linux-ext4 - Re: [PATCH] mke2fs: do not change root dir ownership

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, 13 May 2013 12:12:27 -0400
From:	Mike Frysinger <vapier@...too.org>
To:	"Theodore Ts'o" <tytso@....edu>
Cc:	linux-ext4@...r.kernel.org
Subject: Re: [PATCH] mke2fs: do not change root dir ownership

On Monday 13 May 2013 09:37:31 Theodore Ts'o wrote:
> On Mon, May 06, 2013 at 05:21:56PM -0400, Mike Frysinger wrote:
> > If you use `mke2fs` on a file, the code will automatically chown the root
> > dir to the active uid/gid.  It doesn't do this to any other files though.
> > 
> > I can't see where this would really be desirable: you still need root in
> > order to mount, and the lost+found dir is owned by root.  It means if you
> > want to generate a rootfs as a non-root user, you first have to run it
> > through sudo or manually run `chown 0:0` after you've mounted it.
> 
> Yeah, this was something that we've been doing in e2fsprogs since 0.5b
> (i.e., dating back to 1997).  I agree that the behavior is a bit silly
> and we should probably change it.  It *is* a behavioural change,
> though, so I'm going to make it something that changes in 1.43, as
> opposed to a 1.42.x maintenance release.

np.  as long as it eventually gets fixed, i'm happy :).

> A workarond that I'd recommend (since we will have lots of people
> creating file systems for various mobile/embedded systems, and they
> will have scripts that need to work on existing versions of e2fsprogs)
> is to do something like this:
> 
>    mke2fs -t ext4 /tmp/foo.img 16384
>    debugfs /tmp/foo.img -R "set_inode_field / uid 0"
>    debugfs /tmp/foo.img -R "set_inode_field / gid 0"

nifty.  we already have to mount the fs to populate it, so putting a chown in 
there isn't a big deal in the current setup.

now if only e2fsprogs would integrate the `genext2fs` project :).
-mike

Download attachment "signature.asc " of type "application/pgp-signature" (837 bytes)