Commit 2f303b74 authored by Paolo Bonzini's avatar Paolo Bonzini
Browse files

KVM: Convert kvm_lock back to non-raw spinlock 



In commit e935b837 ("KVM: Convert kvm_lock to raw_spinlock"),
the kvm_lock was made a raw lock.  However, the kvm mmu_shrink()
function tries to grab the (non-raw) mmu_lock within the scope of
the raw locked kvm_lock being held.  This leads to the following:

BUG: sleeping function called from invalid context at kernel/rtmutex.c:659
in_atomic(): 1, irqs_disabled(): 0, pid: 55, name: kswapd0
Preemption disabled at:[<ffffffffa0376eac>] mmu_shrink+0x5c/0x1b0 [kvm]

Pid: 55, comm: kswapd0 Not tainted 3.4.34_preempt-rt
Call Trace:
 [<ffffffff8106f2ad>] __might_sleep+0xfd/0x160
 [<ffffffff817d8d64>] rt_spin_lock+0x24/0x50
 [<ffffffffa0376f3c>] mmu_shrink+0xec/0x1b0 [kvm]
 [<ffffffff8111455d>] shrink_slab+0x17d/0x3a0
 [<ffffffff81151f00>] ? mem_cgroup_iter+0x130/0x260
 [<ffffffff8111824a>] balance_pgdat+0x54a/0x730
 [<ffffffff8111fe47>] ? set_pgdat_percpu_threshold+0xa7/0xd0
 [<ffffffff811185bf>] kswapd+0x18f/0x490
 [<ffffffff81070961>] ? get_parent_ip+0x11/0x50
 [<ffffffff81061970>] ? __init_waitqueue_head+0x50/0x50
 [<ffffffff81118430>] ? balance_pgdat+0x730/0x730
 [<ffffffff81060d2b>] kthread+0xdb/0xe0
 [<ffffffff8106e122>] ? finish_task_switch+0x52/0x100
 [<ffffffff817e1e94>] kernel_thread_helper+0x4/0x10
 [<ffffffff81060c50>] ? __init_kthread_worker+0x

After the previous patch, kvm_lock need not be a raw spinlock anymore,
so change it back.

Reported-by: default avatarPaul Gortmaker <paul.gortmaker@windriver.com>
Cc: kvm@vger.kernel.org
Cc: gleb@redhat.com
Cc: jan.kiszka@siemens.com
Reviewed-by: default avatarGleb Natapov <gleb@redhat.com>
Signed-off-by: default avatarPaolo Bonzini <pbonzini@redhat.com>
parent 4a937f96

Documentation/virtual/kvm/locking.txt

+1 −1
Original line number Diff line number Diff line
@@ -132,7 +132,7 @@ See the comments in spte_has_volatile_bits() and mmu_spte_update(). 
------------



Name:		kvm_lock

Type:		raw_spinlock

Type:		spinlock_t

Arch:		any

Protects:	- vm_list

arch/x86/kvm/mmu.c

+2 −3
Original line number Diff line number Diff line
@@ -4428,7 +4428,7 @@ mmu_shrink_scan(struct shrinker *shrink, struct shrink_control *sc) 
	int nr_to_scan = sc->nr_to_scan;

	unsigned long freed = 0;



	raw_spin_lock(&kvm_lock);

	spin_lock(&kvm_lock);



	list_for_each_entry(kvm, &vm_list, vm_list) {

		int idx;
@@ -4478,9 +4478,8 @@ unlock: 
		break;

	}



	raw_spin_unlock(&kvm_lock);

	spin_unlock(&kvm_lock);

	return freed;



}



static unsigned long

arch/x86/kvm/x86.c

+4 −4
Original line number Diff line number Diff line
@@ -5263,7 +5263,7 @@ static int kvmclock_cpufreq_notifier(struct notifier_block *nb, unsigned long va 


	smp_call_function_single(freq->cpu, tsc_khz_changed, freq, 1);



	raw_spin_lock(&kvm_lock);

	spin_lock(&kvm_lock);

	list_for_each_entry(kvm, &vm_list, vm_list) {

		kvm_for_each_vcpu(i, vcpu, kvm) {

			if (vcpu->cpu != freq->cpu)
@@ -5273,7 +5273,7 @@ static int kvmclock_cpufreq_notifier(struct notifier_block *nb, unsigned long va 
				send_ipi = 1;

