From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:37:57 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: Test from new site. Sender: NetBSD-Admin@cbmuucp.commodore.com This is a test from the new NetBSD-Amiga mailing list site. Please update any macros that you use to point to: NetBSD-Amiga@commodore.com and send all subscribes and unsubscribes to: NetBSD-Admin@commodore.com All spooled messages posted to icecube will be sent here shortly, and aliases will be updated to point to the new address to prevent any missed messages. From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:40:20 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: OOPS! cbmuucp.commodore.com Sender: NetBSD-Admin@cbmuucp.commodore.com ***BRAIN FART ALERT*** it's ... cbmuucp.commodore.com, not commodore.com.. SORRY!! From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:45:15 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: Kernel built, but panics? Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: Date: Wed, 30 Jun 93 07:57:28 DST From: niklas@appli.se (Niklas Hallqvist) To: abair@parsons.sps.mot.com Cc: netbsd-amiga@icecube.rain.com In-Reply-To: <9306291725.AA26969@sol-tx.sps.mot.com> (message from Alan Bair on Tue, 29 Jun 93 12:25:54 CDT) Subject: Re: Kernel built, but panics? >>>>> "Alan" == Alan Bair writes: Alan> I did notice another problem with the Makefile built by Alan> config. For the target subr_mcount.o, it does not give any Alan> commands, like the ${NORMAL_C} that most everything else has, so Alan> it never compiled this file. I just added the ${NORMAL_C} to get Alan> past that. I got that as well, and fixed it just like you did. At te time I thought it was stack size related (I had had a small stack size when I ran config), so I didn't bother to check it again. Well, I just did and that problem is still there. GNU make and dmake handle this case differently, if memory serves me right , GNU make stops but dmake goes on to compile via a default rule. That might be why Markus haven't seen it. Alan> By the way, I seem to be the only one asking about building the Alan> kernel. Has anyone else got it to work such that you can Alan> succesfully boot the new kernel? Oh yes, it's the only way for me as I run GVP on a A2000! Niklas Hallqvist Phone: +46-(0)31-40 75 00 Applitron Datasystem Fax: +46-(0)31-83 39 50 Molndalsvagen 95 Email: niklas@appli.se S-412 63 GOTEBORG, Sweden mcsun!seunet!appli!niklas From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:46:18 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: Recompiling the BSD kernel Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: From: eeh@public.btr.com (Eduardo E. Horvath eeh@btr.com) Subject: Re: Recompiling the BSD kernel To: mw@eunet.ch (Markus Wild) Date: Tue, 29 Jun 1993 21:27:21 -0700 (PDT) Cc: netbsd-amiga@icecube.rain.com In-Reply-To: <199306271342.AA18258@chsun.eunet.ch> from "Markus Wild" at Jun 27, 93 03:42:40 pm X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 228 I finally have everything installed and I'm trying to build a new kernel, but ld.bsd is constantly giving me GURU #8000 0003. Any ideas? I only have 4M Fast RAM, but I appear to have ~1.8M free when the GURU happens. Eduardo From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:47:51 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: /dev/drum group? Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: From: Tero Manninen Subject: /dev/drum group? To: netbsd-amiga@icecube.rain.com (NetBSD mailing list NetBSD mailing list) Date: Thu, 1 Jul 1993 22:14:31 +0300 (EET DST) X-Mailer: ELM [version 2.4 PL21] Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8 Content-Length: 470 Uhm.. what is the correct group for /dev/drum? I'm not sure if my /dev was cloned 100% when I tarred it and transferred to the current root disk, but now I noticed that the group was wheel (permissions for that "r") and for any normal user outside wheel group ps and w didn't work. The /bin/ps is sgid kmem -> wheel group of /dev/drum actually denied the access to it :-) The same was for /usr/bin/w. I changed the /dev/drum to sgid kmem.. all well so far.. ++Tero From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:47:57 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: GVP support delayed... :-( Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: Date: Thu, 1 Jul 93 14:06:33 DST From: niklas@appli.se (Niklas Hallqvist) To: netbsd-amiga@icecube.rain.com Subject: Re: GVP support delayed... :-( I know there are a bunch of people interested in the GVP support I'm adding to NetBSD. Sad to say, I probably won't make it before I leave for vacation. I'll be back August. My AmigaDOS disk got corrupted while I tried to add yet another SCSI disk to my bus. On the corrupted disk is my NetBSD devtree, so I can't do any work until I get it up (the disk, I mean :-). This also means that I won't be able to contribute the SCSI subclassing implementation before August (although it was almost finished, it's no big deal really). If someone wants it sooner, just go ahead and roll your own and send it to Markus. BTW. The 377 kernel on ftp.eunet.ch, why no source or readme, Markus? Niklas From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:49:14 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: NetBSD: Prog for battery clock Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: From: Holger Emden Subject: NetBSD: Prog for battery clock To: netbsd-admin@icecube.rain.com Date: Thu, 1 Jul 1993 06:28:53 +0200 (MET DST) Cc: butabrax@cs.tu-berlin.de (Holger Emden) X-Mailer: ELM [version 2.4 PL21] Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Content-Length: 3540 Hello. First a big THANK YOU for NetBSD. It's fantastic. In one of your mails, i read, that you are looking for a program to read the battery clock. The program below works fine on my A3000, but i don't know, if it works on all Amiga, because in several books, they use different addresses and offsets. But before : I would like to get into your mailing list. Thank you. Here it comes: * Read the internal battery clock of the A3000 * * In a lot of books, the address and offsets in memory are different, * so lets have a look at battclock.resource :-) * * All data is found in the lower 4 bits of a byte. Also, the clock uses * upper and lower digits. start: lea $dc0000,a0 ; get base of battery clock lea buffer(pc),a1 ; destination address moveq #$f,d0 ; mask for lower 4 bits clr.b $37(a0) ; hey clock, wanna read you ... move.b $1b(a0),(a1) ; weekday ( 0-6 / 1-7 ??? ) and.b d0,(a1)+ move.b $23(a0),(a1) ; upper digit of day and.b d0,(a1)+ move.b $1f(a0),(a1) ; lower digit of day and.b d0,(a1)+ move.b $2b(a0),(a1) ; upper digit of month and.b d0,(a1)+ move.b $27(a0),(a1) ; lower digit of month and.b d0,(a1)+ move.b $33(a0),(a1) ; upper digit of jear and.b d0,(a1)+ move.b $2f(a0),(a1) ; lower digit of jear and.b d0,(a1)+ move.b $17(a0),(a1) ; upper digit of hour and.b d0,(a1)+ move.b $13(a0),(a1) ; lower digit of hour and.b d0,(a1)+ move.b $0f(a0),(a1) ; upper digit of minute and.b d0,(a1)+ move.b $0b(a0),(a1) ; lower digit of minute and.b d0,(a1)+ move.b $07(a0),(a1) ; upper digit of second and.b d0,(a1)+ move.b $03(a0),(a1) ; lower digit of second and.b d0,(a1)+ move.b $37(a0),(a1) ; control-byte 1 and.b d0,(a1)+ move.b $3b(a0),(a1) ; control-byte 2 and.b d0,(a1)+ move.b $3f(a0),(a1) ; control-byte 3 and.b d0,(a1) move.b #9,$37(a0) ; ok clock, go on ... lea buffer(pc),a0 ; convert from buffer to lea result(pc),a1 ; result move.b (a0)+,(a1)+ ; weekday moveq #0,d0 move.b (a0)+,d0 ; upper digit of day muls #10,d0 add.b (a0)+,d0 ; lower digit of day move.b d0,(a1)+ ; save it moveq #0,d0 move.b (a0)+,d0 ; upper digit of month muls #10,d0 add.b (a0)+,d0 ; lower digit of month move.b d0,(a1)+ ; save it moveq #0,d0 move.b (a0)+,d0 ; upper digit of jear muls #10,d0 add.b (a0)+,d0 ; lower digit of jear move.b d0,(a1)+ ; save it moveq #0,d0 move.b (a0)+,d0 ; upper digit of hour muls #10,d0 add.b (a0)+,d0 ; lower digit of hour move.b d0,(a1)+ ; save it moveq #0,d0 move.b (a0)+,d0 ; upper digit of minute muls #10,d0 add.b (a0)+,d0 ; lower digit of minute move.b d0,(a1)+ ; save it moveq #0,d0 move.b (a0)+,d0 ; upper digit of second muls #10,d0 add.b (a0)+,d0 ; lower digit of second move.b d0,(a1)+ ; save it * ok, ok, next time i'll use a macro :-) move.b (a0)+,(a1)+ ; control-byte 1 move.b (a0)+,(a1)+ ; control-byte 2 move.b (a0)+,(a1) ; control-byte 3 rts ; we did it :-) buffer: dc.l 0,0,0,0 ; only a buffer result: dc.b 0 ; weekday dc.b 0 ; day dc.b 0 ; month dc.b 0 ; jear dc.b 0 ; hour dc.b 0 ; minute dc.b 0 ; second dc.b 0 ; control-byte 1 dc.b 0 ; control-byte 2 dc.b 0 ; control-byte 3 Thank you very much and i wish you a lot of fun for your work on NetBSD. Holger Emden (butabrax) -- Holger Emden, Technical University of Berlin (Germany), Area : Computer-Engineering, EMail : butabrax@cs.tu-berlin.de From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:49:37 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: vmunix377 ps still broken. Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: From: hoffmann@nutmeg.cs.ntu.edu.au (Arthur Hoffmann) Subject: vmunix377 ps still broken. To: netbsd-admin@icecube.rain.com Date: Wed, 30 Jun 1993 16:17:12 +0930 (CST) X-Mailer: ELM [version 2.4 PL5] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 908 Hi, I just installed vmunix377 and it seemed as if ps and w were working, but when going into multi user mode I still get: ps:kvm_openfiles can't open /dev/drum w doesnt work either. By the way, how come I can't login into root from a user account, It says my password is wrong. When logged out of the user accout it works. I seem to have some problem with passwords anyway, when typing su in a users account then i get told that I am in the wrong group. And vipw wouldn't let me change the root entry. (maybe someone suggests a good book on UNIX for systems operateros :) ). And something else, is there any way to get files from NetBSD to the amiga without a tape? was using term and wanted to download the new kernel when I realized that I had no way to put it onto the amiga. so I had to reboot into Amigados. Thanks for your help. bye. Arthur Hoffmann hoffmann@nutmeg.ntu.edu.au From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:49:52 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: NetBSD for GVP based machines Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: From: Trevor Elbourne Subject: NetBSD for GVP based machines To: netbsd-amiga@icecube.rain.com (Amiga NetBSD mailing list) Date: Tue, 29 Jun 93 17:33:27 EST X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@icecube.rain.com I notice that NetBSD has been ported to run on amigas with GVP series II HD controllers. Is it available yet? Thanks in advance, Trevor. -- +-------------------------------------------------------------------- -+ | "Let me be the first to offer you my sincerest contrafibularities" | | - E. Blackadder | +----------------------------+--------------------------------------- -+ | Trevor Elbourne | Computer and Systems Technology (CaST) | | trevore@vast.unsw.edu.au | Dept of Computer Science & Engineering | | trevore@cse.unsw.edu.au | University of New South Wales | +----------------------------+--------------------------------------- -+ From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:50:10 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: Display and recompiling Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: Date: Wed, 30 Jun 93 08:11:23 DST From: niklas@appli.se (Niklas Hallqvist) To: mw@eunet.ch Cc: abair@parsons.sps.mot.com, netbsd-amiga@icecube.rain.com In-Reply-To: <199306281838.AA06385@chsun.eunet.ch> (mw@eunet.ch) Subject: Re: Display and recompiling >>>>> "Markus" == Markus Wild writes: Markus> Mime-Version: 1.0 Markus> Content-Type: text/plain; charset=US-ASCII Markus> Content-Transfer-Encoding: 7bit Markus> Content-Length: 2214 Alan> Would running Enforcer, Mungwall and sushi have anything to do with Alan> the 2.04 -> NetBSD booting and a bad display? I don't run these with Alan> 1.3. Markus> I'm running enforcer... It has nothing to do with those tools. For me it's the kickstart image that matters. If I warmboot on my NetBSD boot floppy when 37.175 is in effect the console is messy. If I coolboot on my 1.3 ROMs everything is fine. My boot floppy just has ixemul, loadbsd and vmunix on it, no other tools. Niklas Hallqvist Phone: +46-(0)31-40 75 00 Applitron Datasystem Fax: +46-(0)31-83 39 50 Molndalsvagen 95 Email: niklas@appli.se S-412 63 GOTEBORG, Sweden mcsun!seunet!appli!niklas From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:50:35 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: Old GVP Impact controllers Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: Date: Wed, 30 Jun 93 08:15:25 DST From: niklas@appli.se (Niklas Hallqvist) To: erikr@minerva.cis.yale.edu Cc: NetBSD-Amiga@icecube.rain.com In-Reply-To: (message from Erik Rauch on Mon, 28 Jun 1993 13:56:24 -0400 (EDT)) Subject: Re: Old GVP Impact controllers >>>>> "Erik" == Erik Rauch writes: Erik> Mime-Version: 1.0 Erik> Content-Type: TEXT/PLAIN; charset=US-ASCII Niklas> Thanks!!! Once I got the 366 version, it took me about 5 minutes Niklas> to put in the GVP DMA support, and it worked right away, Niklas> unbelievable :-) Markus> Cool! Ok, we now run on more Amigas than just A3000, nice :-) Erik> Would this work on old GVP Impact A2000/2 controllers Erik> (pre-Series II)? Not yet, and will probably not work in the foreseeable future unless someone with a pre-Series II controller like to hack a little. I can provide some help with the hardware interfacing, as I've disassembled gvpscsi.device once, and got some knowledge on these controllers as well when doing that. Niklas From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:51:00 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: Multiserial ports (fwd) Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: From: Alan Bair Subject: Re: Multiserial ports (fwd) To: netbsd-amiga@icecube.rain.com (NetBSD on Amiga) Date: Tue, 29 Jun 93 22:05:27 CDT Cc: thecure@mullian.ee.mu.oz.au X-Mailer: ELM [version 2.3 PL11] I am forwarding this to the mailing list for Matthew. He sent it to me by mistake and has lost the mailing list address: netbsd-amiga@icecube.rain.com Forwarded message: > From: thecure@mullian.ee.mu.OZ.AU (matthew aldous) > This is my reponse to hist questions below: > Sorry I don't this kind of information about either Linux or > NetBSD to be of any help. However, I would suggest you go > with NetBSD, since I suspect this would give you a single > OS for both machine types that is fairly far along on both. > Amiga Linux will probably be awhile, especially with Amiga > NetBSD available. > > Also, this seemed to be addressed directly to me, so it may > not have made it to the NetBSD mailing list. > > His questions: > > Not strictly an Amiga BSD thing, but.. > > I have a friend asking me to install a un*x environment for him > and I'm trying hard to toss up if I should install Linux or > NetBSD ( both on a 386dx platform). The catch for him is that he > has two four port serial cards that have to be working under it > as well. I've succesfully had it working under Linux, and never > tried NetBSD (on 386dx) > > Now, the plan is we both connect up at some stage, and as I have > the a2232 (8 serial ports) I'm hoping at some stage there will > be support for it. > > Does anyone know if the NetBSD (386dx) can support 8 serial ports? > > (If I install an OS on his puter, I'd like the same on mine.. > Should I wait for Linux (ha) or wait for NetBSD) > > (In case it's not clear I have the Amiga, he's got the 386dx) -- Alan Bair AMCU DSCS Motorola, Inc. (Design Software & Mail Stop OE-320 Computer Services) 6501 William Cannon Dr. West Austin, TX 78735-8598 abair@amcu-tx.sps.mot.com From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:51:09 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: ps? Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: From: Magnus.Gronlund@ludd.luth.se Subject: ps? To: netbsd-amiga@icecube.rain.com Date: Tue, 29 Jun 1993 21:36:01 +0200 (MET DST) X-Mailer: ELM [version 2.4 PL17] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 497 What's wrong with ps? Everything's ok the first few times but after that all I get is a long, long wait and finally a segmentation fault ( trapsignal(96, 11, 4, 4, 63a2) ). It doesn't matter what kernel I use. I'm surpriced I haven't read anything about this anywhere, have I missed something in the docs? Over to something positive. I can now announce a new mirror of the NetBSD-Amiga files. They can be found at: ftp.luth.se - "/pub/unix/NetBSD-Amiga" /Magnus Gr|nlund d91-mgd@ludd.luth.se From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:51:19 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: NetBSD for GVP based machines Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: Date: Wed, 30 Jun 93 08:00:38 DST From: niklas@appli.se (Niklas Hallqvist) To: trevore@vast.unsw.edu.au Cc: netbsd-amiga@icecube.rain.com In-Reply-To: <199306290733.AA12792@moule.vast.unsw.edu.au> (message from Trevor Elbourne on Tue, 29 Jun 93 17:33:27 EST) Subject: Re: NetBSD for GVP based machines >>>>> "Trevor" == Trevor Elbourne writes: Trevor> I notice that NetBSD has been ported to run on amigas with GVP Trevor> series II HD controllers. Is it available yet? No, not yet. I hope it will be this week, otherwise I'll post source here (it's not big) for people to play with. I'm leaving for vacation by Friday and won't have access to computers (I think) for the rest of July. Niklas From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:51:22 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: vmunix.366 and compiling on NetBSD Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: Date: Tue, 29 Jun 93 13:14:42 -0600 From: baford@schirf.cs.utah.edu (bryan ford) To: netbsd-amiga@icecube.rain.com Subject: vmunix.366 and compiling on NetBSD First: vmunix.335 works fine for me, but I have been unable to get vmunix.366 to work at all. After it loads and prints the available memory on the AmigaDOS window, it blanks the screen to white (OK), and remains that way forever (not OK). Any ideas? Second: I have successfully compiled vmunix under NetBSD. Markus: there's one line in genassym.c which uses '\T' for tab instead of '\t'; I assume this doesn't cause problems under ixemul, but it does under BSD. Also, the rule lines were missing in the Makefile for "subr_mcount.o"; I don't know where this part of the Makefile originally comes from. Finally, as was already mentioned on this list, the "machine" symlink in the compile directory points to "//include"; it must be changed to "../../include" for BSD. Other than that, I can't remember having any other problems. But there _is_ one slight problem _after_ compiling: the generated vmunix doesn't work (for me anyway). It does precisely the same thing as vmunix.366, though I'm building from the .335 sources (actually .317 plus the stuff in "changes-366"). Again, any ideas? The vmunix compiled under BSD is about 2K shorter than either vmunix.335 or vmunix.366 (Markus's Amiga-compiled versions); I assume this is because BSD is using GCC 2.4.3 while I guess Markus used GCC 2.3.3 on the Amiga. Are there known bugs in 2.4.3 that might be breaking the kernel? Thanks! Bryan From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:51:26 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: ARGH! Help... Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: Date: Tue, 29 Jun 1993 15:48:56 -0400 From: ah360@yfn.ysu.edu (T. Caster) To: netbsd-admin@icecube.rain.com Subject: ARGH! Help... Reply-To: ah360@yfn.ysu.edu OK, first, I'm not no CS major nor a weenie, but i need some help please. First, my config: Amiga 3000 - 8Mb fast & 2 chip RAM 50Mb Quan on id6 and SyQuest 5110 on id0 OpalVision card and GVP MultiI/O extender OS 2.04/2.1 (mismas mixture of both) Problem: (Note: Have been following msg, but not seen answr) Goes into kernel partition ok, but says cannot mount root and syncing disks: trapbad kernel access at 15 Oh, I'm using kernel 366. One other note, I noticed some mention sync on the controller now my system said that is was async. If that's the prob., how do I change it to sync? I have no idea... Here's some more info... info... ... SCSI: async SCSI id 7 ... sd0: SyQuest ... 66700 ... ... Change root dev to sd0a bpf ... (X 5 of these) panic cannot mount root syncing disks... trap bad kernel access at 15 trap type 8 code=4010715 v15 pid=0 pc=00019302 ps 2004 sfc 0601 dfc 0001 ... (then goes into some registers, then does a core dump) I followed all install procedurs to the T and everything seemed OK, part and cp the rootfs and all. I did the whole thing twice. I plan on trying older kernels, but I don't think that's the problem. Since it said it had "syncing" problems, I think it's around there? Please, if anyone can help i'd appreciate it! Also, will there be a way to support GVP I/O Extender so I can add the 2 extra serial ports? On bootup I noticed ser0: [1/3] (or somthing like that) But since I couldn't get access, could see if it recognised. Later - TC -- Techno Caster - ah360@yfn.ysu.edu From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:53:07 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: Kernel built, but panics? Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: From: Alan Bair Subject: Kernel built, but panics? To: netbsd-amiga@icecube.rain.com (NetBSD on Amiga) Date: Tue, 29 Jun 93 12:25:54 CDT X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@icecube.rain.com I have managed to rebuild the kernel using the latest sys source (366?). My gcc, sh, gnumake, etc. setup still has a few problems, so I had to play some games with the Makefile, but I managed to get all the way through to a new vmunix. I did notice another problem with the Makefile built by config. For the target subr_mcount.o, it does not give any commands, like the ${NORMAL_C} that most everything else has, so it never compiled this file. I just added the ${NORMAL_C} to get past that. When I booted vmunix.ARB under 2.1 it came up with a "more" readable font. This time it was very tall, but still had the top and bottom pixels chopped off. It was easy to tell what the letters were suppose to be compared to what I had been getting under 2.1. If I boot under 1.3, I get what looks like a Topaz font, easy to read and I did not manage to build dumpfont and use it. So I don't think the default font is as bad as Markus makes it sound. In any case, on to the next problem. Either way I boot the kernel, it shows some intial information, then pauses while I guess it is probing the SCSI bus. Then it starts to display some more information and it looks like it paniced, since a register dump flys by on the display. This is immediately followed by a reboot of the Amiga. So the question is how do I get the display to slow down or better yet stop before it reboots. I need to see what is being displayed to know what caused the panic. If I can't figure it out, I imagine Markus would like to know what was in the register dump in order to provide some hints. By the way, I seem to be the only one asking about building the kernel. Has anyone else got it to work such that you can succesfully boot the new kernel? -- Alan Bair AMCU DSCS Motorola, Inc. (Design Software & Mail Stop OE-320 Computer Services) 6501 William Cannon Dr. West Austin, TX 78735-8598 abair@amcu-tx.sps.mot.com From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:53:44 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: whine :) Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: From: Juha Kallioinen Subject: whine :) To: netbsd-amiga@icecube.rain.com Date: Tue, 29 Jun 1993 17:07:28 +0200 (EET DST) X-Mailer: ELM [version 2.4 PL17] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 338 Sender: NetBSD-Admin@icecube.rain.com Err this might sound like whining, but I hate the colors of the console :) During those long nightly hours of hacking with NetBSD my poor eyes hurt from that glowing white (and small) text against dark background. I'm not even suggesting anything, just makeing a notice :) Keep up the good work. -Juha -- // kaltsi@vipunen.hut.fi From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:54:15 1993 Subject: Discalimer To: netbsd-amiga@cbmuucp.cbm.commodore.com X-Mailer: ELM [version 2.2 PL0] Sender: NetBSD-Admin@cbmuucp.commodore.com I thought I'd let people know that Commodore isn't officially endorsing Markus Wild's NetBSD port to the Amiga. I'm just letting the list live here because I'm a nice guy. Anyway, keep up the good work, guys! -Ken From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 03:54:49 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: the stuff I uploaded... Sender: NetBSD-Admin@cbmuucp.commodore.com Begin forwarded message: Date: Mon, 28 Jun 93 19:53:59 -0700 From: mykes@shell.portal.com (mike myke schwartz) To: netbsd-amiga@icecube.rain.com Subject: the stuff I uploaded... Sender: NetBSD-Admin@icecube.rain.com I uploaded it to wuarchive.wustl.edu /archive/systems/amiga/NetBSD/incoming: aterm.tar.z aterm.readme ascreen.tar.z ascreen.readme I spent a lot of time/work doing the aterm.readme file... Download it and have alook :-) From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 04:00:54 1993 To: NetBSD-Amiga@cbmuucp.commodore.com Subject: Back messages. Sender: NetBSD-Admin@cbmuucp.commodore.com Sorry for the flood, and I may have overshot what was spooled up, but I believe that should be everything that was posted on to the list before the change over. Your friendly NetBSD-Amiga mailing list coordinator. From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 06:16:10 1993 X-Mailer: //\\miga Electronic Mail (AmiElm 1.19) Content-Length: 702 To: netbsd-amiga@cbmuucp.commodore.com Subject: Re: whine :) Sender: NetBSD-Admin@cbmuucp.commodore.com > From: Juha Kallioinen > > Err this might sound like whining, but I hate the colors of the > console :) During those long nightly hours of hacking with NetBSD > my poor eyes hurt from that glowing white (and small) text against > dark background. Hey. Why not the greenscreen green-on-black look, for those of us who would like to think we're in the Golden Age of Hacking? :) -- Joseph Hillenburg (xterm), NPS Technologies, Bloomington, Indiana UUCP: jph@anaconda.Bloomington.IN.US \ / "My OTHER computer is a Cray III" +----o0o----+ Long live the Hornet Current Project: BezFly 3.0 (Why Surf when you can Fly?) From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 13:46:41 1993 Subject: Re: GVP support delayed... :-( To: billc@icecube.commodore.com (William J. Coldwell) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 591 Sender: NetBSD-Admin@cbmuucp.commodore.com > for vacation. I'll be back August. My AmigaDOS disk got corrupted > while I tried to add yet another SCSI disk to my bus. On the > corrupted > disk is my NetBSD devtree, so I can't do any work until I get it up > (the disk, I mean :-). :-(( > BTW. The 377 kernel on ftp.eunet.ch, why no source or readme, > Markus? Because there were too few changes, really. I'll distribute sources with the next release again. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 13:49:39 1993 Subject: Re: vmunix377 ps still broken. To: billc@icecube.commodore.com (William J. Coldwell) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 634 Sender: NetBSD-Admin@cbmuucp.commodore.com > I just installed vmunix377 and it seemed as if ps and w were working, > > but when going into multi user mode I still get: > ps:kvm_openfiles can't open /dev/drum chgrp kmem /dev/drum > By the way, how come I can't login into root from a user account, It > says my chown root /bin/su /bin/passwd chmod u+s /bin/su /bin/passwd Sorry bout those permission problems, I always used programs as root, and didn't realize there were wrong permission settings... -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 13:53:34 1993 Subject: Re: ps? To: billc@icecube.commodore.com (William J. Coldwell) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 776 Sender: NetBSD-Admin@cbmuucp.commodore.com > What's wrong with ps? Everything's ok the first few times but after > that all > I get is a long, long wait and finally a segmentation fault > ( trapsignal(96, 11, 4, 4, 63a2) ). > It doesn't matter what kernel I use. > I'm surpriced I haven't read anything about this anywhere, have I > missed > something in the docs? Never happend to me.. Well, I'll try to upgrade to the newest kernel sources this weekend, there seems t be a vital fix to the kmalloc-package, perhaps those memory-related problems can be fixed that way? > files. They can be found at: ftp.luth.se - "/pub/unix/NetBSD-Amiga" Cool! -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 13:59:24 1993 Subject: Re: vmunix.366 and compiling on NetBSD To: billc@icecube.commodore.com (William J. Coldwell) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1650 Sender: NetBSD-Admin@cbmuucp.commodore.com > First: vmunix.335 works fine for me, but I have been unable to get vmunix.335 is *evil*! Don't *ever* use any rsd* or rst* devices on it, or you'll trash all over... > vmunix.366 to work at all. After it loads and prints the available > memory on the AmigaDOS window, it blanks the screen to white (OK), > and remains that way forever (not OK). Any ideas? The screen should never ever blank white.. Hmmmmm (noticed config bugs) config will probably change considerably soon, we got some new nice config code :-) > But there _is_ one slight problem _after_ compiling: the generated > vmunix doesn't work (for me anyway). It does precisely the same > thing > as vmunix.366, though I'm building from the .335 sources (actually > .317 plus the stuff in "changes-366"). Again, any ideas? Weird.. Well, I've never seen something like that, perhaps it would be a good idea to insert some serspit('a') to serspit('z') into the sources along the way it takes while booting, and watching on a serial console how far it gets, and WHERE it's causing this white screen. You need a serial terminal for this though. > The vmunix compiled under BSD is about 2K shorter than either > vmunix.335 or vmunix.366 (Markus's Amiga-compiled versions); I assume > this is because BSD is using GCC 2.4.3 while I guess Markus used > GCC 2.3.3 on the Amiga. Are there known bugs in 2.4.3 that might > be breaking the kernel? None that I'd know of.. I'm recompiling 2.4.5 right now, we'll see.. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 3 15:55:12 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: Sender: NetBSD-Admin@cbmuucp.commodore.com SUBSCRIBE ingemar jansson From NetBSD-Admin@cbmuucp.commodore.com Sun Jul 4 01:29:32 1993 Subject: amigados emulator To: netbsd-amiga@icecube.rain.com (NetBSD mailing list NetBSD mailing list) X-Mailer: ELM [version 2.4 PL21] Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8 Content-Length: 514 Sender: NetBSD-Admin@cbmuucp.commodore.com I have changed couple of lines from the Sun3 hack of original amigados emulator for C= SysVr4 and produced emulator seems to run as well as it is written ;-) Only whish now would be that we had a source level debugger for NetBSD :-/ I had no luck to upload this to ftp.eunet.ch because ftp complained the destination directory (incoming) to be in read only filesystem :-( Currently the packet can be found from this address: phoenix.oulu.fi:/pub/amiga This will be removed at some time in near future.. ++Tero From NetBSD-Admin@cbmuucp.commodore.com Sun Jul 4 02:33:39 1993 Subject: safe? To: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@cbmuucp.commodore.com Is NetBSD safe enough to run? ie., not prone to harddrive trashing. Do I need a Unix partition, or can I just use a fake partition (ala Bridgeboard)? Is 8MB ram, '030/25, and 50MB HD partition comfortable enough to run the barebones system? thanks, -sam From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 5 05:49:28 1993 Subject: Booting OK on SyQuest ID 5, Tape ID 2 To: netbsd-amiga@cbmuucp.commodore.com (NetBSD on Amiga) X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@cbmuucp.commodore.com Well I finally built a kernel that boots off my SyQuest with ID 5. My WangTek tape at ID 2 also works just fine. The latest kernel vmunix.377 still does not boot with the above setup. As with all the others and my previous attempts it just ignores my SyQuest at ID 5. The "problem" is in the description of SCSI device IDs for the system to use. I edited the sys/arch/amiga/compile/ARB/ioconf.c file to look like this. struct amiga_device amiga_dinit[] = { /*driver, cdriver, unit, ctlr, slave, addr, dk, flags*/ { &serdriver, D 0x0, 0, -1, -1, C 0x10003, 0, 0x0 }, { &sddriver, &scsidriver, 5, 0, 5, C 0x0, 1, 0x0 }, { &sddriver, &scsidriver, 6, 0, 6, C 0x0, 1, 0x0 }, { &stdriver, &scsidriver, 0, 0, 2, C 0x0, 0, 0x0 }, { &grfdriver, D 0x0, 0, -1, -1, C 0x10007, 0, 0x0 }, 0 }; The key is to have a slave address listed only once and to match the real hardware. The original copy of this file built by config had all the IDs for the disks 0-6, but it also had IDs for tapes at 4 & 5. Since these were latter in the list, I suspect they overlayed the 4 & 5 disk IDs, which took out my SyQuest at ID 5. I have not tried building this kernel again with the changes in the sys/arch/amiga/conf/ARB file to reflect the same devices as above, but I expect it will work. I had tried some changes in the config file previously, but I guess I never took out the tape entry on ID 5, so I never uncovered my SyQuest. I had tried some changes to the tape ID, but never got past the boot to know it was working. As mentioned above the tape at ID 2 worked just fine. I tared off the /etc directory and then untared it under AmigaDOS with no errors. Markus, could you setup your kernel builds to allow disks at 0, 1, 2, 3, 5 & 6 and a tape at 4. This should still fit what you have stated for supported setups, but allow me to boot your released kernels. -- Alan Bair AMCU DSCS Motorola, Inc. (Design Software & Mail Stop OE-320 Computer Services) 6501 William Cannon Dr. West Austin, TX 78735-8598 abair@amcu-tx.sps.mot.com From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 5 13:06:28 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: netbsd-amiga@cbmuucp.commodore.com Subject: Re: vmunix377 ps still broken. Sender: NetBSD-Admin@cbmuucp.commodore.com On Jul 3, 1:34pm, Markus Wild wrote: > Subject: Re: vmunix377 ps still broken. [...] > Sorry bout those permission problems, I always used programs as root, and > didn't realize there were wrong permission settings... Hm, wasn't there the golden rule: Never be supseruser unless you actually want to administrate something ? :-) Even on the Amiga, i believe it's not that good always to be superuser, and it looks bad in mails, too ... *grin* -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 5 19:26:57 1993 Subject: Re: Booting OK on SyQuest ID 5, Tape ID 2 To: abair@parsons.sps.mot.com (Alan Bair) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 733 Sender: NetBSD-Admin@cbmuucp.commodore.com > Markus, could you setup your kernel builds to allow disks at 0, 1, 2, 3, 5 & 6 > and a tape at 4. This should still fit what you have stated for supported > setups, but allow me to boot your released kernels. I might also try to make the tape fully wildcarded? Ie. it will use the first unit that identifies as tape? Experiments for later releases, I just uploaded 390, and this won't have any such changes. Chances are though that 16M machines might finally work, as I replaced a file (vfs__bio.c) that did alot of those TEMP malloc's that caused the 16M hanger. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 5 19:39:04 1993 Subject: Re: vmunix377 ps still broken. To: markus@techfak.uni-bielefeld.de (Markus Illenseer) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 518 Sender: NetBSD-Admin@cbmuucp.commodore.com > Hm, wasn't there the golden rule: Never be supseruser unless you > actually want to administrate something ? :-) What else do I usually do these days.. :-)) > Even on the Amiga, i believe it's not that good always to be superuser, > and it looks bad in mails, too ... *grin* Oh, lets talk about that again when I run uucp and/or sendmail ok? :-)) -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 5 19:50:13 1993 Subject: Slip To: netbsd-amiga@cbmuucp.commodore.com Content-Type: text Content-Length: 890 Sender: NetBSD-Admin@cbmuucp.commodore.com I know that this has been asked before, and it MAY not be a NetBSD Specific enough question to ask here, BUT... How do I get NetBSD to do slip... I am not familiar enough with network set-up to do it by myself, and nobody seems to want to answer this question... So far the only 2 commands I have any clue about are ifconfig and route. And I don't REALLY know how to work either of them... I have tried all sorts of "ifconfig sl0 ..." commands, and route ALWAYStells me that it can't open such-and-such socket and aborts... Also "ping" always tells me "No route to host" except if I ping localhost. I need help. BTW, I am SOOOOO impressed with NetBSD. It crashed on me for the first time in 5 days today! (I NEVER would've thought that I could make it last that long before I had seen it. Thanks Markus) -- Kevin McCarthy signals@krypton.mankato.msus.edu From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 5 22:23:25 1993 Subject: uploaded #390 kernel/sources To: netbsd-amiga@cbmuucp.commodore.com (nba) X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 428 Sender: NetBSD-Admin@cbmuucp.commodore.com Just uploaded new pair of vmunix/bsdsyssrc #390. Changes are mostly added prototypes to the mach-indep parts, and a move to a different vfs_bio source. This *might* eliminate the weird hanger on machines with 16M, so those of you with 16M, please check it out. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 6 07:38:35 1993 Subject: Fix for copper bug To: netbsd-amiga@cbmuucp.commodore.com (NetBSD on Amiga) X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@cbmuucp.commodore.com Well, I think I have a fix for the weird display problem some of use were having in which the characters end up managled in various ways. I have built two kernels, with the only differnce being the fix I will describe below and one has the problem, the other doesn't. I also noticed for the first time the problem where the last character on a line is gone, others have already reported this. I'll see if I can track this one down too. Now on to the fix. The problem is with the copper STOP macro definition in sys/arch/amiga/dev/grf_cc.c. It sets the low order bit to 1 when it should always be 0. There are also some problems with the WAIT macro, but it never appears to be used (yet) so it causes no problems. Here is the original code with my updates intermixed. ------------------------------------------------------------------------- /* copper instructions */ /* ARB Hopefully the address below fits: bits 17-24 with others zero */ #define M0VE(cl, reg, val) \ do { *cl++ = CUST0M_0FS(reg); *cl++ = val; } while (0) /* ARB Major variable usage error & bit settings #define WAIT(cl, vp, hp, bfd, ve, he) \ do { *cl++ = ((vp & 0xff)<<8)|(hp & 0xfe)|1; \ *cl++ = (bfd<<15)|((ve & 0x7f)<<8)|(hp & 0xfe)|1; } while (0) */ #define WAIT(cl, vp, hp, bfd, ve, he) \ do { *cl++ = ((vp & 0xff)<<8)|(hp & 0xfe)|1; \ *cl++ = (bfd<<15)|((ve & 0x7f)<<8)|(he & 0xfe)|0; } while (0) /* ARB Not exactly correct #define ST0P(cl) \ do { *cl++ = 0xffff; *cl++ = 0xffff; } while (0) */ #define ST0P(cl) \ do { *cl++ = 0xffff; *cl++ = 0xfffe; } while (0) ------------------------------------------------------------------------- (Yeah, I know, this should have been in patch format, but I'm lazy) For those of you that didn't like the colors of the display, here is a hint to let you fix this yourself. In the file sys/arch/amiga/dev/grf_ccreg.h change the values of DEF_COL0 and DEF_COL1 to your favorite background and foreground colors. Just make the changes, rebuild and enjoy a nice readable display. Now I don't need to go through AmigaDOS 1.3 to get a good display. I have no idea why this change works since that bit in the copper instruction is not explained. Another one of those mysterious hardware dependencies :) Oh, now that I am booting from AmigaDOS 2.1, the 3 finger reset works fine. It was not working when booted from 1.3. PS: Markus, don't label be as a copper expert. I just got tired of the display problem and started reading the code and 3-4 manuals I have. Then I stumbled on the STOP macro error. At first I thought, No way that could be it, but I gave it a shot and surprise it worked. -- Alan Bair AMCU DSCS Motorola, Inc. (Design Software & Mail Stop OE-320 Computer Services) 6501 William Cannon Dr. West Austin, TX 78735-8598 abair@amcu-tx.sps.mot.com From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 6 07:42:20 1993 Subject: Re: Kernel built, but panics? To: oster@skorpio.usask.ca Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@cbmuucp.commodore.com > > > Hello all! > > > First of all: A BIG THANK YOU to MARKUS for all his GREAT WORK!!!! > > > From: Alan Bair > > Subject: Kernel built, but panics? > > > > This sound familiar... Which version of GCC are you using? (I'm > assuming that > 2.3.3 is the current version - there seemed to be some problems when > I tried > using 2.2.2...) I have 2.3.3 with the lib fixes. Make sure you put the AmigaDOS includes under the gcc: tree. I hadn't done this and ran into problems with missing includes to compile NetBSD. > > > > So the question is how do I get the display to slow down or better > yet stop > > before it reboots. I need to see what is being displayed to know > what caused > > the panic. If I can't figure it out, I imagine Markus would like to > know > > what was in the register dump in order to provide some hints. I don't think I mentioned this in my other posting about how I got a bootable kernel, but to slow things down I just added a bunch of printfs with a few DELAY(10000000) interspursed. This gave me the information I needed and allowed me to jot it down. > > How much Fast RAM do you have? I've heard there were/are some > problems if > you have 16 MB. What "options" did you leave out when you > re-compiled the I only have 8M FAST, so I don't need the loadbsd -k kludge yet :) I left most of the options as they came from Markus. Now that I have a booting kernel, I'll try pruning things out. > kernel. Markus mentioned a couple of them in the "RECOMPILE" text > file (ISOFS is the only one I've removed). The other thing to check > is that the SCSI ID's in the configuration file > /sys/arch/amiga/conf/AMIGA (I think) are setup for your > configuration, > and that the "config" line of that file is what you want. I described the problem and fix in my other posting. However, even though it now boots, I cannot get into multi-user mode. I have fixed the fstab to match my setup. Loaded up the usrbin.tgz files and checked the partitions with fsck. (I keep getting a bad block count on the same inode in root after remounting, but I clear that manually with no apparent problems) When I "exit" the main shell, the multi-user startup begins and it complains about alls kinds of missing commands. Most of these seem to be for daemons that don't really matter now; networking, printer, etc. Then it gives the date and time like it was about to finish and then stops. A little while latter I get a message like the following: init FATAL error: console: interrupted system call It will do this several times. I then try the 3 finger reboot, but only get an almost black screen. It takes a power down to reboot to AmigaDOS. Anybody got any hints? -- Alan Bair AMCU DSCS Motorola, Inc. (Design Software & Mail Stop OE-320 Computer Services) 6501 William Cannon Dr. West Austin, TX 78735-8598 abair@amcu-tx.sps.mot.com From NetBSD-Admin@cbmuucp.commodore.com Wed Jul 7 05:18:33 1993 Subject: Re: Kernel built, but panics? To: abair@parsons.sps.mot.com (Alan Bair) Cc: oster@skorpio.usask.ca, netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL0] Content-Type: text Content-Length: 555 Sender: NetBSD-Admin@cbmuucp.commodore.com > don't really matter now; networking, printer, etc. Then it gives the date and > time like it was about to finish and then stops. A little while latter I get > a message like the following: > > init FATAL error: console: interrupted system call > > It will do this several times. I then try the 3 finger reboot, but only get > an almost black screen. It takes a power down to reboot to AmigaDOS. Anybody > got any hints? You need the usrinclib.tgz(?) archive which has /usr/libexec/getty, among other things. I had this same problem too. - Kyle From NetBSD-Admin@cbmuucp.commodore.com Wed Jul 7 23:21:14 1993 Subject: Multiuser login problems To: kxs5829%ultb.isc.rit.edu@uunet.uucp.commodore.com (K.X. Saunders) Cc: netbsd-amiga@cbmuucp.commodore.com (NetBSD on Amiga) X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@cbmuucp.commodore.com > > time like it was about to finish and then stops. A little while latter I get > > a message like the following: > > > > init FATAL error: console: interrupted system call > > > > It will do this several times. I then try the 3 finger reboot, but only get > > an almost black screen. It takes a power down to reboot to AmigaDOS. Anybody > > got any hints? > > You need the usrinclib.tgz(?) archive which has /usr/libexec/getty, > among other things. I had this same problem too. > > - Kyle > Thanks. I actually ran across your solution by accident when I started loading up the /usr files, but was not aware of why it worked. Now it gets past that, but the login prompt is badly managled and I can only login in as mw. The prompt ends up being multiple lines, spaced across the display and with missing or rearranged letters. However it does accept input and comes back with the normal password prompt. I have tried changing the root password via passwd and editing /etc/passwd, but the login process will not accept it. Anyone have any hints on this weirdness? PS: Is there a problem with the mailing list? The amount of traffic seems to have slowed down and the response time seems to be a day or two compared to several hours last week. If people are just busy, OK, but this also corresponds with the new site switch over, so I thought I would mention it. -- Alan Bair AMCU DSCS Motorola, Inc. (Design Software & Mail Stop OE-320 Computer Services) 6501 William Cannon Dr. West Austin, TX 78735-8598 abair@amcu-tx.sps.mot.com From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 00:26:06 1993 Subject: diff 390-394 and misc.. :-/ To: netbsd-amiga@icecube.rain.com (NetBSD mailing list NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 1985 Sender: NetBSD-Admin@cbmuucp.commodore.com Ok, I just compiled 394 kernel, after some hand made corrections after automatic diff. Namely arch/amiga/dev/grf_ccreg.h and ite.c missed couple of changes (generated the *.rej files). Then ofcourse the arch/amiga/compile/AMIGA/vers.c and version couldn't be patched because I had done some own compilations and the version numbers were not even close what patch was looking for ;-) What comes to colours.. I changed the background to all black and as some mentioned before.. the text colour to green like in the old days :-) I have been running the 390 kernel with 16M for some time now without any ill effects. There is still problems with w command. It shows nothing or garbage to the place where it should show the currently run program at each tty (like -tcsh or trsh, or..). BTW, the correction to /etc/gettytab that was proposed in comp.unix.amiga works :-) Now I saw for the first time a correct login prompt! So, change the default:\ :ap:... (or what was it) to default:\ :np:... (like no-parity) One more thing.. Is there a problem with quite small tar files in tape? I have now been compiling the kernel in netbsd and transferred the compiled kernel to amigados via tape, but in the amigados side I can't read the vmunix tar out of tape unless I make many copies of it.. tar cvf /dev/rst0 vmunix vmunix.2 vmunix.3 (different names because later extractions would overwrite previous, and the last file seems to be always corrupted ?!). I think tar writes eof marks to tape, but I tried to make those by hand too (mt -f /dev/nrst0 weof). With only one vmunix (tar cvf /dev/rst0 vmunix) amigados tar and tape-handler don't find anything from tape, even when I have written multiple tars after each other tar cf /dev/rst0 vmunix; tar cf /dev/nrst0 vmunix; tar cf /dev/nrst0 vmunix; My tape is Archive Viper and there is no problem reading the written tape in the unix side, just in the amigados... Also, I can move amigados written tapes easily to unix... ++Tero From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 02:04:52 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: netbsd on 2000 with 2091 Sender: NetBSD-Admin@cbmuucp.commodore.com I've been looking at the code to see what needs to be done to get NetBSD up with a 2000 and a 2091 controller. The first thing that comes up is the locations of the chips and controller. Can someone with a hardware reference tell me the locations of: CIAAbase = 0xbfe001 CIABbase = 0xbfd000 CUSTOMbase = 0xdff000 chipmem_start = 0x400 the values i listed are those that Markus already is using for the 3000. Did i miss anything? As for getting the address of the controller. There's alot of code in autoconf.c that is never being called. Actually the whole cpu_start() routine in machdep.c doesn't seem to get called. Markus, it looks like you were planning on doing dynamic configuration of the kernel. Has this idea been scrapped? The ConfigDev structure is being passed to start_c(), so i can get the 2091 address from there. But is the dma and everything else the same between the built in 3000 scsi and the 2091? Is there anything else i should be looking to change? I might try to get the dynamic configuration stuff working so we could boot all amiga's. thanks, ...alan ------------------------------------------------------------------------------- Alan Commike Landmark Graphics, Houston Tx. commike@lgc.com (713) 560 - 1249 The ultimate error correcting compiler would generate a Database Management System out of ... nothing! - Josef Moellers ------------------------------------------------------------------------------- From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 03:32:52 1993 Subject: Re: Multiuser login problems To: abair@parsons.sps.mot.com (Alan Bair) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL0] Content-Type: text Content-Length: 917 Sender: NetBSD-Admin@cbmuucp.commodore.com > Thanks. I actually ran across your solution by accident when I started > loading up the /usr files, but was not aware of why it worked. > > Now it gets past that, but the login prompt is badly managled and I can only > login in as mw. The prompt ends up being multiple lines, spaced across the > display and with missing or rearranged letters. However it does accept input > and comes back with the normal password prompt. I have tried changing the > root password via passwd and editing /etc/passwd, but the login process will > not accept it. Anyone have any hints on this weirdness? This is the parity problem I mentioned in news. The default entry in /etc/gettytab has ':ap:' which is "any parity". You want to change that to ':np:' for "no parity". The reason why the password prompt is normal is because getty prompts for the login name and then exec's login with the name as an argument. - Kyle From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 04:02:20 1993 Subject: vmunix.394 To: netbsd-amiga@cbmuucp.commodore.com Content-Type: text Content-Length: 1203 Sender: NetBSD-Admin@cbmuucp.commodore.com I just thought I would say that vmunix.394 (as distributed) DOES NOT work on my A3000(25/6megs ram) The previous versions did. I have my Hard drive full with BSD partitons, and haven't got room for a compiler. (It's a small drive) so I don't have any clue why it's broke. Previous versions would crash on me if I didn't fill up my chip memory with something before I ran it. 394 doesn't run no matter what I try. It also crashes different than other versions. The old kernels would crash to a blu screen with rolling dark grey bars on it. The new kernel just sitts at a light grey screen. No movement or anything. I'm sorry if This isn't much help, I am no kernel hacker, but there is something wrong in the differences between 390 and 394... -- Kevin McCarthy signals@krypton.mankato.msus.edu ------------------------------------------------------------------------------- "If I were a carpenter, I'd hammer on my piglet, I'd collect the $7 and I'd buy a big prosthetic forehead and wear it on my real head." -They Might Be Giants ------------------------------------------------------------------------------- HELLO! I'm a .signature virus! Join in and copy me into yours! From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 04:08:58 1993 Subject: Re: vmunix.394 To: signals@krypton.mankato.msus.edu (Kevin M Mccarthy) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL0] Content-Type: text Content-Length: 957 Sender: NetBSD-Admin@cbmuucp.commodore.com > > I just thought I would say that vmunix.394 (as distributed) DOES NOT work on > my A3000(25/6megs ram) The previous versions did. I have my Hard drive full > with BSD partitons, and haven't got room for a compiler. (It's a small drive) > so I don't have any clue why it's broke. Previous versions would crash on me if > I didn't fill up my chip memory with something before I ran it. 394 doesn't > run no matter what I try. It also crashes different than other versions. The > old kernels would crash to a blu screen with rolling dark grey bars on it. The > new kernel just sitts at a light grey screen. No movement or anything. I'm > sorry if This isn't much help, I am no kernel hacker, but there is something > wrong in the differences between 390 and 394... Well, not to rub it in or anything like that, but I'm happy to say that 394 fixed the display problems the previous versions had. I was suprised to see we had a mouse pointer. - Kyle From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 05:13:29 1993 To: commike@lgc.com Subject: Re: netbsd on 2000 with 2091 Cc: netbsd-amiga@cbmuucp.commodore.com Sender: NetBSD-Admin@cbmuucp.commodore.com Alan said: >The ConfigDev structure is being passed to start_c(), so i can get the >2091 address from there. This is needed for the generic 5380 driver that I'm doing. I've been working from the driver end back rather than tackling start_c(). >Is there anything else i should be looking to change? I might try to get the >dynamic configuration stuff working so we could boot all amiga's. Please do. Life would be much nicer. From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 12:08:52 1993 Subject: Re: diff 390-394 and misc.. :-/ To: nix@stekt.oulu.fi (Tero Manninen) Cc: netbsd-amiga@icecube.rain.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1054 Sender: NetBSD-Admin@cbmuucp.commodore.com > Namely arch/amiga/dev/grf_ccreg.h and ite.c > missed couple of changes (generated the *.rej files). Hmm... strange... I thought I did the diffs against 390... > There is still problems with w command. It shows > nothing or garbage to the place where it should show > the currently run program at each tty (like -tcsh or > trsh, or..). You DID update your /vmunix when using new kernels? Don't forget w and ps have to lookup the symbol offsets in the /vmunix image to know where to seek in /dev/kmem. > Is there a problem with quite small tar files in tape? No.. shouldn't... perhaps you should try using the B option to tar. > but in the amigados side I can't read the vmunix tar > out of tape unless I make many copies of it.. > tar cvf /dev/rst0 vmunix vmunix.2 vmunix.3 > (different names because later extractions would overwrite That's very strange... Tapes are weird :-) -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 12:56:39 1993 Subject: Re: vmunix.394 To: signals@krypton.mankato.msus.edu (Kevin M Mccarthy) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Content-Length: 847 Sender: NetBSD-Admin@cbmuucp.commodore.com Hello > 394 doesn't > run no matter what I try. It also crashes different than other versions. The > old kernels would crash to a blu screen with rolling dark grey bars on it. The > new kernel just sitts at a light grey screen. No movement or anything. I'm > sorry if This isn't much help, I am no kernel hacker, but there is something > wrong in the differences between 390 and 394... > > -- > Kevin McCarthy signals@krypton.mankato.msus.edu I have problems with 394, too. If I start NetBSD with AmigaDOS fully started, I get the same effect - only a grey screen. But without Startup-Sequence NetBSD starts without problem. I never had problems with older versions of NetBSD. ---------------------------------------------------------------------- Michael Suelmann suelmann@forwiss.uni-passau.de From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 13:10:05 1993 Subject: Re: netbsd on 2000 with 2091 To: commike@lgc.com Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 576 Sender: NetBSD-Admin@cbmuucp.commodore.com > As for getting the address of the controller. There's alot of code in > autoconf.c that is never being called. Actually the whole cpu_start() routine > in machdep.c doesn't seem to get called. Markus, it looks like you were > planning on doing dynamic configuration of the kernel. Has this idea been > scrapped? I *will* do that, I just currently lack the resource-map calls, but they should be ready RSN. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 13:50:52 1993 Subject: Re: diff 390-394 and misc.. :-/ To: mw@eunet.ch (Markus Wild) Cc: netbsd-amiga@icecube.rain.com (NetBSD mailing list NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Content-Length: 1019 Sender: NetBSD-Admin@cbmuucp.commodore.com > > There is still problems with w command. It shows > > nothing or garbage to the place where it should show > > the currently run program at each tty (like -tcsh or > > trsh, or..). > > You DID update your /vmunix when using new kernels? Don't forget w and > ps have to lookup the symbol offsets in the /vmunix image to know where > to seek in /dev/kmem. I think I did.. I'll have to copy the 394 kernel to root when I boot next time (this evening). I'll see then if the -O2 -funroll-loops has any effect to any direction :-) How about the other kernel names, the "netbsd" and (what was the other) links that pointed to /vmunix, are these needed? (I think one utility once needed /netbsd ...). I don't mind.. but these are ugly links in the root ;-) > > Is there a problem with quite small tar files in tape? > > No.. shouldn't... perhaps you should try using the B option to tar. As it appears, I'm not the only who has the same problem, and even hardware is different (my Viper and the other A3070). ++Tero From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 16:56:59 1993 Subject: Re: netbsd on 2000 with 2091 To: commike@lgc.com Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 3140 Sender: NetBSD-Admin@cbmuucp.commodore.com > I've been looking at the code to see what needs to be done to get NetBSD up > with a 2000 and a 2091 controller. The first thing that comes up is the > locations of the chips and controller. Can someone with a hardware reference > tell me the locations of: > > CIAAbase = 0xbfe001 > CIABbase = 0xbfd000 > CUSTOMbase = 0xdff000 > chipmem_start = 0x400 I kludged together a kernel that works on my A2500/30 (2630+2091). It does not autoconfig properly, so it assumes that the A2091 is adressed at 0xe90000. As far as I can see, the CUSTOMbase area covers both the 3091 and the ZorroII I/O area, so remapping is not necessary. > > the values i listed are those that Markus already is using for the 3000. > Did i miss anything? > > > As for getting the address of the controller. There's alot of code in > autoconf.c that is never being called. Actually the whole cpu_start() routine > in machdep.c doesn't seem to get called. Markus, it looks like you were > planning on doing dynamic configuration of the kernel. Has this idea been > scrapped? I did add code to recognize an autoconfig'ed 2091 in autoconfig.c and put it in the list of discovered devices, but I have not hacked on the code to match found devices to configuration tables. > > The ConfigDev structure is being passed to start_c(), so i can get the > 2091 address from there. But is the dma and everything else the same between > the built in 3000 scsi and the 2091? > The DMA registers are at different relative addresses and the bits in the DMA control registers are in different places. We need to add conditional code in the DMA drivers wherever registers are accessed. Otherwise everything is the same. > Is there anything else i should be looking to change? I might try to get the > dynamic configuration stuff working so we could boot all amiga's. There are three problems I see with the current driver setup: 1) There is no provision for determining whether we are running on an A3000 or another type of amiga. This means that we don't know if the A3091 driver is valid and should be configured as the primary driver. This information should be passed in by loadbsd. 2) The 3091 driver is hardwired into the kernel and is configured before other devices are even polled. This must be changed. 3) The way the 3091 driver uses global variables means that it is probably not reentrant, and we need a reentrant driver if anyone has an A3000 and a 2091 and want to use both (e.g. for more than 8 drives, etc.). If anyone can explain to me how memory is being mapped in the current (and future) NetBSD kernel I would greatly appreciate it. There currently seems to be 3 different sets of MMAP code, only one of which is in use. None of this is documented, and there is no indication of future trends so I have no idea how to add a driver that will remain compatible with whatever Markus is doing. Oh, BTW, the kernel I've worked on is 366. ========================================================================= Eduardo Horvath eeh@btr.com ..!{decwrl,mips,fernwood}!btr!eeh "Trust me, I am cognizant of what I am doing." - Hammeroid From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 17:35:28 1993 Subject: Re: diff 390-394 and misc.. :-/ To: mw@eunet.ch (Markus Wild) Cc: netbsd-amiga@cbmuucp.commodore.com (NetBSD on Amiga) X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@cbmuucp.commodore.com > > > Namely arch/amiga/dev/grf_ccreg.h and ite.c > > missed couple of changes (generated the *.rej files). > > Hmm... strange... I thought I did the diffs against 390... I suspect your memory is correct, I had no problems running patch against the 390 sources. No I need to start playing with the makefile again to get it to work under my gcc environment :( > > You DID update your /vmunix when using new kernels? Don't forget w and > ps have to lookup the symbol offsets in the /vmunix image to know where > to seek in /dev/kmem. I thought we would need to do that, now you have confirm it. If I failed to do this, would it cause the failure of kmem_openfiles to access nswap? -- Alan Bair AMCU DSCS Motorola, Inc. (Design Software & Mail Stop OE-320 Computer Services) 6501 William Cannon Dr. West Austin, TX 78735-8598 abair@amcu-tx.sps.mot.com From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 17:38:06 1993 Subject: Re: diff 390-394 and misc.. :-/ To: abair@parsons.sps.mot.com (Alan Bair) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 621 Sender: NetBSD-Admin@cbmuucp.commodore.com > > You DID update your /vmunix when using new kernels? Don't forget w and > > ps have to lookup the symbol offsets in the /vmunix image to know where > > to seek in /dev/kmem. > > I thought we would need to do that, now you have confirm it. If I failed to > do this, would it cause the failure of kmem_openfiles to access nswap? No, that was another problem. /dev/kmem did some validity checks that didn't work. Hopefully, this problem should be gone. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 18:31:45 1993 Subject: Re: diff 390-394 and misc.. :-/ To: abair@parsons.sps.mot.com (Alan Bair) Cc: netbsd-amiga@icecube.rain.com (NetBSD mailing list NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 601 Sender: NetBSD-Admin@cbmuucp.commodore.com > > > Namely arch/amiga/dev/grf_ccreg.h and ite.c > > > missed couple of changes (generated the *.rej files). > > > > Hmm... strange... I thought I did the diffs against 390... > > I suspect your memory is correct, I had no problems running patch against the > 390 sources. No I need to start playing with the makefile again to get it > to work under my gcc environment :( The difference may be that I compile my kernel in NetBSD :-) Actually it souldn't matter.. but anyway, I managed to put the rejected patches into sources by hand and the thing even compiled (now I have to try it..). ++Tero From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 21:59:07 1993 Subject: Please change termcap file... To: netbsd-amiga@cbmuucp.commodore.com (nba) X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 479 Sender: NetBSD-Admin@cbmuucp.commodore.com At least if you're using emacs, the current termcap entry for the vt200 (and vt220, vt100. xterm is fine though) is horrible. Please remove /etc/termcap, and copy over /usr/gnu/lib/emacs/19.11/etc/termcap.dat into /etc/termcap instead. Works MUCH better! (So mykes, the bug wasn't in ite.c, but in termcap :-)) -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 8 23:27:02 1993 Subject: gtroff/libg++ on eunet To: netbsd-amiga@cbmuucp.commodore.com (nba) X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 537 Sender: mw@magnolia.eunet.ch Sender: NetBSD-Admin@cbmuucp.commodore.com I just uploaded libg++ and gnu-troff binaries to ftp.eunet.ch, for those of you that wanted to format man pages :-) I hope I got the whole installation, just packaged from /usr/gnu what was newer then a given date. I'll try to make diff's to the standard GNU packages over the weekend, so who wants can recompile the whole thing (reserve some time for it though :-)) -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 9 21:21:27 1993 Subject: bootblock code To: netbsd-amiga@icecube.rain.com (NetBSD mailing list NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 1188 Sender: NetBSD-Admin@cbmuucp.commodore.com Here we are, I asked some time before about possibility to boot directly to netbsd and now I have looked into it a bit.. I did some experiments by reading data out of my disk that has the root partition (/dev/rsd4c). As it reads in there is some space already reserved to the boot code (8192 bytes from start of root partition, and then comes 8192 bytes of super block data). Loading of the /vmunix shouldn't be too big task once you know the file system structure, which I don't do :-( Also, I don't know much about the boot procedure that AmigaOS does when it finds a bootable partition, but I guess it is much like floppy boot (i.e. you can't do dos.library calls, but anything else goes, so you should read the kernel data by exec DoIO() & co. functions (once you have opened the scsi.device)). Starting the loaded vmunix is easy, just do what loadbsd.c does :-) Then there is the question, how do you find the partition where the bootcode itself is? Is the boot partition marked any special way in filesystem resources? (OpenResource()). Maybe I should look for bootcode from other platforms to an example.. (as I did.. i386 code is fine (for the C-part :-)) ++Tero From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 10 02:23:30 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: So what _IS_ NetBSD? Sender: NetBSD-Admin@cbmuucp.commodore.com Ok, I subscribe to the mailing list and I've seen posts about it in comp.unix.amiga, but what is NetBSD? Is it a port of 386BSD to the amiga, or something similar? Can I blow away my unsupported and semi-trashed AmigaUnix 2.05Beta partition and run NetBSD instead? Any info would be appreciated! Thanks!! -Scott L. Scott Emmons CableData Research Center uunet!cdsac!scotte KC6NFP From c5039 Mon Jul 19 20:33:24 1993 Date: Mon, 19 Jul 93 20:33:24 +0200 From: Harald Backert To: c9020 Subject: [hubert.feyrer@rrzc1.rz.uni-regensburg.de: BI-Mailing-List] Status: RO Hallo Hubert, > Irgendwie hab ichs geschafft, meine gesamte Mail zu loeschen! > Koenntest Du mir die Info von Markus wg. der BI-Mailing-List > nochmal zuschicken? Ich tu die in Dein tmp-Verzeichnis, das ist einfacher fuer mich :-) > Im uebrigen, wo ist H45 in der Chemie?! Zwischen den beiden grossen Chemie-Hoersaelen (H43, H44). Also wenn Du an der Chemie-Cafete vorbei gehst, dann noch etwa 10m. > Wie geht's NetBSD? Den 490 Kernel habe ich gerade gesaugt, s'is > aber kein src da, nicht witzig! Im uebrigen habe ich jetzt neben > meinem Evolution-Controller einen A2091 stecken :-) Gestern von etwa 21:00 Uhr bis heute um 2 Uhr haben wir (mtk, ill und ich) Testversionen vom NetBSD getestet, geaendert und geprueft. Heraus gekommen ist ein NetBSD, das schon ziemlich weit kommt. 2091 ist da, alle Platten werden erkannt, auch ser0 und grf0, aber dummerweise crasht meine Kiste, sobald vmunix versucht irgendwas mit dem rootfs anzufangen. Es kommen zwar kurz irgendwelche Meldungen, aber die sind so schnell weg, dass man die nicht lesen kann. Dementsprechend mager ist die Fehlerkorrekturmoeglichkeit, weil niemand genau weiss, warum vmunix crasht. Markus will heute noch die kompletten neuen Quelltexte auf ftp.eunet.ch tun, damit evt. andere den Fehler finden koennen :-) Da Du jetzt einen 2091 hast (Herzlichen Glueckwunsch!), kannst Du auch mitmachen :-) Tschuess, Harald From eeh@public.btr.com Tue Jul 20 16:51:48 1993 Received: from public.btr.com by btr.com (4.0/SMI-4.0-btr.1) id AA24566; Tue, 20 Jul 93 07:50:37 PDT Received: by public.btr.com (4.1/SMI-4.1.public.1) id AA16120; Tue, 20 Jul 93 07:50:34 PDT Date: Tue, 20 Jul 93 07:50:34 PDT From: eeh@public.btr.com (Eduardo E. Horvath eeh@btr.com) Message-Id: <9307201450.AA16120@public.btr.com> To: feyrer@rrzc1.rz.uni-regensburg.de Subject: Re: Success! (was: Re: NetBSD and A2000/2630) Newsgroups: comp.unix.amiga Organization: BTR UNIX BBS and Email hub. For info about BTR contact support@btr.com X-Newsreader: TIN [version 1.2 PL0] Status: RO In article <22gt3qINN5pm@rrzs3.uni-regensburg.de> you wrote: =>I've managed to run NetBSD on an A2000 equipped with A2630 and =>A2091. It works with my own version of the a2091-driver as well as =>with mtk's 490-Kernel. (mtks kernel doesn't have a cursor after =>booting ... ;-) You managed to get #490 to work on your machine? It won't boot on my A2500 (I think it's the new MMU code.) How much RAM do you have? Do you have ECS? Here's my setup: A2500/30: A2000, 1M CHIP + ECS A2630, new ROMs + 4MB 32-bit RAM A2091, no RAM, adress 0x0e90000 -- ========================================================================= Eduardo Horvath eeh@btr.com ..!{decwrl,mips,fernwood}!btr!eeh "Trust me, I am cognizant of what I am doing." - Hammeroid From mw@eunet.ch Tue Jul 20 18:36:36 1993 Received: by chsun.eunet.ch (5.65c8/1.34) id AA06014; Tue, 20 Jul 1993 18:37:34 +0200 From: mw@eunet.ch (Markus Wild) Message-Id: <199307201637.AA06014@chsun.eunet.ch> Subject: Re: A2000+A2091: Success! To: hubert.feyrer@rrzc1.rz.uni-regensburg.de (Hubert Feyrer) Date: Tue, 20 Jul 1993 18:37:33 +0200 (MET DST) In-Reply-To: <9307201526.AA19621@rrzc1.rz.uni-regensburg.de> from "Hubert Feyrer" at Jul 20, 93 05:26:57 pm X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1341 Status: RO Hallo Hubert, > > Nur zu Deiner Info, der 490-Kernel bootet bei mir bis zur Shell > durch, nachdem ich mich mit der hdtoolbox und dem rootfs > herumgeschlagen habe. Wie es aussieht habe ich wohl als erster SUPA! Grmpf, jetzt frag ich mich aber wirklich, was bei den andern nicht lief.. ist ja total unberechenbar das ganze... > Allerdings bekomme ich beim newfs fuer meine usr-Partition > dauernd den Fehler, dass ich zuwenig Zylinder in der Zylinder- > Gruppe habe. Option -c hilft mir da leider auch nicht weiter! Das geschieht bei Partitionen die Parameter von HDToolBox verwenden.. Die schert sich naemlich einen Dreck um die wirklichen Disk-Parameter, und verteilt die Sektoren nach Lust und Laune. Bei mir hat das eine 64 Blocks/1 Kopf Verteilung gegeben auf der LPS105.. Das Problem ist, dass sich das BSD Filesystem gemaess verwendeter Disk-Geometrie optimiert, und die Platte in sogenannte Cylinder-Groups unterteilt. Damit das aber gut funktioniert, sollte die Angabe im RDB einigermassen mit der Wirklichkeit uebereinstimmen :-/ Die einfachste Loesung: lies die Daten aus mit einem SCSI-Tool, und gib sie explizit an bei newfs: -t HEADS -u BLOCKSPERTRACK. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From s_grau@ira.uka.de Tue Jul 20 20:36:26 1993 Received: from ira.uka.de by iraun1.ira.uka.de with SMTP (PP) id <07349-0@iraun1.ira.uka.de>; Tue, 20 Jul 1993 20:36:55 +0200 Received: from ira.uka.de by iraul1.ira.uka.de id <26815-0@iraul1.ira.uka.de>; Tue, 20 Jul 1993 20:36:39 +0200 From: s_grau@ira.uka.de (Guenther Grau) To: feyrer@rrzc1.rz.uni-regensburg.de Subject: NetBSD Date: Tue, 20 Jul 1993 20:36:40 +0200 Sender: s_grau@ira.uka.de Status: RO Hallo Hubert, wir haben uns eben auf irc ueber NetBSD unterhalten (Ich bin Maeuschen). Ich habe von einem Programm gesprochen, dass filetodev ersetzt. Hier eine Mail, die ich bekommen habe (ist auch aus der NetBSD-Mailing-liste): Date: Mon, 12 Jul 1993 14:49:40 -0400 From: Chris Hooper Message-Id: <199307121849.AA14468@astro.tc.mtu.edu> To: NetBSD-Amiga@cbmuucp.commodore.com Subject: re: Using a floptical to install BSD Sender: NetBSD-Admin@cbmuucp.commodore.com David E. Jones wrote: > filetodev 108 4666 foo.tar scsi.device 2 1000 This program always did make me a bit nervous. I rewrote it a couple weeks ago to be a little more user-friendly. If you have the partition mounted under AmigaDOS, you guys might want to check out dcp, which may be found at ftp.mtu.edu in the directory /pub/cdh. Then, you would type, "dcp foo.tar bsdd:" Be warned that it is still beta-ware and a bug would mean dire consequences for other filesystems on your hard drive. - Chris Hooper Programmer, MTU Networking (906) 487-2800=voice cdh@mtu.edu Michigan Technological University 2787=fax Viel Erfolg weiterhin mit NetBSD, Guenther P.S.: Vielleicht komme ich doch noch nach Bielefeld. Ill hat mich auf die Warteliste gesetzt. Falls noch jemand absagt, dann kann ich auch noch kommen. Ich wuerde so gerne ... From s_grau@ira.uka.de Tue Jul 20 20:36:26 1993 Received: from ira.uka.de by iraun1.ira.uka.de with SMTP (PP) id <07349-0@iraun1.ira.uka.de>; Tue, 20 Jul 1993 20:36:55 +0200 Received: from ira.uka.de by iraul1.ira.uka.de id <26815-0@iraul1.ira.uka.de>; Tue, 20 Jul 1993 20:36:39 +0200 From: s_grau@ira.uka.de (Guenther Grau) To: feyrer@rrzc1.rz.uni-regensburg.de Subject: NetBSD Date: Tue, 20 Jul 1993 20:36:40 +0200 Sender: s_grau@ira.uka.de Status: RO Hallo Hubert, wir haben uns eben auf irc ueber NetBSD unterhalten (Ich bin Maeuschen). Ich habe von einem Programm gesprochen, dass filetodev ersetzt. Hier eine Mail, die ich bekommen habe (ist auch aus der NetBSD-Mailing-liste): Date: Mon, 12 Jul 1993 14:49:40 -0400 From: Chris Hooper Message-Id: <199307121849.AA14468@astro.tc.mtu.edu> To: NetBSD-Amiga@cbmuucp.commodore.com Subject: re: Using a floptical to install BSD Sender: NetBSD-Admin@cbmuucp.commodore.com David E. Jones wrote: > filetodev 108 4666 foo.tar scsi.device 2 1000 This program always did make me a bit nervous. I rewrote it a couple weeks ago to be a little more user-friendly. If you have the partition mounted under AmigaDOS, you guys might want to check out dcp, which may be found at ftp.mtu.edu in the directory /pub/cdh. Then, you would type, "dcp foo.tar bsdd:" Be warned that it is still beta-ware and a bug would mean dire consequences for other filesystems on your hard drive. - Chris Hooper Programmer, MTU Networking (906) 487-2800=voice cdh@mtu.edu Michigan Technological University 2787=fax Viel Erfolg weiterhin mit NetBSD, Guenther P.S.: Vielleicht komme ich doch noch nach Bielefeld. Ill hat mich auf die Warteliste gesetzt. Falls noch jemand absagt, dann kann ich auch noch kommen. Ich wuerde so gerne ... From eeh@public.btr.com Wed Jul 21 03:18:58 1993 Received: from public.btr.com by btr.com (4.0/SMI-4.0-btr.1) id AA25093; Tue, 20 Jul 93 18:17:52 PDT Received: by public.btr.com (4.1/SMI-4.1.public.1) id AA17924; Tue, 20 Jul 93 18:17:48 PDT From: eeh@public.btr.com (Eduardo E. Horvath eeh@btr.com) Message-Id: <9307210117.AA17924@public.btr.com> Subject: Re: Success! (was: Re: NetBSD and A2000/2630) To: hubert.feyrer@rrzc1.rz.uni-regensburg.de (Hubert Feyrer) Date: Tue, 20 Jul 1993 18:17:47 -0700 (PDT) In-Reply-To: <9307201512.AA19596@rrzc1.rz.uni-regensburg.de> from "Hubert Feyrer" at Jul 20, 93 05:12:21 pm X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 338 Status: RO > Here's my setup: > > A2000 (1MB Chip RAM, no Fast RAM) > + ECS > + OS2.0 ROM > A2630 (4MB RAM) > A2091 SCSI-Contr. > + Seagate 1096 N (80MB) > Evolution SCSI-Contrl. > + Fujitsu 2624FA (520MB) > 4MB RAM board. ^^^^^^^^^^^^^^ Is this 16-bit or 32-bit, and is it in the Autoconfig address space or beyond 16MB? Thanks, Eduardo From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 23 14:51:42 1993 Subject: kvm_mkdb fails To: netbsd-amiga@cbmuucp.commodore.com (NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 1680 Sender: NetBSD-Admin@cbmuucp.commodore.com Hello, my netbsd was down for about week, and I'm still using quite old kernel (394), since it worked quite well for me and I didn't want to download the 452 source tree (the 394-450 diff had some problems ?) and recompile for my system. Ok, I had one root disk screwup and