beige g3 etch install
Hello all, I have been working with Debian for quite a few years now and using powermac G3 machines with extra network cards in them as routers and firewalls, etc. These machines were set up back in the days of Woody and have been kept uptodate with security updates, but otherwise pretty much left untouched. Due to the issues around the firmware in these machines and because it was useful, I always set them up to boot initially into mac OS9 then use BootX to hand over the Debian. This worked a treat. However, I thought it would be good come up to date by installing Etch instead. I used a free machine that was not actually in use and ran the install using the current network install ISO. Things have changed since the days of woody and it now seems that floppy images (boot image and root image) are no longer used. Hence, I copied the initrd.gz file over to the mac HD and set that as the ram disk for the install. I also copied across the linux kernel and put that into the kernels folder in the system folder. Using that allows me to boot into the installer and using the installer I deleted the previous linux partition (hda7) and swap (hda8) and made new ones. Then installed the base system, etc,etc all the way through to where it runs tasksel. I just leave that at the basic system for now. Following on some more, finally we get to the point of trying to install Quik (which I don't need) and it gives an error anyway, since I have selected ext3 file system that is not supported in quik. I therefore say to carry on without a boot loader. Everything goes fine all the way to rebooting into the new system. However, when I do that, OS9 will not boot up. I just get the flashing disk symbol with a question mark on it. Popping the OS9 CD and booting off that and then running disk setup shows me that the HD has somehow been altered so it is not recognised properly as a mac HD. During the partitioning step, I did not alter anything other than hda7 and 8. I have found that I can reinstall the apple hard disk driver onto the disk and this then gets OS9 up and working. However, I cannot then boot into Debian, since the boot process gets a little way in and then I get a kernel panic at the point where it tries to mount the file system (error about no file system at /dev/hda7). I have done this several times now and the same thing happens every time. The install goes fine but then I end up with a completely unusable machine. I am wondering if I am going about the install process wrongly (ie using the initrd.gz file). I can't find anything really useful in the install manual or using Google. I will probably end up looking really stupid when someone points out an obvious mistake I have made, but I can live with that. If anyone has got etch installed on the beige g3 (its a 266 mhz machine, but I can't tell you the firmware version, etc. Would need to find out how to get at this) and can share their expertise, it would be most appreciated. Thanks in advance, Nick. * Dr. N.R. Helps Medical Research Council Protein Phosphorylation Unit College of Life Sciences MSI/WTB/JBC Complex University of Dundee Dundee DD1 5EH Scotland t: 44 (0)1382 384745 (office) t: 44 (0)1382 388019 (lab) f: 44 (0)1382 388729 e: [EMAIL PROTECTED] w: http://www.dnaseq.co.uk/ w: http://www.dundee.ac.uk/lifesciences/mrcppu/
beige g3 etch install - working
Continuation bottom posted... * Dr. N.R. Helps Medical Research Council Protein Phosphorylation Unit College of Life Sciences MSI/WTB/JBC Complex University of Dundee Dundee DD1 5EH Scotland t: 44 (0)1382 384745 (office) t: 44 (0)1382 388019 (lab) f: 44 (0)1382 388729 e: [EMAIL PROTECTED] w: http://www.dnaseq.co.uk/ w: http://www.dundee.ac.uk/lifesciences/mrcppu/ >>> Nicholas Helps 11/15/07 3:24 PM >>> Hello all, I have been working with Debian for quite a few years now and using powermac G3 machines with extra network cards in them as routers and firewalls, etc. These machines were set up back in the days of Woody and have been kept uptodate with security updates, but otherwise pretty much left untouched. Due to the issues around the firmware in these machines and because it was useful, I always set them up to boot initially into mac OS9 then use BootX to hand over the Debian. This worked a treat. However, I thought it would be good come up to date by installing Etch instead. I used a free machine that was not actually in use and ran the install using the current network install ISO. Things have changed since the days of woody and it now seems that floppy images (boot image and root image) are no longer used. Hence, I copied the initrd.gz file over to the mac HD and set that as the ram disk for the install. I also copied across the linux kernel and put that into the kernels folder in the system folder. Using that allows me to boot into the installer and using the installer I deleted the previous linux partition (hda7) and swap (hda8) and made new ones. Then installed the base system, etc,etc all the way through to where it runs tasksel. I just leave that at the basic system for now. Following on some more, finally we get to the point of trying to install Quik (which I don't need) and it gives an error anyway, since I have selected ext3 file system that is not supported in quik. I therefore say to carry on without a boot