From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 1 15:30:50 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id PAA02276 for ; Thu, 1 Jan 1998 15:30:47 GMT Received: from vger.rutgers.edu ([128.6.190.2]:11345 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <15821-625>; Thu, 1 Jan 1998 17:30:23 +0200 Received: by vger.rutgers.edu id <971014-8509>; Thu, 1 Jan 1998 10:27:29 -0500 Received: from post-10.mail.demon.net ([194.217.242.154]:49055 "HELO post.mail.demon.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970983-8509>; Thu, 1 Jan 1998 10:26:56 -0500 Received: from latrigg.demon.co.uk ([158.152.64.157]) by post.mail.demon.net id aa1021948; 1 Jan 98 15:26 GMT Date: Thu, 01 Jan 1998 11:54:49 +0000 (GMT) From: Chris Pringle Subject: Installing ARM Linux on an A5000 To: linux-arm@vger.rutgers.edu Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII X-Organization: At Home, 01285 713514 X-Mailer: ANT RISCOS Marcel [ver 1.26] X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Problem 1 : Getting ARM Linux to see my Ether3 card How do I get linux to see my Ether3 card on an A5000? It says it cannot see the device anywhere on my system. What extra arguments should I add? Problem 2 : Installing ARM Linux using a Local Hard Disk image How do I install ARM Linux using a local Hard Disk image? If it can read my ADFS Drives do I have to put the files in a image file with long filenames and no limitation to directory size? If so how do I create them, do I use PCem or is the partition type different? From what I can remember DOSDisc's can not have long file names. Problem 3 : Getting ARM Linux to see my Oak SCSI card How do I get linux to see my Oak SCSI 1 card with a PowerROM upgrade (to make it use a SCSI 2 command set). Again it says it cannot see the device anywhere on my system. What extra arguments should I include. I have a SCSI CDROM Drive attached and a SCSI Viper 150 tape streamer. Any solutions to these problems would be appreciated. Could all solutions be sent directly to me as I don't know if I am yet suscribed to the mailing list. Regards, --- Chris Pringle 22, St. Mary's Drive, Fairford, Glos. GL7 4LQ P.S.Check out My web site at http://www.latrigg.demon.co.uk/Chris.homepage.html unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 1 15:54:51 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id PAA02442 for ; Thu, 1 Jan 1998 15:54:50 GMT Received: from vger.rutgers.edu ([128.6.190.2]:2148 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <15707-626>; Thu, 1 Jan 1998 17:55:01 +0200 Received: by vger.rutgers.edu id <970983-8509>; Thu, 1 Jan 1998 10:51:58 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:2275 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <970890-8512>; Thu, 1 Jan 1998 10:51:23 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xnncN-00049Q-00; Thu, 1 Jan 1998 16:37:11 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Chris Pringle cc: linux-arm@vger.rutgers.edu Subject: Re: Installing ARM Linux on an A5000 In-reply-to: Your message of "Thu, 01 Jan 1998 11:54:49 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 01 Jan 1998 16:37:10 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >Problem 1 : Getting ARM Linux to see my Ether3 card >How do I get linux to see my Ether3 card on an A5000? It says it cannot >see the device anywhere on my system. What extra arguments should I add? What are the exact messages you get? My A5000 detects its Ether3 OK. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 1 21:15:14 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id VAA04509 for ; Thu, 1 Jan 1998 21:15:13 GMT Received: from vger.rutgers.edu ([128.6.190.2]:10283 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <15719-627>; Thu, 1 Jan 1998 23:14:37 +0200 Received: by vger.rutgers.edu id <971484-8509>; Thu, 1 Jan 1998 16:11:22 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:1075 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <971395-8512>; Thu, 1 Jan 1998 16:09:54 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id VAA04328; Thu, 1 Jan 1998 21:10:41 GMT From: Matthew Wilcox Message-Id: <199801012110.VAA04328@odie.barnet.ac.uk> Subject: Re: Installing ARM Linux on an A5000 To: cpringle@latrigg.demon.co.uk (Chris Pringle) Date: Thu, 1 Jan 1998 21:10:38 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: from "Chris Pringle" at Jan 1, 98 11:54:49 am X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > Problem 2 : Installing ARM Linux using a Local Hard Disk image > How do I install ARM Linux using a local Hard Disk image? If it can read my > ADFS Drives do I have to put the files in a image file with long filenames > and no limitation to directory size? If so how do I create them, do I use > PCem or is the partition type different? From what I can remember DOSDisc's can > not have long file names. You don't actually *need* long file names. As long as they are unique, the rpm manager can make do. > Problem 3 : Getting ARM Linux to see my Oak SCSI card > How do I get linux to see my Oak SCSI 1 card with a PowerROM upgrade (to make > it use a SCSI 2 command set). Again it says it cannot see the device anywhere > on my system. What extra arguments should I include. I have a SCSI CDROM Drive > attached and a SCSI Viper 150 tape streamer. Ah. Oh dear. The same sort of problem affects RiscBSD.. let me explain as well as I understand the problem: The ROM on a podule contains a manufacturer ID & product code. When you replace the Oak ROM with a PowerROM, the manufacturer & product code change. This is a Bad Thing and Alsystems have received wrist-slappings from the Causality over this before. What needs to happen is that the Oak driver has to be modified to recognise either the original or the new powerrom. I suspect you might be able to help here by telling us what manufacturer and product codes Linux reports during bootup. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 2 00:15:45 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id AAA05674 for ; Fri, 2 Jan 1998 00:15:44 GMT Received: from vger.rutgers.edu ([128.6.190.2]:10348 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12200-623>; Fri, 2 Jan 1998 02:15:04 +0200 Received: by vger.rutgers.edu id <971334-8512>; Thu, 1 Jan 1998 19:05:41 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:2727 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <971193-8512>; Thu, 1 Jan 1998 19:00:33 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xnqPQ-0004Mj-00; Thu, 1 Jan 1998 19:36:00 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Chris Pringle Cc: linux-arm@vger.rutgers.edu Subject: Re: Installing ARM Linux on an A5000 In-reply-to: Your message of "Thu, 01 Jan 1998 16:38:13 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 01 Jan 1998 19:36:00 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >Thanks for replying, it just says "Error > I cannot find the device anywhere on your system" Ah. Can you look at the messages you get when the kernel starts, and see if it seems to be detecting the card then? >I have not tried adding extra arguments because I don't know any. A the moment >I am using a RiscPC boot disk, this may be the problem, but I do not have an >a5k boot disk and do not know where to get one. Any Ideas? I don't know enough about how Russell's installer works (never having used it) to say, really. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sat Jan 3 18:59:01 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id SAA21300 for ; Sat, 3 Jan 1998 18:59:00 GMT Received: from vger.rutgers.edu ([128.6.190.2]:16158 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <16334-627>; Sat, 3 Jan 1998 20:59:02 +0200 Received: by vger.rutgers.edu id <971798-637>; Sat, 3 Jan 1998 13:32:48 -0500 Received: from post-20.mail.demon.net ([194.217.242.27]:59790 "HELO post.mail.demon.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <971418-634>; Sat, 3 Jan 1998 13:28:55 -0500 Received: from latrigg.demon.co.uk ([158.152.64.157]) by post.mail.demon.net id aa2007004; 3 Jan 98 18:04 GMT Date: Sat, 03 Jan 1998 18:04:20 +0000 (GMT) From: Chris Pringle Subject: A5k bootdisk To: linux-arm@vger.rutgers.edu Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII X-Organization: At Home, 01285 713514 X-Mailer: ANT RISCOS Marcel [ver 1.26] X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Does anyone know where I can obtain the a5k bootdisk? Regards, --- Chris Pringle 22, St. Mary's Drive, Fairford, Glos. GL7 4LQ P.S.Check out My web site at http://www.latrigg.demon.co.uk/Chris.homepage.html unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 4 01:56:17 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id BAA23847 for ; Sun, 4 Jan 1998 01:56:16 GMT Received: from vger.rutgers.edu ([128.6.190.2]:8462 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11296-626>; Sun, 4 Jan 1998 03:53:56 +0200 Received: by vger.rutgers.edu id <971367-637>; Sat, 3 Jan 1998 16:37:01 -0500 Received: from dialup.smilie.mpx.com.au ([203.10.66.177]:18498 "EHLO intrepid.infotainment.com.au" ident: "root") by vger.rutgers.edu with ESMTP id <972310-637>; Sat, 3 Jan 1998 14:24:02 -0500 Received: from localhost (smilie@localhost) by intrepid.infotainment.com.au (8.8.5/8.8.5) with SMTP id QAA32582 for ; Fri, 2 Jan 1998 16:57:38 +1100 Date: Fri, 2 Jan 1998 16:57:38 +1100 (EST) From: Anthony Rumble To: linux-arm@vger.rutgers.edu Subject: Getting an EBSA board.. or possibly DNARD board Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO I am very keen in doing some work with linux on an SA110 platform, and was wondering if anyone know of a good source for a one off EBSA or DNARD based board. --- Anthony Rumble Interactive Infotainment Systems P/L ACN: 076 046 252 Phone 02-9798-7604 Mobile 015-955-042 Pager 016-634-997 unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 4 13:52:47 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id NAA27874 for ; Sun, 4 Jan 1998 13:52:46 GMT Received: from vger.rutgers.edu ([128.6.190.2]:48476 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13375-627>; Sun, 4 Jan 1998 15:49:06 +0200 Received: by vger.rutgers.edu id <970980-634>; Sun, 4 Jan 1998 07:15:15 -0500 Received: from methan.in-berlin.de ([160.45.10.13]:3006 "HELO methan.in-berlin.de" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970979-637>; Sun, 4 Jan 1998 07:13:30 -0500 Received: by methan.in-berlin.de (Smail3.2.0.98) from bolzen.in-berlin.de (194.94.235.1) with esmtp id ; Sun, 4 Jan 1998 13:14:24 +0100 (MET) Received: by bolzen.in-berlin.de (Smail3.2) from yoda with uucp id ; Sun, 4 Jan 1998 13:13:54 +0100 (MET) Received: by yoda.in-berlin.de (Smail3.1.28.1 #0) id m0xopqr-000000C; Sun, 4 Jan 98 12:12 GMT Date: Sun, 04 Jan 1998 12:20:57 +0100 From: Stefan Voss To: linux-arm@vger.rutgers.edu Subject: problem with installling over NFS (and ftp) Message-ID: Organization: Individual Network (IN), Berlin, Germany X-Mailer: Messenger v1.10 for RISC OS X-Posting-Agent: RISC OS Newsbase 0.59f X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi folks, I hope this is the right place to post to. In the last few days I tried to install ARMLinux. I decided to go with the NFS option using my Linux laptop as NFS server. Downloading the software, partioning the IDE disk (Quantum Fireball), creating the boot disks and booting with the boot disks worked well. My hard disk and network card (I-Cubed EtherH) were detected correctly. I get the following error after specifying the IP address of the NFS server and NFS directory: Internal error: Oops: 0 CPU: 0 pc : [] lr : [] sp : c1460ec4 ip : c1460f1c fp : c1460f18 r10: 00000000 r9 : 0001c000 r8 : 00018f80 r7 : 0000007a r6 : c1601598 r5 : 00000000 r4 : c17f0be8 r3 : 00000001 r2 : c2049000 r1 : 00000006 r0 : c204ef6c Flags: nZCv IRQs on FIQs on Mode SVC_32 Process insmod (pid: 12, stackpage=c1460000) Stack: Backtrace: Function entered at [] from [] Function entered at [] from [] Function entered at [] from [] Code: pc not in code space My configuration is: Risc PC with SA110, 24 MB RAM, 2 MB VRAM, RISCOS 3.7, PowerTec SCSI card, I-Cubed EtherH, Quantum Fireball 4.3 GB (1 GB RISCOS, 2 GB Linux, 1 GB free) The mounting is the last step that seems to work: darkstar:/root # showmount -e Export list for darkstar: /opt /home/voss darkstar:/root # showmount -a All mount points on darkstar: yoda.in-berlin.de:/home yoda.in-berlin.de:/home/voss yoda.in-berlin.de:/home/voss/RedHat yoda.in-berlin.de:/home/voss/armlinux "darkstar" is the Linux NFS server. "yoda" is the ARMLinux machine. On the NFS server the RedHat directory looks like this: voss@darkstar:/home/voss/RedHat > ls -l total 11 drwxr-xr-x 2 voss users 9216 Dec 21 09:22 RPMS drwxr-xr-x 2 voss users 1024 Dec 21 10:06 base drwxr-xr-x 5 voss users 1024 Dec 21 10:07 instimage voss@darkstar:/home/voss/RedHat > ls -l instimage/usr/bin/insmod -rwxr-xr-x 1 voss users 16064 Dec 21 10:10 instimage/usr/bin/insmod To me it seems as if "insmod" is producing a segmentation fault but I don't know if "insmod" has been executed from the NFS directory or from the boot disk. I also tried installing using ftp. The installation stops just after an sucessful ftp login. I can see this on my ftp server's log file. There's no visible sign of a crash. It just stops with an open ftp session. Any hints? Regards, Stefan -- Stefan Voss (voss@yoda.in-berlin.de) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 4 18:30:32 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id SAA29639 for ; Sun, 4 Jan 1998 18:30:31 GMT Received: from vger.rutgers.edu ([128.6.190.2]:48476 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <15703-626>; Sun, 4 Jan 1998 20:29:05 +0200 Received: by vger.rutgers.edu id <970898-634>; Sun, 4 Jan 1998 08:54:55 -0500 Received: from post-10.mail.demon.net ([194.217.242.154]:43113 "HELO post.mail.demon.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970896-637>; Sun, 4 Jan 1998 08:54:00 -0500 Received: from treblig.demon.co.uk ([194.222.210.232]) by post.mail.demon.net id aa1028460; 4 Jan 98 13:43 GMT Date: Sun, 4 Jan 1998 13:39:46 +0000 (GMT) From: Dave Gilbert X-Sender: gilbertd@tardis.home.dave To: Anthony Rumble cc: linux-arm@vger.rutgers.edu Subject: Re: Getting an EBSA board.. or possibly DNARD board In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO On Fri, 2 Jan 1998, Anthony Rumble wrote: > > I am very keen in doing some work with linux on an SA110 platform, and was > wondering if anyone know of a good source for a one off EBSA or DNARD > based board. What is DNARD? Dave ---------------------------------------------------- Man can not live - Dr. David Alan Gilbert - gro.gilbert @ treblig.org by bread alone. He- ---------------------------------------------------- needs chocolate. - unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 4 18:33:01 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id SAA29647 for ; Sun, 4 Jan 1998 18:33:00 GMT Received: from vger.rutgers.edu ([128.6.190.2]:48476 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <15561-625>; Sun, 4 Jan 1998 20:31:21 +0200 Received: by vger.rutgers.edu id <970940-637>; Sun, 4 Jan 1998 09:01:49 -0500 Received: from dialup.smilie.mpx.com.au ([203.10.66.177]:24745 "EHLO intrepid.infotainment.com.au" ident: "smilie") by vger.rutgers.edu with ESMTP id <970923-634>; Sun, 4 Jan 1998 09:01:11 -0500 Received: from localhost (smilie@localhost) by intrepid.infotainment.com.au (8.8.5/8.8.5) with SMTP id XAA13041; Sun, 4 Jan 1998 23:56:24 +1100 Date: Sun, 4 Jan 1998 23:56:20 +1100 (EST) From: Anthony Rumble To: Dave Gilbert cc: linux-arm@vger.rutgers.edu Subject: Re: Getting an EBSA board.. or possibly DNARD board In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO On Sun, 4 Jan 1998, Dave Gilbert wrote: > On Fri, 2 Jan 1998, Anthony Rumble wrote: > > > I am very keen in doing some work with linux on an SA110 platform, and was > > wondering if anyone know of a good source for a one off EBSA or DNARD > > based board. > > What is DNARD? Digital Network Appliance Reference Design http://www.research.digital.com/SRC/iag/info/new-home.html Basically.. it's Digitals NC in a box.. with ALL the specs given away .. It even includes the specs for the packeging.. the EBSA I think is the Evaluation Board StrongARM.. I could be wrong.. I've asked the peopel within Digital where to get one.. and they haven't been much use so far.. and now that the chips production has now all been sold to Intel...... Anyway.. I have a number of NC projects I want to start working on, which is why I want to try and get one of these beasts ASAP.. Ive mailed every customer on digitals pages who say there working on a production unit where and when I can get one.. I haven't yet gotten any satisfactory responce.. What have I gotta do? Jump up and down with cash in my hand? --- Anthony Rumble Interactive Infotainment Systems P/L ACN: 076 046 252 Phone 02-9798-7604 Mobile 015-955-042 Pager 016-634-997 unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 4 20:24:32 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id UAA30333 for ; Sun, 4 Jan 1998 20:24:30 GMT Received: from vger.rutgers.edu ([128.6.190.2]:48476 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <15719-626>; Sun, 4 Jan 1998 22:18:54 +0200 Received: by vger.rutgers.edu id <970922-634>; Sun, 4 Jan 1998 10:03:55 -0500 Received: from hades.itl.net ([194.145.1.31]:57855 "HELO super.net.uk" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970963-637>; Sun, 4 Jan 1998 10:02:11 -0500 Received: from pingu.itl.net [194.145.15.252] by super.net.uk with smtp (Exim 1.62 #1) id 0xorbY-0005q5-00; Sun, 4 Jan 1998 15:04:45 +0000 Message-Id: <1.5.4.32.19980104145715.00669ae8@digiserve.com> X-Sender: garan@digiserve.com (Unverified) X-Mailer: Windows Eudora Light Version 1.5.4 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sun, 04 Jan 1998 14:57:15 +0000 To: linux-arm@vger.rutgers.edu From: G a r a n Subject: bootloader X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi A butchered copy of the formatter from !MOFS and also of riscbsd_bb and !Partman would finally partition my scsi disc. I've tried wiping the disc many times since, and it won't be partitioned until I use BOTH the changed versions of the above... weird. However, in the FAQ, to install the bootloader I use the -bootkernel argument again but then specify root=/dev/sda3 ( in my case) in the extra arguments. However, I assume acornscsi_mod.o isn't loaded, so how would I load this at the same time, or do I need acornscsi compiled into my kernel? Currently it complains, which is understandable. Any help greatly appreciated. Thanks Garan -- garan @ digiserve . com garanj @ etoncomp . demon . co . uk unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 4 21:32:39 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id VAA30777 for ; Sun, 4 Jan 1998 21:32:37 GMT Received: from vger.rutgers.edu ([128.6.190.2]:48476 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <9246-623>; Sun, 4 Jan 1998 23:32:11 +0200 Received: by vger.rutgers.edu id <971285-637>; Sun, 4 Jan 1998 10:32:38 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1238 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971291-634>; Sun, 4 Jan 1998 10:25:39 -0500 Received: from caramon.armlinux.org (dialup1.cymru.net [163.164.160.65]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id PAA27846; Sun, 4 Jan 1998 15:24:46 GMT Received: from raistlin.armlinux.org (raistlin [192.168.0.3]) by caramon.armlinux.org (8.7.4/8.7.3) with ESMTP id PAA05635; Sun, 4 Jan 1998 15:25:56 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.armlinux.org (8.7.4/8.7.3) id PAA00885; Sun, 4 Jan 1998 15:25:53 GMT Message-Id: <199801041525.PAA00885@raistlin.armlinux.org> Subject: Re: problem with installling over NFS (and ftp) To: voss@yoda.in-berlin.de (Stefan Voss) Date: Sun, 4 Jan 1998 15:25:51 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: from "Stefan Voss" at Jan 4, 98 12:20:57 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Stefan Voss writes: > I get the following error after specifying the IP address of the NFS > server and NFS directory: > > Internal error: Oops: 0 > CPU: 0 > pc : [] > lr : [] You need to get the latest version of the insmod utilities (and installation disks) off the FTP site. PS. The tell tail sign for this problem is the pc=0xc4........ with lr=0xc2........ _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 5 06:55:12 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id GAA01547 for ; Mon, 5 Jan 1998 06:55:11 GMT Received: from vger.rutgers.edu ([128.6.190.2]:48476 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13726-150>; Sun, 4 Jan 1998 23:44:47 +0200 Received: by vger.rutgers.edu id <971660-637>; Sun, 4 Jan 1998 12:13:28 -0500 Received: from post-10.mail.demon.net ([194.217.242.154]:38210 "HELO post.mail.demon.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <971655-637>; Sun, 4 Jan 1998 12:12:51 -0500 Received: from treblig.demon.co.uk ([194.222.210.232]) by post.mail.demon.net id ab1021863; 4 Jan 98 17:12 GMT Date: Sun, 4 Jan 1998 17:09:04 +0000 (GMT) From: Dave Gilbert X-Sender: gilbertd@tardis.home.dave To: Anthony Rumble cc: linux-arm@vger.rutgers.edu Subject: Re: Getting an EBSA board.. or possibly DNARD board In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO On Sun, 4 Jan 1998, Anthony Rumble wrote: > On Sun, 4 Jan 1998, Dave Gilbert wrote: > > > On Fri, 2 Jan 1998, Anthony Rumble wrote: > > > > > I am very keen in doing some work with linux on an SA110 platform, and was > > > wondering if anyone know of a good source for a one off EBSA or DNARD > > > based board. > > > > What is DNARD? > > Digital Network Appliance Reference Design > http://www.research.digital.com/SRC/iag/info/new-home.html Ah - one of those! > I've asked the peopel within Digital where to get one.. and they haven't > been much use so far.. and now that the chips production has now all been > sold to Intel...... Don't worry about that; Digital won't actually sell you anything - or tell you anything useful like where to buy one. In actual fact its rumoured that its impossible to buy anything that Digital make :-) Look at Digital's web page and look for the distributor for your country; phone them up and quote the exact part number for the EBSA-285 from DECs web page - and if your lucky they'll be able to sell you one. Last time I asked they said there was an 8 week leed time - that should be about now! So yes you should be able to get one (that was talking to Eurodis/Bytech - the UK distributor). Prepare to fight your way through to someone who knows. > What have I gotta do? Jump up and down with cash in my hand? YOu think that will help - ha! Dave ---------------------------------------------------- Man can not live - Dr. David Alan Gilbert - gro.gilbert @ treblig.org by bread alone. He- ---------------------------------------------------- needs chocolate. - unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 5 06:55:16 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id GAA01550 for ; Mon, 5 Jan 1998 06:55:13 GMT Received: from vger.rutgers.edu ([128.6.190.2]:48476 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <4953-21270>; Mon, 5 Jan 1998 00:01:13 +0200 Received: by vger.rutgers.edu id <971692-637>; Sun, 4 Jan 1998 15:10:14 -0500 Received: from dialup.smilie.mpx.com.au ([203.10.66.177]:25314 "EHLO intrepid.infotainment.com.au" ident: "smilie") by vger.rutgers.edu with ESMTP id <970976-634>; Sun, 4 Jan 1998 15:09:00 -0500 Received: from localhost (smilie@localhost) by intrepid.infotainment.com.au (8.8.5/8.8.5) with SMTP id GAA14428; Mon, 5 Jan 1998 06:03:37 +1100 Date: Mon, 5 Jan 1998 06:03:36 +1100 (EST) From: Anthony Rumble To: Dave Gilbert cc: linux-arm@vger.rutgers.edu Subject: Re: Getting an EBSA board.. or possibly DNARD board In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO On Sun, 4 Jan 1998, Dave Gilbert wrote: > On Sun, 4 Jan 1998, Anthony Rumble wrote: > > > On Sun, 4 Jan 1998, Dave Gilbert wrote: > > > > > On Fri, 2 Jan 1998, Anthony Rumble wrote: > > > > > > > I am very keen in doing some work with linux on an SA110 platform, and was > > > > wondering if anyone know of a good source for a one off EBSA or DNARD > > > > based board. > > > > > > What is DNARD? > > > > Digital Network Appliance Reference Design > > http://www.research.digital.com/SRC/iag/info/new-home.html > > Ah - one of those! > > > I've asked the peopel within Digital where to get one.. and they haven't > > been much use so far.. and now that the chips production has now all been > > sold to Intel...... > > Don't worry about that; Digital won't actually sell you anything - or > tell you anything useful like where to buy one. In actual fact its > rumoured that its impossible to buy anything that Digital make :-) > > Look at Digital's web page and look for the distributor for your country; > phone them up and quote the exact part number for the EBSA-285 from DECs > web page - and if your lucky they'll be able to sell you one. > Last time I asked they said there was an 8 week leed time - that should be > about now! So yes you should be able to get one (that was talking to > Eurodis/Bytech - the UK distributor). > Prepare to fight your way through to someone who knows. Talking to some "Expert Digital Consumers".. they say it's all in knowing the part number. The problem is not knowing where to buy one, but HOW to buy one... We aparantly don't know how.. If you get the full localised part number (rumered to be at least 32 characters long).. you can order the product.. --- Anthony Rumble Interactive Infotainment Systems P/L ACN: 076 046 252 Phone 02-9798-7604 Mobile 015-955-042 Pager 016-634-997 unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 5 06:55:20 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id GAA01556 for ; Mon, 5 Jan 1998 06:55:18 GMT Received: from vger.rutgers.edu ([128.6.190.2]:14876 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10304-27083>; Mon, 5 Jan 1998 01:56:39 +0200 Received: by vger.rutgers.edu id <970793-12312>; Sun, 4 Jan 1998 18:54:55 -0500 Received: from enterprise.ce.com.au ([203.23.60.66]:12934 "EHLO enterprise.ce.com.au" ident: "SOCKWRITE-65") by vger.rutgers.edu with ESMTP id <970791-12312>; Sun, 4 Jan 1998 18:54:42 -0500 Received: from borg.in.ce.com.au by enterprise.ce.com.au with smtp (Smail3.2.0.90 #2) id m0xoztR-001q1yC; Mon, 5 Jan 1998 10:55:45 +1100 (EST) Date: Mon, 5 Jan 1998 09:55:52 +1100 (EST) From: Anthony Rumble X-Sender: smilie@borg.in.ce.com.au To: linux-arm@vger.rutgers.edu Subject: Corel Video Network computer Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Is this port their own? or based on the current ARM port? I was just wondering.. since they claim it's an SA110 based machine running linux, and aparantly near production? If it's not the Current ARM port, who did their port? How did they do it so quickly without anyone knowing? --- Anthony Rumble Interactive Infotainment Systems P/L ACN: 076 046 252 Phone 02-9798-7604 Mobile 015-955-042 Pager 016-634-997 unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 5 12:50:26 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id MAA04356 for ; Mon, 5 Jan 1998 12:50:25 GMT Received: from vger.rutgers.edu ([128.6.190.2]:35893 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <1256-150>; Mon, 5 Jan 1998 14:48:27 +0200 Received: by vger.rutgers.edu id <970956-12312>; Mon, 5 Jan 1998 07:46:23 -0500 Received: from methan.in-berlin.de ([160.45.10.13]:4730 "HELO methan.in-berlin.de" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970957-12312>; Mon, 5 Jan 1998 07:44:20 -0500 Received: by methan.in-berlin.de (Smail3.2.0.98) from bolzen.in-berlin.de (194.94.235.1) with esmtp id ; Mon, 5 Jan 1998 13:45:10 +0100 (MET) Received: by bolzen.in-berlin.de (Smail3.2) from yoda with uucp id ; Mon, 5 Jan 1998 13:44:53 +0100 (MET) Received: by yoda.in-berlin.de (Smail3.1.28.1 #0) id m0xpCpV-000000C; Mon, 5 Jan 98 12:44 GMT Date: Mon, 05 Jan 1998 13:13:06 +0100 From: Stefan Voss To: linux-arm@vger.rutgers.edu Subject: Re: problem with installling over NFS (and ftp) Message-ID: <2a2aa348%voss@yoda.in-berlin.de> In-Reply-To: <199801041525.PAA00885@raistlin.armlinux.org> Organization: Individual Network (IN), Berlin, Germany X-Mailer: Messenger v1.10 for RISC OS X-Posting-Agent: RISC OS Newsbase 0.59f X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO In message <199801041525.PAA00885@raistlin.armlinux.org>, Russell wrote: > Stefan Voss writes: > > I get the following error after specifying the IP address of the NFS > > server and NFS directory: > > > > Internal error: Oops: 0 > > CPU: 0 > > pc : [] > > lr : [] > > You need to get the latest version of the insmod utilities (and installation > disks) off the FTP site. I'm a bit confused what insmod you are talking about. There's an insmod binary in the instimage installation tree, on the root disk and on the supplemental disk. I downloaded the insmod binary from the instimage installation tree again but this was the same that I already had (according to "diff" and "sum") I also downloaded the root-rpc and supplemental disks again from ftp.arm.uk.linux.org /pub/armlinux/distrib/RedHat/disks and used !WriteDisk to re-create the boot disks. The result however remains the same. I still get the error described above. I also tried a 2.0.30 kernel instead of the latest 2.0.31 but the install programs says that it doesn't know the EtherH device. The boot messages however seem to indicate that the EtherH card is detected correctly. > > PS. The tell tail sign for this problem is the pc=0xc4........ with lr=0xc2........ Which of the insmod binaries is supposed to be executing at the time when the error occurrs? I noticed that the insmod binaries are not the same. If I mount the root disk on my Linux laptop (mount /dev/fd0 /mnt) I get: voss@darkstar:/home/voss > ls -l /mnt/sbin/insmod -rwxr-xr-x 1 root root 106432 Jun 29 1997 /mnt/sbin/insmod If I mount the supplemental disk I get: voss@darkstar:/home/voss > ls -l /mnt/usr/bin/insmod -rwxr-xr-x 1 root root 9396 Jun 29 1997 /mnt/usr/bin/insmod voss@darkstar:/home/voss > file /mnt/usr/bin/insmod /mnt/usr/bin/insmod: gzip compressed data, deflated, last modified: Sun Jun 29 23:54:10 1997, max compression, os: Unix And finally the binary from the instimage tree: voss@darkstar:/home/voss > ls -l RedHat/instimage/usr/bin/insmod -rwxr-xr-x 1 voss users 16064 Dec 21 10:10 RedHat/instimage/usr/bin/insmod voss@darkstar:/home/voss > file RedHat/instimage/usr/bin/insmod RedHat/instimage/usr/bin/insmod: PDP-11 executable not stripped > | | Russell King rmk@ecs.soton.ac.uk --- --- Regards, Stefan -- Stefan Voss (voss@yoda.in-berlin.de) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 5 13:25:33 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id NAA04606 for ; Mon, 5 Jan 1998 13:25:31 GMT Received: from vger.rutgers.edu ([128.6.190.2]:50017 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <19987-150>; Mon, 5 Jan 1998 15:22:58 +0200 Received: by vger.rutgers.edu id <970951-12312>; Mon, 5 Jan 1998 08:20:32 -0500 Received: from server21.digital.fr ([193.56.15.21]:4203 "EHLO server21.digital.fr" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970953-12312>; Mon, 5 Jan 1998 08:20:09 -0500 Received: from mail.vbo.dec.com (mail.vbo.dec.com [16.36.208.34]) by server21.digital.fr (8.7.5/8.7) with ESMTP id OAA16445; Mon, 5 Jan 1998 14:21:29 +0100 (MET) Received: from vbormc.vbo.dec.com (vbormc.vbo.dec.com [16.36.208.94]) by mail.vbo.dec.com (8.8.6/8.7) with ESMTP id OAA24481; Mon, 5 Jan 1998 14:21:29 +0100 (MET) Received: from linux.reo.dec.com (linux.reo.dec.com [16.36.0.79]) by vbormc.vbo.dec.com (8.7.3/8.7) with ESMTP id PAA04306; Mon, 5 Jan 1998 15:04:39 +0100 Received: from linux.reo.dec.com (localhost [127.0.0.1]) by linux.reo.dec.com (8.8.5/8.7.1) with ESMTP id NAA19259; Mon, 5 Jan 1998 13:19:27 GMT Message-Id: <199801051319.NAA19259@linux.reo.dec.com> To: Anthony Rumble cc: linux-arm@vger.rutgers.edu Subject: Re: Corel Video Network computer In-reply-to: Your message of "Mon, 05 Jan 1998 09:55:52 +1100." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 05 Jan 1998 13:19:27 +0000 From: "Ka'Plaagh" X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > > Is this port their own? or based on the current ARM port? > It was based on the 2.0.29 port from Russell. > I was just wondering.. since they claim it's an SA110 based machine > running linux, and aparantly near production? > > If it's not the Current ARM port, who did their port? How did they do it > so quickly without anyone knowing? > They didn't tell too many people about it. They have folded their sources back to Russell - so it should all become mainstream soon. > --- > Anthony Rumble > Interactive Infotainment Systems P/L ACN: 076 046 252 > Phone 02-9798-7604 Mobile 015-955-042 Pager 016-634-997 > > unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu ---------------------------------------------------------------------- David A Rusling Consulting Engineer European Semiconductor Applications Digital Equipment Co Ltd., Engineering PO Box 121, Imperial Way, Worton Grange Reading RG2 0TU Linux, Alpha, StrongArm, PCI Tel: UK-(0)1734-204380 Fax: UK-(0)1734-203133 ---------------------------------------------------------------------- unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 5 13:27:42 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id NAA04610 for ; Mon, 5 Jan 1998 13:27:41 GMT Received: from vger.rutgers.edu ([128.6.190.2]:50017 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10725-27083>; Mon, 5 Jan 1998 15:21:49 +0200 Received: by vger.rutgers.edu id <970951-12312>; Mon, 5 Jan 1998 08:19:50 -0500 Received: from server21.digital.fr ([193.56.15.21]:4193 "EHLO server21.digital.fr" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970959-12312>; Mon, 5 Jan 1998 08:19:34 -0500 Received: from mail.vbo.dec.com (mail.vbo.dec.com [16.36.208.34]) by server21.digital.fr (8.7.5/8.7) with ESMTP id OAA10280; Mon, 5 Jan 1998 14:20:06 +0100 (MET) Received: from vbormc.vbo.dec.com (vbormc.vbo.dec.com [16.36.208.94]) by mail.vbo.dec.com (8.8.6/8.7) with ESMTP id OAA24738; Mon, 5 Jan 1998 14:20:05 +0100 (MET) Received: from linux.reo.dec.com (linux.reo.dec.com [16.36.0.79]) by vbormc.vbo.dec.com (8.7.3/8.7) with ESMTP id PAA04221; Mon, 5 Jan 1998 15:03:16 +0100 Received: from linux.reo.dec.com (localhost [127.0.0.1]) by linux.reo.dec.com (8.8.5/8.7.1) with ESMTP id NAA19244; Mon, 5 Jan 1998 13:18:02 GMT Message-Id: <199801051318.NAA19244@linux.reo.dec.com> To: Anthony Rumble cc: Dave Gilbert , linux-arm@vger.rutgers.edu Subject: Re: Getting an EBSA board.. or possibly DNARD board In-reply-to: Your message of "Sun, 04 Jan 1998 23:56:20 +1100." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 05 Jan 1998 13:18:02 +0000 From: "Ka'Plaagh" X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > On Sun, 4 Jan 1998, Dave Gilbert wrote: > > > On Fri, 2 Jan 1998, Anthony Rumble wrote: > > > > > I am very keen in doing some work with linux on an SA110 platform, and was > > > wondering if anyone know of a good source for a one off EBSA or DNARD > > > based board. > > > > What is DNARD? > > Digital Network Appliance Reference Design > http://www.research.digital.com/SRC/iag/info/new-home.html > > Basically.. it's Digitals NC in a box.. with ALL the specs given away .. > It even includes the specs for the packeging.. > Correct, it's also known as SHARK. > the EBSA I think is the Evaluation Board StrongARM.. I could be wrong.. > Yup, Evaluation Board StrongARM. I guess that they could easily be called SAEBs. > I've asked the peopel within Digital where to get one.. and they haven't > been much use so far.. and now that the chips production has now all been > sold to Intel...... > Tell me about it. By the by, I haven't answered because I only just saw this mail. There again, I don't work at all between Christmas and New Year (ever). > Anyway.. I have a number of NC projects I want to start working on, which > is why I want to try and get one of these beasts ASAP.. > DNARD is hard to get but is being made (not by Digital anymore). > Ive mailed every customer on digitals pages who say there working on a > production unit where and when I can get one.. I haven't yet gotten any > satisfactory responce.. > > What have I gotta do? Jump up and down with cash in my hand? > The latest EBSA is the EBSA-285. It is available in quantity now, all of our suppliers (see the Digital Semiconductor web site). Linux is not yet (quite) running on it but Russell should have his system by now, so you never know. As a system it looks a lot like Corel's 21285 based SA-110 system and very like Causality's upcoming system. > --- > Anthony Rumble > Interactive Infotainment Systems P/L ACN: 076 046 252 > Phone 02-9798-7604 Mobile 015-955-042 Pager 016-634-997 > > unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu ---------------------------------------------------------------------- David A Rusling Consulting Engineer European Semiconductor Applications Digital Equipment Co Ltd., Engineering PO Box 121, Imperial Way, Worton Grange Reading RG2 0TU Linux, Alpha, StrongArm, PCI Tel: UK-(0)1734-204380 Fax: UK-(0)1734-203133 ---------------------------------------------------------------------- unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 5 13:47:26 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id NAA04820 for ; Mon, 5 Jan 1998 13:47:25 GMT Received: from vger.rutgers.edu ([128.6.190.2]:20484 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11780-150>; Mon, 5 Jan 1998 15:41:48 +0200 Received: by vger.rutgers.edu id <970958-12312>; Mon, 5 Jan 1998 08:37:18 -0500 Received: from dialup.smilie.mpx.com.au ([203.10.66.177]:32478 "EHLO intrepid.infotainment.com.au" ident: "smilie") by vger.rutgers.edu with ESMTP id <970960-12312>; Mon, 5 Jan 1998 08:26:52 -0500 Received: from localhost (smilie@localhost) by intrepid.infotainment.com.au (8.8.5/8.8.5) with SMTP id XAA18281; Mon, 5 Jan 1998 23:22:49 +1100 Date: Mon, 5 Jan 1998 23:22:49 +1100 (EST) From: Anthony Rumble Reply-To: Anthony Rumble To: "Ka'Plaagh" cc: linux-arm@vger.rutgers.edu Subject: Re: Getting an EBSA board.. or possibly DNARD board In-Reply-To: <199801051318.NAA19244@linux.reo.dec.