Jump to content

[PLUGIN] Intel iGPU SR-IOV - Support Page

This topic contains 325 posts with an estimated read time of 221 minutes. A summary containing the most significant posts is available with an estimated read time of 5 minutes.
View summary
giganode
in Plugin Support

Featured Replies

I am having a similiar issue to Lunixx, everytime I set the number of VF and restart the system it still says 0 available. I've confirmed my BIOS and above 4G decoding is enabled, as well as SR-IOV. I then checked my system log and found this

  

Mar 18 23:43:35 Nostromo root: plugin: installing: i915-sriov.plg
Mar 18 23:43:35 Nostromo root: Executing hook script: pre_plugin_checks
Mar 18 23:43:35 Nostromo root: plugin: running: anonymous
Mar 18 23:43:35 Nostromo root: plugin: creating: /usr/local/emhttp/plugins/intel-i915-sriov/README.md - from INLINE content
Mar 18 23:43:35 Nostromo root: plugin: checking: /boot/config/plugins/i915-sriov/unraid-i915-sriov-2023.11.22.txz - MD5
Mar 18 23:43:35 Nostromo root: plugin: skipping: /boot/config/plugins/i915-sriov/unraid-i915-sriov-2023.11.22.txz already exists
Mar 18 23:43:35 Nostromo root: plugin: running: upgradepkg --install-new /boot/config/plugins/i915-sriov/unraid-i915-sriov-2023.11.22.txz
Mar 18 23:43:35 Nostromo root: 
Mar 18 23:43:35 Nostromo root: +==============================================================================
Mar 18 23:43:35 Nostromo root: | Installing new package /boot/config/plugins/i915-sriov/unraid-i915-sriov-2023.11.22.txz
Mar 18 23:43:35 Nostromo root: +==============================================================================
Mar 18 23:43:35 Nostromo root: 
Mar 18 23:43:35 Nostromo root: Verifying package unraid-i915-sriov-2023.11.22.txz.
Mar 18 23:43:35 Nostromo root: Installing package unraid-i915-sriov-2023.11.22.txz:
Mar 18 23:43:35 Nostromo root: PACKAGE DESCRIPTION:
Mar 18 23:43:35 Nostromo root: Package unraid-i915-sriov-2023.11.22.txz installed.
Mar 18 23:43:35 Nostromo root: plugin: running: anonymous
Mar 18 23:43:36 Nostromo root: patching file usr/local/emhttp/plugins/dynamix.vm.manager/include/libvirt.php
Mar 18 23:43:36 Nostromo root: Hunk #1 succeeded at 816 (offset 36 lines).
Mar 18 23:43:36 Nostromo root: 
Mar 18 23:43:36 Nostromo root: ------------------------SR-IOV package found locally!-------------------------
Mar 18 23:43:36 Nostromo root: ----------------SR-IOV package already installed, nothing to do----------------
Mar 18 23:43:36 Nostromo root: ---------------------Kernel Module 'i915' already enabled----------------------
Mar 18 23:43:38 Nostromo root: ---Setting VFs to: 2---
Mar 18 23:43:38 Nostromo kernel: pci 0000:00:02.0: no driver bound to device; cannot configure SR-IOV
Mar 18 23:43:38 Nostromo root: 
Mar 18 23:43:38 Nostromo root: -------------------------------------------------
Mar 18 23:43:38 Nostromo root: ---Installation from SR-IOV plugin successful!---
Mar 18 23:43:38 Nostromo root: -------------------------------------------------
Mar 18 23:43:38 Nostromo root: 
Mar 18 23:43:38 Nostromo root: plugin: i915-sriov.plg installed
Mar 18 23:43:38 Nostromo root: plugin: i915-sriov.plg installed
Mar 18 23:43:38 Nostromo root: Executing hook script: post_plugin_checks
Mar 18 23:43:38 Nostromo root: plugin: installing: libvirtwol.plg
Mar 18 23:43:38 Nostromo root: Executing hook script: pre_plugin_checks
Mar 18 23:43:38 Nostromo root: plugin: checking: /boot/config/plugins/libvirtwol/libvirt-python-env-4.10.0-x86_64-1.txz - MD5
Mar 18 23:43:38 Nostromo root: plugin: skipping: /boot/config/plugins/libvirtwol/libvirt-python-env-4.10.0-x86_64-1.txz already exists
Mar 18 23:43:38 Nostromo root: plugin: running: upgradepkg --install-new /boot/config/plugins/libvirtwol/libvirt-python-env-4.10.0-x86_64-1.txz