loader. Everything goes fine all the way to rebooting into the new system. However, when I do that, OS9 will not boot up. I just get the flashing disk symbol with a question mark on it. Popping the OS9 CD and booting off that and then running disk setup shows me that the HD has somehow been altered so it is not recognised properly as a mac HD. During the partitioning step, I did not alter anything other than hda7 and 8. I have found that I can reinstall the apple hard disk driver onto the disk and this then gets OS9 up and working. However, I cannot then boot into Debian, since the boot process gets a little way in and then I get a kernel panic at the point where it tries to mount the file system (error about no file system at /dev/hda7). I have done this several times now and the same thing happens every time. The install goes fine but then I end up with a completely unusable machine. I am wondering if I am going about the install process wrongly (ie using the initrd.gz file). I can't find anything really useful in the install manual or using Google. I will probably end up looking really stupid when someone points out an obvious mistake I have made, but I can live with that. If anyone has got etch installed on the beige g3 (its a 266 mhz machine, but I can't tell you the firmware version, etc. Would need to find out how to get at this) and can share their expertise, it would be most appreciated. Thanks in advance, Nick. So, here is the way I got it to work First off, I have to thank Rick for a very helpful discussion about this. Without his help, I would not have been able to get this baby working! I have OS9.2 on a circa 500 meg HFS+ partition, leaving the rest for debian. I use BootX as a boot loader. I used the netinstall CD for Etch. First boot into OS9 with the Etch CD in the CD drive. Copy the vmlinux kernel from the CD install directory and put it into the kernels folder in the system folder. Copy the initrd.gz image to the BootX folder. Select the BootX application and tell it to use the vmlinux kernel and set the RAM disk to initrd.gz image. In the additional kernel arguments box, type "DEBCONF_PRIORITY=low" (no quotes). Boot into linux. The installer will start. Go through this answering questions. When you get to the partitioning section, Partman will run. You should choose "manual" as your partitioning method. Select the free space and make a linux partition with Ext3 format and choose to mount "/" on it. On my machine this then becomes /dev/hda7/. Also make a swap partition (/dev/hda8). The sizes of these will depend on your HD size. You can also obviously choose to split up your file system so that not everything is under the root (/). I also named my /dev/hda7/ as "debian". Probably makes no difference, but I did notice th
Fwd: beige g3 etch install - working (slight edit)
* Dr. N.R. Helps Medical Research Council Protein Phosphorylation Unit College of Life Sciences MSI/WTB/JBC Complex University of Dundee Dundee DD1 5EH Scotland t: 44 (0)1382 384745 (office) t: 44 (0)1382 388019 (lab) f: 44 (0)1382 388729 e: [EMAIL PROTECTED] w: http://www.dnaseq.co.uk/ w: http://www.dundee.ac.uk/lifesciences/mrcppu/ >>> Nicholas Helps 11/28/07 3:14 PM >>> Continuation bottom posted... (with slight edit due to a typo in the original post) Hello all, I have been working with Debian for quite a few years now and using powermac G3 machines with extra network cards in them as routers and firewalls, etc. These machines were set up back in the days of Woody and have been kept uptodate with security updates, but otherwise pretty much left untouched. Due to the issues around the firmware in these machines and because it was useful, I always set them up to boot initially into mac OS9 then use BootX to hand over the Debian. This worked a treat. However, I thought it would be good come up to date by installing Etch instead. I used a free machine that was not actually in use and ran the install using the current network install ISO. Things have changed since the days of woody and it now seems that floppy images (boot image and root image) are no longer used. Hence, I copied the initrd.gz file over to the mac HD and set that as the ram disk for the install. I also copied across the linux kernel and put that into the kernels folder in the system folder. Using that allows me to boot into the installer and using the installer I deleted the previous linux partition (hda7) and swap (hda8) and made new ones. Then installed the base system, etc,etc all the way through to where it runs tasksel. I just leave that at the basic system for now. Following on some more, finally we get to the point of trying to install Quik (which I don't need) and it gives an error anyway, since I have selected ext3 file system that is not supported in quik. I therefore say to carry on without a boot loader. Everything goes fine all the way to rebooting into the new system. However, when I do that, OS9 will not boot up. I just get the flashing disk symbol with a question mark on it. Popping the OS9 CD and booting off that and then running disk setup shows me that the HD has somehow been altered so it is not recognised properly as a mac HD. During the partitioning step, I did not alter anything other than hda7 and 8. I have found that I can reinstall the apple hard disk driver onto the disk and this