netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH net] rxrpc: Fix untrusted unsigned subtract
@ 2025-09-11 23:06 David Howells
  2025-09-12 18:48 ` Simon Horman
  2025-09-14 20:10 ` patchwork-bot+netdevbpf
  0 siblings, 2 replies; 5+ messages in thread
From: David Howells @ 2025-09-11 23:06 UTC (permalink / raw)
  To: Dan Carpenter, netdev
  Cc: dhowells, Marc Dionne, Jakub Kicinski, David S. Miller,
	Eric Dumazet, Paolo Abeni, Simon Horman, linux-afs, linux-kernel

Fix the following Smatch Smatch static checker warning:

   net/rxrpc/rxgk_app.c:65 rxgk_yfs_decode_ticket()
   warn: untrusted unsigned subtract. 'ticket_len - 10 * 4'

by prechecking the length of what we're trying to extract in two places in
the token and decoding for a response packet.

Also use sizeof() on the struct we're extracting rather specifying the size
numerically to be consistent with the other related statements.

Fixes: 9d1d2b59341f ("rxrpc: rxgk: Implement the yfs-rxgk security class (GSSAPI)")
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Closes: https://lists.infradead.org/pipermail/linux-afs/2025-September/010135.html
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: Jakub Kicinski <kuba@kernel.org>
cc: "David S. Miller" <davem@davemloft.net>
cc: Eric Dumazet <edumazet@google.com>
cc: Paolo Abeni <pabeni@redhat.com>
cc: Simon Horman <horms@kernel.org>
cc: linux-afs@lists.infradead.org
cc: netdev@vger.kernel.org
---
 net/rxrpc/rxgk_app.c |   19 ++++++++++++++-----
 1 file changed, 14 insertions(+), 5 deletions(-)

diff --git a/net/rxrpc/rxgk_app.c b/net/rxrpc/rxgk_app.c
index df684b5a8531..30275cb5ba3e 100644
--- a/net/rxrpc/rxgk_app.c
+++ b/net/rxrpc/rxgk_app.c
@@ -54,6 +54,10 @@ int rxgk_yfs_decode_ticket(struct rxrpc_connection *conn, struct sk_buff *skb,
 
 	_enter("");
 
+	if (ticket_len < 10 * sizeof(__be32))
+		return rxrpc_abort_conn(conn, skb, RXGK_INCONSISTENCY, -EPROTO,
+					rxgk_abort_resp_short_yfs_tkt);
+
 	/* Get the session key length */
 	ret = skb_copy_bits(skb, ticket_offset, tmp, sizeof(tmp));
 	if (ret < 0)
@@ -195,22 +199,23 @@ int rxgk_extract_token(struct rxrpc_connection *conn, struct sk_buff *skb,
 		__be32 token_len;
 	} container;
 
+	if (token_len < sizeof(container))
+		goto short_packet;
+
 	/* Decode the RXGK_TokenContainer object.  This tells us which server
 	 * key we should be using.  We can then fetch the key, get the secret
 	 * and set up the crypto to extract the token.
 	 */
 	if (skb_copy_bits(skb, token_offset, &container, sizeof(container)) < 0)
-		return rxrpc_abort_conn(conn, skb, RXGK_PACKETSHORT, -EPROTO,
-					rxgk_abort_resp_tok_short);
+		goto short_packet;
 
 	kvno		= ntohl(container.kvno);
 	enctype		= ntohl(container.enctype);
 	ticket_len	= ntohl(container.token_len);
 	ticket_offset	= token_offset + sizeof(container);
 
-	if (xdr_round_up(ticket_len) > token_len - 3 * 4)
-		return rxrpc_abort_conn(conn, skb, RXGK_PACKETSHORT, -EPROTO,
-					rxgk_abort_resp_tok_short);
+	if (xdr_round_up(ticket_len) > token_len - sizeof(container))
+		goto short_packet;
 
 	_debug("KVNO %u", kvno);
 	_debug("ENC  %u", enctype);
@@ -285,4 +290,8 @@ int rxgk_extract_token(struct rxrpc_connection *conn, struct sk_buff *skb,
 	 * also come out this way if the ticket decryption fails.
 	 */
 	return ret;
+
+short_packet:
+	return rxrpc_abort_conn(conn, skb, RXGK_PACKETSHORT, -EPROTO,
+				rxgk_abort_resp_tok_short);
 }


^ permalink raw reply related	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2025-09-14 20:10 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-09-11 23:06 [PATCH net] rxrpc: Fix untrusted unsigned subtract David Howells
2025-09-12 18:48 ` Simon Horman
2025-09-12 19:43   ` David Howells
2025-09-14 20:04     ` Jakub Kicinski
2025-09-14 20:10 ` patchwork-bot+netdevbpf

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).