Arch Linux

aw

Member

Registered: 2013-10-04

Posts: 921

Website

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

Should I start from Part 3 and proceed from there? It will be a tad difficult since I cannot boot into a graphical environment, I think i just need to remove pci-stub from the initramfs, and boot without loading the systemd service and kernel module (pci_stub).

You could have skimmed parts 1 & 2 in the time you've been waiting for a reply...

---

http://vfio.blogspot.com
Looking for a more open forum to discuss vfio related uses?  Try https://www.redhat.com/mailman/listinfo/vfio-users

garnerlogan65

Member

Registered: 2014-11-22

Posts: 26

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

That's what I'm about to do. I got my system working again, by undoing all of my changes. Hopefully your method works.

Last edited by garnerlogan65 (2015-06-05 20:46:29)

aw

Member

Registered: 2013-10-04

Posts: 921

Website

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

That's what I'm about to do. I got my system working again, by undoing all of my changes. Hopefully your method works.

EDIT: i have an i7 4960X/Asus Rampage IV Extreme, do I need the ACS patch?

Read part 1 again

---

http://vfio.blogspot.com
Looking for a more open forum to discuss vfio related uses?  Try https://www.redhat.com/mailman/listinfo/vfio-users

garnerlogan65

Member

Registered: 2014-11-22

Posts: 26

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

It seems like it should work fine. I'll let you know what happens.

garnerlogan65

Member

Registered: 2014-11-22

Posts: 26

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

I'm on part 3. A little confused, how do I get this: options vfio-pci ids=1002:ffffffff:ffffffff:ffffffff:00030000:ffff00ff,1002:ffffffff:ffffffff:ffffffff:00040300:ffffffff,10de:ffffffff:ffffffff:ffffffff:00030000:ffff00ff,10de:ffffffff:ffffffff:ffffffff:00040300:fffffff, I know how to get the vendor IDs but I'm confused on how to get this.

aw

Member

Registered: 2013-10-04

Posts: 921

Website

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

I'm on part 3. A little confused, how do I get this: options vfio-pci ids=1002:ffffffff:ffffffff:ffffffff:00030000:ffff00ff,1002:ffffffff:ffffffff:ffffffff:00040300:ffffffff,10de:ffffffff:ffffffff:ffffffff:00030000:ffff00ff,10de:ffffffff:ffffffff:ffffffff:00040300:fffffff, I know how to get the vendor IDs but I'm confused on how to get this.

If pci-stub is a module you can run modinfo pci-stub, otherwise you can look at the code and find this:

Initial PCI IDs to add to the stub driver, format is vendor:device[:subvendor[:subdevice[:class[:class_mask]]]] and multiple comma separated entries can be specified

So vendor:device are the minimum required.  ffffffff is the kernel definition for PCI_ANY_ID, so it matches anything.  So we match any device ID, any subsystem vendor ID and device ID for AMD or Nvidia, then we go on to filter a little further with class codes for multimedia for VGA devices.  You should not need to do anything but possibly use this directly.  If you have two identical 780s with only one to be used by the guest, you probably want to concentrate on the initramfs scripts using the driver_override option.  If they're not identical (different subsystem vendors), then pci-stub is an option.

---

http://vfio.blogspot.com
Looking for a more open forum to discuss vfio related uses?  Try https://www.redhat.com/mailman/listinfo/vfio-users

garnerlogan65

Member

Registered: 2014-11-22

Posts: 26

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

They are identical. So don't use pci_stub? I still need to use vfio-pci with ids right? I don't understand all the different number types.

EDIT: Arch doesn't have 4.1 yet, so I guess I have to make do with pci-stub.

Last edited by garnerlogan65 (2015-06-05 21:35:28)

aw

Member

Registered: 2013-10-04

Posts: 921

Website

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

They are identical. So don't use pci_stub? I still need to use vfio-pci with ids right? I don't understand all the different number types.

EDIT: Arch doesn't have 4.1 yet, so I guess I have to make do with pci-stub.

Start here:

Another issue that users encounter when sequestering devices is what to do when there are multiple devices with the same vendor:device ID and some are intended to be used for the host...

