freebsd-dev/sys/net/if_loop.c
Andrew Gallatin eedb49598b Clear mbuf hashtype on loopback when RSS is enabled.
The hashtype on an outgoing mbuf reflects the correct hash on the
transmit side of the connection.  If this hash persists on loopback,
the receiving RSS/PCBGROUP code will use it to look up the pcbgroup
for the transmit side, which will often not match the pcbgroup for the
receive side of the connection.  This leads to TCP connections
hanging, and dropping the SYN/ACK packet.   This is essentially
the same as having a hardware network card generate mbufs with an
incorrect RSS hash.

There are a number of places which can set the hash on transmit,
so the simplest fix is to simply clear the hash at loopback time.
Clearing the hash allows a new, correct hash to be calculated in
software on the receive side.

Reviewed by:	jtl
Discussed with:	adrian
Sponsored by:	Netflix
2016-10-20 13:48:29 +00:00

452 lines
10 KiB
C

/*-
* 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
* $FreeBSD$
*/
/*
* Loopback interface driver for protocol testing and timing.
*/
#include "opt_inet.h"
#include "opt_inet6.h"
#include "opt_rss.h"
#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>
#include <sys/socket.h>
#include <sys/sockio.h>
#include <sys/sysctl.h>
#include <net/if.h>
#include <net/if_var.h>
#include <net/if_clone.h>
#include <net/if_types.h>
#include <net/netisr.h>
#include <net/route.h>
#include <net/bpf.h>
#include <net/vnet.h>
#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>
#ifdef TINY_LOMTU
#define LOMTU (1024+512)
#elif defined(LARGE_LOMTU)
#define LOMTU 131072
#else
#define LOMTU 16384
#endif
#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);
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 *);
VNET_DEFINE(struct ifnet *, loif); /* Used externally */
#ifdef VIMAGE
static VNET_DEFINE(struct if_clone *, lo_cloner);
#define V_lo_cloner VNET(lo_cloner)
#endif
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_PSEUDO, 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_INIT_IF, SI_ORDER_SECOND,
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);
int
looutput(struct ifnet *ifp, struct mbuf *m, const struct sockaddr *dst,
struct route *ro)
{
u_int32_t af;
#ifdef MAC
int error;
#endif
M_ASSERTPKTHDR(m); /* check if we have the packet header */
#ifdef MAC
error = mac_ifnet_check_transmit(ifp, m);
if (error) {
m_freem(m);
return (error);
}
#endif
if (ro != NULL && ro->ro_flags & (RT_REJECT|RT_BLACKHOLE)) {
m_freem(m);
return (ro->ro_flags & RT_BLACKHOLE ? 0 : EHOSTUNREACH);
}
if_inc_counter(ifp, IFCOUNTER_OPACKETS, 1);
if_inc_counter(ifp, IFCOUNTER_OBYTES, m->m_pkthdr.len);
#ifdef RSS
M_HASHTYPE_CLEAR(m);
#endif
/* BPF writes need to be handled specially. */
if (dst->sa_family == AF_UNSPEC || dst->sa_family == pseudo_AF_HDRCMPLT)
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)
{
int isr;
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 */
u_int32_t af1 = af;
/*
* We need to prepend the address family.
*/
bpf_mtap2(V_loif->if_bpf, &af1, sizeof(af1), m);
}
}
}
/* 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) {
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);
m->m_data -= (mtod(m,vm_offset_t) & 3);
}
#endif
}
/* Deliver to upper layer protocol */
switch (af) {
#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
default:
printf("if_simloop: can't handle af=%d\n", af);
m_freem(m);
return (EAFNOSUPPORT);
}
if_inc_counter(ifp, IFCOUNTER_IPACKETS, 1);
if_inc_counter(ifp, IFCOUNTER_IBYTES, m->m_pkthdr.len);
netisr_queue(isr, m); /* mbuf is free'd on failure. */
return (0);
}
/*
* Process an ioctl request.
*/
/* ARGSUSED */
int
loioctl(struct ifnet *ifp, u_long cmd, caddr_t data)
{
struct ifreq *ifr = (struct ifreq *)data;
int error = 0, mask;
switch (cmd) {
case SIOCSIFADDR:
ifp->if_flags |= IFF_UP;
ifp->if_drv_flags |= IFF_DRV_RUNNING;
/*
* Everything else is done at a higher level.
*/
break;
case SIOCADDMULTI:
case SIOCDELMULTI:
if (ifr == NULL) {
error = EAFNOSUPPORT; /* XXX */
break;
}
switch (ifr->ifr_addr.sa_family) {
#ifdef INET
case AF_INET:
break;
#endif
#ifdef INET6
case AF_INET6:
break;
#endif
default:
error = EAFNOSUPPORT;
break;
}
break;
case SIOCSIFMTU:
ifp->if_mtu = ifr->ifr_mtu;
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;
default:
error = EINVAL;
}
return (error);
}