GeneralThis forum is for non-technical general discussion which can include both Linux and non-Linux topics. Have fun!
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 like the Baytrail Pentiums. I've had like 6 of them. Never had the J1900 (guessing it's that not a j0900) but had multiple N2940's that are essentially the same except slower frequency to hit lower TDP.
Good choice too. As you saw, the Baytrail SOC's are REALLY linux-friendly. Almost everything is Intel and old enough that even slightly aged kernels (like Slackware or Debian use) support it absolutely without issues at this point.
It wasn't a choice. That was what they had. I don't know if you have Computer Exchange in the US. It's a chain that buys hardware, games, videos and so on from members of the public for cash, refurbishes them if necessary and then sells them on to new users. You can't order what you want; you can only buy what they've got. And lately they haven't been selling any towers, only laptops. I bought both my laptops from them. The first (an HP Pavilion I think) had to be returned because it had a dud keyboard, but I still have the Samsung I bought from them later.
I now have a map of the partitioning of this big hard drive and I know which partitions I want to keep for the time being and which can be deleted. My plan is to use fsarchiver (recently installed in Debian) to archive Slackware and later LFS and transfer them onto a memory stick. I'll use SystemRescue to create the partitions I want and unpack the archives onto them, then chroot into Slack and get elilo installed. And then see if it boots!
So this is what I have, what I think it's for, and what I plan to do with it:
sda1 (1 GB ntfs) Windows recovery tools. I'll delete that and create a BIOS boot partition. I probably won't need it as I plan to use UEFI in native mode but it will do no harm to have it in case I have to revert to csm.
sda2 (260 MB fat32) EFI System Partition. I'll keep that.
sda3 (500 MB fat32) secondary EFI partition, intended for Lenovo OneKey emergency boots. Keep for the time being (though OneKey won't work in Linux).
sda4 (128 MB unknown fs) Microsoft reserved partition. Don't know what it's for because I can't mount it. Probably spyware. Get rid of it!
sda5 (438 GB ntfs) Windows. Remove.
sda6 (848 MB ntfs) Windows restore (factory settings). Remove.
sda7 (24 GB ntfs) OneKey dump and restore (snapshots). Remove.
In place of these, I want to create:
sda4 (5 GB swap). That should be enough for sleep, given that RAM is 4 GB. I'm sure this computer can sleep because Lenovo Thinkstations are just laptops in tower cases.
sda5 (100 GB). Data partition.
sda6 (20 GB) Slackware
sda7 (30 GB) LFS
sda8 (30 GB) Leave empty for next LFS.
which still leaves plenty of unallocated space if I decide to clone Debian, but I'm coming down more and more against that. I used to love Debian, but I don't like what it's turning into.
OK, I've deleted the Windows partition and made new partitions for swap, data, Slackware and LFS (X2). I currently have the fsarchived data unpacking into the new data partition. That took about half an hour to pack but it should unpack faster as the new machine has twice the memory and twice the number of cpu cores. If all goes well, I will do the same for Slackware during the next few days. Then of course I will have to make it bootable. AFAIK that means editing /etc/fstab with the new device names, making a new initrd, and installing and configuring elilo. Then ditto for LFS.
Just for an experiment, I tried booting with the F2 key down. This activates the second EFI system partition, which loads Lenovo's OneKey Recovery system. I'm beginning to wonder if that could be adapted in some way. For example, suppose it booted a SystemRescue image...
PS: Wow! 10 minutes. That's power!
Last edited by hazel; 01-17-2019 at 10:11 AM.
Reason: postscript added
Unpacked Slackware onto its partition, edited fstab and created a new initrd. Interestingly, Patrick's advisory script recommended a sheaf of new kernel modules to include. Then I installed elilo and ran eliloconfig. But then I got cold feet and shut down.
I've been reading the many text files that come with elilo. One of them seemed to be saying that I need an extra library package called gnu-efi to make elilo work. Is that the case? I couldn't do an ldd on it because it isn't an ELF program.
PS: I found this on duckduckgo and it looks like a development kit, not a library, so perhaps I don't need it after all.
Last edited by hazel; 01-18-2019 at 09:59 AM.
Reason: Added postscript
Hi hazel,
I'm pretty sure that even with UEFI and GPT partitions one can always fall back on the install media to boot any system for which you know the kernel name or if it would boot with one of those on the install media like "huge.s". The install media boots very quickly to a full stop where one can either proceed with normal installation or redirect to boot any bootable partition. I frankly have no idea why such a simple and eminently useful idea isn't all but universal but then Patrick does champion sub-genius LOL. Anyway there are basic instructions at that 1st full stop and is a wonderful recovery device. FWIW nothing bad happens from a failed boot... no need for warm feet. Just try it and deal with the outcome until it works. You may even get a nice surprise - You're better than you imagined
AFAIR, using shutdown -r on SystemRescue gives you a menu with one option to boot from the hard drive. I'll try that first. But I do have the installation cd (the mini-iso, not the full one) so that's a second string to my bow.
Distribution: Mainly Devuan, antiX, & Void, with Tiny Core, Fatdog, & BSD thrown in.
Posts: 5,484
Rep:
Just catching up with your latest exploits.....
Sounds like a decent bit of hardware, you've picked up, however, I would just have made sure I could boot a live distro, then zapped the whole drive, & partitioned it for how I wanted it, then done fresh installs.
AFAIR, using shutdown -r on SystemRescue gives you a menu with one option to boot from the hard drive. I'll try that first. But I do have the installation cd (the mini-iso, not the full one) so that's a second string to my bow.
That second string plays this tune (it's from memory so it could be off a bit but the CD explains it correctly) great, simple recovery tool
Yay! Slackware boots! I failed to reboot it from SystemRescue but when I did a ctrl-alt-del, I was able to whisk the CD out and up came elilo, followed by Slackware.
I went over the dmesg output carefully. ata1 and ata2 are SATA and these are the hard drive and the cd/dvd drive. ata3 and ata4 are PATA legacy. So that covers the business about Thinkcentres being PATA and Thinkstations SATA. There is actually one controller of each type. It just depends which one they chose to wire up.
I checked the ethernet controller too and noted down the driver (r8169). I shall need to rebuild my LFS kernel and include this driver. One odd thing is that the interface was renamed twice: from eth0 to eth126 to eth1. I've no idea what causes that. Some udev rule presumably. There is definitely no wifi, whatever the manual says. I'm sure either the SystemRescue kernel or the Slackware kernel would have picked it up if it were there.
I plan to switch the two machines: bring the new one downstairs and plug it into my router, and take the old one upstairs. It can stay there until I decide to get rid of it. Once I know I can get online with the new machine, I will transfer LFS.
This reserves eth0 for a particular MAC. The new card must have a different MAC so gets a different name. How do I clear this? Should I edit the file or delete it?
Does the system has an on-board Intel ethernet adapter? I haven't got my old AMD system anymore, but it had an onboard via-rhine2 one and I put in a 2nd (faster) ethernet adaptor in it. On-board I used for the local net, the other (eth1) for getting onto the university network.
If you do NOT have got that adaptor (e1000e, check with lspci) you can delete the file, a new one will be generated on the next reboot. Otherwise you CAN change the NAME= field to <whatever> i.e. eth1
I edited it. If I had known I could just delete it and make the system give me a new rules file, I would have done that. It would have been quicker. Anyway, it's not the reference to the e1000 driver that caused the problem; if you look again, you will see that it's just a comment. What is specified as a condition is the MAC address. I edited that into the correct MAC for this card and rebooted. Now my card is eth0 again, dhcp works, and I am online using the new machine to type this. Mission accomplished.
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.