From a9888628cb2c768202a4530e2816da1889cc3165 Mon Sep 17 00:00:00 2001 From: Ilya Maximets Date: Tue, 9 Sep 2025 18:54:15 +0200 Subject: net: dst_metadata: fix IP_DF bit not extracted from tunnel headers Both OVS and TC flower allow extracting and matching on the DF bit of the outer IP header via OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT in the OVS_KEY_ATTR_TUNNEL and TCA_FLOWER_KEY_FLAGS_TUNNEL_DONT_FRAGMENT in the TCA_FLOWER_KEY_ENC_FLAGS respectively. Flow dissector extracts this information as FLOW_DIS_F_TUNNEL_DONT_FRAGMENT from the tunnel info key. However, the IP_TUNNEL_DONT_FRAGMENT_BIT in the tunnel key is never actually set, because the tunneling code doesn't actually extract it from the IP header. OAM and CRIT_OPT are extracted by the tunnel implementation code, same code also sets the KEY flag, if present. UDP tunnel core takes care of setting the CSUM flag if the checksum is present in the UDP header, but the DONT_FRAGMENT is not handled at any layer. Fix that by checking the bit and setting the corresponding flag while populating the tunnel info in the IP layer where it belongs. Not using __assign_bit as we don't really need to clear the bit in a just initialized field. It also doesn't seem like using __assign_bit will make the code look better. Clearly, users didn't rely on this functionality for anything very important until now. The reason why this doesn't break OVS logic is that it only matches on what kernel previously parsed out and if kernel consistently reports this bit as zero, OVS will only match on it to be zero, which sort of works. But it is still a bug that the uAPI reports and allows matching on the field that is not actually checked in the packet. And this is causing misleading -df reporting in OVS datapath flows, while the tunnel traffic actually has the bit set in most cases. This may also cause issues if a hardware properly implements support for tunnel flag matching as it will disagree with the implementation in a software path of TC flower. Fixes: 7d5437c709de ("openvswitch: Add tunneling interface.") Fixes: 1d17568e74de ("net/sched: cls_flower: add support for matching tunnel control flags") Signed-off-by: Ilya Maximets Reviewed-by: Ido Schimmel Link: https://patch.msgid.link/20250909165440.229890-2-i.maximets@ovn.org Signed-off-by: Jakub Kicinski --- include/net/dst_metadata.h | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/net/dst_metadata.h b/include/net/dst_metadata.h index 4160731dcb6e..1fc2fb03ce3f 100644 --- a/include/net/dst_metadata.h +++ b/include/net/dst_metadata.h @@ -3,6 +3,7 @@ #define __NET_DST_METADATA_H 1 #include +#include #include #include #include @@ -220,9 +221,15 @@ static inline struct metadata_dst *ip_tun_rx_dst(struct sk_buff *skb, int md_size) { const struct iphdr *iph = ip_hdr(skb); + struct metadata_dst *tun_dst; + + tun_dst = __ip_tun_set_dst(iph->saddr, iph->daddr, iph->tos, iph->ttl, + 0, flags, tunnel_id, md_size); - return __ip_tun_set_dst(iph->saddr, iph->daddr, iph->tos, iph->ttl, - 0, flags, tunnel_id, md_size); + if (tun_dst && (iph->frag_off & htons(IP_DF))) + __set_bit(IP_TUNNEL_DONT_FRAGMENT_BIT, + tun_dst->u.tun_info.key.tun_flags); + return tun_dst; } static inline struct metadata_dst *__ipv6_tun_set_dst(const struct in6_addr *saddr, -- cgit v1.2.3 From 2293c57484ae64c9a3c847c8807db8c26a3a4d41 Mon Sep 17 00:00:00 2001 From: "Matthieu Baerts (NGI0)" Date: Fri, 12 Sep 