From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752836Ab3FGGMY (ORCPT ); Fri, 7 Jun 2013 02:12:24 -0400 Received: from shadbolt.e.decadent.org.uk ([88.96.1.126]:50293 "EHLO shadbolt.e.decadent.org.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752633Ab3FGGMN (ORCPT ); Fri, 7 Jun 2013 02:12:13 -0400 Message-ID: <1370585517.4021.94.camel@deadeye.wl.decadent.org.uk> Subject: Re: [ 157/184] inet: add RCU protection to inet->opt From: Ben Hutchings To: Willy Tarreau Cc: linux-kernel@vger.kernel.org, stable@vger.kernel.org, Eric Dumazet , Herbert Xu , "David S. Miller" Date: Fri, 07 Jun 2013 07:11:57 +0100 In-Reply-To: <20130604172136.807153289@1wt.eu> References: <20130604172136.807153289@1wt.eu> Content-Type: multipart/signed; micalg="pgp-sha512"; protocol="application/pgp-signature"; boundary="=-clMxhFgyzB8bodF5to7V" X-Mailer: Evolution 3.4.4-3 Mime-Version: 1.0 X-SA-Exim-Connect-IP: 192.168.4.101 X-SA-Exim-Mail-From: ben@decadent.org.uk X-SA-Exim-Scanned: No (on shadbolt.decadent.org.uk); SAEximRunCond expanded to false Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org --=-clMxhFgyzB8bodF5to7V Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable On Tue, 2013-06-04 at 19:24 +0200, Willy Tarreau wrote: > 2.6.32-longterm review patch. If anyone has any objections, please let m= e know. >=20 > ------------------ >=20 > From: Eric Dumazet >=20 > commit f6d8bd051c391c1c0458a30b2a7abcd939329259 upstream. >=20 > We lack proper synchronization to manipulate inet->opt ip_options >=20 > Problem is ip_make_skb() calls ip_setup_cork() and > ip_setup_cork() possibly makes a copy of ipc->opt (struct ip_options), > without any protection against another thread manipulating inet->opt. >=20 > Another thread can change inet->opt pointer and free old one under us. >=20 > Use RCU to protect inet->opt (changed to inet->inet_opt). >=20 > Instead of handling atomic refcounts, just copy ip_options when > necessary, to avoid cache line dirtying. >=20 > We cant insert an rcu_head in struct ip_options since its included in > skb->cb[], so this patch is large because I had to introduce a new > ip_options_rcu structure. >=20 > Signed-off-by: Eric Dumazet > Cc: Herbert Xu > Signed-off-by: David S. Miller > [dannf/bwh: backported to Debian's 2.6.32] Signed-off-by: Ben Hutchings > Signed-off-by: Willy Tarreau > --- > include/net/inet_sock.h | 14 +++-- > include/net/ip.h | 11 ++-- > net/dccp/ipv4.c | 15 +++--- > net/dccp/ipv6.c | 2 +- > net/ipv4/af_inet.c | 16 ++++-- > net/ipv4/cipso_ipv4.c | 113 ++++++++++++++++++++++------------= ------ > net/ipv4/icmp.c | 23 ++++---- > net/ipv4/inet_connection_sock.c | 8 +-- > net/ipv4/ip_options.c | 38 +++++++------- > net/ipv4/ip_output.c | 50 +++++++++--------- > net/ipv4/ip_sockglue.c | 33 ++++++++---- > net/ipv4/raw.c | 19 +++++-- > net/ipv4/syncookies.c | 4 +- > net/ipv4/tcp_ipv4.c | 33 +++++++----- > net/ipv4/udp.c | 21 ++++++-- > net/ipv6/tcp_ipv6.c | 2 +- > 16 files changed, 235 insertions(+), 167 deletions(-) >=20 > diff --git a/include/net/inet_sock.h b/include/net/inet_sock.h > index 47004f3..cf65e77 100644 > --- a/include/net/inet_sock.h > +++ b/include/net/inet_sock.h > @@ -56,7 +56,15 @@ struct ip_options { > unsigned char __data[0]; > }; > =20 > -#define optlength(opt) (sizeof(struct ip_options) + opt->optlen) > +struct ip_options_rcu { > + struct rcu_head rcu; > + struct ip_options opt; > +}; > + > +struct ip_options_data { > + struct ip_options_rcu opt; > + char data[40]; > +}; > =20 > struct inet_request_sock { > struct request_sock req; > @@ -77,7 +85,7 @@ struct inet_request_sock { > acked : 1, > no_srccheck: 1; > kmemcheck_bitfield_end(flags); > - struct ip_options *opt; > + struct ip_options_rcu *opt; > }; > =20 > static inline struct inet_request_sock *inet_rsk(const struct request_so= ck *sk) > @@ -122,7 +130,7 @@ struct inet_sock { > __be32 saddr; > __s16 uc_ttl; > __u16 cmsg_flags; > - struct ip_options *opt; > + struct ip_options_rcu *inet_opt; > __be16 sport; > __u16 id; > __u8 tos; > diff --git a/include/net/ip.h b/include/net/ip.h > index 69db943..a7d4675 100644 > --- a/include/net/ip.h > +++ b/include/net/ip.h > @@ -54,7 +54,7 @@ struct ipcm_cookie > { > __be32 addr; > int oif; > - struct ip_options *opt; > + struct ip_options_rcu *opt; > union skb_shared_tx shtx; > }; > =20 > @@ -92,7 +92,7 @@ extern int igmp_mc_proc_init(void); > =20 > extern int ip_build_and_send_pkt(struct sk_buff *skb, struct sock *sk, > __be32 saddr, __be32 daddr, > - struct ip_options *opt); > + struct ip_options_rcu *opt); > extern int ip_rcv(struct sk_buff *skb, struct net_device *dev, > struct packet_type *pt, struct net_device *orig_dev); > extern int ip_local_deliver(struct sk_buff *skb); > @@ -362,14 +362,15 @@ extern int ip_forward(struct sk_buff *skb); > * Functions provided by ip_options.c > */ > =20 > -extern void ip_options_build(struct sk_buff *skb, struct ip_options *opt= , __be32 daddr, struct