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]
Date:   Mon, 01 Jun 2020 11:32:06 -0700 (PDT)
From:   David Miller <davem@...emloft.net>
To:     hexie3605@...il.com
Cc:     kuba@...nel.org, madhuparnabhowmik04@...il.com,
        netdev@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] drivers/net/wan/lapbether.c: Fixed kernel panic when
 used with AF_PACKET sockets

From: Xie He <hexie3605@...il.com>
Date: Wed, 27 May 2020 20:21:33 -0700

> When we use "AF_PACKET" sockets to send data directly over LAPB over
> Ethernet using this driver, the kernel will panic because of
> insufficient header space allocated in the "sk_buff" struct.
> 
> The header space needs 18 bytes because:
>   the lapbether driver will remove a pseudo header of 1 byte;
>   the lapb module will prepend the LAPB header of 2 or 3 bytes;
>   the lapbether driver will prepend a length field of 2 bytes and the
> Ethernet header of 14 bytes.
> 
> So -1 + 3 + 16 = 18.
> 
> Signed-off-by: Xie He <hexie3605@...il.com>

This is not the real problem.

The real problem is that this is a stacked, layered, device and the
lapbether driver does not take the inner device's header length into
consideration.  It should take this from the child device's netdev
structure rather than use constants.

Your test case will still fail when lapbether is stacked on top of a
VLAN device or similar, even with your changes.

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