The patch below does not apply to the 5.15-stable tree. If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to stable@vger.kernel.org.
To reproduce the conflict and resubmit, you may use the following commands:
git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-5.15.y git checkout FETCH_HEAD git cherry-pick -x 09355f7abb9fbfc1a240be029837921ea417bf4f # <resolve conflicts, build, test, etc.> git commit -s git send-email --to 'stable@vger.kernel.org' --in-reply-to '2024082641-cauterize-slum-9eb2@gregkh' --subject-prefix 'PATCH 5.15.y' HEAD^..
Possible dependencies:
09355f7abb9f ("mptcp: pm: fullmesh: select the right ID later") b9d69db87fb7 ("mptcp: let the in-kernel PM use mixed IPv4 and IPv6 addresses") 837cf45df163 ("mptcp: fix race in incoming ADD_ADDR option processing") a88c9e496937 ("mptcp: do not block subflows creation on errors") 86e39e04482b ("mptcp: keep track of local endpoint still available for each msk") f7d6a237d742 ("mptcp: fix per socket endpoint accounting") b29fcfb54cd7 ("mptcp: full disconnect implementation") 59060a47ca50 ("mptcp: clean up harmless false expressions") 3ce0852c86b9 ("mptcp: enforce HoL-blocking estimation") 6511882cdd82 ("mptcp: allocate fwd memory separately on the rx and tx path") 765ff425528f ("mptcp: use lockdep_assert_held_once() instead of open-coding it")
thanks,
greg k-h
------------------ original commit in Linus's tree ------------------
From 09355f7abb9fbfc1a240be029837921ea417bf4f Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" matttbe@kernel.org Date: Mon, 19 Aug 2024 21:45:30 +0200 Subject: [PATCH] mptcp: pm: fullmesh: select the right ID later
When reacting upon the reception of an ADD_ADDR, the in-kernel PM first looks for fullmesh endpoints. If there are some, it will pick them, using their entry ID.
It should set the ID 0 when using the endpoint corresponding to the initial subflow, it is a special case imposed by the MPTCP specs.
Note that msk->mpc_endpoint_id might not be set when receiving the first ADD_ADDR from the server. So better to compare the addresses.
Fixes: 1a0d6136c5f0 ("mptcp: local addresses fullmesh") Cc: stable@vger.kernel.org Reviewed-by: Mat Martineau martineau@kernel.org Signed-off-by: Matthieu Baerts (NGI0) matttbe@kernel.org Link: https://patch.msgid.link/20240819-net-mptcp-pm-reusing-id-v1-12-38035d40de5b... Signed-off-by: Jakub Kicinski kuba@kernel.org
diff --git a/net/mptcp/pm_netlink.c b/net/mptcp/pm_netlink.c index d0a80f537fc3..a2e37ab1c40f 100644 --- a/net/mptcp/pm_netlink.c +++ b/net/mptcp/pm_netlink.c @@ -636,6 +636,7 @@ static unsigned int fill_local_addresses_vec(struct mptcp_sock *msk, { struct sock *sk = (struct sock *)msk; struct mptcp_pm_addr_entry *entry; + struct mptcp_addr_info mpc_addr; struct pm_nl_pernet *pernet; unsigned int subflows_max; int i = 0; @@ -643,6 +644,8 @@ static unsigned int fill_local_addresses_vec(struct mptcp_sock *msk, pernet = pm_nl_get_pernet_from_msk(msk); subflows_max = mptcp_pm_get_subflows_max(msk);
+ mptcp_local_address((struct sock_common *)msk, &mpc_addr); + rcu_read_lock(); list_for_each_entry_rcu(entry, &pernet->local_addr_list, list) { if (!(entry->flags & MPTCP_PM_ADDR_FLAG_FULLMESH)) @@ -653,7 +656,13 @@ static unsigned int fill_local_addresses_vec(struct mptcp_sock *msk,
if (msk->pm.subflows < subflows_max) { msk->pm.subflows++; - addrs[i++] = entry->addr; + addrs[i] = entry->addr; + + /* Special case for ID0: set the correct ID */ + if (mptcp_addresses_equal(&entry->addr, &mpc_addr, entry->addr.port)) + addrs[i].id = 0; + + i++; } } rcu_read_unlock();