On Mon, Jan 12, 2015 at 05:18:39PM -0500, Johannes Weiner wrote:
> On Mon, Jan 12, 2015 at 12:30:37PM +0300, Vladimir Davydov wrote:
> > Since try_to_free_mem_cgroup_pages() can now call slab shrinkers, we
> > should initialize reclaim_state and account reclaimed slab pages in
> > scan_control->nr_reclaimed.
> >
> > Signed-off-by: Vladimir Davydov <vdavy...@parallels.com>
> > ---
> > mm/vmscan.c | 33 ++++++++++++++++++++++-----------
> > 1 file changed, 22 insertions(+), 11 deletions(-)
> >
> > diff --git a/mm/vmscan.c b/mm/vmscan.c
> > index 16f3e45742d6..b2c041139a51 100644
> > --- a/mm/vmscan.c
> > +++ b/mm/vmscan.c
> > @@ -367,13 +367,16 @@ static unsigned long do_shrink_slab(struct
> > shrink_control *shrinkctl,
> > * the ->seeks setting of the shrink function, which indicates the
> > * cost to recreate an object relative to that of an LRU page.
> > *
> > - * Returns the number of reclaimed slab objects.
> > + * Returns the number of reclaimed slab objects. The number of reclaimed
> > + * pages is added to *@ret_nr_reclaimed.
> >
> > static unsigned long shrink_slab(gfp_t gfp_mask, int nid,
> > struct mem_cgroup *memcg,
> > unsigned long nr_scanned,
> > - unsigned long nr_eligible)
> > + unsigned long nr_eligible,
> > + unsigned long *ret_nr_reclaimed)
>
> Can't we just return the number of pages directly from this function?
Hmm, we can. That would look better, of course. However, reclaimed_slab
can be 0 even if we reclaimed tons of dentries/inodes, simply because
they are freed by rcu. In this case, we can abort drop_slab beforehand.
Do you think it's OK?
Thanks,
Vladimir
>
> > @@ -426,7 +434,7 @@ void drop_slab_node(int nid)
> > freed = 0;
> > do {
> > freed += shrink_slab(GFP_KERNEL, nid, memcg,
> > - 1000, 1000);
> > + 1000, 1000, &nr_reclaimed);
> > } while ((memcg = mem_cgroup_iter(NULL, memcg, NULL)) != NULL);
> > } while (freed > 10);
>
> This is the only caller that cares about the return value, and it's a
> magic number that could probably be changed to comparing with a magic
> number of pages instead.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
