Jump to content

unRAID plugin for iGPU SR-IOV support


Recommended Posts

4 minutes ago, Daniel15 said:

which CPU are you using?

iam occupied to atm, i have an 13700k in my Server, i have no hiccups what so ever i can use PLEX/encode/VM at the same time

 

 

 

i went from 6.12.3 to 6.12.4 with the old Plugin and used the Patch from ich777, from there i used the new Plugin and thats it. 

 

updated from 6.12.4 to 6.12.5 and 6.12.6 the Plugin did the updating on its own.

Edited by domrockt
Link to comment
1 hour ago, Daniel15 said:

Hmm... I'll have to try and figure out why mine is crashing. I tested five times and could consistently repro the crash on 6.12.5, whereas it never happens with 6.12.4. Unfortunately I don't have a test system with an Intel CPU, just my production system, so I'll probably just have to stick to 6.12.4 for now. 

 

@domrocktand @neunghaha28 - which CPU are you using? 

 

i5 13500

Link to comment

I just built a second UNRAID server with a low power 11900K (bit of a contradiction I know) and an Asrock  Z490D4U-2lT2 motherboard. I want to split the iGPU between Plex for transcoding and one other VM for some very light remote gaming/emulation.

 

Is the SRV IO plugin going to enable me to achieve this set up and is there anything I need to be aware of config wise to get this to work?

 

Just installed the plugin and it says my 11900 is not supported or may be in VF mode. I thought the SRV IO plugin worked with the 11900k and what is VF mode?

Edited by mikeyosm
Link to comment
1 hour ago, mikeyosm said:

Just installed the plugin and it says my 11900 is not supported or may be in VF mode. I thought the SRV IO plugin worked with the 11900k and what is VF mode?

sadly the 11th Series Intel Chips are "hit & miss" ... while way more misses ...

 

no gvt-g support as 11th gen, no sr iov support ... even its stated more or less by Intel ...

 

mentioned several times about 11th gen CPU's .. sorry ...

  • Thanks 1
Link to comment
1 hour ago, alturismo said:

sadly the 11th Series Intel Chips are "hit & miss" ... while way more misses ...

 

no gvt-g support as 11th gen, no sr iov support ... even its stated more or less by Intel ...

 

mentioned several times about 11th gen CPU's .. sorry ...

No problem, bit of a shame. My other UNRAID server has an Nvidia P4 so not too much of an issue for me.

Link to comment

Hi @ich777

 

Tried installing the plugin tonight and got the following on install:

 

plugin: downloading: i915-sriov.plg ... done plugin: downloading: unraid-i915-sriov-2023.04.06.txz ... done +============================================================================== | Installing new package /boot/config/plugins/i915-sriov/unraid-i915-sriov-2023.04.06.txz +============================================================================== Verifying package unraid-i915-sriov-2023.04.06.txz. Installing package unraid-i915-sriov-2023.04.06.txz: PACKAGE DESCRIPTION: Package unraid-i915-sriov-2023.04.06.txz installed. patching file usr/local/emhttp/plugins/dynamix.vm.manager/include/libvirt.php Hunk #1 succeeded at 816 (offset 36 lines). +============================================================================== | WARNING - WARNING - WARNING - WARNING - WARNING - WARNING - WARNING - WARNING | | Don't close this window with the red 'X' in the top right corner | until the 'DONE' button is displayed! | | WARNING - WARNING - WARNING - WARNING - WARNING - WARNING - WARNING - WARNING +============================================================================== ----------Downloading i915-sriov module Package for kernel v6.1.64---------- ---------This could take some time, please don't close this window!----------- --------Can't download i915-sriov module Package for kernel v6.1.64----------- plugin: run failed: '/bin/bash' returned 1 Executing hook script: post_plugin_checks

 

Any assistance you can offer would be appreciated. 

Link to comment
2 hours ago, martinkeat said:

Any assistance you can offer would be appreciated. 

well, you see the error

 

2 hours ago, martinkeat said:

--------Can't download i915-sriov module Package for kernel v6.1.64--------

usually its

 

- you have issues downloading from github

- you have some other DNS in your net running which collides

- ...

also could be (uncommon)

- even broken USB, no spave left on Flash USB

- ...

