PC hangs loading initramfs (ACPI messes up?)

smitropoulos · 2021-02-04 10:26:36

Hello all!

I have been developing a fear of restarts after updates for the past couple of years. For the most part, my PC has been fine after them, but some times it just hangs in (when I used GRUB) `Loading initial ramdisk` or a black screen (not off, just black) after the message that the ramdisk is being loaded (with systemd-boot AND rEFInd).

Since this is a recurring hit or miss situation, I decided to go ahead and try to fix it. I concluded (please give insight) that this is probably a bootloader issue, or I am creating (for some reason) corrupt initram images. In my tries to solve, I have been running mkinitcpio -P for every restart.

some specs:

GPU
05:00.0 VGA compatible controller: NVIDIA Corporation GP104 [GeForce GTX 1070] (rev a1)  

CPU
model name      : Intel(R) Core(TM) i7-5820K CPU @ 3.30GHz

MOBO
Base Board Information
        Manufacturer: Gigabyte Technology Co., Ltd.
        Product Name: X99-UD4-CF

RAM
Corsair Vengeance 8GB@3200Mhz  x 4

I have read seemingly most of the posts in the forum regarding similar issues, and have tried:

Using Grub

Hardware tweaks
* Booting using only one drive, disconnecting my other 5 or 6
* Using different RAM sticks
* Using only one stick
* Messed with the UEFI settings. I have followed the wiki guidelines and disable fast and secure boot always.

Software tweaks
* Trying fallback images. It seems to work more often than not, but still.
* Running linux-lts
* Checked the ESP was booted properly and in the correct order in fstab
* Running EFI from my Windows disk ESP
* Running intel-ucode as the first initrd
* Switch between different nvidia drivers, nvidia, nvidia-lts (for each kernel), nouveau
* Wiping clean my ESP (on the linux disk) and starting again
* All of the different kernel options for booting, nomodset, acpi=off, noquiet, nosplash etc.
* Increasing log levels / using kernel debug mode in hope that something gets recorded regarding the problem
* Installing haveged in case the delay/hung was related to entropy
* Mashing keys while stuck in case the delay/hung was related to entropy

Using systemd-boot

Not much, just installed the day before yesterday. Managed to boot first try using linux. Restarted and it hung.

Using rEFInd

Again not much since I installed that just after the systemd-boot failed yesterday BUT it gave me some information after its first failed attempt to boot.

What I mean by ts first failed attempt to boot; sometimes, hard resetting the PC after a hang, causes it to boot. Most of the times it takes multiple hard resets and booting to a fallback image.

So yesterday I noticed this (disregard the datetime, I just found the same errors on previous logs):

