Toshiba Satellite T1910CS notebook
The following is advice for anyone who runs into the syslinux problem (rescue disk doesn't boot, computer screen displays "boot failed") when trying to install a Debian installation from floppies (maybe some other distributions, as well). The machine here is a Toshiba Satellite T1910CS, 486, 110M HD, 8MB ram . 1) Download resc1400.bin from Debian via ftp. 2) dd that file to a floppy. 3) mount the floppy and copy the files named "linux" (kernel) and "root.bin" (ramdisk image) to a local hard disk or some other medium. I understand that some people prefer to use mtools for this task. I just mounted the floppy as msdos and copied the files out that way. 4) dd "linux" to a floppy. dd "root.bin" to the same floppy using an offset of 720 or so, e.g., dd if=root.bin of=/dev/fd0 bs=1k seek=720 (for a floppy disk at /dev/fd0 . the value for seek needs to be a bit larger than the size of the kernel file being installed). 5) use the rdev command as follows: rdev /dev/fd0 /dev/fd0 rdev -r /dev/fd0 49872 *** The resulting floppy will boot the machine you are trying to install Debian to. You will be asked to insert a root disk. Here you just hit the return key and follow the instructions for the first part of the installation. When you are asked for the driver disk you put a floppy which contains drv1400.bin (also from Debian via ftp). You may have to enter /dev/fd0/ if asked by the program where the device driver is. *** Note: Things should proceed from here without problem except for one important point. When you attempt to install the "base" files (from Devian via ftp) via nfs, hard disk, floppy, etc., you will be asked to insert "the rescue disk" in the floppy drive to start up the process. Here, the program is looking for files in resc1400.bin other than "linux" or "ramdisk". In order to satisfy the program and move on to the next step in the installation, you have to insert a floppy which has the downloaded resc1400.bin dd 'ed to it (one that most people can boot from in the first place - with syslinux in it). Once you do that you can move on. In all you will need a regular rescue disk, a stripped rescue disk (w/ kernel and ramdisk files, prepared w/ redev), a driver disk and persistence. Tony Laszlo, [EMAIL PROTECTED] Jiyugaoka, Tokyo
Toshiba Satellite T1910CS notebook - preferred version?
I am trying to install Debian on a Toshiba Satellite T1910CS, 486, 110M HD, 8MB ram . I won't be running X and will just need to use: vim, pine (+fetchmail/sendmail), lynx, ftp, telnet and pcmcia and parallel port modules. Have gone through the installation steps from floppy to the point where the system boots from the hard disk. Next, I need to get an Corega EtherII PCC-T ethernet card working in order to point dselect to the Packages.gz on the debian ftp or www site. My questions: * In accordance with advice from someone who had installed Linux on one of these beasts, I chose Debian 2.0 distribution (/dists/Debian-2.0/). Is this the preferred version for this machine and these needs? * while uname -a shows that the kernel running is 2.0.36, somehow modules are 2.0.34 . I tried to insmod pcmcia.o , etc. and got an error message that the modules don't machine the kernel. How could this happen and what's the best way to reinstall so it doesn't happen or fix the problem without reinstalling? * To see what would happen, I downloaded the pcmcia-modules deb file (2.0.36) from the Debian-2.0 site (copying it over to the Toshiba via floppies). This I was able to install with dpkg -i ; so, while the modules are 2.0.34, there are two sets of pcmcia modules, 2.0.34 and 2.0.36 . The Ethernet card requires pcnet_cs.0 so I entered the necessary lines in /etc/pcmcia/config and started pcmcia with "/etc/init.d/pcmcia start" (I have this card working on a Linux box that runs with Turbolinux [quite similar to Redhat]). On the Toshiba w/ Debian it's not perfect yet, but the card is being recognized, at least partially. Anything else I need to do? Does the network need to be configured before the card will be recognized properly? * Finally, I downloaded the kernel source from /Debian-2.0/ and tried dpkg -i. It seems that I need to install binutils first, or at the same time. Anything else needed before I can recompile the kernel? bin86, maybe? kernel headers of some kind? Thanks! Tony Laszlo Jiyugaoka, Tokyo
dselect and Packages.gz
I have installed Debian 2.0 (kernel 2.0.34) from floppies and have booted from the hard disk after installing all the base.tgz. Do I need to install the rest of the packages from the old 2.0 archives, or from the current files? I searched for 2.0 files and found the following. ftp.debian.org/debian-archive/dists/Debian-2.0 Dselect, however, says it cannot find Packages.gz when I instruct it to access the site and directories below (or anything other than the current version's packages.gz). What can I do? Thanks. Tony Laszlo Tokyo
swap partition too small
In trying to conserve the little hard disk space (110M) that I have, I have unfortunately made the swap partition too small when installing Debian. I know I should probably pay penance and do the install all over again, but am wondering if there is a way to carve out a portion of the hard disk for swap at this stage of the game. Something like a fips that can shrink a Linux partition. No? Just thought I'd ask before reinstalling... Tony Laszlo Tokyo
Re: swap partition too small
Many thanks to Jeff and Matthew for the advice on my swap dilemma. On Thu, 13 Jul 2000, Jeff Green wrote: > Find someone who owns a copy of pq-magic. The DOS version will run from On Thu, 13 Jul 2000, Matthew Dalton wrote: > (this will give you a file 4Mb large, containing all 0's, in the root > directory) > # dd if=/dev/zero of=/swapfile bs=1k count=4096 > > (format it as a swap file) > # mkswap /swapfile 4096 >
rc.local ?
Please forgive this rather elementary question. I am now trying out potato with the 2.2.17 kernel. I can't figure out which file has taken the place of rc.local. Also, /etc/conf.modules must be deleted and we must now use /etc/modules.conf, right? Finally, I noticed that there is a VM problem with this kernel (from experience and from reports in usenet). I haven't seen a report that this problem has been actually solved yet. Does it still exist in the -prex kernels? I haven't experienced any problems (other than warnings about free_page). Is this going to become a problem in a case where the machine is not doing nothing more stressful than recompiling a kernel? Thanks. Tony Laszlo Tokyo
