pull request: fixes for ovpn 2026-07-20
Subsystems (5)
Antonio Quartulli
maintainercover pull request: fixes for ovpn 2026-07-20
Hi all!
This is a resend of the series of small fixes I sent on 2026-06-08, now rebased on top of the latest net/main.
Changes compared to the previous submission are the addition of the committer Signed-off-by tags that were missing on two patches (as flagged by the SoB checker) and the amending of the commit message about the new clock function being used for keepalive tracking. No code was changed.
There are larger fixes in our queue which we are still working on, therefore please ignore any "previous issue" Sashiko may report.
Please pull or let me know of any issue!
Thanks a lot,
Antonio
The following changes since commit e13caf1c26587434f0b768193100440939c0fb91:
Merge tag 'net-7.2-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net (2026-07-17 10:25:13 -0700)
are available in the Git repository at:
https://github.com/OpenVPN/ovpn-net-next.git ovpn-net-20260720
for you to fetch changes up to f7e6287ccd3abeed9e638b581dc3fdf742106ba3:
ovpn: use monotonic clock for peer keepalive timeouts (2026-07-20 16:29:31 +0200)
----------------------------------------------------------------
Included fixes:
* ensure keepalive timestamps are computed using monotonic source
* avoid UAF in unlock_ovpn() when iterating over release_list
* fix memleak in selftest tool
* ensure reference to peer is acquired before scheduling worker (which may drop the not-yet-taken ref)
* fix refcount leak in case of concurrent TX and RX TCP error
* fix potential refcount unbalance in case of sock release in P2P mode
----------------------------------------------------------------
Marco Baffo (2):
ovpn: fix use after free in unlock_ovpn()
ovpn: use monotonic clock for peer keepalive timeouts
Pavitra Jha (1):
ovpn: fix peer refcount leak in TCP error paths
Qing Ming (1):
ovpn: avoid putting unrelated P2P peer on socket release
Shuvam Pandey (1):
ovpn: hold peer before scheduling keepalive work
longlong yan (1):
selftests/net: ovpn: fix getaddrinfo memory leak in ovpn_parse_remote()
drivers/net/ovpn/io.c | 4 ++-- drivers/net/ovpn/peer.c | 16 +++++++++------- drivers/net/ovpn/tcp.c | 6 ++++-- tools/testing/selftests/net/ovpn/ovpn-cli.c | 4 +++- 4 files changed, 18 insertions(+), 12 deletions(-)
Antonio Quartulli
maintainerPATCH 1/6 ovpn: avoid putting unrelated P2P peer on socket release
From: Qing Ming <[email protected]>
ovpn_peer_release_p2p() is called when an OVPN UDP socket is being destroyed. It checks the currently published P2P peer and releases it only if that peer still uses the socket being destroyed.
A peer replacement can publish a new peer before the old UDP socket is destroyed. When the old socket destruction path runs afterwards, ovpn_peer_release_p2p() observes the new peer through ovpn->peer. Since the new peer uses a different socket, the function takes the socket mismatch branch.
That branch still calls ovpn_peer_put(peer). At this point, however, peer is the currently published replacement peer, not the peer associated with the socket being destroyed. Dropping its reference can free it while ovpn->peer still points to it, leading to later use-after-free accesses from the peer and socket cleanup paths.
KASAN reports this as a slab-use-after-free on the kmalloc-1k ovpn_peer object. In the reproducer, the object is allocated from ovpn_peer_new() via ovpn_nl_peer_new_doit(), and freed through ovpn_peer_release_rcu() from RCU callback processing. Observed access sites include ovpn_peer_remove(), ovpn_socket_release(), ovpn_nl_peer_del_notify(), and unlock_ovpn().
Fix this by returning from the socket mismatch branch without putting the peer.
