freebsd-dev/sys/net/if_loop.c

464 lines
11 KiB
C
Raw Normal View History

/*-
1994-05-24 10:09:53 +00:00
* Copyright (c) 1982, 1986, 1993
* The Regents of the University of California. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* @(#)if_loop.c 8.2 (Berkeley) 1/9/95
1999-08-28 01:08:13 +00:00
* $FreeBSD$
1994-05-24 10:09:53 +00:00
*/
/*
* Loopback interface driver for protocol testing and timing.
*/
#include "opt_inet.h"
#include "opt_inet6.h"
1994-05-24 10:09:53 +00:00
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/mbuf.h>
#include <sys/module.h>
#include <machine/bus.h>
#include <sys/rman.h>
1994-05-24 10:09:53 +00:00
#include <sys/socket.h>
#include <sys/sockio.h>
#include <sys/sysctl.h>
1994-05-24 10:09:53 +00:00
#include <net/if.h>
#include <net/if_var.h>
#include <net/if_clone.h>
1994-05-24 10:09:53 +00:00
#include <net/if_types.h>
#include <net/netisr.h>
#include <net/route.h>
#include <net/bpf.h>
#include <net/vnet.h>
1994-05-24 10:09:53 +00:00
#ifdef INET
#include <netinet/in.h>
#include <netinet/in_var.h>
#endif
#ifdef INET6
#ifndef INET
#include <netinet/in.h>
#endif
#include <netinet6/in6_var.h>
#include <netinet/ip6.h>
#endif
#include <security/mac/mac_framework.h>
1994-05-29 07:43:54 +00:00
#ifdef TINY_LOMTU
1994-05-24 10:09:53 +00:00
#define LOMTU (1024+512)
#elif defined(LARGE_LOMTU)
#define LOMTU 131072
1994-05-29 07:43:54 +00:00
#else
#define LOMTU 16384
1994-05-29 07:43:54 +00:00
#endif
1994-05-24 10:09:53 +00:00
#define LO_CSUM_FEATURES (CSUM_IP | CSUM_TCP | CSUM_UDP | CSUM_SCTP)
#define LO_CSUM_FEATURES6 (CSUM_TCP_IPV6 | CSUM_UDP_IPV6 | CSUM_SCTP_IPV6)
#define LO_CSUM_SET (CSUM_DATA_VALID | CSUM_DATA_VALID_IPV6 | \
CSUM_PSEUDO_HDR | \
CSUM_IP_CHECKED | CSUM_IP_VALID | \
CSUM_SCTP_VALID)
int loioctl(struct ifnet *, u_long, caddr_t);
static void lortrequest(int, struct rtentry *, struct rt_addrinfo *);
int looutput(struct ifnet *ifp, struct mbuf *m,
const struct sockaddr *dst, struct route *ro);
static int lo_clone_create(struct if_clone *, int, caddr_t);
static void lo_clone_destroy(struct ifnet *);
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
VNET_DEFINE(struct ifnet *, loif); /* Used externally */
#ifdef VIMAGE
static VNET_DEFINE(struct if_clone *, lo_cloner);
#define V_lo_cloner VNET(lo_cloner)
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
#endif
Introduce vnet module registration / initialization framework with dependency tracking and ordering enforcement. With this change, per-vnet initialization functions introduced with r190787 are no longer directly called from traditional initialization functions (which cc in most cases inlined to pre-r190787 code), but are instead registered via the vnet framework first, and are invoked only after all prerequisite modules have been initialized. In the long run, this framework should allow us to both initialize and dismantle multiple vnet instances in a correct order. The problem this change aims to solve is how to replay the initialization sequence of various network stack components, which have been traditionally triggered via different mechanisms (SYSINIT, protosw). Note that this initialization sequence was and still can be subtly different depending on whether certain pieces of code have been statically compiled into the kernel, loaded as modules by boot loader, or kldloaded at run time. The approach is simple - we record the initialization