On 02-Oct-23 5:59 PM, Peter Zijlstra wrote: > On Fri, Sep 29, 2023 at 08:50:07PM +0530, Ravi Bangoria wrote: >> On 29-Sep-23 5:23 PM, Peter Zijlstra wrote: > >>> I don't think you need to go that far, host can use PMU just fine as >>> long as it doesn't overlap with a vCPU. Basically, if you force >>> perf_attr::exclude_guest on everything your vCPU can haz the full thing. > > ^ this.. > >> How about keying off based on PMU specific KVM module parameter? Something >> like what Manali has proposed for AMD VIBS? Please see solution 1.1: >> >> https://lore.kernel.org/r/3a6c693e-1ef4-6542-bc90-d4468773b97d@xxxxxxx > > So I hadn't read that, but isn't that more or less what I proposed > above? Yes, it's pretty much same as: "Like I wrote, all we need to do is ensure vCPU tasks will never have a perf-event scheduled that covers guest mode. Currently this would be achievable by having event creation for both: - CPU events without attr::exclude_guest=1, and - task events for vCPU task of interest without attr::exclude_guest=1 error with -EBUSY or something." Thanks, Ravi
