You are not logged in.

#1 2025-05-28 12:47:32

msix29
Member
Registered: 2025-05-28
Posts: 6

(old) AMD GPU failing to load

It works when `radeon.si_support=1`, but I want vulkan support, which seems to only work with `amdgpu`. I tried both `cik` and `si`. For `si`, running `sudo dmesg | grep -i 'amdgpu.*firmware'`, I get:

[    7.666802] [drm:uvd_v3_1_hw_init [amdgpu]] *ERROR* amdgpu: UVD Firmware validate fail (-22).

And when using `si_support=1`, for

sudo dmesg | grep -i amd | grep -e error
journalctl -b | grep -i amdgpu | grep -e error

I get

[    7.667660] amdgpu 0000:01:00.0: amdgpu: Fatal error during GPU init
[    7.874361] amdgpu 0000:01:00.0: probe with driver amdgpu failed with error -22
May 28 14:54:32 Arch kernel: amdgpu 0000:01:00.0: amdgpu: Fatal error during GPU init
May 28 14:54:32 Arch kernel: amdgpu 0000:01:00.0: probe with driver amdgpu failed with error -22

Using `radeon.si_support=1` works but no Vulkan - which is mainly what I need.

Offline

#2 2025-05-28 13:06:24

Klara
Member
Registered: 2025-05-20
Posts: 14

Re: (old) AMD GPU failing to load

What AMD GPU are you using specifically? Can you show us the contents of lspci?

Offline

#3 2025-05-28 19:15:40

msix29
Member
Registered: 2025-05-28
Posts: 6

Re: (old) AMD GPU failing to load

❯ lspci -k | grep -A 3 -i vga
00:02.0 VGA compatible controller: Intel Corporation HD Graphics 530 (rev 06)
    DeviceName: Onboard IGD
    Subsystem: Hewlett-Packard Company Device 80d5
    Kernel driver in use: i915
--
01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Venus XTX [Radeon HD 8890M / R9 M275X/M375X] (rev 83)
    Subsystem: Hewlett-Packard Company Device 80d5
    Kernel modules: radeon, amdgpu
01:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Oland/Hainan/Cape Verde/Pitcairn HDMI Audio [Radeon HD 7000 Series]

The first is my iGPU - supports vulkan but ofc not the desirable performance.

Offline

#4 2025-05-28 20:35:41

cryptearth
Member
Registered: 2024-02-03
Posts: 1,447

Re: (old) AMD GPU failing to load

so this is a laptop - exact model would be helpful
according to https://www.techpowerup.com/gpu-specs/amd-venus.g135 that gpu seems to support vulkan
my guess: it's one of those old chips which on paper were planed to support features but only properly work with windows because linux support never got that far
my laptop is the same: its gt710m is supposed to be supported by nvid-390 - but even when tinker around to use it nvid chip only something doesn't properly work anyway - so I just use win10 because the drivers installed by win update after install do work properly
my advice: you can try to tinker with your system and MAY find some specific setup it MAY work - or you can save yourself getting gray hair and just use win7 or win10 and the proper windows drivers
it's just "this hardware from an era when linux driver development weren't up to what they are today"

Offline

#5 2025-05-28 22:20:28

msix29
Member
Registered: 2025-05-28
Posts: 6

Re: (old) AMD GPU failing to load

It's the HP Zbook 15 G3. And yeah, it does support vulkan on Windows 10 (which I used to main, before it fully nuked itself, lol). By "use the proper windows drivers", you mean switching back to Windows, right? I tried various setups and none worked, though if I found one I'll be sure to send it here.

Offline

#6 2025-05-29 12:11:35

Lone_Wolf
Administrator
From: Netherlands, Europe
Registered: 2005-10-04
Posts: 13,708

Re: (old) AMD GPU failing to load

That gpu seems to be GCN 1 which should work with amdgpu si_support .

Since this looks like a intel + amd hybrid setup , you may have to tweak the modules in the initramfs to ensure things are initialised in the correct order.

Please post the contents your /etc/mkinitcpio.conf .


Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.

clean chroot building not flexible enough ?
Try clean chroot manager by graysky

Online

#7 2025-05-29 13:13:17

msix29
Member
Registered: 2025-05-28
Posts: 6

Re: (old) AMD GPU failing to load

My /etc/mkinitcpio.conf is unchanged, only:

MODULES=(amdgpu)

But here:

# vim:set ft=sh
# MODULES
# The following modules are loaded before any boot hooks are
# run.  Advanced users may wish to specify all system modules
# in this array.  For instance:
#     MODULES=(usbhid xhci_hcd)
MODULES=(amdgpu)

# BINARIES
# This setting includes any additional binaries a given user may
# wish into the CPIO image.  This is run last, so it may be used to
# override the actual binaries included by a given hook
# BINARIES are dependency parsed, so you may safely ignore libraries
BINARIES=()

# FILES
# This setting is similar to BINARIES above, however, files are added
# as-is and are not parsed in any way.  This is useful for config files.
FILES=()

# HOOKS
# This is the most important setting in this file.  The HOOKS control the
# modules and scripts added to the image, and what happens at boot time.
# Order is important, and it is recommended that you do not change the
# order in which HOOKS are added.  Run 'mkinitcpio -H <hook name>' for
# help on a given hook.
# 'base' is _required_ unless you know precisely what you are doing.
# 'udev' is _required_ in order to automatically load modules
# 'filesystems' is _required_ unless you specify your fs modules in MODULES
# Examples:
##   This setup specifies all modules in the MODULES setting above.
##   No RAID, lvm2, or encrypted root is needed.
#    HOOKS=(base)
#
##   This setup will autodetect all modules for your system and should
##   work as a sane default
#    HOOKS=(base udev autodetect modconf block filesystems fsck)
#
##   This setup will generate a 'full' image which supports most systems.
##   No autodetection is done.
#    HOOKS=(base udev modconf block filesystems fsck)
#
##   This setup assembles a mdadm array with an encrypted root file system.
##   Note: See 'mkinitcpio -H mdadm_udev' for more information on RAID devices.
#    HOOKS=(base udev modconf keyboard keymap consolefont block mdadm_udev encrypt filesystems fsck)
#
##   This setup loads an lvm2 volume group.
#    HOOKS=(base udev modconf block lvm2 filesystems fsck)
#
##   This will create a systemd based initramfs which loads an encrypted root filesystem.
#    HOOKS=(base systemd autodetect modconf kms keyboard sd-vconsole sd-encrypt block filesystems fsck)
#
##   NOTE: If you have /usr on a separate partition, you MUST include the
#    usr and fsck hooks.
HOOKS=(base udev autodetect microcode modconf kms keyboard keymap consolefont block filesystems fsck)

# COMPRESSION
# Use this to compress the initramfs image. By default, zstd compression
# is used for Linux ≥ 5.9 and gzip compression is used for Linux < 5.9.
# Use 'cat' to create an uncompressed image.
#COMPRESSION="zstd"
#COMPRESSION="gzip"
#COMPRESSION="bzip2"
#COMPRESSION="lzma"
#COMPRESSION="xz"
#COMPRESSION="lzop"
#COMPRESSION="lz4"

# COMPRESSION_OPTIONS
# Additional options for the compressor
#COMPRESSION_OPTIONS=()

# MODULES_DECOMPRESS
# Decompress loadable kernel modules and their firmware during initramfs
# creation. Switch (yes/no).
# Enable to allow further decreasing image size when using high compression
# (e.g. xz -9e or zstd --long --ultra -22) at the expense of increased RAM usage
# at early boot.
# Note that any compressed files will be placed in the uncompressed early CPIO
# to avoid double compression.
#MODULES_DECOMPRESS="no"

I think this might be needed, it's in /etc/default/grub

GRUB_CMDLINE_LINUX_DEFAULT="loglevel=3 quiet radeon.si_support=0 amdgpu.si_support=1 amdgpu.runpm=0"

This produces the init error I sent above (here it is again):

❯ journalctl -b | grep -i amdgpu | grep -e error
May 28 14:54:32 Arch kernel: amdgpu 0000:01:00.0: amdgpu: Fatal error during GPU init
May 28 14:54:32 Arch kernel: amdgpu 0000:01:00.0: probe with driver amdgpu failed with error -22

Just found out how to use the code blocks! Thought it would be backticks, lol.

Last edited by msix29 (2025-05-29 13:14:51)

Offline

#8 2025-05-29 13:26:07

Lone_Wolf
Administrator
From: Netherlands, Europe
Registered: 2005-10-04
Posts: 13,708

Re: (old) AMD GPU failing to load

On many laptops (and desktop hybrid graphics) the integrated card needs to be initialised fully before any dedicated card.


In /etc/mkinitcpio.conf
Remove kms from the HOOKS line
add i915 to the MODULES line
remove amdgpu from the MODULES line

This should help to get the integrated intel card the time to initialise before the amd card starts initialising .
Note that graphics initialisation takes time and Display Managers often display stuff before graphics initialisation has finished .
to be sure you're not bitten by this, give the laptop 2 minutes or so to allow things to settle after booting.

Then login and run as root / with root rights

# journalctl -b | curl -F 'file=@-' 0x0.st

Post the link it outputs.


Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.

clean chroot building not flexible enough ?
Try clean chroot manager by graysky

Online

#9 2025-05-29 16:07:25

msix29
Member
Registered: 2025-05-28
Posts: 6

Re: (old) AMD GPU failing to load

Did it all, gave it more than 15 minutes. Still same error.

http://0x0.st/83At.txt

Offline

#10 2025-05-29 16:50:08

cryptearth
Member
Registered: 2024-02-03
Posts: 1,447

Re: (old) AMD GPU failing to load

msix29 wrote:

By "use the proper windows drivers", you mean switching back to Windows, right?

that's pretty much what I suggest, yes
windows should work out required drivers by itself by just letting driver-autoinstall-from-win-update do its job - but HP should offer proper maybe more update-to-date drivers
point is: this device was designed with windows in mind - likely has a Win8 badge on it - and it's from a time when proper driver development on linux just had started - so most of the hardware from that era got only up to a state: "it properly generates a stable image - job done" - noone really put effort into getting 3D working proper - and even if main focus was on openGL only with vulkan still years in the future
maybe you'll be able to get it working somehow with what exist on the internet - maybe you have to get your hands dirty and tinker with the drivers yourself - and I wish you the best luck for that - but even if you get there: performance will be bad anyway - and dealing with all that just for the sake of getting it done?
I gave up on mine (let alone since my cat dropped my laptop and killed the screen) - I just try to save you from a lot of headaches as honestly I have the feeling you likely end there as well (unless you get lucky and get it to work just by the fact because its amd instead of nvid)

sorry I can't help here - good luck - please keep us updated if you manage to get it working

Offline

#11 2025-05-29 16:57:21

Lone_Wolf
Administrator
From: Netherlands, Europe
Registered: 2005-10-04
Posts: 13,708

