Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Vitaly Kuznetsov <vkuznets@redhat.com>
To:     kvm@vger.kernel.org
Cc:     Paolo Bonzini <pbonzini@redhat.com>,
        =?UTF-8?q?Radim=20Kr=C4=8Dm=C3=A1=C5=99?= <rkrcmar@redhat.com>,
        Roman Kagan <rkagan@virtuozzo.com>,
        "K. Y. Srinivasan" <kys@microsoft.com>,
        Haiyang Zhang <haiyangz@microsoft.com>,
        Stephen Hemminger <sthemmin@microsoft.com>,
        "Michael Kelley (EOSG)" <Michael.H.Kelley@microsoft.com>,
        Jim Mattson <jmattson@google.com>,
        Liran Alon <liran.alon@oracle.com>,
        linux-kernel@vger.kernel.org
Subject: [PATCH v5 10/12] x86/kvm/nVMX: nested state migration for Enlightened VMCS
Date:   Thu, 13 Sep 2018 19:05:20 +0200
Message-Id: <20180913170522.24876-11-vkuznets@redhat.com>
In-Reply-To: <20180913170522.24876-1-vkuznets@redhat.com>
References: <20180913170522.24876-1-vkuznets@redhat.com>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Add support for get/set of nested state when Enlightened VMCS is in use.
A new KVM_STATE_NESTED_EVMCS flag to indicate eVMCS on the vCPU was enabled
is added.

Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
---
 arch/x86/include/uapi/asm/kvm.h |  1 +
 arch/x86/kvm/vmx.c              | 82 +++++++++++++++++++++++++++++++----------
 arch/x86/kvm/x86.c              |  6 ++-
 3 files changed, 67 insertions(+), 22 deletions(-)

diff --git a/arch/x86/include/uapi/asm/kvm.h b/arch/x86/include/uapi/asm/kvm.h
index 86299efa804a..0b45bd0ea13b 100644
--- a/arch/x86/include/uapi/asm/kvm.h
+++ b/arch/x86/include/uapi/asm/kvm.h
@@ -380,6 +380,7 @@ struct kvm_sync_regs {
 
 #define KVM_STATE_NESTED_GUEST_MODE	0x00000001
 #define KVM_STATE_NESTED_RUN_PENDING	0x00000002
+#define KVM_STATE_NESTED_EVMCS		0x00000004
 
 #define KVM_STATE_NESTED_SMM_GUEST_MODE	0x00000001
 #define KVM_STATE_NESTED_SMM_VMXON	0x00000002
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 25a25fff8dd9..04d23aebbb17 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -1603,7 +1603,8 @@ static int nested_enable_evmcs(struct kvm_vcpu *vcpu,
 	 * maximum supported version. KVM supports versions from 1 to
 	 * KVM_EVMCS_VERSION.
 	 */
-	*vmcs_version = (KVM_EVMCS_VERSION << 8) | 1;
+	if (vmcs_version)
+		*vmcs_version = (KVM_EVMCS_VERSION << 8) | 1;
 
 	vmx->nested.msrs.pinbased_ctls_high &= ~EVMCS1_UNSUPPORTED_PINCTRL;
 	vmx->nested.msrs.entry_ctls_high &= ~EVMCS1_UNSUPPORTED_VMENTRY_CTRL;
@@ -9353,7 +9354,8 @@ static int handle_vmptrld(struct kvm_vcpu *vcpu)
  * This is an equivalent of the nested hypervisor executing the vmptrld
  * instruction.
  */
-static int nested_vmx_handle_enlightened_vmptrld(struct kvm_vcpu *vcpu)
+static int nested_vmx_handle_enlightened_vmptrld(struct kvm_vcpu *vcpu,
+						 bool from_launch)
 {
 	struct vcpu_vmx *vmx = to_vmx(vcpu);
 	struct hv_vp_assist_page assist_page;
@@ -9404,8 +9406,9 @@ static int nested_vmx_handle_enlightened_vmptrld(struct kvm_vcpu *vcpu)
 		 * present in struct hv_enlightened_vmcs, ...). Make sure there
 		 * are no leftovers.
 		 */
-		memset(vmx->nested.cached_vmcs12, 0,
-		       sizeof(*vmx->nested.cached_vmcs12));
+		if (from_launch)
+			memset(vmx->nested.cached_vmcs12, 0,
+			       sizeof(*vmx->nested.cached_vmcs12));
 
