linux-kernel - Re: debugfs_remove() vs. anything that is dynamic

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:	Fri, 04 Apr 2008 17:41:24 +0200
From:	Johannes Berg <johannes@...solutions.net>
To:	Greg KH <greg@...ah.com>
Cc:	Linux Kernel list <linux-kernel@...r.kernel.org>
Subject: Re: debugfs_remove() vs. anything that is dynamic

> If you worry about this type of interaction, use debugfs_create_file,
> which takes a fileops, and set your module owner in there so that the
> reference count will not allow your module from being removed.
>
> Also remember, you have to be root to unload modules, so if you are
> doing that, and you have debugfs files open, you should know better :)

That really was just an example, we have per-wireless-device debugfs
code users can easily trigger a debugfs_remove() by unplugging a usb
netdevice for example. Also, that means that anything that is dynamic
would have the lifetime rules imposed by debugfs which is rather
awkward.

The current code for simple_attr_open copies the in i_private pointer:

attr->data = inode->i_private;

if, instead, it would keep a reference to that, like

attr->dataptr = &inode->i_private;

we could NULL out that pointer on debugfs_remove() and have
simple_attr_read() just return -ENOENT.

However, if nobody else is concerned about this, I'll just remove all
the wireless debugfs code instead, I just don't want to allow crashing
it that way.

johannes

Download attachment "signature.asc" of type "application/pgp-signature" (829 bytes)