AMD Tahiti PRO [Radeon HD 7950/8950 OEM / R9 280] Black screen during and after install ZORIN OS 16

Bourne · 14 September 2021 09:57

I wrote little up the same link with another. The one user answear it working perfectly without problems.

AMDGPU - ArchWikiand_Sea_Islands(CIK)_support
HD7950 is of a "Southern Islands" GPU family, and as such supports two different kernel drivers, an old radeon and a new amdgpu . Check the link above on how to force one or another on your card...
recently i installed the driver for 3 radeon vii on an ubuntu server and that was dead easy. i am less sure for this old card tho. thank you all.
There is guide if some place you stuck then we can try go forward.

Bourne · 14 September 2021 10:03

AMDGPU - ArchWiki here is full installation guide step by step.

GaiusMarius · 14 September 2021 10:36

Following the tuto step by step though i encountered this issue : where can i make adjustments to my kernel as they mention to do:

"When building or compiling a kernel, CONFIG_DRM_AMDGPU_SI=Y and/or CONFIG_DRM_AMDGPU_CIK=Y should be be set in the config."

Cheers

Bourne · 14 September 2021 11:03

GaiusMarius · 14 September 2021 11:11

Manjaro?

Bourne · 14 September 2021 12:01

https://bugs.freedesktop.org/show_bug.cgi?id=60879
Have I missed anything out AMDGPU vs Kernel 5.8 Issues? Pop 20.10 - Linux, macOS and Everything Not-Windows - Linus Tech Tips idk if put this here.

GaiusMarius · 14 September 2021 16:37

Created.

When i run:
lspci -v | grep driver:

I get this output:

	Kernel driver in use: pcieport
	Kernel driver in use: uhci_hcd
	Kernel driver in use: uhci_hcd
	Kernel driver in use: uhci_hcd
	Kernel driver in use: ehci-pci
	Kernel driver in use: snd_hda_intel
	Kernel driver in use: pcieport
	Kernel driver in use: pcieport
	Kernel driver in use: pcieport
	Kernel driver in use: uhci_hcd
	Kernel driver in use: uhci_hcd
	Kernel driver in use: uhci_hcd
	Kernel driver in use: ehci-pci
	Kernel driver in use: lpc_ich
	Kernel driver in use: ahci
	Kernel driver in use: snd_hda_intel
	Kernel driver in use: sky2
	Kernel driver in use: pata_marvell
	Kernel driver in use: skge
	Kernel driver in use: firewire_ohci

Dont see any amdgpu isnt this weird?

and when i run:
sudo lshw -c video

the output is:
*-display UNCLAIMED
description: VGA compatible controller
product: Tahiti PRO [Radeon HD 7950/8950 OEM / R9 280]
vendor: Advanced Micro Devices, Inc. [AMD/ATI]
physical id: 0
bus info: pci@0000:01:00.0
version: 00
width: 64 bits
clock: 33MHz
capabilities: pm pciexpress msi vga_controller cap_list
configuration: latency=0
resources: memory:d0000000-dfffffff memory:fe880000-fe8bffff ioport:b000(size=256) memory:c0000-dffff

Cheers

GaiusMarius · 14 September 2021 17:39

When i run this
dmesg | grep -i amdgpu

to check drivers loaded during boot th output is:
[ 0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-5.11.0-34-generic root=UUID=d6af0f40-f6ce-463e-b7cb-57efad91fa18 ro quiet splash modprobe.blacklist=radeon amdgpu.si_support=1 amdgpu.cik_support=0 amdgpu.dc=1 amdgpu.dpm=1
[ 0.067570] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-5.11.0-34-generic root=UUID=d6af0f40-f6ce-463e-b7cb-57efad91fa18 ro quiet splash modprobe.blacklist=radeon amdgpu.si_support=1 amdgpu.cik_support=0 amdgpu.dc=1 amdgpu.dpm=1
[ 1.443656] [drm] amdgpu kernel modesetting enabled.
[ 1.443803] amdgpu: Topology: Add CPU node
[ 1.443889] amdgpu 0000:01:00.0: vgaarb: deactivate vga console
[ 1.444813] amdgpu 0000:01:00.0: amdgpu: Trusted Memory Zone (TMZ) feature not supported
[ 1.460307] amdgpu 0000:01:00.0: amdgpu: Fetched VBIOS from ROM BAR
[ 1.460310] amdgpu: ATOM BIOS: 113-AD38900-101
[ 1.460404] amdgpu 0000:01:00.0: amdgpu: VRAM: 3072M 0x000000F400000000 - 0x000000F4BFFFFFFF (3072M used)
[ 1.460408] amdgpu 0000:01:00.0: amdgpu: GART: 1024M 0x000000FF00000000 - 0x000000FF3FFFFFFF
[ 1.460551] [drm] amdgpu: 3072M of VRAM memory ready
[ 1.460555] [drm] amdgpu: 3072M of GTT memory ready.
[ 1.461155] amdgpu 0000:01:00.0: amdgpu: PCIE GART of 1024M enabled (table at 0x000000F400900000).
[ 1.461772] [drm] amdgpu: dpm initialized
[ 1.461803] amdgpu 0000:01:00.0: Direct firmware load for amdgpu/tahiti_uvd.bin failed with error -2
[ 1.461807] amdgpu 0000:01:00.0: amdgpu: amdgpu_uvd: Can't load firmware "amdgpu/tahiti_uvd.bin"
[ 1.461813] [drm:amdgpu_device_ip_init [amdgpu]] ERROR sw_init of IP block <uvd_v3_1> failed -2
[ 1.462140] amdgpu 0000:01:00.0: amdgpu: amdgpu_device_ip_init failed
[ 1.462148] amdgpu 0000:01:00.0: amdgpu: Fatal error during GPU init
[ 1.462182] amdgpu: probe of 0000:01:00.0 failed with error -2

