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;
MIME-Version: 1.0
References: <20230810021640.252033-1-ming.lei@redhat.com>
In-Reply-To: <20230810021640.252033-1-ming.lei@redhat.com>
From:   Guangwu Zhang <guazhang@redhat.com>
Date:   Fri, 11 Aug 2023 18:59:25 +0800
Message-ID: <CAGS2=Ypt3U6AiR77B-WDUMB5atGt9YsuQoEn+PgZVmcH0n8tXg@mail.gmail.com>
Subject: Re: [PATCH] lib/group_cpus.c: avoid to acquire cpu hotplug lock in group_cpus_evenly
To:     Ming Lei <ming.lei@redhat.com>
Cc:     Thomas Gleixner <tglx@linutronix.de>, linux-kernel@vger.kernel.org,
        Keith Busch <kbusch@kernel.org>,
        linux-nvme@lists.infradead.org, linux-block@lists.infradead.org,
        Yi Zhang <yi.zhang@redhat.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Hi Ming,
I don't find any IO block after apply the patch.

Tested-by: Guangwu Zhang <guazhang@redhat.com>

Ming Lei <ming.lei@redhat.com> =E4=BA=8E2023=E5=B9=B48=E6=9C=8810=E6=97=A5=
=E5=91=A8=E5=9B=9B 10:22=E5=86=99=E9=81=93=EF=BC=9A
>
> group_cpus_evenly() could be part of storage driver's error handler,
> such as nvme driver, when may happen during CPU hotplug, in which
> storage queue has to drain its pending IOs because all CPUs associated
> with the queue are offline and the queue is becoming inactive. And
> handling IO needs error handler to provide forward progress.
>
> Then dead lock is caused:
>
> 1) inside CPU hotplug handler, CPU hotplug lock is held, and blk-mq's
> handler is waiting for inflight IO
>
> 2) error handler is waiting for CPU hotplug lock
>
> 3) inflight IO can't be completed in blk-mq's CPU hotplug handler because
> error handling can't provide forward progress.
>
> Solve the deadlock by not holding CPU hotplug lock in group_cpus_evenly()=
,
> in which two stage spreads are taken: 1) the 1st stage is over all presen=
t
> CPUs; 2) the end stage is over all other CPUs.
>
> Turns out the two stage spread just needs consistent 'cpu_present_mask', =
and
> remove the CPU hotplug lock by storing it into one local cache. This way
> doesn't change correctness, because all CPUs are still covered.
>
> Cc: Keith Busch <kbusch@kernel.org>
> Cc: linux-nvme@lists.infradead.org
> Cc: linux-block@lists.infradead.org
> Reported-by: Yi Zhang <yi.zhang@redhat.com>
> Reported-by: Guangwu Zhang <guazhang@redhat.com>
> Signed-off-by: Ming Lei <ming.lei@redhat.com>
> ---
>  lib/group_cpus.c | 22 ++++++++++++++++------
>  1 file changed, 16 insertions(+), 6 deletions(-)
>
> diff --git a/lib/group_cpus.c b/lib/group_cpus.c
> index aa3f6815bb12..15006e79196f 100644
> --- a/lib/group_cpus.c
> +++ b/lib/group_cpus.c
> @@ -348,6 +348,7 @@ struct cpumask *group_cpus_evenly(unsigned int numgrp=
s)
>  {
>         unsigned int curgrp =3D 0, nr_present =3D 0, nr_others =3D 0;
>         cpumask_var_t *node_to_cpumask;
> +       cpumask_var_t local_cpu_present_mask;
>         cpumask_var_t nmsk, npresmsk;
>         int ret =3D -ENOMEM;
>         struct cpumask *masks =3D NULL;
> @@ -355,6 +356,16 @@ struct cpumask *group_cpus_evenly(unsigned int numgr=
ps)
>         if (!zalloc_cpumask_var(&nmsk, GFP_KERNEL))
>                 return NULL;
>
> +       if (!zalloc_cpumask_var(&local_cpu_present_mask, GFP_KERNEL))
> +               goto fail_local_pres_mask;
> +
> +       /*
> +        * Make a local cache of 'cpu_present_mask', so the two stages
> +        * spread can observe consistent 'cpu_present_mask' without holdi=
ng
> +        * cpu hotplug lock.
> +        */
> +       cpumask_copy(local_cpu_present_mask, cpu_present_mask);
> +
>         if (!zalloc_cpumask_var(&npresmsk, GFP_KERNEL))
>                 goto fail_nmsk;
>
> @@ -366,13 +377,11 @@ struct cpumask *group_cpus_evenly(unsigned int numg=
rps)
>         if (!masks)
>                 goto fail_node_to_cpumask;
>
> -       /* Stabilize the cpumasks */
> -       cpus_read_lock();
>         build_node_to_cpumask(node_to_cpumask);
>
>         /* grouping present CPUs first */
>         ret =3D __group_cpus_evenly(curgrp, numgrps, node_to_cpumask,
> -                                 cpu_present_mask, nmsk, masks);
> +                                 local_cpu_present_mask, nmsk, masks);
>         if (ret < 0)
>                 goto fail_build_affinity;
>         nr_present =3D ret;
> @@ -387,15 +396,13 @@ struct cpumask *group_cpus_evenly(unsigned int numg=
rps)
>                 curgrp =3D 0;
>         else
>                 curgrp =3D nr_present;
> -       cpumask_andnot(npresmsk, cpu_possible_mask, cpu_present_mask);
> +       cpumask_andnot(npresmsk, cpu_possible_mask, local_cpu_present_mas=
k);
>         ret =3D __group_cpus_evenly(curgrp, numgrps, node_to_cpumask,
>                                   npresmsk, nmsk, masks);
>         if (ret >=3D 0)
>                 nr_others =3D ret;
>
>   fail_build_affinity:
> -       cpus_read_unlock();
> -
>         if (ret >=3D 0)
>                 WARN_ON(nr_present + nr_others < numgrps);
>
> @@ -406,6 +413,9 @@ struct cpumask *group_cpus_evenly(unsigned int numgrp=
s)
>         free_cpumask_var(npresmsk);
>
>   fail_nmsk:
> +       free_cpumask_var(local_cpu_present_mask);
> +
> + fail_local_pres_mask:
>         free_cpumask_var(nmsk);
>         if (ret < 0) {
>                 kfree(masks);
> --
> 2.40.1
>


--=20
Guangwu Zhang
Thanks