 

Crucially the part that caught my attention was "pci 0000:00:02.0: no driver bound to device; cannot configure SR-IOV" so I checked the new System Drivers in Tools and in there I could see that the i915 driver is listed as disabled. I found a file called i915.conf with the text "blacklist i915" present in /boot/config/modprobe.d, so I deleted the file and rebooted. However the plugin still failed to work. I checked and the file i915.conf is still there, so I deleted it and this time created a new i915.conf that was blank with touch and restarted again. This time the file was overwritten with a new file that contained the text "blacklist i915" again. I also found a blank copy of i915.conf in /etc/modprobe.d. I have deleted the file multiple times, but each time I restart it is there again. There are no scripts that create this file, so I don't know where it is coming from.

 

I've also tried going into the Tools section of the GUI, opening the System Drivers section and trying to remove he Blacklist from there. I have edited the modprobe config, as well as deleting it ad finishing by pressing the "Rebuild Modules" button. But each time the i915 file is removed, only to reappear when I reboot.

 

Any help would be appreciated as I feel like I am going around in circles here.

Edited by Deadboy01
Added more steps

  • Quote
    • Replies 324
    • Views 45k
    • Created
    • Last Reply

    Top Posters In This Topic

    Popular Days

    Most Popular Posts

    • ich777
      ich777

      Just a little heads up for all i915 SRIOV users...   When the first beta/RC from Unraid 6.13.0 or 7.0 drops it will most likely be the case that SRIOV is not supported there. This is ca

    • alturismo
      alturismo

      to all on unraid 7 beta2, plugin updated and working again   so anyone who wanted to wait for unraid 7 due sr iov ... enjoy now  

    • ich777
      ich777

      The plugin is the same and I'm still compiling the plugin packages for this new plugin for @giganode.   Uninstall my plugin Shutdown all VMs Install the plugin from the CA

    Posted Images

    4 hours ago, Deadboy01 said:

    But each time the i915 file is removed, only to reappear when I reboot.

    may rather post a diagnostics ...

     

    and the modprobe file should be there, please dont remove it, you probably also use the intel gpu top plugin or so which also automatically create it.

  • Quote
    On 3/19/2024 at 4:35 AM, alturismo said:

    you probably also use the intel gpu top plugin or so which also automatically create it.

    I don't use the intel gpu top plugin, so that is not the cause. I upgraded the server yesterday and previously it was using a Haswell Xeon CPU, so this is the first time I have had an intel iGPU present.

     

    On 3/19/2024 at 4:35 AM, alturismo said:

    may rather post a diagnostics ...

    Attached is the diagnostics file.

     

    Edited by Deadboy01
    Removed diagnostic file

  • Quote
    5 hours ago, Deadboy01 said:

    I don't use the intel gpu top plugin, so that is not the cause.

    i see, may try installing it, also dont see any config error from your side ...

     

    may 1 point, is your primary GPU in BIOS settled to internal iGPU ?

     

    may check this too ... if the NV PCIe device is setted as primary ... may thats the reason, something i would look for primary.

     

    Rest is looking ok, besides the error you see and mentioned yourself, should look otherwise ...

  • Quote
    On 3/19/2024 at 4:37 PM, alturismo said:

    i see, may try installing it, also dont see any config error from your side ...

     

    may 1 point, is your primary GPU in BIOS settled to internal iGPU ?

     

    may check this too ... if the NV PCIe device is setted as primary ... may thats the reason, something i would look for primary.

     

    Rest is looking ok, besides the error you see and mentioned yourself, should look otherwise ...

