From patchwork Sat Apr 13 00:08:47 2013 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Scott Wood X-Patchwork-Id: 236290 Return-Path: X-Original-To: incoming@patchwork.ozlabs.org Delivered-To: patchwork-incoming@bilbo.ozlabs.org Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 6720B2C00AB for ; Sat, 13 Apr 2013 10:09:20 +1000 (EST) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754654Ab3DMAJM (ORCPT ); Fri, 12 Apr 2013 20:09:12 -0400 Received: from co1ehsobe001.messaging.microsoft.com ([216.32.180.184]:18474 "EHLO co1outboundpool.messaging.microsoft.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752577Ab3DMAI6 (ORCPT ); Fri, 12 Apr 2013 20:08:58 -0400 Received: from mail82-co1-R.bigfish.com (10.243.78.252) by CO1EHSOBE009.bigfish.com (10.243.66.72) with Microsoft SMTP Server id 14.1.225.23; Sat, 13 Apr 2013 00:08:57 +0000 Received: from mail82-co1 (localhost [127.0.0.1]) by mail82-co1-R.bigfish.com (Postfix) with ESMTP id 2041C8A02F4; Sat, 13 Apr 2013 00:08:57 +0000 (UTC) X-Forefront-Antispam-Report: CIP:70.37.183.190; KIP:(null); UIP:(null); IPV:NLI; H:mail.freescale.net; RD:none; EFVD:NLI X-SpamScore: 0 X-BigFish: VS0(zzzz1f42h1fc6h1ee6h1de0h1fdah1202h1e76h1d1ah1d2ahzz8275bhz2dh2a8h668h839hd24he5bhf0ah1288h12a5h12a9h12bdh12e5h137ah139eh13b6h1441h1504h1537h162dh1631h1758h1898h18e1h1946h19b5h1ad9h1b0ah1155h) Received: from mail82-co1 (localhost.localdomain [127.0.0.1]) by mail82-co1 (MessageSwitch) id 1365811734825226_19087; Sat, 13 Apr 2013 00:08:54 +0000 (UTC) Received: from CO1EHSMHS006.bigfish.com (unknown [10.243.78.249]) by mail82-co1.bigfish.com (Postfix) with ESMTP id C73B77C006F; Sat, 13 Apr 2013 00:08:54 +0000 (UTC) Received: from mail.freescale.net (70.37.183.190) by CO1EHSMHS006.bigfish.com (10.243.66.16) with Microsoft SMTP Server (TLS) id 14.1.225.23; Sat, 13 Apr 2013 00:08:54 +0000 Received: from tx30smr01.am.freescale.net (10.81.153.31) by 039-SN1MMR1-001.039d.mgd.msft.net (10.84.1.13) with Microsoft SMTP Server (TLS) id 14.2.328.11; Sat, 13 Apr 2013 00:08:53 +0000 Received: from snotra.am.freescale.net ([10.214.86.208]) by tx30smr01.am.freescale.net (8.14.3/8.14.0) with ESMTP id r3D08l0j029657; Fri, 12 Apr 2013 17:08:52 -0700 From: Scott Wood To: Alexander Graf CC: , , , Scott Wood Subject: [PATCH v4 6/6] kvm/ppc/mpic: add KVM_CAP_IRQ_MPIC Date: Fri, 12 Apr 2013 19:08:47 -0500 Message-ID: <1365811727-24431-7-git-send-email-scottwood@freescale.com> X-Mailer: git-send-email 1.7.10.4 In-Reply-To: <1365811727-24431-1-git-send-email-scottwood@freescale.com> References: <1364954273-18196-1-git-send-email-scottwood@freescale.com> <1365811727-24431-1-git-send-email-scottwood@freescale.com> MIME-Version: 1.0 X-OriginatorOrg: freescale.com Sender: kvm-ppc-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: kvm-ppc@vger.kernel.org Enabling this capability connects the vcpu to the designated in-kernel