Re: [RFC] rework mem_cgroup iterator

[Glauber Costa <glommer@parallels.com>]

On 11/14/2012 10:41 PM, Tejun Heo wrote:
> Hello, Glauber.
> 
> On Wed, Nov 14, 2012 at 05:17:51PM +0100, Glauber Costa wrote:
>> Why can't we reuse the scheduler iterator and move it to kernel/cgroup.c
>> ? It already exists, provide sane ordering, and only relies on parent
>> information - which cgroup core already have - to do the walk.
> 
> Hmmm... we can but I personally much prefer for_each_*() iterators
> over callback based ones.  It's just much easier to share states
> across an iteration and follow the logic.  walk_tg_tree_from() does
> have the benefit of being able to combine pre and post visits in the
> same walk, which doesn't seem to have any user at the moment.
> 
> Thanks.
> 

Is there any particular reason why we can't do the other way around
then, and use a for_each_*() for sched walks? Without even consider what
I personally prefer, what I really don't like is to have two different
cgroup walkers when it seems like we could very well have just one.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>