[v10,21/39] KVM: nSVM: hyper-v: Enable L2 TLB flush

Commit Message

Vitaly Kuznetsov Sept. 21, 2022, 3:24 p.m. UTC

Implement Hyper-V L2 TLB flush for nSVM. The feature needs to be enabled
both in extended 'nested controls' in VMCB and VP assist page.
According to Hyper-V TLFS, synthetic vmexit to L1 is performed with
- HV_SVM_EXITCODE_ENL exit_code.
- HV_SVM_ENL_EXITCODE_TRAP_AFTER_FLUSH exit_info_1.

Note: VP assist page is cached in 'struct kvm_vcpu_hv' so
recalc_intercepts() doesn't need to read from guest's memory. KVM
needs to update the case upon each VMRUN and after svm_set_nested_state
(svm_get_nested_state_pages()) to handle the case when the guest got
migrated while L2 was running.

Reviewed-by: Maxim Levitsky <[e-mail]>
Signed-off-by: Vitaly Kuznetsov <[e-mail]>
---
 arch/x86/kvm/svm/hyperv.c |  7 +++++++
 arch/x86/kvm/svm/hyperv.h | 30 ++++++++++++++++++++++++++++++
 arch/x86/kvm/svm/nested.c | 36 ++++++++++++++++++++++++++++++++++--
 3 files changed, 71 insertions(+), 2 deletions(-)

Comments

Sean Christopherson Sept. 21, 2022, 9:31 p.m. UTC | #1

On Wed, Sep 21, 2022, Vitaly Kuznetsov wrote:
> diff --git a/arch/x86/kvm/svm/hyperv.h b/arch/x86/kvm/svm/hyperv.h
> index dd2e393f84a0..7b01722838bf 100644
> --- a/arch/x86/kvm/svm/hyperv.h
> +++ b/arch/x86/kvm/svm/hyperv.h
> @@ -33,6 +33,9 @@ struct hv_enlightenments {
>   */
>  #define VMCB_HV_NESTED_ENLIGHTENMENTS VMCB_SW
>  
> +#define HV_SVM_EXITCODE_ENL 0xF0000000
> +#define HV_SVM_ENL_EXITCODE_TRAP_AFTER_FLUSH   (1)

Same as the synthetic VMX exit reason, these should go in hyperv-tlfs.h.  Keeping
these out of KVM also helps avoid the need for svm/hyperv.h.

https://learn.microsoft.com/en-us/virtualization/hyper-v-on-windows/tlfs/nested-virtualization#synthetic-vm-exit

> +
>  static inline void nested_svm_hv_update_vm_vp_ids(struct kvm_vcpu *vcpu)
>  {
>  	struct vcpu_svm *svm = to_svm(vcpu);
> @@ -48,6 +51,33 @@ static inline void nested_svm_hv_update_vm_vp_ids(struct kvm_vcpu *vcpu)
>  	hv_vcpu->nested.vp_id = hve->hv_vp_id;
>  }
>  
> +static inline bool

Strongly prefer 'int' with 0/-errno over a boolean.  Hrm, maybe add a prep patch
to convert kvm_hv_get_assist_page() to return 0/-errno?  That way this can still
return kvm_hv_get_assist_page() directly.

> nested_svm_hv_update_vp_assist(struct kvm_vcpu *vcpu)

Maybe s/update/verify since there isn't a true update anywhere?

> +{
> +	if (!to_hv_vcpu(vcpu))

This check isn't necessary, it's covered by kvm_hv_assist_page_enabled().

> +		return true;
> +
> +	if (!kvm_hv_assist_page_enabled(vcpu))
> +		return true;
> +
> +	return kvm_hv_get_assist_page(vcpu);

As mentioned earlier, I think this belongs in arch/x86/kvm/hyperv.h.

Patch

diff --git a/arch/x86/kvm/svm/hyperv.c b/arch/x86/kvm/svm/hyperv.c
index 911f51021af1..088f6429b24c 100644
--- a/arch/x86/kvm/svm/hyperv.c
+++ b/arch/x86/kvm/svm/hyperv.c
@@ -8,4 +8,11 @@ 
 
 void svm_hv_inject_synthetic_vmexit_post_tlb_flush(struct kvm_vcpu *vcpu)
 {
+	struct vcpu_svm *svm = to_svm(vcpu);
+
+	svm->vmcb->control.exit_code = HV_SVM_EXITCODE_ENL;
+	svm->vmcb->control.exit_code_hi = 0;
+	svm->vmcb->control.exit_info_1 = HV_SVM_ENL_EXITCODE_TRAP_AFTER_FLUSH;
+	svm->vmcb->control.exit_info_2 = 0;
+	nested_svm_vmexit(svm);
 }
diff --git a/arch/x86/kvm/svm/hyperv.h b/arch/x86/kvm/svm/hyperv.h
index dd2e393f84a0..7b01722838bf 100644
--- a/arch/x86/kvm/svm/hyperv.h
+++ b/arch/x86/kvm/svm/hyperv.h
@@ -33,6 +33,9 @@  struct hv_enlightenments {
  */
 #define VMCB_HV_NESTED_ENLIGHTENMENTS VMCB_SW
 
