freebsd-skq/sys/net/if_enc.c
Andrew Turner 5f901c92a8 Use the new VNET_DEFINE_STATIC macro when we are defining static VNET
variables.

Reviewed by:	bz
Sponsored by:	DARPA, AFRL
Differential Revision:	https://reviews.freebsd.org/D16147
2018-07-24 16:35:52 +00:00

446 lines
12 KiB
C

/*-
* SPDX-License-Identifier: BSD-2-Clause-FreeBSD
*
* Copyright (c) 2006 The FreeBSD Project.
* Copyright (c) 2015 Andrey V. Elsukov <ae@FreeBSD.org>
* 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.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR 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 AUTHOR 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.
*
* $FreeBSD$
*/
#include "opt_inet.h"
#include "opt_inet6.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/hhook.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/mbuf.h>
#include <sys/module.h>
#include <machine/bus.h>
#include <sys/rman.h>
#include <sys/socket.h>
#include <sys/sockio.h>
#include <sys/sysctl.h>
#include <net/if.h>
#include <net/if_enc.h>
#include <net/if_var.h>
#include <net/if_clone.h>
#include <net/if_types.h>
#include <net/pfil.h>
#include <net/route.h>
#include <net/netisr.h>
#include <net/bpf.h>
#include <net/vnet.h>
#include <netinet/in.h>
#include <netinet/in_systm.h>
#include <netinet/ip.h>
#include <netinet/ip_var.h>
#include <netinet/in_var.h>
#ifdef INET6
#include <netinet/ip6.h>
#include <netinet6/ip6_var.h>
#endif
#include <netipsec/ipsec.h>
#include <netipsec/xform.h>
#define ENCMTU (1024+512)
/* XXX this define must have the same value as in OpenBSD */
#define M_CONF 0x0400 /* payload was encrypted (ESP-transport) */
#define M_AUTH 0x0800 /* payload was authenticated (AH or ESP auth) */
#define M_AUTH_AH 0x2000 /* header was authenticated (AH) */
struct enchdr {
u_int32_t af;
u_int32_t spi;
u_int32_t flags;
};
struct enc_softc {
struct ifnet *sc_ifp;
};
VNET_DEFINE_STATIC(struct enc_softc *, enc_sc);
#define V_enc_sc VNET(enc_sc)
VNET_DEFINE_STATIC(struct if_clone *, enc_cloner);
#define V_enc_cloner VNET(enc_cloner)
static int enc_ioctl(struct ifnet *, u_long, caddr_t);
static int enc_output(struct ifnet *, struct mbuf *,
const struct sockaddr *, struct route *);
static int enc_clone_create(struct if_clone *, int, caddr_t);
static void enc_clone_destroy(struct ifnet *);
static int enc_add_hhooks(struct enc_softc *);
static void enc_remove_hhooks(struct enc_softc *);
static const char encname[] = "enc";
#define IPSEC_ENC_AFTER_PFIL 0x04
/*
* Before and after are relative to when we are stripping the
* outer IP header.
*
* AFTER_PFIL flag used only for bpf_mask_*. It enables BPF capturing
* after PFIL hook execution. It might be useful when PFIL hook does
* some changes to the packet, e.g. address translation. If PFIL hook
* consumes mbuf, nothing will be captured.