 

are you using any DNS solutions at home, are you may blocked due DNS setup ...

  • Like 1
Link to comment

Hi,
unfortunately I have the same error as martinkeat.
The i915 plugin worked fine, but after the update to Unraid 6.12.6 the VF/drivers could no longer be bound. My syslog contains only the following entry:

Dec 11 00:50:13 unraid root: plugin: creating: /boot/config/plugins/i915-sriov//unraid-i915-sriov-2023.04.06.txz - downloading from URL https://raw.githubusercontent.com/zhtengw/unraid-i915-sriov/master/packages/unraid-i915-sriov-2023.04.06.txz
Dec 11 00:50:14 unraid root: plugin: running: upgradepkg --install-new /boot/config/plugins/i915-sriov//unraid-i915-sriov-2023.04.06.txz

 

I am concerned about the two "//" in the path or is this unfounded? It seems to me that I have a corrupt installation. Are there any instructions on how to manually remove the module etc.? Despite the statement "allready installed", I can't find an i915 entry under /boot/config/plugins, for example.

 

Many thanks for your tips and help. I have no DNS problem, a wget https://raw.githubusercontent.com/zhtengw/unraid-i915-sriov/master/packages/unraid-i915-sriov-2023.04.06.txz is successfully possible via ssh, also write access to /boot/config/

Link to comment
9 hours ago, knatzkopp said:

unfortunately I have the same error as martinkeat.

On 12/10/2023 at 1:55 AM, martinkeat said:

Tried installing the plugin tonight and got the following on install:

It looks like both of you are using the old plugin, since the plugin date from my fork is: 2023.11.22 and not 2023.04.06

Please look at the first recommended post in this thread on top, uninstall the old plugin and install my fork since @zhtengw seems like to be not active anymore and that's why I forked the plugin (I explained this in the recommended post).

  • Like 1
Link to comment
1 hour ago, ich777 said:

It looks like both of you are using the old plugin, since the plugin date from my fork is: 2023.11.22 and not 2023.04.06

Please look at the first recommended post in this thread on top, uninstall the old plugin and install my fork since @zhtengw seems like to be not active anymore and that's why I forked the plugin (I explained this in the recommended post).

Oh sorry, as we like to say in Germany: I can't see the wood for the trees... Didn't follow the pinned message. After an install & reboot everything is great again.
Thanks for the quick help -> reading always helps, should just do it :)

  • Like 2
Link to comment
On 11/23/2023 at 2:53 PM, Dotmaster5 said:

 

Hello alturismo,

 

I had the same Problems like you, i can not start Parsec or Sunshine because the IGPU was not able to output to an Display. I Try it with an HDMI Display connected, but it was not recognized inside the VM so the GPU can not set as output device. The IGPU has no fault, no error 43 or something, it works and was shown in the Device Manager and as a working GPU in the Task Manager. When i start a Program, like Ungine Heaven, it shows a Load in the Task Manager and also in the GPU Statistics on Unraid so basicly the IGPU was working correctly. A Short test with a Docker Container ( Steam Headless ) was working, it shows the connected Display and works nice. I think the Problem has something to do with the Parstrough to the VM, maybe something is missing so that the Display information was not passing to the VM  / Intel IGPU Driver. I have some more Servers with Nvidia GPU´s (Grid K2, Tesla M10, Tesla P4) but i never seen such Problems with the GPU Pastrough to a VM.

 

My CPU is a I3-N305 and the System is a Beelink EQ12 Pro

 

 

So I was at the exact point you were at last night. I googled a bit more, and I came across this post by scoops22 on reddit (attached to this reply). Using the older intel driver and the virtual display from MikeTheTech worked!

 

I can now use Sunshine to stream from my 12600k iGPU. Thank you everyone involved in making this work!