    When I did this I get the error message "No intel graphics card present" from the Intel Graphics SR-IOV page under Settings

     

    22 hours ago, ich777 said:

    Please update the BIOS, it's from 2022

    I cannot believe I missed this, updating the BIOS has resolved another issue I was facing, but I am still experiencing the same issues with regards to the SR-IOV plugin.

  • Quote
    • Author
    On 3/19/2024 at 12:58 AM, Deadboy01 said:

    I am having a similiar issue to Lunixx, everytime I set the number of VF and restart the system it still says 0 available. I've confirmed my BIOS and above 4G decoding is enabled, as well as SR-IOV. I then checked my system log and found this

      

    Mar 18 23:43:35 Nostromo root: plugin: installing: i915-sriov.plg
Mar 18 23:43:35 Nostromo root: Executing hook script: pre_plugin_checks
Mar 18 23:43:35 Nostromo root: plugin: running: anonymous
Mar 18 23:43:35 Nostromo root: plugin: creating: /usr/local/emhttp/plugins/intel-i915-sriov/README.md - from INLINE content
Mar 18 23:43:35 Nostromo root: plugin: checking: /boot/config/plugins/i915-sriov/unraid-i915-sriov-2023.11.22.txz - MD5
Mar 18 23:43:35 Nostromo root: plugin: skipping: /boot/config/plugins/i915-sriov/unraid-i915-sriov-2023.11.22.txz already exists
Mar 18 23:43:35 Nostromo root: plugin: running: upgradepkg --install-new /boot/config/plugins/i915-sriov/unraid-i915-sriov-2023.11.22.txz
Mar 18 23:43:35 Nostromo root: 
Mar 18 23:43:35 Nostromo root: +==============================================================================
Mar 18 23:43:35 Nostromo root: | Installing new package /boot/config/plugins/i915-sriov/unraid-i915-sriov-2023.11.22.txz
Mar 18 23:43:35 Nostromo root: +==============================================================================
Mar 18 23:43:35 Nostromo root: 
Mar 18 23:43:35 Nostromo root: Verifying package unraid-i915-sriov-2023.11.22.txz.
Mar 18 23:43:35 Nostromo root: Installing package unraid-i915-sriov-2023.11.22.txz:
Mar 18 23:43:35 Nostromo root: PACKAGE DESCRIPTION:
Mar 18 23:43:35 Nostromo root: Package unraid-i915-sriov-2023.11.22.txz installed.
Mar 18 23:43:35 Nostromo root: plugin: running: anonymous
Mar 18 23:43:36 Nostromo root: patching file usr/local/emhttp/plugins/dynamix.vm.manager/include/libvirt.php
Mar 18 23:43:36 Nostromo root: Hunk #1 succeeded at 816 (offset 36 lines).
Mar 18 23:43:36 Nostromo root: 
Mar 18 23:43:36 Nostromo root: ------------------------SR-IOV package found locally!-------------------------
Mar 18 23:43:36 Nostromo root: ----------------SR-IOV package already installed, nothing to do----------------
Mar 18 23:43:36 Nostromo root: ---------------------Kernel Module 'i915' already enabled----------------------
Mar 18 23:43:38 Nostromo root: ---Setting VFs to: 2---
Mar 18 23:43:38 Nostromo kernel: pci 0000:00:02.0: no driver bound to device; cannot configure SR-IOV
Mar 18 23:43:38 Nostromo root: 
Mar 18 23:43:38 Nostromo root: -------------------------------------------------
Mar 18 23:43:38 Nostromo root: ---Installation from SR-IOV plugin successful!---
Mar 18 23:43:38 Nostromo root: -------------------------------------------------
Mar 18 23:43:38 Nostromo root: 
Mar 18 23:43:38 Nostromo root: plugin: i915-sriov.plg installed
Mar 18 23:43:38 Nostromo root: plugin: i915-sriov.plg installed
Mar 18 23:43:38 Nostromo root: Executing hook script: post_plugin_checks
Mar 18 23:43:38 Nostromo root: plugin: installing: libvirtwol.plg
Mar 18 23:43:38 Nostromo root: Executing hook script: pre_plugin_checks
Mar 18 23:43:38 Nostromo root: plugin: checking: /boot/config/plugins/libvirtwol/libvirt-python-env-4.10.0-x86_64-1.txz - MD5
Mar 18 23:43:38 Nostromo root: plugin: skipping: /boot/config/plugins/libvirtwol/libvirt-python-env-4.10.0-x86_64-1.txz already exists
Mar 18 23:43:38 Nostromo root: plugin: running: upgradepkg --install-new /boot/config/plugins/libvirtwol/libvirt-python-env-4.10.0-x86_64-1.txz