rtable *rt, int is_frag); > +extern void ip_options_build(struct sk_buff *skb, struct ip_options *opt= , > + __be32 daddr, struct rtable *rt, int is_frag); > extern int ip_options_echo(struct ip_options *dopt, struct sk_buff *skb)= ; > extern void ip_options_fragment(struct sk_buff *skb); > extern int ip_options_compile(struct net *net, > struct ip_options *opt, struct sk_buff *skb); > -extern int ip_options_get(struct net *net, struct ip_options **optp, > +extern int ip_options_get(struct net *net, struct ip_options_rcu **optp, > unsigned char *data, int optlen); > -extern int ip_options_get_from_user(struct net *net, struct ip_options *= *optp, > +extern int ip_options_get_from_user(struct net *net, struct ip_options_r= cu **optp, > unsigned char __user *data, int optlen); > extern void ip_options_undo(struct ip_options * opt); > extern void ip_forward_options(struct sk_buff *skb); > diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c > index d14c0a3..cef3656 100644 > --- a/net/dccp/ipv4.c > +++ b/net/dccp/ipv4.c > @@ -47,6 +47,7 @@ int dccp_v4_connect(struct sock *sk, struct sockaddr *u= addr, int addr_len) > __be32 daddr, nexthop; > int tmp; > int err; > + struct ip_options_rcu *inet_opt; > =20 > dp->dccps_role =3D DCCP_ROLE_CLIENT; > =20 > @@ -57,10 +58,12 @@ int dccp_v4_connect(struct sock *sk, struct sockaddr = *uaddr, int addr_len) > return -EAFNOSUPPORT; > =20 > nexthop =3D daddr =3D usin->sin_addr.s_addr; > - if (inet->opt !=3D NULL && inet->opt->srr) { > + > + inet_opt =3D inet->inet_opt; > + if (inet_opt !=3D NULL && inet_opt->opt.srr) { > if (daddr =3D=3D 0) > return -EINVAL; > - nexthop =3D inet->opt->faddr; > + nexthop =3D inet_opt->opt.faddr; > } > =20 > tmp =3D ip_route_connect(&rt, nexthop, inet->saddr, > @@ -75,7 +78,7 @@ int dccp_v4_connect(struct sock *sk, struct sockaddr *u= addr, int addr_len) > return -ENETUNREACH; > } > =20 > - if (inet->opt =3D=3D NULL || !inet->opt->srr) > + if (inet_opt =3D=3D NULL || !inet_opt->opt.srr) > daddr =3D rt->rt_dst; > =20 > if (inet->saddr =3D=3D 0) > @@ -86,8 +89,8 @@ int dccp_v4_connect(struct sock *sk, struct sockaddr *u= addr, int addr_len) > inet->daddr =3D daddr; > =20 > inet_csk(sk)->icsk_ext_hdr_len =3D 0; > - if (inet->opt !=3D NULL) > - inet_csk(sk)->icsk_ext_hdr_len =3D inet->opt->optlen; > + if (inet_opt) > + inet_csk(sk)->icsk_ext_hdr_len =3D inet_opt->opt.optlen; > /* > * Socket identity is still unknown (sport may be zero). > * However we set state to DCCP_REQUESTING and not releasing socket > @@ -397,7 +400,7 @@ struct sock *dccp_v4_request_recv_sock(struct sock *s= k, struct sk_buff *skb, > newinet->daddr =3D ireq->rmt_addr; > newinet->rcv_saddr =3D ireq->loc_addr; > newinet->saddr =3D ireq->loc_addr; > - newinet->opt =3D ireq->opt; > + newinet->inet_opt =3D ireq->opt; > ireq->opt =3D NULL; > newinet->mc_index =3D inet_iif(skb); > newinet->mc_ttl =3D ip_hdr(skb)->ttl; > diff --git a/net/dccp/ipv6.c b/net/dccp/ipv6.c > index 9ed1962..2f11de7 100644 > --- a/net/dccp/ipv6.c > +++ b/net/dccp/ipv6.c > @@ -600,7 +600,7 @@ static struct sock *dccp_v6_request_recv_sock(struct = sock *sk, > =20 > First: no IPv4 options. > */ > - newinet->opt =3D NULL; > + newinet->inet_opt =3D NULL; > =20 > /* Clone RX bits */ > newnp->rxopt.all =3D np->rxopt.all; > diff --git a/net/ipv4/af_inet.c b/net/ipv4/af_inet.c > index a289878..d1992a4 100644 > --- a/net/ipv4/af_inet.c > +++ b/net/ipv4/af_inet.c > @@ -152,7 +152,7 @@ void inet_sock_destruct(struct sock *sk) > WARN_ON(sk->sk_wmem_queued); > WARN_ON(sk->sk_forward_alloc); > =20 > - kfree(inet->opt); > + kfree(inet->inet_opt); > dst_release(sk->sk_dst_cache); > sk_refcnt_debug_dec(sk); > } > @@ -1065,9 +1065,11 @@ static int inet_sk_reselect_saddr(struct sock *sk) > __be32 old_saddr =3D inet->saddr; > __be32 new_saddr; > __be32 daddr =3D inet->daddr; > + struct ip_options_rcu *inet_opt; > =20 > - if (inet->opt && inet->opt->srr) > - daddr =3D inet->opt->faddr; > + inet_opt =3D inet->inet_opt; > + if (inet_opt && inet_opt->opt.srr) > + daddr =3D inet_opt->opt.faddr; > =20 > /* Query new route. */ > err =3D ip_route_connect(&rt, daddr, 0, > @@ -1109,6 +1111,7 @@ int inet_sk_rebuild_header(struct sock *sk) > struct inet_sock *inet =3D inet_sk(sk); > struct rtable *rt =3D (struct rtable *)__sk_dst_check(sk, 0); > __be32 daddr; > + struct ip_options_rcu *inet_opt; > int err; > =20 > /* Route is OK, nothing to do. */ > @@ -1116,9 +1119,12 @@ int inet_sk_rebuild_header(struct sock *sk) > return 0; > =20 > /* Reroute. */ > + rcu_read_lock(); > + inet_opt =3D rcu_dereference(inet->inet_opt); > daddr =3D inet->daddr; > - if (inet->opt && inet->opt->srr) > - daddr =3D inet->opt->faddr; > + if (inet_opt && inet_opt->opt.srr) > + daddr =3D inet_opt->opt.faddr; > + rcu_read_unlock(); > { > struct flowi fl =3D { > .oif =3D sk->sk_bound_dev_if, > diff --git a/net/ipv4/cipso_ipv4.c b/net/ipv4/cipso_ipv4.c > index 10f8f8d..b6d06d6 100644 > --- a/net/ipv4/cipso_ipv4.c > +++ b/net/ipv4/cipso_ipv4.c > @@ -1860,6 +1860,11 @@ static int cipso_v4_genopt(unsigned char *buf, u32= buf_len, > return CIPSO_V4_HDR_LEN + ret_val; > } > =20 > +static void opt_kfree_rcu(struct rcu_head *head) > +{ > + kfree(container_of(head, struct ip_options_rcu, rcu)); > +} > + > /** > * cipso_v4_sock_setattr - Add a CIPSO option to a socket > * @sk: the socket > @@ -1882,7 +1887,7 @@ int cipso_v4_sock_setattr(struct sock *sk, > unsigned char *buf =3D NULL; > u32 buf_len; > u32 opt_len; > - struct ip_options *opt =3D NULL; > + struct ip_options_rcu *old, *opt =3D NULL; > struct inet_sock *sk_inet; > struct inet_connection_sock *sk_conn; > =20 > @@ -1918,22 +1923,25 @@ int cipso_v4_sock_setattr(struct sock *sk, > ret_val =3D -ENOMEM; > goto socket_setattr_failure; > } > - memcpy(opt->__data, buf, buf_len); > - opt->optlen =3D opt_len; > - opt->cipso =3D sizeof(struct iphdr); > + memcpy(opt->opt.__data, buf, buf_len); > + opt->opt.optlen =3D opt_len; > + opt->opt.cipso =3D sizeof(struct iphdr); > kfree(buf); > buf =3D NULL; > =20 > sk_inet =3D inet_sk(sk); > + > + old =3D sk_inet->inet_opt; > if (sk_inet->is_icsk) { > sk_conn =3D inet_csk(sk); > - if (sk_inet->opt) > - sk_conn->icsk_ext_hdr_len -=3D sk_inet->opt->optlen; > - sk_conn->icsk_ext_hdr_len +=3D opt->optlen; > + if (old) > + sk_conn->icsk_ext_hdr_len -=3D old->opt.optlen; > + sk_conn->icsk_ext_hdr_len +=3D opt->opt.optlen; > sk_conn->icsk_sync_mss(sk, sk_conn->icsk_pmtu_cookie); > } > - opt =3D xchg(&sk_inet->opt, opt); > - kfree(opt); > + rcu_assign_pointer(sk_inet->inet_opt, opt); > + if (old) > + call_rcu(&old->rcu, opt_kfree_rcu); > =20 > return 0; > =20 > @@ -1963,7 +1971,7 @@ int cipso_v4_req_setattr(struct request_sock *req, > unsigned char *buf =3D NULL; > u32 buf_len; > u32 opt_len; > - struct ip_options *opt =3D NULL; > + struct ip_options_rcu *opt =3D NULL; > struct inet_request_sock *req_inet; > =20 > /* We allocate the maximum CIPSO option size here so we are probably > @@ -1991,15 +1999,16 @@ int cipso_v4_req_setattr(struct request_sock *req= , > ret_val =3D -ENOMEM; > goto req_setattr_failure; > } > - memcpy(opt->__data, buf, buf_len); > - opt->optlen =3D opt_len; > - opt->cipso =3D sizeof(struct iphdr); > + memcpy(opt->opt.__data, buf, buf_len); > + opt->opt.optlen =3D opt_len; > + opt->opt.cipso =3D sizeof(struct iphdr); > kfree(buf); > buf =3D NULL; > =20 > req_inet =3D inet_rsk(req); > opt =3D xchg(&req_inet->opt, opt); > - kfree(opt); > + if (opt) > + call_rcu(&opt->rcu, opt_kfree_rcu); > =20 > return 0; > =20 > @@ -2019,34 +2028,34 @@ req_setattr_failure: > * values on failure. > * > */ > -int cipso_v4_delopt(struct ip_options **opt_ptr) > +int cipso_v4_delopt(struct ip_options_rcu **opt_ptr) > { > int hdr_delta =3D 0; > - struct ip_options *opt =3D *opt_ptr; > + struct ip_options_rcu *opt =3D *opt_ptr; > =20 > - if (opt->srr || opt->rr || opt->ts || opt->router_alert) { > + if (opt->opt.srr || opt->opt.rr || opt->opt.ts || opt->opt.router_alert= ) { > u8 cipso_len; > u8 cipso_off; > unsigned char *cipso_ptr; > int iter; > int optlen_new; > =20 > - cipso_off =3D opt->cipso - sizeof(struct iphdr); > - cipso_ptr =3D &opt->__data[cipso_off]; > + cipso_off =3D opt->opt.cipso - sizeof(struct iphdr); > + cipso_ptr =3D &opt->opt.__data[cipso_off]; > cipso_len =3D cipso_ptr[1]; > =20 > - if (opt->srr > opt->cipso) > - opt->srr -=3D cipso_len; > - if (opt->rr > opt->cipso) > - opt->rr -=3D cipso_len; > - if (opt->ts > opt->cipso) > - opt->ts -=3D cipso_len; > - if (opt->router_alert > opt->cipso) > - opt->router_alert -=3D cipso_len; > - opt->cipso =3D 0; > + if (opt->opt.srr > opt->opt.cipso) > + opt->opt.srr -=3D cipso_len; > + if (opt->opt.rr > opt->opt.cipso) > + opt->opt.rr -=3D cipso_len; > + if (opt->opt.ts > opt->opt.cipso) > + opt->opt.ts -=3D cipso_len; > + if (opt->opt.router_alert > opt->opt.cipso) > + opt->opt.router_alert -=3D cipso_len; > + opt->opt.cipso =3D 0; > =20 > memmove(cipso_ptr, cipso_ptr + cipso_len, > - opt->optlen - cipso_off - cipso_len); > + opt->opt.optlen - cipso_off - cipso_len); > =20 > /* determining the new total option length is tricky because of > * the padding necessary, the only thing i can think to do at > @@ -2055,21 +2064,21 @@ int cipso_v4_delopt(struct ip_options **opt_ptr) > * from there we can determine the new total option length */ > iter =3D 0; > optlen_new =3D 0; > - while (iter < opt->optlen) > - if (opt->__data[iter] !=3D IPOPT_NOP) { > - iter +=3D opt->__data[iter + 1]; > + while (iter < opt->opt.optlen) > + if (opt->opt.__data[iter] !=3D IPOPT_NOP) { > + iter +=3D opt->opt.