Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Wed, 29 Jul 2020 12:49:43 +0300
From:   Vladimir Oltean <olteanv@gmail.com>
To:     Kurt Kanzenbach <kurt.kanzenbach@linutronix.de>
Cc:     Alison Wang <alison.wang@nxp.com>, catalin.marinas@arm.com,
        will@kernel.org, paulmck@kernel.org, mw@semihalf.com,
        leoyang.li@nxp.com, vladimir.oltean@nxp.com,
        linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org,
        Thomas Gleixner <tglx@linutronix.de>,
        Anna-Maria Gleixner <anna-maria@linutronix.de>
Subject: Re: [RFC PATCH] arm64: defconfig: Disable fine-grained task level
 IRQ time accounting
Message-ID: <20200729094943.lsmhsqlnl7rlnl6f@skbuf>
References: <20200729033934.22349-1-alison.wang@nxp.com>
 <877dumbtoi.fsf@kurt>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <877dumbtoi.fsf@kurt>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Wed, Jul 29, 2020 at 10:40:29AM +0200, Kurt Kanzenbach wrote:
> Hi Alison,
> 
> On Wed Jul 29 2020, Alison Wang wrote:
> > In the current arm64 defconfig, CONFIG_IRQ_TIME_ACCOUNTING is enabled as
> > default. According to my tests on NXP's LayerScape and i.MX platforms,
> > the system hangs when running the command "stress-ng --hrtimers 1" with
> > CONFIG_IRQ_TIME_ACCOUNTING enabled. Disabling this option, the issue
> > disappears. CONFIG_IRQ_TIME_ACCOUNTING causes serious performance impact
> > when running hrtimer stress test at the same time.
> 
> I think instead of disabling the option for all arm64 devices, it might
> be better to analyze the root-cause why the hrtimer test hangs when this
> option is enabled.
> 
> +Cc hrtimer maintainers: Thomas and Anna-Maria
> 
> Thanks,
> Kurt

For more context, here is my original report of the issue:
https://lkml.org/lkml/2020/6/4/1062

Just like you, I could not reproduce the RCU stalls and system hang on a
5.6-rt kernel, just on mainline and derivatives, using the plain
defconfig.

The issue is not specific to Layerscape or i.MX8, but rather I was able
to see the same behavior on Marvell Armada 37xx as well as Qualcomm
MSM8976.

So, while of course I agree that disabling IRQ time accounting for arm64
isn't a real solution, it isn't by far an exaggerated proposal either.
Nonetheless, the patch is just a RFC and should be treated as such. We
are at a loss when it comes to debugging this any further and we would
appreciate some pointers.

Thanks,
-Vladimir