AquaWolf

Solved it by using OVMF TPU Bios and reinstalling Windows.

Ok I tried the GPU Passthrough with a Ubuntu VM and there is everything working the difference between the Ubuntu and Windows is that im using in Ubuntu OVMF and for Windows SeaBios. I tried to recreate the Windows VM with a OVMF but im only getting a shell console when starting with only VNC graphics to test if the VM is booting.

Hey there, i tried the upgrade form 6.9 to 6.11 today hoping everything will go right. Actually i have now the Problem my GPU passthrough is not working anymore with my Windows VM. I had the same issue with 6.10.RC4 I think. When I'm starting the Windows VM with VNC the vm starts and boots into windows. Does anybody has an idea what could be wrong after upgrade? tower-diagnostics-20221017-1316.zip windowsvm.txt

Ok I tried to restore from the backup before Unraid upgrade to rc2. I was not able to start there the VM in beginning, but after deleting the libvirt.img repasting the xml for the vm the passthrough worked. I noticed that I don't had all thes option in the syslinux settings. Then I rerolled the backup of the rc2 and tried there also with and without all these syslinux settings but the gpu passthrough won't work. I also did a delete of libvirt img on rc2 but no success. I'm a bit out of options... perhaps i have to rollback the unraid version and try to fix the mac vlan issue of docker another way

I have a complete backup of the usb boot device, should I restore it? I'm having then again the problem with the docker net.

no sadfully it does not change it.

I have system diagnostics from before, but honestly after the VMs did not started I tried many combinations to try fixing the passthrough problem. Here is an old diagnostics file where the VMs worked with passthrough but there was an issue with docker mac-vlan what has been solved now by ip-vlan implementation. I also tried rollback to the old Unraid version but there the passthrough is also not working anymore. tower-diagnostics-20220206-2004.zip

Regarding driver I installed the most current livirt driver. (removed it before) I did this with the build-in vnc client from unraid. Regarding correctness of the bios rom file I really don't understand why it worked before on lower unraid version, there i had no problem with Error 43 or booting. yeha I extracted with a windows (dual) boot and gpuz and removed all the lines before U� until the start of the rom file the selected is removed in the fixed rom file. I did extracted the rom yesterday. Edit: On addition, when selecting the bios from config gui there is a empty block so no access to the filesystem via gui i have to paste the path manually:

yep did this already this BAR error is gone now. My Syslinux config looks actually like this: default menu.c32 menu title Lime Technology, Inc. prompt 0 timeout 50 label Unraid OS menu default kernel /bzimage append iommu=pt pci=noaer video=vesafb:off video=efifb:off pcie_acs_override=downstream initrd=/bzroot label Unraid OS GUI Mode kernel /bzimage append pcie_acs_override=downstream initrd=/bzroot,/bzroot-gui label Unraid OS Safe Mode (no plugins, no GUI) kernel /bzimage append initrd=/bzroot unraidsafemode label Unraid OS GUI Safe Mode (no plugins) kernel /bzimage append initrd=/bzroot,/bzroot-gui unraidsafemode label Memtest86+ kernel /memtest tower-diagnostics-20220222-0933.zip

Hey there on 6.9.2 Unraid my both windows 10 VMs worked like they should both were working with GPU passthrough with a gtx 1070 ti. After upgrade the gpu passthrough is not working anymore. I tried it with several boot settings found here in the forum but no luck yet. When have a rom selected for the GPU the machine gets somehow a ip address from my router, but when I try to connect via rdp I get a timeout error. I also cannot ping the machine. When I remove the line of the rom the machine boots up correctly I can rdp then on it, but the graphics card has a 43 error. Actually I'm a bit lost of option what I could to to get these machines working again. Does anybody got an advice? Kind regards tower-diagnostics-20220221-2155.zip

Regarding GPU Passthrough I noticed that after removing the Graphics ROM BIOS the VM is booting correctly. When I try to pick the BIOS file from the GUI it only shows a blank selection could this be somehow related?