 

 

 

 

Screenshot 2023-12-12 125633.png

Edited by Quadrupole
screenshot
  • Thanks 1
Link to comment
7 hours ago, Quadrupole said:

and the virtual display from MikeTheTech worked!

thanks alot for the "wink", tested here now and yes, also working here on the n100 platform now, same behaviour as gvt-g with hw de/encoding, very nice and thanks sgain, this made me fire up the plugin and the vm again to give it a try ;) and worked out ;)

Link to comment

may 1 addition i see here, while connected to the VM everything is smooth and running fine, but once i quit the connection and the VM is in idle ... CPU usage is increasing to 100 % here.

 

Once i reconnect to the VM (either way, Parsec, VNC, ...) the VM returns to normal state and CPU usage is back to normal again.

 

just as note from my side here

 

in use

image.thumb.png.4571d07b3a9bcdf073dbeb2afb6ef350.png

 

disconnected and in idle

image.thumb.png.6bc0d8489404013d91a107607ea18c00.png

Link to comment
  • 2 weeks later...

I've been having some issues after the latest update.  6.12.6.
One of the issues I am having now is my /dev/dri has disappeared. 
 

Note:  I do not have an HDMI dummy plug installed, I didn't have one prior to upgrade don't know if that matters.
Note:  I do not know if any bios settings got reset during reboots, it MIGHT be disabled in bios I can not check as I am remote for a while.

System: 13700k with the 770 iGPU

Radeon Card for VM
The system devices tab shows my iGPU and that it is in the kernel.

INTEL_GPU_TOP also shows no devices.

When I install the SR_IOV plugin I get an error up top saying a process is still running and then eventually the GUI hangs.  Don't know if that's related but there it is.

Tried with both 
/boot/config/modprobe.d/i915.conf  - empty file

and

/boot/config/modprobe.d/i915.conf - blacklist i915

 

 

root@brain:/dev# lsmod | grep -i 'i915'
i915                 2527232  1
iosf_mbi               20480  2 i915,intel_rapl_common
drm_buddy              16384  2 amdgpu,i915
i2c_algo_bit           16384  2 amdgpu,i915
ttm                    77824  3 amdgpu,drm_ttm_helper,i915
drm_display_helper    135168  2 amdgpu,i915
drm_kms_helper        167936  3 drm_display_helper,amdgpu,i915
intel_gtt              24576  1 i915
drm                   495616  9 gpu_sched,drm_kms_helper,drm_display_helper,drm_buddy,amdgpu,drm_ttm_helper,i915,ttm
video                  61440  2 amdgpu,i915
i2c_core               86016  8 drm_kms_helper,i2c_algo_bit,drm_display_helper,i2c_smbus,amdgpu,i2c_i801,i915,drm
backlight              20480  5 video,drm_display_helper,amdgpu,i915,drm

 

root@brain:/dev# modinfo i915
filename:       /lib/modules/6.1.64-Unraid/kernel/drivers/gpu/drm/i915/i915.ko.xz
license:        GPL and additional rights
description:    Intel Graphics
author:         Intel Corporation
author:         Tungsten Graphics, Inc.
import_ns:      DMA_BUF

 

 

root@brain:/dev# lspci -k
00:00.0 Host bridge: Intel Corporation Device a703 (rev 01)
        DeviceName: Onboard - Other
        Subsystem: Micro-Star International Co., Ltd. [MSI] Device 7d32
00:01.0 PCI bridge: Intel Corporation Device a70d (rev 01)
        Subsystem: Micro-Star International Co., Ltd. [MSI] Device 7d32
        Kernel driver in use: pcieport
00:02.0 VGA compatible controller: Intel Corporation Raptor Lake-S GT1 [UHD Graphics 770] (rev 04)
        DeviceName: Onboard - Video
        Subsystem: Micro-Star International Co., Ltd. [MSI] Device 7d32
        Kernel driver in use: i915
        Kernel modules: i915
00:06.0 PCI bridge: Intel Corporation Raptor Lake PCIe 4.0 Graphics Port (rev 01)
        Kernel driver in use: pcieport
00:08.0 System peripheral: Intel Corporation GNA Scoring Accelerator module (rev 01)
        DeviceName: Onboard - Other


 

 

 

Current setup:  INTEL_GPU_TOP installed, SR_IOV not installed, /dev/dri missing

 

image.thumb.png.dd16009e6e59643b46250759364dfc0e.png

image.thumb.png.8ba9e88f6e6b658ab2a10720b16e425d.png

 

 

 

 

Thank you in advance to anyone looking at this.

