Hi I work with the developers on upstream TDX support. Now that the TDX support in kvm-coco-queue is getting more mature, we thought it's the time to do regression test on normal VM and it might be useful to start sharing the results of our testing on the list. This is in addition to the normal suite of automated tests that run in our CI. Except the issue[0] which was already reported on upstream KVM, no other new issues were hit, especially no known regression bugs in the "MMU part 1", "VM/vCPU creation" or "MMU part 2" patches. Details -------------------------- Test Environment CPU: Sapphire Rapids/Emerald Rapids OVMF: https://github.com/tianocore/edk2/commit/cf8241faccc176dd857e5d42e84f3a2f0fc8d16c Host/Guest Kernel: https://git.kernel.org/pub/scm/virt/kvm/kvm.git/log/?h=kvm-coco-queue QEMU: https://github.com/intel-staging/qemu-tdx/tree/tdx-qemu-upstream-v6.1-fix_kvm_hypercall_return_value Tested features or cases: - Basic boot - Boot multiple VMs - Boot various distros(including Windows) - Boot VM with huge resource/complex cpu topology - Stress boot - Memory hotplug/unplug, memory with NX hugepage on - Memory workload in high/low memory VMs with NX hugepage on - Device passthrough(NIC) - Live migration - Nested - Intel key instructions - 5 level paging - Bus lock debug exception - PMU/vPMU - SGX - Workload(kernel build) in VM [0] https://bugzilla.kernel.org/show_bug.cgi?id=219588 Best Regards Fan
