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;
Date:   Wed, 24 Aug 2022 06:46:02 -0400
From:   Johannes Weiner <hannes@cmpxchg.org>
To:     Chengming Zhou <zhouchengming@bytedance.com>
Cc:     tj@kernel.org, mkoutny@suse.com, surenb@google.com,
        gregkh@linuxfoundation.org, corbet@lwn.net, mingo@redhat.com,
        peterz@infradead.org, songmuchun@bytedance.com,
        cgroups@vger.kernel.org, linux-doc@vger.kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH v3 07/10] sched/psi: add PSI_IRQ to track IRQ/SOFTIRQ
 pressure
Message-ID: <YwYBasgyIU0iQgL3@cmpxchg.org>
References: <20220824081829.33748-1-zhouchengming@bytedance.com>
 <20220824081829.33748-8-zhouchengming@bytedance.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20220824081829.33748-8-zhouchengming@bytedance.com>
Precedence: bulk

On Wed, Aug 24, 2022 at 04:18:26PM +0800, Chengming Zhou wrote:
> @@ -903,6 +903,36 @@ void psi_task_switch(struct task_struct *prev, struct task_struct *next,
>  	}
>  }
>  
> +#ifdef CONFIG_IRQ_TIME_ACCOUNTING
> +void psi_account_irqtime(struct task_struct *task, u32 delta)
> +{
> +	int cpu = task_cpu(task);
> +	void *iter = NULL;
> +	struct psi_group *group;
> +	struct psi_group_cpu *groupc;
> +	u64 now;
> +
> +	if (!task->pid)
> +		return;
> +
> +	now = cpu_clock(cpu);
> +
> +	while ((group = iterate_groups(task, &iter))) {
> +		groupc = per_cpu_ptr(group->pcpu, cpu);
> +
> +		write_seqcount_begin(&groupc->seq);
> +
> +		record_times(groupc, now);
> +		groupc->times[PSI_IRQ_FULL] += delta;
> +
> +		write_seqcount_end(&groupc->seq);
> +
> +		if (group->poll_states & (1 << PSI_IRQ_FULL))
> +			psi_schedule_poll_work(group, 1);
> +	}

Shouldn't this kick avgs_work too? If the CPU is otherwise idle,
times[PSI_IRQ_FULL] would overflow after two missed averaging runs.

avgs_work should probably also self-perpetuate when PSI_IRQ_FULL is in
changed_states. (Looking at that code, I think it can be simplified:
delete nonidle and do `if (changed_states) schedule_delayed_work()`.)