*/
VNET_DEFINE_STATIC(int, filter_mask_in) = IPSEC_ENC_BEFORE;
VNET_DEFINE_STATIC(int, bpf_mask_in) = IPSEC_ENC_BEFORE;
VNET_DEFINE_STATIC(int, filter_mask_out) = IPSEC_ENC_BEFORE;
VNET_DEFINE_STATIC(int, bpf_mask_out) = IPSEC_ENC_BEFORE | IPSEC_ENC_AFTER;
#define V_filter_mask_in VNET(filter_mask_in)
#define V_bpf_mask_in VNET(bpf_mask_in)
#define V_filter_mask_out VNET(filter_mask_out)
#define V_bpf_mask_out VNET(bpf_mask_out)
static SYSCTL_NODE(_net, OID_AUTO, enc, CTLFLAG_RW, 0, "enc sysctl");
static SYSCTL_NODE(_net_enc, OID_AUTO, in, CTLFLAG_RW, 0, "enc input sysctl");
static SYSCTL_NODE(_net_enc, OID_AUTO, out, CTLFLAG_RW, 0, "enc output sysctl");
SYSCTL_INT(_net_enc_in, OID_AUTO, ipsec_filter_mask,
CTLFLAG_RW | CTLFLAG_VNET, &VNET_NAME(filter_mask_in), 0,
"IPsec input firewall filter mask");
SYSCTL_INT(_net_enc_in, OID_AUTO, ipsec_bpf_mask,
CTLFLAG_RW | CTLFLAG_VNET, &VNET_NAME(bpf_mask_in), 0,
"IPsec input bpf mask");
SYSCTL_INT(_net_enc_out, OID_AUTO, ipsec_filter_mask,
CTLFLAG_RW | CTLFLAG_VNET, &VNET_NAME(filter_mask_out), 0,
"IPsec output firewall filter mask");
SYSCTL_INT(_net_enc_out, OID_AUTO, ipsec_bpf_mask,
CTLFLAG_RW | CTLFLAG_VNET, &VNET_NAME(bpf_mask_out), 0,
"IPsec output bpf mask");
static void
enc_clone_destroy(struct ifnet *ifp)
{
struct enc_softc *sc;
sc = ifp->if_softc;
KASSERT(sc == V_enc_sc, ("sc != ifp->if_softc"));
bpfdetach(ifp);
if_detach(ifp);
if_free(ifp);
free(sc, M_DEVBUF);
V_enc_sc = NULL;
}
static int
enc_clone_create(struct if_clone *ifc, int unit, caddr_t params)
{
struct ifnet *ifp;
struct enc_softc *sc;
sc = malloc(sizeof(struct enc_softc), M_DEVBUF,
M_WAITOK | M_ZERO);
ifp = sc->sc_ifp = if_alloc(IFT_ENC);
if (ifp == NULL) {
free(sc, M_DEVBUF);
return (ENOSPC);
}
if (V_enc_sc != NULL) {
if_free(ifp);
free(sc, M_DEVBUF);
return (EEXIST);
}
V_enc_sc = sc;
if_initname(ifp, encname, unit);
ifp->if_mtu = ENCMTU;
ifp->if_ioctl = enc_ioctl;
ifp->if_output = enc_output;
ifp->if_softc = sc;
if_attach(ifp);
bpfattach(ifp, DLT_ENC, sizeof(struct enchdr));
return (0);
}
static int
enc_output(struct ifnet *ifp, struct mbuf *m, const struct sockaddr *dst,
struct route *ro)
{
m_freem(m);
return (0);
}
static int
enc_ioctl(struct ifnet *ifp, u_long cmd, caddr_t data)
{
if (cmd != SIOCSIFFLAGS)
return (EINVAL);
if (ifp->if_flags & IFF_UP)
ifp->if_drv_flags |= IFF_DRV_RUNNING;
else
ifp->if_drv_flags &= ~IFF_DRV_RUNNING;
return (0);
}
static void
enc_bpftap(struct ifnet *ifp, struct mbuf *m, const struct secasvar *sav,
int32_t hhook_type, uint8_t enc, uint8_t af)
{
struct enchdr hdr;
if (hhook_type == HHOOK_TYPE_IPSEC_IN &&
(enc & V_bpf_mask_in) == 0)
return;
else if (hhook_type == HHOOK_TYPE_IPSEC_OUT &&
(enc & V_bpf_mask_out) == 0)
return;
if (bpf_peers_present(ifp->if_bpf) == 0)
return;
hdr.af = af;
hdr.spi = sav->spi;
hdr.flags = 0;
if (sav->alg_enc != SADB_EALG_NONE)
hdr.flags |= M_CONF;
if (sav->alg_auth != SADB_AALG_NONE)
hdr.flags |= M_AUTH;
bpf_mtap2(ifp->if_bpf, &hdr, sizeof(hdr), m);
}
/*
* One helper hook function is used by any hook points.
