From patchwork Fri May 6 10:45:28 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Andre Przywara X-Patchwork-Id: 67274 Delivered-To: patch@linaro.org Received: by 10.140.92.199 with SMTP id b65csp255434qge; Fri, 6 May 2016 04:03:36 -0700 (PDT) X-Received: by 10.66.172.165 with SMTP id bd5mr28234654pac.128.1462532616594; Fri, 06 May 2016 04:03:36 -0700 (PDT) Return-Path: Received: from bombadil.infradead.org (bombadil.infradead.org. [2001:1868:205::9]) by mx.google.com with ESMTPS id c1si17807727pas.37.2016.05.06.04.03.36 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Fri, 06 May 2016 04:03:36 -0700 (PDT) Received-SPF: pass (google.com: best guess record for domain of linux-arm-kernel-bounces+patch=linaro.org@lists.infradead.org designates 2001:1868:205::9 as permitted sender) client-ip=2001:1868:205::9; Authentication-Results: mx.google.com; spf=pass (google.com: best guess record for domain of linux-arm-kernel-bounces+patch=linaro.org@lists.infradead.org designates 2001:1868:205::9 as permitted sender) smtp.mailfrom=linux-arm-kernel-bounces+patch=linaro.org@lists.infradead.org Received: from localhost ([127.0.0.1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.80.1 #2 (Red Hat Linux)) id 1aydWU-0005Av-K4; Fri, 06 May 2016 11:02:18 +0000 Received: from foss.arm.com ([217.140.101.70]) by bombadil.infradead.org with esmtp (Exim 4.80.1 #2 (Red Hat Linux)) id 1aydHU-0004QB-UB for linux-arm-kernel@lists.infradead.org; Fri, 06 May 2016 10:46:59 +0000 Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.72.51.249]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id F29A292A; Fri, 6 May 2016 03:46:13 -0700 (PDT) Received: from e104803-lin.lan (unknown [10.1.203.153]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPSA id BC0BC3F252; Fri, 6 May 2016 03:46:04 -0700 (PDT) From: Andre Przywara To: Marc Zyngier , Christoffer Dall Subject: [PATCH v3 15/55] KVM: arm/arm64: vgic-new: Implement virtual IRQ injection Date: Fri, 6 May 2016 11:45:28 +0100 Message-Id: <1462531568-9799-16-git-send-email-andre.przywara@arm.com> X-Mailer: git-send-email 2.7.3 In-Reply-To: <1462531568-9799-1-git-send-email-andre.przywara@arm.com> References: <1462531568-9799-1-git-send-email-andre.przywara@arm.com> X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20160506_034649_398151_21D9C26D X-CRM114-Status: GOOD ( 23.75 ) X-Spam-Score: -9.0 (---------) X-Spam-Report: SpamAssassin version 3.4.0 on bombadil.infradead.org summary: Content analysis details: (-9.0 points) pts rule name description ---- ---------------------- -------------------------------------------------- -5.0 RCVD_IN_DNSWL_HI RBL: Sender listed at http://www.dnswl.org/, high trust [217.140.101.70 listed in list.dnswl.org] -0.0 SPF_PASS SPF: sender matches SPF record -2.1 RP_MATCHES_RCVD Envelope sender domain matches handover relay domain -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: 2.1.20 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: linux-arm-kernel@lists.infradead.org, kvmarm@lists.cs.columbia.edu, kvm@vger.kernel.org, Eric Auger MIME-Version: 1.0 Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+patch=linaro.org@lists.infradead.org From: Christoffer Dall Provide a vgic_queue_irq() function which decides whether a given IRQ needs to be queued to a VCPU's ap_list. This should be called whenever an IRQ becomes pending or enabled, either as a result of userspace injection, from in-kernel emulated devices like the architected timer or from MMIO accesses to the distributor emulation. Also provides the necessary functions to allow userland to inject an IRQ to a guest. Since this is the first code that starts using our locking mechanism, we add some (hopefully) clear documentation of our locking strategy and requirements along with this patch. [Andre: refactor out vgic_queue_irq_unlock()] Signed-off-by: Christoffer Dall Signed-off-by: Andre Przywara --- Changelog RFC..v1: - add spinlock checks protected by CONFIG_DEBUG_SPINLOCK - add more comments to vgic_target_oracle - remove BUG_ON() if IRQ is neither edge or level - rename vgic_queue_irq() to vgic_queue_irq_unlock() - simplify initial check in vgic_queue_irq_unlock() - extend retry loop to ask the oracle again - minor comment fixes Changelog v1 .. v2: - move most IRQ injection code into vgic_update_irq_pending() include/kvm/vgic/vgic.h | 3 + virt/kvm/arm/vgic/vgic.c | 209 +++++++++++++++++++++++++++++++++++++++++++++++ virt/kvm/arm/vgic/vgic.h | 1 + 3 files changed, 213 insertions(+) -- 2.7.3 _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel Reviewed-by: Eric Auger diff --git a/include/kvm/vgic/vgic.h b/include/kvm/vgic/vgic.h index 39933ee..2bfb42c 100644 --- a/include/kvm/vgic/vgic.h +++ b/include/kvm/vgic/vgic.h @@ -181,6 +181,9 @@ struct vgic_cpu { u64 live_lrs; }; +int kvm_vgic_inject_irq(struct kvm *kvm, int cpuid, unsigned int intid, + bool level); + #define irqchip_in_kernel(k) (!!((k)->arch.vgic.in_kernel)) #define vgic_initialized(k) (false) #define vgic_ready(k) ((k)->arch.vgic.ready) diff --git a/virt/kvm/arm/vgic/vgic.c b/virt/kvm/arm/vgic/vgic.c index fb45537..92b78a0 100644 --- a/virt/kvm/arm/vgic/vgic.c +++ b/virt/kvm/arm/vgic/vgic.c @@ -19,8 +19,31 @@ #include "vgic.h" +#define CREATE_TRACE_POINTS +#include "../trace.h" + +#ifdef CONFIG_DEBUG_SPINLOCK +#define DEBUG_SPINLOCK_BUG_ON(p) BUG_ON(p) +#else +#define DEBUG_SPINLOCK_BUG_ON(p) +#endif + struct vgic_global __section(.hyp.text) kvm_vgic_global_state; +/* + * Locking order is always: + * vgic_cpu->ap_list_lock + * vgic_irq->irq_lock + * + * (that is, always take the ap_list_lock before the struct vgic_irq lock). + * + * When taking more than one ap_list_lock at the same time, always take the + * lowest numbered VCPU's ap_list_lock first, so: + * vcpuX->vcpu_id < vcpuY->vcpu_id: + * spin_lock(vcpuX->arch.vgic_cpu.ap_list_lock); + * spin_lock(vcpuY->arch.vgic_cpu.ap_list_lock); + */ + struct vgic_irq *vgic_get_irq(struct kvm *kvm, struct kvm_vcpu *vcpu, u32 intid) { @@ -39,3 +62,189 @@ struct vgic_irq *vgic_get_irq(struct kvm *kvm, struct kvm_vcpu *vcpu, WARN(1, "Looking up struct vgic_irq for reserved INTID"); return NULL; } + +/** + * kvm_vgic_target_oracle - compute the target vcpu for an irq + * + * @irq: The irq to route. Must be already locked. + * + * Based on the current state of the interrupt (enabled, pending, + * active, vcpu and target_vcpu), compute the next vcpu this should be + * given to. Return NULL if this shouldn't be injected at all. + * + * Requires the IRQ lock to be held. + */ +static struct kvm_vcpu *vgic_target_oracle(struct vgic_irq *irq) +{ + DEBUG_SPINLOCK_BUG_ON(!spin_is_locked(&irq->irq_lock)); + + /* If the interrupt is active, it must stay on the current vcpu */ + if (irq->active) + return irq->vcpu; + + /* + * If the IRQ is not active but enabled and pending, we should direct + * it to its configured target VCPU. + */ + if (irq->enabled && irq->pending) + return irq->target_vcpu; + + /* If neither active nor pending and enabled, then this IRQ should not + * be queued to any VCPU. + */ + return NULL; +} + +/* + * Only valid injection if changing level for level-triggered IRQs or for a + * rising edge. + */ +static bool vgic_validate_injection(struct vgic_irq *irq, bool level) +{ + switch (irq->config) { + case VGIC_CONFIG_LEVEL: + return irq->line_level != level; + case VGIC_CONFIG_EDGE: + return level; + } + + return false; +} + +/* + * Check whether an IRQ needs to (and can) be queued to a VCPU's ap list. + * Do the queuing if necessary, taking the right locks in the right order. + * Returns true when the IRQ was queued, false otherwise. + * + * Needs to be entered with the