Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751063AbWADJEU (ORCPT ); Wed, 4 Jan 2006 04:04:20 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751609AbWADJET (ORCPT ); Wed, 4 Jan 2006 04:04:19 -0500 Received: from linux01.gwdg.de ([134.76.13.21]:24810 "EHLO linux01.gwdg.de") by vger.kernel.org with ESMTP id S1751063AbWADJES (ORCPT ); Wed, 4 Jan 2006 04:04:18 -0500 Date: Wed, 4 Jan 2006 10:03:12 +0100 (MET) From: Jan Engelhardt To: =?ISO-8859-2?Q?Tomasz_K=B3oczko?= cc: Adrian Bunk , Olivier Galibert , Alistair John Strachan , Tomasz Torcz , Andi Kleen , perex@suse.cz, alsa-devel@alsa-project.org, James@superbug.demon.co.uk, sailer@ife.ee.ethz.ch, linux-sound@vger.kernel.org, zab@zabbo.net, kyle@parisc-linux.org, parisc-linux@lists.parisc-linux.org, jgarzik@pobox.com, Thorsten Knabe , zwane@commfireservices.com, zaitcev@yahoo.com, linux-kernel@vger.kernel.org Subject: Re: [2.6 patch] schedule obsolete OSS drivers for removal In-Reply-To: Message-ID: References: <20050726150837.GT3160@stusta.de> <200601031629.21765.s0348365@sms.ed.ac.uk> <20060103170316.GA12249@dspnet.fr.eu.org> <200601031716.13409.s0348365@sms.ed.ac.uk> <20060103192449.GA26030@dspnet.fr.eu.org> <20060103193736.GG3831@stusta.de> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2343 Lines: 61 > 2) ALSA API is to complicated: most applications opens single sound > stream. Seconded. > 3) ALSA kernel architecture is to complicated. This main reason why > configuring sound on Linux is SO COMPLICATED. From my system: > > # lsmod | grep ^snd | wc -l > 19 > Why would you, or the "Desktop User Joe" using Torvalds-advertised KDE, care about how many modules are loaded? Splitting code up to make things more modular is A Good Choice most of the times. There is, however, one point in which you have reason: if the overall modular structure is bigger than a mono one, then it's indeed "complicated". > ALSA is also requires much more bigger code size than OSS variant > (count all snd* modules size, jackd and libasound and compare this with > OSS modules and user space OSS library size). Simillar is on allocated > memory in all system sound components. > jackd does not count - ALSA works without it. > Many task switches incurred by the current ALSA architecture > add quickly up to perceivalble deleys during processing. In many cases > sound must be handled with RT piorites so all code sise must possible > small for handle this with minimal latency. > > 4) ALSA mixing model is UNSECURE by design because all mixing sound > streams (for example with diffrent sampling rates) are performed > in user space. > I'd rather have libasound segfault rather than my kernel oopsing, in case someone forgot a NULL check. > Also using jackd also "creates problems" with RTing this proccess and > much more bigger problems on configure stage (for joe user). > All this can be handled in secure and proprer RT prioprities > ONLY on kernel space (so all gaming mith jackd or so is plain waste > of time). Only on kernel level is possible correctly handle all this. > With ALSA you can't extend in esy way for example SELinux for prevent > intercept sound streem from microphone by remote user. Current OSS API > is much more better for SELinux. > Why ? because all mixing on OSS is performed on kernel level. > AFAICS, OSS does not do mixing at all in its current state. Jan Engelhardt -- - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/