[PATCH] Netfilter: Fix integer overflow in net/ipv6/netfilter/ip6_tables.c (fwd)

[PATCH] Netfilter: Fix integer overflow in net/ipv6/netfilter/ip6_tables.c (fwd)

[James Morris]

---------- Forwarded message ----------
Date: Sat, 20 Mar 2010 22:32:40 +0800
From: wzt.wzt@gmail.com
To: linux-kernel@vger.kernel.org
Cc: netfilter-devel@vger.kernel.org, kaber@trash.net
Subject: [PATCH] Netfilter: Fix integer overflow in
    net/ipv6/netfilter/ip6_tables.c

The get.size field in the get_entries() interface is not bounded
correctly. The size is used to determine the total entry size.
The size is bounded, but can overflow and so the size checks may
not be sufficient to catch invalid size. Fix it by catching size
values that would cause overflows before calculating the size.

Signed-off-by: Zhitong Wang <zhitong.wangzt@alibaba-inc.com>

---
 net/ipv4/netfilter/ip_tables.c  |    4 ++++
 net/ipv6/netfilter/ip6_tables.c |    4 ++++
 2 files changed, 8 insertions(+), 0 deletions(-)

diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c
index 4e7c719..6abd3d2 100644
--- a/net/ipv4/netfilter/ip_tables.c
+++ b/net/ipv4/netfilter/ip_tables.c
@@ -1164,6 +1164,10 @@ get_entries(struct net *net, struct ipt_get_entries __user *uptr, int *len)
 	}
 	if (copy_from_user(&get, uptr, sizeof(get)) != 0)
 		return -EFAULT;
+
+	if (get.size >= INT_MAX / sizeof(struct ipt_get_entries))
+		return -EINVAL;
+
 	if (*len != sizeof(struct ipt_get_entries) + get.size) {
 		duprintf("get_entries: %u != %zu\n",
 			 *len, sizeof(get) + get.size);
diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c
index 0b4557e..5185822 100644
--- a/net/ipv6/netfilter/ip6_tables.c
+++ b/net/ipv6/netfilter/ip6_tables.c
@@ -1190,6 +1190,10 @@ get_entries(struct net *net, struct ip6t_get_entries __user *uptr, int *len)
 	}
 	if (copy_from_user(&get, uptr, sizeof(get)) != 0)
 		return -EFAULT;
+
+	if (get.size >= INT_MAX / sizeof(struct ip6t_get_entries))
+		return -EINVAL;
+
 	if (*len != sizeof(struct ip6t_get_entries) + get.size) {
 		duprintf("get_entries: %u != %zu\n",
 			 *len, sizeof(get) + get.size);
-- 
1.6.5.3

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Re: [PATCH] Netfilter: Fix integer overflow in net/ipv6/netfilter/ip6_tables.c (fwd)

[Eric Dumazet]

Le lundi 22 mars 2010 à 08:50 +1100, James Morris a écrit :
> ---------- Forwarded message ----------
> Date: Sat, 20 Mar 2010 22:32:40 +0800
> From: wzt.wzt@gmail.com
> To: linux-kernel@vger.kernel.org
> Cc: netfilter-devel@vger.kernel.org, kaber@trash.net
> Subject: [PATCH] Netfilter: Fix integer overflow in
>     net/ipv6/netfilter/ip6_tables.c
> 
> The get.size field in the get_entries() interface is not bounded
> correctly. The size is used to determine the total entry size.
> The size is bounded, but can overflow and so the size checks may
> not be sufficient to catch invalid size. Fix it by catching size
> values that would cause overflows before calculating the size.
> 
> Signed-off-by: Zhitong Wang <zhitong.wangzt@alibaba-inc.com>
> 


Unless I am wrong, this patch is not necessary, since we do a bit
later :


                if (get.size == private->size)
                        ret = copy_entries_to_user(private->size,
                                                   t, uptr->entrytable);
                else {
                        duprintf("get_entries: I've got %u not %u!\n",
                                 private->size, get.size);
                        ret = -EAGAIN;
                }

So if get.size doesnt match private->size (kernel certified), we exit
with an error. No calculation, no overflow involved.