Re: [PATCH 2/2] KVM: x86: Forbid KVM_SET_CPUID{,2} after KVM_RUN

[Sean Christopherson <seanjc@google.com>]

On Fri, Jan 14, 2022, Maxim Levitsky wrote:
> On Thu, 2022-01-13 at 22:33 +0000, Sean Christopherson wrote:
> > On Mon, Jan 03, 2022, Igor Mammedov wrote:
> > > On Mon, 03 Jan 2022 09:04:29 +0100
> > > Vitaly Kuznetsov <vkuznets@redhat.com> wrote:
> > > 
> > > > Paolo Bonzini <pbonzini@redhat.com> writes:
> > > > 
> > > > > On 12/27/21 18:32, Igor Mammedov wrote:  
> > > > > > > Tweaked and queued nevertheless, thanks.  
> > > > > > it seems this patch breaks VCPU hotplug, in scenario:
> > > > > > 
> > > > > >    1. hotunplug existing VCPU (QEMU stores VCPU file descriptor in parked cpus list)
> > > > > >    2. hotplug it again (unsuspecting QEMU reuses stored file descriptor when recreating VCPU)
> > > > > > 
> > > > > > RHBZ:https://bugzilla.redhat.com/show_bug.cgi?id=2028337#c11
> > > > > >   
> > > > > 
> > > > > The fix here would be (in QEMU) to not call KVM_SET_CPUID2 again. 
> > > > > However, we need to work around it in KVM, and allow KVM_SET_CPUID2 if 
> > > > > the data passed to the ioctl is the same that was set before.  
> > > > 
> > > > Are we sure the data is going to be *exactly* the same? In particular,
> > > > when using vCPU fds from the parked list, do we keep the same
> > > > APIC/x2APIC id when hotplugging? Or can we actually hotplug with a
> > > > different id?
> > > 
> > > If I recall it right, it can be a different ID easily.
> > 
> > No, it cannot.  KVM doesn't provide a way for userspace to change the APIC ID of
> > a vCPU after the vCPU is created.  x2APIC flat out disallows changing the APIC ID,
> > and unless there's magic I'm missing, apic_mmio_write() => kvm_lapic_reg_write()
> > is not reachable from userspace.
> 
> So after all, it is true that vcpu_id == initial APIC_ID,
> and if we don't let guest change it, it will be always like that?

Except for kvm_apic_set_state(), which I forgot existed, yes.

> You said that its not true in the other mail in the thread. 

I was wrong, I was thinking that userspace could reach kvm_lapic_reg_write(), but
I forgot that there would be no connection without x2apic.  But I forgot about
kvm_apic_set_state()...

> I haven't checked it in the code yet, as I never was much worried about
> userspace changing, but I will check it soon.
> 
> I did a quick look and I see that at least the userspace can call
> 'kvm_apic_set_state' and it contains snapshot of all apic registers,
> including apic id.  However it would be very easy to add a check there and
> fail if userspace attempts to set APIC_ID != vcpu_id.

Yeah, hopefully that doesn't break any userspace.  I can't imagine it would,
because if the guest disabled and re-enabled the APIC, kvm_lapic_set_base() would
restore the APIC ID to vcpu_id.

With luck, that's the last hole we need to close...