Merge branch 'kvm-sev-move-context' into kvm-master

Add support for AMD SEV and SEV-ES intra-host migration support.  Intra
host migration provides a low-cost mechanism for userspace VMM upgrades.

In the common case for intra host migration, we can rely on the normal
ioctls for passing data from one VMM to the next. SEV, SEV-ES, and other
confidential compute environments make most of this information opaque, and
render KVM ioctls such as "KVM_GET_REGS" irrelevant.  As a result, we need
the ability to pass this opaque metadata from one VMM to the next. The
easiest way to do this is to leave this data in the kernel, and transfer
ownership of the metadata from one KVM VM (or vCPU) to the next.  In-kernel
hand off makes it possible to move any data that would be
unsafe/impossible for the kernel to hand directly to userspace, and
cannot be reproduced using data that can be handed to userspace.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

1 files changed, 255 insertions, 47 deletions

diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c
index 1964b9a174be..531613f758ba 100644
--- a/arch/x86/kvm/svm/sev.c
+++ b/arch/x86/kvm/svm/sev.c
@@ -120,16 +120,26 @@ static bool __sev_recycle_asids(int min_asid, int max_asid)
 	return true;
 }
 
+static int sev_misc_cg_try_charge(struct kvm_sev_info *sev)
+{
+	enum misc_res_type type = sev->es_active ? MISC_CG_RES_SEV_ES : MISC_CG_RES_SEV;
+	return misc_cg_try_charge(type, sev->misc_cg, 1);
+}
+
+static void sev_misc_cg_uncharge(struct kvm_sev_info *sev)
+{
+	enum misc_res_type type = sev->es_active ? MISC_CG_RES_SEV_ES : MISC_CG_RES_SEV;
+	misc_cg_uncharge(type, sev->misc_cg, 1);
+}
+
 static int sev_asid_new(struct kvm_sev_info *sev)
 {
 	int asid, min_asid, max_asid, ret;
 	bool retry = true;
-	enum misc_res_type type;
 
-	type = sev->es_active ? MISC_CG_RES_SEV_ES : MISC_CG_RES_SEV;
 	WARN_ON(sev->misc_cg);
 	sev->misc_cg = get_current_misc_cg();
-	ret = misc_cg_try_charge(type, sev->misc_cg, 1);
+	ret = sev_misc_cg_try_charge(sev);
 	if (ret) {
 		put_misc_cg(sev->misc_cg);
 		sev->misc_cg = NULL;
@@ -162,7 +172,7 @@ again:
 
 	return asid;
 e_uncharge:
-	misc_cg_uncharge(type, sev->misc_cg, 1);
+	sev_misc_cg_uncharge(sev);
 	put_misc_cg(sev->misc_cg);
 	sev->misc_cg = NULL;
 	return ret;
@@ -179,7 +189,6 @@ static void sev_asid_free(struct kvm_sev_info *sev)
 {
 	struct svm_cpu_data *sd;
 	int cpu;
-	enum misc_res_type type;
 
 	mutex_lock(&sev_bitmap_lock);
 
@@ -192,8 +201,7 @@ static void sev_asid_free(struct kvm_sev_info *sev)
 
 	mutex_unlock(&sev_bitmap_lock);
 
-	type = sev->es_active ? MISC_CG_RES_SEV_ES : MISC_CG_RES_SEV;
-	misc_cg_uncharge(type, sev->misc_cg, 1);
+	sev_misc_cg_uncharge(sev);
 	put_misc_cg(sev->misc_cg);
 	sev->misc_cg = NULL;
 }
@@ -590,7 +598,7 @@ static int sev_es_sync_vmsa(struct vcpu_svm *svm)
 	 * traditional VMSA as it has been built so far (in prep
 	 * for LAUNCH_UPDATE_VMSA) to be the initial SEV-ES state.
 	 */
-	memcpy(svm->vmsa, save, sizeof(*save));
+	memcpy(svm->sev_es.vmsa, save, sizeof(*save));
 
 	return 0;
 }
@@ -612,11 +620,11 @@ static int __sev_launch_update_vmsa(struct kvm *kvm, struct kvm_vcpu *vcpu,
 	 * the VMSA memory content (i.e it will write the same memory region
 	 * with the guest's key), so invalidate it first.
 	 */
-	clflush_cache_range(svm->vmsa, PAGE_SIZE);
+	clflush_cache_range(svm->sev_es.vmsa, PAGE_SIZE);
 
 	vmsa.reserved = 0;
 	vmsa.handle = to_kvm_svm(kvm)->sev_info.handle;
-	vmsa.address = __sme_pa(svm->vmsa);
+	vmsa.address = __sme_pa(svm->sev_es.vmsa);
 	vmsa.len = PAGE_SIZE;
 	ret = sev_issue_cmd(kvm, SEV_CMD_LAUNCH_UPDATE_VMSA, &vmsa, error);
 	if (ret)
@@ -1536,6 +1544,204 @@ static bool cmd_allowed_from_miror(u32 cmd_id)
 	return false;
 }
 