then gets OS9 up and working. However, I cannot then boot into Debian, since the boot process gets a little way in and then I get a kernel panic at the point where it tries to mount the file system (error about no file system at /dev/hda7). I have done this several times now and the same thing happens every time. The install goes fine but then I end up with a completely unusable machine. I am wondering if I am going about the install process wrongly (ie using the initrd.gz file). I can't find anything really useful in the install manual or using Google. I will probably end up looking really stupid when someone points out an obvious mistake I have made, but I can live with that. If anyone has got etch installed on the beige g3 (its a 266 mhz machine, but I can't tell you the firmware version, etc. Would need to find out how to get at this) and can share their expertise, it would be most appreciated. Thanks in advance, Nick. So, here is the way I got it to work First off, I have to thank Rick for a very helpful discussion about this. Without his help, I would not have been able to get this baby working! I have OS9.2 on a circa 500 meg HFS+ partition, leaving the rest for debian. I use BootX as a boot loader. I used the netinstall CD for Etch. First boot into OS9 with the Etch CD in the CD drive. Copy the vmlinux kernel from the CD install directory and put it into the kernels folder in the system folder. Copy the initrd.gz image to the BootX folder. Select the BootX application and tell it to use the vmlinux kernel and set the RAM disk to initrd.gz image. In the additional kernel arguments box, type "DEBCONF_PRIORITY=low" (no quotes). Boot into linux. The installer will start. Go through this answering questions. When you get to the partitioning section, Partman will run. You should choose "manual" as your partitioning method. Select the free space and make a linux partition with Ext3 format and choose to mount "/" on it. On my machine this then becomes /dev/hda7/. Also make a swap partition (/dev/hda8). The sizes of these will depend on your HD size. You can also obviously choose to split up your file system so that not everything is under the root (/). I also named my /dev/hda7/ as "
Re: error: eth1: switching to forced 10bt
Hello Peter, Wolfgang has given you lots of info and I am not sure I can help much. However... This is an info message about your network and it appears that the linux kernel keeps trying to reconfigure your ethernet interface (eth1) to 100 baseT (that is the bt notation) and then to 10 baseT speed. If nothing is plugged into the port, this would seem odd, since speed changes normally only occur when something changes on the network (eg you plug a network cable into the socket and the port changes to full duplex). I am also wondering why on a standard PPC computer your default network port is eth1. Normally, if you only have one port (do you have more than one? Are they/it plugged into anything?), it would be eth0... To be able to make changes without the messages causing problems, you can change to another console. Use the keyboard command (press the alt key and, while holding down alt, press the right arrow key). You can now log into a second console where the message will not keep coming up (to get back to the other console, press "alt-left arrow). There are four consoles you can have on the go at one time; try not to get confused about which you are working in! Try some of the commands that Wolfgang suggested. Also a simple "ifconfig" command with no further arguments would be useful. That will print out on screen what you current network port configuration is. That output would be useful to see. Google is definitely your friend when it comes to linux problems. Almost certainly someone has seen this before and asked the same question on a forum. Google might just find that thread and your answer for you. Cheers, Nick. * Dr. N.R. Helps Medical Research Council Protein Phosphorylation Unit College of Life Sciences MSI/WTB/JBC Complex University of Dundee Dundee DD1 5EH Scotland t: 44 (0)1382 384745 (office) t: 44 (0)1382 388019 (lab) f: 44 (0)1382 388729 e: [EMAIL PROTECTED] w: http://www.dnaseq.co.uk/ w: http://www.dundee.ac.uk/lifesciences/mrcppu/ >>> Peter O'Doherty <[EMAIL PROTECTED]> 01/12/08 10:22 AM >>> Thanks a lot for your help Wolfgang. When I try sysv- rc ...etc I get the following error: - bash sysv- rc- conf: command not found Please bear with me as I am a total linux newbie so I have no idea how to find these scripts (/etc/init.b/ or /etc/rcX.d). In any case these messages (about eth1) keep appearing without giving you the chance to type uninterrupted. (Is there a connection here with XServer too as the first message I got was that it was not working so I chose the option to disable it?) Thanks, Peter >> I'd try this: >> >> sysv- rc- conf networking off >> >> more on that: >> http://wolfgangpfeiffer.com/foolinglinux.html#rc >> >> But I'm not sure, because I just realized my networking routines, >> here >> on Debian/unstable, seem to react weird: >> >> # ifdown eth0 >> ifdown: interface eth0 not configured >> >> But when running 'ifconfig', eth0 still is showing up ... >> >> so you might want having a look to the scripts in /etc/init.b/ or in >> /etc/rcX.d if the 'sysv- rc- conf' incantation does not help ...
