From 36393395536064e483b73d173f6afc103eadfbc4 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Thu, 14 Jun 2012 22:21:46 -0700 Subject: ipv4: Handle PMTU in all ICMP error handlers. With ip_rt_frag_needed() removed, we have to explicitly update PMTU information in every ICMP error handler. Create two helper functions to facilitate this. 1) ipv4_sk_update_pmtu() This updates the PMTU when we have a socket context to work with. 2) ipv4_update_pmtu() Raw version, used when no socket context is available. For this interface, we essentially just pass in explicit arguments for the flow identity information we would have extracted from the socket. And you'll notice that ipv4_sk_update_pmtu() is simply implemented in terms of ipv4_update_pmtu() Note that __ip_route_output_key() is used, rather than something like ip_route_output_flow() or ip_route_output_key(). This is because we absolutely do not want to end up with a route that does IPSEC encapsulation and the like. Instead, we only want the route that would get us to the node described by the outermost IP header. Reported-by: Steffen Klassert Signed-off-by: David S. Miller --- net/ipv4/udp.c | 1 + 1 file changed, 1 insertion(+) (limited to 'net/ipv4/udp.c') diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index eaca73644e7..db017efb76e 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -615,6 +615,7 @@ void __udp4_lib_err(struct sk_buff *skb, u32 info, struct udp_table *udptable) break; case ICMP_DEST_UNREACH: if (code == ICMP_FRAG_NEEDED) { /* Path MTU discovery */ + ipv4_sk_update_pmtu(skb, sk, info); if (inet->pmtudisc != IP_PMTUDISC_DONT) { err = EMSGSIZE; harderr = 1; -- cgit v1.2.3-70-g09d2 From 22911fc581f6a241e2897a7a8603e97344a6ec82 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 27 Jun 2012 00:23:44 +0000 Subject: net: skb_free_datagram_locked() doesnt drop all packets dropwatch wrongly diagnose all received UDP packets as drops. This patch removes trace_kfree_skb() done in skb_free_datagram_locked(). Locations calling skb_free_datagram_locked() should do it on their own. As a result, drops are accounted on the right function. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/core/datagram.c | 1 - net/ipv4/udp.c | 5 ++++- net/ipv6/udp.c | 8 +++++--- net/sunrpc/svcsock.c | 12 ++++++------ 4 files changed, 15 insertions(+), 11 deletions(-) (limited to 'net/ipv4/udp.c') diff --git a/net/core/datagram.c b/net/core/datagram.c index ae6acf6a3de..0337e2b7686 100644 --- a/net/core/datagram.c +++ b/net/core/datagram.c @@ -248,7 +248,6 @@ void skb_free_datagram_locked(struct sock *sk, struct sk_buff *skb) unlock_sock_fast(sk, slow); /* skb is now orphaned, can be freed outside of locked section */ - trace_kfree_skb(skb, skb_free_datagram_locked); __kfree_skb(skb); } EXPORT_SYMBOL(skb_free_datagram_locked); diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index db017efb76e..ee37d47d472 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -108,6 +108,7 @@ #include #include #include +#include #include "udp_impl.h" struct udp_table udp_table __read_mostly; @@ -1220,8 +1221,10 @@ try_again: goto csum_copy_err; } - if (err) + if (unlikely(err)) { + trace_kfree_skb(skb, udp_recvmsg); goto out_free; + } if (!peeked) UDP_INC_STATS_USER(sock_net(sk), diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index 051ad481973..1ecd1024948 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -48,6 +48,7 @@ #include #include +#include #include "udp_impl.h" int ipv6_rcv_saddr_equal(const struct sock *sk, const struct sock *sk2) @@ -385,15 +386,16 @@ try_again: if (skb_csum_unnecessary(skb)) err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr), - msg->msg_iov, copied ); + msg->msg_iov, copied); else { err = skb_copy_and_csum_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov); if (err == -EINVAL) goto csum_copy_err; } - if (err) + if (unlikely(err)) { + trace_kfree_skb(skb, udpv6_recvmsg); goto out_free; - + } if (!peeked) { if (is_udp4) UDP_INC_STATS_USER(sock_net(sk), diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index a6de09de5d2..18bc130255a 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -43,6 +43,7 @@ #include #include #include +#include #include #include @@ -619,6 +620,8 @@ static int svc_udp_recvfrom(struct svc_rqst *rqstp) if (!svc_udp_get_dest_address(rqstp, cmh)) { net_warn_ratelimited("svc: received unknown control message %d/%d; dropping RPC reply datagram\n", cmh->cmsg_level, cmh->cmsg_type); +out_free: + trace_kfree_skb(skb, svc_udp_recvfrom); skb_free_datagram_locked(svsk->sk_sk, skb); return 0; } @@ -630,8 +633,7 @@ static int svc_udp_recvfrom(struct svc_rqst *rqstp) if (csum_partial_copy_to_xdr(&rqstp->rq_arg, skb)) { local_bh_enable(); /* checksum error */ - skb_free_datagram_locked(svsk->sk_sk, skb); - return 0; + goto out_free; } local_bh_enable(); skb_free_datagram_locked(svsk->sk_sk, skb); @@ -640,10 +642,8 @@ static int svc_udp_recvfrom(struct svc_rqst *rqstp) rqstp->rq_arg.head[0].iov_base = skb->data + sizeof(struct udphdr); rqstp->rq_arg.head[0].iov_len = len; - if (skb_checksum_complete(skb)) { - skb_free_datagram_locked(svsk->sk_sk, skb); - return 0; - } + if (skb_checksum_complete(skb)) + goto out_free; rqstp->rq_xprt_ctxt = skb; } -- cgit v1.2.3-70-g09d2 From 55be7a9c6074f749d617a7fc1914c9a23505438c Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Wed, 11 Jul 2012 21:27:49 -0700 Subject: ipv4: Add redirect support to all protocol icmp error handlers. Signed-off-by: David S. Miller --- net/dccp/ipv4.c | 11 +++++++++++ net/ipv4/ah4.c | 18 +++++++++++++----- net/ipv4/esp4.c | 18 +++++++++++++----- net/ipv4/ip_gre.c | 9 ++++++++- net/ipv4/ipcomp.c | 18 +++++++++++++----- net/ipv4/ipip.c | 9 +++++++++ net/ipv4/ping.c | 1 + net/ipv4/raw.c | 2 ++ net/ipv4/tcp_ipv4.c | 11 +++++++++++ net/ipv4/udp.c | 3 +++ net/ipv4/xfrm4_policy.c | 10 ++++++++++ net/sctp/input.c | 16 ++++++++++++++++ 12 files changed, 110 insertions(+), 16 deletions(-) (limited to 'net/ipv4/udp.c') diff --git a/net/dccp/ipv4.c b/net/dccp/ipv4.c index 3eb76b5f221..8f41a319085 100644 --- a/net/dccp/ipv4.c +++ b/net/dccp/ipv4.c @@ -195,6 +195,14 @@ static inline void dccp_do_pmtu_discovery(struct sock *sk, } /* else let the usual retransmit timer handle it */ } +static void dccp_do_redirect(struct sk_buff *skb, struct sock *sk) +{ + struct dst_entry *dst = __sk_dst_check(sk, 0); + + if (dst && dst->ops->redirect) + dst->ops->redirect(dst, skb); +} + /* * This routine is called by the ICMP module when it gets some sort of error * condition. If err < 0 then the socket should be closed and the error @@ -259,6 +267,9 @@ static void dccp_v4_err(struct sk_buff *skb, u32 info) } switch (type) { + case ICMP_REDIRECT: + dccp_do_redirect(skb, sk); + goto out; case ICMP_SOURCE_QUENCH: /* Just silently ignore these. */ goto out; diff --git a/net/ipv4/ah4.c b/net/ipv4/ah4.c index 916d5ecaf6c..a0d8392491c 100644 --- a/net/ipv4/ah4.c +++ b/net/ipv4/ah4.c @@ -398,17 +398,25 @@ static void ah4_err(struct sk_buff *skb, u32 info) struct ip_auth_hdr *ah = (struct ip_auth_hdr *)(skb->data+(iph->ihl<<2)); struct xfrm_state *x; - if (icmp_hdr(skb)->type != ICMP_DEST_UNREACH || - icmp_hdr(skb)->code != ICMP_FRAG_NEEDED) + switch (icmp_hdr(skb)->type) { + case ICMP_DEST_UNREACH: + if (icmp_hdr(skb)->code != ICMP_FRAG_NEEDED) + return; + case ICMP_REDIRECT: + break; + default: return; + } x = xfrm_state_lookup(net, skb->mark, (const xfrm_address_t *)&iph->daddr, ah->spi, IPPROTO_AH, AF_INET); if (!x) return; - pr_debug("pmtu discovery on SA AH/%08x/%08x\n", - ntohl(ah->spi), ntohl(iph->daddr)); - ipv4_update_pmtu(skb, net, info, 0, 0, IPPROTO_AH, 0); + + if (icmp_hdr(skb)->type == ICMP_DEST_UNREACH) + ipv4_update_pmtu(skb, net, info, 0, 0, IPPROTO_AH, 0); + else + ipv4_redirect(skb, net, 0, 0, IPPROTO_AH, 0); xfrm_state_put(x); } diff --git a/net/ipv4/esp4.c b/net/ipv4/esp4.c index 7b95b49a36c..b61e9deb7c7 100644 --- a/net/ipv4/esp4.c +++ b/net/ipv4/esp4.c @@ -484,17 +484,25 @@ static void esp4_err(struct sk_buff *skb, u32 info) struct ip_esp_hdr *esph = (struct ip_esp_hdr *)(skb->data+(iph->ihl<<2)); struct xfrm_state *x; - if (icmp_hdr(skb)->type != ICMP_DEST_UNREACH || - icmp_hdr(skb)->code != ICMP_FRAG_NEEDED) + switch (icmp_hdr(skb)->type) { + case ICMP_DEST_UNREACH: + if (icmp_hdr(skb)->code != ICMP_FRAG_NEEDED) + return; + case ICMP_REDIRECT: + break; + default: return; + } x = xfrm_state_lookup(net, skb->mark, (const xfrm_address_t *)&iph->daddr, esph->spi, IPPROTO_ESP, AF_INET); if (!x) return; - NETDEBUG(KERN_DEBUG "pmtu discovery on SA ESP/%08x/%08x\n", - ntohl(esph->spi), ntohl(iph->daddr)); - ipv4_update_pmtu(skb, net, info, 0, 0, IPPROTO_ESP, 0); + + if (icmp_hdr(skb)->type == ICMP_DEST_UNREACH) + ipv4_update_pmtu(skb, net, info, 0, 0, IPPROTO_ESP, 0); + else + ipv4_redirect(skb, net, 0, 0, IPPROTO_ESP, 0); xfrm_state_put(x); } diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c index 594cec35ac4..0c3123566d7 100644 --- a/net/ipv4/ip_gre.c +++ b/net/ipv4/ip_gre.c @@ -528,6 +528,9 @@ static void ipgre_err(struct sk_buff *skb, u32 info) if (code != ICMP_EXC_TTL) return; break; + + case ICMP_REDIRECT: + break; } rcu_read_lock(); @@ -543,7 +546,11 @@ static void ipgre_err(struct sk_buff *skb, u32 info) t->parms.link, 0, IPPROTO_GRE, 0); goto out; } - + if (type == ICMP_REDIRECT) { + ipv4_redirect(skb, dev_net(skb->dev), t->parms.link, 0, + IPPROTO_GRE, 0); + goto out; + } if (t->parms.iph.daddr == 0 || ipv4_is_multicast(t->parms.iph.daddr)) goto out; diff --git a/net/ipv4/ipcomp.c b/net/ipv4/ipcomp.c index b91375482d8..d3ab47e19a8 100644 --- a/net/ipv4/ipcomp.c +++ b/net/ipv4/ipcomp.c @@ -31,18 +31,26 @@ static void ipcomp4_err(struct sk_buff *skb, u32 info) struct ip_comp_hdr *ipch = (struct ip_comp_hdr *)(skb->data+(iph->ihl<<2)); struct xfrm_state *x; - if (icmp_hdr(skb)->type != ICMP_DEST_UNREACH || - icmp_hdr(skb)->code != ICMP_FRAG_NEEDED) + switch (icmp_hdr(skb)->type) { + case ICMP_DEST_UNREACH: + if (icmp_hdr(skb)->code != ICMP_FRAG_NEEDED) + return; + case ICMP_REDIRECT: + break; + default: return; + } spi = htonl(ntohs(ipch->cpi)); x = xfrm_state_lookup(net, skb->mark, (const xfrm_address_t *)&iph->daddr, spi, IPPROTO_COMP, AF_INET); if (!x) return; - NETDEBUG(KERN_DEBUG "pmtu discovery on SA IPCOMP/%08x/%pI4\n", - spi, &iph->daddr); - ipv4_update_pmtu(skb, net, info, 0, 0, IPPROTO_COMP, 0); + + if (icmp_hdr(skb)->type == ICMP_DEST_UNREACH) + ipv4_update_pmtu(skb, net, info, 0, 0, IPPROTO_COMP, 0); + else + ipv4_redirect(skb, net, 0, 0, IPPROTO_COMP, 0); xfrm_state_put(x); } diff --git a/net/ipv4/ipip.c b/net/ipv4/ipip.c index 715338a1b20..c2d0e6d8baa 100644 --- a/net/ipv4/ipip.c +++ b/net/ipv4/ipip.c @@ -360,6 +360,8 @@ static int ipip_err(struct sk_buff *skb, u32 info) if (code != ICMP_EXC_TTL) return 0; break; + case ICMP_REDIRECT: + break; } err = -ENOENT; @@ -376,6 +378,13 @@ static int ipip_err(struct sk_buff *skb, u32 info) goto out; } + if (type == ICMP_REDIRECT) { + ipv4_redirect(skb, dev_net(skb->dev), t->dev->ifindex, 0, + IPPROTO_IPIP, 0); + err = 0; + goto out; + } + if (t->parms.iph.daddr == 0) goto out; diff --git a/net/ipv4/ping.c b/net/ipv4/ping.c index 340fcf29a96..6232d476f37 100644 --- a/net/ipv4/ping.c +++ b/net/ipv4/ping.c @@ -387,6 +387,7 @@ void ping_err(struct sk_buff *skb, u32 info) break; case ICMP_REDIRECT: /* See ICMP_SOURCE_QUENCH */ + ipv4_sk_redirect(skb, sk); err = EREMOTEIO; break; } diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index 659ddfb1094..ff0f071969e 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -218,6 +218,8 @@ static void raw_err(struct sock *sk, struct sk_buff *skb, u32 info) if (type == ICMP_DEST_UNREACH && code == ICMP_FRAG_NEEDED) ipv4_sk_update_pmtu(skb, sk, info); + else if (type == ICMP_REDIRECT) + ipv4_sk_redirect(skb, sk); /* Report error on raw socket, if: 1. User requested ip_recverr. diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 01545a3fc0f..087a8488843 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -321,6 +321,14 @@ static void do_pmtu_discovery(struct sock *sk, const struct iphdr *iph, u32 mtu) } /* else let the usual retransmit timer handle it */ } +static void do_redirect(struct sk_buff *skb, struct sock *sk) +{ + struct dst_entry *dst = __sk_dst_check(sk, 0); + + if (dst && dst->ops->redirect) + dst->ops->redirect(dst, skb); +} + /* * This routine is called by the ICMP module when it gets some * sort of error condition. If err < 0 then the socket should @@ -394,6 +402,9 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info) } switch (type) { + case ICMP_REDIRECT: + do_redirect(icmp_skb, sk); + goto out; case ICMP_SOURCE_QUENCH: /* Just silently ignore these. */ goto out; diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index ee37d47d472..b4c3582a991 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -630,6 +630,9 @@ void __udp4_lib_err(struct sk_buff *skb, u32 info, struct udp_table *udptable) err = icmp_err_convert[code].errno; } break; + case ICMP_REDIRECT: + ipv4_sk_redirect(skb, sk); + break; } /* diff --git a/net/ipv4/xfrm4_policy.c b/net/ipv4/xfrm4_policy.c index 87d3fcc302d..258ebd7b268 100644 --- a/net/ipv4/xfrm4_policy.c +++ b/net/ipv4/xfrm4_policy.c @@ -202,6 +202,15 @@ static void xfrm4_update_pmtu(struct dst_entry *dst, u32 mtu) path->ops->update_pmtu(path, mtu); } +static void xfrm4_redirect(struct dst_entry *dst, struct sk_buff *skb) +{ + struct xfrm_dst *xdst = (struct xfrm_dst *)dst; + struct dst_entry *path = xdst->route; + + if (path->ops->redirect) + path->ops->redirect(path, skb); +} + static void xfrm4_dst_destroy(struct dst_entry *dst) { struct xfrm_dst *xdst = (struct xfrm_dst *)dst; @@ -225,6 +234,7 @@ static struct dst_ops xfrm4_dst_ops = { .protocol = cpu_to_be16(ETH_P_IP), .gc = xfrm4_garbage_collect, .update_pmtu = xfrm4_update_pmtu, + .redirect = xfrm4_redirect, .cow_metrics = dst_cow_metrics_generic, .destroy = xfrm4_dst_destroy, .ifdown = xfrm4_dst_ifdown, diff --git a/net/sctp/input.c b/net/sctp/input.c index 80564fe0302..9fb4247f9a9 100644 --- a/net/sctp/input.c +++ b/net/sctp/input.c @@ -423,6 +423,18 @@ void sctp_icmp_frag_needed(struct sock *sk, struct sctp_association *asoc, sctp_retransmit(&asoc->outqueue, t, SCTP_RTXR_PMTUD); } +static void sctp_icmp_redirect(struct sock *sk, struct sctp_transport *t, + struct sk_buff *skb) +{ + struct dst_entry *dst; + + if (!t) + return; + dst = sctp_transport_dst_check(t); + if (dst && dst->ops->redirect) + dst->ops->redirect(dst, skb); +} + /* * SCTP Implementer's Guide, 2.37 ICMP handling procedures * @@ -628,6 +640,10 @@ void sctp_v4_err(struct sk_buff *skb, __u32 info) err = EHOSTUNREACH; break; + case ICMP_REDIRECT: + sctp_icmp_redirect(sk, transport, skb); + err = 0; + break; default: goto out_unlock; } -- cgit v1.2.3-70-g09d2