[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CADYN=9+-er+j_7MQDrVbs8Kt5zAz5EdEjszeJWuHjP1K1GjWFg@mail.gmail.com>
Date: Fri, 16 Feb 2024 11:24:26 +0100
From: Anders Roxell <anders.roxell@...aro.org>
To: Tejun Heo <tj@...nel.org>
Cc: Lai Jiangshan <jiangshanlai@...il.com>, linux-kernel@...r.kernel.org,
torvalds@...ux-foundation.org, Naresh Kamboju <naresh.kamboju@...aro.org>,
Peter Zijlstra <peterz@...radead.org>, Valentin Schneider <vschneid@...hat.com>,
Sebastian Andrzej Siewior <bigeasy@...utronix.de>, Stephen Rothwell <sfr@...b.auug.org.au>
Subject: Re: [PATCH v2 wq/for-6.9] workqueue, irq_work: Build fix for !CONFIG_IRQ_WORK
On Fri, 16 Feb 2024 at 06:10, Tejun Heo <tj@...nel.org> wrote:
>
> 2f34d7337d98 ("workqueue: Fix queue_work_on() with BH workqueues") added
> irq_work usage to workqueue; however, it turns out irq_work is actually
> optional and the change breaks build on configuration which doesn't have
> CONFIG_IRQ_WORK enabled.
>
> Fix build by making workqueue use irq_work only when CONFIG_SMP and enabling
> CONFIG_IRQ_WORK when CONFIG_SMP is set. It's reasonable to argue that it may
> be better to just always enable it. However, this still saves a small bit of
> memory for tiny UP configs and also the least amount of change, so, for now,
> let's keep it conditional.
>
> Verified to do the right thing for x86_64 allnoconfig and defconfig, and
> aarch64 allnoconfig, allnoconfig + prink disable (SMP but nothing selects
> IRQ_WORK) and a modified aarch64 Kconfig where !SMP and nothing selects
> IRQ_WORK.
>
> v2: `depends on SMP` leads to Kconfig warnings when CONFIG_IRQ_WORK is
> selected by something else when !CONFIG_SMP. Use `def_bool y if SMP`
> instead.
>
> Signed-off-by: Tejun Heo <tj@...nel.org>
> Reported-by: Naresh Kamboju <naresh.kamboju@...aro.org>
> Fixes: 2f34d7337d98 ("workqueue: Fix queue_work_on() with BH workqueues")
> Cc: Stephen Rothwell <sfr@...b.auug.org.au>
> ---
> Hello,
>
> Unfortunately, the previous patch triggers Kconfig warnings when IRQ_WORK is
> selected by something else but !CONFIG_SMP. This one seems to do the right
> thing in all cases.
>
> Naresh, Anders, can you please test it again?
I applied this patch on yesterdays next tag next-20240215 and I was able to
build the arm64 tinyconfig.
Tested-by: Anders Roxell <anders.roxell@...aro.org>
>
> Thanks.
>
> init/Kconfig | 2 +-
> kernel/workqueue.c | 24 +++++++++++++++---------
> 2 files changed, 16 insertions(+), 10 deletions(-)
>
> diff --git a/init/Kconfig b/init/Kconfig
> index 8df18f3a9748..0d21c9e0398f 100644
> --- a/init/Kconfig
> +++ b/init/Kconfig
> @@ -106,7 +106,7 @@ config CONSTRUCTORS
> bool
>
> config IRQ_WORK
> - bool
> + def_bool y if SMP
>
> config BUILDTIME_TABLE_SORT
> bool
> diff --git a/kernel/workqueue.c b/kernel/workqueue.c
> index 04e35dbe6799..6ae441e13804 100644
> --- a/kernel/workqueue.c
> +++ b/kernel/workqueue.c
> @@ -1209,6 +1209,20 @@ static struct irq_work *bh_pool_irq_work(struct worker_pool *pool)
> return &per_cpu(bh_pool_irq_works, pool->cpu)[high];
> }
>
> +static void kick_bh_pool(struct worker_pool *pool)
> +{
> +#ifdef CONFIG_SMP
> + if (unlikely(pool->cpu != smp_processor_id())) {
> + irq_work_queue_on(bh_pool_irq_work(pool), pool->cpu);
> + return;
> + }
> +#endif
> + if (pool->attrs->nice == HIGHPRI_NICE_LEVEL)
> + raise_softirq_irqoff(HI_SOFTIRQ);
> + else
> + raise_softirq_irqoff(TASKLET_SOFTIRQ);
> +}
> +
> /**
> * kick_pool - wake up an idle worker if necessary
> * @pool: pool to kick
> @@ -1227,15 +1241,7 @@ static bool kick_pool(struct worker_pool *pool)
> return false;
>
> if (pool->flags & POOL_BH) {
> - if (likely(pool->cpu == smp_processor_id())) {
> - if (pool->attrs->nice == HIGHPRI_NICE_LEVEL)
> - raise_softirq_irqoff(HI_SOFTIRQ);
> - else
> - raise_softirq_irqoff(TASKLET_SOFTIRQ);
> - } else {
> - irq_work_queue_on(bh_pool_irq_work(pool), pool->cpu);
> - }
> -
> + kick_bh_pool(pool);
> return true;
> }
>
> --
> 2.43.2
>
Powered by blists - more mailing lists