Ok I nearly lost hope, since not being able to figure it out.

PLEASE. SEND. HELP.

_ First: this might gonna get a long one, but I‘m desperately looking for help!
Second: I‘m a total newb on Linux, so I have really limited Linux know-how.

Specs:
Asus ROG Strix G15DS-R7700X088W
AMD Ryzen 7 7700X
NVIDIA GeForce RTX 3070
2x 1TB SSDs; 1x M2 NVME with W11 running, 1x SATA

Goal:
Running Dual Boot with W11 on the first, M2 SSD (already running fine) and Linux (Nobara preferred) on the second, SATA SSD

Distros I tried:
Nobara 39
Fedora 39
Fedora 38
Ubuntu 22.04
Pop!_OS

Problem I run into:
I can‘t boot even from the LiveUSB without the „acpi=off“ option. If I do, I get just a black Screen (with Backlight still on) or, if I get into the Grub options first, there‘s only „booting command list“ visible but nothing else happens (even with „quiet“ disabled, no info on the Screen at all).
One thing I noticed, since my Keyboard, Mouse and Mousemat (Razerfly) have lighting, when I try to boot without the acpi=off, they go dark. And stay dark. With acpi=off the keyboard alone goes dark but then lights up again after 2-3 seconds.
If I run it with acpi=off, I can boot and install, but I then have to boot every time with acpi=off. This leads to the graphics driver not being recognized by the OS and running always in 1024x768 „software rendering“ resolution (even with proper drivers installed and enabled and nouveau on blacklist). So just let „acpi=off“ enabled isn‘t an option.

I did, after researching for several hours, try with various other options (nomodeset, acpi=ht, pci=biosirq, noapic, nolapic, and so on, tried a ton of those) but nothing did the trick - always black screen of death without acpi=off.

I did update my BIOS to the latest Version (306), did try every possibilty of options enabled/disabled (Fast Boot, Secure Boot, IOMMU, acpi settings in BIOS, secondary on-board Graphics,…) with no change.

Since I ran out of options (in relation to my google and reddit search skills), knowledge (total newb on Linux) and possibility to ask friends (that know more about linux than me), I‘m desperate enough to ask for help.

You are my last hope, before giving up on Linux with my PC.

If someone has an idea I could try or even a solution, I‘d be endlessly thankful!

If I missed some info or something is needed, don‘t hesitate to as for specific details._

#linuxquestions

  • Varen@kbin.socialOP
    1·
    9 months ago

    Thank you.

    The only way I got it working was again with acpi=off and with the 3 I just land on the CLI?

        • onlinepersonaEnglish
          1·
          9 months ago

          Could you share them? Anything that looks like an error would be good. If you can, can you export the output of dmseg (prints the kernel logs)?

          dmesg > /tmp/kernel.log
curl --upload-file ./kernel.log https://transfer.sh/kernel.log

          That will output a link to the uploaded kernel log that you can share here. It should work with the one that started with acpi=off. Since your system won’t start without it, you can just take a picture after it crashes if it has text output. Basically, we need to see the error in order to be able to help you.

          Edit: if you have an HDMI capture card and a second computer, a recording of the HDMI output would be great.

          • Varen@kbin.socialOP
            1·
            9 months ago

            Sure. The whole output or something like dmesg -T --level=emerg,alert,crit,err,warn ?

            • onlinepersonaEnglish
              2·
              9 months ago

              I don’t think it reveals anything private, so the entire output should be fine. If you’re uncomfortable, then --level=emerg,alert,crit,err,warn should probably do. -T is unncessary, IMO.

              • Varen@kbin.socialOP
                2·
                9 months ago

                there you go:
                https://pastebin.com/kNFqya2L

                taken with acpi=off boot.
                Since without acpi=off it’s not possible to boot and there is absolutely no output after grub (as if the kernel wouldn’t load at all) there simply nothing to show …

                • onlinepersonaEnglish
                  1·
                  9 months ago

                  [ 9.101103] mt7921e 0000:09:00.0: can’t find IRQ for PCI INT A; please try using pci=biosirq

                  Have you tried pci=biosirq? From the description of the kernel parameter I don’t think it’s the right solution, but it might be worth trying.

                  biosirq		[X86-32] Use PCI BIOS calls to get the interrupt
				routing table. These calls are known to be buggy
				on several machines and they hang the machine
				when used, but on other computers it's the only
				way to get the interrupt routing table. Try
				this option if the kernel is unable to allocate
				IRQs or discover secondary PCI buses on your
				motherboard.

                  This one below though, might be the most important clue to a solution.

                  [   12.149945] NVRM: The NVIDIA probe routine failed for 1 device(s).
[   12.150113] NVRM: None of the NVIDIA devices were initialized.
[   14.289240] NVRM: Can't find an IRQ for your NVIDIA card!
[   14.294129] NVRM: Please check your BIOS settings.
[   14.294303] NVRM: [Plug & Play OS] should be set to NO
[   14.294475] NVRM: [Assign IRQ to VGA] should be set to YES

                  Maybe check your BIOS settings for anything that resembles those last 2 lines?

                  CC BY-NC-SA 4.0

                  • Varen@kbin.socialOP
                    2·
                    9 months ago

                    Yep, saw this entries already and tried with it. To no change unfortunately.
                    Did some research and also found the suggest to try with
                    acpi=noirq noapic
                    but also tried to no change.

                    Regarding the BIOS options, I did also check for these options, bug they simply aren‘t there 🤷🏻‍♂️ but will look again after those. Got told that this error might be due to the acpi=off thingy