Skip to content
View in the app

A better way to browse. Learn more.

Unraid

A full-screen app on your home screen with push notifications, badges and more.

To install this app on iOS and iPadOS
  1. Tap the Share icon in Safari
  2. Scroll the menu and tap Add to Home Screen.
  3. Tap Add in the top-right corner.
To install this app on Android
  1. Tap the 3-dot menu (⋮) in the top-right corner of the browser.
  2. Tap Add to Home screen or Install app.
  3. Confirm by tapping Install.

Unable to start virtual machines - Operation not permitted

Featured Replies

Goodday all,

 

my name is Jim and I am new to the Lime Technology. I am excited to start with this great product.

 

Plz if this is already solved of this is the wrong thread plz tell so and point me out

 

I have a problem starting VM’s that are created through the Web gui.

 

Let me first list down my setup:

 

Hardware:

Mobo Asus Rampage III Extreme (both

Mem         12GB (3 x 4GB)

CPU          i7-970 hexacore

GFX         AMD HD 6970 x 3

HDD1 OCZ Revo 120GB PCI-E

HDD2 WD Green 1TB

BOOT USB 16GB

 

unRaid config:

No parity

Single disk raid

1 share bridged br0

 

Software:

unRaid v6

Trial license (want to try it first ofc)

 

VM:

Windows 8.1

Installed through VNC

 

After we created a  virtual machine and we want to start the machine we get the following error.

 

vfio: failed to set iommu for container: Operation not permitted

vfio: failed to setup container for group # (depends on which VGA we passtrough but always same error)

vfio: failed to get group #

Device initialization failed

Device 'vfio-pci' could not be initialized

 

If we choose an other GFX card the message is the same execpt for the different device number (GFX).

 

I think it's has something to do with the GFX card. I am able to install and start a virtual machine with VNC mode but I am not ablo to start it with direct i/o.

 

Did anybody else ran into this problem? Is it my hardware configuration? What is the minimum setup for two GFX cards?

 

Thank you in advanced, I realy would like to start using this awesome software!

 

Greetings,

 

Jim

 

  • Author

Thank you CHBMB, I will let you know today if it's a succes. I think the extra article of Alex Williamson will help a lot.

  • Author

I tried the ACS on and off... It didn't make a different. The bios change from SeaBIOS to OFMV didn't work either. Adjusting machine type from i440fx to Q35 also to no avail.

And the manual ROM files also didn't work.

 

In the syslog I get the message "BIOS IS BROKEN"... below a excerpt of the error:

 

Jan 20 19:28:52 Tower kernel: WARNING: CPU: 6 PID: 1 at drivers/iommu/intel-iommu.c:4744 init_dmars+0x258/0x454()

Jan 20 19:28:52 Tower kernel: Your BIOS is broken; DMA routed to ISOCH DMAR unit but no TLB space.

Jan 20 19:28:52 Tower kernel: BIOS vendor: American Megatrends Inc.; Ver: 1502  ; Product Version: System Version

Jan 20 19:28:52 Tower kernel: Modules linked in:

Jan 20 19:28:52 Tower kernel: CPU: 6 PID: 1 Comm: swapper/0 Tainted: G          I    4.1.15-unRAID #1

Jan 20 19:28:52 Tower kernel: Hardware name: System manufacturer System Product Name/Rampage III Extreme, BIOS 1502    10/03/2011

Jan 20 19:28:52 Tower kernel: 0000000000000009 ffff880351137d38 ffffffff815f1ad0 0000000000001d2c

Jan 20 19:28:52 Tower kernel: ffff880351137d88 ffff880351137d78 ffffffff8104775b ffff88035093c400

Jan 20 19:28:52 Tower kernel: ffffffff818d6767 ffff880350ece000 0000000000000000 ffff88034d636180

Jan 20 19:28:52 Tower kernel: Call Trace:

Jan 20 19:28:52 Tower kernel: [<ffffffff815f1ad0>] dump_stack+0x4c/0x6e

Jan 20 19:28:52 Tower kernel: [<ffffffff8104775b>] warn_slowpath_common+0x97/0xb1

Jan 20 19:28:52 Tower kernel: [<ffffffff818d6767>] ? init_dmars+0x258/0x454

Jan 20 19:28:52 Tower kernel: [<ffffffff810477b6>] warn_slowpath_fmt+0x41/0x43

Jan 20 19:28:52 Tower kernel: [<ffffffff818d6767>] init_dmars+0x258/0x454

Jan 20 19:28:52 Tower kernel: [<ffffffff818d6bf8>] intel_iommu_init+0x295/0x585

Jan 20 19:28:52 Tower kernel: [<ffffffff818ac2a8>] pci_iommu_init+0x13/0x3e

Jan 20 19:28:52 Tower kernel: [<ffffffff818ac295>] ? memblock_find_dma_reserve+0x146/0x146

Jan 20 19:28:52 Tower kernel: [<ffffffff81002145>] do_one_initcall+0xe9/0x178

Jan 20 19:28:52 Tower kernel: [<ffffffff818a405d>] kernel_init_freeable+0x17c/0x204

Jan 20 19:28:52 Tower kernel: [<ffffffff818a3807>] ? initcall_blacklist+0xa5/0xa5

