While testing the blamed commit below, I was able to miss (!)
packetdrill failures in the fastopen test-cases.

On passive fastopen the child socket is created by incoming TCP MPC syn,
allow for both MPC_SYN and MPC_ACK header.

Fixes: 724b00c12957 ("mptcp: refine opt_mp_capable determination")
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
---
v1 -> v2:
 - keep it simple don't bother handling DSS OoO pkts (matttbe)
---
 net/mptcp/subflow.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c
index 1117d1e84274..0dcb721c89d1 100644
--- a/net/mptcp/subflow.c
+++ b/net/mptcp/subflow.c
@@ -783,7 +783,8 @@ static struct sock *subflow_syn_recv_sock(const struct sock *sk,
 		 * options.
 		 */
 		mptcp_get_options(skb, &mp_opt);
-		if (!(mp_opt.suboptions & OPTION_MPTCP_MPC_ACK))
+		if (!(mp_opt.suboptions &
+		      (OPTION_MPTCP_MPC_SYN | OPTION_MPTCP_MPC_ACK)))
 			fallback = true;
 
 	} else if (subflow_req->mp_join) {
-- 
2.43.0

Hi Paolo,

On 15/01/2024 19:31, Paolo Abeni wrote:
> While testing the blamed commit below, I was able to miss (!)
> packetdrill failures in the fastopen test-cases.
> 
> On passive fastopen the child socket is created by incoming TCP MPC syn,
> allow for both MPC_SYN and MPC_ACK header.
> 
> Fixes: 724b00c12957 ("mptcp: refine opt_mp_capable determination")
> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
> ---
> v1 -> v2:
>  - keep it simple don't bother handling DSS OoO pkts (matttbe)

Thank you for the v2!

I just applied it in our tree (fix for -net):

New patches for t/upstream-net and t/upstream:
- 8560e02c1efc: mptcp: relax check on MPC passive fallback
- Results: 8cd75bfd0036..dd080a859960 (export-net)
- Results: 25ad5c88ddc7..2d5218531f82 (export)

Tests are now in progress:

https://cirrus-ci.com/github/multipath-tcp/mptcp_net-next/export-net/20240116T172013
https://cirrus-ci.com/github/multipath-tcp/mptcp_net-next/export/20240116T172013

Cheers,
Matt
-- 
Sponsored by the NGI0 Core fund.

Hi Paolo,

On 15/01/2024 19:31, Paolo Abeni wrote:
> While testing the blamed commit below, I was able to miss (!)
> packetdrill failures in the fastopen test-cases.
> 
> On passive fastopen the child socket is created by incoming TCP MPC syn,
> allow for both MPC_SYN and MPC_ACK header.
> 
> Fixes: 724b00c12957 ("mptcp: refine opt_mp_capable determination")
> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
> ---
> v1 -> v2:
>  - keep it simple don't bother handling DSS OoO pkts (matttbe)

Thank you for the v2, looks good to me (if the CI is also OK with it)

Reviewed-by: Matthieu Baerts <matttbe@kernel.org>

Cheers,
Matt
-- 
Sponsored by the NGI0 Core fund.