That does not rely on anything new in 4.1, the driver_override interface has been available for some time.

---

http://vfio.blogspot.com
Looking for a more open forum to discuss vfio related uses?  Try https://www.redhat.com/mailman/listinfo/vfio-users

garnerlogan65

Member

Registered: 2014-11-22

Posts: 26

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

Followed the rest of part 3. Here are the contents of /etc/dracut.conf.d/local.conf (http://pastebin.com/B7TcbCeF), and my /etc/modprobe.d/local.conf (http://pastebin.com/1qCFMacS), I had to make the vfio script in the /etc/ folder since I do not have an /sbin directory for some reason. Obviously as your instructions state dracut creates "initramfs-uname -r.img," am I supposed to replace the other initramfs in my bootloader with this one? Also, upon generating the new initramfs, dracut threw some errors "Failed to install module vfio" and it also failed to install vfio_iommu_type1. Did I do something wrong here?

aw

Member

Registered: 2013-10-04

Posts: 921

Website

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

Followed the rest of part 3. Here are the contents of /etc/dracut.conf.d/local.conf (http://pastebin.com/B7TcbCeF), and my /etc/modprobe.d/local.conf (http://pastebin.com/1qCFMacS), I had to make the vfio script in the /etc/ folder since I do not have an /sbin directory for some reason. Obviously as your instructions state dracut creates "initramfs-uname -r.img," am I supposed to replace the other initramfs in my bootloader with this one? Also, upon generating the new initramfs, dracut threw some errors "Failed to install module vfio" and it also failed to install vfio_iommu_type1. Did I do something wrong here?

Yes, you're building an initramfs to replace the original.  Is vfio built as a module or static?  (grep VFIO /boot/config-`uname -r`)  I generally wouldn't recommend it static, we'll need to find some other way to get the script to run in the initramfs if it's not a module.

---

http://vfio.blogspot.com
Looking for a more open forum to discuss vfio related uses?  Try https://www.redhat.com/mailman/listinfo/vfio-users

garnerlogan65

Member

Registered: 2014-11-22

Posts: 26

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

Doing "grep VFIO" on the normal initramfs, returns no output. It is built as a module (modinfo vfio), didn't realize you were the author I literally thought the image was supposed to be called "initramfs-uname -r", I have since fixed it. Should I try booting with it?

Rommy

Member

Registered: 2014-10-17

Posts: 51

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

There are no non-working versions.  Follow the guide series in my blog below for what I think is the current best setup.

Thank you very much :-)

Bronek

Member

From: London

Registered: 2014-02-14

Posts: 119

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

In the first post of my recent how-to series I talk about the hardware I use, not that I'd recommend it, but to talk about why it works for my config - http://vfio.blogspot.com/2015/05/vfio-g … dware.html

I won't be as generous as Duelist, skip AMD platforms.  They've admittedly given up on the single core performance race, you don't know what you're getting on an APU system for isolation, and their latest performance/workstation class chipset is quickly coming up on its four year anniversary since launch.
. . .

If you really want to assign a disk controller to a VM, you're into risky territory, mostly because the cheap ones are all crap.  If you try to assign an integrated controller, the boot ROM for it may be integrated into the system BIOS, so you'll have no way to boot the VM unless you can extract it... and hope it supports UEFI if you're going the OVMF route with the VM.  You may have better luck with a discrete HBA, but you're still in the unknown/crap space.  Personally I find virtio-scsi w/ multi-queue and a fast backing setup to meet my needs.

