From: Zachary Amsden Date: Fri, 20 Aug 2010 08:07:28 +0000 (-1000) Subject: KVM: x86: Add clock sync request to hardware enable X-Git-Url: http://git.lede-project.org./?a=commitdiff_plain;h=ca84d1a24c376e0841f35db08dab7b829c8c0b1e;p=openwrt%2Fstaging%2Fblogic.git KVM: x86: Add clock sync request to hardware enable If there are active VCPUs which are marked as belonging to a particular hardware CPU, request a clock sync for them when enabling hardware; the TSC could be desynchronized on a newly arriving CPU, and we need to recompute guests system time relative to boot after a suspend event. This covers both cases. Note that it is acceptable to take the spinlock, as either no other tasks will be running and no locks held (BSP after resume), or other tasks will be guaranteed to drop the lock relatively quickly (AP on CPU_STARTING). Noting we now get clock synchronization requests for VCPUs which are starting up (or restarting), it is tempting to attempt to remove the arch/x86/kvm/x86.c CPU hot-notifiers at this time, however it is not correct to do so; they are required for systems with non-constant TSC as the frequency may not be known immediately after the processor has started until the cpufreq driver has had a chance to run and query the chipset. Updated: implement better locking semantics for hardware_enable Removed the hack of dropping and retaking the lock by adding the semantic that we always hold kvm_lock when hardware_enable is called. The one place that doesn't need to worry about it is resume, as resuming a frozen CPU, the spinlock won't be taken. Signed-off-by: Zachary Amsden Signed-off-by: Marcelo Tosatti --- diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 4ff0c271f125..d0764a258047 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -5533,7 +5533,15 @@ int kvm_arch_vcpu_reset(struct kvm_vcpu *vcpu) int kvm_arch_hardware_enable(void *garbage) { + struct kvm *kvm; + struct kvm_vcpu *vcpu; + int i; + kvm_shared_msr_cpu_online(); + list_for_each_entry(kvm, &vm_list, vm_list) + kvm_for_each_vcpu(i, vcpu, kvm) + if (vcpu->cpu == smp_processor_id()) + kvm_request_guest_time_update(vcpu); return kvm_x86_ops->hardware_enable(garbage); } diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 5186e728c53e..da117a6b1e2e 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -1961,7 +1961,9 @@ static int kvm_cpu_hotplug(struct notifier_block *notifier, unsigned long val, case CPU_STARTING: printk(KERN_INFO "kvm: enabling virtualization on CPU%d\n", cpu); + spin_lock(&kvm_lock); hardware_enable(NULL); + spin_unlock(&kvm_lock); break; } return NOTIFY_OK; @@ -2168,8 +2170,10 @@ static int kvm_suspend(struct sys_device *dev, pm_message_t state) static int kvm_resume(struct sys_device *dev) { - if (kvm_usage_count) + if (kvm_usage_count) { + WARN_ON(spin_is_locked(&kvm_lock)); hardware_enable(NULL); + } return 0; }