[vfio-users] Windows 10 Guest Boots 1 in 5 times

Did you tried to disable Hibernate? I recall reading that there were issues with it and Windows 10 inside a VM. Try with this: Right Click the Windows Menu Open a Command Prompt (Admin) powercfg.exe /hibernate off ___ vfio-users mailing list vfio-user

[vfio-users] My guide for installing Arch Linux + QEMU-KVM-VFIO

Around 2 years ago I did a guide for installing Arch Linux with Xen to do VGA Passthrough. I think not more than 3 or 4 people actually used it while it was up-to-date, then it got completely obsolete as KVM-VFIO outclassed Xen for VGA Passthrough. I decided to overhaul my old guide (Which is cu

Re: [vfio-users] Running 2 VMs with assigned GPUs causes both GPUs to crash, unless all virtual disks exist on the same physicakl media as /

Do you know if that issue is reproducible in other systems? The Nehalem generation Chipsets had issues with Interrupt Remapping, so when I see people with these systems having issues when using the IOMMU, I would point my finger there first. You may want to try disabling Interrupt Remapping for

Re: [vfio-users] Mediated device interface + nvidia

You WILL see it on consumer grade GPUs on a massive scale, as any Intel platform with a Broadwell or Skylake Processor with IGP will have access to Intel GVT-g. While the raw power of the Intel IGPs is a fraction of a bigger Radeon or GeForce, they are rather on par with features, and as long th

Re: [vfio-users] Mediated device interface + nvidia

> This guide still indicates support for 4th generation (aka > Haswell), but I don't know how accurate it is: > > https://github.com/01org/gvt-linux/wiki/GVTg_Setup_Guide Probabily a typo or oversight. Here is a quote straight from one of the Intel devs from less than a month ago on iGVT-g Mailin

[vfio-users] Intel LGA 1151 Kaby Lake Chipsets and PCIe ACS

As everyone knows by now, the Skylake refresh, Kaby Lake, was released a few days ago as the new 7th gen Core Processors. This includes the Union Point Chipsets (200 Series) for LGA 1151, which are the B250, Q250, Q270, H270, Z270, and possibily at least one more Chipset for Xeons E3 1200v6, whi

[vfio-users] Lots of info for AMD Ryzen early adopters

I spend the last days trying to get as much info as I could about Ryzen from authoritative sources. Reelevant for us (Virtualization and Passthrough) are the followings: 1) GENERAL RYZEN SUPPORT IN LINUX You need a really recent Linux Kernel to get Ryzen to work at all. Popular mainstream Li

[vfio-users] Some questions for passthrough gpu card to windows 7 32bit vm

> Hi, > I'm trying to use PCI passthrough to give an NVIDA GPU to a windows vm > with qemu-kvm. I have a Windows 7 64bit VM working well under Fedora > 25 with an Nvidia GPU GTX 750Ti. But when I tried Windows 7 32bit, the > OVMF UEFI could not load the Windows ISO. Main issue: No 32 Bits Windows

Re: [vfio-users] Blank output on passed card. i915 host gtx1060 vm

I had a weird bug with my Radeon 5770 where OVMF would display either a black screen or totally garbage output if I had less than 2 GiB in the VM. You're using just 1, try with 2. Try also with Chipset Q35 instead of 440FX (-M q35). Also, for correctness sake, add a PCIe Root Port and attach th

Re: [vfio-users] NVIDIA driver cause to black screen and can't launch desktop environment

>> I run QEMU with the following command: >> -device vfio-pci,host=01:00.0,x-vga=on -device vfio-pci,host=01:00.1 You are missing the PCIe Root Bridge, and attaching the GPU to it. Linux is picky about that. Try with this: -device ioh3420,chassis=0,id=testbus \ -device vfio-pci,host=01:00.0,

Re: [vfio-users] Linux iommu issue with peer-to-peer dma transfers between NVidia GTX 1080s

I recall that there were some issues (Not necessarily yours) with P2P in nVidia cards related to ACS. While anyone that cares about IOMMU Groups will love ACS due to the better granularity, it broke what you want to use, and you may want to check if you can disable it. Check these: https://www

[vfio-users] Linux iommu issue with peer-to-peer dma transfers between NVidia GTX 1080s

I forgot to add this Thread, it has far more info regarding the nVidia cards P2P issue when ACS is enabled: https://devtalk.nvidia.com/default/topic/883054/cuda-programming-and-performance/multi-gpu-peer-to-peer-access-failing-on-tesla-k80-/1

[vfio-users] UEFI adoption in GPU oproms [was: passthrough not working.]

