Messages in this thread | | | Date | Mon, 29 Jun 2015 17:58:53 +0900 | From | Sergey Senozhatsky <> | Subject | Re: [RFC][PATCHv3 4/7] zsmalloc: introduce zs_can_compact() function |
| |
On (06/29/15 15:45), Minchan Kim wrote: [..] > > +/* > > + * Make sure that we actually can compact this class, > > + * IOW if migration will release at least one szpage. > > zspage,
ok.
> > + * > > + * Should be called under class->lock > > Please comment about return.
ok.
> > + */ > > +static unsigned long zs_can_compact(struct size_class *class) > > +{ > > + /* > > + * Calculate how many unused allocated objects we > > + * have and see if we can free any zspages. Otherwise, > > + * compaction can just move objects back and forth w/o > > + * any memory gain. > > + */ > > + unsigned long obj_wasted = zs_stat_get(class, OBJ_ALLOCATED) - > > + zs_stat_get(class, OBJ_USED); > > + > > I want to check one more thing. > > We could have lots of ZS_ALMOST_FULL but no ZS_ALMOST_EMPTY. > In this implementation, compaction cannot have a source so > it would better to bail out. > IOW, > > if (!zs_stat_get(class, CLASS_ALMOST_EMPTY)) > return 0;
ok.
-ss
| |