Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <c00a1b621850fedaea3da01ab178e91b79f1dce5.camel@redhat.com>
Subject: Re: [PATCH 7/7] mm/page_alloc: Replace local_lock with normal
 spinlock
From:   Nicolas Saenz Julienne <nsaenzju@redhat.com>
To:     Mel Gorman <mgorman@techsingularity.net>
Cc:     Andrew Morton <akpm@linux-foundation.org>,
        Marcelo Tosatti <mtosatti@redhat.com>,
        Vlastimil Babka <vbabka@suse.cz>,
        Michal Hocko <mhocko@kernel.org>,
        Hugh Dickins <hughd@google.com>,
        LKML <linux-kernel@vger.kernel.org>,
        Linux-MM <linux-mm@kvack.org>
Date:   Tue, 21 Jun 2022 11:31:41 +0200
In-Reply-To: <20220621092950.GF15453@techsingularity.net>
References: <20220613125622.18628-1-mgorman@techsingularity.net>
         <20220613125622.18628-8-mgorman@techsingularity.net>
         <04709b2d0dc702c9bf50f57cde125b07cdf54363.camel@redhat.com>
         <20220621092950.GF15453@techsingularity.net>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.2 (3.44.2-1.fc36) 
MIME-Version: 1.0
Precedence: bulk

On Tue, 2022-06-21 at 10:29 +0100, Mel Gorman wrote:
> On Fri, Jun 17, 2022 at 11:39:03AM +0200, Nicolas Saenz Julienne wrote:
> > Hi Mel,
> >=20
> > On Mon, 2022-06-13 at 13:56 +0100, Mel Gorman wrote:
> > > @@ -3446,12 +3490,16 @@ void free_unref_page(struct page *page, unsig=
ned int order)
> > >  		migratetype =3D MIGRATE_MOVABLE;
> > >  	}
> > > =20
> > > -	local_lock_irqsave(&pagesets.lock, flags);
> > > -	freed_pcp =3D free_unref_page_commit(page, migratetype, order, fals=
e);
> > > -	local_unlock_irqrestore(&pagesets.lock, flags);
> > > -
> > > -	if (unlikely(!freed_pcp))
> > > +	zone =3D page_zone(page);
> > > +	pcp_trylock_prepare(UP_flags);
> >=20
> > Now that you're calling the *_irqsave() family of function you can drop
> > pcp_trylock_prepare/finish()
> >=20
> > For the record in UP:
> >=20
> > #define spin_trylock_irqsave(lock, flags) \
> > ({ \
> > 	local_irq_save(flags); \
> > 	1;
> > })
> >=20
>=20
> The missing patch that is deferred for a later release uses spin_trylock
> so unless that is never merged because there is an unfixable flaw in it,
> I'd prefer to leave the preparation in place.
>=20
> > > +	pcp =3D pcpu_spin_trylock_irqsave(struct per_cpu_pages, lock, zone-=
>per_cpu_pageset, flags);
> > > +	if (pcp) {
> > > +		free_unref_page_commit(pcp, zone, page, migratetype, order);
> > > +		pcp_spin_unlock_irqrestore(pcp, flags);
> > > +	} else {
> > >  		free_one_page(page_zone(page), page, pfn, order, migratetype, FPI_=
NONE);
> > > +	}
> > > +	pcp_trylock_finish(UP_flags);
> > >  }
> > > =20
> > >  /*
> >=20
> > As Vlastimil mentioned elsewhere, I also wonder if it makes sense to ju=
st
> > bypass patch #5. Especially as its intent isn't true anymore:
> >=20
> > "As preparation for dealing with both of those problems, protect the li=
sts
> > with a spinlock.  The IRQ-unsafe version of the lock is used because IR=
Qs
> > are already disabled by local_lock_irqsave.  spin_trylock is used in
> > preparation for a time when local_lock could be used instead of
> > lock_lock_irqsave."
> >=20
>=20
> It's still true, the patch just isn't included as I wanted them to be
> separated by time so a bisection that points to it is "obvious" instead
> of pointing at the whole series as being a potential problem.

Understood, I jumped straight into the code and missed your comment in the
cover letter.

Thanks!

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