sequence established by the traditional mechanisms whenever vnet_mod_register() is called for a particular vnet module. The vnet_mod_register_multi() variant allows a single initializer function to be registered multiple times but with different arguments - currently this is only used in kern/uipc_domain.c by net_add_domain() with different struct domain * as arguments, which allows for protosw-registered initialization routines to be invoked in a correct order by the new vnet initialization framework. For the purpose of identifying vnet modules, each vnet module has to have a unique ID, which is statically assigned in sys/vimage.h. Dynamic assignment of vnet module IDs is not supported yet. A vnet module may specify a single prerequisite module at registration time by filling in the vmi_dependson field of its vnet_modinfo struct with the ID of the module it depends on. Unless specified otherwise, all vnet modules depend on VNET_MOD_NET (container for ifnet list head, rt_tables etc.), which thus has to and will always be initialized first. The framework will panic if it detects any unresolved dependencies before completing system initialization. Detection of unresolved dependencies for vnet modules registered after boot (kldloaded modules) is not provided. Note that the fact that each module can specify only a single prerequisite may become problematic in the long run. In particular, INET6 depends on INET being already instantiated, due to TCP / UDP structures residing in INET container. IPSEC also depends on INET, which will in turn additionally complicate making INET6-only kernel configs a reality. The entire registration framework can be compiled out by turning on the VIMAGE_GLOBALS kernel config option. Reviewed by: bz Approved by: julian (mentor)
2009-04-11 05:58:58 +00:00
static struct if_clone *lo_cloner;
static const char loname[] = "lo";
static void
lo_clone_destroy(struct ifnet *ifp)
{
#ifndef VIMAGE
/* XXX: destroying lo0 will lead to panics. */
KASSERT(V_loif != ifp, ("%s: destroying lo0", __func__));
#endif
bpfdetach(ifp);
if_detach(ifp);
if_free(ifp);
}
static int
lo_clone_create(struct if_clone *ifc, int unit, caddr_t params)
{
struct ifnet *ifp;
ifp = if_alloc(IFT_LOOP);
if (ifp == NULL)
return (ENOSPC);
if_initname(ifp, loname, unit);
ifp->if_mtu = LOMTU;
ifp->if_flags = IFF_LOOPBACK | IFF_MULTICAST;
ifp->if_ioctl = loioctl;
ifp->if_output = looutput;
ifp->if_snd.ifq_maxlen = ifqmaxlen;
ifp->if_capabilities = ifp->if_capenable =
IFCAP_HWCSUM | IFCAP_HWCSUM_IPV6;
ifp->if_hwassist = LO_CSUM_FEATURES | LO_CSUM_FEATURES6;
if_attach(ifp);
bpfattach(ifp, DLT_NULL, sizeof(u_int32_t));
if (V_loif == NULL)
V_loif = ifp;
return (0);
}
static void
vnet_loif_init(const void *unused __unused)
{
#ifdef VIMAGE
lo_cloner = if_clone_simple(loname, lo_clone_create, lo_clone_destroy,
1);
V_lo_cloner = lo_cloner;
#else
lo_cloner = if_clone_simple(loname, lo_clone_create, lo_clone_destroy,
1);
#endif
}
VNET_SYSINIT(vnet_loif_init, SI_SUB_PROTO_IFATTACHDOMAIN, SI_ORDER_ANY,
vnet_loif_init, NULL);
#ifdef VIMAGE
static void
vnet_loif_uninit(const void *unused __unused)
{
if_clone_detach(V_lo_cloner);
V_loif = NULL;
}
VNET_SYSUNINIT(vnet_loif_uninit, SI_SUB_PROTO_IFATTACHDOMAIN, SI_ORDER_ANY,
vnet_loif_uninit, NULL);
#endif
static int
loop_modevent(module_t mod, int type, void *data)
{
switch (type) {
case MOD_LOAD:
break;
case MOD_UNLOAD:
