[PATCH v2 3/4] KVM: arm64: Use config_lock to protect data ordered against KVM_RUN

[Oliver Upton]

There are various bits of VM-scoped data that can only be configured
before the first call to KVM_RUN, such as the hypercall bitmaps and
the PMU. As these fields are protected by the kvm->lock and accessed
while holding vcpu->mutex, this is yet another example of lock
inversion.

Change out the kvm->lock for kvm->arch.config_lock in all of these
instances. Opportunistically simplify the locking mechanics of the
PMU configuration by holding the config_lock for the entirety of
kvm_arm_pmu_v3_set_attr().

Note that this also addresses a couple of bugs. There is an unguarded
read of the PMU version in KVM_ARM_VCPU_PMU_V3_FILTER which could race
with KVM_ARM_VCPU_PMU_V3_SET_PMU. Additionally, until now writes to the
per-vCPU vPMU irq were not serialized VM-wide, meaning concurrent calls
to KVM_ARM_VCPU_PMU_V3_IRQ could lead to a false positive in
pmu_irq_is_valid().

Signed-off-by: Oliver Upton <oliver.upton at linux.dev>
---
 arch/arm64/kvm/arm.c        |  4 ++--
 arch/arm64/kvm/guest.c      |  2 ++
 arch/arm64/kvm/hypercalls.c |  4 ++--
 arch/arm64/kvm/pmu-emul.c   | 23 ++++++-----------------
 4 files changed, 12 insertions(+), 21 deletions(-)

diff --git a/arch/arm64/kvm/arm.c b/arch/arm64/kvm/arm.c
index 1478bec52767..a4793961b73d 100644
--- a/arch/arm64/kvm/arm.c
+++ b/arch/arm64/kvm/arm.c
@@ -624,9 +624,9 @@ int kvm_arch_vcpu_run_pid_change(struct kvm_vcpu *vcpu)
 	if (kvm_vm_is_protected(kvm))
 		kvm_call_hyp_nvhe(__pkvm_vcpu_init_traps, vcpu);
 
-	mutex_lock(&kvm->lock);
+	mutex_lock(&kvm->arch.config_lock);
 	set_bit(KVM_ARCH_FLAG_HAS_RAN_ONCE, &kvm->arch.flags);
-	mutex_unlock(&kvm->lock);
+	mutex_unlock(&kvm->arch.config_lock);
 
 	return ret;
 }
diff --git a/arch/arm64/kvm/guest.c b/arch/arm64/kvm/guest.c
index 07444fa22888..481c79cf22cd 100644
--- a/arch/arm64/kvm/guest.c
+++ b/arch/arm64/kvm/guest.c
@@ -957,7 +957,9 @@ int kvm_arm_vcpu_arch_set_attr(struct kvm_vcpu *vcpu,
 
 	switch (attr->group) {
 	case KVM_ARM_VCPU_PMU_V3_CTRL:
+		mutex_lock(&vcpu->kvm->arch.config_lock);
 		ret = kvm_arm_pmu_v3_set_attr(vcpu, attr);
+		mutex_unlock(&vcpu->kvm->arch.config_lock);
 		break;
 	case KVM_ARM_VCPU_TIMER_CTRL:
 		ret = kvm_arm_timer_set_attr(vcpu, attr);
diff --git a/arch/arm64/kvm/hypercalls.c b/arch/arm64/kvm/hypercalls.c
index 5da884e11337..fbdbf4257f76 100644
--- a/arch/arm64/kvm/hypercalls.c
+++ b/arch/arm64/kvm/hypercalls.c
@@ -377,7 +377,7 @@ static int kvm_arm_set_fw_reg_bmap(struct kvm_vcpu *vcpu, u64 reg_id, u64 val)
 	if (val & ~fw_reg_features)
 		return -EINVAL;
 
-	mutex_lock(&kvm->lock);
+	mutex_lock(&kvm->arch.config_lock);
 
 	if (test_bit(KVM_ARCH_FLAG_HAS_RAN_ONCE, &kvm->arch.flags) &&
 	    val != *fw_reg_bmap) {
@@ -387,7 +387,7 @@ static int kvm_arm_set_fw_reg_bmap(struct kvm_vcpu *vcpu, u64 reg_id, u64 val)
 
