From mboxrd@z Thu Jan 1 00:00:00 1970 From: Xin Long Subject: Re: [PATCH net] sctp: do not check port in sctp_inet6_cmp_addr Date: Thu, 12 Apr 2018 00:16:58 +0800 Message-ID: References: <340aad3be762046ca9d02e54edba5bfefa2f4e71.1523451485.git.lucien.xin@gmail.com> <20180411143607.GA4141@hmswarspite.think-freely.org> <20180411145910.GC3711@localhost.localdomain> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Cc: Neil Horman , network dev , linux-sctp@vger.kernel.org, davem To: Marcelo Ricardo Leitner Return-path: Received: from mail-it0-f67.google.com ([209.85.214.67]:50836 "EHLO mail-it0-f67.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752735AbeDKQQ7 (ORCPT ); Wed, 11 Apr 2018 12:16:59 -0400 In-Reply-To: <20180411145910.GC3711@localhost.localdomain> Sender: netdev-owner@vger.kernel.org List-ID: On Wed, Apr 11, 2018 at 10:59 PM, Marcelo Ricardo Leitner wrote: > On Wed, Apr 11, 2018 at 10:36:07AM -0400, Neil Horman wrote: >> On Wed, Apr 11, 2018 at 08:58:05PM +0800, Xin Long wrote: >> > pf->cmp_addr() is called before binding a v6 address to the sock. It >> > should not check ports, like in sctp_inet_cmp_addr. >> > >> > But sctp_inet6_cmp_addr checks the addr by invoking af(6)->cmp_addr, >> > sctp_v6_cmp_addr where it also compares the ports. >> > >> > This would cause that setsockopt(SCTP_SOCKOPT_BINDX_ADD) could bind >> > multiple duplicated IPv6 addresses after Commit 40b4f0fd74e4 ("sctp: >> > lack the check for ports in sctp_v6_cmp_addr"). >> > >> > This patch is to remove af->cmp_addr called in sctp_inet6_cmp_addr, >> > but do the proper check for both v6 addrs and v4mapped addrs. >> > >> > Fixes: 40b4f0fd74e4 ("sctp: lack the check for ports in sctp_v6_cmp_addr") >> > Reported-by: Jianwen Ji >> > Signed-off-by: Xin Long >> > --- >> > net/sctp/ipv6.c | 27 ++++++++++++++++++++++++--- >> > 1 file changed, 24 insertions(+), 3 deletions(-) >> > >> > diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c >> > index f1fc48e..be4b72c 100644 >> > --- a/net/sctp/ipv6.c >> > +++ b/net/sctp/ipv6.c >> > @@ -846,8 +846,8 @@ static int sctp_inet6_cmp_addr(const union sctp_addr *addr1, >> > const union sctp_addr *addr2, >> > struct sctp_sock *opt) >> > { >> > - struct sctp_af *af1, *af2; >> > struct sock *sk = sctp_opt2sk(opt); >> > + struct sctp_af *af1, *af2; >> > >> > af1 = sctp_get_af_specific(addr1->sa.sa_family); >> > af2 = sctp_get_af_specific(addr2->sa.sa_family); >> > @@ -863,10 +863,31 @@ static int sctp_inet6_cmp_addr(const union sctp_addr *addr1, >> > if (sctp_is_any(sk, addr1) || sctp_is_any(sk, addr2)) >> > return 1; >> > >> > - if (addr1->sa.sa_family != addr2->sa.sa_family) >> > + if (addr1->sa.sa_family != addr2->sa.sa_family) { >> > + if (addr1->sa.sa_family == AF_INET && >> > + addr2->sa.sa_family == AF_INET6 && >> > + ipv6_addr_v4mapped(&addr2->v6.sin6_addr)) >> > + if (addr2->v6.sin6_addr.s6_addr32[3] == >> > + addr1->v4.sin_addr.s_addr) >> > + return 1; >> > + if (addr2->sa.sa_family == AF_INET && >> > + addr1->sa.sa_family == AF_INET6 && >> > + ipv6_addr_v4mapped(&addr1->v6.sin6_addr)) >> > + if (addr1->v6.sin6_addr.s6_addr32[3] == >> > + addr2->v4.sin_addr.s_addr) >> > + return 1; >> > + return 0; >> > + } >> > + >> > + if (!ipv6_addr_equal(&addr1->v6.sin6_addr, &addr2->v6.sin6_addr)) >> > + return 0; >> > + >> > + if ((ipv6_addr_type(&addr1->v6.sin6_addr) & IPV6_ADDR_LINKLOCAL) && >> > + addr1->v6.sin6_scope_id && addr2->v6.sin6_scope_id && >> > + addr1->v6.sin6_scope_id != addr2->v6.sin6_scope_id) >> > return 0; >> > >> > - return af1->cmp_addr(addr1, addr2); >> > + return 1; >> > } >> > >> > /* Verify that the provided sockaddr looks bindable. Common verification, >> > -- >> > 2.1.0 >> > >> This looks correct to me, but is it worth duplicating the comparison code like >> this from the cmp_addr function? It might be more worthwhile to add a flag to >> the cmp_addr method to direct weather it needs to check port values or not. >> That way you could continue to use the cmp_addr function here. > > Adding a flag into sctp_v6_cmp_addr will get us a terrible code to > read. It's already not one of the best looking part of it. Maybe > still duplicate part of it it, but at 'af' level? As in: > - af->cmp_addr > - af->cmp_addr_port > What do you think of: static int sctp_v6_cmp_addr(const union sctp_addr *addr1, const union sctp_addr *addr2) { return __sctp_v6_cmp_addr(addr1, addr2) && addr1->v6.sin_port == addr2->v6.sin_port; } (v6.sin_port and v4.sin_port have the same offset in union sctp_addr, we've exploited this in many places in SCTP) From mboxrd@z Thu Jan 1 00:00:00 1970 From: Xin Long Date: Wed, 11 Apr 2018 16:16:58 +0000 Subject: Re: [PATCH net] sctp: do not check port in sctp_inet6_cmp_addr Message-Id: List-Id: References: <340aad3be762046ca9d02e54edba5bfefa2f4e71.1523451485.git.lucien.xin@gmail.com> <20180411143607.GA4141@hmswarspite.think-freely.org> <20180411145910.GC3711@localhost.localdomain> In-Reply-To: <20180411145910.GC3711@localhost.localdomain> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Marcelo Ricardo Leitner Cc: Neil Horman , network dev , linux-sctp@vger.kernel.org, davem On Wed, Apr 11, 2018 at 10:59 PM, Marcelo Ricardo Leitner wrote: > On Wed, Apr 11, 2018 at 10:36:07AM -0400, Neil Horman wrote: >> On Wed, Apr 11, 2018 at 08:58:05PM +0800, Xin Long wrote: >> > pf->cmp_addr() is called before binding a v6 address to the sock. It >> > should not check ports, like in sctp_inet_cmp_addr. >> > >> > But sctp_inet6_cmp_addr checks the addr by invoking af(6)->cmp_addr, >> > sctp_v6_cmp_addr where it also compares the ports. >> > >> > This would cause that setsockopt(SCTP_SOCKOPT_BINDX_ADD) could bind >> > multiple duplicated IPv6 addresses after Commit 40b4f0fd74e4 ("sctp: >> > lack the check for ports in sctp_v6_cmp_addr"). >> > >> > This patch is to remove af->cmp_addr called in sctp_inet6_cmp_addr, >> > but do the proper check for both v6 addrs and v4mapped addrs. >> > >> > Fixes: 40b4f0fd74e4 ("sctp: lack the check for ports in sctp_v6_cmp_addr") >> > Reported-by: Jianwen Ji >> > Signed-off-by: Xin Long >> > --- >> > net/sctp/ipv6.c | 27 ++++++++++++++++++++++++--- >> > 1 file changed, 24 insertions(+), 3 deletions(-) >> > >> > diff --git a/net/sctp/ipv6.c b/net/sctp/ipv6.c >> > index f1fc48e..be4b72c 100644 >> > --- a/net/sctp/ipv6.c >> > +++ b/net/sctp/ipv6.c >> > @@ -846,8 +846,8 @@ static int sctp_inet6_cmp_addr(const union sctp_addr *addr1, >> > const union sctp_addr *addr2, >> > struct sctp_sock *opt) >> > { >> > - struct sctp_af *af1, *af2; >> > struct sock *sk = sctp_opt2sk(opt); >> > + struct sctp_af *af1, *af2; >> > >> > af1 = sctp_get_af_specific(addr1->sa.sa_family); >> > af2 = sctp_get_af_specific(addr2->sa.sa_family); >> > @@ -863,10 +863,31 @@ static int sctp_inet6_cmp_addr(const union sctp_addr *addr1, >> > if (sctp_is_any(sk, addr1) || sctp_is_any(sk, addr2)) >> > return 1; >> > >> > - if (addr1->sa.sa_family != addr2->sa.sa_family) >> > + if (addr1->sa.sa_family != addr2->sa.sa_family) { >> > + if (addr1->sa.sa_family = AF_INET && >> > + addr2->sa.sa_family = AF_INET6 && >> > + ipv6_addr_v4mapped(&addr2->v6.sin6_addr)) >> > + if (addr2->v6.sin6_addr.s6_addr32[3] = >> > + addr1->v4.sin_addr.s_addr) >> > + return 1; >> > + if (addr2->sa.sa_family = AF_INET && >> > + addr1->sa.sa_family = AF_INET6 && >> > + ipv6_addr_v4mapped(&addr1->v6.sin6_addr)) >> > + if (addr1->v6.sin6_addr.s6_addr32[3] = >> > + addr2->v4.sin_addr.s_addr) >> > + return 1; >> > + return 0; >> > + } >> > + >> > + if (!ipv6_addr_equal(&addr1->v6.sin6_addr, &addr2->v6.sin6_addr)) >> > + return 0; >> > + >> > + if ((ipv6_addr_type(&addr1->v6.sin6_addr) & IPV6_ADDR_LINKLOCAL) && >> > + addr1->v6.sin6_scope_id && addr2->v6.sin6_scope_id && >> > + addr1->v6.sin6_scope_id != addr2->v6.sin6_scope_id) >> > return 0; >> > >> > - return af1->cmp_addr(addr1, addr2); >> > + return 1; >> > } >> > >> > /* Verify that the provided sockaddr looks bindable. Common verification, >> > -- >> > 2.1.0 >> > >> This looks correct to me, but is it worth duplicating the comparison code like >> this from the cmp_addr function? It might be more worthwhile to add a flag to >> the cmp_addr method to direct weather it needs to check port values or not. >> That way you could continue to use the cmp_addr function here. > > Adding a flag into sctp_v6_cmp_addr will get us a terrible code to > read. It's already not one of the best looking part of it. Maybe > still duplicate part of it it, but at 'af' level? As in: > - af->cmp_addr > - af->cmp_addr_port > What do you think of: static int sctp_v6_cmp_addr(const union sctp_addr *addr1, const union sctp_addr *addr2) { return __sctp_v6_cmp_addr(addr1, addr2) && addr1->v6.sin_port = addr2->v6.sin_port; } (v6.sin_port and v4.sin_port have the same offset in union sctp_addr, we've exploited this in many places in SCTP)