From patchwork Mon Aug 21 20:35:30 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: =?utf-8?b?UmFkaW0gS3LEjW3DocWZ?= X-Patchwork-Id: 804163 Return-Path: X-Original-To: incoming-imx@patchwork.ozlabs.org Delivered-To: patchwork-incoming-imx@bilbo.ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=lists.infradead.org (client-ip=65.50.211.133; helo=bombadil.infradead.org; envelope-from=linux-arm-kernel-bounces+incoming-imx=patchwork.ozlabs.org@lists.infradead.org; receiver=) Authentication-Results: ozlabs.org; dkim=pass (2048-bit key; unprotected) header.d=lists.infradead.org header.i=@lists.infradead.org header.b="WalaGkrh"; dkim-atps=neutral Received: from bombadil.infradead.org (bombadil.infradead.org [65.50.211.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ozlabs.org (Postfix) with ESMTPS id 3xblv05tT8z9t16 for ; Tue, 22 Aug 2017 06:40:28 +1000 (AEST) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20170209; h=Sender: Content-Transfer-Encoding:Content-Type:Cc:List-Subscribe:List-Help:List-Post: List-Archive:List-Unsubscribe:List-Id:MIME-Version:References:In-Reply-To: Message-Id:Date:Subject:To:From:Reply-To:Content-ID:Content-Description: Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID: List-Owner; bh=Y/saYSYzNn8RaX4u2+i8uMiKRlk7slmYuLhUbJA7cpw=; b=WalaGkrh3VK05J GifPEI7taulJj/nsJcSfmWFGRGAQHjutMbNNmtSAsiFKdEaptjIalis2m1znxivxS9/GmxK3i1RtG 5asWhYp91dZ5478IhXm57FnzKAqUFg+691wsvXW+7Jo5HvmiF6WBMoKbxBJRVTUf7RIJvDUf8bvH5 r/dbxAHWrsZ2iz7BNhhhGm/+Ihf40eHYjspJ3ZZZnFOtZgVZjmFJDhUk5WfCNI5Cr59AV7D66fgLo BQWv0yK+3b21fEEpJKrC1NK+vLdVJFq3ixcs7a0sliM2gDKPqxQjYU8AvSKte9uh3zeZBQEIiOffE laVv+rFVNVBev1487XVg==; Received: from localhost ([127.0.0.1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.87 #1 (Red Hat Linux)) id 1djtUn-0005qW-8Q; Mon, 21 Aug 2017 20:40:25 +0000 Received: from mx1.redhat.com ([209.132.183.28]) by bombadil.infradead.org with esmtps (Exim 4.87 #1 (Red Hat Linux)) id 1djtTd-0003QN-TH for linux-arm-kernel@lists.infradead.org; Mon, 21 Aug 2017 20:39:16 +0000 Received: from smtp.corp.redhat.com (int-mx01.intmail.prod.int.phx2.redhat.com [10.5.11.11]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 629BE61467; Mon, 21 Aug 2017 20:38:56 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mx1.redhat.com 629BE61467 Authentication-Results: ext-mx10.extmail.prod.ext.phx2.redhat.com; dmarc=none (p=none dis=none) header.from=redhat.com Authentication-Results: ext-mx10.extmail.prod.ext.phx2.redhat.com; spf=fail smtp.mailfrom=rkrcmar@redhat.com Received: from flask (unknown [10.43.2.80]) by smtp.corp.redhat.com (Postfix) with SMTP id D7E4F5C460; Mon, 21 Aug 2017 20:38:46 +0000 (UTC) Received: by flask (sSMTP sendmail emulation); Mon, 21 Aug 2017 22:38:46 +0200 From: =?UTF-8?q?Radim=20Kr=C4=8Dm=C3=A1=C5=99?