linux-kernel - Re: dcache: remove trylock loops (was Re: [BUG] lock_parent() breakage when used from shrink_dentry

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] [day] [month] [year] [list]

Message-ID: <20180428001003.GA14907@ZenIV.linux.org.uk>
Date:   Sat, 28 Apr 2018 01:10:03 +0100
From:   Al Viro <viro@...IV.linux.org.uk>
To:     NeilBrown <neilb@...e.com>
Cc:     Linus Torvalds <torvalds@...ux-foundation.org>,
        linux-fsdevel <linux-fsdevel@...r.kernel.org>,
        Christoph Hellwig <hch@....de>,
        Thomas Gleixner <tglx@...utronix.de>,
        Peter Zijlstra <peterz@...radead.org>,
        Sebastian Andrzej Siewior <bigeasy@...utronix.de>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Eric Biederman <ebiederm@...ssion.com>
Subject: Re: dcache: remove trylock loops (was Re: [BUG] lock_parent()
 breakage when used from shrink_dentry_list())

On Tue, Mar 13, 2018 at 12:12:51PM +1100, NeilBrown wrote:

> >	     * selinux inode_doinit_with_dentry() (two call sites, very alike)
> 
> I'm less sure about this one, but I think it probably wants
> d_find_any_alias() as well.  Like cap_inode_getsecurity() it only wants
> a dentry so that it can pass something to __vfs_getxattr(),
> and that only wants a dentry so it can pass something to ->get.
> 
> Possibly we should rename d_find_alias() to d_find_hashed_alias() so that
> people need to make a conscious choice between d_find_hashed_alias() and
> d_find_any_alias() ??

FWIW, it *is* a bug; this
                        /*
                         * this is can be hit on boot when a file is accessed
                         * before the policy is loaded.  When we load policy we
                         * may find inodes that have no dentry on the
                         * sbsec->isec_head list.  No reason to complain as these
                         * will get fixed up the next time we go through
                         * inode_doinit with a dentry, before these inodes could
                         * be used again by userspace.
                         */
in selinux/hooks.c is flat-out wrong now.  Sure, if you first load selinux
policy after exporting something over NFS or letting attacker play with
open-by-fhandle, you are past any help, but still...

I disagree about going for d_find_any_alias() from the very beginning, BTW -
we need to try it in case of d_find_alias() failure, but sufficiently crappy
filesystem can bloody well fail to access xattrs via disconnected dentry.