfo *ii;
|
1994-05-24 10:09:53 +00:00
|
|
|
struct in_aliasreq *ifra = (struct in_aliasreq *)data;
|
2009-06-25 11:52:33 +00:00
|
|
|
int error, hostIsNew, iaIsNew, maskIsNew;
|
2007-03-29 21:39:22 +00:00
|
|
|
int iaIsFirst;
|
2011-12-21 12:39:08 +00:00
|
|
|
u_long ocmd = cmd;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Pre-10.x compat: OSIOCAIFADDR passes a shorter
|
|
|
|
* struct in_aliasreq, without ifra_vhid.
|
|
|
|
*/
|
|
|
|
if (cmd == OSIOCAIFADDR)
|
|
|
|
cmd = SIOCAIFADDR;
|
2001-11-30 14:00:55 +00:00
|
|
|
|
2008-10-26 19:17:25 +00:00
|
|
|
ia = NULL;
|
2007-03-29 21:39:22 +00:00
|
|
|
iaIsFirst = 0;
|
2001-11-30 14:00:55 +00:00
|
|
|
iaIsNew = 0;
|
2007-03-29 21:39:22 +00:00
|
|
|
allhosts_addr.s_addr = htonl(INADDR_ALLHOSTS_GROUP);
|
1994-05-24 10:09:53 +00:00
|
|
|
|
2009-04-23 21:41:37 +00:00
|
|
|
/*
|
|
|
|
* Filter out ioctls we implement directly; forward the rest on to
|
|
|
|
* in_lifaddr_ioctl() and ifp->if_ioctl().
|
|
|
|
*/
|
1999-12-22 19:13:38 +00:00
|
|
|
switch (cmd) {
|
2009-04-23 21:41:37 +00:00
|
|
|
case SIOCGIFADDR:
|
|
|
|
case SIOCGIFBRDADDR:
|
|
|
|
case SIOCGIFDSTADDR:
|
|
|
|
case SIOCGIFNETMASK:
|
2011-11-21 14:10:13 +00:00
|
|
|
case SIOCDIFADDR:
|
|
|
|
break;
|
|
|
|
case SIOCAIFADDR:
|
|
|
|
/*
|
|
|
|
* ifra_addr must be present and be of INET family.
|
|
|
|
* ifra_broadaddr and ifra_mask are optional.
|
|
|
|
*/
|
|
|
|
if (ifra->ifra_addr.sin_len != sizeof(struct sockaddr_in) ||
|
|
|
|
ifra->ifra_addr.sin_family != AF_INET)
|
|
|
|
return (EINVAL);
|
|
|
|
if (ifra->ifra_broadaddr.sin_len != 0 &&
|
2011-11-22 19:39:27 +00:00
|
|
|
(ifra->ifra_broadaddr.sin_len !=
|
|
|
|
sizeof(struct sockaddr_in) ||
|
2011-11-21 14:10:13 +00:00
|
|
|
ifra->ifra_broadaddr.sin_family != AF_INET))
|
|
|
|
return (EINVAL);
|
2011-12-21 12:39:08 +00:00
|
|
|
#if 0
|
|
|
|
/*
|
|
|
|
* ifconfig(8) in pre-10.x doesn't set sin_family for the
|
|
|
|
* mask. The code is disabled for the 10.x timeline, to
|
|
|
|
* make SIOCAIFADDR compatible with 9.x ifconfig(8).
|
|
|
|
* The code should be enabled in 11.x
|
|
|
|
*/
|
2011-11-21 14:10:13 +00:00
|
|
|
if (ifra->ifra_mask.sin_len != 0 &&
|
|
|
|
(ifra->ifra_mask.sin_len != sizeof(struct sockaddr_in) ||
|
|
|
|
ifra->ifra_mask.sin_family != AF_INET))
|
|
|
|
return (EINVAL);
|
2011-12-21 12:39:08 +00:00
|
|
|
#endif
|
2011-11-21 14:10:13 +00:00
|
|
|
break;
|
2009-04-23 21:41:37 +00:00
|
|
|
case SIOCSIFADDR:
|
|
|
|
case SIOCSIFBRDADDR:
|
|
|
|
case SIOCSIFDSTADDR:
|
|
|
|
case SIOCSIFNETMASK:
|
2012-01-16 09:53:24 +00:00
|
|
|
/* We no longer support that old commands. */
|
|
|
|
return (EINVAL);
|
2009-04-23 21:41:37 +00:00
|
|
|
|
1999-12-22 19:13:38 +00:00
|
|
|
case SIOCALIFADDR:
|
2006-11-06 13:42:10 +00:00
|
|
|
if (td != NULL) {
|
|
|
|
error = priv_check(td, PRIV_NET_ADDIFADDR);
|
|
|
|
if (error)
|
|
|
|
return (error);
|
|
|
|
}
|
2008-10-26 19:17:25 +00:00
|
|
|
if (ifp == NULL)
|
|
|
|
return (EINVAL);
|
2006-11-06 13:42:10 +00:00
|
|
|
return in_lifaddr_ioctl(so, cmd, data, ifp, td);
|
|
|
|
|
1999-12-22 19:13:38 +00:00
|
|
|
case SIOCDLIFADDR:
|
2006-11-06 13:42:10 +00:00
|
|
|
if (td != NULL) {
|
|
|
|
error = priv_check(td, PRIV_NET_DELIFADDR);
|
|
|
|
if (error)
|
|
|
|
return (error);
|
|
|
|
}
|
2008-10-26 19:17:25 +00:00
|
|
|
if (ifp == NULL)
|
|
|
|
return (EINVAL);
|
2006-11-06 13:42:10 +00:00
|
|
|
return in_lifaddr_ioctl(so, cmd, data, ifp, td);
|
|
|
|
|
1999-12-22 19:13:38 +00:00
|
|
|
case SIOCGLIFADDR:
|
2008-10-26 19:17:25 +00:00
|
|
|
if (ifp == NULL)
|
|
|
|
return (EINVAL);
|
2001-09-12 08:38:13 +00:00
|
|
|
return in_lifaddr_ioctl(so, cmd, data, ifp, td);
|
2009-04-23 21:41:37 +00:00
|
|
|
|
|
|
|
default:
|
|
|
|
if (ifp == NULL || ifp->if_ioctl == NULL)
|
|
|
|
return (EOPNOTSUPP);
|
|
|
|
return ((*ifp->if_ioctl)(ifp, cmd, data));
|
1999-12-22 19:13:38 +00:00
|
|
|
}
|
|
|
|
|
2009-04-23 21:41:37 +00:00
|
|
|
if (ifp == NULL)
|
|
|
|
return (EADDRNOTAVAIL);
|
|
|
|
|
2009-04-24 09:54:46 +00:00
|
|
|
/*
|
|
|
|
* Security checks before we get involved in any work.
|
|
|
|
*/
|
|
|
|
switch (cmd) {
|
|
|
|
case SIOCAIFADDR:
|
|
|
|
if (td != NULL) {
|
|
|
|
error = priv_check(td, PRIV_NET_ADDIFADDR);
|
|
|
|
if (error)
|
|
|
|
return (error);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case SIOCDIFADDR:
|
|
|
|
if (td != NULL) {
|
|
|
|
error = priv_check(td, PRIV_NET_DELIFADDR);
|
|
|
|
if (error)
|
|
|
|
return (error);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
|
|
|
* Find address for this interface, if it exists.
|
1996-03-15 17:08:07 +00:00
|
|
|
*
|
2009-04-23 21:41:37 +00:00
|
|
|
* If an alias address was specified, find that one instead of the
|
|
|
|
* first one on the interface, if possible.
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
2009-04-23 21:41:37 +00:00
|
|
|
dst = ((struct sockaddr_in *)&ifr->ifr_addr)->sin_addr;
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RLOCK();
|
2009-04-23 21:41:37 +00:00
|
|
|
LIST_FOREACH(iap, INADDR_HASH(dst.s_addr), ia_hash) {
|
|
|
|
if (iap->ia_ifp == ifp &&
|
|
|
|
iap->ia_addr.sin_addr.s_addr == dst.s_addr) {
|
|
|
|
if (td == NULL || prison_check_ip4(td->td_ucred,
|
|
|
|
&dst) == 0)
|
|
|
|
ia = iap;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2009-06-23 20:19:09 +00:00
|
|
|
if (ia != NULL)
|
|
|
|
ifa_ref(&ia->ia_ifa);
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RUNLOCK();
|
2009-04-23 21:41:37 +00:00
|
|
|
if (ia == NULL) {
|
2012-01-05 19:00:36 +00:00
|
|
|
IF_ADDR_RLOCK(ifp);
|
2009-04-23 21:41:37 +00:00
|
|
|
TAILQ_FOREACH(ifa, &ifp->if_addrhead, ifa_link) {
|
|
|
|
iap = ifatoia(ifa);
|
|
|
|
if (iap->ia_addr.sin_family == AF_INET) {
|
|
|
|
if (td != NULL &&
|
|
|
|
prison_check_ip4(td->td_ucred,
|
|
|
|
&iap->ia_addr.sin_addr) != 0)
|
|
|
|
continue;
|
|
|
|
ia = iap;
|
2001-09-29 04:34:11 +00:00
|
|
|
break;
|
|
|
|
}
|
2009-04-23 21:41:37 +00:00
|
|
|
}
|
2009-06-23 20:19:09 +00:00
|
|
|
if (ia != NULL)
|
|
|
|
ifa_ref(&ia->ia_ifa);
|
2012-01-05 19:00:36 +00:00
|
|
|
IF_ADDR_RUNLOCK(ifp);
|
2001-09-29 04:34:11 +00:00
|
|
|
}
|
2009-04-23 21:41:37 +00:00
|
|
|
if (ia == NULL)
|
|
|
|
iaIsFirst = 1;
|
1994-05-24 10:09:53 +00:00
|
|
|
|
2009-04-25 23:02:57 +00:00
|
|
|
error = 0;
|
1994-05-24 10:09:53 +00:00
|
|
|
switch (cmd) {
|
|
|
|
case SIOCAIFADDR:
|
|
|
|
case SIOCDIFADDR:
|
2011-11-24 22:43:48 +00:00
|
|
|
if (ifra->ifra_addr.sin_family == AF_INET) {
|
2009-06-23 20:19:09 +00:00
|
|
|
struct in_ifaddr *oia;
|
|
|
|
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RLOCK();
|
2001-02-04 13:13:25 +00:00
|
|
|
for (oia = ia; ia; ia = TAILQ_NEXT(ia, ia_link)) {
|
1995-04-25 19:50:20 +00:00
|
|
|
if (ia->ia_ifp == ifp &&
|
|
|
|
ia->ia_addr.sin_addr.s_addr ==
|
|
|
|
ifra->ifra_addr.sin_addr.s_addr)
|
|
|
|
break;
|
|
|
|
}
|
2009-06-23 20:19:09 +00:00
|
|
|
if (ia != NULL && ia != oia)
|
|
|
|
ifa_ref(&ia->ia_ifa);
|
|
|
|
if (oia != NULL && ia != oia)
|
|
|
|
ifa_free(&oia->ia_ifa);
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RUNLOCK();
|
1995-05-30 08:16:23 +00:00
|
|
|
if ((ifp->if_flags & IFF_POINTOPOINT)
|
1995-04-25 19:50:20 +00:00
|
|
|
&& (cmd == SIOCAIFADDR)
|
|
|
|
&& (ifra->ifra_dstaddr.sin_addr.s_addr
|
|
|
|
== INADDR_ANY)) {
|
2009-04-25 23:02:57 +00:00
|
|
|
error = EDESTADDRREQ;
|
2009-06-23 20:19:09 +00:00
|
|
|
goto out;
|
1995-04-25 19:50:20 +00:00
|
|
|
}
|
1994-05-24 10:09:53 +00:00
|
|
|
}
|
2009-04-25 23:02:57 +00:00
|
|
|
if (cmd == SIOCDIFADDR && ia == NULL) {
|
|
|
|
error = EADDRNOTAVAIL;
|
2009-06-23 20:19:09 +00:00
|
|
|
goto out;
|
2009-04-25 23:02:57 +00:00
|
|
|
}
|
2008-10-26 19:17:25 +00:00
|
|
|
if (ia == NULL) {
|
1996-12-13 21:29:07 +00:00
|
|
|
ia = (struct in_ifaddr *)
|
2009-04-25 23:02:57 +00:00
|
|
|
malloc(sizeof *ia, M_IFADDR, M_NOWAIT |
|
|
|
|
M_ZERO);
|
|
|
|
