Recent content by tristank

I'm hitting the same issue on kernel 5.15.39-1-pve. $ dmesg | grep mpt2sas_cm0 [ 1.058716] mpt2sas_cm0: 64 BIT PCI BUS DMA ADDRESSING SUPPORTED, total mem (32471524 kB) [ 1.068129] mpt2sas_cm0: CurrentHostPageSize is 0: Setting default host page size to 4k [ 1.068140] mpt2sas_cm0...

Thanks this helped. I also download the ibt-0041-0041.sfi firmware for bluetooth to work but I'm still getting an error saying Failed to load Intel Firmware file intel/ibt-0041-0041.sfi (-2).

I'm also struggeling with same issue on iGPU passthrough aka gvt-d. I had to patch the kernel in order to workaround a FLR error. I'm not getting any errors in Windows guests (q35) or Ubuntu guests (q35) but the sad thing is that the screen is black. I do have hdmi dummy plug connected to the...

Has this been fixed? Because I think I'm having the same issue on a Z490M Pro4 with a i9-10850K when I'm trying to use SR-IOV with a Intel 82576. ``` [ 480.881090] BUG: kernel NULL pointer dereference, address: 0000000000000008 [ 480.881108] #PF: supervisor read access in kernel mode [...

As of right now it is still not possible to install WIndows 7 on a VirtIO Block device because the drivers (virtio-win 0.1.215) will not be recognized. You will also have to make sure to uncheck "pre-enroll-keys" or to disable secure boot in the guests uefi (hit F2 during boot to get into the...

I think this only possible with LXD or KVM not with LXC since you can't pass a PCIe device to a LXC container. For docker there is pipewire. Correct me if I'm wrong.

On ZFS installation systemd-boot is in use for booting not grub. So as it turns out you can define the default boot entry by selecting the entry in the systemd-boot menu and hit "d" for default. Would have been nice to to see this in the wiki-page. Quote from here.

I just followed your instructions but this won't work for a zfs installation with efistub. Any suggestions how to do this? Also there seems to be missmatch between the instructions in the image and the example. In the image you recommend the menuentry gnulinux-simple-******** and in the example...

There goes my hope for a doing this in convenient way through pve-efiboot-tool.

Does someone know how to boot into a specfic pve-kernel by default?

I just downgraded the kernel to 5.11.22-7-pve and can confirm that it is now possible to passthrough the Intel HD 500 iGPU to Ubuntu 21.10 guest VM. $ apt install pve-kernel-5.11.22-7-pve $ pve-efiboot-tool kernel list $ pve-efiboot-tool kernel add 5.11.22-7-pve $ update-initramfs -u -k all &&...

For me it was as easy as: $ apt install pve-kernel-5.11.22-7-pve $ pve-efiboot-tool kernel list $ pve-efiboot-tool kernel add 5.11.22-7-pve $ update-initramfs -u -k all && pve-efiboot-tool refresh EDIT: I was wrong, this commands just installs a specific kernel. Sadly there seems to be no...

Ich kämpfe zur Zeit auch mit dem Proxmox Kernel, es sieht so aus als könnte er die Ursache für ein Problem mit GVT-d sein.

It depends. The device-id has to match the one actually used by your iGPU. You can check that with ./rom-parser rom.bin. For me this was nesecarry because the device-id from the dump was 0406 and not 5a85 like reported by lspci when running Proxmox with UEFI enabled. lspci -nnk | grep -A4 VGA...

Ich kann diesen Blog-Post empfehlen. Dort hat sich jemand intensiv mit der Frage auseinandergesetzt. Es gibt verschiedene Wege das zu erreichen: a) OVM in LXC-Container und Storage per Bindmount durchreichen b) OVM in einer QEMU/KVM VM und Storage per 9pfs mount durchreichen c) OVM in einer...