You'll notice the following that caught my attention:
vgaarb: deactivate vga console
Trusted Memory Zone (TMZ) feature not supported
Direct firmware load for amdgpu/tahiti_uvd.bin failed with error -2
amdgpu_uvd: Can't load firmware "amdgpu/tahiti_uvd.bin"
[drm:amdgpu_device_ip_init [amdgpu]] ERROR sw_init of IP block <uvd_v3_1> failed -2
amdgpu_device_ip_init failed
Fatal error during GPU init
robe of 0000:01:00.0 failed with error -2

Not sure if this pinpoints a problem that we can solve though..

Bourne · 14 September 2021 17:51

You created many good job if you see your graphic card now is more installed correctly.
Here some more helps from Linus Torvalds

github.com

torvalds/linux/blob/master/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c

/*
 * Copyright 2011 Advanced Micro Devices, Inc.
 * All Rights Reserved.
 *
 * Permission is hereby granted, free of charge, to any person obtaining a
 * copy of this software and associated documentation files (the
 * "Software"), to deal in the Software without restriction, including
 * without limitation the rights to use, copy, modify, merge, publish,
 * distribute, sub license, and/or sell copies of the Software, and to
 * permit persons to whom the Software is furnished to do so, subject to
 * the following conditions:
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
 * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
 * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
 * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
 * USE OR OTHER DEALINGS IN THE SOFTWARE.
 *

This file has been truncated. show original

Bourne · 14 September 2021 18:03

So this information what you post on terminal you missing some files.
https://www.phoronix.com/scan.php?page=news_item&px=AMDGPU-UVD-SI-Firmware

Here you can find firmware what you need
https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/amdgpu
You can clone git://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git and copy the missing files to /lib/firmware/amdgpu .
Here little guide on answear 17"

Bourne · 14 September 2021 18:14

Can't load firmware "amdgpu/tahiti_uvd.bin" I found what you missing on this webside then you just must go with instruction there.
Here is on ubuntu 20.04 solved problem with this.

Bourne · 14 September 2021 18:50

https://lkml.org/lkml/2021/9/13/254
here something on Trusted Memory Zone (TMZ) feature not supported

GaiusMarius · 14 September 2021 19:08

Copied all the missing firmware for amdgpu in the folder /lib/firmware/amdgpu/ then i rebooted.

When i run the command:
dmesg | grep -i amdgpu

I get the same output is seems:
[ 0.000000] Command line: BOOT_IMAGE=/boot/vmlinuz-5.11.0-34-generic root=UUID=d6af0f40-f6ce-463e-b7cb-57efad91fa18 ro quiet splash modprobe.blacklist=radeon amdgpu.si_support=1 amdgpu.cik_support=0 amdgpu.dc=1 amdgpu.dpm=1 vt.handoff=7
[ 0.067710] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-5.11.0-34-generic root=UUID=d6af0f40-f6ce-463e-b7cb-57efad91fa18 ro quiet splash modprobe.blacklist=radeon amdgpu.si_support=1 amdgpu.cik_support=0 amdgpu.dc=1 amdgpu.dpm=1 vt.handoff=7
[ 1.480049] [drm] amdgpu kernel modesetting enabled.
[ 1.480180] amdgpu: Topology: Add CPU node
[ 1.480254] fb0: switching to amdgpudrmfb from VESA VGA
[ 1.480329] amdgpu 0000:01:00.0: vgaarb: deactivate vga console
[ 1.480465] amdgpu 0000:01:00.0: amdgpu: Trusted Memory Zone (TMZ) feature not supported
[ 1.496278] amdgpu 0000:01:00.0: amdgpu: Fetched VBIOS from ROM BAR
[ 1.496281] amdgpu: ATOM BIOS: 113-AD38900-101
[ 1.496365] amdgpu 0000:01:00.0: amdgpu: VRAM: 3072M 0x000000F400000000 - 0x000000F4BFFFFFFF (3072M used)
[ 1.496368] amdgpu 0000:01:00.0: amdgpu: GART: 1024M 0x000000FF00000000 - 0x000000FF3FFFFFFF
[ 1.496514] [drm] amdgpu: 3072M of VRAM memory ready
[ 1.496517] [drm] amdgpu: 3072M of GTT memory ready.
[ 1.496977] amdgpu 0000:01:00.0: amdgpu: PCIE GART of 1024M enabled (table at 0x000000F4007E9000).
[ 1.497609] [drm] amdgpu: dpm initialized
[ 1.497633] amdgpu 0000:01:00.0: Direct firmware load for amdgpu/tahiti_uvd.bin failed with error -2
[ 1.497637] amdgpu 0000:01:00.0: amdgpu: amdgpu_uvd: Can't load firmware "amdgpu/tahiti_uvd.bin"
[ 1.497639] [drm:amdgpu_device_ip_init [amdgpu]] ERROR sw_init of IP block <uvd_v3_1> failed -2
[ 1.497917] amdgpu 0000:01:00.0: amdgpu: amdgpu_device_ip_init failed
[ 1.497922] amdgpu 0000:01:00.0: amdgpu: Fatal error during GPU init
[ 1.497949] amdgpu: probe of 0000:01:00.0 failed with error -2