* + from hhook_type we can determine the packet direction:
* HHOOK_TYPE_IPSEC_IN or HHOOK_TYPE_IPSEC_OUT;
* + from hhook_id we can determine address family: AF_INET or AF_INET6;
* + udata contains pointer to enc_softc;
* + ctx_data contains pointer to struct ipsec_ctx_data.
*/
static int
enc_hhook(int32_t hhook_type, int32_t hhook_id, void *udata, void *ctx_data,
void *hdata, struct osd *hosd)
{
struct ipsec_ctx_data *ctx;
struct enc_softc *sc;
struct ifnet *ifp, *rcvif;
struct pfil_head *ph;
int pdir;
sc = (struct enc_softc *)udata;
ifp = sc->sc_ifp;
if ((ifp->if_flags & IFF_UP) == 0)
return (0);
ctx = (struct ipsec_ctx_data *)ctx_data;
/* XXX: wrong hook point was used by caller? */
if (ctx->af != hhook_id)
return (EPFNOSUPPORT);
enc_bpftap(ifp, *ctx->mp, ctx->sav, hhook_type, ctx->enc, ctx->af);
switch (hhook_type) {
case HHOOK_TYPE_IPSEC_IN:
if (ctx->enc == IPSEC_ENC_BEFORE) {
/* Do accounting only once */
if_inc_counter(ifp, IFCOUNTER_IPACKETS, 1);
if_inc_counter(ifp, IFCOUNTER_IBYTES,
(*ctx->mp)->m_pkthdr.len);
}
if ((ctx->enc & V_filter_mask_in) == 0)
return (0); /* skip pfil processing */
pdir = PFIL_IN;
break;
case HHOOK_TYPE_IPSEC_OUT:
if (ctx->enc == IPSEC_ENC_BEFORE) {
/* Do accounting only once */
if_inc_counter(ifp, IFCOUNTER_OPACKETS, 1);
if_inc_counter(ifp, IFCOUNTER_OBYTES,
(*ctx->mp)->m_pkthdr.len);
}
if ((ctx->enc & V_filter_mask_out) == 0)
return (0); /* skip pfil processing */
pdir = PFIL_OUT;
break;
default:
return (EINVAL);
}
switch (hhook_id) {
#ifdef INET
case AF_INET:
ph = &V_inet_pfil_hook;
break;
#endif
#ifdef INET6
case AF_INET6:
ph = &V_inet6_pfil_hook;
break;
#endif
default:
ph = NULL;
}
if (ph == NULL || !PFIL_HOOKED(ph))
return (0);
/* Make a packet looks like it was received on enc(4) */
rcvif = (*ctx->mp)->m_pkthdr.rcvif;
(*ctx->mp)->m_pkthdr.rcvif = ifp;
if (pfil_run_hooks(ph, ctx->mp, ifp, pdir, 0, ctx->inp) != 0 ||
*ctx->mp == NULL) {
*ctx->mp = NULL; /* consumed by filter */
return (EACCES);
}
(*ctx->mp)->m_pkthdr.rcvif = rcvif;
enc_bpftap(ifp, *ctx->mp, ctx->sav, hhook_type,
IPSEC_ENC_AFTER_PFIL, ctx->af);
return (0);
}
static int
enc_add_hhooks(struct enc_softc *sc)
{
struct hookinfo hki;
int error;
error = EPFNOSUPPORT;
hki.hook_func = enc_hhook;
hki.hook_helper = NULL;
hki.hook_udata = sc;
#ifdef INET
hki.hook_id = AF_INET;
hki.hook_type = HHOOK_TYPE_IPSEC_IN;
error = hhook_add_hook(V_ipsec_hhh_in[HHOOK_IPSEC_INET],
&hki, HHOOK_WAITOK);
if (error != 0)
return (error);
hki.hook_type = HHOOK_TYPE_IPSEC_OUT;
error = hhook_add_hook(V_ipsec_hhh_out[HHOOK_IPSEC_INET],