Re: error: eth1: switching to forced 10bt
Hello Peter, Makes sense. Certainly on OSX, firewire ports are included in the list of network interfaces. I have never put Debian onto anything more modern than a Beige G3. Since these don't have firewire, all my installs have shown the built in ethernet port to be eth0 and additional PCI ethernet cards as eth1, etc. I guess I should start using more uptodate hardware ;-) Nick. * Dr. N.R. Helps Medical Research Council Protein Phosphorylation Unit College of Life Sciences MSI/WTB/JBC Complex University of Dundee Dundee DD1 5EH Scotland t: 44 (0)1382 384745 (office) t: 44 (0)1382 388019 (lab) f: 44 (0)1382 388729 e: [EMAIL PROTECTED] w: http://www.dnaseq.co.uk/ w: http://www.dundee.ac.uk/lifesciences/mrcppu/ >>> Peter Rooney <[EMAIL PROTECTED]> 01/17/08 9:11 AM >>> Dear Nicholas, This happens on some of my machine also, (in particular, an iBook G4 and a Blue & White G3) I recall from reading some messages at boot time text that the firewire port gets called eth0. Peter ( Rooney ) Nicholas Helps wrote: [ snip ] > I am also wondering why on a standard PPC computer your default network port > is eth1. Normally, if you only have one port (do you have more than one? Are they/it plugged into anything?), it would be eth0... [ snip ] -- To UNSUBSCRIBE, email to debian- powerpc- [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: G3 Beige Tower install help
Hello Stephen / all, Here is a post I put on some time ago about this (Stephen, this is the post I mentioned). There is a preamble and then the step by step that worked for me. I have added a little bit about the space you will leave for Debian on your HD. This was not in the original post. You don't want to format that space using OS9, just leave it as free space and let the Debian installer do that for you. Good luck... Nick. Hello all, I have been working with Debian for quite a few years now and using powermac G3 machines with extra network cards in them as routers and firewalls, etc. These machines were set up back in the days of Woody and have been kept uptodate with security updates, but otherwise pretty much left untouched. Due to the issues around the firmware in these machines and because it was useful, I always set them up to boot initially into mac OS9 then use BootX to hand over the Debian. This worked a treat. However, I thought it would be good come up to date by installing Etch instead. I used a free machine that was not actually in use and ran the install using the current network install ISO. Things have changed since the days of woody and it now seems that floppy images (boot image and root image) are no longer used. Hence, I copied the initrd.gz file over to the mac HD and set that as the ram disk for the install. I also copied across the linux kernel and put that into the kernels folder in the system folder. Using that allows me to boot into the installer and using the installer I deleted the previous linux partition (hda7) and swap (hda8) and made new ones. Then installed the base system, etc,etc all the way through to where it runs tasksel. I just leave that at the basic system for now. Following on some more, finally we get to the point of trying to install Quik (which I don't need) and it gives an error anyway, since I have selected ext3 file system that is not supported in quik. I therefore say to carry on without a boot loader. Everything goes fine all the way to rebooting into the new system. However, when I do that, OS9 will not boot up. I just get the flashing disk symbol with a question mark on it. Popping the OS9 CD and booting off that and then running disk setup shows me that the HD has somehow been altered so it is not recognised properly as a mac HD. During the partitioning step, I did not alter anything other than hda7 and 8. I have found that I can reinstall the apple hard disk driver onto the disk and this then gets OS9 up and working. However, I cannot then boot into Debian, since the boot process gets a little way in and then I get a kernel panic at the point where it tries to mount the file system (error about no file system at /dev/hda7). I have done this several times now and the same thing happens every time. The install goes fine but then I end up with a completely unusable machine. I am wondering if I am going about the install process wrongly (ie using the initrd.gz file). I can't find anything really useful in the install manual or using Google. I will probably end up looking really stupid when someone points out an obvious mistake I have made, but I can live with that. If anyone has got etch installed on the beige g3 (its a 266 mhz machine, but I can't tell you the firmware version, etc. Would need to find out how to get at this) and can share their expertise, it would be most appreciated. Thanks in advance, Nick. So, here is the way I got it to work First off, I have to thank Rick for a very helpful discussion about this. Without his help, I would not have been able to get this baby working! I have OS9.2 on a circa 500 meg HFS+ partition, leaving the rest for debian (I will add here something I didn't mention in the original post - don't format this space using the OS9 CD, just leave it as "free space"). I use BootX as a boot loader. I used the netinstall CD for Etch. First boot into OS9 with the Etch CD in the CD drive. Copy the vmlinux kernel from the CD install directory and put it into the kernels folder in the system folder. Copy the initrd.gz image to the BootX folder. Select the BootX application and tell it to use the vmlinux kernel and set the RAM disk to initrd.gz image. In the additional kernel arguments box, type "DEBCONF_PRIORITY=low" (no quotes). Boot into linux. The installer will start. Go through this answering questions. When you get to the partitioning section, Partman will run. You should choose "manual" as your partitioning method. Select the free space and make a linux partition with Ext3 format and choose to mount "/" on it. On my machine this then becomes /dev/hda7/. Also make a swap partition (/dev/hda8). The sizes of these will depend on your HD size. You can also obviously choose to split up your file system so that not everything is under the root (/). I also named my /dev/hda7/ as "debian". Probably makes no difference, but I did notice that the HFS+
