Re: [PATCH v2] KVM: Move VM's worker kthreads back to the original cgroups before exiting.

[Paolo Bonzini <pbonzini@redhat.com>]

On 1/18/22 21:39, Tejun Heo wrote:
> So, these are normally driven by the !populated events. That's how everyone
> else is doing it. If you want to tie the kvm workers lifetimes to kvm
> process, wouldn't it be cleaner to do so from kvm side? ie. let kvm process
> exit wait for the workers to be cleaned up.

It does.  For example kvm_mmu_post_init_vm's call to
kvm_vm_create_worker_thread is matched with the call to
kthread_stop in kvm_mmu_pre_destroy_vm.
  
According to Vpin, the problem is that there's a small amount of time
between the return from kthread_stop and the point where the cgroup
can be removed.  My understanding of the race is the following:

user process			kthread			management
------------			-------			----------
							wait4()
exit_task_work()
   ____fput()
     kvm_mmu_pre_destroy_vm()
       kthread_stop();
         wait_for_completion();
				exit_signals()
				  /* set PF_EXITING */
				exit_mm()
				  exit_mm_release()
				    complete_vfork_done()
				      complete();
cgroup_exit()
   cgroup_set_move_task()
     css_set_update_populated()
exit_notify()
   do_notify_parent()
							<wakeup>
							rmdir()
							  cgroup_destroy_locked()
							    cgroup_is_populated()
							    return -EBUSY
				cgroup_exit()
				  cgroup_set_move_task()
				    css_set_update_populated()

I cannot find the code that makes it possible to rmdir a cgroup
if PF_EXITING is set.

Paolo