&hki, HHOOK_WAITOK);
if (error != 0)
return (error);
#endif
#ifdef INET6
hki.hook_id = AF_INET6;
hki.hook_type = HHOOK_TYPE_IPSEC_IN;
error = hhook_add_hook(V_ipsec_hhh_in[HHOOK_IPSEC_INET6],
&hki, HHOOK_WAITOK);
if (error != 0)
return (error);
hki.hook_type = HHOOK_TYPE_IPSEC_OUT;
error = hhook_add_hook(V_ipsec_hhh_out[HHOOK_IPSEC_INET6],
&hki, HHOOK_WAITOK);
if (error != 0)
return (error);
#endif
return (error);
}
static void
enc_remove_hhooks(struct enc_softc *sc)
{
struct hookinfo hki;
hki.hook_func = enc_hhook;
hki.hook_helper = NULL;
hki.hook_udata = sc;
#ifdef INET
hki.hook_id = AF_INET;
hki.hook_type = HHOOK_TYPE_IPSEC_IN;
hhook_remove_hook(V_ipsec_hhh_in[HHOOK_IPSEC_INET], &hki);
hki.hook_type = HHOOK_TYPE_IPSEC_OUT;
hhook_remove_hook(V_ipsec_hhh_out[HHOOK_IPSEC_INET], &hki);
#endif
#ifdef INET6
hki.hook_id = AF_INET6;
hki.hook_type = HHOOK_TYPE_IPSEC_IN;
hhook_remove_hook(V_ipsec_hhh_in[HHOOK_IPSEC_INET6], &hki);
hki.hook_type = HHOOK_TYPE_IPSEC_OUT;
hhook_remove_hook(V_ipsec_hhh_out[HHOOK_IPSEC_INET6], &hki);
#endif
}
static void
vnet_enc_init(const void *unused __unused)
{
V_enc_sc = NULL;
V_enc_cloner = if_clone_simple(encname, enc_clone_create,
enc_clone_destroy, 1);
}
VNET_SYSINIT(vnet_enc_init, SI_SUB_PSEUDO, SI_ORDER_ANY,
vnet_enc_init, NULL);
static void
vnet_enc_init_proto(void *unused __unused)
{
KASSERT(V_enc_sc != NULL, ("%s: V_enc_sc is %p\n", __func__, V_enc_sc));
if (enc_add_hhooks(V_enc_sc) != 0)
enc_clone_destroy(V_enc_sc->sc_ifp);
}
VNET_SYSINIT(vnet_enc_init_proto, SI_SUB_PROTO_IFATTACHDOMAIN, SI_ORDER_ANY,
vnet_enc_init_proto, NULL);
static void
vnet_enc_uninit(const void *unused __unused)
{
KASSERT(V_enc_sc != NULL, ("%s: V_enc_sc is %p\n", __func__, V_enc_sc));
if_clone_detach(V_enc_cloner);
}
VNET_SYSUNINIT(vnet_enc_uninit, SI_SUB_INIT_IF, SI_ORDER_ANY,
vnet_enc_uninit, NULL);
/*
* The hhook consumer needs to go before ip[6]_destroy are called on
* SI_ORDER_THIRD.
*/
static void
vnet_enc_uninit_hhook(const void *unused __unused)
{
KASSERT(V_enc_sc != NULL, ("%s: V_enc_sc is %p\n", __func__, V_enc_sc));
enc_remove_hhooks(V_enc_sc);
}
VNET_SYSUNINIT(vnet_enc_uninit_hhook, SI_SUB_PROTO_DOMAIN, SI_ORDER_FOURTH,
vnet_enc_uninit_hhook, NULL);
static int
enc_modevent(module_t mod, int type, void *data)
{
switch (type) {
case MOD_LOAD:
case MOD_UNLOAD:
break;
default:
return (EOPNOTSUPP);
}
return (0);
}
static moduledata_t enc_mod = {
"if_enc",
enc_modevent,
0
};
DECLARE_MODULE(if_enc, enc_mod, SI_SUB_PSEUDO, SI_ORDER_ANY);
MODULE_VERSION(if_enc, 1);