Cannot initialize GSP firmware RM

[hrushirajg23]

NVIDIA Open GPU Kernel Modules Version

580.95.05

Please confirm this issue does not happen with the proprietary driver (of the same version). This issue tracker is only for bugs specific to the open kernel driver.

• I confirm that this does not happen with the proprietary driver package.

Operating System and Version

No LSB modules are available. Description: Ubuntu 24.04.3 LTS

Kernel Release

Linux vicharak 6.1.75-axon #axon SMP Tue Nov 11 13:04:42 IST 2025 aarch64 aarch64 aarch64 GNU/Linux

Please confirm you are running a stable release kernel (e.g. not a -rc). We do not accept bug reports for unreleased kernels.

• I am running on a stable kernel release.

Hardware: GPU

geforce rtx 3080

Describe the bug

when this issue is solved I face the following issue, randomly by some trial and error I face this issue with open drivers.

[ 10.842101] NVRM: s_vbiosPatchInterfaceData: Found pIntFaceHdr entry count: 0, expected 2
[ 10.842107] NVRM: s_vbiosPatchInterfaceData: too few interface entires found for FWSEC cmd 0x15
[ 10.842112] NVRM: s_prepareForFwsec_TU102: Falcon ucode from hs
[ 10.842114] NVRM: s_prepareForFwsec_TU102: failed to prepare interface data for FWSEC cmd 0x15: 0x25
[ 10.842116] NVRM: s_prepareForFwsec_TU102: (note: VBIOS version 94.02.71.40.83)
[ 10.842119] NVRM: nvCheckOkFailedNoLog: Check failed: Invalid data passed [NV_ERR_INVALID_DATA] (0x00000025) returned from kgspPrepareForBootstrap_HAL(pGpu, pKernelGsp, KGSP_BOOT_MODE_NORMAL) @ kernel_gsp.c:3664
[ 10.842171] NVRM: RmInitAdapter: Cannot initialize GSP firmware RM
[ 10.843941] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x62:0x25:2015)
[ 10.845693] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0

would you let me the reason behind the "Cannot initialize GSP firmware RM" issue ?

With the proprietary modules I face only the following issue.
[ 10.843941] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x62:0x25:2015)
[ 10.845693] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0

To Reproduce

occurs on every boot

Bug Incidence

Always

nvidia-bug-report.log.gz

nvidia-bug-report.log.gz

More Info

No response

[MacroMelon]

I belive I'm facing a potentially simillar issue (version 580.105.08):

My RTX 3080 mobile (device - Lenovo T15g Gen 2i, Intel i9 + rtx3080 + Intel iGPU with hybrid graphics, running the latest BIOS they provide) didn't work right (nvidia-smi returns No devices were found but the kernal modules and drivers were loaded correctly according to lspci -k -d ::03xx)

After days of trying to diagnose the issue, I found that it was because the GSP wasn't being initialised properly - mainly from dmesg logs:

NVRM: nvAssertOkFailedNoLog: Assertion failed: Failure: Generic Error [NV_ERR_GENERIC] (0x0000FFFF) returned from RPC_HDR->rpc_result @ kernel_gsp.c:4999
NVRM: nvAssertOkFailedNoLog: Assertion failed: Failure: Generic Error [NV_ERR_GENERIC] (0x0000FFFF) returned from kgspWaitForRmInitDone(pGpu, pKernelGsp) @ kernel_gsp_tu102.c:570
NVRM: _kgspBootGspRm: unexpected WPR2 already up, cannot proceed with booting GSP
NVRM: _kgspBootGspRm: (the GPU is likely in a bad state and may need to be reset) 
NVRM: RmInitAdapter: Cannot initialize GSP firmware RM
NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x62:0x40:2015)
NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0

After which I switched to the proprietary nvidia package (not nivida-open), and the issue persisted, but now the logs were just:

NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x62:0x40:2661)
NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0

I "solved" (read: avoided) the issue by disabling the GSP firmware entirely by creating the file /etc/modprobe.d/nvidia.conf containing:

options nvidia NVreg_EnableGpuFirmware=0

and then rebuilding initramfs using mkinitcpio -P

Although this "solves" the issue and everything works for me now, its less than ideal because I would like to use the nvidia-open drivers and also leverage the benefits of the CPU offload to the GSP firmware.

[awsms]

@MacroMelon sounds like the 580.95.05 regression that affects Legion laptops (tracked as 5622970 internally). Still no fixes or news announced whatsoever. Can you try downgrading to 580.82.09?
https://forums.developer.nvidia.com/t/no-hdmi-output-on-nobara-linux-42-from-lenovo-legion-7i-w-rtx-3080/347774

[amrit1711]

Hi All,
We are tracking this issue internally with bug 5053788.
Shall share further feedback once it is available.

[MacroMelon]

sounds like the 580.95.05 regression that affects Legion laptops (tracked as 5622970 internally). Still no fixes or news announced whatsoever. Can you try downgrading to 580.82.09?

@awsms Ooh that seems like a good shout! Thanks for bringing it to my attention - unfortunately I use this pc for work and have a few projects going on right now, I'm a bit hesistant to mess with it at the moment. Once I get a bit of breathing room I'll defenitely give it shot! Thanks :D
(also I don't really understand how downgrading works with pacman yet)

We are tracking this issue internally with bug 5053788.
Shall share further feedback once it is available.

Nice! Hoping for the best!

[solskogen]

I got this problem as well on NVIDIA GeForce RTX 2060, and it was caused by using a kernel that was using 16K_PAGES. When I changed to CONFIG_ARM64_4K_PAGES, the problems went away.

[awsms]

@MacroMelon they were able to reproduce it. now let's see wait for a fix to come.

[MacroMelon]

@MacroMelon sounds like the 580.95.05 regression that affects Legion laptops (tracked as 5622970 internally). Still no fixes or news announced whatsoever. Can you try downgrading to 580.82.09? https://forums.developer.nvidia.com/t/no-hdmi-output-on-nobara-linux-42-from-lenovo-legion-7i-w-rtx-3080/347774

Just an update on this, I did in fact try it by installing the 580.82.09 driver via the nvidia-580xx-dkms packages from the AUR, and unfortunately it's still the same (works when GSP firmware is disabled, doesn't work if enabled). Strange, but hopefully the information helps?

[edit]

Just as some context, the only reason I did this was because I was kinda forced to as the main arch repository dropped the nvidia proprietary package, so it was either I go ahead and mess with older driver versions or not be able to update any other packages via packman.

It's a bit bleh because I had to force remove the nivida-utils (v590xx) package, which lib32-nvidia-utils depends on (which a bunch of other stuff depends on) and break the dependancies (unless, afaik, I want to go on a whole side quest of manually finding and installing the right versions of everything), so I really hope I don't run into any issues becuase of it :|

It seems fine so far! :D