if (ia == NULL) {
|
|
|
|
error = ENOBUFS;
|
2009-06-23 20:19:09 +00:00
|
|
|
goto out;
|
2009-04-25 23:02:57 +00:00
|
|
|
}
|
|
|
|
|
2002-12-18 11:46:59 +00:00
|
|
|
ifa = &ia->ia_ifa;
|
2009-06-21 19:30:33 +00:00
|
|
|
ifa_init(ifa);
|
1996-12-13 21:29:07 +00:00
|
|
|
ifa->ifa_addr = (struct sockaddr *)&ia->ia_addr;
|
|
|
|
ifa->ifa_dstaddr = (struct sockaddr *)&ia->ia_dstaddr;
|
|
|
|
ifa->ifa_netmask = (struct sockaddr *)&ia->ia_sockmask;
|
2002-12-18 11:46:59 +00:00
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
ia->ia_sockmask.sin_len = 8;
|
2001-10-30 15:57:20 +00:00
|
|
|
ia->ia_sockmask.sin_family = AF_INET;
|
1994-05-24 10:09:53 +00:00
|
|
|
if (ifp->if_flags & IFF_BROADCAST) {
|
|
|
|
ia->ia_broadaddr.sin_len = sizeof(ia->ia_addr);
|
|
|
|
ia->ia_broadaddr.sin_family = AF_INET;
|
|
|
|
}
|
|
|
|
ia->ia_ifp = ifp;
|
2005-10-28 20:29:42 +00:00
|
|
|
|
2009-06-23 20:19:09 +00:00
|
|
|
ifa_ref(ifa); /* if_addrhead */
|
2012-01-05 19:00:36 +00:00
|
|
|
IF_ADDR_WLOCK(ifp);
|
2009-04-19 22:16:19 +00:00
|
|
|
TAILQ_INSERT_TAIL(&ifp->if_addrhead, ifa, ifa_link);
|
2012-01-05 19:00:36 +00:00
|
|
|
IF_ADDR_WUNLOCK(ifp);
|
2009-06-23 20:19:09 +00:00
|
|
|
ifa_ref(ifa); /* in_ifaddrhead */
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_WLOCK();
|
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
|
|
|
TAILQ_INSERT_TAIL(&V_in_ifaddrhead, ia, ia_link);
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_WUNLOCK();
|
2001-11-30 14:00:55 +00:00
|
|
|
iaIsNew = 1;
|
1994-05-24 10:09:53 +00:00
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case SIOCGIFADDR:
|
|
|
|
case SIOCGIFNETMASK:
|
|
|
|
case SIOCGIFDSTADDR:
|
|
|
|
case SIOCGIFBRDADDR:
|
2009-04-25 23:02:57 +00:00
|
|
|
if (ia == NULL) {
|
|
|
|
error = EADDRNOTAVAIL;
|
2009-06-23 20:19:09 +00:00
|
|
|
goto out;
|
2009-04-25 23:02:57 +00:00
|
|
|
}
|
1994-05-24 10:09:53 +00:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2009-04-25 23:02:57 +00:00
|
|
|
/*
|
2009-06-23 20:19:09 +00:00
|
|
|
* Most paths in this switch return directly or via out. Only paths
|
|
|
|
* that remove the address break in order to hit common removal code.
|
2009-04-25 23:02:57 +00:00
|
|
|
*/
|
|
|
|
switch (cmd) {
|
1994-05-24 10:09:53 +00:00
|
|
|
case SIOCGIFADDR:
|
|
|
|
*((struct sockaddr_in *)&ifr->ifr_addr) = ia->ia_addr;
|
2009-06-23 20:19:09 +00:00
|
|
|
goto out;
|
1994-05-24 10:09:53 +00:00
|
|
|
|
|
|
|
case SIOCGIFBRDADDR:
|
2009-04-25 23:02:57 +00:00
|
|
|
if ((ifp->if_flags & IFF_BROADCAST) == 0) {
|
|
|
|
error = EINVAL;
|
2009-06-23 20:19:09 +00:00
|
|
|
goto out;
|
2009-04-25 23:02:57 +00:00
|
|
|
}
|
1994-05-24 10:09:53 +00:00
|
|
|
*((struct sockaddr_in *)&ifr->ifr_dstaddr) = ia->ia_broadaddr;
|
2009-06-23 20:19:09 +00:00
|
|
|
goto out;
|
1994-05-24 10:09:53 +00:00
|
|
|
|
|
|
|
case SIOCGIFDSTADDR:
|
2009-04-25 23:02:57 +00:00
|
|
|
if ((ifp->if_flags & IFF_POINTOPOINT) == 0) {
|
|
|
|
error = EINVAL;
|
2009-06-23 20:19:09 +00:00
|
|
|
goto out;
|
2009-04-25 23:02:57 +00:00
|
|
|
}
|
1994-05-24 10:09:53 +00:00
|
|
|
*((struct sockaddr_in *)&ifr->ifr_dstaddr) = ia->ia_dstaddr;
|
2009-06-23 20:19:09 +00:00
|
|
|
goto out;
|
1994-05-24 10:09:53 +00:00
|
|
|
|
|
|
|
case SIOCGIFNETMASK:
|
|
|
|
*((struct sockaddr_in *)&ifr->ifr_addr) = ia->ia_sockmask;
|
2009-06-23 20:19:09 +00:00
|
|
|
goto out;
|
1994-05-24 10:09:53 +00:00
|
|
|
|
|
|
|
case SIOCAIFADDR:
|
|
|
|
maskIsNew = 0;
|
|
|
|
hostIsNew = 1;
|
|
|
|
error = 0;
|
2011-11-24 22:46:11 +00:00
|
|
|
if (ifra->ifra_addr.sin_addr.s_addr ==
|
2011-11-21 14:10:13 +00:00
|
|
|
ia->ia_addr.sin_addr.s_addr)
|
|
|
|
hostIsNew = 0;
|
1994-05-24 10:09:53 +00:00
|
|
|
if (ifra->ifra_mask.sin_len) {
|
2012-07-31 11:31:12 +00:00
|
|
|
/*
|
2009-09-15 01:01:03 +00:00
|
|
|
* QL: XXX
|
|
|
|
* Need to scrub the prefix here in case
|
|
|
|
* the issued command is SIOCAIFADDR with
|
|
|
|
* the same address, but with a different
|
|
|
|
* prefix length. And if the prefix length
|
2012-07-31 11:31:12 +00:00
|
|
|
* is the same as before, then the call is
|
2009-09-15 01:01:03 +00:00
|
|
|
* un-necessarily executed here.
|
|
|
|
*/
|
2011-05-29 02:21:35 +00:00
|
|
|
in_ifscrub(ifp, ia, LLE_STATIC);
|
1994-05-24 10:09:53 +00:00
|
|
|
ia->ia_sockmask = ifra->ifra_mask;
|
2001-10-30 15:57:20 +00:00
|
|
|
ia->ia_sockmask.sin_family = AF_INET;
|
1994-05-24 10:09:53 +00:00
|
|
|
ia->ia_subnetmask =
|
2012-07-31 11:31:12 +00:00
|
|
|
ntohl(ia->ia_sockmask.sin_addr.s_addr);
|
1994-05-24 10:09:53 +00:00
|
|
|
maskIsNew = 1;
|
|
|
|
}
|
|
|
|
if ((ifp->if_flags & IFF_POINTOPOINT) &&
|
|
|
|
(ifra->ifra_dstaddr.sin_family == AF_INET)) {
|
2011-05-29 02:21:35 +00:00
|
|
|
in_ifscrub(ifp, ia, LLE_STATIC);
|
1994-05-24 10:09:53 +00:00
|
|
|
ia->ia_dstaddr = ifra->ifra_dstaddr;
|
|
|
|
maskIsNew = 1; /* We lie; but the effect's the same */
|
|
|
|
}
|
2011-11-21 20:07:12 +00:00
|
|
|
if (hostIsNew || maskIsNew)
|
2012-01-16 09:53:24 +00:00
|
|
|
error = in_ifinit(ifp, ia, &ifra->ifra_addr, maskIsNew,
|
|
|
|
(ocmd == cmd ? ifra->ifra_vhid : 0));
|
2001-11-30 14:00:55 +00:00
|
|
|
if (error != 0 && iaIsNew)
|
2010-01-08 17:49:24 +00:00
|
|
|
break;
|
2001-11-30 14:00:55 +00:00
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
if ((ifp->if_flags & IFF_BROADCAST) &&
|
2011-11-21 14:10:13 +00:00
|
|
|
ifra->ifra_broadaddr.sin_len)
|
1994-05-24 10:09:53 +00:00
|
|
|
ia->ia_broadaddr = ifra->ifra_broadaddr;
|
2007-03-29 21:39:22 +00:00
|
|
|
if (error == 0) {
|
2009-03-09 22:54:17 +00:00
|
|
|
ii = ((struct in_ifinfo *)ifp->if_afdata[AF_INET]);
|
2009-03-09 17:53:05 +00:00
|
|
|
if (iaIsFirst &&
|
|
|
|
(ifp->if_flags & IFF_MULTICAST) != 0) {
|
|
|
|
error = in_joingroup(ifp, &allhosts_addr,
|
|
|
|
NULL, &ii->ii_allhosts);
|
|
|
|
}
|
2004-02-26 04:27:55 +00:00
|
|
|
EVENTHANDLER_INVOKE(ifaddr_event, ifp);
|
2007-03-29 21:39:22 +00:00
|
|
|
}
|
2009-06-23 20:19:09 +00:00
|
|
|
goto out;
|
1994-05-24 10:09:53 +00:00
|
|
|
|
|
|
|
case SIOCDIFADDR:
|
net/route.c:
A route generated from an RTF_CLONING route had the RTF_WASCLONED flag
set but did not have a reference to the parent route, as documented in
the rtentry(9) manpage. This prevented such routes from being deleted
when their parent route is deleted.
Now, for example, if you delete an IP address from a network interface,
all ARP entries that were cloned from this interface route are flushed.
This also has an impact on netstat(1) output. Previously, dynamically
created ARP cache entries (RTF_STATIC flag is unset) were displayed as
part of the routing table display (-r). Now, they are only printed if
the -a option is given.
netinet/in.c, netinet/in_rmx.c:
When address is removed from an interface, also delete all routes that
point to this interface and address. Previously, for example, if you
changed the address on an interface, outgoing IP datagrams might still
use the old address. The only solution was to delete and re-add some
routes. (The problem is easily observed with the route(8) command.)
Note, that if the socket was already bound to the local address before
this address is removed, new datagrams generated from this socket will
still be sent from the old address.
PR: kern/20785, kern/21914
Reviewed by: wollman (the idea)
2001-03-15 14:52:12 +00:00
|
|
|
/*
|
|
|
|
* in_ifscrub kills the interface route.
|
|
|
|
*/
|
2011-05-20 19:12:20 +00:00
|
|
|
in_ifscrub(ifp, ia, LLE_STATIC);
|
2009-04-25 23:02:57 +00:00
|
|
|
|
net/route.c:
A route generated from an RTF_CLONING route had the RTF_WASCLONED flag
set but did not have a reference to the parent route, as documented in
the rtentry(9) manpage. This prevented such routes from being deleted
when their parent route is deleted.
Now, for example, if you delete an IP address from a network interface,
all ARP entries that were cloned from this interface route are flushed.