= To: linux-kernel@vger.kernel.org, kvm@vger.kernel.org, linux-mips@linux-mips.org, kvm-ppc@vger.kernel.org, linux-s390@vger.kernel.org, linux-arm-kernel@lists.infradead.org Subject: [PATCH RFC v3 9/9] KVM: split kvm->vcpus into chunks Date: Mon, 21 Aug 2017 22:35:30 +0200 Message-Id: <20170821203530.9266-10-rkrcmar@redhat.com> In-Reply-To: <20170821203530.9266-1-rkrcmar@redhat.com> References: <20170821203530.9266-1-rkrcmar@redhat.com> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.11 X-Greylist: Sender IP whitelisted, not delayed by milter-greylist-4.5.16 (mx1.redhat.com [10.5.110.39]); Mon, 21 Aug 2017 20:38:56 +0000 (UTC) X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20170821_133914_132724_32C8908B X-CRM114-Status: GOOD ( 18.35 ) X-Spam-Score: -6.9 (------) X-Spam-Report: SpamAssassin version 3.4.1 on bombadil.infradead.org summary: Content analysis details: (-6.9 points) pts rule name description ---- ---------------------- -------------------------------------------------- -5.0 RCVD_IN_DNSWL_HI RBL: Sender listed at http://www.dnswl.org/, high trust [209.132.183.28 listed in list.dnswl.org] 0.0 HK_RANDOM_FROM From username looks random 0.0 HK_RANDOM_ENVFROM Envelope sender username looks random -0.0 SPF_PASS SPF: sender matches SPF record -0.0 RP_MATCHES_RCVD Envelope sender domain matches handover relay domain -0.0 SPF_HELO_PASS SPF: HELO matches SPF record -1.9 BAYES_00 BODY: Bayes spam probability is 0 to 1% [score: 0.0000] X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.21 Precedence: list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: Christoffer Dall , James Hogan , David Hildenbrand , Marc Zyngier , Cornelia Huck , Paul Mackerras , Christian Borntraeger , Paolo Bonzini , Alexander Graf Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+incoming-imx=patchwork.ozlabs.org@lists.infradead.org List-Id: linux-imx-kernel.lists.patchwork.ozlabs.org This allows us to have high KVM_VCPU_MAX without wasting too much space with small guests. RCU is a viable alternative now that we do not have to protect the kvm_for_each_vcpu() loop. Suggested-by: David Hildenbrand Signed-off-by: Radim Krčmář --- arch/mips/kvm/mips.c | 2 +- arch/x86/kvm/vmx.c | 2 +- include/linux/kvm_host.h | 27 ++++++++++++++++++++------- virt/kvm/kvm_main.c | 27 +++++++++++++++++++++++---- 4 files changed, 45 insertions(+), 13 deletions(-) diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c index c841cb434486..7d452163dcef 100644 --- a/arch/mips/kvm/mips.c +++ b/arch/mips/kvm/mips.c @@ -488,7 +488,7 @@ int kvm_vcpu_ioctl_interrupt(struct kvm_vcpu *vcpu, if (irq->cpu == -1) dvcpu = vcpu; else - dvcpu = vcpu->kvm->vcpus[irq->cpu]; + dvcpu = kvm_get_vcpu(vcpu->kvm, irq->cpu); if (intr == 2 || intr == 3 || intr == 4) { kvm_mips_callbacks->queue_io_int(dvcpu, irq); diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index ae0f04e26fec..2b92c2de2b3a 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -11741,7 +11741,7 @@ static int vmx_update_pi_irte(struct kvm *kvm, unsigned int host_irq, if (!kvm_arch_has_assigned_device(kvm) || !irq_remapping_cap(IRQ_POSTING_CAP) || - !kvm_vcpu_apicv_active(kvm->vcpus[0])) + !kvm_vcpu_apicv_active(kvm_get_vcpu(kvm, 0))) return 0; idx = srcu_read_lock(&kvm->irq_srcu); diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h index 5417dac55272..5cc3ca8b92b3 100644 --- a/include/linux/kvm_host.h +++ b/include/linux/kvm_host.h @@ -388,12 +388,16 @@ struct kvm_memslots { int used_slots; }; +#define KVM_VCPUS_CHUNK_SIZE 