Passthrough working on Ubuntu but not on Win10


fxf

Recommended Posts

Hi,
I'm trailing unraid for private use on my desktop/server.

 

So far it's running, but the pass-through is almost not working. I'm writing almost, since I have pass-through for 1 minute when trying an Ubunt installation. Windows 10 is not working at all (black screen).

Tried the generic Openelec with no luck, as well.

Is this commonly seen, that some VM's will pass-through while others wont, with same settings/GPU?

 

BR fx

Edited by fxf
Link to comment

Well, considering it's a very general question and not (yet) a technically question, so the lack of technical information should make sense.

You try to force me to jump directly to error-tracing, before even establishing whether it's a general issue or a specific case.

 

 

Link to comment
1 hour ago, fxf said:

Well, considering it's a very general question and not (yet) a technically question, so the lack of technical information should make sense.

You try to force me to jump directly to error-tracing, before even establishing whether it's a general issue or a specific case.

 

 

 

and yet, here you are with no closer explanation for your problem. this path you've chosen is working out quite well for you.

 

the reason for asking for diagnostics is to give the ability for others on this forum to provide you with as specific a resolution as possible on the first attempt, and not after 47 guesses wasting your time and ours. 

 

But if you'd like a guess about the problem based on what you gave us, here are a few: bad gpu, bad wiring, mouse in the computer eating your wires, misconfigured syslinux.cfg, hardware error, lack of iommu support, lack of virt support, monitor power cable not plugged in, one-D-ten-T user, gpu not seated properly, monitor display cable not plugged in, hdmi cable plugged in wrong port on gpu, bad cable from server to monitor, ghost in the machine, bad install on multiple vms, misconfiguration of drivers for windows, gpu too old, motherboard too old, cpu too old, misconfigured ovmf, or other user error yet to be determined at this time due to lack of proper diagnostic files.

 

start with those. 

  • Like 1
Link to comment
3 hours ago, fxf said:

Well, considering it's a very general question and not (yet) a technically question, so the lack of technical information should make sense.

You try to force me to jump directly to error-tracing, before even establishing whether it's a general issue or a specific case.

 

 

I'll be frank. Your attitude for sure won't get you any help.

Link to comment
5 hours ago, fxf said:

establishing whether it's a general issue or a specific case.

If you are using a hardware list and software configuration that is working well for others, then it's a specific case. If your hardware list is not known to work well, it's a general issue. Since you didn't provide a comprehensive hardware and software configuration list with your original question, asking for current diagnostics provides you with an easy way of getting some of that information to us.

 

No info provided leads to no good guesses, only smart alec comments.

Link to comment

OKAY OKAY ! Sorry!

Was just wondering if there often was issues with WIN10. 

 

Since this is not on the table, I'll check my HDMI cable... since there is consensus about this being a possible cause for only Ubuntu working with pass-through, but not WIN10 :) 

 

That being said, I'll try getting the diagnostics.

Link to comment
Model: Custom
M/B: ASRock - X79 Extreme6/GB
CPU: Intel® Core™ i7-3820 CPU @ 3.60GHz
HVM: Enabled
IOMMU: Enabled
Cache: 256 kB, 1024 kB, 10240 kB
Memory: 32 GB (max. installable capacity 512 GB)
Network: bond0: fault-tolerance (active-backup), mtu 1500 
 eth0: 1000 Mb/s, full duplex, mtu 1500
Kernel: Linux 4.14.49-unRAID x86_64
Link to comment
1 hour ago, fxf said:

Hm funny it's not on the INFO list, but my GPU is a NVIDIA 1060 6GB.

1060 is fine.

 

this call trace isn't helping you out:

 

