The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

«‹›1033 »

Saukkis: May 16, 2003; Unless I'm on the inside curve pointing straight at oncoming traffic the high beams stay on and I laugh at your puny protest flashes.
I am Most Important Man. Most Important Man in the World.

F_Shit_Fitzgerald posted:

Is there a way to protect a password from being deleted (accidentally or maliciously) from pass? I like the package; it's a pretty secure way to store passwords but anyone could go into your system and just do pass rm [site name] if they wanted to gently caress you over.

e: I guess I could alias it, but someone savvy enough could just bypass that...

Sure, it's called backups.

# ? Aug 20, 2023 02:42

Adbot: ADBOT LOVES YOU

# ? May 24, 2024 12:29

spiritual bypass: Feb 19, 2008; Grimey Drawer

F_Shit_Fitzgerald posted:

Is there a way to protect a password from being deleted (accidentally or maliciously) from pass? I like the package; it's a pretty secure way to store passwords but anyone could go into your system and just do pass rm [site name] if they wanted to gently caress you over.

e: I guess I could alias it, but someone savvy enough could just bypass that...

Push your passwords to a remote git host. Require a passphrase for the ssh key to that you use to authenticate

# ? Aug 20, 2023 02:42

cruft: Oct 25, 2007

you use a password manager that allows you to type "pass rm $site" and it deletes passwords? That doesn't sound very secure to me, that sounds like a text file.

Is the file it uses at least chmod 0600, like .netrc?

e: oh, it uses cached credentials and is the standard wotzit for XDG.

Anyway I'm going to fall back on "if someone has access to your login, you have 99 other problems"

# ? Aug 20, 2023 02:53

F_Shit_Fitzgerald: Feb 2, 2017

Yeah. Fortunately, I'm the only one who uses my Linux machine, so the only possible snafu would be me doing something stupid (which happens pretty often!). The remote repository is a great idea; I'll probably do like I did for my dotfiles and make a git bare repository. Thanks!

# ? Aug 20, 2023 02:56

Mr. Crow: May 22, 2008; Snap City mayor for life

cruft posted:

you use a password manager that allows you to type "pass rm $site" and it deletes passwords? That doesn't sound very secure to me, that sounds like a text file.

Is the file it uses at least chmod 0600, like .netrc?

e: oh, it uses cached credentials and is the standard wotzit for XDG.

Anyway I'm going to fall back on "if someone has access to your login, you have 99 other problems"

assuming you found it but https://www.passwordstore.org/

nthing making it a git repo is probably your best bet.

# ? Aug 20, 2023 02:58

Inceltown: Aug 6, 2019

Pass has git baked in and is the expected way to use it. It makes it really easy to synch them with your phone too.

# ? Aug 20, 2023 03:00

Klyith: Aug 3, 2007; GBS Pledge Week

Two basic, useful questions to frequently ask yourself:
1) What happens when the drive(s) in my PC die with zero warning?
2) What happens when I am standing outside my burning house in my underwear and all my poo poo is gone?

That covers a lot of stuff -- backups of data you care about, passwords for things you need, recovery for online accounts, etc.

Getting into the weeds of something specific like pass is thinking too small.

# ? Aug 20, 2023 03:23

Mr. Crow: May 22, 2008; Snap City mayor for life

I'm brainfarting how to do this, I have a USB headset that weirdly gets recognized as an sr0 CD-ROM device as well the USB device; how do I block this in udev from being loaded as a the cdrom?

# ? Aug 23, 2023 16:58

lobsterminator: Oct 16, 2012

Mr. Crow posted:

I'm brainfarting how to do this, I have a USB headset that weirdly gets recognized as an sr0 CD-ROM device as well the USB device; how do I block this in udev from being loaded as a the cdrom?

Alternative solution: Have you tried using a CD burner software and writing a music CD into it? Maybe it will start playing music.