printf("loop module unload - not possible for this module type\n");
return (EINVAL);
default:
return (EOPNOTSUPP);
}
return (0);
}
static moduledata_t loop_mod = {
"if_lo",
loop_modevent,
0
};
DECLARE_MODULE(if_lo, loop_mod, SI_SUB_PROTO_IFATTACHDOMAIN, SI_ORDER_ANY);
1994-05-24 10:09:53 +00:00
int
looutput(struct ifnet *ifp, struct mbuf *m, const struct sockaddr *dst,
struct route *ro)
{
u_int32_t af;
struct rtentry *rt = NULL;
#ifdef MAC
int error;
#endif
M_ASSERTPKTHDR(m); /* check if we have the packet header */
if (ro != NULL)
rt = ro->ro_rt;
#ifdef MAC
error = mac_ifnet_check_transmit(ifp, m);
if (error) {
m_freem(m);
return (error);
}
#endif
if (rt && rt->rt_flags & (RTF_REJECT|RTF_BLACKHOLE)) {
m_freem(m);
return (rt->rt_flags & RTF_BLACKHOLE ? 0 :
rt->rt_flags & RTF_HOST ? EHOSTUNREACH : ENETUNREACH);
}
ifp->if_opackets++;
ifp->if_obytes += m->m_pkthdr.len;
/* BPF writes need to be handled specially. */
if (dst->sa_family == AF_UNSPEC)
bcopy(dst->sa_data, &af, sizeof(af));
else
af = dst->sa_family;
#if 1 /* XXX */
switch (af) {
case AF_INET:
if (ifp->if_capenable & IFCAP_RXCSUM) {
m->m_pkthdr.csum_data = 0xffff;
m->m_pkthdr.csum_flags = LO_CSUM_SET;
}
m->m_pkthdr.csum_flags &= ~LO_CSUM_FEATURES;
break;
case AF_INET6:
#if 0
/*
* XXX-BZ for now always claim the checksum is good despite
* any interface flags. This is a workaround for 9.1-R and
* a proper solution ought to be sought later.
*/
if (ifp->if_capenable & IFCAP_RXCSUM_IPV6) {
m->m_pkthdr.csum_data = 0xffff;
m->m_pkthdr.csum_flags = LO_CSUM_SET;
}
#else
m->m_pkthdr.csum_data = 0xffff;
m->m_pkthdr.csum_flags = LO_CSUM_SET;
#endif
m->m_pkthdr.csum_flags &= ~LO_CSUM_FEATURES6;
break;
default:
printf("looutput: af=%d unexpected\n", af);
m_freem(m);
return (EAFNOSUPPORT);
}
#endif
return (if_simloop(ifp, m, af, 0));
}
/*
* if_simloop()
*
* This function is to support software emulation of hardware loopback,
* i.e., for interfaces with the IFF_SIMPLEX attribute. Since they can't
* hear their own broadcasts, we create a copy of the packet that we
* would normally receive via a hardware loopback.
*
* This function expects the packet to include the media header of length hlen.
*/
int
if_simloop(struct ifnet *ifp, struct mbuf *m, int af, int hlen)
1994-05-24 10:09:53 +00:00
{
int isr;
1994-05-24 10:09:53 +00:00
M_ASSERTPKTHDR(m);
m_tag_delete_nonpersistent(m);
m->m_pkthdr.rcvif = ifp;
#ifdef MAC
mac_ifnet_create_mbuf(ifp, m);
#endif
/*
* Let BPF see incoming packet in the following manner:
* - Emulated packet loopback for a simplex interface
* (net/if_ethersubr.c)
* -> passes it to ifp's BPF
* - IPv4/v6 multicast packet loopback (netinet(6)/ip(6)_output.c)
* -> not passes it to any BPF
* - Normal packet loopback from myself to myself (net/if_loop.c)
* -> passes to lo0's BPF (even in case of IPv6, where ifp!=lo0)
*/
if (hlen > 0) {
if (bpf_peers_present(ifp->if_bpf)) {
bpf_mtap(ifp->if_bpf, m);
}
} else {
if (bpf_peers_present(V_loif->if_bpf)) {
if ((m->m_flags & M_MCAST) == 0 || V_loif == ifp) {
/* XXX beware sizeof(af) != 4 */
2008-08-01 09:41:45 +00:00
u_int32_t af1 = af;
/*
* We need to prepend the address family.
*/
bpf_mtap2(V_loif->if_bpf, &af1, sizeof(af1), m);
}
}
1994-05-24 10:09:53 +00:00
}
/* Strip away media header */
if (hlen > 0) {
m_adj(m, hlen);
#ifndef __NO_STRICT_ALIGNMENT
/*
* Some archs do not like unaligned data, so
* we move data down in the first mbuf.