MPIC. Using explicit connections between vcpus and irqchips allows for flexibility, but the main benefit at the moment is that it simplifies the code -- KVM doesn't need vm-global state to remember which MPIC object is associated with this vm, and it doesn't need to care about ordering between irqchip creation and vcpu creation. Signed-off-by: Scott Wood --- v4: - add vcpu->arch.irq_cpu_id, instead of incorrectly using vcpu_id now that the numberspaces are separate - pass the vcpu to mpic code when disconnecting from the vcpu - minor bugfixes on error paths --- Documentation/virtual/kvm/api.txt | 8 ++++ arch/powerpc/include/asm/kvm_host.h | 9 ++++ arch/powerpc/include/asm/kvm_ppc.h | 4 +- arch/powerpc/kvm/booke.c | 4 ++ arch/powerpc/kvm/mpic.c | 82 ++++++++++++++++++++++++++++++++--- arch/powerpc/kvm/powerpc.c | 30 +++++++++++++ include/uapi/linux/kvm.h | 1 + 7 files changed, 130 insertions(+), 8 deletions(-) diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt index d52f3f9..4c326ae 100644 --- a/Documentation/virtual/kvm/api.txt +++ b/Documentation/virtual/kvm/api.txt @@ -2728,3 +2728,11 @@ to receive the topmost interrupt vector. When disabled (args[0] == 0), behavior is as if this facility is unsupported. When this capability is enabled, KVM_EXIT_EPR can occur. + +6.6 KVM_CAP_IRQ_MPIC + +Architectures: ppc +Parameters: args[0] is the MPIC device fd + args[1] is the MPIC CPU number for this vcpu + +This capability connects the vcpu to an in-kernel MPIC device. diff --git a/arch/powerpc/include/asm/kvm_host.h b/arch/powerpc/include/asm/kvm_host.h index 7e7aef9..36368c9 100644 --- a/arch/powerpc/include/asm/kvm_host.h +++ b/arch/powerpc/include/asm/kvm_host.h @@ -375,6 +375,11 @@ struct kvmppc_booke_debug_reg { u64 dac[KVMPPC_BOOKE_MAX_DAC]; }; +#define KVMPPC_IRQ_DEFAULT 0 +#define KVMPPC_IRQ_MPIC 1 + +struct openpic; + struct kvm_vcpu_arch { ulong host_stack; u32 host_pid; @@ -554,6 +559,10 @@ struct kvm_vcpu_arch { unsigned long magic_page_pa; /* phys addr to map the magic page to */ unsigned long magic_page_ea; /* effect. addr to map the magic page to */ + int irq_type; /* one of KVM_IRQ_* */ + int irq_cpu_id; + struct openpic *mpic; /* KVM_IRQ_MPIC */ + #ifdef CONFIG_KVM_BOOK3S_64_HV struct kvm_vcpu_arch_shared shregs; diff --git a/arch/powerpc/include/asm/kvm_ppc.h b/arch/powerpc/include/asm/kvm_ppc.h index 3b63b97..b25d475 100644 --- a/arch/powerpc/include/asm/kvm_ppc.h +++ b/arch/powerpc/include/asm/kvm_ppc.h @@ -248,7 +248,6 @@ int kvmppc_set_one_reg(struct kvm_vcpu *vcpu, u64 id, union kvmppc_one_reg *); void kvmppc_set_pid(struct kvm_vcpu *vcpu, u32 pid); struct openpic; -void kvmppc_mpic_put(struct openpic *opp); #ifdef CONFIG_KVM_BOOK3S_64_HV static inline void kvmppc_set_xics_phys(int cpu, unsigned long addr) @@ -276,6 +275,9 @@ static inline void kvmppc_set_epr(struct kvm_vcpu *vcpu, u32 epr) } void kvmppc_mpic_set_epr(struct