Derek

brain-diagnostics-20231223-1641.zip

Link to comment
2 hours ago, dirkinthedark said:

For what its worth, after I unbound my AMD card from VRIO, deleted that, removed all plugins (Intel), and removed the modprobe.d for i915 it showed up again.

In the Diagnostics you've provided you have a Trace here:

Dec 23 12:02:29 brain kernel: i915 0000:00:02.0: [drm] VT-d active for gfx access
Dec 23 12:02:29 brain kernel: BUG: unable to handle page fault for address: 0000000fffffffe8
Dec 23 12:02:29 brain kernel: #PF: supervisor read access in kernel mode
Dec 23 12:02:29 brain kernel: #PF: error_code(0x0000) - not-present page
Dec 23 12:02:29 brain kernel: PGD 103b1c067 P4D 103b1c067 PUD 0 
Dec 23 12:02:29 brain kernel: Oops: 0000 [#1] PREEMPT SMP NOPTI
Dec 23 12:02:29 brain kernel: CPU: 8 PID: 1125 Comm: udevd Tainted: P           O       6.1.64-Unraid #1
Dec 23 12:02:29 brain kernel: Hardware name: Micro-Star International Co., Ltd. MS-7D32/MAG Z690 TOMAHAWK WIFI DDR4 (MS-7D32), BIOS 1.F0 11/02/2023
Dec 23 12:02:29 brain kernel: RIP: 0010:kernfs_root+0x0/0x14
Dec 23 12:02:29 brain kernel: Code: 89 cb 44 8b 6c 24 30 e8 24 32 fb ff 48 8b bd 58 02 00 00 48 89 da 5b 48 89 c6 5d 4c 89 e1 45 89 e8 41 5c 41 5d e9 66 ff ff ff <48> 8b 47 08 48 85 c0 48 0f 45 f8 48 8b 47 50 c3 cc cc cc cc 0f 1f
Dec 23 12:02:29 brain kernel: RSP: 0018:ffffc9000096fa88 EFLAGS: 00010286
Dec 23 12:02:29 brain kernel: RAX: 0000000000000000 RBX: 0000000fffffffe0 RCX: ffffffff813f52b3
Dec 23 12:02:29 brain kernel: RDX: 0000000000000000 RSI: ffffffff81e987e8 RDI: 0000000fffffffe0
Dec 23 12:02:29 brain kernel: RBP: ffffffff81e986a0 R08: 0000000000000000 R09: ffffffff829513f0
Dec 23 12:02:29 brain kernel: R10: 00003fffffffffff R11: fefefefefefefeff R12: ffff888107c25848
Dec 23 12:02:29 brain kernel: R13: ffff8881011be000 R14: ffff888103809b50 R15: ffff8881011be0d0
Dec 23 12:02:29 brain kernel: FS:  000014e276d8e240(0000) GS:ffff889fff600000(0000) knlGS:0000000000000000
Dec 23 12:02:29 brain kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Dec 23 12:02:29 brain kernel: CR2: 0000000fffffffe8 CR3: 0000000103992000 CR4: 0000000000750ee0
Dec 23 12:02:29 brain kernel: PKRU: 55555554
Dec 23 12:02:29 brain kernel: Call Trace:
Dec 23 12:02:29 brain kernel: <TASK>
Dec 23 12:02:29 brain kernel: ? __die_body+0x1a/0x5c
Dec 23 12:02:29 brain kernel: ? page_fault_oops+0x329/0x376
Dec 23 12:02:29 brain kernel: ? do_user_addr_fault+0x12e/0x48d
Dec 23 12:02:29 brain kernel: ? exc_page_fault+0xfb/0x11d
Dec 23 12:02:29 brain kernel: ? asm_exc_page_fault+0x22/0x30
Dec 23 12:02:29 brain kernel: ? __blk_mq_run_hw_queue+0x73/0x73