This also has an impact on netstat(1) output. Previously, dynamically
created ARP cache entries (RTF_STATIC flag is unset) were displayed as
part of the routing table display (-r). Now, they are only printed if
the -a option is given.
netinet/in.c, netinet/in_rmx.c:
When address is removed from an interface, also delete all routes that
point to this interface and address. Previously, for example, if you
changed the address on an interface, outgoing IP datagrams might still
use the old address. The only solution was to delete and re-add some
routes. (The problem is easily observed with the route(8) command.)
Note, that if the socket was already bound to the local address before
this address is removed, new datagrams generated from this socket will
still be sent from the old address.
PR: kern/20785, kern/21914
Reviewed by: wollman (the idea)
2001-03-15 14:52:12 +00:00
|
|
|
/*
|
|
|
|
* in_ifadown gets rid of all the rest of
|
|
|
|
* the routes. This is not quite the right
|
|
|
|
* thing to do, but at least if we are running
|
|
|
|
* a routing process they will come back.
|
|
|
|
*/
|
2001-05-11 14:37:34 +00:00
|
|
|
in_ifadown(&ia->ia_ifa, 1);
|
2004-02-26 04:27:55 +00:00
|
|
|
EVENTHANDLER_INVOKE(ifaddr_event, ifp);
|
2001-11-30 14:00:55 +00:00
|
|
|
error = 0;
|
1994-05-24 10:09:53 +00:00
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
2009-04-23 21:41:37 +00:00
|
|
|
panic("in_control: unsupported ioctl");
|
1994-05-24 10:09:53 +00:00
|
|
|
}
|
2001-11-30 14:00:55 +00:00
|
|
|
|
A major overhaul of the CARP implementation. The ip_carp.c was started
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
|
|
|
if (ia->ia_ifa.ifa_carp)
|
|
|
|
(*carp_detach_p)(&ia->ia_ifa);
|
|
|
|
|
2012-01-05 19:00:36 +00:00
|
|
|
IF_ADDR_WLOCK(ifp);
|
2010-10-16 19:53:22 +00:00
|
|
|
/* Re-check that ia is still part of the list. */
|
|
|
|
TAILQ_FOREACH(ifa, &ifp->if_addrhead, ifa_link) {
|
|
|
|
if (ifa == &ia->ia_ifa)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
if (ifa == NULL) {
|
|
|
|
/*
|
|
|
|
* If we lost the race with another thread, there is no need to
|
|
|
|
* try it again for the next loop as there is no other exit
|
|
|
|
* path between here and out.
|
|
|
|
*/
|
2012-01-05 19:00:36 +00:00
|
|
|
IF_ADDR_WUNLOCK(ifp);
|
2010-10-16 19:53:22 +00:00
|
|
|
error = EADDRNOTAVAIL;
|
|
|
|
goto out;
|
|
|
|
}
|
2001-11-30 14:00:55 +00:00
|
|
|
TAILQ_REMOVE(&ifp->if_addrhead, &ia->ia_ifa, ifa_link);
|
2012-01-05 19:00:36 +00:00
|
|
|
IF_ADDR_WUNLOCK(ifp);
|
2012-08-19 11:54:02 +00:00
|
|
|
ifa_free(&ia->ia_ifa); /* if_addrhead */
|
2009-06-25 11:52:33 +00:00
|
|
|
|
|
|
|
IN_IFADDR_WLOCK();
|
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
|
|
|
TAILQ_REMOVE(&V_in_ifaddrhead, ia, ia_link);
|
2009-06-23 20:19:09 +00:00
|
|
|
|
2011-11-21 14:10:13 +00:00
|
|
|
LIST_REMOVE(ia, ia_hash);
|
|
|
|
IN_IFADDR_WUNLOCK();
|
|
|
|
/*
|
|
|
|
* If this is the last IPv4 address configured on this
|
|
|
|
* interface, leave the all-hosts group.
|
|
|
|
* No state-change report need be transmitted.
|
|
|
|
*/
|
|
|
|
IFP_TO_IA(ifp, iap);
|
|
|
|
if (iap == NULL) {
|
|
|
|
ii = ((struct in_ifinfo *)ifp->if_afdata[AF_INET]);
|
|
|
|
IN_MULTI_LOCK();
|
|
|
|
if (ii->ii_allhosts) {
|
|
|
|
(void)in_leavegroup_locked(ii->ii_allhosts, NULL);
|
|
|
|
ii->ii_allhosts = NULL;
|
|
|
|
}
|
|
|
|
IN_MULTI_UNLOCK();
|
2009-06-25 11:52:33 +00:00
|
|
|
} else
|
2011-11-21 14:10:13 +00:00
|
|
|
ifa_free(&iap->ia_ifa);
|
|
|
|
|
2009-06-25 11:52:33 +00:00
|
|
|
ifa_free(&ia->ia_ifa); /* in_ifaddrhead */
|
2009-06-23 20:19:09 +00:00
|
|
|
out:
|
|
|
|
if (ia != NULL)
|
|
|
|
ifa_free(&ia->ia_ifa);
|
2009-04-25 23:02:57 +00:00
|
|
|
return (error);
|
1994-05-24 10:09:53 +00:00
|
|
|
}
|
|
|
|
|
1999-12-22 19:13:38 +00:00
|
|
|
/*
|
|
|
|
* SIOC[GAD]LIFADDR.
|
|
|
|
* SIOCGLIFADDR: get first address. (?!?)
|
|
|
|
* SIOCGLIFADDR with IFLR_PREFIX:
|
|
|
|
* get first address that matches the specified prefix.
|
|
|
|
* SIOCALIFADDR: add the specified address.
|
|
|
|
* SIOCALIFADDR with IFLR_PREFIX:
|
|
|
|
* EINVAL since we can't deduce hostid part of the address.
|
|
|
|
* SIOCDLIFADDR: delete the specified address.
|
|
|
|
* SIOCDLIFADDR with IFLR_PREFIX:
|
|
|
|
* delete the first address that matches the specified prefix.
|
|
|
|
* return values:
|
|
|
|
* EINVAL on invalid parameters
|
|
|
|
* EADDRNOTAVAIL on prefix match failed/specified address not found
|
|
|
|
* other values may be returned from in_ioctl()
|
|
|
|
*/
|
|
|
|
static int
|
2007-05-10 15:58:48 +00:00
|
|
|
in_lifaddr_ioctl(struct socket *so, u_long cmd, caddr_t data,
|
|
|
|
struct ifnet *ifp, struct thread *td)
|
1999-12-22 19:13:38 +00:00
|
|
|
{
|
|
|
|
struct if_laddrreq *iflr = (struct if_laddrreq *)data;
|
|
|
|
struct ifaddr *ifa;
|
|
|
|
|
|
|
|
/* sanity checks */
|
2008-10-26 19:17:25 +00:00
|
|
|
if (data == NULL || ifp == NULL) {
|
1999-12-22 19:13:38 +00:00
|
|
|
panic("invalid argument to in_lifaddr_ioctl");
|
|
|
|
/*NOTRECHED*/
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (cmd) {
|
|
|
|
case SIOCGLIFADDR:
|
|
|
|
/* address must be specified on GET with IFLR_PREFIX */
|
|
|
|
if ((iflr->flags & IFLR_PREFIX) == 0)
|
|
|
|
break;
|
|
|
|
/*FALLTHROUGH*/
|
|
|
|
case SIOCALIFADDR:
|
|
|
|
case SIOCDLIFADDR:
|
|
|
|
/* address must be specified on ADD and DELETE */
|
2000-01-13 14:52:53 +00:00
|
|
|
if (iflr->addr.ss_family != AF_INET)
|
2008-10-26 19:17:25 +00:00
|
|
|
return (EINVAL);
|
2000-01-13 14:52:53 +00:00
|
|
|
if (iflr->addr.ss_len != sizeof(struct sockaddr_in))
|
2008-10-26 19:17:25 +00:00
|
|
|
return (EINVAL);
|
1999-12-22 19:13:38 +00:00
|
|
|
/* XXX need improvement */
|
2000-01-13 14:52:53 +00:00
|
|
|
if (iflr->dstaddr.ss_family
|
|
|
|
&& iflr->dstaddr.ss_family != AF_INET)
|
2008-10-26 19:17:25 +00:00
|
|
|
return (EINVAL);
|
2000-01-13 14:52:53 +00:00
|
|
|
if (iflr->dstaddr.ss_family
|
|
|
|
&& iflr->dstaddr.ss_len != sizeof(struct sockaddr_in))
|
2008-10-26 19:17:25 +00:00
|
|
|
return (EINVAL);
|
1999-12-22 19:13:38 +00:00
|
|
|
break;
|
|
|
|
default: /*shouldn't happen*/
|
2008-10-26 19:17:25 +00:00
|
|
|
return (EOPNOTSUPP);
|
1999-12-22 19:13:38 +00:00
|
|
|
}
|
|
|
|
if (sizeof(struct in_addr) * 8 < iflr->prefixlen)
|
2008-10-26 19:17:25 +00:00
|
|
|
return (EINVAL);
|
1999-12-22 19:13:38 +00:00
|
|
|
|
|
|
|
switch (cmd) {
|
|
|
|
case SIOCALIFADDR:
|
|
|
|
{
|
|
|
|
struct in_aliasreq ifra;
|
|
|
|
|
|
|
|
if (iflr->flags & IFLR_PREFIX)
|
2008-10-26 19:17:25 +00:00
|
|
|
return (EINVAL);
|
1999-12-22 19:13:38 +00:00
|
|
|
|
2012-01-04 13:23:51 +00:00
|
|
|
/* copy args to in_aliasreq, perform ioctl(SIOCAIFADDR). */
|
1999-12-22 19:13:38 +00:00
|
|
|
bzero(&ifra, sizeof(ifra));
|
|
|
|
bcopy(iflr->iflr_name, ifra.ifra_name,
|
|
|
|
sizeof(ifra.ifra_name));
|
|
|
|
|
2000-01-13 14:52:53 +00:00
|
|
|
bcopy(&iflr->addr, &ifra.ifra_addr, iflr->addr.ss_len);
|
1999-12-22 19:13:38 +00:00
|
|
|
|
2000-01-13 14:52:53 +00:00
|
|
|
if (iflr->dstaddr.ss_family) { /*XXX*/
|
1999-12-22 19:13:38 +00:00
|
|
|
bcopy(&iflr->dstaddr, &ifra.ifra_dstaddr,
|
2000-01-13 14:52:53 +00:00
|
|
|
iflr->dstaddr.ss_len);
|
1999-12-22 19:13:38 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
ifra.ifra_mask.sin_family = AF_INET;
|
|
|
|
ifra.ifra_mask.sin_len = sizeof(struct sockaddr_in);
|
|
|
|
in_len2mask(&ifra.ifra_mask.sin_addr, iflr->prefixlen);
|
|
|
|
|
2008-10-26 19:17:25 +00:00
|
|
|
return (in_control(so, SIOCAIFADDR, (caddr_t)&ifra, ifp, td));
|
1999-12-22 19:13:38 +00:00
|
|
|
}
|
|
|
|
case SIOCGLIFADDR:
|
|
|
|
case SIOCDLIFADDR:
|
|
|
|
{
|
|
|
|
struct in_ifaddr *ia;
|
|
|
|
struct in_addr mask, candidate, match;
|
|
|
|
struct sockaddr_in *sin;
|
|
|
|
|
|
|
|
bzero(&mask, sizeof(mask));
|
2007-06-17 00:31:24 +00:00
|
|
|
bzero(&match, sizeof(match));
|
1999-12-22 19:13:38 +00:00
|
|
|
if (iflr->flags & IFLR_PREFIX) {
|
|
|
|
/* lookup a prefix rather than address. */
|
|
|
|
in_len2mask(&mask, iflr->prefixlen);
|
|
|
|
|
|
|
|
sin = (struct sockaddr_in *)&iflr->addr;
|
|
|
|
match.s_addr = sin->sin_addr.s_addr;
|
|
|
|
match.s_addr &= mask.s_addr;
|
|
|
|
|
|
|
|
/* if you set extra bits, that's wrong */
|
|
|
|
if (match.s_addr != sin->sin_addr.s_addr)
|
2008-10-26 19:17:25 +00:00
|
|
|
return (EINVAL);
|
1999-12-22 19:13:38 +00:00
|
|
|
|
|
|
|
} else {
|
2007-06-17 00:31:24 +00:00
|
|
|
/* on getting an address, take the 1st match */
|
|
|
|
/* on deleting an address, do exact match */
|
|
|
|
if (cmd != SIOCGLIFADDR) {
|
1999-12-22 19:13:38 +00:00
|
|
|
in_len2mask(&mask, 32);
|
|
|
|
sin = (struct sockaddr_in *)&iflr->addr;
|
|
|
|