had to reinstall the rootfs. I believe that system worked after that, but now after one week (and some fsck's since my reboot doesn't work (16M machine)) /etc/rc reported on kvm_mkdb that it couldn't build the database. kvm_mkdb: Unknown error: 0 (or some such) and a temporary file *.tmp is left to /var/run directory. System works after that, but ofcourse kernel variables can't be found by ps & co. dev_mkdb works, /vmunix is readable (copied to /dev/null as a test :-) and I have no clue why the kvm_mkdb fails. Then something completely different.. I have done some boot code to read the /vmunix out of bsd root partition. It works for the first few blocks, until it starts to read the first level indirection blocks (after 12 direct blocks) but crashes right after that (Bus error in amigados :-) I'll have to try to "debug" it in netbsd. First thing will be ofcourse something like what loadbsd does currently, but instead of reading amigados file, the boot loader will read the kernel out of netbsd root partition. I have had some experiments with amigados boot blocks in hard disk partitions and the step to that level isn't big after the first working loader (I think and hope :-). BTW, the boot block code seems to run in such early stage of amigaos bootstrap, that there is no intuition screen. Is there any easy way to debug boot code (instead of sending data to serial port)? ++Tero From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 23 18:13:23 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: netbsd-amiga@cbmuucp.commodore.com Subject: Reboot and other stuff Sender: NetBSD-Admin@cbmuucp.commodore.com Just a question, it looked like 'reboot' does not automatically do a sync under NetBSD, is that right ? Then, /dev/reboot seems not to work for me, actually, is the kernel restarted and does netBSD reboots into the single user mode ? I would like to have an account 'shutdown', but i am unable to do so, because both, reboot and shutdown are sending messages to the initd and this one cancels because it has no output stream ? In the rc.local, the very first command is somthing like 'stty'; this one always segfaults (trapsignal) and leaves a nice core.stty behind, what's wrong here ? ObWhinie: I want my cursor back ! :-) -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 23 19:53:27 1993 Subject: kvm_mkdb and libc hash functions To: netbsd-amiga@cbmuucp.commodore.com (NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 831 Sender: NetBSD-Admin@cbmuucp.commodore.com Ok, I couldn't detect what was wrong with the kvm_mkdb command in /etc/rc, so I fetched the sources and tried to recompile, and later to add some debugging output... ...but the kvm_mkdb sources that I got from ftp.funet.fi /pub/unix/NetBSD/... directory (actually i386 binary/ sources distribution) didn't link! I think the source code should be common? Linker wanted to have hash_open() from libc, but there wasn't.. so, let's take some more sources, now usr/lib/libc/db/hash/* dynahash.c contained the wanted function, and that wanted hfunc.c with it, now the kvm_mkdb linked fine, but crashed right away when run :-/ What's wrong? In case my libc.a is corrupted, I'll download it after this writing and test again, but I fear it won't help.. (as whining has been done here before, so will I: "Where is the gdb?" :-) ++Tero From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 23 20:39:09 1993 Subject: Re: kvm_mkdb and libc hash functions To: nix@stekt.oulu.fi (Tero Manninen) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 831 Sender: NetBSD-Admin@cbmuucp.commodore.com > command in /etc/rc, so I fetched the sources and tried > to recompile, and later to add some debugging output... > ...but > the kvm_mkdb sources that I got from ftp.funet.fi > /pub/unix/NetBSD/... directory (actually i386 binary/ > sources distribution) didn't link! > I think the source code should be common? If you do such things, *always* get the sources from the NetBSD-current tree! Things have change *quite* a bit since the early days, especially the db-package has undergone several incompatible changes. I get my sources from sun-lamp.cs.berkeley.edu, perhaps you find some near mirror, dunno. > "Where is the gdb?" :-) I'll give it another try :-) -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 23 20:41:42 1993 Subject: Re: NetBSD and Retina To: jonesjg@dg-rtp.dg.com (Greg Jones) Cc: netbsd-amiga@cbmuucp.commodore.com (nba) X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 913 Sender: NetBSD-Admin@cbmuucp.commodore.com > Why have the board come up in the lower resolution, then have an ioctl to > the device to change the resolution. Someone could write a 'preferences' > command that does the ioctl to change the resolutions. And how/where should that program react? Nah, I guess I'll have the generic kernel present a resolution menu with lets say 10 or so different resolutions, where the user can chose one. If you go and compile your own kernel, you can default it on a certain resolution, and you won't get the question on boot anymore. I hate to say it, but that's how pc-linux does it... > Will I need a retna, to get X? Well, *I*'ll concentrate on a Retina server, but this doesn't mean there will be no custom chips server (mono anyway). -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From DCG9367@tntech.edu Fri Jul 23 22:07:36 1993 Received: from tntech.edu by tntech.edu (PMDF V4.2-13 #3874) id <01H0VWVPJGA8AH3K1R@tntech.edu>; Fri, 23 Jul 1993 15:08:16 CDT Date: Fri, 23 Jul 1993 15:08:16 -0500 (CDT) From: DCG9367@tntech.edu Subject: NetBSD on GVp controller To: feyrer@rrzc1.rz.uni-regensburg.de Message-Id: <01H0VWVPLLG2AH3K1R@tntech.edu> X-Vms-To: IN%"feyrer@rrzc1.rz.uni-regensburg.de" Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Content-Transfer-Encoding: 7BIT Status: RO I saw a message you posted to comp.unix.amiga a few days ago about getting NetBSD to work on your 2000. I was wondering if you know if it will work with GVP Series II controlles yet. I have an A500 with a CSA 030/882 Derringer and a GVP Series II controller, 13MB RAM and 450MB HD space. I have tried kernel #490 from ftp.eunet.ch(146.228.10.16) but it doesn't get very far. I get a few messages like: configuring hardware device [2/8]: device not found (or something like that) it does this for about 3 devices and then finds and configures the graphics and serial devices, but never any scsi devices. Then it says no suitable root found and stops. I have my root set up, but its not even identifying the scsi devices available and I have three of them. Any help would be greatly appreciated. Could you post this to comp.unix.amiga for me, I can read the messages through Gopher but I can't post. I can be reached at: dgreene@servdtnnsh.er.usgs.gov or dcg9367@tntech.edu Thanks. David From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 23 22:55:26 1993 Subject: Re: kvm_mkdb and libc hash functions To: mw@eunet.ch (Markus Wild) Cc: netbsd-amiga@cbmuucp.commodore.com (NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 1217 Sender: NetBSD-Admin@cbmuucp.commodore.com > > > command in /etc/rc, so I fetched the sources and tried > > to recompile, and later to add some debugging output... > > ...but Here is the story. Somehow my /vmunix was corrupted at the end, where string "database" is, now kvm_mkdb looked the correct offset to that point from the beginning of /vmunix and lseek to it. The data in the point contained only zero, program read an integer from the position that it thought that would contain the size of string space and decremented 4 bytes (the size of the size-var itself). Now that the first read var was zero, decrementing 4 got -4... then, try to malloc -4 bytes ;-) As a result of this /vmunix corruption I discovered that when malloc fails to allocate memory, it doesn't set the errno variable! Than's why the kvm_mkdb gave so informatic error message (kvm_mkdb: Unknown error: 0). Maybe someone thought that in vm malloc never fails ;-) ;-) > *always* get the sources from the > NetBSD-current tree! Things have change *quite* a bit since > the early days, Point taken. The sun-lamp site seems to be much of it's time unreachable, (down I guess, since neighbor hosts respond to ping) and I couldn't find the NetBSD-current mirrors anywhere.. ++Tero From NetBSD-Admin@cbmuucp.commodore.com Sun Jul 25 13:18:12 1993 Subject: uploaded kernel #500 To: netbsd-amiga@cbmuucp.commodore.com (nba) X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 2028 Sender: NetBSD-Admin@cbmuucp.commodore.com I just uploaded sources and binary for #500 (trarah:-)). I tried to integrate the sources I got for GVP11 support, as well as implementing limited multicontroller support. It's currently limited in that you can have more than one say GVP controller, but you can't have a 2091 and a GVP, or plugh a GVP into your A3000. This is because of the one single scsi.c file I'm using for all three controllers (if you really fell you HAVE to use different controllers, go take a look a fix :-)). These sources have been compiled under BSD, I hope I didn't break things to still compile under AmigaDOS. Note: you'll have to recompile config, for both environments, as the former version didn't handle device names containing digits (only restriction now is that the device name must be terminated with a letter, so the last digit(s) can indicate the number of such configured devices). The process of reintroducing multicontroller support (surprise, surprise, this already was in the template hp300 sources..:-)) brought support for LUNs back (if ever somebody would like to have LUNs...), and I *tried* to support drives that don't have a standard RDB, so you CDROM freaks out there might be able to mount your drives now. Keep in mind that partition C still means the whole drive, and for CDROMs that would be the partition to try to mount the ISO fs on. I just got an idea how I could get rid of the nasty bus-device-resetting, but I didn't want to delay the release of sources any longer, so that will have to wait awhile. Note that the "no-cursor" bug in the cc-driver is probably still there, I didn't feel like wading thru that black copper-magic.. :-) Other than that, have fun. And I'd sure like to hear whether the current version shows different behavior on non-A3000, would like to see more than one A2000/2091 combo that boots right off... pleaze :-) -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Sun Jul 25 16:54:38 1993 Subject: no more bus-device-reset.. To: netbsd-amiga@cbmuucp.commodore.com (nba) X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 838 Sender: NetBSD-Admin@cbmuucp.commodore.com Ok, see subject, boots are MUCH faster now, you'll notice if you have some drives on the bus, and each one taking its time to complete a reset command. That no longer happens. #508 instead negotiates for sync mode itself, so we're not only getting rid of resetting the devices, we also get support for sync transfers for free. I like it :-) A 508 kernel is on ftp.eunet.ch, as well as diffs (3k:-)) from 500 sources to 508. Enjoy:) -Markus PS: for those with drives that misbehave if sent sync-negotiation messages, look at scsi.c, there's a yet unneeded array called inhibit_sync, enter controller/unit of which device doesn't want to be sent sync-negotiation messages, ever. -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Sun Jul 25 21:39:06 1993 Subject: Re: uploaded kernel #500 To: mw@eunet.ch (Markus Wild) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1990 Sender: NetBSD-Admin@cbmuucp.commodore.com > > I just uploaded sources and binary for #500 (trarah:-)). I tried to > integrate the sources I got for GVP11 support, as well as > implementing limited multicontroller support. It's currently limited > in that you can have more than one say GVP controller, but you can't > have a 2091 and a GVP, or plugh a GVP into your A3000. This is > because of the one single scsi.c file I'm using for all three > controllers (if you really fell you HAVE to use different > controllers, go take a look a fix :-)). I have already looked into this and may actually work on it if 1) there is ever a lull long enough to do this between updates to my source trees, and 2) someone else doesn't beat me to the punch. > These sources have been compiled under BSD, I hope I didn't break > things to still compile under AmigaDOS. Note: you'll have to > recompile config, for both environments, as the former version didn't > handle device names containing digits (only restriction now is that > the device name must be terminated with a letter, so the last > digit(s) can indicate the number of such configured devices). There doesn't appear to be a lex (or flex)in the current NetBSD. > Other than that, have fun. And I'd sure like to hear whether the > current version shows different behavior on non-A3000, would like > to see more than one A2000/2091 combo that boots right off... pleaze :-) I downloaded the latest kernel (vmunix.508) and it still won't boot on my A2500. Something is still broken with the new VM setup. I'll keep hacking on it, but could you take a look to see if you do anything A3000 specific like expecting FAST RAM at above the 16M mark, expecting more than 4MB of contiguous RAM, or doing an unsafe bcopy of the kernel image over the kernel image? Eduardo ----- ========================================================================= Eduardo Horvath eeh@btr.com ..!{decwrl,mips,fernwood}!btr!eeh "Trust me, I am cognizant of what I am doing." - Hammeroid From NetBSD-Admin@cbmuucp.commodore.com Sun Jul 25 23:09:12 1993 To: eeh@public.btr.com (Eduardo E. Horvath eeh@btr.com) Cc: mw@eunet.ch (Markus Wild), netbsd-amiga@cbmuucp.commodore.com Subject: Re: uploaded kernel #500 <9307251937.AA02370@public.btr.com> X-Phone: (510) 549-3563 X-Mts: smtp Sender: NetBSD-Admin@cbmuucp.commodore.com > There doesn't appear to be a lex (or flex)in the current NetBSD. [2] sun-lamp # ls ~ftp/pub/NetBSD-current/src/gnu/usr.bin/lex COPYING ecs.c lex.1 nfa.c tblcmp.c CVS/ flexdef.h lex.skel parse.y yylex.c Makefile flexdoc.1 lib/ pathnames.h ccl.c gen.c main.c scan.l dfa.c initscan.c misc.c sym.c chris From NetBSD-Admin@cbmuucp.commodore.com Sun Jul 25 23:36:52 1993 Subject: Re: uploaded kernel #500 To: cgd@postgres.Berkeley.EDU Cc: eeh@public.btr.com, netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1329 Sender: NetBSD-Admin@cbmuucp.commodore.com > > There doesn't appear to be a lex (or flex)in the current NetBSD. > > [2] sun-lamp # ls ~ftp/pub/NetBSD-current/src/gnu/usr.bin/lex > COPYING ecs.c lex.1 nfa.c tblcmp.c > CVS/ flexdef.h lex.skel parse.y yylex.c > Makefile flexdoc.1 lib/ pathnames.h > ccl.c gen.c main.c scan.l > dfa.c initscan.c misc.c sym.c > > chris Oh, welcome to netbsd-amiga, Chris! This remark makes something clear I think I never really stressed: the binary distribution on ftp.eunet.ch is my *first* shot at binaries, the first thing I did when I was able to boot into BSD and run the gcc I generated under Amigados. These are far from complete, and far from optimal too. They also lack compiled man-pages, because at that time, I simply didn't have groff up to compile them. So, if you're really into BSD hacking, I'd strongly suggest you get yourself a copy of the current source distribution from sun-lamp.cs.berkeley.edu. I'll upload a new binary set soon, when I'm able to really compile thru (ie. say "make" in /usr/src :-)). -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 00:31:12 1993 To: eeh@public.btr.com Subject: Re: uploaded kernel #500 Cc: netbsd-amiga@cbmuucp.commodore.com Sender: NetBSD-Admin@cbmuucp.commodore.com > > >> These sources have been compiled under BSD, I hope I didn't break >> things to still compile under AmigaDOS. Note: you'll have to >> recompile config, for both environments, as the former version didn't >> handle device names containing digits (only restriction now is that >> the device name must be terminated with a letter, so the last >> digit(s) can indicate the number of such configured devices). > >There doesn't appear to be a lex (or flex)in the current NetBSD. > OK looks like a good time to list what I've "ported". Lex/flex was needed by myself for nethack. I got it and it compiled fine. I got it and other NetBSD src fm: ftp.uu.net under: systems/unix/bsd-sources [or somthing like that] I've also compiled fine the following: flex (under gnu dir) mail.local (found the orig in libexec later, but it still works) banner (Oh; happy, happy, joy, joy) tout (a time out prog I got fm a book) fish ( a go fish game) I think that's it... couldn't get sendmail working tho... While I'm here. Mr Babble (or Babylon...) said he has no intent to do any work of porting his drivers. So lost that one. If any- one else has the GVP I/O help me !!! - TC -- Techno Caster - ah360@yfn.ysu.edu From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 01:19:02 1993 Subject: Re: uploaded kernel #500 To: ah360@yfn.ysu.edu Cc: netbsd-amiga@cbmuucp.commodore.com (NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 713 Sender: NetBSD-Admin@cbmuucp.commodore.com Ok, I have now tried to comple the 508.. there appears to be problems with the supplied usr.sbin/config, because the generated a3000scsi.h, a2091scsi.h and gvp11scsi.h contain garbage :-/ a3000scsi.h should contain line: #define NA3000SCSI 1 but it contains: #define NA^S^P^P^PSCSI 1 Same thing to the other number conversations in 2091/gvp11 files. See that ^S := '3' and ^P := '0' :-/ Also when I reconfig and for example a3000scsi.h file exists, the config doesn't regenerate it, not until I remove it manually. Hmm.. just tested, the config seems to just append the new changes if I have changed the definition manually: #define NA3000SCSI 1 becomes: #define NA3000SCSI 1 #define NA^S^P^P^PSCSI 1 ++Tero From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 03:52:37 1993 Subject: vmunix.508 doesn't see my drives! To: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL21] Content-Type: text Content-Length: 531 Sender: NetBSD-Admin@cbmuucp.commodore.com Subject line sez it all. Here's what I get: dma0: A3000 32bit DMA scsi0: , scsi id 7 a3000scsi0 [1/1] sbic_wait TIMEO @1088 with asr=x0 csr=x41 scsi0: target 6 now synchronous, period=248 ns, offset=8. Realtime clock A3000 rtclock0 [1/4] grf0: 640x400 monochrome custom chips display grf0: [1/7] ser0: [1/3] no suitable root And there it stops. Vmunix.500 worked OK, except for no cursor. My hardware: A3000/16, 4MB RAM Quantum Prodrive 105S at SCSI ID 6 Insite Floptical 21MB removable at SCSI ID 2 WD33C93 00-04 PROTO :-( From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 04:07:01 1993 Subject: Re: vmunix.508 doesn't see my drives! To: netbsd-amiga@cbmuucp.commodore.com Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 581 Sender: NetBSD-Admin@cbmuucp.commodore.com > Subject line sez it all. Here's what I get: > no suitable root > > And there it stops. Vmunix.500 worked OK, except for no cursor. > My hardware: i got the same thing the first time i used 508 - then i gave it the three fingered salute and ran loadbsd again, and it came up okay. odd. -- paul theodoropoulos pt@crl.com diogenes@well.sf.ca.us System: A3000\25MHz\2Mchip\8Mfast\1int.floppy\CBM A2286BB\Maxtor 213M: scsi 6 8M rootfs,58M usr,14M swap,120M amigados\Maxtor 80M: IDE, 40M PC, 40M amiga,jdisk.device\Emerald (Teac) 60M tape drive - st600 type - scsi 4 From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 07:14:12 1993 Subject: Which tape drive? To: netbsd-amiga@cbmuucp.commodore.com (Amiga NetBSD mailing list) X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@cbmuucp.commodore.com I'm looking for a tape drive, can anyone recommend one? (I don't know anything about them). Also I've been away for a while and lost track of whats going on. Is there support for GVP controllers in the lastest kernel (the FAQ seems to indicate that there is) or are people still compiling custom setups for their GVP based systems? I guess I'm asking can I run NetBSD at the moment? My system: A2000 + 2.05 ROMS 1 MEG chip ram GVP GFORCE 030 + 68882 + 5 MEGS FAST RAM Quantum 120 MEG HD Seagate 240 MEG HD Retina (2 MEG) Thanks, Trevor. -- +----------------------------+----------------------------------------+ | Trevor Elbourne | Computer and Systems Technology (CaST) | | trevore@vast.unsw.edu.au | Dept of Computer Science & Engineering | | trevore@cse.unsw.edu.au | University of New South Wales | +----------------------------+----------------------------------------+ From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 09:39:21 1993 To: ah360@yfn.ysu.edu, nix@stekt.oulu.fi Subject: Re: uploaded kernel #500 Cc: netbsd-amiga@cbmuucp.commodore.com Sender: NetBSD-Admin@cbmuucp.commodore.com re: NA^S^P^P^P^PSCSI sounds like toupper() is broken. i honestly don't know whether this is fixed in the netbsd-current sources, but you might want to get libc and the include files (at minimum) from sun-lamp, and compile, install, etc., them, and recompile config with them. chris From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 09:43:55 1993 To: ah360@yfn.ysu.edu Subject: Re: uploaded kernel #500 Cc: netbsd-amiga@cbmuucp.commodore.com Sender: NetBSD-Admin@cbmuucp.commodore.com umm, as much as i hate to say it: i'd advise *against* getting NetBSD sources from ftp.uu.net... (or any other place which doesn't carry -current sources) on sun-lamp.cs.berkeley.edu there's a list of ftp sites. those which are listed as carrying NetBSD-current are updated regularly (i hope!) from the NetBSD-current tree on lamp... since NetBSD 0.8 (which is that latest uunet carries, i believe), we've upgraded virtually *all* of the GNU tools, and made *so* many bug fixes to the user-level utilties that using 0.8 code is *really* not worthwhile... "party on, dudes!" chris From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 14:06:48 1993 Subject: Re: uploaded kernel #500 To: nix@stekt.oulu.fi (Tero Manninen) Cc: ah360@yfn.ysu.edu, netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 653 Sender: NetBSD-Admin@cbmuucp.commodore.com > a3000scsi.h should contain line: > #define NA3000SCSI 1 > but it contains: > #define NA^S^P^P^PSCSI 1 Guess why I distributed include/ctype.h ? Yep:-) Please move that over to /usr/include, and recompile. Also remove all *.h files in compile/AMIGA before a new config attempt, or you get the symptoms you described with a valid line following the former corrupt line. The reason for the bug? Former toupper() and tolower() macros were not safe to be alled with autoincrement mode.. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 14:20:47 1993 Subject: Re: vmunix.508 doesn't see my drives! To: dejones@gandalf.ca (David E. Jones) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 2009 Sender: NetBSD-Admin@cbmuucp.commodore.com > sbic_wait TIMEO @1088 with asr=x0 csr=x41 Not nice.. this says (csr=x41): unexpected disconnect by target! > scsi0: target 6 now synchronous, period=248 ns, offset=8. Since it is followed by this line, I guess (pure guess though...) your drive didn't like the sync handshake as part of the ID handshake, or it plain doesn't want to take any commands after setting sync parameters, ie. you'd have to re-select the drive afterwards. > no suitable root > > And there it stops. Vmunix.500 worked OK, except for no cursor. No wonder.. after the scsi0: target .. line it should have printed a line identifying your drive, ie. saying something like sd6: QUANTUM... Since it didn't, it wasn't even able to read the IDENDTIFY string from the device. Only difference between 500 and 508 is sync-support, so your drive apparently didn't like sync setup... If you have the space and the environment to recompile the kernel, have a look at scsi.c, there's this yet unused array called "inhibit_sync". Put a 1 in the slot corresponding to your Quantum, and retry the boot. Note: although this trick inhibits setup of sync handshake, it will not - as former versions of the kernel did - reset the drive. The reason behind: I guess if the drive has problems with sync mode under BSD, it has problems with sync mode under ADOS too, so ADOS won't have initialized it to sync mode, and then we don't have to reset it to clear sync mode either. > Quantum Prodrive 105S at SCSI ID 6 Is this a predecessor of the LP105S? Just asking, because the kernel works fine on my LP105S. > Insite Floptical 21MB removable at SCSI ID 2 Was there a medium inside the floptical when booting? Just wondering why the drive doesn't show up in the scan. > WD33C93 00-04 PROTO :-( Hmm.. it would be easy to just blame the proto chip, I think it's something nastier.. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 14:23:24 1993 Subject: Re: vmunix.508 doesn't see my drives! To: pt@crl.com (paul theodoropoulos) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL11] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 532 Sender: NetBSD-Admin@cbmuucp.commodore.com > i got the same thing the first time i used 508 - then i gave it the > three fingered salute and ran loadbsd again, and it came up okay. > odd. Very strange. Did you get the identical symptoms, eg. same TIMEO at same line, with same csr value? The csr value is very important, as it tells me in what (wrong) state the controller chip was when the timeout occured. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 15:11:53 1993 Subject: Re: uploaded kernel #500 To: mw@eunet.ch (Markus Wild) Cc: netbsd-amiga@cbmuucp.commodore.com (NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 527 Sender: NetBSD-Admin@cbmuucp.commodore.com > > > a3000scsi.h should contain line: > > #define NA3000SCSI 1 > > but it contains: > > #define NA^S^P^P^PSCSI 1 > > Guess why I distributed include/ctype.h ? Yep:-) I did move, but after the config compilation :-/ Then something about the a3000/a2091/gvp11 compilation.. I tried to compile only with a3000 (commented out the others in the configuration file) and the result: a2091/gvp11 code is still compiled :-/ There should be some #if (NA2091SCSI > 0) and same for gvp lines around the dev/scsi.c, I think.. ++Tero From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 21:13:13 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: netbsd-amiga@cbmuucp.commodore.com (nba) Subject: Re: uploaded kernel #500 Sender: NetBSD-Admin@cbmuucp.commodore.com On Jul 25, 1:18pm, Markus Wild wrote: > Subject: uploaded kernel #500 [...] > Note that the "no-cursor" bug in the cc-driver is probably still > there, I didn't feel like wading thru that black copper-magic.. :-) Well, actually it behaves just strange. The cursor is visible when some programms disbale them (i.e. more, ls) and vice-versa. Does this give you a hint ? Probably it's not the copper-magic, but a false #ifdef ? > Other than that, have fun. And I'd sure like to hear whether the > current version shows different behavior on non-A3000, would like > to see more than one A2000/2091 combo that boots right off... pleaze :-) Ok, testing it *NOW* *Going to beaver and start #508* *Back* No success, the console driver still fails to initialise, wheras #299 does it right. I guess Byranosaurus (Bryan ! Where are you ?! Please fix that ! :-), did messed something up ? My A2000 has 2MB Chip-RAM and the latest Agnus, if that helps... -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Mon Jul 26 21:23:32 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: ah360@yfn.ysu.edu, eeh@public.btr.com Subject: GVP (was: Re: uploaded kernel #500) Cc: netbsd-amiga@cbmuucp.commodore.com Sender: NetBSD-Admin@cbmuucp.commodore.com On Jul 25, 6:30pm, T. Caster wrote: > While I'm here. Mr Babble (or Babylon...) said he has no intent > to do any work of porting his drivers. So lost that one. If any- > one else has the GVP I/O help me !!! - TC Mr Babel. And his mail site is Babylon. Well yes, he is too busy to deal with NetBSD. His comment is that anyone having experience with the 16551 serial controller can probably do a port. Doesn't the A2232 have a 16551, too ? -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 00:32:10 1993 Subject: Re: GVP (was: Re: uploaded kernel #500) To: markus@techfak.uni-bielefeld.de (Markus Illenseer) Cc: ah360@yfn.ysu.edu, eeh@public.btr.com, netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 599 Sender: NetBSD-Admin@cbmuucp.commodore.com > anyone having experience with the 16551 serial controller can > probably do a port. Hm.. is this the plain chip used in gazillions of PC RS232 boards? Is it the FIFO derivate of those chips? If yes, we already have a driver for it.. > Doesn't the A2232 have a 16551, too ? Hmm.. as far as I know, the A2232 has a cpu on the board, that you somehow how to download code there first. I don't know what UARTs are on the board. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 00:40:10 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: netbsd-amiga@cbmuucp.commodore.com Subject: Re: GVP (was: Re: uploaded kernel #500) Sender: NetBSD-Admin@cbmuucp.commodore.com On Jul 27, 12:32am, Markus Wild wrote: > Subject: Re: GVP (was: Re: uploaded kernel #500) > > anyone having experience with the 16551 serial controller can > > probably do a port. > > Hm.. is this the plain chip used in gazillions of PC RS232 boards? Is it > the FIFO derivate of those chips? If yes, we already have a driver for it.. I have no idea right now, last time i looked into a PC i got a shock anyway, those motherboards became ... babyboards ... And those Multi-IO cards have everything on: IDE, serial, parallel, joystick... > > > Doesn't the A2232 have a 16551, too ? > > Hmm.. as far as I know, the A2232 has a cpu on the board, that you somehow > how to download code there first. I don't know what UARTs are on the board. I just checked: they have 6502-devirate and 6551. -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 01:16:42 1993 To: markus@TechFak.Uni-Bielefeld.DE, netbsd-amiga@cbmuucp.commodore.com Subject: Re: uploaded kernel #500 Sender: NetBSD-Admin@cbmuucp.commodore.com > No success, the console driver still fails to initialise, wheras > #299 does it right. > > I guess Byranosaurus (Bryan ! Where are you ?! Please fix that ! :-), > did messed something up ? It's quite possible, although offhand I can't think of what might be causing problems. How exactly does it fail - does the screen just stay blank forever, or what? If it's a problem with my magic copper lists, you can disable them by commenting out the install_copper_list (or something like that) line in the initialization routine in ite_cc.c. See if that at least gets you a display. Bryan From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 01:56:38 1993 X-Mailer: //\\miga Electronic Mail (AmiElm 1.19) Content-Length: 1117 To: netbsd-amiga@cbmuucp.commodore.com Subject: Re: vmunix.508 doesn't see my drives! (fwd) Sender: NetBSD-Admin@cbmuucp.commodore.com On Jul 26, Markus Wild wrote: > > sbic_wait TIMEO @1088 with asr=x0 csr=x41 > > Not nice.. this says (csr=x41): unexpected disconnect by > target! > > > scsi0: target 6 now synchronous, period=248 ns, offset=8. It does the same thing to me. But sometimes it just times out with period=208ns, offset=12, and then just has the TIMEO mentioned above. Sometimes it can even identify one of my drives. And just three minutes ago it went on timing out after it once timed out for every drive but then with asr=x20 and csr=xff. My drives support and use synchronous transfers under AmigaDOS, they are Maxtor 213something at ID 0 and 1. When testing the above I did not have anything else like the tape drive turned on. > > WD33C93 00-04 PROTO :-( > > Hmm.. it would be easy to just blame the proto chip, I think > it's something nastier.. Maybe it IS the PROTO chip - I still got it as well (is there anyone who wants to sell the AMD scsi chip in Germany?). Rudy --- Rudolf Neuhaus home: ->rudy@dorn.ruhr.de 44137 Dortmund (FRG) work: rudy@iml.fhg.de Tel. +49(0) 231/134404 uni : uphd02@zx1.hrz.uni-dortmund.de From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 08:10:09 1993 To: baford@schirf.cs.utah.edu (bryan ford) Subject: Re: uploaded kernel #500 Cc: netbsd-amiga@cbmuucp.commodore.com Sender: NetBSD-Admin@cbmuucp.commodore.com From: baford@schirf.cs.utah.edu (bryan ford) >It's quite possible, although offhand I can't think of what might >be causing problems. How exactly does it fail - does the screen >just stay blank forever, or what? Mine does.. 2500/20 w/2091 >If it's a problem with my magic copper lists, you can disable them >by commenting out the install_copper_list (or something like that) >line in the initialization routine in ite_cc.c. See if that at >least gets you a display. I'm not in any position to compile it :-(.. Markus, what Make are you supposed to use with the Amiga gcc port to compile this? Manx' make, SAS' smake, and Dillon's DMake all choke on it... I would _*love*_ to be able to cross compile this on the NeXT.. and if anyone has been able to, I'd like to hear from you. From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 08:42:09 1993 Subject: Re: vmunix.508 doesn't see my drives! (fwd) To: bsdlist@dorn.ruhr.de (Rudolf Neuhaus) Cc: netbsd-amiga@cbmuucp.commodore.com (NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 845 Sender: NetBSD-Admin@cbmuucp.commodore.com > > > sbic_wait TIMEO @1088 with asr=x0 csr=x41 > > > > Not nice.. this says (csr=x41): unexpected disconnect by > > target! > > > > > scsi0: target 6 now synchronous, period=248 ns, offset=8. > > It does the same thing to me. But sometimes it just times out with > period=208ns, offset=12, and then just has the TIMEO mentioned above. Count me too on this. I have A3000, WD-PROTO, Quantum PD425S and LP105S drives, both refuse to work with synchro enabled in the scsi.c After I disabled the sync (initisied the somethign*inhibit* table in the scsi.c and now I can boot into 508, although without sync transfer. In the debug output, csr=x41 if I remember correctly.. Btw, the infamous cursor bug is in my system too.. I better stick with the good old 450 for now :-) The console in 508 seems to be much faster, only if it worked.. ++Tero From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 16:18:59 1993 Subject: Re: uploaded kernel #500 To: billc@icecube.rain.com (William J. Coldwell) Cc: baford@schirf.cs.utah.edu, netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@cbmuucp.commodore.com > > I'm not in any position to compile it :-(.. Markus, what Make are you > supposed to use with the Amiga gcc port to compile this? Manx' make, > SAS' smake, and Dillon's DMake all choke on it... > > I would _*love*_ to be able to cross compile this on the NeXT.. and > if anyone has been able to, I'd like to hear from you. > Use gnumake, but make sure your stack is at least 300K. I picked up a pre-compiled version on one of the Amiga MACH FTP sites. You may also need pdksh unless you edit the Makefiles to work with the AmigaDOS shell. -- Alan Bair AMCU DSCS Motorola, Inc. (Design Software & Mail Stop OE-320 Computer Services) 6501 William Cannon Dr. West Austin, TX 78735-8598 abair@amcu-tx.sps.mot.com From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 16:37:00 1993 Subject: Re: uploaded kernel #500 To: markus@TechFak.Uni-Bielefeld.DE (Markus Illenseer) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1680 Sender: NetBSD-Admin@cbmuucp.commodore.com > > On Jul 25, 1:18pm, Markus Wild wrote: > > Subject: uploaded kernel #500 > [...] > > Note that the "no-cursor" bug in the cc-driver is probably still > > there, I didn't feel like wading thru that black copper-magic.. :-) > > Well, actually it behaves just strange. The cursor is visible > when some programms disbale them (i.e. more, ls) and vice-versa. > > Does this give you a hint ? Probably it's not the copper-magic, but > a false #ifdef ? > > > Other than that, have fun. And I'd sure like to hear whether the > > current version shows different behavior on non-A3000, would like > > to see more than one A2000/2091 combo that boots right off... pleaze :-) > > Ok, testing it *NOW* *Going to beaver and start #508* > > *Back* > > No success, the console driver still fails to initialise, wheras > #299 does it right. > > I guess Byranosaurus (Bryan ! Where are you ?! Please fix that ! :-), > did messed something up ? > > My A2000 has 2MB Chip-RAM and the latest Agnus, if that helps... I still say you're barking up the wrong tree. All evidence on my machine (A2500/30) points to MMU problems. I would be able to verify this with the #508 kernel if I hadn't compiled it under BSD. Does anyone know why Ados can't read from tapes cut on NetBSD? I'm using BTN and GNU tar. BSD can read its own tapes with no problem, but when I try to read my Archive Viper from Ados it returns with no data, even if I try to cat from the tape. Eduardo ========================================================================= Eduardo Horvath eeh@btr.com ..!