# ? Aug 23, 2023 17:04

pseudorandom name: May 6, 2007

Does the supposed CDROM contain Windows drivers?

# ? Aug 23, 2023 19:33

Vulture Culture: Jul 14, 2003; I was never enjoying it. I only eat it for the nutrients.

Mr. Crow posted:

I'm brainfarting how to do this, I have a USB headset that weirdly gets recognized as an sr0 CD-ROM device as well the USB device; how do I block this in udev from being loaded as a the cdrom?

I've seen this before. If you mount sr0 as iso9660 and cat its contents into MPlayer, you'll see a black and white video and then your phone will start ringing

# ? Aug 23, 2023 19:43

cruft: Oct 25, 2007

Mr. Crow posted:

I'm brainfarting how to do this, I have a USB headset that weirdly gets recognized as an sr0 CD-ROM device as well the USB device; how do I block this in udev from being loaded as a the cdrom?

You have to make a higher-priority rule that matches the thing and creates the right node for it.

That's all I remember. I'd start with https://wiki.archlinux.org/title/Android_Debug_Bridge#Adding_udev_rules, which I guess is the only other thing I remember about this.

# ? Aug 23, 2023 19:44

Mr. Crow: May 22, 2008; Snap City mayor for life

So I was messing around with it today and gave up, this is what I ended up with. I actually ended up resolving it somewhat but then I couldn't pass the USB device to VMs in virt manager and thats where I basically said whatever.

code:

> cat /etc/udev/rules.d/65-yealink.rules
# udev rules to disable yealink being detected as an sr device
# this file is '65' as it needs to run after hwdb modalias is set by 
#   /usr/lib/udev/rules.d/50-udev-default.rules
# but before we load the kernel module defined
#   /usr/lib/udev/rules.d/80-drivers.rules

ACTION!="add", GOTO="yealink_rules_end"

## Yealink WH66

# this disables the mass storage driver but prevents virt-manager from sharing with vms
# MODALIAS pulled from 'udevadm info /sys/device/path'
#SUBSYSTEM!="usb", GOTO="yealink_rules_end"
#ACTION=="add", ENV{MODALIAS}=="usb:v6993pB02Bd0101dc00dsc00dp00ic08isc06ip50in05", RUN="/bin/sh -c 'echo 0 >/sys/\$devpath/authorized'"

# this runs but doesn't actually stop the driver from loading
#ACTION=="add", SUBSYSTEM=="scsi", ENV{MODALIAS}=="scsi:t-0x05", RUN:="/usr/bin/true"

# this might prevent it from showing up in the UI??
#SUBSYSTEM=="block", ENV{ID_VENDOR_ID}=="6993", ENV{UDISKS_IGNORE}="1"

# random settings
#ENV{ID_USB_VENDOR_ID}=="6993",
#ATTRS{idVendor}=="6993", ATTRS{idProduct}=="b02b"

LABEL="yealink_rules_end"

I don't really understand USB but this thing has like half a dozen or more interface descriptors in lsusb, I could disable the one that ends up loading the cdrom but then I have the issue above so :shrug:

I hadn't ever messed with udevadm test before but it's kind of cool, I was using it to check that my changes were being loaded and seeing what other rules were taking affect, I think every time I've ever used udevs before I was either copying something or doing something simple like changing ownership, was kinda fun messing around with it.

Mr. Crow fucked around with this message at 03:33 on Aug 24, 2023

# ? Aug 24, 2023 03:29

pseudorandom name: May 6, 2007

You can probably somehow do echo "6993 B02B" > /sys/bus/usb/drivers/usb-storage/remove_id at boot.

edit: nevermind, this only works for device IDs manually added via new_id

The good news is that the CDROM devices also show up on Windows, which means it's all working correctly.

pseudorandom name fucked around with this message at 04:52 on Aug 24, 2023

# ? Aug 24, 2023 04:45

The Merkinman: Apr 22, 2007; I sell only quality merkins. What is a merkin you ask? Why, it's a wig for your genitals!

