Received-SPF: pass (google.com: domain of linux-kernel+bounces-75904-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Precedence: bulk
MIME-Version: 1.0
References: <20240221174333.700197-1-tj@kernel.org> <20240221174333.700197-7-tj@kernel.org>
In-Reply-To: <20240221174333.700197-7-tj@kernel.org>
From: Lai Jiangshan <jiangshanlai@gmail.com>
Date: Thu, 22 Feb 2024 12:36:29 +0800
Message-ID: <CAJhGHyAOzXVwGy5T-KXpwXgcWp6jLaTGGzspPqQU4Dw+x1GqUg@mail.gmail.com>
Subject: Re: [PATCH 6/7] workqueue: Allow cancel_work_sync() and
 disable_work() from atomic contexts on BH work items
To: Tejun Heo <tj@kernel.org>
Cc: torvalds@linux-foundation.org, linux-kernel@vger.kernel.org, 
	allen.lkml@gmail.com, kernel-team@meta.com
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hello

On Thu, Feb 22, 2024 at 1:43=E2=80=AFAM Tejun Heo <tj@kernel.org> wrote:

> @@ -4077,11 +4076,37 @@ static bool __flush_work(struct work_struct *work=
, bool from_cancel)
>
>         rcu_read_lock();
>         pool =3D start_flush_work(work, &barr, from_cancel);
> +       is_bh =3D pool && (pool->flags & POOL_BH);
>         rcu_read_unlock();
>         if (!pool)
>                 return false;
>
> -       wait_for_completion(&barr.done);
> +       if (is_bh && from_cancel) {

Can "*work_data_bits(work) & WORK_OFFQ_BH" be used here?
If so, the previous patch will not be needed.

Thanks.
Lai

> +               /*
> +                * We're flushing a BH work item which is being canceled.=
 It
> +                * must have been executing during start_flush_work() and=
 can't
> +                * currently be queued. If @work is still executing, we k=
now it
> +                * is running in the BH context and thus can be busy-wait=
ed.
> +                *
> +                * On RT, prevent a live lock when %current preempted sof=
t
> +                * interrupt processing or prevents ksoftirqd from runnin=
g by
> +                * keeping flipping BH. If the BH work item runs on a dif=
ferent
> +                * CPU then this has no effect other than doing the BH
> +                * disable/enable dance for nothing. This is copied from
> +                * kernel/softirq.c::tasklet_unlock_spin_wait().
> +                */
> +               while (!try_wait_for_completion(&barr.done)) {
> +                       if (IS_ENABLED(CONFIG_PREEMPT_RT)) {
> +                               local_bh_disable();
> +                               local_bh_enable();
> +                       } else {
> +                               cpu_relax();
> +                       }
> +               }
> +       } else {
> +               wait_for_completion(&barr.done);
> +       }
> +
>         destroy_work_on_stack(&barr.work);
>         return true;
>  }