It looks like no one here heared about updGOP. I posted it a few times here already... http://www.win-raid.com/t892f16-AMD-and-Nvidia-GOP-update-No-requests-DIY.html Make your own UEFI GOP ROM. The tool was intended to be used to create ROMs for flashing the cards themselves, but it works even b

[vfio-users] q35, pcie 1x PCH, IOMMU confusion and freezing-host with no error log

You are not telling enough info. Things that would be useful: Your Motherboard and Processor If the FireWire card is a Motherboard onboard Device or a discrete card, and which it is Full lspci -nn and -tvnn and IOMMU Groups, for completion sake Full VM launch settings. What else is in that pve-q3

Re: [vfio-users] GPU driver crashes when running a second VM if either VM has a virtual disk stored on physical media other than the root disk. Tested on three X58 chipset MBs

The Nehalem era X58, 5500 and 5520 Chipsets had a notoriously broken Interrupt Remapping implementation: https://support.citrix.com/article/CTX136517 (Love that symptoms list) https://www.netiq.com/support/kb/doc.php?id=7014344 https://serverfault.com/questions/745593/does-disabling-vt-d-and-inter

Re: [vfio-users] ACS and IOMMU groups on Skylake/Sunrise

Wasn't there a year ago or two a guy that had a Skylake Motherboard from the Gigabyte brand that had all the Chipset PCIe Slots grouped together because the Firmware didn't exposed the ACS Control Bits or something along those lines? It was this one: https://www.redhat.com/archives/vfio-users/20

Re: [vfio-users] Not enough IRQs in 82093AA IOAPIC

Note that the IOAPIC being used depends on the Chipset. 440FX may use an optional 82093AA IOAPIC, which is what is in the PDF that you linked earlier (It used to be a standalone chip and usually only found on Dual Processor Motherboards. IOAPIC was extremely rare on consumer 440FX Single Process

Re: [vfio-users] Not enough IRQs in 82093AA IOAPIC

I have something that could be useful for you, albeit chances are that it is a bit too little too late: https://habr.com/ru/post/501912/ I think that everything related to PCI Interrupts gets rather chaotic when you get to PCI Express. As PCIe is a superset of PCI that is backwards compatibl

[vfio-users] An epic wall of text with questions and comments from a Xen user

Hello. I'm replying your mail with comments: > The UEFI GOP needs some minimal ROM capacity in order for the manufacturer to > provide it for us. I found that also when I was browsing for the UEFI for my > friend’s MSI GTX 560 Ti Lightning. Based on random comments found in Google, I don't exp

[vfio-users] USB enumeration (uPD720201)

Its extremely probable than your issue is unrelated to VFIO, Passthrough, or whatever. Besides than that USB Controller was one of the worst of the lot according to some googling, there are also fake/counterfeit versions of it of even lower quality that had multitude of issues: http://club.dx.co

[vfio-users] usb does not wake up after G1/S3 standby

What Hardware it is? Haswell Rev. C1 Chipsets were know to have a USB S3 resume bug, so if you have a LGA 1150 based system, its probable that you are affected. ___ vfio-users mailing list vfio-users@redhat.com

[vfio-users] Supermicro C7Z170-OCE, a rather interesing Skylake Motherboard for VFIO if PEX8747 downstream ACS works

For your information, here is a Motherboard that has a rather unique feature. The Supermicro C7Z170-OCE supports Skylake and has all the 16 PCIe Lanes coming from the Processor PCIe Controller going through the included PLX PEX8747 chip, as you can see in the Manual at Page 23 / 1 - 9 : http://w

Re: [vfio-users] Supermicro C7Z170-OCE, a rather interesing Skylake Motherboard for VFIO if PEX8747 downstream ACS works

So basically, I would like someone to confirm if PEX8747 ACS support works downstream, even on Intel consumer platforms. No. Whether or not a PCIe switch supports ACS is irrelevant when it's placed in a configuration downstream of a PCIe root port th

[vfio-users] Some questions regarding VGA Arbitration, BIOS and UEFI Boot

I'm sure I asked this before and was expecting AW tehnical response, but instead he said "TL;DR" since it was in a ridiculous long mail with a multitude of other questions. But now I need a response since I plan to do this next week, so I have to try again. I have readed this numerous times an

[vfio-users] Some questions regarding VGA Arbitration, BIOS and UEFI Boot

Switched from plain text to enriched text format. I ALWAYS forget when mailing here that the archive does not automatically warp lines if using plain text. I'm sure I asked this before and was expecting AW tehnical response, but instead he said "TL;DR" since

Re: [vfio-users] Some questions regarding VGA Arbitration, BIOS and UEFI Boot

> I think I was in a similar situation before. I am using Haswell CPU > with its integrated graphics for host OS booted on UEFI board. I have > VGA arbitration disabled in the kernel. I am using a Radeon 7870 to > passthrough to the guest, and this card did not have UEFI firmware > available. I obt

[vfio-users] Tianocore Splash Screen Not Displayed on Passed-through GPU.

