From 13b52cd44670e3359055e9918d0e766d89836425 Mon Sep 17 00:00:00 2001 From: Brian Haley Date: Fri, 23 Apr 2010 11:26:08 +0000 Subject: IPv6: Add dontfrag argument to relevant functions Add dontfrag argument to relevant functions for IPV6_DONTFRAG support, as well as allowing the value to be passed-in via ancillary cmsg data. Signed-off-by: Brian Haley Signed-off-by: David S. Miller --- net/ipv6/raw.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) (limited to 'net/ipv6/raw.c') diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index 8763b1a0814..44a84ea9b3e 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -733,6 +733,7 @@ static int rawv6_sendmsg(struct kiocb *iocb, struct sock *sk, int addr_len = msg->msg_namelen; int hlimit = -1; int tclass = -1; + int dontfrag = -1; u16 proto; int err; @@ -811,7 +812,8 @@ static int rawv6_sendmsg(struct kiocb *iocb, struct sock *sk, memset(opt, 0, sizeof(struct ipv6_txoptions)); opt->tot_len = sizeof(struct ipv6_txoptions); - err = datagram_send_ctl(sock_net(sk), msg, &fl, opt, &hlimit, &tclass); + err = datagram_send_ctl(sock_net(sk), msg, &fl, opt, &hlimit, + &tclass, &dontfrag); if (err < 0) { fl6_sock_release(flowlabel); return err; @@ -880,6 +882,9 @@ static int rawv6_sendmsg(struct kiocb *iocb, struct sock *sk, if (tclass < 0) tclass = np->tclass; + if (dontfrag < 0) + dontfrag = np->dontfrag; + if (msg->msg_flags&MSG_CONFIRM) goto do_confirm; @@ -890,7 +895,7 @@ back_from_confirm: lock_sock(sk); err = ip6_append_data(sk, ip_generic_getfrag, msg->msg_iov, len, 0, hlimit, tclass, opt, &fl, (struct rt6_info*)dst, - msg->msg_flags); + msg->msg_flags, dontfrag); if (err) ip6_flush_pending_frames(sk); -- cgit v1.2.3-70-g09d2 From 4b340ae20d0e2366792abe70f46629e576adaf5e Mon Sep 17 00:00:00 2001 From: Brian Haley Date: Fri, 23 Apr 2010 11:26:09 +0000 Subject: IPv6: Complete IPV6_DONTFRAG support Finally add support to detect a local IPV6_DONTFRAG event and return the relevant data to the user if they've enabled IPV6_RECVPATHMTU on the socket. The next recvmsg() will return no data, but have an IPV6_PATHMTU as ancillary data. Signed-off-by: Brian Haley Signed-off-by: David S. Miller --- include/linux/ipv6.h | 2 ++ include/net/ipv6.h | 2 ++ net/ipv6/af_inet6.c | 3 ++ net/ipv6/datagram.c | 87 +++++++++++++++++++++++++++++++++++++++++++++++++++ net/ipv6/ip6_output.c | 24 +++++++++----- net/ipv6/raw.c | 3 ++ net/ipv6/udp.c | 3 ++ 7 files changed, 116 insertions(+), 8 deletions(-) (limited to 'net/ipv6/raw.c') diff --git a/include/linux/ipv6.h b/include/linux/ipv6.h index 1976942cf6f..2ab5509f6d4 100644 --- a/include/linux/ipv6.h +++ b/include/linux/ipv6.h @@ -257,6 +257,7 @@ struct inet6_skb_parm { }; #define IP6CB(skb) ((struct inet6_skb_parm*)((skb)->cb)) +#define IP6CBMTU(skb) ((struct ip6_mtuinfo *)((skb)->cb)) static inline int inet6_iif(const struct sk_buff *skb) { @@ -366,6 +367,7 @@ struct ipv6_pinfo { struct ipv6_txoptions *opt; struct sk_buff *pktoptions; + struct sk_buff *rxpmtu; struct { struct ipv6_txoptions *opt; u8 hop_limit; diff --git a/include/net/ipv6.h b/include/net/ipv6.h index 7ab6323e631..eba5cc00325 