Re: G3 Beige Tower install help
* Dr. N.R. Helps Medical Research Council Protein Phosphorylation Unit College of Life Sciences MSI/WTB/JBC Complex University of Dundee Dundee DD1 5EH Scotland t: 44 (0)1382 384745 (office) t: 44 (0)1382 388019 (lab) f: 44 (0)1382 388729 e: [EMAIL PROTECTED] w: http://www.dnaseq.co.uk/ w: http://www.ppu.mrc.ac.uk >>> Stephen Allen <[EMAIL PROTECTED]> 09/08/08 11:38 PM >>> On Mon, Sep 08, 2008 at 10:41:09AM +0100, Nicholas Helps wrote: > Hello Stephen / all, Hi Nicholas: > So, here is the way I got it to work [ ...] Excellent instructions, thanks very much Nicholas ! I'm writing this from my G3 Beige booted into Lenny- Debian using mutt. :) I'm a happy camper. I think you have just written the best SxS I've ever seen. You should write technical manuals. The thing about updating the MacOS disk driver is all it took, and I was able to boot into the install I had done from yesterday. Yay !! Thanks to *Everyone* who helped here. Amazing that there are people around that remember the incantations one has to do to get these Old World boxen to work with a current Linux. Nicolas, do you mind if I add these instructions to the Debian Wiki ? I'll be sure to give full attribution. Cheers. -- Regards, S.D.Allen - Toronto -- To UNSUBSCRIBE, email to debian- powerpc- [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] Hello Stephen (and everyone else), Being polite this time ;-) Glad the step by step worked. Please make sure you acknowledge Rick as well, without this help, I would never have gotten my install working either The list works! Having read the other posts about the expert installer setting, I would add that in my experience setting the installer to expert did not allow me to avoid the installer trying to install quik. On older versions (eg Woody) even without putting the installer in expert mode one could stop this (and the install manual explained how). With Etch, the installer just pushed ahead regardless of what I tried to do to stop it. Fortunately, since quik does not work with ext3 file systems, it aborted. Hope you enjoy your new system. I have found Debian to run extremely well on these systems. Best wishes, Nick. The University of Dundee is a registered Scottish charity, No: SC015096 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: G3 Beige Tower install with BootX - help
>>> Rick Thomas <[EMAIL PROTECTED]> 09/09/08 8:49 AM >>> > > Hello all, > > I have been working with Debian for quite a few years now and using > powermac G3 machines with extra network cards in them as routers > and firewalls, etc. These machines were set up back in the days of > Woody and have been kept uptodate with security updates, but > otherwise pretty much left untouched. Due to the issues around the > firmware in these machines and because it was useful, I always set > them up to boot initially into mac OS9 then use BootX to hand over > the Debian. This worked a treat. > > However, I thought it would be good come up to date by installing > Etch instead. I'll bring this up to Lenny. [Rick] > I used a free machine that was not actually in use and ran the > install using the current network install ISO. I used the Lenny Beta2 "businesscard" install disk. > Things have changed since the days of woody and it now seems that > floppy images (boot image and root image) are no longer used. > Hence, I copied the initrd.gz file over to the mac HD and set that > as the ram disk for the install. I also copied across the linux > kernel and put that into the kernels folder in the system folder. In a little more detail: I drag- n- dropped the "vmlinux" and "initrd.gz" files from the Debian install CD (they are in the "install/powerpc" directory) into the MacOS9 "System Folder:Linux Kernels" folder. Then I chose those two as the "Kernel" and "use specified RAM disk" options in BootX. I set the "More kernel arguments" box to "DEBCONF_PRIORITY=low" (without the quotes) to specify that I wanted to run the installer in "expert" mode, so I could do a couple of non- default tricks needed by OldWorld Macs using BootX. You may or may not want to check the "No video driver" checkbox, depending on your video hardware. This is the equivalent of the "video=ofonly" in yaboot. You'll have to experiment to find out which option works for you. You may (probably will) want to change the "Ramdisk size" option to a larger number. I use 32768. The default is 8192. > Using that allows me to boot into the installer where I answered the questions in the usual way, until I got to the step "Load Installer Components from CD". There I chose the option to install "hfs- modules": HFS filesystem support. We'll need them later on when we copy the new customized kernel and initrd to the MacOS9 partition. This is the first "non- default trick" for BootX installation, for which we need to be in "expert mode". There will be another occasion later. > and using the installer partitioner in the installer > I deleted the previous linux partition (hda7) and swap (hda8) and > made new ones. I used the "manual partitioning" option and created "root" and "swap" partitions. Note that these two partitions should have single digit numbers. Otherwise, during the reboot following the installation, the Linux boot process will hang "waiting for the root partition". I suspect this is a bug somewhere in the code that decodes the kernel arguments. Putting this restriction another way, the