What OS do you have installed, and at what point do you see something on screen? If you only see when Windows starts, it would look like if it was running in Secondary VGA Passthrough mode, since it has to wait for the OS Drivers to get the Video Card running instead of loading the PCI Option RO

Re: [vfio-users] Some questions regarding VGA Arbitration, BIOS and UEFI Boot

Great answers, as expected. Some smaller miniquestions/comments followups >> 1) The host using UEFI or legacy BIOS and VGA support is irrelevant. In an >> ideal world, maybe this wouldn't be the case, but Xorg wants exclusive >> access to VGA even though it probably

[vfio-users] Disk Performance

I spend a lot of time experimenting what works and what doesn't, through I never benchmarked anything to see performance differences.The main issue is that QEMU has a load of old syntaxis and parameters mixed with new ones that you must use to get the latest features, and documentation is extrem

[vfio-users] Disk Performance‏

Errata: If the SCSI Controller has id=scsi0, the Bus name is scsi0.0 This works: -drive if=none,id=drive0,format=raw,file=/dev/vg0/w10x64-device virtio-scsi-pci,id=scsi0-device scsi-hd,drive=drive0,bus=scsi0.0 (Or you can omit bus=)__

Re: [vfio-users] Disk Performance

o-blk-pci,drive=ssd2,request-merging=on,iothread=iothread2,modern-pio-notify=on,config-wce=off On Sat, Apr 2, 2016 at 9:56 PM, Zir Blazer wrote: I spend a lot of time experimenting what works and what doesn't, through I never benchmarked anything to see performance differences.The main i

Re: [vfio-users] GPU Pass-through for Linux guest

If I recall correctly, doing VGA Passthrough in Linux presented different issues that those that you're used to in Windows. The major difference was that Windows and the Video Card Drivers doesn't mind if you do something like attaching a 2014 PCI Express Video Card directly to the 1996 440FX PC

[vfio-users] Radeon 5770 Passthrough works in Chipset 440FX but not in Q35

Finally I had both the time and will to start migrating from Xen to standalone QEMU. I want to install everything (Both host OS and guest OS) from scratch in my already-not-so-new SSD. Sadly, I hitted a hard wall, and after a lot of hours testing different options, I was out of ideas. Just me

[vfio-users] Why does the audio from VM crackle?

@Arjen What you suggest is something that previously happened to me in Xen around two years ago. Xen 4.3 worked fine with PCI and VGA Passthrough, but when 4.4 was released, they somehow wrecked it. The Passthroughed Sound Card audio was horrendous, and the Video Card was a no-go - good thing is

[vfio-users] Intel x2APIC and APIC Virtualization (APICv or vAPIC)

With Ivy Bridge-E, Intel introduced their APIC Virtualization extension, APICv (Rather contradictory, on other sources its called vAPIC). APICv seems to be built on top of x2APIC, so it requires it to be working on the host to be used. You can check that in the host by using: dmesg | grep x2apic

Re: [vfio-users] Intel x2APIC and APIC Virtualization (APICv or vAPIC)

> I'd say that at least some consumer product support apicv, as it happens with > PCIe ACS probably. My i7-5930K reports it HAS support for it, and it is > enabled by default. LGA 2011-3 Processors like your Ci7 5930K are Haswell-E, NOT standard Haswell, they are not the same. There may be Core

Re: [vfio-users] Radeon 5770 Passthrough works in Chipset 440FX but not in Q35

Allright, I have a lot more info since I spend several hours gathering logs. At the end, my logging style became ridiculous messy since I couldn't keep track of several variables and settings in each case plus some edge cases, so isn't 100% scientific, but I got general info about what happens a

Re: [vfio-users] Radeon 5770 Passthrough works in Chipset 440FX but not in Q35

Some more testing... If instead of... -cpu host -and some other parameters like the HyperV Enlightnements- I use... -cpu qemu64 There are no MSR errors in Windows 10. This includes Passthrough with 440FX. It STILL doesn't work in Q35.I also tested that with GParted. Instead of the multiple mentio

[vfio-users] Problems with R9 290 audio

The ioh3420 is named root.1. You have the Radeon (01:00.0) attached to it, but the HDMI (01:00.1) is attached to pcie.0, which is the default name of the Q35 Bus. Put it at root.1 too.Also, the addr for the HDMI has to be at Function 1 of the Radeon. -device ioh3420,bus=pcie.0,addr=1c.0,multifu

[vfio-users] Binding Radeon R9 380X 4gb to the vfio driver hangs system on recent kernels.

If it hard freezes when executing the vfio bind script for the Radeon, its probable than the crash happens either when it gets unbinded from the radeon Driver, or attached to vfio-pci in a dirty state. I don't see it blacklisted anywhere, so try doing so so it doesn't get initialized and try aga