		}

	}

	raw_spin_unlock(&kvm_lock);

	spin_unlock(&kvm_lock);



	if (freq->old < freq->new && send_ipi) {

		/*
@@ -5426,12 +5426,12 @@ static void pvclock_gtod_update_fn(struct work_struct *work) 
	struct kvm_vcpu *vcpu;

	int i;



	raw_spin_lock(&kvm_lock);

	spin_lock(&kvm_lock);

	list_for_each_entry(kvm, &vm_list, vm_list)

		kvm_for_each_vcpu(i, vcpu, kvm)

			set_bit(KVM_REQ_MASTERCLOCK_UPDATE, &vcpu->requests);

	atomic_set(&kvm_guest_has_master_clock, 0);

	raw_spin_unlock(&kvm_lock);

	spin_unlock(&kvm_lock);

}



static DECLARE_WORK(pvclock_gtod_work, pvclock_gtod_update_fn);

include/linux/kvm_host.h

+1 −1
Original line number Diff line number Diff line
@@ -142,7 +142,7 @@ struct kvm; 
struct kvm_vcpu;

extern struct kmem_cache *kvm_vcpu_cache;



extern raw_spinlock_t kvm_lock;

extern spinlock_t kvm_lock;

extern struct list_head vm_list;



struct kvm_io_range {

virt/kvm/kvm_main.c

+9 −9
Original line number Diff line number Diff line
@@ -70,7 +70,7 @@ MODULE_LICENSE("GPL"); 
 * 		kvm->lock --> kvm->slots_lock --> kvm->irq_lock

 */



DEFINE_RAW_SPINLOCK(kvm_lock);

DEFINE_SPINLOCK(kvm_lock);

static DEFINE_RAW_SPINLOCK(kvm_count_lock);

LIST_HEAD(vm_list);
@@ -491,9 +491,9 @@ static struct kvm *kvm_create_vm(unsigned long type) 
	if (r)

		goto out_err;



	raw_spin_lock(&kvm_lock);

	spin_lock(&kvm_lock);

	list_add(&kvm->vm_list, &vm_list);

	raw_spin_unlock(&kvm_lock);

	spin_unlock(&kvm_lock);



	return kvm;
@@ -582,9 +582,9 @@ static void kvm_destroy_vm(struct kvm *kvm) 
	struct mm_struct *mm = kvm->mm;



	kvm_arch_sync_events(kvm);

	raw_spin_lock(&kvm_lock);

	spin_lock(&kvm_lock);

	list_del(&kvm->vm_list);

	raw_spin_unlock(&kvm_lock);

	spin_unlock(&kvm_lock);

	kvm_free_irq_routing(kvm);

	for (i = 0; i < KVM_NR_BUSES; i++)

		kvm_io_bus_destroy(kvm->buses[i]);
@@ -3054,10 +3054,10 @@ static int vm_stat_get(void *_offset, u64 *val) 
	struct kvm *kvm;



	*val = 0;

	raw_spin_lock(&kvm_lock);

	spin_lock(&kvm_lock);

	list_for_each_entry(kvm, &vm_list, vm_list)

		*val += *(u32 *)((void *)kvm + offset);

	raw_spin_unlock(&kvm_lock);

	spin_unlock(&kvm_lock);

	return 0;

}
@@ -3071,12 +3071,12 @@ static int vcpu_stat_get(void *_offset, u64 *val) 
	int i;



	*val = 0;

	raw_spin_lock(&kvm_lock);

	spin_lock(&kvm_lock);

	list_for_each_entry(kvm, &vm_list, vm_list)

		kvm_for_each_vcpu(i, vcpu, kvm)

			*val += *(u32 *)((void *)vcpu + offset);



	raw_spin_unlock(&kvm_lock);

	spin_unlock(&kvm_lock);

	return 0;

}

CRA Git | Maintained and supported by SUSTech CRA and CCSE