+#define HV_SVM_EXITCODE_ENL 0xF0000000
+#define HV_SVM_ENL_EXITCODE_TRAP_AFTER_FLUSH   (1)
+
 static inline void nested_svm_hv_update_vm_vp_ids(struct kvm_vcpu *vcpu)
 {
 	struct vcpu_svm *svm = to_svm(vcpu);
@@ -48,6 +51,33 @@  static inline void nested_svm_hv_update_vm_vp_ids(struct kvm_vcpu *vcpu)
 	hv_vcpu->nested.vp_id = hve->hv_vp_id;
 }
 
+static inline bool nested_svm_hv_update_vp_assist(struct kvm_vcpu *vcpu)
+{
+	if (!to_hv_vcpu(vcpu))
+		return true;
+
+	if (!kvm_hv_assist_page_enabled(vcpu))
+		return true;
+
+	return kvm_hv_get_assist_page(vcpu);
+}
+
+static inline bool nested_svm_l2_tlb_flush_enabled(struct kvm_vcpu *vcpu)
+{
+	struct vcpu_svm *svm = to_svm(vcpu);
+	struct hv_enlightenments *hve =
+		(struct hv_enlightenments *)svm->nested.ctl.reserved_sw;
+	struct kvm_vcpu_hv *hv_vcpu = to_hv_vcpu(vcpu);
+
+	if (!hv_vcpu)
+		return false;
+
+	if (!hve->hv_enlightenments_control.nested_flush_hypercall)
+		return false;
+
+	return hv_vcpu->vp_assist_page.nested_control.features.directhypercall;
+}
+
 void svm_hv_inject_synthetic_vmexit_post_tlb_flush(struct kvm_vcpu *vcpu);
 
 #endif /* __ARCH_X86_KVM_SVM_HYPERV_H__ */
diff --git a/arch/x86/kvm/svm/nested.c b/arch/x86/kvm/svm/nested.c
index b8df2f4f880e..e0484028dc70 100644
--- a/arch/x86/kvm/svm/nested.c
+++ b/arch/x86/kvm/svm/nested.c
@@ -149,8 +149,12 @@  void recalc_intercepts(struct vcpu_svm *svm)
 		vmcb_clr_intercept(c, INTERCEPT_VINTR);
 	}
 
-	/* We don't want to see VMMCALLs from a nested guest */
-	vmcb_clr_intercept(c, INTERCEPT_VMMCALL);
+	/*
+	 * We want to see VMMCALLs from a nested guest only when Hyper-V L2 TLB
+	 * flush feature is enabled.
+	 */
+	if (!nested_svm_l2_tlb_flush_enabled(&svm->vcpu))
+		vmcb_clr_intercept(c, INTERCEPT_VMMCALL);
 
 	for (i = 0; i < MAX_INTERCEPT; i++)
 		c->intercepts[i] |= g->intercepts[i];
@@ -473,6 +477,17 @@  static void nested_save_pending_event_to_vmcb12(struct vcpu_svm *svm,
 
 static void nested_svm_transition_tlb_flush(struct kvm_vcpu *vcpu)
 {
+	/*
+	 * KVM_REQ_HV_TLB_FLUSH flushes entries from either L1's VP_ID or
+	 * L2's VP_ID upon request from the guest. Make sure we check for
+	 * pending entries for the case when the request got misplaced (e.g.
+	 * a transition from L2->L1 happened while processing L2 TLB flush
+	 * request or vice versa). kvm_hv_vcpu_flush_tlb() will not flush
+	 * anything if there are no requests in the corresponding buffer.
+	 */
+	if (to_hv_vcpu(vcpu))
+		kvm_make_request(KVM_REQ_HV_TLB_FLUSH, vcpu);
+
 	/*
 	 * TODO: optimize unconditional TLB flush/MMU sync.  A partial list of
 	 * things to fix before this can be conditional:
@@ -824,6 +839,12 @@  int nested_svm_vmrun(struct kvm_vcpu *vcpu)
 		return 1;
 	}
 
+	/* This fails when VP assist page is enabled but the supplied GPA is bogus */
+	if (!nested_svm_hv_update_vp_assist(vcpu)) {
+		kvm_inject_gp(vcpu, 0);
+		return 1;
+	}
+
 	vmcb12_gpa = svm->vmcb->save.rax;
 	ret = kvm_vcpu_map(vcpu, gpa_to_gfn(vmcb12_gpa), &map);
 	if (ret == -EINVAL) {
@@ -1413,6 +1434,7 @@  static int svm_check_nested_events(struct kvm_vcpu *vcpu)
 int nested_svm_exit_special(struct vcpu_svm *svm)
 {
 	u32 exit_code = svm->vmcb->control.exit_code;
+	struct kvm_vcpu *vcpu = &svm->vcpu;
 
 	switch (exit_code) {
 	case SVM_EXIT_INTR:
@@ -1431,6 +1453,13 @@  int nested_svm_exit_special(struct vcpu_svm *svm)
 			return NESTED_EXIT_HOST;
 		break;
 	}
+	case SVM_EXIT_VMMCALL:
+		/* Hyper-V L2 TLB flush hypercall is handled by L0 */
+		if (guest_hv_cpuid_has_l2_tlb_flush(vcpu) &&
+		    nested_svm_l2_tlb_flush_enabled(vcpu) &&
+		    kvm_hv_is_tlb_flush_hcall(vcpu))
+			return NESTED_EXIT_HOST;
+		break;
 	default:
 		break;
 	}
@@ -1711,6 +1740,9 @@  static bool svm_get_nested_state_pages(struct kvm_vcpu *vcpu)
 		return false;
 	}
 
+	if (!nested_svm_hv_update_vp_assist(vcpu))
+		return false;
+
 	return true;
 }