--- drivers/net/ovpn/peer.c | 1 - 1 file changed, 1 deletion(-)
drivers/net/ovpn/peer.c +0 -1
@@ -1167,7 +1167,6 @@ static void ovpn_peer_release_p2p(struct ovpn_priv *ovpn, struct sock *sk, ovpn_sock =rcu_access_pointer (peer->sock);if (!ovpn_sock || ovpn_sock->sk != sk) {spin_unlock_bh (&ovpn->lock); -ovpn_peer_put (peer);return ; } }
Signature · 1 line
2.54.0
Re: [PATCH net 1/6] ovpn: avoid putting unrelated P2P peer on socket release
This series was applied to netdev/net.git (main) by Antonio Quartulli <[email protected]>:
Hello:
This series was applied to netdev/net.git (main)
by Antonio Quartulli <[email protected]>:
Quoted 6 lines · On Mon, 20 Jul 2026 16:41:26 +0200 you wrote:
From: Qing Ming <[email protected]>
ovpn_peer_release_p2p()is called when an OVPN UDP socket is being destroyed. It checks the currently published P2P peer and releases it only if that peer still uses the socket being destroyed.
Quoted 7 lines
A peer replacement can publish a new peer before the old UDP socket is destroyed. When the old socket destruction path runs afterwards,
ovpn_peer_release_p2p()observes the new peer throughovpn->peer. Since the new peer uses a different socket, the function takes the socket mismatch branch.[...]
Here is the summary with links:
- [net,1/6] ovpn: avoid putting unrelated P2P peer on socket release
https://git.kernel.org/netdev/net/c/b52c5103f64e - [net,2/6] ovpn: fix peer refcount leak in TCP error paths https://git.kernel.org/netdev/net/c/63bbe18fc030 - [net,3/6] ovpn: hold peer before scheduling keepalive work https://git.kernel.org/netdev/net/c/a4710ae2e7e3 - [net,4/6] selftests/net: ovpn: fix getaddrinfo memory leak in ovpn_parse_remote() https://git.kernel.org/netdev/net/c/0bd9cfebc1c9 - [net,5/6] ovpn: fix use after free in unlock_ovpn() https://git.kernel.org/netdev/net/c/e1ad6fe5db71 - [net,6/6] ovpn: use monotonic clock for peer keepalive timeouts https://git.kernel.org/netdev/net/c/f7e6287ccd3a
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html
Antonio Quartulli
maintainerPATCH 2/6 ovpn: fix peer refcount leak in TCP error paths
In reply to Antonio QuartulliFrom: Pavitra Jha <[email protected]>
When either the TCP RX or TX error path calls ovpn_peer_hold() followed by schedule_work(&peer->tcp.defer_del_work), and the work item is already pending from the other path, schedule_work() returns false and the work runs only once. Since ovpn_tcp_peer_del_work() calls ovpn_peer_put() exactly once, the extra reference taken by the losing path is never dropped, leaking the peer object.
The race window:
CPU0 (strparser/RX error): CPU1 (tcp_tx_work/TX error):
ovpn_peer_hold() <- refcnt+1 ovpn_peer_hold() <- refcnt+2
schedule_work() <- queued schedule_work() <- NO-OP
(work already pending)
ovpn_tcp_peer_del_work runs:
ovpn_peer_del()
ovpn_peer_put() <- refcnt+1
<- peer never freedFix by checking the return value of schedule_work() in both paths and calling ovpn_peer_put() to drop the extra reference if the work was already pending. ovpn_peer_hold() is kept unconditional in the TX path as it cannot fail at that point.