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO On Mon, 5 Jan 1998, Ka'Plaagh wrote: > > I've asked the peopel within Digital where to get one.. and they haven't > > been much use so far.. and now that the chips production has now all been > > sold to Intel...... > > > > Tell me about it. By the by, I haven't answered because I only just saw this > mail. There again, I don't work at all between Christmas and New Year (ever). I have some Order numbers.. Im not sure if there the correct ones.. I have EC-R6MSB-TE for the EBSA-285 Evaluation Board and EC-R8JLB-TE for the DNARD.. Are these correct? > > Anyway.. I have a number of NC projects I want to start working on, which > > is why I want to try and get one of these beasts ASAP.. > > > > DNARD is hard to get but is being made (not by Digital anymore). > > > Ive mailed every customer on digitals pages who say there working on a > > production unit where and when I can get one.. I haven't yet gotten any > > satisfactory responce.. > > > > What have I gotta do? Jump up and down with cash in my hand? > > The latest EBSA is the EBSA-285. It is available in quantity now, all of our > suppliers (see the Digital Semiconductor web site). Linux is not yet (quite) > running on it but Russell should have his system by now, so you never know. > As a system it looks a lot like Corel's 21285 based SA-110 system and very like > Causality's upcoming system. Would be very keen to be kept in the loop regarding availability.. Im in Sydney Australia, so it may be even harder for me to get them.. I haven't approached the local distributer yet.. Im afraid he will have no idea what im after.. If you could verify the ordering numbers.... that may help.. --- Anthony Rumble Interactive Infotainment Systems P/L ACN: 076 046 252 Phone 02-9798-7604 Mobile 015-955-042 Pager 016-634-997 unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 5 18:38:11 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id SAA06815 for ; Mon, 5 Jan 1998 18:38:10 GMT Received: from vger.rutgers.edu ([128.6.190.2]:22098 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10355-21270>; Mon, 5 Jan 1998 20:34:55 +0200 Received: by vger.rutgers.edu id <970914-12312>; Mon, 5 Jan 1998 13:26:09 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:1919 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <970991-12312>; Mon, 5 Jan 1998 13:24:10 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id SAA06765; Mon, 5 Jan 1998 18:25:12 GMT From: Matthew Wilcox Message-Id: <199801051825.SAA06765@odie.barnet.ac.uk> Subject: Re: new patch To: linux@arm.uk.linux.org (Russell King - ARM Linux Admin) Date: Mon, 5 Jan 1998 18:25:11 +0000 (GMT) Cc: Philip.Blundell@pobox.com, linux-arm@vger.rutgers.edu In-Reply-To: <199712302352.XAA10579@raistlin.armlinux.org> from "Russell King - ARM Linux Admin" at Dec 30, 97 11:52:39 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > > BTW, for those of you using the new GCC, something that has been discovered > by Dave Gilbert: > > When compiling the kernel, turn off some of the optimisations, specifically > the one that allows GCC/EGCS to assume that all structures are word aligned. > GCC will try to load a short using a LDR with a shift. EG, to access bytes > 12 and 13 only of a structure, GCC will use by default: > ldr r0, [r1, #14] > mov r0, r0, lsr #16 > > This means that if r1 isn't word aligned (it certainly isn't for IP ethernet > packets in this case), the LDR picks up the wrong data. > > Does anyone know if this is a legal optimisation that the compiler is making? All you need to add is -mshort-load-bytes to the command line. Phil told me about this one already. > BTW, don't say that it should be word aligned. If that structure was word > aligned, then the IP structures won't be, which will be even worse. Blame > Xerox for that one! Cuh, those Xerox guys, hey? Fancy not optimising a network protocol for a 32 bit CPU due to be designed in 20 years time.. ;-) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From Philip.Blundell@pobox.com Mon Jan 5 21:38:13 1998 Return-Path: Received: from paddington.london.uk.eu.org (exim@tazenda.demon.co.uk [158.152.220.239]) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id VAA08055 for ; Mon, 5 Jan 1998 21:38:11 GMT Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xpL8J-0000Bv-00; Mon, 5 Jan 1998 22:36:31 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Matthew Wilcox cc: linux@arm.uk.linux.org (Russell King - ARM Linux Admin), linux-arm@vger.rutgers.edu Subject: Re: new patch In-reply-to: Your message of "Mon, 05 Jan 1998 18:25:11 GMT." <199801051825.SAA06765@odie.barnet.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 05 Jan 1998 22:36:31 +0000 From: Philip Blundell Message-Id: Status: RO >All you need to add is -mshort-load-bytes to the command line. Phil told >me about this one already. Actually, that won't fix it in all cases (though I think it will for the specific one of the Ethernet code). According to Richard Earnshaw it can still go wrong. In the short term we need to either use -mshort-load-bytes if it works, and fix the kernel if not. In the longer term we might be able to persuade gcc to generate the right code automatically. p. From Philip.Blundell@pobox.com Mon Jan 5 21:52:00 1998 Return-Path: Received: from paddington.london.uk.eu.org (exim@tazenda.demon.co.uk [158.152.220.239]) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id VAA08241 for ; Mon, 5 Jan 1998 21:51:57 GMT Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xpLLi-0000Dq-00; Mon, 5 Jan 1998 22:50:22 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Russell King - ARM Linux Admin cc: willy@odie.barnet.ac.uk, linux-arm@vger.rutgers.edu Subject: Re: new patch In-reply-to: Your message of "Tue, 30 Dec 1997 23:52:39 GMT." <199712302352.XAA10579@raistlin.armlinux.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 05 Jan 1998 22:50:22 +0000 From: Philip Blundell Message-Id: Status: RO >Does anyone know if this is a legal optimisation that the compiler is making? The consensus among the gcc people is that it is, because the ARM ABI is allowed to define a minimum alignment (and has done so, to 32). The optimisation does produce noticeably better code in the vast majority of cases. Changing this alignment now would not only cause a performance hit for all binaries, but it would make gcc incompatible with armcc, and with previous versions of gcc (in other words, all libraries would need recompiling). A way to signal to gcc that a structure is unaligned in this way is on the cards in the future, but we don't have it yet. In the short term we need to find a workaround such as using -mshort-load-bytes on the offending source files, or fixing them to use explicit byte ops. p. From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 5 21:54:52 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id VAA08245 for ; Mon, 5 Jan 1998 21:54:51 GMT Received: from vger.rutgers.edu ([128.6.190.2]:24145 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12566-27418>; Mon, 5 Jan 1998 23:52:54 +0200 Received: by vger.rutgers.edu id <970977-12312>; Mon, 5 Jan 1998 16:42:35 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:1396 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <970985-12312>; Mon, 5 Jan 1998 16:38:09 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id VAA08060; Mon, 5 Jan 1998 21:39:03 GMT From: Matthew Wilcox Message-Id: <199801052139.VAA08060@odie.barnet.ac.uk> Subject: Re: bootloader To: garan@digiserve.com (G a r a n) Date: Mon, 5 Jan 1998 21:39:02 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: <1.5.4.32.19980104145715.00669ae8@digiserve.com> from "G a r a n" at Jan 4, 98 02:57:15 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > A butchered copy of the formatter from !MOFS and also of riscbsd_bb and > !Partman would finally partition my scsi disc. I've tried wiping the disc > many times since, and it won't be partitioned until I use BOTH the changed > versions of the above... weird. Very. Would you mind writing down exactly what you did and posting it here? That way it will go into my archives & anyone else who suffers from the same problem can have it fixed more easily. > However, in the FAQ, to install the bootloader I use the -bootkernel > argument again but then specify root=/dev/sda3 ( in my case) in the extra > arguments. However, I assume acornscsi_mod.o isn't loaded, so how would I > load this at the same time, or do I need acornscsi compiled into my kernel? > Currently it complains, which is understandable. You need it compiled in. The best thing to do would be to boot off floppy, insmod the acornscsi.o and then roll your own kernel. It's not terribly hard, don't be frightened. (1) Download the virgin 2.0.31 source from your nearest mirror. (2) Unpack it (tar -zxf linux-2.0.31.tar.gz) (3) Apply the latest patch on the ftp site (cd linux && zcat ../linux-patch.gz | patch -p1) (4) If you're using a RiscPC, leave it as-is, else edit arch/arm/Makefile (5) Type `make config', ensuring you leave `enable versioning on module symbols' ON. Else the FPE won't load. Enable various other things as you see fit. Pressing ? gives you help on that option. Don't make things as modules until you really nkow what you're doing. (6) make dep && make clean && make zImage [cup-o-coffee time] (7) cp arch/arm/zImage /wherever/you/want/it/vmlinuz (8) edit /etc/boot.conf (9) run loadmap -v (10) reboot. Enjoy life. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From Philip.Blundell@pobox.com Mon Jan 5 22:02:39 1998 Return-Path: Received: from paddington.london.uk.eu.org (exim@tazenda.demon.co.uk [158.152.220.239]) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id WAA08289 for ; Mon, 5 Jan 1998 22:02:37 GMT Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xpLW3-0000Fg-00; Mon, 5 Jan 1998 23:01:03 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Matthew Wilcox cc: linux-arm@vger.rutgers.edu Subject: Re: bootloader In-reply-to: Your message of "Mon, 05 Jan 1998 21:39:02 GMT." <199801052139.VAA08060@odie.barnet.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 05 Jan 1998 23:01:02 +0000 From: Philip Blundell Message-Id: Status: RO >[cup-o-coffee time] As a matter of passing interest, how long does it take to compile a kernel on a RiscPC? p. (a member of the Gilbert school of benchmarking) From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 5 22:17:54 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id WAA08502 for ; Mon, 5 Jan 1998 22:17:53 GMT Received: from vger.rutgers.edu ([128.6.190.2]:22122 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12297-27418>; Tue, 6 Jan 1998 00:16:18 +0200 Received: by vger.rutgers.edu id <970958-12312>; Mon, 5 Jan 1998 17:03:38 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:1449 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <971001-12312>; Mon, 5 Jan 1998 16:53:56 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id VAA08250; Mon, 5 Jan 1998 21:55:02 GMT From: Matthew Wilcox Message-Id: <199801052155.VAA08250@odie.barnet.ac.uk> Subject: Re: problem with installling over NFS (and ftp) To: voss@yoda.in-berlin.de (Stefan Voss) Date: Mon, 5 Jan 1998 21:55:02 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: <2a2aa348%voss@yoda.in-berlin.de> from "Stefan Voss" at Jan 5, 98 01:13:06 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > > In message <199801041525.PAA00885@raistlin.armlinux.org>, > Russell wrote: > > > Stefan Voss writes: > > > I get the following error after specifying the IP address of the NFS > > > server and NFS directory: > > > > > > Internal error: Oops: 0 > > > CPU: 0 > > > pc : [] > > > lr : [] > > > > You need to get the latest version of the insmod utilities (and installation > > disks) off the FTP site. > > I'm a bit confused what insmod you are talking about. There's an insmod > binary in the instimage installation tree, on the root disk and on the > supplemental disk. I downloaded the insmod binary from the instimage > installation tree again but this was the same that I already had > (according to "diff" and "sum") > > I also downloaded the root-rpc and supplemental disks again from > ftp.arm.uk.linux.org /pub/armlinux/distrib/RedHat/disks and used > !WriteDisk to re-create the boot disks. > > The result however remains the same. I still get the error described > above. I also tried a 2.0.30 kernel instead of the latest 2.0.31 but the > install programs says that it doesn't know the EtherH device. The boot > messages however seem to indicate that the EtherH card is detected > correctly. The boot messages are misleading. These are merely the id strings out of the podule ROMs. It does *not* mean that the kernel recognises the card. IIRC, i reported a problem to russell with the EtherLan500s which he then fixed in 2.0.31. The insmod binary that is contained in an RPM works.. but I dunno whether that would work or not before the rest of the system is installed. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 5 22:33:30 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id WAA08564 for ; Mon, 5 Jan 1998 22:33:29 GMT Received: from vger.rutgers.edu ([128.6.190.2]:27145 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12402-27418>; Tue, 6 Jan 1998 00:31:37 +0200 Received: by vger.rutgers.edu id <970965-12312>; Mon, 5 Jan 1998 17:13:04 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:1206 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <971013-12312>; Mon, 5 Jan 1998 17:01:39 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xpLW3-0000Fg-00; Mon, 5 Jan 1998 23:01:03 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Matthew Wilcox cc: linux-arm@vger.rutgers.edu Subject: Re: bootloader In-reply-to: Your message of "Mon, 05 Jan 1998 21:39:02 GMT." <199801052139.VAA08060@odie.barnet.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 05 Jan 1998 23:01:02 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >[cup-o-coffee time] As a matter of passing interest, how long does it take to compile a kernel on a RiscPC? p. (a member of the Gilbert school of benchmarking) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 5 22:49:07 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id WAA08754 for ; Mon, 5 Jan 1998 22:49:06 GMT Received: from vger.rutgers.edu ([128.6.190.2]:2849 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <1816-150>; Tue, 6 Jan 1998 00:47:24 +0200 Received: by vger.rutgers.edu id <970972-12312>; Mon, 5 Jan 1998 17:40:08 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:1554 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <971016-12312>; Mon, 5 Jan 1998 17:27:15 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id WAA08527; Mon, 5 Jan 1998 22:28:30 GMT From: Matthew Wilcox Message-Id: <199801052228.WAA08527@odie.barnet.ac.uk> Subject: Re: new patch To: Philip.Blundell@pobox.com (Philip Blundell) Date: Mon, 5 Jan 1998 22:28:29 +0000 (GMT) Cc: linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu In-Reply-To: from "Philip Blundell" at Jan 5, 98 10:50:22 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >> EG, to access bytes >> 12 and 13 only of a structure, GCC will use by default: >> ldr r0, [r1, #14] >> mov r0, r0, lsr #16 > > >Does anyone know if this is a legal optimisation that the compiler is making? > > The consensus among the gcc people is that it is, because the ARM ABI is > allowed to define a minimum alignment (and has done so, to 32). The > optimisation does produce noticeably better code in the vast majority of cases. > > A way to signal to gcc that a structure is unaligned in this way is on the > cards in the future, but we don't have it yet. In the short term we need to > find a workaround such as using -mshort-load-bytes on the offending source > files, or fixing them to use explicit byte ops. Can I ask for a specific example here? Are we talking about something along the lines of: struct foo { int a; short b; short c; char d; short e; short f; } ie which of these will be wrongly accessed? And what alignment will GCC give them? I don't think that -mshort-load-bytes is the correct way to differentiate the two cases, I think -munaligned-loads versus -mno-unaligned-loads is the correct name for the switch - simply because I don't think shorts should be loaded as bytes! It's /at least/ as efficient to load them as words and then shift them twice as it is to load as bytes and orr them together. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From alan@snowcrash.cymru.net Mon Jan 5 23:08:53 1998 Return-Path: Received: from snowcrash.cymru.net (snowcrash.cymru.net [163.164.160.3]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id XAA08786 for ; Mon, 5 Jan 1998 23:08:52 GMT Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id XAA25647; Mon, 5 Jan 1998 23:08:55 GMT From: Alan Cox Message-Id: <199801052308.XAA25647@snowcrash.cymru.net> Subject: Re: new patch To: Philip.Blundell@pobox.com (Philip Blundell) Date: Mon, 5 Jan 1998 23:08:53 +0000 (GMT) Cc: willy@odie.barnet.ac.uk, linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu In-Reply-To: from "Philip Blundell" at Jan 5, 98 10:36:31 pm Content-Type: text Status: RO > In the short term we need to either use -mshort-load-bytes if it works, and > fix the kernel if not. In the longer term we might be able to persuade gcc to > generate the right code automatically. Fixing the kernel is a non option for every case I think. What I've suggested to Russell is this for later ARM chips o Compile with gcc 2.8 without -mshort-load-bytes o In kernel space take exceptions on unaligned load/store (since there should be almost none) o Instrument the ones we get and write little unaligned load macros for the ones that show a lot (ie ether_type_trans) Fixing gcc is hard - it means fixing the __attribute(packed)) bug in the ARM so that it works properly and potentially adding an __attribute(unaligned) to the compiler. By taking alignment traps now gcc wont generate them all the time its realistic to make AX.25 and IPX work on the ARM From dg@cogency.co.uk Tue Jan 6 09:54:44 1998 Return-Path: Received: from mailhost.cogency.co.uk (man01.cogency.co.uk [194.201.189.33] (may be forged)) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id JAA12578 for ; Tue, 6 Jan 1998 09:54:42 GMT Received: from cogency.co.uk by mailhost.cogency.co.uk (SMI-8.6/PIPEX simple 1.27) id JAA11650; Tue, 6 Jan 1998 09:54:17 GMT Sender: dg@cogency.co.uk Message-ID: <34B1FEAF.54D6FE9D@cogency.co.uk> Date: Tue, 06 Jan 1998 09:51:43 +0000 From: David Alan Gilbert Organization: Cogency Technology Inc X-Mailer: Mozilla 4.04 [en] (X11; I; SunOS 5.5 sun4m) MIME-Version: 1.0 To: Alan Cox CC: Philip Blundell , willy@odie.barnet.ac.uk, linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu Subject: Re: new patch References: <199801052308.XAA25647@snowcrash.cymru.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Status: RO Alan Cox wrote: > Fixing the kernel is a non option for every case I think. What I've > suggested to Russell is this for later ARM chips But what of the older ARMs. Dave > unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu -- ---------------------------------------------------------------------- - Dr. David Alan Gilbert - WARNING! This is a beta release .signature- - Work: dg @ cogency.co.uk - 0161-428-9444 - - Home: gro.gilbert @ treblig.org - - ---------------------------------------------------------------------- From alan@snowcrash.cymru.net Tue Jan 6 10:03:30 1998 Return-Path: Received: from snowcrash.cymru.net (snowcrash.cymru.net [163.164.160.3]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id KAA12598 for ; Tue, 6 Jan 1998 10:03:29 GMT Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id KAA03621; Tue, 6 Jan 1998 10:03:35 GMT From: Alan Cox Message-Id: <199801061003.KAA03621@snowcrash.cymru.net> Subject: Re: new patch To: dg@cogency.co.uk (David Alan Gilbert) Date: Tue, 6 Jan 1998 10:03:33 +0000 (GMT) Cc: alan@cymru.net, Philip.Blundell@pobox.com, willy@odie.barnet.ac.uk, linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu In-Reply-To: <34B1FEAF.54D6FE9D@cogency.co.uk> from "David Alan Gilbert" at Jan 6, 98 09:51:43 am Content-Type: text Status: RO > > > Fixing the kernel is a non option for every case I think. What I've > > suggested to Russell is this for later ARM chips > > But what of the older ARMs. You would appear to end up with the same situation as before. Those cases explicitly handled work, the others dont. The code (especially networking) was never designed to cope with a CPU as weak as the early arm designs are at handling arbitary alignments. You could argue thats a design weakness and I wouldnt argue that hard but I will take clean patches that dont impact other platforms.. Alan From dg@cogency.co.uk Tue Jan 6 10:04:25 1998 Return-Path: Received: from mailhost.cogency.co.uk (man01.cogency.co.uk [194.201.189.33] (may be forged)) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id KAA12602 for ; Tue, 6 Jan 1998 10:04:23 GMT Received: from cogency.co.uk by mailhost.cogency.co.uk (SMI-8.6/PIPEX simple 1.27) id KAA11656; Tue, 6 Jan 1998 10:05:01 GMT Sender: dg@cogency.co.uk Message-ID: <34B20134.E0B49CB5@cogency.co.uk> Date: Tue, 06 Jan 1998 10:02:28 +0000 From: David Alan Gilbert Organization: Cogency Technology Inc X-Mailer: Mozilla 4.04 [en] (X11; I; SunOS 5.5 sun4m) MIME-Version: 1.0 To: Matthew Wilcox CC: Philip Blundell , linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu Subject: Re: new patch References: <199801052228.WAA08527@odie.barnet.ac.uk> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Status: RO Matthew Wilcox wrote: > Can I ask for a specific example here? Are we talking about something along > the lines of: > > struct foo > { > int a; > short b; > short c; > char d; > short e; > short f; > } > > ie which of these will be wrongly accessed? And what alignment will GCC > give them? OK - your case above is fine - it will work OK. But here is the case which started this whole ball rolling: struct ethhdr { char addr1[6]; char addr2[6]; short proto; }; This is the header on an ethernet packet; when it comes in off the wire its put into a buffer with this structure starting on a 2 byte alginment address; thats done so that the data which follows (probably a lump of Tcp (or is it IP?)) will be word aligned. In the depths of the ethernet code there is something which accesses the 'proto' field from this 2 byte aligned pointer. Unfortunatly the ARM GCC assumes that ALL structures are word aligned and generates code to access the short using a LDR which relies on the word alignment - it thus gets a dud value for 'proto' and the incoming packet is thrown away because it can't figure out what type of packet it is. Why doesn't this cause a problem for other Linux's? All other Linux GCC's assume that struct's are algined to the size of the largest component within the struct (upto word size?) and thus they would only assume this was short aligned and generate appropriate code. (I, like most C programmers I suspect, had presumed that this was what was ment to happen in C but the GCC guys have been telling us that the standard actually allows the C implementation to specify this alignment). Why hasn't this caused problems before for ARM Linux? Earlier versions of GCC didn't happen to have optimisations which excercised the problem. Russ had to do some hacks to the isofs stuff to make it work because..... Structures get padded: struct x { short a; }; sizeof(x)=4 and thus if you happen to have a datastructure described using collections of small structures you get padding all over the place. > I don't think that -mshort-load-bytes is the correct way to differentiate > the two cases, I think -munaligned-loads versus -mno-unaligned-loads is the > correct name for the switch - simply because I don't think shorts should > be loaded as bytes! That isn't what's happening. The -mshort-load-bytes is turning off the new optimisation effectivly and causing GCC to emit code which doesn't rely on the word alignment IN THIS CASE. Its not a general soloution to the unalignment. Dave -- ---------------------------------------------------------------------- - Dr. David Alan Gilbert - WARNING! This is a beta release .signature- - Work: dg @ cogency.co.uk - 0161-428-9444 - - Home: gro.gilbert @ treblig.org - - ---------------------------------------------------------------------- From alan@snowcrash.cymru.net Tue Jan 6 10:28:16 1998 Return-Path: Received: from snowcrash.cymru.net (snowcrash.cymru.net [163.164.160.3]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id KAA12786 for ; Tue, 6 Jan 1998 10:28:15 GMT Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id KAA04289; Tue, 6 Jan 1998 10:28:29 GMT From: Alan Cox Message-Id: <199801061028.KAA04289@snowcrash.cymru.net> Subject: Re: new patch To: dg@cogency.co.uk (David Alan Gilbert) Date: Tue, 6 Jan 1998 10:28:26 +0000 (GMT) Cc: willy@odie.barnet.ac.uk, Philip.Blundell@pobox.com, linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu In-Reply-To: <34B20134.E0B49CB5@cogency.co.uk> from "David Alan Gilbert" at Jan 6, 98 10:02:28 am Content-Type: text Status: RO > put into a buffer with this structure starting on a 2 byte alginment > address; thats done so that the data which follows (probably a lump > of Tcp (or is it IP?)) will be word aligned. Actually dont rely on this. Some cards cannot do non 32bit aligned DMA and those cards will land the packet otherwise aligned. > assume that struct's are algined to the size of the largest component within > the struct (upto word size?) and thus they would only assume this was short > aligned and generate appropriate code. (I, like most C programmers I suspect, Not always - on byte aligned packets - eg IP over AX.25 the alpha and sparc will sometimes fault and emulate the faulting instruction. This is faster than filling the generic cases with byte loads > had presumed that this was what was ment to happen in C but the GCC guys have > been telling us that the standard actually allows the C implementation to > specify this alignment). GCC is correct in chosing to align its fields. Its broken in that it doesnt honour "packed" attributes on the ARM. Alan From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 6 11:07:48 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id LAA13069 for ; Tue, 6 Jan 1998 11:07:47 GMT Received: from vger.rutgers.edu ([128.6.190.2]:11792 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11710-27083>; Tue, 6 Jan 1998 13:04:56 +0200 Received: by vger.rutgers.edu id <970913-12312>; Tue, 6 Jan 1998 06:02:38 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:1626 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <970948-12312>; Tue, 6 Jan 1998 06:01:24 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id LAA13035; Tue, 6 Jan 1998 11:02:35 GMT From: Matthew Wilcox Message-Id: <199801061102.LAA13035@odie.barnet.ac.uk> Subject: Re: bootloader To: Philip.Blundell@pobox.com (Philip Blundell) Date: Tue, 6 Jan 1998 11:02:32 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: from "Philip Blundell" at Jan 5, 98 11:01:02 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > > >[cup-o-coffee time] > > As a matter of passing interest, how long does it take to compile a kernel on > a RiscPC? > > p. (a member of the Gilbert school of benchmarking) With an ARM6, 1 hour 15 minutes timein user, 1 hour 50 time-elapsed, but I was fiddling with glibc snapshots at the same time + using it as an NFS server, so 1 hour 20 seems about right. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From pb@nexus.co.uk Tue Jan 6 11:12:53 1998 Return-Path: Received: from globe (nexusel.demon.co.uk [158.152.30.195]) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id LAA13150 for ; Tue, 6 Jan 1998 11:12:51 GMT Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xpWsQ-00008S-00; Tue, 6 Jan 1998 11:08:54 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xpWsP-0005HH-00; Tue, 6 Jan 1998 11:08:53 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Matthew Wilcox cc: linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu Subject: Re: new patch In-reply-to: Your message of "Mon, 05 Jan 1998 22:28:29 GMT." <199801052228.WAA08527@odie.barnet.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Tue, 06 Jan 1998 11:08:53 +0000 From: Philip Blundell Message-Id: Status: RO >Can I ask for a specific example here? Are we talking about something along >the lines of: > >struct foo >{ > int a; > short b; > short c; > char d; > short e; > short f; >} > >ie which of these will be wrongly accessed? And what alignment will GCC >give them? In that case, all will be well. The structure must have 32-bit alignment by anybody's standards because of the `int' member. >be loaded as bytes! It's /at least/ as efficient to load them as words >and then shift them twice as it is to load as bytes and orr them together. The problem arises because you don't know what will happen if you load a word at arbitrary alignment. p. From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 6 11:15:52 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id LAA13307 for ; Tue, 6 Jan 1998 11:15:51 GMT Received: from vger.rutgers.edu ([128.6.190.2]:18962 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12278-21270>; Tue, 6 Jan 1998 13:14:39 +0200 Received: by vger.rutgers.edu id <970800-12312>; Tue, 6 Jan 1998 06:12:12 -0500 Received: from nexusel.demon.co.uk ([158.152.30.195]:1359 "HELO globe" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970810-12312>; Tue, 6 Jan 1998 06:11:46 -0500 Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xpWbB-00006r-00; Tue, 6 Jan 1998 10:51:05 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xpWbA-0005Fz-00; Tue, 6 Jan 1998 10:51:04 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Alan Cox cc: linux-arm@vger.rutgers.edu Subject: Re: new patch In-reply-to: Your message of "Mon, 05 Jan 1998 23:08:53 GMT." <199801052308.XAA25647@snowcrash.cymru.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Tue, 06 Jan 1998 10:51:03 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >Fixing the kernel is a non option for every case I think. What I've >suggested to Russell is this for later ARM chips > >o Compile with gcc 2.8 without -mshort-load-bytes >o In kernel space take exceptions on unaligned load/store (since there > should be almost none) That's no good for ARM3 though. >o Instrument the ones we get and write little unaligned load > macros for the ones that show a lot (ie ether_type_trans) If you don't use -mshort-load-bytes GCC is basically being told that the MMU will not trap unaligned accesses. I believe that without that flag it can deliberately misalign loads in order to take advantage of the slightly idiomatic ARM behaviour. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From pb@nexus.co.uk Tue Jan 6 11:59:57 1998 Return-Path: Received: from globe (nexusel.demon.co.uk [158.152.30.195]) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id LAA13550 for ; Tue, 6 Jan 1998 11:59:56 GMT Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xpXbu-0000D5-00; Tue, 6 Jan 1998 11:55:54 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xpXbt-0005KH-00; Tue, 6 Jan 1998 11:55:53 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Matthew Wilcox Subject: Re: new patch In-reply-to: Your message of "Tue, 06 Jan 1998 11:18:12 GMT." <199801061118.LAA13312@odie.barnet.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Tue, 06 Jan 1998 11:55:52 +0000 From: Philip Blundell Message-Id: Status: RO >Yes, but I wasn't suggesting that. In the case where you /do/ know the >alignment to be on a 32-bit boundary, doing a load from a non-aligned address >is silly if the MMU is going to take an exception. Equally, loading a pair >of bytes is not a good idea. So loading a word where you *do* know the >alignment and then shifting &| masking is the right thing to do. That's true, and that's more or less what GCC is trying to do. The reason it gets upset is that if you violate the structures-must-be-32-bit-aligned rule, it has no way to know whether an element of that structure is going to be aligned or not at the time it generates code. The only two choices you have are to generate optimistic code which is fast in the aligned case and fails in the non-aligned case, and to generate pessimistic code which always works but is slow. GCC takes the former approach, and I think this is best. It would be possible, albeit very hard, to make the compiler smarter about this - for example it could in theory assume that any structure containing an element with size 4 or greater, or any automatic, would be aligned. p. From linux@arm.uk.linux.org Tue Jan 6 21:14:36 1998 Return-Path: Received: from snowcrash.cymru.net (snowcrash.cymru.net [163.164.160.3]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id VAA17513 for ; Tue, 6 Jan 1998 21:14:36 GMT Received: from caramon.armlinux.org (dialup3.cymru.net [163.164.160.67]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id VAA17144; Tue, 6 Jan 1998 21:14:37 GMT Received: from raistlin.armlinux.org (raistlin [192.168.0.3]) by caramon.armlinux.org (8.7.4/8.7.3) with ESMTP id XAA02500; Mon, 5 Jan 1998 23:15:07 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.armlinux.org (8.7.4/8.7.3) id XAA00927; Mon, 5 Jan 1998 23:15:04 GMT Message-Id: <199801052315.XAA00927@raistlin.armlinux.org> Subject: Re: new patch To: Philip.Blundell@pobox.com (Philip Blundell) Date: Mon, 5 Jan 1998 23:15:04 +0000 (GMT) Cc: linux@arm.uk.linux.org, willy@odie.barnet.ac.uk, linux-arm@vger.rutgers.edu, linux@treblig.org In-Reply-To: from "Philip Blundell" at Jan 5, 98 10:50:22 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Status: RO Philip Blundell writes: > >Does anyone know if this is a legal optimisation that the compiler is making? > > The consensus among the gcc people is that it is, because the ARM ABI is > allowed to define a minimum alignment (and has done so, to 32). The > optimisation does produce noticeably better code in the vast majority of cases. Someone ought to look into the __packed__ attribute in GCC - according to Alan Cox, it is supposed to tell the compiler to ignore any of these stupid alignments, for instance, have a look at the changes in fs/isofs/rock.[ch], which originally was something like (this is from memory): struct times { char time[7]; }; struct RR_TF { char c1; char c2; struct times[3]; } Now, on the disk, the layout is: Byte Contents +0 c1 +1 c2 +2 times[0].time[0] ... +8 times[0].time[6] +9 times[1].time[0] ... and so on. However, if left to ARM GCC, then times[0].time[0] will be at +4, and times[1].time[0] at +12. Alan claims that if you use __packed__ on these, it should tell GCC that you want everything to be as small as possible, and that would be correct. However, ARM GCC appears to ignore it. Can someone confirm both the claim, and my observed action of GCC, and confirm if ARM GCC is correct/wrong. (And I hope that this doesn't cause another war-of-the compilers). _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 6 23:24:19 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id XAA18280 for ; Tue, 6 Jan 1998 23:24:18 GMT Received: from vger.rutgers.edu ([128.6.190.2]:59498 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <4131-21270>; Wed, 7 Jan 1998 00:27:54 +0200 Received: by vger.rutgers.edu id <971450-12312>; Tue, 6 Jan 1998 16:11:16 -0500 Received: from post-10.mail.demon.net ([194.217.242.154]:37163 "HELO post.mail.demon.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <971081-12312>; Tue, 6 Jan 1998 15:51:03 -0500 Received: from succexpr.demon.co.uk ([158.152.208.192]) by post.mail.demon.net id aa1026057; 6 Jan 98 20:48 GMT Message-ID: <34B2802E.446B9B3D@causality.com> Date: Tue, 06 Jan 1998 19:04:14 +0000 From: "Neil A. Carson" Organization: Causality Limited (London, UK) X-Mailer: Mozilla 3.01 (X11; I; NetBSD 1.2G arm32) MIME-Version: 1.0 To: David Alan Gilbert CC: linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu Subject: Re: new patch References: <199801052228.WAA08527@odie.barnet.ac.uk> <34B20134.E0B49CB5@cogency.co.uk> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO David Alan Gilbert wrote: > Why hasn't this caused problems before for ARM Linux? Earlier versions of GCC > didn't happen to have optimisations which excercised the problem. Russ had > to do some hacks to the isofs stuff to make it work because..... > > Structures get padded: > > struct x { > short a; > }; > > sizeof(x)=4 > > and thus if you happen to have a datastructure described using collections > of small structures you get padding all over the place. Dunnow if it's of any relevance or not; we hit this with netbsd/arm a while ago, and made some compiler changes which are now part of the gcc three, but recognised as not being for general use, I think. Could Linux not adopt this approach too whih I think would solve any problems? Or maybe I'm getting the wrong end of the stick, adn the problems are future ones not present ones. N. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Wed Jan 7 14:14:08 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id OAA23919 for ; Wed, 7 Jan 1998 14:14:07 GMT Received: from vger.rutgers.edu ([128.6.190.2]:29504 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <1094-21270>; Wed, 7 Jan 1998 16:12:48 +0200 Received: by vger.rutgers.edu id <971405-12312>; Wed, 7 Jan 1998 05:04:29 -0500 Received: from [194.201.189.33] ([194.201.189.33]:51006 "HELO mailhost.cogency.co.uk" ident: "SOCKWRITE-65") by vger.rutgers.edu with SMTP id <971499-12312>; Wed, 7 Jan 1998 04:51:00 -0500 Received: from cogency.co.uk by mailhost.cogency.co.uk (SMI-8.6/PIPEX simple 1.27) id JAA12312; Wed, 7 Jan 1998 09:53:11 GMT Message-ID: <34B34FED.B6E02D75@cogency.co.uk> Date: Wed, 07 Jan 1998 09:50:37 +0000 From: David Alan Gilbert Organization: Cogency Technology Inc X-Mailer: Mozilla 4.04 [en] (X11; I; SunOS 5.5 sun4m) MIME-Version: 1.0 To: "Neil A. Carson" CC: linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu Subject: Re: new patch References: <199801052228.WAA08527@odie.barnet.ac.uk> <34B20134.E0B49CB5@cogency.co.uk> <34B2802E.446B9B3D@causality.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Neil A. Carson wrote: > Dunnow if it's of any relevance or not; we hit this with netbsd/arm a > while ago, and made some compiler changes which are now part of the gcc > three, but recognised as not being for general use, I think. Could Linux > not adopt this approach too whih I think would solve any problems? Or > maybe I'm getting the wrong end of the stick, adn the problems are > future ones not present ones. Which version of the compiler has the mods (or is it a patch you have)? Dave -- ---------------------------------------------------------------------- - Dr. David Alan Gilbert - WARNING! This is a beta release .signature- - Work: dg @ cogency.co.uk - 0161-428-9444 - - Home: gro.gilbert @ treblig.org - - ---------------------------------------------------------------------- unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 8 20:09:41 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id UAA03264 for ; Thu, 8 Jan 1998 20:09:40 GMT Received: from vger.rutgers.edu ([128.6.190.2]:12385 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <2330-21270>; Thu, 8 Jan 1998 22:04:41 +0200 Received: by vger.rutgers.edu id <971505-12312>; Thu, 8 Jan 1998 10:41:48 -0500 Received: from post-10.mail.demon.net ([194.217.242.154]:41308 "HELO post.mail.demon.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <971500-12312>; Thu, 8 Jan 1998 05:54:45 -0500 Received: from ewrcsdra.demon.co.uk ([158.152.81.3]) by post.mail.demon.net id aa1023859; 8 Jan 98 10:50 GMT Received: from localhost (rsimpson@localhost) by ewrcsdra.demon.co.uk (8.7.5/v3.2) with SMTP id LAA11232; Thu, 8 Jan 1998 11:14:25 GMT Date: Thu, 8 Jan 1998 11:14:24 +0000 (GMT) From: Richard Simpson To: "Carl A. Waldspurger" cc: linux-arm@vger.rutgers.edu Subject: Re: debugger for ARM/Linux? In-Reply-To: <9801072343.AA29494@rowdy.pa.dec.