netdev - Re: CVE-2014-9900 fix is not upstream

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: <1472128800.24772.11.camel@sipsolutions.net>
Date:   Thu, 25 Aug 2016 14:40:00 +0200
From:   Johannes Berg <johannes@...solutions.net>
To:     Hannes Frederic Sowa <hannes@...essinduktion.org>,
        Lennart Sorensen <lsorense@...lub.uwaterloo.ca>,
        Al Viro <viro@...IV.linux.org.uk>
Cc:     Joe Perches <joe@...ches.com>, David Miller <davem@...emloft.net>,
        ben@...adent.org.uk, luis.henriques@...onical.com,
        avijitnsec@...eaurora.org, netdev@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: CVE-2014-9900 fix is not upstream


> If we want to go down this route, probably the only option is to add
> __attribute__((pack)) those structs to just have no padding at all,
> thus breaking uapi.
> 

We could also spell out the padding bytes as reserved, i.e. instead of

struct ethtool_wolinfo {
        __u32   cmd;
        __u32   supported;
        __u32   wolopts;
        __u8    sopass[SOPASS_MAX];     // 6, actually
};

we could do

struct ethtool_wolinfo {
        __u32   cmd;
        __u32   supported;
        __u32   wolopts;
        __u8    sopass[SOPASS_MAX];     // 6, actually
	__u8	reserved[2];
};

and then the compiler has to properly treat it, since it's no longer
unnamed padding.

Maybe somebody can come up with a smart BUILD_BUG_ON() to ensure such
structs have no padding.

That would allow us to keep the C99 initializers (which is nice) and
not have to worry about this.

johannes