Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20190514131654.25463-1-oleksandr@redhat.com> <20190514131654.25463-4-oleksandr@redhat.com>
 <20190514132249.h233crdsz3b7akys@atomlin.usersys.com>
In-Reply-To: <20190514132249.h233crdsz3b7akys@atomlin.usersys.com>
From:   Timofey Titovets <nefelim4ag@gmail.com>
Date:   Wed, 15 May 2019 03:48:50 +0300
Message-ID: <CAGqmi77dtid9M8fZuWimeiWMw8r9Awu579mo8UsaVGTECwxRwA@mail.gmail.com>
Subject: Re: [PATCH RFC v2 3/4] mm/ksm: introduce force_madvise knob
To:     Aaron Tomlin <atomlin@redhat.com>
Cc:     Oleksandr Natalenko <oleksandr@redhat.com>,
        Linux Kernel <linux-kernel@vger.kernel.org>,
        Kirill Tkhai <ktkhai@virtuozzo.com>,
        Vlastimil Babka <vbabka@suse.cz>,
        Michal Hocko <mhocko@suse.com>,
        Matthew Wilcox <willy@infradead.org>,
        Pavel Tatashin <pasha.tatashin@soleen.com>,
        Grzegorz Halat <ghalat@redhat.com>, linux-mm@kvack.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

LGTM

Reviewed-by: Timofey Titovets <nefelim4ag@gmail.com>

=D0=B2=D1=82, 14 =D0=BC=D0=B0=D1=8F 2019 =D0=B3. =D0=B2 16:22, Aaron Tomlin=
 <atomlin@redhat.com>:
>
> On Tue 2019-05-14 15:16 +0200, Oleksandr Natalenko wrote:
> > Present a new sysfs knob to mark task's anonymous memory as mergeable.
> >
> > To force merging task's VMAs, its PID is echoed in a write-only file:
> >
> >    # echo PID > /sys/kernel/mm/ksm/force_madvise
> >
> > Force unmerging is done similarly, but with "minus" sign:
> >
> >    # echo -PID > /sys/kernel/mm/ksm/force_madvise
> >
> > "0" or "-0" can be used to control the current task.
> >
> > To achieve this, previously introduced ksm_enter()/ksm_leave() helpers
> > are used in the "store" handler.
> >
> > Signed-off-by: Oleksandr Natalenko <oleksandr@redhat.com>
> > ---
> >  mm/ksm.c | 68 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
> >  1 file changed, 68 insertions(+)
> >
> > diff --git a/mm/ksm.c b/mm/ksm.c
> > index e9f3901168bb..22c59fb03d3a 100644
> > --- a/mm/ksm.c
> > +++ b/mm/ksm.c
> > @@ -2879,10 +2879,77 @@ static void wait_while_offlining(void)
> >
> >  #define KSM_ATTR_RO(_name) \
> >       static struct kobj_attribute _name##_attr =3D __ATTR_RO(_name)
> > +#define KSM_ATTR_WO(_name) \
> > +     static struct kobj_attribute _name##_attr =3D __ATTR_WO(_name)
> >  #define KSM_ATTR(_name) \
> >       static struct kobj_attribute _name##_attr =3D \
> >               __ATTR(_name, 0644, _name##_show, _name##_store)
> >
> > +static ssize_t force_madvise_store(struct kobject *kobj,
> > +                                  struct kobj_attribute *attr,
> > +                                  const char *buf, size_t count)
> > +{
> > +     int err;
> > +     pid_t pid;
> > +     bool merge =3D true;
> > +     struct task_struct *tsk;
> > +     struct mm_struct *mm;
> > +     struct vm_area_struct *vma;
> > +
> > +     err =3D kstrtoint(buf, 10, &pid);
> > +     if (err)
> > +             return -EINVAL;
> > +
> > +     if (pid < 0) {
> > +             pid =3D abs(pid);
> > +             merge =3D false;
> > +     }
> > +
> > +     if (!pid && *buf =3D=3D '-')
> > +             merge =3D false;
> > +
> > +     rcu_read_lock();
> > +     if (pid) {
> > +             tsk =3D find_task_by_vpid(pid);
> > +             if (!tsk) {
> > +                     err =3D -ESRCH;
> > +                     rcu_read_unlock();
> > +                     goto out;
> > +             }
> > +     } else {
> > +             tsk =3D current;
> > +     }
> > +
> > +     tsk =3D tsk->group_leader;
> > +
> > +     get_task_struct(tsk);
> > +     rcu_read_unlock();
> > +
> > +     mm =3D get_task_mm(tsk);
> > +     if (!mm) {
> > +             err =3D -EINVAL;
> > +             goto out_put_task_struct;
> > +     }
> > +     down_write(&mm->mmap_sem);
> > +     vma =3D mm->mmap;
> > +     while (vma) {
> > +             if (merge)
> > +                     ksm_enter(vma->vm_mm, vma, &vma->vm_flags);
> > +             else
> > +                     ksm_leave(vma, vma->vm_start, vma->vm_end, &vma->=
vm_flags);
> > +             vma =3D vma->vm_next;
> > +     }
> > +     up_write(&mm->mmap_sem);
> > +     mmput(mm);
> > +
> > +out_put_task_struct:
> > +     put_task_struct(tsk);
> > +
> > +out:
> > +     return err ? err : count;
> > +}
> > +KSM_ATTR_WO(force_madvise);
> > +
> >  static ssize_t sleep_millisecs_show(struct kobject *kobj,
> >                                   struct kobj_attribute *attr, char *bu=
f)
> >  {
> > @@ -3185,6 +3252,7 @@ static ssize_t full_scans_show(struct kobject *ko=
bj,
> >  KSM_ATTR_RO(full_scans);
> >
> >  static struct attribute *ksm_attrs[] =3D {
> > +     &force_madvise_attr.attr,
> >       &sleep_millisecs_attr.attr,
> >       &pages_to_scan_attr.attr,
> >       &run_attr.attr,
>
> Looks fine to me.
>
> Reviewed-by: Aaron Tomlin <atomlin@redhat.com>
>
> --
> Aaron Tomlin


--=20
Have a nice day,
Timofey.