On Thu, Aug 28, 2025 at 02:15:53PM +0200, Florian Westphal wrote:

Pablo Neira Ayuso [off-list ref] wrote:

quoted

On Thu, Aug 28, 2025 at 11:14:35AM +0200, Fabian Bläse wrote:

quoted

The icmp_ndo_send function was originally introduced to ensure proper
rate limiting when icmp_send is called by a network device driver,
where the packet's source address may have already been transformed
by SNAT.

However, the original implementation only considers the
IP_CT_DIR_ORIGINAL direction for SNAT and always replaced the packet's
source address with that of the original-direction tuple. This causes
two problems:

1. For SNAT:
   Reply-direction packets were incorrectly translated using the source
   address of the CT original direction, even though no translation is
   required.

2. For DNAT:
   Reply-direction packets were not handled at all. In DNAT, the original
   direction's destination is translated. Therefore, in the reply
   direction the source address must be set to the reply-direction
   source, so rate limiting works as intended.

Fix this by using the connection direction to select the correct tuple
for source address translation, and adjust the pre-checks to handle
reply-direction packets in case of DNAT.

Additionally, wrap the `ct->status` access in READ_ONCE(). This avoids
possible KCSAN reports about concurrent updates to `ct->status`.

I think such concurrent update cannot not happen, NAT bits are only
set for the first packet of a connection, which sets up the nat
configuration, so READ_ONCE() can go away.

Yes, the NAT bits stay in place but not other flags in ->status, e.g.
DYING, ASSURED, etc.

So I believe its needed, concurrent update of ->status is possible and
KCSAN would warn.  Other spots either use READ_ONCE or use test_bit().

There are a more checks for ct->status & NAT_MASK in the tree that I
can see, if you are correct, then maybe a new helper function to check
for NAT_MASK is needed.

Anyway, as for this patch, READ_ONCE should not harm.