[SOLVED] Can't log into system, several issues (keyboard doesn't even work, mounting, modules) after current update (jun 11)
SlackwareThis Forum is for the discussion of Slackware Linux.
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.
Can't log into system, several issues (keyboard doesn't even work, mounting, modules) after current update (jun 11)
I don't even know where to begin, so I'll just list the issues I was able to glean, myself.
(Can't really scroll the tty without a working keyboard after all)
Edit:
I initially forgot to mention: I only changed/updated the kernel after the update and a reboot later that day.
Turned on the PC next day, no keyboard, no drive UUIDs being recognized on the same kernel that worked the day prior.
Edit:I am booting using 'legacy bios' not UEFI.
/dev/shm and /dev/pts don't have mount points.
uuids for my harddrives are not recognized
(despite being correct, checked it with a liveslack stick)
alsa seems to have troubles
keyboard doesn't work, if I unplug it and put it back in, all the mode lights stay on (numlock, caps, scroll)
'failed to find sysfs'
kmod tripping over some holders it can't find for the realtek r8168 module (which is one of my biggest problems in general, it seems it only works with the realtek drivers I have to compile myself)
when I use direct device pathes like /dev/sda1 they get recognized but things still work
grub-makecfg goes nuts with device nodes it can't find
(and a recent change also disabled the os prober, hot diggity damn bad timing I am SO GLAD I found a backup in vims backup dir no less)
(...and I wish I kept a backup of my system, too! ...frack!)
Anyway, there's a bunch of stuff wrong, I think the last update before this one I did was around march?
I'd have to check the /var/log/removed_packages I guess
What I have tried so far (other than using direct device paths):
I edited inittab to run a script that ran slackpkg update, but slackpkg plus seems to have some issue with GPGs or something and other problems
(used a different kernel that loads my internet but keyboard is still off)
grabbed Alien Bobs current ISO to do a slackpkg reinstall slackware64 in case I messed up the update process, but that didn't help either
I install-new'd and clean-system'd
(blacklisting my custom packages to get rid of any 'removed' packages, but there were none)
danced around with module blacklisting for the ethernet card and alsa-loopback, didn't work.
'slackpkg new-config' checks, but nothing out of the ordinary
single user mode (init 1)
...if only the keyboard would work at least...
it *seems* that the system kind of works through the script I put in instead of the 'login' prompt in the first tty but without a keyboard or an X session for that matter, not easy.
Got any ideas?
Last edited by Nang Dabbit; 06-13-2021 at 08:22 PM.
For me sounds like something bad is happening to kernel. Tried some different boot options? The first came to my mind is maintenance mode: init 1. Did you messed with kernel versions perhaps? Last time I looked at there were two different kernel versions accessible 5.10.x and from extra/ 5.12.x. If also slackpkg was updated then as I remember first update slackpg then again with new version run update.
Edit: If you can run Liveslak try to chroot to system on hard drive. I won't help much here. It is long time I had to chroot anywhere. People around for sure will help. It issue command itself is simple but there is issue with mount /proc and /sys. I mean /proc and /sys coming from live system have to be accessible within chroot environment.
For me sounds like something bad is happening to kernel. Tried some different boot options? The first came to my mind is maintenance mode: init 1. Did you messed with kernel versions perhaps? Last time I looked at there were two different kernel versions accessible 5.10.x and from extra/ 5.12.x. If also slackpkg was updated then as I remember first update slackpg then again with new version run update.
Thanks for the reply.
It is likely that the kernel also has something to do with it, in the backup grub config there was an entry with a kernel I compiled myself many moons ago.
That worked, but only as far as having my keyboard working, no ethernet driver, after all the kernel packages are different, including the sources/headers (although I think headers don't matter as much? Not sure.), and the whole slew of kmod stuff and filesystem things still persisted.
I also tried init 1, in fact that was the first thing I did, I think.
I added that to my list of "what I tried" in the first post.
Last edited by Nang Dabbit; 06-13-2021 at 05:57 AM.
Do a backup of your own data, then do a clean install.
Next time I hope you will be aware that doing updates on a half of year old slackware-current installation is a really bad idea.
IF you are not willingly to follow thoughtfully the updates of -current, applying them daily, I believe that's better to stay on a stable release of your choice.
Last edited by LuckyCyborg; 06-13-2021 at 06:00 AM.
Do a backup of your own data, then do a clean install.
Next time I hope you will be aware that doing updates on a half of year old slackware-current installation is a really bad idea.
