From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S935764AbYD1PgS (ORCPT ); Mon, 28 Apr 2008 11:36:18 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S934077AbYD1Pf5 (ORCPT ); Mon, 28 Apr 2008 11:35:57 -0400 Received: from saraswathi.solana.com ([198.99.130.12]:55633 "EHLO saraswathi.solana.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933996AbYD1Pf4 (ORCPT ); Mon, 28 Apr 2008 11:35:56 -0400 Date: Mon, 28 Apr 2008 11:29:40 -0400 From: Jeff Dike To: WANG Cong Cc: akpm@osdl.org, linux-kernel@vger.kernel.org, user-mode-linux-devel@lists.sourceforge.net Subject: Re: [PATCH 6/19] UML - hppfs fixes Message-ID: <20080428152940.GA7465@c2.user-mode-linux.org> References: <20080425175608.GA11182@c2.user-mode-linux.org> <20080426.163132.194697674.xiyou.wangcong@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20080426.163132.194697674.xiyou.wangcong@gmail.com> User-Agent: Mutt/1.5.17 (2007-11-01) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sat, Apr 26, 2008 at 04:31:32PM +0800, WANG Cong wrote: > > + rem = copy_to_user(buf, &data->contents[off], count); > > + *ppos += count - rem; > > + if (rem > 0) > > + return -EFAULT; > > Could you please explain why check 'rem' after using it here? Actually, this isn't as wrong as it looks. copy_to_user returns what hasn't been copied (rem == remaining). So, I think the mistake is to return -EFAULT here. Returning the short count would be right, unless rem == count, in which case we return -EFAULT. Jeff -- Work email - jdike at linux dot intel dot com