Re: [Qemu-devel] [PATCH v2 3/6] net: make netclient re-entrant with refcnt

[liu ping fan]

On Wed, Jul 3, 2013 at 3:49 PM, Stefan Hajnoczi <address@hidden> wrote:
> On Wed, Jul 03, 2013 at 11:41:19AM +0800, liu ping fan wrote:
>> On Mon, Jul 1, 2013 at 7:50 PM, Stefan Hajnoczi <address@hidden> wrote:
>> > On Thu, Jun 20, 2013 at 05:14:56PM +0800, liu ping fan wrote:
>> >> On Tue, Jun 18, 2013 at 8:41 PM, Stefan Hajnoczi <address@hidden> wrote:
>> >> > On Thu, Jun 13, 2013 at 05:03:03PM +0800, Liu Ping Fan wrote:
>> >> >> @@ -1109,6 +1146,7 @@ void net_cleanup(void)
>> >> >>              qemu_del_net_client(nc);
>> >> >>          }
>> >> >>      }
>> >> >> +    qemu_mutex_destroy(&net_clients_lock);
>> >> >
>> >> > Why is it okay to iterate over net_clients here without the lock?
>> >>
>> >>  atexit(&net_cleanup); So no other racers exist.
>> >
>> > What about dataplane?  The device may not be reset when net_cleanup runs.
>> >
>> Does the func registered by atexit run after all of the other threads 
>> terminate?
>
> I imagine that atexit(3) runs while detached threads are still alive,
> but I'm not sure about the exact rules.  The pthread specification links
> I found online didn't state the rules.
>
Haha, finally, got some hint for this.  pthread_exit(3) says:
       After  the  last  thread  in  a  process terminates, the
process terminates as by calling exit(3) with an exit status of zero;
thus, process-shared
       resources are released and functions registered using atexit(3)
are called.

Regards,
Pingfan

> Stefan