 	}
 	return 1;
@@ -11142,6 +11145,15 @@ static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu)
 	}
 
 	if (vmx->nested.need_vmcs12_sync) {
+		/*
+		 * hv_evmcs may end up being not mapped after migration (when
+		 * L2 was running), map it here to make sure vmcs12 changes are
+		 * properly reflected.
+		 */
+		if (vmx->nested.enlightened_vmcs_enabled &&
+		    !vmx->nested.hv_evmcs)
+			nested_vmx_handle_enlightened_vmptrld(vcpu, false);
+
 		if (vmx->nested.hv_evmcs) {
 			copy_vmcs12_to_enlightened(vmx);
 			/* All fields are clean */
@@ -13155,7 +13167,7 @@ static int nested_vmx_run(struct kvm_vcpu *vcpu, bool launch)
 	if (!nested_vmx_check_permission(vcpu))
 		return 1;
 
-	if (!nested_vmx_handle_enlightened_vmptrld(vcpu))
+	if (!nested_vmx_handle_enlightened_vmptrld(vcpu, true))
 		return 1;
 
 	if (!vmx->nested.hv_evmcs && !nested_vmx_check_vmcs12(vcpu))
@@ -14366,6 +14378,20 @@ static int enable_smi_window(struct kvm_vcpu *vcpu)
 	return 0;
 }
 
+static inline int vmx_has_valid_vmcs12(struct kvm_vcpu *vcpu)
+{
+	struct vcpu_vmx *vmx = to_vmx(vcpu);
+
+	/*
+	 * In case we do two consecutive get/set_nested_state()s while L2 was
+	 * running hv_evmcs may end up not being mapped (we map it from
+	 * nested_vmx_run()/vmx_vcpu_run()). Check is_guest_mode() as we always
+	 * have vmcs12 if it is true.
+	 */
+	return is_guest_mode(vcpu) || vmx->nested.current_vmptr != -1ull ||
+		vmx->nested.hv_evmcs;
+}
+
 static int vmx_get_nested_state(struct kvm_vcpu *vcpu,
 				struct kvm_nested_state __user *user_kvm_nested_state,
 				u32 user_data_size)
@@ -14386,16 +14412,15 @@ static int vmx_get_nested_state(struct kvm_vcpu *vcpu,
 	vmx = to_vmx(vcpu);
 	vmcs12 = get_vmcs12(vcpu);
 
-	/* FIXME: Enlightened VMCS is currently unsupported */
-	if (vmx->nested.hv_evmcs)
-		return -ENOTSUPP;
+	if (nested_vmx_allowed(vcpu) && vmx->nested.enlightened_vmcs_enabled)
+		kvm_state.flags |= KVM_STATE_NESTED_EVMCS;
 
 	if (nested_vmx_allowed(vcpu) &&
 	    (vmx->nested.vmxon || vmx->nested.smm.vmxon)) {
 		kvm_state.vmx.vmxon_pa = vmx->nested.vmxon_ptr;
 		kvm_state.vmx.vmcs_pa = vmx->nested.current_vmptr;
 
-		if (vmx->nested.current_vmptr != -1ull) {
+		if (vmx_has_valid_vmcs12(vcpu)) {
 			kvm_state.size += VMCS12_SIZE;
 
 			if (is_guest_mode(vcpu) &&
@@ -14424,20 +14449,24 @@ static int vmx_get_nested_state(struct kvm_vcpu *vcpu,
 	if (copy_to_user(user_kvm_nested_state, &kvm_state, sizeof(kvm_state)))
 		return -EFAULT;
 
-	if (vmx->nested.current_vmptr == -1ull)
+	if (!vmx_has_valid_vmcs12(vcpu))
 		goto out;
 