Sep 10 14:15:43 Tower kernel: vfio-pci 0000:02:00.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=io+mem:owns=io+mem
Sep 10 14:15:44 Tower kernel: irq 16: nobody cared (try booting with the "irqpoll" option)
Sep 10 14:15:44 Tower kernel: CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.14.49-unRAID #1
Sep 10 14:15:44 Tower kernel: Hardware name: To Be Filled By O.E.M. To Be Filled By O.E.M./X79 Extreme6/GB, BIOS P1.30 05/16/2012
Sep 10 14:15:44 Tower kernel: Call Trace:
Sep 10 14:15:44 Tower kernel: <IRQ>
Sep 10 14:15:44 Tower kernel: dump_stack+0x5d/0x79
Sep 10 14:15:44 Tower kernel: __report_bad_irq+0x32/0xac
Sep 10 14:15:44 Tower kernel: note_interrupt+0x1d4/0x225
Sep 10 14:15:44 Tower kernel: handle_irq_event_percpu+0x39/0x3f
Sep 10 14:15:44 Tower kernel: handle_irq_event+0x33/0x51
Sep 10 14:15:44 Tower kernel: handle_fasteoi_irq+0x8c/0xfc
Sep 10 14:15:44 Tower kernel: handle_irq+0x1c/0x1f
Sep 10 14:15:44 Tower kernel: do_IRQ+0x3b/0xbb
Sep 10 14:15:44 Tower kernel: common_interrupt+0x7d/0x7d
Sep 10 14:15:44 Tower kernel: </IRQ>
Sep 10 14:15:44 Tower kernel: RIP: 0010:poll_idle+0x4d/0x58
Sep 10 14:15:44 Tower kernel: RSP: 0018:ffffffff81c03ec0 EFLAGS: 00000246 ORIG_RAX: ffffffffffffffad
Sep 10 14:15:44 Tower kernel: RAX: 0000000080200000 RBX: 0000000000000000 RCX: 000000000000012d
Sep 10 14:15:44 Tower kernel: RDX: 0000000000000000 RSI: ffffffff81c56480 RDI: ffff88083fc28440
Sep 10 14:15:44 Tower kernel: RBP: ffff88083fc28440 R08: 0000069eb7774340 R09: 0000000000000018
Sep 10 14:15:44 Tower kernel: R10: ffffffff81c03e90 R11: 0000000000000000 R12: ffffffff81c56480
Sep 10 14:15:44 Tower kernel: R13: 0000000000000000 R14: ffffffff81c56498 R15: 0000018e873e2d67
Sep 10 14:15:44 Tower kernel: cpuidle_enter_state+0x9d/0x135
Sep 10 14:15:44 Tower kernel: do_idle+0x11a/0x179
Sep 10 14:15:44 Tower kernel: cpu_startup_entry+0x18/0x1a
Sep 10 14:15:44 Tower kernel: start_kernel+0x3dc/0x3e4
Sep 10 14:15:44 Tower kernel: secondary_startup_64+0xa5/0xb0
Sep 10 14:15:44 Tower kernel: handlers:
Sep 10 14:15:44 Tower kernel: [<ffffffff8141aae7>] usb_hcd_irq
Sep 10 14:15:44 Tower kernel: Disabling IRQ #16
Sep 10 14:16:25 Tower kernel: vfio-pci 0000:02:00.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=io+mem:owns=io+mem
Sep 10 14:16:25 Tower kernel: br0: port 2(vnet0) entered blocking state
Sep 10 14:16:25 Tower kernel: br0: port 2(vnet0) entered disabled state
Sep 10 14:16:25 Tower kernel: device vnet0 entered promiscuous mode
Sep 10 14:16:25 Tower kernel: br0: port 2(vnet0) entered blocking state
Sep 10 14:16:25 Tower kernel: br0: port 2(vnet0) entered forwarding state
Sep 10 14:16:25 Tower kernel: kvm: SMP vm created on host with unstable TSC; guest TSC will not be reliable
Sep 10 14:16:26 Tower kernel: vfio_ecap_init: 0000:02:00.0 hiding ecap 0x19@0x900
Sep 10 14:16:26 Tower acpid: input device has been disconnected, fd 8
Sep 10 14:16:26 Tower acpid: input device has been disconnected, fd 6
Sep 10 14:16:26 Tower acpid: input device has been disconnected, fd 7
Sep 10 14:16:26 Tower acpid: input device has been disconnected, fd 4

 

 

I'm not good with call traces, but I suspect that your card is conflicting with something else in the bios. You can try moving it to a different slot, or see what IRQ it's on in your bios and adjust accordingly. But before that, look into updating your bios because its from: 05/16/2012. It may help resolve several other errors that appear in your syslogs and/or resolve the issue of failure to passthrough/lockup on it's own.

 

after that, since you only have 1 gpu, you may have to follow the steps in the video below

 

Edited by 1812
Link to comment

Hm thanks!! - good point with the BIOS upgrade. I actually thought it was the latest. I'll try to flash it !

I already tried moving the GPU to an other slot, which doesn't changes anything. I haven't touched the IRQ's though... I'll try that after a BIOS upgrade.

Link to comment
1 hour ago, fxf said:

Hm thanks!! - good point with the BIOS upgrade. I actually thought it was the latest. I'll try to flash it !

I already tried moving the GPU to an other slot, which doesn't changes anything. I haven't touched the IRQ's though... I'll try that after a BIOS upgrade. 

Try booting unRAID in Legacy Mode. I have seen a post reporting UEFI mode writing to memory area that conflicts with the GPU.

Link to comment
1 hour ago, testdasi said:

Try booting unRAID in Legacy Mode. I have seen a post reporting UEFI mode writing to memory area that conflicts with the GPU.

This is interesting because I saw something similar once about a month ago when I had  uefi set in unRaid but legacy was still enabled in bios. Setting bios to uefi only fixed it. I thought it was just the quirk of one of my servers.

Link to comment

Oh the benefit of diagnostics.

  • Your motherboard is ASRock - X79 Extreme6
  • Your CPU is i7-3820
  • Your lspci says you only have 1 VGA device - that is the GTX 1060

That says you don't have on-board / integrated graphics and the GTX 1060 is your only and primary GPU. Yet I don't find any romfile (aka vbios) line in your qemu file.

 

Passing through an NVIDIA GPU as primary without vbios is a futile endeavour.

 

You may want to try this:

  • Boot into Windows barebone, run gpuz and dump a vbios
  • Watch Ed's video (aka Spaceinvader One) that @1812 replied originally and follow Ed's instructions to edit the vbios (follow it to the t, don't try to skip any step)
  • Change your VM config and use the rom file that you edited and see how it goes from there
Link to comment

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...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.