__data[iter + 1]; > optlen_new =3D iter; > } else > iter++; > - hdr_delta =3D opt->optlen; > - opt->optlen =3D (optlen_new + 3) & ~3; > - hdr_delta -=3D opt->optlen; > + hdr_delta =3D opt->opt.optlen; > + opt->opt.optlen =3D (optlen_new + 3) & ~3; > + hdr_delta -=3D opt->opt.optlen; > } else { > /* only the cipso option was present on the socket so we can > * remove the entire option struct */ > *opt_ptr =3D NULL; > - hdr_delta =3D opt->optlen; > - kfree(opt); > + hdr_delta =3D opt->opt.optlen; > + call_rcu(&opt->rcu, opt_kfree_rcu); > } > =20 > return hdr_delta; > @@ -2086,15 +2095,15 @@ int cipso_v4_delopt(struct ip_options **opt_ptr) > void cipso_v4_sock_delattr(struct sock *sk) > { > int hdr_delta; > - struct ip_options *opt; > + struct ip_options_rcu *opt; > struct inet_sock *sk_inet; > =20 > sk_inet =3D inet_sk(sk); > - opt =3D sk_inet->opt; > - if (opt =3D=3D NULL || opt->cipso =3D=3D 0) > + opt =3D sk_inet->inet_opt; > + if (opt =3D=3D NULL || opt->opt.cipso =3D=3D 0) > return; > =20 > - hdr_delta =3D cipso_v4_delopt(&sk_inet->opt); > + hdr_delta =3D cipso_v4_delopt(&sk_inet->inet_opt); > if (sk_inet->is_icsk && hdr_delta > 0) { > struct inet_connection_sock *sk_conn =3D inet_csk(sk); > sk_conn->icsk_ext_hdr_len -=3D hdr_delta; > @@ -2112,12 +2121,12 @@ void cipso_v4_sock_delattr(struct sock *sk) > */ > void cipso_v4_req_delattr(struct request_sock *req) > { > - struct ip_options *opt; > + struct ip_options_rcu *opt; > struct inet_request_sock *req_inet; > =20 > req_inet =3D inet_rsk(req); > opt =3D req_inet->opt; > - if (opt =3D=3D NULL || opt->cipso =3D=3D 0) > + if (opt =3D=3D NULL || opt->opt.cipso =3D=3D 0) > return; > =20 > cipso_v4_delopt(&req_inet->opt); > @@ -2187,14 +2196,18 @@ getattr_return: > */ > int cipso_v4_sock_getattr(struct sock *sk, struct netlbl_lsm_secattr *se= cattr) > { > - struct ip_options *opt; > + struct ip_options_rcu *opt; > + int res =3D -ENOMSG; > =20 > - opt =3D inet_sk(sk)->opt; > - if (opt =3D=3D NULL || opt->cipso =3D=3D 0) > - return -ENOMSG; > - > - return cipso_v4_getattr(opt->__data + opt->cipso - sizeof(struct iphdr)= , > - secattr); > + rcu_read_lock(); > + opt =3D rcu_dereference(inet_sk(sk)->inet_opt); > + if (opt && opt->opt.cipso) > + res =3D cipso_v4_getattr(opt->opt.__data + > + opt->opt.cipso - > + sizeof(struct iphdr), > + secattr); > + rcu_read_unlock(); > + return res; > } > =20 > /** > diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c > index 5bc13fe..859d781 100644 > --- a/net/ipv4/icmp.c > +++ b/net/ipv4/icmp.c > @@ -107,8 +107,7 @@ struct icmp_bxm { > __be32 times[3]; > } data; > int head_len; > - struct ip_options replyopts; > - unsigned char optbuf[40]; > + struct ip_options_data replyopts; > }; > =20 > /* An array of errno for error messages from dest unreach. */ > @@ -362,7 +361,7 @@ static void icmp_reply(struct icmp_bxm *icmp_param, s= truct sk_buff *skb) > struct inet_sock *inet; > __be32 daddr; > =20 > - if (ip_options_echo(&icmp_param->replyopts, skb)) > + if (ip_options_echo(&icmp_param->replyopts.opt.opt, skb)) > return; > =20 > sk =3D icmp_xmit_lock(net); > @@ -376,10 +375,10 @@ static void icmp_reply(struct icmp_bxm *icmp_param,= struct sk_buff *skb) > daddr =3D ipc.addr =3D rt->rt_src; > ipc.opt =3D NULL; > ipc.shtx.flags =3D 0; > - if (icmp_param->replyopts.optlen) { > - ipc.opt =3D &icmp_param->replyopts; > - if (ipc.opt->srr) > - daddr =3D icmp_param->replyopts.faddr; > + if (icmp_param->replyopts.opt.opt.optlen) { > + ipc.opt =3D &icmp_param->replyopts.opt; > + if (ipc.opt->opt.srr) > + daddr =3D icmp_param->replyopts.opt.opt.faddr; > } > { > struct flowi fl =3D { .nl_u =3D { .ip4_u =3D > @@ -516,7 +515,7 @@ void icmp_send(struct sk_buff *skb_in, int type, int = code, __be32 info) > IPTOS_PREC_INTERNETCONTROL) : > iph->tos; > =20 > - if (ip_options_echo(&icmp_param.replyopts, skb_in)) > + if (ip_options_echo(&icmp_param.replyopts.opt.opt, skb_in)) > goto out_unlock; > =20 >=20 > @@ -532,15 +531,15 @@ void icmp_send(struct sk_buff *skb_in, int type, in= t code, __be32 info) > icmp_param.offset =3D skb_network_offset(skb_in); > inet_sk(sk)->tos =3D tos; > ipc.addr =3D iph->saddr; > - ipc.opt =3D &icmp_param.replyopts; > + ipc.opt =3D &icmp_param.replyopts.opt; > ipc.shtx.flags =3D 0; > =20 > { > struct flowi fl =3D { > .nl_u =3D { > .ip4_u =3D { > - .daddr =3D icmp_param.replyopts.srr ? > - icmp_param.replyopts.faddr : > + .daddr =3D icmp_param.replyopts.opt.opt.srr ? > + icmp_param.replyopts.opt.opt.faddr : > iph->saddr, > .saddr =3D saddr, > .tos =3D RT_TOS(tos) > @@ -629,7 +628,7 @@ route_done: > room =3D dst_mtu(&rt->u.dst); > if (room > 576) > room =3D 576; > - room -=3D sizeof(struct iphdr) + icmp_param.replyopts.optlen; > + room -=3D sizeof(struct iphdr) + icmp_param.replyopts.opt.opt.optlen; > room -=3D sizeof(struct icmphdr); > =20 > icmp_param.data_len =3D skb_in->len - icmp_param.offset; > diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_s= ock.c > index 537731b..a3bf986 100644 > --- a/net/ipv4/inet_connection_sock.c > +++ b/net/ipv4/inet_connection_sock.c > @@ -356,11 +356,11 @@ struct dst_entry *inet_csk_route_req(struct sock *s= k, > { > struct rtable *rt; > const struct inet_request_sock *ireq =3D inet_rsk(req); > - struct ip_options *opt =3D inet_rsk(req)->opt; > + struct ip_options_rcu *opt =3D inet_rsk(req)->opt; > struct flowi fl =3D { .oif =3D sk->sk_bound_dev_if, > .nl_u =3D { .ip4_u =3D > - { .daddr =3D ((opt && opt->srr) ? > - opt->faddr : > + { .daddr =3D ((opt && opt->opt.srr) ? > + opt->opt.faddr : > ireq->rmt_addr), > .saddr =3D ireq->loc_addr, > .tos =3D RT_CONN_FLAGS(sk) } }, > @@ -374,7 +374,7 @@ struct dst_entry *inet_csk_route_req(struct sock *sk, > security_req_classify_flow(req, &fl); > if (ip_route_output_flow(net, &rt, &fl, sk, 0)) > goto no_route; > - if (opt && opt->is_strictroute && rt->rt_dst !