 	/*
 	 * When running L2, the authoritative vmcs12 state is in the
 	 * vmcs02. When running L1, the authoritative vmcs12 state is
-	 * in the shadow vmcs linked to vmcs01, unless
-	 * sync_shadow_vmcs is set, in which case, the authoritative
-	 * vmcs12 state is in the vmcs12 already.
+	 * in the shadow or enlightened vmcs linked to vmcs01. In case
+	 * need_vmcs12_sync is set the authoritative vmcs12 state is in
+	 * the vmcs12 already.
 	 */
-	if (is_guest_mode(vcpu))
+	if (is_guest_mode(vcpu)) {
 		sync_vmcs12(vcpu, vmcs12);
-	else if (enable_shadow_vmcs && !vmx->nested.need_vmcs12_sync)
-		copy_shadow_to_vmcs12(vmx);
+	} else if (!vmx->nested.need_vmcs12_sync) {
+		if (vmx->nested.hv_evmcs)
+			copy_enlightened_to_vmcs12(vmx);
+		else if (enable_shadow_vmcs)
+			copy_shadow_to_vmcs12(vmx);
+	}
 
 	if (copy_to_user(user_kvm_nested_state->data, vmcs12, sizeof(*vmcs12)))
 		return -EFAULT;
@@ -14465,6 +14494,9 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
 	if (kvm_state->format != 0)
 		return -EINVAL;
 
+	if (kvm_state->flags & KVM_STATE_NESTED_EVMCS)
+		nested_enable_evmcs(vcpu, NULL);
+
 	if (!nested_vmx_allowed(vcpu))
 		return kvm_state->vmx.vmxon_pa == -1ull ? 0 : -EINVAL;
 
@@ -14507,11 +14539,21 @@ static int vmx_set_nested_state(struct kvm_vcpu *vcpu,
 	if (kvm_state->size < sizeof(kvm_state) + sizeof(*vmcs12))
 		return 0;
 
-	if (kvm_state->vmx.vmcs_pa == kvm_state->vmx.vmxon_pa ||
-	    !page_address_valid(vcpu, kvm_state->vmx.vmcs_pa))
-		return -EINVAL;
+	if (kvm_state->vmx.vmcs_pa != -1ull) {
+		if (kvm_state->vmx.vmcs_pa == kvm_state->vmx.vmxon_pa ||
+		    !page_address_valid(vcpu, kvm_state->vmx.vmcs_pa))
+			return -EINVAL;
 
-	set_current_vmptr(vmx, kvm_state->vmx.vmcs_pa);
+		set_current_vmptr(vmx, kvm_state->vmx.vmcs_pa);
+	} else if (kvm_state->flags & KVM_STATE_NESTED_EVMCS) {
+		/*
+		 * Sync eVMCS upon entry as we may not have
+		 * HV_X64_MSR_VP_ASSIST_PAGE set up yet.
+		 */
+		vmx->nested.need_vmcs12_sync = true;
+	} else {
+		return -EINVAL;
+	}
 
 	if (kvm_state->vmx.smm.flags & KVM_STATE_NESTED_SMM_VMXON) {
 		vmx->nested.smm.vmxon = true;
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 0157f480d4ad..8c180fbb9e1d 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -4053,11 +4053,13 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 			return -EINVAL;
 
 		if (kvm_state.flags &
-		    ~(KVM_STATE_NESTED_RUN_PENDING | KVM_STATE_NESTED_GUEST_MODE))
+		    ~(KVM_STATE_NESTED_RUN_PENDING | KVM_STATE_NESTED_GUEST_MODE
+		      | KVM_STATE_NESTED_EVMCS))
 			return -EINVAL;
 
 		/* nested_run_pending implies guest_mode.  */
-		if (kvm_state.flags == KVM_STATE_NESTED_RUN_PENDING)
+		if ((kvm_state.flags & KVM_STATE_NESTED_RUN_PENDING)
+		    && !(kvm_state.flags & KVM_STATE_NESTED_GUEST_MODE))
 			return -EINVAL;
 
 		r = kvm_x86_ops->set_nested_state(vcpu, user_kvm_nested_state, &kvm_state);
-- 
2.14.4