Sorry for slight off-topic; I'm planning to build a kvm VM running large ZFS filesystem; this filesystem works best when it has direct access to backing storage. The idea is to use vfio and pass an LSI HBA (either build into SuperMicro motherboard, or a discrete one) to VM, and attach all ZFS disks to it. Guest would be using mpt2sas driver. However for boot and root filesystem of guest I would use regular block device with virtio driver, rather than passed-through HBA. I suppose in this case there should be no need to worry about BIOS of HBA (no boot device) and my only worry should about tuning the machine for ZFS? BTW, apparently only little effort is required to use ZFS under kvm, e.g. https://github.com/zfsonlinux/zfs/issue … -106493033

emacsomancer

Member

Registered: 2014-09-20

Posts: 211

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

My relevant computer specs:

• ASRock Z97 Anniversary motherboard (manual lists it as VT-d capable)

• Intel Core i5-4460

• EVGA NVIDIA GeForce GTX 960

• 16G ram

I'm considering getting either another GTX 960 or else a GTX 970 (of some sort). (Or else getting a 720, 730, 750 for use with the Linux host, and move the GTX 960 for use by Windows]. For the purpose of running games in a Windows VM within an Arch host.

As far as I can tell all of my parts of VT-d capable/enabled. (I can enable VT-d in my motherboard bios, I noticed).   

Before I buy another GPU I wanted to check to make sure it is likely I would be able to get GPU passthrough to a Windows VM working with this setup. (Any concerns with GTX 960 or 970? or recommendations?)

[I'm assuming there's no easy way to dynamically switch from NVIDIA to integrated video to free up my NVIDIA card when I want to run Windows.]

Last edited by emacsomancer (2015-06-06 21:04:45)

lersek

Member

Registered: 2015-03-15

Posts: 38

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

... UEFI GOP can't go beyond 1024x768

This is false in general.

without device-specific drivers.

This is self-contradictory :)

How does gigabyte Z87X-UD3H output 1920x1080 in UEFI setup? Intel-specific hacks?

GOP is shorthand for GRAPHICS_OUTPUT_PROTOCOL. It is an interface (an abstraction) that is defined by the UEFI specification. UEFI fimware video drivers produce it. It can flexibly express any mode list that the actual driver decides to expose, based on video RAM and any other internal sw & hw factors. So this is why the first quote is false.

The second quote makes no sense because GOPs are exposed by, and are the standard interface of, device-specific UEFI video drivers. A UEFI GOP does not exist without a device-specific UEFI driver; at worst you can say that the actual driver is low quality or generic, because, while it recognizes the video card, it cannot utilize its full capabilities.

One nice bit about the GOP design is that the runtime OS (a UEFI OS) can inherit the linear framebuffer characteristics (start address, size, pixel format, resolution etc; just writing from memory, not checking the spec) from the GOP, and continue using the video mode that has been last selected while in UEFI. There won't be any acceleration, but you can have a nice (fixed) color depth and (fixed) resolution in the UEFI runtime OS without device-specific drivers. (In order to select a different video mode, you have to reboot into UEFI, or install the OS-level device-specific video driver.)

The three occurrences of the word "inherit" in the OVMF whitepaper all relate to this matter.

WildyLion

Member

Registered: 2015-06-07

Posts: 5

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

Hi.

I'm kinda new here, but I managed to set up KVM VGA passthrough using VFIO with Windows 8.1 as a guest. The thing is, my VM seems to have some timing issues despite all the hyper-v enlightenments turned on.
E.g. I'm able to play Witcher 3, but with post-processing settings turned down a notch from what I'm usually playing on bare metal at. I'm passing through a Xonar U7 audio card via USB controller passthrough and the sound seems to click sometimes, indicating something is wrong with the timing.
I'm also passing a USB hub containing my mouse, trackball and keyboard to the same guest using a hardware ATEN KVM device that switches the USB ports to the needed root hub.

My configuration is:

* AMD FX-8350
* ASUS M5A99FX PRO R 2.0
* 16 GB RAM
* Sapphire R9 290 Vapor-X as a passthrough card (Slot 3), hosts a Dell 2415 via DP
* ATI Radeon HD 4650 as a boot VGA (Slot 1), hosts both of my monitors (Dell 2415 and 2412)
* Boot options: pci-stub.ids=1002:67be,1002:aac8 iommu=pt iommu=1
* Using libvirt 1.2.15, qemu 2.3.0 and stock kernel (4.0.4-2 at present)
* VM config is here: http://pastebin.ca/3020403

Could anybody please give me a hint on where to look next?

Edit: Maybe I should reinstall everything yet again and use Q35? Will that solve my issues?

