linux-kernel - Re: [PATCH 2/3] mm, memcg: inline mem_cgroup_{charge/uncharge} to improve disabled memcg config

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAJuCfpHTT6v2Gjph=BMHPRQ76DRY8q3+4UEs0gaOkm3cLvJKDw@mail.gmail.com>
Date:   Fri, 9 Jul 2021 10:17:48 -0700
From:   Suren Baghdasaryan <surenb@...gle.com>
To:     Johannes Weiner <hannes@...xchg.org>
Cc:     Tejun Heo <tj@...nel.org>, Michal Hocko <mhocko@...nel.org>,
        vdavydov.dev@...il.com, Andrew Morton <akpm@...ux-foundation.org>,
        Shakeel Butt <shakeelb@...gle.com>,
        Roman Gushchin <guro@...com>, songmuchun@...edance.com,
        Yang Shi <shy828301@...il.com>, alexs@...nel.org,
        alexander.h.duyck@...ux.intel.com, richard.weiyang@...il.com,
        Vlastimil Babka <vbabka@...e.cz>, Jens Axboe <axboe@...nel.dk>,
        Joonsoo Kim <iamjoonsoo.kim@....com>,
        David Hildenbrand <david@...hat.com>,
        Matthew Wilcox <willy@...radead.org>, apopple@...dia.com,
        Minchan Kim <minchan@...nel.org>, linmiaohe@...wei.com,
        LKML <linux-kernel@...r.kernel.org>,
        cgroups mailinglist <cgroups@...r.kernel.org>,
        linux-mm <linux-mm@...ck.org>,
        kernel-team <kernel-team@...roid.com>
Subject: Re: [PATCH 2/3] mm, memcg: inline mem_cgroup_{charge/uncharge} to
 improve disabled memcg config

On Fri, Jul 9, 2021 at 8:18 AM Suren Baghdasaryan <surenb@...gle.com> wrote:
>
> On Fri, Jul 9, 2021 at 7:48 AM Johannes Weiner <hannes@...xchg.org> wrote:
> >
> > On Thu, Jul 08, 2021 at 05:05:08PM -0700, Suren Baghdasaryan wrote:
> > > Inline mem_cgroup_{charge/uncharge} and mem_cgroup_uncharge_list functions
> > > functions to perform mem_cgroup_disabled static key check inline before
> > > calling the main body of the function. This minimizes the memcg overhead
> > > in the pagefault and exit_mmap paths when memcgs are disabled using
> > > cgroup_disable=memory command-line option.
> > > This change results in ~0.4% overhead reduction when running PFT test
> > > comparing {CONFIG_MEMCG=n} against {CONFIG_MEMCG=y, cgroup_disable=memory}
> > > configurationon on an 8-core ARM64 Android device.
> > >
> > > Signed-off-by: Suren Baghdasaryan <surenb@...gle.com>
> >
> > Sounds reasonable to me as well. One comment:
> >
> > > @@ -693,13 +693,59 @@ static inline bool mem_cgroup_below_min(struct mem_cgroup *memcg)
> > >               page_counter_read(&memcg->memory);
> > >  }
> > >
> > > -int mem_cgroup_charge(struct page *page, struct mm_struct *mm, gfp_t gfp_mask);
> > > +struct mem_cgroup *get_mem_cgroup_from_mm(struct mm_struct *mm);
> > > +
> > > +int __mem_cgroup_charge(struct page *page, struct mem_cgroup *memcg,
> > > +                     gfp_t gfp);
> > > +/**
> > > + * mem_cgroup_charge - charge a newly allocated page to a cgroup
> > > + * @page: page to charge
> > > + * @mm: mm context of the victim
> > > + * @gfp_mask: reclaim mode
> > > + *
> > > + * Try to charge @page to the memcg that @mm belongs to, reclaiming
> > > + * pages according to @gfp_mask if necessary. if @mm is NULL, try to
> > > + * charge to the active memcg.
> > > + *
> > > + * Do not use this for pages allocated for swapin.
> > > + *
> > > + * Returns 0 on success. Otherwise, an error code is returned.
> > > + */
> > > +static inline int mem_cgroup_charge(struct page *page, struct mm_struct *mm,
> > > +                                 gfp_t gfp_mask)
> > > +{
> > > +     struct mem_cgroup *memcg;
> > > +     int ret;
> > > +
> > > +     if (mem_cgroup_disabled())
> > > +             return 0;
> > > +
> > > +     memcg = get_mem_cgroup_from_mm(mm);
> > > +     ret = __mem_cgroup_charge(page, memcg, gfp_mask);
> > > +     css_put(&memcg->css);
> > > +
> > > +     return ret;
> >
> > Why not do
> >
> > int __mem_cgroup_charge(struct page *page, struct mm_struct *mm,
> >                         gfp_t gfp_mask);
> >
> > static inline int mem_cgroup_charge(struct page *page, struct mm_struct *mm,
> >                                     gfp_t gfp_mask)
> > {
> >         if (mem_cgroup_disabled())
> >                 return 0;
> >
> >         return __mem_cgroup_charge(page, memcg, gfp_mask);
> > }
> >
> > like in the other cases as well?
> >
> > That would avoid inlining two separate function calls into all the
> > callsites...
> >
> > There is an (internal) __mem_cgroup_charge() already, but you can
> > rename it charge_memcg().
>
> Sounds good. I'll post an updated version with your suggestion.
> Thanks for the review, Johannes!

Posted v2 just for this patch at
https://lore.kernel.org/patchwork/patch/1455550 . Please let me know
if you want me to resent the whole patchset instead of just this
patch.
>
> >