From: Xin Long <lucien.xin@gmail.com>
To: Jakub Kicinski <kuba@kernel.org>
Cc: network dev <netdev@vger.kernel.org>, davem <davem@davemloft.net>,
Dan Carpenter <dan.carpenter@oracle.com>,
Andreas Roeseler <andreas.a.roeseler@gmail.com>
Subject: Re: [PATCH net] icmp: fix icmp_ext_echo_iio parsing in icmp_build_probe
Date: Thu, 14 Oct 2021 11:39:16 +0800 [thread overview]
Message-ID: <CADvbK_eTAEE7dKK7gtxUfuJONHULF5GW+bZmAXM7V06LxKOB_w@mail.gmail.com> (raw)
In-Reply-To: <20211013093118.691255c7@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com>
On Thu, Oct 14, 2021 at 12:31 AM Jakub Kicinski <kuba@kernel.org> wrote:
>
> On Tue, 12 Oct 2021 04:43:07 -0400 Xin Long wrote:
> > In icmp_build_probe(), the icmp_ext_echo_iio parsing should be done
> > step by step and skb_header_pointer() return value should always be
> > checked, this patch fixes 3 places in there:
> >
> > - On case ICMP_EXT_ECHO_CTYPE_NAME, it should only copy ident.name
> > from skb by skb_header_pointer(), its len is ident_len. Besides,
> > the return value of skb_header_pointer() should always be checked.
> >
> > - On case ICMP_EXT_ECHO_CTYPE_INDEX, move ident_len check ahead of
> > skb_header_pointer(), and also do the return value check for
> > skb_header_pointer().
> >
> > - On case ICMP_EXT_ECHO_CTYPE_ADDR, before accessing iio->ident.addr.
> > ctype3_hdr.addrlen, skb_header_pointer() should be called first,
> > then check its return value and ident_len.
> > On subcases ICMP_AFI_IP and ICMP_AFI_IP6, also do check for ident.
> > addr.ctype3_hdr.addrlen and skb_header_pointer()'s return value.
> > On subcase ICMP_AFI_IP, the len for skb_header_pointer() should be
> > "sizeof(iio->extobj_hdr) + sizeof(iio->ident.addr.ctype3_hdr) +
> > sizeof(struct in_addr)" or "ident_len".
> >
> > Fixes: d329ea5bd884 ("icmp: add response to RFC 8335 PROBE messages")
> > Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
> > Signed-off-by: Xin Long <lucien.xin@gmail.com>
>
> > diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
> > index 8b30cadff708..818c79266c48 100644
> > --- a/net/ipv4/icmp.c
> > +++ b/net/ipv4/icmp.c
> > @@ -1061,38 +1061,48 @@ bool icmp_build_probe(struct sk_buff *skb, struct icmphdr *icmphdr)
> > dev = NULL;
> > switch (iio->extobj_hdr.class_type) {
> > case ICMP_EXT_ECHO_CTYPE_NAME:
> > - iio = skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(_iio), &_iio);
> > if (ident_len >= IFNAMSIZ)
> > goto send_mal_query;
> > + iio = skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(iio->extobj_hdr) +
> > + ident_len, &_iio);
> > + if (!iio)
> > + goto send_mal_query;
> > memset(buff, 0, sizeof(buff));
> > memcpy(buff, &iio->ident.name, ident_len);
> > dev = dev_get_by_name(net, buff);
> > break;
> > case ICMP_EXT_ECHO_CTYPE_INDEX:
> > - iio = skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(iio->extobj_hdr) +
> > - sizeof(iio->ident.ifindex), &_iio);
> > if (ident_len != sizeof(iio->ident.ifindex))
> > goto send_mal_query;
> > + iio = skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(iio->extobj_hdr) +
> > + ident_len, &_iio);
> > + if (!iio)
> > + goto send_mal_query;
> > dev = dev_get_by_index(net, ntohl(iio->ident.ifindex));
> > break;
> > case ICMP_EXT_ECHO_CTYPE_ADDR:
> > - if (ident_len != sizeof(iio->ident.addr.ctype3_hdr) +
> > - iio->ident.addr.ctype3_hdr.addrlen)
> > + iio = skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(iio->extobj_hdr) +
> > + sizeof(iio->ident.addr.ctype3_hdr), &_iio);
> > + if (!iio || ident_len != sizeof(iio->ident.addr.ctype3_hdr) +
> > + iio->ident.addr.ctype3_hdr.addrlen)
> > goto send_mal_query;
> > switch (ntohs(iio->ident.addr.ctype3_hdr.afi)) {
> > case ICMP_AFI_IP:
> > + if (iio->ident.addr.ctype3_hdr.addrlen != sizeof(struct in_addr))
> > + goto send_mal_query;
> > iio = skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(iio->extobj_hdr) +
> > - sizeof(struct in_addr), &_iio);
> > - if (ident_len != sizeof(iio->ident.addr.ctype3_hdr) +
> > - sizeof(struct in_addr))
> > + ident_len, &_iio);
> > + if (!iio)
> > goto send_mal_query;
> > dev = ip_dev_find(net, iio->ident.addr.ip_addr.ipv4_addr);
> > break;
> > #if IS_ENABLED(CONFIG_IPV6)
> > case ICMP_AFI_IP6:
> > - iio = skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(_iio), &_iio);
> > - if (ident_len != sizeof(iio->ident.addr.ctype3_hdr) +
> > - sizeof(struct in6_addr))
> > + if (iio->ident.addr.ctype3_hdr.addrlen != sizeof(struct in6_addr))
> > + goto send_mal_query;
> > + iio = skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(iio->extobj_hdr) +
> > + ident_len, &_iio);
> > + if (!iio)
> > goto send_mal_query;
> > dev = ipv6_stub->ipv6_dev_find(net, &iio->ident.addr.ip_addr.ipv6_addr, dev);
> > dev_hold(dev);
>
> If I'm reading this right we end up with the same skb_header_pointer
> call 4 times, every path ends up calling:
>
> skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(iio->extobj_hdr) + ident_len, &_iio);
>
> and looks like the skb does not get modified in between so these calls
> are equivalent.
