[PATCH net] ipv4: restrict IPOPT_SSRR and IPOPT_LSRR options

[PATCH net] ipv4: restrict IPOPT_SSRR and IPOPT_LSRR options

[Eric Dumazet]

This patch restricts setting Loose Source and Record Route (LSRR)
and Strict Source and Record Route (SSRR) IP options to users
with CAP_NET_RAW capability.

This prevents unprivileged applications from forcing packets to route
through attacker-controlled nodes to leak TCP ISN and possibly other
protocol information.

While LSRR and SSRR are commonly filtered in many network environments,
they may still be supported and forwarded along some network paths.

RFC 7126 (Recommendations on Filtering of IPv4 Packets Containing
IPv4 Options) recommend to drop these options in 4.3 and 4.4.

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Reported-by: Tamir Shahar <tamirthesis@gmail.com>
Reported-by: Amit Klein <aksecurity@gmail.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 net/ipv4/ip_options.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/net/ipv4/ip_options.c b/net/ipv4/ip_options.c
index be8815ce3ac242372eeae4a97091cda26d40ceb0..ac0d147c4b8cc347839a044adc43897faffd95c8 100644
--- a/net/ipv4/ip_options.c
+++ b/net/ipv4/ip_options.c
@@ -283,6 +283,10 @@ int __ip_options_compile(struct net *net,
 		switch (*optptr) {
 		case IPOPT_SSRR:
 		case IPOPT_LSRR:
+			if (!skb && !ns_capable(net->user_ns, CAP_NET_RAW)) {
+				pp_ptr = optptr;
+				goto error;
+			}
 			if (optlen < 3) {
 				pp_ptr = optptr + 1;
 				goto error;
-- 
2.54.0.1013.g208068f2d8-goog

Re: [PATCH net] ipv4: restrict IPOPT_SSRR and IPOPT_LSRR options

[Ido Schimmel]

On Mon, Jun 01, 2026 at 12:41:57PM +0000, Eric Dumazet wrote:
> This patch restricts setting Loose Source and Record Route (LSRR)
> and Strict Source and Record Route (SSRR) IP options to users
> with CAP_NET_RAW capability.
> 
> This prevents unprivileged applications from forcing packets to route
> through attacker-controlled nodes to leak TCP ISN and possibly other
> protocol information.
> 
> While LSRR and SSRR are commonly filtered in many network environments,
> they may still be supported and forwarded along some network paths.
> 
> RFC 7126 (Recommendations on Filtering of IPv4 Packets Containing
> IPv4 Options) recommend to drop these options in 4.3 and 4.4.
> 
> Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
> Reported-by: Tamir Shahar <tamirthesis@gmail.com>
> Reported-by: Amit Klein <aksecurity@gmail.com>
> Signed-off-by: Eric Dumazet <edumazet@google.com>
> ---
>  net/ipv4/ip_options.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/net/ipv4/ip_options.c b/net/ipv4/ip_options.c
> index be8815ce3ac242372eeae4a97091cda26d40ceb0..ac0d147c4b8cc347839a044adc43897faffd95c8 100644
> --- a/net/ipv4/ip_options.c
> +++ b/net/ipv4/ip_options.c
> @@ -283,6 +283,10 @@ int __ip_options_compile(struct net *net,
>  		switch (*optptr) {
>  		case IPOPT_SSRR:
>  		case IPOPT_LSRR:
> +			if (!skb && !ns_capable(net->user_ns, CAP_NET_RAW)) {
> +				pp_ptr = optptr;
> +				goto error;
> +			}

This might be a problem for netfilter which calls __ip_options_compile()
without an skb from ipv4_find_option().

AFAICT, user space can include IP options either using the IP_OPTIONS
socket options or using the IP_RETOPTS cmsg, but the latter is not
applicable to TCP.

If we put the check in do_ip_setsockopt(), we create a weird situation
for non-TCP sockets: They can't set these IP options using IP_OPTIONS,
but can set them via the IP_RETOPTS cmsg.

So maybe we should just put the check in ip_options_get() which is
called from both paths? Something like:

diff --git a/net/ipv4/ip_options.c b/net/ipv4/ip_options.c
index be8815ce3ac2..09d745112c15 100644
--- a/net/ipv4/ip_options.c
+++ b/net/ipv4/ip_options.c
@@ -530,6 +530,10 @@ int ip_options_get(struct net *net, struct ip_options_rcu **optp,
                kfree(opt);
                return -EINVAL;
        }
+       if (opt->opt.srr && !ns_capable(net->user_ns, CAP_NET_RAW)) {
+               kfree(opt);
+               return -EPERM;
+       }
        kfree(*optp);
        *optp = opt;
        return 0;