     

    Crucially the part that caught my attention was "pci 0000:00:02.0: no driver bound to device; cannot configure SR-IOV" so I checked the new System Drivers in Tools and in there I could see that the i915 driver is listed as disabled. I found a file called i915.conf with the text "blacklist i915" present in /boot/config/modprobe.d, so I deleted the file and rebooted. However the plugin still failed to work. I checked and the file i915.conf is still there, so I deleted it and this time created a new i915.conf that was blank with touch and restarted again. This time the file was overwritten with a new file that contained the text "blacklist i915" again. I also found a blank copy of i915.conf in /etc/modprobe.d. I have deleted the file multiple times, but each time I restart it is there again. There are no scripts that create this file, so I don't know where it is coming from.

     

    I've also tried going into the Tools section of the GUI, opening the System Drivers section and trying to remove he Blacklist from there. I have edited the modprobe config, as well as deleting it ad finishing by pressing the "Rebuild Modules" button. But each time the i915 file is removed, only to reappear when I reboot.

     

    Any help would be appreciated as I feel like I am going around in circles here.

     

    40 minutes ago, Deadboy01 said:

    When I did this I get the error message "No intel graphics card present" from the Intel Graphics SR-IOV page under Settings

     

    I cannot believe I missed this, updating the BIOS has resolved another issue I was facing, but I am still experiencing the same issues with regards to the SR-IOV plugin.

     

    Sorry for the late answer. Can you please do the following:

     

    uninstall all intel igpu related plugins -> full shutdown of the server -> start server and install intel_gpu_top and afterwards install the sriov plugin -> go to sriov plugin settings page and hit "enable now" and set VFs to 2 and save to file -> reboot

     

    Don't touch any files manually. after the reboot check log files etc and repost diagnostics please.

     

    Please also test in terminal and post the outcome:

      

    intel_gpu_top -d sriov

     

     

  • Quote
    On 3/20/2024 at 7:33 PM, giganode said:

     

     

    Sorry for the late answer. Can you please do the following:

     

    uninstall all intel igpu related plugins -> full shutdown of the server -> start server and install intel_gpu_top and afterwards install the sriov plugin -> go to sriov plugin settings page and hit "enable now" and set VFs to 2 and save to file -> reboot

     

    Don't touch any files manually. after the reboot check log files etc and repost diagnostics please.

     

    Please also test in terminal and post the outcome:

      

    intel_gpu_top -d sriov

     

     

    Thanks for responding. I followed all of the steps listed, and have gotten the same result. I've checked the logs and everything seems to be the same, the error message "no driver bound to device; cannot configure SR-IOV" is still there. I have attached the new diagnostic file as requested (the device name has changed as I am still in the process of setting up the new server, but it is the same hardware).

     

    I ran the command and received the following output. I double checked the BIOS and SRIOV is definitely enabled.

    Quote

    Requested device sriov not found!

     

     

    Edited by Deadboy01
    Removed diagnostic file

  • Quote

    SUCCESS! 🥳

     

    I went down a bit of a rabbit hole, so bear with me. After running the command "intel_gpu_top -d sriov" I looked a little more into the "Intel-GPU-TOP" plugin as this is the first time I have had an iGPU in an UnRAID server. I tried running the command "intel_gpu_top" but received an error message along the lines of no i915 device present (I don't remember the exact error). I then removed the "Intel SR-IOV" plugin and rebooted to try and isolate the problem, but got the same error. Next I tried to install radeontop and got an error saying "------Something went wrong! Can't enable-------" during the build process. So I checked ich777's support forum for that plugin and found someone experiencing the same issue. The resolution there was to remove "nomodeset" from the syslinux.config. I removed it and rebooted.

