KVM: arm64: Return guest memory from EL2 via dedicated teardown memcache

Rather than relying on the host to free the previously-donated pKVM
hypervisor VM pages explicitly on teardown, introduce a dedicated
teardown memcache which allows the host to reclaim guest memory
resources without having to keep track of all of the allocations made by
the pKVM hypervisor at EL2.

Tested-by: Vincent Donnefort <vdonnefort@google.com>
Co-developed-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Fuad Tabba <tabba@google.com>
Signed-off-by: Quentin Perret <qperret@google.com>
Signed-off-by: Will Deacon <will@kernel.org>
[maz: dropped __maybe_unused from unmap_donated_memory_noclear()]
Signed-off-by: Marc Zyngier <maz@kernel.org>
Link: https://lore.kernel.org/r/20221110190259.26861-21-will@kernel.org

3 files changed, 33 insertions, 13 deletions

diff --git a/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h b/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h
index 420b87e755a4..b7bdbe63deed 100644
--- a/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h
+++ b/arch/arm64/kvm/hyp/include/nvhe/mem_protect.h
@@ -76,7 +76,7 @@ void handle_host_mem_abort(struct kvm_cpu_context *host_ctxt);
 
 int hyp_pin_shared_mem(void *from, void *to);
 void hyp_unpin_shared_mem(void *from, void *to);
-void reclaim_guest_pages(struct pkvm_hyp_vm *vm);
+void reclaim_guest_pages(struct pkvm_hyp_vm *vm, struct kvm_hyp_memcache *mc);
 int refill_memcache(struct kvm_hyp_memcache *mc, unsigned long min_pages,
 		    struct kvm_hyp_memcache *host_mc);
 
diff --git a/arch/arm64/kvm/hyp/nvhe/mem_protect.c b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
index 0162afba6dc4..94cd48f7850e 100644
--- a/arch/arm64/kvm/hyp/nvhe/mem_protect.c
+++ b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
@@ -260,19 +260,24 @@ int kvm_guest_prepare_stage2(struct pkvm_hyp_vm *vm, void *pgd)
 	return 0;
 }
 
-void reclaim_guest_pages(struct pkvm_hyp_vm *vm)
+void reclaim_guest_pages(struct pkvm_hyp_vm *vm, struct kvm_hyp_memcache *mc)
 {
-	void *pgd = vm->pgt.pgd;
-	unsigned long nr_pages;
-
-	nr_pages = kvm_pgtable_stage2_pgd_size(vm->kvm.arch.vtcr) >> PAGE_SHIFT;
+	void *addr;
 
+	/* Dump all pgtable pages in the hyp_pool */
 	guest_lock_component(vm);
 	kvm_pgtable_stage2_destroy(&vm->pgt);
 	vm->kvm.arch.mmu.pgd_phys = 0ULL;
 	guest_unlock_component(vm);
 
-	WARN_ON(__pkvm_hyp_donate_host(hyp_virt_to_pfn(pgd), nr_pages));
+	/* Drain the hyp_pool into the memcache */
+	addr = hyp_alloc_pages(&vm->pool, 0);
+	while (addr) {
+		memset(hyp_virt_to_page(addr), 0, sizeof(struct hyp_page));
+		push_hyp_memcache(mc, addr, hyp_virt_to_phys);
+		WARN_ON(__pkvm_hyp_donate_host(hyp_virt_to_pfn(addr), 1));
+		addr = hyp_alloc_pages(&vm->pool, 0);
+	}
 }
 
 int __pkvm_prot_finalize(void)
diff --git a/arch/arm64/kvm/hyp/nvhe/pkvm.c b/arch/arm64/kvm/hyp/nvhe/pkvm.c
index 0768307566d4..81835c2f4c5a 100644
--- a/arch/arm64/kvm/hyp/nvhe/pkvm.c
+++ b/arch/arm64/kvm/hyp/nvhe/pkvm.c
@@ -393,7 +393,7 @@ static void unmap_donated_memory(void *va, size_t size)
 	__unmap_donated_memory(va, size);
 }
 
-static void __maybe_unused unmap_donated_memory_noclear(void *va, size_t size)
+static void unmap_donated_memory_noclear(void *va, size_t size)
 {
 	if (!va)
 		return;
@@ -527,8 +527,21 @@ unlock:
 	return ret;
 }
 
+static void
+teardown_donated_memory(struct kvm_hyp_memcache *mc, void *addr, size_t size)
+{
+	size = PAGE_ALIGN(size);
+	memset(addr, 0, size);
+
+	for (void *start = addr; start < addr + size; start += PAGE_SIZE)
+		push_hyp_memcache(mc, start, hyp_virt_to_phys);
+
+	unmap_donated_memory_noclear(addr, size);
+}
+
 int __pkvm_teardown_vm(pkvm_handle_t handle)
 {
+	struct kvm_hyp_memcache *mc;
 	struct pkvm_hyp_vm *hyp_vm;
 	struct kvm *host_kvm;
 	unsigned int idx;
@@ -547,25 +560,27 @@ int __pkvm_teardown_vm(pkvm_handle_t handle)
 		goto err_unlock;
 	}
 
+	host_kvm = hyp_vm->host_kvm;
+
 	/* Ensure the VMID is clean before it can be reallocated */
 	__kvm_tlb_flush_vmid(&hyp_vm->kvm.arch.mmu);
 	remove_vm_table_entry(handle);
 	hyp_spin_unlock(&vm_table_lock);
 
 	/* Reclaim guest pages (including page-table pages) */
-	reclaim_guest_pages(hyp_vm);
+	mc = &host_kvm->arch.pkvm.teardown_mc;
+	reclaim_guest_pages(hyp_vm, mc);
 	unpin_host_vcpus(hyp_vm->vcpus, hyp_vm->nr_vcpus);
 
-	/* Return the metadata pages to the host */
+	/* Push the metadata pages to the teardown memcache */
 	for (idx = 0; idx < hyp_vm->nr_vcpus; ++idx) {
 		struct pkvm_hyp_vcpu *hyp_vcpu = hyp_vm->vcpus[idx];
 
-		unmap_donated_memory(hyp_vcpu, sizeof(*hyp_vcpu));
+		teardown_donated_memory(mc, hyp_vcpu, sizeof(*hyp_vcpu));
 	}
 
-	host_kvm = hyp_vm->host_kvm;
 	vm_size = pkvm_get_hyp_vm_size(hyp_vm->kvm.created_vcpus);
-	unmap_donated_memory(hyp_vm, vm_size);
+	teardown_donated_memory(mc, hyp_vm, vm_size);
 	hyp_unpin_shared_mem(host_kvm, host_kvm + 1);
 	return 0;