IRQ lock already held, but will return + * with all locks dropped. + */ +bool vgic_queue_irq_unlock(struct kvm *kvm, struct vgic_irq *irq) +{ + struct kvm_vcpu *vcpu; + + DEBUG_SPINLOCK_BUG_ON(!spin_is_locked(&irq->irq_lock)); + +retry: + vcpu = vgic_target_oracle(irq); + if (irq->vcpu || !vcpu) { + /* + * If this IRQ is already on a VCPU's ap_list, then it + * cannot be moved or modified and there is no more work for + * us to do. + * + * Otherwise, if the irq is not pending and enabled, it does + * not need to be inserted into an ap_list and there is also + * no more work for us to do. + */ + spin_unlock(&irq->irq_lock); + return false; + } + + /* + * We must unlock the irq lock to take the ap_list_lock where + * we are going to insert this new pending interrupt. + */ + spin_unlock(&irq->irq_lock); + + /* someone can do stuff here, which we re-check below */ + + spin_lock(&vcpu->arch.vgic_cpu.ap_list_lock); + spin_lock(&irq->irq_lock); + + /* + * Did something change behind our backs? + * + * There are two cases: + * 1) The irq became pending or active behind our backs and/or + * the irq->vcpu field was set correspondingly when putting + * the irq on an ap_list. Then drop the locks and return. + * 2) Someone changed the affinity on this irq behind our + * backs and we are now holding the wrong ap_list_lock. + * Then drop the locks and try the new VCPU. + */ + if (irq->vcpu || !(irq->pending && irq->enabled)) { + spin_unlock(&irq->irq_lock); + spin_unlock(&vcpu->arch.vgic_cpu.ap_list_lock); + return false; + } + + if (irq->target_vcpu != vcpu) { + spin_unlock(&irq->irq_lock); + spin_unlock(&vcpu->arch.vgic_cpu.ap_list_lock); + + spin_lock(&irq->irq_lock); + goto retry; + } + + list_add_tail(&irq->ap_list, &vcpu->arch.vgic_cpu.ap_list_head); + irq->vcpu = vcpu; + + spin_unlock(&irq->irq_lock); + spin_unlock(&vcpu->arch.vgic_cpu.ap_list_lock); + + kvm_vcpu_kick(vcpu); + + return true; +} + +static int vgic_update_irq_pending(struct kvm *kvm, int cpuid, + unsigned int intid, bool level, + bool mapped_irq) +{ + struct kvm_vcpu *vcpu; + struct vgic_irq *irq; + int ret; + + trace_vgic_update_irq_pending(cpuid, intid, level); + + vcpu = kvm_get_vcpu(kvm, cpuid); + if (!vcpu && intid < VGIC_NR_PRIVATE_IRQS) + return -EINVAL; + + irq = vgic_get_irq(kvm, vcpu, intid); + if (!irq) + return -EINVAL; + + if (irq->hw != mapped_irq) + return -EINVAL; + + spin_lock(&irq->irq_lock); + + if (!vgic_validate_injection(irq, level)) { + /* Nothing to see here, move along... */ + spin_unlock(&irq->irq_lock); + return 0; + } + + if (irq->config == VGIC_CONFIG_LEVEL) { + irq->line_level = level; + irq->pending = level || irq->soft_pending; + } else { + irq->pending = true; + } + + vgic_queue_irq_unlock(kvm, irq); + + return 0; +} + +/** + * kvm_vgic_inject_irq - Inject an IRQ from a device to the vgic + * @kvm: The VM structure pointer + * @cpuid: The CPU for PPIs + * @intid: The INTID to inject a new state to. + * @level: Edge-triggered: true: to trigger the interrupt + * false: to ignore the call + * Level-sensitive true: raise the input signal + * false: lower the input signal + * + * The VGIC is not concerned with devices being active-LOW or active-HIGH for + * level-sensitive interrupts. You can think of the level parameter as 1 + * being HIGH and 0 being LOW and all devices being active-HIGH. + */ +int kvm_vgic_inject_irq(struct kvm *kvm, int cpuid, unsigned int intid, + bool level) +{ + return vgic_update_irq_pending(kvm, cpuid, intid, level, false); +} diff --git a/virt/kvm/arm/vgic/vgic.h b/virt/kvm/arm/vgic/vgic.h index 61b8d22..c625767 100644 --- a/virt/kvm/arm/vgic/vgic.h +++ b/virt/kvm/arm/vgic/vgic.h @@ -18,5 +18,6 @@ struct vgic_irq *vgic_get_irq(struct kvm *kvm, struct kvm_vcpu *vcpu, u32 intid); +bool vgic_queue_irq_unlock(struct kvm *kvm, struct vgic_irq *irq); #endif