     

    After that I found that "intel_gpu_top" worked and I was able to install radeontop without errors. Then I reinstalled the "Intel SR-IOV" plugin, enabled it and rebooted. After that it seems to be working as expected now. I've had this server for 6+ years, so cannot remember when or why I added "nomodeset", but I've gone through so many hardware changes that it is probably not needed anymore.

     

    TLDR, check your syslinux.config and if you have "nomodeset" in there, remove it.

    Edited by Deadboy01

  • Quote
    • Like 1
    • Upvote 1

    Anyone having issues with their Unraid Server crashing when using this plugin? I think it has to do with the VM locking up at somepoint and then causing the Unraid Server to lock up. It's kinda been imposible to grab logs since I have no idea when it is going to happen, but this past time the VM crashed and I caught it. My WebUI became unusable, but luckily I have a BLIKVM hooked up to my system and I had booted into the GUI mode and was able to pull some info. Unfortunately I had to take screenshots and couldn't copy and paste logs. A VM log is also screenshotted. I get those errors on the VM but performance seems fine besides the crash at some point. 

    Screenshot 2024-03-26 at 2.12.27 PM.png

    Screenshot 2024-03-26 at 2.12.48 PM.png

    Screenshot 2024-03-26 at 2.34.23 PM.png

  • Quote

    im facing a strange issue, if i reboot the server, my windows vm runs fine with sr-iov with my i5 14500 

     

    but if i quit the vm, my whole gpu dissapears, no gpu on device list and plugin shows no gpu found until i reboot the server

     

     

    i915 0000:00:02.1: [drm] *ERROR* tlb invalidation response timed out for seqno 23:o

      

    ar 29 05:57:58 NAS kernel: vfio-pci 0000:00:02.1: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=io+mem:owns=none
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: Running in SR-IOV VF mode
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: [drm] GT0: GUC: interface version 0.1.4.1
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: [drm] VT-d active for gfx access
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: [drm] Using Transparent Hugepages
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: [drm] GT0: GUC: interface version 0.1.4.1
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: GuC firmware PRELOADED version 1.4 submission:SR-IOV VF
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: HuC firmware PRELOADED
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: [drm] Protected Xe Path (PXP) protected content support initialized
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: [drm] PMU not supported for this GPU.
Mar 29 05:57:58 NAS kernel: sdd: sdd1 sdd2 sdd3 sdd4
Mar 29 05:57:58 NAS kernel: [drm] Initialized i915 1.6.0 20201103 for 0000:00:02.1 on minor 1
Mar 29 05:57:58 NAS kernel: ata6.00: Enabling discard_zeroes_data
Mar 29 05:57:58 NAS kernel: sdd: sdd1 sdd2 sdd3 sdd4
Mar 29 05:57:58 NAS usb_manager: Info: rc.usb_manager  vm_action Windows 11 stopped end -
Mar 29 05:57:59 NAS kernel: i915 0000:00:02.1: [drm] *ERROR* tlb invalidation response timed out for seqno 23
Mar 29 05:57:59 NAS kernel: i915 0000:00:02.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=none:owns=io+mem
Mar 29 05:57:59 NAS kernel: i915 0000:00:02.2: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=none:owns=none
Mar 29 05:57:59 NAS kernel: pci 0000:00:02.1: Removing from iommu group 19
Mar 29 05:57:59 NAS kernel: i915 0000:00:02.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=io+mem:owns=io+mem
Mar 29 05:57:59 NAS kernel: pci 0000:00:02.2: Removing from iommu group 20
Mar 29 05:58:00 NAS unassigned.devices: Disk with ID 'Samsung_SSD_860_EVO_500GB_S3Z2NB0K660578V (dev1)' is not set to auto mount.
Mar 29 05:58:00 NAS unassigned.devices: Disk with ID 'Samsung_SSD_860_EVO_500GB_S3Z2NB0K660578V (dev1)' is not set to auto mount.
Mar 29 05:58:00 NAS unassigned.devices: Disk with ID 'Samsung_SSD_860_EVO_500GB_S3Z2NB0K660578V (dev1)' is not set to auto mount.
Mar 29 05:58:00 NAS unassigned.devices: Partition '/dev/sdd2' does not have a file system and cannot be mounted.
Mar 29 05:58:01 NAS kernel: i915 0000:00:02.0: Disabled 2 VFs
Mar 29 05:58:01 NAS kernel: Console: switching to colour dummy device 80x25
Mar 29 05:58:01 NAS acpid: input device has been disconnected, fd 11
Mar 29 05:58:01 NAS kernel: pci 0000:00:02.0: Removing from iommu group 0