I have hardware compatibility question:
So I used to use a custom built desktop, with parts over a decade old. The motherboard is an ASUS M4A78 Pro.
Because of the Pandemic I got a webcam: Logitech, Inc. C920 HD Pro Webcam.
It worked fine with my desktop and its ancient motherboard, but I don't recall if I had to do anything other than plug it in.
As the desktop is quite old, I got a Dell XPS 9310 (2020? 2021?), which has since been upgraded to Ubuntu 22.04.3 LTS.
Unfortunately the webcam doesn't seem to work properly on the laptop. At first I thought it was the ThinkPad Dock I was using, but even plugging the webcam into the laptop directly, albeit with a USB-A -> USB-C converter, it has the same issue. I see video through software in something like Cheese, but the preview appears as though it takes like the first second and loops it. If I record video it just comes out as a still image.

# ? Aug 24, 2023 16:50

mystes: May 31, 2006

F_Shit_Fitzgerald posted:

Is there a way to protect a password from being deleted (accidentally or maliciously) from pass? I like the package; it's a pretty secure way to store passwords but anyone could go into your system and just do pass rm [site name] if they wanted to gently caress you over.

e: I guess I could alias it, but someone savvy enough could just bypass that...

Anyone could just delete your entire password store from any other password manager. It's arguably bad that pass leaks information about what sites you have accounts on, but worrying about the fact that an attacker could delete a specific password with it doesn't really make sense to me.

# ? Aug 25, 2023 22:27

cruft: Oct 25, 2007

I have a lot of source code on this computer, and some of it performs trigonometry calculations. How can I safeguard against an attacker changing the value of Pi?

# ? Aug 25, 2023 22:59

Tesseraction: Apr 5, 2009

Remember that in a pinch you can define it as 3.

# ? Aug 25, 2023 23:03

Vulture Culture: Jul 14, 2003; I was never enjoying it. I only eat it for the nutrients.

cruft posted:

I have a lot of source code on this computer, and some of it performs trigonometry calculations. How can I safeguard against an attacker changing the value of Pi?

actual threat model for those of us working in regulatory environments

# ? Aug 25, 2023 23:44

Volguus: Mar 3, 2009

Tesseraction posted:

Remember that in a pinch you can define it as 3.

The real value is 3.2 :fsmug:

# ? Aug 26, 2023 00:01

F_Shit_Fitzgerald: Feb 2, 2017

I thought it was 22/7...

# ? Aug 26, 2023 00:19

cruft: Oct 25, 2007

Vulture Culture posted:

actual threat model for those of us working in regulatory environments

This isn't too terribly distant from one of the main plot points in The Three Body Problem by Cixin Liu. Great read: I recommend it, and so did Barack Obama!

# ? Aug 26, 2023 01:27

Volguus: Mar 3, 2009

cruft posted:

This isn't too terribly distant from one of the main plot points in The Three Body Problem by Cixin Liu. Great read: I recommend it, and so did Barack Obama!

The only advice I'd have for that book (or the trilogy) is to be patient. It'll all be revealed in due time.

# ? Aug 26, 2023 01:43

ziasquinn: Jan 1, 2006; Fallen Rib

hmm I dropped off halfway thru book 2 should I pick it up again??

# ? Aug 26, 2023 02:41

BlankSystemDaemon: Mar 13, 2009

Tesseraction posted:

Remember that in a pinch you can define it as 3.

Just memorize it to 40 places, that's all the precision you'll ever need.

# ? Aug 26, 2023 10:46

AlexDeGruven: Jun 29, 2007; Watch me pull my dongle out of this tiny box

Wasn't it something like 15 places that was enough to accurately estimate the circumference of the universe to a crazy precision?

# ? Aug 26, 2023 14:36

xzzy: Mar 5, 2009

Google maps give lat/long coordinates with 7 decimals, which is like 1 centimeter precision. Just in case you want to give directions to a dime.

