Re: [RFC PATCH] rcu: introduce kfree_rcu()

["Paul E. McKenney" <paulmck@linux.vnet.ibm.com>]

On Thu, Sep 18, 2008 at 06:56:11PM +0200, Manfred Spraul wrote:
> Paul E. McKenney wrote:
>> On Thu, Sep 18, 2008 at 12:18:28PM +0800, Lai Jiangshan wrote:
>>   
>>> sometimes a rcu callback is just calling kfree() to free a struct's 
>>> memory
>>> (we say this callback is a trivial callback.).
>>> this patch introduce kfree_rcu() to do these things directly, easily.
>>>     
>>
>> Interesting!  Please see questions and comments below.
>>
>>   
>>> There are 4 reasons that we need kfree_rcu():
>>>
>>> 1) unloadable modules:
>>>    a module(rcu callback is defined in this module) using rcu must
>>>    call rcu_barrier() when unload. rcu_barrier() will increase
>>>    the system's overhead(the more cpus the worse) and
>>>    rcu_barrier() is very time-consuming. if all rcu callback defined
>>>    in this module are trivial callback, we can just call kfree_rcu()
>>>    instead, save a rcu_barrier() when unload.
>>>     
> Hmm: why is rcu_barrier() sufficient to prevent races?
> Offlining a cpu reorders rcu callbacks - rcu_barrier() can return before 
> all previous call_rcu() callbacks were called.

The rcu_barrier() family of functions registers a callback on each CPU,
and waits until all these callbacks have been invoked.  The CPU offlining
process preserves the order of the callbacks that were registered on a
given CPU.  Thus, when rcu_barrier() returns, all RCU callbacks previously
registered are guaranteed to have already been invoked, regardless of
what CPUs might have been offlined and onlined in the meantime.

							Thanx, Paul