Messages in this thread | | | Date | Fri, 1 Aug 2008 14:10:32 -0700 | From | "Paul E. McKenney" <> | Subject | Re: [RFC][PATCH 2/2] rcu classic: new algorithm for callbacks-processing(v2) |
| |
On Fri, Jul 18, 2008 at 04:09:30PM +0200, Ingo Molnar wrote: > > * Lai Jiangshan <laijs@cn.fujitsu.com> wrote: > > > This is v2, it's a little deference from v1 that I > > had send to lkml. > > use ACCESS_ONCE > > use rcu_batch_after/rcu_batch_before for batch # comparison. > > > > rcutorture test result: > > (hotplugs: do cpu-online/offline once per second) > > > > No CONFIG_NO_HZ: OK, 12hours > > No CONFIG_NO_HZ, hotplugs: OK, 12hours > > CONFIG_NO_HZ=y: OK, 24hours > > CONFIG_NO_HZ=y, hotplugs: Failed. > > (Failed also without my patch applied, exactly the same bug occurred, > > http://lkml.org/lkml/2008/7/3/24) > > thanks, i've applied this to tip/core/rcu - but it would be nice have an > ack/nak from Paul as well, before we can proceed further.
I like the general approach -- simplification is a good thing.
So, for the moment:
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
I need to do a more detailed review (e.g., the memory-barrier changes and the incorporation of some grace-period processing into call_rcu()), after which I will either ask for changes or give a Reviewed-by.
I need to do this more-detailed review before we ship to mainline. (Gak, already two weeks since Jiangshan sent this!!! My apologies!!!)
Thanx, Paul
| |