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Russell has supplied me with a copy of gdb, which he claims works, for inclusion in the soon to be released CD-ROM. Unfortunately, I have not yet had a chance to try it. If and when Russell gives the go ahead I will upload it to the FTP site. Richard Simpson Farnborough, Hants, Uk Fax: 01252 392118 rsimpson@ewrcsdra.demon.co.uk On Wed, 7 Jan 1998, Carl A. Waldspurger wrote: > > Hi. I've been successfully running ARM/Linux, but can't seem to find > a port of gdb. Does anyone know where I can find a port of gdb and/or > any other debugging tools for this platform? > > Thanks, > > Carl Waldspurger > caw@pa.dec.com > unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu > unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 8 21:07:30 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id VAA03665 for ; Thu, 8 Jan 1998 21:07:29 GMT Received: from vger.rutgers.edu ([128.6.190.2]:16675 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10495-21270>; Thu, 8 Jan 1998 22:41:33 +0200 Received: by vger.rutgers.edu id <971497-12312>; Thu, 8 Jan 1998 10:50:09 -0500 Received: from vacim.scg.man.ac.uk ([130.88.94.96]:4788 "EHLO vacim.scg.man.ac.uk" ident: "root") by vger.rutgers.edu with ESMTP id <971775-12312> convert rfc822-to-8bit; Thu, 8 Jan 1998 07:21:43 -0500 Received: by vacim.scg.man.ac.uk id m0xqH0O-000DmoC (Debian Smail-3.2 1996-Jul-4 #2); Thu, 8 Jan 1998 12:24:12 +0000 (GMT) Message-Id: From: richard@vacim.scg.man.ac.uk (S. Richard Grint) Subject: Re: libgcc.a In-Reply-To: from Philip Blundell at "Jan 7, 98 05:57:12 pm" To: linux-arm@vger.rutgers.edu Date: Thu, 8 Jan 1998 12:24:12 +0000 (GMT) X-Mailer: ELM [version 2.4ME+ PL31 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8BIT X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > >cross-compile the compiler, it says I must find libgcc.a, which I don have > >and am unable to find anywhere. > > What version of GCC are you using? There's an assembler source file for > libgcc1.a supplied and recent versions know how to use this automatically. > > p. > GCC 2.7.2.2 - It may know how to make libgcc1.a automatically, but when cross compiling, it wont ...extracted from the Makefile: # Specify the rule for actually making libgcc.a, LIBGCC = libgcc.a # and the rule for installing it. INSTALL_LIBGCC = install-libgcc # Specify the rule for actually making libgcc1.a. # The value may be empty; that means to do absolutely nothing # with or for libgcc1.a. LIBGCC1 = libgcc1.a # Specify the rule for making libgcc1.a for a cross-compiler. # The default rule assumes that libgcc1.a is supplied by the user. CROSS_LIBGCC1 = libgcc1.cross Do I need to edit the makefile to get it to try and compile libgcc1.a for me? If so how?? Rich unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 9 01:20:50 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id BAA05450 for ; Fri, 9 Jan 1998 01:20:49 GMT Received: from vger.rutgers.edu ([128.6.190.2]:51507 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12881-21270>; Fri, 9 Jan 1998 03:11:36 +0200 Received: by vger.rutgers.edu id <971105-241>; Thu, 8 Jan 1998 17:46:34 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:1104 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <971457-12312>; Thu, 8 Jan 1998 16:25:18 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xqQGJ-0000BN-00; Thu, 8 Jan 1998 22:17:15 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: richard@vacim.scg.man.ac.uk (S. Richard Grint) cc: linux-arm@vger.rutgers.edu Subject: Re: libgcc.a In-reply-to: Your message of "Thu, 08 Jan 1998 12:24:12 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 08 Jan 1998 22:17:15 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >Do I need to edit the makefile to get it to try and compile libgcc1.a for >me? If so how?? In config/arm/t-linux, add these lines: # Since libgcc1 is an assembler file, we can build it automatically for the # cross-compiler. CROSS_LIBGCC1 = libgcc1-asm.a LIBGCC1 = libgcc1-asm.a (The last one may be there already) p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 9 06:38:41 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id GAA07218 for ; Fri, 9 Jan 1998 06:38:40 GMT Received: from vger.rutgers.edu ([128.6.190.2]:56434 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <1459-21270>; Fri, 9 Jan 1998 07:54:38 +0200 Received: by vger.rutgers.edu id <970956-241>; Thu, 8 Jan 1998 22:52:34 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:2048 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <970989-241>; Thu, 8 Jan 1998 21:03:25 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id CAA05662; Fri, 9 Jan 1998 02:03:15 GMT From: Matthew Wilcox Message-Id: <199801090203.CAA05662@odie.barnet.ac.uk> Subject: Re: libgcc.a To: richard@vacim.scg.man.ac.uk (S. Richard Grint) Date: Fri, 9 Jan 1998 02:03:14 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: from "S. Richard Grint" at Jan 8, 98 12:24:12 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > > > >cross-compile the compiler, it says I must find libgcc.a, which I don have > > >and am unable to find anywhere. > > > > What version of GCC are you using? There's an assembler source file for > > libgcc1.a supplied and recent versions know how to use this automatically. > > > > p. > > > GCC 2.7.2.2 - It may know how to make libgcc1.a automatically, but when > cross compiling, it wont ...extracted from the Makefile: [deletia] > Do I need to edit the makefile to get it to try and compile libgcc1.a for > me? If so how?? I told you where to get libgcc.a from already.. why are you still asking? unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 9 08:36:14 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id IAA07909 for ; Fri, 9 Jan 1998 08:36:12 GMT Received: from vger.rutgers.edu ([128.6.190.2]:6995 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <16455-21270>; Fri, 9 Jan 1998 10:30:53 +0200 Received: by vger.rutgers.edu id <971047-241>; Fri, 9 Jan 1998 03:07:13 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:1051 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <971282-241>; Fri, 9 Jan 1998 02:43:01 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xqa0F-00007f-00; Fri, 9 Jan 1998 08:41:19 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Russ King cc: linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu Subject: PLIP In-reply-to: Your message of "Thu, 08 Jan 1998 18:17:23 GMT." <11036.9801081817@diana.ecs.soton.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 09 Jan 1998 08:41:18 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >It should be able to work on RiscPCs without any problem (but I haven't been >able to confirm this). However, A5000 machines definitely won't work, Presumably you mean that the interrupt sense is inverted, rather than the line itself. Does IOC not give you a way to change that? In any case, I think there are enough other machines with this problem that PLIP could beneficially have an `other end has broken ACK' mode. Various people are working on PLIP for extended and broken hardware at the moment, so let's wait and see what they come up with. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 9 10:36:00 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id KAA08753 for ; Fri, 9 Jan 1998 10:35:57 GMT Received: from vger.rutgers.edu ([128.6.190.2]:25100 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <17116-21270>; Fri, 9 Jan 1998 12:33:52 +0200 Received: by vger.rutgers.edu id <970906-241>; Fri, 9 Jan 1998 05:32:00 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:2028 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970896-241>; Fri, 9 Jan 1998 05:31:44 -0500 Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id KAA10874; Fri, 9 Jan 1998 10:32:18 GMT From: Alan Cox Message-Id: <199801091032.KAA10874@snowcrash.cymru.net> Subject: Re: PLIP To: Philip.Blundell@pobox.com (Philip Blundell) Date: Fri, 9 Jan 1998 10:32:16 +0000 (GMT) Cc: rmk@ecs.soton.ac.uk, linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu In-Reply-To: from "Philip Blundell" at Jan 9, 98 08:41:18 am Content-Type: text X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > Presumably you mean that the interrupt sense is inverted, rather than the line > itself. Does IOC not give you a way to change that? Thats actually in common with the m68K Amiga hacked up plip tricks. Yes it would be good to have an "inverted ack" mode. Alan unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 9 13:01:46 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id NAA09738 for ; Fri, 9 Jan 1998 13:01:45 GMT Received: from vger.rutgers.edu ([128.6.190.2]:31091 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11505-27083>; Fri, 9 Jan 1998 15:00:41 +0200 Received: by vger.rutgers.edu id <970800-241>; Fri, 9 Jan 1998 07:58:58 -0500 Received: from out2.ibm.net ([165.87.194.229]:4150 "EHLO out2.ibm.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970810-241>; Fri, 9 Jan 1998 07:58:33 -0500 Received: from cvs.com.au (slip-129-37-30-176.fl.us.ibm.net [129.37.30.176]) by out2.ibm.net (8.8.5/8.6.9) with ESMTP id MAA47196 for ; Fri, 9 Jan 1998 12:59:17 GMT Message-ID: <34B61FA0.474CB30A@cvs.com.au> Date: Fri, 09 Jan 1998 08:01:20 -0500 From: Charles Esson Reply-To: charlese@cvs.com.au Organization: Colour Vision Systems X-Mailer: Mozilla 4.04 [en] (WinNT; I) MIME-Version: 1.0 To: "linux-arm@vger.rutgers.edu" Subject: Elf standard Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Does anyone know if the ELF standard get sorted out for the ARM. If so does anyone know where I can get a copy. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 9 17:01:59 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id RAA11455 for ; Fri, 9 Jan 1998 17:01:58 GMT Received: from vger.rutgers.edu ([128.6.190.2]:63289 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <16835-150>; Fri, 9 Jan 1998 18:58:51 +0200 Received: by vger.rutgers.edu id <970881-241>; Fri, 9 Jan 1998 11:53:20 -0500 Received: from monarch.ja.net ([128.86.16.37]:34328 "EHLO monarch.ja.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970835-241>; Fri, 9 Jan 1998 11:52:50 -0500 Received: from sun3.nsfnet-relay.ac.uk by monarch.ja.net with SPAM SMTP (PP); Fri, 9 Jan 1998 13:54:47 +0000 Received: from bright.ecs.soton.ac.uk by sun3.nsfnet-relay.ac.uk with NRS SMTP (PP); Fri, 9 Jan 1998 13:54:44 +0000 Received: from diana.ecs.soton.ac.uk by bright.ecs.soton.ac.uk; Fri, 9 Jan 98 13:52:13 GMT From: Russ King Message-Id: <22474.9801091354@diana.ecs.soton.ac.uk> Subject: Re: PLIP To: Philip.Blundell@pobox.com (Philip Blundell) Date: Fri, 9 Jan 1998 13:54:03 +0000 (GMT) Cc: linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu In-Reply-To: from "Philip Blundell" at Jan 9, 98 08:41:18 am X-Mailer: ELM [version 2.4 PL24] Content-Type: text X-Orcpt: rfc822;linux-arm@vger.rutgers.edu X-Envid: [/PRMD=UK.AC/ADMD= /C=GB/;<22474.9801091354@diana.ecs.soto] Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Philip Blundell writes: > Presumably you mean that the interrupt sense is inverted, rather than the line > itself. Does IOC not give you a way to change that? No, the ACK itself is inverted, not the interrupt sense. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 9 18:11:39 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id SAA11867 for ; Fri, 9 Jan 1998 18:11:38 GMT Received: from vger.rutgers.edu ([128.6.190.2]:19252 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <17249-21270>; Fri, 9 Jan 1998 20:05:21 +0200 Received: by vger.rutgers.edu id <970883-241>; Fri, 9 Jan 1998 12:51:20 -0500 Received: from nexusel.demon.co.uk ([158.152.30.195]:2734 "HELO globe" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970907-241>; Fri, 9 Jan 1998 12:38:20 -0500 Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xqiLW-0000aB-00; Fri, 9 Jan 1998 17:35:50 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xqiLV-0007LK-00; Fri, 9 Jan 1998 17:35:49 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Russ King cc: linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu Subject: Re: PLIP In-reply-to: Your message of "Fri, 09 Jan 1998 13:54:03 GMT." <22474.9801091354@diana.ecs.soton.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 09 Jan 1998 17:35:48 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO In message <22474.9801091354@diana.ecs.soton.ac.uk>, Russ King writes: >Philip Blundell writes: >> Presumably you mean that the interrupt sense is inverted, rather than the li >ne >> itself. Does IOC not give you a way to change that? > >No, the ACK itself is inverted, not the interrupt sense. Ah. Oh well, I think the same is still true. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sat Jan 10 12:53:49 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id MAA19137 for ; Sat, 10 Jan 1998 12:53:43 GMT Received: from vger.rutgers.edu ([128.6.190.2]:321 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <16387-21270>; Sat, 10 Jan 1998 14:53:33 +0200 Received: by vger.rutgers.edu id <970956-241>; Sat, 10 Jan 1998 07:51:31 -0500 Received: from mailer3.bham.ac.uk ([147.188.128.54]:48345 "HELO mailer3.bham.ac.uk" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970896-241>; Sat, 10 Jan 1998 07:51:10 -0500 Received: from bham.ac.uk by mailer3.bham.ac.uk with SMTP (XT-PP) with ESMTP; Sat, 10 Jan 1998 12:52:05 +0000 Received: from eee.bham.ac.uk by bham.ac.uk with SMTP (XT-PP); Sat, 10 Jan 1998 12:52:04 +0000 Received: from nampa by eee.bham.ac.uk (SMI-8.6/SMI-SVR4) id MAA28576; Sat, 10 Jan 1998 12:51:59 GMT Message-ID: <34B76EED.3869@eee.bham.ac.uk> Date: Sat, 10 Jan 1998 12:51:57 +0000 From: "I.V.McLoughlin" Organization: The University of Birmingham X-Mailer: Mozilla 3.02 (X11; I; SunOS 5.5.1 sun4u) MIME-Version: 1.0 To: linux-arm@vger.rutgers.edu Subject: ARMlinux vs. RiscBSD Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi, I don't mean to be contravercial-my question is a serious one. I'm a user of Solaris (on Suns) , RiscBSD (on my RPC) and Linux (only on PC atm). RiscBSD is a great product and works well on my SA-RPC, but I really need to know how this compares against ARM Linux. Is there anybody that has used both RiscBSD 1.3 and a current version of ARM Linux: what features do both have do they lack any features. Some questions in particular - can ARM Linux users access the ADFS hard discs from within Linux, and can they read/write the Linux partition from ADFS too? Which is a more 'efficient' implementation? Does ARM Linux use shared libraries? Is it a stable platform for commercial code development? I think we're lucky to have multiple UNIX versions available for our Acorns - and I congratulate the developers who've made this possible. Multiple versions give the user more choice and bring Acorns out more into the 'real world of computing'. But it does make the choice a little complicated from the users point of view, so any help that people can give me will be appreciated. Perhaps off-list replies would be better unless others also require this information. tia, Ian McLoughlin -- Dr. I.V.McLoughlin [for Chinese, S.E.Asia, Speech & DSP] |_ School of Electronic and Electrical Engineering, | The University of Birmingham, Edgbaston, Birmingham. |_ Simoco Telecommunications Ltd., Cambridge |_ http://sun1.bham.ac.uk/i.v.mcloughlin unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 11 01:52:19 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id BAA24008 for ; Sun, 11 Jan 1998 01:52:17 GMT Received: from vger.rutgers.edu ([128.6.190.2]:2396 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <20520-27418>; Sun, 11 Jan 1998 03:45:05 +0200 Received: by vger.rutgers.edu id <970953-241>; Sat, 10 Jan 1998 20:38:48 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:1933 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <970983-241>; Sat, 10 Jan 1998 20:38:12 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id BAA23770; Sun, 11 Jan 1998 01:38:12 GMT From: Matthew Wilcox Message-Id: <199801110138.BAA23770@odie.barnet.ac.uk> Subject: Re: ARMlinux vs. RiscBSD To: mcloughi@eee.bham.ac.uk (I.V.McLoughlin) Date: Sun, 11 Jan 1998 01:38:12 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: <34B76EED.3869@eee.bham.ac.uk> from "I.V.McLoughlin" at Jan 10, 98 12:51:57 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > I don't mean to be contravercial-my question is a serious > one. Understood. Bear in mind that any reply you get is liable to be ill-informed, and/or based upon the prejudices of the respondent. I do not except myself from this. > RiscBSD is a great product and works well on my SA-RPC, > but I really need to know how this compares against > ARM Linux. RiscBSD is somewhat more mature than ARMLinux. This means it has a greater installed base of users and supports more hardware. Against that, there's vastly more software available specifically for Linux than NetBSD. Personally, I deplore this, but that's the way of life it seems ;-( > Some questions in particular - can ARM Linux users access > the ADFS hard discs from within Linux, and can they read/write > the Linux partition from ADFS too? The ADFS hard disc is readable, but not writable. This isn't something that's fundamentally missing, it's simply that no-one's implemented it yet. Thanks to Forrey (Phil Norman), there's an ext2 partition reader for RISC OS, but again it's not writable. It's also not a proper filing system, it's very much like SparkPlug. > Which is a more 'efficient' implementation? Does ARM Linux > use shared libraries? Is it a stable platform for commercial > code development? ARM Linux has better performance, I've been told. Yes, ARMLinux has shared libraries, but they're based around the libc4. Until we have ELF, we cannot migrate to libc6. Work is in progress on this. There are a few problems with libc4 that I simply haven't bothered to trace, because they will go away when libc 6 is introduced. I'd call it stable. It's frozen solid on me once in the last week, and I don't exactly treat it gently. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From neil@causality.com Sun Jan 11 02:24:24 1998 Return-Path: Received: from mailhub.nc.com (proxy@mailhub.nc.com [207.88.25.3]) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id CAA24187 for ; Sun, 11 Jan 1998 02:24:21 GMT Received: (from proxy@localhost) by mailhub.nc.com (8.6.12/8.6.9) id SAA31643; Sat, 10 Jan 1998 18:25:36 -0800 Received: from riscbsd1.client.nc.com(172.17.8.88) by mailhub.nc.com via smap (V2.0) id xma031628; Sat, 10 Jan 98 18:25:16 -0800 Sender: neil@nc.com Message-ID: <34B82E5B.41C67EA6@causality.com> Date: Sun, 11 Jan 1998 02:28:43 +0000 From: "Neil A. Carson" Organization: Causality Limited (London, UK) X-Mailer: Mozilla 3.01 (X11; I; NetBSD 1.2G arm32) MIME-Version: 1.0 To: Matthew Wilcox CC: "I.V.McLoughlin" , linux-arm@vger.rutgers.edu Subject: Re: ARMlinux vs. RiscBSD References: <199801110138.BAA23770@odie.barnet.ac.uk> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Status: RO Matthew Wilcox wrote: > Understood. Bear in mind that any reply you get is liable to be > ill-informed, and/or based upon the prejudices of the respondent. > I do not except myself from this. Neither do I :-) but I try my best. > RiscBSD is somewhat more mature than ARMLinux. This means it has a > greater installed base of users and supports more hardware. Against that, > there's vastly more software available specifically for Linux than NetBSD. > Personally, I deplore this, but that's the way of life it seems ;-( I wouldn't agree with this; any changes in software compiling that need to be made are normally because of the ARM, not because of the host OS. IN a similar vein to all the Linux distributions, there are all the NetBSD 'packages' (around two gigabytes of source) that build straight away. In any case, we'll be seeing StarOffice for NetBSD/arm32 as and when StarDivision finish it off :-) and maybe also Netscape when the Communicator port is finished (Nav 3 attempt failed, but the Communicator licencing is different). > The ADFS hard disc is readable, but not writable. This isn't something > that's fundamentally missing, it's simply that no-one's implemented it > yet. > Thanks to Forrey (Phil Norman), there's an ext2 partition reader for > RISC OS, but again it's not writable. It's also not a proper filing > system, it's very much like SparkPlug. Yep, on RiscBSD there's no way of getting at the ADFS partition from inside it (although someone has written such a utility) but you do have the full read/write from RiscOS. > > Which is a more 'efficient' implementation? Does ARM Linux > > use shared libraries? Is it a stable platform for commercial > > code development? > > ARM Linux has better performance, I've been told. Yes, ARMLinux has shared > libraries, but they're based around the libc4. Until we have ELF, we cannot > migrate to libc6. Work is in progress on this. There are a few problems > with libc4 that I simply haven't bothered to trace, because they will go > away when libc 6 is introduced. ARM Linux forks quite a bit faster than RiscBSD (about 10X :-) ). This is because the Mach VM in RiscBSD was written completely without virtual split caches in mind. Other than that, performance is pretty much the same---both systems being hit by the context switch time on StrongARM which figures largely in 'internal latency' benchmarks. RiscBSD's shared library system is better, not requiring fixed addresses and using PIC code instead (which can occasionally cause problems compiling programs without support for the older Linux library model). Well I think I've tried to be unbiased :-) Regards, Neil From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 11 03:30:07 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id DAA24584 for ; Sun, 11 Jan 1998 03:30:05 GMT Received: from vger.rutgers.edu ([128.6.190.2]:52026 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11848-27083>; Sun, 11 Jan 1998 05:27:20 +0200 Received: by vger.rutgers.edu id <970994-241>; Sat, 10 Jan 1998 22:24:59 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:1222 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <971003-241>; Sat, 10 Jan 1998 22:24:25 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id DAA24573; Sun, 11 Jan 1998 03:23:06 GMT From: Matthew Wilcox Message-Id: <199801110323.DAA24573@odie.barnet.ac.uk> Subject: Re: ARMlinux vs. RiscBSD To: neil@causality.com (Neil A. Carson) Date: Sun, 11 Jan 1998 03:23:05 +0000 (GMT) Cc: mcloughi@eee.bham.ac.uk, linux-arm@vger.rutgers.edu In-Reply-To: <34B82E5B.41C67EA6@causality.com> from "Neil A. Carson" at Jan 11, 98 02:28:43 am X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > > RiscBSD is somewhat more mature than ARMLinux. This means it has a > > greater installed base of users and supports more hardware. Against that, > > there's vastly more software available specifically for Linux than NetBSD. > > Personally, I deplore this, but that's the way of life it seems ;-( > > I wouldn't agree with this; any changes in software compiling that need > to be made are normally because of the ARM, not because of the host OS. Possibly. I've certainly seen a lot of problems with programs assuming signed chars. Since both ports use GCC, and gcc plain chars default to unsigned on the ARM (for efficiency reasons), this causes programs to not compile out of the box on either OS. The most common example I've seen of this in Linux utilities is comparing the `int' return from getopt to a variable of type `char'. If you ever see a warning about comparison always being 0 or 1 due to limited range of data type, suspect this sort of problem and investigate. > IN a similar vein to all the Linux distributions, there are all the > NetBSD 'packages' (around two gigabytes of source) that build straight > away. In any case, we'll be seeing StarOffice for NetBSD/arm32 as and > when StarDivision finish it off :-) and maybe also Netscape when the > Communicator port is finished (Nav 3 attempt failed, but the > Communicator licencing is different). Hmm. Anyone fancy writing aBCS? Or forming a working committee to discuss an arm-open binary standard? ;-) > ARM Linux forks quite a bit faster than RiscBSD (about 10X :-) ). This > is because the Mach VM in RiscBSD was written completely without virtual > split caches in mind. Other than that, performance is pretty much the > same---both systems being hit by the context switch time on StrongARM > which figures largely in 'internal latency' benchmarks. According to a friend who's run 2 ARM610 RiscPCs side by side, ARMLinux feels more responsive in practice. > RiscBSD's shared > library system is better, not requiring fixed addresses and using PIC > code instead (which can occasionally cause problems compiling programs > without support for the older Linux library model). I agree. The RiscBSD 1.3 library system is definitely superior to the current a.out based ARM Linux system. However, once ELF happens and we move to glibc, I believe the Linux system to be superior. I have to admit that's more a gut feeling that 100% based on experiment ;-) > Well I think I've tried to be unbiased :-) Except you didn't include your sig declaring you work for Causality ;-) (though you'd've had to have your head under a rock for the past 2 years to not realise Neil was an original memeber of the RiscBSD project ;-) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From neil@causality.com Sun Jan 11 04:51:21 1998 Return-Path: Received: from mailhub.nc.com (proxy@mailhub.nc.com [207.88.25.3]) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id EAA25122 for ; Sun, 11 Jan 1998 04:51:19 GMT Received: (from proxy@localhost) by mailhub.nc.com (8.6.12/8.6.9) id UAA05326; Sat, 10 Jan 1998 20:52:34 -0800 Received: from riscbsd1.client.nc.com(172.17.8.88) by mailhub.nc.com via smap (V2.0) id xma005321; Sat, 10 Jan 98 20:52:31 -0800 Sender: neil@nc.com Message-ID: <34B850DF.167EB0E7@causality.com> Date: Sun, 11 Jan 1998 04:55:59 +0000 From: "Neil A. Carson" Organization: Causality Limited (London, UK) X-Mailer: Mozilla 3.01 (X11; I; NetBSD 1.2G arm32) MIME-Version: 1.0 To: Matthew Wilcox CC: mcloughi@eee.bham.ac.uk, linux-arm@vger.rutgers.edu Subject: Re: ARMlinux vs. RiscBSD References: <199801110323.DAA24573@odie.barnet.ac.uk> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Status: RO Matthew Wilcox wrote: > Hmm. Anyone fancy writing aBCS? Or forming a working committee to > discuss an arm-open binary standard? ;-) :-) > > ARM Linux forks quite a bit faster than RiscBSD (about 10X :-) ). This > > is because the Mach VM in RiscBSD was written completely without virtual > > split caches in mind. Other than that, performance is pretty much the > > same---both systems being hit by the context switch time on StrongARM > > which figures largely in 'internal latency' benchmarks. > > According to a friend who's run 2 ARM610 RiscPCs side by side, ARMLinux > feels more responsive in practice. You noticed I wrote 10X not 10%. Binaries will run at the same speed on both, but Linux will load stuff a damn site quicker and also handle processes that fork better (eg PostGres server forks on each new connection, though we are working with a VM guy to try and fix this). In general this adds to the feel of responsiveness of the system, making Linux much more 'snappy.' Some things will detract---RiscBSD's Floating Point is several times faster than Linux's for example, but I think overall Linux is reckoned to be more responsive when running stuff. I think in X with some programs already running you'd be hard pushed to notice any difference. Also, one has to consider scalability in these things---whereas under light load a Linux system will often outperform most other counterparts, certain areas of eg the VM system will scale with L^2 in terms of their efficiency (where L is the load on the system for example) whereas on *BSD this often tends to be a flatter (though still not completely linear, obviously), more proportional relationship. This is why FreeBSD and NetBSD tend to be very popular in heavy server applications, but less in workstation applications. Though FreeBSD offers similar figures to Linux at the lower end as well :) offering stuff like kernel threads that RiscBSD currently lacks. > > RiscBSD's shared > > library system is better, not requiring fixed addresses and using PIC > > code instead (which can occasionally cause problems compiling programs > > without support for the older Linux library model). > > I agree. The RiscBSD 1.3 library system is definitely superior to the > current a.out based ARM Linux system. However, once ELF happens and we > move to glibc, I believe the Linux system to be superior. I have to > admit that's more a gut feeling that 100% based on experiment ;-) ELF eases stuff like C++ debugging, and when gcc 2.8 makes it I believe will offer significant advantages. We've done most of the stuff to make ELF possible in our a.out-PIC stuff, but we thought we'd wait on you chaps to do the last leg :-) The main thing for us was to get a proper mechanism going with the current toolchain (as NetBSD never supported the fixed address stuff) but we stopped there because of other committments. So Linux will have a definate lead on this until we catch up. > > Well I think I've tried to be unbiased :-) > > Except you didn't include your sig declaring you work for Causality ;-) > > (though you'd've had to have your head under a rock for the past 2 years > to not realise Neil was an original memeber of the RiscBSD project ;-) Indeed! Just over three years in fact, how time flies :-) Though I try to think I can speak about these things with my head screwed on properly and without bias: I have ARM Linux installed myself (though I prefer to use VMS than any UNIX anyway, OK go on laugh :) ) and have used it for some stuff, and it is more snappy. Really it's horses for courses, as I said in a recent article on the two OSs. If you want solidity, filesystem reliability, fast FP, a fully working toolchain/C++ (yes I know we broke C++ recently, but a fix is going out now!), easiy install etc go for RiscBSD. If you want something that runs as fast as possible, has the lowest memory requirements, and is more popular/funky, choose ARMLinux. If you want both, buy a PC and put FreeBSD on it :-) Regards, Neil From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 11 19:58:26 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id TAA30616 for ; Sun, 11 Jan 1998 19:58:21 GMT Received: from vger.rutgers.edu ([128.6.190.2]:56673 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <20488-27083>; Sun, 11 Jan 1998 21:17:25 +0200 Received: by vger.rutgers.edu id <971267-241>; Sun, 11 Jan 1998 08:36:13 -0500 Received: from post-20.mail.demon.net ([194.217.242.27]:40561 "HELO post.mail.demon.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <971245-241>; Sun, 11 Jan 1998 08:36:00 -0500 Received: from latrigg.demon.co.uk ([158.152.64.157]) by post.mail.demon.net id aa2002011; 11 Jan 98 13:33 GMT Date: Sun, 11 Jan 1998 13:33:48 +0000 (GMT) From: Chris Pringle Subject: Installing ARM Linux on an A5000 To: linux-arm@vger.rutgers.edu In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII X-Organization: At Home, 01285 713514 X-Mailer: ANT RISCOS Marcel [ver 1.26] X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO On the RedHat installer when it asks me where the RedHat RPM's are, it only gives me one device to choose from which is /dev/hda3, my linux partition. Is the A5000 kernel cabable of mounting ADFS partitions? Regards, --- Chris Pringle 22, St. Mary's Drive, Fairford, Glos. GL7 4LQ P.S.Check out My web site at http://www.latrigg.demon.co.uk/Chris.homepage.html unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 11 20:00:30 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id UAA30622 for ; Sun, 11 Jan 1998 20:00:29 GMT Received: from vger.rutgers.edu ([128.6.190.2]:56673 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <20473-27418>; Sun, 11 Jan 1998 21:19:24 +0200 Received: by vger.rutgers.edu id <971275-241>; Sun, 11 Jan 1998 08:36:51 -0500 Received: from post-20.mail.demon.net ([194.217.242.27]:40722 "HELO post.mail.demon.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <971274-241>; Sun, 11 Jan 1998 08:36:21 -0500 Received: from latrigg.demon.co.uk ([158.152.64.157]) by post.mail.demon.net id aa2002021; 11 Jan 98 13:33 GMT Date: Sun, 11 Jan 1998 13:32:57 +0000 (GMT) From: Chris Pringle Subject: ARM Linux on an A5000 with Ether3 To: linux-arm@vger.rutgers.edu In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII X-Organization: At Home, 01285 713514 X-Mailer: ANT RISCOS Marcel [ver 1.26] X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO How do I get the Ether3 card to work on the A5000. At the moment I am using the latest A5000 kernel with a RiscPC boot disk as the A5000 boot disk is as far as I know unavailable. What extra arguments do I need to add to get it to work. Regards, --- Chris Pringle 22, St. Mary's Drive, Fairford, Glos. GL7 4LQ P.S.Check out My web site at http://www.latrigg.demon.co.uk/Chris.homepage.html unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 11 20:26:33 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id UAA30800 for ; Sun, 11 Jan 1998 20:26:32 GMT Received: from vger.rutgers.edu ([128.6.190.2]:56673 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <16475-27083>; Sun, 11 Jan 1998 21:56:49 +0200 Received: by vger.rutgers.edu id <971340-241>; Sun, 11 Jan 1998 09:13:05 -0500 Received: from hermes.ex.ac.uk ([144.173.6.14]:11756 "EHLO exeter.ac.uk" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971063-241>; Sun, 11 Jan 1998 09:08:59 -0500 Received: from hebe [144.173.6.23] by hermes via SMTP (OAA03795); Sun, 11 Jan 1998 14:09:06 GMT Date: Sun, 11 Jan 1998 14:09:05 +0000 (GMT) From: Phil Norman X-Sender: py95pcfn@hebe To: Chris Pringle cc: linux-arm@vger.rutgers.edu Subject: Re: Installing ARM Linux on an A5000 In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO On Thu, 1 Jan 1998, Chris Pringle wrote: > Problem 2 : Installing ARM Linux using a Local Hard Disk image > How do I install ARM Linux using a local Hard Disk image? If it can read my > ADFS Drives do I have to put the files in a image file with long filenames > and no limitation to directory size? If so how do I create them, do I use > PCem or is the partition type different? From what I can remember DOSDisc's can > not have long file names. If you put them in an image file the ADFS linux read-only FS won't read them (although I've heard there is an x-files archive reader somewhere). If you use LongFileNames, linux will see all the files with long names as (C)JPT_xxx where xxx is a number. If you need more than 77 RPMs to be installed, the best way is to zip them all up (with their long filenames) on ADFS and then get linux to unzip the zip file from your adfs partition onto your ext2 partition, and then install from there. -- Phil Norman, mailing from Exeter Uni. Programs available for download from my web page. email: p.c.f.norman@ex.ac.uk web: http://newton.ex.ac.uk/general/ug/norman unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 11 21:11:50 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id VAA31059 for ; Sun, 11 Jan 1998 21:11:49 GMT Received: from vger.rutgers.edu ([128.6.190.2]:56673 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <17800-27083>; Sun, 11 Jan 1998 22:35:20 +0200 Received: by vger.rutgers.edu id <971348-241>; Sun, 11 Jan 1998 09:29:15 -0500 Received: from hermes.ex.ac.uk ([144.173.6.14]:12262 "EHLO exeter.ac.uk" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971083-241>; Sun, 11 Jan 1998 09:28:10 -0500 Received: from hebe [144.173.6.23] by hermes via SMTP (OAA04205); Sun, 11 Jan 1998 14:28:29 GMT Date: Sun, 11 Jan 1998 14:28:29 +0000 (GMT) From: Phil Norman X-Sender: py95pcfn@hebe To: ARM Linux Subject: ExeterFS is dead! Long live IscaFS! Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi! I haven't fixed all the bugs in ExeterFS; I've killed it off instead. Since so many people were moaning about how there should be an image filing system ext2 reader I've written one. It's called IscaFS and, like ExeterFS, includes full source code. However, there are a few differences in IscaFS: 1: It's written in hand-coded ARM assembly language. 2: It uses the same block cacheing method as linux does, and so is a little faster than the frankly snail-like ExeterFS. 3: It is an image filing system, so you can do all those funky things you can do with all filing systems, and you don't need to worry about a bug-ridden wimp front-end. 4: It handles floppy discs 'invisibly' (just as DOSFS handles DOS floppies without you noticing it). 5: It handles ext2 partitions by using pseudo image files. A program is supplied with the IscaFS release which creates them for you. It also has only two bugs (AFAIK). The first is shared with ExeterFS in that it doesn't understand what a symbolic link is. Double-click on a symblink if you REALLY want to, but don't expect anything more than toflatoflatoflatofla (I would include the real characters but they're high ASCII, so the emailing it wouldn't work). The second bug is a very weird one and occurs when IscaFS tries to get the contents of a directory in response to an OS_GBPB,10-like call. It's probably something to do with the offset, but it's only happened when I used my cunning IscaFS directory-contents-getting testing method (using the FilerAction count facility). It works for the most part, but just occasionally goes horribly wrong. Sometimes it tries to read a part of the disc that doesn't exist, and sometimes when reading directory contents it reads files that aren't there. It only happens occasionally, which is even worse from a debugging point of view. I might turn IscaFS into a read/write image filing system one day, but first I'm going to ensure that ALL the bugs have been removed from the read-only version. If anyone finds a bug, PLEASE let me know. I hope someone finds this useful. Bye for now, Phil P.S. You can get IscaFS with full source code from my web page (in the sig). PLEASE ENSURE THAT YOU READ THE !ReadMe FILE _FIRST_. -- Phil Norman, mailing from Exeter Uni. Programs available for download from my web page. email: p.c.f.norman@ex.ac.uk web: http://newton.ex.ac.uk/general/ug/norman unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 12 00:13:42 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id AAA32458 for ; Mon, 12 Jan 1998 00:13:40 GMT Received: from vger.rutgers.edu ([128.6.190.2]:56673 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <7094-27083>; Mon, 12 Jan 1998 02:06:17 +0200 Received: by vger.rutgers.edu id <971678-241>; Sun, 11 Jan 1998 11:00:55 -0500 Received: from mail.tin.it ([194.243.154.39]:49315 "EHLO fes3.cs.tin.it" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971646-241>; Sun, 11 Jan 1998 10:57:36 -0500 Received: from riscpc (Monza4-48.tin.it [195.31.128.47]) by fes3.cs.tin.it (8.8.4/8.8.4) with SMTP id QAA08547 for ; Sun, 11 Jan 1998 16:58:56 +0100 (MET) Date: Sun, 11 Jan 1998 12:00:48 +0100 From: Sergio Monesi To: linux-arm@vger.rutgers.edu Subject: Re: ARMlinux vs. RiscBSD Message-ID: <4806BA67DE%msergio@tin.it> In-Reply-To: <34B850DF.167EB0E7@causality.com> Organization: None, AFAIK... X-Mailer: Messenger v1.11 for RISC OS X-Editor: Zap, using ZapEmail =?ISO-8859-1?Q?0.21=DF?