public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Pavel Machek <pavel@ucw.cz>
To: Miklos Szeredi <miklos@szeredi.hu>
Cc: akpm@linux-foundation.org, hch@infradead.org, serue@us.ibm.com,
	viro@ftp.linux.org.uk, ebiederm@xmission.com, kzak@redhat.com,
	linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
	containers@lists.osdl.org, util-linux-ng@vger.kernel.org
Subject: Re: [patch 7/9] unprivileged mounts: allow unprivileged fuse mounts
Date: Wed, 9 Jan 2008 12:33:25 +0100	[thread overview]
Message-ID: <20080109113325.GC9735@elf.ucw.cz> (raw)
In-Reply-To: <E1JCWxm-0005Fn-36@pomaz-ex.szeredi.hu>

Hi!

> > > AFAIR there were two security vulnerabilities in fuse's history, one
> > > of them an information leak in the kernel module, and the other one an
> > > mtab corruption issue in the fusermount utility.  I don't think this
> > > is such a bad track record.
> > 
> > Not bad indeed. But I'd consider 'kill -9 not working' to be DoS
> > vulnerability...
> 
> The worst that can happen is that a sysadmin doesn't read the docs
> (likely) before enabling fuse on a multiuser system, and is surprised
> by a user doing funny things.  And _then_ has to go read the docs, or
> google for some info.  This is basically how things normally work, and
> I don't consider it a DoS.

No, this is not normal. Kill -9 has been estabilished long time ago,
and we should not be documenting its now-brokenness in
Documentation/filesystems/fuse.txt .

For example, my /etc/inittab currently has:

kb::kbrequest:/etc/rc/rc.reboot 2 0

#
# This file handles system shutdown and reboot.
#

PATH=/sbin:/bin:/usr/sbin:/usr/bin
sync &

# Kill all processes.
wall System is going down NOW\!
echo -n -e "\rSystem is going down: processes."
killall5 -15
echo -n "."
sleep 1
echo -n ". "
killall5 -9

# Before unmounting file systems write a reboot record to wtmp.
echo -n "wtmp "
halt -w

# Swap needs to be unmounted because otherwise busy filesystems
remain.
echo -n "swap "
swapoff -a
swapoff /c/swap
swapoff /c/swap2

# Unmount file systems
echo -n "umount."
umount -a || (
	sync &
	echo -n "umount-retry."
	sleep 1
	umount -a || sulogin
)
echo -n ". "
mount -n -o remount,ro /

# Now halt or reboot.
if [ "$2" = "0" ] ; then
	swapoff -a
	echo "halted."
	halt -p -f
else
	echo "rebooting..."
	reboot -d -f
fi

...this will break with FUSE enabled, right? (Minor security hole by
allowing users to stop c-a-delete, where none existed before?)

I'm currently suspending by 'echo "mem" > /sys/power/state'. How
should I do that _safely_ with FUSE enabled?

If I want to get rid of nasty user in multiuser system, I do 
su nastyuser 'kill -9 -1' . How do I do the equivalent with FUSE
enabled? (Without affecting other users?)

Load average was never really meaningful number, but with FUSE
enabled, users can set it to 666 without actually eating any CPU.

SIGSTOP used to work, allowing you to prevent user processes from
working while you examine them. Now SIGSTOP can be delayed for
arbitrary time.

Heck, imagine malicious user process misbehaves. Before FUSE, you
could at least attach it with gdb to look what it is doing. Now you
can't.

I really believe FUSE vs. signals needs fixing. Either that, or
updating all the manpages

man 1 kill:
-       KILL       9   exit      this signal may not be blocked
+       KILL       9   exit      this signal may not be blocked, except by FUSE user mount

								Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

  reply	other threads:[~2008-01-09 11:33 UTC|newest]

