linux-kernel - Re: [PATCH 2/2] dev_ioctl: split out SIOC?IFMAP ioctls

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] [day] [month] [year] [list]

Date:   Thu, 1 Oct 2020 17:00:03 +0200
From:   Arnd Bergmann <arnd@...db.de>
To:     Christoph Hellwig <hch@...radead.org>
Cc:     "David S. Miller" <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>,
        Jiri Pirko <jiri@...lanox.com>,
        Taehee Yoo <ap420073@...il.com>,
        Eric Dumazet <edumazet@...gle.com>,
        Alexei Starovoitov <ast@...nel.org>,
        Andrew Lunn <andrew@...n.ch>, Jens Axboe <axboe@...nel.dk>,
        Networking <netdev@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 2/2] dev_ioctl: split out SIOC?IFMAP ioctls

On Tue, Sep 29, 2020 at 7:53 PM Christoph Hellwig <hch@...radead.org> wrote:
> On Fri, Sep 25, 2020 at 02:28:29PM +0200, Arnd Bergmann wrote:

> > Do you mean we should check that the (larger) user space size
> > remains what it is for future changes, or that the (smaller)
> > kernel size remains the same on all kernels, or maybe both?
>
> I had something like:
>
>         BUILD_BUG_ON(sizeof(struct ifmap) >
>                      sizeof(struct ifreq) - IFNAMSIZ);
>
> plus a suitable comment in mind.

But that condition is true on all 64-bit architectures, which is the
fundamental issue I'm working around. I can try to capture that
better in the comment though.

My expectation here is that passing the smaller 'ifreq' structure
to ndo_do_ioctl() is safe as long as all drivers use only the
remaining members of ifr_ifru that all fit into the first 16 bytes.
Do you see a problem with that assumption?

      Arnd