= X-Posting-Agent: RISC OS Newsbase 0.59h X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO In message <34B850DF.167EB0E7@causality.com> "Neil A. Carson" wrote: > You noticed I wrote 10X not 10%. Binaries will run at the same speed on > both, but Linux will load stuff a damn site quicker and also handle > processes that fork better (eg PostGres server forks on each new > connection, though we are working with a VM guy to try and fix this). In > general this adds to the feel of responsiveness of the system, making > Linux much more 'snappy.' Apart from this fork problem (try to run "find -exec grep" or something similar on RiscBSD!), RiscBSD's disc caching seems to be far inferior compared to ARMLinux and, actually, it doesn't seem to work at all on my machine (24Mb DRAM)! In fact, the second time I run, say, "ls -lR" (or even something more complex with some pipe) on Linux I got an instant output and no disc access while there is no noticeable difference between the first and second run of that on RiscBSD. Maybe I should configure more buffers (I think it is possible to do this from the bootloader) but I suppose that Linux uses a dynamically resizeable buffer area... Also, I got no problems connecting to Internet using PPP on ARMLinux but I remember that I had many problems with RiscBSD when I tried ages ago (1.1 release probably). > I think in X with some programs already running you'd be hard pushed to > notice any difference. "find / -exec grep ..." in an Xterm? :-) > If you want [...] a fully working toolchain/C++ [...] go for RiscBSD. I tried to compile Lesstif and Afterstep on ARMLinux without much success (especially with the former). I don't know if these work as expected on RiscBSD... About 'solidity' and 'filesystem reliability', I wouldn't say this for RiscBSD, last time I tried it with 12Mb of RAM many programs crashed (eg. gcc, X) due to incorrect handling of many page faults and sometimes the kernel crashed as well. This doesn't seem to happen with more memory but I wouldn't rate the system as 'stable' with such a bug in the kernel... Please correct me if I am wrong and this has been fixed recently! I never tried ARMLinux with 12Mb of RAM so I can't comment on it... Sergio -- Sergio Monesi... \ . . . \ Cracking RC5-64 with a StrongARM RiscPC mailto:msergio@tin.it \ . . . \ http://rc5.distributed.net/ http://bounce.to/sergio/ \ . . . \ Join the Wild Wereboar Team (#1308)! unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 12 07:56:35 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id HAA02213 for ; Mon, 12 Jan 1998 07:56:22 GMT Received: from vger.rutgers.edu ([128.6.190.2]:25356 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11509-27418>; Mon, 12 Jan 1998 08:21:14 +0200 Received: by vger.rutgers.edu id <971062-241>; Mon, 12 Jan 1998 01:17:48 -0500 Received: from mailhub.nc.com ([207.88.25.3]:24679 "HELO mailhub.nc.com" ident: "proxy") by vger.rutgers.edu with SMTP id <971070-241>; Mon, 12 Jan 1998 01:16:53 -0500 Received: (from proxy@localhost) by mailhub.nc.com (8.6.12/8.6.9) id WAA08094; Sun, 11 Jan 1998 22:18:16 -0800 Received: from riscbsd1.client.nc.com(172.17.8.88) by mailhub.nc.com via smap (V2.0) id xma008074; Sun, 11 Jan 98 22:17:52 -0800 Message-ID: <34B9B3BB.41C67EA6@causality.com> Date: Mon, 12 Jan 1998 06:10:03 +0000 From: "Neil A. Carson" Organization: Causality Limited (London, UK) X-Mailer: Mozilla 3.01 (X11; I; NetBSD 1.2G arm32) MIME-Version: 1.0 To: Sergio Monesi CC: linux-arm@vger.rutgers.edu Subject: Re: ARMlinux vs. RiscBSD References: <4806BA67DE%msergio@tin.it> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Sergio Monesi wrote: > In fact, the second time I run, say, "ls -lR" (or even something more complex > with some pipe) on Linux I got an instant output and no disc access while > there is no noticeable difference between the first and second run of that on > RiscBSD. There wouldn't be. As I've said before, these things aren't clean+dry that simple---it's different philosophies behind the OSs. What happens, is that the FFS filesystem synchronously writes through critical meta-data to the filesystem in order to help guarantee a consistent state in case the FS crashes due to power failure. This makes the system more robust, but slower on certain operations (opening, closing, creating and deleting a file---reads and writes of data will be the same). Changing allocated buffers won't help. What you really want, again, is FreeBSDs new IOVM system :) > Also, I got no problems connecting to Internet using PPP on ARMLinux but I > remember that I had many problems with RiscBSD when I tried ages ago (1.1 > release probably). 1.1 is ages old, to be fair. ppp works perfectly fine with RiscBSD, and many users have modems connected with it. > > I think in X with some programs already running you'd be hard pushed to > > notice any difference. > > "find / -exec grep ..." in an Xterm? :-) :-) > > If you want [...] a fully working toolchain/C++ [...] go for RiscBSD. > > I tried to compile Lesstif and Afterstep on ARMLinux without much success > (especially with the former). I don't know if these work as expected on > RiscBSD... Lesstif (also QT, KDE etc) certainly do, I haven't tried afterstep. There shouldn't be a problem running with 12MB RAM, and the X server is pretty stable these days and supports several different bpps---and things like font realisation are much faster because of the much better FP capability. What really matters is not "what knocks what on what," but really the differing philosophies and planned uses of the OSs. I've never been a great fan of OS advocacy (apart from the BeOS!), certainly not as regards UNIXs anyway. My advice is for a single user meddling system run Linux, for a file server run RiscBSD, and for the best of all worlds use FreeBSD on a PC. Regards, Neil unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 12 09:46:32 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id JAA02942 for ; Mon, 12 Jan 1998 09:46:30 GMT Received: from vger.rutgers.edu ([128.6.190.2]:20568 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <16480-150>; Mon, 12 Jan 1998 11:40:54 +0200 Received: by vger.rutgers.edu id <970912-241>; Mon, 12 Jan 1998 04:38:11 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1840 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970903-241>; Mon, 12 Jan 1998 04:34:07 -0500 Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id JAA00600; Mon, 12 Jan 1998 09:35:06 GMT From: Alan Cox Message-Id: <199801120935.JAA00600@snowcrash.cymru.net> Subject: Re: ARMlinux vs. RiscBSD To: neil@causality.com (Neil A. Carson) Date: Mon, 12 Jan 1998 09:35:04 +0000 (GMT) Cc: msergio@tin.it, linux-arm@vger.rutgers.edu In-Reply-To: <34B9B3BB.41C67EA6@causality.com> from "Neil A. Carson" at Jan 12, 98 06:10:03 am Content-Type: text X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > state in case the FS crashes due to power failure. This makes the system > more robust, but slower on certain operations (opening, closing, > creating and deleting a file---reads and writes of data will be the > same). Actually there are fairly comprehensive proofs that synchronous metadata increases the probability of corruption because it produces cases where the metadata (ie the file system structure) is correct but the data within the files was never written. What we both want is a logging ext2/ufs. As far as I am aware both are in progress. Its one of the important items for an embedded system unless you want to include a UPS instead ;) Alan unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 12 14:53:01 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id OAA05171 for ; Mon, 12 Jan 1998 14:53:00 GMT Received: from vger.rutgers.edu ([128.6.190.2]:61041 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13426-27083>; Mon, 12 Jan 1998 16:40:45 +0200 Received: by vger.rutgers.edu id <970979-241>; Mon, 12 Jan 1998 09:34:00 -0500 Received: from out4.ibm.net ([165.87.194.239]:2609 "EHLO out4.ibm.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971051-241>; Mon, 12 Jan 1998 09:23:46 -0500 Received: from cvs.com.au (slip166-72-68-78.fl.us.ibm.net [166.72.68.78]) by out4.ibm.net (8.8.5/8.6.9) with ESMTP id OAA39334 for ; Mon, 12 Jan 1998 14:25:34 GMT Message-ID: <34BA2862.FB3A9B98@cvs.com.au> Date: Mon, 12 Jan 1998 09:27:46 -0500 From: Charles Esson Reply-To: charlese@cvs.com.au Organization: Colour Vision Systems X-Mailer: Mozilla 4.04 [en] (WinNT; I) MIME-Version: 1.0 To: "linux-arm@vger.rutgers.edu" Subject: Re: ARMlinux vs. RiscBSD References: <199801120935.JAA00600@snowcrash.cymru.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Alan Cox wrote: > > state in case the FS crashes due to power failure. This makes the system > > more robust, but slower on certain operations (opening, closing, > > creating and deleting a file---reads and writes of data will be the > > same). > > Actually there are fairly comprehensive proofs that synchronous metadata > increases the probability of corruption because it produces cases where the > metadata (ie the file system structure) is correct but the data within the > files was never written. > Dam right, a correct directory might be nice for the kernel but it is of little use to the application, if the data is rubbish. The big differences are: 1) The license terms, read both carefully if you want to use it for commercial purposes, and pick the one you prefer. Your choice. 2) The standard the systems are based on, I think this is of little relevance but some might think otherwise. Your prejudice 3) Popularity of linux and riskBSD in general. Do a web search to form your own opinion. 4) Development activity. I think riskBSD were the winner, but I could be wrong, I think linux is catching up, now if Causality where to swap to Linux. Me I picked linux, given time I think the ARM version will catch up to RiscBSD, and there is a greater chance the people I hire and the customers I sell to will of heard of it. It was not a technical choice, technicaly I don't like unix or its derivatives for real time control. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 12 15:30:31 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id PAA05405 for ; Mon, 12 Jan 1998 15:30:29 GMT Received: from vger.rutgers.edu ([128.6.190.2]:22071 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10297-150>; Mon, 12 Jan 1998 17:23:30 +0200 Received: by vger.rutgers.edu id <970994-241>; Mon, 12 Jan 1998 10:19:58 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1652 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970963-241>; Mon, 12 Jan 1998 10:17:54 -0500 Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id OAA05910; Mon, 12 Jan 1998 14:54:58 GMT From: Alan Cox Message-Id: <199801121454.OAA05910@snowcrash.cymru.net> Subject: Re: ARMlinux vs. RiscBSD To: charlese@cvs.com.au Date: Mon, 12 Jan 1998 14:54:55 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: <34BA2862.FB3A9B98@cvs.com.au> from "Charles Esson" at Jan 12, 98 09:27:46 am Content-Type: text X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > It was not a technical choice, technicaly I don't like unix or its derivatives > for real time control. Just out of curiousity have you ever looked at the RTLinux work at NMT, putting Linux (or in theory any other OS if tweaked) on top of a very small hard real time kernel. It runs hard real time tasks in the RT kernel and they can talk via queues etc to the non hard RT Unix on top of it Alan unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 12 18:46:14 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id SAA06840 for ; Mon, 12 Jan 1998 18:46:13 GMT Received: from vger.rutgers.edu ([128.6.190.2]:36474 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <6844-150>; Mon, 12 Jan 1998 20:26:12 +0200 Received: by vger.rutgers.edu id <971186-241>; Mon, 12 Jan 1998 13:11:23 -0500 Received: from mail.tin.it ([194.243.154.39]:35023 "EHLO fes3.cs.tin.it" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971296-241>; Mon, 12 Jan 1998 12:59:41 -0500 Received: from riscpc (root@Monza7-133.tin.it [195.103.249.132]) by fes3.cs.tin.it (8.8.4/8.8.4) with SMTP id SAA14585; Mon, 12 Jan 1998 18:40:06 +0100 (MET) Date: Mon, 12 Jan 1998 12:01:18 +0100 From: Sergio Monesi To: "Neil A. Carson" Cc: linux-arm@vger.rutgers.edu Subject: Re: ARMlinux vs. RiscBSD Message-ID: <48073E4995%msergio@tin.it> In-Reply-To: <34B9B3BB.41C67EA6@causality.com> Organization: None, AFAIK... X-Mailer: Messenger v1.11 for RISC OS X-Editor: Zap, using ZapEmail =?ISO-8859-1?Q?0.21=DF?= X-Posting-Agent: RISC OS Newsbase 0.59h X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO I'm sorry if I seemed to criticize RiscBSD without knowing what I was speaking about... However, I think that most Acorn users prefers to have a UNIX tuned for a low-level workstation, not for a top-level server which the RiscPC obviously isn't, so Linux seems to be the right OS for us... In message <34B9B3BB.41C67EA6@causality.com> "Neil A. Carson" wrote: > Sergio Monesi wrote: > > In fact, the second time I run, say, "ls -lR" (or even something more > > complex with some pipe) on Linux I got an instant output and no disc > > access while there is no noticeable difference between the first and > > second run of that on RiscBSD. > There wouldn't be. As I've said before, these things aren't clean+dry > that simple---it's different philosophies behind the OSs. What happens, > is that the FFS filesystem synchronously writes through critical > meta-data to the filesystem in order to help guarantee a consistent > state in case the FS crashes due to power failure. This makes the system > more robust, but slower on certain operations (opening, closing, > creating and deleting a file---reads and writes of data will be the > same). OK, I understand this, but where are the write/create/delete operations in "ls -lR"? I can't see any risk in buffering the data used by "ls -lR", whatever happens during its execution, the filesystem integrity cannot be compromised. Incidentally, the slowness of the RiscPC internal IDE impacts heavily on RiscBSD, let alone the SCSI drivers that seems to be terribly slow (from what I read on the mailing list)... How do you hope to use RiscBSD as a file server? > Changing allocated buffers won't help. What you really want, again, is > FreeBSDs new IOVM system :) Since you seem to appreciate almost every feature of FreeBSD, why didn't you (actually the RiscBSD kernel team) port it instead of NetBSD? Would it be hard to switch to FreeBSD now? What about porting to NetBSD all those nice peculiarities of FreeBSD? Cheers, Sergio -- Sergio Monesi... \ . . . \ Cracking RC5-64 with a StrongARM RiscPC mailto:msergio@tin.it \ . . . \ http://rc5.distributed.net/ http://bounce.to/sergio/ \ . . . \ Join the Wild Wereboar Team (#1308)! unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 12 19:56:58 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id TAA07271 for ; Mon, 12 Jan 1998 19:56:57 GMT Received: from vger.rutgers.edu ([128.6.190.2]:13424 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12762-21270>; Mon, 12 Jan 1998 21:41:15 +0200 Received: by vger.rutgers.edu id <970964-241>; Mon, 12 Jan 1998 13:51:20 -0500 Received: from mailhub.nc.com ([207.88.25.3]:28807 "HELO mailhub.nc.com" ident: "proxy") by vger.rutgers.edu with SMTP id <971118-241>; Mon, 12 Jan 1998 13:32:19 -0500 Received: (from proxy@localhost) by mailhub.nc.com (8.6.12/8.6.9) id KAA11025; Mon, 12 Jan 1998 10:33:28 -0800 Received: from riscbsd1.client.nc.com(172.17.8.88) by mailhub.nc.com via smap (V2.0) id xma011014; Mon, 12 Jan 98 10:33:21 -0800 Message-ID: <34BA61DA.41C67EA6@causality.com> Date: Mon, 12 Jan 1998 18:32:58 +0000 From: "Neil A. Carson" Organization: Causality Limited (London, UK) X-Mailer: Mozilla 3.01 (X11; I; NetBSD 1.2G arm32) MIME-Version: 1.0 To: Alan Cox CC: msergio@tin.it, linux-arm@vger.rutgers.edu Subject: Re: ARMlinux vs. RiscBSD References: <199801120935.JAA00600@snowcrash.cymru.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Alan Cox wrote: > Actually there are fairly comprehensive proofs that synchronous metadata > increases the probability of corruption because it produces cases where the > metadata (ie the file system structure) is correct but the data within the > files was never written. I think it's swings and roundabouts: McKusick (author of original FFS) is presently licencing some pretty cunning buffer cache code, which constructs a directed graph of the synchronous dependencies in the buffer cache, coalescing this ordering where possible and then doing the writes in order where required. He's currently making a bomb off this (it's pretty complex apparently), licencing it to Sun, companies with big servers, and such folks---there's clearly plenty of evidence to prove the opposite too which is of course why the thing was written that way in the first instance. So there's clearly a lot of people out there who've proven these proofs otherwise. I wouldn't have thought the chance of corruption would go up, because the filesystem will still be dirty anyway so stuff would be constructed automatically upon fsck. What I believe is that it reduces the likelyhood of manual intervention being required at fsck time (ie major inconsistencies that are not automatically solvable, rather than more minor ones) so that more files tend to be lost+founded or i-nodes cleared. > What we both want is a logging ext2/ufs. As far as I am aware both are in > progress. Its one of the important items for an embedded system unless you > want to include a UPS instead ;) When I've finished some stuff, I was going to write a journaling any-UNIX filesystem providing some features similar to SGIs xfs (if someone is also doing this now, let me know to save me wasting my effort later on :) ). Some recent research into log-structured filesystems (I believe there was a Linux log-structured project) by the FreeBSD guys---as they've finally got one of them working at last :) --- have suggested that these are in fact a waste of time compared to McKusicks funky stuff, apart from the decreased fsck time of course. Well, all interesting stuff :-) though I'll shut up as filesystem design is probably not a topic for this list! Cheers, Neil unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 12 19:57:19 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id TAA07275 for ; Mon, 12 Jan 1998 19:57:18 GMT Received: from vger.rutgers.edu ([128.6.190.2]:13424 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12666-27418>; Mon, 12 Jan 1998 21:41:01 +0200 Received: by vger.rutgers.edu id <970977-241>; Mon, 12 Jan 1998 13:50:56 -0500 Received: from mailhub.nc.com ([207.88.25.3]:28802 "HELO mailhub.nc.com" ident: "proxy") by vger.rutgers.edu with SMTP id <971095-241>; Mon, 12 Jan 1998 13:31:54 -0500 Received: (from proxy@localhost) by mailhub.nc.com (8.6.12/8.6.9) id KAA11028; Mon, 12 Jan 1998 10:33:28 -0800 Received: from riscbsd1.client.nc.com(172.17.8.88) by mailhub.nc.com via smap (V2.0) id xma011015; Mon, 12 Jan 98 10:33:22 -0800 Message-ID: <34BA62BC.167EB0E7@causality.com> Date: Mon, 12 Jan 1998 18:36:44 +0000 From: "Neil A. Carson" Organization: Causality Limited (London, UK) X-Mailer: Mozilla 3.01 (X11; I; NetBSD 1.2G arm32) MIME-Version: 1.0 To: charlese@cvs.com.au CC: "linux-arm@vger.rutgers.edu" Subject: Re: ARMlinux vs. RiscBSD References: <199801120935.JAA00600@snowcrash.cymru.net> <34BA2862.FB3A9B98@cvs.com.au> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Charles Esson wrote: > 4) Development activity. I think riskBSD were the winner, but I could be wrong, > I think linux is catching up, now if Causality where to swap to Linux. We tend to be quite agnostic about stuff I think---well I try anyway. I use loads of different OSs on a regular basis, all of which have merits and problems. If we were starting again now (we've been developing 'BSD for over 3.5 years now, after all :) ) we'd probably have chosen Linux, but at the time it was completely unportable which stopped us. If we were starting again in another three months time, we'd probably choose FreeBSD, as the current state of FreeBSD-current has one of the most advanced VMs (which will work extremely well with ARM) of any free OS. In a similar light, though, we find ourselves doing stuff with CPUs other than ARMs now, so the ARM is not 100% priority---more 75% and falling :-( Cheers, Neil unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 12 20:02:14 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id UAA07282 for ; Mon, 12 Jan 1998 20:02:13 GMT Received: from vger.rutgers.edu ([128.6.190.2]:55051 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11176-27083>; Mon, 12 Jan 1998 21:54:37 +0200 Received: by vger.rutgers.edu id <971081-241>; Mon, 12 Jan 1998 14:22:29 -0500 Received: from out4.ibm.net ([165.87.194.239]:1183 "EHLO out4.ibm.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971036-241>; Mon, 12 Jan 1998 13:44:22 -0500 Received: from cvs.com.au (slip166-72-68-78.fl.us.ibm.net [166.72.68.78]) by out4.ibm.net (8.8.5/8.6.9) with ESMTP id SAB04682 for ; Mon, 12 Jan 1998 18:46:08 GMT Message-ID: <34BA6573.3074431A@cvs.com.au> Date: Mon, 12 Jan 1998 13:48:19 -0500 From: Charles Esson Reply-To: charlese@cvs.com.au Organization: Colour Vision Systems X-Mailer: Mozilla 4.04 [en] (WinNT; I) MIME-Version: 1.0 To: "linux-arm@vger.rutgers.edu" Subject: Re: ARMlinux vs. RiscBSD References: <199801121454.OAA05910@snowcrash.cymru.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Alan Cox wrote: > It runs hard real time tasks in the RT kernel and they can talk via queues > etc to the non hard RT Unix on top of it. No I havn't I will take a look at it. Thanks. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 12 21:08:45 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id VAA07730 for ; Mon, 12 Jan 1998 21:08:44 GMT Received: from vger.rutgers.edu ([128.6.190.2]:18012 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12406-27418>; Mon, 12 Jan 1998 23:06:17 +0200 Received: by vger.rutgers.edu id <970866-241>; Mon, 12 Jan 1998 14:51:43 -0500 Received: from mailhub.nc.com ([207.88.25.3]:30064 "HELO mailhub.nc.com" ident: "proxy") by vger.rutgers.edu with SMTP id <971361-241>; Mon, 12 Jan 1998 14:28:22 -0500 Received: (from proxy@localhost) by mailhub.nc.com (8.6.12/8.6.9) id LAA14837; Mon, 12 Jan 1998 11:30:16 -0800 Received: from riscbsd1.client.nc.com(172.17.8.88) by mailhub.nc.com via smap (V2.0) id xma014815; Mon, 12 Jan 98 11:29:58 -0800 Message-ID: <34BA6FE3.FF6D5DF@causality.com> Date: Mon, 12 Jan 1998 19:32:51 +0000 From: "Neil A. Carson" Organization: Causality Limited (London, UK) X-Mailer: Mozilla 3.01 (X11; I; NetBSD 1.2G arm32) MIME-Version: 1.0 To: Sergio Monesi CC: linux-arm@vger.rutgers.edu Subject: Re: ARMlinux vs. RiscBSD References: <48073E4995%msergio@tin.it> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Sergio Monesi wrote: > However, I think that most Acorn users prefers to have a UNIX tuned for a > low-level workstation, not for a top-level server which the RiscPC obviously > isn't, so Linux seems to be the right OS for us... In the majority of cases, I'd agree with that! > OK, I understand this, but where are the write/create/delete operations in > "ls -lR"? I can't see any risk in buffering the data used by "ls -lR", > whatever happens during its execution, the filesystem integrity cannot be > compromised. It's the update of the inode-last-accessed times. Strictly this doesn't need to be done sychronously, but if you don't it creates problems when trying to do other things synchronously, so... There is an asynchronous mount option, which is nearly fully implemented now, which can (shortly) be used instead. > Incidentally, the slowness of the RiscPC internal IDE impacts heavily on > RiscBSD, let alone the SCSI drivers that seems to be terribly slow (from what > I read on the mailing list)... How do you hope to use RiscBSD as a file > server? RapIDE :) A lot of companies use a machine in this config for serving smallish networks of NCs (including Xemplar's standard small-network solutions, Acorn's clients etc) and, because most serving doesn't involve forking, it does pretty well. > Since you seem to appreciate almost every feature of FreeBSD, why didn't you > (actually the RiscBSD kernel team) port it instead of NetBSD? Would it be > hard to switch to FreeBSD now? What about porting to NetBSD all those nice > peculiarities of FreeBSD? FreeBSD would involve a fair bit of work, but it's being considered. Really FreeBSD has next to no BSD left in it anymore :-) Again initally it was unportable, like Linux---the main reason that we did NetBSD was it was the only portable OS at the time, and even now it still tends to be easier to maintain architecture ports than other free OSs. N. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 12 21:36:06 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id VAA07911 for ; Mon, 12 Jan 1998 21:36:05 GMT Received: from vger.rutgers.edu ([128.6.190.2]:18012 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10774-27083>; Mon, 12 Jan 1998 23:26:19 +0200 Received: by vger.rutgers.edu id <971178-241>; Mon, 12 Jan 1998 15:51:04 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1504 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971135-241>; Mon, 12 Jan 1998 15:00:54 -0500 Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id UAA11083; Mon, 12 Jan 1998 20:02:52 GMT From: Alan Cox Message-Id: <199801122002.UAA11083@snowcrash.cymru.net> Subject: Re: ARMlinux vs. RiscBSD To: neil@causality.com (Neil A. Carson) Date: Mon, 12 Jan 1998 20:02:51 +0000 (GMT) Cc: alan@cymru.net, msergio@tin.it, linux-arm@vger.rutgers.edu In-Reply-To: <34BA61DA.41C67EA6@causality.com> from "Neil A. Carson" at Jan 12, 98 06:32:58 pm Content-Type: text X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > the filesystem will still be dirty anyway so stuff would be constructed > automatically upon fsck. What I believe is that it reduces the What occurs is that the Linux case you lose the last few writes, the BSD case you lose the last few data writes > likelyhood of manual intervention being required at fsck time (ie major > inconsistencies that are not automatically solvable, rather than more > minor ones) so that more files tend to be lost+founded or i-nodes > cleared. Actually that isnt the case. It requires your fsck is a fraction smarter but not significantly. The main issue is deleted files. > Some recent research into log-structured filesystems (I believe there > was a Linux log-structured project) by the FreeBSD guys---as they've Yep. And theres also a journalled ext2fs project which is fairly similar to McKusicks work but a lot simpler in its approach. (NTfs btw is very similar in its algorithms to McKusicks work) > are in fact a waste of time compared to McKusicks funky stuff, apart > from the decreased fsck time of course. For ARM stuff the important thing to most folks is that if you flip the switch you cannto damage any data. Alan unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 12 21:50:19 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id VAA08086 for ; Mon, 12 Jan 1998 21:50:18 GMT Received: from vger.rutgers.edu ([128.6.190.2]:18012 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <6963-150>; Mon, 12 Jan 1998 23:33:29 +0200 Received: by vger.rutgers.edu id <971255-241>; Mon, 12 Jan 1998 15:55:15 -0500 Received: from mauve.csi.cam.ac.uk ([131.111.8.38]:42177 "HELO mauve.csi.cam.ac.uk" ident: "exim") by vger.rutgers.edu with SMTP id <971197-241>; Mon, 12 Jan 1998 15:22:22 -0500 Received: from jma24.robinson.cam.ac.uk [131.111.217.212] (exim) by mauve.csi.cam.ac.uk with smtp (Exim 1.82 #1) id 0xrqOr-0003qk-00; Mon, 12 Jan 1998 20:23:57 +0000 Received: from jma24.robinson.cam.ac.uk [127.0.0.1] (jma24) by jma24.robinson.cam.ac.uk with esmtp (Exim 1.73 #5) id 0xrqP8-0000xF-00; Mon, 12 Jan 1998 20:24:14 +0000 Date: Mon, 12 Jan 1998 20:24:14 +0000 (GMT) From: John Appleby Subject: ARM-Linux for the A5000 To: linux-arm@vger.rutgers.edu cc: rmk@ecs.soton.ac.uk MIME-Version: 1.0 Content-Type: TEXT/plain; CHARSET=US-ASCII Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hey everyone, I have an A5000 which I would like to run Linux on... I understand that the root disk isn't quite ready and isn't the same as the RISC-PC root disk. Is there a way of installing by hand to get it to boot, or a development root disk? There doesn't seem to be ANYTHING about this on the web... Cheers, ---- John Appleby, Pourquoi est-ce qu'un reve Robinson College, ne soit jamais realise? Cambridge University, Simone de Beauvoir - Les Belles Images UK. email: jma24@cam.ac.uk Tel.: +44 1223 460954 www: http://www.apples.net Fax.: +44 1344 22245 ntalk: jma24@cyril.apples.net -------------------------------------------------------------------------- Please do NOT send me ANY commercial email. I'm asking nicely. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 13 00:54:20 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id AAA09261 for ; Tue, 13 Jan 1998 00:54:19 GMT Received: from vger.rutgers.edu ([128.6.190.2]:25966 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <6531-27083>; Tue, 13 Jan 1998 02:43:23 +0200 Received: by vger.rutgers.edu id <971225-241>; Mon, 12 Jan 1998 18:53:58 -0500 Received: from mailhub.nc.com ([207.88.25.3]:5365 "HELO mailhub.nc.com" ident: "proxy") by vger.rutgers.edu with SMTP id <971239-241>; Mon, 12 Jan 1998 18:08:18 -0500 Received: (from proxy@localhost) by mailhub.nc.com (8.6.12/8.6.9) id PAA32426; Mon, 12 Jan 1998 15:10:06 -0800 Received: from riscbsd1.client.nc.com(172.17.8.88) by mailhub.nc.com via smap (V2.0) id xma032387; Mon, 12 Jan 98 15:09:37 -0800 Message-ID: <34BAA2AC.4487EB71@causality.com> Date: Mon, 12 Jan 1998 23:09:32 +0000 From: "Neil A. Carson" Organization: Causality Limited (London, UK) X-Mailer: Mozilla 3.01 (X11; I; NetBSD 1.2G arm32) MIME-Version: 1.0 To: Alan Cox CC: msergio@tin.it, linux-arm@vger.rutgers.edu Subject: Re: ARMlinux vs. RiscBSD References: <199801122002.UAA11083@snowcrash.cymru.net> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Alan Cox wrote: > > Some recent research into log-structured filesystems (I believe there > > was a Linux log-structured project) by the FreeBSD guys---as they've > > Yep. And theres also a journalled ext2fs project which is fairly > similar to McKusicks work but a lot simpler in its approach. (NTfs btw > is very similar in its algorithms to McKusicks work) NTFS? As in the filesystem for Windows NT? That uses a Journal for preserving the synchronous nature of its operations. Of course the journaled ext2 will be slower than the current one---about half way between the speed of FFS and the existing ext2, but with much greater reliability and fsck times than either. Hmm I wonder if you could do what McKusick does, and splat out the directed partial ordering to disc just before doing most of the ops, thus gaining most of the speed of ext2 and the reliabiltity of a journaled FS? Food for thought :-) Is there a URL for the journaled ext2FS project? I'd quite like to do some porting over of this as/when it's nearing completion. Regards, Neil unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 13 08:50:25 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id IAA12065 for ; Tue, 13 Jan 1998 08:50:24 GMT Received: from vger.rutgers.edu ([128.6.190.2]:59722 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <2595-21270>; Tue, 13 Jan 1998 10:48:40 +0200 Received: by vger.rutgers.edu id <970965-241>; Tue, 13 Jan 1998 03:44:55 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:1257 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <970800-241>; Tue, 13 Jan 1998 03:38:09 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xs2mi-0000GI-00; Tue, 13 Jan 1998 09:37:24 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: John Appleby cc: linux-arm@vger.rutgers.edu, rmk@ecs.soton.ac.uk Subject: Re: ARM-Linux for the A5000 In-reply-to: Your message of "Mon, 12 Jan 1998 20:24:14 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Tue, 13 Jan 1998 09:37:23 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >I have an A5000 which I would like to run Linux on... I understand that >the root disk isn't quite ready and isn't the same as the RISC-PC root >disk. > >Is there a way of installing by hand to get it to boot, or a >development root disk? There doesn't seem to be ANYTHING about this on >the web... The way forward is probably NFS root. You can cross-compile an A5000 kernel. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 13 19:27:22 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id TAA16607 for ; Tue, 13 Jan 1998 19:27:21 GMT Received: from vger.rutgers.edu ([128.6.190.2]:61225 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10339-27083>; Tue, 13 Jan 1998 21:21:27 +0200 Received: by vger.rutgers.edu id <970884-241>; Tue, 13 Jan 1998 10:22:52 -0500 Received: from dwmw2.robinson.cam.ac.uk ([131.111.217.153]:4166 "HELO imladris.demon.co.uk" ident: "exim") by vger.rutgers.edu with SMTP id <971031-241>; Tue, 13 Jan 1998 10:22:31 -0500 Received: from localhost (dwmw2.robinson.cam.ac.uk) [127.0.0.1] (dwmw2) by imladris.demon.co.uk with esmtp (Exim 1.82 #9) id 0xs8Ad-0003MQ-00; Tue, 13 Jan 1998 15:22:27 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Philip Blundell Cc: John Appleby , linux-arm@vger.rutgers.edu, rmk@ecs.soton.ac.uk Subject: Re: ARM-Linux for the A5000 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Tue, 13 Jan 1998 15:22:26 +0000 From: David Woodhouse Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Philip.Blundell@pobox.com said: > The way forward is probably NFS root. You can cross-compile an A5000 > kernel. > p. If we're not too stupid to make ourselves a cross-compiler, that is :( After a few hours of attempting to compile first gcc-2.7.2.2 with the patches on the FTP site, and then egcs-1.0.1, we're thoroughly p***ed off with the whole thing. It there anyone out there who could mail me one, please? If you can, please mail it directly to dwmw2@dwmw2.robinson.cam.ac.uk -- ---- ---- ---- David Woodhouse, Robinson College, CB3 9AN, England. (+44) 0976 658355 Dave@imladris.demon.co.uk http://dwmw2.robinson.cam.ac.uk finger pgp@dwmw2.robinson.cam.ac.uk for PGP key. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 13 21:08:32 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id VAA17274 for ; Tue, 13 Jan 1998 21:08:27 GMT Received: from vger.rutgers.edu ([128.6.190.2]:55059 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10320-21270>; Tue, 13 Jan 1998 23:06:35 +0200 Received: by vger.rutgers.edu id <970881-241>; Tue, 13 Jan 1998 11:16:24 -0500 Received: from nexusel.demon.co.uk ([158.152.30.195]:8316 "HELO globe" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <971565-241>; Tue, 13 Jan 1998 11:15:19 -0500 Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xs8qj-0004Nv-00; Tue, 13 Jan 1998 16:05:57 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xs8qi-0000Gv-00; Tue, 13 Jan 1998 16:05:56 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: David Woodhouse cc: linux-arm@vger.rutgers.edu Subject: Re: ARM-Linux for the A5000 In-reply-to: Your message of "Tue, 13 Jan 1998 15:22:26 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Tue, 13 Jan 1998 16:05:56 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >If we're not too stupid to make ourselves a cross-compiler, that is :( >After a few hours of attempting to compile first gcc-2.7.2.2 with the patches >on the FTP site, and then egcs-1.0.1, we're thoroughly p***ed off with the >whole thing. > >It there anyone out there who could mail me one, please? > >If you can, please mail it directly to dwmw2@dwmw2.robinson.cam.ac.uk The patches on the ftp site are quite old - I suggest you go for egcs until gcc 2.8 is officially released. What problem are you having? Don't forget that you need binutils installed first, and you also need the Linux kernel headers to be in place to build libgcc.a. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 13 21:08:33 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id VAA17277 for ; Tue, 13 Jan 1998 21:08:32 GMT Received: from vger.rutgers.edu ([128.6.190.2]:55059 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10849-27083>; Tue, 13 Jan 1998 23:04:31 +0200 Received: by vger.rutgers.edu id <971148-241>; Tue, 13 Jan 1998 11:11:30 -0500 Received: from hermes.ex.ac.uk ([144.173.6.14]:2861 "EHLO exeter.ac.uk" ident: "TIMEDOUT") by vger.rutgers.edu with ESMTP id <971648-241>; Tue, 13 Jan 1998 11:10:48 -0500 Received: from hebe [144.173.6.23] by hermes via SMTP (QAA20742); Tue, 13 Jan 1998 16:11:10 GMT Date: Tue, 13 Jan 1998 16:11:09 +0000 (GMT) From: Phil Norman X-Sender: py95pcfn@hebe To: ARM Linux Subject: Secondary partition tables Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi! RiscBSD and Linux cannot exist on the same disc because their secondary partition tables are different. Wouldn't it be a far better idea to have a standard partition table that both could use, with room for expansion so that future OSses could use the same system too (including RiscOS filing systems in fact)? The Linux partition table caters for that possibility and has a very simple method of storing partitions together with an identifying word. I'm not sure what RiscBSD uses, but it's probably something similar. How about BSD and Linux people talking about which should be used by both, since this would simplify things considerably and, in fact, would allow both OSses to use the SAME partitions (if certain partitions would be used for data files, images etc). I think that linux can read UFS (or whatever it's called) partitions and BSD can probably read ext2 partitions too, so this would be a far more flexible system. Anyway, I'll leave the kernelly people to ponder that one. Bye for now, Phil -- Phil Norman, mailing from Exeter Uni. Programs available for download from my web page. email: p.c.f.norman@ex.ac.uk web: http://newton.ex.ac.uk/general/ug/norman unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 13 22:56:52 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id WAA18687 for ; Tue, 13 Jan 1998 22:56:42 GMT Received: from vger.rutgers.edu ([128.6.190.2]:55059 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11319-27418>; Wed, 14 Jan 1998 00:51:23 +0200 Received: by vger.rutgers.edu id <971031-241>; Tue, 13 Jan 1998 14:33:20 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:1529 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <971202-241>; Tue, 13 Jan 1998 14:28:18 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id TAA16612; Tue, 13 Jan 1998 19:30:01 GMT From: Matthew Wilcox Message-Id: <199801131930.TAA16612@odie.barnet.ac.uk> Subject: repeatable hang To: linux@arm.uk.linux.org Date: Tue, 13 Jan 1998 19:29:59 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO I have a condition which will hang ARMLinux every time I try it. I can work around it, but it's a bit bizarre... I telnet to an i586-linux box running kernel 2.0.33 from my arm-linux box running 2.0.31. Not particularly hacked around. I then run nvi on the i586-linux box (because it's debian and I hadn't changed it to vim yet). I then use the cursor keys to move to the end of the file... and the arm linux box hangs. The only other bit of information that might prove useful or interesting is that this is over an Ether3 card. Anyone got any ideas? unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 13 23:11:43 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id XAA18884 for ; Tue, 13 Jan 1998 23:11:41 GMT Received: from vger.rutgers.edu ([128.6.190.2]:55059 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12247-150>; Wed, 14 Jan 1998 00:53:10 +0200 Received: by vger.rutgers.edu id <970810-241>; Tue, 13 Jan 1998 14:40:50 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1824 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971414-241>; Tue, 13 Jan 1998 14:37:17 -0500 Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id TAA32664; Tue, 13 Jan 1998 19:39:23 GMT From: Alan Cox Message-Id: <199801131939.TAA32664@snowcrash.cymru.net> Subject: Re: ARM-Linux for the A5000 To: Dave@imladris.demon.co.uk (David Woodhouse) Date: Tue, 13 Jan 1998 19:39:20 +0000 (GMT) Cc: Philip.Blundell@pobox.com, jma24@cam.ac.uk, linux-arm@vger.rutgers.edu, rmk@ecs.soton.ac.uk In-Reply-To: from "David Woodhouse" at Jan 13, 98 03:22:26 pm Content-Type: text X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > on the FTP site, and then egcs-1.0.1, we're thoroughly p***ed off with the > whole thing. Nod. I had horrible problems and my compiler looked like it was writing wrong code afterwards. bintuils seems to cross wonderfully. I'd happily provide ftp space for a cross arm compiler set that works. Bonus points if it can write the bogoid acorn binary format too ;) Alan unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From alan@snowcrash.cymru.net Wed Jan 14 00:23:06 1998 Return-Path: Received: from snowcrash.cymru.net (snowcrash.cymru.net [163.164.160.3]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id AAA20357 for ; Wed, 14 Jan 1998 00:23:05 GMT Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id AAA04122; Wed, 14 Jan 1998 00:23:20 GMT From: Alan Cox Message-Id: <199801140023.AAA04122@snowcrash.cymru.net> Subject: Re: ARM-Linux for the A5000 To: willy@odie.barnet.ac.uk (Matthew Wilcox) Date: Wed, 14 Jan 1998 00:23:14 +0000 (GMT) Cc: alan@cymru.net, Dave@imladris.demon.co.uk, Philip.Blundell@pobox.com, jma24@cam.ac.uk, linux-arm@vger.rutgers.edu, rmk@ecs.soton.ac.uk In-Reply-To: <199801132322.XAA19196@odie.barnet.ac.uk> from "Matthew Wilcox" at Jan 13, 98 11:22:41 pm Content-Type: text Status: RO Acorns linker may cope with AOFF from a.out but nto everyone working on arm linux will have access to acorn kit as the ARM moves on, and AOFF is needed for stuff like the ebsa285 loader. So aout2aoff or similar is needed Alan From owner-linux-arm-outgoing@vger.rutgers.edu Wed Jan 14 02:32:21 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id CAA21083 for ; Wed, 14 Jan 1998 02:32:20 GMT Received: from vger.rutgers.edu ([128.6.190.2]:55059 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <6239-27083>; Wed, 14 Jan 1998 02:25:25 +0200 Received: by vger.rutgers.edu id <971080-241>; Tue, 13 Jan 1998 18:21:54 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:1216 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <971070-241>; Tue, 13 Jan 1998 18:21:24 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id XAA19196; Tue, 13 Jan 1998 23:22:45 GMT From: Matthew Wilcox Message-Id: <199801132322.XAA19196@odie.barnet.ac.uk> Subject: Re: ARM-Linux for the A5000 To: alan@cymru.net (Alan Cox) Date: Tue, 13 Jan 1998 23:22:41 +0000 (GMT) Cc: Dave@imladris.demon.co.uk, Philip.Blundell@pobox.com, jma24@cam.ac.uk, linux-arm@vger.rutgers.edu, rmk@ecs.soton.ac.uk In-Reply-To: <199801131939.TAA32664@snowcrash.cymru.net> from "Alan Cox" at Jan 13, 98 07:39:20 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > > > on the FTP site, and then egcs-1.0.1, we're thoroughly p***ed off with the > > whole thing. > > Nod. I had horrible problems and my compiler looked like it was writing > wrong code afterwards. bintuils seems to cross wonderfully. I'd happily > provide ftp space for a cross arm compiler set that works. Bonus points > if it can write the bogoid acorn binary format too ;) Umm.. I'll look at the cross-code that gcc 2.8 is producing and see if I can spot the difference. It's not necessary to write AOF code - Acorn's linker copes with a.out input files. OTOH, since gcc 2.8 isn't compiling my kernel right yet, there's prolly other problems that ought to be resolved before we let it run free. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Wed Jan 14 05:22:30 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id FAA22124 for ; Wed, 14 Jan 1998 05:22:29 GMT Received: from vger.rutgers.edu ([128.6.190.2]:22539 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <17175-27083>; Wed, 14 Jan 1998 05:36:33 +0200 Received: by vger.rutgers.edu id <970796-241>; Tue, 13 Jan 1998 22:33:00 -0500 Received: from dwmw2.robinson.cam.ac.uk ([131.111.217.153]:2132 "HELO imladris.demon.co.uk" ident: "exim") by vger.rutgers.edu with SMTP id <970861-241>; Tue, 13 Jan 1998 22:32:19 -0500 Received: from localhost (dwmw2.robinson.cam.ac.uk) [127.0.0.1] (dwmw2) by imladris.demon.co.uk with esmtp (Exim 1.82 #9) id 0xsJbB-0000rI-00; Wed, 14 Jan 1998 03:34:37 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Alan Cox Cc: Philip.Blundell@pobox.com, jma24@cam.ac.uk, linux-arm@vger.rutgers.edu, rmk@ecs.soton.ac.uk Subject: ARM Cross-compiler (Was "Re: ARM-Linux for the A5000") Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 14 Jan 1998 03:34:37 +0000 From: David Woodhouse Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Thanks to all who offered advice. I now have a functional egcs-1.0.1 and binutils-2.8.1.0.15 for the ARM target, running on i586-linux-glibc. The whole lot's at: ftp://dwmw2.robinson.cam.ac.uk/pub/arm/arm-egcs-et-utils.bin.tar.gz (28Mb of it) It compiles the kernel that's sitting in the same directory, although I've not booted it to check that it's generating correct code. -- ---- ---- ---- David Woodhouse, Robinson College, CB3 9AN, England. (+44) 0976 658355 Dave@imladris.demon.co.uk http://dwmw2.robinson.cam.ac.uk finger pgp@dwmw2.robinson.cam.ac.uk for PGP key. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From pb@nexus.co.uk Wed Jan 14 09:29:56 1998 Return-Path: Received: from globe (nexusel.demon.co.uk [158.152.30.195]) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id JAA23596 for ; Wed, 14 Jan 1998 09:29:54 GMT Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xsP3u-0005TQ-00; Wed, 14 Jan 1998 09:24:38 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xsP3t-0000yF-00; Wed, 14 Jan 1998 09:24:37 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Alan Cox cc: willy@odie.barnet.ac.uk (Matthew Wilcox), Dave@imladris.demon.co.uk, jma24@cam.ac.uk, linux-arm@vger.rutgers.edu, rmk@ecs.soton.ac.uk Subject: Re: ARM-Linux for the A5000 In-reply-to: Your message of "Wed, 14 Jan 1998 00:23:14 GMT." <199801140023.AAA04122@snowcrash.cymru.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 14 Jan 1998 09:24:37 +0000 From: Philip Blundell Message-Id: Status: RO In message <199801140023.AAA04122@snowcrash.cymru.net>, Alan Cox writes: >Acorns linker may cope with AOFF from a.out but nto everyone working on arm >linux will have access to acorn kit as the ARM moves on, and AOFF is needed >for stuff like the ebsa285 loader. So aout2aoff or similar is needed Do you just need to be able to generate executable AOF from an a.out binary? If so it shouldn't be hard - it's something I've thought about before (because I wanted to be able to cross-compile Acorn applications). Linkable AOF is probably a bit harder. p. From alan@snowcrash.cymru.net Wed Jan 14 09:45:09 1998 Return-Path: Received: from snowcrash.cymru.net (snowcrash.cymru.net [163.164.160.3]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id JAA23776 for ; Wed, 14 Jan 1998 09:45:08 GMT Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id JAA12081; Wed, 14 Jan 1998 09:45:29 GMT From: Alan Cox Message-Id: <199801140945.JAA12081@snowcrash.cymru.net> Subject: Re: ARM-Linux for the A5000 To: pb@nexus.co.uk (Philip Blundell) Date: Wed, 14 Jan 1998 09:45:26 +0000 (GMT) Cc: alan@cymru.net, willy@odie.barnet.ac.uk, Dave@imladris.demon.co.uk, jma24@cam.ac.uk, linux-arm@vger.rutgers.edu, rmk@ecs.soton.ac.uk In-Reply-To: from "Philip Blundell" at Jan 14, 98 09:24:37 am Content-Type: text Status: RO > Do you just need to be able to generate executable AOF from an a.out binary? Or ELF , or arbitary binary image. I wrote the AOF headers for the EBSA285 test app in hex instead > If so it shouldn't be hard - it's something I've thought about before (because > I wanted to be able to cross-compile Acorn applications). aout2aoff for folks with stuff like the EBSA boards that expect that format might be a good move. It seems fairly easy to arrange From dg@cogency.co.uk Wed Jan 14 10:02:03 1998 Return-Path: Received: from mailhost.cogency.co.uk (man01.cogency.co.uk [194.201.189.33] (may be forged)) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id KAA23825 for ; Wed, 14 Jan 1998 10:02:01 GMT Received: from cogency.co.uk by mailhost.cogency.co.uk (SMI-8.6/PIPEX simple 1.27) id KAA17829; Wed, 14 Jan 1998 10:02:55 GMT Sender: dg@cogency.co.uk Message-ID: <34BC8CAB.A312D419@cogency.co.uk> Date: Wed, 14 Jan 1998 10:00:11 +0000 From: David Alan Gilbert Organization: Cogency Technology Inc X-Mailer: Mozilla 4.04 [en] (X11; I; SunOS 5.5 sun4m) MIME-Version: 1.0 To: Matthew Wilcox Subject: Re: repeatable hang References: <199801131930.TAA16612@odie.barnet.ac.uk> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Status: RO Hi Matthew, How hard is the hang? Do things like the capslock button still work on the Arc? Does it ping? Dave -- ---------------------------------------------------------------------- - Dr. David Alan Gilbert - WARNING! This is a beta release .signature- - Work: dg @ cogency.co.uk - 0161-428-9444 - - Home: gro.gilbert @ treblig.org - - ---------------------------------------------------------------------- From owner-linux-arm-outgoing@vger.rutgers.edu Wed Jan 14 10:19:40 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id KAA23992 for ; Wed, 14 Jan 1998 10:19:39 GMT Received: from vger.rutgers.edu ([128.6.190.2]:46970 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <17058-27083>; Wed, 14 Jan 1998 12:13:33 +0200 Received: by vger.rutgers.edu id <970810-241>; Wed, 14 Jan 1998 05:07:10 -0500 Received: from [194.201.189.33] ([194.201.189.33]:60751 "HELO mailhost.cogency.co.uk" ident: "SOCKWRITE-65") by vger.rutgers.edu with SMTP id <970802-241>; Wed, 14 Jan 1998 05:06:52 -0500 Received: from cogency.co.uk by mailhost.cogency.co.uk (SMI-8.6/PIPEX simple 1.27) id KAA17857; Wed, 14 Jan 1998 10:09:42 GMT Message-ID: <34BC8E42.66D2E791@cogency.co.uk> Date: Wed, 14 Jan 1998 10:06:58 +0000 From: David Alan Gilbert Organization: Cogency Technology Inc X-Mailer: Mozilla 4.04 [en] (X11; I; SunOS 5.5 sun4m) MIME-Version: 1.0 To: linux-arm@vger.rutgers.edu Subject: Re:Cross compilation References: <199801132322.XAA19196@odie.barnet.ac.uk> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi, I have cross compilation going fairly reliably from an Alpha producing ARM kernels; the way I look at is if I can do it from an Alpha it should be dead easy to do it from a 32 bit machine. I'll post my current list of instructions in a few days time. I'm using EGCS-971225 with a small patch (which is really an Alpha only problem). For me (from Alpha) Binutils-....18 produces a kernel which doesn't boot but ....15 works - I haven't figured out why yet. So far I have it working for producing old Arc (and should do A5K kernels) and I have produced a RiscPC kernel - but I know there are those having problems producing RPC kernels. Instructions in a couple of days (when I am doing email from home). Dave -- ---------------------------------------------------------------------- - Dr. David Alan Gilbert - WARNING! This is a beta release .signature- - Work: dg @ cogency.co.uk - 0161-428-9444 - - Home: gro.gilbert @ treblig.org - - ---------------------------------------------------------------------- unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From pb@nexus.co.uk Wed Jan 14 11:35:53 1998 Return-Path: Received: from globe (nexusel.demon.co.uk [158.152.30.195]) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id LAA24477 for ; Wed, 14 Jan 1998 11:35:52 GMT Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xsR73-0005iY-00; Wed, 14 Jan 1998 11:36:01 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xsR72-00017r-00; Wed, 14 Jan 1998 11:36:00 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Alan Cox cc: willy@odie.barnet.ac.uk, Dave@imladris.demon.co.uk, jma24@cam.ac.uk, linux-arm@vger.rutgers.edu, rmk@ecs.soton.ac.uk Subject: Re: ARM-Linux for the A5000 In-reply-to: Your message of "Wed, 14 Jan 1998 09:45:26 GMT." <199801140945.JAA12081@snowcrash.cymru.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 14 Jan 1998 11:36:00 +0000 From: Philip Blundell Message-Id: Status: RO In message <199801140945.JAA12081@snowcrash.cymru.net>, Alan Cox writes: >> Do you just need to be able to generate executable AOF from an a.out binary? > >Or ELF , or arbitary binary image. I wrote the AOF headers for the EBSA285 >test app in hex instead Right. Well, objcopy will convert ELF/a.out/whatever to raw binary for you, so all we really need is the trivial bit of code to wrap it in an AOF header. I'll knock something up a bit later. p. From owner-linux-arm-outgoing@vger.rutgers.edu Wed Jan 14 15:00:55 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id PAA25915 for ; Wed, 14 Jan 1998 15:00:54 GMT Received: from vger.rutgers.edu ([128.6.190.2]:32061 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <20677-6684>; Wed, 14 Jan 1998 16:43:14 +0200 Received: by vger.rutgers.edu id <970974-241>; Wed, 14 Jan 1998 09:07:22 -0500 Received: from methan.in-berlin.de ([160.45.10.13]:4553 "HELO methan.in-berlin.de" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <971047-241>; Wed, 14 Jan 1998 08:46:49 -0500 Received: by methan.in-berlin.de (Smail3.2.0.98) from bolzen.in-berlin.de (194.94.235.1) with esmtp id ; Wed, 14 Jan 1998 14:49:13 +0100 (MET) Received: by bolzen.in-berlin.de (Smail3.2) from yoda with uucp id ; Wed, 14 Jan 1998 14:48:58 +0100 (MET) Received: by yoda.in-berlin.de (Smail3.1.28.1 #0) id m0xsTjX-000000C; Wed, 14 Jan 98 13:23 GMT Date: Wed, 14 Jan 1998 14:12:16 +0100 From: Stefan Voss To: linux-arm@vger.rutgers.edu Subject: Still insmod install problem Message-ID: Organization: Individual Network (IN), Berlin, Germany X-Mailer: Messenger v1.11 for RISC OS X-Posting-Agent: RISC OS Newsbase 0.59f X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi folks, I'm still having insmod crashing at the beginning of the install process. I tried older kernels and root disks but older kernels don't recognize my EtherH card and older root disks show the same error. Can some kind soul send (or put online) an up-to-date *statically* linked insmod binary? I'm going to copy the new binary over to one on the root disk to see if that makes any difference. Another question: Does installing from a "local hard disk" mean an ext2 file system or can an installation be done from an ADFS partition? Regards, Stefan -- Stefan Voss (voss@yoda.in-berlin.de) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From Nicholas.Clark@liverpool.ac.uk Wed Jan 14 17:10:02 1998 Return-Path: Received: from pump2.york.ac.uk (pump2.york.ac.uk [144.32.128.12]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id RAA26742 for ; Wed, 14 Jan 1998 17:10:01 GMT Received: from mailhub.liverpool.ac.uk (mail.liv.ac.uk [138.253.100.84]) by pump2.york.ac.uk (8.8.7/8.8.7) with SMTP id RAA19034 for ; Wed, 14 Jan 1998 17:10:32 GMT Received: from uxe.liv.ac.uk by mail.liv.ac.uk with Local-SMTP (PP) with ESMTP; Wed, 14 Jan 1998 17:10:25 +0000 To: mrw103@york.ac.uk (Matthew Wilcox) Subject: Re: GCC problems References: <19980105.170033.93@lavenham.demon.co.uk> <68rndb$p06$1@pump1.york.ac.uk> Content-Transfer-Encoding: 8bit X-use-a-real-newsreader: From: Spider plant breeding program Date: 14 Jan 1998 17:10:25 +0000 In-Reply-To: mrw103@york.ac.uk's message of "5 Jan 1998 22:38:35 GMT" Message-ID: Lines: 30 X-Mailer: Gnus v5.5/XEmacs 20.3 - "Vatican City" Status: RO mrw103@york.ac.uk (Matthew Wilcox) writes: > Stuart Thomson (st@nospam.demon.co.uk) wrote: > : I've just downloaded GNU/GCC, and have tried to compile the Hello World > : program. However, the compiler gives an error as follows: > > [...] > : ld /tmp/cc0001731.o -lgcc -LGCC: -LUnix: -lUnixLib -lgcc > : ld fatal error : ld returned 33 exit status > : gcc: Fatal compiler error: program ld > > I can tell you what's wrong. I don't know how to fix it, and I've read > the RISC OS GCC FAQ. The compiler is passing `-lgcc' to the linker, but > drlink prefers `-l gcc'. I worked around this by using gcc -c to compile > the .o file then issuing the link step myself. Is it the problem fixed in Nick's patch at http://www.btinternet.com/~nick.burrett/ ? Nick PS That URL is the answer to a FAQ. :-) -- #!perl -wlpi[finger.liv.ac.uk] # If it doesn't work see doio.c line ~256 BEGIN{$_="use SocketYIN;sockeXPZSOCK_STREAM,~proto'tcp'and\$|=connecXpack'S na4x8',AZ79,\$;=~host qq$^Ior die\$!;print'/w nickc\r'YOUT;print\$^I";s/X/t STDIN,/g;s/Y/;select STD/g;s/Z/F_INET,/g;s/~(\w*)/get$1byname/g;eval||die} From linux@arm.uk.linux.org Wed Jan 14 18:19:35 1998 Return-Path: Received: from snowcrash.cymru.net (snowcrash.cymru.net [163.164.160.3]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id SAA27300 for ; Wed, 14 Jan 1998 18:19:32 GMT Received: from caramon.armlinux.org (dialup1.cymru.net [163.164.160.65]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id SAA22579; Wed, 14 Jan 1998 18:19:41 GMT Received: (from linux@localhost) by caramon.armlinux.org (8.7.4/8.7.3) id AAA13036; Wed, 14 Jan 1998 00:27:18 GMT From: Russell King - ARM Linux Admin Message-Id: <199801140027.AAA13036@caramon.armlinux.org> Subject: Re: repeatable hang To: willy@odie.barnet.ac.uk (Matthew Wilcox) Date: Wed, 14 Jan 1998 00:27:17 +0000 (GMT) Cc: linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu In-Reply-To: <199801131930.TAA16612@odie.barnet.ac.uk> from "Matthew Wilcox" at Jan 13, 98 07:29:59 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Status: RO Matthew Wilcox writes: > I telnet to an i586-linux box running kernel 2.0.33 from my arm-linux box > running 2.0.31. Not particularly hacked around. I then run nvi on the > i586-linux box (because it's debian and I hadn't changed it to vim yet). Can you strace the nvi session and send me the output (preferably with a suitable strace -s value to allow all the write's to be captured). _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | From rmk@ecs.soton.ac.uk Wed Jan 14 18:19:36 1998 Return-Path: Received: from snowcrash.cymru.net (snowcrash.cymru.net [163.164.160.3]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id SAA27302 for ; Wed, 14 Jan 1998 18:19:34 GMT Received: from caramon.armlinux.org (dialup1.cymru.net [163.164.160.65]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id SAA22583; Wed, 14 Jan 1998 18:19:46 GMT Received: (from rmk@localhost) by caramon.armlinux.org (8.7.4/8.7.3) id AAA12991; Wed, 14 Jan 1998 00:15:56 GMT From: Russell King Message-Id: <199801140015.AAA12991@caramon.armlinux.org> Subject: Re: ARM-Linux for the A5000 To: willy@odie.barnet.ac.uk (Matthew Wilcox) Date: Wed, 14 Jan 1998 00:15:54 +0000 (GMT) Cc: alan@cymru.net, Dave@imladris.demon.co.uk, Philip.Blundell@pobox.com, jma24@cam.ac.uk, linux-arm@vger.rutgers.edu, rmk@diana.ecs.soton.ac.uk In-Reply-To: <199801132322.XAA19196@odie.barnet.ac.uk> from "Matthew Wilcox" at Jan 13, 98 11:22:41 pm X-Location: london.england.earth.mulky-way.universe Reply-To: rmk@ecs.soton.ac.uk X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Status: RO Matthew Wilcox writes: > Umm.. I'll look at the cross-code that gcc 2.8 is producing and see if > I can spot the difference. It's not necessary to write AOF code - Acorn's > linker copes with a.out input files. If you're using GCC 2.8, then PLEASE make sure that you are using the correct libgcc.a - if you're using -mapcs-32, then you *must* use the APCS-32 version of libgcc.a, otherwise the ApCS-26 version. This has already caused someone problems, so if other people report it, I'll add it to the FAQ. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | From Philip.Blundell@pobox.com Wed Jan 14 22:13:24 1998 Return-Path: Received: from paddington.london.uk.eu.org (exim@tazenda.demon.co.uk [158.152.220.239]) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id WAA28790 for ; Wed, 14 Jan 1998 22:13:19 GMT Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xsbvU-00009X-00; Wed, 14 Jan 1998 23:08:48 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: rmk@ecs.soton.ac.uk cc: willy@odie.barnet.ac.uk (Matthew Wilcox), alan@cymru.net, Dave@imladris.demon.co.uk, jma24@cam.ac.uk, linux-arm@vger.rutgers.edu, rmk@diana.ecs.soton.ac.uk Subject: Re: ARM-Linux for the A5000 In-reply-to: Your message of "Wed, 14 Jan 1998 00:15:54 GMT." <199801140015.AAA12991@caramon.armlinux.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 14 Jan 1998 23:08:48 +0000 From: Philip Blundell Message-Id: Status: RO >Matthew Wilcox writes: >> Umm.. I'll look at the cross-code that gcc 2.8 is producing and see if >> I can spot the difference. It's not necessary to write AOF code - Acorn's >> linker copes with a.out input files. > >If you're using GCC 2.8, then PLEASE make sure that you are using the correct >libgcc.a - if you're using -mapcs-32, then you *must* use the APCS-32 version >of libgcc.a, otherwise the ApCS-26 version. This has already caused someone >problems, so if other people report it, I'll add it to the FAQ. It's quite hard to use the wrong version - nobody should have the path to libgcc.a coded in anywhere. p. From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 15 02:36:27 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id CAA30737 for ; Thu, 15 Jan 1998 02:36:26 GMT Received: from vger.rutgers.edu ([128.6.190.2]:46660 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <21304-18750>; Thu, 15 Jan 1998 02:51:18 +0200 Received: by vger.rutgers.edu id <971294-241>; Wed, 14 Jan 1998 16:41:53 -0500 Received: from dwmw2.robinson.cam.ac.uk ([131.111.217.153]:4233 "HELO imladris.demon.co.uk" ident: "exim") by vger.rutgers.edu with SMTP id <971297-241> convert rfc822-to-8bit; Wed, 14 Jan 1998 16:40:08 -0500 Received: from localhost (dwmw2.robinson.cam.ac.uk) [127.0.0.1] (dwmw2) by imladris.demon.co.uk with esmtp (Exim 1.82 #9) id 0xsaZy-0004GP-00; Wed, 14 Jan 1998 21:42:30 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: John Appleby Cc: linux-arm@vger.rutgers.edu, rmk@ecs.soton.ac.uk Subject: Re: ARM-Linux for the A5000 Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7BIT Date: Wed, 14 Jan 1998 21:42:30 +0000 From: David Woodhouse Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO jma24@cam.ac.uk said: > I have an A5000 which I would like to run Linux on... I understand > that the root disk isn't quite ready and isn't the same as the RISC-PC > root disk. What's wrong with the root-a5k disk? Also, are the RPMS on the FTP site compiled for ARM3, or will they all need recompiling? -- ---- ---- ---- David Woodhouse, Robinson College, CB3 9AN, England. (+44) 0976 658355 Dave@imladris.demon.co.uk http://dwmw2.robinson.cam.ac.uk finger pgp@dwmw2.robinson.cam.ac.uk for PGP key. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 15 04:48:25 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id EAA31594 for ; Thu, 15 Jan 1998 04:48:23 GMT Received: from vger.rutgers.edu ([128.6.190.2]:46660 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <2118-24815>; Thu, 15 Jan 1998 03:25:26 +0200 Received: by vger.rutgers.edu id <971051-241>; Wed, 14 Jan 1998 19:36:56 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1494 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971056-241>; Wed, 14 Jan 1998 19:35:58 -0500 Received: from caramon.armlinux.org (dialup1.cymru.net [163.164.160.65]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id AAA28686; Thu, 15 Jan 1998 00:38:23 GMT Received: from raistlin.armlinux.org (raistlin [192.168.0.3]) by caramon.armlinux.org (8.7.4/8.7.3) with ESMTP id VAA13835; Wed, 14 Jan 1998 21:35:16 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.armlinux.org (8.7.4/8.7.3) id VAA00593; Wed, 14 Jan 1998 21:35:01 GMT Message-Id: <199801142135.VAA00593@raistlin.armlinux.org> Subject: Re: Still insmod install problem To: voss@yoda.in-berlin.de (Stefan Voss) Date: Wed, 14 Jan 1998 21:35:00 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: from "Stefan Voss" at Jan 14, 98 02:12:16 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Stefan Voss writes: > I'm still having insmod crashing at the beginning of the install process. > I tried older kernels and root disks but older kernels don't recognize my > EtherH card and older root disks show the same error. Use the latest root disk - that binary will work. If it doesn't, then none of the others will. > Another question: Does installing from a "local hard disk" mean an ext2 > file system or can an installation be done from an ADFS partition? It means a DOS/ADFS/ext2 filesystem, however, ADFS does have the limit of 77 filenames per directory. (maybe someone could do something like add the X-files to the linux adfs? Or how about having a set of kerneld-loadable modules aka RiscOS's Image Filing System? I'm sure that it should be possible (don't ask me to do it tho - I've got enough on at the moment). _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 15 06:29:00 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id GAA32121 for ; Thu, 15 Jan 1998 06:28:59 GMT Received: from vger.rutgers.edu ([128.6.190.2]:46660 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <21461-18750>; Thu, 15 Jan 1998 03:04:20 +0200 Received: by vger.rutgers.edu id <971722-241>; Wed, 14 Jan 1998 17:34:07 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:1211 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <971183-241>; Wed, 14 Jan 1998 17:26:51 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xsc6q-0000BI-00; Wed, 14 Jan 1998 23:20:32 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: David Woodhouse cc: John Appleby , linux-arm@vger.rutgers.edu, rmk@ecs.soton.ac.uk Subject: Re: ARM-Linux for the A5000 In-reply-to: Your message of "Wed, 14 Jan 1998 21:42:30 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 14 Jan 1998 23:20:31 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >jma24@cam.ac.uk said: >> I have an A5000 which I would like to run Linux on... I understand >> that the root disk isn't quite ready and isn't the same as the RISC-PC >> root disk. > >What's wrong with the root-a5k disk? > >Also, are the RPMS on the FTP site compiled for ARM3, or will they all need >recompiling? User applications run in 26-bit mode on all CPUs, I believe. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 15 14:39:17 1998 Return-Path: Received: from nic.funet.fi ([128.214.248.6] (may be forged)) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id OAA00935 for ; Thu, 15 Jan 1998 14:39:16 GMT Received: from vger.rutgers.edu ([128.6.190.2]:46690 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <2658-24815>; Thu, 15 Jan 1998 15:52:42 +0200 Received: by vger.rutgers.edu id <971500-241>; Thu, 15 Jan 1998 04:09:29 -0500 Received: from irwell.zetnet.co.uk ([194.247.47.48]:18607 "EHLO irwell.zetnet.co.uk" ident: "root") by vger.rutgers.edu with ESMTP id <971506-241>; Thu, 15 Jan 1998 04:08:45 -0500 Received: from Michael (man-185.dialup.zetnet.co.uk [194.247.40.235]) by irwell.zetnet.co.uk (8.8.7/8.8.5) with SMTP id JAA07612 for ; Thu, 15 Jan 1998 09:11:37 GMT Date: Thu, 15 Jan 1998 09:11:57 +0000 (GMT) From: Michael Howard Subject: Installation Problems To: Linux Mailing List Illegal-Object: Syntax error in Message-ID: value found on vger.rutgers.edu: Message-ID: ^-illegal end of message identification MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII X-Organization: Organisation name, location. Telephone/Fax? X-Mailer: ANT RISCOS Marcel [ver 1.26] X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Message-Id: <19980115090939Z971500-241+10420@vger.rutgers.edu> Status: RO Hi All, As a newcomer to arm-linux and this mailing list, can someone help with my installation problem. I have been attempting to install linux on my RiscPC 700 (not strongarm) for the past few days without success. I can't get past the 'Find Installation Files' screen. Only one device is presented in the dialog box, /dev/hdb3, which is the Native Linux partition but Redhat/base and RedHat/RPMS are on the Filecore partition. I'm unable to change the devive presented in the box. I'm using the latest RedHat tree from the ftp site. One anomoly I noticed during the installation was on the partition screen, when I listed the partitions the Linux Table partition was reported as starting at '201600' which is correct, but ending at '1064447' which is the end of the disc. One further point, due to the constraints of ADFS, I have the RPMS on my hard drive as an X-Files image, is this an acceptable medium for install purposes? If not, how do I know which files can be left out of RPMS to create a single ADFS directory? I apologise if these problems are old hat or if I'm just being a little thick and missing something simple but any help or advice would be appreciated (I am fairly capable in DRS/NX). Cheers, Mick. -- Mick Howard Warrington unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 15 15:12:33 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id PAA01198 for ; Thu, 15 Jan 1998 15:12:32 GMT Received: from vger.rutgers.edu ([128.6.190.2]:46690 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11584-24815>; Thu, 15 Jan 1998 15:58:42 +0200 Received: by vger.rutgers.edu id <971540-241>; Thu, 15 Jan 1998 04:26:49 -0500 Received: from hermes.ex.ac.uk ([144.173.6.14]:4007 "EHLO exeter.ac.uk" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971533-241>; Thu, 15 Jan 1998 04:26:31 -0500 Received: from hebe [144.173.6.23] by hermes via SMTP (JAA16674); Thu, 15 Jan 1998 09:28:53 GMT Date: Thu, 15 Jan 1998 09:28:53 +0000 (GMT) From: Phil Norman X-Sender: py95pcfn@hebe To: Russell King - ARM Linux Admin cc: linux-arm@vger.rutgers.edu Subject: Re: Still insmod install problem In-Reply-To: <199801142135.VAA00593@raistlin.armlinux.org> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO On Wed, 14 Jan 1998, Russell King - ARM Linux Admin wrote: > It means a DOS/ADFS/ext2 filesystem, however, ADFS does have the limit of > 77 filenames per directory. (maybe someone could do something like add the > X-files to the linux adfs? Or how about having a set of kerneld-loadable > modules aka RiscOS's Image Filing System? I'm sure that it should be > possible (don't ask me to do it tho - I've got enough on at the moment). I'd love to do some hacking to get linux adfs working with x-files (and LongFileNames too), but unfortunately I'm having huge problems compiling the kernel. I can compile little programs like splitf/joinf, HelloWorld and even a fully working MUD, but I just can't get the kernel to work. I do all the makeing correctly and all the installing, but when the kernel gets to the point where it reads the IDE discs the drive 'hda' times out. I think it sends some dodgy messages to the IDE interface which causes it to fail, since when I reboot into RiscOS, it doesn't understand ADFS::4 anymore (it recognises it again after a reset). The kernel options I configured were, AFAICR, Extended IDE off Old IDE on I assume this is correct, since I'm using the original IDE interface supplied with the first (RiscOS 3.5) RiscPCs (the one that needs the daughter board for CDROMs). I've been having a few other problems too. Several binaries refuse to work. The MUD I mentioned above, on exit, gives a nasty data abort-style error message, as does 'loadmap'. Is this a StrongARM problem (I say this since Chris Johns also mentioned he had problems with some binaries crashing, and he has a StrongARM too)? I've installed all the latest RPMs and nothing seems to fix this. So, although I'd love to do loads of kernel hackery (esp. filesystems), I can't until I can compile my own kernels. So can anyone help PLEASE? Bye for now, Phil -- Phil Norman, mailing from Exeter Uni. Programs available for download from my web page. email: p.c.f.norman@ex.ac.uk web: http://newton.ex.ac.uk/general/ug/norman unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 15 15:22:10 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id PAA01370 for ; Thu, 15 Jan 1998 15:22:09 GMT Received: from vger.rutgers.edu ([128.6.190.2]:46690 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11755-18750>; Thu, 15 Jan 1998 16:15:29 +0200 Received: by vger.rutgers.edu id <971559-241>; Thu, 15 Jan 1998 05:25:16 -0500 Received: from methan.in-berlin.de ([160.45.10.13]:2362 "HELO methan.in-berlin.de" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <971555-241>; Thu, 15 Jan 1998 05:25:03 -0500 Received: by methan.in-berlin.de (Smail3.2.0.98) from bolzen.in-berlin.de (194.94.235.1) with esmtp id ; Thu, 15 Jan 1998 11:27:44 +0100 (MET) Received: by bolzen.in-berlin.de (Smail3.2) from yoda with uucp id ; Thu, 15 Jan 1998 11:27:25 +0100 (MET) Received: by yoda.in-berlin.de (Smail3.1.28.1 #0) id m0xsnU5-000000C; Thu, 15 Jan 98 10:29 GMT Date: Thu, 15 Jan 1998 11:17:02 +0100 From: Stefan Voss To: linux-arm@vger.rutgers.edu Subject: Re: Still insmod install problem Message-ID: In-Reply-To: <199801142135.VAA00593@raistlin.armlinux.org> Organization: Individual Network (IN), Berlin, Germany X-Mailer: Messenger v1.11 for RISC OS X-Posting-Agent: RISC OS Newsbase 0.59f X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO In message <199801142135.VAA00593@raistlin.armlinux.org>, Russell King wrote: > [...] > > Another question: Does installing from a "local hard disk" mean an ext2 > > file system or can an installation be done from an ADFS partition? > > It means a DOS/ADFS/ext2 filesystem, however, ADFS does have the limit of > 77 filenames per directory. (maybe someone could do something like add the Under what name is an ADFS partition known in the Linux install program? The boot messages show an ADFS partition as 'hda1' but the install program only offers the Linux native partitions (hda3 - hda7) as place of the installation files. > [...] > | | Russell King rmk@ecs.soton.ac.uk --- --- Regards, Stefan -- Stefan Voss (voss@yoda.in-berlin.de) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 15 16:00:50 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id QAA01619 for ; Thu, 15 Jan 1998 16:00:49 GMT Received: from vger.rutgers.edu ([128.6.190.2]:46690 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11177-12619>; Thu, 15 Jan 1998 17:05:29 +0200 Received: by vger.rutgers.edu id <971182-241>; Thu, 15 Jan 1998 08:04:07 -0500 Received: from out4.ibm.net ([165.87.194.239]:3258 "EHLO out4.ibm.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971199-241>; Thu, 15 Jan 1998 08:01:16 -0500 Received: from cvs.com.au (slip-32-100-15-228.fl.us.ibm.net [32.100.15.228]) by out4.ibm.net (8.8.5/8.6.9) with ESMTP id NAA52200; Thu, 15 Jan 1998 13:04:00 GMT Message-ID: <34BE09C7.7B404C35@cvs.com.au> Date: Thu, 15 Jan 1998 08:06:15 -0500 From: Charles Esson Reply-To: charlese@cvs.com.au Organization: Colour Vision Systems X-Mailer: Mozilla 4.04 [en] (WinNT; I) MIME-Version: 1.0 To: "linux-arm@vger.rutgers.edu" , Russell King - ARM Linux Admin Subject: Re: Elf standard References: <34B61FA0.474CB30A@cvs.com.au> <34B626B8.91AE8999@cogency.co.uk> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO David Alan Gilbert wrote: > Charles Esson wrote: > > > > Does anyone know if the ELF standard got sorted out for the ARM. If so > > does anyone know where I can get a copy. > > ARM have documented a standard for ARM ELF executables; see the > ARM SDK documentation (which is commercial). > I don't think they have defined anything for object files; but when > they were asked since ARM Linux needs one for Linux 2.1 they said > something like that basically if someone wanted to produce a standard they'd > look at it and if they liked it they would say OK. > Russ has got ELF going (enough for a kernel build) - so ask him for details. No doubt all will think it strange but I am working on a FORTH for the SA1100. I intend to use this to boot linux, and to debug hardware. I would like the FORTH to understand ELF binaries so it can generate them and look at them. To this end I need a copy of the standard. I have ordered an ARM development kit from DEC ( I hope this is the SDK ). Do you have something developed for the object files as indicated by David. Regards Charles Esson unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 15 17:07:05 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id RAA02067 for ; Thu, 15 Jan 1998 17:07:03 GMT Received: from vger.rutgers.edu ([128.6.190.2]:15160 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10584-12619>; Thu, 15 Jan 1998 18:21:08 +0200 Received: by vger.rutgers.edu id <970881-241>; Thu, 15 Jan 1998 11:15:59 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:2043 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970856-241>; Thu, 15 Jan 1998 11:04:43 -0500 Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id QAA12282; Thu, 15 Jan 1998 16:07:32 GMT From: Alan Cox Message-Id: <199801151607.QAA12282@snowcrash.cymru.net> Subject: Re: Elf standard To: charlese@cvs.com.au Date: Thu, 15 Jan 1998 16:07:30 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu, linux@arm.uk.linux.org In-Reply-To: <34BE09C7.7B404C35@cvs.com.au> from "Charles Esson" at Jan 15, 98 08:06:15 am Content-Type: text X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > No doubt all will think it strange but I am working on a FORTH for the SA1100. I > intend to use this to boot linux, and to debug hardware. I would like the FORTH This may be useful anyway - its sure not strange. The Sun prom is written in forth, and there is a standard for initialising I/O cards using forth stuff that much non PC stuff uses. I think PowerPC for one. Alan unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Thu Jan 15 19:37:25 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id TAA03248 for ; Thu, 15 Jan 1998 19:37:23 GMT Received: from vger.rutgers.edu ([128.6.190.2]:9521 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13379-6684>; Thu, 15 Jan 1998 21:28:28 +0200 Received: by vger.rutgers.edu id <970983-241>; Thu, 15 Jan 1998 14:09:37 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:1028 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <970979-241>; Thu, 15 Jan 1998 14:04:56 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xscLj-0000D8-00; Wed, 14 Jan 1998 23:35:55 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: linux-arm@vger.rutgers.edu Subject: gcc 2.8.0 released Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 14 Jan 1998 23:35:55 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi. GCC 2.8.0 has been released and is on the usual FTP sites. It now supports the `arm-linuxaout' and `arm-netbsd' targets out of the box. However: - the supplied config.guess reportedly doesn't grok ARM Linux systems. You need to specify the system name by hand. - you need a recent binutils (2.8.1.0.15 or better; 2.8.1.0.18 is reported to have some problems). - ARM Linux/ELF is not supported at all in this release - there are problems with some parts of the kernel code on non-StrongARM machines, to do with unaligned shorts. You may need to add `-mshort-load-bytes' to your CFLAGS when compiling the kernel as a work around. Enjoy, and apologies to those who have been waiting a long time for it. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 16 08:03:23 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id IAA07952 for ; Fri, 16 Jan 1998 08:03:23 GMT Received: from vger.rutgers.edu ([128.6.190.2]:19282 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <1362-12619>; Fri, 16 Jan 1998 10:01:29 +0200 Received: by vger.rutgers.edu id <970856-241>; Fri, 16 Jan 1998 02:54:09 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:1054 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <971005-241>; Fri, 16 Jan 1998 02:50:37 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xt7VI-0000Bm-00; Fri, 16 Jan 1998 08:51:52 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: linux-arm@vger.rutgers.edu Subject: binutils 2.8.1.0.19 is released (fwd) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 16 Jan 1998 08:51:52 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi. HJ has made a new version of binutils which includes my previous patch. Dave, do you still have problems with compiling the kernel with this version? p. ------- Forwarded Message From: hjl@lucon.org (H.J. Lu) Subject: binutils 2.8.1.0.19 is released. To: Philip.Blundell@pobox.com (Philip Blundell) Date: Thu, 15 Jan 1998 14:39:14 -0800 (PST) Cc: libc-linux@gnu.org (GNU C Library), linux-gcc@vger.rutgers.edu (linuxgcc), kjahds@kjahds.com (Kenneth Albanowski), lmfken@lmf.ericsson.se (Kenneth Osterberg), ian@lasermoon.co.uk, mat@lcs.mit.edu (Mat Hostetter), doughera@lafcol.lafayette.edu (Andy Dougherty), brian@mathworks.com (Brian Bourgault), john@etools.com (John W. Christy), craig@metrolink.com (Craig Groeschel), imp@village.org (Warner Losh), robf@Willows.Com (Rob Farnum), meissner@cygnus.com (Michael Meissner), rfg@monkeys.com (Ron Guilmette), roell@xinside.com (Thomas Roell), burley@gnu.org (Craig Burley), linux-binutils-in@polstra.com (John Polstra), Shimon@i-Connect.Net (Simon Shapiro), galenh@micron.net (Galen Hazelwood), ralf@mailhost.uni-koblenz.de (Ralf Baechle) In-Reply-To: from "Philip Blundell" at Dec 27, 97 12:08:27 pm X-Mailer: ELM [version 2.4 PL25 PGP2] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit > Here's a patch against binutils 2.8.1.0.18. It contains the initial > support for ELF and PIC code on the ARM. It also contains a workaround for a Thanks. I changed Linux/ARM/ELF target from arm-linuxelf to arm-linux*. > bug that causes objdump to crash when disassembling ARM code. > > Please include it in the next binutils release. > H.J. - ---- This is the beta release of binutils 2.8.1.0.19 for Linux, which is based on the binutils 980114 snapshot plus linux/PPC patch and Linux/ARM support. This release should work on linux/x86. But there may be some problems on Linux/alpha. I was told as was ok, but ld and/or dynamic linker were not right on Linux/alpha. I couldn't finish "make check" on Linux/alpha with the libg++ addon. There are some ELF bug fixes. But I don't know if it fixes all the bugs reported on linux/sparc, Linux/MIPS and linux/PPC. I'd like to hear reports on them. Please report any bugs related to binutils 2.8.1.0.19 to hjl@gnu.ai.mit.edu. Changes from binutils 2.8.1.0.18: 1. Update from the binutils gas-980114. 2. The ARM/ELF support. 3. Fix lib*.so.* support. Use the release number in soname. Changes from binutils 2.8.1.0.17: 1. Update from the binutils gas-971211. 2. Add lib*.so.* support. But the soname is lib*.so.2. Changes from binutils 2.8.1.0.16: 1. Update from the binutils gas-971105. 2. Fix ld test for linux. 3. Fix the 2.8.1.0.16 patch. Changes from binutils 2.8.1.0.15: 1. Update from the binutils gas-971103. 2. ELF/Alpha seems to work fine, including versioning. 3. ELF/PPC patches from Geoff Keating . IMPORTANT: The .plt format that the Alpha was using was not thread safe. So I changed it. The new format is *not* binary compatible with the old, thus you must use glibc 2.0.4 to get an ld.so that can understand it. Note that the new ld.so knows how to deal with the old .plt format, so old libraries will still work, but they should be relinked eventually for performance reasons. The file list: 1. binutils-2.8.1.0.19.tar.gz. Source code. 2. binutils-2.8.1.0.18-2.8.1.0.19.diff.gz. Patch against the previous beta source code. 3. binutils-2.8.1.0.19.bin.tar.gz. Precompiled Linux/x86 binaries for libc 5. 4. binutils-2.8.1.0.19-glibc.bin.tar.gz. Precompiled Linux/x86 binaries for libc 6 (glibc 2.0.5c or above). The primary ftp sites for the compiler/C library are tsx-11.mit.edu under pub/linux/packages/GCC and sunsite.unc.edu under pub/Linux/GCC. The beta directory is in private/tofu under the GCC directory. Most of my stuff can also be found at ftp://ftp.yggdrasil.com/private/hjl To install the binary package, please follow the procedure very closely. Please backup/save all the files you are instructed to delete and you should do gzip -dc binutils-2.8.1.0.19.bin.tar.gz | tar tvvf - or gzip -dc binutils-2.8.1.0.19-glibc.bin.tar.gz | tar tvvf - to see what is in there. Please do back up before you remove things. To install for libc 5, PLEASE DO 1. su root 2. cd / 3. rm -f /usr/bin/encaps /usr/bin/nm 4. gzip -dc binutils-2.8.1.0.19.bin.tar.gz | tar xvvf - 5. ldconfig To install for libc 6 (glibc 2.0.5c or above), PLEASE DO 1. su root 2. cd / 3. rm -f /usr/bin/encaps /usr/bin/nm 4. gzip -dc binutils-2.8.1.0.19-glibc.bin.tar.gz | tar xvvf - 5. ldconfig Now you have the new gas/binutils under /usr/bin and /usr/ix86-linuxaout/bin. You have to use /usr/ix86-linuxaout/bin/as and /usr/ix86-linuxaout/bin/ld -m i386linux if you want to use a.out as and ld directly. I have changed the target names for Linux/x86 and Linux/x86 (a.out) to ix86-linux and ix86-linuxaout respectively. The precompiled binaries are installed under /usr/ix86-linux and /usr/ix86-linuxaout. You should make appropriate symbolic links if you have a different name for the Linux/x86 target, like i586-unknown-linux. If you have an old linux library in the a.out format and you cannot obtain the newer version in the ELF format for whatever reason, you can try "objcopy --remove-leading-char" on the a.out library and see if it can link with your code in ELF. For Sybase client libraries, I did # cd sybase/lib # for f in *.a do objcopy --remove-leading-char $f done It seems to compile/link fine with libc 5.4.2x. But I don't know if it really works or not. The resulting binaries should work if the libc ABI used by the old library is unchanged in the new libc. According to Bruce Milner : Just a quick note about your example using objcopy for sybase. The sybase a.out libraries "are" dependent on a changed feature of the libc ABI. The ctype macros changed to a byte order independent format (little endian). In order to use the sybase libraries, one needs to rename the the ctype table variables in the sybase library files and link in an old version of the libc's "C-ctype.o ctype-info.o" files. thain@sunquest.sunquest.com put a ELF'ized version of these libraries on ftp://mudshark.sunquest.com/pub/ctlib-linux-elf Thanks. H.J. Lu hjl@gnu.org 01/15/98 ------- End of Forwarded Message unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 16 10:05:34 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id KAA08700 for ; Fri, 16 Jan 1998 10:05:32 GMT Received: from vger.rutgers.edu ([128.6.190.2]:20823 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <2473-18750>; Fri, 16 Jan 1998 12:00:37 +0200 Received: by vger.rutgers.edu id <970955-241>; Fri, 16 Jan 1998 04:55:16 -0500 Received: from methan.in-berlin.de ([160.45.10.13]:1550 "HELO methan.in-berlin.de" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970948-241>; Fri, 16 Jan 1998 04:55:08 -0500 Received: by methan.in-berlin.de (Smail3.2.0.98) from bolzen.in-berlin.de (194.94.235.1) with esmtp id ; Fri, 16 Jan 1998 10:57:50 +0100 (MET) Received: by bolzen.in-berlin.de (Smail3.2) from yoda with uucp id ; Fri, 16 Jan 1998 10:57:27 +0100 (MET) Received: by yoda.in-berlin.de (Smail3.1.28.1 #0) id m0xt9PZ-000000C; Fri, 16 Jan 98 09:54 GMT Date: Fri, 16 Jan 1998 10:50:14 +0100 From: Stefan Voss To: linux-arm@vger.rutgers.edu Subject: Re: Still insmod install problem Message-ID: In-Reply-To: Organization: Individual Network (IN), Berlin, Germany X-Mailer: Messenger v1.11 for RISC OS X-Posting-Agent: RISC OS Newsbase 0.59f X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO In message , Phil Norman wrote: > [...] > > I've been having a few other problems too. Several binaries refuse to > work. The MUD I mentioned above, on exit, gives a nasty data abort-style > error message, as does 'loadmap'. Is this a StrongARM problem (I say > this since Chris Johns also mentioned he had problems with some binaries > crashing, and he has a StrongARM too)? I've installed all the latest > RPMs and nothing seems to fix this. Just FYI my installation problem isn't a StrongARM problem. I just put my old ARM 610 back in and got the same error (insmod crashing) as before. > [...] > Bye for now, > Phil > Regards, Stefan -- Stefan Voss (voss@yoda.in-berlin.de) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 16 10:41:20 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id KAA08929 for ; Fri, 16 Jan 1998 10:41:19 GMT Received: from vger.rutgers.edu ([128.6.190.2]:27146 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <17294-12619>; Fri, 16 Jan 1998 12:40:24 +0200 Received: by vger.rutgers.edu id <970884-241>; Fri, 16 Jan 1998 05:35:21 -0500 Received: from nexusel.demon.co.uk ([158.152.30.195]:4188 "HELO globe" ident: "SOCKFAULT1") by vger.rutgers.edu with SMTP id <970800-241>; Fri, 16 Jan 1998 05:31:42 -0500 Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xt8tt-0001JM-00; Fri, 16 Jan 1998 10:21:21 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xt8ts-00061s-00; Fri, 16 Jan 1998 10:21:20 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: hjl@gnu.org Cc: linux-arm@vger.rutgers.edu Subject: binutils 2.8.1.0.19 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 16 Jan 1998 10:21:20 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi. Pat Beirne sent me this small patch. It is needed for BFD to generate working ELF code on the ARM. HJ, is it possible to get hold of the gas snapshot source that you based your release on? I'd like to make a patch to send to the gas2 people for inclusion in the master sources. Although I thought I was subscribed to the gas2 mailing list I don't think I've ever had a message from it, so either it's not working or my subscription didn't happen. p. Index: bfd/ChangeLog.linux =================================================================== RCS file: /usr/repository/gnu/binutils/bfd/ChangeLog.linux,v retrieving revision 1.3 diff -u -r1.3 ChangeLog.linux --- ChangeLog.linux 1998/01/16 10:00:01 1.3 +++ ChangeLog.linux 1998/01/16 10:06:27 @@ -1,3 +1,7 @@ +Fri Jan 16 10:06:19 1998 Philip Blundell + + * elf32-arm.c: Use REL relocs not RELA. + Thu Dec 25 23:42:21 1997 Philip Blundell * config.bfd: (arm-*-linux*, arm-*-elf, arm-*-netbsd): New Index: bfd/elf32-arm.c =================================================================== RCS file: /usr/repository/gnu/binutils/bfd/elf32-arm.c,v retrieving revision 1.1 diff -u -r1.1 elf32-arm.c --- elf32-arm.c 1998/01/16 10:00:05 1.1 +++ elf32-arm.c 1998/01/16 10:06:14 @@ -23,7 +23,7 @@ #include "libbfd.h" #include "elf-bfd.h" -#define USE_RELA /* we want RELA relocations, not REL */ +#define USE_REL /* we want REL relocations, not RELA */ /* ARM relocations */ enum arm_reloc_type unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 16 11:40:08 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id LAA09391 for ; Fri, 16 Jan 1998 11:40:07 GMT Received: from vger.rutgers.edu ([128.6.190.2]:52037 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11368-18750>; Fri, 16 Jan 1998 13:39:19 +0200 Received: by vger.rutgers.edu id <970800-241>; Fri, 16 Jan 1998 06:34:44 -0500 Received: from pecan.acorn.co.uk ([136.170.131.22]:53353 "HELO pecan.acorn.co.uk" ident: "TIMEDOUT") by vger.rutgers.edu with SMTP id <970863-241>; Fri, 16 Jan 1998 06:34:06 -0500 Received: from ether120.acorn.co.uk by pecan.acorn.co.uk (SMI-8.6/SMI-SVR4) id LAA13122; Fri, 16 Jan 1998 11:37:03 GMT Date: Fri, 16 Jan 1998 11:37:01 +0000 (GMT) From: John Vickers Subject: Re: binutils 2.8.1.0.19 To: Philip Blundell Cc: linux-arm@vger.rutgers.edu In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII X-Organization: Organisation name, location. Telephone/Fax? X-Mailer: ANT RISCOS Marcel [ver 1.26] X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi. On Fri 16 Jan, Philip Blundell wrote: > Pat Beirne sent me this small patch. It is needed for BFD to generate > working ELF code on the ARM. [...] What compiler are you using for arm-elf ? John Vickers. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 16 12:10:55 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id MAA09585 for ; Fri, 16 Jan 1998 12:10:54 GMT Received: from vger.rutgers.edu ([128.6.190.2]:23650 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <17195-6684>; Fri, 16 Jan 1998 14:07:05 +0200 Received: by vger.rutgers.edu id <970881-241>; Fri, 16 Jan 1998 07:02:55 -0500 Received: from server21.digital.fr ([193.56.15.21]:1280 "EHLO server21.digital.fr" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970863-241>; Fri, 16 Jan 1998 07:02:36 -0500 Received: from mail.vbo.dec.com (mail.vbo.dec.com [16.36.208.34]) by server21.digital.fr (8.7.5/8.7) with ESMTP id NAA20447; Fri, 16 Jan 1998 13:04:34 +0100 (MET) Received: from vbormc.vbo.dec.com (vbormc.vbo.dec.com [16.36.208.94]) by mail.vbo.dec.com (8.8.6/8.7) with ESMTP id NAA06973; Fri, 16 Jan 1998 13:04:33 +0100 (MET) Received: from linux.reo.dec.com (linux.reo.dec.com [16.36.0.79]) by vbormc.vbo.dec.com (8.7.3/8.7) with ESMTP id NAA28395; Fri, 16 Jan 1998 13:45:20 +0100 Received: from linux.reo.dec.com (localhost [127.0.0.1]) by linux.reo.dec.com (8.8.5/8.7.1) with ESMTP id MAA29271; Fri, 16 Jan 1998 12:02:37 GMT Message-Id: <199801161202.MAA29271@linux.reo.dec.com> To: charlese@cvs.com.au cc: "linux-arm@vger.rutgers.edu" , Russell King - ARM Linux Admin Subject: Re: Elf standard In-reply-to: Your message of "Thu, 15 Jan 1998 08:06:15 EST." <34BE09C7.7B404C35@cvs.com.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 16 Jan 1998 12:02:37 +0000 From: "Ka'Plaagh" X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO I don't think that it's strange, we used forth on the EBSA285. I presume that you're using the work that Neal Crook did (he also did the EBSA110 forth stuff) Dave ---------------------------------------------------------------------- David A Rusling Consulting Engineer European Semiconductor Applications Digital Equipment Co Ltd., Engineering PO Box 121, Imperial Way, Worton Grange Reading RG2 0TU Linux, Alpha, StrongArm, PCI Tel: UK-(0)1734-204380 Fax: UK-(0)1734-203133 ---------------------------------------------------------------------- unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Fri Jan 16 13:50:35 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id NAA10700 for ; Fri, 16 Jan 1998 13:50:34 GMT Received: from vger.rutgers.edu ([128.6.190.2]:33375 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11975-18750>; Fri, 16 Jan 1998 15:48:53 +0200 Received: by vger.rutgers.edu id <970883-241>; Fri, 16 Jan 1998 08:41:43 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:6739 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <970893-241>; Fri, 16 Jan 1998 08:39:52 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id NAA10482; Fri, 16 Jan 1998 13:42:20 GMT From: Matthew Wilcox Message-Id: <199801161342.NAA10482@odie.barnet.ac.uk> Subject: Re: Still insmod install problem To: P.C.F.Norman@exeter.ac.uk (Phil Norman) Date: Fri, 16 Jan 1998 13:42:19 +0000 (GMT) Cc: linux@arm.uk.linux.org, linux-arm@vger.rutgers.edu In-Reply-To: from "Phil Norman" at Jan 15, 98 09:28:53 am X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > The kernel options I configured were, AFAICR, > Extended IDE off > Old IDE on This is incorrect. > I assume this is correct, since I'm using the original IDE interface > supplied with the first (RiscOS 3.5) RiscPCs (the one that needs the > daughter board for CDROMs). It's not a matter of age of system, it's age of driver. The older one is smaller, but less featureful. It (apparently) does not work with some systems. Since almost everyone uses the new driver, I suspect that is the one that Russell worked on porting. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sat Jan 17 19:20:01 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id TAA23373 for ; Sat, 17 Jan 1998 19:20:00 GMT Received: from vger.rutgers.edu ([128.6.190.2]:43785 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11330-24815>; Sat, 17 Jan 1998 21:17:45 +0200 Received: by vger.rutgers.edu id <970866-241>; Sat, 17 Jan 1998 14:04:54 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1028 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970964-241>; Sat, 17 Jan 1998 13:52:32 -0500 Received: from caramon.armlinux.org (dialup3.cymru.net [163.164.160.67]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id SAA27249; Sat, 17 Jan 1998 18:55:56 GMT Received: from raistlin.armlinux.org (raistlin [192.168.0.3]) by caramon.armlinux.org (8.7.4/8.7.3) with ESMTP id WAA17722; Fri, 16 Jan 1998 22:45:56 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.armlinux.org (8.7.4/8.7.3) id SAA09060; Sat, 17 Jan 1998 18:57:02 GMT Message-Id: <199801171857.SAA09060@raistlin.armlinux.org> Subject: Re: ARM-Linux for the A5000 To: Dave@imladris.demon.co.uk (David Woodhouse) Date: Sat, 17 Jan 1998 18:57:01 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: from "David Woodhouse" at Jan 16, 98 11:35:29 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO David Woodhouse writes: > init=/bin/bash causes it to hang. > init=/bin/ash works. Thereafter, running bash causes an "Illegal instruction" Bash *needs* the FPEmulator module loaded in order to run. > > *All* applications should run on an ARM3, and up. Which binaries > > don't run? > > Illegal instruction: bash, talk, make See above - these also need the FPE. > Can I (cross-)compile standard mount utils and use them on the ARM? What about > mkswap'ing on an i386 - would the resulting partition be usable on the ARM? mkswap is quite different on the ARM - the swap file is page size dependent. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sat Jan 17 20:47:11 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id UAA23968 for ; Sat, 17 Jan 1998 20:47:10 GMT Received: from vger.rutgers.edu ([128.6.190.2]:52320 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10761-24815>; Sat, 17 Jan 1998 22:46:29 +0200 Received: by vger.rutgers.edu id <970912-241>; Sat, 17 Jan 1998 15:41:43 -0500 Received: from ba1.augsburg.baynet.de ([194.95.202.130]:20827 "EHLO ba1.augsburg.baynet.de" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970903-241>; Sat, 17 Jan 1998 15:41:24 -0500 Received: from ba471101 (dial053.augsburg.baynet.de [194.95.202.23]) by ba1.augsburg.baynet.de (8.8.8/8.8.8) with SMTP id VAA30323 for ; Sat, 17 Jan 1998 21:45:02 +0100 Date: Sat, 17 Jan 1998 14:50:00 +0100 From: Richard Atterer To: linux-arm@vger.rutgers.edu Subject: Re: Still insmod install problem Message-ID: <4809E0E960%atterer@informatik.tu-muenchen.de> In-Reply-To: Reply-To: atterer@augsburg.baynet.de Organization: Home X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO In message Phil Norman wrote: > I'd love to do some hacking to get linux adfs working with x-files (and > LongFileNames too), but unfortunately I'm having huge problems compiling > the kernel. BTW: raFS supports long filenames, too! I don't know anything about how Linux accesses RiscOS filing systems, but raFS is no image filing system - it's a "normal" filing system just like ADFS. Will this make things easier? I'm an absolute Linux newbie (I don't even program in C;), so I don't know whether this is of any relevance. [snip rest] Cheers, Richard -- __ _ |_) /| Richard Atterer | \/¯| http://home.augsburg.baynet.de/richard.atterer/ raFS V1.10 ¯ ´` ¯ .. "Bother", said Pooh, as he saw "FileCore in use" unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sat Jan 17 21:01:26 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id VAA23975 for ; Sat, 17 Jan 1998 21:01:25 GMT Received: from vger.rutgers.edu ([128.6.190.2]:9581 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <9432-6684>; Sat, 17 Jan 1998 23:00:33 +0200 Received: by vger.rutgers.edu id <971083-241>; Sat, 17 Jan 1998 15:55:10 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1897 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971070-241>; Sat, 17 Jan 1998 15:48:46 -0500 Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id UAA28993; Sat, 17 Jan 1998 20:51:50 GMT From: Alan Cox Message-Id: <199801172051.UAA28993@snowcrash.cymru.net> Subject: Re: Still insmod install problem To: atterer@augsburg.baynet.de Date: Sat, 17 Jan 1998 20:51:48 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: <4809E0E960%atterer@informatik.tu-muenchen.de> from "Richard Atterer" at Jan 17, 98 02:50:00 pm Content-Type: text X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > I don't know anything about how Linux accesses RiscOS filing systems, but > raFS is no image filing system - it's a "normal" filing system just like > ADFS. Will this make things easier? Linux doesnt actually care. The Linux model of a filing system is basically generic file operations at the top, a block device at the bottom. "Image" file systems are handled by a device that takes a file as its bottom layer and makes it a block device. Alan unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sat Jan 17 23:49:47 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id XAA25187 for ; Sat, 17 Jan 1998 23:49:46 GMT Received: from vger.rutgers.edu ([128.6.190.2]:5745 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10843-6684>; Sun, 18 Jan 1998 01:44:15 +0200 Received: by vger.rutgers.edu id <970958-241>; Sat, 17 Jan 1998 18:39:36 -0500 Received: from post-20.mail.demon.net ([194.217.242.27]:53840 "HELO post.mail.demon.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970964-241>; Sat, 17 Jan 1998 18:36:46 -0500 Received: from treblig.demon.co.uk ([194.222.210.232]) by post.mail.demon.net id aa2024666; 17 Jan 98 23:36 GMT Date: Sat, 17 Jan 1998 23:33:04 +0000 (GMT) From: Dave Gilbert X-Sender: gilbertd@tardis.home.dave To: linux-arm@vger.rutgers.edu Subject: Cross compile instructions Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi, As promised here are my cross compile instructions; some of the finer points may only apply to using an Alpha as a host but they shouldn't cause a problem crossing from anything else. -------------------------------------------------------------------- Cross Building ARM kernels from Alphas (linux@treblig.org) ---------------------------------------------------------- (The techniques in this file probably also allow you to cross from other platforms as well, reports of success and failure would be appreciated) I'm using the EGCS compiler; its a modified version of GCC; its more uptodate than GCC 2.7.2.3 (which won't do the job) - its not the same as GCC 2.8.0. I'm using EGCS snapshot 971225. It can be found on ftp.cygnus.com/egcs/snapshots ------------------------------------------------------------------------------ Then apply this fix: The patch I have is a small one by Richard Earnshaw (of ARM), it fixes a problem with generating some constants on 64 bit machines. (The patch here is hand copied!) Its in egcs/gcc/explow.c around line 120 originally reads: if (GET_CODE (XEXP (x, 1)) == CONST_INT) return plus_constant (XEXP (x, 0), c + INTVAL (XEXP (x, 1))); Change it to: if (GET_CODE (XEXP (x, 1)) == CONST_INT) return plus_constant (XEXP (x, 0), ((c + INTVAL (XEXP (x, 1))) & GET_MODE_MASK (mode))); ------------------------------------------------------------------------------ Binutils - before building the compiler you need a working set of binutils; I used version 2.8.1.0.15 (anything earlier than 2.8.1.0.15 will probably break). >> 2.8.1.0.18,2.8.1.0.19 is broken << Configure it with: ./configure arm-unknown-linuxaout --prefix=/home/arm/tools that should build easily. Make install In /home/arm/tools/bin you'll find some (most?) binaries with no prefix and some starting with arm-unknown-linuxaout- - use a shell script to rename all files to begin with arm-unknown-linuxaout- - thus my bin directory looks like: [gilbertd@tardis bin]$ ls arm-unknown-linuxaout-addr2line arm-unknown-linuxaout-objcopy arm-unknown-linuxaout-ar arm-unknown-linuxaout-objdump arm-unknown-linuxaout-as arm-unknown-linuxaout-protoize arm-unknown-linuxaout-c++ arm-unknown-linuxaout-ranlib arm-unknown-linuxaout-c++filt arm-unknown-linuxaout-size arm-unknown-linuxaout-g++ arm-unknown-linuxaout-strings arm-unknown-linuxaout-g77 arm-unknown-linuxaout-strip arm-unknown-linuxaout-gasp arm-unknown-linuxaout-unprotoize arm-unknown-linuxaout-gcc gcov arm-unknown-linuxaout-gprof protoize arm-unknown-linuxaout-ld unprotoize arm-unknown-linuxaout-nm (Although some of these won't be there yet for you!) ------------------------------------------------------------------------------ Put /home/arm/tools/bin in your path and Now configure EGCS with: ./configure --target=arm-unknown-linuxaout --prefix=/home/arm/tools --with-gnu -ld --host=alpha-local-linux-gnu Next you need to create /home/arm/tools/arm-unknown-linuxaout/include and populate it with the include directory off an ARM Linux system; all stdio.h etc is needed and make symlinks from include/linux and include/asm and include/scsi etc. to whereever your ARM Linux kernel source is, in the same way that you would when building a kenel native. Also float.h is needed to build some of the stuff (objc only I think??) do make LANGUAGES="C" make install LANGUAGES="C" (I had some problems here - but I can't remember what - tell me if you hit any - I think they were from me not having copied the includes in yet). One reported problem is that if you don't have a _G_config.h in your include directory it fails building an st-dummy.c - hmm. As far as I can tell it is the install of GCC which creates it - try doing a make install and then doing it again ? You should find even if this fails that the build has finished in the egcs/gcc directory - and that that can be installed. IF you are really lucky it will let you build for other languages as well. ------------------------------------------------------------------------------ Now time for the kernel.... Edit linux/include/asm-arm/a.out.h, add a #include near the top (NOTE this is actually wrong but works - trying to find safe soloution for this!) and change all the types in the struct exec to __u32 (from unsigned and unsigned long). Edit the Makefile to set ARCH = arm edit the CROSS_COMPILE= line to: CROSS_COMPILE = /home/arm/tools/bin/arm-unknown-linuxaout- Edit arch/arm/Makefile, type make xconfig, set it up as you want and then type: make and then make Image (make zImage not done yet). ------------------------------------------------------------------------------- In egcs-971225 you'l get an error building gcc/cse.c (multiple defaults in a case statement at line 751 and 755; 751 is a macro, 755 is a default: followed by a break; Remove the default: break from cse.c ------------------------------------------------------------------------------- There is an erm...disagreement... between the Linux kernel and the newer C compilers about an alignment issue - add -mshort-load-bytes to the CFLAGS below. (This probably also stops the use of the fancy new halfword instructions which don't work on StrongARM RiscPCs) [This is a temporary hack - a proper fix will arrive later - 11/01/98] --------------------------------------------- Change arch/arm/makefile so that the section is updated to: ifeq ($(PROCESSOR),armo) CFLAGS_PROC += -mapcs-26 -D__arm2__ -mshort-load-bytes TEXTADDR = 0x02080000 ZTEXTADDR = 0x01800000 ZRELADDR = 0x02080000 HEAD := arch/arm/kernel/head.o endif ifeq ($(PROCESSOR),armv) CFLAGS_PROC += -mapcs-32 -D__arm6__ -mshort-load-bytes TEXTADDR = 0xC0008000 ZTEXTADDR = 0x00008000 ZRELADDR = 0x00008000 HEAD := arch/arm/kernel/head32.o endif you might also like to add -mcpu-strongarm if you have one (untested!) ------------------------------------------------------------------------------- To build Linux on a non-linux sysytem (e.g. DEC UNIX) you need gmake,gfind and bash and you need to edit the top level makefile to use them instead of the standard system utilities. Another problem is genksyms; you need this to be able to add version information to the kernel for modules - and you need that to be able to load the floating point emulator module. You can get this in the Linux modutils package (available from ftp://ftp.kernel.org/pub/linux/kernel/v2.1/modutils-2.1.55.tar.gz) I'm told that this works nicely. -------------------------------------------------------------------- / Dr. David Alan Gilbert | Running Linux on Alpha & ARM \ \ gro.gilbert @ treblig.org | ------- Happy in hex ------- / \____________________________|___ http://www.treblig.demon.co.uk __/ unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 18 09:20:17 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id JAA28579 for ; Sun, 18 Jan 1998 09:20:16 GMT Received: from vger.rutgers.edu ([128.6.190.2]:20528 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13998-12619>; Sun, 18 Jan 1998 11:19:50 +0200 Received: by vger.rutgers.edu id <971327-241>; Sun, 18 Jan 1998 04:15:10 -0500 Received: from methan.in-berlin.de ([160.45.10.13]:2848 "HELO methan.in-berlin.de" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <971333-241>; Sun, 18 Jan 1998 04:14:20 -0500 Received: by methan.in-berlin.de (Smail3.2.0.98) from bolzen.in-berlin.de (194.94.235.1) with esmtp id ; Sun, 18 Jan 1998 10:17:55 +0100 (MET) Received: by bolzen.in-berlin.de (Smail3.2) from yoda with uucp id ; Sun, 18 Jan 1998 10:17:25 +0100 (MET) Received: by yoda.in-berlin.de (Smail3.1.28.1 #0) id m0xtrn1-000000C; Sun, 18 Jan 98 09:17 GMT Date: Sun, 18 Jan 1998 10:11:53 +0100 From: Stefan Voss To: linux-arm@vger.rutgers.edu Subject: Re: Installation Problems Message-ID: <13494ba48%voss@yoda.in-berlin.de> In-Reply-To: Organization: Individual Network (IN), Berlin, Germany X-Mailer: Messenger v1.11 for RISC OS X-Posting-Agent: RISC OS Newsbase 0.59f X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO In message , Phil Norman wrote: > [...] > I've written a complete installation guide for installing armlinux from > the Clan CD (which has the RPMs in an x-files archive too). You can > download this from my web site (URL in the sig). I'm afraid it's several > meg, but it does contain everything you'll need to install armlinux. The > guide will work just as well if you've downloaded the RPMs and haven't got > them from the Clan CD. By following Phil's install guide I was able to copy the RPMS to the native Linux partitions and do a local hard disk install. Thanks Phil! The most recent kernel still had insmod crashing so I had to use an earlier kernel (from Oct 26th). As described in Phil's install guide the install program says "Couldn't find a kernel" when it wants to configure lilo. At what location is the install program looking for a kernel? It shouldn't be a problem to copy a kernel manually to the location where the install program can find it. After that the install program could finish its lilo configuration. Another problem occurrs when I try to start Linux by doubleclicking on !Linux. I used !LinConfig to set the kernel location to '*ADFS::4.0'. When I doubleclick on !Linux I get the error message: "invalid bmap magicin first sector of partition". I suppose that this error occurrs because the lilo setup couldn't be finished. I tried the "Override kernel" option in !Linux.Choices and in !Boot.Choices.Linux.Config to use a kernel from RiscOS but I got the same error. > > Best of luck, > Phil > Regards, Stefan -- Stefan Voss (voss@yoda.in-berlin.de) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 18 11:59:04 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id LAA29674 for ; Sun, 18 Jan 1998 11:59:03 GMT Received: from vger.rutgers.edu ([128.6.190.2]:39991 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13185-18750>; Sun, 18 Jan 1998 13:58:33 +0200 Received: by vger.rutgers.edu id <971483-241>; Sun, 18 Jan 1998 06:53:57 -0500 Received: from irwell.zetnet.co.uk ([194.247.47.48]:28758 "EHLO irwell.zetnet.co.uk" ident: "root") by vger.rutgers.edu with ESMTP id <971484-241>; Sun, 18 Jan 1998 06:53:27 -0500 Received: from Michael (man-126.dialup.zetnet.co.uk [194.247.40.161]) by irwell.zetnet.co.uk (8.8.7/8.8.5) with SMTP id LAA14053; Sun, 18 Jan 1998 11:57:12 GMT Date: Sun, 18 Jan 1998 11:57:41 +0000 (GMT) From: Michael Howard Subject: Re: Installation Problems To: Stefan Voss Cc: Linux Mailing List In-Reply-To: <13494ba48%voss@yoda.in-berlin.de> Illegal-Object: Syntax error in Message-ID: value found on vger.rutgers.edu: Message-ID: ^-illegal end of message identification MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII X-Organization: Organisation name, location. Telephone/Fax? X-Mailer: ANT RISCOS Marcel [ver 1.26] X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Message-Id: <19980118115402Z971483-241+22019@vger.rutgers.edu> Status: RO > Another problem occurrs when I try to start Linux by doubleclicking on > !Linux. I used !LinConfig to set the kernel location to '*ADFS::4.0'. When > I doubleclick on !Linux I get the error message: > > "invalid bmap magicin first sector of partition". I get exactly the same message but I'm using my 2nd IDE drive, 5. > > I suppose that this error occurrs because the lilo setup couldn't be > finished. I tried the "Override kernel" option in !Linux.Choices and in > !Boot.Choices.Linux.Config to use a kernel from RiscOS but I got the same > error. > I tried it too with the same result. Additionally I had to manually edit the Choices file as !LinConfig fails to save a Choices file. During installation ony a third of the RPMS files installed correctly. Any ideas as to why? They came of the Clan CD. I'll keep trying. Cheers, Mick. -- Mick Howard Warrington unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 18 15:28:13 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id PAA31073 for ; Sun, 18 Jan 1998 15:28:12 GMT Received: from vger.rutgers.edu ([128.6.190.2]:48135 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12591-12619>; Sun, 18 Jan 1998 17:27:08 +0200 Received: by vger.rutgers.edu id <971790-241>; Sun, 18 Jan 1998 10:21:51 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:2003 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971792-241>; Sun, 18 Jan 1998 10:19:42 -0500 Received: from caramon.armlinux.org (dialup2.cymru.net [163.164.160.66]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id PAA10656 for ; Sun, 18 Jan 1998 15:23:33 GMT Received: from raistlin.armlinux.org (raistlin [192.168.0.3]) by caramon.armlinux.org (8.7.4/8.7.3) with ESMTP id PAA18666 for ; Sun, 18 Jan 1998 15:14:21 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.armlinux.org (8.7.4/8.7.3) id PAA04307 for linux-arm@vger.rutgers.edu; Sun, 18 Jan 1998 15:14:17 GMT Message-Id: <199801181514.PAA04307@raistlin.armlinux.org> Subject: Insmod crashing problems... To: linux-arm@vger.rutgers.edu Date: Sun, 18 Jan 1998 15:14:16 +0000 (GMT) X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi. It concerns me that some people are having a problem with insmod crashing. There was a known problem with early insmod versions whereby if you had >=24MB of RAM, then when insmod is used, it doesn't perform all the fixup's between the kernel and the module correctly. This can be identified by a page fault / oops which looks like: pc: [<0xc4......>] lr: [<0xc2......>] ... Process insmod .... This is because the ARM's b/bl instructions can only address +/- 16MB from the current location. The latest insmod fixes this problem, and should be on the latest installation disks on the FTP site. Can the people who are having problems with insmod please check that they are using the latest disks, and the latest insmod package. If they are, then can they urgently mail me. Thanks. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 18 22:15:09 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id WAA02028 for ; Sun, 18 Jan 1998 22:15:08 GMT Received: from vger.rutgers.edu ([128.6.190.2]:54128 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <17055-12619>; Mon, 19 Jan 1998 00:13:43 +0200 Received: by vger.rutgers.edu id <970912-241>; Sun, 18 Jan 1998 17:07:21 -0500 Received: from methan.in-berlin.de ([160.45.10.13]:1626 "HELO methan.in-berlin.de" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970986-241>; Sun, 18 Jan 1998 16:58:12 -0500 Received: by methan.in-berlin.de (Smail3.2.0.98) from bolzen.in-berlin.de (194.94.235.1) with esmtp id ; Sun, 18 Jan 1998 23:02:14 +0100 (MET) Received: by bolzen.in-berlin.de (Smail3.2) from yoda with uucp id ; Sun, 18 Jan 1998 23:01:52 +0100 (MET) Received: by yoda.in-berlin.de (Smail3.1.28.1 #0) id m0xu3ki-000000C; Sun, 18 Jan 98 22:03 GMT Date: Sun, 18 Jan 1998 23:01:36 +0100 From: Stefan Voss To: linux-arm@vger.rutgers.edu Subject: Re: Installation Problems Message-ID: <30c191a48%voss@yoda.in-berlin.de> In-Reply-To: <199801181510.PAA04292@raistlin.armlinux.org> Organization: Individual Network (IN), Berlin, Germany X-Mailer: Messenger v1.11 for RISC OS X-Posting-Agent: RISC OS Newsbase 0.59f X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO In message <199801181510.PAA04292@raistlin.armlinux.org> you wrote: > Stefan Voss writes: > > The most recent kernel still had insmod crashing so I had to use an > > earlier kernel (from Oct 26th). > > What is the problem? Does insmod give an illegal instruction? It's still the crash that has pc=0xc4 and lr=0xc2 > > [...] > > | | Russell King rmk@ecs.soton.ac.uk --- --- Regards, Stefan -- Stefan Voss (voss@yoda.in-berlin.de) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 18 22:15:18 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id WAA02031 for ; Sun, 18 Jan 1998 22:15:17 GMT Received: from vger.rutgers.edu ([128.6.190.2]:54128 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <17319-12619>; Mon, 19 Jan 1998 00:14:25 +0200 Received: by vger.rutgers.edu id <970958-241>; Sun, 18 Jan 1998 17:07:28 -0500 Received: from methan.in-berlin.de ([160.45.10.13]:1624 "HELO methan.in-berlin.de" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970974-241>; Sun, 18 Jan 1998 16:58:07 -0500 Received: by methan.in-berlin.de (Smail3.2.0.98) from bolzen.in-berlin.de (194.94.235.1) with esmtp id ; Sun, 18 Jan 1998 23:02:02 +0100 (MET) Received: by bolzen.in-berlin.de (Smail3.2) from yoda with uucp id ; Sun, 18 Jan 1998 23:01:52 +0100 (MET) Received: by yoda.in-berlin.de (Smail3.1.28.1 #0) id m0xu2Pn-000000C; Sun, 18 Jan 98 20:37 GMT Date: Sun, 18 Jan 1998 21:26:37 +0100 From: Stefan Voss To: linux-arm@vger.rutgers.edu Subject: Re: Insmod crashing problems... Message-ID: <33f89a48%voss@yoda.in-berlin.de> In-Reply-To: <199801181514.PAA04307@raistlin.armlinux.org> Organization: Individual Network (IN), Berlin, Germany X-Mailer: Messenger v1.11 for RISC OS X-Posting-Agent: RISC OS Newsbase 0.59f X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO In message <199801181514.PAA04307@raistlin.armlinux.org>, Russell King wrote: > [...] > > This can be identified by a page fault / oops which looks like: > > pc: [<0xc4......>] > lr: [<0xc2......