[SOLVED] amdgpu(?) problem on HP ProBook 455 G7

frenex · 2020-11-19 11:08:10

Hello,
I have HP ProBook 455 G7 (it has ryzen 5 pro 4650u with integrated vega graphics) and after systemd messages, screen just goes blank. It works with nomodeset and, several times, graphics were loaded correctly without it and I could start X11, but I couldn't figure out what was different those times. Also, perhaps it is worth mentioning, that even though the screen is blank, I can still log in and, for example, reboot the machine. I tried using different kernel parameters, but none I tried made it work. I have updated linux to 5.9.9 and it seems to be loading correctly more frequently (in about 4/10 cases).
Here are journalctl entries related to amdgpu, when graphics don't load:

 Nov 20 02:36:52 HP-ProBook kernel: [drm] amdgpu kernel modesetting enabled. 
Nov 20 02:36:52 HP-ProBook kernel: amdgpu: Topology: Add CPU node 
Nov 20 02:36:52 HP-ProBook kernel: fb0: switching to amdgpudrmfb from EFI VGA 
Nov 20 02:36:52 HP-ProBook kernel: amdgpu 0000:05:00.0: vgaarb: deactivate vga console 
Nov 20 02:36:52 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: Trusted Memory Zone (TMZ) feature disabled as experimental (default) 
Nov 20 02:36:52 HP-ProBook kernel: amdgpu 0000:05:00.0: Invalid PCI ROM header signature: expecting 0xaa55, got 0xffff 
Nov 20 02:36:52 HP-ProBook kernel: amdgpu: ATOM BIOS: SWBRT58350.001 
Nov 20 02:36:52 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: VRAM: 512M 0x000000F400000000 - 0x000000F41FFFFFFF (512M used) 
Nov 20 02:36:52 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: GART: 1024M 0x0000000000000000 - 0x000000003FFFFFFF 
Nov 20 02:36:52 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: AGP: 267419648M 0x000000F800000000 - 0x0000FFFFFFFFFFFF 
Nov 20 02:36:52 HP-ProBook kernel: [drm] amdgpu: 512M of VRAM memory ready 
Nov 20 02:36:52 HP-ProBook kernel: [drm] amdgpu: 3072M of GTT memory ready. 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: SMU is initialized successfully! 
Nov 20 02:36:53 HP-ProBook kernel: [drm:amdgpu_dm_init.isra.0.cold [amdgpu]] *ERROR* amdgpu: failed to initialize hdcp_workqueue. 
Nov 20 02:36:53 HP-ProBook kernel: snd_hda_intel 0000:05:00.1: bound 0000:05:00.0 (ops amdgpu_dm_audio_component_bind_ops [amdgpu]) 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu: Topology: Add dGPU node [0x1636:0x1002] 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: SE 1, SH per SE 2, CU per SH 18, active_cu_number 26 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring gfx uses VM inv eng 0 on hub 0 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 1 on hub 0 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 4 on hub 0 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 5 on hub 0 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 6 on hub 0 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 7 on hub 0 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 8 on hub 0 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 9 on hub 0 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 10 on hub 0 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring kiq_2.1.0 uses VM inv eng 11 on hub 0 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring sdma0 uses VM inv eng 0 on hub 1 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring vcn_dec uses VM inv eng 1 on hub 1 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring vcn_enc0 uses VM inv eng 4 on hub 1 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring vcn_enc1 uses VM inv eng 5 on hub 1 
Nov 20 02:36:53 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring jpeg_dec uses VM inv eng 6 on hub 1 
Nov 20 02:36:53 HP-ProBook kernel: [drm] Initialized amdgpu 3.39.0 20150101 for 0000:05:00.0 on minor 0

And here are the logs for when graphics do load correctly:

