From mboxrd@z Thu Jan  1 00:00:00 1970
From: Nick Piggin <npiggin@suse.de>
Subject: Re: [patch v2] fcntl: return -EFAULT if copy_to_user fails
Date: Thu, 3 Jun 2010 23:10:31 +1000
Message-ID: <20100603131031.GM6822@laptop>
References: <20100603100402.GR5483@bicker>
 <4C07826A.6060302@oss.ntt.co.jp>
 <20100603103542.GV5483@bicker>
 <4C07990A.8080508@fusionio.com>
 <4C079D34.5010500@oss.ntt.co.jp>
 <1275568683.2456.33.camel@edumazet-laptop>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: QUOTED-PRINTABLE
Cc: Takuya Yoshikawa <yoshikawa.takuya@oss.ntt.co.jp>,
	Jens Axboe <jaxboe@fusionio.com>,
	Dan Carpenter <error27@gmail.com>,
	Matthew Wilcox <matthew@wil.cx>,
	Alexander Viro <viro@zeniv.linux.org.uk>,
	Andrew Morton <akpm@linux-foundation.org>,
	Oleg Nesterov <oleg@redhat.com>,
	Greg Kroah-Hartman <gregkh@suse.de>,
	Peter Zijlstra <a.p.zijlstra@chello.nl>,
	linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org
To: Eric Dumazet <eric.dumazet@gmail.com>
Return-path: <linux-kernel-owner@vger.kernel.org>
Content-Disposition: inline
In-Reply-To: <1275568683.2456.33.camel@edumazet-laptop>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: linux-fsdevel.vger.kernel.org

On Thu, Jun 03, 2010 at 02:38:03PM +0200, Eric Dumazet wrote:
> Le jeudi 03 juin 2010 =E0 21:16 +0900, Takuya Yoshikawa a =E9crit :
> > (2010/06/03 20:59), Jens Axboe wrote:
> > > On 2010-06-03 12:35, Dan Carpenter wrote:
> > >> copy_to_user() returns the number of bytes remaining, but we wan=
t to
> > >> return -EFAULT.
> > >> 	ret =3D fcntl(fd, F_SETOWN_EX, NULL);
> > >> With the original code ret would be 8 here.
> > >>
> > >> V2: Takuya Yoshikawa pointed out a similar issue in f_getown_ex(=
)
> > >
> > > Pretty basic bug, how long has this been there?
> >=20
> > IIUC, from the beginning, when these were introduced.
>=20
> Maybe copy_to_user() was changed sometime to return a partial count
> instead of EFAULT ?

I think it's been like that since first introduced. Some functions
do need to know in order to do partial copies.

=20
> I do think we should have a set of helper functions, instead of
> spreading special EFAULT cases in one housand places...
>=20
> This is really ugly.
>=20
> static inline int sec_copy_to_user(arg1, arg2, arg3)
> {
> 	int res =3D copy_to_user(arg1, arg2, arg3);
>=20
> 	return (res > 0) ? -EFAULT : res;
> }

It would be unfortunate if it adds more confusion. I'd prefer to have
a sufficiently different name. memcpy_to_user/memcpy_from_user
perhaps?