Jan 20 19:28:52 Tower kernel: [<ffffffff815e4da5>] ? rest_init+0x89/0x89

Jan 20 19:28:52 Tower kernel: [<ffffffff815e4dae>] kernel_init+0x9/0xd5

Jan 20 19:28:52 Tower kernel: [<ffffffff815f75d2>] ret_from_fork+0x42/0x70

Jan 20 19:28:52 Tower kernel: [<ffffffff815e4da5>] ? rest_init+0x89/0x89

Jan 20 19:28:52 Tower kernel: ---[ end trace 2470d3252461daf9 ]---

Jan 20 19:28:52 Tower kernel: IOMMU: hardware identity mapping for device 0000:00:1b.0

Jan 20 19:28:52 Tower kernel: IOMMU: Setting RMRR:

Jan 20 19:28:52 Tower kernel: IOMMU: Setting identity map for device 0000:00:1a.0 [0x9f7dc000 - 0x9f7dbfff]

Jan 20 19:28:52 Tower kernel: ------------[ cut here ]------------

 

 

I hope to solve this puzzle....

 

Greetings ,

 

Jim

 

In the syslog I get the message "BIOS IS BROKEN"... below a excerpt of the error:

 

 

I've seen a post with this in before as well, but can't remember where, and am pretty sure jonp ended up commenting.  Not terribly helpful I know, but do some searching and hopefully it'll turn up..

I've seen a post with this in before as well, but can't remember where, and am pretty sure jonp ended up commenting.  Not terribly helpful I know, but do some searching and hopefully it'll turn up..

 

I'm bored, :P here http://lime-technology.com/forum/index.php?topic=45530.msg435107#msg435107

 

Jim beat you to it...  ;D

 

I just refreshed and seen that.. Stupid work!  >:(

I wonder, have you tried with just two graphics cards, and perhaps with the second graphics card in the slot farthest from the first? Just a hunch.

I wonder, have you tried with just two graphics cards, and perhaps with the second graphics card in the slot farthest from the first? Just a hunch.

 

As far as I'm aware, passthrough of two gfx cards isn't supported, although may be in the future? 

 

I'm of the distinct opinion that despite LinusTechTips and all the other "Wow, this is amazing and easy" videos out there, that virtualisation on Unraid can be problematic and nothing has yet convinced me that a large source of the problems is that some hardware just doesn't play nice.....

His motherboard doesn't have onboard video, so one graphics card would be for UnRaid, whilst the second would be for his virtual machine. In his hardware summary he lists his graphics card x 3 which suggests three graphics cards and an attempt to passthrough two, which as you say, is not supported, hence my suggestion of removing one graphics card and trying with only two.

  • Author

Hey we made some progres.

atm we are using 2 6970 and a lame old x1500 (X1500 for unraid and 1 for a vm win 8)

We got it to work with following sysconfig:

 

default /syslinux/menu.c32

menu title Lime Technology

prompt 0

timeout 50

label unRAID OS

  menu default

  kernel /bzimage

  append vfio_iommu_type1.allow_unsafe_interrupts=1 iommu=pt initrd=/bzroot

label unRAID OS Safe Mode (no plugins)

  kernel /bzimage

  append initrd=/bzroot unraidsafemode

label Memtest86+

  kernel /memtest

 

We are going to test later if the unsafe interupts is needed or not.

His motherboard doesn't have onboard video, so one graphics card would be for UnRaid, whilst the second would be for his virtual machine. In his hardware summary he lists his graphics card x 3 which suggests three graphics cards and an attempt to passthrough two, which as you say, is not supported, hence my suggestion of removing one graphics card and trying with only two.

 

Ah, gotcha...

Glad to hear you are making progress.

  • Author

We dont know exactly why but after some testing.

This config works:

 

default /syslinux/menu.c32

menu title Lime Technology

prompt 0

timeout 50

label unRAID OS

  menu default

  kernel /bzimage

  append vfio_iommu_type1.allow_unsafe_interrupts=1 initrd=/bzroot

label unRAID OS Safe Mode (no plugins)

  kernel /bzimage

  append iommu=pt initrd=/bzroot

  append initrd=/bzroot unraidsafemode

label Memtest86+

  kernel /memtest

  • Author

Well we got it working and it works pretty wel.

 

What I notice is that when you do a GFX benchmark on two VM's the score is lower then doing the same on one VM.

 

I thought the scores would probably end near to eachother because I thought the systems (GFX's) are working alone. So how come my GFX score is about 50% when doing a single VM GFX benchmark?  I understand that there is always some lost in performance as you share the lanes on the motherboard and 1 CPU.

 

Is this maybe a PCI-X lane issue? Or is it normal behaviour?

 

Thank you in advanced. The config files an screenshots will follow this weekend for other users to use.

  • 2 weeks later...

Archived

This topic is now archived and is closed to further replies.

Account

Navigation

Search

Search

Configure browser push notifications

Chrome (Android)
  1. Tap the lock icon next to the address bar.
  2. Tap Permissions → Notifications.
  3. Adjust your preference.
Chrome (Desktop)
  1. Click the padlock icon in the address bar.
  2. Select Site settings.
  3. Find Notifications and adjust your preference.