Nov 20 02:28:33 HP-ProBook kernel: [drm] amdgpu kernel modesetting enabled. 
Nov 20 02:28:33 HP-ProBook kernel: amdgpu: Topology: Add CPU node 
Nov 20 02:28:33 HP-ProBook kernel: fb0: switching to amdgpudrmfb from EFI VGA 
Nov 20 02:28:33 HP-ProBook kernel: amdgpu 0000:05:00.0: vgaarb: deactivate vga console 
Nov 20 02:28:33 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: Trusted Memory Zone (TMZ) feature disabled as experimental (default) 
Nov 20 02:28:33 HP-ProBook kernel: amdgpu: ATOM BIOS: SWBRT58350.001 
Nov 20 02:28:33 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: VRAM: 512M 0x000000F400000000 - 0x000000F41FFFFFFF (512M used) 
Nov 20 02:28:33 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: GART: 1024M 0x0000000000000000 - 0x000000003FFFFFFF 
Nov 20 02:28:33 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: AGP: 267419648M 0x000000F800000000 - 0x0000FFFFFFFFFFFF 
Nov 20 02:28:33 HP-ProBook kernel: [drm] amdgpu: 512M of VRAM memory ready 
Nov 20 02:28:33 HP-ProBook kernel: [drm] amdgpu: 3072M of GTT memory ready. 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: SMU is initialized successfully! 
Nov 20 02:28:34 HP-ProBook kernel: snd_hda_intel 0000:05:00.1: bound 0000:05:00.0 (ops amdgpu_dm_audio_component_bind_ops [amdgpu]) 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu: Topology: Add dGPU node [0x1636:0x1002] 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: SE 1, SH per SE 2, CU per SH 18, active_cu_number 26 
Nov 20 02:28:34 HP-ProBook kernel: fbcon: amdgpudrmfb (fb0) is primary device 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: [drm] fb0: amdgpudrmfb frame buffer device 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring gfx uses VM inv eng 0 on hub 0 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 1 on hub 0 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 4 on hub 0 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 5 on hub 0 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 6 on hub 0 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 7 on hub 0 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 8 on hub 0 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 9 on hub 0 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 10 on hub 0 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring kiq_2.1.0 uses VM inv eng 11 on hub 0 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring sdma0 uses VM inv eng 0 on hub 1 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring vcn_dec uses VM inv eng 1 on hub 1 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring vcn_enc0 uses VM inv eng 4 on hub 1 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring vcn_enc1 uses VM inv eng 5 on hub 1 
Nov 20 02:28:34 HP-ProBook kernel: amdgpu 0000:05:00.0: amdgpu: ring jpeg_dec uses VM inv eng 6 on hub 1 
Nov 20 02:28:34 HP-ProBook kernel: [drm] Initialized amdgpu 3.39.0 20150101 for 0000:05:00.0 on minor 0 
Nov 20 02:28:34 HP-ProBook systemd[1]: Starting Load/Save Screen Backlight Brightness of backlight:amdgpu_bl0... 
Nov 20 02:28:34 HP-ProBook systemd[1]: Finished Load/Save Screen Backlight Brightness of backlight:amdgpu_bl0. 
Nov 20 02:28:34 HP-ProBook audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 msg='unit=systemd-backlight@backlight:amdgpu_bl0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'

So, as you can see, these lines are different:

Nov 20 02:36:52 HP-ProBook kernel: amdgpu 0000:05:00.0: Invalid PCI ROM header signature: expecting 0xaa55, got 0xffff
Nov 20 02:36:53 HP-ProBook kernel: [drm:amdgpu_dm_init.isra.0.cold [amdgpu]] *ERROR* amdgpu: failed to initialize hdcp_workqueue.

Can anything be done about that? Thanks.

Last edited by frenex (2020-11-27 12:15:27)

Glaeqen · 2020-11-23 22:02:44

The same laptop. Ended up here after some digging. With `nomodeset` enabled, it at least starts but limited to 1024x768 which is unusable. I have the very same problem. Not sure what causes it; at first glance it looks like `amdgpu` in fact. PCI ROM header is most likely negligible but the other one? Not sure. Timing problem in a driver?