Yep 👍 Does anybody got an idea why the gpu passthrough VMs not working anymore?

yep recreated docker image, now it's working (have to restore all my containers from Apps Tab now) Here the diagnostics. tower-diagnostics-20220216-1401.zip

I Upgrade to rc2 Docker is not working starting anymore seems to be a Problem with the docker image Feb 16 13:32:59 Tower emhttpd: shcmd (68): /usr/local/sbin/mount_image '/mnt/user/system/docker/docker.img' /var/lib/docker 100 Feb 16 13:32:59 Tower kernel: loop2: detected capacity change from 0 to 209715200 Feb 16 13:32:59 Tower kernel: BTRFS: device fsid b7558f03-4ad1-4884-b9b7-9fe63d3900b9 devid 1 transid 2830106 /dev/loop2 scanned by mount (4507) Feb 16 13:32:59 Tower kernel: BTRFS info (device loop2): flagging fs with big metadata feature Feb 16 13:32:59 Tower kernel: BTRFS info (device loop2): using free space tree Feb 16 13:32:59 Tower kernel: BTRFS info (device loop2): has skinny extents Feb 16 13:32:59 Tower kernel: BTRFS info (device loop2): bdev /dev/loop2 errs: wr 0, rd 0, flush 0, corrupt 4, gen 0 Feb 16 13:32:59 Tower kernel: BTRFS info (device loop2): enabling ssd optimizations Feb 16 13:32:59 Tower kernel: BTRFS info (device loop2): cleaning free space cache v1 Feb 16 13:32:59 Tower kernel: BTRFS warning (device loop2): checksum verify failed on 28567109632 wanted 0x92b333b4 found 0x5563533e level 0 Feb 16 13:32:59 Tower kernel: BTRFS warning (device loop2): checksum verify failed on 28567109632 wanted 0x92b333b4 found 0x5563533e level 0 Feb 16 13:32:59 Tower kernel: BTRFS: error (device loop2) in btrfs_set_free_space_cache_v1_active:3992: errno=-5 IO failure Feb 16 13:32:59 Tower kernel: BTRFS error (device loop2): commit super ret -30 Feb 16 13:32:59 Tower root: mount: /var/lib/docker: can't read superblock on /dev/loop2. Feb 16 13:32:59 Tower kernel: BTRFS error (device loop2): open_ctree failed Feb 16 13:32:59 Tower root: mount error And my VMs with GPU Passthrough are not working anymore I'm getting just a few lines in syslog: Feb 16 13:38:13 Tower kernel: br0: port 2(vnet1) entered blocking state Feb 16 13:38:13 Tower kernel: br0: port 2(vnet1) entered disabled state Feb 16 13:38:13 Tower kernel: device vnet1 entered promiscuous mode Feb 16 13:38:13 Tower kernel: br0: port 2(vnet1) entered blocking state Feb 16 13:38:13 Tower kernel: br0: port 2(vnet1) entered forwarding state Feb 16 13:38:15 Tower avahi-daemon[4465]: Joining mDNS multicast group on interface vnet1.IPv6 with address fe80::fc54:ff:fe30:3c6e. Feb 16 13:38:15 Tower avahi-daemon[4465]: New relevant interface vnet1.IPv6 for mDNS. Feb 16 13:38:15 Tower avahi-daemon[4465]: Registering new address record for fe80::fc54:ff:fe30:3c6e on vnet1.*. Feb 16 13:38:19 Tower kernel: vfio-pci 0000:43:00.0: vfio_ecap_init: hiding ecap 0x19@0x900 Feb 16 13:38:19 Tower kernel: vfio-pci 0000:43:00.0: No more image in the PCI ROM Feb 16 13:38:19 Tower kernel: vfio-pci 0000:41:00.0: vfio_ecap_init: hiding ecap 0x1e@0x110 Feb 16 13:38:19 Tower kernel: vfio-pci 0000:41:00.0: vfio_ecap_init: hiding ecap 0x19@0x300

Ok I'll give that a try this applies to me I'll report back this weekend.