     

    Edited by Encore

  • Quote
    • Author
    On 3/26/2024 at 7:35 PM, PaulW08 said:

    Anyone having issues with their Unraid Server crashing when using this plugin? I think it has to do with the VM locking up at somepoint and then causing the Unraid Server to lock up. It's kinda been imposible to grab logs since I have no idea when it is going to happen, but this past time the VM crashed and I caught it. My WebUI became unusable, but luckily I have a BLIKVM hooked up to my system and I had booted into the GUI mode and was able to pull some info. Unfortunately I had to take screenshots and couldn't copy and paste logs. A VM log is also screenshotted. I get those errors on the VM but performance seems fine besides the crash at some point. 

    Screenshot 2024-03-26 at 2.12.27 PM.png

    Screenshot 2024-03-26 at 2.12.48 PM.png

    Screenshot 2024-03-26 at 2.34.23 PM.png

     

    For now you can ignore the VFIO_MAP_DMA errors.

     

    Is the vm running 24/7?

     

    Please post Diagnostics.

     

  • Quote
    • Author
    14 hours ago, Encore said:

    im facing a strange issue, if i reboot the server, my windows vm runs fine with sr-iov with my i5 14500 

     

    but if i quit the vm, my whole gpu dissapears, no gpu on device list and plugin shows no gpu found until i reboot the server

     

     

    i915 0000:00:02.1: [drm] *ERROR* tlb invalidation response timed out for seqno 23:o

      

    ar 29 05:57:58 NAS kernel: vfio-pci 0000:00:02.1: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=io+mem:owns=none
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: Running in SR-IOV VF mode
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: [drm] GT0: GUC: interface version 0.1.4.1
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: [drm] VT-d active for gfx access
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: [drm] Using Transparent Hugepages
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: [drm] GT0: GUC: interface version 0.1.4.1
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: GuC firmware PRELOADED version 1.4 submission:SR-IOV VF
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: HuC firmware PRELOADED
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: [drm] Protected Xe Path (PXP) protected content support initialized
Mar 29 05:57:58 NAS kernel: i915 0000:00:02.1: [drm] PMU not supported for this GPU.
Mar 29 05:57:58 NAS kernel: sdd: sdd1 sdd2 sdd3 sdd4
Mar 29 05:57:58 NAS kernel: [drm] Initialized i915 1.6.0 20201103 for 0000:00:02.1 on minor 1
Mar 29 05:57:58 NAS kernel: ata6.00: Enabling discard_zeroes_data
Mar 29 05:57:58 NAS kernel: sdd: sdd1 sdd2 sdd3 sdd4
Mar 29 05:57:58 NAS usb_manager: Info: rc.usb_manager  vm_action Windows 11 stopped end -
Mar 29 05:57:59 NAS kernel: i915 0000:00:02.1: [drm] *ERROR* tlb invalidation response timed out for seqno 23
Mar 29 05:57:59 NAS kernel: i915 0000:00:02.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=none:owns=io+mem
Mar 29 05:57:59 NAS kernel: i915 0000:00:02.2: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=none:owns=none
Mar 29 05:57:59 NAS kernel: pci 0000:00:02.1: Removing from iommu group 19
Mar 29 05:57:59 NAS kernel: i915 0000:00:02.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=io+mem:owns=io+mem
Mar 29 05:57:59 NAS kernel: pci 0000:00:02.2: Removing from iommu group 20
Mar 29 05:58:00 NAS unassigned.devices: Disk with ID 'Samsung_SSD_860_EVO_500GB_S3Z2NB0K660578V (dev1)' is not set to auto mount.
Mar 29 05:58:00 NAS unassigned.devices: Disk with ID 'Samsung_SSD_860_EVO_500GB_S3Z2NB0K660578V (dev1)' is not set to auto mount.
Mar 29 05:58:00 NAS unassigned.devices: Disk with ID 'Samsung_SSD_860_EVO_500GB_S3Z2NB0K660578V (dev1)' is not set to auto mount.
Mar 29 05:58:00 NAS unassigned.devices: Partition '/dev/sdd2' does not have a file system and cannot be mounted.
Mar 29 05:58:01 NAS kernel: i915 0000:00:02.0: Disabled 2 VFs
Mar 29 05:58:01 NAS kernel: Console: switching to colour dummy device 80x25
Mar 29 05:58:01 NAS acpid: input device has been disconnected, fd 11
Mar 29 05:58:01 NAS kernel: pci 0000:00:02.0: Removing from iommu group 0