root and swap partitions should each be chosen from hda7, hda8, or hda9 (assuming your MacOS9 partition is hda6, as it usually will be.) If either of them are hda10 or greater, you'll have problems later. You can use two- digit partition numbers for things like "/home" and "/usr", it's just the root that is restricted. The "guided partitioning" will try to create an ext2 "boot" partition. This is necessary for the quik bootloader, but completely *un*necessary for BootX. In fact, it's actually undesirable because I've recently discovered that the default size (8 MB) for the "boot" partition is too small -- kernel and initrd have grown since quik was written. So including /boot as just a directory in the root partition allows it to have arbitrarily large contents. While you're there in the partitioner, make a note of the partition numbers of your MacOS9 and root partitions. You'll want them later. > Then installed the base system, etc,etc all the way through to > where it runs tasksel. I just leave that at the basic system for now. > Following on some more, finally we get to the point of having finished "Select and install software" where the next thing it would want to do is "Install quik on a hard disk". You do *not* want to install quik. You've got MacOS9 and BootX to do that job. So skip over that line and continue with "Continue without boot loader". It will tell you that you need to use the vmlinux from the "boot" partition and set the kernel parameter "root=" to the root partition that you just installed into. Write down the root partition number (if you didn't do so during the partitioning step) you'll need it later. At this point you should switch to a different console (hit the "alt" and "F2" keys) and do t
starmax 3000 boot settings
Hello all, This is no doubt old hat to all you dyed in the wool linux people, but I'm fairly new to the game and so I am having a few problems. I'll briefly tell you what I have done and where I have gotten stuck at. Downloaded woody CD images (first two only for the time being - CD#1 is NONUS version) and burned ISO images. Downloaded the two boot install images and made floppies. Backup up all my Mac OS stuff off my internal IDE (1.2GB) HD. Inserted floppy 1 and booted. Inserted floppy 2 at request and got to install menu. Chose language, chose kbd, etc and made it all the way through to make system bootable. As far I could tell everything was fine and I installed the base system and kernel. I partitioned the HD so that there is a partition map at hda1(few KB), a linux native at hda2 (1.1GB) and a linux swap at hda3 (100MB). Once I made the system bootable (using quik), I restarted. The first time I did this all I got was a blank screen. I then went off and read about quik and open firmware. I also tried zapping the pram. This didn't help. I then booted back into the boot floppies and once everything got going, I opened a second consol and used nvsetenv to look at and alter the firmware: Set input/output devices to kbd and screen. Auto-boot left set at true. Set boot-device to ata/[EMAIL PROTECTED]:0 no boot-file set On restart I get a white screen with some text on it about quik second stage boot, woody.., then a boot: prompt. I tried entering "linux", "boot", "vmlinux". It kept asking for a path to the kernel. So I tried things like dev/hda2/vmlinux, etc. None of these worked. Back into the install process again and using a second consol showed that there does not appear to be a linux kernel on the HD. From what I understand it should be at the root (/) directory. All there is at that place is "rclinux". There is no "vmlinux". From what I read, this is what the kernel should be called. The install manual mentions very little about anything like this and basically says quik will set up everything. Presumably I am missing something very simple. I did see a "quirk" of "quik" for the starmax where it said what the boot-device should be set to (as above) and somewhere else that is said you usually do not need to set a boot-file. To be honest, I am getting lost in all the firmware and quik stuff. As I said above, I can't find vmlinux anywhere on my HD and I also can't find a quik.conf file. Without either of these, I would imagine that linux can't boot. It might also explain why all the common suggestions for solving problems by entering "default values" (like debian, or linux, or vmlinux) don't work. Can anyone give me some pointers and indications as to how to proceed and why vmlinux and quik.conf don't appear to be on the HD even though the installer should have installed them? Many thanks in advance, Nick.
starmax 3000 install problems
Hello all, Thanks for the comments on this. What solved it was redoing the "make system bootable" bit again (this sorted out the boot-device setting) and then just entering "Linux" at the boot: prompt. I did have to go back and start the whole install process again prior to this, since the installer got confused otherwise and asked for log-in password that had not yet been set... After all this, I get to the point where I can boot back into linux and it starts off the HD. However, one other little problem is that X will not start. It chucks up some errors about not being able to start the server, then bombs out and goes to consol mode. Maybe I have not set all the properties right during the installation process. I am thinking it might be screen res or colour depth (I also mucked up the mouse setting and used a PS2 mouse not the setting that is listed in the installer... whoops!). Anyone got some ideas as to how to alter this (? x.conf file) and what the sort of settings might be? I have what came with the starmax in terms of display hardware (1MB of video ram, the standard built-in display adaptor) and a 15" apple multiscan monitor. The mouse is the standard ADB mouse supplied with the system. Cheers, Nick.