kvm_vcpu *vcpu); +int kvmppc_mpic_connect_vcpu(struct kvm_device *dev, struct kvm_vcpu *vcpu, + u32 cpu); +void kvmppc_mpic_disconnect_vcpu(struct openpic *opp, struct kvm_vcpu *vcpu); int kvm_vcpu_ioctl_config_tlb(struct kvm_vcpu *vcpu, struct kvm_config_tlb *cfg); diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c index cff53d4..0097912 100644 --- a/arch/powerpc/kvm/booke.c +++ b/arch/powerpc/kvm/booke.c @@ -430,6 +430,10 @@ static int kvmppc_booke_irqprio_deliver(struct kvm_vcpu *vcpu, if (update_epr == true) { if (vcpu->arch.epr_flags & KVMPPC_EPR_USER) kvm_make_request(KVM_REQ_EPR_EXIT, vcpu); + else if (vcpu->arch.epr_flags & KVMPPC_EPR_KERNEL) { + BUG_ON(vcpu->arch.irq_type != KVMPPC_IRQ_MPIC); + kvmppc_mpic_set_epr(vcpu); + } } new_msr &= msr_mask; diff --git a/arch/powerpc/kvm/mpic.c b/arch/powerpc/kvm/mpic.c index 8ed4072..a231e0c 100644 --- a/arch/powerpc/kvm/mpic.c +++ b/arch/powerpc/kvm/mpic.c @@ -114,7 +114,7 @@ static struct fsl_mpic_info fsl_mpic_42 = { static int get_current_cpu(void) { struct kvm_vcpu *vcpu = current->thread.kvm_vcpu; - return vcpu ? vcpu->vcpu_id : -1; + return vcpu ? vcpu->arch.irq_cpu_id : -1; } static int openpic_cpu_write_internal(void *opaque, gpa_t addr, @@ -244,7 +244,7 @@ static void mpic_irq_raise(struct openpic *opp, struct irq_dest *dst, return; } - pr_debug("%s: cpu %d output %d\n", __func__, dst->vcpu->vcpu_id, + pr_debug("%s: cpu %d output %d\n", __func__, dst->vcpu->arch.irq_cpu_id, output); if (output != ILR_INTTGT_INT) /* TODO */ @@ -262,7 +262,7 @@ static void mpic_irq_lower(struct openpic *opp, struct irq_dest *dst, return; } - pr_debug("%s: cpu %d output %d\n", __func__, dst->vcpu->vcpu_id, + pr_debug("%s: cpu %d output %d\n", __func__, dst->vcpu->arch.irq_cpu_id, output); if (output != ILR_INTTGT_INT) /* TODO */ @@ -1160,6 +1160,20 @@ static uint32_t openpic_iack(struct openpic *opp, struct irq_dest *dst, return retval; } +void kvmppc_mpic_set_epr(struct kvm_vcpu *vcpu) +{ + struct openpic *opp = vcpu->arch.mpic; + int cpu = vcpu->arch.irq_cpu_id; + unsigned long flags; + + spin_lock_irqsave(&opp->lock, flags); + + if ((opp->gcr & opp->mpic_mode_mask) == GCR_MODE_PROXY) + kvmppc_set_epr(vcpu, openpic_iack(opp, &opp->dst[cpu], cpu)); + + spin_unlock_irqrestore(&opp->lock, flags); +} + static int openpic_cpu_read_internal(void *opaque, gpa_t addr, u32 *ptr, int idx) { @@ -1426,10 +1440,10 @@ static void map_mmio(struct openpic *opp) static void unmap_mmio(struct openpic *opp) { - BUG_ON(opp->mmio_mapped); - opp->mmio_mapped = false; - - kvm_io_bus_unregister_dev(opp->kvm, KVM_MMIO_BUS, &opp->mmio); + if (opp->mmio_mapped) { + opp->mmio_mapped = false; + kvm_io_bus_unregister_dev(opp->kvm, KVM_MMIO_BUS, &opp->mmio); + } } static int set_base_addr(struct openpic *opp, struct kvm_device_attr *attr) @@ -1688,3 +1702,57 @@ struct kvm_device_ops kvm_mpic_ops = { .get_attr = mpic_get_attr, .has_attr = mpic_has_attr, }; + +int kvmppc_mpic_connect_vcpu(struct kvm_device *dev, struct kvm_vcpu *vcpu, + u32 cpu) +{ + struct openpic *opp = dev->private; + int ret = 0; + + if (dev->ops != &kvm_mpic_ops) + return -EPERM; + if (opp->kvm != vcpu->kvm) + return -EPERM; + if (cpu < 0 || cpu >= MAX_CPU) + return -EPERM; + + spin_lock_irq(&opp->lock); + + if (opp->dst[cpu].vcpu) { + ret = -EEXIST; + goto out; + } + if (vcpu->arch.irq_type) { + ret = -EBUSY; + goto out; + } + + opp->dst[cpu].vcpu = vcpu; + opp->nb_cpus = max(opp->nb_cpus, cpu + 1); + + vcpu->arch.mpic = opp; + vcpu->arch.irq_cpu_id = cpu; + vcpu->arch.irq_type = KVMPPC_IRQ_MPIC; + + /* This might need to be changed if GCR gets extended */ + if (opp->mpic_mode_mask == GCR_MODE_PROXY) + vcpu->arch.epr_flags |= KVMPPC_EPR_KERNEL; + + kvm_device_get(dev); +out: + spin_unlock_irq(&opp->lock); + return ret; +} + +/* + * This should only happen immediately before the mpic is destroyed, + * so we shouldn't need to worry about anything still trying to + * access the vcpu pointer. + */ +void kvmppc_mpic_disconnect_vcpu(struct openpic *opp, struct kvm_vcpu *vcpu) +{ + BUG_ON(!opp->dst[vcpu->arch.irq_cpu_id].vcpu); + + opp->dst[vcpu->arch.irq_cpu_id].vcpu = NULL; + kvm_device_put(opp->dev); +} diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c index c9a2972..1d3888b 100644 --- a/arch/powerpc/kvm/powerpc.c +++ b/arch/powerpc/kvm/powerpc.c @@ -25,6 +25,7 @@ #include #include #include +#include #include #include #include @@ -327,6 +328,9 @@ int kvm_dev_ioctl_check_extension(long ext) #if defined(CONFIG_KVM_E500V2) || defined(CONFIG_KVM_E500MC) case KVM_CAP_SW_TLB: #endif +#ifdef CONFIG_KVM_MPIC + case KVM_CAP_IRQ_MPIC: +#endif r = 1; break; case KVM_CAP_COALESCED_MMIO: @@ -460,6 +464,13 @@ void kvm_arch_vcpu_free(struct kvm_vcpu *vcpu) tasklet_kill(&vcpu->arch.tasklet); kvmppc_remove_vcpu_debugfs(vcpu); + + switch (vcpu->arch.irq_type) { + case KVMPPC_IRQ_MPIC: + kvmppc_mpic_disconnect_vcpu(vcpu->arch.mpic, vcpu); + break; + } + kvmppc_core_vcpu_free(vcpu); } @@ -794,6 +805,25 @@ static int kvm_vcpu_ioctl_enable_cap(struct kvm_vcpu *vcpu, break; } #endif +#ifdef CONFIG_KVM_MPIC + case KVM_CAP_IRQ_MPIC: { + struct file *filp; + struct kvm_device *dev; + + r = -EBADF; + filp = fget(cap->args[0]); + if (!filp) + break; + + r = -EPERM; + dev = kvm_device_from_filp(filp); + if (dev) + r = kvmppc_mpic_connect_vcpu(dev, vcpu, cap->args[1]); + + fput(filp); + break; + } +#endif default: r = -EINVAL; break; diff --git a/include/uapi/linux/kvm.h b/include/uapi/linux/kvm.h index 76963ec..5b3248a 100644 --- a/include/uapi/linux/kvm.h +++ b/include/uapi/linux/kvm.h @@ -669,6 +669,7 @@ struct kvm_ppc_smmu_info { #define KVM_CAP_ARM_PSCI 87 #define KVM_CAP_ARM_SET_DEVICE_ADDR 88 #define KVM_CAP_DEVICE_CTRL 89 +#define KVM_CAP_IRQ_MPIC 90 #ifdef KVM_CAP_IRQ_ROUTING