2025 14:52:21 +0200 Subject: mptcp: pm: nl: announce deny-join-id0 flag During the connection establishment, a peer can tell the other one that it cannot establish new subflows to the initial IP address and port by setting the 'C' flag [1]. Doing so makes sense when the sender is behind a strict NAT, operating behind a legacy Layer 4 load balancer, or using anycast IP address for example. When this 'C' flag is set, the path-managers must then not try to establish new subflows to the other peer's initial IP address and port. The in-kernel PM has access to this info, but the userspace PM didn't. The RFC8684 [1] is strict about that: (...) therefore the receiver MUST NOT try to open any additional subflows toward this address and port. So it is important to tell the userspace about that as it is responsible for the respect of this flag. When a new connection is created and established, the Netlink events now contain the existing but not currently used 'flags' attribute. When MPTCP_PM_EV_FLAG_DENY_JOIN_ID0 is set, it means no other subflows to the initial IP address and port -- info that are also part of the event -- can be established. Link: https://datatracker.ietf.org/doc/html/rfc8684#section-3.1-20.6 [1] Fixes: 702c2f646d42 ("mptcp: netlink: allow userspace-driven subflow establishment") Reported-by: Marek Majkowski Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/532 Reviewed-by: Mat Martineau Signed-off-by: Matthieu Baerts (NGI0) Link: https://patch.msgid.link/20250912-net-mptcp-pm-uspace-deny_join_id0-v1-2-40171884ade8@kernel.org Signed-off-by: Jakub Kicinski --- include/uapi/linux/mptcp.h | 2 ++ include/uapi/linux/mptcp_pm.h | 4 ++-- 2 files changed, 4 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/uapi/linux/mptcp.h b/include/uapi/linux/mptcp.h index 67d015df8893..5fd5b4cf75ca 100644 --- a/include/uapi/linux/mptcp.h +++ b/include/uapi/linux/mptcp.h @@ -31,6 +31,8 @@ #define MPTCP_INFO_FLAG_FALLBACK _BITUL(0) #define MPTCP_INFO_FLAG_REMOTE_KEY_RECEIVED _BITUL(1) +#define MPTCP_PM_EV_FLAG_DENY_JOIN_ID0 _BITUL(0) + #define MPTCP_PM_ADDR_FLAG_SIGNAL (1 << 0) #define MPTCP_PM_ADDR_FLAG_SUBFLOW (1 << 1) #define MPTCP_PM_ADDR_FLAG_BACKUP (1 << 2) diff --git a/include/uapi/linux/mptcp_pm.h b/include/uapi/linux/mptcp_pm.h index 6ac84b2f636c..7359d34da446 100644 --- a/include/uapi/linux/mptcp_pm.h +++ b/include/uapi/linux/mptcp_pm.h @@ -16,10 +16,10 @@ * good time to allocate memory and send ADD_ADDR if needed. Depending on the * traffic-patterns it can take a long time until the MPTCP_EVENT_ESTABLISHED * is sent. Attributes: token, family, saddr4 | saddr6, daddr4 | daddr6, - * sport, dport, server-side. + * sport, dport, server-side, [flags]. * @MPTCP_EVENT_ESTABLISHED: A MPTCP connection is established (can start new * subflows). Attributes: token, family, saddr4 | saddr6, daddr4 | daddr6, - * sport, dport, server-side. + * sport, dport, server-side, [flags]. * @MPTCP_EVENT_CLOSED: A MPTCP connection has stopped. Attribute: token. * @MPTCP_EVENT_ANNOUNCED: A new address has been announced by the peer. * Attributes: token, rem_id, family, daddr4 | daddr6 [, dport]. -- cgit v1.2.3 From 6b4be64fd9fec16418f365c2d8e47a7566e9eba5 Mon Sep 17 00:00:00 2001 From: Jianbo Liu Date: Mon, 15 Sep 2025 15:24:32 +0300 Subject: net/mlx5e: Harden uplink netdev access against device unbind The function mlx5_uplink_netdev_get() gets the uplink netdevice pointer from