Re: Installing Debian(Woody) on a 7200, again
>To:debian-powerpc@lists.debian.org >Subject: Installing Debian(Woody) on a 7200, again >Reply-to: [EMAIL PROTECTED] >From: "Lorenzo Thurman" <[EMAIL PROTECTED]> >Cc:Date: Thu, 27 Mar 2003 09:35:36 -0500 (EST) > > >I tried the boot floppy 5 times with disks made from three differnet computers >and it still will not work. The >same thing happens everytime, the boot process starts, but after miboot kicks >in, the s >reen goes black and loses sync, then nothing. Anyone h >ave any ideas? I thought Debian was supposed to have a better installer, so >far this has been a waste of time. >I've used Mandrake and LinuxPPC in the past and they worked w/o too much >trouble on thi > computer, so I can't understand why Debian has to be >so difficult. > > > > >___ >Join Excite! - http://www.excite.com >The most personalized portal on the Web! > > >-- >To UNSUBSCRIBE, email to [EMAIL PROTECTED] >with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > > Hello Lorenzo, I completed an install on a 7200 last night with no problems at all (not what you wanted to hear ;-). Anyway, I wonder if you have some nonstandard hardware installed - especially video hardware - or have set some firmware settings in the past that cause the installer problems. You could try resetting the firmware before installation and that might help. I'm only a newbie myself, but the 7200 seems to be one of the powermac architectures that is very well supported by debian and should not require any "tweeks" to get the install to work. Best of luck, Nick.
fbset on woody
Hi all, Does fbset work properly with woody on a starmax 3000? I have just put debian onto this machine and after configuring the xserver during installation (800 x 600 with 16 bit colour) got horrible colours when the desktop environment started (it looked like it may 16 colours - ie not a bit depth of 16, but 16 actual colours). I then went back and used the dpkg reconfigure tool for xserver to change this to 8 bit colour (256 colours) and got something acceptable at 1024 x 768 res. However, I would like to be able to change these settings without having to go back and reconfigure every setting in the xserver (ie without using dpkg reconfigure xserver xfree86). I found a help page on a web site about debian (not THE debian www site) that referred to fbset for setting frame buffer variables. This seems to be the right tool to use. It was not installed as part of the standard install, but apt-get did find it on the distribution CD and managed to install it. I ran it and told it to set the x and y resolutions to "-yres 800 -xres 600" (a suggested resolution). The screen went blank and I had to do a restart. It does seem strange that it quotes yres as 800, since I always thought that screen res would be (eg) 640 x 480, 800 x 600, etc with the larger figure being the xres. I also tried it that way round and the screen went blank again. The info on the help page did say that "whacky" settings would cause the screen to blank. However, the above resolutions don't seem "whacky" to me (at least one way round - whichever is the "right" way). Anyone got any ideas, or another (standard?) way to alter xserver resolutions in debian on the mac? Thanks in advance, Nick.
ppp...problem
Hi all, I have woody on a starmax 3000 with a GV 56K external modem. All works fine from within mac OS (dual boot OS9 and woody on same machine), but I can't get a ppp connection from debian. I used pppconfig to set up the necessary configuration files and then type "pon" from a shell script. The modem picks up, makes noises, it gets a "link" to the ISP and then after about 1 minute the connection drops. During this time I cannot ping anything (eg the name servers in the ISP) or use a www browser to access any sites (I get an error "server not found"). I have also tried wvdial and this is more informative. Everything seems fine all the way through to "initiating pppd, pid set to...". ie it dials, picks up, authenticates. If I use top to see active processes, I can't see pppd listed and so it almost looks like pppd fails to start up. If I just type pppd at the shell prompt, I get a message something like "the remote device requires to authenticate, but I could not find an appropriate secret (password) to use". The setting in the pppconfig file is for "noaouth", which I believe means that pppd will not require an authentication. I also wonder if it has anything to do with my hosts.allow/deny files. However, I set hosts.allow to ALL:ALL before the other entry and then tried again and there was still no joy. The correct name servers and search domain are listed in the appropriate file (that I forget the name of right now) - although this would not stop me pinging ip addresses. I have found very little trouble shooting info anywhere about this and all the configuration tutorials etc, don't seem to mention this apparent problem. I'm rather stuck at this point and hope that someone else on the list might have an idea or two. Thanks in advance, Nick.