# ? Aug 26, 2023 15:18

Klyith: Aug 3, 2007; GBS Pledge Week

AlexDeGruven posted:

Wasn't it something like 15 places that was enough to accurately estimate the circumference of the universe to a crazy precision?

"39 digits of pi would suffice to calculate the circumference of the known universe to the width of a hydrogen atom"

Hmmm...

code:

Language 	Command 	Digits 	Value
Java 	Math.PI 	16 	3.141592653589793
C/C++ 	M_PI 	21 	3.14159265358979323846
Python 	math.pi 	21 	3.14159265358979323846
Swift 3.0 	Double.PI 	>=16 	3.141592653589793�
Go 	Pi 	63 	3.14159265358979323846264338327950288419716939937510582097494459

the universe must be running on Go

# ? Aug 26, 2023 15:38

BlankSystemDaemon: Mar 13, 2009

AlexDeGruven posted:

Wasn't it something like 15 places that was enough to accurately estimate the circumference of the universe to a crazy precision?

If memory serves, that gets you a human-scale approximation of the circumference of the visible universe, which is probably good enough for most people.
Just under 40 gets you so close, that you're gonna need to measure at sub-atomic scales in order to check.

EDIT: Welp, beaten..

Klyith posted:

Hmmm...

code:

Language 	Command 	Digits 	Value
Java 	Math.PI 	16 	3.141592653589793
C/C++ 	M_PI 	21 	3.14159265358979323846
Python 	math.pi 	21 	3.14159265358979323846
Swift 3.0 	Double.PI 	>=16 	3.141592653589793�
Go 	Pi 	63 	3.14159265358979323846264338327950288419716939937510582097494459

the universe must be running on Go

Computing pi, along with factorizing primes, is one of the go-to for number-crunching tests on computers, so I'd hope that systems programming languages would do a bit better than 20 digits.
Where are those from?

BlankSystemDaemon fucked around with this message at 17:01 on Aug 26, 2023

# ? Aug 26, 2023 16:58

Klyith: Aug 3, 2007; GBS Pledge Week

BlankSystemDaemon posted:

Computing pi, along with factorizing primes, is one of the go-to for number-crunching tests on computers, so I'd hope that systems programming languages would do a bit better than 20 digits.
Where are those from?

A random rear end webpage on google. It's a joke so like don't take it seriously.

edit: easiest fact-check, Python docs say:
math.pi
The mathematical constant π = 3.141592�, to available precision. (my emphasis)

code:

~ > python
>>> import math
>>> print(math.pi)
3.141592653589793
>>> import numpy
>>> print(numpy.float128(math.pi))
3.141592653589793116

edit 2: wait a minute, that's not a float128. Also the 116 on the end is wrong. :what:

I guess float128 isn't real.

edit 3: yeah evidently numpy.float128 only works if you're on a platform that has actual 128bit registers, so not x86-64

Klyith fucked around with this message at 17:36 on Aug 26, 2023

# ? Aug 26, 2023 17:15

Subjunctive: Sep 12, 2006; ✨sparkle and shine✨

BlankSystemDaemon posted:

Computing pi, along with factorizing primes, is one of the go-to for number-crunching tests on computers, so I'd hope that systems programming languages would do a bit better than 20 digits.
Where are those from?

Sure, but those programs that are crunching pi aren�t using the standard library�s definition of pi, pretty much as the whole point of the exercise. Any of those definitions are going to give more precision than is needed for systems programming tasks outside of astrophysics, I think.

But nonetheless, Rust just fills the f64

�std::f64::consts::PI� posted:

pub const PI: f64 = 3.14159265358979323846264338327950288f64;

# ? Aug 26, 2023 17:59

Vavrek: Mar 2, 2013; I like your style hombre, but this is no laughing matter. Assault on a police officer. Theft of police property. Illegal possession of a firearm. FIVE counts of attempted murder. That comes to... 29 dollars and 40 cents. Cash, cheque, or credit card?

