From patchwork Wed Apr 17 07:45:25 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Yong-Xuan Wang X-Patchwork-Id: 13632957 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 05C72C4345F for ; Wed, 17 Apr 2024 07:45:47 +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:MIME-Version:List-Subscribe:List-Help: List-Post:List-Archive:List-Unsubscribe:List-Id: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=ZY4Co3icbjfO2TDQ900kYnK8vp+94hH9r7NOFYEMiXc=; b=scpPXb74QplFYJ iHFdp3PoJIC6Tl6hGr22Lh50xcIhGh4U/zJcptq1OaigcDm2MTX33JawaNPbVRApbNOFGVJvnoYqe 2PQHSZaBPMEfkYkVY/ZY9TklVa8A2o9BUFN08HttoHq4cUwaZ9jiNcHdwX+yvuXL+F7+8ZAkjDgN5 CqBssgGLrHvutD+OY4pQngnKqWjFqHcEYqy4Om8Zw6ftc904L2wbI3ZKMUgyUjpxWu0fT2W7Be/6P QpoaRgZXv24q+6o6hRue4ki5++xDJyzftFGw1RUYE3mSL67AMKeLUgNxcyk/MGR8jnWSJbHV/nawd 9JSU7Nr5L1p6vAlISKzg==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.97.1 #2 (Red Hat Linux)) id 1rwzzD-0000000F5G9-2dph; Wed, 17 Apr 2024 07:45:43 +0000 Received: from mail-pl1-x630.google.com ([2607:f8b0:4864:20::630]) by bombadil.infradead.org with esmtps (Exim 4.97.1 #2 (Red Hat Linux)) id 1rwzzA-0000000F5EQ-22JK for linux-riscv@lists.infradead.org; Wed, 17 Apr 2024 07:45:42 +0000 Received: by mail-pl1-x630.google.com with SMTP id d9443c01a7336-1e5c7d087e1so35528455ad.0 for ; Wed, 17 Apr 2024 00:45:39 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=sifive.com; s=google; t=1713339939; x=1713944739; darn=lists.infradead.org; h=references:in-reply-to:message-id:date:subject:cc:to:from:from:to :cc:subject:date:message-id:reply-to; bh=St7UoB0o1VZp6BFYuL7tbSNr44kxdvXND12E5lAvbIw=; b=bZPaBw+FmrzXSjcDSMimb0yUzEzDZLCuqsFs5fCrYCXkYQdmfU3mhpyVQ77DcWYUZn AhtDvoYj0Vl+x64caKbwI0eEEQIDxwkBMVezDJSYpMYUWrNfmW+Mdm6yF/tVOWM+iX8s bKCm448e97wf167/1w+In9tXnnZHkLN/ZzkQEjKOY10Z5E01nObNOtZv/VR97LzVXq3K LM/Blyy+7e5Fh0Mz2Gm68B1e2f3HKEpgeiMYGDIerhF0eAeJAqTyISn/JX4ApO6+7/fV gavzVooajKxN67BrY23WMwrw63UC3ADxSlohwHDXS3ruZBC+R+hKMXDWiKz74bC6cxip zGxw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1713339939; x=1713944739; h=references:in-reply-to:message-id:date:subject:cc:to:from :x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=St7UoB0o1VZp6BFYuL7tbSNr44kxdvXND12E5lAvbIw=; b=hLP4RpxGNJamYEKE8VgDdCE+waQqBtDIhDkUKR97UQyRsn8mC8Vh0KzbKYpsdoUF1k r3+OBzhc5U8EPI1mA1tdozY2UhPHTNbnw9KjOj3OYZOpZvcN32cNRKsnr7nGnKvfbX2a nj8Epo/PS8QzfUnMS9HhpKkunvrDvswSp9WF8pkyKDNLAcfazoFPSEImcN5/1Nml3Pdp gOIbldA7R+tqRRFcQEupJTwiS6GtuZFMrZ9GzJQUuEZhB7ID6xm01eE9nk9TvVEb7+VF 6OhZvc7WHUmTm3LFPm7mj6sCgVqoiWVx1Yxfn88abDQ8e1rv1Wzay2Heiv04UTRGOQ2w vI2A== X-Gm-Message-State: AOJu0YxHorCumfclJSGaDz0WPbxnyygrf6CUtSBe4KYVbqCWtwurgLPL f2RG3ei60X/b/58DAMRgnjESil4QREGa85GcoMOZVHc3GChHEopfnbN8HmW7R/4TaL91lp91d+l uddRUwe/X1vkvQlJzmgtCylFJ8Pg1u/ToavBM9WioPtOoBMJetGZfNO4u0wvVtK7prXPiM45EdD +/M0l7xuR1extDrimr68rFTtfj+bhNAGu01IkzvnLoVMiY3UGuIVDaP+4CyA== X-Google-Smtp-Source: AGHT+IFHSsF9cxqiNY1Nu+bx/qodN9CUjIwlbNj3jKBIDWTLMlrbov3NEnKLK0AuzqvNxhhd9qpAyQ== X-Received: by 2002:a17:902:f7c5:b0:1e0:c567:bb42 with SMTP id h5-20020a170902f7c500b001e0c567bb42mr13147740plw.59.1713339938562; Wed, 17 Apr 2024 00:45:38 -0700 (PDT) Received: from hsinchu26.internal.sifive.com (59-124-168-89.hinet-ip.hinet.net. [59.124.168.89]) by smtp.gmail.com with ESMTPSA id g21-20020a170902c39500b001e7b7a79340sm3166065plg.267.2024.04.17.00.45.36 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 17 Apr 2024 00:45:38 -0700 (PDT) From: Yong-Xuan Wang To: linux-riscv@lists.infradead.org, kvm-riscv@lists.infradead.org Cc: greentime.hu@sifive.com, vincent.chen@sifive.com, Yong-Xuan Wang , Anup Patel , Atish Patra , Paul Walmsley , Palmer Dabbelt , Albert Ou , kvm@vger.kernel.org, linux-kernel@vger.kernel.org Subject: [PATCH v2 1/2] RISCV: KVM: Introduce mp_state_lock to avoid lock inversion in SBI_EXT_HSM_HART_START Date: Wed, 17 Apr 2024 15:45:25 +0800 Message-Id: <20240417074528.16506-2-yongxuan.wang@sifive.com> X-Mailer: git-send-email 2.17.1 In-Reply-To: <20240417074528.16506-1-yongxuan.wang@sifive.com> References: <20240417074528.16506-1-yongxuan.wang@sifive.com> X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20240417_004540_560579_3381EB5F X-CRM114-Status: GOOD ( 19.90 ) X-BeenThere: linux-riscv@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: "linux-riscv" Errors-To: linux-riscv-bounces+linux-riscv=archiver.kernel.org@lists.infradead.org Documentation/virt/kvm/locking.rst advises that kvm->lock should be acquired outside vcpu->mutex and kvm->srcu. However, when KVM/RISC-V handling SBI_EXT_HSM_HART_START, the lock ordering is vcpu->mutex, kvm->srcu then kvm->lock. Although the lockdep checking no longer complains about this after commit f0f44752f5f6 ("rcu: Annotate SRCU's update-side lockdep dependencies"), it's necessary to replace kvm->lock with a new dedicated lock to ensure only one hart can execute the SBI_EXT_HSM_HART_START call for the target hart simultaneously. Additionally, this patch also rename "power_off" to "mp_state" with two possible values. The vcpu->mp_state_lock also protects the access of vcpu->mp_state. Signed-off-by: Yong-Xuan Wang Reviewed-by: Anup Patel --- arch/riscv/include/asm/kvm_host.h | 7 ++-- arch/riscv/kvm/vcpu.c | 56 ++++++++++++++++++++++++------- arch/riscv/kvm/vcpu_sbi.c | 7 ++-- arch/riscv/kvm/vcpu_sbi_hsm.c | 23 ++++++++----- 4 files changed, 68 insertions(+), 25 deletions(-) diff --git a/arch/riscv/include/asm/kvm_host.h b/arch/riscv/include/asm/kvm_host.h index 484d04a92fa6..64d35a8c908c 100644 --- a/arch/riscv/include/asm/kvm_host.h +++ b/arch/riscv/include/asm/kvm_host.h @@ -252,8 +252,9 @@ struct kvm_vcpu_arch { /* Cache pages needed to program page tables with spinlock held */ struct kvm_mmu_memory_cache mmu_page_cache; - /* VCPU power-off state */ - bool power_off; + /* VCPU power state */ + struct kvm_mp_state mp_state; + spinlock_t mp_state_lock; /* Don't run the VCPU (blocked) */ bool pause; @@ -375,7 +376,9 @@ void kvm_riscv_vcpu_flush_interrupts(struct kvm_vcpu *vcpu); void kvm_riscv_vcpu_sync_interrupts(struct kvm_vcpu *vcpu); bool kvm_riscv_vcpu_has_interrupts(struct kvm_vcpu *vcpu, u64 mask); void kvm_riscv_vcpu_power_off(struct kvm_vcpu *vcpu); +void __kvm_riscv_vcpu_power_on(struct kvm_vcpu *vcpu); void kvm_riscv_vcpu_power_on(struct kvm_vcpu *vcpu); +bool kvm_riscv_vcpu_stopped(struct kvm_vcpu *vcpu); void kvm_riscv_vcpu_sbi_sta_reset(struct kvm_vcpu *vcpu); void kvm_riscv_vcpu_record_steal_time(struct kvm_vcpu *vcpu); diff --git a/arch/riscv/kvm/vcpu.c b/arch/riscv/kvm/vcpu.c index b5ca9f2e98ac..70937f71c3c4 100644 --- a/arch/riscv/kvm/vcpu.c +++ b/arch/riscv/kvm/vcpu.c @@ -102,6 +102,8 @@ int kvm_arch_vcpu_create(struct kvm_vcpu *vcpu) struct kvm_cpu_context *cntx; struct kvm_vcpu_csr *reset_csr = &vcpu->arch.guest_reset_csr; + spin_lock_init(&vcpu->arch.mp_state_lock); + /* Mark this VCPU never ran */ vcpu->arch.ran_atleast_once = false; vcpu->arch.mmu_page_cache.gfp_zero = __GFP_ZERO; @@ -201,7 +203,7 @@ void kvm_arch_vcpu_unblocking(struct kvm_vcpu *vcpu) int kvm_arch_vcpu_runnable(struct kvm_vcpu *vcpu) { return (kvm_riscv_vcpu_has_interrupts(vcpu, -1UL) && - !vcpu->arch.power_off && !vcpu->arch.pause); + !kvm_riscv_vcpu_stopped(vcpu) && !vcpu->arch.pause); } int kvm_arch_vcpu_should_kick(struct kvm_vcpu *vcpu) @@ -429,26 +431,50 @@ bool kvm_riscv_vcpu_has_interrupts(struct kvm_vcpu *vcpu, u64 mask) return kvm_riscv_vcpu_aia_has_interrupts(vcpu, mask); } -void kvm_riscv_vcpu_power_off(struct kvm_vcpu *vcpu) +static void __kvm_riscv_vcpu_power_off(struct kvm_vcpu *vcpu) { - vcpu->arch.power_off = true; + vcpu->arch.mp_state.mp_state = KVM_MP_STATE_STOPPED; kvm_make_request(KVM_REQ_SLEEP, vcpu); kvm_vcpu_kick(vcpu); } -void kvm_riscv_vcpu_power_on(struct kvm_vcpu *vcpu) +void kvm_riscv_vcpu_power_off(struct kvm_vcpu *vcpu) +{ + spin_lock(&vcpu->arch.mp_state_lock); + __kvm_riscv_vcpu_power_off(vcpu); + spin_unlock(&vcpu->arch.mp_state_lock); +} + +void __kvm_riscv_vcpu_power_on(struct kvm_vcpu *vcpu) { - vcpu->arch.power_off = false; + vcpu->arch.mp_state.mp_state = KVM_MP_STATE_RUNNABLE; kvm_vcpu_wake_up(vcpu); } +void kvm_riscv_vcpu_power_on(struct kvm_vcpu *vcpu) +{ + spin_lock(&vcpu->arch.mp_state_lock); + __kvm_riscv_vcpu_power_on(vcpu); + spin_unlock(&vcpu->arch.mp_state_lock); +} + +bool kvm_riscv_vcpu_stopped(struct kvm_vcpu *vcpu) +{ + bool ret; + + spin_lock(&vcpu->arch.mp_state_lock); + ret = vcpu->arch.mp_state.mp_state == KVM_MP_STATE_STOPPED; + spin_unlock(&vcpu->arch.mp_state_lock); + + return ret; +} + int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu, struct kvm_mp_state *mp_state) { - if (vcpu->arch.power_off) - mp_state->mp_state = KVM_MP_STATE_STOPPED; - else - mp_state->mp_state = KVM_MP_STATE_RUNNABLE; + spin_lock(&vcpu->arch.mp_state_lock); + *mp_state = vcpu->arch.mp_state; + spin_unlock(&vcpu->arch.mp_state_lock); return 0; } @@ -458,17 +484,21 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu, { int ret = 0; + spin_lock(&vcpu->arch.mp_state_lock); + switch (mp_state->mp_state) { case KVM_MP_STATE_RUNNABLE: - vcpu->arch.power_off = false; + vcpu->arch.mp_state.mp_state = KVM_MP_STATE_RUNNABLE; break; case KVM_MP_STATE_STOPPED: - kvm_riscv_vcpu_power_off(vcpu); + __kvm_riscv_vcpu_power_off(vcpu); break; default: ret = -EINVAL; } + spin_unlock(&vcpu->arch.mp_state_lock); + return ret; } @@ -584,11 +614,11 @@ static void kvm_riscv_check_vcpu_requests(struct kvm_vcpu *vcpu) if (kvm_check_request(KVM_REQ_SLEEP, vcpu)) { kvm_vcpu_srcu_read_unlock(vcpu); rcuwait_wait_event(wait, - (!vcpu->arch.power_off) && (!vcpu->arch.pause), + (!kvm_riscv_vcpu_stopped(vcpu)) && (!vcpu->arch.pause), TASK_INTERRUPTIBLE); kvm_vcpu_srcu_read_lock(vcpu); - if (vcpu->arch.power_off || vcpu->arch.pause) { + if (kvm_riscv_vcpu_stopped(vcpu) || vcpu->arch.pause) { /* * Awaken to handle a signal, request to * sleep again later. diff --git a/arch/riscv/kvm/vcpu_sbi.c b/arch/riscv/kvm/vcpu_sbi.c index 72a2ffb8dcd1..1851fc979bd2 100644 --- a/arch/riscv/kvm/vcpu_sbi.c +++ b/arch/riscv/kvm/vcpu_sbi.c @@ -138,8 +138,11 @@ void kvm_riscv_vcpu_sbi_system_reset(struct kvm_vcpu *vcpu, unsigned long i; struct kvm_vcpu *tmp; - kvm_for_each_vcpu(i, tmp, vcpu->kvm) - tmp->arch.power_off = true; + kvm_for_each_vcpu(i, tmp, vcpu->kvm) { + spin_lock(&vcpu->arch.mp_state_lock); + tmp->arch.mp_state.mp_state = KVM_MP_STATE_STOPPED; + spin_unlock(&vcpu->arch.mp_state_lock); + } kvm_make_all_cpus_request(vcpu->kvm, KVM_REQ_SLEEP); memset(&run->system_event, 0, sizeof(run->system_event)); diff --git a/arch/riscv/kvm/vcpu_sbi_hsm.c b/arch/riscv/kvm/vcpu_sbi_hsm.c index 7dca0e9381d9..115a6c6525fd 100644 --- a/arch/riscv/kvm/vcpu_sbi_hsm.c +++ b/arch/riscv/kvm/vcpu_sbi_hsm.c @@ -18,12 +18,18 @@ static int kvm_sbi_hsm_vcpu_start(struct kvm_vcpu *vcpu) struct kvm_cpu_context *cp = &vcpu->arch.guest_context; struct kvm_vcpu *target_vcpu; unsigned long target_vcpuid = cp->a0; + int ret = 0; target_vcpu = kvm_get_vcpu_by_id(vcpu->kvm, target_vcpuid); if (!target_vcpu) return SBI_ERR_INVALID_PARAM; - if (!target_vcpu->arch.power_off) - return SBI_ERR_ALREADY_AVAILABLE; + + spin_lock(&target_vcpu->arch.mp_state_lock); + + if (target_vcpu->arch.mp_state.mp_state != KVM_MP_STATE_STOPPED) { + ret = SBI_ERR_ALREADY_AVAILABLE; + goto out; + } reset_cntx = &target_vcpu->arch.guest_reset_context; /* start address */ @@ -34,14 +40,18 @@ static int kvm_sbi_hsm_vcpu_start(struct kvm_vcpu *vcpu) reset_cntx->a1 = cp->a2; kvm_make_request(KVM_REQ_VCPU_RESET, target_vcpu); - kvm_riscv_vcpu_power_on(target_vcpu); + __kvm_riscv_vcpu_power_on(target_vcpu); + +out: + spin_unlock(&target_vcpu->arch.mp_state_lock); + return 0; } static int kvm_sbi_hsm_vcpu_stop(struct kvm_vcpu *vcpu) { - if (vcpu->arch.power_off) + if (kvm_riscv_vcpu_stopped(vcpu)) return SBI_ERR_FAILURE; kvm_riscv_vcpu_power_off(vcpu); @@ -58,7 +68,7 @@ static int kvm_sbi_hsm_vcpu_get_status(struct kvm_vcpu *vcpu) target_vcpu = kvm_get_vcpu_by_id(vcpu->kvm, target_vcpuid); if (!target_vcpu) return SBI_ERR_INVALID_PARAM; - if (!target_vcpu->arch.power_off) + if (!kvm_riscv_vcpu_stopped(target_vcpu)) return SBI_HSM_STATE_STARTED; else if (vcpu->stat.generic.blocking) return SBI_HSM_STATE_SUSPENDED; @@ -71,14 +81,11 @@ static int kvm_sbi_ext_hsm_handler(struct kvm_vcpu *vcpu, struct kvm_run *run, { int ret = 0; struct kvm_cpu_context *cp = &vcpu->arch.guest_context; - struct kvm *kvm = vcpu->kvm; unsigned long funcid = cp->a6; switch (funcid) { case SBI_EXT_HSM_HART_START: - mutex_lock(&kvm->lock); ret = kvm_sbi_hsm_vcpu_start(vcpu); - mutex_unlock(&kvm->lock); break; case SBI_EXT_HSM_HART_STOP: ret = kvm_sbi_hsm_vcpu_stop(vcpu);