From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from cthulhu.engr.sgi.com (cthulhu.engr.sgi.com [192.26.80.2]) by neteng.engr.sgi.com (971110.SGI.8.8.8/960327.SGI.AUTOCF) via SMTP id WAA179436 for ; Mon, 26 Jan 1998 22:48:01 -0800 (PST) Return-Path: Received: (from majordomo-owner@localhost) by cthulhu.engr.sgi.com (950413.SGI.8.6.12/960327.SGI.AUTOCF) id WAA08679 for linux-list; Mon, 26 Jan 1998 22:45:19 -0800 Received: from sgi.sgi.com (sgi.engr.sgi.com [192.26.80.37]) by cthulhu.engr.sgi.com (950413.SGI.8.6.12/960327.SGI.AUTOCF) via ESMTP id WAA08672; Mon, 26 Jan 1998 22:45:17 -0800 Received: from informatik.uni-koblenz.de (mailhost.uni-koblenz.de [141.26.4.1]) by sgi.sgi.com (950413.SGI.8.6.12/970507) via ESMTP id WAA06608; Mon, 26 Jan 1998 22:45:15 -0800 env-from (ralf@uni-koblenz.de) From: ralf@uni-koblenz.de Received: from uni-koblenz.de (pmport-18.uni-koblenz.de [141.26.249.18]) by informatik.uni-koblenz.de (8.8.8/8.8.8) with ESMTP id HAA29092; Tue, 27 Jan 1998 07:45:13 +0100 (MET) Received: (from ralf@localhost) by uni-koblenz.de (8.8.7/8.8.7) id HAA06851; Tue, 27 Jan 1998 07:41:55 +0100 Message-ID: <19980127074155.62514@uni-koblenz.de> Date: Tue, 27 Jan 1998 07:41:55 +0100 To: "William J. Earl" Cc: Alex deVries , SGI Linux Subject: Re: More on modules... References: <199801270151.RAA09867@fir.engr.sgi.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.85e In-Reply-To: <199801270151.RAA09867@fir.engr.sgi.com>; from William J. Earl on Mon, Jan 26, 1998 at 05:51:40PM -0800 Sender: owner-linux@cthulhu.engr.sgi.com Precedence: bulk On Mon, Jan 26, 1998 at 05:51:40PM -0800, William J. Earl wrote: > > But, I use the term 'work' with a grain of salt; the modules appear to > > load, but I get errors like: > > Illegal instruction at c000c0d0 ra=88034ed4 > > Unable to handle kernel paging request at virtual address c0008113, epc == > > c000c0d8, ra == 88034ed4 > > Killed > ... > > I wonder if the kernel is properly synchronzing the I-cache with > the D-cache after loading the module. In general, you need to > writeback the primary D-cache and invalidate the primary I-cache for > the range of addresses occupied by the driver (or simply > index-writeback-invalidate all of the D-cache and index-invalidate all > of the I-cache, if the driver is larger than the cache size). Yes, we do that properly. Alex, when compiling modules, does the compiler pass the flag -mlong-jumps? Ralf