Hi,

I'm on unraid 7.0.1 and I tried to install the IT87 driver since I have  gigabyte b860m Eagle for my intel ultra 235 .

Even after installing and installing it the fan controller won't appear in dynamix fan controller

Do I have to wait for the unraid 7.1 update or should I open a ticket in the IT87 github page?

 

I have 2 intel arc a770 and am using unraid 7.1 which should have a kernel that support intel gpus properly

I still can't get ollama to work though, what am i doing wrong?
for context the devices are recognized, i've tried no card number, card1 and card2 as those are the gpus not the apu

Edited April 5, 20251 yr by ceverson

4 hours ago, ceverson said:

I still can't get ollama to work though

Sorry but I don‘t ise ollama and that is not the support thread for ollama, I would recommend that you create a post in the support thread from the container.

 

I even didn‘t know that you can use a Intel ARC card for ollama.

Does Kernel 6.12.x mean that fan speed and temp can be reported from Intel Arc Axxx and Bxxx cards?

6 hours ago, Presjar said:

Does Kernel 6.12.x mean that fan speed and temp can be reported from Intel Arc Axxx and Bxxx cards?

I think you should ask that in the Plugin Support Thread from GPU Statistics.

11 hours ago, Presjar said:

Does Kernel 6.12.x mean that fan speed and temp can be reported from Intel Arc Axxx and Bxxx cards?

Fan speed should show.

 

Fix output for some igpus.

i915 Add Temp of cards if provided when on kernel 6.14 or will display N/A

 

XE fans in 6.16

 

https://www.phoronix.com/news/Intel-Xe-Linux-6.16-Fan-Speeds

 

Thank you both. @SimonF @ich777

Edited April 21, 20251 yr by Presjar

root@Tower:~# intel_gpu_top
No device filter specified and no discrete/integrated i915 devices found

 

@ich777 - Hi, I am new to unraid and unable to get the plugin to work for my iGPU on the Intel Ultra 5 245K. 

 

As someone has previously mentioned when the system boots up i see the information for a brief second on GPU Stats and then it disappears. I get errors when trying to use iGPU with Frigate saying that the iGPU doesn't exist.

 

Any help will be awesome,

 

Cheers

tower-diagnostics-20250427-1334.zip

32 minutes ago, RyMas said:

Any help will be awesome

Please upgrade to Unraid 7.1.0-rc.1 and your iGPU should show up.

Hi,I’m currently using Unraid 7.0.1 with a MCX311A-XCAT 10G single-port NIC (07:00.0 Ethernet controller: Mellanox Technologies MT27500 Family [ConnectX-3]).

 

While trying to use the Mellanox Firmware Tools plugin, I ran into an issue with the latest version (2025.02.10)—it fails to recognize the NIC. After some testing, I found that the problem is due to the mstflint version being too new. My card only works properly with mstflint version 4.18.0 or earlier.

 

So I downgraded to the 2021.11.30 version of Mellanox Firmware Tools, which includes the older mstflint. However, when I try to run it, I get the following error:

 

mstflint: error while loading shared libraries: libcrypto.so.1.1: cannot open shared object file: No such file or directory

 

After reading earlier replies, I confirmed that Unraid 7 uses OpenSSL 3 by default, which seems to be causing the incompatibility.

 

Now I’m in a dilemma:

• The latest tools don’t work with my NIC.
• The older tools require OpenSSL 1.1, which Unraid 7 no longer provides.

Would it be possible to add support for older NICs like the ConnectX-3 in future versions of the plugin, or provide a build with backward-compatible mstflint and OpenSSL 1.1 included?

16 minutes ago, Tardisyuan said:

While trying to use the Mellanox Firmware Tools plugin, I ran into an issue with the latest version (2025.02.10)—it fails to recognize the NIC. After some testing, I found that the problem is due to the mstflint version being too new. My card only works properly with mstflint version 4.18.0 or earlier.

Please always post your Diagnostics when you have hardware related issues.