    Please post Diagnostics.

  • Quote
    34 minutes ago, giganode said:

     

    Don't bind the gpu to vfio.

     

    Did you manually add 

      

    modprobe i915

     

    to your /boot/config/go file?

     

     

    yep, i removed the vfio

     

    and yes, im not 100% sure, modprobe i915 was added by me long time ago - dk if i added these with my old cpu, switched from i5 9400 to i5 14500 maybe it was for the old passtrough 

    Edited by Encore

  • Quote
    • Author
    yep, i removed the vfio
     
    and yes, im not 100% sure, modprobe i915 was added by me long time ago - dk if i added these with my old cpu, switched from i5 9400 to i5 14500 maybe it was for the old passtrough 

    Please remove the line and try again
  • Quote
    2 hours ago, giganode said:


    Please remove the line and try again

    nope, after VM quit host gpu is missing , need to reboot to bring it back.

     

    removed the line in go file, rebooted -> started VM (VF was visible, passtrough worked) - after VM quit = gpu missing 

     

    fresh diagnostic after vm quit

     

    bios is the newest, ME is newest

     

     

     

    nas-diagnostics-20240330-0222.zip

    Edited by Encore

  • Quote
    5 hours ago, Encore said:

    fresh diagnostic after vm quit

     

    may some tryouts

     

    go file, /sbin/sysctl -p /boot/config/sysctl.conf ... may disable this as test

    i915 file, may remove your extra entries guc, fbc

    Win VM, may try q35 mashine type instead i440

     

    14th gen may not yet fully supported ... wait ...

  • Quote
    • Like 1
    6 hours ago, alturismo said:

    may some tryouts

     

    go file, /sbin/sysctl -p /boot/config/sysctl.conf ... may disable this as test

    i915 file, may remove your extra entries guc, fbc

    Win VM, may try q35 mashine type instead i440

     

    14th gen may not yet fully supported ... wait ...

    edit:

     

    nm - crash again

     

    fresh diag

      