+static int sev_lock_for_migration(struct kvm *kvm)
+{
+	struct kvm_sev_info *sev = &to_kvm_svm(kvm)->sev_info;
+
+	/*
+	 * Bail if this VM is already involved in a migration to avoid deadlock
+	 * between two VMs trying to migrate to/from each other.
+	 */
+	if (atomic_cmpxchg_acquire(&sev->migration_in_progress, 0, 1))
+		return -EBUSY;
+
+	mutex_lock(&kvm->lock);
+
+	return 0;
+}
+
+static void sev_unlock_after_migration(struct kvm *kvm)
+{
+	struct kvm_sev_info *sev = &to_kvm_svm(kvm)->sev_info;
+
+	mutex_unlock(&kvm->lock);
+	atomic_set_release(&sev->migration_in_progress, 0);
+}
+
+
+static int sev_lock_vcpus_for_migration(struct kvm *kvm)
+{
+	struct kvm_vcpu *vcpu;
+	int i, j;
+
+	kvm_for_each_vcpu(i, vcpu, kvm) {
+		if (mutex_lock_killable(&vcpu->mutex))
+			goto out_unlock;
+	}
+
+	return 0;
+
+out_unlock:
+	kvm_for_each_vcpu(j, vcpu, kvm) {
+		if (i == j)
+			break;
+
+		mutex_unlock(&vcpu->mutex);
+	}
+	return -EINTR;
+}
+
+static void sev_unlock_vcpus_for_migration(struct kvm *kvm)
+{
+	struct kvm_vcpu *vcpu;
+	int i;
+
+	kvm_for_each_vcpu(i, vcpu, kvm) {
+		mutex_unlock(&vcpu->mutex);
+	}
+}
+
+static void sev_migrate_from(struct kvm_sev_info *dst,
+			      struct kvm_sev_info *src)
+{
+	dst->active = true;
+	dst->asid = src->asid;
+	dst->handle = src->handle;
+	dst->pages_locked = src->pages_locked;
+
+	src->asid = 0;
+	src->active = false;
+	src->handle = 0;
+	src->pages_locked = 0;
+
+	if (dst->misc_cg != src->misc_cg)
+		sev_misc_cg_uncharge(src);
+
+	put_misc_cg(src->misc_cg);
+	src->misc_cg = NULL;
+
+	INIT_LIST_HEAD(&dst->regions_list);
+	list_replace_init(&src->regions_list, &dst->regions_list);
+}
+
+static int sev_es_migrate_from(struct kvm *dst, struct kvm *src)
+{
+	int i;
+	struct kvm_vcpu *dst_vcpu, *src_vcpu;
+	struct vcpu_svm *dst_svm, *src_svm;
+
+	if (atomic_read(&src->online_vcpus) != atomic_read(&dst->online_vcpus))
+		return -EINVAL;
+
+	kvm_for_each_vcpu(i, src_vcpu, src) {
+		if (!src_vcpu->arch.guest_state_protected)
+			return -EINVAL;
+	}
+
+	kvm_for_each_vcpu(i, src_vcpu, src) {
+		src_svm = to_svm(src_vcpu);
+		dst_vcpu = kvm_get_vcpu(dst, i);
+		dst_svm = to_svm(dst_vcpu);
+
+		/*
+		 * Transfer VMSA and GHCB state to the destination.  Nullify and
+		 * clear source fields as appropriate, the state now belongs to
+		 * the destination.
+		 */
+		memcpy(&dst_svm->sev_es, &src_svm->sev_es, sizeof(src_svm->sev_es));
+		dst_svm->vmcb->control.ghcb_gpa = src_svm->vmcb->control.ghcb_gpa;
+		dst_svm->vmcb->control.vmsa_pa = src_svm->vmcb->control.vmsa_pa;
+		dst_vcpu->arch.guest_state_protected = true;
+
+		memset(&src_svm->sev_es, 0, sizeof(src_svm->sev_es));
+		src_svm->vmcb->control.ghcb_gpa = INVALID_PAGE;
+		src_svm->vmcb->control.vmsa_pa = INVALID_PAGE;
+		src_vcpu->arch.guest_state_protected = false;
+	}
+	to_kvm_svm(src)->sev_info.es_active = false;
+	to_kvm_svm(dst)->sev_info.es_active = true;
+
+	return 0;
+}
+
+int svm_vm_migrate_from(struct kvm *kvm, unsigned int source_fd)
+{
+	struct kvm_sev_info *dst_sev = &to_kvm_svm(kvm)->sev_info;
+	struct kvm_sev_info *src_sev;
+	struct file *source_kvm_file;
+	struct kvm *source_kvm;
+	int ret;
+
+	ret = sev_lock_for_migration(kvm);
+	if (ret)
+		return ret;
+
+	if (sev_guest(kvm)) {
+		ret = -EINVAL;
+		goto out_unlock;
+	}
+
+	source_kvm_file = fget(source_fd);
+	if (!file_is_kvm(source_kvm_file)) {
+		ret = -EBADF;
+		goto out_fput;
+	}
+
+	source_kvm = source_kvm_file->private_data;
+	ret = sev_lock_for_migration(source_kvm);
+	if (ret)
+		goto out_fput;
+
+	if (!sev_guest(source_kvm)) {
+		ret = -EINVAL;
+		goto out_source;
+	}
+
+	src_sev = &to_kvm_svm(source_kvm)->sev_info;
+	dst_sev->misc_cg = get_current_misc_cg();
+	if (dst_sev->misc_cg != src_sev->misc_cg) {
+		ret = sev_misc_cg_try_charge(dst_sev);
+		if (ret)
+			goto out_dst_put_cgroup;
+	}
+
+	ret = sev_lock_vcpus_for_migration(kvm);
+	if (ret)
+		goto out_dst_cgroup;
+	ret = sev_lock_vcpus_for_migration(source_kvm);
+	if (ret)
+		goto out_dst_vcpu;
+
+	if (sev_es_guest(source_kvm)) {
+		ret = sev_es_migrate_from(kvm, source_kvm);
+		if (ret)
+			goto out_source_vcpu;
+	}
+	sev_migrate_from(dst_sev, src_sev);
+	kvm_vm_dead(source_kvm);
+	ret = 0;
+
+out_source_vcpu:
+	sev_unlock_vcpus_for_migration(source_kvm);
+out_dst_vcpu:
+	sev_unlock_vcpus_for_migration(kvm);
+out_dst_cgroup:
+	if (ret < 0) {
+		sev_misc_cg_uncharge(dst_sev);
+out_dst_put_cgroup:
+		put_misc_cg(dst_sev->misc_cg);
+		dst_sev->misc_cg = NULL;
+	}
+out_source:
+	sev_unlock_after_migration(source_kvm);
+out_fput:
+	if (source_kvm_file)
+		fput(source_kvm_file);
+out_unlock:
+	sev_unlock_after_migration(kvm);
+	return ret;
+}
+
 int svm_mem_enc_op(struct kvm *kvm, void __user *argp)
 {
 	struct kvm_sev_cmd sev_cmd;
@@ -2038,16 +2244,16 @@ void sev_free_vcpu(struct kvm_vcpu *vcpu)
 	svm = to_svm(vcpu);
 
