netdev - Re: [PATCH 1/4] Ethernet drivers in 3.14-rc3 kernel: fix 3 buffer overflows triggered by hardware devices

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <CAP6odji314zs8cOmYdvYgFcoJy5ggqPXfNJMrDH8R1m6a3Y_cA@mail.gmail.com>
Date:	Fri, 21 Feb 2014 19:12:08 -0800
From:	Grant Grundler <grantgrundler@...il.com>
To:	Alon Nafta <alon@...vatecore.com>
Cc:	netdev <netdev@...r.kernel.org>
Subject: Re: [PATCH 1/4] Ethernet drivers in 3.14-rc3 kernel: fix 3 buffer
 overflows triggered by hardware devices

On Fri, Feb 21, 2014 at 4:02 PM, Alon Nafta <alon@...vatecore.com> wrote:
> From: Alon Nafta <alon@...vatecore.com>
>
> Linux Kernel contains multiple overflow conditions that are triggered as
> hardware-supplied inputs are not properly validated when parsing Ethernet
> packets. This may allow a local attacker to cause an overflow, resulting in
> a denial of service or potentially allowing the execution of arbitrary code.
>
> The programmatic error resides in the use of an integer type to describe
> packet length, without proper validation for negative values. In all three
> (3) bugs this patch fixes, a value of 0x30000 for the hardware signal, named
> status, will result in the value of 0xffffffff for pkt_len, and an
> allocation of a socket buffer with size of 0x1. This result in an overflow
> when data is copied into that buffer.
>
> Signed-off-by: Alon Nafta <alon@...vatecore.com>

LGTM.  Thanks! :)

Reviewed-by: Grant Grundler <grundler@...isc-linux.org>

> ---
> diff -uprN -X linux-3.14-rc3/Documentation/dontdiff
> linux-3.14-rc3-orig/drivers/net/ethernet/dec/tulip/de4x5.c
> linux-3.14-rc3/drivers/net/ethernet/dec/tulip/de4x5.c
> --- linux-3.14-rc3-orig/drivers/net/ethernet/dec/tulip/de4x5.c 2014-02-20
> 17:59:14.704084300 -0800
> +++ linux-3.14-rc3/drivers/net/ethernet/dec/tulip/de4x5.c 2014-02-20
> 18:23:08.987749400 -0800
> @@ -1635,8 +1635,8 @@ de4x5_rx(struct net_device *dev)
>   if (status & RD_OF)           lp->pktStats.rx_overflow++;
>       } else {                          /* A valid frame received */
>   struct sk_buff *skb;
> - short pkt_len = (short)(le32_to_cpu(lp->rx_ring[entry].status)
> -                             >> 16) - 4;
> + short pkt_len = (short)((le32_to_cpu(lp->rx_ring[entry].status)
> +                             >> 16) - 4) & 0x7fff;
>
>   if ((skb = de4x5_alloc_rx_buff(dev, entry, pkt_len)) == NULL) {
>       printk("%s: Insufficient memory; nuking packet.\n",
> diff -uprN -X linux-3.14-rc3/Documentation/dontdiff
> linux-3.14-rc3-orig/drivers/net/ethernet/dec/tulip/winbond-840.c
> linux-3.14-rc3/drivers/net/ethernet/dec/tulip/winbond-840.c
> --- linux-3.14-rc3-orig/drivers/net/ethernet/dec/tulip/winbond-840.c
> 2014-02-20 17:59:14.757666100 -0800
> +++ linux-3.14-rc3/drivers/net/ethernet/dec/tulip/winbond-840.c 2014-02-20
> 18:22:19.419612200 -0800
> @@ -1218,7 +1218,7 @@ static int netdev_rx(struct net_device *
>   } else {
>   struct sk_buff *skb;
>   /* Omit the four octet CRC from the length. */
> - int pkt_len = ((status >> 16) & 0x7ff) - 4;
> + int pkt_len = ((status >> 16) - 4) & 0x7ff;
>
>  #ifndef final_version
>   if (debug > 4)
> diff -uprN -X linux-3.14-rc3/Documentation/dontdiff
> linux-3.14-rc3-orig/drivers/net/ethernet/smsc/epic100.c
> linux-3.14-rc3/drivers/net/ethernet/smsc/epic100.c
> --- linux-3.14-rc3-orig/drivers/net/ethernet/smsc/epic100.c 2014-02-20
> 17:59:17.844045500 -0800
> +++ linux-3.14-rc3/drivers/net/ethernet/smsc/epic100.c 2014-02-20
> 18:21:13.196237400 -0800
> @@ -1172,7 +1172,7 @@ static int epic_rx(struct net_device *de
>   } else {
>   /* Malloc up new buffer, compatible with net-2e. */
>   /* Omit the four octet CRC from the length. */
> - short pkt_len = (status >> 16) - 4;
> + short pkt_len = ((status >> 16) - 4) & 0x7fff;
>   struct sk_buff *skb;
>
>   if (pkt_len > PKT_BUF_SZ - 4) {
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html