LMDE V5 and Intel gpu a780

[magifix]

I had contact this week with Intel support. It appears that the type of gpu detected is not correct, it is an Intel uhd 770 gpu and not an arc a780. The procedure recommended by Intel is based on a kernel 5.17:

https://dgpu-docs.intel.com/installatio ... y-arc.html

LMDE Elsie has a kernel 5.10 by default and 6.0 with the backport. The procedure is therefore inapplicable.
The situation is not perfect (no hw acceleration), but the resolution is correct, so I will wait until the next version.

[SMG]

I had contact this week with Intel support. It appears that the type of gpu detected is not correct, it is an Intel uhd 770 gpu and not an arc a780. The procedure recommended by Intel is based on a kernel 5.17:

https://dgpu-docs.intel.com/installatio ... y-arc.html

That is not kernel 5.17, but kernel 5.17-oem which is a specific kernel build available in the Ubuntu repositories. If you would install LM21, you could install the 5.17-oem kernel with sudo apt-get install linux-oem-22.04.

OEM Kernels are slightly different from the kernels normally used in Ubuntu/Linux Mint.

The situation is not perfect (no hw acceleration), but the resolution is correct, so I will wait until the next version.

No hardware acceleration means your CPU is doing extra work.

[magifix]

I can't thank you enough for the quality of your support. For the moment I want to stay on LMDE, the gpu performance issue is currently secondary. But I have taken note of the solution.

[magifix]

Today I got a update for the backport kernel and was hoping that it may solve the hw acceleration problem:

old: 6.0.3-1~bpo11+1 new: 6.0.12-1~bpo11+1

Unfortunatly it just added a new problem. Now at reboot it says:

A failure may be imminent and cause unpredictable fail.
Press F1 to run SETUP.

If I go to the setup and discard changes, then I am able to boot the old kernel.

How can I diagnose the problem and/or go back to the working backport?

[SMG]

Today I got a update for the backport kernel and was hoping that it may solve the hw acceleration problem:

old: 6.0.3-1~bpo11+1 new: 6.0.12-1~bpo11+1

Unfortunatly it just added a new problem. Now at reboot it says:

A failure may be imminent and cause unpredictable fail.
Press F1 to run SETUP.

The information I found indicates that is related to your hardware (hard drive) and not the operating system. Do you know if you have the latest firmware for your drive?
[SUPPORT] Please back-up your data and replace your hard disk drive. A failure may be imminent.

If I go to the setup and discard changes, then I am able to boot the old kernel.

How can I diagnose the problem and/or go back to the working backport?

What brand drive do you have? I found issues with Samsung and well as the prior link I provided for Crucial.

[magifix]

You point the problem, when looking with the disk utility, it says that one of my disks is having problems soon. It is a 1TB Samsung disk. This is not the boot disk, but it drives me to the bios. I was confuse because it appends just after updating the bpo kernel at reboot. So now let's organize a new disk.
Thank you so much, you saved my chrimas time! I wish you all the best.

[jwiz]

If you are trying to run the latest kernel you should also consider getting the lastest firmware (intel/i915) from https://git.kernel.org/pub/scm/linux/ke ... rmware.git

[magifix]

I'm not familiar with linux firmware update, should it be done with fwupdate ( fwupdmgr update)?

[magifix]

After looking around, I found that fwupd is for devices firmware, and linux firmware may be updated this way:

https://tutorialforlinux.com/2021/01/12 ... t-guide/2/

Is it safe to replace the whole directory or just the part i915?

[jwiz]

I would only replace the i915 (gfx firmware) folder to see if that make a difference.
Do a backup of the original i915 or simply rename it prior to replacing.