I'm pretty sure I didn't get this from this thread, and so this isn't a repost here, but even if so:

Non-Euclidean Doom: what happens to a game when pi is not 3.14159�
https://www.youtube.com/watch?v=_ZSFRWJCUY4

As was said by a goon when I first saw it: highly recommended if "what happens if you recompile Doom with a different value for pi?" sounds like something you might enjoy. Not recommended otherwise.

The background is that while pi isn't directly written anywhere in Doom's code, there are a lot of precomputed geometric calculation results which are, and working backward from those let people realize that the value for pi was off at the tenth digit. Because John Carmack had entered it from memory and got that digit wrong. The video is about what happens when you fix that error. And what happens if you set pi to 3. Or 2, or 4.

# ? Aug 26, 2023 18:28

BlankSystemDaemon: Mar 13, 2009

Vavrek posted:

I'm pretty sure I didn't get this from this thread, and so this isn't a repost here, but even if so:

Non-Euclidean Doom: what happens to a game when pi is not 3.14159�
https://www.youtube.com/watch?v=_ZSFRWJCUY4

As was said by a goon when I first saw it: highly recommended if "what happens if you recompile Doom with a different value for pi?" sounds like something you might enjoy. Not recommended otherwise.

The background is that while pi isn't directly written anywhere in Doom's code, there are a lot of precomputed geometric calculation results which are, and working backward from those let people realize that the value for pi was off at the tenth digit. Because John Carmack had entered it from memory and got that digit wrong. The video is about what happens when you fix that error. And what happens if you set pi to 3. Or 2, or 4.

I love this kind of poo poo, it's great - thanks for the recommendation.

Tom7/Suckerpinch on YouTube also does videos like this, and publishes papers at the conference dedicated to computational heresy called SIGBOVIK.

# ? Aug 26, 2023 20:50

FunkyAl: Mar 28, 2010; Your vitals soar.

I have a cudy wireless web adapter and I've been having trouble with it. I have been using some drivers I got for it online and installing them with sudo make/install, but whenever I restart the drivers are gone and when Linux updates there is usually some command line error that prevents the driver from installing. Sometimes when this happens one of the drivers will work after not working for no clear reason, sometimes the adapter shows up without needing the install. But, I haven't made any progress this past time, is there another way, preferably without having to reinstall every time I restart?

# ? Aug 27, 2023 00:58

Hadlock: Nov 9, 2004

xzzy posted:

Google maps give lat/long coordinates with 7 decimals, which is like 1 centimeter precision. Just in case you want to give directions to a dime.

I actually used this the other day to find a usgs gps mark

# ? Aug 27, 2023 04:54

CaptainSarcastic: Jul 6, 2013

FunkyAl posted:

I have a cudy wireless web adapter and I've been having trouble with it. I have been using some drivers I got for it online and installing them with sudo make/install, but whenever I restart the drivers are gone and when Linux updates there is usually some command line error that prevents the driver from installing. Sometimes when this happens one of the drivers will work after not working for no clear reason, sometimes the adapter shows up without needing the install. But, I haven't made any progress this past time, is there another way, preferably without having to reinstall every time I restart?

See if you can figure out the actual chipset being used and install the firmware for that. It's probably Realtek or Atheros or something. What does the output of lsusb say?

