Re: [PATCH v3 26/27] KVM: nSVM: Use different ASIDs for L1 and L2

[Sean Christopherson <seanjc@google.com>]

On Mon, Oct 26, 2026, Yosry Ahmed wrote:
> On Mon, Oct 26, 2026 at 10:33 AM Sean Christopherson <seanjc@google.com> wrote:
> >
> > On Thu, Oct 22, 2026, Yosry Ahmed wrote:
> > > On Thu, Oct 22, 2026 at 3:44 PM Sean Christopherson <seanjc@google.com> wrote:
> > > >
> > > > On Fri, Oct 16, 2026, Jim Mattson wrote:
> > > > > On Mon, Oct 12, 2026 at 4:37 PM Yosry Ahmed <yosry@kernel.org> wrote:
> > > > > Moreover, does this change the semantics of PV TLB flush? In
> > > > > record_steal_time(), we have:
> > > > >
> > > > > if (st_preempted & KVM_VCPU_FLUSH_TLB)
> > > > >         kvm_vcpu_flush_tlb_guest(vcpu);
> > > > >
> > > > > IIUC, an idle vCPU that wakes into L2 will flush only L2's ASID. In
> > > > > the past, with a shared ASID, L1's ASID was always flushed on emulated
> > > > > VM-exit anyway. Now, L1's ASID will be unaffected. Though I can't find
> > > > > any real documentation on PV TLB flush, it *seems* like the remote
> > > > > vCPU would want L1's ASID to be flushed. OTOH, this change makes nSVM
> > > > > behave just like nVMX.
> > > >
> > > > That's a bug that absolutely needs to be fixed, probably relatively urgently.  We
> > > > should also document that passing through the steal_time shared information to L2
> > > > to utilize PV TLB flushes for L2 is explicitly not supported.
> > > >
> > > > I hate to re-introduce deferred actions for L2 => L1 things, but that seems like
> > > > the only sane way to handle this, because handling the !tdp_enabled case, where
> > > > KVM needs to sync the current root, requires knowing what MMU is going to be in
> > > > use when L1 regains control.  Yuck.
> > >
> > > Are you implying that *you* will be re-introducing deferred actions? :P
> >
> > Heh, probably not this calendar year.
> 
> Are you opposed to this series landing before that is fixed, keeping
> in mind that it will expose nSVM to the pre-existing bug?

Nah, though I should just fix the damn bug.  It's not hard, I just don't want to
do it because I don't want to reintroduce a deferred request.  Untested, but I
think this is all that's needed.

diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h
index 0378577d9d8b..696aae38f619 100644
--- a/arch/x86/include/asm/kvm_host.h
+++ b/arch/x86/include/asm/kvm_host.h
@@ -879,6 +879,7 @@ struct kvm_vcpu_arch {
 
 	struct {
 		u8 preempted;
+		bool flush_tlb_pending;
 		u64 msr_val;
 		u64 last_steal;
 		struct gfn_to_hva_cache cache;
diff --git a/arch/x86/kvm/regs.h b/arch/x86/kvm/regs.h
index 97c17b120a57..efa54f9f9e9a 100644
--- a/arch/x86/kvm/regs.h
+++ b/arch/x86/kvm/regs.h
@@ -469,6 +469,11 @@ static inline void leave_guest_mode(struct kvm_vcpu *vcpu)
 		kvm_make_request(KVM_REQ_LOAD_EOI_EXITMAP, vcpu);
 	}
 
+	if (vcpu->arch.st.flush_tlb_pending) {
+		kvm_make_request(KVM_REQ_TLB_FLUSH_GUEST, vcpu);
+		vcpu->arch.st.flush_tlb_pending = false;
+	}
+
 	vcpu->stat.guest_mode = 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 32563a91fba9..15eb033b8e17 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -2117,8 +2117,12 @@ static void record_steal_time(struct kvm_vcpu *vcpu)
 
 		trace_kvm_pv_tlb_flush(vcpu->vcpu_id,
 				       st_preempted & KVM_VCPU_FLUSH_TLB);
-		if (st_preempted & KVM_VCPU_FLUSH_TLB)
-			kvm_vcpu_flush_tlb_guest(vcpu);
+		if (st_preempted & KVM_VCPU_FLUSH_TLB) {
+			if (!is_guest_mode(vcpu))
+				kvm_vcpu_flush_tlb_guest(vcpu);
+			else
+				vcpu->arch.st.flush_tlb_pending = true;
+		}
 
 		if (!user_access_begin(st, sizeof(*st)))
 			goto dirty;