[magifix]

I tried to replace the i915 directory under /lib/firmware. After the startup I still have these messages, although the new files are now present:

Code: Select all

drwxr-xr-x  2 root root    4096 14 déc 13:42 i915
drwxr-xr-x  2 root root    4096 23 déc 14:18 i915.20221224

ls -l i915/tgl_guc_70.1.1.bin
-rw-r--r-- 1 root root 277440 14 déc 13:42 i915/tgl_guc_70.1.1.bin

dmesg output:

Code: Select all

[    1.404742] fb0: switching to i915 from EFI VGA
[    1.404778] Console: switching to colour dummy device 80x25
[    1.404794] i915 0000:00:02.0: vgaarb: deactivate vga console
[    1.404819] i915 0000:00:02.0: [drm] Using Transparent Hugepages
[    1.405367] i915 0000:00:02.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=io+mem:owns=io+mem
[    1.405393] i915 0000:00:02.0: firmware: direct-loading firmware i915/adls_dmc_ver2_01.bin
[    1.405774] i915 0000:00:02.0: [drm] Finished loading DMC firmware i915/adls_dmc_ver2_01.bin (v2.1)
[    1.406212] i915 0000:00:02.0: firmware: failed to load i915/tgl_guc_70.1.1.bin (-2)
[    1.406214] firmware_class: See https://wiki.debian.org/Firmware for information about missing firmware
[    1.406217] i915 0000:00:02.0: firmware: failed to load i915/tgl_guc_70.1.1.bin (-2)
[    1.406220] i915 0000:00:02.0: firmware: failed to load i915/tgl_guc_69.0.3.bin (-2)
[    1.406223] i915 0000:00:02.0: firmware: failed to load i915/tgl_guc_69.0.3.bin (-2)
[    1.406224] i915 0000:00:02.0: GuC firmware i915/tgl_guc_70.1.1.bin: fetch failed with error -2
[    1.406225] i915 0000:00:02.0: Please file a bug on drm/i915; see https://gitlab.freedesktop.org/drm/intel/-/wikis/How-to-file-i915-bugs for details.
[    1.406226] i915 0000:00:02.0: [drm] GuC firmware(s) can be downloaded from https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/i915
[    1.478153] i915 0000:00:02.0: [drm] GuC firmware i915/tgl_guc_70.1.1.bin version 0.0
[    1.478167] i915 0000:00:02.0: GuC initialization failed -2
[    1.478169] i915 0000:00:02.0: Enabling uc failed (-5)
[    1.478170] i915 0000:00:02.0: Failed to initialize GPU, declaring it wedged!
[    1.478329] i915 0000:00:02.0: [drm:add_taint_for_CI [i915]] CI tainted:0x9 by intel_gt_init+0xb6/0x310 [i915]
[    1.542660] usb 1-2: new high-speed USB device number 2 using xhci_hcd
[    1.613206] [drm] Initialized i915 1.6.0 20201103 for 0000:00:02.0 on minor 0
[    1.613986] ACPI: video: Video Device [GFX0] (multi-head: yes  rom: no  post: no)
[    1.614356] input: Video Bus as /devices/LNXSYSTM:00/LNXSYBUS:00/PNP0A08:00/LNXVIDEO:00/input/input3
[    1.629136] fbcon: i915drmfb (fb0) is primary device

And the inxi output gives the same result as before:

Code: Select all

Graphics:  Device-1: Intel driver: i915 v: kernel 
           Display: x11 server: X.Org 1.20.11 driver: loaded: modesetting unloaded: fbdev,vesa resolution: 1680x1050~60Hz 
           OpenGL: renderer: llvmpipe (LLVM 11.0.1 256 bits) v: 4.5 Mesa 20.3.5 

[SMG]

I tried to replace the i915 directory under /lib/firmware. After the startup I still have these messages, although the new files are now present:

Code: Select all

drwxr-xr-x  2 root root    4096 14 déc 13:42 i915
drwxr-xr-x  2 root root    4096 23 déc 14:18 i915.20221224

ls -l i915/tgl_guc_70.1.1.bin
-rw-r--r-- 1 root root 277440 14 déc 13:42 i915/tgl_guc_70.1.1.bin

I'm confused as to why you are checking for Tiger Lake (11th-gen Intel) when your system is Raptor Lake (13th-gen Intel).

tgl = Tiger Lake

Getting the firmware for Tiger Lake, or even Alder Lake (12th-gen Intel) is not going to help your Raptor Lake system.

[magifix]

I wish you all a Merry Christmas and a peaceful time.