Thread overview: 62+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-01-08 11:35 [patch 0/9] mount ownership and unprivileged mount syscall (v6) Miklos Szeredi
2008-01-08 11:35 ` [patch 1/9] unprivileged mounts: add user mounts to the kernel Miklos Szeredi
2008-01-08 21:34   ` Pavel Machek
2008-01-08 21:47   ` Pavel Machek
2008-01-14 21:46   ` Serge E. Hallyn
2008-01-08 11:35 ` [patch 2/9] unprivileged mounts: allow unprivileged umount Miklos Szeredi
2008-01-14 21:48   ` Serge E. Hallyn
2008-01-08 11:35 ` [patch 3/9] unprivileged mounts: account user mounts Miklos Szeredi
2008-01-08 18:18   ` Dave Hansen
2008-01-08 19:18     ` Miklos Szeredi
2008-01-14 21:53   ` Serge E. Hallyn
2008-01-08 11:35 ` [patch 4/9] unprivileged mounts: propagate error values from clone_mnt Miklos Szeredi
2008-01-14 22:23   ` Serge E. Hallyn
2008-01-15 10:15     ` Miklos Szeredi
2008-01-08 11:35 ` [patch 5/9] unprivileged mounts: allow unprivileged bind mounts Miklos Szeredi
2008-01-08 18:12   ` Dave Hansen
2008-01-08 19:08     ` Miklos Szeredi
2008-01-08 19:15       ` Dave Hansen
2008-01-08 20:44       ` Szabolcs Szakacsits
2008-01-09 12:45       ` Jan Engelhardt
2008-01-09 13:26         ` Karel Zak
2008-01-09 13:32           ` Miklos Szeredi
2008-01-08 18:26   ` Dave Hansen
2008-01-08 19:21     ` Miklos Szeredi
2008-01-10  4:47   ` Serge E. Hallyn
2008-01-14 22:42   ` Serge E. Hallyn
2008-01-08 11:35 ` [patch 6/9] unprivileged mounts: allow unprivileged mounts Miklos Szeredi
2008-01-09 11:11   ` Karel Zak
2008-01-09 12:41     ` Miklos Szeredi
2008-01-14 22:58   ` Serge E. Hallyn
2008-01-08 11:35 ` [patch 7/9] unprivileged mounts: allow unprivileged fuse mounts Miklos Szeredi
2008-01-08 21:46   ` Pavel Machek
2008-01-08 22:42     ` Miklos Szeredi
2008-01-08 22:58       ` Pavel Machek
2008-01-09  9:11         ` Miklos Szeredi
2008-01-09 11:33           ` Pavel Machek [this message]
2008-01-09 13:16             ` Miklos Szeredi
2008-01-09 13:35               ` Pavel Machek
2008-01-09 13:48                 ` Miklos Szeredi
2008-01-09 14:00                   ` Pavel Machek
2008-01-09 14:14                     ` Miklos Szeredi
2008-01-08 23:56       ` Nigel Cunningham
2008-01-09  8:47         ` Miklos Szeredi
2008-01-09  9:29           ` Nigel Cunningham
2008-01-09 11:12           ` Pavel Machek
2008-01-09  9:19         ` Szabolcs Szakacsits
2008-01-14 23:24   ` Serge E. Hallyn
2008-01-15 10:29     ` Miklos Szeredi
2008-01-15 13:35       ` Serge E. Hallyn
2008-01-08 11:35 ` [patch 8/9] unprivileged mounts: propagation: inherit owner from parent Miklos Szeredi
2008-01-14 23:13   ` Serge E. Hallyn
2008-01-15 10:39     ` Miklos Szeredi
2008-01-15 14:21       ` Serge E. Hallyn
2008-01-15 14:37         ` Miklos Szeredi
2008-01-15 14:59           ` Serge E. Hallyn
2008-01-08 11:35 ` [patch 9/9] unprivileged mounts: add "no submounts" flag Miklos Szeredi
2008-01-14 23:39   ` Serge E. Hallyn
2008-01-15 10:41     ` Miklos Szeredi
2008-01-15 10:53       ` A. C. Censi
2008-01-15 10:58         ` Miklos Szeredi
2008-01-15 13:47           ` Serge E. Hallyn
2008-01-16  9:43             ` Miklos Szeredi

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20080109113325.GC9735@elf.ucw.cz \
    --to=pavel@ucw.cz \
    --cc=akpm@linux-foundation.org \
    --cc=containers@lists.osdl.org \
    --cc=ebiederm@xmission.com \
    --cc=hch@infradead.org \
    --cc=kzak@redhat.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=miklos@szeredi.hu \
    --cc=serue@us.ibm.com \
    --cc=util-linux-ng@vger.kernel.org \
    --cc=viro@ftp.linux.org.uk \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox