netdev - Re: off-by-one in DecodeQ931

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

Message-ID: <20160425152905.GA8621@strlen.de>
Date:	Mon, 25 Apr 2016 17:29:05 +0200
From:	Florian Westphal <fw@...len.de>
To:	Toby DiPasquale <toby@...g.net>
Cc:	pablo@...filter.org, Patrick McHardy <kaber@...sh.net>,
	kadlec@...ckhole.kfki.hu, davem@...emloft.net,
	netfilter-devel@...r.kernel.org, coreteam@...filter.org,
	netdev@...r.kernel.org
Subject: Re: off-by-one in DecodeQ931

Toby DiPasquale <toby@...g.net> wrote:
> I was reviewing the H.323 conntrack helper in the kernel when I came
> across what appears to be an off-by-one error in the DecodeQ931
> function. The MessageType field of the Q931 record is assigned and p
> is incremented, but the corresponding decrement to sz is missing,
> leading the sz variable to be one more than it should be. This patch
> decrements sz so it is the proper value going into the parsing of the
> information elements.
> 
> Signed-off-by: Toby DiPasquale <toby@...g.net>

Looks correct, BUT

> diff --git a/net/netfilter/nf_conntrack_h323_asn1.c
> b/net/netfilter/nf_conntrack_h323_asn1.c
> index bcd5ed6..68b1557 100644
> --- a/net/netfilter/nf_conntrack_h323_asn1.c
> +++ b/net/netfilter/nf_conntrack_h323_asn1.c
> @@ -849,6 +849,7 @@ int DecodeQ931(unsigned char *buf, size_t sz, Q931 *q931)
>         if (sz < 1)
>                 return H323_ERROR_BOUND;

sz can be 1

>         q931->MessageType = *p++;
> +       sz--;

sz is now 0

>         PRINT("MessageType = %02X\n", q931->MessageType);
>         if (*p & 0x80) {
>                 p++;
>                 sz--;

-> sz (size_t) will underflow here

I'd suggest to change the if (sz < 1) to if (sz < 2) to
resolve this, the while loop below has to be taken anyway.