match.s_addr = sin->sin_addr.s_addr;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2012-01-05 19:00:36 +00:00
|
|
|
IF_ADDR_RLOCK(ifp);
|
1999-12-22 19:13:38 +00:00
|
|
|
TAILQ_FOREACH(ifa, &ifp->if_addrhead, ifa_link) {
|
2012-01-04 13:23:51 +00:00
|
|
|
if (ifa->ifa_addr->sa_family != AF_INET)
|
1999-12-22 19:13:38 +00:00
|
|
|
continue;
|
2007-06-17 00:31:24 +00:00
|
|
|
if (match.s_addr == 0)
|
1999-12-22 19:13:38 +00:00
|
|
|
break;
|
2012-01-04 13:29:26 +00:00
|
|
|
sin = (struct sockaddr_in *)&ifa->ifa_addr;
|
|
|
|
candidate.s_addr = sin->sin_addr.s_addr;
|
1999-12-22 19:13:38 +00:00
|
|
|
candidate.s_addr &= mask.s_addr;
|
|
|
|
if (candidate.s_addr == match.s_addr)
|
|
|
|
break;
|
|
|
|
}
|
2012-01-04 13:26:56 +00:00
|
|
|
if (ifa != NULL)
|
|
|
|
ifa_ref(ifa);
|
2012-01-05 19:00:36 +00:00
|
|
|
IF_ADDR_RUNLOCK(ifp);
|
2008-10-26 19:17:25 +00:00
|
|
|
if (ifa == NULL)
|
|
|
|
return (EADDRNOTAVAIL);
|
1999-12-22 19:13:38 +00:00
|
|
|
ia = (struct in_ifaddr *)ifa;
|
|
|
|
|
|
|
|
if (cmd == SIOCGLIFADDR) {
|
|
|
|
/* fill in the if_laddrreq structure */
|
|
|
|
bcopy(&ia->ia_addr, &iflr->addr, ia->ia_addr.sin_len);
|
|
|
|
|
|
|
|
if ((ifp->if_flags & IFF_POINTOPOINT) != 0) {
|
|
|
|
bcopy(&ia->ia_dstaddr, &iflr->dstaddr,
|
|
|
|
ia->ia_dstaddr.sin_len);
|
|
|
|
} else
|
|
|
|
bzero(&iflr->dstaddr, sizeof(iflr->dstaddr));
|
|
|
|
|
|
|
|
iflr->prefixlen =
|
|
|
|
in_mask2len(&ia->ia_sockmask.sin_addr);
|
|
|
|
|
|
|
|
iflr->flags = 0; /*XXX*/
|
2012-01-04 13:26:56 +00:00
|
|
|
ifa_free(ifa);
|
1999-12-22 19:13:38 +00:00
|
|
|
|
2008-10-26 19:17:25 +00:00
|
|
|
return (0);
|
1999-12-22 19:13:38 +00:00
|
|
|
} else {
|
|
|
|
struct in_aliasreq ifra;
|
|
|
|
|
2012-01-04 13:23:51 +00:00
|
|
|
/* fill in_aliasreq and do ioctl(SIOCDIFADDR) */
|
1999-12-22 19:13:38 +00:00
|
|
|
bzero(&ifra, sizeof(ifra));
|
|
|
|
bcopy(iflr->iflr_name, ifra.ifra_name,
|
|
|
|
sizeof(ifra.ifra_name));
|
|
|
|
|
|
|
|
bcopy(&ia->ia_addr, &ifra.ifra_addr,
|
|
|
|
ia->ia_addr.sin_len);
|
|
|
|
if ((ifp->if_flags & IFF_POINTOPOINT) != 0) {
|
|
|
|
bcopy(&ia->ia_dstaddr, &ifra.ifra_dstaddr,
|
|
|
|
ia->ia_dstaddr.sin_len);
|
|
|
|
}
|
|
|
|
bcopy(&ia->ia_sockmask, &ifra.ifra_dstaddr,
|
|
|
|
ia->ia_sockmask.sin_len);
|
2012-01-04 13:26:56 +00:00
|
|
|
ifa_free(ifa);
|
1999-12-22 19:13:38 +00:00
|
|
|
|
2008-10-26 19:17:25 +00:00
|
|
|
return (in_control(so, SIOCDIFADDR, (caddr_t)&ifra,
|
|
|
|
ifp, td));
|
1999-12-22 19:13:38 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-10-26 19:17:25 +00:00
|
|
|
return (EOPNOTSUPP); /*just for safety*/
|
1999-12-22 19:13:38 +00:00
|
|
|
}
|
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
|
|
|
* Delete any existing route for an interface.
|
|
|
|
*/
|
1997-02-13 19:46:45 +00:00
|
|
|
void
|
2011-05-20 19:12:20 +00:00
|
|
|
in_ifscrub(struct ifnet *ifp, struct in_ifaddr *ia, u_int flags)
|
1994-05-24 10:09:53 +00:00
|
|
|
{
|
2007-05-10 15:58:48 +00:00
|
|
|
|
2011-05-20 19:12:20 +00:00
|
|
|
in_scrubprefix(ia, flags);
|
1994-05-24 10:09:53 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Initialize an interface's internet address
|
|
|
|
* and routing table entry.
|
|
|
|
*/
|
1995-11-14 20:34:56 +00:00
|
|
|
static int
|
2007-05-10 15:58:48 +00:00
|
|
|
in_ifinit(struct ifnet *ifp, struct in_ifaddr *ia, struct sockaddr_in *sin,
|
2012-01-16 09:53:24 +00:00
|
|
|
int masksupplied, int vhid)
|
1994-05-24 10:09:53 +00:00
|
|
|
{
|
|
|
|
register u_long i = ntohl(sin->sin_addr.s_addr);
|
2012-12-24 21:35:48 +00:00
|
|
|
int flags, error = 0;
|
1994-05-24 10:09:53 +00:00
|
|
|
|
2011-11-21 14:10:13 +00:00
|
|
|
IN_IFADDR_WLOCK();
|
|
|
|
if (ia->ia_addr.sin_family == AF_INET)
|
2002-10-22 22:50:38 +00:00
|
|
|
LIST_REMOVE(ia, ia_hash);
|
1994-05-24 10:09:53 +00:00
|
|
|
ia->ia_addr = *sin;
|
2011-11-21 14:10:13 +00:00
|
|
|
LIST_INSERT_HEAD(INADDR_HASH(ia->ia_addr.sin_addr.s_addr),
|
|
|
|
ia, ia_hash);
|
|
|
|
IN_IFADDR_WUNLOCK();
|
|
|
|
|
A major overhaul of the CARP implementation. The ip_carp.c was started
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
|
|
|
if (vhid > 0) {
|
|
|
|
if (carp_attach_p != NULL)
|
|
|
|
error = (*carp_attach_p)(&ia->ia_ifa, vhid);
|
|
|
|
else
|
|
|
|
error = EPROTONOSUPPORT;
|
|
|
|
}
|
|
|
|
if (error)
|
|
|
|
return (error);
|
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
2013-01-03 10:21:28 +00:00
|
|
|
* Give the interface a chance to initialize
|
|
|
|
* if this is its first address,
|
|
|
|
* and to validate the address if necessary.
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
2011-11-21 14:10:13 +00:00
|
|
|
if (ifp->if_ioctl != NULL &&
|
|
|
|
(error = (*ifp->if_ioctl)(ifp, SIOCSIFADDR, (caddr_t)ia)) != 0)
|
2005-06-02 00:04:08 +00:00
|
|
|
/* LIST_REMOVE(ia, ia_hash) is done in in_control */
|
|
|
|
return (error);
|
2011-11-21 14:10:13 +00:00
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
2011-10-15 16:28:06 +00:00
|
|
|
* Be compatible with network classes, if netmask isn't supplied,
|
|
|
|
* guess it based on classes.
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
2011-12-06 20:55:20 +00:00
|
|
|
if (!masksupplied) {
|
2011-10-15 16:28:06 +00:00
|
|
|
if (IN_CLASSA(i))
|
|
|
|
ia->ia_subnetmask = IN_CLASSA_NET;
|
|
|
|
else if (IN_CLASSB(i))
|
|
|
|
ia->ia_subnetmask = IN_CLASSB_NET;
|
|
|
|
else
|
|
|
|
ia->ia_subnetmask = IN_CLASSC_NET;
|
1994-05-24 10:09:53 +00:00
|
|
|
ia->ia_sockmask.sin_addr.s_addr = htonl(ia->ia_subnetmask);
|
2011-10-15 16:28:06 +00:00
|
|
|
}
|
1994-05-24 10:09:53 +00:00
|
|
|
ia->ia_subnet = i & ia->ia_subnetmask;
|
|
|
|
in_socktrim(&ia->ia_sockmask);
|
2012-12-24 21:35:48 +00:00
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
/*
|
|
|
|
* Add route for the network.
|
|
|
|
*/
|
2012-12-24 21:35:48 +00:00
|
|
|
flags = RTF_UP;
|
1994-05-24 10:09:53 +00:00
|
|
|
ia->ia_ifa.ifa_metric = ifp->if_metric;
|
|
|
|
if (ifp->if_flags & IFF_BROADCAST) {
|
2011-10-15 18:41:25 +00:00
|
|
|
if (ia->ia_subnetmask == IN_RFC3021_MASK)
|
|
|
|
ia->ia_broadaddr.sin_addr.s_addr = INADDR_BROADCAST;
|
|
|
|
else
|
|
|
|
ia->ia_broadaddr.sin_addr.s_addr =
|
|
|
|
htonl(ia->ia_subnet | ~ia->ia_subnetmask);
|
1994-05-24 10:09:53 +00:00
|
|
|
} else if (ifp->if_flags & IFF_LOOPBACK) {
|
2004-11-17 23:14:03 +00:00
|
|
|
ia->ia_dstaddr = ia->ia_addr;
|
1994-05-24 10:09:53 +00:00
|
|
|
flags |= RTF_HOST;
|
|
|
|
} else if (ifp->if_flags & IFF_POINTOPOINT) {
|
|
|
|
if (ia->ia_dstaddr.sin_family != AF_INET)
|
|
|
|
return (0);
|
|
|
|
flags |= RTF_HOST;
|
|
|
|
}
|
A major overhaul of the CARP implementation. The ip_carp.c was started
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
|
|
|
if (!vhid && (error = in_addprefix(ia, flags)) != 0)
|
2004-11-12 20:53:51 +00:00
|
|
|
return (error);
|
2001-11-30 14:00:55 +00:00
|
|
|
|
2009-05-14 05:27:09 +00:00
|
|
|
if (ia->ia_addr.sin_addr.s_addr == INADDR_ANY)
|
|
|
|
return (0);
|
|
|
|
|
2011-11-21 14:10:13 +00:00
|
|
|
if (ifp->if_flags & IFF_POINTOPOINT &&
|
|
|
|
ia->ia_dstaddr.sin_addr.s_addr == ia->ia_addr.sin_addr.s_addr)
|
2010-02-02 20:38:30 +00:00
|
|
|
return (0);
|
|
|
|
|
2009-05-12 07:41:20 +00:00
|
|
|
/*
|
|
|
|
* add a loopback route to self
|
|
|
|
*/
|
A major overhaul of the CARP implementation. The ip_carp.c was started
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
|
|
|
if (V_useloopback && !vhid && !(ifp->if_flags & IFF_LOOPBACK)) {
|
2009-12-30 21:35:34 +00:00
|
|
|
struct route ia_ro;
|
|
|
|
|
|
|
|
bzero(&ia_ro, sizeof(ia_ro));
|
|
|
|
*((struct sockaddr_in *)(&ia_ro.ro_dst)) = ia->ia_addr;
|
2012-02-03 13:08:44 +00:00
|
|
|
rtalloc_ign_fib(&ia_ro, 0, RT_DEFAULT_FIB);
|
2009-12-30 21:35:34 +00:00
|
|
|
if ((ia_ro.ro_rt != NULL) && (ia_ro.ro_rt->rt_ifp != NULL) &&
|
|
|
|
(ia_ro.ro_rt->rt_ifp == V_loif)) {
|
|
|
|
RT_LOCK(ia_ro.ro_rt);
|
|
|
|
RT_ADDREF(ia_ro.ro_rt);
|
|
|
|
RTFREE_LOCKED(ia_ro.ro_rt);
|
|
|
|
} else
|
2012-07-31 11:31:12 +00:00
|
|
|
error = ifa_add_loopback_route((struct ifaddr *)ia,
|
|
|
|
(struct sockaddr *)&ia->ia_addr);
|
2009-12-30 21:35:34 +00:00
|
|
|
if (error == 0)
|
|
|
|
ia->ia_flags |= IFA_RTSELF;
|
|
|
|
if (ia_ro.ro_rt != NULL)
|
|
|
|
RTFREE(ia_ro.ro_rt);
|
|
|
|
}
|
2009-05-12 07:41:20 +00:00
|
|
|
|
1994-05-24 10:09:53 +00:00
|
|
|
return (error);
|
|
|
|
}
|
|
|
|
|
2004-11-12 20:53:51 +00:00
|
|
|
#define rtinitflags(x) \
|
|
|
|
((((x)->ia_ifp->if_flags & (IFF_LOOPBACK | IFF_POINTOPOINT)) != 0) \
|
|
|
|
? RTF_HOST : 0)
|
2009-12-30 22:13:01 +00:00
|
|
|
|
|
|
|
/*
|
2012-07-31 11:31:12 +00:00
|
|
|
* Generate a routing message when inserting or deleting
|
2009-12-30 22:13:01 +00:00
|
|
|
* an interface address alias.