128 +#define KVM_VCPUS_CHUNKS_NUM \ + (round_up(KVM_MAX_VCPUS, KVM_VCPUS_CHUNK_SIZE) / KVM_VCPUS_CHUNK_SIZE) + struct kvm { spinlock_t mmu_lock; struct mutex slots_lock; struct mm_struct *mm; /* userspace tied to this vm */ struct kvm_memslots __rcu *memslots[KVM_ADDRESS_SPACE_NUM]; - struct kvm_vcpu *vcpus[KVM_MAX_VCPUS]; + struct kvm_vcpu **vcpus[KVM_VCPUS_CHUNKS_NUM]; struct list_head vcpu_list; /* @@ -484,14 +488,23 @@ static inline struct kvm_io_bus *kvm_get_bus(struct kvm *kvm, enum kvm_bus idx) !refcount_read(&kvm->users_count)); } -static inline struct kvm_vcpu *kvm_get_vcpu(struct kvm *kvm, int i) +static inline struct kvm_vcpu *__kvm_get_vcpu(struct kvm *kvm, int id) { - /* Pairs with smp_wmb() in kvm_vm_ioctl_create_vcpu, in case - * the caller has read kvm->online_vcpus before (as is the case - * for kvm_for_each_vcpu, for example). + return kvm->vcpus[id / KVM_VCPUS_CHUNK_SIZE][id % KVM_VCPUS_CHUNK_SIZE]; +} + +static inline struct kvm_vcpu *kvm_get_vcpu(struct kvm *kvm, int id) +{ + if (id >= atomic_read(&kvm->online_vcpus)) + return NULL; + + /* + * Pairs with smp_wmb() in kvm_vm_ioctl_create_vcpu. Ensures that the + * pointers leading to an online vcpu are valid. */ smp_rmb(); - return kvm->vcpus[i]; + + return __kvm_get_vcpu(kvm, id); } #define kvm_for_each_vcpu(vcpup, kvm) \ @@ -514,7 +527,7 @@ static inline struct kvm_vcpu *kvm_get_vcpu_by_id(struct kvm *kvm, int id) if (id < 0) return NULL; - if (id < KVM_MAX_VCPUS) + if (id < atomic_read(&kvm->online_vcpus)) vcpu = kvm_get_vcpu(kvm, id); if (vcpu && vcpu->vcpu_id == id) return vcpu; diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 6cec58cad6c7..f9d68ec332c6 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -759,11 +759,14 @@ void kvm_free_vcpus(struct kvm *kvm) mutex_lock(&kvm->lock); - i = atomic_read(&kvm->online_vcpus); + i = round_up(atomic_read(&kvm->online_vcpus), KVM_VCPUS_CHUNK_SIZE) / + KVM_VCPUS_CHUNK_SIZE; atomic_set(&kvm->online_vcpus, 0); - while (i--) + while (i--) { + kfree(kvm->vcpus[i]); kvm->vcpus[i] = NULL; + } mutex_unlock(&kvm->lock); } @@ -2480,6 +2483,8 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, u32 id) { int r; struct kvm_vcpu *vcpu; + struct kvm_vcpu **vcpusp; + unsigned chunk, offset; if (id >= KVM_MAX_VCPU_ID) return -EINVAL; @@ -2517,8 +2522,22 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, u32 id) vcpu->vcpus_idx = atomic_read(&kvm->online_vcpus); - BUG_ON(kvm->vcpus[vcpu->vcpus_idx]); + chunk = vcpu->vcpus_idx / KVM_VCPUS_CHUNK_SIZE; + offset = vcpu->vcpus_idx % KVM_VCPUS_CHUNK_SIZE; + if (!kvm->vcpus[chunk]) { + kvm->vcpus[chunk] = kzalloc(KVM_VCPUS_CHUNK_SIZE * sizeof(**kvm->vcpus), + GFP_KERNEL); + if (!kvm->vcpus[chunk]) { + r = -ENOMEM; + goto unlock_vcpu_destroy; + } + + BUG_ON(offset != 0); + } + + vcpusp = &kvm->vcpus[chunk][offset]; + BUG_ON(*vcpusp); /* Now it's all set up, let userspace reach it */ kvm_get_kvm(kvm); @@ -2528,7 +2547,7 @@ static int kvm_vm_ioctl_create_vcpu(struct kvm *kvm, u32 id) goto unlock_vcpu_destroy; } - kvm->vcpus[atomic_read(&kvm->online_vcpus)] = vcpu; + *vcpusp = vcpu; list_add_tail_rcu(&vcpu->vcpu_list, &kvm->vcpu_list); /*