# ? Aug 27, 2023 07:43

notwithoutmyanus: Mar 17, 2009

So, I am running ubuntu and jsut tried to do an apt-get upgrade after updating, and I get this. It looks like there's possibly a dpkg issue involving AMDgpu drivers or something? How can I troubleshoot/resolve this? I think at one point things were bugging out and I ended up doing a dpkg remove on amdgpu - but googling this sort of issue didn't find me anything (probably because I don't know which part to troubleshoot outside of the dpkg error)

code:

sudo apt-get upgrade -y
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Calculating upgrade... Done
The following packages have been kept back:
  gjs libgjs0g
0 upgraded, 0 newly installed, 0 to remove and 2 not upgraded.
4 not fully installed or removed.
After this operation, 0 B of additional disk space will be used.
Setting up linux-headers-6.2.0-31-generic (6.2.0-31.31~22.04.1) ...
/etc/kernel/header_postinst.d/dkms:
 * dkms: running auto installation service for kernel 6.2.0-31-generic
Error! Could not locate dkms.conf file.
File: /var/lib/dkms/amdgpu/6.0.5-1620033.22.04/source/dkms.conf does not exist.
   ...fail!
run-parts: /etc/kernel/header_postinst.d/dkms exited with return code 4
dpkg: error processing package linux-headers-6.2.0-31-generic (--configure):
 installed linux-headers-6.2.0-31-generic package post-installation script subprocess returned error exit status 1
Setting up linux-image-6.2.0-31-generic (6.2.0-31.31~22.04.1) ...
dpkg: dependency problems prevent configuration of linux-headers-generic-hwe-22.04:
 linux-headers-generic-hwe-22.04 depends on linux-headers-6.2.0-31-generic; however:
  Package linux-headers-6.2.0-31-generic is not configured yet.

dpkg: error processing package linux-headers-generic-hwe-22.04 (--configure):
 dependency problems - leaving unconfigured
dpkg: dependency problems prevent configuration of linux-generic-hwe-22.04:
 linux-generic-hwe-22.04 depends on linux-headers-generic-hwe-22.04 (= 6.2.0.31.31~22.04.8); however:
  Package linux-headers-generic-hwe-22.04 is not configured yet.
dpkg: error processing package linux-generic-hwe-22.04 (--configure):
dependency problems - leaving unconfigured
Processing triggers for linux-image-6.2.0-31-generic (6.2.0-31.31~22.04.1) ...
/etc/kernel/postinst.d/dkms:
 * dkms: running auto installation service for kernel 6.2.0-31-generic
Error! Could not locate dkms.conf file.
File: /var/lib/dkms/amdgpu/6.0.5-1620033.22.04/source/dkms.conf does not exist.
   ...fail!
run-parts: /etc/kernel/postinst.d/dkms exited with return code 4
dpkg: error processing package linux-image-6.2.0-31-generic (--configure):
 installed linux-image-6.2.0-31-generic package post-installation script subprocess returned error exit status 1
No apport report written because MaxReports is reached already
                                                              Errors were encountered while processing:
 linux-headers-6.2.0-31-generic
 linux-headers-generic-hwe-22.04
 linux-generic-hwe-22.04
 linux-image-6.2.0-31-generic
E: Sub-process /usr/bin/dpkg returned an error code (1)

# ? Aug 28, 2023 19:54

Woolie Wool: Jun 2, 2006

Recently I saw that the colored lighting software renderer had been added to the Quake II source port YamagiQ2. I decided to install it on my ThinkPad (my real PC is packed away because I am about to move), and both for nostalgia and to be easy on the feeble Sandy Bridge CPU I tried running it in 640x480, but as it turns out this renderer likes to crash. And when it crashes in 640x480 it completely trashes my KDE display settings, not only locking the screen in 640x480 but making the desktop totally unresponsive, even after a reboot. I ended up having to reboot into single user mode to get a tty to rm -rf the kscreen config folder (trying to get into a TTY from the SDDM login screen just made SDDM pop back up after a couple of frames).

