Re: [PATCH 1/2] memcg: move memcg_{alloc,free}_cache_params to slab_common.c

From: Vladimir Davydov
Date: Tue Sep 23 2014 - 03:31:26 EST


On Mon, Sep 22, 2014 at 04:08:25PM -0400, Johannes Weiner wrote:
> On Thu, Sep 18, 2014 at 07:50:19PM +0400, Vladimir Davydov wrote:
> > The only reason why they live in memcontrol.c is that we get/put css
> > reference to the owner memory cgroup in them. However, we can do that in
> > memcg_{un,}register_cache.
> >
> > So let's move them to slab_common.c and make them static.
> >
> > Signed-off-by: Vladimir Davydov <vdavydov@xxxxxxxxxxxxx>
> > Cc: Johannes Weiner <hannes@xxxxxxxxxxx>
> > Cc: Michal Hocko <mhocko@xxxxxxx>
> > Cc: Christoph Lameter <cl@xxxxxxxxx>
>
> Cool, so you get rid of the back-and-forth between memcg and slab, and
> thereby also shrink the public memcg interface.

It should be mentioned that we still call memcg_update_array_size()
(defined at memcontrol.c) from memcg_update_all_caches()
(slab_common.c), because we must hold the slab_mutex while updating
memcg_limited_groups_array_size. However, I'm going to remove this
requirement and get rid of memcg_update_array_size() too. This is what
"[PATCH -mm 10/14] memcg: add rwsem to sync against memcg_caches arrays
relocation", which is a part of my "Per memcg slab shrinkers" patch set,
does.

Thanks,
Vladimir
--
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/