On Mon, Apr 09, 2018 at 02:58:18PM +0200, Christoffer Dall wrote: > Hi Mark, > > [Sorry for late reply] > > On Fri, Mar 09, 2018 at 02:28:38PM +0000, Mark Rutland wrote: > > On Tue, Feb 06, 2018 at 01:38:47PM +0100, Christoffer Dall wrote: > > > On Mon, Nov 27, 2017 at 04:38:04PM +0000, Mark Rutland wrote: > > > > When pointer authentication is supported, a guest may wish to use it. > > > > This patch adds the necessary KVM infrastructure for this to work, with > > > > a semi-lazy context switch of the pointer auth state. > > > > > > > > When we schedule a vcpu, > > > > > > That's not quite what the code does, the code only does this when we > > > schedule back a preempted or blocked vcpu thread. > > > > Does that only leave the case of the vCPU being scheduled for the first > > time? Or am I missing something else? > > > > [...] > > In the current patch, you're only calling kvm_arm_vcpu_ptrauth_disable() > from kvm_arch_sched_in() which is only called on the preempt notifier > patch, which leaves out every time we enter the guest from userspace and > therefore also the initial run of the vCPU (assuming there's no > preemption in the kernel prior to running the first time). > > vcpu_load() takes care of all the cases. I see. > > > I still find this decision to begin trapping again quite arbitrary, and > > > would at least prefer this to be in vcpu_load (which would make the > > > behavior match the commit text as well). > > > > Sure, done. > > > > > My expectation would be that if a guest is running software with pointer > > > authentication enabled, then it's likely to either keep using the > > > feature, or not use it at all, so I would make this a one-time flag. > > > > I think it's likely that some applications will use ptrauth while others > > do not. Even if the gust OS supports ptrauth, KVM may repeatedly preempt > > an application that doesn't use it, and we'd win in that case. > > > > There are also some rarer cases, like kexec in a guest from a > > ptrauth-aware kernel to a ptrauth-oblivious one. > > > > I don't have strong feelings either way, and I have no data. > > I think your intuition sounds sane, and let's reset the flag on every > vcpu_load, and we can always revisit when we have hardware and data if > someone reports a performance issue. Cool. I've switched to vcpu_load() locally, and will use that in v3. Thanks, Mark. _______________________________________________ kvmarm mailing list kvmarm@xxxxxxxxxxxxxxxxxxxxx https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