 

16 minutes ago, Tardisyuan said:

So I downgraded to the 2021.11.30 version of Mellanox Firmware Tools, which includes the older mstflint. However, when I try to run it, I get the following error:

This version is not supported on Unraid 7, the first version that supported Unraid 7 was: 2024.07.11 which is based on mstflint v4.28.0

 

16 minutes ago, Tardisyuan said:

Would it be possible to add support for older NICs like the ConnectX-3 in future versions of the plugin, or provide a build with backward-compatible mstflint and OpenSSL 1.1 included?

I'll look into that but I'm not sure since I flashed a this exact same card about a year ago and it was working just fine.

 

Where did you have that information from that this card isn't supported anymore by mstflint?

39 minutes ago, ich777 said:

Please always post your Diagnostics when you have hardware related issues.

I’ve now attached the Diagnostics log generated and plugin info screenshot while using plugin version 2025.02.10 as requested.

unraid-diagnostics-20250430-1814.zip

 

 

40 minutes ago, ich777 said:

Where did you have that information from that this card isn't supported anymore by mstflint?

 

Regarding your question about why I believe the card is no longer supported by newer versions of mstflint:

I actually built a custom Docker environment using a Dockerfile for testing purposes. Here’s what I found:

• When using mstflint v4.20.0, I encountered the exact same error as shown in the Diagnostics—Failed to recognize device (or similar).
• When I downgraded to mstflint v4.18.0 within the same environment, everything worked perfectly: the NIC was detected and firmware tools operated as expected.

 

This leads me to believe that support for the ConnectX-3 (MCX311A-XCAT) may have been dropped or broken starting from mstflint v4.20.0.

 

Here’s the simplified Dockerfile I used for testing:

Dockerfile

9 minutes ago, Tardisyuan said:

I actually built a custom Docker environment using a Dockerfile for testing purposes. Here’s what I found:

So to speak you have no evidence that it is really related to the version, correct?

I see no evidence in the source that it was removed.

 

I report back when I have time to test.

 

 

BTW I would also recommend that you update the BIOS since you are a few versions behind and you have a lot of these messages in your syslog which are related to a Firmware bug:

Apr 30 01:14:19 Unraid kernel: ACPI BIOS Error (bug): Failure creating named object [\_SB.PC00.PEG1.PEGP._DSM.USRG], AE_ALREADY_EXISTS (20230628/dsfield-184)
Apr 30 01:14:19 Unraid kernel: ACPI Error: AE_ALREADY_EXISTS, CreateBufferField failure (20230628/dswload2-477)
Apr 30 01:14:19 Unraid kernel: ACPI Error: Aborting method \_SB.PC00.PEG1.PEGP._DSM due to previous error (AE_ALREADY_EXISTS) (20230628/psparse-529)
Apr 30 01:14:20 Unraid kernel: ACPI BIOS Error (bug): Failure creating named object [\_SB.PC00.PEG1.PEGP._DSM.USRG], AE_ALREADY_EXISTS (20230628/dsfield-184)
Apr 30 01:14:20 Unraid kernel: ACPI Error: AE_ALREADY_EXISTS, CreateBufferField failure (20230628/dswload2-477)
Apr 30 01:14:20 Unraid kernel: ACPI Error: Aborting method \_SB.PC00.PEG1.PEGP._DSM due to previous error (AE_ALREADY_EXISTS) (20230628/psparse-529)

 

2 hours ago, ich777 said:

So to speak you have no evidence that it is really related to the version, correct?

I see no evidence in the source that it was removed.

 

I report back when I have time to test.

 

 

BTW I would also recommend that you update the BIOS since you are a few versions behind and you have a lot of these messages in your syslog which are related to a Firmware bug:

Apr 30 01:14:19 Unraid kernel: ACPI BIOS Error (bug): Failure creating named object [\_SB.PC00.PEG1.PEGP._DSM.USRG], AE_ALREADY_EXISTS (20230628/dsfield-184)
Apr 30 01:14:19 Unraid kernel: ACPI Error: AE_ALREADY_EXISTS, CreateBufferField failure (20230628/dswload2-477)
Apr 30 01:14:19 Unraid kernel: ACPI Error: Aborting method \_SB.PC00.PEG1.PEGP._DSM due to previous error (AE_ALREADY_EXISTS) (20230628/psparse-529)
Apr 30 01:14:20 Unraid kernel: ACPI BIOS Error (bug): Failure creating named object [\_SB.PC00.PEG1.PEGP._DSM.USRG], AE_ALREADY_EXISTS (20230628/dsfield-184)
Apr 30 01:14:20 Unraid kernel: ACPI Error: AE_ALREADY_EXISTS, CreateBufferField failure (20230628/dswload2-477)
Apr 30 01:14:20 Unraid kernel: ACPI Error: Aborting method \_SB.PC00.PEG1.PEGP._DSM due to previous error (AE_ALREADY_EXISTS) (20230628/psparse-529)

 

You’re right — I don’t have solid evidence that it’s strictly version-related, just based on my limited testing results. Thanks a lot for taking the time to look into it!

 

As for the BIOS, I did try updating it, but unfortunately the issue still persists. I appreciate the reminder though — I’ll keep looking for possible solutions on that front.

ich777,

 

any chance you can create an intel npu driver plugin for unraid 7.1? thanks

 

https://github.com/intel/linux-npu-driver/releases

11 minutes ago, carlks said:

any chance you can create an intel npu driver plugin for unraid 7.1? thanks

I have to look into that, this will probably take me a few days.

From a first glance each NPU generation needs it's own driver.

 

Can you give me an example what software you want to use with that?

23 minutes ago, ich777 said:

I have to look into that, this will probably take me a few days.

From a first glance each NPU generation needs it's own driver.

 

Can you give me an example what software you want to use with that?

Hoping to test it with Scrypted NVR's openvino object detection.

 

https://docs.scrypted.app/server-hardware.html#:~:text=Scrypted NVR can leverage Intel's NPU.

3 minutes ago, carlks said:

Hoping to test it with Scrypted NVR's openvino object detection.

Okay, please give me possible one or two weeks since this seems a bit more complicated than I thought and I also need hardware to test on.

10 minutes ago, ich777 said:

Okay, please give me possible one or two weeks since this seems a bit more complicated than I thought and I also need hardware to test on.

Thanks, appreciate the help. Let me know if you need someone to test it. I have access to a Core Ultra 9 285K unraid system.

Edited April 30, 20251 yr by carlks

10 hours ago, carlks said:

Thanks, appreciate the help. Let me know if you need someone to test it. I have access to a Core Ultra 9 285K unraid system.

The NPU driver module itself needs to be built directly into Unraid to work, I already forwarded the request and I let you know when I know more.

 

However when the driver is built into Unraid it should be activated automatically and you should have the device path:

/dev/accel/accel0

directly on Unraid and of course this will mean that it is probably built into one of the next releases.

I'm not 100% sure if it's enough to have the Libraries, Compile and OpenVino components in the Docker container but from what I read that should be enough, so nothing else should be needed.

On 4/30/2025 at 10:35 PM, carlks said:

Core Ultra 9 285K

The driver is now included in Unraid 7.1.0-rc.4 please test if it's working, I assume you don't have to modprobe intel_vpu for it to work because it should load the driver on boot automatically.

 

To check if the driver works you can simply do the following from a Unraid terminal:

ls -l /dev/accel/

and you should see a directory named "accel0" in there.

On 4/30/2025 at 12:53 PM, Tardisyuan said:

it’s strictly version-related

It seems that you are correct and older cards are not supported any more on newer versions from mft-tools

 

I have to think about how to continue since this would possibly mean a complete overhaul from the plugin and how it works.

4 hours ago, ich777 said:

The driver is now included in Unraid 7.1.0-rc.4 please test if it's working, I assume you don't have to modprobe intel_vpu for it to work because it should load the driver on boot automatically.

 

To check if the driver works you can simply do the following from a Unraid terminal:

ls -l /dev/accel/

and you should see a directory named "accel0" in there.

Updated to rc4 but no luck with that directory.

 

1 hour ago, carlks said:

Updated to rc4 but no luck with that directory.