=3D rt->rt_gateway) > + if (opt && opt->opt.is_strictroute && rt->rt_dst !=3D rt->rt_gateway) > goto route_err; > return &rt->u.dst; > =20 > diff --git a/net/ipv4/ip_options.c b/net/ipv4/ip_options.c > index 94bf105..8a95972 100644 > --- a/net/ipv4/ip_options.c > +++ b/net/ipv4/ip_options.c > @@ -35,7 +35,7 @@ > * saddr is address of outgoing interface. > */ > =20 > -void ip_options_build(struct sk_buff * skb, struct ip_options * opt, > +void ip_options_build(struct sk_buff *skb, struct ip_options *opt, > __be32 daddr, struct rtable *rt, int is_frag) > { > unsigned char *iph =3D skb_network_header(skb); > @@ -82,9 +82,9 @@ void ip_options_build(struct sk_buff * skb, struct ip_o= ptions * opt, > * NOTE: dopt cannot point to skb. > */ > =20 > -int ip_options_echo(struct ip_options * dopt, struct sk_buff * skb) > +int ip_options_echo(struct ip_options *dopt, struct sk_buff *skb) > { > - struct ip_options *sopt; > + const struct ip_options *sopt; > unsigned char *sptr, *dptr; > int soffset, doffset; > int optlen; > @@ -94,10 +94,8 @@ int ip_options_echo(struct ip_options * dopt, struct s= k_buff * skb) > =20 > sopt =3D &(IPCB(skb)->opt); > =20 > - if (sopt->optlen =3D=3D 0) { > - dopt->optlen =3D 0; > + if (sopt->optlen =3D=3D 0) > return 0; > - } > =20 > sptr =3D skb_network_header(skb); > dptr =3D dopt->__data; > @@ -156,7 +154,7 @@ int ip_options_echo(struct ip_options * dopt, struct = sk_buff * skb) > dopt->optlen +=3D optlen; > } > if (sopt->srr) { > - unsigned char * start =3D sptr+sopt->srr; > + unsigned char *start =3D sptr+sopt->srr; > __be32 faddr; > =20 > optlen =3D start[1]; > @@ -499,19 +497,19 @@ void ip_options_undo(struct ip_options * opt) > } > } > =20 > -static struct ip_options *ip_options_get_alloc(const int optlen) > +static struct ip_options_rcu *ip_options_get_alloc(const int optlen) > { > - return kzalloc(sizeof(struct ip_options) + ((optlen + 3) & ~3), > + return kzalloc(sizeof(struct ip_options_rcu) + ((optlen + 3) & ~3), > GFP_KERNEL); > } > =20 > -static int ip_options_get_finish(struct net *net, struct ip_options **op= tp, > - struct ip_options *opt, int optlen) > +static int ip_options_get_finish(struct net *net, struct ip_options_rcu = **optp, > + struct ip_options_rcu *opt, int optlen) > { > while (optlen & 3) > - opt->__data[optlen++] =3D IPOPT_END; > - opt->optlen =3D optlen; > - if (optlen && ip_options_compile(net, opt, NULL)) { > + opt->opt.__data[optlen++] =3D IPOPT_END; > + opt->opt.optlen =3D optlen; > + if (optlen && ip_options_compile(net, &opt->opt, NULL)) { > kfree(opt); > return -EINVAL; > } > @@ -520,29 +518,29 @@ static int ip_options_get_finish(struct net *net, s= truct ip_options **optp, > return 0; > } > =20 > -int ip_options_get_from_user(struct net *net, struct ip_options **optp, > +int ip_options_get_from_user(struct net *net, struct ip_options_rcu **op= tp, > unsigned char __user *data, int optlen) > { > - struct ip_options *opt =3D ip_options_get_alloc(optlen); > + struct ip_options_rcu *opt =3D ip_options_get_alloc(optlen); > =20 > if (!opt) > return -ENOMEM; > - if (optlen && copy_from_user(opt->__data, data, optlen)) { > + if (optlen && copy_from_user(opt->opt.__data, data, optlen)) { > kfree(opt); > return -EFAULT; > } > return ip_options_get_finish(net, optp, opt, optlen); > } > =20 > -int ip_options_get(struct net *net, struct ip_options **optp, > +int ip_options_get(struct net *net, struct ip_options_rcu **optp, > unsigned char *data, int optlen) > { > - struct ip_options *opt =3D ip_options_get_alloc(optlen); > + struct ip_options_rcu *opt =3D ip_options_get_alloc(optlen); > =20 > if (!opt) > return -ENOMEM; > if (optlen) > - memcpy(opt->__data, data, optlen); > + memcpy(opt->opt.