Last edited by WildyLion (2015-06-07 11:35:23)

jurie

Member

Registered: 2015-06-07

Posts: 1

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

Hi there,

I'm using ASRock Z97 Extreme6 and a Intel i5-4690K with an old GTX 670.
I followed the steps from https://www.youtube.com/watch?v=Tq345_v2-Ew  but cant see to get it working?
My setup/ config is here http://pastebin.com/kp3ES6Sn

The virtual machine boots, but I cant see any output on my GTX 670 card?
Should I just enable i915 VGA arbiter patch as a module "i915" ?

Could someone please point me in the right direction please?
Thanks

Nesousx

Member

Registered: 2012-03-27

Posts: 46

Website

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

Hi,

I am having issues setting up sound in my Guest VM. Guest is running under Win 10, here is the "launcher":

qemu-system-x86_64 -name winten -enable-kvm -m 16384 -cpu host,kvm=off  \
-smp 4,sockets=1,cores=4,threads=1 \
-drive if=pflash,format=raw,readonly,file=/usr/share/edk2.git/ovmf-x64/OVMF_CODE.fd \
-drive if=pflash,format=raw,file=/usr/share/edk2.git/ovmf-x64/OVMF_VARS.fd \
-device vfio-pci,host=01:00.0 -device vfio-pci,host=01:00.1 -device vfio-pci,host=00:1d.0 \
-vga none \
-device virtio-scsi-pci,id=scsi \
-drive file=/dev/guest/win10,id=win,format=raw,if=none -device scsi-hd,drive=win \
-drive file=/dev/guest/win7,id=data,format=raw,if=none -device scsi-hd,drive=data \
-net nic,model=virtio,macaddr=52:54:00:7c:3c:42 -net bridge,br=br0 \
-soundhw hda \
-nographic

Everything seems OK. However, sounds is "lagging" / distorted. I have tried several stuff from first post, like :

QEMU_PA_SAMPLES=128 QEMU_AUDIO_DRV=pa qemu-system-x86_64...

.. changing sample size from 32 to 8192.. didn't help.

Please also note that I see the following message while running the launch script :

pulseaudio: pa_context_connect() failed
pulseaudio: Reason: Connection refused
pulseaudio: Failed to initialize PA contextaudio: Could not init `pa' audio driver

And :

QEMU_ALSA_DAC_BUFFER_SIZE=512 QEMU_ALSA_DAC_PERIOD_SIZE=170 QEMU_AUDIO_DRV=alsa qemu-system-x86_64...

Also changing buffer size and period size from several values.

I always end up in lagging audio.

I remember having this issue months ago (on a Win 7 host), then I fixed it, but I can't recall what I did. Maybe it was something about using pulse system wide. But I am not willing to do it again unless I am really sure it comes from here.

Lately I have been using  a USB sound card shared from Host to Guest and it worked fine. However, I had issue in the host, since I had to reselect sound card every time I plug it out / back in, which bored me.

If anyone has any insight on how to debug this, it would be gladly appreciated. Otherwise, it might be better to find a fix (such as this : http://alsa.opensrc.org/Udev, I believe) for the USB sound card on host, but this will be off topic.

Thanks a lot in advance

Last edited by Nesousx (2015-06-07 16:39:53)

supertrampx

Member

Registered: 2015-06-07

Posts: 11

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

Hello!
I have successfully gotten VGA-passthrough with qemu working on my second monitor with my 970gtx on my Sr-2 classified.
Currently I am having troubles getting qemu to recognize a boot device, I have searched for MANY hours, so perhaps someone could point me in the right direction? Perhaps I am sleepy and missed something.
Currently using this vm script. I have left the disk and cd devices undefined here, as I have no clue what is wrong and would like to start from a clear slate. QEMU boots fine on my second card, and passthrough is working.

