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: <20200901124615.137200-1-pasha.tatashin@soleen.com>
 <20200902140851.GJ4617@dhcp22.suse.cz> <CA+CK2bBZdN56fmsC2jyY_ju8rQfG2-9hForf1CEdcUVL1+wrrA@mail.gmail.com>
 <74f2341a-7834-3e37-0346-7fbc48d74df3@suse.cz> <20200902151306.GL4617@dhcp22.suse.cz>
In-Reply-To: <20200902151306.GL4617@dhcp22.suse.cz>
From:   Pavel Tatashin <pasha.tatashin@soleen.com>
Date:   Wed, 2 Sep 2020 11:40:53 -0400
Message-ID: <CA+CK2bDVq5UTJfgFuhzPrAGbrFOMcR=Cm1A0nue7M2z+yMac3Q@mail.gmail.com>
Subject: Re: [PATCH] mm/memory_hotplug: drain per-cpu pages again during
 memory offline
To:     Michal Hocko <mhocko@suse.com>
Cc:     Vlastimil Babka <vbabka@suse.cz>,
        LKML <linux-kernel@vger.kernel.org>,
        Andrew Morton <akpm@linux-foundation.org>,
        linux-mm <linux-mm@kvack.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

> > >> But this means that the page is not isolated and so it could be reused
> > >> for something else. No?
> > >
> > > The page is in a movable zone, has zero references, and the section is
> > > isolated (i.e. set_pageblock_migratetype(page, MIGRATE_ISOLATE);) is
> > > set. The page should be offlinable, but it is lost in a pcp list as
> > > that list is never drained again after the first failure to migrate
> > > all pages in the range.
> >
> > Yeah. To answer Michal's "it could be reused for something else" - yes, somebody
> > could allocate it from the pcplist before we do the extra drain. But then it
> > becomes "visible again" and the loop in __offline_pages() should catch it by
> > scan_movable_pages() - do_migrate_range(). And this time the pageblock is
> > already marked as isolated, so the page (freed by migration) won't end up on the
> > pcplist again.
>
> So the page block is marked MIGRATE_ISOLATE but the allocation itself
> could be used for non migrateable objects. Or does anything prevent that
> from happening?

Vlastimil is right, we could allocate from pcplist, if someone
requests allocation, nothing from what I can tell prevents that, and
we will immediately migrate that page in do_migrate_range().

> We really do depend on isolation to not allow reuse when offlining.

Once a page is isolated it is not re-used but here this page is not
isolated because of the race between adding to pcp and isolation.
Draining the second time on a failure fixes the race.

Pasha