Keyboard not working

farzammag · 2023-08-07 22:09:54

So i made 2 new discoveries:

after a clean shutdown (cutting power fully) i could sucessfully launch into a live USB of arch (a bit older, didnt check the exact kernel, but it worked, thats the point, not a hardware issue)

another discovery is that the error i pointed out, actually occured in the live usb: https://drive.google.com/file/d/1CWzRvb … drive_link
Also the same error you pointed out: https://drive.google.com/file/d/1jUrOpb … drive_link

Sorry for the ugly screenshots, there were no otherways.

That was a relief, at least now i dont have to fight with hardware issues now, but now im just more interested to know, what actually is going on, at this point after 3 days of struggle and troubleshooting, it would just be easier to reinstall the whole system, but again, i am really interested to know what is happening, and how a faulty system (or maybe kernel or firmware) is causing this issue in grub, live usb and bios if i dont fully drain power and boot directly into them.

Another thing: i enabled core-testing and updated the system, hoped that would change anything, didnt make any difference, tried to reboot a bunch of times for a good boot, no chance, so i just rolled back to a snap shot prior to that.

seth · 2023-08-08 13:51:37

after a clean shutdown (cutting power fully) i could sucessfully launch into a live USB of arch

Consistently? You had random success before…

Did you try the LTS kernel?

farzammag · 2023-08-08 14:06:20

seth wrote:

after a clean shutdown (cutting power fully) i could sucessfully launch into a live USB of arch
Consistently? You had random success before…
Did you try the LTS kernel?

yes, booting directly from bios into live usb works constantly.

A reboot from my current system into uefi doesnt work, appareantly it just locks the keyboard.
I have tested the LTS kernel on this installation.

I have an extra NVME lying around, i will install a new system just with lts kernel and give an update.

farzammag · 2023-08-08 19:03:28

So hi there, the random nvme i mentioned actually had a very very old windows installation on it (from the looks of it its from october 2021)

Sad news is, keyboard worked, i removed my old disk, replacing with the new one, wanting to install a full new system for testing, i noticed there is a windows installation on here, after several reboots to make sure this was no accident, it worked.

The first boot didnt work tho, despite totally replacing the disk, the first boot hanged, but after i shut down totally, it worked, i also tested it a bunch of times as mentioned, another thing that i noticed:

Someone gave me the idea that the problem may be caused by mkinitcpio, so i put the keyboard hook before autodetect and rebuild it, didnt help, although the problem doesnt really seem to be caused by that

and also not the kernel, because testing the lts kernel (i had both kernels selectable in grub) didnt help either, i also rolled back immidiately after the update, problem wasnt solved either, so now im getting the idea its being caused way earlier, perhaps in the bootloader, although it works in grub after a clean shutdown, when i select uefi under grub options, it totally stops working, also when i select any kernel or snapshots, it also doesnt work, but booting directly into bios with f2, doesnt cause that issue bios (keyboard works, unlike when i go into that with grub) and also when i directly go into a usb live media of arch (with kernel 6.2.* i think, it worked too), and now with windows with a totally bootmanager, its fine as well.

perhaps we should take a look into grub/uefi.

Last edited by farzammag (2023-08-08 19:19:49)

farzammag · 2023-08-16 19:13:23

Here another succesful boot using a usb stick, im installing a new system, but i noticed a lot of Bios acpi bug and errors?

https://0x0.st/HLa6.txt

seth · 2023-08-16 19:32:05

ACPI errors are sorta normal, also the keyboard works in that context.
Functionality when booting from an external device usually yells "hibernating windows" and your BIOS interference having an impact supports that…
I'd not be overly confident that the re-installed system will work - maybe this is a timing problem/race condition.