Can you please post your Diagnostics?

I am running into an error where after installing my B580.

I've made sure REBAR is on and am running 7.1.0r4

 

I do not see a dev/dri/ folder and am having issues running intel_gpu_top as well saying no devices are being found
 

dmesg | grep -i xe
[    0.000000] NX (Execute Disable) protection: active
[    0.000780] MTRR map: 7 entries (3 fixed + 4 variable; max 20), built from 9 variable MTRRs
[    0.060049] [mem 0xc0000000-0xefffffff] available for PCI devices
[    0.642444] ... fixed-purpose events:   0
[    0.812153] pci 0000:00:01.2:   bridge window [io  0xe000-0xffff]
[    0.813681] pci 0000:00:03.1:   bridge window [mem 0xd0000000-0xe07fffff 64bit pref]
[    0.828297] pci 0000:02:00.0:   bridge window [io  0xe000-0xffff]
[    0.832724] pci 0000:03:04.0:   bridge window [io  0xe000-0xefff]
[    0.842791] pci 0000:06:00.0: BAR 2 [io  0xe000-0xe01f]
[    0.852462] pci 0000:0b:00.0: BAR 0 [mem 0xe0000000-0xe07fffff 64bit pref]
[    0.900815] pci 0000:03:04.0:   bridge window [io  0xe000-0xefff]
[    0.902034] pci 0000:02:00.0:   bridge window [io  0xe000-0xffff]
[    0.902362] pci 0000:00:01.2:   bridge window [io  0xe000-0xffff]
[    0.903680] pci 0000:00:03.1:   bridge window [mem 0xd0000000-0xe07fffff 64bit pref]
[    0.904591] pci 0000:03:02.0: bridge window [mem 0x00000000-0x000fffff] to [bus 04] requires relaxed alignment rules
[    0.906827] pci 0000:03:04.0:   bridge window [io  0xe000-0xefff]
[    0.908046] pci 0000:02:00.0:   bridge window [io  0xe000-0xffff]
[    0.908373] pci 0000:00:01.2:   bridge window [io  0xe000-0xffff]
[    0.909690] pci 0000:00:03.1:   bridge window [mem 0xd0000000-0xe07fffff 64bit pref]
[    0.911900] pci 0000:02:00.0: bridge window [mem 0x00000000-0x007fffff] to [bus 03-0a] requires relaxed alignment rules
[    0.924715] pci 0000:03:04.0:   bridge window [io  0xe000-0xefff]
[    0.926719] pci 0000:02:00.0:   bridge window [io  0xe000-0xffff]
[    0.926940] pci 0000:00:01.2:   bridge window [io  0xe000-0xffff]
[    0.928257] pci 0000:00:03.1:   bridge window [mem 0xd0000000-0xe07fffff 64bit pref]
[    0.931721] pci 0000:03:04.0:   bridge window [io  0xe000-0xefff]
[    0.933652] pci 0000:02:00.0:   bridge window [io  0xe000-0xffff]
[    0.933983] pci 0000:00:01.2:   bridge window [io  0xe000-0xffff]
[    0.935299] pci 0000:00:03.1:   bridge window [mem 0xd0000000-0xe07fffff 64bit pref]
[    0.936586] pci_bus 0000:02: resource 0 [io  0xe000-0xffff]
[    0.936796] pci_bus 0000:03: resource 0 [io  0xe000-0xffff]
[    0.937321] pci_bus 0000:06: resource 0 [io  0xe000-0xefff]
[    0.938061] pci_bus 0000:0b: resource 2 [mem 0xd0000000-0xe07fffff 64bit pref]
[  116.445145] RAPL PMU: API unit is 2^-32 Joules, 1 fixed counters, 163840 ms ovfl timer
[  117.405533] xe 0000:0d:00.0: vgaarb: deactivate vga console
[  298.546975] xe 0000:0d:00.0: [drm] *ERROR* PCODE initialization timedout after: 3 min
[  298.546984] xe 0000:0d:00.0: probe with driver xe failed with error -110

Any help would be much appreciated

 

tower-diagnostics-20250502-1343.zip