This looks much clearer.
I was worrying that ident_len might be too big, and ident_len should be verified
before calling skb_header_pointer(ident_len).
But it should be okay as skb_header_pointer() will return NULL if ident_len is
too big.
Will post v2 with the suggestion, thanks!
>
> So why don't we instead consolidate the paths:
>
> diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
> index 8b30cadff708..efa2ec1a85bf 100644
> --- a/net/ipv4/icmp.c
> +++ b/net/ipv4/icmp.c
> @@ -1057,11 +1057,15 @@ bool icmp_build_probe(struct sk_buff *skb, struct icmphdr *icmphdr)
> if (ntohs(iio->extobj_hdr.length) <= sizeof(iio->extobj_hdr))
> goto send_mal_query;
> ident_len = ntohs(iio->extobj_hdr.length) - sizeof(iio->extobj_hdr);
> + iio = skb_header_pointer(skb, sizeof(_ext_hdr),
> + sizeof(iio->extobj_hdr) + ident_len, &_iio);
> + if (!iio)
> + goto send_mal_query;
> +
> status = 0;
> dev = NULL;
> switch (iio->extobj_hdr.class_type) {
> case ICMP_EXT_ECHO_CTYPE_NAME:
> - iio = skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(_iio), &_iio);
> if (ident_len >= IFNAMSIZ)
> goto send_mal_query;
> memset(buff, 0, sizeof(buff));
> @@ -1069,8 +1073,6 @@ bool icmp_build_probe(struct sk_buff *skb, struct icmphdr *icmphdr)
> dev = dev_get_by_name(net, buff);
> break;
> case ICMP_EXT_ECHO_CTYPE_INDEX:
> - iio = skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(iio->extobj_hdr) +
> - sizeof(iio->ident.ifindex), &_iio);
> if (ident_len != sizeof(iio->ident.ifindex))
> goto send_mal_query;
> dev = dev_get_by_index(net, ntohl(iio->ident.ifindex));
> @@ -1081,18 +1083,13 @@ bool icmp_build_probe(struct sk_buff *skb, struct icmphdr *icmphdr)
> goto send_mal_query;
> switch (ntohs(iio->ident.addr.ctype3_hdr.afi)) {
> case ICMP_AFI_IP:
> - iio = skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(iio->extobj_hdr) +
> - sizeof(struct in_addr), &_iio);
> - if (ident_len != sizeof(iio->ident.addr.ctype3_hdr) +
> - sizeof(struct in_addr))
> + if (iio->ident.addr.ctype3_hdr.addrlen != sizeof(struct in_addr))
> goto send_mal_query;
> dev = ip_dev_find(net, iio->ident.addr.ip_addr.ipv4_addr);
> break;
> #if IS_ENABLED(CONFIG_IPV6)
> case ICMP_AFI_IP6:
> - iio = skb_header_pointer(skb, sizeof(_ext_hdr), sizeof(_iio), &_iio);
> - if (ident_len != sizeof(iio->ident.addr.ctype3_hdr) +
> - sizeof(struct in6_addr))
> + if (iio->ident.addr.ctype3_hdr.addrlen != sizeof(struct in6_addr))
> goto send_mal_query;
> dev = ipv6_stub->ipv6_dev_find(net, &iio->ident.addr.ip_addr.ipv6_addr, dev);
> dev_hold(dev);
> --
> 2.31.1
>
next prev parent reply other threads:[~2021-10-14 3:39 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-10-12 8:43 [PATCH net] icmp: fix icmp_ext_echo_iio parsing in icmp_build_probe Xin Long
2021-10-13 16:31 ` Jakub Kicinski
2021-10-14 3:39 ` Xin Long [this message]
2021-10-14 3:59 ` Xin Long
2021-10-14 4:07 ` Xin Long
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CADvbK_eTAEE7dKK7gtxUfuJONHULF5GW+bZmAXM7V06LxKOB_w@mail.gmail.com \
--to=lucien.xin@gmail.com \
--cc=andreas.a.roeseler@gmail.com \
--cc=dan.carpenter@oracle.com \
--cc=davem@davemloft.net \
--cc=kuba@kernel.org \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).