|
|
|
|
*/
|
2012-07-31 11:31:12 +00:00
|
|
|
static void in_addralias_rtmsg(int cmd, struct in_addr *prefix,
|
2009-12-30 22:13:01 +00:00
|
|
|
struct in_ifaddr *target)
|
|
|
|
{
|
|
|
|
struct route pfx_ro;
|
|
|
|
struct sockaddr_in *pfx_addr;
|
|
|
|
struct rtentry msg_rt;
|
|
|
|
|
|
|
|
/* QL: XXX
|
|
|
|
* This is a bit questionable because there is no
|
|
|
|
* additional route entry added/deleted for an address
|
|
|
|
* alias. Therefore this route report is inaccurate.
|
|
|
|
*/
|
|
|
|
bzero(&pfx_ro, sizeof(pfx_ro));
|
|
|
|
pfx_addr = (struct sockaddr_in *)(&pfx_ro.ro_dst);
|
|
|
|
pfx_addr->sin_len = sizeof(*pfx_addr);
|
|
|
|
pfx_addr->sin_family = AF_INET;
|
|
|
|
pfx_addr->sin_addr = *prefix;
|
|
|
|
rtalloc_ign_fib(&pfx_ro, 0, 0);
|
|
|
|
if (pfx_ro.ro_rt != NULL) {
|
|
|
|
msg_rt = *pfx_ro.ro_rt;
|
|
|
|
|
|
|
|
/* QL: XXX
|
|
|
|
* Point the gateway to the new interface
|
2012-07-31 11:31:12 +00:00
|
|
|
* address as if a new prefix route entry has
|
|
|
|
* been added through the new address alias.
|
|
|
|
* All other parts of the rtentry is accurate,
|
2009-12-30 22:13:01 +00:00
|
|
|
* e.g., rt_key, rt_mask, rt_ifp etc.
|
|
|
|
*/
|
2012-07-31 11:31:12 +00:00
|
|
|
msg_rt.rt_gateway = (struct sockaddr *)&target->ia_addr;
|
|
|
|
rt_newaddrmsg(cmd, (struct ifaddr *)target, 0, &msg_rt);
|
2009-12-30 22:13:01 +00:00
|
|
|
RTFREE(pfx_ro.ro_rt);
|
|
|
|
}
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2004-11-12 20:53:51 +00:00
|
|
|
/*
|
2007-06-17 00:31:24 +00:00
|
|
|
* Check if we have a route for the given prefix already or add one accordingly.
|
2004-11-12 20:53:51 +00:00
|
|
|
*/
|
A major overhaul of the CARP implementation. The ip_carp.c was started
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
|
|
|
int
|
2007-05-10 15:58:48 +00:00
|
|
|
in_addprefix(struct in_ifaddr *target, int flags)
|
2004-11-12 20:53:51 +00:00
|
|
|
{
|
|
|
|
struct in_ifaddr *ia;
|
2005-10-22 14:50:27 +00:00
|
|
|
struct in_addr prefix, mask, p, m;
|
2004-11-12 20:53:51 +00:00
|
|
|
int error;
|
|
|
|
|
2007-06-17 00:31:24 +00:00
|
|
|
if ((flags & RTF_HOST) != 0) {
|
2004-11-12 20:53:51 +00:00
|
|
|
prefix = target->ia_dstaddr.sin_addr;
|
2007-06-17 00:31:24 +00:00
|
|
|
mask.s_addr = 0;
|
|
|
|
} else {
|
2004-11-12 20:53:51 +00:00
|
|
|
prefix = target->ia_addr.sin_addr;
|
|
|
|
mask = target->ia_sockmask.sin_addr;
|
|
|
|
prefix.s_addr &= mask.s_addr;
|
|
|
|
}
|
|
|
|
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RLOCK();
|
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
|
|
|
TAILQ_FOREACH(ia, &V_in_ifaddrhead, ia_link) {
|
2005-10-22 14:50:27 +00:00
|
|
|
if (rtinitflags(ia)) {
|
2011-08-10 06:17:06 +00:00
|
|
|
p = ia->ia_dstaddr.sin_addr;
|
2004-11-12 20:53:51 +00:00
|
|
|
|
2005-10-22 14:50:27 +00:00
|
|
|
if (prefix.s_addr != p.s_addr)
|
|
|
|
continue;
|
|
|
|
} else {
|
|
|
|
p = ia->ia_addr.sin_addr;
|
|
|
|
m = ia->ia_sockmask.sin_addr;
|
|
|
|
p.s_addr &= m.s_addr;
|
|
|
|
|
|
|
|
if (prefix.s_addr != p.s_addr ||
|
|
|
|
mask.s_addr != m.s_addr)
|
|
|
|
continue;
|
|
|
|
}
|
2004-11-12 20:53:51 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If we got a matching prefix route inserted by other
|
|
|
|
* interface address, we are done here.
|
|
|
|
*/
|
2005-08-18 10:34:30 +00:00
|
|
|
if (ia->ia_flags & IFA_ROUTE) {
|
2010-03-09 01:11:45 +00:00
|
|
|
#ifdef RADIX_MPATH
|
2012-07-31 11:31:12 +00:00
|
|
|
if (ia->ia_addr.sin_addr.s_addr ==
|
2010-09-04 16:06:01 +00:00
|
|
|
target->ia_addr.sin_addr.s_addr) {
|
|
|
|
IN_IFADDR_RUNLOCK();
|
2010-03-09 01:11:45 +00:00
|
|
|
return (EEXIST);
|
2010-09-04 16:06:01 +00:00
|
|
|
} else
|
2010-03-09 01:11:45 +00:00
|
|
|
break;
|
|
|
|
#endif
|
A major overhaul of the CARP implementation. The ip_carp.c was started
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
|
|
|
if (V_nosameprefix) {
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RUNLOCK();
|
2005-08-18 10:34:30 +00:00
|
|
|
return (EEXIST);
|
2009-06-25 11:52:33 +00:00
|
|
|
} else {
|
2009-12-30 22:13:01 +00:00
|
|
|
in_addralias_rtmsg(RTM_ADD, &prefix, target);
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RUNLOCK();
|
2005-08-18 10:34:30 +00:00
|
|
|
return (0);
|
2009-06-25 11:52:33 +00:00
|
|
|
}
|
2005-08-18 10:34:30 +00:00
|
|
|
}
|
2004-11-12 20:53:51 +00:00
|
|
|
}
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RUNLOCK();
|
2004-11-12 20:53:51 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* No-one seem to have this prefix route, so we try to insert it.
|
|
|
|
*/
|
|
|
|
error = rtinit(&target->ia_ifa, (int)RTM_ADD, flags);
|
|
|
|
if (!error)
|
|
|
|
target->ia_flags |= IFA_ROUTE;
|
2008-10-26 19:17:25 +00:00
|
|
|
return (error);
|
2004-11-12 20:53:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If there is no other address in the system that can serve a route to the
|
|
|
|
* same prefix, remove the route. Hand over the route to the new address
|
|
|
|
* otherwise.
|
|
|
|
*/
|
A major overhaul of the CARP implementation. The ip_carp.c was started
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
|
|
|
int
|
2011-05-20 19:12:20 +00:00
|
|
|
in_scrubprefix(struct in_ifaddr *target, u_int flags)
|
2004-11-12 20:53:51 +00:00
|
|
|
{
|
|
|
|
struct in_ifaddr *ia;
|
2011-12-13 06:56:43 +00:00
|
|
|
struct in_addr prefix, mask, p, m;
|
2009-12-30 21:35:34 +00:00
|
|
|
int error = 0;
|
2009-05-20 21:07:15 +00:00
|
|
|
struct sockaddr_in prefix0, mask0;
|
2004-11-12 20:53:51 +00:00
|
|
|
|
2009-07-27 17:08:06 +00:00
|
|
|
/*
|
|
|
|
* Remove the loopback route to the interface address.
|
|
|
|
* The "useloopback" setting is not consulted because if the
|
|
|
|
* user configures an interface address, turns off this
|
|
|
|
* setting, and then tries to delete that interface address,
|
|
|
|
* checking the current setting of "useloopback" would leave
|
|
|
|
* that interface address loopback route untouched, which
|
|
|
|
* would be wrong. Therefore the interface address loopback route
|
|
|
|
* deletion is unconditional.