Jan 04 08:43:12 Blastoise kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT1.SPT0._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Jan 04 08:43:12 Blastoise kernel: ACPI Error: Aborting method \_SB.PCI0.SAT1.SPT0._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Jan 04 08:43:12 Blastoise kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT1.SPT3._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Jan 04 08:43:12 Blastoise kernel: ACPI Error: Aborting method \_SB.PCI0.SAT1.SPT3._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Jan 04 08:43:12 Blastoise kernel: ata5.00: ATA-9: WDC WD20EZRZ-00Z5HB0, 80.00A80, max UDMA/133
Jan 04 08:43:12 Blastoise kernel: ata5.00: 3907029168 sectors, multi 16: LBA48 NCQ (depth 32), AA
Jan 04 08:43:12 Blastoise kernel: ata8.00: HPA detected: current 1953523055, native 1953525168
Jan 04 08:43:12 Blastoise kernel: ata8.00: ATA-9: WDC WD10EZEX-08M2NA0, 01.01A01, max UDMA/100
Jan 04 08:43:12 Blastoise kernel: ata8.00: 1953523055 sectors, multi 16: LBA48 NCQ (depth 32), AA
Jan 04 08:43:12 Blastoise kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT1.SPT0._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Jan 04 08:43:12 Blastoise kernel: ACPI Error: Aborting method \_SB.PCI0.SAT1.SPT0._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Jan 04 08:43:12 Blastoise kernel: ata5.00: configured for UDMA/133
Jan 04 08:43:12 Blastoise kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT1.SPT3._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Jan 04 08:43:12 Blastoise kernel: ACPI Error: Aborting method \_SB.PCI0.SAT1.SPT3._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Jan 04 08:43:12 Blastoise kernel: ata8.00: configured for UDMA/100
Jan 04 08:43:12 Blastoise kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT1.SPT2._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Jan 04 08:43:12 Blastoise kernel: ACPI Error: Aborting method \_SB.PCI0.SAT1.SPT2._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Jan 04 08:43:12 Blastoise kernel: ata7.00: supports DRM functions and may not be fully accessible
Jan 04 08:43:12 Blastoise kernel: ata7.00: ATA-9: Samsung SSD 750 EVO 250GB, MAT01B6Q, max UDMA/133
Jan 04 08:43:12 Blastoise kernel: ata7.00: 488397168 sectors, multi 1: LBA48 NCQ (depth 32), AA
Jan 04 08:43:12 Blastoise kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT1.SPT2._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Jan 04 08:43:12 Blastoise kernel: ACPI Error: Aborting method \_SB.PCI0.SAT1.SPT2._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Jan 04 08:43:12 Blastoise kernel: ata7.00: supports DRM functions and may not be fully accessible
Jan 04 08:43:12 Blastoise kernel: ata7.00: configured for UDMA/133
Jan 04 08:43:12 Blastoise kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT1.SPT1._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Jan 04 08:43:12 Blastoise kernel: ACPI Error: Aborting method \_SB.PCI0.SAT1.SPT1._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)
Jan 04 08:43:12 Blastoise kernel: ata6.00: ATA-8: OCZ-VERTEX3, 2.25, max UDMA/133
Jan 04 08:43:12 Blastoise kernel: ata6.00: 234441648 sectors, multi 16: LBA48 NCQ (depth 32), AA
Jan 04 08:43:12 Blastoise kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.SAT1.SPT1._GTF.DSSP], AE_NOT_FOUND (20200925/psargs-330)
Jan 04 08:43:12 Blastoise kernel: ACPI Error: Aborting method \_SB.PCI0.SAT1.SPT1._GTF due to previous error (AE_NOT_FOUND) (20200925/psparse-529)

I immediately used acpi=off and it magically worked. The PC booted fast, without problems 10 times in a row. Do you think it has anything to do with ACPI? Is this a common problem?

Last edited by smitropoulos (2021-02-04 10:27:46)

V1del · 2021-02-04 10:41:30

Bootloader is going to be largely irrelevant on this. What is "not booting" for you? What should come up a graphical session? A commonish issue is a race condition in between the xorg initialization and the graphics card being up, which can be alleviated by adding the relevant modules to your initramfs: https://wiki.archlinux.org/index.php/NV … de_setting

Various ACPI table errors are usually normal and not necessarily a cause for concern, and completely disabling acpi might have unwanted side effects, what you can try here in order to fix the ACPI errors in the log is doing a UEFI/motherboard firmware update and hoping that things have been fixed here.

smitropoulos · 2021-02-04 10:53:00

V1del wrote:

Bootloader is going to be largely irrelevant on this. What is "not booting" for you? What should come up a graphical session? A commonish issue is a race condition in between the xorg initialization and the graphics card being up, which can be alleviated by adding the relevant modules to your initramfs: https://wiki.archlinux.org/index.php/NV … de_setting
Various ACPI table errors are usually normal and not necessarily a cause for concern, and completely disabling acpi might have unwanted side effects, what you can try here in order to fix the ACPI errors in the log is doing a UEFI/motherboard firmware update and hoping that things have been fixed here.

Firstly thanks for the answer,

A successful boot is: booting to console, logging in and using startx, so I guess the race condition is not the problem. Also, I think you suggested it in another post, I tried to switch TTYs when stuck with no luck .

Forgot to mention I have already gone through the KMS steps with no results.

As for the mobo firmware, nothing has been officially released since 2018.

In case you do find this interesting, I can post more logs any day

Last edited by smitropoulos (2021-02-04 11:14:49)

smitropoulos · 2021-02-05 07:50:18

An update:

Since it was time for an upgrade already, I opted for a B550 asus mobo and a Ryzen 5 5600x. The rest of the hardware remains the same. Absolutely no problems whatsoever apart from the need to reinstall nvidia drivers and amd-ucode instead of intel-ucode. Radical troubleshooting procedure, I know.

I have encountered many X99 platform posts in the forums, making me think that the platform, not as successful as others were, did not receive the love it needed from manufacturers, and hence the multiple errors. The CPU's it supported, were a happy middle between server grade commercial use and home PC, supporting ECC memory among others, and thus not been preferred by many. Not to mention it was the first platform to support ddr4.

All in all, Arch now runs no problem.

Arch Linux

#1 2021-02-04 10:26:36