 	if (vcpu->arch.guest_state_protected)
-		sev_flush_guest_memory(svm, svm->vmsa, PAGE_SIZE);
-	__free_page(virt_to_page(svm->vmsa));
+		sev_flush_guest_memory(svm, svm->sev_es.vmsa, PAGE_SIZE);
+	__free_page(virt_to_page(svm->sev_es.vmsa));
 
-	if (svm->ghcb_sa_free)
-		kfree(svm->ghcb_sa);
+	if (svm->sev_es.ghcb_sa_free)
+		kfree(svm->sev_es.ghcb_sa);
 }
 
 static void dump_ghcb(struct vcpu_svm *svm)
 {
-	struct ghcb *ghcb = svm->ghcb;
+	struct ghcb *ghcb = svm->sev_es.ghcb;
 	unsigned int nbits;
 
 	/* Re-use the dump_invalid_vmcb module parameter */
@@ -2073,7 +2279,7 @@ static void dump_ghcb(struct vcpu_svm *svm)
 static void sev_es_sync_to_ghcb(struct vcpu_svm *svm)
 {
 	struct kvm_vcpu *vcpu = &svm->vcpu;
-	struct ghcb *ghcb = svm->ghcb;
+	struct ghcb *ghcb = svm->sev_es.ghcb;
 