Last edited by Glaeqen (2020-11-23 22:03:12)

Lone_Wolf · 2020-11-24 14:03:45

Are you running Latest fimware ?

If yes, we'll need the full log of a boot without nomodeset .

journalctl -b -somenumber can be used to retrieve logs from previous boots.

Glaeqen · 2020-11-25 07:58:12

Lone_Wolf wrote:

Are you running Latest fimware ?

Link you provided leads to "We were unable to find drivers for your product." page. What do you mean by firmware? BIOS? Shouldn't everything be provided either in a kernel or as a Arch package?

If yes, we'll need the full log of a boot without nomodeset.

Not sure about the firmware thing but there goes the gist. I provided "good" boot and "bad" boot in separate files.

In "bad" file on line 1279 you can find

kernel: [drm:dm_hw_init.cold [amdgpu]] *ERROR* amdgpu: failed to initialize hdcp_workqueue.

Lone_Wolf · 2020-11-25 14:03:06

Glaeqen wrote:

Link you provided leads to "We were unable to find drivers for your product." page. What do you mean by firmware? BIOS? Shouldn't everything be provided either in a kernel or as a Arch package?

HP site detects you use linux and they don't have drivers for linux. Just choose windows 10 64-bit as OS.

Firmware is what hardware devices system uses to initialise / boot up. it's often called BIOS / UEFI / EFI but those are specific implementations of firmware for motherboards.

Normally the manufacturer provides and updates it, some manufacturers do support using fwupd to update firmware.

There are some error messages in the bad boot log, but they don't appear to be critical.

Nov 24 00:15:04 hp-probook-445-g7 systemd[1]: Received SIGINT.

That is definitely a bad thing and suggest low-level issues (like firmware).

The amdgpu kernel module is inialised rather late, try configuring early KMS .

You are using zen kernel, have you tried with the stock linux kernel ?

Last edited by Lone_Wolf (2020-11-25 14:03:31)

frenex · 2020-11-25 14:22:28

I did update BIOS after I got the laptop (it has an option for a network update), but It didn't help. I also tried configuring early KMS, but it did nothing as well, instead, systemd messages weren't shown at all.

Last edited by frenex (2020-11-25 16:27:41)

mcloaked · 2020-11-25 18:00:21

I have a recent HP Envy X360 15-0000na and when I first installed arch I had problems with the boot - on my machine the one change that made a big difference was adding a parameter for acpi_backlight=vendor to the kernel line for boot, as per

rw root=UUID=xxxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx PARTUUID=xxxxxxxxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxx initrd=boot\amd-ucode.img acpi_backlight=vendor quiet initrd=\boot\initramfs-linux.img

I don't know if you have tried that but may be worth seeing if that makes any difference?

frenex · 2020-11-25 18:20:49

mcloaked wrote:

I have a recent HP Envy X360 15-0000na and when I first installed arch I had problems with the boot - on my machine the one change that made a big difference was adding a parameter for acpi_backlight=vendor to the kernel line for boot

I tried that but, unfortunately, the problem still persists.

Glaeqen · 2020-11-25 18:23:14

frenex wrote:

mcloaked wrote:
I have a recent HP Envy X360 15-0000na and when I first installed arch I had problems with the boot - on my machine the one change that made a big difference was adding a parameter for acpi_backlight=vendor to the kernel line for boot
I tried that but, unfortunately, the problem still persists.

Same.

icar · 2020-11-25 18:30:15

acpi_backlight=video is how I'm running my 3700U. Maybe you guys could try that.
Also, in /etc/mkinitcpio.conf put in MODULES=(amdgpu) for early load. (remember to rebuild them afterwards)

Last edited by icar (2020-11-25 18:34:04)

frenex · 2020-11-25 18:41:08

icar wrote:

acpi_backlight=video is how I'm running my 3700U. Maybe you guys could try that.
Also, in /etc/mkinitcpio.conf put in MODULES=(amdgpu) for early load. (remember to rebuild them afterwards)

Doesn't work. And, as I mentioned earlier, early KMS just makes screen go black right after bootloader.

