From: Andrey Misiurov <amisiurov@cloudlinux.com>
To: Andrew Lunn <andrew+netdev@lunn.ch>,
"David S . Miller" <davem@davemloft.net>,
Eric Dumazet <edumazet@kernel.org>,
Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>
Cc: Wojciech Drewek <wojciech.drewek@intel.com>,
Simon Horman <horms@kernel.org>,
netdev@vger.kernel.org, linux-kernel@vger.kernel.org,
Andrey Misiurov <amisiurov@cloudlinux.com>
Subject: [PATCH net] pfcp: drop packets until the device is initialised
Date: Thu, 19 Nov 2026 13:29:46 +0100 [thread overview]
Message-ID: <20261119122946.1720693-1-amisiurov@cloudlinux.com> (raw)
pfcp_newlink() creates the PFCP UDP socket (port 8805) and publishes it
(sk_user_data = pfcp, encap_rcv = pfcp_encap_recv) before
register_netdevice() runs ndo_init, which is where pfcp->dev is set. A
datagram that arrives in that window passes the "pfcp != NULL" check in
pfcp_encap_recv() and then dereferences a NULL pfcp->dev:
BUG: kernel NULL pointer dereference, address: 0000000000000108
#PF: supervisor read access in kernel mode
Oops: 0000 [#1] SMP PTI
CPU: 0 PID: 26516 Comm: pfcp_race
RIP: 0010:pfcp_encap_recv+0xeb/0x1d0 [pfcp]
Call Trace:
<IRQ>
udp_queue_rcv_one_skb+0x24a/0x560
udp_queue_rcv_skb+0x4f/0x80
udp_unicast_rcv_skb+0x78/0xa0
__udp4_lib_rcv+0x519/0x5f0
udp_rcv+0x25/0x40
ip_protocol_deliver_rcu+0xd4/0x210
ip_local_deliver_finish+0x86/0x110
ip_local_deliver+0x65/0x110
ip_rcv+0x17d/0x190
__netif_receive_skb_one_core+0x8d/0xa0
The window is wide when register_netdevice() fails before ndo_init (e.g. an
invalid name): the socket stays live until pfcp_del_sock() on the error
path. A narrower window also exists on the normal success path.
Drop the packet while pfcp->dev is not yet set or the device is not
running.
Fixes: 76c8764ef36a ("pfcp: add PFCP module")
Signed-off-by: Andrey Misiurov <amisiurov@cloudlinux.com>
---
Found during a backport review.
Reproduced on ubuntu-resolute 7.0.0-42: flood UDP :8805 while creating a
pfcp link with an invalid name (register_netdevice() fails before ndo_init)
-> the NULL-deref above (kdump vmcore). With the fix the same run stays clean.
drivers/net/pfcp.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/pfcp.c b/drivers/net/pfcp.c
--- a/drivers/net/pfcp.c
+++ b/drivers/net/pfcp.c
@@ -61,7 +61,7 @@
goto drop;
pfcp = rcu_dereference_sk_user_data(sk);
- if (unlikely(!pfcp))
+ if (unlikely(!pfcp || !pfcp->dev || !netif_running(pfcp->dev)))
goto drop;
unparsed = pfcp_hdr(skb);
reply other threads:[~2026-11-19 12:29 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20261119122946.1720693-1-amisiurov@cloudlinux.com \
--to=amisiurov@cloudlinux.com \
--cc=andrew+netdev@lunn.ch \
--cc=davem@davemloft.net \
--cc=edumazet@kernel.org \
--cc=horms@kernel.org \
--cc=kuba@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=wojciech.drewek@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®