|
|
|
|
*/
|
2009-05-14 05:27:09 +00:00
|
|
|
if ((target->ia_addr.sin_addr.s_addr != INADDR_ANY) &&
|
2009-12-30 21:35:34 +00:00
|
|
|
!(target->ia_ifp->if_flags & IFF_LOOPBACK) &&
|
|
|
|
(target->ia_flags & IFA_RTSELF)) {
|
|
|
|
struct route ia_ro;
|
|
|
|
int freeit = 0;
|
|
|
|
|
|
|
|
bzero(&ia_ro, sizeof(ia_ro));
|
|
|
|
*((struct sockaddr_in *)(&ia_ro.ro_dst)) = target->ia_addr;
|
|
|
|
rtalloc_ign_fib(&ia_ro, 0, 0);
|
|
|
|
if ((ia_ro.ro_rt != NULL) && (ia_ro.ro_rt->rt_ifp != NULL) &&
|
|
|
|
(ia_ro.ro_rt->rt_ifp == V_loif)) {
|
|
|
|
RT_LOCK(ia_ro.ro_rt);
|
|
|
|
if (ia_ro.ro_rt->rt_refcnt <= 1)
|
|
|
|
freeit = 1;
|
2011-10-07 18:01:34 +00:00
|
|
|
else if (flags & LLE_STATIC) {
|
2009-12-30 21:35:34 +00:00
|
|
|
RT_REMREF(ia_ro.ro_rt);
|
2011-10-07 18:01:34 +00:00
|
|
|
target->ia_flags &= ~IFA_RTSELF;
|
|
|
|
}
|
2009-12-30 21:35:34 +00:00
|
|
|
RTFREE_LOCKED(ia_ro.ro_rt);
|
|
|
|
}
|
2011-05-20 19:12:20 +00:00
|
|
|
if (freeit && (flags & LLE_STATIC)) {
|
2009-12-30 21:35:34 +00:00
|
|
|
error = ifa_del_loopback_route((struct ifaddr *)target,
|
2012-07-31 11:31:12 +00:00
|
|
|
(struct sockaddr *)&target->ia_addr);
|
2011-05-20 19:12:20 +00:00
|
|
|
if (error == 0)
|
|
|
|
target->ia_flags &= ~IFA_RTSELF;
|
|
|
|
}
|
2011-10-07 22:22:19 +00:00
|
|
|
if ((flags & LLE_STATIC) &&
|
|
|
|
!(target->ia_ifp->if_flags & IFF_NOARP))
|
2011-05-20 19:12:20 +00:00
|
|
|
/* remove arp cache */
|
|
|
|
arp_ifscrub(target->ia_ifp, IA_SIN(target)->sin_addr.s_addr);
|
2009-05-12 07:41:20 +00:00
|
|
|
}
|
|
|
|
|
2011-12-13 06:56:43 +00:00
|
|
|
if (rtinitflags(target)) {
|
2004-11-12 20:53:51 +00:00
|
|
|
prefix = target->ia_dstaddr.sin_addr;
|
2011-12-13 06:56:43 +00:00
|
|
|
mask.s_addr = 0;
|
|
|
|
} else {
|
2004-11-12 20:53:51 +00:00
|
|
|
prefix = target->ia_addr.sin_addr;
|
|
|
|
mask = target->ia_sockmask.sin_addr;
|
|
|
|
prefix.s_addr &= mask.s_addr;
|
|
|
|
}
|
|
|
|
|
2009-12-30 22:13:01 +00:00
|
|
|
if ((target->ia_flags & IFA_ROUTE) == 0) {
|
|
|
|
in_addralias_rtmsg(RTM_DELETE, &prefix, target);
|
|
|
|
return (0);
|
|
|
|
}
|
|
|
|
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RLOCK();
|
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
|
|
|
TAILQ_FOREACH(ia, &V_in_ifaddrhead, ia_link) {
|
2011-12-13 06:56:43 +00:00
|
|
|
if (rtinitflags(ia)) {
|
2004-11-12 20:53:51 +00:00
|
|
|
p = ia->ia_dstaddr.sin_addr;
|
2011-12-13 06:56:43 +00:00
|
|
|
|
|
|
|
if (prefix.s_addr != p.s_addr)
|
|
|
|
continue;
|
|
|
|
} else {
|
2004-11-12 20:53:51 +00:00
|
|
|
p = ia->ia_addr.sin_addr;
|
2011-12-13 06:56:43 +00:00
|
|
|
m = ia->ia_sockmask.sin_addr;
|
|
|
|
p.s_addr &= m.s_addr;
|
|
|
|
|
|
|
|
if (prefix.s_addr != p.s_addr ||
|
|
|
|
mask.s_addr != m.s_addr)
|
|
|
|
continue;
|
2004-11-12 20:53:51 +00:00
|
|
|
}
|
|
|
|
|
2011-12-13 06:56:43 +00:00
|
|
|
if ((ia->ia_ifp->if_flags & IFF_UP) == 0)
|
2004-11-12 20:53:51 +00:00
|
|
|
continue;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If we got a matching prefix address, move IFA_ROUTE and
|
|
|
|
* the route itself to it. Make sure that routing daemons
|
|
|
|
* get a heads-up.
|
|
|
|
*/
|
A major overhaul of the CARP implementation. The ip_carp.c was started
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
2011-12-16 12:16:56 +00:00
|
|
|
if ((ia->ia_flags & IFA_ROUTE) == 0) {
|
2011-03-21 14:19:40 +00:00
|
|
|
ifa_ref(&ia->ia_ifa);
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RUNLOCK();
|
2011-05-29 02:21:35 +00:00
|
|
|
error = rtinit(&(target->ia_ifa), (int)RTM_DELETE,
|
2004-11-12 20:53:51 +00:00
|
|
|
rtinitflags(target));
|
2011-05-29 02:21:35 +00:00
|
|
|
if (error == 0)
|
|
|
|
target->ia_flags &= ~IFA_ROUTE;
|
|
|
|
else
|
|
|
|
log(LOG_INFO, "in_scrubprefix: err=%d, old prefix delete failed\n",
|
|
|
|
error);
|
2004-11-12 20:53:51 +00:00
|
|
|
error = rtinit(&ia->ia_ifa, (int)RTM_ADD,
|
|
|
|
rtinitflags(ia) | RTF_UP);
|
|
|
|
if (error == 0)
|
|
|
|
ia->ia_flags |= IFA_ROUTE;
|
2011-05-29 02:21:35 +00:00
|
|
|
else
|
|
|
|
log(LOG_INFO, "in_scrubprefix: err=%d, new prefix add failed\n",
|
|
|
|
error);
|
2011-03-21 14:19:40 +00:00
|
|
|
ifa_free(&ia->ia_ifa);
|
2008-10-26 19:17:25 +00:00
|
|
|
return (error);
|
2004-11-12 20:53:51 +00:00
|
|
|
}
|
|
|
|
}
|
2009-06-25 11:52:33 +00:00
|
|
|
IN_IFADDR_RUNLOCK();
|
2004-11-12 20:53:51 +00:00
|
|
|
|
2009-05-20 21:07:15 +00:00
|
|
|
/*
|
|
|
|
* remove all L2 entries on the given prefix
|
|
|
|
*/
|
|
|
|
bzero(&prefix0, sizeof(prefix0));
|
|
|
|
prefix0.sin_len = sizeof(prefix0);
|
|
|
|
prefix0.sin_family = AF_INET;
|
|
|
|
prefix0.sin_addr.s_addr = target->ia_subnet;
|
|
|
|
bzero(&mask0, sizeof(mask0));
|
|
|
|
mask0.sin_len = sizeof(mask0);
|
|
|
|
mask0.sin_family = AF_INET;
|
|
|
|
mask0.sin_addr.s_addr = target->ia_subnetmask;
|
2012-07-31 11:31:12 +00:00
|
|
|
lltable_prefix_free(AF_INET, (struct sockaddr *)&prefix0,
|
|
|
|
(struct sockaddr *)&mask0, flags);
|
2009-05-20 21:07:15 +00:00
|
|
|
|
2004-11-12 20:53:51 +00:00
|
|
|
/*
|
|
|
|
* As no-one seem to have this prefix, we can remove the route.
|
|
|
|
*/
|
2011-05-29 02:21:35 +00:00
|
|
|
error = rtinit(&(target->ia_ifa), (int)RTM_DELETE, rtinitflags(target));
|
|
|
|
if (error == 0)
|
|
|
|
target->ia_flags &= ~IFA_ROUTE;
|
|
|
|
else
|
|
|
|
log(LOG_INFO, "in_scrubprefix: err=%d, prefix delete failed\n", error);
|
|
|
|
return (error);
|
2004-11-12 20:53:51 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
#undef rtinitflags
|
1994-05-24 10:09:53 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Return 1 if the address might be a local broadcast address.
|
|
|
|
*/
|
1994-05-25 09:21:21 +00:00
|
|
|
int
|
2007-05-10 15:58:48 +00:00
|
|
|
in_broadcast(struct in_addr in, struct ifnet *ifp)
|
1994-05-24 10:09:53 +00:00
|
|
|
{
|
|
|
|
register struct ifaddr *ifa;
|
|
|
|
u_long t;
|
|
|
|
|
|
|
|
if (in.s_addr == INADDR_BROADCAST ||
|
|
|
|
in.s_addr == INADDR_ANY)
|
2008-10-26 19:17:25 +00:00
|
|
|
return (1);
|
1994-05-24 10:09:53 +00:00
|
|
|
if ((ifp->if_flags & IFF_BROADCAST) == 0)
|
2008-10-26 19:17:25 +00:00
|
|
|
return (0);
|
1994-05-24 10:09:53 +00:00
|
|
|
t = ntohl(in.s_addr);
|
|
|
|
/*
|
|
|
|
* Look through the list of addresses for a match
|
|
|
|
* with a broadcast address.
|
|
|
|
*/
|
|
|
|
#define ia ((struct in_ifaddr *)ifa)
|
2001-03-16 20:00:53 +00:00
|
|
|
TAILQ_FOREACH(ifa, &ifp->if_addrhead, ifa_link)
|
1994-05-24 10:09:53 +00:00
|
|
|
if (ifa->ifa_addr->sa_family == AF_INET &&
|
|
|
|
(in.s_addr == ia->ia_broadaddr.sin_addr.s_addr ||
|
|
|
|
/*
|
2011-10-15 18:41:25 +00:00
|
|
|
* Check for old-style (host 0) broadcast, but
|
|
|
|
* taking into account that RFC 3021 obsoletes it.
|
1994-05-24 10:09:53 +00:00
|
|
|
*/
|
2012-07-31 11:31:12 +00:00
|
|
|
(ia->ia_subnetmask != IN_RFC3021_MASK &&
|
|
|
|
t == ia->ia_subnet)) &&
|
1996-01-08 20:59:06 +00:00
|
|
|
/*
|
|
|
|
* Check for an all one subnetmask. These
|
|
|
|
* only exist when an interface gets a secondary
|
|
|
|
* address.
|
|
|
|
*/
|
2012-07-31 11:31:12 +00:00
|
|
|
ia->ia_subnetmask != (u_long)0xffffffff)
|
2008-10-26 19:17:25 +00:00
|
|
|
return (1);
|
1994-05-24 10:09:53 +00:00
|
|
|
return (0);
|
|
|
|
#undef ia
|
|
|
|
}
|
2007-03-20 00:36:10 +00:00
|
|
|
|
2009-03-09 17:53:05 +00:00
|
|
|
/*
|
|
|
|
* On interface removal, clean up IPv4 data structures hung off of the ifnet.
|
|
|
|
*/
|
|
|
|
void
|
|
|
|
in_ifdetach(struct ifnet *ifp)
|
|
|
|
{
|
|
|
|
|
|
|
|
in_pcbpurgeif0(&V_ripcbinfo, ifp);
|
|
|
|
in_pcbpurgeif0(&V_udbinfo, ifp);
|
|
|
|
in_purgemaddrs(ifp);
|
|
|
|
}
|
|
|
|
|
2006-09-28 10:04:07 +00:00
|
|
|
/*
|
2007-03-20 00:36:10 +00:00
|
|
|
* Delete all IPv4 multicast address records, and associated link-layer
|
|
|
|
* multicast address records, associated with ifp.
|
2009-03-09 17:53:05 +00:00
|
|
|
* XXX It looks like domifdetach runs AFTER the link layer cleanup.
|
2009-03-17 14:41:54 +00:00
|
|
|
* XXX This should not race with ifma_protospec being set during
|
|
|
|
* a new allocation, if it does, we have bigger problems.
|
2006-09-28 10:04:07 +00:00
|
|
|
*/
|
2007-03-20 00:36:10 +00:00
|
|
|
static void
|
|
|
|
in_purgemaddrs(struct ifnet *ifp)
|
2006-09-28 10:04:07 +00:00
|
|
|
{
|
2009-03-09 17:53:05 +00:00
|
|
|
LIST_HEAD(,in_multi) purgeinms;
|
|
|
|
struct in_multi *inm, *tinm;
|
|
|
|
struct ifmultiaddr *ifma;
|
2006-09-28 10:04:07 +00:00
|
|
|
|
2009-03-09 17:53:05 +00:00
|
|
|
LIST_INIT(&purgeinms);
|
2006-09-28 10:04:07 +00:00
|
|
|
IN_MULTI_LOCK();
|
2009-03-09 17:53:05 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Extract list of in_multi associated with the detaching ifp
|
|
|
|
* which the PF_INET layer is about to release.
|
|
|
|
* We need to do this as IF_ADDR_LOCK() may be re-acquired
|
|
|
|
* by code further down.
