Slackware - InstallationThis forum is for the discussion of installation issues with Slackware.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
I'm quite used to installing NetBSD by breaking into a shell from the installer and taking it from there. It allows me to use GPT for partitioning the disk among other advantages.
I presume there's a way of doing the same with Slackware but it's not staring me in the face at the moment. Directing me to a guide would be useful.
The rationale behind this is that it is awkward to install Slackware in a Xen domU (guest, in layman's terms) using the installer. Simply unpacking the sets and fixing fstab and whatever else would make life a lot easier.
I don't understand what you are trying to do. Do you not want to boot off the ISO at all?
I'm a little muddled myself. It's been a while since I needed to do this. As far as I remember, I was not presented with a console while setting up a paravirtualised Linux domU in Xen (a virtual guest that runs at near native speed). I got around it at the time by installing Slackware as normal to an fdisk partition on the real disk, booting back into Xen and setting up this Slackware as a virtual machine from there. (Very convenient, as it allowed me to run the same Slackware as either a virtual or real machine.)
This option is no longer available to me, since I need GPT partitioning and there are too many roadblocks preventing me from going down the same route I went down years ago. So I thought it would be useful to install Slackware perhaps from the host, by unpacking the sets, entering the chroot and doing all the configuration from there. A little bit like how you can break out of the Slackware installer to do LVM, RAID, LUKS, etc., except that you don't boot the installer at all.
I suppose you are essentially looking for a way to "install" Slackware packages to a directory? Since Slackware packages are just fancy tarballs, you could conceivable uncompress your pkgtools, tar, core-utils, and probably a few others to a directory, then you could chroot into it and then use pkgtools to install the rest there (reinstalling the packages you unpacked so they show up properly with pkgtools.
The problem I foresee is just trying to find out what packages are required to get pkgtools to work (at least Installpkg/upgradepkg, since you wouldn't really need the dialog provided by pkgtools).
Once everything is installed to the chroot, you could then run your normal post-install scripts like netconfig, timeconfig, etc, followed by editing your lilo.conf to ensure everything is listed properly, and then finally installing it to the MBR.
I almost wonder if it would be easier to install a system elsewhere (maybe a VM), then copy everything over to the new chroot (maybe creating a giant tarball and then extracting it on the new computer).
I suppose you are essentially looking for a way to "install" Slackware packages to a directory? Since Slackware packages are just fancy tarballs, you could conceivable uncompress your pkgtools, tar, core-utils, and probably a few others to a directory, then you could chroot into it and then use pkgtools to install the rest there (reinstalling the packages you unpacked so they show up properly with pkgtools.
The problem I foresee is just trying to find out what packages are required to get pkgtools to work (at least Installpkg/upgradepkg, since you wouldn't really need the dialog provided by pkgtools).
This is what I do with NetBSD. The question is, why extract just the pkgtools? Why not extract the whole lot, since the rest of the Slackware packages are also tarballs?
Quote:
Once everything is installed to the chroot, you could then run your normal post-install scripts like netconfig, timeconfig, etc, followed by editing your lilo.conf to ensure everything is listed properly, and then finally installing it to the MBR.
I almost wonder if it would be easier to install a system elsewhere (maybe a VM), then copy everything over to the new chroot (maybe creating a giant tarball and then extracting it on the new computer).
That's another way of doing it, and I'm weighing up the options. Things get a little complicated installing into a paravirtualized guest in Xen. HVM guests fine, because you get an SDL console, but PV guests not so straightforward. The console is tricky to get working for starters.
This is what I do with NetBSD. The question is, why extract just the pkgtools? Why not extract the whole lot, since the rest of the Slackware packages are also tarballs?
I suppose you could, but then pkgtools would not be aware of any of the installed packages, plus there's all the doinst.sh files that are run to set up symlinks, handle .new files, etc.
I suppose if you reinstalled all packages after you extracted them so pkgtool is aware and doinst.sh can be run, but that just seems like double the work, so it'd probably be easier to find just the packages needed for pkgtool.
Do you want to install Slackware on an empty partition from a .iso on your running system? That's what I want to do.
Here's how I propose to go about it (I haven't done this before)
First attempt (succeeded already)
1. Extract the initrd image to the file system (e.g. in /tmp)
2. chroot into that image
3. mount /proc
4. run setup
This isn't a full solution, because I don't yet have any media to install from. I'll try that next and post again afterward.
Here are the commands to get this far. slackware-14.2-install-dvd.iso is mounted at /home/dunc/slkdvd
cd /tmp
rm -rf initrd
mkdir initrd
cd initrd
cat /home/dunc/slkdvd/isolinux/initrd.img|gzip -d|cpio -i
chroot . bin/bash
mount /proc
usr/lib/setup/setup
This brings up the light blue menu on the dark blue background. An error message flashes up on the way - will investigate that next
Currently, there's no paravirtual CDROM driver available. So in PV guest, you can only add a iso image as a disk and cannot change this disk at runtime.
Not every distribution provides the xen-tools package for an automated PV creation and configuration. Alpine Linux is one of these distro and therefore provides detailed instructions in order to install and start a PV domU manually
*** New step: hard-link the iso image into the initrd created in my previous post:
cd /tmp/initrd
ln /usr/iso/slackware-14.2-iso/slackware-14.2-install-dvd.iso
*** UNMOUNT the partition to be installed into
umount /dev/sda3
chroot . bin/bash
mount /proc
mkdir iso # Don't use cdrom
mount -oro,loop slackware-14.2-install-dvd.iso iso
usr/lib/setup/setup
---------------------------
Now back in the menu.
I selected swap partitions. Only able to do that because the main system doesn't have any.
Otherwise you'd have to do this once booted into the system you're about to install.
Select /dev/sda3 and quick format, ext3.
Decline offers to use more Linux partitions or add others.
Now at the SOURCE MEDIA SELECTION prompt, select "Install from a pre-mounted directory" and specify "/iso/slackware"
accept defaults in PACKAGE SERIES SELECTION dialogue
Select full installation, and it starts
You can check progress by doing a chroot from another xterm. The mounts therein are not visible to the main system (but added swap disks are).
Back on the main screen, once the install finishes, I skip network config (can copy from main system) and lilo.
Unmount everything before exiting chroot (otherwise /sda3 will nor be cleanly umounted at reboot time).
Since I installed vmlinuz on the main system already, I can boot appending "root=/dev/sda3": it will use /sda3/lib/modules/4.4.14-smp but they are the same as in /lib/modules.
Tried a reboot with root=/dev/sda3 and everything in the newly installed system is OK.
Last edited by duncan_roe; 07-31-2016 at 03:57 AM.
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.