What happens when you add

i8042.noaux=1 i8042.nomux=1 i8042.kbdreset=1 i8042.notimeout=1 i8042.dumbkbd=1 i8042.reset

to the https://wiki.archlinux.org/title/Kernel_parameters ?

farzammag · 2023-08-28 11:03:18

Hey, sorry for the late replay, i had a busy week, can i ask if i have to add them all at the same time or add them 1 by one and test them?

seth wrote:

ACPI errors are sorta normal, also the keyboard works in that context.
Functionality when booting from an external device usually yells "hibernating windows" and your BIOS interference having an impact supports that…
I'd not be overly confident that the re-installed system will work - maybe this is a timing problem/race condition.
What happens when you add
i8042.noaux=1 i8042.nomux=1 i8042.kbdreset=1 i8042.notimeout=1 i8042.dumbkbd=1 i8042.reset
to the https://wiki.archlinux.org/title/Kernel_parameters ?

seth · 2023-08-28 11:09:44

All at once.
If it does something, you can bisect down the relevant one(s)

farzammag · 2023-09-08 08:05:23

seth wrote:

All at once.
If it does something, you can bisect down the relevant one(s)

as it took me around 3 weeks of trouble shooting, seems like the problem is solved by its own, but i have a strong feeling it somehow has to do with plasma desktop, im using gnome now, all fine and good, last night i moved to kde because well i liked it,´and suddenly the keyboard issue was there as well.

i reinstalled with gnome and its working now
im not sure if it makes sense to troubleshoot this any further, i thank you for your help

seth · 2023-09-08 08:13:03

Assumung the errors in #14 are relevant the only way KDE could have impacted that is through powerdevil

Once in sddm, keyboard input was unresponsive

Are you still using SDDM or did you also switch to GDM?

However

i reinstalled with…

You could just try to install plasma and start it from GDM and see what happens, becuase it's far more likely that the re-installation was more instrumental here than any DE switch.

farzammag · 2023-09-08 09:32:17

seth wrote:

Assumung the errors in #14 are relevant the only way KDE could have impacted that is through powerdevil
Once in sddm, keyboard input was unresponsive
Are you still using SDDM or did you also switch to GDM?
However
i reinstalled with…
You could just try to install plasma and start it from GDM and see what happens, becuase it's far more likely that the re-installation was more instrumental here than any DE switch.

i really couldnt think that a software could affect the hardware like that even after the reboot after bios, out of curiosity i will reinstall sddm, enable it, later the desktop itself to find out what exactly was happening.
ill let you know about more today.
will you still need the journal? or should i rather look for xorg logs and so on?

and now that i remember i think i know why i had that random successful boot, it was because i disabled sddm to just go into tty directly!

Last edited by farzammag (2023-09-08 09:34:55)

seth · 2023-09-08 12:48:05

If we ignore the device flicker in #14 and focus on the DM I'd concern DM and DE being on X11 and wayland (or vv.)
In theroy anything could (effectively) "xintput disable <device>" or "evtest --grab <device>" and it become unavailable.
In practice that would be considered malicious

Do/did you run any keyloggers etc?

farzammag · 2023-09-08 13:45:52

technically i did not install anything from any source beside official pacman mirrors, and only used those with ssl

and also i had this issue on a freshly installed system as soon as i installed kde so i dont think this is rather a malware thing

i also really never thought xorg input could make a usb device disappear from lsusb, im still convinced its just a bug

im at work now, as soon as i arrive ill test plasma and sddm on my new system

maybe i fucked up somewhere in my old system and we will never find out what caused it, but i hope not, at this point i already have a solution, i just use gnome, but out of curiosity id still want to solve this problem

seth · 2023-09-08 13:55:55

Disappearing from lsusb means the device fell off the bus and this would be (liekly) the HW or the kernel, not anything in userspace.
However, usb autosuspend might trigger that; some HW doesn't react well to this kind of power saving.
gnome may simply keep the device/bus busy and implicitly prevent any autosuspend?

farzammag · 2023-09-08 14:14:23

Okay, i will try to disable usb autosusbend in bios, and also in kde power manager (i think it was upower?)
i hope i can reproduce the problem in the first place

but shouldnt usb auto suspend be actually for external usbs? does it really have an impact on internal hardware?

seth · 2023-09-08 14:34:27

does it really have an impact on internal hardware?