 	/*
 	 * The GHCB protocol so far allows for the following data
@@ -2093,7 +2299,7 @@ static void sev_es_sync_from_ghcb(struct vcpu_svm *svm)
 {
 	struct vmcb_control_area *control = &svm->vmcb->control;
 	struct kvm_vcpu *vcpu = &svm->vcpu;
-	struct ghcb *ghcb = svm->ghcb;
+	struct ghcb *ghcb = svm->sev_es.ghcb;
 	u64 exit_code;
 
 	/*
@@ -2140,7 +2346,7 @@ static int sev_es_validate_vmgexit(struct vcpu_svm *svm)
 	struct ghcb *ghcb;
 	u64 exit_code = 0;
 
-	ghcb = svm->ghcb;
+	ghcb = svm->sev_es.ghcb;
 
 	/* Only GHCB Usage code 0 is supported */
 	if (ghcb->ghcb_usage)
@@ -2258,33 +2464,34 @@ vmgexit_err:
 
 void sev_es_unmap_ghcb(struct vcpu_svm *svm)
 {
-	if (!svm->ghcb)
+	if (!svm->sev_es.ghcb)
 		return;
 
-	if (svm->ghcb_sa_free) {
+	if (svm->sev_es.ghcb_sa_free) {
 		/*
 		 * The scratch area lives outside the GHCB, so there is a
 		 * buffer that, depending on the operation performed, may
 		 * need to be synced, then freed.
 		 */
-		if (svm->ghcb_sa_sync) {
+		if (svm->sev_es.ghcb_sa_sync) {
 			kvm_write_guest(svm->vcpu.kvm,
-					ghcb_get_sw_scratch(svm->ghcb),
-					svm->ghcb_sa, svm->ghcb_sa_len);
-			svm->ghcb_sa_sync = false;
+					ghcb_get_sw_scratch(svm->sev_es.ghcb),
+					svm->sev_es.ghcb_sa,
+					svm->sev_es.ghcb_sa_len);
+			svm->sev_es.ghcb_sa_sync = false;
 		}
 
-		kfree(svm->ghcb_sa);
-		svm->ghcb_sa = NULL;
-		svm->ghcb_sa_free = false;
+		kfree(svm->sev_es.ghcb_sa);
+		svm->sev_es.ghcb_sa = NULL;
+		svm->sev_es.ghcb_sa_free = false;
 	}
 
-	trace_kvm_vmgexit_exit(svm->vcpu.vcpu_id, svm->ghcb);
+	trace_kvm_vmgexit_exit(svm->vcpu.vcpu_id, svm->sev_es.ghcb);
 
 	sev_es_sync_to_ghcb(svm);
 
-	kvm_vcpu_unmap(&svm->vcpu, &svm->ghcb_map, true);
-	svm->ghcb = NULL;
+	kvm_vcpu_unmap(&svm->vcpu, &svm->sev_es.ghcb_map, true);
+	svm->sev_es.ghcb = NULL;
 }
 
 void pre_sev_run(struct vcpu_svm *svm, int cpu)
@@ -2314,7 +2521,7 @@ void pre_sev_run(struct vcpu_svm *svm, int cpu)
 static bool setup_vmgexit_scratch(struct vcpu_svm *svm, bool sync, u64 len)
 {
 	struct vmcb_control_area *control = &svm->vmcb->control;
-	struct ghcb *ghcb = svm->ghcb;
+	struct ghcb *ghcb = svm->sev_es.ghcb;
 	u64 ghcb_scratch_beg, ghcb_scratch_end;
 	u64 scratch_gpa_beg, scratch_gpa_end;
 	void *scratch_va;
@@ -2350,7 +2557,7 @@ static bool setup_vmgexit_scratch(struct vcpu_svm *svm, bool sync, u64 len)
 			return false;
 		}
 
-		scratch_va = (void *)svm->ghcb;
+		scratch_va = (void *)svm->sev_es.ghcb;
 		scratch_va += (scratch_gpa_beg - control->ghcb_gpa);
 	} else {
 		/*
@@ -2380,12 +2587,12 @@ static bool setup_vmgexit_scratch(struct vcpu_svm *svm, bool sync, u64 len)
 		 * the vCPU next time (i.e. a read was requested so the data
 		 * must be written back to the guest memory).
 		 */
-		svm->ghcb_sa_sync = sync;
-		svm->ghcb_sa_free = true;
+		svm->sev_es.ghcb_sa_sync = sync;
+		svm->sev_es.ghcb_sa_free = true;
 	}
 