mdev->mlx5e_res.uplink_netdev. However, the netdevice can be removed and its pointer cleared when unbound from the mlx5_core.eth driver. This results in a NULL pointer, causing a kernel panic. BUG: unable to handle page fault for address: 0000000000001300 at RIP: 0010:mlx5e_vport_rep_load+0x22a/0x270 [mlx5_core] Call Trace: mlx5_esw_offloads_rep_load+0x68/0xe0 [mlx5_core] esw_offloads_enable+0x593/0x910 [mlx5_core] mlx5_eswitch_enable_locked+0x341/0x420 [mlx5_core] mlx5_devlink_eswitch_mode_set+0x17e/0x3a0 [mlx5_core] devlink_nl_eswitch_set_doit+0x60/0xd0 genl_family_rcv_msg_doit+0xe0/0x130 genl_rcv_msg+0x183/0x290 netlink_rcv_skb+0x4b/0xf0 genl_rcv+0x24/0x40 netlink_unicast+0x255/0x380 netlink_sendmsg+0x1f3/0x420 __sock_sendmsg+0x38/0x60 __sys_sendto+0x119/0x180 do_syscall_64+0x53/0x1d0 entry_SYSCALL_64_after_hwframe+0x4b/0x53 Ensure the pointer is valid before use by checking it for NULL. If it is valid, immediately call netdev_hold() to take a reference, and preventing the netdevice from being freed while it is in use. Fixes: 7a9fb35e8c3a ("net/mlx5e: Do not reload ethernet ports when changing eswitch mode") Signed-off-by: Jianbo Liu Reviewed-by: Cosmin Ratiu Reviewed-by: Jiri Pirko Reviewed-by: Dragos Tatulea Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/1757939074-617281-2-git-send-email-tariqt@nvidia.com Signed-off-by: Jakub Kicinski --- include/linux/mlx5/driver.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h index 8c5fbfb85749..10fe492e1fed 100644 --- a/include/linux/mlx5/driver.h +++ b/include/linux/mlx5/driver.h @@ -663,6 +663,7 @@ struct mlx5e_resources { bool tisn_valid; } hw_objs; struct net_device *uplink_netdev; + netdevice_tracker tracker; struct mutex uplink_netdev_lock; struct mlx5_crypto_dek_priv *dek_priv; }; -- cgit v1.2.3 From 87ebb628a5acb892eba41ef1d8989beb8f036034 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 17 Sep 2025 13:53:37 +0000 Subject: net: clear sk->sk_ino in sk_set_socket(sk, NULL) Andrei Vagin reported that blamed commit broke CRIU. Indeed, while we want to keep sk_uid unchanged when a socket is cloned, we want to clear sk->sk_ino. Otherwise, sock_diag might report multiple sockets sharing the same inode number. Move the clearing part from sock_orphan() to sk_set_socket(sk, NULL), called both from sock_orphan() and sk_clone_lock(). Fixes: 5d6b58c932ec ("net: lockless sock_i_ino()") Closes: https://lore.kernel.org/netdev/aMhX-VnXkYDpKd9V@google.com/ Closes: https://github.com/checkpoint-restore/criu/issues/2744 Reported-by: Andrei Vagin Signed-off-by: Eric Dumazet Acked-by: Andrei Vagin Link: https://patch.msgid.link/20250917135337.1736101-1-edumazet@google.com Signed-off-by: Jakub Kicinski --- include/net/sock.h | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/include/net/sock.h b/include/net/sock.h index fb13322a11fc..2e14283c5be1 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -2061,6 +2061,9 @@ static inline void sk_set_socket(struct sock *sk, struct socket *sock) if (sock) { WRITE_ONCE(sk->sk_uid, SOCK_INODE(sock)->i_uid); WRITE_ONCE(sk->sk_ino, SOCK_INODE(sock)->i_ino); + } else { + /* Note: sk_uid is unchanged. */ + WRITE_ONCE(sk->sk_ino, 0); } } @@ -2082,8 +2085,6 @@ static inline void sock_orphan(struct sock *sk) sock_set_flag(sk, SOCK_DEAD); sk_set_socket(sk, NULL); sk->sk_wq = NULL; - /* Note: sk_uid is unchanged. */ - WRITE_ONCE(sk->sk_ino, 0); write_unlock_bh(&sk->sk_callback_lock); } -- cgit v1.2.3