svm->ir_list_lock can be taken during __avic_vcpu_put which can be called from schedule() via kvm_sched_out. Therefore use a raw spinlock instead. This fixes the following lockdep warning: [ 728.022965] ============================= [ 728.027438] [ BUG: Invalid wait context ] [ 728.031911] 6.12.0-146.1640_2124176644.el10.x86_64+debug #1 Not tainted [ 728.039294] ----------------------------- [ 728.043765] qemu-kvm/38299 is trying to lock: [ 728.048624] ff11000239725600 (&svm->ir_list_lock){....}-{3:3}, at: __avic_vcpu_put+0xfd/0x300 [kvm_amd] [ 728.059135] other info that might help us debug this: [ 728.064768] context-{5:5} [ 728.067688] 2 locks held by qemu-kvm/38299: [ 728.072352] #0: ff11000239723ba8 (&vcpu->mutex){+.+.}-{4:4}, at: kvm_vcpu_ioctl+0x240/0xe00 [kvm] [ 728.082425] #1: ff11000b906056d8 (&rq->__lock){-.-.}-{2:2}, at: raw_spin_rq_lock_nested+0x2e/0x130 [ 728.092540] stack backtrace: [ 728.095758] CPU: 1 UID: 0 PID: 38299 Comm: qemu-kvm Kdump: loaded Not tainted 6.12.0-146.1640_2124176644.el10.x86_64+debug #1 PREEMPT(voluntary) [ 728.095763] Hardware name: AMD Corporation QUARTZ/QUARTZ, BIOS RQZ100AB 09/14/2023 [ 728.095766] Call Trace: [ 728.095769] [ 728.095775] dump_stack_lvl+0x6f/0xb0 [ 728.095782] __lock_acquire+0x921/0xb80 [ 728.095787] ? mark_held_locks+0x40/0x70 [ 728.095793] lock_acquire.part.0+0xbe/0x270 [ 728.095797] ? __avic_vcpu_put+0xfd/0x300 [kvm_amd] [ 728.095811] ? srso_alias_return_thunk+0x5/0xfbef5 [ 728.095815] ? rcu_is_watching+0x15/0xb0 [ 728.095818] ? srso_alias_return_thunk+0x5/0xfbef5 [ 728.095821] ? lock_acquire+0x120/0x170 [ 728.095824] ? __avic_vcpu_put+0xfd/0x300 [kvm_amd] [ 728.095836] _raw_spin_lock_irqsave+0x46/0x90 [ 728.095840] ? __avic_vcpu_put+0xfd/0x300 [kvm_amd] [ 728.095850] __avic_vcpu_put+0xfd/0x300 [kvm_amd] [ 728.095866] svm_vcpu_put+0xfa/0x130 [kvm_amd] [ 728.095877] kvm_arch_vcpu_put+0x48c/0x790 [kvm] [ 728.095944] kvm_sched_out+0x161/0x1c0 [kvm] [ 728.095994] prepare_task_switch+0x36b/0xf60 [ 728.095998] ? rcu_is_watching+0x15/0xb0 [ 728.096004] __schedule+0x4f7/0x1890 [ 728.096010] ? __pfx___schedule+0x10/0x10 [ 728.096012] ? srso_alias_return_thunk+0x5/0xfbef5 [ 728.096019] ? __pfx_vcpu_enter_guest.constprop.0+0x10/0x10 [kvm] [ 728.096067] ? srso_alias_return_thunk+0x5/0xfbef5 [ 728.096069] ? find_held_lock+0x32/0x90 [ 728.096072] ? local_clock_noinstr+0xd/0xe0 [ 728.096080] schedule+0xd4/0x260 [ 728.096083] xfer_to_guest_mode_handle_work+0x54/0xc0 [ 728.096089] vcpu_run+0x69a/0xa70 [kvm] [ 728.096136] kvm_arch_vcpu_ioctl_run+0xdc0/0x17e0 [kvm] [ 728.096185] kvm_vcpu_ioctl+0x39f/0xe00 [kvm] Signed-off-by: Maxim Levitsky --- arch/x86/kvm/svm/avic.c | 16 ++++++++-------- arch/x86/kvm/svm/svm.h | 2 +- 2 files changed, 9 insertions(+), 9 deletions(-) diff --git a/arch/x86/kvm/svm/avic.c b/arch/x86/kvm/svm/avic.c index f286b5706d7c..2814e7e24e89 100644 --- a/arch/x86/kvm/svm/avic.c +++ b/arch/x86/kvm/svm/avic.c @@ -788,7 +788,7 @@ int avic_init_vcpu(struct vcpu_svm *svm) struct kvm_vcpu *vcpu = &svm->vcpu; INIT_LIST_HEAD(&svm->ir_list); - spin_lock_init(&svm->ir_list_lock); + raw_spin_lock_init(&svm->ir_list_lock); if (!enable_apicv || !irqchip_in_kernel(vcpu->kvm)) return 0; @@ -816,9 +816,9 @@ static void svm_ir_list_del(struct kvm_kernel_irqfd *irqfd) if (!vcpu) return; - spin_lock_irqsave(&to_svm(vcpu)->ir_list_lock, flags); + raw_spin_lock_irqsave(&to_svm(vcpu)->ir_list_lock, flags); list_del(&irqfd->vcpu_list); - spin_unlock_irqrestore(&to_svm(vcpu)->ir_list_lock, flags); + raw_spin_unlock_irqrestore(&to_svm(vcpu)->ir_list_lock, flags); } int avic_pi_update_irte(struct kvm_kernel_irqfd *irqfd, struct kvm *kvm, @@ -855,7 +855,7 @@ int avic_pi_update_irte(struct kvm_kernel_irqfd *irqfd, struct kvm *kvm, * list of IRQs being posted to the vCPU, to ensure the IRTE * isn't programmed with stale pCPU/IsRunning information. */ - guard(spinlock_irqsave)(&svm->ir_list_lock); + guard(raw_spinlock_irqsave)(&svm->ir_list_lock); /* * Update the target pCPU for IOMMU doorbells if the vCPU is @@ -972,7 +972,7 @@ static void __avic_vcpu_load(struct kvm_vcpu *vcpu, int cpu, * up-to-date entry information, or that this task will wait until * svm_ir_list_add() completes to set the new target pCPU. */ - spin_lock_irqsave(&svm->ir_list_lock, flags); + raw_spin_lock_irqsave(&svm->ir_list_lock, flags); entry = svm->avic_physical_id_entry; WARN_ON_ONCE(entry & AVIC_PHYSICAL_ID_ENTRY_IS_RUNNING_MASK); @@ -997,7 +997,7 @@ static void __avic_vcpu_load(struct kvm_vcpu *vcpu, int cpu, avic_update_iommu_vcpu_affinity(vcpu, h_physical_id, action); - spin_unlock_irqrestore(&svm->ir_list_lock, flags); + raw_spin_unlock_irqrestore(&svm->ir_list_lock, flags); } void avic_vcpu_load(struct kvm_vcpu *vcpu, int cpu) @@ -1035,7 +1035,7 @@ static void __avic_vcpu_put(struct kvm_vcpu *vcpu, enum avic_vcpu_action action) * or that this task will wait until svm_ir_list_add() completes to * mark the vCPU as not running. */ - spin_lock_irqsave(&svm->ir_list_lock, flags); + raw_spin_lock_irqsave(&svm->ir_list_lock, flags); avic_update_iommu_vcpu_affinity(vcpu, -1, action); @@ -1059,7 +1059,7 @@ static void __avic_vcpu_put(struct kvm_vcpu *vcpu, enum avic_vcpu_action action) svm->avic_physical_id_entry = entry; - spin_unlock_irqrestore(&svm->ir_list_lock, flags); + raw_spin_unlock_irqrestore(&svm->ir_list_lock, flags); } void avic_vcpu_put(struct kvm_vcpu *vcpu) diff --git a/arch/x86/kvm/svm/svm.h b/arch/x86/kvm/svm/svm.h index e4b04f435b3d..dbd87a87046a 100644 --- a/arch/x86/kvm/svm/svm.h +++ b/arch/x86/kvm/svm/svm.h @@ -329,7 +329,7 @@ struct vcpu_svm { * back into remapped mode). */ struct list_head ir_list; - spinlock_t ir_list_lock; + raw_spinlock_t ir_list_lock; struct vcpu_sev_es_state sev_es; -- 2.49.0