Re: [PATCH v7 4/6] rcu: Add RCU stall diagnosis information

From: "Leizhen (ThunderTown)" <thunder.leizhen@huawei.com>
To: Frederic Weisbecker <frederic@kernel.org>
Cc: "Paul E . McKenney" <paulmck@kernel.org>,
	Neeraj Upadhyay <quic_neeraju@quicinc.com>,
	Josh Triplett <josh@joshtriplett.org>,
	"Steven Rostedt" <rostedt@goodmis.org>,
	Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
	Lai Jiangshan <jiangshanlai@gmail.com>,
	Joel Fernandes <joel@joelfernandes.org>, <rcu@vger.kernel.org>,
	<linux-kernel@vger.kernel.org>, Robert Elliott <elliott@hpe.com>
Subject: Re: [PATCH v7 4/6] rcu: Add RCU stall diagnosis information
Date: Thu, 17 Nov 2022 21:25:44 +0800	[thread overview]
Message-ID: <54aa07c4-b7e3-1a0e-e6a3-444573caba16@huawei.com> (raw)
In-Reply-To: <20221117122238.GC839309@lothringen>

On 2022/11/17 20:22, Frederic Weisbecker wrote:
> On Thu, Nov 17, 2022 at 09:57:18AM +0800, Leizhen (ThunderTown) wrote:
>>
>>
>> On 2022/11/17 6:39, Frederic Weisbecker wrote:
>>> On Fri, Nov 11, 2022 at 09:07:07PM +0800, Zhen Lei wrote:
>>>> @@ -262,6 +279,8 @@ struct rcu_data {
>>>>  	short rcu_onl_gp_flags;		/* ->gp_flags at last online. */
>>>>  	unsigned long last_fqs_resched;	/* Time of last rcu_resched(). */
>>>>  	unsigned long last_sched_clock;	/* Jiffies of last rcu_sched_clock_irq(). */
>>>> +	struct rcu_snap_record snap_record; /* Snapshot of core stats at half of */
>>>> +					    /* the first RCU stall timeout */
>>>
>>> This should be under #ifdef CONFIG_RCU_CPU_STALL_CPUTIME
>>
>> This will not work for now because we also support boot option
>> rcupdate.rcu_cpu_stall_cputime.
> 
> I'm confused. If CONFIG_RCU_CPU_STALL_CPUTIME=n then rcupdate.rcu_cpu_stall_cputime has
> no effect, right?

No, rcupdate.rcu_cpu_stall_cputime override CONFIG_RCU_CPU_STALL_CPUTIME. Because
the default value of CONFIG_RCU_CPU_STALL_CPUTIME is n, so in most cases, we need
rcupdate.rcu_cpu_stall_cputime as the escape route.

If CONFIG_RCU_CPU_STALL_CPUTIME=y is default, your suggestion is more appropriate.

> 
> Thanks.
> 
>>
>>>
>>>> +static void print_cpu_stat_info(int cpu)
>>>> +{
>>>> +	struct rcu_snap_record rsr, *rsrp;
>>>> +	struct rcu_data *rdp = per_cpu_ptr(&rcu_data, cpu);
>>>> +	struct kernel_cpustat *kcsp = &kcpustat_cpu(cpu);
>>>> +
>>>> +	if (!rcu_cpu_stall_cputime)
>>>> +		return;
>>>> +
>>>> +	rsrp = &rdp->snap_record;
>>>> +	if (rsrp->gp_seq != rdp->gp_seq)
>>>> +		return;
>>>> +
>>>> +	rsr.cputime_irq     = kcpustat_field(kcsp, CPUTIME_IRQ, cpu);
>>>> +	rsr.cputime_softirq = kcpustat_field(kcsp, CPUTIME_SOFTIRQ, cpu);
>>>> +	rsr.cputime_system  = kcpustat_field(kcsp, CPUTIME_SYSTEM, cpu);
>>>> +
>>>> +	pr_err("\t         hardirqs   softirqs   csw/system\n");
>>>> +	pr_err("\t number: %8ld %10d %12lld\n",
>>>> +		kstat_cpu_irqs_sum(cpu) - rsrp->nr_hardirqs,
>>>> +		kstat_cpu_softirqs_sum(cpu) - rsrp->nr_softirqs,
>>>> +		nr_context_switches_cpu(cpu) - rsrp->nr_csw);
>>>> +	pr_err("\tcputime: %8lld %10lld %12lld   ==> %lld(ms)\n",
>>>> +		div_u64(rsr.cputime_irq - rsrp->cputime_irq, NSEC_PER_MSEC),
>>>> +		div_u64(rsr.cputime_softirq - rsrp->cputime_softirq, NSEC_PER_MSEC),
>>>> +		div_u64(rsr.cputime_system - rsrp->cputime_system, NSEC_PER_MSEC),
>>>> +		jiffies64_to_msecs(jiffies - rsrp->jiffies));
>>>
>>> jiffies_to_msecs() should be enough.
>>
>> OK, thanks.
>>
>>>
>>> Thanks.
>>>
>>> .
>>>
>>
>> -- 
>> Regards,
>>   Zhen Lei
> .
> 

-- 
Regards,
  Zhen Lei