Back on the technical subject, I do not know why in my case during the boot, the os looks for a file called tgl_guc_xxx, but you are perfectly right, according to this article the memonic tgl corresponds to tiger lake:

https://01.org/linuxgraphics/downloads/firmware

so there is no point in looking in that direction

According to this other article, in my case the required filename should be rpl_guc_xxx:

https://www.phoronix.com/news/Intel-Rap ... e-Graphics

Unfortunatly there is no rpl_guc file in the latest linux-firmware archive I downloaded. Then if I would found an archive with the right file in there, according to the information I found, I should still run the command below to make the driver available at boot time:

Code: Select all

update-initramfs -c -k all

I must conclude the following to solve my problem:

- determine which kernel version is needed
- have the right graphics card drivers under /lib/firmware/i915/rpl_guc*
- regenerate the initramfs file (update-initramfs -c -k all)

Unfortunately for raptor lake I can't find any driver and I don't know from which kernel it will be supported. So for now let's wait and see.

[jwiz]

If you are into tinkering, copy the adlp_xxx entries in the i915 folder and rename them into rpl_xxx and see what happens as both use the same UHD 770 graphics.

No guarantees and tinker at your own risk.

[SMG]

Back on the technical subject, I do not know why in my case during the boot, the os looks for a file called tgl_guc_xxx

I believe it is because the default is to look for all of them. A similar type of warning happens for those using AMD graphics--it checks for the firmware for all GPUs and not just the GPU in the current computer.

Unfortunatly there is no rpl_guc file in the latest linux-firmware archive I downloaded.

The information I found indicates Raptor Lakes graphics support starts with the 5.19 kernel. I was not able to find anything specific about firmware needing to be installed separately from the kernel.

It may just be the rest of the graphics stack for LMDE5 is just too old to work properly with your iGPU? (Just a guess on my part.)

You could try booting a live session of Ubuntu 22.10 which has the 5.19 kernel to see what the inxi -Gx results show. That might give us an idea of what is needed.

[magifix]

I tried the Ubuntu 22.10 live DVD. There is no inxi, but I got these informations:

Code: Select all

lshw -c video

  *-display                 
       description: VGA compatible controller
       product: Raptor Lake-S UHD Graphics
       vendor: Intel Corporation
       physical id: 2
       bus info: pci@0000:00:02.0
       logical name: /dev/fb0
       version: 04
       width: 64 bits
       clock: 33MHz
       capabilities: pciexpress msi pm vga_controller bus_master cap_list rom fb
       configuration: depth=32 driver=i915 latency=0 mode=1680x1050 resolution=1680,1050 visual=truecolor xres=1680 yres=1050
       resources: iomemory:600-5ff iomemory:400-3ff irq:179 memory:6002000000-6002ffffff memory:4000000000-400fffffff ioport:5000(size=64) memory:c0000-dffff memory:4010000000-4016ffffff memory:4020000000-40ffffffff

Code: Select all

lspci -nn | grep -E 'VGA|Display'
00:02.0 VGA compatible controller [0300]: Intel Corporation Raptor Lake-S UHD Graphics [8086:a780] (rev 04)

and under /lib/firmware/i915 there is no rpl file, just about the same as under LMDE but newer versions.

One can see that Ubuntu 22.10 can detect the right gpu. No idea how to check the hw acceleration on this OS.

[SMG]

One can see that Ubuntu 22.10 can detect the right gpu. No idea how to check the hw acceleration on this OS.

I thought it did have inxi, but you can check the acceleration by running the following in a terminal.

Code: Select all

glxinfo -B

[magifix]

In the Ubuntu live DVD environment, inxi and glxinfo are unfortunatly not available. But as we have seen kernel 5.19 is a step further with Raptor Lake gpu integration. I was expecting that LMDE backport with kernel version 6.x may be so far, but this seems not to be the case, and I think it is not in our hand.

[SMG]

In the Ubuntu live DVD environment, inxi and glxinfo are unfortunatly not available.

I would not have guessed that, but I found several sources which indicate glxinfo must be installed on Ubuntu with:

Code: Select all

sudo apt-get install mesa-utils

[herbie643]

I am really late to the party but perhaps a Liquoriix kernel might be worth a try. I have run that kernel with no problems at all. To install just google liquorix and on their site they have easiest ever instructions to in the kernel.