From patchwork Wed Sep 14 08:34:55 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Will Deacon X-Patchwork-Id: 12975854 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id CD621ECAAD8 for ; Wed, 14 Sep 2022 08:54:00 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender: Content-Transfer-Encoding:Content-Type:List-Subscribe:List-Help:List-Post: List-Archive:List-Unsubscribe:List-Id:MIME-Version:References:In-Reply-To: Message-Id:Date:Subject:Cc:To:From:Reply-To:Content-ID:Content-Description: Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID: List-Owner; bh=9SctE6J3zRIgrbAwW5oDmzypELYpLOs4TsaeT/DeFWI=; b=KBwoxaggrZEHJH Tz/MtqU4vNTRBApR1T7KtAg8eks0m3jqERE+dupbIjpq5wK0bZ/4nRLATESB5MnJ5FXn8Chus753f YzSxyEJDPSqA3GKU2HNyZBXrAw8ss2pz0PDucCnSEryGsnUc7ZzknjLeswk1DsKZv47ubApgMihtT MPUtfPqRx4O/0miPYw+SVK8PMBln1TtEG2Tsx0wFWzb9EvgRyHoRSEdS/hBTX9S5vPH4lgQ8YpRfb x4jzu7zIeaVkKif2TX3qS0Hq7DDt/B9TP5cKckci1t8x7FgBd3MSylVQ9SS9+Iwp6UqvN8RrLrYAQ z3CCFSqMS3qYplG52w2Q==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1oYO8O-00Dj2F-V6; Wed, 14 Sep 2022 08:52:41 +0000 Received: from ams.source.kernel.org ([2604:1380:4601:e00::1]) by bombadil.infradead.org with esmtps (Exim 4.94.2 #2 (Red Hat Linux)) id 1oYNsd-00DcRf-2H for linux-arm-kernel@lists.infradead.org; Wed, 14 Sep 2022 08:36:31 +0000 Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id 71341B81689; Wed, 14 Sep 2022 08:36:21 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 2B057C4347C; Wed, 14 Sep 2022 08:36:17 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1663144580; bh=MTvQFSZG4vCk/H/YW0N6L6nwP5YUiWUxFHraR/uYdFw=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=XEvdyst9lDq77hhSi7TMv2GqCwunp0NUX7th58GNsjaDevVrDYdt45aj+HYJvzoFB KwScNa1RN9OY3NfoDK6dL5DSB4P3WwcdiQZ72MXnxt/aPRgLNTge59ywyC4vr7UaDj VWT2Rpw3NVjGJf8+u/Km+SUPdvc1SgZ4USdaEqirCp4XnNyqyI346kOhXNd5iKTeTe O5V1TQD1PQvtkJ7rdtmsR+ET1tCVJ11IaTbJ6s16f3GIvvQvDhIx31z653AyfnBc+9 xQZOSjq2lO8pKwTt6RhW453fKa19LJGeS8fgXZNXC3h3xA90nxgtDNZ2tzR0swJmxj 0sBL2hsMmdAfA== From: Will Deacon To: kvmarm@lists.cs.columbia.edu Cc: Will Deacon , Sean Christopherson , Vincent Donnefort , Alexandru Elisei , Catalin Marinas , James Morse , Chao Peng , Quentin Perret , Suzuki K Poulose , Mark Rutland , Fuad Tabba , Oliver Upton , Marc Zyngier , kernel-team@android.com, kvm@vger.kernel.org, linux-arm-kernel@lists.infradead.org Subject: [PATCH v3 20/25] KVM: arm64: Return guest memory from EL2 via dedicated teardown memcache Date: Wed, 14 Sep 2022 09:34:55 +0100 Message-Id: <20220914083500.5118-21-will@kernel.org> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20220914083500.5118-1-will@kernel.org> References: <20220914083500.5118-1-will@kernel.org> MIME-Version: 1.0 X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20220914_013623_498151_49FD63BD X-CRM114-Status: GOOD ( 19.56 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org From: Quentin Perret 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. Signed-off-by: Quentin Perret Signed-off-by: Will Deacon --- arch/arm64/include/asm/kvm_host.h | 7 +---- arch/arm64/kvm/hyp/include/nvhe/mem_protect.h | 2 +- arch/arm64/kvm/hyp/nvhe/mem_protect.c | 17 ++++++---- arch/arm64/kvm/hyp/nvhe/pkvm.c | 20 ++++++++++-- arch/arm64/kvm/pkvm.c | 31 ++++--------------- 5 files changed, 36 insertions(+), 41 deletions(-) diff --git a/arch/arm64/include/asm/kvm_host.h b/arch/arm64/include/asm/kvm_host.h index 93f25231b247..362fe3073970 100644 --- a/arch/arm64/include/asm/kvm_host.h +++ b/arch/arm64/include/asm/kvm_host.h @@ -177,12 +177,7 @@ typedef unsigned int pkvm_handle_t; struct kvm_protected_vm { pkvm_handle_t handle; struct mutex vm_lock; - - struct { - void *pgd; - void *vm; - void *vcpus[KVM_MAX_VCPUS]; - } hyp_donations; + struct kvm_hyp_memcache teardown_mc; }; struct kvm_arch { 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 27b16a6b85bb..ffa56a89acdb 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 6ff78118a140..09e221ea6184 100644 --- a/arch/arm64/kvm/hyp/nvhe/pkvm.c +++ b/arch/arm64/kvm/hyp/nvhe/pkvm.c @@ -539,8 +539,21 @@ int __pkvm_init_vcpu(pkvm_handle_t handle, struct kvm_vcpu *host_vcpu, 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; unsigned int idx; int err; @@ -563,7 +576,8 @@ int __pkvm_teardown_vm(pkvm_handle_t handle) hyp_spin_unlock(&vm_table_lock); /* Reclaim guest pages (including page-table pages) */ - reclaim_guest_pages(hyp_vm); + mc = &hyp_vm->host_kvm->arch.pkvm.teardown_mc; + reclaim_guest_pages(hyp_vm, mc); unpin_host_vcpus(hyp_vm->vcpus, hyp_vm->nr_vcpus); /* Push the metadata pages to the teardown memcache */ @@ -572,10 +586,10 @@ int __pkvm_teardown_vm(pkvm_handle_t handle) 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)); } - unmap_donated_memory(hyp_vm, hyp_vm->donated_memory_size); + teardown_donated_memory(mc, hyp_vm, hyp_vm->donated_memory_size); return 0; err_unlock: diff --git a/arch/arm64/kvm/pkvm.c b/arch/arm64/kvm/pkvm.c index 754632a608e3..a9953db08592 100644 --- a/arch/arm64/kvm/pkvm.c +++ b/arch/arm64/kvm/pkvm.c @@ -147,8 +147,6 @@ static int __pkvm_create_hyp_vm(struct kvm *host_kvm) 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); @@ -167,12 +165,12 @@ static int __pkvm_create_hyp_vm(struct kvm *host_kvm) 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) + if (ret) { + free_pages_exact(hyp_vcpu, hyp_vcpu_sz); goto destroy_vm; + } } return 0; @@ -201,30 +199,13 @@ int pkvm_create_hyp_vm(struct kvm *host_kvm) 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) + 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); + host_kvm->arch.pkvm.handle = 0; + free_hyp_memcache(&host_kvm->arch.pkvm.teardown_mc); } int pkvm_init_host_vm(struct kvm *host_kvm)