linux-kernel - Re: [patch 5/9] unprivileged mounts: allow unprivileged bind mounts

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]

Message-Id: <1199816786.9834.70.camel@localhost>
Date:	Tue, 08 Jan 2008 10:26:26 -0800
From:	Dave Hansen <haveblue@...ibm.com>
To:	Miklos Szeredi <miklos@...redi.hu>
Cc:	akpm@...ux-foundation.org, hch@...radead.org, serue@...ibm.com,
	viro@....linux.org.uk, ebiederm@...ssion.com, kzak@...hat.com,
	linux-fsdevel@...r.kernel.org, containers@...ts.osdl.org,
	util-linux-ng@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [patch 5/9] unprivileged mounts: allow unprivileged bind mounts

On Tue, 2008-01-08 at 12:35 +0100, Miklos Szeredi wrote:
> @@ -510,10 +533,16 @@ static struct vfsmount *clone_mnt(struct
>                                         int flag)
>  {
>         struct super_block *sb = old->mnt_sb;
> -       struct vfsmount *mnt = alloc_vfsmnt(old->mnt_devname);
> +       struct vfsmount *mnt;
> 
> +       if (flag & CL_SETUSER) {
> +               int err = reserve_user_mount();
> +               if (err)
> +                       return ERR_PTR(err);
> +       }
> +       mnt = alloc_vfsmnt(old->mnt_devname);
>         if (!mnt)
> -               return ERR_PTR(-ENOMEM);
> +               goto alloc_failed;
> 
>         mnt->mnt_flags = old->mnt_flags;
>         atomic_inc(&sb->s_active); 

I think there's a little race here.  We could have several users racing
to get to this point when nr_user_mounts==max_user_mounts-1.  One user
wins the race and gets their mount reserved.  The others get the error
out of reserve_user_mount(), and return.

But, the winner goes on to error out on some condition further down in
clone_mnt() and never actually instantiates the mount.

Do you think this is a problem?

I think just about the one solution is to block new mounters on a
sleepable lock until the race winner actually finishes their mount
operation.  

-- Dave

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/