Received-SPF: softfail (google.com: domain of transitioning linux-kernel-owner@vger.kernel.org does not designate 23.128.96.19 as permitted sender) client-ip=23.128.96.19;
Message-ID: <9b04479e3810dbbc3cb2ebd68a3ab3fd034e125e.camel@redhat.com>
Subject: Re: [PATCH 0/6] Drain remote per-cpu directly v3
From:   Nicolas Saenz Julienne <nsaenzju@redhat.com>
To:     Mel Gorman <mgorman@techsingularity.net>,
        Andrew Morton <akpm@linux-foundation.org>
Cc:     Marcelo Tosatti <mtosatti@redhat.com>,
        Vlastimil Babka <vbabka@suse.cz>,
        Michal Hocko <mhocko@kernel.org>,
        LKML <linux-kernel@vger.kernel.org>,
        Linux-MM <linux-mm@kvack.org>
Date:   Fri, 13 May 2022 14:24:47 +0200
In-Reply-To: <20220512085043.5234-1-mgorman@techsingularity.net>
References: <20220512085043.5234-1-mgorman@techsingularity.net>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.1 (3.44.1-1.fc36) 
MIME-Version: 1.0
Precedence: bulk

Hi Mel,

On Thu, 2022-05-12 at 09:50 +0100, Mel Gorman wrote:
> Changelog since v2
> o More conversions from page->lru to page->[pcp_list|buddy_list]
> o Additional test results in changelogs
>=20
> Changelog since v1
> o Fix unsafe RT locking scheme
> o Use spin_trylock on UP PREEMPT_RT
>=20
> This series has the same intent as Nicolas' series "mm/page_alloc:
> Remote
> per-cpu lists drain support" -- avoid interference of a high priority
> task due to a workqueue item draining per-cpu page lists. While many
> workloads can tolerate a brief interruption, it may be cause a real-
> time
> task runnning on a NOHZ_FULL CPU to miss a deadline and at minimum,
> the draining in non-deterministic.
>=20
> Currently an IRQ-safe local_lock protects the page allocator per-cpu
> lists.
> The local_lock on its own prevents migration and the IRQ disabling
> protects
> from corruption due to an interrupt arriving while a page allocation
> is
> in progress. The locking is inherently unsafe for remote access
> unless
> the CPU is hot-removed.
>=20
> This series adjusts the locking. A spinlock is added to struct
> per_cpu_pages to protect the list contents while local_lock_irq
> continues
> to prevent migration and IRQ reentry. This allows a remote CPU to
> safely
> drain a remote per-cpu list.
>=20
> This series is a partial series. Follow-on work should allow the
> local_irq_save to be converted to a local_irq to avoid IRQs being
> disabled/enabled in most cases. Consequently, there are some TODO
> comments
> highlighting the places that would change if local_irq was used.
> However,
> there are enough corner cases that it deserves a series on its own
> separated by one kernel release and the priority right now is to
> avoid
> interference of high priority tasks.
>=20
> Patch 1 is a cosmetic patch to clarify when page->lru is storing
> buddy pages
> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0and when it is storing pe=
r-cpu pages.
>=20
> Patch 2 shrinks per_cpu_pages to make room for a spin lock. Strictly
> speaking
> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0this is not necessary but=
 it avoids per_cpu_pages consuming
> another
> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0cache line.
>=20
> Patch 3 is a preparation patch to avoid code duplication.
>=20
> Patch 4 is a simple micro-optimisation that improves code flow
> necessary for
> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0a later patch to avoid co=
de duplication.
>=20
> Patch 5 uses a spin_lock to protect the per_cpu_pages contents while
> still
> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0relying on local_lock to =
prevent migration, stabilise the pcp
> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0lookup and prevent IRQ re=
entrancy.
>=20
> Patch 6 remote drains per-cpu pages directly instead of using a
> workqueue.
>=20
> =C2=A0include/linux/mm_types.h |=C2=A0=C2=A0 5 +
> =C2=A0include/linux/mmzone.h=C2=A0=C2=A0 |=C2=A0 12 +-
> =C2=A0mm/page_alloc.c=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0 | 348 +++++++++++++++++++++++++------------
> --

Thanks for the series!

I'm testing the series ATM, both with on vanilla and RT kernels. I'll have =
the
results by Monday.

Regards,

--=20
Nicol=C3=A1s S=C3=A1enz