linux-kernel - Re: [PATCH] mlx4: use snprintf() instead of sprintf() for safety

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]

Message-ID: <Y33AeQzuXFjnr+t/@unreal>
Date:   Wed, 23 Nov 2022 08:40:57 +0200
From:   Leon Romanovsky <leon@...nel.org>
To:     Jakub Kicinski <kuba@...nel.org>
Cc:     Peter Kosyh <pkosyh@...dex.ru>, Tariq Toukan <tariqt@...dia.com>,
        "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>,
        Paolo Abeni <pabeni@...hat.com>, netdev@...r.kernel.org,
        linux-rdma@...r.kernel.org, linux-kernel@...r.kernel.org,
        lvc-project@...uxtesting.org
Subject: Re: [PATCH] mlx4: use snprintf() instead of sprintf() for safety

On Tue, Nov 22, 2022 at 12:12:23PM -0800, Jakub Kicinski wrote:
> On Tue, 22 Nov 2022 16:48:15 +0200 Leon Romanovsky wrote:
> > On Tue, Nov 22, 2022 at 04:04:53PM +0300, Peter Kosyh wrote:
> > > Use snprintf() to avoid the potential buffer overflow. Although in the
> > > current code this is hardly possible, the safety is unclean.  
> > 
> > Let's fix the tools instead. The kernel code is correct.
> 
> I'm guessing the code is correct because port can't be a high value?

Yes, port value is provided as input to mlx4_init_port_info() and it is
capped by MLX4_MAX_PORTS, which is 2.

> Otherwise, if I'm counting right, large enough port representation
> (e.g. 99999999) could overflow the string. If that's the case - how
> would they "fix the tool" to know the port is always a single digit?

I may admit that I don't know how hard or easy to implement it, but it
will be great if tool would be able to understand that dev->caps.num_ports
are not really dynamic values, but constant ones.

However, I don't mind if we merge it.

Thanks,
Reviewed-by: Leon Romanovsky <leonro@...dia.com>