From: Andrew Zaborowski <andrew.zaborowski at intel.com>
To: ell at lists.01.org
Subject: [PATCH 17/17] icmp6: Ignore bad prefixes per RFC4862
Date: Sat, 14 May 2022 00:47:13 +0200 [thread overview]
Message-ID: <20220513224713.1447773-17-andrew.zaborowski@intel.com> (raw)
In-Reply-To: 20220513224713.1447773-1-andrew.zaborowski@intel.com
[-- Attachment #1: Type: text/plain, Size: 2130 bytes --]
"Silently ignore" Prefix Information options with a preferred time
longer than valid time and those matching the link-local prefix as
mandated by RFC4862 Section 5.5.3. These prefixes are also invalid
according to RFC4861.
---
ell/icmp6.c | 32 ++++++++++++++++++++++++++++++--
1 file changed, 30 insertions(+), 2 deletions(-)
diff --git a/ell/icmp6.c b/ell/icmp6.c
index ab2fe9c..96ba1ec 100644
--- a/ell/icmp6.c
+++ b/ell/icmp6.c
@@ -45,6 +45,7 @@
#include "time-private.h"
#include "queue.h"
#include "net.h"
+#include "net-private.h"
#include "netlink.h"
#include "rtnl.h"
#include "missing.h"
@@ -817,7 +818,27 @@ struct l_icmp6_router *_icmp6_router_parse(const struct nd_router_advert *ra,
i->onlink = true;
i->valid_lifetime = l_get_be32(opts + 4);
i->preferred_lifetime = l_get_be32(opts + 8);
- memcpy(i->address, opts + 16, 16);
+
+ /*
+ * Only the initial Prefix Length bits of the prefix
+ * are valid. The remaining bits "MUST" be ignored
+ * by the receiver.
+ */
+ memcpy(i->address, net_prefix_from_ipv6(opts + 16,
+ i->prefix_len), 16);
+
+ /*
+ * For SLAAC (RFC4862) we need to "silently ignore"
+ * routes with a preferred lifetime longer than valid
+ * lifetime, and those with the link-local prefix.
+ * Since it makes sense, do it regardless of SLAAC.
+ */
+ if (i->preferred_lifetime > i->valid_lifetime)
+ break;
+
+ if (i->prefix_len >= 10 &&
+ IN6_IS_ADDR_LINKLOCAL(i->address))
+ break;
n_routes += 1;
break;
@@ -869,7 +890,14 @@ struct l_icmp6_router *_icmp6_router_parse(const struct nd_router_advert *ra,
i->onlink = false;
i->preference = preference;
i->valid_lifetime = l_get_be32(opts + 4);
- memcpy(i->address, opts + 8, (i->prefix_len + 7) / 8);
+
+ /*
+ * Only the initial Prefix Length bits of the prefix
+ * are valid. The remaining bits "MUST" be ignored
+ * by the receiver.
+ */
+ memcpy(i->address, net_prefix_from_ipv6(opts + 8,
+ i->prefix_len), 16);
n_routes += 1;
break;
--
2.32.0
next reply other threads:[~2022-05-13 22:47 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-05-13 22:47 Andrew Zaborowski [this message]
-- strict thread matches above, loose matches on Subject: below --
2022-05-18 19:02 [PATCH 17/17] icmp6: Ignore bad prefixes per RFC4862 Denis Kenzior
2022-05-13 14:55 Andrew Zaborowski
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=20220513224713.1447773-17-andrew.zaborowski@intel.com \
--to=ell@lists.linux.dev \
/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 read-only IMAP folder(s) and NNTP newsgroup(s).