-	svm->ghcb_sa = scratch_va;
-	svm->ghcb_sa_len = len;
+	svm->sev_es.ghcb_sa = scratch_va;
+	svm->sev_es.ghcb_sa_len = len;
 
 	return true;
 }
@@ -2504,15 +2711,15 @@ int sev_handle_vmgexit(struct kvm_vcpu *vcpu)
 		return -EINVAL;
 	}
 
-	if (kvm_vcpu_map(vcpu, ghcb_gpa >> PAGE_SHIFT, &svm->ghcb_map)) {
+	if (kvm_vcpu_map(vcpu, ghcb_gpa >> PAGE_SHIFT, &svm->sev_es.ghcb_map)) {
 		/* Unable to map GHCB from guest */
 		vcpu_unimpl(vcpu, "vmgexit: error mapping GHCB [%#llx] from guest\n",
 			    ghcb_gpa);
 		return -EINVAL;
 	}
 
-	svm->ghcb = svm->ghcb_map.hva;
-	ghcb = svm->ghcb_map.hva;
+	svm->sev_es.ghcb = svm->sev_es.ghcb_map.hva;
+	ghcb = svm->sev_es.ghcb_map.hva;
 
 	trace_kvm_vmgexit_enter(vcpu->vcpu_id, ghcb);
 
@@ -2535,7 +2742,7 @@ int sev_handle_vmgexit(struct kvm_vcpu *vcpu)
 		ret = kvm_sev_es_mmio_read(vcpu,
 					   control->exit_info_1,
 					   control->exit_info_2,
-					   svm->ghcb_sa);
+					   svm->sev_es.ghcb_sa);
 		break;
 	case SVM_VMGEXIT_MMIO_WRITE:
 		if (!setup_vmgexit_scratch(svm, false, control->exit_info_2))
@@ -2544,7 +2751,7 @@ int sev_handle_vmgexit(struct kvm_vcpu *vcpu)
 		ret = kvm_sev_es_mmio_write(vcpu,
 					    control->exit_info_1,
 					    control->exit_info_2,
-					    svm->ghcb_sa);
+					    svm->sev_es.ghcb_sa);
 		break;
 	case SVM_VMGEXIT_NMI_COMPLETE:
 		ret = svm_invoke_exit_handler(vcpu, SVM_EXIT_IRET);
@@ -2604,7 +2811,8 @@ int sev_es_string_io(struct vcpu_svm *svm, int size, unsigned int port, int in)
 	if (!setup_vmgexit_scratch(svm, in, bytes))
 		return -EINVAL;
 
-	return kvm_sev_es_string_io(&svm->vcpu, size, port, svm->ghcb_sa, count, in);
+	return kvm_sev_es_string_io(&svm->vcpu, size, port, svm->sev_es.ghcb_sa,
+				    count, in);
 }
 
 void sev_es_init_vmcb(struct vcpu_svm *svm)
@@ -2619,7 +2827,7 @@ void sev_es_init_vmcb(struct vcpu_svm *svm)
 	 * VMCB page. Do not include the encryption mask on the VMSA physical
 	 * address since hardware will access it using the guest key.
 	 */
-	svm->vmcb->control.vmsa_pa = __pa(svm->vmsa);
+	svm->vmcb->control.vmsa_pa = __pa(svm->sev_es.vmsa);
 
 	/* Can't intercept CR register access, HV can't modify CR registers */
 	svm_clr_intercept(svm, INTERCEPT_CR0_READ);
@@ -2691,8 +2899,8 @@ void sev_vcpu_deliver_sipi_vector(struct kvm_vcpu *vcpu, u8 vector)
 	struct vcpu_svm *svm = to_svm(vcpu);
 
 	/* First SIPI: Use the values as initially set by the VMM */
-	if (!svm->received_first_sipi) {
-		svm->received_first_sipi = true;
+	if (!svm->sev_es.received_first_sipi) {
+		svm->sev_es.received_first_sipi = true;
 		return;
 	}
 
@@ -2701,8 +2909,8 @@ void sev_vcpu_deliver_sipi_vector(struct kvm_vcpu *vcpu, u8 vector)
 	 * the guest will set the CS and RIP. Set SW_EXIT_INFO_2 to a
 	 * non-zero value.
 	 */
-	if (!svm->ghcb)
+	if (!svm->sev_es.ghcb)
 		return;
 
-	ghcb_set_sw_exit_info_2(svm->ghcb, 1);
+	ghcb_set_sw_exit_info_2(svm->sev_es.ghcb, 1);
 }