--- drivers/net/ovpn/tcp.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/net/ovpn/tcp.c +4 -2
@@ -151,7 +151,8 @@ static void ovpn_tcp_rcv(struct strparser *strp, struct sk_buff *skb)/* take reference for deferred peer deletion. should never fail */ if (WARN_ON (!ovpn_peer_hold (peer))) goto err_nopeer; -schedule_work (&peer->tcp.defer_del_work); +if (!schedule_work (&peer->tcp.defer_del_work)) +ovpn_peer_put (peer);ovpn_dev_dstats_rx_dropped (peer->ovpn->dev); err_nopeer:kfree_skb (skb);
@@ -283,7 +284,8 @@ static void ovpn_tcp_send_sock(struct ovpn_peer *peer, struct sock *sk)* stream therefore we abort the connection */ ovpn_peer_hold (peer); -schedule_work (&peer->tcp.defer_del_work); +if (!schedule_work (&peer->tcp.defer_del_work)) +ovpn_peer_put (peer);/* we bail out immediately and keep tx_in_progress set * to true. This way we prevent more TX attempts
Signature · 1 line
2.54.0
Antonio Quartulli
maintainerPATCH 3/6 ovpn: hold peer before scheduling keepalive work
In reply to Antonio QuartulliFrom: Shuvam Pandey <[email protected]>
ovpn_peer_keepalive_send() passes its peer reference to ovpn_xmit_special(), which ultimately drops it. The keepalive scheduler currently queues the work first and takes the reference only after schedule_work() reports that the work was queued.
Once schedule_work() queues the item, another CPU may run the worker before the caller gets to ovpn_peer_hold(). In that case the worker can consume a reference that was not acquired for it, corrupting the peer lifetime accounting.
Take the peer reference before queueing the work and drop it again when the work was already pending.
--- drivers/net/ovpn/peer.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/net/ovpn/peer.c +4 -2
@@ -1284,8 +1284,10 @@ static time64_t ovpn_peer_keepalive_work_single(struct ovpn_peer *peer,netdev_dbg (peer->ovpn->dev,"sending keepalive to peer %u\n" , peer->id); -if (schedule_work (&peer->keepalive_work)) -ovpn_peer_hold (peer); +if (WARN_ON (!ovpn_peer_hold (peer))) +return 0 ; +if (!schedule_work (&peer->keepalive_work)) +ovpn_peer_put (peer); }if (next_run1 < next_run2)
Signature · 1 line
2.54.0
Antonio Quartulli
maintainerPATCH 4/6 selftests/net: ovpn: fix getaddrinfo memory leak in ovpn_parse_remote()
In reply to Antonio QuartulliFrom: longlong yan <[email protected]>
The ovpn_parse_remote() function has two memory management issues:
1. When both 'host' and 'vpnip' are non-NULL, the first getaddrinfo() allocation is leaked because 'result' is overwritten by the second getaddrinfo() call without freeing the first allocation.
2. When both 'host' and 'vpnip' are NULL, 'result' is an uninitialized stack variable passed to freeaddrinfo(), which is undefined behavior.
Fix by initializing 'result' to NULL and calling freeaddrinfo() after the first getaddrinfo() result is consumed.
--- tools/testing/selftests/net/ovpn/ovpn-cli.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
tools/testing/selftests/net/ovpn/ovpn-cli.c +3 -1
@@ -1785,7 +1785,7 @@ static int ovpn_parse_remote(struct ovpn_ctx *ovpn, const char *host, const char *service, const char *vpnip) { int ret; -struct addrinfo *result; +struct addrinfo *result =NULL ; struct addrinfo hints = { .ai_family = ovpn->sa_family, .ai_socktype = SOCK_DGRAM,
@@ -1809,6 +1809,8 @@ static int ovpn_parse_remote(struct ovpn_ctx *ovpn, const char *host, }memcpy (&ovpn->remote, result->ai_addr, result->ai_addrlen); +freeaddrinfo (result); + result =NULL ; }if (vpnip) {
Signature · 1 line
2.54.0
Antonio Quartulli
maintainerPATCH 5/6 ovpn: fix use after free in unlock_ovpn()
In reply to Antonio QuartulliFrom: Marco Baffo <[email protected]>
unlock_ovpn() iterates over the release_list using llist_for_each_entry() and drops the peer reference inside the loop body via ovpn_peer_put().