#!/bin/bash

configfile=/etc/vfio-pci1.cfg

vfiobind() {
    dev="$1"
        vendor=$(cat /sys/bus/pci/devices/$dev/vendor)
        device=$(cat /sys/bus/pci/devices/$dev/device)
        if [ -e /sys/bus/pci/devices/$dev/driver ]; then
                echo $dev > /sys/bus/pci/devices/$dev/driver/unbind
        fi
        echo $vendor $device > /sys/bus/pci/drivers/vfio-pci/new_id

}

modprobe vfio-pci

cat $configfile | while read line;do
    echo $line | grep ^# >/dev/null 2>&1 && continue
        vfiobind $line
done

#Startup Windows 7 Guest with VGA-Passthrough
qemu-system-x86_64 -name win7 -enable-kvm -m 1024 -cpu host,kvm=off \
-smp 4,sockets=1,cores=4,threads=1 \
-bios /home/coyote/seabios/out/bios.bin \
-device vfio-pci,host=04:00.0,x-vga=on -device vfio-pci,host=04:00.1 \
-vga none \

exit 0

Nesousx

Member

Registered: 2012-03-27

Posts: 46

Website

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

Hello!
I have successfully gotten VGA-passthrough with qemu working on my second monitor with my 970gtx on my Sr-2 classified.
Currently I am having troubles getting qemu to recognize a boot device, I have searched for MANY hours, so perhaps someone could point me in the right direction? Perhaps I am sleepy and missed something.
Currently using this vm script. I have left the disk and cd devices undefined here, as I have no clue what is wrong and would like to start from a clear slate. QEMU boots fine on my second card, and passthrough is working.

#!/bin/bash

configfile=/etc/vfio-pci1.cfg

vfiobind() {
    dev="$1"
        vendor=$(cat /sys/bus/pci/devices/$dev/vendor)
        device=$(cat /sys/bus/pci/devices/$dev/device)
        if [ -e /sys/bus/pci/devices/$dev/driver ]; then
                echo $dev > /sys/bus/pci/devices/$dev/driver/unbind
        fi
        echo $vendor $device > /sys/bus/pci/drivers/vfio-pci/new_id

}

modprobe vfio-pci

cat $configfile | while read line;do
    echo $line | grep ^# >/dev/null 2>&1 && continue
        vfiobind $line
done

#Startup Windows 7 Guest with VGA-Passthrough
qemu-system-x86_64 -name win7 -enable-kvm -m 1024 -cpu host,kvm=off \
-smp 4,sockets=1,cores=4,threads=1 \
-bios /home/coyote/seabios/out/bios.bin \
-device vfio-pci,host=04:00.0,x-vga=on -device vfio-pci,host=04:00.1 \
-vga none \

exit 0

Hi,

You don't seem to be using any "hard drive" in your "starter script". You should probably have a line like :

-drive file=/dev/guest/win10,id=win,format=raw,if=none -device scsi-hd,drive=win

Have you checked Aw's Blog (http://vfio.blogspot.fr/), especially the 5 parts tutorial on VFIO GPU? Plenty of advices there.

You might also want to re-read complety first post.

EDIT : you said, that you left the CD and HDD devices undefined. But I believe we actually want to see the full config file in order to help you.

Last edited by Nesousx (2015-06-07 17:20:36)

supertrampx

Member

Registered: 2015-06-07

Posts: 11

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

"I have left the disk and cd devices undefined here, as I have no clue what is wrong"

Thanks for reply but I said this already

I will check that link, thank you, other suggestions would be great!

Nesousx

Member

Registered: 2012-03-27

Posts: 46

Website

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

"I have left the disk and cd devices undefined here, as I have no clue what is wrong"

Thanks for reply but I said this already

I will check that link, thank you, other suggestions would be great!

Check my edit.

supertrampx

Member

Registered: 2015-06-07

Posts: 11

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

Ok, well what is a very common way to set this up? Could I use an iso made from a windows cd?

Currently I have created a windows iso from the windows 7 ultimate disk. and prepared a raw windows img using the dd command.
do I need to specify a bus device? I have recieved no errors other than no bootable device found...

/home/coyote/windows7.iso
/home/coyote/win7.img
/home/coyote/virtio.iso "windows drivers"

Many thanks in advance.

Nesousx

Member

Registered: 2012-03-27

Posts: 46

Website

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

Ok, well what is a very common way to set this up? Could I use an iso made from a windows cd?