|
|
|
|
*/
|
2012-01-05 19:00:36 +00:00
|
|
|
IF_ADDR_RLOCK(ifp);
|
2009-03-09 17:53:05 +00:00
|
|
|
TAILQ_FOREACH(ifma, &ifp->if_multiaddrs, ifma_link) {
|
2009-03-17 14:41:54 +00:00
|
|
|
if (ifma->ifma_addr->sa_family != AF_INET ||
|
|
|
|
ifma->ifma_protospec == NULL)
|
2009-03-09 17:53:05 +00:00
|
|
|
continue;
|
2009-03-17 14:41:54 +00:00
|
|
|
#if 0
|
|
|
|
KASSERT(ifma->ifma_protospec != NULL,
|
|
|
|
("%s: ifma_protospec is NULL", __func__));
|
|
|
|
#endif
|
2009-03-09 17:53:05 +00:00
|
|
|
inm = (struct in_multi *)ifma->ifma_protospec;
|
|
|
|
LIST_INSERT_HEAD(&purgeinms, inm, inm_link);
|
2006-09-28 10:04:07 +00:00
|
|
|
}
|
2012-01-05 19:00:36 +00:00
|
|
|
IF_ADDR_RUNLOCK(ifp);
|
2005-09-18 17:36:28 +00:00
|
|
|
|
2009-03-09 17:53:05 +00:00
|
|
|
LIST_FOREACH_SAFE(inm, &purgeinms, inm_link, tinm) {
|
|
|
|
LIST_REMOVE(inm, inm_link);
|
2009-04-24 22:11:53 +00:00
|
|
|
inm_release_locked(inm);
|
2009-03-09 17:53:05 +00:00
|
|
|
}
|
|
|
|
igmp_ifdetach(ifp);
|
2005-09-18 17:36:28 +00:00
|
|
|
|
2009-03-09 17:53:05 +00:00
|
|
|
IN_MULTI_UNLOCK();
|
2005-09-18 17:36:28 +00:00
|
|
|
}
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
|
|
|
|
struct in_llentry {
|
|
|
|
struct llentry base;
|
|
|
|
struct sockaddr_in l3_addr4;
|
|
|
|
};
|
|
|
|
|
2012-02-23 18:21:37 +00:00
|
|
|
/*
|
|
|
|
* Deletes an address from the address table.
|
|
|
|
* This function is called by the timer functions
|
|
|
|
* such as arptimer() and nd6_llinfo_timer(), and
|
|
|
|
* the caller does the locking.
|
|
|
|
*/
|
|
|
|
static void
|
|
|
|
in_lltable_free(struct lltable *llt, struct llentry *lle)
|
|
|
|
{
|
|
|
|
LLE_WUNLOCK(lle);
|
|
|
|
LLE_LOCK_DESTROY(lle);
|
|
|
|
free(lle, M_LLTABLE);
|
|
|
|
}
|
|
|
|
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
static struct llentry *
|
|
|
|
in_lltable_new(const struct sockaddr *l3addr, u_int flags)
|
|
|
|
{
|
|
|
|
struct in_llentry *lle;
|
|
|
|
|
2012-04-10 06:52:39 +00:00
|
|
|
lle = malloc(sizeof(struct in_llentry), M_LLTABLE, M_NOWAIT | M_ZERO);
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
if (lle == NULL) /* NB: caller generates msg */
|
|
|
|
return NULL;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* For IPv4 this will trigger "arpresolve" to generate
|
|
|
|
* an ARP request.
|
|
|
|
*/
|
2010-11-30 15:57:00 +00:00
|
|
|
lle->base.la_expire = time_uptime; /* mark expired */
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
lle->l3_addr4 = *(const struct sockaddr_in *)l3addr;
|
|
|
|
lle->base.lle_refcnt = 1;
|
2012-02-23 18:21:37 +00:00
|
|
|
lle->base.lle_free = in_lltable_free;
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
LLE_LOCK_INIT(&lle->base);
|
2012-08-02 13:57:49 +00:00
|
|
|
callout_init_rw(&lle->base.la_timer, &lle->base.lle_lock,
|
|
|
|
CALLOUT_RETURNUNLOCKED);
|
|
|
|
|
|
|
|
return (&lle->base);
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
}
|
|
|
|
|
2009-05-20 21:07:15 +00:00
|
|
|
#define IN_ARE_MASKED_ADDR_EQUAL(d, a, m) ( \
|
|
|
|
(((ntohl((d)->sin_addr.s_addr) ^ (a)->sin_addr.s_addr) & (m)->sin_addr.s_addr)) == 0 )
|
|
|
|
|
|
|
|
static void
|
2012-07-31 11:31:12 +00:00
|
|
|
in_lltable_prefix_free(struct lltable *llt, const struct sockaddr *prefix,
|
|
|
|
const struct sockaddr *mask, u_int flags)
|
2009-05-20 21:07:15 +00:00
|
|
|
{
|
|
|
|
const struct sockaddr_in *pfx = (const struct sockaddr_in *)prefix;
|
|
|
|
const struct sockaddr_in *msk = (const struct sockaddr_in *)mask;
|
|
|
|
struct llentry *lle, *next;
|
2012-07-31 11:31:12 +00:00
|
|
|
int i;
|
2010-11-12 22:03:02 +00:00
|
|
|
size_t pkts_dropped;
|
2009-05-20 21:07:15 +00:00
|
|
|
|
2012-08-02 13:57:49 +00:00
|
|
|
IF_AFDATA_WLOCK(llt->llt_ifp);
|
2012-07-31 11:31:12 +00:00
|
|
|
for (i = 0; i < LLTBL_HASHTBL_SIZE; i++) {
|
2009-05-20 21:07:15 +00:00
|
|
|
LIST_FOREACH_SAFE(lle, &llt->lle_head[i], lle_next, next) {
|
2012-07-31 11:31:12 +00:00
|
|
|
/*
|
2011-05-20 19:12:20 +00:00
|
|
|
* (flags & LLE_STATIC) means deleting all entries
|
2012-07-31 11:31:12 +00:00
|
|
|
* including static ARP entries.
|
2011-05-20 19:12:20 +00:00
|
|
|
*/
|
2012-07-31 11:31:12 +00:00
|
|
|
if (IN_ARE_MASKED_ADDR_EQUAL(satosin(L3_ADDR(lle)),
|
|
|
|
pfx, msk) && ((flags & LLE_STATIC) ||
|
|
|
|
!(lle->la_flags & LLE_STATIC))) {
|
2009-05-20 21:07:15 +00:00
|
|
|
LLE_WLOCK(lle);
|
2012-08-02 13:57:49 +00:00
|
|
|
if (callout_stop(&lle->la_timer))
|
2010-04-11 16:04:08 +00:00
|
|
|
LLE_REMREF(lle);
|
2010-11-12 22:03:02 +00:00
|
|
|
pkts_dropped = llentry_free(lle);
|
|
|
|
ARPSTAT_ADD(dropped, pkts_dropped);
|
2009-05-20 21:07:15 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2012-08-02 13:57:49 +00:00
|
|
|
IF_AFDATA_WUNLOCK(llt->llt_ifp);
|
2009-05-20 21:07:15 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
static int
|
2009-12-30 21:35:34 +00:00
|
|
|
in_lltable_rtcheck(struct ifnet *ifp, u_int flags, const struct sockaddr *l3addr)
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
{
|
|
|
|
struct rtentry *rt;
|
|
|
|
|
|
|
|
KASSERT(l3addr->sa_family == AF_INET,
|
|
|
|
("sin_family %d", l3addr->sa_family));
|
|
|
|
|
|
|
|
/* XXX rtalloc1 should take a const param */
|
|
|
|
rt = rtalloc1(__DECONST(struct sockaddr *, l3addr), 0, 0);
|
2011-07-08 09:38:33 +00:00
|
|
|
|
2011-10-03 19:06:55 +00:00
|
|
|
if (rt == NULL)
|
|
|
|
return (EINVAL);
|
|
|
|
|
2011-07-08 09:38:33 +00:00
|
|
|
/*
|
|
|
|
* If the gateway for an existing host route matches the target L3
|
2011-10-03 19:06:55 +00:00
|
|
|
* address, which is a special route inserted by some implementation
|
|
|
|
* such as MANET, and the interface is of the correct type, then
|
|
|
|
* allow for ARP to proceed.
|
2011-07-08 09:38:33 +00:00
|
|
|
*/
|
2011-10-03 19:51:18 +00:00
|
|
|
if (rt->rt_flags & RTF_GATEWAY) {
|
2011-10-10 17:41:11 +00:00
|
|
|
if (!(rt->rt_flags & RTF_HOST) || !rt->rt_ifp ||
|
2012-07-31 11:31:12 +00:00
|
|
|
rt->rt_ifp->if_type != IFT_ETHER ||
|
|
|
|
(rt->rt_ifp->if_flags & (IFF_NOARP | IFF_STATICARP)) != 0 ||
|
|
|
|
memcmp(rt->rt_gateway->sa_data, l3addr->sa_data,
|
|
|
|
sizeof(in_addr_t)) != 0) {
|
2011-10-10 17:41:11 +00:00
|
|
|
RTFREE_LOCKED(rt);
|
|
|
|
return (EINVAL);
|
|
|
|
}
|
2011-10-03 19:51:18 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2012-07-31 11:31:12 +00:00
|
|
|
* Make sure that at least the destination address is covered
|
|
|
|
* by the route. This is for handling the case where 2 or more
|
2011-10-03 19:51:18 +00:00
|
|
|
* interfaces have the same prefix. An incoming packet arrives
|
|
|
|
* on one interface and the corresponding outgoing packet leaves
|
|
|
|
* another interface.
|
|
|
|
*/
|
2011-10-25 04:06:29 +00:00
|
|
|
if (!(rt->rt_flags & RTF_HOST) && rt->rt_ifp != ifp) {
|
2011-10-10 17:41:11 +00:00
|
|
|
const char *sa, *mask, *addr, *lim;
|
2011-10-03 19:51:18 +00:00
|
|
|
int len;
|
|
|
|
|
2011-10-10 17:41:11 +00:00
|
|
|
mask = (const char *)rt_mask(rt);
|
2011-10-25 04:06:29 +00:00
|
|
|
/*
|
|
|
|
* Just being extra cautious to avoid some custom
|
|
|
|
* code getting into trouble.
|
|
|
|
*/
|
|
|
|
if (mask == NULL) {
|
|
|
|
RTFREE_LOCKED(rt);
|
|
|
|
return (EINVAL);
|
|
|
|
}
|
|
|
|
|
|
|
|
sa = (const char *)rt_key(rt);
|
2011-10-10 17:41:11 +00:00
|
|
|
addr = (const char *)l3addr;
|
|
|
|
len = ((const struct sockaddr_in *)l3addr)->sin_len;
|
2011-10-03 19:51:18 +00:00
|
|
|
lim = addr + len;
|
|
|
|
|
|
|
|
for ( ; addr < lim; sa++, mask++, addr++) {
|
|
|
|
if ((*sa ^ *addr) & *mask) {
|
2009-10-23 18:27:34 +00:00
|
|
|
#ifdef DIAGNOSTIC
|
2011-10-03 19:51:18 +00:00
|
|
|
log(LOG_INFO, "IPv4 address: \"%s\" is not on the network\n",
|
|
|
|
inet_ntoa(((const struct sockaddr_in *)l3addr)->sin_addr));
|
2009-10-02 01:45:11 +00:00
|
|
|
#endif
|
2011-10-10 17:41:11 +00:00
|
|
|
RTFREE_LOCKED(rt);
|
|
|
|
return (EINVAL);
|
2011-10-03 19:51:18 +00:00
|
|
|
}
|
|
|
|
}
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
}
|
2011-10-03 19:51:18 +00:00
|
|
|
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
RTFREE_LOCKED(rt);
|
2011-10-10 17:41:11 +00:00
|
|
|
return (0);
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Return NULL if not found or marked for deletion.
|
|
|
|
* If found return lle read locked.