IF you are not willingly to follow thoughtfully the updates of -current, applying them daily, I believe that's better to stay on a stable release of your choice.
Guess I have been lucky everytime before, then.
A clean install is a nightmare scenario, which I will go for, but only if there really isn't any other way (within reason).
Quote:
Originally Posted by igadoter
Sorry for late edit. Please reread my post.
Yeah, I used chroot to edit config files and use the Slackware current ISO Alien Bob so generously provides on his website to reinstall all the packages.
Maybe take fresh copy of kernel (huge) rewrite existed one in /boot - rerun lilo. The same essentially you should do with corresponding modules - just be sure kernel runs. At the first it can panic but it would mean it works. If you can't reinstall kernel directly you can do a trick to install small version of live edition - at the end installer runs boot loader. Hopefully it will recognize both systems. Once you fix you main installation you will remove that rescue system.
I'm not getting any kernel panics, the keyboard merely seems to not work.
I also used the kernel before the update, which I had backed up in another location and that one didn't work either, even though it did before.
I understand but everything is fine when you run liveslak? You may try to put some commands into rc.local - it is last script being executed. I mean do chmod -x for every service in /etc/rc.d/* except rc.local - you can say put there dmesg command - maybe with redirection - you can read contents later.
The other version what I posted above is to arrange say 2GB free space on hard drive to install there -current, only packages from a/ set - and use the kernel from that new installation to run main installation. This will give answer what's the issue - kernel or something else? By the way if you run generic kernel did you regenerate init ram disk? I got idea that you are just stucked in that temporary environment. I mean somehow kernel can not go beyond init ram disk. Something holds it. But it is only idea.
Well, here is the output of dmesg, ran from rc.local
It does not seem to contain any of the really interesting messages, though.
Too long to post, too.
Edit: I'm running a huge kernel for the time being.
Edit:
Since this is the kernel I used before the update, which has both the ethernet driver and the nvidia one compiled, I'm going to try an X session...
Edit:
That did not work out, hm.
Last edited by Nang Dabbit; 06-13-2021 at 07:35 AM.
Looks good. I would tell process eats resources - but it does not have much sense as you the same problems even in level 1. Try to allow syslog and look at it. If even in level 1 keyboard is slow - seems it is kernel issue - way here seems to be kernel diagnostic. Boot system and assure all services are disabled except syslog. There is trick to open virtual console for output only: there you can point tail -f /var/log/syslog > /dev/ttyS1 - I am scratching my head - my knowledge about kernel diagnostic is rather limited. In level 1 for me there is nothing which would slow down keyboard - some modules are in conflict possible. Seems that running run level 1 does not kill modules. I had discrete graphic card and onboard graphic - and kernel loaded both modules - causing total mess. Because seems only you has that it is related to your hardware configuration. Try to attach usb keyboard.
I'm not sure if it's really about draining resources especially since I can run programs via rc.local etc.
The keyboard is just flat out 'dead'.
And there's mounting and module errors, etc.
For the time being I'll just use windows and wait for the next current update, unless there's more suggestions coming in here.
Thanks for your time so far, though.
I've littered the init scripts with sleeps to hopefully see more details, and I could read that part much easier, but I swear that either suppressed some of the messages, or they were part of the really quickly scrolling stuff before it.
I made screenshots, with a camera, but noone likes to look at those, so here's a summary, somewhat abbreviated/paraphrased
Code:
rodata_test: all successful
proc mounted on /proc
tmpfs ounted on /run
loop: module loaded
Initialized LVM (Logical Volume Manager):
Failed to find sysfs mount point (/dev/sda1 was passed as the root drive from grub, btw)
testing root filesystem status: r/o
/dev/sda1: clean
Remounting root with r/w enabled
mount: /dev/sda1 mounted on /
Updating module dependency list for 5.10.27
modprobe: FATAL: Module r8168 not found in module directory (the ethernet driver module, different kernel, same results tho)
fuse: init (API version 7.32)
Configuring kernel params: /sbin/sysctl -e (rest of the line is out of the picture)
mount: /dev/pts: mount point doesn't exist
mount: /dev/shm: ditto
----
jump to somewhere in runlevel 3:
----
/etc/rc.d/rc.inet1: line 316 /proc/sys/net/ipv6 (line goes off picture)
no file or directory
ditto for line 324
Device 'eth0' does not exist
And the rest was basically normal from there on, without me being able to use the keyboard, still
I'm going to try the 14.2 stable kernel now.
Last edited by Nang Dabbit; 06-13-2021 at 09:32 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.