From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751385AbWHDFCa (ORCPT ); Fri, 4 Aug 2006 01:02:30 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751386AbWHDFC3 (ORCPT ); Fri, 4 Aug 2006 01:02:29 -0400 Received: from fsmlabs.com ([168.103.115.128]:1999 "EHLO fsmlabs.com") by vger.kernel.org with ESMTP id S1751385AbWHDFC3 (ORCPT ); Fri, 4 Aug 2006 01:02:29 -0400 From: Cort Dougan Date: Thu, 3 Aug 2006 23:02:00 -0600 To: linux-kernel@vger.kernel.org Cc: zwane@fsmlabs.com Subject: automated Linux Regression suite Message-ID: <20060804050200.GA22923@hq.fsmlabs.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.4.1i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I saw the talk by Greg Kroah-Hartman at http://www.kroah.com/log/linux/ols_2006_keynote.html and wanted to correct a mistake (perhaps itself a myth). A regression suite, fully automated, has existed for Linux since I was running the PPC tree in the 90's. I still use an extensive and very much upgraded version here at FSMLabs. There will be a C/C++ Journal article on it in the next few months but there is a link describing it and how to re-implement it here: http://www2.fsmlabs.com/~cort/papers/sparky/sparky.pdf http://www2.fsmlabs.com/~cort/papers/sparky/sparky.nohead.html It tests FreeBSD, NetBSD and Linux for PowerPC, x86, x86-64, FRV, MIPS, ARM, XScale and IXP. It tests user-land, kernel functionality, RTLinux and RTCore along with all of our realtime drivers. To this day we're expanding it's coverage and capability. It took one long weekend to write the first version of this system.