{decwrl,mips,fernwood}!btr!eeh "Trust me, I am cognizant of what I am doing." - Hammeroid From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 17:11:50 1993 Subject: Re: uploaded kernel #500 To: eeh@public.btr.com (Eduardo E. Horvath eeh@btr.com) Cc: netbsd-amiga@cbmuucp.commodore.com (NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 749 Sender: NetBSD-Admin@cbmuucp.commodore.com > using BTN and GNU tar. BSD can read its own tapes with no problem, > but when I try to read my Archive Viper from Ados it returns with no > data, even if I try to cat from the tape. I don't know why, but I you do like this: cd / ln vmunix vmunix.foo tar cf /dev/rst0 vmunix vmunix.foo vmunix.foo vmunix.foo then you can extract the vmunix in amigados: gtar xf tape: vmunix Also, much better way to move vmunix to amigados is to mount the bsd partitions to ados :-) There is this nice BFFS (bsd ffs) for amiga, it's read-only, but works quite well for the purpose. You can't do 'loadbsd bsd-root:vmunix' but still have to copy the file to amiga partition (why??) and at mounting time the filesystem generates couple of enforcer hits.. ++Tero From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 18:08:39 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: netbsd-amiga@cbmuucp.commodore.com (NetBSD mailing list ) Subject: Re: vmunix.508 doesn't see my drives! (fwd) Sender: NetBSD-Admin@cbmuucp.commodore.com On Jul 27, 9:41am, Tero Manninen wrote: [...] > Count me too on this. > I have A3000, WD-PROTO, Quantum PD425S and LP105S drives, both > refuse to work with synchro enabled in the scsi.c > After I disabled the sync (initisied the somethign*inhibit* table > in the scsi.c and now I can boot into 508, although without sync > transfer. Aha :-) I will check that one, too... I always thought the LPS105 is capable to do asynchronous transfer... > Btw, the infamous cursor bug is in my system too.. I better stick > with the good old 450 for now :-) > The console in 508 seems to be much faster, only if it worked.. I stick for 490 and try to fix the bug... *grmpf*. -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 19:01:09 1993 Subject: Re: uploaded kernel #500 To: abair@amcu-tx.sps.mot.com (Alan Bair) Cc: billc@icecube.rain.com, baford@schirf.cs.utah.edu, netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 491 Sender: NetBSD-Admin@cbmuucp.commodore.com > > I'm not in any position to compile it :-(.. Markus, what Make are you > > supposed to use with the Amiga gcc port to compile this? Manx' make, > > SAS' smake, and Dillon's DMake all choke on it... What make should one use to compile BSD? I'd suggest BSD-make:-) That's the make included with the gcc233 distribution:-) -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 19:06:57 1993 Subject: Re: vmunix.508 doesn't see my drives! (fwd) To: markus@techfak.uni-bielefeld.de (Markus Illenseer) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 888 Sender: NetBSD-Admin@cbmuucp.commodore.com > On Jul 27, 9:41am, Tero Manninen wrote: > [...] > > Count me too on this. > > I have A3000, WD-PROTO, Quantum PD425S and LP105S drives, both > > refuse to work with synchro enabled in the scsi.c > > After I disabled the sync (initisied the somethign*inhibit* table > > in the scsi.c and now I can boot into 508, although without sync > > transfer. > > Aha :-) I will check that one, too... > I always thought the LPS105 is capable to do asynchronous transfer... VERY fishy.. I have an lps105, and it works *fine* with the 508 kernel, so I really start to look suspiciously at my old PROTO chip on the table.. Perhaps the replacement with AMD did the trick? I don't feel like replacing the chips to verify though... -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 19:13:15 1993 Subject: Re: uploaded kernel #500 To: nix@stekt.oulu.fi (Tero Manninen) Cc: eeh@public.btr.com, netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL17] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 266 Sender: NetBSD-Admin@cbmuucp.commodore.com > > I don't know why, but I you do like this: > cd / > ln vmunix vmunix.foo > tar cf /dev/rst0 vmunix vmunix.foo vmunix.foo vmunix.foo It is only the last file that gets trashed (EOF missing or?) so you only need one vmunix.foo or any file of no value. /Peter From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 20:47:12 1993 Subject: Re: vmunix.508 doesn't see my drives! (fwd) To: markus@TechFak.Uni-Bielefeld.DE (Markus Illenseer) Cc: netbsd-amiga@cbmuucp.commodore.com (NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 318 Sender: NetBSD-Admin@cbmuucp.commodore.com > > Btw, the infamous cursor bug is in my system too.. I better stick > > with the good old 450 for now :-) > > The console in 508 seems to be much faster, only if it worked.. > > I stick for 490 and try to fix the bug... *grmpf*. Ah, but there are no 490 sources available any more, I must have missed it. ++Tero From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 20:57:46 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: netbsd-amiga@cbmuucp.commodore.com Subject: Re: uploaded kernel #500 Sender: NetBSD-Admin@cbmuucp.commodore.com On Jul 27, 7:15am, Eduardo E. Horvath eeh@btr.com wrote: [...] > I still say you're barking up the wrong tree. All evidence on my > machine (A2500/30) points to MMU problems. I would be able to verify > this with the #508 kernel if I hadn't compiled it under BSD. I am sure i bark, whoever he needs it :-) The old #299 _does_ work on my A2500/30, of course it does not recognize the SCSI-adaptor and so on, but the console does work. So i believe the console-init fails. -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 21:00:23 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: netbsd-amiga@cbmuucp.commodore.com Subject: Re: vmunix.508 doesn't see my drives! (fwd) Sender: NetBSD-Admin@cbmuucp.commodore.com On Jul 27, 7:06pm, Markus Wild wrote: [...] > VERY fishy.. I have an lps105, and it works *fine* with the 508 kernel, > so I really start to look suspiciously at my old PROTO chip on the table.. > Perhaps the replacement with AMD did the trick? I don't feel like replacing > the chips to verify though... And i don't feel like to _buy_ an AMD if it's not worth it :-) If the PROTO does but synchronous, i will use it until i get a new SCSI-Chip... But currently i don't have it. -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 21:06:12 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: netbsd-amiga@cbmuucp.commodore.com (NetBSD mailing list) Subject: Re: vmunix.508 doesn't see my drives! (fwd) Sender: NetBSD-Admin@cbmuucp.commodore.com On Jul 27, 9:46pm, Tero Manninen wrote: [...] > Ah, but there are no 490 sources available any more, > I must have missed it. I meant, i stick on #490 and try to deal with the #508 sources to find out the bugs in #508 :-) -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Tue Jul 27 21:20:11 1993 Subject: Re: vmunix.508 doesn't see my drives! (fwd) To: markus@TechFak.Uni-Bielefeld.DE (Markus Illenseer) Cc: netbsd-amiga@cbmuucp.commodore.com (NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 365 Sender: NetBSD-Admin@cbmuucp.commodore.com > And i don't feel like to _buy_ an AMD if it's not worth it :-) > If the PROTO does but synchronous, i will use it until i get a > new SCSI-Chip... But currently i don't have it. Out of curiosity, how much better performance sync transfer gives? Ordering just one chip may be expensive, and if the difference is not that big, there is no use to do it. ++Tero From NetBSD-Admin@cbmuucp.commodore.com Wed Jul 28 17:07:57 1993 To: netbsd-amiga@cbmuucp.commodore.com Cc: amigaman@esu.edu Subject: A2000 & GVP X-Mts: smtp Sender: NetBSD-Admin@cbmuucp.commodore.com Hi, I was away for a week and a few days, and a few kernels and a new FAQ popped out! You guys are really flying! My question is, though, does the new kernel work on A2000s equipped with GVP Series II Combo accelerators? The FAQ and some messages would seem to indicate that it >>might<<. I spent two days wrestling with my HDs and finally cleared off a 105M and two 80Ms, and set up 3 BSD partitions (as per the new FAQ) on the 105, and I get "no suitable root" Let me describe my setup: A2000 4.x Motherboard, 1M Chip GVP Combo Accelerator (50Mhz 68030/68882) 8MB 32-bit RAM (mapped at 0x01000000-0x017FFFFF) 8MB 16-bit RAM GVP Series II SCSI Controller GVP IO-Extender The BSDR, BSDS, and BSDD partitions are on a Quantum LPS105 at SCSI ID 5. I have a Wangtek WT150 at ID 4, and the two other drives I hope to use are at ID 3 and 6. However, when I "loadbsd vmunix.508" (or .500, for that matter) I get: (Grey screen, green pointer) Copyright ... Copyright ... NetBSD 0.8a () #508: Sun Jul 25 16:18:41 1993 AMIGA/A2000 MC68030 CPU+MMU, 25Mhz MC68882 FPU) real mem = 8388608 avail mem = 6070272 using 102 buffers containing 835584 bytes of memory grf0: 640x400 monochrome custom chip display grf0 [1/7] ser0 [1/3] no suitable root (just stops here...) I don't see it recognizing that there is a SCSI controller built into the Combo accelerator, nor do I see it recognizing that I have an additional GVP controller in the system. Does 508 support GVP? Would my second controller be confusing it? Why does it only see 8M of RAM? (I think it's the 32-bit RAM, the message on AmigaDOS seemed to indicate that it was at 0x01000000, though it disappeared quick) Can I ever get it to see all 16M? Would I want to? (The other 8 is only 16-bit, but it's a shame to waste it!) Why does it say it's 25Mhz? How many licks DOES it take to get to the center of a Tootsie-Roll Tootsie-Pop? ;) )Russ From NetBSD-Admin@cbmuucp.commodore.com Wed Jul 28 17:10:22 1993 Subject: Re: A2000 & GVP To: amigaman@esu.edu (Russ Miranda - Amiga Specialist) Cc: netbsd-amiga@cbmuucp.commodore.com, amigaman@esu.edu X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 898 Sender: NetBSD-Admin@cbmuucp.commodore.com > 8MB 32-bit RAM (mapped at 0x01000000-0x017FFFFF) > 8MB 16-bit RAM Where is the 16bit ram? For BSD to recognize 16M, the memory must be contigous, which in your case doesn't seem possible. > I don't see it recognizing that there is a SCSI controller built into > the Combo accelerator, nor do I see it recognizing that I have an 500/508 have a typo that makes them ignore GVP controllers, although the support really is in the kernel. To fix, load the kernel in your favorite binary editor, and replace: "GVPIIscsi" by "gvpscsi^@^@" with ^@ being ASCII-NUL (0). Watch case! Take care not to change the length of the binary. > 16-bit, but it's a shame to waste it!) Why does it say it's 25Mhz? Because it doesn't know:-) -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Wed Jul 28 21:53:52 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: "Russ Miranda - Amiga Specialist" , netbsd-amiga@cbmuucp.commodore.com Subject: Re: A2000 & GVP Sender: NetBSD-Admin@cbmuucp.commodore.com On Jul 28, 10:32am, "Russ Miranda - Amiga Specialist" wrote: > Subject: A2000 & GVP > Hi, > I was away for a week and a few days, and a few kernels and a new > FAQ popped out! You guys are really flying! My question is, though, > does the new kernel work on A2000s equipped with GVP Series II Combo > accelerators? The FAQ and some messages would seem to indicate that it > >>might<<. I spent two days wrestling with my HDs and finally cleared > off a 105M and two 80Ms, and set up 3 BSD partitions (as per the new > FAQ) on the 105, and I get "no suitable root" Let me describe my > setup: The 'no suitable root' error seems to be specific for the #508 vmunix. If available, get the #490 or #450, just to test, if your config works. Anyway, it looks good for you, you apparently have a working config, but your drive is not accepted by the kernel. [...] > grf0: 640x400 monochrome custom chip display > grf0 [1/7] > ser0 [1/3] > no suitable root Ops, looks like the controller is not accepted. See mw's mail. -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 00:24:31 1993 Subject: Re: uploaded kernel #500 To: markus@TechFak.Uni-Bielefeld.DE (Markus Illenseer) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1344 Sender: NetBSD-Admin@cbmuucp.commodore.com > > On Jul 27, 7:15am, Eduardo E. Horvath eeh@btr.com wrote: > [...] > > I still say you're barking up the wrong tree. All evidence on my > > machine (A2500/30) points to MMU problems. I would be able to verify > > this with the #508 kernel if I hadn't compiled it under BSD. > > I am sure i bark, whoever he needs it :-) > The old #299 _does_ work on my A2500/30, of course it does not > recognize the SCSI-adaptor and so on, but the console does work. > So i believe the console-init fails. > > > -- > Markus Illenseer Full signature available on request. > That's right. The MMU problem started with #450. Does #452 boot on your machine? I believe the problem may be with systems that have less than 4MB of contiguous FAST RAM. I'm still chasing shadows, but if you can compile your own kernel, poke different values in $dff180 at various places in amiga_init(). It woill change the background color. You can prove me wrong if the background remains the color you set at the very end of the function. On my system, the last color I get is what I set just before enabling the MMU. Eduardo ========================================================================= Eduardo Horvath eeh@btr.com ..!{decwrl,mips,fernwood}!btr!eeh "Trust me, I am cognizant of what I am doing." - Hammeroid From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 03:20:07 1993 Subject: shared libraries & DaggeX To: netbsd-amiga@cbmuucp.commodore.com (Amiga NetBSD mailing list) X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@cbmuucp.commodore.com Hiya all, Couple of queries; does amiga NetBSD use shared libraries? Helps to know this when deciding how much hard disk space to devote to BSD. Secondly can DaggeX be used as a local X server under NetBSD or is it an ADos X server? Thanks, Trevor. -- +---------------------------------------------------------------------+ | "Let me be the first to offer you my sincerest contrafibularities" | | - E. Blackadder | +----------------------------+----------------------------------------+ | Trevor Elbourne | Computer and Systems Technology (CaST) | | trevore@vast.unsw.edu.au | Dept of Computer Science & Engineering | | trevore@cse.unsw.edu.au | University of New South Wales | +----------------------------+----------------------------------------+ From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 07:00:39 1993 To: netbsd-amiga@cbmuucp.commodore.com Cc: amigaman@esu.edu Subject: A2000 & GVP (the saga continues) X-Mts: smtp Sender: NetBSD-Admin@cbmuucp.commodore.com Hi again, After newzapping vmunix.500, I got it to recognize my GVP controller, and talk to the drives. However... all is not peachy. The SCSI bus locks up while booting. Here's what happens: (imagine normal bootup is beginning... I am omitting all the copyrights, etc) [...] dma0: GVPII DMA scsi0: async, scsi id 7 gvpscsi0 [2017/11] sd0: QUANTUM LPS105S 910109405 rev , 205561 512 byte blocks sd0 at gvpscsi0, slave 0 sd1: QUANTUM LPS105S 910109405 rev , 205561 512 byte blocks sd1 at gvpscsi0, slave 1 sd2: MAXTOR LXT-213S rev 4.20, 415436 512 byte blocks sd2 at gvpscsi0, slave 2 sd3: QUANTUM P80S 980-80-94xx rev , 164058 512 byte blocks sd3 at gvpscsi0, slave 3 st0: WANGTEK >5150ES SCSI ES41< rev st0 at gvpscsi0, slave 4 sd5: MAXTOR LXT-213S rev 4.20, 415436 512 byte blocks sd5 at gvpscsi0, slave 5 sd6: QUANTUM P80S 980-80-94xx rev , 164058 512 byte blocks sd6 at gvpscsi0, slave 6 root on sd0 bpf: sl0 attached bpf: sl1 attached bpf: sl2 attached bpf: sl3 attached bpf: lo0 attached At this point, the SCSI bus locks up. The activity light on the controller comes on and stays on. The console still functions - while I was taking down this information, it blanked out on me. The pointer still moves with the mouse. (oh, and get this, my cursor is visable and the colors are fine by me!) The lockup seems to me to be the same sort of thing I get when I assign two devices the same SCSI id, but I checked that! I previously had the root on sd5, but it said "root on sd0" and then it couldn't mount root (it was on sd5!) so I swapped the IDs to see what would happen, and it went for a little bit longer before dying on me. Root IS now on sd0, as is my swap partition, and my BSDD. Has anyone had and luck (good or bad) with any of the drives above? There are two Quantum LPS105S drives, two Maxtor LXT-213S drives, and two Quantum P80S drives. I'm going to start trying removing drives and seeing if that helps, but I'm open to suggestion! Thanks in advance, once again! )Russ From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 08:46:02 1993 Subject: A2000 & GVP (the saga continues) (fwd) To: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL21] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 461 Sender: NetBSD-Admin@cbmuucp.commodore.com > From: "Russ Miranda - Amiga Specialist" > > Hi again, > After newzapping vmunix.500, I got it to recognize my GVP controller, > and talk to the drives. However... all is not peachy. The SCSI bus locks > up while booting. Here's what happens: > ... Interesting... I have the exact same problem with both 'ZAPed 500 and 508. I thought it was a termination problem, but I guess not! Similar setup. -- Ryan Brooks ryan@chaos.mcs.mu.edu From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 09:35:37 1993 Subject: Re: shared libraries & DaggeX To: trevore@vast.unsw.edu.au (Trevor Elbourne) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 852 Sender: NetBSD-Admin@cbmuucp.commodore.com > Couple of queries; does amiga NetBSD use shared libraries? Helps to Not yet. Someone is working on implementing a free version of SunOS style shared libraries, but it's not yet done. I agree I'd really want shared libraries when we go X... (hm, how good/bad is the accessibility of Sun3 among the Amiga community? It should be possible to - as an intermeditate solution - write a crt0.c that loads the - sun - ld.so, and thus load sunos shared libraries. However, these binaries would of course not be distributable, everyone would have to find "his" personal sun3... > Secondly can DaggeX be used as a local X server under NetBSD or is it > an ADos X server? It's an ADOS server. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 09:44:49 1993 Subject: Re: A2000 & GVP (the saga continues) To: amigaman@esu.edu (Russ Miranda - Amiga Specialist) Cc: netbsd-amiga@cbmuucp.commodore.com, amigaman@esu.edu X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 818 Sender: NetBSD-Admin@cbmuucp.commodore.com > root on sd0 *this* is already something that shouldn't be here.. 500/508 are GENERIC kernels, and they should never say "root on", but "changing root to". Since you got "root on", something is very fishy.. If you feel like it, the crucial things are spread over amiga/autoconf.c, amiga/swapgeneric.c and dev/sd.c. > At this point, the SCSI bus locks up. The activity light on the controller I think because there is no swap-space due to the above bug.. > Has anyone had and luck (good or bad) with any of the drives above? > There are two Quantum LPS105S drives, two Maxtor LXT-213S drives, and My lps105 works fine for me, no problem at all. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 10:01:15 1993 Subject: Re: Reboot and other stuff To: markus@techfak.uni-bielefeld.de (Markus Illenseer) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 985 Sender: NetBSD-Admin@cbmuucp.commodore.com > Just a question, it looked like 'reboot' does not automatically > do a sync under NetBSD, is that right ? /dev/reboot does, but I don't trust it;-) > Then, /dev/reboot seems not to work for me, actually, is the kernel > restarted and does netBSD reboots into the single user mode ? cp /vmunix /dev/reboot should restart the kernel you copied there. Requires newer generation kernel though, that supports /dev/reboot. > I would like to have an account 'shutdown', but i am unable to > do so, because both, reboot and shutdown are sending messages to > the initd and this one cancels because it has no output stream ? Hmmm... > In the rc.local, the very first command is somthing like 'stty'; > this one always segfaults (trapsignal) and leaves a nice core.stty > behind, what's wrong here ? Recompile stty... -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 10:09:37 1993 Subject: Re: shared libraries & DaggeX To: mw@eunet.ch (Markus Wild) Cc: netbsd-amiga@cbmuucp.commodore.com (NetBSD mailing list) X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 8bit Content-Length: 289 Sender: NetBSD-Admin@cbmuucp.commodore.com > (hm, how good/bad is the accessibility of > Sun3 among the Amiga community? It should be possible to - as an > intermeditate solution - write a crt0.c that loads the - sun - ld.so, and > thus load sunos shared libraries. I {know,have accesss to} one Sun-3 in our university :-) ++Tero From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 11:52:36 1993 Subject: Re: vmunix.508 doesn't see my drives! (fwd) To: nix@stekt.oulu.fi (Tero Manninen) Cc: bsdlist@dorn.ruhr.de, netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1140 Sender: NetBSD-Admin@cbmuucp.commodore.com > I have A3000, WD-PROTO, Quantum PD425S and LP105S drives, both > refuse to work with synchro enabled in the scsi.c > After I disabled the sync (initisied the somethign*inhibit* table > in the scsi.c and now I can boot into 508, although without sync > transfer. > > In the debug output, csr=x41 if I remember correctly.. Sudden disconnect too.. Ok, lets discuss this. Anyone knowing SCSI specs well.. I'm currently sending sending Identify, followed by the Sync-req message, in one go (thus one MsgOut phase). I thought this is valid according to SCSI specs, isn't it? What seems to happen in those cases where a disconnect error occurs (csr==0x41) is that the target sends, after the acknowledged sync message (ie. I get the sync request back like the protocol requests, meaning that the target accepted it) a command-complete message and disconnects. In my eyes, it should proceed to command-phase, who's right? Is it valid behavior for the target to just "hang up" ? -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 14:21:14 1993 Subject: SCSI snafus To: mw@eunet.ch (Markus Wild) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL21] Content-Type: text Content-Length: 546 Sender: NetBSD-Admin@cbmuucp.commodore.com > Sudden disconnect too.. Ok, lets discuss this. Anyone knowing SCSI specs > well.. I'm currently sending sending Identify, followed by the > Sync-req message, in one go (thus one MsgOut phase). I thought this is > valid according to SCSI specs, isn't it? What seems to happen in those Question: what happens if you don't get a message out phase? The KludgeMach 33c93 driver expects message out phase. However, my floptical goes immediately to command phase, skipping message out. How do you handle sync negotiation with something like this? From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 14:51:11 1993 Subject: Re: SCSI snafus To: dejones@gandalf.ca (David E. Jones) Cc: netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 988 Sender: NetBSD-Admin@cbmuucp.commodore.com > Question: what happens if you don't get a message out phase? Hmm.. probably lockup.. Note that I don't request an additional msgout phase in the BSD driver (contrary to my mach driver), I'm sending the sync request together with the identification message. Drives usually grant you this first msgout phase immediately, since without, you woudn't be able to identify yourself, and thus no disc/recon protocol would be supported. > The KludgeMach 33c93 driver expects message out phase. However, my floptical > goes immediately to command phase, skipping message out. How do you handle > sync negotiation with something like this? Not currently a problem with this driver, see above. But I do think it's a problem if a target all of a sudden decides to quit (command-complete/disc) without going to command-phase... -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 19:29:03 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: netbsd-amiga@cbmuucp.commodore.com Subject: Re: Reboot and other stuff Sender: NetBSD-Admin@cbmuucp.commodore.com On Jul 29, 10:00am, Markus Wild wrote: [...] > > In the rc.local, the very first command is somthing like 'stty'; > > this one always segfaults (trapsignal) and leaves a nice core.stty > > behind, what's wrong here ? > > Recompile stty... Recompile ? It's in the rootfs ... *grmbl* Noone else has this problem with stty ? -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 19:30:46 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: bsdlist@dorn.ruhr.de, netbsd-amiga@cbmuucp.commodore.com Subject: Re: vmunix.508 doesn't see my drives! (fwd) Sender: NetBSD-Admin@cbmuucp.commodore.com On Jul 29, 11:52am, Markus Wild wrote: [...] > message and disconnects. In my eyes, it should proceed to command-phase, who's > right? Is it valid behavior for the target to just "hang up" ? You are the initiator, aren't you ? Thus you should ask the SLAVE's specs, no ? Hm... -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Thu Jul 29 19:51:40 1993 Subject: Re: vmunix.508 doesn't see my drives! To: markus@TechFak.Uni-Bielefeld.DE (Markus Illenseer) Cc: bsdlist@dorn.ruhr.de, netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 680 Sender: NetBSD-Admin@cbmuucp.commodore.com > > message and disconnects. In my eyes, it should proceed to command-phase, who's > > right? Is it valid behavior for the target to just "hang up" ? > > You are the initiator, aren't you ? Thus you should ask the SLAVE's > specs, no ? Hm... Huh? If you're referring to sync-handshake: both, initiator and target are allowed to start a sync-handshake. The other party is to either reject the message, or to reply with same or inferior parameters, where upon both parties agree on these parameters and go sync. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 30 00:10:18 1993 Subject: Re: uploaded kernel #500 To: mw@eunet.ch (Markus Wild) Cc: abair@amcu-tx.sps.mot.com, billc@icecube.rain.com, baford@schirf.cs.utah.edu, netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL22] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 635 Sender: NetBSD-Admin@cbmuucp.commodore.com I think I've found where netbsd crashes on my machine. the line: tc = 0x82d08b00; consistantly crashes my machine. I have tried moving it before the first pmove, I have tried making it static data, and I have even tried initializing it in the declaration. So far nothing has worked. I don't understand why storing a value in a variable should cause this problem, but it does, and it does consistantly. Any ideas? Eduardo ========================================================================= Eduardo Horvath eeh@btr.com ..!{decwrl,mips,fernwood}!btr!eeh "Trust me, I am cognizant of what I am doing." - Hammeroid From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 30 00:11:00 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: Markus' amiga setup ? Sender: NetBSD-Admin@cbmuucp.commodore.com Hey Markus, Could you post a copy of your configuration dir tree for building kernels on the Amiga side? (dir opt a) Having some problems getting my setup just like yours... thanks -- Ryan Brooks ryan@chaos.mcs.mu.edu From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 30 10:47:09 1993 X-Mailer: Mail User's Shell (7.1.0 4/25/90) To: netbsd-amiga@cbmuucp.commodore.com Subject: #513 works Sender: NetBSD-Admin@cbmuucp.commodore.com Subject says it all. Version #513 works flawlessly with my A3000 and my LPS105. Printed text shows that the drive is running in synchronous mode. And the drives are recognized quite fast, booting is fun now :-) Haven't tested it on my A2000 yet. ObFlame: I want my cursor back ! :-) Bryan.... *looking sharply at him* -- Markus Illenseer Full signature available on request. From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 30 16:29:50 1993 X400-Originator: /DD.ID=1619692/G=Hamish/I=HI/S=Macdonald/@bnr.ca X400-Mts-Identifier: [/PRMD=BNR/ADMD=TELECOM.CANADA/C=CA/;bcars735.b.295:30.06.93.14.26.29] X400-Content-Type: P2-1984 (2) Content-Identifier: Re: vmunix.50... Sender: "Hamish (H.I.) Macdonald" To: mw@eunet.ch Cc: markus@techfak.uni-bielefeld.de, bsdlist@dorn.ruhr.de, netbsd-amiga@cbmuucp.commodore.com Subject: Re: vmunix.508 doesn't see my drives! Sender: NetBSD-Admin@cbmuucp.commodore.com >> > right? Is it valid behavior for the target to just "hang up" ? >> > You are the initiator, aren't you ? Thus you should ask the SLAVE's >> specs, no ? Hm... >>>>> On Thu Jul 29 14:09:00 1993, >>>>> mw@eunet.ch wrote: mw> Huh? If you're referring to sync-handshake: both, initiator and mw> target are allowed to start a sync-handshake. The other party is mw> to either reject the message, or to reply with same or inferior mw> parameters, where upon both parties agree on these parameters and mw> go sync. I suspect that if the target wants to disconnect after the sync-handshake, it is probably perfectly valid. The only type of phase that the initiator can start is the BUS_FREE phase (by releasing the BSY signal), or the MESSAGE_OUT phase (by asserting ATN). Other phases are under the targets control, no? I think if it wants to disconnect, it's allowed. I suppose in this situation you should just restart; i.e. issue the select again, and just send the IDENTIFY message this time. From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 30 16:47:08 1993 Subject: Re: vmunix.508 doesn't see my drives! To: hamish@bnr.ca (Hamish) Cc: markus@techfak.uni-bielefeld.de, bsdlist@dorn.ruhr.de, netbsd-amiga@cbmuucp.commodore.com X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 669 Sender: NetBSD-Admin@cbmuucp.commodore.com > I suspect that if the target wants to disconnect after the > sync-handshake, it is probably perfectly valid. The only type of Maybe.. it's sure not reasonable though... > disconnect, it's allowed. I suppose in this situation you should just > restart; i.e. issue the select again, and just send the IDENTIFY > message this time. That's what I'm currently trying out in #516, lets see if it fixes that problem. If you get other problems, try enabling _sync_debug = 1 to see more details about it. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 30 18:09:41 1993 Subject: vmunix.516 - nothing happens To: netbsd-amiga@cbmuucp.commodore.com (Amiga NetBSD mailing list) X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@cbmuucp.commodore.com I tried booting the 516 kernel (first time I have attempted boot NetBSD) and nothing happened. When I ran loadbsd vmunix.516 in the shell it said "4 megs fast blah blah ...". I then got a grey screen and green cross (mouse pointer I assume) but that was it!! nothing else happened, the machine just sat there. What is wrong? My setup: A2000 + 1MEG chip GVP GFORCE '030 + 5MEGS FAST Quantum LP120S (116MEG) HD at SCSI ID0 I have setup the disk correctly (ROOT, SWAP, USR and HOME), filetodev the rootfs and changed GVPIIscsi to gvpscsi^@^@ in the kernel. Thanks muchly, Trevor. -- +---------------------------------------------------------------------+ | "Let me be the first to offer you my sincerest contrafibularities" | | - E. Blackadder | +----------------------------+----------------------------------------+ | Trevor Elbourne | Computer and Systems Technology (CaST) | | trevore@vast.unsw.edu.au | Dept of Computer Science & Engineering | | trevore@cse.unsw.edu.au | University of New South Wales | +----------------------------+----------------------------------------+ From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 30 19:01:02 1993 To: netbsd-amiga@cbmuucp.commodore.com Subject: NetBSD 508 SCSI and console Sender: NetBSD-Admin@cbmuucp.commodore.com I just got BSD 508 compiled and running. Two gotchas... 1. I had to disable SCSI sync transfers. It appears this is the same problem other people are having. I have a Quantum LP52S, a P105S (I think), and a 200-meg Connor drive which claims it's a "DEC R224" (probably OEM'd). Once the Quantum P105S was successfully detected on bootup (it's slave 0), but the DEC and everything after it has never been initialized correctly. The typical error is "sbic_wait TIMEO @1088..." 2. Markus W: You _broke_ my beautiful console code!!! I don't get a cursor; presumably the same thing that's happening to Markus I. I'll look into it in two or three days if no one else beets me to it (very busy weekend coming :-) ). Otherwise, looks good! Bryan P.S. There also seems to be a problem with my /dev/reboot code, which probably existed from the start. Loading a kernel with a very different size than the kernel already loaded seems to not work. This is not very important, since it usually works, but I'll try to look into it sometime soon anyway... If anyone has ideas on what might be happening, I'm all ears. From NetBSD-Admin@cbmuucp.commodore.com Fri Jul 30 19:22:38 1993 Subject: Re: vmunix.516 - nothing happens To: trevore@vast.unsw.edu.au (Trevor Elbourne) Cc: netbsd-amiga@cbmuucp.commodore.com (NetBSD-Amiga mailing list) X-Mailer: ELM [version 2.3 PL0] Sender: NetBSD-Admin@cbmuucp.commodore.com Trevor Elbourne: > > I tried booting the 516 kernel (first time I have attempted boot > NetBSD) and nothing happened. When I ran loadbsd vmunix.516 in the ... > My setup: > > A2000 + 1MEG chip > GVP GFORCE '030 + 5MEGS FAST > Quantum LP120S (116MEG) HD at SCSI ID0 Almost identical to my setup, I have HD at SCSI ID#4 though (330 MB Micropolis 1684) and CD-ROM at ID#5. > I have setup the disk correctly (ROOT, SWAP, USR and HOME), filetodev > the rootfs and changed GVPIIscsi to gvpscsi^@^@ in the kernel. This is the problem: you don't need to patch that string anymore. Try to boot with the original #516 kernel and tell if you get to the prompt! My machine finds now the root (thanks to mtk's frantic debugging yesterday :) and starts booting but hangs after it says 'bpf: lo0 attached'. -- / / most stupid ayrjola@hut.fi / / dinosaurs ___________________/ / ought to starve --------------------' From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 31 03:33:10 1993 Subject: Re: vmunix.516 - nothing happens To: ayrjola@hila.hut.fi (Ari Yrj|l{) Cc: netbsd-amiga@cbmuucp.commodore.com (Amiga NetBSD mailing list) X-Mailer: ELM [version 2.3 PL11] Sender: NetBSD-Admin@cbmuucp.commodore.com > Trevor Elbourne: > > > > I tried booting the 516 kernel (first time I have attempted boot > > NetBSD) and nothing happened. When I ran loadbsd vmunix.516 in the > ... > > My setup: > > > > A2000 + 1MEG chip > > GVP GFORCE '030 + 5MEGS FAST > > Quantum LP120S (116MEG) HD at SCSI ID0 > > Almost identical to my setup, I have HD at SCSI ID#4 though > (330 MB Micropolis 1684) and CD-ROM at ID#5. > > > > I have setup the disk correctly (ROOT, SWAP, USR and HOME), filetodev > > the rootfs and changed GVPIIscsi to gvpscsi^@^@ in the kernel. > > This is the problem: you don't need to patch that string anymore. > Try to boot with the original #516 kernel and tell if you get > to the prompt! My machine finds now the root (thanks to mtk's > frantic debugging yesterday :) and starts booting but hangs > after it says 'bpf: lo0 attached'. > > > -- > / / most stupid > ayrjola@hut.fi / / dinosaurs > ___________________/ / ought to starve > --------------------' I tried using 516 both with and without changing that GVPIIscsi string, neither worked and both yielded the same result :( Trevor. -- +----------------------------+----------------------------------------+ | Trevor Elbourne | Computer and Systems Technology (CaST) | | trevore@vast.unsw.edu.au | Dept of Computer Science & Engineering | | trevore@cse.unsw.edu.au | University of New South Wales | +----------------------------+----------------------------------------+ From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 31 05:07:39 1993 To: netbsd-amiga@cbmuucp.commodore.com Cc: amigaman@esu.edu Subject: Trouble w/vmunix.516 X-Mts: smtp Sender: NetBSD-Admin@cbmuucp.commodore.com Hi all, I saw 516 up there today, and snatched it. When booting, I got the following: [...boring stuff omitted...] GVPIIscsi0 [2017/11] scsi0: target 0 now synchronous, period=248ns, offset=8. sbic_wait TIMEO @630 with asr=x0 csr=x41 sbic_wait TIMEO @630 with asr=x0 csr=x41 sbic_wait TIMEO @630 with asr=x0 csr=x41 [you get the picture, this continues ad infinitum, or until a key is pressed] I tried it several times, and the first time it seemed to actually work on ID 0, and did this sbic_wait TIMEO stuff on ID 1. Subsequent times it never got past ID 0. I could not repeat the first attempt, I have no idea what was different. Once again, that's a GVP Series II Combo Accelerator (and controller) and the relevant drives are (IDs 0 and 1) Quantum LP105Ss, although there are 7 devices total on the SCSI bus (plus the controller itself). Hope this helps someone! )Russ From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 31 13:54:39 1993 Subject: Re: Trouble w/vmunix.516 To: amigaman@esu.edu (Russ Miranda - Amiga Specialist) Cc: netbsd-amiga@cbmuucp.commodore.com, amigaman@esu.edu X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1600 Sender: NetBSD-Admin@cbmuucp.commodore.com > sbic_wait TIMEO @630 with asr=x0 csr=x41 > sbic_wait TIMEO @630 with asr=x0 csr=x41 > sbic_wait TIMEO @630 with asr=x0 csr=x41 > [you get the picture, this continues ad infinitum, or until a key is pressed] What is happening is the target disconnects unmotivated (0x41), the driver decides that this might be a target that disconnects after sync negotiation and retries selection, then the target disconnects again, etc. The one thing I don't understand is WHY it disconnects... Please try to enable some debugging, I'd recommend you start by setting _sync_debug to 1, and see whether you get some more information. Then try setting _scsi_debug to 1, and later to 2. You should get tons of information with _scsi_debug at 2. For setting those variables, see README.binpatch and the binpatch program. > Once again, that's a GVP Series II Combo Accelerator (and controller) and the > relevant drives are (IDs 0 and 1) Quantum LP105Ss, although there are 7 devices > total on the SCSI bus (plus the controller itself). Hmm.. perhaps there's a reason why gvpscsi.device doesn't support sync scsi.. I can hardly accept it, because it uses the same controller chip as the 2091 and the 3000, and THOSE do sync like a charm. I unfortunately declared the inhibit_sync array in scsi.c static, so the symbol doesn't show up in vmunix. Unless you're really desperate, I won't walk you thru offset-addition to symbols that ARE in vmunix.... -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 31 14:59:43 1993 To: netbsd-amiga@cbmuucp.commodore.com Cc: amigaman@esu.edu Subject: GVP & vmunix.516 revisited. X-Mts: smtp Sender: NetBSD-Admin@cbmuucp.commodore.com OK, I turned on the debugging, and here's what I got... interestingly enough, it appears that it was once again able to succeed on ID 0, and failed on ID 1. Maybe the contrast is helpful!! [...] find_busslaves: for GVPIIscsi0 looking for sd0 at slave 0...ixfer_in fail: l2 i80 w49940 scsi0: target 0 now synchronous, period = 248ns, offset=8. ixfer_in fail: l6 i80 w49982 sd0: QUANTUM LP105S 910109405 rev , 205561 512 byte blocks A @196 - 16660 = 16464 B @16758 - 82908 = 66150 D @83006 - 205408 = 122402 total 3 BSD-partitions found sd0 at GVPscsi0, slave 0 looking for sd1 at slave 1...ixfer_in fail: l2 i80 w49930 scsi0: target 1 now synchronous, period 248ns, offset-8 ixfer_in fail: l6 i80 w49981 sbic_wait TIMEO @630 with asr=x0, csr=x41 sbic_wait TIMEO @630 with asr=x0, csr=x41 ["""" continues like this] I was also happy to see it found my partitions! ;) Good luck, I'll be away until Sunday night. )Russ From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 31 15:11:42 1993 Subject: new config binary for amigados To: netbsd-amiga@cbmuucp.commodore.com (nba) X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 473 Sender: NetBSD-Admin@cbmuucp.commodore.com Just uploaded new config binary (compiled sources as distributed in #500 sources) for use under amigados. This one replaces bin/config on eunet, and is now packed, thus bin/config.gz. Might be helpful for people compiling BSD kernels under AmigaDOS without having tools like bison/yacc and flex installed. -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH From NetBSD-Admin@cbmuucp.commodore.com Sat Jul 31 21:07:30 1993 Subject: Re: GVP & vmunix.516 revisited. To: amigaman@esu.edu (Russ Miranda - Amiga Specialist) Cc: netbsd-amiga@cbmuucp.commodore.com, amigaman@esu.edu X-Mailer: ELM [version 2.4 PL23alpha] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 417 Sender: NetBSD-Admin@cbmuucp.commodore.com > OK, I turned on the debugging, and here's what I got... interestingly enough, > it appears that it was once again able to succeed on ID 0, and failed on ID 1. Hmm.. you didn't seem to enable sync debugging, set _sync_debug = 1 _scsi_debug = 2 -Markus -- CHUUG/EUnet Switzerland Markus Wild Zweierstrasse 35 Tel: +41 1 291 45 80 mw@eunet.ch CH-8004 Zuerich Fax: +41 1 291 46 42 S=mw;P=EUnet;A=EUnet;C=CH