>] > ... > Process insmod .... > > This is because the ARM's b/bl instructions can only address +/- 16MB from > the current location. > > The latest insmod fixes this problem, and should be on the latest installation > disks on the FTP site. > > Can the people who are having problems with insmod please check that they are > using the latest disks, and the latest insmod package. If they are, then can > they urgently mail me. Just an observation from my install session: I had the above insmod crash only with the latest (2.0.31) kernel from the ftp site. With an earlier kernel (Oct 26th) insmod runs just fine. The crash also only happens when installing. Insmod doesn't crash when booting into multiuser mode. In this case it "only" fails because of unresolved symbols when using the latest kernel. And yes I used the latest disks. I downloaded them twice to be sure. > | | Russell King rmk@ecs.soton.ac.uk --- --- Regards, Stefan -- Stefan Voss (voss@yoda.in-berlin.de) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 18 22:15:22 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id WAA02034 for ; Sun, 18 Jan 1998 22:15:21 GMT Received: from vger.rutgers.edu ([128.6.190.2]:54128 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <17363-6684>; Mon, 19 Jan 1998 00:14:29 +0200 Received: by vger.rutgers.edu id <971028-241>; Sun, 18 Jan 1998 17:09:16 -0500 Received: from methan.in-berlin.de ([160.45.10.13]:1628 "HELO methan.in-berlin.de" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <971070-241>; Sun, 18 Jan 1998 16:58:25 -0500 Received: by methan.in-berlin.de (Smail3.2.0.98) from bolzen.in-berlin.de (194.94.235.1) with esmtp id ; Sun, 18 Jan 1998 23:02:26 +0100 (MET) Received: by bolzen.in-berlin.de (Smail3.2) from yoda with uucp id ; Sun, 18 Jan 1998 23:01:51 +0100 (MET) Received: by yoda.in-berlin.de (Smail3.1.28.1 #0) id m0xu2Bs-000000C; Sun, 18 Jan 98 20:23 GMT Date: Sun, 18 Jan 1998 21:11:06 +0100 From: Stefan Voss To: linux-arm@vger.rutgers.edu Subject: Re: Installation Problems Message-ID: <64a387a48%voss@yoda.in-berlin.de> In-Reply-To: <19980118115402Z971483-241+22019@vger.rutgers.edu> Organization: Individual Network (IN), Berlin, Germany X-Mailer: Messenger v1.11 for RISC OS X-Posting-Agent: RISC OS Newsbase 0.59f X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO In message <19980118115402Z971483-241+22019@vger.rutgers.edu>, Michael Howard wrote: > > Another problem occurrs when I try to start Linux by doubleclicking on > > !Linux. I used !LinConfig to set the kernel location to '*ADFS::4.0'. When > > I doubleclick on !Linux I get the error message: > > > > "invalid bmap magicin first sector of partition". > > I get exactly the same message but I'm using my 2nd IDE drive, 5. I found the solution to this problem in the FAQ. I had to - copy a kernel to /vmlinux - create /etc/boot.conf with an appropriate entry, see the boot.conf man page - Create the following link: ln -s /dev/hda3 /dev/root /dev/hda3 is my Linux root partition. If you are using drive 5 it should probably be "ln -s /dev/hdb3 /dev/root" for you. - loadmap -v After that I can boot into Linux just by doubleclicking !Liunx. > > [...] > > Cheers, Mick. > Regards, Stefan -- Stefan Voss (voss@yoda.in-berlin.de) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 18 22:31:31 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id WAA02111 for ; Sun, 18 Jan 1998 22:31:30 GMT Received: from vger.rutgers.edu ([128.6.190.2]:45067 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13121-6684>; Mon, 19 Jan 1998 00:29:55 +0200 Received: by vger.rutgers.edu id <970990-241>; Sun, 18 Jan 1998 17:19:38 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1400 "EHLO snowcrash.cymru.net" ident: "TIMEDOUT") by vger.rutgers.edu with ESMTP id <970966-241>; Sun, 18 Jan 1998 17:19:10 -0500 Received: from caramon.armlinux.org (dialup1.cymru.net [163.164.160.65]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id WAA17875 for ; Sun, 18 Jan 1998 22:22:40 GMT Received: (from rmk@localhost) by caramon.armlinux.org (8.7.4/8.7.3) id WAA19844 for linux-arm@vger.rutgers.edu; Sun, 18 Jan 1998 22:25:39 GMT From: Russell King Message-Id: <199801182225.WAA19844@caramon.armlinux.org> Subject: Release of Kernel source code for 2.0.33 and 2.1.78 To: linux-arm@vger.rutgers.edu Date: Sun, 18 Jan 1998 22:25:37 +0000 (GMT) X-Location: london.england.earth.mulky-way.universe Reply-To: rmk@ecs.soton.ac.uk X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi all! I have just finished uploading the source code for kernel versions 2.0.33 and 2.1.78. 2.0.33 has been tested on the EBSA, RiscPC and A5000. (Dave, can you try it on the ARC please?) 2.1.78 requires an ELF compiler and binutils. I will be releasing patches for binutils 2.8 and gcc 2.7.2.2 shortly for this purpose. A new copy of the modutils will be required, as well as a new floating point emulator module, also to be released shortly. Any bugs found with these should be emailed to (with all relevent info) rmk@ecs.soton.ac.uk. Any other questions/problems should be emailed to linux@arm.uk.linux.org or the mailing list. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sun Jan 18 22:51:22 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id WAA02311 for ; Sun, 18 Jan 1998 22:51:21 GMT Received: from vger.rutgers.edu ([128.6.190.2]:38435 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <16846-6684>; Mon, 19 Jan 1998 00:47:38 +0200 Received: by vger.rutgers.edu id <970889-241>; Sun, 18 Jan 1998 17:37:38 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:2056 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <970871-241>; Sun, 18 Jan 1998 17:31:31 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xu4DJ-0002SL-00; Sun, 18 Jan 1998 23:33:13 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Dave Gilbert Cc: linux-arm@vger.rutgers.edu Subject: New Binutils, New Danger In-reply-to: Your message of "Sun, 18 Jan 1998 17:42:02 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 18 Jan 1998 23:33:13 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > Its adr's in hand created code I think: Hi Dave. Can you try the appended patch to gas? p. Sun Jan 18 22:33:27 1998 Philip Blundell * config/tc-arm.c (md_apply_fix3): Cast offsets correctly to work on 64-bit machines. --- config/tc-arm.c~ Fri Dec 26 13:43:55 1997 +++ config/tc-arm.c Sun Jan 18 22:31:59 1998 @@ -5096,8 +5096,8 @@ /* If the instruction will fail, see if we can fix things up by changing the opcode. */ - if (newval == FAIL - && (newval = negate_data_op (&temp, value)) == FAIL) + if ((int)newval == FAIL + && (int)(newval = negate_data_op (&temp, value)) == FAIL) { as_bad_where (fixP->fx_file, fixP->fx_line, "invalid constant after fixup\n"); unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 00:20:36 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id AAA03092 for ; Mon, 19 Jan 1998 00:20:35 GMT Received: from vger.rutgers.edu ([128.6.190.2]:3452 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12994-18750>; Mon, 19 Jan 1998 02:18:35 +0200 Received: by vger.rutgers.edu id <970930-241>; Sun, 18 Jan 1998 19:13:34 -0500 Received: from dwmw2.robinson.cam.ac.uk ([131.111.217.153]:3882 "HELO imladris.demon.co.uk" ident: "exim") by vger.rutgers.edu with SMTP id <970934-241>; Sun, 18 Jan 1998 19:11:36 -0500 Received: from localhost (dwmw2.robinson.cam.ac.uk) [127.0.0.1] (dwmw2) by imladris.demon.co.uk with esmtp (Exim 1.82 #9) id 0xu4sO-0004dl-00; Mon, 19 Jan 1998 00:15:40 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: rmk@ecs.soton.ac.uk Cc: linux-arm@vger.rutgers.edu Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 19 Jan 1998 00:15:40 +0000 From: David Woodhouse Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO rmk@ecs.soton.ac.uk said: > I have just finished uploading the source code for kernel versions > 2.0.33 and 2.1.78. > 2.0.33 has been tested on the EBSA, RiscPC and A5000. (Dave, can you > try it on the ARC please?) On the A5000, cross-compiling using: jma24 /mnt/tmp/root/usr/src/linux # arm-unknown-linuxaout-gcc -v Reading specs from /usr/local/lib/gcc-lib/arm-unknown-linuxaout/2.7.2.2/specs gcc version 2.7.2.2 jma24 /mnt/tmp/root/usr/src/linux # arm-unknown-linuxaout-as -v GNU assembler version 2.7 (arm-unknown-linuxaout), using BFD version 2.7 I needed this patch. --- pgtable.h.orig Sun Jan 18 23:29:11 1998 +++ pgtable.h Sun Jan 18 23:33:22 1998 @@ -277,7 +277,7 @@ extern __inline__ pmd_t mk_pmd (pte_t *ptep) { pmd_t pmd; - pmd_val(pmd) = virt_to_phys((void *)ptep) | _PAGE_TABLE; + pmd_val(pmd) = (unsigned long)virt_to_phys((void *)ptep) | _PAGE_TABLE; return pmd; } However, I still get: arm-unknown-linuxaout-gcc -D__KERNEL__ -I/mnt/tmp/root/usr/src/linux/include -m3 -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -fno-strength-reduce -pipe -D__ASSEMBLY__ -c -o entry-armo.o entry-armo.S entry-armo.S: Assembler messages: entry-armo.S:388: Error: Register or shift expression expected -- statement `teqp pc,$0x08000003' entry-armo.S:406: Error: Register or shift expression expected -- statement `teqp pc,$0x00000003' make[1]: *** [entry-armo.o] Error 1 make[1]: Leaving directory `/mnt/tmp/root/usr/src/linux/arch/arm/kernel' make: *** [linuxsubdirs] Error 2 These are in the ENABLEIRQS and DISABLEIRQS macros from assembler.h, and I've absolutely no clue how to go about fixing it. Other than that: Is it possible to compile the FP-emulation into the kernel? Does anyone have a mkswap or swapon binary for ARM3? -- ---- ---- ---- David Woodhouse, Robinson College, CB3 9AN, England. (+44) 0976 658355 Dave@imladris.demon.co.uk http://dwmw2.robinson.cam.ac.uk finger pgp@dwmw2.robinson.cam.ac.uk for PGP key. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 08:06:03 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id IAA05806 for ; Mon, 19 Jan 1998 08:06:01 GMT Received: from vger.rutgers.edu ([128.6.190.2]:51486 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10445-24815>; Mon, 19 Jan 1998 10:04:57 +0200 Received: by vger.rutgers.edu id <970801-241>; Mon, 19 Jan 1998 03:00:01 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:1111 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <970791-241>; Mon, 19 Jan 1998 02:59:45 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xuD5e-0000C5-00; Mon, 19 Jan 1998 09:01:54 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: David Woodhouse cc: rmk@ecs.soton.ac.uk, linux-arm@vger.rutgers.edu Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 In-reply-to: Your message of "Mon, 19 Jan 1998 00:15:40 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 19 Jan 1998 09:01:53 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > Is it possible to compile the FP-emulation into the kernel? No, because as I understand it the ARM FPE is under a binary-only licence. This is a bit of an unhappy state of affairs, but until somebody writes us a better emulator we're rather stuck with it. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 08:06:20 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id IAA05809 for ; Mon, 19 Jan 1998 08:06:19 GMT Received: from vger.rutgers.edu ([128.6.190.2]:51486 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11019-6684>; Mon, 19 Jan 1998 10:05:05 +0200 Received: by vger.rutgers.edu id <970791-241>; Mon, 19 Jan 1998 03:00:05 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:1109 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <970799-241>; Mon, 19 Jan 1998 02:59:46 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xuD4Y-0000BH-00; Mon, 19 Jan 1998 09:00:46 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: David Woodhouse cc: rmk@ecs.soton.ac.uk, linux-arm@vger.rutgers.edu Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 In-reply-to: Your message of "Mon, 19 Jan 1998 00:15:40 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 19 Jan 1998 09:00:46 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >m3 -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -fno-strength-reduce -pi >pe -D__ASSEMBLY__ -c -o entry-armo.o entry-armo.S >entry-armo.S: Assembler messages: >entry-armo.S:388: Error: Register or shift expression expected -- statement `t >eqp pc,$0x08000003' > >entry-armo.S:406: Error: Register or shift expression expected -- statement `t >eqp pc,$0x00000003' Your binutils may be too old to grok `$'-style expressions. I know Russell tried to make this a config option but it looks like some slipped through. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 09:24:43 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id JAA06364 for ; Mon, 19 Jan 1998 09:24:42 GMT Received: from vger.rutgers.edu ([128.6.190.2]:22373 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10366-18750>; Mon, 19 Jan 1998 11:23:32 +0200 Received: by vger.rutgers.edu id <970791-241>; Mon, 19 Jan 1998 04:18:47 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1205 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970799-241>; Mon, 19 Jan 1998 04:18:26 -0500 Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id JAA27289; Mon, 19 Jan 1998 09:22:28 GMT From: Alan Cox Message-Id: <199801190922.JAA27289@snowcrash.cymru.net> Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 To: Philip.Blundell@pobox.com (Philip Blundell) Date: Mon, 19 Jan 1998 09:22:26 +0000 (GMT) Cc: Dave@imladris.demon.co.uk, rmk@ecs.soton.ac.uk, linux-arm@vger.rutgers.edu In-Reply-To: from "Philip Blundell" at Jan 19, 98 09:01:53 am Content-Type: text X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > No, because as I understand it the ARM FPE is under a binary-only licence. > This is a bit of an unhappy state of affairs, but until somebody writes us a > better emulator we're rather stuck with it. That shouldnt be very hard to do. The 2.1.x FPU emulation used on the ultrasparc is designed to give a good basic FPU framework for any box. Also writing an FPU emulator is generally not hard, but sometimes a little tedious. You can often start with emulate_fmul(float a, float b) { return a*b; } and compile all these routines with gcc -msoft-float ;) unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 10:14:33 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id KAA07102 for ; Mon, 19 Jan 1998 10:14:32 GMT Received: from vger.rutgers.edu ([128.6.190.2]:32292 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10895-6684>; Mon, 19 Jan 1998 12:12:50 +0200 Received: by vger.rutgers.edu id <970813-241>; Mon, 19 Jan 1998 05:04:22 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:26712 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <970801-241>; Mon, 19 Jan 1998 05:03:47 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id KAA06825; Mon, 19 Jan 1998 10:07:41 GMT From: Matthew Wilcox Message-Id: <199801191007.KAA06825@odie.barnet.ac.uk> Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 To: rmk@ecs.soton.ac.uk Date: Mon, 19 Jan 1998 10:07:40 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: <199801182225.WAA19844@caramon.armlinux.org> from "Russell King" at Jan 18, 98 10:25:37 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > 2.0.33 has been tested on the EBSA, RiscPC and A5000. (Dave, can you > try it on the ARC please?) I've got a gcc pre-2.8 snapshot (think it's december 25th) and binutils 2.8.1.0.15, so I set the CONFIG_NEW_BINUTILS option. For some reason in arch/arm/{mm,lib}/Makefile, the test against CONFIG (ifdef) doesn't work, and it tries to do the old thing. I replaced it with the test that's used in arch/arm/Makefile (ifeq y) which works. I didn't make clean (otherwise I wouldn't have been able to report this now..) but after fixing that, proc-arm6,7.S and proc-sa110.S fail to include the constants.h file - I changed this to #include "../lib/constants.h" After that it compiled fine. I haven't tested the resulting kernel, due to me bodging my startup files (oops). More reports later.. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 10:20:20 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id KAA07149 for ; Mon, 19 Jan 1998 10:20:19 GMT Received: from vger.rutgers.edu ([128.6.190.2]:9776 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11187-12619>; Mon, 19 Jan 1998 12:18:54 +0200 Received: by vger.rutgers.edu id <970812-241>; Mon, 19 Jan 1998 05:13:23 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:26742 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <970843-241>; Mon, 19 Jan 1998 05:10:54 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id KAA06944; Mon, 19 Jan 1998 10:13:28 GMT From: Matthew Wilcox Message-Id: <199801191013.KAA06944@odie.barnet.ac.uk> Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 To: alan@cymru.net (Alan Cox) Date: Mon, 19 Jan 1998 10:13:27 +0000 (GMT) Cc: Philip.Blundell@pobox.com, Dave@imladris.demon.co.uk, rmk@ecs.soton.ac.uk, linux-arm@vger.rutgers.edu In-Reply-To: <199801190922.JAA27289@snowcrash.cymru.net> from "Alan Cox" at Jan 19, 98 09:22:26 am X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > > > No, because as I understand it the ARM FPE is under a binary-only licence. > > This is a bit of an unhappy state of affairs, but until somebody writes us a > > better emulator we're rather stuck with it. > > That shouldnt be very hard to do. The 2.1.x FPU emulation used on the > ultrasparc is designed to give a good basic FPU framework for any box. Interesting.. As an aside, the current implementation of most of libm uses calls to the FPE to do much of the work. glibc2's maths routines work in fixed-point, so this would mean that any replacement FPE would not have to be optimised to the limit since it will be used less in the future. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 12:04:01 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id MAA08358 for ; Mon, 19 Jan 1998 12:04:00 GMT Received: from vger.rutgers.edu ([128.6.190.2]:9582 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13225-12619>; Mon, 19 Jan 1998 14:01:45 +0200 Received: by vger.rutgers.edu id <970821-241>; Mon, 19 Jan 1998 06:54:19 -0500 Received: from hermes.ex.ac.uk ([144.173.6.14]:31683 "EHLO exeter.ac.uk" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970809-241>; Mon, 19 Jan 1998 06:51:23 -0500 Received: from hebe [144.173.6.23] by hermes via SMTP (LAA13848); Mon, 19 Jan 1998 11:55:26 GMT Date: Mon, 19 Jan 1998 11:55:26 +0000 (GMT) From: Phil Norman X-Sender: py95pcfn@hebe To: linux-arm@vger.rutgers.edu Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO On Mon, 19 Jan 1998, Philip Blundell wrote: > > Is it possible to compile the FP-emulation into the kernel? > > No, because as I understand it the ARM FPE is under a binary-only licence. > This is a bit of an unhappy state of affairs, but until somebody writes us a > better emulator we're rather stuck with it. Wouldn't it be possible to release it as a gcc binary form (ie you have a file like this: { 0xe7 0x77 0x77 0x77 .... } (don't compile that one btw - it's an undefined instruction (useful for debugging)). I've seen quite a few i386-linux programs distributed with bits of source code like that. It makes for a fairly large source file, but it'd allow gcc to compile the thing into the linux kernel. Phil -- Phil Norman, mailing from Exeter Uni. Programs available for download from my web page. email: p.c.f.norman@ex.ac.uk web: http://newton.ex.ac.uk/general/ug/norman unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 12:56:54 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id MAA08805 for ; Mon, 19 Jan 1998 12:56:53 GMT Received: from vger.rutgers.edu ([128.6.190.2]:29216 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13816-18750>; Mon, 19 Jan 1998 14:51:03 +0200 Received: by vger.rutgers.edu id <970837-241>; Mon, 19 Jan 1998 07:44:08 -0500 Received: from lenzie.cent.gla.ac.uk ([130.209.34.18]:44885 "EHLO lenzie.cent.gla.ac.uk" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970813-241>; Mon, 19 Jan 1998 07:41:58 -0500 Received: from localhost (9606585c@localhost sender 9606585c) by lenzie.cent.gla.ac.uk (8.8.4/UK-2.2a/cent-sparc) with SMTP id MAA22068; Mon, 19 Jan 1998 12:45:44 GMT Date: Mon, 19 Jan 1998 12:45:43 +0000 (GMT) From: James Craig <9606585c@udcf.gla.ac.uk> X-Sender: 9606585c@lenzie.cent.gla.ac.uk To: Phil Norman cc: linux-arm@vger.rutgers.edu Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO On Mon, 19 Jan 1998, Phil Norman wrote: > On Mon, 19 Jan 1998, Philip Blundell wrote: > > > > Is it possible to compile the FP-emulation into the kernel? > > > > No, because as I understand it the ARM FPE is under a binary-only licence. > > This is a bit of an unhappy state of affairs, but until somebody writes us a > > better emulator we're rather stuck with it. > > Wouldn't it be possible to release it as a gcc binary form (ie you have a > file like this: > { 0xe7 0x77 0x77 0x77 .... } > > (don't compile that one btw - it's an undefined instruction (useful for > debugging)). I've seen quite a few i386-linux programs distributed with > bits of source code like that. It makes for a fairly large source file, > but it'd allow gcc to compile the thing into the linux kernel. There's a few similar bits like that in the alpha kernel for much the same job; It's not a big deal. The thing is though, why bother when it means you've got another bit of code in the kernel that's just as good modular (since it's probably not needed for *everything* - for example, embedded ARM systems might not need FP at all). I think it might be an idea for all the linux kernel stuff to be modular, and include a semi-decent way of loading modules up from within a kernel image (this goes for i386 too). initrd has the right idea, but it's not ideal; We just want modules loaded up by the kernel at boot time - using a whole FS and RAM disk etc. for that is overkill. The other point is that if it worked that way (I guess I'm suggesting something a bit like the RISC OS ROM here...) then we could rmmod modules that were included with the kernel at boot-time and load up a replacement. The best of both worlds - everything loaded at boot-time without needing an FS, *and* modularity. :) -- James Craig <9606585c@udcf.gla.ac.uk> unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 13:19:35 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id NAA08998 for ; Mon, 19 Jan 1998 13:19:34 GMT Received: from vger.rutgers.edu ([128.6.190.2]:9281 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13391-18750>; Mon, 19 Jan 1998 15:18:21 +0200 Received: by vger.rutgers.edu id <970807-241>; Mon, 19 Jan 1998 08:12:51 -0500 Received: from monarch.ja.net ([128.86.16.37]:59480 "EHLO monarch.ja.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970799-241>; Mon, 19 Jan 1998 08:12:39 -0500 Received: from sun3.nsfnet-relay.ac.uk by monarch.ja.net with SPAM SMTP (PP); Mon, 19 Jan 1998 13:02:42 +0000 Received: from bright.ecs.soton.ac.uk by sun3.nsfnet-relay.ac.uk with NRS SMTP (PP); Mon, 19 Jan 1998 13:02:36 +0000 Received: from diana.ecs.soton.ac.uk by bright.ecs.soton.ac.uk; Mon, 19 Jan 98 13:00:05 GMT From: Russ King Message-Id: <22056.9801191301@diana.ecs.soton.ac.uk> Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 To: Dave@imladris.demon.co.uk (David Woodhouse) Date: Mon, 19 Jan 1998 13:01:51 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: from "David Woodhouse" at Jan 19, 98 00:15:40 am X-Mailer: ELM [version 2.4 PL24] Content-Type: text X-Orcpt: rfc822;linux-arm@vger.rutgers.edu X-Envid: [/PRMD=UK.AC/ADMD= /C=GB/;<22056.9801191301@diana.ecs.soto] Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO David Woodhouse writes: > --- pgtable.h.orig Sun Jan 18 23:29:11 1998 > +++ pgtable.h Sun Jan 18 23:33:22 1998 > @@ -277,7 +277,7 @@ > extern __inline__ pmd_t mk_pmd (pte_t *ptep) > { > pmd_t pmd; > - pmd_val(pmd) = virt_to_phys((void *)ptep) | _PAGE_TABLE; > + pmd_val(pmd) = (unsigned long)virt_to_phys((void *)ptep) | _PAGE_TABLE; > return pmd; > } Change this to __virt_to_phys - it returns an unsigned long from an unsigned long. > arm-unknown-linuxaout-gcc -D__KERNEL__ -I/mnt/tmp/root/usr/src/linux/include -m3 -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -fno-strength-reduce -pipe -D__ASSEMBLY__ -c -o entry-armo.o entry-armo.S > entry-armo.S: Assembler messages: > entry-armo.S:388: Error: Register or shift expression expected -- statement `teqp pc,$0x08000003' > > entry-armo.S:406: Error: Register or shift expression expected -- statement `teqp pc,$0x00000003' > > make[1]: *** [entry-armo.o] Error 1 > make[1]: Leaving directory `/mnt/tmp/root/usr/src/linux/arch/arm/kernel' > make: *** [linuxsubdirs] Error 2 Which version of the binutils are you using? > Other than that: > Is it possible to compile the FP-emulation into the kernel? Definitely not. The FPEmulator is not part of the kernel since it is a copy of the Acorn FP Emulator module in the RiscOS PRMs (modified under license). > Does anyone have a mkswap or swapon binary for ARM3? The standard ones work. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 13:20:31 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id NAA09006 for ; Mon, 19 Jan 1998 13:20:30 GMT Received: from vger.rutgers.edu ([128.6.190.2]:9281 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <14240-6684>; Mon, 19 Jan 1998 15:18:55 +0200 Received: by vger.rutgers.edu id <970821-241>; Mon, 19 Jan 1998 08:13:51 -0500 Received: from nexusel.demon.co.uk ([158.152.30.195]:12542 "HELO globe" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970812-241>; Mon, 19 Jan 1998 08:13:40 -0500 Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xuH3K-0004ZH-00; Mon, 19 Jan 1998 13:15:46 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xuH3D-0003j9-00; Mon, 19 Jan 1998 13:15:39 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Phil Norman cc: linux-arm@vger.rutgers.edu Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 In-reply-to: Your message of "Mon, 19 Jan 1998 11:55:26 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 19 Jan 1998 13:15:38 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >Wouldn't it be possible to release it as a gcc binary form (ie you have a >file like this: >{ 0xe7 0x77 0x77 0x77 .... } > >(don't compile that one btw - it's an undefined instruction (useful for >debugging)). I've seen quite a few i386-linux programs distributed with >bits of source code like that. It makes for a fairly large source file, >but it'd allow gcc to compile the thing into the linux kernel. No; I think you run into trouble with the GPL if you do that. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 13:21:25 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id NAA09012 for ; Mon, 19 Jan 1998 13:21:24 GMT Received: from vger.rutgers.edu ([128.6.190.2]:9281 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13393-12619>; Mon, 19 Jan 1998 15:19:05 +0200 Received: by vger.rutgers.edu id <970812-241>; Mon, 19 Jan 1998 08:13:54 -0500 Received: from nexusel.demon.co.uk ([158.152.30.195]:12539 "HELO globe" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970799-241>; Mon, 19 Jan 1998 08:13:33 -0500 Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xuH4c-0004ZP-00; Mon, 19 Jan 1998 13:17:06 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xuH4U-0003k5-00; Mon, 19 Jan 1998 13:16:58 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: James Craig <9606585c@udcf.gla.ac.uk> cc: Phil Norman , linux-arm@vger.rutgers.edu Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 In-reply-to: Your message of "Mon, 19 Jan 1998 12:45:43 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 19 Jan 1998 13:16:58 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >There's a few similar bits like that in the alpha kernel for much the same >job; It's not a big deal. The thing is though, why bother when it means >you've got another bit of code in the kernel that's just as good modular >(since it's probably not needed for *everything* - for example, embedded The same code takes up less space in the kernel than as a module - quite noticeably so on ARM2/ARM3 systems. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 13:34:54 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id NAA09091 for ; Mon, 19 Jan 1998 13:34:53 GMT Received: from vger.rutgers.edu ([128.6.190.2]:47440 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13818-18750>; Mon, 19 Jan 1998 15:33:12 +0200 Received: by vger.rutgers.edu id <970806-241>; Mon, 19 Jan 1998 08:19:57 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1701 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970843-241>; Mon, 19 Jan 1998 08:16:48 -0500 Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id NAA31643; Mon, 19 Jan 1998 13:21:00 GMT From: Alan Cox Message-Id: <199801191321.NAA31643@snowcrash.cymru.net> Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 To: P.C.F.Norman@exeter.ac.uk (Phil Norman) Date: Mon, 19 Jan 1998 13:20:57 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: from "Phil Norman" at Jan 19, 98 11:55:26 am Content-Type: text X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > debugging)). I've seen quite a few i386-linux programs distributed with > bits of source code like that. It makes for a fairly large source file, > but it'd allow gcc to compile the thing into the linux kernel. The license intentionally protects Linux from such misuse unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 13:54:56 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id NAA09280 for ; Mon, 19 Jan 1998 13:54:53 GMT Received: from vger.rutgers.edu ([128.6.190.2]:37224 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12027-6684>; Mon, 19 Jan 1998 15:50:30 +0200 Received: by vger.rutgers.edu id <970856-241>; Mon, 19 Jan 1998 08:42:10 -0500 Received: from cptsu5.univ-mrs.fr ([139.124.7.5]:4370 "EHLO cptsu5.univ-mrs.fr" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970843-241>; Mon, 19 Jan 1998 08:41:07 -0500 Received: from cptsu1.univ-mrs.fr (cptsu1.univ-mrs.fr [139.124.7.1]) by cptsu5.univ-mrs.fr (8.8.8/jtpda-5.2.9.1) with ESMTP id OAA17983 for ; Mon, 19 Jan 1998 14:45:54 +0100 (MET) Received: by cptsu1.univ-mrs.fr (8.7.6/jtpda-5.0) id OAA00336 for linux-arm@vger.rutgers.edu; Mon, 19 Jan 1998 14:38:49 GMT Date: Mon, 19 Jan 1998 14:38:49 GMT From: penne@cptsu5.univ-mrs.fr Message-Id: <199801191438.OAA00336@cptsu1.univ-mrs.fr> To: linux-arm@vger.rutgers.edu Subject: installation problem X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hello there, I am a futur user of arm-linux. I got a problem at the moment during the installation : I planed to install linux from a hard drive prepared on a PC. So I put the RedHat distribution and used the installer. All went OK up to a point, just after the install programm asked me if I wanted to configure LAN (and I replied no), then it told me that it could not find any kernel. I don't know at all WHERE am I ssupposed to put the kernel in the install tree, and it seems to be specified nowhere ... Please help ! Vincent Penne. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From pb@nexus.co.uk Mon Jan 19 15:39:54 1998 Return-Path: Received: from globe (nexusel.demon.co.uk [158.152.30.195]) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id PAA10003 for ; Mon, 19 Jan 1998 15:39:52 GMT Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xuJCO-0004oZ-00; Mon, 19 Jan 1998 15:33:16 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xuJCG-0004pH-00; Mon, 19 Jan 1998 15:33:08 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Matthew Wilcox Cc: linux-arm@vger.rutgers.edu Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 In-reply-to: Your message of "Mon, 19 Jan 1998 10:07:40 GMT." <199801191007.KAA06825@odie.barnet.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 19 Jan 1998 15:33:06 +0000 From: Philip Blundell Message-Id: Status: RO >I've got a gcc pre-2.8 snapshot (think it's december 25th) and binutils BTW, I sent the config.guess patch to the gcc2 people just now. I've appended a copy of it below in case anybody else is having trouble with gcc-2.8.0 not correctly detecting their system type. p. Mon Jan 19 15:30:12 1998 Philip Blundell * config.guess: Add support for Linux/ARM. Index: config.guess =================================================================== RCS file: /usr/repository/gnu/gcc/config.guess,v retrieving revision 1.3 diff -u -r1.3 config.guess --- config.guess 1997/11/17 11:57:41 1.3 +++ config.guess 1997/12/01 14:17:11 @@ -493,6 +493,12 @@ echo `echo ${UNAME_MACHINE}|sed -e 's,[-/].*$,,'`-unknown-gnu`echo ${UNAME_RELEASE}|sed -e 's,/.*$,,'` exit 0 ;; *:Linux:*:*) + # uname on the ARM produces all sorts of strangeness, and we need to + # filter it out. + case "$UNAME_MACHINE" in + arm* | sa110*) UNAME_MACHINE="arm" ;; + esac + # The BFD linker knows what the default object file format is, so # first see if it will tell us. ld_help_string=`ld --help 2>&1` @@ -506,6 +512,7 @@ i?86linux) echo "${UNAME_MACHINE}-pc-linux-gnuaout" ; exit 0 ;; i?86coff) echo "${UNAME_MACHINE}-pc-linux-gnucoff" ; exit 0 ;; sparclinux) echo "${UNAME_MACHINE}-unknown-linux-gnuaout" ; exit 0 ;; + armlinux) echo "${UNAME_MACHINE}-unknown-linux-gnuaout" ; exit 0 ;; m68klinux) echo "${UNAME_MACHINE}-unknown-linux-gnuaout" ; exit 0 ;; elf32ppc) echo "powerpc-unknown-linux-gnu" ; exit 0 ;; esac From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 18:24:48 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id SAA11233 for ; Mon, 19 Jan 1998 18:24:47 GMT Received: from vger.rutgers.edu ([128.6.190.2]:31292 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10492-6684>; Mon, 19 Jan 1998 20:21:24 +0200 Received: by vger.rutgers.edu id <970875-11737>; Mon, 19 Jan 1998 13:08:20 -0500 Received: from nexusel.demon.co.uk ([158.152.30.195]:14543 "HELO globe" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <970862-11737>; Mon, 19 Jan 1998 13:01:45 -0500 Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xuLOc-00055s-00; Mon, 19 Jan 1998 17:54:02 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xuLOU-0007ER-00; Mon, 19 Jan 1998 17:53:54 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: rmk@ecs.soton.ac.uk Cc: linux-arm@vger.rutgers.edu Subject: parport changes Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 19 Jan 1998 17:53:54 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Hi Russell. Can you explain why the following patches are needed? If they're necessary for parport to work on ARM then it would be nice to have them sent to Tim Waugh, who maintains it at the moment p. diff -urN /mnt/linux/drivers/misc/parport_pc.c linux/drivers/misc/parport_pc.c --- /mnt/linux/drivers/misc/parport_pc.c Sat Jan 17 20:08:30 1998 +++ linux/drivers/misc/parport_pc.c Sat Dec 27 23:39:24 1997 @@ -11,10 +11,6 @@ #include #include -#include -#include -#include - #include #include #include @@ -23,6 +19,10 @@ #include #include #include + +#include +#include +#include #include diff -urN /mnt/linux/drivers/misc/parport_procfs.c linux/drivers/misc/parport_procfs.c --- /mnt/linux/drivers/misc/parport_procfs.c Sun Dec 7 00:18:11 1997 +++ linux/drivers/misc/parport_procfs.c Sat Dec 27 23:41:13 1997 @@ -10,9 +10,6 @@ #include #include #include -#include -#include - #include #include #include @@ -21,6 +18,9 @@ #include #include + +#include +#include #include unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 18:52:08 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id SAA11494 for ; Mon, 19 Jan 1998 18:52:07 GMT Received: from vger.rutgers.edu ([128.6.190.2]:32870 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10493-6684>; Mon, 19 Jan 1998 20:49:01 +0200 Received: by vger.rutgers.edu id <970799-11737>; Mon, 19 Jan 1998 13:42:27 -0500 Received: from odie.barnet.ac.uk ([194.82.202.98]:31009 "EHLO odie.barnet.ac.uk" ident: "willy") by vger.rutgers.edu with ESMTP id <970856-11737>; Mon, 19 Jan 1998 13:39:18 -0500 Received: (from willy@localhost) by odie.barnet.ac.uk (8.8.6/8.8.6) id SAA11328; Mon, 19 Jan 1998 18:43:22 GMT From: Matthew Wilcox Message-Id: <199801191843.SAA11328@odie.barnet.ac.uk> Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 To: rmk@ecs.soton.ac.uk Date: Mon, 19 Jan 1998 18:43:20 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: <199801182225.WAA19844@caramon.armlinux.org> from "Russell King" at Jan 18, 98 10:25:37 pm X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > I have just finished uploading the source code for kernel versions 2.0.33 > and 2.1.78. I compiled ether3 and etherh as modules. EtherH works no better with my 500 than it did under 2.0.31: RX packets 0 errors 0 dropped 4318 overruns 0 TX packets 9 errors 0 dropped 0 overruns 0 IRQ 33 Base addr 0x1000 Memory 80001200-d4 Every time I call ifconfig, the number of dropped packets reported increments by 162, irrespective of how much time I allow to elapse. But Ether3 is broken: loading device `eth1'... eth1: ether3 found at 80000000, IRQ 32 ether address 00:02:07:00:dd:a0 eth1: RAM failed with (00 instead of 5A) at 0x0001-0x0002 eth1: RAM failed with (00 instead of 5A) at 0x0003-0x0004 eth1: RAM failed with (00 instead of 5A) at 0x0005-0x0006 eth1: RAM failed with (00 instead of 5A) at 0x0007-0x0008 eth1: RAM failed with (00 instead of 5A) at 0x0009-0x000A Initialisation of ether3 failed. The fact that it's reporting 2-byte values makes me suspicious this may be a short-load-bytes problem.. I'll try the released gcc 2.8.0 later. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From rmk@ecs.soton.ac.uk Mon Jan 19 22:54:54 1998 Return-Path: Received: from snowcrash.cymru.net (snowcrash.cymru.net [163.164.160.3]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id WAA13364 for ; Mon, 19 Jan 1998 22:54:53 GMT Received: from caramon.armlinux.org (dialup1.cymru.net [163.164.160.65]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id WAA09456 for ; Mon, 19 Jan 1998 22:55:02 GMT Received: from raistlin.armlinux.org (raistlin [192.168.0.3]) by caramon.armlinux.org (8.7.4/8.7.3) with ESMTP id WAA22926 for ; Mon, 19 Jan 1998 22:35:05 GMT From: Russell King Received: (from rmk@localhost) by raistlin.armlinux.org (8.7.4/8.7.3) id WAA00792 for willy@odie.barnet.ac.uk; Mon, 19 Jan 1998 22:34:46 GMT Message-Id: <199801192234.WAA00792@raistlin.armlinux.org> Subject: Re: Release of Kernel source code for 2.0.33 and 2.1.78 To: willy@odie.barnet.ac.uk (Matthew Wilcox) Date: Mon, 19 Jan 1998 22:34:46 +0000 (GMT) In-Reply-To: <199801191843.SAA11328@odie.barnet.ac.uk> from "Matthew Wilcox" at Jan 19, 98 06:43:20 pm X-Location: london.england.earth.mulky-way.universe Reply-To: rmk@ecs.soton.ac.uk X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Status: RO Matthew Wilcox writes: > Every time I call ifconfig, the number of dropped packets reported > increments by 162, irrespective of how much time I allow to elapse. I think that this is a hardware bug in the NE2000 clone - the data sheet for a NE2000 clone (and the 8390.c driver) both say that it's supposed to clear the counter registers on read, however the etherh cards do not. > loading device `eth1'... > eth1: ether3 found at 80000000, IRQ 32 ether address 00:02:07:00:dd:a0 > eth1: RAM failed with (00 instead of 5A) at 0x0001-0x0002 > eth1: RAM failed with (00 instead of 5A) at 0x0003-0x0004 > eth1: RAM failed with (00 instead of 5A) at 0x0005-0x0006 > eth1: RAM failed with (00 instead of 5A) at 0x0007-0x0008 > eth1: RAM failed with (00 instead of 5A) at 0x0009-0x000A > Initialisation of ether3 failed. I saw this with 2.1.78 and thought that it was a problem with that until I found it - the io.h file is broken ;( I'm amazed that it didn't cause more problems. In include/asm/arch-{arc,a5k}/io.h, there is #defined __out*. The problem is that the __outw expansion does not shift the value into the upper word as well as the lower. I'll get some new patches out tomorrow to fix this (hopefully during the day if poss). _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 23:18:42 1998 Return-Path: Received: from nic.funet.fi ([128.214.248.6] (may be forged)) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id XAA13543 for ; Mon, 19 Jan 1998 23:18:40 GMT Received: from vger.rutgers.edu ([128.6.190.2]:1664 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <9874-12619>; Tue, 20 Jan 1998 01:16:33 +0200 Received: by vger.rutgers.edu id <970837-11737>; Mon, 19 Jan 1998 18:09:10 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1566 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970894-11737>; Mon, 19 Jan 1998 17:51:10 -0500 Received: from caramon.armlinux.org (dialup1.cymru.net [163.164.160.65]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id WAA09469; Mon, 19 Jan 1998 22:55:16 GMT Received: from raistlin.armlinux.org (raistlin [192.168.0.3]) by caramon.armlinux.org (8.7.4/8.7.3) with ESMTP id WAA22905; Mon, 19 Jan 1998 22:26:56 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.armlinux.org (8.7.4/8.7.3) id VAA00366; Mon, 19 Jan 1998 21:22:06 GMT Message-Id: <199801192122.VAA00366@raistlin.armlinux.org> Subject: Re: installation problem To: penne@cptsu5.univ-mrs.fr Date: Mon, 19 Jan 1998 21:22:05 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: <199801191438.OAA00336@cptsu1.univ-mrs.fr> from "penne@cpt.univ-mrs.fr" at Jan 19, 98 02:38:49 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO penne@cpt.univ-mrs.fr writes: >... > configure LAN (and I replied no), then it told me that it could not find > any kernel. I don't know at all WHERE am I ssupposed to put the kernel in > the install tree, and it seems to be specified nowhere ... Please help ! The answer to your problem can be found in question 6 on http://www.arm.uk.linux.org/armlinux/faq.html _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Mon Jan 19 23:27:52 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id XAA13569 for ; Mon, 19 Jan 1998 23:27:51 GMT Received: from vger.rutgers.edu ([128.6.190.2]:48401 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <9268-12619>; Tue, 20 Jan 1998 01:26:10 +0200 Received: by vger.rutgers.edu id <970809-11737>; Mon, 19 Jan 1998 18:21:02 -0500 Received: from mail1.digital.com ([204.123.2.50]:2205 "EHLO mail1.digital.com" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970839-11737>; Mon, 19 Jan 1998 18:18:17 -0500 Received: from pobox1.pa.dec.com (pobox1.pa.dec.com [16.1.240.19]) by mail1.digital.com (8.8.8/8.8.8/WV1.0c) with SMTP id PAA09064; Mon, 19 Jan 1998 15:22:31 -0800 (PST) Received: from stowe.pa.dec.com by pobox1.pa.dec.com (5.65v3.2/1.1.10.5/07Nov97-1157AM) id AA08071; Mon, 19 Jan 1998 15:22:30 -0800 Received: by stowe.pa.dec.com; (5.65v3.2/1.1.8.2/06Jun96-0357PM) id AA22151; Mon, 19 Jan 1998 15:22:18 -0800 Date: Mon, 19 Jan 1998 15:22:18 -0800 Message-Id: <9801192322.AA22151@stowe.pa.dec.com> From: Deborah Wallach To: rusling@linux.reo.dec.com Cc: charlese@cvs.com.au, linux-arm@vger.rutgers.edu, brakmo@pa.dec.com, caw@pa.dec.com In-Reply-To: <199801161438.OAA01011@linux.reo.dec.com> (rusling@linux.reo.dec.com) Subject: SA1100 linux [was Re: FORTH] Address: Digital Equipment Corporation; Western Research Laboratory; 250 University Avenue; Palo Alto, CA 94301; USA Phone: 650-617-3315 X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO We've been working on a Linux port to the SA1100 (specifically, to Brutus, the SA1100 evaluation board). We have a preliminary version up and running right now (console, keyboard, serial ports only), based on Russell's August 2.0.30 distribution. The source code still needs cleaning, and we're nowhere near ready for distribution yet, but if anyone else is interested in developing code for Brutus and wants to start with working code, please contact us. Deborah Wallach, Carl Waldspurger, Larry Brakmo {kerr, caw, brakmo}@pa.dec.com Western Research Laboratory Digital Equipment Corporation unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From pb@nexus.co.uk Mon Jan 26 14:45:05 1998 Return-Path: Received: from globe (nexusel.demon.co.uk [158.152.30.195]) by odie.barnet.ac.uk (8.8.6/8.8.6) with SMTP id OAA25524 for ; Mon, 26 Jan 1998 14:45:03 GMT Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xwpa4-0005bM-00; Mon, 26 Jan 1998 14:32:08 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xwpa3-0003TR-00; Mon, 26 Jan 1998 14:32:07 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Matthew Wilcox cc: rusling@linux.reo.dec.com (Ka'Plaagh), linux-arm@vger.rutgers.edu Subject: Re: Building a toolset for ARM/Linux In-reply-to: Your message of "Mon, 26 Jan 1998 13:29:12 GMT." <199801261329.NAA24931@odie.barnet.ac.