Re: [RFC][PATCH 1/6] memcg: free all at rmdir

[KAMEZAWA Hiroyuki]

On Thu, 13 Nov 2008 06:50:40 +0530
Balbir Singh <balbir@xxxxxxxxxxxxxxxxxx> wrote:

> KAMEZAWA Hiroyuki wrote:
> > On Wed, 12 Nov 2008 16:07:58 -0800
> > Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote:
> > 
> >> On Wed, 12 Nov 2008 12:26:56 +0900
> >> KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx> wrote:
> >>
> >>> +5.1 on_rmdir
> >>> +set behavior of memcg at rmdir (Removing cgroup) default is "drop".
> >>> +
> >>> +5.1.1 drop
> >>> +       #echo on_rmdir drop > memory.attribute
> >>> +       This is default. All pages on the memcg will be freed.
> >>> +       If pages are locked or too busy, they will be moved up to the parent.
> >>> +       Useful when you want to drop (large) page caches used in this memcg.
> >>> +       But some of in-use page cache can be dropped by this.
> >>> +
> >>> +5.1.2 keep
> >>> +       #echo on_rmdir keep > memory.attribute
> >>> +       All pages on the memcg will be moved to its parent.
> >>> +       Useful when you don't want to drop page caches used in this memcg.
> >>> +       You can keep page caches from some library or DB accessed by this
> >>> +       memcg on memory.
> >> Would it not be more useful to implement a per-memcg version of
> >> /proc/sys/vm/drop_caches?  (One without drop_caches' locking bug,
> >> hopefully).
> >>
> >> If we do this then we can make the above "keep" behaviour non-optional,
> >> and the operator gets to choose whether or not to drop the caches
> >> before doing the rmdir.
> >>
> >> Plus, we get a new per-memcg drop_caches capability.  And it's a nicer
> >> interface, and it doesn't have the obvious races which on_rmdir has,
> >> etc.
> >>
> >> hm?
> >>
> > In my plan, I'll add
> > 
> > memory.shrink_usage interface to do and allows
> > 
> > #echo 0M > memory.shrink_memory_usage
> > (you may swap tasks out if there is task..)
> > 
> > to drop pages.
> > 
> 
> So, shrink_memory_usage is just for dropping caches? I don't understand the part
> about swap tasks out.
> 
No, just for shrinking usage. It can also drops ANON.


> > Balbir, how do you think ? I've already removed "force_empty".
> 
> Have you? Won't that go against API/ABI compatibility guidelines. I would
> recommend cc'ing linux-api as well. Sorry, I missed the patch that removes
> force_empty. Me culpa.
> 
account_move did that....

That was only for debug and is a hole for security of resource management.
So, removed.

Thanks,
-Kame

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/