From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263836AbTKRXKb (ORCPT ); Tue, 18 Nov 2003 18:10:31 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263839AbTKRXKa (ORCPT ); Tue, 18 Nov 2003 18:10:30 -0500 Received: from dp.samba.org ([66.70.73.150]:56979 "EHLO lists.samba.org") by vger.kernel.org with ESMTP id S263836AbTKRXKW (ORCPT ); Tue, 18 Nov 2003 18:10:22 -0500 From: Rusty Russell To: Andrew Morton From: Chris Wright Cc: James Morris , sds@epoch.ncsc.mil, aviro@redhat.com, linux-kernel@vger.kernel.org, russell@coker.com.au Subject: Re: Fw: Re: [PATCH][RFC] Remove CLONE_FILES from init kernel thread creation In-reply-to: Your message of "Mon, 17 Nov 2003 13:32:40 -0800." <20031117133240.04e9c8e8.akpm@osdl.org> Date: Tue, 18 Nov 2003 15:42:23 +1100 Message-Id: <20031118231021.D61362C242@lists.samba.org> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org In message <20031117133240.04e9c8e8.akpm@osdl.org> you write: > > Rusty? Any idea why call_usermodehelper() uses CLONE_FILES? Your guess is correct about consolidation. But there is no bug AFAICT. The "normal" (non-waiting case) uses "CLONE_VFORK | SIGCHLD". The "waiting" case uses "CLONE_KERNEL", but then that thread does another fork with just "SIGCHLD". ie. the actual thread which does the exec never uses CLONE_FILES. Hope that clarifies, Rusty. -- Anyone who quotes me in their sig is an idiot. -- Rusty Russell.