If this drops the last reference, the peer is eventually freed. However, llist_for_each_entry() reads peer->release_entry.next in the loop advance expression, which runs after the body. By that time the peer may have already been freed, resulting in a use after free when advancing to the next list entry.
Fix this by using llist_for_each_entry_safe(), which caches the next pointer before executing the loop body.
--- drivers/net/ovpn/peer.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/net/ovpn/peer.c +3 -2
@@ -26,11 +26,12 @@ static void unlock_ovpn(struct ovpn_priv *ovpn, struct llist_head *release_list) __releases(&ovpn->lock) { -struct ovpn_peer *peer; +struct ovpn_peer *peer, *next;spin_unlock_bh (&ovpn->lock); -llist_for_each_entry (peer, release_list->first, release_entry) { +llist_for_each_entry_safe (peer, next, release_list->first, + release_entry) {ovpn_socket_release (peer);ovpn_peer_put (peer); }
Signature · 1 line
2.54.0
Antonio Quartulli
maintainerPATCH 6/6 ovpn: use monotonic clock for peer keepalive timeouts
In reply to Antonio QuartulliFrom: Marco Baffo <[email protected]>
Replace ktime_get_real_seconds() with the monotonic ktime_get_boottime_seconds() to ensure the keepalive mechanism is robust against system clock modifications.
Right now, the driver uses ktime_get_real_seconds() to track peer timeouts, relying on the system wall-clock.
An administrative time adjustment or an NTP sync that steps the clock forward can cause now' to instantly exceed last_recv + timeout'.
When this occurs, the driver artificially expires healthy peers. Depending on the OpenVPN user-space configuration, this triggers a premature tunnel restart (if --keepalive or --ping-restart is used) or a complete disconnection of the client (if --ping-exit is used).
--- drivers/net/ovpn/io.c | 4 ++-- drivers/net/ovpn/peer.c | 4 ++-- 2 files changed, 4 insertions(+), 4 deletions(-)
drivers/net/ovpn/io.c +2 -2
@@ -142,7 +142,7 @@ void ovpn_decrypt_post(void *data, int ret) }/* keep track of last received authenticated packet for keepalive */ -WRITE_ONCE (peer->last_recv,ktime_get_real_seconds ()); +WRITE_ONCE (peer->last_recv,ktime_get_boottime_seconds ());rcu_read_lock (); sock =rcu_dereference (peer->sock);
@@ -294,7 +294,7 @@ void ovpn_encrypt_post(void *data, int ret)ovpn_peer_stats_increment_tx (&peer->link_stats, orig_len);/* keep track of last sent packet for keepalive */ -WRITE_ONCE (peer->last_sent,ktime_get_real_seconds ()); +WRITE_ONCE (peer->last_sent,ktime_get_boottime_seconds ());/* skb passed down the stack - don't free it */ skb =NULL ; err_unlock:
drivers/net/ovpn/peer.c +2 -2
@@ -45,7 +45,7 @@ static void unlock_ovpn(struct ovpn_priv *ovpn,*/ void ovpn_peer_keepalive_set (struct ovpn_peer *peer,u32 interval,u32 timeout) { -time64_t now =ktime_get_real_seconds (); +time64_t now =ktime_get_boottime_seconds (); netdev_dbg(peer->ovpn->dev, "scheduling keepalive for peer %u: interval=%u timeout=%u\n",
@@ -1359,7 +1359,7 @@ void ovpn_peer_keepalive_work(struct work_struct *work) { struct ovpn_priv *ovpn = container_of(work, struct ovpn_priv, keepalive_work.work); -time64_t next_run =0 , now =ktime_get_real_seconds (); +time64_t next_run =0 , now =ktime_get_boottime_seconds ();LLIST_HEAD (release_list);spin_lock_bh (&ovpn->lock);
Signature · 1 line
2.54.0