[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <20231115100857.24659-1-oneukum@suse.com>
Date: Wed, 15 Nov 2023 11:08:57 +0100
From: Oliver Neukum <oneukum@...e.com>
To: dmitry.bezrukov@...antia.com,
marcinguy@...il.com,
davem@...emloft.net,
edumazet@...gle.com,
kuba@...nel.org,
pabeni@...hat.com,
linux-usb@...r.kernel.org,
netdev@...r.kernel.org
Cc: Oliver Neukum <oneukum@...e.com>
Subject: [PATCH] usb: aqc111: check packet for fixup for true limit
If a device sends a packet that is inbetween 0
and sizeof(u64) the value passed to skb_trim()
as length will wrap around ending up as some very
large value.
The driver will then proceed to parse the header
located at that position, which will either oops or
process some random value.
The fix is to check against sizeof(u64) rather than
0, which the driver currently does. The issue exists
since the introduction of the driver.
Signed-off-by: Oliver Neukum <oneukum@...e.com>
---
drivers/net/usb/aqc111.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/net/usb/aqc111.c b/drivers/net/usb/aqc111.c
index a017e9de2119..7b8afa589a53 100644
--- a/drivers/net/usb/aqc111.c
+++ b/drivers/net/usb/aqc111.c
@@ -1079,17 +1079,17 @@ static int aqc111_rx_fixup(struct usbnet *dev, struct sk_buff *skb)
u16 pkt_count = 0;
u64 desc_hdr = 0;
u16 vlan_tag = 0;
- u32 skb_len = 0;
+ u32 skb_len;
if (!skb)
goto err;
- if (skb->len == 0)
+ skb_len = skb->len;
+ if (skb_len < sizeof(desc_hdr))
goto err;
- skb_len = skb->len;
/* RX Descriptor Header */
- skb_trim(skb, skb->len - sizeof(desc_hdr));
+ skb_trim(skb, skb_len - sizeof(desc_hdr));
desc_hdr = le64_to_cpup((u64 *)skb_tail_pointer(skb));
/* Check these packets */
--
2.42.0
Powered by blists - more mailing lists