| View previous topic :: View next topic |
| Author |
Message |
Helsy7
n00b
Joined: 17 Nov 2017
Posts: 7
|
 Posted: Sun Feb 25, 2018 3:12 am Post subject: Kernel panic when I try to unload amdgpu kernel module |
 |
|
Hi, I cannot unload the amdgpu kernel module.
When I try this the terminal or tty hangs and I have to open another tty or terminal.
The AMD graphics card is a RX480 8GB Reference with a modified BIOS from a RX580 8GB Reference.
I used the first BIOS from this thread: http://www.overclock.net/forum/67-amd-ati/1634872-bios-mod-rx480-rx580-conversions-how-rx470-rx480-rx570-rx580.html
In the first PCIe Slot of my mainboard ASUS B350 PRIME is a Nvidia GT 630 and this card is used for boot splash screen / grub / tty / X server.
I'm able to blacklist the amdgpu module but that does not solve my problem because I want to be able to load and unload the amdgpu module.
I do not need to start the X server to get this problem.
please help me with this problem
Hardware:
CPU: AMD 1800X
GPU: RX480
Mainboard: ASUS B350 PRIME
RAM: 2x 8GB G.SKILL TridentZ 3200Mhz CL14
No Overclocking
Software:
Kernel: linux-4.15.5-gentoo
Profile: amd64/17.0/desktop
interesting part of dmesg:
| Code: | [ 54.665693] [drm] amdgpu: finishing device.
[ 54.665704] WARNING: CPU: 11 PID: 3506 at drivers/gpu/drm/drm_crtc.c:108 drm_crtc_force_disable+0x3a/0x50
[ 54.665704] Modules linked in: amdkfd amdgpu(-) nouveau mfd_core chash ttm efivarfs
[ 54.665709] CPU: 11 PID: 3506 Comm: rmmod Not tainted 4.15.5-gentoo #1
[ 54.665709] Hardware name: System manufacturer System Product Name/PRIME B350-PLUS, BIOS 3805 02/05/2018
[ 54.665711] RIP: 0010:drm_crtc_force_disable+0x3a/0x50
[ 54.665711] RSP: 0018:ffffb22fc1f57d68 EFLAGS: 00010286
[ 54.665712] RAX: ffffffffc04589e0 RBX: ffff9e4e5c8f9000 RCX: 0000000000000000
[ 54.665713] RDX: ffffb22fc1f57d68 RSI: ffff9e4e5c8f9000 RDI: ffffb22fc1f57d98
[ 54.665713] RBP: ffff9e4e5b378b68 R08: ffff9e4e6eee3880 R09: ffff9e4e5c6dc9c0
[ 54.665714] R10: ffffb22fc1f57ce0 R11: 0000000000000001 R12: ffff9e4e5b378800
[ 54.665714] R13: ffffffffc04e72d0 R14: ffff9e4e5bc81900 R15: 0000000000000000
[ 54.665715] FS: 00007fbeab936740(0000) GS:ffff9e4e6eec0000(0000) knlGS:0000000000000000
[ 54.665716] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 54.665716] CR2: 00007fbeab190bf0 CR3: 0000000416980000 CR4: 00000000003406e0
[ 54.665716] Call Trace:
[ 54.665719] drm_crtc_force_disable_all+0x49/0x70
[ 54.665729] amdgpu_device_fini+0x1ac/0x1c0 [amdgpu]
[ 54.665737] amdgpu_driver_unload_kms+0x45/0x90 [amdgpu]
[ 54.665738] drm_dev_unregister+0x3a/0xe0
[ 54.665745] amdgpu_pci_remove+0x14/0x40 [amdgpu]
[ 54.665747] pci_device_remove+0x36/0xb0
[ 54.665749] device_release_driver_internal+0x155/0x220
[ 54.665750] driver_detach+0x2d/0x60
[ 54.665751] bus_remove_driver+0x47/0xa0
[ 54.665752] pci_unregister_driver+0x1d/0x90
[ 54.665764] amdgpu_exit+0x11/0x3e1 [amdgpu]
[ 54.665766] SyS_delete_module+0x166/0x220
[ 54.665768] do_syscall_64+0x5b/0x100
[ 54.665770] entry_SYSCALL_64_after_hwframe+0x3d/0xa2 |
full dmesg:
https://pastebin.com/NvNrr8uh
emerge --info --verbose
https://pastebin.com/HCdDufa2
kernel .config
https://pastebin.com/yKkPtepn |
|
| Back to top |
|
 |
The_Document
Apprentice
Joined: 03 Feb 2018
Posts: 275
|
| Posted: Sun Feb 25, 2018 8:11 am Post subject: Re: Kernel panic when I try to unload amdgpu kernel module |
|
|
| I haven't built gpu drivers as modules simply because it seems absurd to do so. The amdgpu driver is relatively new and might not have support for module unloading. Also what do you issue to unload it? If Im not mistaken it should claim to be in use and not permit you to unload it. |
|
| Back to top |
|
|
|
Kernel & Hardware