__data, data, optlen); > return ip_options_get_finish(net, optp, opt, optlen); > } > =20 > diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c > index 44b7910..7dde039 100644 > --- a/net/ipv4/ip_output.c > +++ b/net/ipv4/ip_output.c > @@ -137,14 +137,14 @@ static inline int ip_select_ttl(struct inet_sock *i= net, struct dst_entry *dst) > * > */ > int ip_build_and_send_pkt(struct sk_buff *skb, struct sock *sk, > - __be32 saddr, __be32 daddr, struct ip_options *opt) > + __be32 saddr, __be32 daddr, struct ip_options_rcu *opt) > { > struct inet_sock *inet =3D inet_sk(sk); > struct rtable *rt =3D skb_rtable(skb); > struct iphdr *iph; > =20 > /* Build the IP header. */ > - skb_push(skb, sizeof(struct iphdr) + (opt ? opt->optlen : 0)); > + skb_push(skb, sizeof(struct iphdr) + (opt ? opt->opt.optlen : 0)); > skb_reset_network_header(skb); > iph =3D ip_hdr(skb); > iph->version =3D 4; > @@ -160,9 +160,9 @@ int ip_build_and_send_pkt(struct sk_buff *skb, struct= sock *sk, > iph->protocol =3D sk->sk_protocol; > ip_select_ident(iph, &rt->u.dst, sk); > =20 > - if (opt && opt->optlen) { > - iph->ihl +=3D opt->optlen>>2; > - ip_options_build(skb, opt, daddr, rt, 0); > + if (opt && opt->opt.optlen) { > + iph->ihl +=3D opt->opt.optlen>>2; > + ip_options_build(skb, &opt->opt, daddr, rt, 0); > } > =20 > skb->priority =3D sk->sk_priority; > @@ -312,9 +312,10 @@ int ip_queue_xmit(struct sk_buff *skb, int ipfragok) > { > struct sock *sk =3D skb->sk; > struct inet_sock *inet =3D inet_sk(sk); > - struct ip_options *opt =3D inet->opt; > + struct ip_options_rcu *inet_opt =3D NULL; > struct rtable *rt; > struct iphdr *iph; > + int res; > =20 > /* Skip all of this if the packet is already routed, > * f.e. by something like SCTP. > @@ -325,13 +326,15 @@ int ip_queue_xmit(struct sk_buff *skb, int ipfragok= ) > =20 > /* Make sure we can route this packet. */ > rt =3D (struct rtable *)__sk_dst_check(sk, 0); > + rcu_read_lock(); > + inet_opt =3D rcu_dereference(inet->inet_opt); > if (rt =3D=3D NULL) { > __be32 daddr; > =20 > /* Use correct destination address if we have options. */ > daddr =3D inet->daddr; > - if(opt && opt->srr) > - daddr =3D opt->faddr; > + if (inet_opt && inet_opt->opt.srr) > + daddr =3D inet_opt->opt.faddr; > =20 > { > struct flowi fl =3D { .oif =3D sk->sk_bound_dev_if, > @@ -359,11 +362,11 @@ int ip_queue_xmit(struct sk_buff *skb, int ipfragok= ) > skb_dst_set(skb, dst_clone(&rt->u.dst)); > =20 > packet_routed: > - if (opt && opt->is_strictroute && rt->rt_dst !=3D rt->rt_gateway) > + if (inet_opt && inet_opt->opt.is_strictroute && rt->rt_dst !=3D rt->rt_= gateway) > goto no_route; > =20 > /* OK, we know where to send it, allocate and build IP header. */ > - skb_push(skb, sizeof(struct iphdr) + (opt ? opt->optlen : 0)); > + skb_push(skb, sizeof(struct iphdr) + (inet_opt ? inet_opt->opt.optlen := 0)); > skb_reset_network_header(skb); > iph =3D ip_hdr(skb); > *((__be16 *)iph) =3D htons((4 << 12) | (5 << 8) | (inet->tos & 0xff)); > @@ -377,9 +380,9 @@ packet_routed: > iph->daddr =3D rt->rt_dst; > /* Transport layer set skb->h.foo itself. */ > =20 > - if (opt && opt->optlen) { > - iph->ihl +=3D opt->optlen >> 2; > - ip_options_build(skb, opt, inet->daddr, rt, 0); > + if (inet_opt && inet_opt->opt.optlen) { > + iph->ihl +=3D inet_opt->opt.optlen >> 2; > + ip_options_build(skb, &inet_opt->opt, inet->daddr, rt, 0); > } > =20 > ip_select_ident_more(iph, &rt->u.dst, sk, > @@ -387,10 +390,12 @@ packet_routed: > =20 > skb->priority =3D sk->sk_priority; > skb->mark =3D sk->sk_mark; > - > - return ip_local_out(skb); > + res =3D ip_local_out(skb); > + rcu_read_unlock(); > + return res; > =20 > no_route: > + rcu_read_unlock(); > IP_INC_STATS(sock_net(sk), IPSTATS_MIB_OUTNOROUTES); > kfree_skb(skb); > return -EHOSTUNREACH; > @@ -809,7 +814,7 @@ int ip_append_data(struct sock *sk, > /* > * setup for corking. > */ > - opt =3D ipc->opt; > + opt =3D ipc->opt ? &ipc->opt->opt : NULL; > if (opt) { > if (inet->cork.opt =3D=3D NULL) { > inet->cork.opt =3D kmalloc(sizeof(struct ip_options) + 40, sk->sk_al= location); > @@ -1367,26 +1372,23 @@ void ip_send_reply(struct sock *sk, struct sk_buf= f *skb, struct ip_reply_arg *ar > unsigned int len) > { > struct inet_sock *inet =3D inet_sk(sk); > - struct { > - struct ip_options opt; > - char data[40]; > - } replyopts; > + struct ip_options_data replyopts; > struct ipcm_cookie ipc; > __be32 daddr; > struct rtable *rt =3D skb_rtable(skb); > =20 > - if (ip_options_echo(&replyopts.opt, skb)) > + if (ip_options_echo(&replyopts.opt.opt, skb)) > return; > =20 > daddr =3D ipc.addr =3D rt->rt_src; > ipc.opt =3D NULL; > ipc.shtx.flags =3D 0; > =20 > - if (replyopts.opt.optlen) { > + if (replyopts.opt.opt.optlen) { > ipc.opt =3D &replyopts.opt; > =20 > - if (ipc.opt->srr) > - daddr =3D replyopts.opt.faddr; > + if (replyopts.opt.opt.srr) > + daddr =3D replyopts.opt.opt.faddr; > } > =20 > { > diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c > index 184a7ad..099e6c3 100644 > --- a/net/ipv4/ip_sockglue.c > +++ b/net/ipv4/ip_sockglue.c > @@ -434,6 +434,11 @@ out: > } > =20 >=20 > +static void opt_kfree_rcu(struct rcu_head *head) > +{ > + kfree(container_of(head, struct ip_options_rcu, rcu)); > +} > + > /* > * Socket option code for IP. This is the end of the line after any > * TCP,UDP etc options on an IP socket. > @@ -479,13 +484,15 @@ static int do_ip_setsockopt(struct sock *sk, int le= vel, > switch (optname) { > case IP_OPTIONS: > { > - struct ip_options *opt =3D NULL; > + struct ip_options_rcu *old, *opt =3D NULL; > + > if (optlen > 40 || optlen < 0) > goto e_inval; > err =3D ip_options_get_from_user(sock_net(sk), &opt, > optval, optlen); > if (err) > break; > + old =3D inet->inet_opt; > if (inet->is_icsk) { > struct inet_connection_sock *icsk =3D inet_csk(sk); > #if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) > @@ -494,17 +501,18 @@ static int do_ip_setsockopt(struct sock *sk, int le= vel, > (TCPF_LISTEN | TCPF_CLOSE)) && > inet->daddr !