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 26 Jan 1998 14:32:06 +0000 From: Philip Blundell Message-Id: Status: RO >Get gcc 2.8.0 and build its libgcc, then slot that into the 2.7.2.3 instead That _may_ work, but it's not a good idea. The code in libgcc is quite tightly tied to the compiler, and it's not a good idea to mix and match. The way to get libgcc built by default is to add the following lines to config/arm/t-linux if they're not already there: # Since libgcc1 is an assembler file, we can build it automatically for the # cross-compiler. CROSS_LIBGCC1 = libgcc1-asm.a LIBGCC1 = libgcc1-asm.a p. From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 27 14:13:20 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id OAA12115 for ; Tue, 27 Jan 1998 14:13:17 GMT Received: from vger.rutgers.edu ([128.6.190.2]:46105 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13571-621>; Tue, 27 Jan 1998 16:00:57 +0200 Received: by vger.rutgers.edu id <970933-257>; Tue, 27 Jan 1998 08:53:33 -0500 Received: from out5.ibm.net ([165.87.194.245]:1452 "EHLO out5.ibm.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970910-257>; Tue, 27 Jan 1998 08:44:27 -0500 Received: from cvs.com.au (slip-129-37-30-84.fl.us.ibm.net [129.37.30.84]) by out5.ibm.net (8.8.5/8.6.9) with ESMTP id NAA108294 for ; Tue, 27 Jan 1998 13:46:15 GMT Message-ID: <34CDE5B3.7B17DEE4@cvs.com.au> Date: Tue, 27 Jan 1998 08:48:35 -0500 From: Charles Esson Reply-To: charlese@cvs.com.au Organization: Colour Vision Systems X-Mailer: Mozilla 4.04 [en] (WinNT; I) MIME-Version: 1.0 To: "linux-arm@vger.rutgers.edu" Subject: Crystal Ball Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Digital sells semiconductor division. Netscape make source available to net. Digital sell to Compact. We are all working on the wrong products, crystal balls, I tell you thats a product this industry needs. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 27 16:35:38 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id QAA13166 for ; Tue, 27 Jan 1998 16:35:35 GMT Received: from vger.rutgers.edu ([128.6.190.2]:60720 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <11091-619>; Tue, 27 Jan 1998 18:32:49 +0200 Received: by vger.rutgers.edu id <971029-257>; Tue, 27 Jan 1998 11:25:14 -0500 Received: from hermes.ex.ac.uk ([144.173.6.14]:7696 "EHLO exeter.ac.uk" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971031-257>; Tue, 27 Jan 1998 11:15:35 -0500 Received: from hebe [144.173.6.23] by hermes via SMTP (QAA13271); Tue, 27 Jan 1998 16:17:04 GMT Date: Tue, 27 Jan 1998 16:17:04 +0000 (GMT) From: Phil Norman X-Sender: py95pcfn@hebe To: ARM Linux Subject: Little X problem Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO I would use Xwindows a lot more than I do - tis a nice thing. However, I'm having big problems switching to a virtual terminal when in X. I've already tried all combinations of Ctrl/Shift/Alt and F-keys and nothing seems to work. I've seen it done on an i386-linux box using Ctrl-Fsomething (or Shift-Ctrl-Fsomething - not sure), but that was probably using XFree86 which is a different version. Anyone know how I can switch VTs under the armlinux Xwindows? Thanks in advance, Phil P.S. Now exams are over I'm going to do a bit of work on IscaFS. I'm hoping to get v0.02 out within a week or so which will hopefully have all the bugs fixed and some more graceful handling of symbolic links. -- Phil Norman, mailing from Exeter Uni. Programs available for download from my web page. email: p.c.f.norman@ex.ac.uk web: http://newton.ex.ac.uk/general/ug/norman unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 27 17:25:25 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id RAA13545 for ; Tue, 27 Jan 1998 17:25:24 GMT Received: from vger.rutgers.edu ([128.6.190.2]:48733 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <10869-617>; Tue, 27 Jan 1998 19:23:41 +0200 Received: by vger.rutgers.edu id <971138-257>; Tue, 27 Jan 1998 12:12:23 -0500 Received: from puce.csi.cam.ac.uk ([131.111.8.40]:53560 "HELO puce.csi.cam.ac.uk" ident: "exim") by vger.rutgers.edu with SMTP id <971156-257>; Tue, 27 Jan 1998 11:58:26 -0500 Received: from jma24 by puce.csi.cam.ac.uk with smtp (Exim 1.73 #1) id 0xxEN3-0003yk-00; Tue, 27 Jan 1998 17:00:21 +0000 Date: Tue, 27 Jan 1998 17:00:21 +0000 (GMT) From: "J.M. Appleby" X-Sender: jma24@puce.csi.cam.ac.uk To: Phil Norman cc: ARM Linux Subject: Re: Little X problem In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO On Tue, 27 Jan 1998, Phil Norman wrote: > I would use Xwindows a lot more than I do - tis a nice thing. However, > I'm having big problems switching to a virtual terminal when in X. I've > already tried all combinations of Ctrl/Shift/Alt and F-keys and nothing > seems to work. I've seen it done on an i386-linux box using > Ctrl-Fsomething (or Shift-Ctrl-Fsomething - not sure), but that was > probably using XFree86 which is a different version. Anyone know how I > can switch VTs under the armlinux Xwindows? Have you tried chvt? Usually does the trick... J unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 27 23:47:15 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id XAA16494 for ; Tue, 27 Jan 1998 23:47:14 GMT Received: from vger.rutgers.edu ([128.6.190.2]:59203 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13303-617>; Wed, 28 Jan 1998 01:46:35 +0200 Received: by vger.rutgers.edu id <971043-257>; Tue, 27 Jan 1998 18:17:20 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1519 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971592-257>; Tue, 27 Jan 1998 17:53:27 -0500 Received: from caramon.armlinux.org (dialup1.cymru.net [163.164.160.65]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id WAA24639; Tue, 27 Jan 1998 22:54:33 GMT Received: from raistlin.armlinux.org (raistlin [192.168.0.3]) by caramon.armlinux.org (8.7.4/8.7.3) with ESMTP id UAA06126; Tue, 27 Jan 1998 20:21:34 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.armlinux.org (8.7.4/8.7.3) id UAA00366; Tue, 27 Jan 1998 20:21:23 GMT Message-Id: <199801272021.UAA00366@raistlin.armlinux.org> Subject: Re: Crystal Ball To: charlese@cvs.com.au Date: Tue, 27 Jan 1998 20:21:23 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: <34CDE5B3.7B17DEE4@cvs.com.au> from "Charles Esson" at Jan 27, 98 08:48:35 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Charles Esson writes: DEC sues Intel. > Digital sells semiconductor division. to Intel in return for a bank... > Netscape make source available to net. Not yet they haven't. > Digital sell to Compact. Erm, yes, well. > We are all working on the wrong products, crystal balls, I tell you > thats a product this industry needs. Crystal balls? Why? What's wrong with a time machine? _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Tue Jan 27 23:59:42 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id XAA16507 for ; Tue, 27 Jan 1998 23:59:41 GMT Received: from vger.rutgers.edu ([128.6.190.2]:32337 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <6370-619>; Wed, 28 Jan 1998 01:57:59 +0200 Received: by vger.rutgers.edu id <971397-257>; Tue, 27 Jan 1998 18:17:45 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1513 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971296-257>; Tue, 27 Jan 1998 17:53:31 -0500 Received: from caramon.armlinux.org (dialup1.cymru.net [163.164.160.65]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id WAA24658; Tue, 27 Jan 1998 22:55:14 GMT Received: from raistlin.armlinux.org (raistlin [192.168.0.3]) by caramon.armlinux.org (8.7.4/8.7.3) with ESMTP id UAA06119; Tue, 27 Jan 1998 20:19:12 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.armlinux.org (8.7.4/8.7.3) id UAA00353; Tue, 27 Jan 1998 20:19:01 GMT Message-Id: <199801272019.UAA00353@raistlin.armlinux.org> Subject: Re: Little X problem To: jma24@hermes.cam.ac.uk (J.M. Appleby) Date: Tue, 27 Jan 1998 20:18:59 +0000 (GMT) Cc: P.C.F.Norman@exeter.ac.uk, linux-arm@vger.rutgers.edu In-Reply-To: from "J.M. Appleby" at Jan 27, 98 05:00:21 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO J.M. Appleby writes: > > I would use Xwindows a lot more than I do - tis a nice thing. However, > > I'm having big problems switching to a virtual terminal when in X. I've > > already tried all combinations of Ctrl/Shift/Alt and F-keys and nothing > > seems to work. I've seen it done on an i386-linux box using > > Ctrl-Fsomething (or Shift-Ctrl-Fsomething - not sure), but that was > > probably using XFree86 which is a different version. Anyone know how I > > can switch VTs under the armlinux Xwindows? > > Have you tried chvt? Usually does the trick... Xwindows != XFree86. XFree86 has VT switching support. Xwindows does not. Attempting to force a VT switch will be very stupid as Xwindows does not support the necessary functions, and will corrupt the screen. Your best bet is to wait until the integration of ARM Linux with the main kernel source tree - chances are that the whole ARM X interface may well change as a result of several people's work. Hence to remove yet-another- X-port which does something in a different way I don't think that it would be a good idea to think about XF86 yet. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Wed Jan 28 01:52:39 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id BAA17257 for ; Wed, 28 Jan 1998 01:52:38 GMT Received: from vger.rutgers.edu ([128.6.190.2]:45126 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13558-620>; Wed, 28 Jan 1998 03:50:03 +0200 Received: by vger.rutgers.edu id <971435-232>; Tue, 27 Jan 1998 19:45:00 -0500 Received: from dwmw2.robinson.cam.ac.uk ([131.111.217.153]:1425 "HELO imladris.demon.co.uk" ident: "exim") by vger.rutgers.edu with SMTP id <971584-257>; Tue, 27 Jan 1998 19:19:49 -0500 Received: from localhost (dwmw2.robinson.cam.ac.uk) [127.0.0.1] (dwmw2) by imladris.demon.co.uk with esmtp (Exim 1.82 #9) id 0xxLG1-0000Fn-00; Wed, 28 Jan 1998 00:21:33 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: Russell King - ARM Linux Admin Cc: jma24@hermes.cam.ac.uk (J.M. Appleby), P.C.F.Norman@exeter.ac.uk, linux-arm@vger.rutgers.edu Subject: Re: Little X problem Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 28 Jan 1998 00:21:31 +0000 From: David Woodhouse Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO linux@arm.uk.linux.org said: > Your best bet is to wait until the integration of ARM Linux with the > main kernel source tree - chances are that the whole ARM X interface > may well change as a result of several people's work. Hence to remove > yet-another- X-port which does something in a different way I don't > think that it would be a good idea to think about XF86 yet. Anyone working on a KGI port to Linux/ARM? -- ---- ---- ---- David Woodhouse, Robinson College, CB3 9AN, England. (+44) 0976 658355 Dave@imladris.demon.co.uk http://dwmw2.robinson.cam.ac.uk finger pgp@dwmw2.robinson.cam.ac.uk for PGP key. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Wed Jan 28 09:50:09 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id JAA20157 for ; Wed, 28 Jan 1998 09:50:08 GMT Received: from vger.rutgers.edu ([128.6.190.2]:3386 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <2682-617>; Wed, 28 Jan 1998 11:47:49 +0200 Received: by vger.rutgers.edu id <971022-232>; Wed, 28 Jan 1998 04:43:27 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1324 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971408-232>; Wed, 28 Jan 1998 04:37:51 -0500 Received: (from alan@localhost) by snowcrash.cymru.net (8.8.7/8.7.1) id JAA01454; Wed, 28 Jan 1998 09:38:38 GMT From: Alan Cox Message-Id: <199801280938.JAA01454@snowcrash.cymru.net> Subject: Re: Little X problem To: Dave@imladris.demon.co.uk (David Woodhouse) Date: Wed, 28 Jan 1998 09:38:35 +0000 (GMT) Cc: linux@arm.uk.linux.org, jma24@hermes.cam.ac.uk, P.C.F.Norman@exeter.ac.uk, linux-arm@vger.rutgers.edu In-Reply-To: from "David Woodhouse" at Jan 28, 98 00:21:31 am Content-Type: text X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO > > may well change as a result of several people's work. Hence to remove > > yet-another- X-port which does something in a different way I don't > > think that it would be a good idea to think about XF86 yet. > > Anyone working on a KGI port to Linux/ARM? Given the current non-state of the KGI code its probably better to wait. Im sure GGI/KGI will produce something useful but its very unlikely to be in quite the style it is now. Far more useful to get the generic frame buffer consoles and Xserver working. Thats the same X server as the m68K uses for example. Alan unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Wed Jan 28 13:53:58 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id NAA22169 for ; Wed, 28 Jan 1998 13:53:56 GMT Received: from vger.rutgers.edu ([128.6.190.2]:33080 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <2935-619>; Wed, 28 Jan 1998 15:52:29 +0200 Received: by vger.rutgers.edu id <971156-232>; Wed, 28 Jan 1998 08:50:21 -0500 Received: from server21.digital.fr ([193.56.15.21]:3223 "EHLO server21.digital.fr" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <970985-232>; Wed, 28 Jan 1998 08:49:50 -0500 Received: from mail.vbo.dec.com (mail.vbo.dec.com [16.36.208.34]) by server21.digital.fr (8.7.5/8.7) with ESMTP id OAA10842 for ; Wed, 28 Jan 1998 14:50:12 +0100 (MET) Received: from vbormc.vbo.dec.com (vbormc.vbo.dec.com [16.36.208.94]) by mail.vbo.dec.com (8.8.6/8.7) with ESMTP id OAA26857 for ; Wed, 28 Jan 1998 14:50:06 +0100 (MET) Received: from linux.reo.dec.com (linux.reo.dec.com [16.36.0.79]) by vbormc.vbo.dec.com (8.7.3/8.7) with ESMTP id PAA04806 for ; Wed, 28 Jan 1998 15:27:49 +0100 Received: from linux.reo.dec.com (localhost [127.0.0.1]) by linux.reo.dec.com (8.8.5/8.7.1) with ESMTP id NAA32153 for ; Wed, 28 Jan 1998 13:48:07 GMT Message-Id: <199801281348.NAA32153@linux.reo.dec.com> To: linux-arm@vger.rutgers.edu Subject: gcc-2.7.2.2 Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 28 Jan 1998 13:48:07 +0000 From: "Ka'Plaagh" X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO All, following help from the list, I have managed to build gcc-2.7.2.2. I've written down the steps for myself (rotten memory), so here they are for anyone who wants to do the same thing... Dave -------- HOWTO Build GNU Cross Tools for (Strong)ARM ------------------------------------------- David A Rusling (david.rusling@digital.com) January 1998 Overview -------- This HOWTO describes, step by step, how to build cross compiler version of GCC-2.7.2.2 and binutils-2.7 on an Intel Linux system. This results in a set of ARM compilers, assemblers, linkers and so on that will run on an Intel Linux system but generate A.OUT format ARM object files and executable images. This is the set of tools currently being used for Linux kernel development in the ARM world. I did the following on a Red Hat 5.0 Intel Linux system, I cannot comment on what other systems this would work on. Step 1: Find the Sources ------------------------ You will need a number of source trees: binutils-2.7.tar.gz - the master repository for GNU tools is ftp://prep.ai.mit.edu:/pub/gnu binutils-arm-2.7.gz - ARM Linux specific patch file for binutils-2.7. This can be found on the ARM Linux tools ftp/web site: http://www.arm.uk.linux.org/~rmk/armlinux/tools.html gcc-2.7.2.2.tar.gz - Again you can get this from ftp://prep.ai.mit.edu:/pub/gnu crossgcc-gcc-2.7.2.2.patch - This is a set of patches to more easily allow you to build cross compilation tools. You can find this on Cygnus's website ftp://ftp.cygnus.com:/pub/embedded/crossgcc gcc-2.7.2.2.patch - ARM Linux specific patch file for gcc-2.7.2.2, again you can find this on http://www.arm.uk.linux.org/~rmk/armlinux/tools.html Step 2: Unpack the Sources -------------------------- I suggest that you have a nicely separate area for these sources. I used /usr/src/ARM for the sources and the kits were all in /usr/kits. I will assume those directories in these instructions: [a] binutils-2.7 $ cd /usr/src/ARM $ tar -zxvf /usr/kits/binutils-2.7.tar.gz $ cd binutils-2.7 $ cp /usr/kits/binutils-arm-2.7.gz . ; gunzip binutils-arm-2.7.gz $ patch -p1 < binutils-arm-2.7 [g] gcc-2.7.2.2 $ cd /usr/src/ARM $ tar -zxvf /usr/kits/gcc-2.7.2.2.tar.gz $ patch -p1 < /usr/kits/crossgcc-gcc-2.7.2.2.patch $ patch -p1 < gcc-2.7.2.2.patch NOTE: When applying the patch gcc-2.7.2.2.patch do NOT answer yes to the 'previously applied patch' question. Step 3: Make Separate Build Areas --------------------------------- You can build the tools within their source trees but I prefer not too, it just gets too cluttered. So, make some build areas: $ mkdir build-binutils build-gcc-2.7.2.2 Step 4: Understand configure ---------------------------- All GNU tools use a shell script called 'configure' to (yes you've guessed it) configure themselves. Basically you pass this script the appropriate information. This information is: [a] The type of target that this cross compiler will build for (arm-unknown-linuxaout, it's a tuple architecture-manufacturer-os) [b] Where this set of tools will be installed on your system (I assume /usr/local/ARM but you can put it anywhere) [c] The type of host, actually the configure script works this out for itself (usually). Step 5: Build Binutils-2.7 -------------------------- $ cd /usr/src/ARM/build-binutils $ ../binutils-2.7/configure --target=arm-unknown-linux --prefix=/usr/local/ARM -v $ make all You need to be superuser to install the built binutils: $ make install This puts them into /usr/local/ARM/bin and /usr/local/ARM/arm-unknown-linuxaout Step 6: Build Gcc-2.7.2.2 ------------------------- [a] There is a little gotcha. First you have got to edit a line in /usr/src/ARM/gcc-2.7.2.2/config/arm/t-linux Change the line MULTILIB_OPTIONS = m2/m3/m6/m7 to MULTILIB_OPTIONS = m2/m3/m6 [b] Configure $ cd /usr/src/ARM/build-gcc-2.7.2.2 $ ../gcc-2.7.2.2/configure --target=arm-unknown-linux --prefix=/usr/local/ARM -v [c] Build it: $ make all This will fail when it is trying to make glibc for the cross tools (ie a C library containing ARM object code). To make this you need to type $ make libgcc1-asm.a $ mv libgcc1-asm.a libgcc1.a After you have done that you can continue the build: $ make all Finally, you need to be superuser to install it: $make install Step 7: Sit Back and Admire Your Work ------------------------------------- You should now have the following tools in your system (this is not a complete list, just the highlights): /usr/local/ARM/arm-unknown-linuxout/bin -rwxr-xr-x 2 root users 569727 Jan 26 11:50 ar -rwxr-xr-x 2 root users 1018501 Jan 26 11:50 as -rwxr-xr-x 1 root users 112622 Jan 26 19:23 gcc -rwxr-xr-x 2 root users 930542 Jan 26 11:50 ld -rwxr-xr-x 2 root users 578883 Jan 26 11:50 nm -rwxr-xr-x 2 root users 569726 Jan 26 11:50 ranlib -rwxr-xr-x 2 root users 976042 Jan 26 11:50 strip /usr/local/ARM/arm-unknown-linuxaout/include -rw-r--r-- 1 root users 1486 Jan 26 19:22 assert.h /usr/local/ARM/include -rw-r--r-- 1 root users 4393 Jan 26 11:50 ansidecl.h -rw-r--r-- 1 root users 89253 Jan 26 11:50 bfd.h -rw-r--r-- 1 root users 18948 Jan 26 11:50 bfdlink.h -rw-r--r-- 1 root users 19750 Jan 26 11:50 obstack.h /usr/local/ARM/bin -rwxr-xr-x 2 root users 569727 Jan 26 11:50 arm-unknown-linuxaout-ar -rwxr-xr-x 2 root users 1018501 Jan 26 11:50 arm-unknown-linuxaout-as -rwxr-xr-x 2 root users 17954 Jan 26 19:22 arm-unknown-linuxaout-c++ -rwxr-xr-x 1 root users 60864 Jan 26 11:50 arm-unknown-linuxaout-c++filt -rwxr-xr-x 2 root users 17954 Jan 26 19:22 arm-unknown-linuxaout-g++ -rwxr-xr-x 1 root users 150511 Jan 26 11:50 arm-unknown-linuxaout-gasp -rwxr-xr-x 1 root users 112622 Jan 26 19:23 arm-unknown-linuxaout-gcc -rwxr-xr-x 2 root users 930542 Jan 26 11:50 arm-unknown-linuxaout-ld -rwxr-xr-x 2 root users 578883 Jan 26 11:50 arm-unknown-linuxaout-nm -rwxr-xr-x 1 root users 976043 Jan 26 11:50 arm-unknown-linuxaout-objcopy -rwxr-xr-x 1 root users 1002142 Jan 26 11:50 arm-unknown-linuxaout-objdump -rwxr-xr-x 2 root users 569726 Jan 26 11:50 arm-unknown-linuxaout-ranlib -rwxr-xr-x 1 root users 532558 Jan 26 11:50 arm-unknown-linuxaout-size -rwxr-xr-x 1 root users 525393 Jan 26 11:50 arm-unknown-linuxaout-strings -rwxr-xr-x 2 root users 976042 Jan 26 11:50 arm-unknown-linuxaout-strip -rwxr-xr-x 1 root users 87807 Jan 26 19:22 protoize -rwxr-xr-x 1 root users 77237 Jan 26 19:22 unprotoize /usr/local/ARM/lib drwxrwsr-x 3 root users 1024 Jan 26 19:22 gcc-lib -rw-rw-r-- 1 root users 1358676 Jan 26 11:50 libarm-unknown-linuxaout-bfd.a -rw-rw-r-- 1 root users 86518 Jan 26 11:50 libarm-unknown-linuxaout-opcodes.a -rw-rw-r-- 1 root users 156468 Jan 26 11:50 libiberty.a ---------------------------------------------------------------------- David A Rusling Consulting Engineer European Semiconductor Applications Digital Equipment Co Ltd., Engineering PO Box 121, Imperial Way, Worton Grange Reading RG2 0TU Linux, Alpha, StrongArm, PCI Tel: UK-(0)1734-204380 Fax: UK-(0)1734-203133 ---------------------------------------------------------------------- unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Wed Jan 28 17:36:41 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id RAA23703 for ; Wed, 28 Jan 1998 17:36:39 GMT Received: from vger.rutgers.edu ([128.6.190.2]:35962 "EHLO vger.rutgers.edu" ident: "TIMEDOUT") by nic.funet.fi with ESMTP id <15355-621>; Wed, 28 Jan 1998 19:35:05 +0200 Received: by vger.rutgers.edu id <971498-232>; Wed, 28 Jan 1998 12:17:22 -0500 Received: from nexusel.demon.co.uk ([158.152.30.195]:10861 "HELO globe" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with SMTP id <971504-232>; Wed, 28 Jan 1998 12:09:18 -0500 Received: from spring.nexus.co.uk [192.0.0.3] (root) by globe with smtp (Exim 1.73 #1) id 0xxazT-0001XN-00; Wed, 28 Jan 1998 17:09:31 +0000 Received: from localhost (spring.nexus.co.uk) [127.0.0.1] (pb) by spring.nexus.co.uk with esmtp (Exim 1.82 #1) id 0xxazR-0004NB-00; Wed, 28 Jan 1998 17:09:29 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: "Ka'Plaagh" cc: linux-arm@vger.rutgers.edu Subject: Re: gcc-2.7.2.2 In-reply-to: Your message of "Wed, 28 Jan 1998 13:48:07 GMT." <199801281348.NAA32153@linux.reo.dec.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 28 Jan 1998 17:09:27 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO In message <199801281348.NAA32153@linux.reo.dec.com>, "Ka'Plaagh" writes: >This will fail when it is trying to make glibc for the cross tools (ie a C >library containing ARM object code). To make this you need to type That's not glibc - you mean libgcc. They're not the same thing. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Wed Jan 28 18:38:10 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id SAA24085 for ; Wed, 28 Jan 1998 18:38:09 GMT Received: from vger.rutgers.edu ([128.6.190.2]:3937 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <1736-619>; Wed, 28 Jan 1998 20:36:20 +0200 Received: by vger.rutgers.edu id <971206-232>; Wed, 28 Jan 1998 13:31:05 -0500 Received: from server21.digital.fr ([193.56.15.21]:3740 "EHLO server21.digital.fr" ident: "TIMEDOUT") by vger.rutgers.edu with ESMTP id <971209-232>; Wed, 28 Jan 1998 13:21:31 -0500 Received: from mail.vbo.dec.com (mail.vbo.dec.com [16.36.208.34]) by server21.digital.fr (8.7.5/8.7) with ESMTP id TAA24496; Wed, 28 Jan 1998 19:22:25 +0100 (MET) Received: from vbormc.vbo.dec.com (vbormc.vbo.dec.com [16.36.208.94]) by mail.vbo.dec.com (8.8.6/8.7) with ESMTP id TAA15677; Wed, 28 Jan 1998 19:22:25 +0100 (MET) Received: from linux.reo.dec.com (linux.reo.dec.com [16.36.0.79]) by vbormc.vbo.dec.com (8.7.3/8.7) with ESMTP id UAA24576; Wed, 28 Jan 1998 20:00:21 +0100 Received: from linux.reo.dec.com (localhost [127.0.0.1]) by linux.reo.dec.com (8.8.5/8.7.1) with ESMTP id SAA06558; Wed, 28 Jan 1998 18:20:42 GMT Message-Id: <199801281820.SAA06558@linux.reo.dec.com> To: Philip Blundell cc: linux-arm@vger.rutgers.edu Subject: Re: gcc-2.7.2.2 In-reply-to: Your message of "Wed, 28 Jan 1998 17:09:27 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 28 Jan 1998 18:20:41 +0000 From: "Ka'Plaagh" X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO duh, thanks, I thought one thing and wrote another... Dave ---------------------------------------------------------------------- David A Rusling Consulting Engineer European Semiconductor Applications Digital Equipment Co Ltd., Engineering PO Box 121, Imperial Way, Worton Grange Reading RG2 0TU Linux, Alpha, StrongArm, PCI Tel: UK-(0)1734-204380 Fax: UK-(0)1734-203133 ---------------------------------------------------------------------- unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Wed Jan 28 23:58:17 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id XAA26559 for ; Wed, 28 Jan 1998 23:58:15 GMT Received: from vger.rutgers.edu ([128.6.190.2]:15645 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13853-620>; Thu, 29 Jan 1998 01:57:53 +0200 Received: by vger.rutgers.edu id <971472-232>; Wed, 28 Jan 1998 18:35:02 -0500 Received: from snowcrash.cymru.net ([163.164.160.3]:1063 "EHLO snowcrash.cymru.net" ident: "NO-IDENT-SERVICE") by vger.rutgers.edu with ESMTP id <971314-232>; Wed, 28 Jan 1998 17:53:48 -0500 Received: from caramon.armlinux.org (dialup1.cymru.net [163.164.160.65]) by snowcrash.cymru.net (8.8.7/8.7.1) with ESMTP id WAA16813; Wed, 28 Jan 1998 22:54:24 GMT Received: from raistlin.armlinux.org (raistlin [192.168.0.3]) by caramon.armlinux.org (8.7.4/8.7.3) with ESMTP id WAA08268; Wed, 28 Jan 1998 22:54:41 GMT From: Russell King - ARM Linux Admin Received: (from linux@localhost) by raistlin.armlinux.org (8.7.4/8.7.3) id WAA00732; Wed, 28 Jan 1998 22:22:14 GMT Message-Id: <199801282222.WAA00732@raistlin.armlinux.org> Subject: Re: Little X problem To: pkoning@xedia.com (Paul Koning) Date: Wed, 28 Jan 1998 22:22:12 +0000 (GMT) Cc: linux-arm@vger.rutgers.edu In-Reply-To: <9801281502.AA12440@kona.> from "Paul Koning" at Jan 28, 98 10:02:52 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO Paul Koning writes: > I'm puzzled. XWindows is a type of windowing system, of which XFree86 > is an instance. So to say that XWindows doesn't have VT switching > support seems odd. Did you mean "The XFree86 implementation of > XWindows has VT switching support, but the Fubar42 version of XWindows > does not" ? XFree86 is {X server + XFree code} - it is a specific hardware `type' in that it has it's own ddx backend for X. The X server is made up of various modules, one of them is ddx - the `device dependent' end. XFree86 is one such backend that happens to support VT switching. XNest is another DDX, as is the vfb, virtual frame buffer. To give you more of an idea, the standard X distribution comes with the following 'device drivers' (minus the armlinux entry of course): hw: drwxr-xr-x 2 src src 1024 May 24 1997 armlinux/ drwxr-xr-x 3 src src 1024 Apr 21 1994 dec/ drwxr-xr-x 5 src src 1024 Apr 21 1994 hp/ drwxr-xr-x 6 src src 1024 Apr 21 1994 ibm/ drwxr-xr-x 2 src src 1024 Apr 21 1994 macII/ drwxr-xr-x 2 src src 1024 Apr 21 1994 omron/ drwxr-xr-x 2 src src 1024 Apr 4 1995 sun/ drwxr-xr-x 2 src src 1024 Apr 21 1994 sunAmoeba/ drwxr-xr-x 3 src src 1024 Apr 21 1994 svga/ drwxr-xr-x 2 src src 1024 Apr 5 1995 vfb/ drwx------ 11 src src 1024 Apr 1 1995 xfree86/ drwxr-xr-x 2 src src 1024 Apr 21 1994 xnest/ Thus, if you have an Xserver built with, say, the macII ddx, then you will not get VT switching. _____ |_____| ------------------------------------------------- ---+---+- | | Russell King rmk@ecs.soton.ac.uk --- --- | | | | http://www.arm.uk.linux.org/~rmk/home.html / / | | +-+-+ --- -+- / | THE developer of ARM Linux |+| /|\ / | | | --- | +-+-+ ------------------------------------------------- /\\\ | unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sat Jan 31 11:04:46 1998 Return-Path: Received: from nic.funet.fi ([128.214.248.6] (may be forged)) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id LAA17345 for ; Sat, 31 Jan 1998 11:04:45 GMT Received: from vger.rutgers.edu ([128.6.190.2]:48189 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <12874-617>; Sat, 31 Jan 1998 13:04:51 +0200 Received: by vger.rutgers.edu id <971093-206>; Sat, 31 Jan 1998 06:02:51 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:1051 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <971086-206>; Sat, 31 Jan 1998 06:02:31 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xybb1-0000AF-00; Sat, 31 Jan 1998 12:00:27 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: linux-arm@vger.rutgers.edu Subject: binutils 2.8.1.0.20 is released (fwd) Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 31 Jan 1998 12:00:27 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO This version has the ELF patches I posted to this list a few weeks ago. It ought also to have the 64-bit ADR fix but I haven't checked that yet. Please let me know if there are any remaining problems. p. ------- Forwarded Message From: hjl@lucon.org (H.J. Lu) Subject: binutils 2.8.1.0.20 is released. To: Philip.Blundell@pobox.com Date: Fri, 30 Jan 1998 18:46:22 -0800 (PST) Cc: kjahds@kjahds.com (Kenneth Albanowski), lmfken@lmf.ericsson.se (Kenneth Osterberg), ian@lasermoon.co.uk, mat@lcs.mit.edu (Mat Hostetter), doughera@lafcol.lafayette.edu (Andy Dougherty), brian@mathworks.com (Brian Bourgault), john@etools.com (John W. Christy), craig@metrolink.com (Craig Groeschel), imp@village.org (Warner Losh), robf@willows.com (Rob Farnum), meissner@cygnus.com (Michael Meissner), rfg@monkeys.com (Ron Guilmette), roell@xinside.com (Thomas Roell), burley@gnu.org (Craig Burley), linux-binutils-in@polstra.com (John Polstra), Shimon@i-Connect.Net (Simon Shapiro), galenh@micron.net (Galen Hazelwood), ralf@mailhost.uni-koblenz.de (Ralf Baechle), linux-gcc@vger.rutgers.edu (linuxgcc) X-Mailer: ELM [version 2.4 PL25 PGP2] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Hi, sunsite.unc.edu is full. Please use other ftp sites. Thanks. - -- H.J. Lu (hjl@gnu.org) - --- This is the beta release of binutils 2.8.1.0.20 for Linux, which is based on the binutils 980130 snapshot plus linux/PPC patch and Linux/ARM support. There was some problem on Linux/alpha. It turned out it was a bug in the alpha dynamic linker in glibc 2.0.6. The current glibc 2.1 and glibc 2.0.7 should be ok. This release should work on linux/x86. There are some ELF bug fixes. But I don't know if it fixes all the bugs reported on linux/sparc, Linux/MIPS and linux/PPC. I'd like to hear reports on them. Please report any bugs related to binutils 2.8.1.0.20 to hjl@gnu.org. Changes from binutils 2.8.1.0.19: 1. Update from the binutils gas-980130. 2. Fix the ARM/ELF support. 3. Fix a.out symbol aliase. 4. Fix -r with .gnu.linkonce.d* and .gnu.linkonce.t* sections on ELF objects. Changes from binutils 2.8.1.0.18: 1. Update from the binutils gas-980114. 2. The ARM/ELF support. 3. Fix lib*.so.* support. Us the release number in soname. Changes from binutils 2.8.1.0.17: 1. Update from the binutils gas-971211. 2. Add lib*.so.* support. But the soname is lib*.so.2. Changes from binutils 2.8.1.0.16: 1. Update from the binutils gas-971105. 2. Fix ld test for linux. 3. Fix the 2.8.1.0.16 patch. Changes from binutils 2.8.1.0.15: 1. Update from the binutils gas-971103. 2. ELF/Alpha seems to work fine, including versioning. 3. ELF/PPC patches from Geoff Keating . IMPORTANT: The .plt format that the Alpha was using was not thread safe. So I changed it. The new format is *not* binary compatible with the old, thus you must use glibc 2.0.4 to get an ld.so that can understand it. Note that the new ld.so knows how to deal with the old .plt format, so old libraries will still work, but they should be relinked eventually for performance reasons. The file list: 1. binutils-2.8.1.0.20.tar.gz. Source code. 2. binutils-2.8.1.0.19-2.8.1.0.20.diff.gz. Patch against the previous beta source code. 3. binutils-2.8.1.0.20.bin.tar.gz. Precompiled Linux/x86 binaries for libc 5. 4. binutils-2.8.1.0.20-glibc.bin.tar.gz. Precompiled Linux/x86 binaries for libc 6 (glibc 2.0.5c or above). The primary ftp sites for the compiler/C library are tsx-11.mit.edu under pub/linux/packages/GCC and sunsite.unc.edu under pub/Linux/GCC. The beta directory is in private/tofu under the GCC directory. Most of my stuff can also be found at ftp://ftp.yggdrasil.com/private/hjl To install the binary package, please follow the procedure very closely. Please backup/save all the files you are instructed to delete and you should do gzip -dc binutils-2.8.1.0.20.bin.tar.gz | tar tvvf - or gzip -dc binutils-2.8.1.0.20-glibc.bin.tar.gz | tar tvvf - to see what is in there. Please do back up before you remove things. To install for libc 5, PLEASE DO 1. su root 2. cd / 3. rm -f /usr/bin/encaps /usr/bin/nm 4. gzip -dc binutils-2.8.1.0.20.bin.tar.gz | tar xvvf - 5. ldconfig To install for libc 6 (glibc 2.0.5c or above), PLEASE DO 1. su root 2. cd / 3. rm -f /usr/bin/encaps /usr/bin/nm 4. gzip -dc binutils-2.8.1.0.20-glibc.bin.tar.gz | tar xvvf - 5. ldconfig Now you have the new gas/binutils under /usr/bin and /usr/ix86-linuxaout/bin. You have to use /usr/ix86-linuxaout/bin/as and /usr/ix86-linuxaout/bin/ld -m i386linux if you want to use a.out as and ld directly. I have changed the target names for Linux/x86 and Linux/x86 (a.out) to ix86-linux and ix86-linuxaout respectively. The precompiled binaries are installed under /usr/ix86-linux and /usr/ix86-linuxaout. You should make appropriate symbolic links if you have a different name for the Linux/x86 target, like i586-unknown-linux. If you have an old linux library in the a.out format and you cannot obtain the newer version in the ELF format for whatever reason, you can try "objcopy --remove-leading-char" on the a.out library and see if it can link with your code in ELF. For Sybase client libraries, I did # cd sybase/lib # for f in *.a do objcopy --remove-leading-char $f done It seems to compile/link fine with libc 5.4.2x. But I don't know if it really works or not. The resulting binaries should work if the libc ABI used by the old library is unchanged in the new libc. According to Bruce Milner : Just a quick note about your example using objcopy for sybase. The sybase a.out libraries "are" dependent on a changed feature of the libc ABI. The ctype macros changed to a byte order independent format (little endian). In order to use the sybase libraries, one needs to rename the the ctype table variables in the sybase library files and link in an old version of the libc's "C-ctype.o ctype-info.o" files. thain@sunquest.sunquest.com put a ELF'ized version of these libraries on ftp://mudshark.sunquest.com/pub/ctlib-linux-elf Thanks. H.J. Lu hjl@gnu.org 01/31/98 ------- End of Forwarded Message unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu From owner-linux-arm-outgoing@vger.rutgers.edu Sat Jan 31 11:24:14 1998 Return-Path: Received: from nic.funet.fi (nic.funet.fi [128.214.248.6]) by odie.barnet.ac.uk (8.8.6/8.8.6) with ESMTP id LAA17533 for ; Sat, 31 Jan 1998 11:24:13 GMT Received: from vger.rutgers.edu ([128.6.190.2]:3158 "EHLO vger.rutgers.edu" ident: "root") by nic.funet.fi with ESMTP id <13004-621>; Sat, 31 Jan 1998 13:23:56 +0200 Received: by vger.rutgers.edu id <971094-206>; Sat, 31 Jan 1998 06:22:00 -0500 Received: from tazenda.demon.co.uk ([158.152.220.239]:1118 "HELO paddington.london.uk.eu.org" ident: "exim") by vger.rutgers.edu with SMTP id <971086-206>; Sat, 31 Jan 1998 06:21:51 -0500 Received: from paddington.london.uk.eu.org [127.0.0.1] (phil) by paddington.london.uk.eu.org with esmtp (Exim 1.71 #4) id 0xybuz-0000Cl-00; Sat, 31 Jan 1998 12:21:05 +0000 X-Mailer: exmh version 2.0zeta 7/24/97 To: linux-arm@vger.rutgers.edu Subject: Re: binutils 2.8.1.0.20 is released (fwd) In-reply-to: Your message of "Sat, 31 Jan 1998 12:00:27 GMT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 31 Jan 1998 12:21:05 +0000 From: Philip Blundell Message-Id: X-Orcpt: rfc822;linux-arm@vger.rutgers.edu Sender: owner-linux-arm@vger.rutgers.edu Precedence: bulk Status: RO >ought also to have the 64-bit ADR fix but I haven't checked that yet. Please Ian's fix for this problem is indeed also included. p. unsubscribe: body of `unsubscribe linux-arm' to majordomo@vger.rutgers.edu