Dec 23 12:02:29 brain kernel: ? kernfs_vfs_xattr_set+0x41/0x41
Dec 23 12:02:29 brain kernel: kernfs_find_and_get_ns+0x1c/0x5c
Dec 23 12:02:29 brain kernel: sysfs_unmerge_group+0x16/0x4d
Dec 23 12:02:29 brain kernel: dpm_sysfs_remove+0x1e/0x52
Dec 23 12:02:29 brain kernel: device_del+0xa4/0x31d
Dec 23 12:02:29 brain kernel: ? i915_ggtt_probe_hw+0x593/0x5be [i915]
Dec 23 12:02:29 brain kernel: platform_device_del+0x21/0x70
Dec 23 12:02:29 brain kernel: platform_device_unregister+0xf/0x19
Dec 23 12:02:29 brain kernel: sysfb_disable+0x2b/0x54
Dec 23 12:02:29 brain kernel: aperture_remove_conflicting_pci_devices+0x1e/0x82
Dec 23 12:02:29 brain kernel: i915_driver_probe+0x83f/0xc19 [i915]
Dec 23 12:02:29 brain kernel: ? slab_free_freelist_hook.constprop.0+0x3b/0xaf
Dec 23 12:02:29 brain kernel: local_pci_probe+0x3d/0x81
Dec 23 12:02:29 brain kernel: pci_device_probe+0x197/0x1eb
Dec 23 12:02:29 brain kernel: ? sysfs_do_create_link_sd+0x71/0xb7
Dec 23 12:02:29 brain kernel: really_probe+0x115/0x282
Dec 23 12:02:29 brain kernel: __driver_probe_device+0xc0/0xf2
Dec 23 12:02:29 brain kernel: driver_probe_device+0x1f/0x77
Dec 23 12:02:29 brain kernel: ? __device_attach_driver+0x97/0x97
Dec 23 12:02:29 brain kernel: __driver_attach+0xd7/0xee
Dec 23 12:02:29 brain kernel: ? __device_attach_driver+0x97/0x97
Dec 23 12:02:29 brain kernel: bus_for_each_dev+0x6e/0xa7
Dec 23 12:02:29 brain kernel: bus_add_driver+0xd8/0x1d0
Dec 23 12:02:29 brain kernel: driver_register+0x99/0xd7
Dec 23 12:02:29 brain kernel: i915_init+0x1f/0x7f [i915]
Dec 23 12:02:29 brain kernel: ? 0xffffffffa0ab7000
Dec 23 12:02:29 brain kernel: do_one_initcall+0x82/0x19f
Dec 23 12:02:29 brain kernel: ? kmalloc_trace+0x43/0x52
Dec 23 12:02:29 brain kernel: do_init_module+0x4b/0x1d4
Dec 23 12:02:29 brain kernel: __do_sys_init_module+0xb6/0xf9
Dec 23 12:02:29 brain kernel: do_syscall_64+0x68/0x81
Dec 23 12:02:29 brain kernel: entry_SYSCALL_64_after_hwframe+0x64/0xce
Dec 23 12:02:29 brain kernel: RIP: 0033:0x14e2772a0dfa
Dec 23 12:02:29 brain kernel: Code: 48 8b 0d 21 20 0d 00 f7 d8 64 89 01 48 83 c8 ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 49 89 ca b8 af 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d ee 1f 0d 00 f7 d8 64 89 01 48
Dec 23 12:02:29 brain kernel: RSP: 002b:00007ffdde819638 EFLAGS: 00000246 ORIG_RAX: 00000000000000af
Dec 23 12:02:29 brain kernel: RAX: ffffffffffffffda RBX: 0000000000468c20 RCX: 000014e2772a0dfa
Dec 23 12:02:29 brain kernel: RDX: 000014e277395aad RSI: 00000000004b1868 RDI: 000014e276476010
Dec 23 12:02:29 brain kernel: RBP: 000014e277395aad R08: 0000000000000007 R09: 0000000000464e30
Dec 23 12:02:29 brain kernel: R10: 0000000000000005 R11: 0000000000000246 R12: 000014e276476010