Re: (old) AMD GPU failing to load

May 29 18:49:44 Arch kernel: [drm] amdgpu kernel modesetting enabled.
May 29 18:49:44 Arch kernel: amdgpu: vga_switcheroo: detected switching method \_SB_.PCI0.GFX0.ATPX handle
May 29 18:49:44 Arch kernel: amdgpu: ATPX version 1, functions 0x00000003
May 29 18:49:44 Arch kernel: amdgpu: ATPX Hybrid Graphics
May 29 18:49:44 Arch kernel: amdgpu: Virtual CRAT table created for CPU
May 29 18:49:44 Arch kernel: amdgpu: Topology: Add CPU node
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: enabling device (0006 -> 0007)
May 29 18:49:44 Arch kernel: [drm] initializing kernel modesetting (VERDE 0x1002:0x6820 0x103C:0x80D5 0x83).
May 29 18:49:44 Arch kernel: [drm] register mmio base: 0xDC000000
May 29 18:49:44 Arch kernel: [drm] register mmio size: 262144
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: detected ip block number 0 <si_common>
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: detected ip block number 1 <gmc_v6_0>
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: detected ip block number 2 <si_ih>
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: detected ip block number 3 <gfx_v6_0>
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: detected ip block number 4 <si_dma>
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: detected ip block number 5 <si_dpm>
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: detected ip block number 6 <dce_v6_0>
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: detected ip block number 7 <uvd_v3_1>
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: Fetched VBIOS from ATRM
May 29 18:49:44 Arch kernel: amdgpu: ATOM BIOS: BR46525.001
May 29 18:49:44 Arch kernel: kfd kfd: amdgpu: VERDE  not supported in kfd
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: Trusted Memory Zone (TMZ) feature not supported
May 29 18:49:44 Arch kernel: [drm] GPU posting now...
May 29 18:49:44 Arch kernel: [drm] PCIE gen 3 link speeds already enabled
May 29 18:49:44 Arch kernel: [drm] vm size is 64 GB, 2 levels, block size is 10-bit, fragment size is 9-bit
May 29 18:49:44 Arch kernel: Adding 4194300k swap on /dev/sda4.  Priority:-2 extents:1 across:4194300k 
May 29 18:49:44 Arch kernel: videodev: Linux video capture interface: v2.00
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: VRAM: 2048M 0x000000F400000000 - 0x000000F47FFFFFFF (2048M used)
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: GART: 1024M 0x000000FF00000000 - 0x000000FF3FFFFFFF
May 29 18:49:44 Arch kernel: [drm] Detected VRAM RAM=2048M, BAR=256M
May 29 18:49:44 Arch kernel: [drm] RAM width 128bits GDDR5
May 29 18:49:44 Arch kernel: [drm] amdgpu: 2048M of VRAM memory ready
May 29 18:49:44 Arch kernel: [drm] amdgpu: 7934M of GTT memory ready.
May 29 18:49:44 Arch kernel: [drm] GART: num cpu pages 262144, num gpu pages 262144
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: PCIE GART of 1024M enabled (table at 0x000000F400000000).
May 29 18:49:44 Arch kernel: FAT-fs (sda3): Volume was not properly unmounted. Some data may be corrupt. Please run fsck.
May 29 18:49:44 Arch kernel: iwlwifi 0000:02:00.0: base HW address: 28:16:ad:38:31:8c, OTP minor version: 0x0
May 29 18:49:44 Arch kernel: ieee80211 phy0: Selected rate control algorithm 'iwl-mvm-rs'
May 29 18:49:44 Arch kernel: iwlwifi 0000:02:00.0 wlp2s0: renamed from wlan0
May 29 18:49:44 Arch kernel: [drm] Internal thermal controller with fan control
May 29 18:49:44 Arch kernel: [drm] amdgpu: dpm initialized
May 29 18:49:44 Arch kernel: [drm] amdgpu atom DIG backlight initialized
May 29 18:49:44 Arch kernel: [drm] AMDGPU Display Connectors
May 29 18:49:44 Arch kernel: [drm] Connector 0:
May 29 18:49:44 Arch kernel: [drm]   eDP-2
May 29 18:49:44 Arch kernel: [drm]   HPD3
May 29 18:49:44 Arch kernel: [drm]   DDC: 0x1954 0x1954 0x1955 0x1955 0x1956 0x1956 0x1957 0x1957
May 29 18:49:44 Arch kernel: [drm]   Encoders:
May 29 18:49:44 Arch kernel: [drm]     LCD1: INTERNAL_UNIPHY2
May 29 18:49:44 Arch kernel: [drm] Connector 1:
May 29 18:49:44 Arch kernel: [drm]   DP-4
May 29 18:49:44 Arch kernel: [drm]   HPD4
May 29 18:49:44 Arch kernel: [drm]   DDC: 0x1958 0x1958 0x1959 0x1959 0x195a 0x195a 0x195b 0x195b
May 29 18:49:44 Arch kernel: [drm]   Encoders:
May 29 18:49:44 Arch kernel: [drm]     DFP1: INTERNAL_UNIPHY2
May 29 18:49:44 Arch kernel: [drm] Connector 2:
May 29 18:49:44 Arch kernel: [drm]   DP-5
May 29 18:49:44 Arch kernel: [drm]   HPD1
May 29 18:49:44 Arch kernel: [drm]   DDC: 0x194c 0x194c 0x194d 0x194d 0x194e 0x194e 0x194f 0x194f
May 29 18:49:44 Arch kernel: [drm]   Encoders:
May 29 18:49:44 Arch kernel: [drm]     DFP2: INTERNAL_UNIPHY1
May 29 18:49:44 Arch kernel: [drm] Connector 3:
May 29 18:49:44 Arch kernel: [drm]   DP-6
May 29 18:49:44 Arch kernel: [drm]   HPD2
May 29 18:49:44 Arch kernel: [drm]   DDC: 0x1950 0x1950 0x1951 0x1951 0x1952 0x1952 0x1953 0x1953
May 29 18:49:44 Arch kernel: [drm]   Encoders:
May 29 18:49:44 Arch kernel: [drm]     DFP3: INTERNAL_UNIPHY1
May 29 18:49:44 Arch kernel: [drm] Found UVD firmware Version: 64.0 Family ID: 13
May 29 18:49:44 Arch kernel: usb 1-7: Found UVC 1.00 device HP HD Camera (05c8:0383)
May 29 18:49:44 Arch kernel: usbcore: registered new interface driver uvcvideo
May 29 18:49:44 Arch kernel: [drm:uvd_v3_1_hw_init [amdgpu]] *ERROR* amdgpu: UVD Firmware validate fail (-22).
May 29 18:49:44 Arch kernel: [drm:amdgpu_device_init.cold [amdgpu]] *ERROR* hw_init of IP block <uvd_v3_1> failed -22
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: amdgpu_device_ip_init failed
May 29 18:49:44 Arch kernel: amdgpu 0000:01:00.0: amdgpu: Fatal error during GPU init

A big part of the initialisation goes fine, but it chokes on the UVD firmware .

the amdgpu kernel module does have a parameter ip_block_mask that might help, but that's beyond my skill level.
You'll have to wait for someone with a higher kernel/hardware skill level .


Disliking systemd intensely, but not satisfied with alternatives so focusing on taming systemd.

clean chroot building not flexible enough ?
Try clean chroot manager by graysky

Online

#12 2025-05-29 17:50:00

msix29
Member
Registered: 2025-05-28
Posts: 6

Re: (old) AMD GPU failing to load

Thanks to both of you.

Offline

Board footer

Powered by FluxBB