100644 --- a/include/net/ipv6.h +++ b/include/net/ipv6.h @@ -578,9 +578,11 @@ extern int ip6_datagram_connect(struct sock *sk, struct sockaddr *addr, int addr_len); extern int ipv6_recv_error(struct sock *sk, struct msghdr *msg, int len); +extern int ipv6_recv_rxpmtu(struct sock *sk, struct msghdr *msg, int len); extern void ipv6_icmp_error(struct sock *sk, struct sk_buff *skb, int err, __be16 port, u32 info, u8 *payload); extern void ipv6_local_error(struct sock *sk, int err, struct flowi *fl, u32 info); +extern void ipv6_local_rxpmtu(struct sock *sk, struct flowi *fl, u32 mtu); extern int inet6_release(struct socket *sock); extern int inet6_bind(struct socket *sock, struct sockaddr *uaddr, diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index 3192aa02ba5..d2df3144429 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -417,6 +417,9 @@ void inet6_destroy_sock(struct sock *sk) if ((skb = xchg(&np->pktoptions, NULL)) != NULL) kfree_skb(skb); + if ((skb = xchg(&np->rxpmtu, NULL)) != NULL) + kfree_skb(skb); + /* Free flowlabels */ fl6_free_socklist(sk); diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c index f5076d349b1..5959230bc6c 100644 --- a/net/ipv6/datagram.c +++ b/net/ipv6/datagram.c @@ -278,6 +278,45 @@ void ipv6_local_error(struct sock *sk, int err, struct flowi *fl, u32 info) kfree_skb(skb); } +void ipv6_local_rxpmtu(struct sock *sk, struct flowi *fl, u32 mtu) +{ + struct ipv6_pinfo *np = inet6_sk(sk); + struct ipv6hdr *iph; + struct sk_buff *skb; + struct ip6_mtuinfo *mtu_info; + + if (!np->rxopt.bits.rxpmtu) + return; + + skb = alloc_skb(sizeof(struct ipv6hdr), GFP_ATOMIC); + if (!skb) + return; + + skb_put(skb, sizeof(struct ipv6hdr)); + skb_reset_network_header(skb); + iph = ipv6_hdr(skb); + ipv6_addr_copy(&iph->daddr, &fl->fl6_dst); + + mtu_info = IP6CBMTU(skb); + if (!mtu_info) { + kfree_skb(skb); + return; + } + + mtu_info->ip6m_mtu = mtu; + mtu_info->ip6m_addr.sin6_family = AF_INET6; + mtu_info->ip6m_addr.sin6_port = 0; + mtu_info->ip6m_addr.sin6_flowinfo = 0; + mtu_info->ip6m_addr.sin6_scope_id = fl->oif; + ipv6_addr_copy(&mtu_info->ip6m_addr.sin6_addr, &ipv6_hdr(skb)->daddr); + + __skb_pull(skb, skb_tail_pointer(skb) - skb->data); + skb_reset_transport_header(skb); + + skb = xchg(&np->rxpmtu, skb); + kfree_skb(skb); +} + /* * Handle MSG_ERRQUEUE */ @@ -381,6 +420,54 @@ out: return err; } +/* + * Handle IPV6_RECVPATHMTU + */ +int ipv6_recv_rxpmtu(struct sock *sk, struct msghdr *msg, int len) +{ + struct ipv6_pinfo *np = inet6_sk(sk); + struct sk_buff *skb; + struct sockaddr_in6 *sin; + struct ip6_mtuinfo mtu_info; + int err; + int copied; + + err = -EAGAIN; + skb = xchg(&np->rxpmtu, NULL); + if (skb == NULL) + goto out; + + copied = skb->len; + if (copied > len) { + msg->msg_flags |= MSG_TRUNC; + copied = len; + } + err = skb_copy_datagram_iovec(skb, 0, msg->msg_iov, copied); + if (err) + goto out_free_skb; + + sock_recv_timestamp(msg, sk, skb); + + memcpy(&mtu_info, IP6CBMTU(skb), sizeof(mtu_info)); + + sin = (struct sockaddr_in6 *)msg->msg_name; + if (sin) { + sin->sin6_family = AF_INET6; + sin->sin6_flowinfo = 0; + sin->sin6_port = 0; + sin->sin6_scope_id = mtu_info.ip6m_addr.sin6_scope_id; + ipv6_addr_copy(&sin->sin6_addr, &mtu_info.ip6m_addr.sin6_addr); + } + + put_cmsg(msg, SOL_IPV6, IPV6_PATHMTU, sizeof(mtu_info), &mtu_info); + + err = copied; + +out_free_skb: + kfree_skb(skb); +out: + return err; +} int datagram_recv_ctl(struct sock *sk, struct msghdr *msg, struct sk_buff *skb) diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index 54d43dd1f08..61e2bef5609 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -1219,15 +1219,23 @@ int ip6_append_data(struct sock *sk, int getfrag(void *from, char *to, */ inet->cork.length += length; - if (((length > mtu) && (sk->sk_protocol == IPPROTO_UDP)) && - (rt->u.dst.dev->features & NETIF_F_UFO)) { + if (length > mtu) { + int proto = sk->sk_protocol; + if (dontfrag && (proto == IPPROTO_UDP || proto == IPPROTO_RAW)){ + ipv6_local_rxpmtu(sk, fl, mtu-exthdrlen); + return -EMSGSIZE; + } - err = ip6_ufo_append_data(sk, getfrag, from, length, hh_len, - fragheaderlen, transhdrlen, mtu, - flags); - if (err) - goto error; - return 0; + if (proto == IPPROTO_UDP && + (rt->u.dst.dev->features & NETIF_F_UFO)) { + + err = ip6_ufo_append_data(sk, getfrag, from, length, + hh_len, fragheaderlen, + transhdrlen, mtu, flags); + if (err) + goto error; + return 0; + } } if ((skb = skb_peek_tail(&sk->sk_write_queue)) == NULL) diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index 44a84ea9b3e..85627386cb0 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -461,6 +461,9 @@ static int rawv6_recvmsg(struct kiocb *iocb, struct sock *sk, if (flags & MSG_ERRQUEUE) return ipv6_recv_error(sk, msg, len); + if (np->rxpmtu && np->rxopt.bits.rxpmtu) + return ipv6_recv_rxpmtu(sk, msg, len); + skb = skb_recv_datagram(sk, flags, noblock, &err); if (!skb) goto out; diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index 39e3665d946..2850e35cee3 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -335,6 +335,9 @@ int udpv6_recvmsg(struct kiocb *iocb, struct sock *sk, if (flags & MSG_ERRQUEUE) return ipv6_recv_error(sk, msg, len); + if (np->rxpmtu && np->rxopt.bits.rxpmtu) + return ipv6_recv_rxpmtu(sk, msg, len); + try_again: skb = __skb_recv_datagram(sk, flags | (noblock ? MSG_DONTWAIT : 0), &peeked, &err); -- cgit v1.2.3-70-g09d2 From f84af32cbca70a3c6d30463dc08c7984af11c277 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 28 Apr 2010 15:31:51 -0700 Subject: net: ip_queue_rcv_skb() helper When queueing a skb to socket, we can immediately release its dst if target socket do not use IP_CMSG_PKTINFO. tcp_data_queue() can drop dst too. This to benefit from a hot cache line and avoid the receiver, possibly on another cpu, to dirty this cache line himself. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- include/net/ip.h | 1 + net/ipv4/ip_sockglue.c | 16 ++++++++++++++++ net/ipv4/raw.c | 2 +- net/ipv4/tcp_input.c | 1 + net/ipv4/udp.c | 2 +- net/ipv6/raw.c | 2 +- net/ipv6/udp.c | 2 +- 7 files changed, 22 insertions(+), 4 deletions(-) (limited to 'net/ipv6/raw.c') diff --git a/include/net/ip.h b/include/net/ip.h index a84ceb69268..8149b77cea9 100644 --- a/include/net/ip.h +++ b/include/net/ip.h @@ -393,6 +393,7 @@ extern int ip_options_rcv_srr(struct sk_buff *skb); * Functions provided by ip_sockglue.c */ +extern int ip_queue_rcv_skb(struct sock *sk, struct sk_buff *skb); extern void ip_cmsg_recv(struct msghdr *msg, struct sk_buff *skb); extern int ip_cmsg_send(struct net *net, struct msghdr *msg, struct ipcm_cookie *ipc); diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c index b0aa0546a3b..ce231780a2b 100644 --- a/net/ipv4/ip_sockglue.c +++ b/net/ipv4/ip_sockglue.c @@ -954,6 +954,22 @@ e_inval: return -EINVAL; } +/** + * ip_queue_rcv_skb - Queue an skb into sock receive queue + * @sk: socket + * @skb: buffer + * + * Queues an skb into socket receive queue. If IP_CMSG_PKTINFO option + * is not set, we drop skb dst entry now, while dst cache line is hot. + */ +int ip_queue_rcv_skb(struct sock *sk, struct sk_buff *skb) +{ + if (!(inet_sk(sk)->cmsg_flags & IP_CMSG_PKTINFO)) + skb_dst_drop(skb); + return sock_queue_rcv_skb(sk, skb); +} +EXPORT_SYMBOL(ip_queue_rcv_skb); + int ip_setsockopt(struct sock *sk, int level, int optname, char __user *optval, unsigned int optlen) { diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index cc6f097fbd5..52ef5af78a4 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -290,7 +290,7 @@ static int raw_rcv_skb(struct sock * sk, struct sk_buff * skb) { /* Charge it to the socket. */ - if (sock_queue_rcv_skb(sk, skb) < 0) { + if (ip_queue_rcv_skb(sk, skb) < 0) { kfree_skb(skb); return NET_RX_DROP; } diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index ae3ec15fb63..e82162c211b 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -4367,6 +4367,7 @@ static void tcp_data_queue(struct sock *sk, struct sk_buff *skb) if (TCP_SKB_CB(skb)->seq == TCP_SKB_CB(skb)->end_seq) goto drop; + skb_dst_drop(skb); __skb_pull(skb, th->doff * 4); TCP_ECN_accept_cwr(tp, skb); diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index 1f86965ba7d..4560b291180 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -1264,7 +1264,7 @@ static int __udp_queue_rcv_skb(struct sock *sk, struct sk_buff *skb) if (inet_sk(sk)->inet_daddr) sock_rps_save_rxhash(sk, skb->rxhash); - rc = sock_queue_rcv_skb(sk, skb); + rc = ip_queue_rcv_skb(sk, skb); if (rc < 0) { int is_udplite = IS_UDPLITE(sk); diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index 85627386cb0..0e3d2dd9207 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -381,7 +381,7 @@ static inline int rawv6_rcv_skb(struct sock * sk, struct sk_buff * skb) } /* Charge it to the socket. */ - if (sock_queue_rcv_skb(sk, skb) < 0) { + if (ip_queue_rcv_skb(sk, skb) < 0) { kfree_skb(skb); return NET_RX_DROP; } diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index 91c60f0090a..79359c8380b 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -514,7 +514,7 @@ int udpv6_queue_rcv_skb(struct sock * sk, struct sk_buff *skb) goto drop; } - if ((rc = sock_queue_rcv_skb(sk, skb)) < 0) { + if ((rc = ip_queue_rcv_skb(sk, skb)) < 0) { /* Note that an ENOMEM error is charged twice */ if (rc == -ENOMEM) UDP6_INC_STATS_BH(sock_net(sk), -- cgit v1.2.3-70-g09d2