Dec 23 12:02:29 brain kernel: R13: 0000000000000000 R14: 0000000000468600 R15: 0000000000000000
Dec 23 12:02:29 brain kernel: </TASK>
Dec 23 12:02:29 brain kernel: Modules linked in: amdgpu zfs(PO+) i915(+) zunicode(PO) zzstd(O) intel_rapl_msr intel_rapl_common x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel zlua(O) zavl(PO) icp(PO) kvm gpu_sched drm_ttm_helper iosf_mbi drm_buddy i2c_algo_bit ttm crct10dif_pclmul crc32_pclmul zcommon(PO) crc32c_intel drm_display_helper ghash_clmulni_intel sha512_ssse3 sha256_ssse3 btusb sha1_ssse3 znvpair(PO) btrtl drm_kms_helper mei_hdcp aesni_intel intel_gtt mei_pxp btbcm crypto_simd cryptd spl(O) drm wmi_bmof mxm_wmi rapl btintel intel_cstate bluetooth intel_uncore mpt3sas(+) nvme agpgart i2c_i801 mei_me syscopyarea i2c_smbus ahci sysfillrect raid_class ecdh_generic sysimgblt video tpm_crb igc i2c_core mei nvme_core scsi_transport_sas joydev libahci ecc fb_sys_fops thermal fan tpm_tis tpm_tis_core tpm wmi intel_pmc_core backlight acpi_pad acpi_tad button unix
Dec 23 12:02:29 brain kernel: CR2: 0000000fffffffe8
Dec 23 12:02:29 brain kernel: ---[ end trace 0000000000000000 ]---
Dec 23 12:02:29 brain kernel: RIP: 0010:kernfs_root+0x0/0x14
Dec 23 12:02:29 brain kernel: Code: 89 cb 44 8b 6c 24 30 e8 24 32 fb ff 48 8b bd 58 02 00 00 48 89 da 5b 48 89 c6 5d 4c 89 e1 45 89 e8 41 5c 41 5d e9 66 ff ff ff <48> 8b 47 08 48 85 c0 48 0f 45 f8 48 8b 47 50 c3 cc cc cc cc 0f 1f
Dec 23 12:02:29 brain kernel: RSP: 0018:ffffc9000096fa88 EFLAGS: 00010286
Dec 23 12:02:29 brain kernel: RAX: 0000000000000000 RBX: 0000000fffffffe0 RCX: ffffffff813f52b3
Dec 23 12:02:29 brain kernel: RDX: 0000000000000000 RSI: ffffffff81e987e8 RDI: 0000000fffffffe0
Dec 23 12:02:29 brain kernel: RBP: ffffffff81e986a0 R08: 0000000000000000 R09: ffffffff829513f0
Dec 23 12:02:29 brain kernel: R10: 00003fffffffffff R11: fefefefefefefeff R12: ffff888107c25848
Dec 23 12:02:29 brain kernel: R13: ffff8881011be000 R14: ffff888103809b50 R15: ffff8881011be0d0
Dec 23 12:02:29 brain kernel: FS:  000014e276d8e240(0000) GS:ffff889fff600000(0000) knlGS:0000000000000000
Dec 23 12:02:29 brain kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Dec 23 12:02:29 brain kernel: CR2: 0000000fffffffe8 CR3: 0000000103992000 CR4: 0000000000750ee0
Dec 23 12:02:29 brain kernel: PKRU: 55555554

 

This usually indicated a bug in the module itself and nothing I can do about, however it seems that everything is now working correct?

However it should be not necessary to unbind you AMD card from VFIO.

  • Like 1
Link to comment
Guest
This topic is now closed to further replies.
×
×
  • Create New...