Great failure state handing, loving windows 98 handles a fullscreen program in a non-native video mode crashing more gracefully than this :negative:

# ? Aug 29, 2023 02:19

Tesseraction: Apr 5, 2009

notwithoutmyanus posted:

code:

sudo apt-get upgrade -y
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Calculating upgrade... Done
The following packages have been kept back:
  gjs libgjs0g
0 upgraded, 0 newly installed, 0 to remove and 2 not upgraded.
4 not fully installed or removed.
After this operation, 0 B of additional disk space will be used.
Setting up linux-headers-6.2.0-31-generic (6.2.0-31.31~22.04.1) ...
/etc/kernel/header_postinst.d/dkms:
 * dkms: running auto installation service for kernel 6.2.0-31-generic
Error! Could not locate dkms.conf file.
File: /var/lib/dkms/amdgpu/6.0.5-1620033.22.04/source/dkms.conf does not exist.
   ...fail!
run-parts: /etc/kernel/header_postinst.d/dkms exited with return code 4
dpkg: error processing package linux-headers-6.2.0-31-generic (--configure):
 installed linux-headers-6.2.0-31-generic package post-installation script subprocess returned error exit status 1
Setting up linux-image-6.2.0-31-generic (6.2.0-31.31~22.04.1) ...
dpkg: dependency problems prevent configuration of linux-headers-generic-hwe-22.04:
 linux-headers-generic-hwe-22.04 depends on linux-headers-6.2.0-31-generic; however:
  Package linux-headers-6.2.0-31-generic is not configured yet.

dpkg: error processing package linux-headers-generic-hwe-22.04 (--configure):
 dependency problems - leaving unconfigured
dpkg: dependency problems prevent configuration of linux-generic-hwe-22.04:
 linux-generic-hwe-22.04 depends on linux-headers-generic-hwe-22.04 (= 6.2.0.31.31~22.04.8); however:
  Package linux-headers-generic-hwe-22.04 is not configured yet.
dpkg: error processing package linux-generic-hwe-22.04 (--configure):
dependency problems - leaving unconfigured
Processing triggers for linux-image-6.2.0-31-generic (6.2.0-31.31~22.04.1) ...
/etc/kernel/postinst.d/dkms:
 * dkms: running auto installation service for kernel 6.2.0-31-generic
Error! Could not locate dkms.conf file.
File: /var/lib/dkms/amdgpu/6.0.5-1620033.22.04/source/dkms.conf does not exist.
   ...fail!
run-parts: /etc/kernel/postinst.d/dkms exited with return code 4
dpkg: error processing package linux-image-6.2.0-31-generic (--configure):
 installed linux-image-6.2.0-31-generic package post-installation script subprocess returned error exit status 1
No apport report written because MaxReports is reached already
                                                              Errors were encountered while processing:
 linux-headers-6.2.0-31-generic
 linux-headers-generic-hwe-22.04
 linux-generic-hwe-22.04
 linux-image-6.2.0-31-generic
E: Sub-process /usr/bin/dpkg returned an error code (1)

Does this AskUbuntu answer solve your issue? https://askubuntu.com/a/231908

# ? Aug 29, 2023 15:22

Adbot: ADBOT LOVES YOU

# ? May 24, 2024 12:29

hifi: Jul 25, 2012

notwithoutmyanus posted:

So, I am running ubuntu and jsut tried to do an apt-get upgrade after updating, and I get this. It looks like there's possibly a dpkg issue involving AMDgpu drivers or something? How can I troubleshoot/resolve this? I think at one point things were bugging out and I ended up doing a dpkg remove on amdgpu - but googling this sort of issue didn't find me anything (probably because I don't know which part to troubleshoot outside of the dpkg error)

It's a dkms issue where it's going thru every folder in /var/lib/dkms/ looking for kernel modules to build for the kernel you just installed. If you removed an amd gpu package then there's probably crap left over there that dpkg wasn't or stopped tracking, specifically you probably have to delete /var/lib/dkms/amdgpu/6.0.5-1620033.22.04/ but poke around and check what kernel version(s) you have, amdgpu version, and if anything is just straight up a broken symlink for instance.

# ? Aug 29, 2023 16:24

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

«‹›1033 »