linux-kernel - Re: [PATCH 1/2] vfs: export symbol 'getname' and 'putname'

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:	Wed, 22 Apr 2015 06:40:48 +0000
From:	"Drokin, Oleg" <oleg.drokin@...el.com>
To:	Greg Kroah-Hartman <gregkh@...uxfoundation.org>
CC:	Christoph Hellwig <hch@...radead.org>,
	Boqun Feng <boqun.feng@...il.com>,
	"<devel@...verdev.osuosl.org>" <devel@...verdev.osuosl.org>,
	"Dilger, Andreas" <andreas.dilger@...el.com>,
	"<linux-kernel@...r.kernel.org>" <linux-kernel@...r.kernel.org>,
	Al Viro <viro@...iv.linux.org.uk>,
	"<linux-fsdevel@...r.kernel.org>" <linux-fsdevel@...r.kernel.org>,
	Jan Kara <jack@...e.cz>
Subject: Re: [PATCH 1/2] vfs: export symbol 'getname' and 'putname'

On Apr 22, 2015, at 2:31 AM, Greg Kroah-Hartman wrote:

> On Tue, Apr 21, 2015 at 10:53:11PM -0700, Christoph Hellwig wrote:
>> Nak on exporting symbols for broken staging code.  Please get rid of
>> the ioctls looking up path names in horrible ways in the lustre code.
> 
> I agree with Christoph, we shouldn't be doing this.  Let's fix lustre
> "properly", which should be possible, right?

Well, sort of.

For the first ioctl we clearly can do an fd pass and
get info from there (need to also teach the tools about the new ioctl,
so I cannot submit a patch right away, but I'll get on it).

In the second case the usecase is a bit more involved.
It deals with the problem of having directory entry pointing to a non-existing
directory (so basically a name only, but we do know it's supposed to be
a directory, just on another server), so I doubt we can even open it to get
the fd.
Now, perhaps we can just allow regular rmdir to ignore the error and kill
the bogus entry anyway, but there was this thought that we might want to alert
the user there's something more fundamentally broken going on here,
and so they would need to call this certain ioctl called if they just would like
to kill the entry anyway as opposed to calling say fsck to make sure there's
nothing else broken.
I am not entirely sure of this idea myself, but it probably made sense
at some point.
I see there's quite a dislike for this approach, so we can remove it if there's
no better option.

Bye,
    Oleg--
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/