 	WRITE_ONCE(*fw_reg_bmap, val);
 out:
-	mutex_unlock(&kvm->lock);
+	mutex_unlock(&kvm->arch.config_lock);
 	return ret;
 }
 
diff --git a/arch/arm64/kvm/pmu-emul.c b/arch/arm64/kvm/pmu-emul.c
index c243b10f3e15..82991d89c2ea 100644
--- a/arch/arm64/kvm/pmu-emul.c
+++ b/arch/arm64/kvm/pmu-emul.c
@@ -875,7 +875,7 @@ static int kvm_arm_pmu_v3_set_pmu(struct kvm_vcpu *vcpu, int pmu_id)
 	struct arm_pmu *arm_pmu;
 	int ret = -ENXIO;
 
-	mutex_lock(&kvm->lock);
+	lockdep_assert_held(&kvm->arch.config_lock);
 	mutex_lock(&arm_pmus_lock);
 
 	list_for_each_entry(entry, &arm_pmus, entry) {
@@ -895,7 +895,6 @@ static int kvm_arm_pmu_v3_set_pmu(struct kvm_vcpu *vcpu, int pmu_id)
 	}
 
 	mutex_unlock(&arm_pmus_lock);
-	mutex_unlock(&kvm->lock);
 	return ret;
 }
 
@@ -903,22 +902,20 @@ int kvm_arm_pmu_v3_set_attr(struct kvm_vcpu *vcpu, struct kvm_device_attr *attr)
 {
 	struct kvm *kvm = vcpu->kvm;
 
+	lockdep_assert_held(&kvm->arch.config_lock);
+
 	if (!kvm_vcpu_has_pmu(vcpu))
 		return -ENODEV;
 
 	if (vcpu->arch.pmu.created)
 		return -EBUSY;
 
-	mutex_lock(&kvm->lock);
 	if (!kvm->arch.arm_pmu) {
 		/* No PMU set, get the default one */
 		kvm->arch.arm_pmu = kvm_pmu_probe_armpmu();
-		if (!kvm->arch.arm_pmu) {
-			mutex_unlock(&kvm->lock);
+		if (!kvm->arch.arm_pmu)
 			return -ENODEV;
-		}
 	}
-	mutex_unlock(&kvm->lock);
 
 	switch (attr->attr) {
 	case KVM_ARM_VCPU_PMU_V3_IRQ: {
@@ -962,19 +959,13 @@ int kvm_arm_pmu_v3_set_attr(struct kvm_vcpu *vcpu, struct kvm_device_attr *attr)
 		     filter.action != KVM_PMU_EVENT_DENY))
 			return -EINVAL;
 
-		mutex_lock(&kvm->lock);
-
-		if (test_bit(KVM_ARCH_FLAG_HAS_RAN_ONCE, &kvm->arch.flags)) {
-			mutex_unlock(&kvm->lock);
+		if (test_bit(KVM_ARCH_FLAG_HAS_RAN_ONCE, &kvm->arch.flags))
 			return -EBUSY;
-		}
 
 		if (!kvm->arch.pmu_filter) {
 			kvm->arch.pmu_filter = bitmap_alloc(nr_events, GFP_KERNEL_ACCOUNT);
-			if (!kvm->arch.pmu_filter) {
-				mutex_unlock(&kvm->lock);
+			if (!kvm->arch.pmu_filter)
 				return -ENOMEM;
-			}
 
 			/*
 			 * The default depends on the first applied filter.
@@ -993,8 +984,6 @@ int kvm_arm_pmu_v3_set_attr(struct kvm_vcpu *vcpu, struct kvm_device_attr *attr)
 		else
 			bitmap_clear(kvm->arch.pmu_filter, filter.base_event, filter.nevents);
 
-		mutex_unlock(&kvm->lock);
-
 		return 0;
 	}
 	case KVM_ARM_VCPU_PMU_V3_SET_PMU: {
-- 
2.40.0.rc1.284.g88254d51c5-goog