Does the desired keybaord actually show up in lsusb right now?
Internal keyboards are often not wired through USB but i8042 or i2c - otherwise USB is a bus, not a plug.
So all devices are subject to autosuspend, regardless of the cable.

farzammag · 2023-09-08 15:17:12

yes, when it is recognized it is under lsusb, then ill try to take a look under upower, other power managers and under bios for autosuspend

farzammag · 2023-09-08 19:30:27

seth wrote:

Assumung the errors in #14 are relevant the only way KDE could have impacted that is through powerdevil
Once in sddm, keyboard input was unresponsive
Are you still using SDDM or did you also switch to GDM?
However
i reinstalled with…
You could just try to install plasma and start it from GDM and see what happens, becuase it's far more likely that the re-installation was more instrumental here than any DE switch.

unfortunately i couldnt reproduce the problem, i assume something is wrong with my installation process as i tweaked a lot of things
it was a seup i had since 2021
this time i used archinstall for a quick install and testing

but again the previous system was also a fresh install but it still broke :

i assume archinstall did things better than i did

although yesterday i didnt do alot of tweaking
here my journal of current boot
https://0x0.st/HfBT.txt

Last edited by farzammag (2023-09-08 19:33:13)

seth · 2023-09-09 06:28:42

Sep 08 21:23:58 maghouli kernel: usb 1-9: New USB device found, idVendor=048d, idProduct=c100, bcdDevice= 0.44
Sep 08 21:23:58 maghouli kernel: usb 1-9: New USB device strings: Mfr=1, Product=2, SerialNumber=0
Sep 08 21:23:58 maghouli kernel: usb 1-9: Product: ITE Device(8910)
Sep 08 21:23:58 maghouli kernel: usb 1-9: Manufacturer: ITE Tech. Inc.
…
Sep 08 21:23:59 maghouli kernel: hid-generic 0003:048D:C100.0004: input,hiddev96,hidraw3: USB HID v1.10 Keyboard [ITE Tech. Inc. ITE Device(8910)] on usb-0000:00:14.0-9/input0

The flicker in #14 seems very relevant now.

farzammag · 2023-09-09 06:38:06

seth wrote:

Sep 08 21:23:58 maghouli kernel: usb 1-9: New USB device found, idVendor=048d, idProduct=c100, bcdDevice= 0.44
Sep 08 21:23:58 maghouli kernel: usb 1-9: New USB device strings: Mfr=1, Product=2, SerialNumber=0
Sep 08 21:23:58 maghouli kernel: usb 1-9: Product: ITE Device(8910)
Sep 08 21:23:58 maghouli kernel: usb 1-9: Manufacturer: ITE Tech. Inc.
…
Sep 08 21:23:59 maghouli kernel: hid-generic 0003:048D:C100.0004: input,hiddev96,hidraw3: USB HID v1.10 Keyboard [ITE Tech. Inc. ITE Device(8910)] on usb-0000:00:14.0-9/input0

The flicker in #14 seems very relevant now.

what does it mean exactly?

seth · 2023-09-09 06:47:21

Aug 07 19:46:02 thunderkeg kernel: usb 1-9: new full-speed USB device number 5 using xhci_hcd
Aug 07 19:46:02 thunderkeg kernel: usb 1-9: device descriptor read/64, error -71
Aug 07 19:46:03 thunderkeg kernel: usb 1-9: device descriptor read/64, error -71
Aug 07 19:46:03 thunderkeg kernel: usb 1-9: new full-speed USB device number 6 using xhci_hcd
Aug 07 19:46:03 thunderkeg kernel: usb 1-9: device descriptor read/64, error -71
Aug 07 19:46:03 thunderkeg kernel: usb 1-9: device descriptor read/64, error -71
Aug 07 19:46:07 thunderkeg kernel: usb 1-9: new full-speed USB device number 8 using xhci_hcd
Aug 07 19:46:07 thunderkeg kernel: usb 1-9: Device not responding to setup address.
Aug 07 19:46:07 thunderkeg kernel: usb 1-9: Device not responding to setup address.
Aug 07 19:46:08 thunderkeg kernel: usb 1-9: device not accepting address 8, error -71
Aug 07 19:46:08 thunderkeg kernel: usb 1-9: new full-speed USB device number 9 using xhci_hcd
Aug 07 19:46:08 thunderkeg kernel: usb 1-9: Device not responding to setup address.
Aug 07 19:46:08 thunderkeg kernel: usb 1-9: Device not responding to setup address.
Aug 07 19:46:08 thunderkeg kernel: usb 1-9: device not accepting address 9, error -71

The device doesn't respond - for whatever reason.
Typically HW, maybe a race condition between some devices on the bus, by rare chance a bug in xhci_hcd.

farzammag · 2023-09-09 06:51:24

seth wrote:

Aug 07 19:46:02 thunderkeg kernel: usb 1-9: new full-speed USB device number 5 using xhci_hcd
Aug 07 19:46:02 thunderkeg kernel: usb 1-9: device descriptor read/64, error -71
Aug 07 19:46:03 thunderkeg kernel: usb 1-9: device descriptor read/64, error -71
Aug 07 19:46:03 thunderkeg kernel: usb 1-9: new full-speed USB device number 6 using xhci_hcd
Aug 07 19:46:03 thunderkeg kernel: usb 1-9: device descriptor read/64, error -71
Aug 07 19:46:03 thunderkeg kernel: usb 1-9: device descriptor read/64, error -71
Aug 07 19:46:07 thunderkeg kernel: usb 1-9: new full-speed USB device number 8 using xhci_hcd
Aug 07 19:46:07 thunderkeg kernel: usb 1-9: Device not responding to setup address.
Aug 07 19:46:07 thunderkeg kernel: usb 1-9: Device not responding to setup address.
Aug 07 19:46:08 thunderkeg kernel: usb 1-9: device not accepting address 8, error -71
Aug 07 19:46:08 thunderkeg kernel: usb 1-9: new full-speed USB device number 9 using xhci_hcd
Aug 07 19:46:08 thunderkeg kernel: usb 1-9: Device not responding to setup address.
Aug 07 19:46:08 thunderkeg kernel: usb 1-9: Device not responding to setup address.
Aug 07 19:46:08 thunderkeg kernel: usb 1-9: device not accepting address 9, error -71

The device doesn't respond - for whatever reason.
Typically HW, maybe a race condition between some devices on the bus, by rare chance a bug in xhci_hcd.

yes i understood that, but it still does work and shows up under lsusb
another reason this could be problematic on my previous setups was maybe because i installed some xhci_pci firmware off of aur? maybe not the brightest move
ill be trying to reproduce the problem on the weekend

seth · 2023-09-09 07:03:26

Please don't full quote, you're bloating the thread w/ redundant data.
There's no such thing as xhci_pci firmware (on the AUR and I'm not sure what it would be in theory)

farzammag · 2023-09-09 07:06:09

https://aur.archlinux.org/packages/mkinitcpio-firmware

i think it was this, maybe im wrong, but i remember i installed it because i got the missing firmware errors

seth · 2023-09-09 07:11:38

oPtIoNaL FiRmWaRe fOr tHe dEfAuLt lInUx kErNeL To gEt rId oF ThE AnNoYiNg wArNiNg: PoSsIbLy mIsSiNg fIrMwArE FoR MoDuLe: MeSsAgEs

That package is as idiotic as it gets, but I'd be surprised if it can also knock out your hardware.

Arch Linux

#26 2023-08-07 22:09:54

Re: Keyboard not working

#27 2023-08-08 13:51:37

Re: Keyboard not working

#28 2023-08-08 14:06:20

Re: Keyboard not working

#29 2023-08-08 19:03:28

Re: Keyboard not working

#30 2023-08-16 19:13:23

Re: Keyboard not working

#31 2023-08-16 19:32:05

Re: Keyboard not working

#32 2023-08-28 11:03:18

Re: Keyboard not working

#33 2023-08-28 11:09:44

Re: Keyboard not working

#34 2023-09-08 08:05:23

Re: Keyboard not working

#35 2023-09-08 08:13:03

Re: Keyboard not working

#36 2023-09-08 09:32:17

Re: Keyboard not working

#37 2023-09-08 12:48:05

Re: Keyboard not working

#38 2023-09-08 13:45:52

Re: Keyboard not working

#39 2023-09-08 13:55:55

Re: Keyboard not working

#40 2023-09-08 14:14:23

Re: Keyboard not working

#41 2023-09-08 14:34:27

Re: Keyboard not working

#42 2023-09-08 15:17:12

Re: Keyboard not working

#43 2023-09-08 19:30:27

Re: Keyboard not working

#44 2023-09-09 06:28:42

Re: Keyboard not working

#45 2023-09-09 06:38:06

Re: Keyboard not working

#46 2023-09-09 06:47:21

Re: Keyboard not working

#47 2023-09-09 06:51:24

Re: Keyboard not working

#48 2023-09-09 07:03:26

Re: Keyboard not working

#49 2023-09-09 07:06:09

Re: Keyboard not working

#50 2023-09-09 07:11:38

Re: Keyboard not working

Board footer