On 7/26/22 10:27, Eric Dumazet wrote:

On Tue, Jul 26, 2022 at 9:06 AM Eric Dumazet [off-list ref] wrote:

quoted

On Tue, Jul 26, 2022 at 8:16 AM Leonard Crestez [off-list ref] wrote:

quoted

Tests are mostly copied from tcp_md5 with minor changes.

It covers VRF support but only based on binding multiple servers: not
multiple keys bound to different interfaces.

Also add a specific -t tcp_authopt to run only these tests specifically.

Thanks for the test.

Could you amend the existing TCP MD5 test to make sure dual sockets
mode is working ?

Apparently, if we have a dual stack listener socket (AF_INET6),
correct incoming IPV4 SYNs are dropped.

quoted

  If this is the case, fixing MD5 should happen first ;

I remember looking into this and my conclusion was that ipv4-mapped-ipv6 
is not worth supporting for AO, at least not in the initial version.

Instead I just wrote a test to check that ipv4-mapped-ipv6 fails for AO:
https://github.com/cdleonard/tcp-authopt-test/blob/main/tcp_authopt_test/test_verify_capture.py#L191

On a closer look it does appear that support existed for 
ipv4-mapped-ipv6 in TCP-MD5 but my test didn't actually exercise it 
correctly so the test had to be fixed.


Do you think it makes sense to add support for ipv4-mapped-ipv6 for AO? 
It's not particularly difficult to test, it was skipped due to a lack of 
application use case and to keep the initial series smaller.

Adding support for this later as a separate commit should be fine. Since 
ivp4-mapped-ipv6 addresses shouldn't appear on the wire giving them 
special treatment "later" should raise no compatibility concerns.

quoted

I think that we are very late in the cycle (linux-5.19 should be
released in 5 days), and your patch set should not be merged so late.

This was posted in order to get code reviews, I'm not actually expecting 
inclusion.

--
Regards,
Leonard