Currently I have created a windows iso from the windows 7 ultimate disk. and prepared a raw windows img using the dd command.
do I need to specify a bus device? I have recieved no errors other than no bootable device found...

/home/coyote/windows7.iso
/home/coyote/win7.img
/home/coyote/virtio.iso "windows drivers"

Many thanks in advance.

What you are posting doesn't look anything like it was suggested by OP and I have no idea how you are using this in your script.

Here is a working config, try to modify it to your needs (drives / disks part) :

qemu-system-x86_64 -name winten -enable-kvm -m 16384 -cpu host,kvm=off  \
-smp 4,sockets=1,cores=4,threads=1 \
-drive if=pflash,format=raw,readonly,file=/usr/share/edk2.git/ovmf-x64/OVMF_CODE.fd \
-drive if=pflash,format=raw,file=/usr/share/edk2.git/ovmf-x64/OVMF_VARS.fd \
-device vfio-pci,host=01:00.0 -device vfio-pci,host=01:00.1 -device vfio-pci,host=00:1d.0 \
-vga none \
-device virtio-scsi-pci,id=scsi \
-drive file=/dev/guest/win10,id=win,format=raw,if=none -device scsi-hd,drive=win \
-drive file=/dev/guest/win7,id=data,format=raw,if=none -device scsi-hd,drive=data \
-drive file=/home/nesousx/Isos/virtio-win-0.1-81.iso,id=virtiocd,if=none -device ide-cd,bus=ide.1,drive=virtiocd \
-drive file=/home/nesousx/Isos/Win_7_Pro_x64.iso,id=isocd,if=none -device scsi-cd,drive=isocd \
-net nic,model=virtio,macaddr=52:54:00:7c:3c:42 -net bridge,br=br0 \
-soundhw hda \
-nographic

First 2 drives statements are "system" and "data" drives, they are LVM volumes. Then the 2 others are iso files (one for Win 7, one for the virtio drivers). Also, if you are using a "just created" iso, make sure it is a working and really boots.

Edit : try this from your original script :

#Startup Windows 7 Guest with VGA-Passthrough
qemu-system-x86_64 -name win7 -enable-kvm -m 1024 -cpu host,kvm=off \
-smp 4,sockets=1,cores=4,threads=1 \
-bios /home/coyote/seabios/out/bios.bin \
-device vfio-pci,host=04:00.0,x-vga=on -device vfio-pci,host=04:00.1 \
-drive file=/home/coyote/win7.img,id=win,format=raw,if=none -device scsi-hd,drive=win \
-drive file=/home/coyote/virtio.iso,id=virtiocd,if=none -device ide-cd,bus=ide.1,drive=virtiocd \
-drive file=/home/coyote/windows7.iso,id=isocd,if=none -device scsi-cd,drive=isocd \
-vga none

Last edited by Nesousx (2015-06-07 21:39:40)

Duelist

Member

Registered: 2014-09-22

Posts: 358

Re: KVM VGA-Passthrough using the new vfio-vga support in kernel =>3.9

The second quote makes no sense because GOPs are exposed by, and are the standard interface of, device-specific UEFI video drivers. A UEFI GOP does not exist without a device-specific UEFI driver; at worst you can say that the actual driver is low quality or generic, because, while it recognizes the video card, it cannot utilize its full capabilities.

Who writes those UEFI drivers? Firmware-engineers(ASUS, Gigabyte, guys who produce the motherboards, etc.), vendors of the devices(EVGA, Sapphire, guys that don't produce motherboards etc.) or AMD/Intel themselves? Where they are stored? Are they hidden somewhere in the firmware(UEFI), or in the device ROM? If it's an intel iGPU, where that ROM with the device driver is located, as i can't recall intel GPUs being listed as a regular PCI device?
And..
Why does emulated QXL can't go beyond 1024x768... even with an extended amount of VRAM given?

---

The forum rules prohibit requesting support for distributions other than arch.
I gave up. It was too late.
What I was trying to do.
The reference about VFIO and KVM VGA passthrough.