=3D LOOPBACK4_IPV6)) { > #endif > - if (inet->opt) > - icsk->icsk_ext_hdr_len -=3D inet->opt->optlen; > + if (old) > + icsk->icsk_ext_hdr_len -=3D old->opt.optlen; > if (opt) > - icsk->icsk_ext_hdr_len +=3D opt->optlen; > + icsk->icsk_ext_hdr_len +=3D opt->opt.optlen; > icsk->icsk_sync_mss(sk, icsk->icsk_pmtu_cookie); > #if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE) > } > #endif > } > - opt =3D xchg(&inet->opt, opt); > - kfree(opt); > + rcu_assign_pointer(inet->inet_opt, opt); > + if (old) > + call_rcu(&old->rcu, opt_kfree_rcu); > break; > } > case IP_PKTINFO: > @@ -1032,12 +1040,15 @@ static int do_ip_getsockopt(struct sock *sk, int = level, int optname, > case IP_OPTIONS: > { > unsigned char optbuf[sizeof(struct ip_options)+40]; > - struct ip_options * opt =3D (struct ip_options *)optbuf; > + struct ip_options *opt =3D (struct ip_options *)optbuf; > + struct ip_options_rcu *inet_opt; > + > + inet_opt =3D inet->inet_opt; > opt->optlen =3D 0; > - if (inet->opt) > - memcpy(optbuf, inet->opt, > - sizeof(struct ip_options)+ > - inet->opt->optlen); > + if (inet_opt) > + memcpy(optbuf, &inet_opt->opt, > + sizeof(struct ip_options) + > + inet_opt->opt.optlen); > release_sock(sk); > =20 > if (opt->optlen =3D=3D 0) > diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c > index ab996f9..07ab583 100644 > --- a/net/ipv4/raw.c > +++ b/net/ipv4/raw.c > @@ -459,6 +459,7 @@ static int raw_sendmsg(struct kiocb *iocb, struct soc= k *sk, struct msghdr *msg, > __be32 saddr; > u8 tos; > int err; > + struct ip_options_data opt_copy; > =20 > err =3D -EMSGSIZE; > if (len > 0xFFFF) > @@ -519,8 +520,18 @@ static int raw_sendmsg(struct kiocb *iocb, struct so= ck *sk, struct msghdr *msg, > saddr =3D ipc.addr; > ipc.addr =3D daddr; > =20 > - if (!ipc.opt) > - ipc.opt =3D inet->opt; > + if (!ipc.opt) { > + struct ip_options_rcu *inet_opt; > + > + rcu_read_lock(); > + inet_opt =3D rcu_dereference(inet->inet_opt); > + if (inet_opt) { > + memcpy(&opt_copy, inet_opt, > + sizeof(*inet_opt) + inet_opt->opt.optlen); > + ipc.opt =3D &opt_copy.opt; > + } > + rcu_read_unlock(); > + } > =20 > if (ipc.opt) { > err =3D -EINVAL; > @@ -529,10 +540,10 @@ static int raw_sendmsg(struct kiocb *iocb, struct s= ock *sk, struct msghdr *msg, > */ > if (inet->hdrincl) > goto done; > - if (ipc.opt->srr) { > + if (ipc.opt->opt.srr) { > if (!daddr) > goto done; > - daddr =3D ipc.opt->faddr; > + daddr =3D ipc.opt->opt.faddr; > } > } > tos =3D RT_CONN_FLAGS(sk); > diff --git a/net/ipv4/syncookies.c b/net/ipv4/syncookies.c > index a6e0e07..0a94b64 100644 > --- a/net/ipv4/syncookies.c > +++ b/net/ipv4/syncookies.c > @@ -309,10 +309,10 @@ struct sock *cookie_v4_check(struct sock *sk, struc= t sk_buff *skb, > * the ACK carries the same options again (see RFC1122 4.2.3.8) > */ > if (opt && opt->optlen) { > - int opt_size =3D sizeof(struct ip_options) + opt->optlen; > + int opt_size =3D sizeof(struct ip_options_rcu) + opt->optlen; > =20 > ireq->opt =3D kmalloc(opt_size, GFP_ATOMIC); > - if (ireq->opt !=3D NULL && ip_options_echo(ireq->opt, skb)) { > + if (ireq->opt !=3D NULL && ip_options_echo(&ireq->opt->opt, skb)) { > kfree(ireq->opt); > ireq->opt =3D NULL; > } > diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c > index 6a4e832..d746d3b3 100644 > --- a/net/ipv4/tcp_ipv4.c > +++ b/net/ipv4/tcp_ipv4.c > @@ -152,6 +152,7 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr *= uaddr, int addr_len) > __be32 daddr, nexthop; > int tmp; > int err; > + struct ip_options_rcu *inet_opt; > =20 > if (addr_len < sizeof(struct sockaddr_in)) > return -EINVAL; > @@ -160,10 +161,11 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr= *uaddr, int addr_len) > return -EAFNOSUPPORT; > =20 > nexthop =3D daddr =3D usin->sin_addr.s_addr; > - if (inet->opt && inet->opt->srr) { > + inet_opt =3D inet->inet_opt; > + if (inet_opt && inet_opt->opt.srr) { > if (!daddr) > return -EINVAL; > - nexthop =3D inet->opt->faddr; > + nexthop =3D inet_opt->opt.faddr; > } > =20 > tmp =3D ip_route_connect(&rt, nexthop, inet->saddr, > @@ -181,7 +183,7 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr *= uaddr, int addr_len) > return -ENETUNREACH; > } > =20 > - if (!inet->opt || !inet->opt->srr) > + if (!inet_opt || !inet_opt->opt.srr) > daddr =3D rt->rt_dst; > =20 > if (!inet->saddr) > @@ -215,8 +217,8 @@ int tcp_v4_connect(struct sock *sk, struct sockaddr *= uaddr, int addr_len) > inet->daddr =3D daddr; > =20 > inet_csk(sk)->icsk_ext_hdr_len =3D 0; > - if (inet->opt) > - inet_csk(sk)->icsk_ext_hdr_len =3D