    Mar 30 13:48:48 NAS kernel: vfio-pci 0000:00:02.1: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=io+mem:owns=none
Mar 30 13:48:48 NAS kernel: i915 0000:00:02.1: Running in SR-IOV VF mode
Mar 30 13:48:48 NAS kernel: i915 0000:00:02.1: [drm] GT0: GUC: interface version 0.1.4.1
Mar 30 13:48:48 NAS kernel: i915 0000:00:02.1: [drm] VT-d active for gfx access
Mar 30 13:48:48 NAS kernel: i915 0000:00:02.1: [drm] Using Transparent Hugepages
Mar 30 13:48:48 NAS kernel: i915 0000:00:02.1: [drm] GT0: GUC: interface version 0.1.4.1
Mar 30 13:48:48 NAS kernel: i915 0000:00:02.1: GuC firmware PRELOADED version 1.4 submission:SR-IOV VF
Mar 30 13:48:48 NAS kernel: i915 0000:00:02.1: HuC firmware PRELOADED
Mar 30 13:48:48 NAS kernel: sdc: sdc1 sdc2 sdc3 sdc4
Mar 30 13:48:48 NAS kernel: i915 0000:00:02.1: [drm] Protected Xe Path (PXP) protected content support initialized
Mar 30 13:48:48 NAS kernel: i915 0000:00:02.1: [drm] PMU not supported for this GPU.
Mar 30 13:48:48 NAS kernel: [drm] Initialized i915 1.6.0 20201103 for 0000:00:02.1 on minor 1
Mar 30 13:48:48 NAS kernel: ata6.00: Enabling discard_zeroes_data
Mar 30 13:48:48 NAS kernel: sdc: sdc1 sdc2 sdc3 sdc4
Mar 30 13:48:48 NAS usb_manager: Info: rc.usb_manager  vm_action Windows 11 stopped end -
Mar 30 13:48:49 NAS kernel: i915 0000:00:02.1: [drm] *ERROR* tlb invalidation response timed out for seqno 23
Mar 30 13:48:49 NAS kernel: i915 0000:00:02.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=none:owns=io+mem
Mar 30 13:48:49 NAS kernel: i915 0000:00:02.2: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=none:owns=none
Mar 30 13:48:49 NAS kernel: pci 0000:00:02.1: Removing from iommu group 19
Mar 30 13:48:49 NAS kernel: i915 0000:00:02.0: vgaarb: changed VGA decodes: olddecodes=none,decodes=io+mem:owns=io+mem
Mar 30 13:48:49 NAS kernel: pci 0000:00:02.2: Removing from iommu group 20
Mar 30 13:48:50 NAS unassigned.devices: Disk with ID 'Samsung_SSD_860_EVO_500GB_S3Z2NB0K660578V (dev2)' is not set to auto mount.
### [PREVIOUS LINE REPEATED 2 TIMES] ###
Mar 30 13:48:50 NAS unassigned.devices: Partition '/dev/sdc2' does not have a file system and cannot be mounted.
Mar 30 13:48:51 NAS kernel: i915 0000:00:02.0: Disabled 2 VFs
Mar 30 13:48:51 NAS kernel: Console: switching to colour dummy device 80x25
Mar 30 13:48:51 NAS acpid: input device has been disconnected, fd 11
Mar 30 13:48:51 NAS kernel: pci 0000:00:02.0: Removing from iommu group 0

    nas-diagnostics-20240330-1350.zip

     

     

    https://github.com/strongtz/i915-sriov-dkms/issues/118

     

    maybe memory leak on kernel? (just copy paste this, i have no glue :D )

    Edited by Encore

  • Quote

    Hello little problem there.

    I tried multiple tweaks but nothing works.

     

    Rocking a 12700 with a Z690 PG Riptide (last bios). I would like to passthrough my UHD770 and i'm not able to create VF's.

     

    I attached the logs when i try to "enable" VF's. Even if i restart the server nothing happens.

    I that directly linked with my Motherboard ?

     

    Also, i cannot activate Above 4G Decoding in bios, if i do it, cannot boot to Unraid. Would that be an issue for SR-IOV ?

     

    Thanks a lot by advance

    Capture d'écran 2024-03-30 213719.png

  • Quote
    41 minutes ago, LuttyMiix said:

    Also, i cannot activate Above 4G Decoding in bios, if i do it, cannot boot to Unraid. Would that be an issue for SR-IOV ?

     

    possible ... and did you try to set a lower number of VF's ?

     

    error saying not enough ressources ...

  • Quote
    5 minutes ago, alturismo said:

    possible ... and did you try to set a lower number of VF's ?

     

    error saying not enough ressources ...

    Tried 1 -2 - 7 nothing works. I'm figuring about how to activate 4G decoding. Don't find a clue on what is missing to pass boot with it activated

  • Quote
    Just now, LuttyMiix said:

    Don't find a clue on what is missing to pass boot with it activated

    basically, UEFI boot needs to be activated to use above 4G and rBAR support

  • Quote
    • Upvote 1
    35 minutes ago, alturismo said:

    basically, UEFI boot needs to be activated to use above 4G and rBAR support

    Wow ok i need to check that. Maybe th'tas what i miss hehe

  • Quote

    • Join the conversation

    You can post now and register later. If you have an account, sign in now to post with your account.
    Note: Your post will require moderator approval before it will be visible.

    Guest
    Reply to this topic...
    Go to topic listing