linux-kernel - Re: [PATCH] vfs: fix linkat error for unprivileged AT_EMPTY

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: <CAJfpegtQJ1MRWe21PTjfpNh2OWEZdgFXuTGTOWQHGLyyk9eGpQ@mail.gmail.com>
Date:	Thu, 25 Sep 2014 16:31:22 +0200
From:	Miklos Szeredi <miklos@...redi.hu>
To:	Al Viro <viro@...iv.linux.org.uk>
Cc:	Linus Torvalds <torvalds@...ux-foundation.org>,
	"Aneesh Kumar K.V" <aneesh.kumar@...ux.vnet.ibm.com>,
	Kernel Mailing List <linux-kernel@...r.kernel.org>,
	Linux-Fsdevel <linux-fsdevel@...r.kernel.org>
Subject: Re: [PATCH] vfs: fix linkat error for unprivileged AT_EMPTY_PATH

On Wed, Sep 24, 2014 at 9:00 PM, Al Viro <viro@...iv.linux.org.uk> wrote:
> On Wed, Sep 24, 2014 at 05:25:23PM +0200, Miklos Szeredi wrote:
>> From: Miklos Szeredi <mszeredi@...e.cz>
>>
>> Return proper error value for linkat(..., AT_EMPTY_PATH) without enough
>> privileges.
>>
>> I guess ENOENT was used here, because without AT_EMPTY_PATH that's what
>> we'd return for an empty path.  But it is wrong for AT_EMPTY_PATH, since we
>> might not even be passing an empty path, we are simply complaining about
>> lack of privs for which EPERM is the proper error.
>
> Umm...  Are you sure that nothing in userland is checking for that
> return value?  I agree that EPERM would make more sense, but...

How could I be sure?

But does it even make sense to check for that error value?  I don't
think it is, since we have never allowed unprivileged AT_EMPTY_PATH
for linkat  (except by bb2314b4799649 which was reverted before being
included in a release).

So if when (if ever) we do allow that, *then* testing for an error
value will make sense.  But surely if one is writing code which can't
even be tested, then it won't come as a big surprise if it will
eventually fail...

Thanks,
Miklos
--
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/