*/
if (mtod(m, vm_offset_t) & 3) {
2000-06-02 19:06:33 +00:00
KASSERT(hlen >= 3, ("if_simloop: hlen too small"));
bcopy(m->m_data,
(char *)(mtod(m, vm_offset_t)
- (mtod(m, vm_offset_t) & 3)),
m->m_len);
2004-07-28 06:59:55 +00:00
m->m_data -= (mtod(m,vm_offset_t) & 3);
}
#endif
}
1994-05-24 10:09:53 +00:00
/* Deliver to upper layer protocol */
switch (af) {
1994-05-24 10:09:53 +00:00
#ifdef INET
case AF_INET:
isr = NETISR_IP;
break;
#endif
#ifdef INET6
case AF_INET6:
m->m_flags |= M_LOOP;
isr = NETISR_IPV6;
break;
#endif
1994-05-24 10:09:53 +00:00
default:
printf("if_simloop: can't handle af=%d\n", af);
1994-05-24 10:09:53 +00:00
m_freem(m);
return (EAFNOSUPPORT);
}
ifp->if_ipackets++;
ifp->if_ibytes += m->m_pkthdr.len;
netisr_queue(isr, m); /* mbuf is free'd on failure. */
1994-05-24 10:09:53 +00:00
return (0);
}
/* ARGSUSED */
1995-12-09 20:47:15 +00:00
static void
lortrequest(int cmd, struct rtentry *rt, struct rt_addrinfo *info)
1994-05-24 10:09:53 +00:00
{
RT_LOCK_ASSERT(rt);
rt->rt_mtu = rt->rt_ifp->if_mtu;
1994-05-24 10:09:53 +00:00
}
/*
* Process an ioctl request.
*/
/* ARGSUSED */
int
loioctl(struct ifnet *ifp, u_long cmd, caddr_t data)
1994-05-24 10:09:53 +00:00
{
struct ifaddr *ifa;
struct ifreq *ifr = (struct ifreq *)data;
int error = 0, mask;
1994-05-24 10:09:53 +00:00
switch (cmd) {
case SIOCSIFADDR:
ifp->if_flags |= IFF_UP;
ifp->if_drv_flags |= IFF_DRV_RUNNING;
1994-05-24 10:09:53 +00:00
ifa = (struct ifaddr *)data;
ifa->ifa_rtrequest = lortrequest;
1994-05-24 10:09:53 +00:00
/*
* Everything else is done at a higher level.
*/
break;
case SIOCADDMULTI:
case SIOCDELMULTI:
if (ifr == 0) {
error = EAFNOSUPPORT; /* XXX */
break;
}
switch (ifr->ifr_addr.sa_family) {
#ifdef INET
case AF_INET:
break;
#endif
#ifdef INET6
case AF_INET6:
break;
#endif
1994-05-24 10:09:53 +00:00
default:
error = EAFNOSUPPORT;
break;
}
break;
1994-08-08 12:09:04 +00:00
case SIOCSIFMTU:
ifp->if_mtu = ifr->ifr_mtu;
1994-08-08 12:09:04 +00:00
break;
case SIOCSIFFLAGS:
break;
case SIOCSIFCAP:
mask = ifp->if_capenable ^ ifr->ifr_reqcap;
if ((mask & IFCAP_RXCSUM) != 0)
ifp->if_capenable ^= IFCAP_RXCSUM;
if ((mask & IFCAP_TXCSUM) != 0)
ifp->if_capenable ^= IFCAP_TXCSUM;
if ((mask & IFCAP_RXCSUM_IPV6) != 0) {
#if 0
ifp->if_capenable ^= IFCAP_RXCSUM_IPV6;
#else
error = EOPNOTSUPP;
break;
#endif
}
if ((mask & IFCAP_TXCSUM_IPV6) != 0) {
#if 0
ifp->if_capenable ^= IFCAP_TXCSUM_IPV6;
#else
error = EOPNOTSUPP;
break;
#endif
}
ifp->if_hwassist = 0;
if (ifp->if_capenable & IFCAP_TXCSUM)
ifp->if_hwassist = LO_CSUM_FEATURES;
#if 0
if (ifp->if_capenable & IFCAP_TXCSUM_IPV6)
ifp->if_hwassist |= LO_CSUM_FEATURES6;
#endif
break;
1994-05-24 10:09:53 +00:00
default:
error = EINVAL;
}
return (error);
}