Re: [PATCH ovpn net 1/9] ovpn: skip rehash for peers already removed from by_id

2026-05-27, 01:18:42 +0200, Antonio Quartulli wrote:
> From: Antonio Quartulli <[email protected]>
> 
> ovpn_nl_peer_set_doit() resolves the target peer via
> ovpn_peer_get_by_id() before taking ovpn->lock. In the window between
> the lookup (which only takes a refcount) and the subsequent
> spin_lock_bh(&ovpn->lock), a concurrent OVPN_CMD_PEER_DEL, keepalive
> expiry, or socket teardown can take ovpn->lock first, run
> ovpn_peer_remove() to unhash the peer from all four tables (by_id,
> by_vpn_addr4/6, by_transp_addr) and release the lock. set_doit then
> acquires ovpn->lock and calls ovpn_peer_hash_vpn_ip(), which
> re-inserts the now-removed peer back into the rehashing tables.
> 
> The same race affects the float path: ovpn_peer_endpoints_update()
> holds only a refcount and acquires ovpn->lock very late (after async
> AEAD decrypt and a netlink notification), then rehashes the peer
> in the by_transp_addr table.
> 
> The resurrected peer becomes reachable again from the RX lookup
> (ovpn_peer_get_by_transp_addr) and the TX VPN-IP lookup, even though
> userspace believes it is gone. Once the data-path refcount drops the
> peer is freed via call_rcu while the hash entries embedded in it
> remain linked, opening a UAF window.
> 
> Bail out of the rehash when hash_entry_id is unhashed, mirroring
> the sentinel already used by ovpn_peer_remove() to detect the
> already-removed state. The check is safe under ovpn->lock, which
> serializes every mutation of hash_entry_id, and is a no-op for the
> add path because ovpn_peer_add_mp() inserts hash_entry_id before
> calling ovpn_peer_hash_vpn_ip().
> 
> Fixes: 1d36a36f6d53 ("ovpn: implement peer add/get/dump/delete via netlink")
> Signed-off-by: Antonio Quartulli <[email protected]>
> ---
>  drivers/net/ovpn/peer.c | 17 +++++++++++++++++
>  1 file changed, 17 insertions(+)

This looks correct to me, just one small thing:

> diff --git a/drivers/net/ovpn/peer.c b/drivers/net/ovpn/peer.c
> index a09d61296425..a472ffe3016b 100644
> --- a/drivers/net/ovpn/peer.c
> +++ b/drivers/net/ovpn/peer.c
> @@ -307,6 +307,16 @@ void ovpn_peer_endpoints_update(struct ovpn_peer *peer, struct sk_buff *skb)
>  			return;
>  		}
>  
> +		/* peer may have been concurrently removed between the caller's
> +		 * initial lookup and our acquisition of ovpn->lock; skip the
> +		 * rehash so we don't re-insert a removed peer
> +		 */
> +		if (unlikely(hlist_unhashed(&peer->hash_entry_id))) {
> +			spin_unlock_bh(&peer->lock);
> +			spin_unlock_bh(&peer->ovpn->lock);
> +			return;
> +		}

I'd add an "unlock2" label (or "unlock_both") to unlock both (maybe
just after the "hlist_nulls_add_head_rcu", before the existing
unlock). The unlock/return is a bit verbose and we have it 3 times
after this patch.

-- 
Sabrina