From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753579AbYJVNj4 (ORCPT ); Wed, 22 Oct 2008 09:39:56 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752210AbYJVNje (ORCPT ); Wed, 22 Oct 2008 09:39:34 -0400 Received: from vsmtp04.dti.ne.jp ([202.216.231.139]:63948 "EHLO vsmtp04.dti.ne.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752046AbYJVNjd (ORCPT ); Wed, 22 Oct 2008 09:39:33 -0400 From: hooanon05@yahoo.co.jp To: ttb@tentacle.dhs.org, rml@novell.com cc: linux-kernel@vger.kernel.org Subject: lock free inotify_ioctl()? Date: Wed, 22 Oct 2008 22:39:30 +0900 Message-ID: <7353.1224682770@jrobl> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hello John McCutchan and Robert Love, I have a question about inotify_ioctl(). I am investigating if a user process misses the inotify event or gets wrong event by ioctl(fd, FIONREAD, &pending) in linux-2.6.27, and found it doesn't acquire dev->ev_mutex lock before accessing dev->queue_size. In SMP system, is it safe and sure too? Is the mutex lock unnecessary? Junjiro R. Okajima