https://bugzilla.kernel.org/show_bug.cgi?id=217307 --- Comment #7 from vkuznets@xxxxxxxxxx --- Assuming this is not a KVM/QEMU regression, I'd suggest to explore two options: 1) Change "-cpu host" to a named CPU model. I don't see "alderlake" CPU models in QEMU so I'd start with something like "Skylake-Client-v4". Remove all other CPU options you have, like "rtm=off,mpx=off,host-cache-info=on,l3-cache=off". Try to find the exact CPU option which breaks things. There were similar but reversed (works with '-cpu host', doesn't work with a named model) issues in the past, e.g. https://lore.kernel.org/qemu-devel/20220308113445.859669-21-pbonzini@xxxxxxxxxx/ 2) Try disabling certain Hyper-V enlightenments, start with "hv-evmcs". In theory, things should work (slowly, but still) without any Hyper-V enlightenments. -- You may reply to this email to add a comment. You are receiving this mail because: You are watching the assignee of the bug.
