Trouble Xfce login and video 9070 XT amdgpu - DP

Desktop Environments XFCE
1

i have some error on boot and just after login Xfce ,
i may have some errors Kernels during boot between Grub , and before before xfce starting that can result in fail screen ( lost signal) , that do appears less since 6.15.9 ( patch drm scheduler )

on other side since few days , start xfxe or just after login xfce i can have a freeze about 5 or more sec , sometimes it’s ok , but last this morning screen black but mouse working.

in the last case error betwen xorg & admpgu appears , even a restart lightdm redo same error

see send log journal & inxi
inxi
https://0x0.st/8Fmj.txt

process sigterm
https://0x0.st/8Fme.txt

black screen xfce
https://0x0.st/8Fm2.txt

another this morning
https://0x0.st/8F3F.txt

2

Hey, you have posted the same link with inxi info three times.

I’ve looked at the last one with your journal, and it’s not showing much. There are a couple of segfaults/coredumps when restarting lightdm/xfce, but nothing relevant IMO.

Can you post your xorg log?

3

i will check in evening and update link
link updated , added
https://0x0.st/8C1p.txt

4

Here it is

août 12 07:15:10 endv kernel: amdgpu 0000:03:00.0: amdgpu: [gfxhub] page fault (src_id:0 ring:24 vmid:1 pasid:32772)
août 12 07:15:10 endv kernel: amdgpu 0000:03:00.0: amdgpu:  in process Xorg pid 1032 thread Xorg:cs0 pid 1033)
août 12 07:15:10 endv kernel: amdgpu 0000:03:00.0: amdgpu:   in page starting at address 0x0000800106c10000 from client 10
août 12 07:15:10 endv kernel: amdgpu 0000:03:00.0: amdgpu: GCVM_L2_PROTECTION_FAULT_STATUS:0x00101031
août 12 07:15:10 endv kernel: amdgpu 0000:03:00.0: amdgpu:          Faulty UTCL2 client ID: TCP (0x8)
août 12 07:15:10 endv kernel: amdgpu 0000:03:00.0: amdgpu:          MORE_FAULTS: 0x1
août 12 07:15:10 endv kernel: amdgpu 0000:03:00.0: amdgpu:          WALKER_ERROR: 0x0
août 12 07:15:10 endv kernel: amdgpu 0000:03:00.0: amdgpu:          PERMISSION_FAULTS: 0x3
août 12 07:15:10 endv kernel: amdgpu 0000:03:00.0: amdgpu:          MAPPING_ERROR: 0x0
août 12 07:15:10 endv kernel: amdgpu 0000:03:00.0: amdgpu:          RW: 0x0
août 12 07:15:10 endv kernel: amdgpu 0000:03:00.0: amdgpu: Ring gfx_0.0.0 reset succeeded

I’ve found this same issue here https://gitlab.freedesktop.org/drm/amd/-/issues/3067

The suggested fix is to add amdgpu.ppfeaturemask=0xf7fff to kernel params

5

i will check jake , i have an another clue , that can explain olders errors :
this motherboard is using PCI video card and 4 Nvme ,
i have checked bus speed on boot linux this is not x16 , not x8 , not x4 but x1 ( i dont know why also iGpu part is also on x1 )

may be i have to force all in x4 x4 x4 x4 Pci , unless one nvme is only x 1

update:
the auto mode for “bifurcation PCI Xhci” was long and try to give more in bandwidth for Gpu pci with 4 nmve , the result was lag on boot , and lost display DP from many points ( drrm , amdgpu , kernel , xorg ) just during boot ( or reboot ) or after login.

if you see some nvme forcing to suspend in boot , this is not good.

now with better balance bandwith pci , boot are really faster with less errors.
recent changes , update kernel 6.16.1 ,amd firmware and xf86-video-amdgpu for this recent video card amd 9070 XT brings greater stability.

6

I see it has 1 pci-e 5.0 x16 slot, 1 pci-e 4.0 x 4 slot, 1 pci-e 4.0 x4 slot. The SATA ports are 3 pci-e 5.0 m.2 and 1 pci-e 4.0 m.2

Not sure why it would be 1x? I assume you are not running RAID and is set to AHCI? I agree with @jake99 that the kernel parameter may be the answer.

7

This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.