From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1750854AbVHWQyD (ORCPT ); Tue, 23 Aug 2005 12:54:03 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1750905AbVHWQyD (ORCPT ); Tue, 23 Aug 2005 12:54:03 -0400 Received: from fmr21.intel.com ([143.183.121.13]:2998 "EHLO scsfmr001.sc.intel.com") by vger.kernel.org with ESMTP id S1750854AbVHWQyC (ORCPT ); Tue, 23 Aug 2005 12:54:02 -0400 Date: Tue, 23 Aug 2005 12:57:32 -0400 From: Benjamin LaHaise To: Jakub Jelinek Cc: Suparna Bhattacharya , Ulrich Drepper , Ingo Molnar , akpm@osdl.org, linux-aio@kvack.org, linux-kernel@vger.kernel.org Subject: Re: Linux AIO status & todo Message-ID: <20050823165732.GD4630@linux.intel.com> References: <20050823074438.GA4586@in.ibm.com> <20050823095609.GZ7403@devserv.devel.redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20050823095609.GZ7403@devserv.devel.redhat.com> User-Agent: Mutt/1.4.1i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Aug 23, 2005 at 05:56:09AM -0400, Jakub Jelinek wrote: > POSIX AIO needs to handle SIGEV_NONE, SIGEV_SIGNAL and SIGEV_THREAD > notification. Obviously kernel shouldn't create threads for SIGEV_THREAD > itself, as kernel shouldn't hardcode all the implementation details how a > thread can be created. But it would be good if AIO signalling e.g. handled > both SIGEV_SIGNAL and SIGEV_SIGNAL | SIGEV_THREAD_ID, with the same usage as > e.g. timer_* syscalls. If kernel makes sure SI_ASYNCIO si_code is set in > the notification signal siginfos, glibc could even use just one helper > thread for timer_*/[al]io_* and maybe in the future other SIGEV_THREAD notification. The signal patch from Sebastien should handle the SIGEV_foo. The patch at http://www.kvack.org/~bcrl/patches/aio-2.6.13-rc6-B1/817_sigevent.diff has the latest changes from me and should do what is needed. -ben