On 6 Dec 2022, at 14:59, Dave Voutila <d...@sisu.io> wrote: > The functional changes between releases include driving more vm exits to > be handled by vmd and not vmm in the kernel. vmd is already quite > chatty, so it's time to revisit which messages add value. > > vcpu_ext_eptviolation sounds more ominous than it should at this > point. It's the primary page fault handler for both Intel EPT and AMD > SVM.
OK, so should I interpret the above as “ignore the error messages, all is well?” The reason I am asking is that on a different system, powered by an Intel Atom (hence showing the “slow L1TF mitigation” message at boot) I do not see these log messages at all (OpenBSD 7.2, three VMs, all OpenBSD 7.2). If I can ignore them that’s fine by me. >> The error messages increase when there is traffic to the VMs, in >> particular to fetest-0 (a web site running nginx) and ns-0 (a >> nameserver running BIND9). >> > > Is it usually a single vm in your topology or all vm's that start > producing the noise? All the VMs alternatively produce the noise. Cheers, Arrigo