|
|
|
|
*/
|
|
|
|
static struct llentry *
|
|
|
|
in_lltable_lookup(struct lltable *llt, u_int flags, const struct sockaddr *l3addr)
|
|
|
|
{
|
|
|
|
const struct sockaddr_in *sin = (const struct sockaddr_in *)l3addr;
|
|
|
|
struct ifnet *ifp = llt->llt_ifp;
|
|
|
|
struct llentry *lle;
|
|
|
|
struct llentries *lleh;
|
|
|
|
u_int hashkey;
|
|
|
|
|
|
|
|
IF_AFDATA_LOCK_ASSERT(ifp);
|
|
|
|
KASSERT(l3addr->sa_family == AF_INET,
|
|
|
|
("sin_family %d", l3addr->sa_family));
|
|
|
|
|
|
|
|
hashkey = sin->sin_addr.s_addr;
|
|
|
|
lleh = &llt->lle_head[LLATBL_HASH(hashkey, LLTBL_HASHMASK)];
|
|
|
|
LIST_FOREACH(lle, lleh, lle_next) {
|
2012-07-31 11:31:12 +00:00
|
|
|
struct sockaddr_in *sa2 = satosin(L3_ADDR(lle));
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
if (lle->la_flags & LLE_DELETED)
|
|
|
|
continue;
|
2009-01-03 00:27:28 +00:00
|
|
|
if (sa2->sin_addr.s_addr == sin->sin_addr.s_addr)
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
if (lle == NULL) {
|
2009-10-23 18:27:34 +00:00
|
|
|
#ifdef DIAGNOSTIC
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
if (flags & LLE_DELETE)
|
2012-08-01 09:00:26 +00:00
|
|
|
log(LOG_INFO, "interface address is missing from cache = %p in delete\n", lle);
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
#endif
|
|
|
|
if (!(flags & LLE_CREATE))
|
|
|
|
return (NULL);
|
|
|
|
/*
|
|
|
|
* A route that covers the given address must have
|
|
|
|
* been installed 1st because we are doing a resolution,
|
|
|
|
* verify this.
|
|
|
|
*/
|
|
|
|
if (!(flags & LLE_IFADDR) &&
|
2009-12-30 21:35:34 +00:00
|
|
|
in_lltable_rtcheck(ifp, flags, l3addr) != 0)
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
goto done;
|
|
|
|
|
|
|
|
lle = in_lltable_new(l3addr, flags);
|
|
|
|
if (lle == NULL) {
|
|
|
|
log(LOG_INFO, "lla_lookup: new lle malloc failed\n");
|
|
|
|
goto done;
|
|
|
|
}
|
|
|
|
lle->la_flags = flags & ~LLE_CREATE;
|
|
|
|
if ((flags & (LLE_CREATE | LLE_IFADDR)) == (LLE_CREATE | LLE_IFADDR)) {
|
|
|
|
bcopy(IF_LLADDR(ifp), &lle->ll_addr, ifp->if_addrlen);
|
|
|
|
lle->la_flags |= (LLE_VALID | LLE_STATIC);
|
|
|
|
}
|
|
|
|
|
|
|
|
lle->lle_tbl = llt;
|
|
|
|
lle->lle_head = lleh;
|
2012-08-02 13:57:49 +00:00
|
|
|
lle->la_flags |= LLE_LINKED;
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
LIST_INSERT_HEAD(lleh, lle, lle_next);
|
|
|
|
} else if (flags & LLE_DELETE) {
|
|
|
|
if (!(lle->la_flags & LLE_IFADDR) || (flags & LLE_IFADDR)) {
|
|
|
|
LLE_WLOCK(lle);
|
2012-08-02 13:57:49 +00:00
|
|
|
lle->la_flags |= LLE_DELETED;
|
2012-06-19 07:34:13 +00:00
|
|
|
EVENTHANDLER_INVOKE(lle_event, lle, LLENTRY_DELETED);
|
2009-10-23 18:27:34 +00:00
|
|
|
#ifdef DIAGNOSTIC
|
2013-04-21 21:28:38 +00:00
|
|
|
log(LOG_INFO, "ifaddr cache = %p is deleted\n", lle);
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
#endif
|
2013-04-21 21:28:38 +00:00
|
|
|
if ((lle->la_flags &
|
|
|
|
(LLE_STATIC | LLE_IFADDR)) == LLE_STATIC)
|
|
|
|
llentry_free(lle);
|
|
|
|
else
|
|
|
|
LLE_WUNLOCK(lle);
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
}
|
|
|
|
lle = (void *)-1;
|
2012-08-01 09:00:26 +00:00
|
|
|
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
}
|
2008-12-28 21:18:01 +00:00
|
|
|
if (LLE_IS_VALID(lle)) {
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
if (flags & LLE_EXCLUSIVE)
|
|
|
|
LLE_WLOCK(lle);
|
|
|
|
else
|
|
|
|
LLE_RLOCK(lle);
|
|
|
|
}
|
|
|
|
done:
|
|
|
|
return (lle);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
in_lltable_dump(struct lltable *llt, struct sysctl_req *wr)
|
|
|
|
{
|
|
|
|
#define SIN(lle) ((struct sockaddr_in *) L3_ADDR(lle))
|
|
|
|
struct ifnet *ifp = llt->llt_ifp;
|
|
|
|
struct llentry *lle;
|
|
|
|
/* XXX stack use */
|
|
|
|
struct {
|
|
|
|
struct rt_msghdr rtm;
|
2013-01-31 08:55:21 +00:00
|
|
|
struct sockaddr_in sin;
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
struct sockaddr_dl sdl;
|
|
|
|
} arpc;
|
|
|
|
int error, i;
|
|
|
|
|
2009-08-25 09:52:38 +00:00
|
|
|
LLTABLE_LOCK_ASSERT();
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
|
|
|
|
error = 0;
|
|
|
|
for (i = 0; i < LLTBL_HASHTBL_SIZE; i++) {
|
|
|
|
LIST_FOREACH(lle, &llt->lle_head[i], lle_next) {
|
|
|
|
struct sockaddr_dl *sdl;
|
2012-08-01 09:00:26 +00:00
|
|
|
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
/* skip deleted entries */
|
2009-10-15 06:12:04 +00:00
|
|
|
if ((lle->la_flags & LLE_DELETED) == LLE_DELETED)
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
continue;
|
2009-01-09 21:57:49 +00:00
|
|
|
/* Skip if jailed and not a valid IP of the prison. */
|
2009-02-05 14:06:09 +00:00
|
|
|
if (prison_if(wr->td->td_ucred, L3_ADDR(lle)) != 0)
|
2009-01-09 21:57:49 +00:00
|
|
|
continue;
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
/*
|
|
|
|
* produce a msg made of:
|
|
|
|
* struct rt_msghdr;
|
2013-01-31 08:55:21 +00:00
|
|
|
* struct sockaddr_in; (IPv4)
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
* struct sockaddr_dl;
|
|
|
|
*/
|
|
|
|
bzero(&arpc, sizeof(arpc));
|
|
|
|
arpc.rtm.rtm_msglen = sizeof(arpc);
|
2009-01-09 10:58:59 +00:00
|
|
|
arpc.rtm.rtm_version = RTM_VERSION;
|
|
|
|
arpc.rtm.rtm_type = RTM_GET;
|
|
|
|
arpc.rtm.rtm_flags = RTF_UP;
|
|
|
|
arpc.rtm.rtm_addrs = RTA_DST | RTA_GATEWAY;
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
arpc.sin.sin_family = AF_INET;
|
|
|
|
arpc.sin.sin_len = sizeof(arpc.sin);
|
|
|
|
arpc.sin.sin_addr.s_addr = SIN(lle)->sin_addr.s_addr;
|
|
|
|
|
|
|
|
/* publish */
|
2013-01-31 08:55:21 +00:00
|
|
|
if (lle->la_flags & LLE_PUB)
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
arpc.rtm.rtm_flags |= RTF_ANNOUNCE;
|
|
|
|
|
|
|
|
sdl = &arpc.sdl;
|
|
|
|
sdl->sdl_family = AF_LINK;
|
|
|
|
sdl->sdl_len = sizeof(*sdl);
|
|
|
|
sdl->sdl_index = ifp->if_index;
|
|
|
|
sdl->sdl_type = ifp->if_type;
|
2009-10-15 06:12:04 +00:00
|
|
|
if ((lle->la_flags & LLE_VALID) == LLE_VALID) {
|
|
|
|
sdl->sdl_alen = ifp->if_addrlen;
|
|
|
|
bcopy(&lle->ll_addr, LLADDR(sdl), ifp->if_addrlen);
|
|
|
|
} else {
|
|
|
|
sdl->sdl_alen = 0;
|
|
|
|
bzero(LLADDR(sdl), ifp->if_addrlen);
|
|
|
|
}
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
|
|
|
|
arpc.rtm.rtm_rmx.rmx_expire =
|
|
|
|
lle->la_flags & LLE_STATIC ? 0 : lle->la_expire;
|
2008-12-26 19:45:24 +00:00
|
|
|
arpc.rtm.rtm_flags |= (RTF_HOST | RTF_LLDATA);
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
if (lle->la_flags & LLE_STATIC)
|
|
|
|
arpc.rtm.rtm_flags |= RTF_STATIC;
|
|
|
|
arpc.rtm.rtm_index = ifp->if_index;
|
|
|
|
error = SYSCTL_OUT(wr, &arpc, sizeof(arpc));
|
|
|
|
if (error)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return error;
|
|
|
|
#undef SIN
|
|
|
|
}
|
|
|
|
|
|
|
|
void *
|
|
|
|
in_domifattach(struct ifnet *ifp)
|
2009-03-09 17:53:05 +00:00
|
|
|
{
|
|
|
|
struct in_ifinfo *ii;
|
|
|
|
struct lltable *llt;
|
|
|
|
|
|
|
|
ii = malloc(sizeof(struct in_ifinfo), M_IFADDR, M_WAITOK|M_ZERO);
|
|
|
|
|
|
|
|
llt = lltable_init(ifp, AF_INET);
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
if (llt != NULL) {
|
2009-05-20 21:07:15 +00:00
|
|
|
llt->llt_prefix_free = in_lltable_prefix_free;
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
llt->llt_lookup = in_lltable_lookup;
|
|
|
|
llt->llt_dump = in_lltable_dump;
|
|
|
|
}
|
2009-03-09 17:53:05 +00:00
|
|
|
ii->ii_llt = llt;
|
|
|
|
|
|
|
|
ii->ii_igmp = igmp_domifattach(ifp);
|
|
|
|
|
|
|
|
return ii;
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
void
|
2009-03-09 17:53:05 +00:00
|
|
|
in_domifdetach(struct ifnet *ifp, void *aux)
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
{
|
2009-03-09 17:53:05 +00:00
|
|
|
struct in_ifinfo *ii = (struct in_ifinfo *)aux;
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
|
2009-03-09 17:53:05 +00:00
|
|
|
igmp_domifdetach(ifp);
|
|
|
|
lltable_free(ii->ii_llt);
|
|
|
|
free(ii, M_IFADDR);
|
This main goals of this project are:
1. separating L2 tables (ARP, NDP) from the L3 routing tables
2. removing as much locking dependencies among these layers as
possible to allow for some parallelism in the search operations
3. simplify the logic in the routing code,
The most notable end result is the obsolescent of the route
cloning (RTF_CLONING) concept, which translated into code reduction
in both IPv4 ARP and IPv6 NDP related modules, and size reduction in
struct rtentry{}. The change in design obsoletes the semantics of
RTF_CLONING, RTF_WASCLONE and RTF_LLINFO routing flags. The userland
applications such as "arp" and "ndp" have been modified to reflect
those changes. The output from "netstat -r" shows only the routing
entries.
Quite a few developers have contributed to this project in the
past: Glebius Smirnoff, Luigi Rizzo, Alessandro Cerri, and
Andre Oppermann. And most recently:
- Kip Macy revised the locking code completely, thus completing
the last piece of the puzzle, Kip has also been conducting
active functional testing
- Sam Leffler has helped me improving/refactoring the code, and
provided valuable reviews
- Julian Elischer setup the perforce tree for me and has helped
me maintaining that branch before the svn conversion
2008-12-15 06:10:57 +00:00
|
|
|
}
|