KVM: arm64: Instantiate pKVM hypervisor VM and vCPU structures from EL1

With the pKVM hypervisor at EL2 now offering hypercalls to the host for
creating and destroying VM and vCPU structures, plumb these in to the
existing arm64 KVM backend to ensure that the hypervisor data structures
are allocated and initialised on first vCPU run for a pKVM guest.

In the host, 'struct kvm_protected_vm' is introduced to hold the handle
of the pKVM VM instance as well as to track references to the memory
donated to the hypervisor so that it can be freed back to the host
allocator following VM teardown. The stage-2 page-table, hypervisor VM
and vCPU structures are allocated separately so as to avoid the need for
a large physically-contiguous allocation in the host at run-time.

Tested-by: Vincent Donnefort <vdonnefort@google.com>
Signed-off-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221110190259.26861-14-will@kernel.org

1 files changed, 138 insertions, 0 deletions

diff --git a/arch/arm64/kvm/pkvm.c b/arch/arm64/kvm/pkvm.c
index 71493136e59c..8c443b915e43 100644
--- a/arch/arm64/kvm/pkvm.c
+++ b/arch/arm64/kvm/pkvm.c
@@ -6,6 +6,7 @@
 
 #include <linux/kvm_host.h>
 #include <linux/memblock.h>
+#include <linux/mutex.h>
 #include <linux/sort.h>
 
 #include <asm/kvm_pkvm.h>
@@ -94,3 +95,140 @@ void __init kvm_hyp_reserve(void)
 	kvm_info("Reserved %lld MiB at 0x%llx\n", hyp_mem_size >> 20,
 		 hyp_mem_base);
 }
+
+/*
+ * Allocates and donates memory for hypervisor VM structs at EL2.
+ *
+ * Allocates space for the VM state, which includes the hyp vm as well as
+ * the hyp vcpus.
+ *
+ * Stores an opaque handler in the kvm struct for future reference.
+ *
+ * Return 0 on success, negative error code on failure.
+ */
+static int __pkvm_create_hyp_vm(struct kvm *host_kvm)
+{
+	size_t pgd_sz, hyp_vm_sz, hyp_vcpu_sz;
+	struct kvm_vcpu *host_vcpu;
+	pkvm_handle_t handle;
+	void *pgd, *hyp_vm;
+	unsigned long idx;
+	int ret;
+
+	if (host_kvm->created_vcpus < 1)
+		return -EINVAL;
+
+	pgd_sz = kvm_pgtable_stage2_pgd_size(host_kvm->arch.vtcr);
+
+	/*
+	 * The PGD pages will be reclaimed using a hyp_memcache which implies
+	 * page granularity. So, use alloc_pages_exact() to get individual
+	 * refcounts.
+	 */
+	pgd = alloc_pages_exact(pgd_sz, GFP_KERNEL_ACCOUNT);
+	if (!pgd)
+		return -ENOMEM;
+
+	/* Allocate memory to donate to hyp for vm and vcpu pointers. */
+	hyp_vm_sz = PAGE_ALIGN(size_add(PKVM_HYP_VM_SIZE,
+					size_mul(sizeof(void *),
+						 host_kvm->created_vcpus)));
+	hyp_vm = alloc_pages_exact(hyp_vm_sz, GFP_KERNEL_ACCOUNT);
+	if (!hyp_vm) {
+		ret = -ENOMEM;
+		goto free_pgd;
+	}
+
+	/* Donate the VM memory to hyp and let hyp initialize it. */
+	ret = kvm_call_hyp_nvhe(__pkvm_init_vm, host_kvm, hyp_vm, pgd);
+	if (ret < 0)
+		goto free_vm;
+
+	handle = ret;
+
+	host_kvm->arch.pkvm.handle = handle;
+	host_kvm->arch.pkvm.hyp_donations.pgd = pgd;
+	host_kvm->arch.pkvm.hyp_donations.vm = hyp_vm;
+
+	/* Donate memory for the vcpus at hyp and initialize it. */
+	hyp_vcpu_sz = PAGE_ALIGN(PKVM_HYP_VCPU_SIZE);
+	kvm_for_each_vcpu(idx, host_vcpu, host_kvm) {
+		void *hyp_vcpu;
+
+		/* Indexing of the vcpus to be sequential starting at 0. */
+		if (WARN_ON(host_vcpu->vcpu_idx != idx)) {
+			ret = -EINVAL;
+			goto destroy_vm;
+		}
+
+		hyp_vcpu = alloc_pages_exact(hyp_vcpu_sz, GFP_KERNEL_ACCOUNT);
+		if (!hyp_vcpu) {
+			ret = -ENOMEM;
+			goto destroy_vm;
+		}
+
+		host_kvm->arch.pkvm.hyp_donations.vcpus[idx] = hyp_vcpu;
+
+		ret = kvm_call_hyp_nvhe(__pkvm_init_vcpu, handle, host_vcpu,
+					hyp_vcpu);
+		if (ret)
+			goto destroy_vm;
+	}
+
+	return 0;
+
+destroy_vm:
+	pkvm_destroy_hyp_vm(host_kvm);
+	return ret;
+free_vm:
+	free_pages_exact(hyp_vm, hyp_vm_sz);
+free_pgd:
+	free_pages_exact(pgd, pgd_sz);
+	return ret;
+}
+
+int pkvm_create_hyp_vm(struct kvm *host_kvm)
+{
+	int ret = 0;
+
+	mutex_lock(&host_kvm->lock);
+	if (!host_kvm->arch.pkvm.handle)
+		ret = __pkvm_create_hyp_vm(host_kvm);
+	mutex_unlock(&host_kvm->lock);
+
+	return ret;
+}
+
+void pkvm_destroy_hyp_vm(struct kvm *host_kvm)
+{
+	unsigned long idx, nr_vcpus = host_kvm->created_vcpus;
+	size_t pgd_sz, hyp_vm_sz;
+
+	if (host_kvm->arch.pkvm.handle)
+		WARN_ON(kvm_call_hyp_nvhe(__pkvm_teardown_vm,
+					  host_kvm->arch.pkvm.handle));
+
+	host_kvm->arch.pkvm.handle = 0;
+
+	for (idx = 0; idx < nr_vcpus; ++idx) {
+		void *hyp_vcpu = host_kvm->arch.pkvm.hyp_donations.vcpus[idx];
+
+		if (!hyp_vcpu)
+			break;
+
+		free_pages_exact(hyp_vcpu, PAGE_ALIGN(PKVM_HYP_VCPU_SIZE));
+	}
+
+	hyp_vm_sz = PAGE_ALIGN(size_add(PKVM_HYP_VM_SIZE,
+					size_mul(sizeof(void *), nr_vcpus)));
+	pgd_sz = kvm_pgtable_stage2_pgd_size(host_kvm->arch.vtcr);
+
+	free_pages_exact(host_kvm->arch.pkvm.hyp_donations.vm, hyp_vm_sz);
+	free_pages_exact(host_kvm->arch.pkvm.hyp_donations.pgd, pgd_sz);
+}
+
+int pkvm_init_host_vm(struct kvm *host_kvm)
+{
+	mutex_init(&host_kvm->lock);
+	return 0;
+}