Netdev List
 help / color / mirror / Atom feed
From: Tristan Madani <tristmd@gmail.com>
To: Pablo Neira Ayuso <pablo@netfilter.org>, Florian Westphal <fw@strlen.de>
Cc: Phil Sutter <phil@nwl.cc>, Christian Brauner <brauner@kernel.org>,
	"David S . Miller" <davem@davemloft.net>,
	Eric Dumazet <edumazet@kernel.org>,
	Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Simon Horman <horms@kernel.org>,
	netfilter-devel@vger.kernel.org, coreteam@netfilter.org,
	netdev@vger.kernel.org,
	Tristan Madani <tristan@talencesecurity.com>,
	stable@vger.kernel.org
Subject: [PATCH nf 1/5] netfilter: xt_owner: pin file with get_file_rcu() before reading f_cred
Date: Tue, 13 Oct 2026 23:43:24 +0000	[thread overview]
Message-ID: <20261013234329.2797868-2-tristmd@gmail.com> (raw)
In-Reply-To: <20261013234329.2797868-1-tristmd@gmail.com>

From: Tristan Madani <tristan@talencesecurity.com>

Since commit 0ede61d8589cc ("file: convert to SLAB_TYPESAFE_BY_RCU"),
struct file objects can have their slab slot reused while RCU readers
still hold pointers to them. The xt_owner match reads file->f_cred to
check fsuid/fsgid/group_info in the packet path without pinning the
file. If the file is freed and the slab slot is reused between reading
sock->file and dereferencing f_cred, the credential check will operate
on stale or unrelated credential data.

Use get_file_rcu() to take a reference on the file before accessing its
credentials, and fput() to release it on all return paths.

Fixes: 0ede61d8589cc ("file: convert to SLAB_TYPESAFE_BY_RCU")
Cc: stable@vger.kernel.org
Signed-off-by: Tristan Madani <tristan@talencesecurity.com>
---
 net/netfilter/xt_owner.c | 27 ++++++++++++++++++---------
 1 file changed, 18 insertions(+), 9 deletions(-)

diff --git a/net/netfilter/xt_owner.c b/net/netfilter/xt_owner.c
index 8f2e57b2a5868..6386b52147d87 100644
--- a/net/netfilter/xt_owner.c
+++ b/net/netfilter/xt_owner.c
@@ -65,8 +65,8 @@ owner_mt(const struct sk_buff *skb, struct xt_action_param *par)
 	const struct xt_owner_match_info *info = par->matchinfo;
 	struct sock *sk = skb_to_full_sk(skb);
 	struct net *net = xt_net(par);
-	const struct socket *sock;
-	const struct file *filp;
+	struct socket *sock;
+	struct file *filp;
 
 	if (!sk || !READ_ONCE(sk->sk_socket) || !net_eq(net, sock_net(sk)))
 		return (info->match ^ info->invert) == 0;
@@ -77,13 +77,17 @@ owner_mt(const struct sk_buff *skb, struct xt_action_param *par)
 		 */
 		return false;
 
-	/* The sk pointer remains valid as long as the skb is. The sk_socket and
-	 * file pointer may become NULL if the socket is closed. Both structures
-	 * (including file->cred) are RCU freed which means they can be accessed
-	 * within a RCU read section.
+	/* The sk pointer remains valid as long as the skb is. The
+	 * sk_socket and file pointer may become NULL if the socket is
+	 * closed. Since struct file uses SLAB_TYPESAFE_BY_RCU, pin the
+	 * file with get_file_rcu() to ensure f_cred is not from a
+	 * reused slab slot.
 	 */
 	sock = READ_ONCE(sk->sk_socket);
-	filp = sock ? READ_ONCE(sock->file) : NULL;
+	if (sock)
+		filp = get_file_rcu(&sock->file);
+	else
+		filp = NULL;
 	if (filp == NULL)
 		return ((info->match ^ info->invert) &
 		       (XT_OWNER_UID | XT_OWNER_GID)) == 0;
@@ -94,8 +98,10 @@ owner_mt(const struct sk_buff *skb, struct xt_action_param *par)
 
 		if ((uid_gte(filp->f_cred->fsuid, uid_min) &&
 		     uid_lte(filp->f_cred->fsuid, uid_max)) ^
-		    !(info->invert & XT_OWNER_UID))
+		    !(info->invert & XT_OWNER_UID)) {
+			fput(filp);
 			return false;
+		}
 	}
 
 	if (info->match & XT_OWNER_GID) {
@@ -120,10 +126,13 @@ owner_mt(const struct sk_buff *skb, struct xt_action_param *par)
 			}
 		}
 
-		if (match ^ !(info->invert & XT_OWNER_GID))
+		if (match ^ !(info->invert & XT_OWNER_GID)) {
+			fput(filp);
 			return false;
+		}
 	}
 
+	fput(filp);
 	return true;
 }
 
-- 
2.47.3


  reply	other threads:[~2026-10-13 23:43 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-13 23:43 [PATCH nf 0/5] netfilter: pin file with get_file_rcu() before reading f_cred Tristan Madani
2026-10-13 23:43 ` Tristan Madani [this message]
2026-10-13 23:43 ` [PATCH nf 2/5] netfilter: nft_meta: " Tristan Madani
2026-10-13 23:43 ` [PATCH nf 3/5] netfilter: nfnetlink_log: " Tristan Madani
2026-10-13 23:43 ` [PATCH nf 4/5] netfilter: nfnetlink_queue: " Tristan Madani
2026-10-13 23:43 ` [PATCH nf 5/5] netfilter: nf_log_syslog: " Tristan Madani
2026-10-13 23:50 ` [PATCH nf 0/5] netfilter: " netdev-bot+sinfo
2026-10-14  9:09 ` [syzbot ci] " syzbot ci

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=20261013234329.2797868-2-tristmd@gmail.com \
    --to=tristmd@gmail.com \
    --cc=brauner@kernel.org \
    --cc=coreteam@netfilter.org \
    --cc=davem@davemloft.net \
    --cc=edumazet@kernel.org \
    --cc=fw@strlen.de \
    --cc=horms@kernel.org \
    --cc=kuba@kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=netfilter-devel@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=pablo@netfilter.org \
    --cc=phil@nwl.cc \
    --cc=stable@vger.kernel.org \
    --cc=tristan@talencesecurity.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