I doubled checked if tahiti_uvd.bin is in the folder and it is so im not sure if this is the reason for the errors

What is vgaarb: deactivate vga console? Maybe this is the root cause for the other errors?

Cheers

Bourne · 14 September 2021 19:14

Documentation/vgaarbiter.txt [LWN.net] here is articles about vgaa what he doing.
This is little weird why firmware not update. did you used command with update and upgrade sudo apt update and upgrade?
Trusted Memory Zone Support Coming To AMD APUs in Linux Kernel 5.6 - LinuxReviews about TMZ

GaiusMarius · 14 September 2021 21:29

Do you think i can install the proprietary driver for ubuntu 20.04 on Zorin16?

https://www.amd.com/en/support/kb/release-notes/rn-amdgpu-unified-linux-20-20

GaiusMarius · 14 September 2021 23:07

Just did some other changes and now its back to black screen on boot

when i use this command to check the boot log:

journalctl --boot=efd3a93a65e44b7b8350ffbef1d430fe

I noticed highlighted in red text the following output:

kernel: kfd kfd: TAHITI  not supported in kfd
[drm] BIOS signature incorrect 20 7
...
kernel: [drm:uvd_v3_1_hw_init [amdgpu]] *ERROR* amdgpu: UVD Firmware validate fail (-22).
kernel: [drm:amdgpu_device_ip_init [amdgpu]] *ERROR* hw_init of IP block <uvd_v3_1> failed -22
kernel: amdgpu 0000:01:00.0: amdgpu: amdgpu_device_ip_init failed
kernel: amdgpu 0000:01:00.0: amdgpu: Fatal error during GPU init

any idea what kfd means? Guess we found the problem.

kfd or bios or firmware

Cheers

Aravisian · 14 September 2021 23:13

This may be interesting:

GaiusMarius · 14 September 2021 23:54

Created /etc/udev/rules.d/10-amdgpu.rules (also /etc/udev/rules.d/10-amdgpu.bak)

with this as input:
KERNEL="kfd", GROUP="video", MODE="0660"

i then ran:
sudo update-initramfs -u

then rebooted but still with the same behavior (=black screen)

GaiusMarius · 15 September 2021 00:01

The latest boot log with black screen:

kernel: kfd kfd: TAHITI not supported in kfd
kernel: [drm] BIOS signature incorrect 0 0
kernel: resource sanity check: requesting [mem 0x000c0000-0x000dffff], which spans more than PCI Bus 0000:00 [mem 0x000d0000-0x000dffff window]
kernel: caller pci_map_rom+0x71/0x1a0 mapping multiple BARs
...

kernel: [drm:uvd_v3_1_hw_init [amdgpu]] ERROR amdgpu: UVD Firmware validate fail (-110).
kernel: [drm:amdgpu_device_ip_init [amdgpu]] ERROR hw_init of IP block <uvd_v3_1> failed -110
kernel: amdgpu 0000:01:00.0: amdgpu: amdgpu_device_ip_init failed
kernel: clocksource: Switched to clocksource tsc
kernel: amdgpu 0000:01:00.0: amdgpu: Fatal error during GPU init
kernel: ------------[ cut here ]------------
kernel: kernel BUG at mm/slub.c:4124!

with some changes:
kfd the same + Bios signature still incorrect but changed + UVD Firmware new error -110...

Bourne · 15 September 2021 05:41

But this is another configuration laptop.
One what I understand on your laptop must working two graphics. One from processor and another from motherboard. If they not installed then you get a black screen. In bios you have both activated? I know on windows was the same problem one user wrote to me long time ago.