Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20201202182725.265020-1-shy828301@gmail.com> <20201202182725.265020-3-shy828301@gmail.com>
 <550518d6-fd50-72be-7c84-71153b470cfd@hisilicon.com>
In-Reply-To: <550518d6-fd50-72be-7c84-71153b470cfd@hisilicon.com>
From:   Yang Shi <shy828301@gmail.com>
Date:   Fri, 11 Dec 2020 11:20:19 -0800
Message-ID: <CAHbLzkr16gAYRpLceusLRtJQxx50Wxq1f3fUoGaYHC5-6U1K5g@mail.gmail.com>
Subject: Re: [PATCH 2/9] mm: vmscan: use nid from shrink_control for tracepoint
To:     "Xiaqing (A)" <saberlily.xia@hisilicon.com>
Cc:     Roman Gushchin <guro@fb.com>, Kirill Tkhai <ktkhai@virtuozzo.com>,
        Shakeel Butt <shakeelb@google.com>,
        Dave Chinner <david@fromorbit.com>,
        Johannes Weiner <hannes@cmpxchg.org>,
        Michal Hocko <mhocko@suse.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        Linux MM <linux-mm@kvack.org>,
        Linux FS-devel Mailing List <linux-fsdevel@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Liu Yi <daniel.liuyi@hisilicon.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Wed, Dec 2, 2020 at 7:13 PM Xiaqing (A) <saberlily.xia@hisilicon.com> wr=
ote:
>
>
>
> On 2020/12/3 2:27, Yang Shi wrote:
> > The tracepoint's nid should show what node the shrink happens on, the s=
tart tracepoint
> > uses nid from shrinkctl, but the nid might be set to 0 before end trace=
point if the
> > shrinker is not NUMA aware, so the traceing log may show the shrink hap=
pens on one
> > node but end up on the other node.  It seems confusing.  And the follow=
ing patch
> > will remove using nid directly in do_shrink_slab(), this patch also hel=
ps cleanup
> > the code.
> >
> > Signed-off-by: Yang Shi <shy828301@gmail.com>
> > ---
> >   mm/vmscan.c | 2 +-
> >   1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/mm/vmscan.c b/mm/vmscan.c
> > index 7d6186a07daf..457ce04eebf2 100644
> > --- a/mm/vmscan.c
> > +++ b/mm/vmscan.c
> > @@ -533,7 +533,7 @@ static unsigned long do_shrink_slab(struct shrink_c=
ontrol *shrinkctl,
> >       new_nr =3D atomic_long_add_return(next_deferred,
> >                                       &shrinker->nr_deferred[nid]);
> >
> > -     trace_mm_shrink_slab_end(shrinker, nid, freed, nr, new_nr, total_=
scan);
> > +     trace_mm_shrink_slab_end(shrinker, shrinkctl->nid, freed, nr, new=
_nr, total_scan);
>
> Hi=EF=BC=8C Yang
>
> When I read this patch, I wondered why you modified it so much until I re=
ad patch6. Could you merge
> this patch into patch6?

Sorry for the late reply. It could be, but I was inclined to think
this is a bug and we might need backport it to stable, so I leave it
as a standalone patch.

>
> Thanks!
>
> >       return freed;
> >   }
> >
>