Glaeqen · 2020-11-26 21:46:39

frenex wrote:

icar wrote:
acpi_backlight=video is how I'm running my 3700U. Maybe you guys could try that.
Also, in /etc/mkinitcpio.conf put in MODULES=(amdgpu) for early load. (remember to rebuild them afterwards)
Doesn't work. And, as I mentioned earlier, early KMS just makes screen go black right after bootloader.

Same. In every configuration, with every kernel variant behaviour is exactly the same.

So, where is the best place to report the bug? Someone mentioned earlier that SIGINT received by systemd looks concerning but I'm not really sure if that's the root of the issue. Amdgpu looks at fault but, again, who knows.

icar · 2020-11-26 23:04:07

amdgpu.gpu_recovery=1 could help, although it might probably be a iommu problem that could be worked around with some kernel parameters. I remember reading something about this in Reddit. I've been trying to find it with no luck. It was something like iommu=pt or pm or similar.

Either way, this could also be a problem with the UEFI settings. Make sure you have Legacy Mode / CSM explicitly disabled and Secure Boot off.

I've also found amdgpu.dpm=1 (dynamic power management) to try, although I have no experience with that.

There are also people on other forums talking about RAM timings, try to see if you can change them (lower probably).

frenex · 2020-11-27 12:14:47

icar wrote:

amdgpu.gpu_recovery=1 could help, although it might probably be a iommu problem that could be worked around with some kernel parameters. I remember reading something about this in Reddit. I've been trying to find it with no luck. It was something like iommu=pt or pm or similar.
Either way, this could also be a problem with the UEFI settings. Make sure you have Legacy Mode / CSM explicitly disabled and Secure Boot off.
I've also found amdgpu.dpm=1 (dynamic power management) to try, although I have no experience with that.
There are also people on other forums talking about RAM timings, try to see if you can change them (lower probably).

Wow! Turning Legacy Boot off did the trick! On my desktop PC I have it turned on and it works perfectly, so I thought it doesn't matter - I just checked that Secure Boot is turned off - but it does. Thanks a lot for the hint!

Last edited by frenex (2020-11-27 12:16:04)

Lone_Wolf · 2020-11-27 13:06:35

On atleast some firmwares allowing Legacy Boot / CSM enable settings that block functionality needed for 64-bit linux.
(No idea if windows 64-bit can work with legacy boot / CSM active).

Does your desktop have an intel processor ?

Last edited by Lone_Wolf (2020-11-27 13:07:01)

frenex · 2020-11-27 13:30:53

Lone_Wolf wrote:

On atleast some firmwares allowing Legacy Boot / CSM enable settings that block functionality needed for 64-bit linux.
(No idea if windows 64-bit can work with legacy boot / CSM active).
Does your desktop have an intel processor ?

It does have an Intel processor.

And, btw, I installed windows 10 on this laptop to check if it would work, when I first encountered that problem with Arch, and it did.

Last edited by frenex (2020-11-27 13:32:05)

Arch Linux

#1 2020-11-19 11:08:10

[SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#2 2020-11-23 22:02:44

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#3 2020-11-24 14:03:45

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#4 2020-11-25 07:58:12

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#5 2020-11-25 14:03:06

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#6 2020-11-25 14:22:28

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#7 2020-11-25 18:00:21

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#8 2020-11-25 18:20:49

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#9 2020-11-25 18:23:14

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#10 2020-11-25 18:30:15

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#11 2020-11-25 18:41:08

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#12 2020-11-26 21:46:39

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#13 2020-11-26 23:04:07

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#14 2020-11-27 12:14:47

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#15 2020-11-27 13:06:35

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

#16 2020-11-27 13:30:53

Re: [SOLVED] amdgpu(?) problem on HP ProBook 455 G7

Board footer