Re: ATI MACH64 - wrong colors
>Date: Sun, 20 Apr 2003 02:28:20 +0200 >To:debian-powerpc@lists.debian.org >Subject: ATI MACH64 - wrong colors >From: Marco Wilka <[EMAIL PROTECTED]> > >Hi, > >I am using Debian unstable on a PowerMac 4400 with an ATI MACH64 VT >graphics adapter and have the following problem when >using XFree86: > > - the colors are only correct when in 8-Bit-Mode > - 16 bit depth/16 bit fbbpp yields in "psychedelic" color shifts > - when using 24 bit depth, red shows up as blue, and everything > that is meant to be blue appears as brown > - (15 bpp doesn't work at all) > >This problem occurs with kernel image 2.2.20 and 2.4.18 as well as >with xserver version 4.1 and 4.2. Kernel command line arguments >don't seem to affect my xserver's behaviour in any way. If I start >with BootX instead of quik and use the "No video driver" option, >the screen output is rendered unreadable. > >Was anyone successful to get the xserver working in more than >8 bit on a stock 4400? Anything I could find on the web on that >topic didn't lead to anything so far, so any help would be >appreciated. > >Thanks in advance, > >Marco > > > >-- >To UNSUBSCRIBE, email to [EMAIL PROTECTED] >with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > > Hi Marco, The Starmax 3000 series is basically the same as the 4400 and also has the same problem (my 3000 also only gives anything reasonable at 8 bit colour). Nobody seems to have a solution to this yet. Nick.
Re: Installation problems on StarMax
Hi there, Chris is right. They are oldworld and quik will work (at least on my 3000/180). You need to set (at least initially) openfirmware to not autoboot. You also need to set it to use the keyboard and monitor as input/output for OF (the default on mine is to use the serial port, which is NOT a good idea unless you want to try to get a second computer to act as a consol - don't even think about going there). You should also set the boot path and default kernel image name. You can change all these settings from a second screen when you are running off the boot floppies. Load them as if you were doing a new install and get all the way to the screen about keyboard configuration. Then use the keyboard combo of "alt right arrow" to activate a new screen. You can now use "nvsetenv" to alter OF settings. This is a linux application provided with the boot floppies and loaded into the ram disk. I can't remember all the settings off the top of my head now and I am not at the machine. I'll get the details together and post a step by step tomorrow... Nick. >Date: Sun, 1 Jun 2003 14:10:47 -0700 >From: Thanasis Kinias <[EMAIL PROTECTED]> >To:debian-powerpc@lists.debian.org >Subject: Installation problems on StarMax > >Greetings, > >I'm installing Woody on a StarMax 5000/300. It's got an IDE HDD and >SCSI CD-ROM and ZIP. It will not boot from the CD-ROM drive, so I have >to use floppies to install. I've gotten as far as ``Make System >Bootable''; unfortunately, after the installer puts Quik on the system, >it is not bootable. > >As I understand it, StarMaxes are considered NewWorld machines -- they >should use Yaboot then, not Quik, right? (I'm really not very familiar >with PowerPC hardware, having only done one installation on an iMac >before.) There is no boot floppy for NewWorld that I can find, only a >rescue.bin which is not bootable (shouldn't it be though?). > >I assume it is trying to install a useless Quik because I used the >powermac instead of new-powermac, but I couldn't find another option. > >If I am correct that this box needs Yaboot and not Quik, how do I get >that installed? ATM the box is not bootable except via floppy, and the >only bootable floppy I has just dumps me right to the installation. > >Any assistance anyone can provide would be much appreciated. > >Regards, >-- >Pax vobiscum; pax cum omnibus. > >Thanasis Kinias >tkinias at asu.edu >Doctoral Student, Department of History >Arizona State University >Tempe, Arizona, U.S.A. > > >-- >To UNSUBSCRIBE, email to [EMAIL PROTECTED] >with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > > * Dr. N.R. Helps Medical Research Council Protein Phosphorylation Unit School of Life Sciences MSI/WTB Complex University of Dundee Dundee DD1 5EH Scotland tel: 44 (0)1382 344239/8019 44 (0)7989 197916 (mobile) fax: 44 (0)1382 223778 email: [EMAIL PROTECTED] Web Pages: http://www.dnaseq.co.uk/ http://www.dundee.ac.uk/lifesciences/mrcppu/
crash investigation
Hi all, As a veritable newbie (at least in comparison) to Linux, I would like some suggestions as to how one goes about investigating system crashes. I have set up a couple of systems which generally run fine. However, today one of my debian systems stopped responding and had to be "hard" rebooted (ie press the power button - Ouch..!). This machine is run from the command prompt with no GUI and has a very limited set of software running on it (essentially it is running as a firewall and dhcp server and nothing else). I am keen to find out why the crash might have happened. I have looked at various log files in /var/log/, but none seem to suggest anything obvious. Could someone with more experience suggest other information that might be help an investigation? The machine has not crashed before and so maybe it is just a "one off". However, without being able to ascertain (as far as one might be able to) what caused it, I am in the dark. Any suggestions welcome (even if you think it is obvious or "in the manual", because I haven't read all the manuals yet... ;-) Thanks, Nick.
