[arch-general] Problematic boot on 202104 iso
Hello, While installing a new machine, the 202104 iso failed to complete the boot process, stalling with a blinking cursor at 0.003625 __common_interrupt: 1.55 No irg handler for vector 0.003625 __common_interrupt: 2.55 No irq handler for vector 0.003625 __common_interrupt: 3.55 No irq handler for vector 0.003625 __common_interrupt: 4.55 No irq handler for vector 0.003625 __common_interrupt: 5.55 No irq handler for vector 0.003625 __common_interrupt: 6.55 No irqg handler for vector 0.003625 __common_interrupt: 7.55 No irq handler for vector 0.003625 __common_interrupt: 8.55 No irqg handler for vector 0.003625 __common_interrupt: 9.55 No irq handler for vector 0.003625 __common_interrupt: 10.55 No irq handler for vector :: running early hook [udev | Starting version 247 .4-2-arch :: running early hook [archiso_pxe_nbd]l :: rumning hook [udevl :: Triggering uevents... It's a desktop PC having UEFI firmware, with CSM disabled, secure boot disabled and fast boot off. The same result is obtained while booting from an internal DVD drive and from a USB stick. Noise and lights from the devices suggest that the booting process is completing without updating the display; that's just a guess however. I could boot from 202011 on DVD, and from 201907 from a USB stick. It may be host dependent; reporting FYI, if it could ever be reproduced and eventually fixed. Regards.
0.003625 __common_interrupt: 1.55 No irg handler for vector 0.003625 __common_interrupt: 2.55 No irq handler for vector 0.003625 __common_interrupt: 3.55 No irq handler for vector 0.003625 __common_interrupt: 4.55 No irq handler for vector 0.003625 __common_interrupt: 5.55 No irq handler for vector 0.003625 __common_interrupt: 6.55 No irqg handler for vector 0.003625 __common_interrupt: 7.55 No irq handler for vector 0.003625 __common_interrupt: 8.55 No irqg handler for vector 0.003625 __common_interrupt: 9.55 No irq handler for vector 0.003625 __common_interrupt: 10.55 No irq handler for vector
I am assuming that you are running a recent AMD Ryzen chipset. This issue was also discussed in this forum post: https://bbs.archlinux.org/ viewtopic.php?id=256227 This is an issue with the UEFI. It can be fixed by updating to a UEFI version that includes AGESA 1.2.0.1 as it was officially fixed in that version. Kind regards, -- Sefa Eyeoglu https://scrumplex.net
Le dimanche 18 avril 2021 16:50:43 CEST Sefa Eyeoglu via arch-general a écrit :
This is an issue with the UEFI. It can be fixed by updating to a UEFI version that includes AGESA 1.2.0.1 as it was officially fixed in that version.
Yes, it's an ASUS TUF X570-PLUS GAMING motherboard. Grasped the latest firmware and will update ASAP. Thank you very much.
Le dimanche 18 avril 2021 16:50:43 CEST Sefa Eyeoglu via arch-general a écrit :
This is an issue with the UEFI. It can be fixed by updating to a UEFI version that includes AGESA 1.2.0.1 as it was officially fixed in that version.
After updating the motherboard firmware, the interrupt error no longer occurs. But with 202104 the boot screen does not update after :: running early hook [udev] Starting version 247.4-2-arch :: running early hook [archiso_pxe_nbd] :: running hook [udev] :: Triggering uevents... _ The boot process does complete in the background, commands blindly typed in are executed : loadkeys fr systemctl reboot No problems are seen with 202102 iso. It happens with one machine only, 202104 iso can boot two other machines normally. As it is host specific, there is probably no valid reason to spend more time on it. More debugging would be of interest if other users report similarly. Regards.
SET via arch-general <arch-general@lists.archlinux.org> wrote:
Le dimanche 18 avril 2021 16:50:43 CEST Sefa Eyeoglu via arch-general a ?crit :
This is an issue with the UEFI. It can be fixed by updating to a UEFI version that includes AGESA 1.2.0.1 as it was officially fixed in that version.
After updating the motherboard firmware, the interrupt error no longer occurs.
But with 202104 the boot screen does not update after
:: running early hook [udev] Starting version 247.4-2-arch :: running early hook [archiso_pxe_nbd] :: running hook [udev] :: Triggering uevents... _
The boot process does complete in the background, commands blindly typed in are executed : loadkeys fr systemctl reboot
No problems are seen with 202102 iso.
It happens with one machine only, 202104 iso can boot two other machines normally. As it is host specific, there is probably no valid reason to spend more time on it.
I am confused about lack of valid reason to spend more time on it. Are those two other machines also have ASUS TUF X570-PLUS GAMING motherboard? I don't have such hardware, and can not help with debuging. -- u34
More debugging would be of interest if other users report similarly.
Regards.
Le lundi 19 avril 2021 20:32:50 CEST u34--- via arch-general a écrit :
Are those two other machines also have...
No, one is a desktop PC with an older MSI 970 GAMING mobo; the other one is an ASUS S505ZA laptop.
I am confused about lack of valid reason to spend more time on it. Well, I meant I don't wish to waste developer time debuging this if it is felt as a niche dysfunction. If someone wants to get more information for a fix, I am ready to follow precise instructions and feedback the results. Now it concerns my work machine, so it may be a lengthy workout as I can't reboot at any time everyday. But I'm ready.
Le lundi 19 avril 2021 20:32:50 CEST u34--- via arch-general a écrit :
I am confused about lack of valid reason to spend more time on it.
I could compare dmesg output from 202102 and 202104 (blindly typing commands). Significant errors in 202104 are : [ 4.383429] amdgpu 0000:0a:00.0: Direct firmware load for amdgpu/ navy_flounder_sos.bin failed with error -2 [ 4.383431] amdgpu 0000:0a:00.0: amdgpu: failed to init sos firmware [ 4.383433] [drm:psp_sw_init [amdgpu]] *ERROR* Failed to load psp firmware! [ 4.383543] [drm:amdgpu_device_init.cold [amdgpu]] *ERROR* sw_init of IP block <psp> failed -2 [ 4.383668] amdgpu 0000:0a:00.0: amdgpu: amdgpu_device_ip_init failed [ 4.383672] amdgpu 0000:0a:00.0: amdgpu: Fatal error during GPU init [ 4.383688] amdgpu: probe of 0000:0a:00.0 failed with error -2 [ 4.427485] random: fast init done The host has a RADEON RX 6700 XT video card. It seems that the amdgpu module enforces the presence of this chip's firmware in 202104, but not in 202102. I guess the next install ISO would include navy_flounder* firmware and this should no more happen. For daily work, I did indeed copy navy_flounder* in the firmware directory following phoronix reading. At least, it's no longer a mystery. Regards.
participants (3)
-
Sefa Eyeoglu
-
SET
-
u34＠net9.ga