inet->opt->optlen; > + if (inet_opt) > + inet_csk(sk)->icsk_ext_hdr_len =3D inet_opt->opt.optlen; > =20 > tp->rx_opt.mss_clamp =3D 536; > =20 > @@ -802,17 +804,18 @@ static void syn_flood_warning(struct sk_buff *skb) > /* > * Save and compile IPv4 options into the request_sock if needed. > */ > -static struct ip_options *tcp_v4_save_options(struct sock *sk, > - struct sk_buff *skb) > +static struct ip_options_rcu *tcp_v4_save_options(struct sock *sk, > + struct sk_buff *skb) > { > - struct ip_options *opt =3D &(IPCB(skb)->opt); > - struct ip_options *dopt =3D NULL; > + const struct ip_options *opt =3D &(IPCB(skb)->opt); > + struct ip_options_rcu *dopt =3D NULL; > =20 > if (opt && opt->optlen) { > - int opt_size =3D optlength(opt); > + int opt_size =3D sizeof(*dopt) + opt->optlen; > + > dopt =3D kmalloc(opt_size, GFP_ATOMIC); > if (dopt) { > - if (ip_options_echo(dopt, skb)) { > + if (ip_options_echo(&dopt->opt, skb)) { > kfree(dopt); > dopt =3D NULL; > } > @@ -1362,6 +1365,7 @@ struct sock *tcp_v4_syn_recv_sock(struct sock *sk, = struct sk_buff *skb, > #ifdef CONFIG_TCP_MD5SIG > struct tcp_md5sig_key *key; > #endif > + struct ip_options_rcu *inet_opt; > =20 > if (sk_acceptq_is_full(sk)) > goto exit_overflow; > @@ -1382,13 +1386,14 @@ struct sock *tcp_v4_syn_recv_sock(struct sock *sk= , struct sk_buff *skb, > newinet->daddr =3D ireq->rmt_addr; > newinet->rcv_saddr =3D ireq->loc_addr; > newinet->saddr =3D ireq->loc_addr; > - newinet->opt =3D ireq->opt; > + inet_opt =3D ireq->opt; > + rcu_assign_pointer(newinet->inet_opt, inet_opt); > ireq->opt =3D NULL; > newinet->mc_index =3D inet_iif(skb); > newinet->mc_ttl =3D ip_hdr(skb)->ttl; > inet_csk(newsk)->icsk_ext_hdr_len =3D 0; > - if (newinet->opt) > - inet_csk(newsk)->icsk_ext_hdr_len =3D newinet->opt->optlen; > + if (inet_opt) > + inet_csk(newsk)->icsk_ext_hdr_len =3D inet_opt->opt.optlen; > newinet->id =3D newtp->write_seq ^ jiffies; > =20 > tcp_mtup_init(newsk); > diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c > index 8e28770..af559e0 100644 > --- a/net/ipv4/udp.c > +++ b/net/ipv4/udp.c > @@ -592,6 +592,7 @@ int udp_sendmsg(struct kiocb *iocb, struct sock *sk, = struct msghdr *msg, > int err, is_udplite =3D IS_UDPLITE(sk); > int corkreq =3D up->corkflag || msg->msg_flags&MSG_MORE; > int (*getfrag)(void *, char *, int, int, int, struct sk_buff *); > + struct ip_options_data opt_copy; > =20 > if (len > 0xFFFF) > return -EMSGSIZE; > @@ -663,22 +664,32 @@ int udp_sendmsg(struct kiocb *iocb, struct sock *sk= , struct msghdr *msg, > free =3D 1; > connected =3D 0; > } > - if (!ipc.opt) > - ipc.opt =3D inet->opt; > + if (!ipc.opt) { > + struct ip_options_rcu *inet_opt; > + > + rcu_read_lock(); > + inet_opt =3D rcu_dereference(inet->inet_opt); > + if (inet_opt) { > + memcpy(&opt_copy, inet_opt, > + sizeof(*inet_opt) + inet_opt->opt.optlen); > + ipc.opt =3D &opt_copy.opt; > + } > + rcu_read_unlock(); > + } > =20 > saddr =3D ipc.addr; > ipc.addr =3D faddr =3D daddr; > =20 > - if (ipc.opt && ipc.opt->srr) { > + if (ipc.opt && ipc.opt->opt.srr) { > if (!daddr) > return -EINVAL; > - faddr =3D ipc.opt->faddr; > + faddr =3D ipc.opt->opt.faddr; > connected =3D 0; > } > tos =3D RT_TOS(inet->tos); > if (sock_flag(sk, SOCK_LOCALROUTE) || > (msg->msg_flags & MSG_DONTROUTE) || > - (ipc.opt && ipc.opt->is_strictroute)) { > + (ipc.opt && ipc.opt->opt.is_strictroute)) { > tos |=3D RTO_ONLINK; > connected =3D 0; > } > diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c > index faae6df..1b25191 100644 > --- a/net/ipv6/tcp_ipv6.c > +++ b/net/ipv6/tcp_ipv6.c > @@ -1391,7 +1391,7 @@ static struct sock * tcp_v6_syn_recv_sock(struct so= ck *sk, struct sk_buff *skb, > =20 > First: no IPv4 options. > */ > - newinet->opt =3D NULL; > + newinet->inet_opt =3D NULL; > newnp->ipv6_fl_list =3D NULL; > =20 > /* Clone RX bits */ --=20 Ben Hutchings Theory and practice are closer in theory than in practice. - John Levine, moderator of comp.compilers --=-clMxhFgyzB8bodF5to7V Content-Type: application/pgp-signature; name="signature.asc" Content-Description: This is a digitally signed message part -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) iQIVAwUAUbF5ree/yOyVhhEJAQrzBhAAp07BSIAH6HKlInZN3eCx0xVDUIQ8R7Tl AMnAwxe+6TuWQB/eN0XeMzilEhtnmmbEP6NF/TaEeCGLBUTlwQHmISXr+sq+R0An gBNQuuGDVylpOwQq9DobW0G4AegJO7a8OrzxFtN/F2orb0G0w0x29w8oeM819GNf AtZAsXC6Oe3k+YVb0uGVx7thYWccaRBiUfj1QZdhlfv3ofk8EKBeqsuouj8Fu4al O1bg0OqI7w53XjcjTKw3uSBkFfdiTHbMgRmzb5FiaAh3t7HFE7sTcUuVw4DfzS6n 9P3Znfcz55VuJdSFrD4U77S5I8uIFPK2fwwrzFIkzVeBcP8gWRlm/pcfpwwnLqO+ BnSl5GXglwwYg/jNnBw6P148DOFu1JGCZdpzSJkUWdZNWbRSxtl6waovEmwMHPnQ afUNiKVYj0u4fuKN5b2wICvgSVxS4sPEi2pFyPTMXgtkTLo79hJb/Wkqd6H48em/ sUb0lWrSIpDQ5UXPUASrHdnK+rt6KQvUJ9Peo6Eh3DrIjXZaM5oTcYQLfvZ7YeTN 62VlEAgRbi0jrT6t2BfrXD10EZN0dsy9Ums1BB8Uo8hvbdvKkw1Fwv+GEVMJN+6t +dw+nFP4Eh7WlN7unvB4WOKxyFCNxhJR6ZPRojz4LtRLfr7c90oa3YkPDAjjVQks /fh4pSjwtLc= =vu9K -----END PGP SIGNATURE----- --=-clMxhFgyzB8bodF5to7V--