Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754523AbdHIQ5E (ORCPT ); Wed, 9 Aug 2017 12:57:04 -0400 Received: from mail.linuxfoundation.org ([140.211.169.12]:39800 "EHLO mail.linuxfoundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754350AbdHIQzS (ORCPT ); Wed, 9 Aug 2017 12:55:18 -0400 From: Greg Kroah-Hartman To: linux-kernel@vger.kernel.org Cc: Greg Kroah-Hartman , stable@vger.kernel.org, Sam Edwards , Marc Haber , Paolo Abeni , Eric Dumazet , "David S. Miller" Subject: [PATCH 4.12 098/106] udp6: fix socket leak on early demux Date: Wed, 9 Aug 2017 09:53:22 -0700 Message-Id: <20170809164530.934276850@linuxfoundation.org> X-Mailer: git-send-email 2.14.0 In-Reply-To: <20170809164515.714288642@linuxfoundation.org> References: <20170809164515.714288642@linuxfoundation.org> User-Agent: quilt/0.65 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3821 Lines: 127 4.12-stable review patch. If anyone has any objections, please let me know. ------------------ From: Paolo Abeni [ Upstream commit c9f2c1ae123a751d4e4f949144500219354d5ee1 ] When an early demuxed packet reaches __udp6_lib_lookup_skb(), the sk reference is retrieved and used, but the relevant reference count is leaked and the socket destructor is never called. Beyond leaking the sk memory, if there are pending UDP packets in the receive queue, even the related accounted memory is leaked. In the long run, this will cause persistent forward allocation errors and no UDP skbs (both ipv4 and ipv6) will be able to reach the user-space. Fix this by explicitly accessing the early demux reference before the lookup, and properly decreasing the socket reference count after usage. Also drop the skb_steal_sock() in __udp6_lib_lookup_skb(), and the now obsoleted comment about "socket cache". The newly added code is derived from the current ipv4 code for the similar path. v1 -> v2: fixed the __udp6_lib_rcv() return code for resubmission, as suggested by Eric Reported-by: Sam Edwards Reported-by: Marc Haber Fixes: 5425077d73e0 ("net: ipv6: Add early demux handler for UDP unicast") Signed-off-by: Paolo Abeni Acked-by: Eric Dumazet Signed-off-by: David S. Miller Signed-off-by: Greg Kroah-Hartman --- include/net/udp.h | 1 + net/ipv4/udp.c | 3 ++- net/ipv6/udp.c | 27 ++++++++++++++++++--------- 3 files changed, 21 insertions(+), 10 deletions(-) --- a/include/net/udp.h +++ b/include/net/udp.h @@ -265,6 +265,7 @@ static inline struct sk_buff *skb_recv_u } void udp_v4_early_demux(struct sk_buff *skb); +void udp_sk_rx_dst_set(struct sock *sk, struct dst_entry *dst); int udp_get_port(struct sock *sk, unsigned short snum, int (*saddr_cmp)(const struct sock *, const struct sock *)); --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -1762,7 +1762,7 @@ drop: /* For TCP sockets, sk_rx_dst is protected by socket lock * For UDP, we use xchg() to guard against concurrent changes. */ -static void udp_sk_rx_dst_set(struct sock *sk, struct dst_entry *dst) +void udp_sk_rx_dst_set(struct sock *sk, struct dst_entry *dst) { struct dst_entry *old; @@ -2120,6 +2120,7 @@ void udp_destroy_sock(struct sock *sk) encap_destroy(sk); } } +EXPORT_SYMBOL(udp_sk_rx_dst_set); /* * Socket option code for UDP --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -291,11 +291,7 @@ static struct sock *__udp6_lib_lookup_sk struct udp_table *udptable) { const struct ipv6hdr *iph = ipv6_hdr(skb); - struct sock *sk; - sk = skb_steal_sock(skb); - if (unlikely(sk)) - return sk; return __udp6_lib_lookup(dev_net(skb->dev), &iph->saddr, sport, &iph->daddr, dport, inet6_iif(skb), udptable, skb); @@ -798,6 +794,24 @@ int __udp6_lib_rcv(struct sk_buff *skb, if (udp6_csum_init(skb, uh, proto)) goto csum_error; + /* Check if the socket is already available, e.g. due to early demux */ + sk = skb_steal_sock(skb); + if (sk) { + struct dst_entry *dst = skb_dst(skb); + int ret; + + if (unlikely(sk->sk_rx_dst != dst)) + udp_sk_rx_dst_set(sk, dst); + + ret = udpv6_queue_rcv_skb(sk, skb); + sock_put(sk); + + /* a return value > 0 means to resubmit the input */ + if (ret > 0) + return ret; + return 0; + } + /* * Multicast receive code */ @@ -806,11 +820,6 @@ int __udp6_lib_rcv(struct sk_buff *skb, saddr, daddr, udptable, proto); /* Unicast */ - - /* - * check socket cache ... must talk to Alan about his plans - * for sock caches... i'll skip this for now. - */ sk = __udp6_lib_lookup_skb(skb, uh->source, uh->dest, udptable); if (sk) { int ret;