Re:Re: patch for virtual machine oriented scheduling(1)

Re:Re: patch for virtual machine oriented scheduling(1)

[alex]

>Sorry for that something is wrong with my code posted earlier.
>Just now, I found a bug in it, which was introduced by my porting my
>work(which is developed under KVM83 and using svn depository) to KVM's
>git depository by hand.  The bug is : you can not shutdown SMP guest
>by killing the related qemu process.

the bug is fixed. it seems that once a spinlock is acquired,
Linux2.6.29 prohibits scheduling out
the spinlock holder. so, patch 7 has to be modified:

static void kvm_destroy_vm(struct kvm *kvm)
 {
+       int i;
       struct mm_struct *mm = kvm->mm;

       kvm_arch_sync_events(kvm);
+       if(!kvm->is_paused_by_controller)
+           vm_pause(kvm);
+       for(i = 0; i < KVM_MAX_VCPUS; ++i) {
+           if (kvm->vcpus[i]) {
+               sched_destroy_vcpu(kvm->vcpus[i]);
+           }
+       }
+       sched_destroy_vm(kvm);

       spin_lock(&kvm_lock);
       list_del(&kvm->vm_list);
       spin_unlock(&kvm_lock);

Regards,
alex.